Volume 140 Nuttiber 3 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY Editorial Board ARTHUR L. COLWIN, Queens College, New York DONALD P. COSTELLO, University of North Carolina PHILIP B. DUNHAM, Syracuse University FRANK M. FISHER, JR., Rice University CATHERINE HENLEY, University of North Carolina MEREDITH L. JONES, Smithsonian Institution W. D. RUSSELL-HUNTER, Syracuse University Managing Editor ROBERT K. JOSEPHSON, Case Western Reserve University CHARLES B. METZ, University of Miami HOWARD A. SCHNEIDERMAN, University of California, Irvine MELVIN SPIEGEL, Dartmouth College STEPHEN A. WAINWRIGHT, Duke University CARROLL M. WILLIAMS, Harvard University JUN 71 Printed aUd-feSued by LANCASTER PRESS, Inc. PRINCE &. LEMON STS. LANCASTER, PA. ;v.\ -* ii THE BIOLOGICAL lin.i.KTix is issued six times a year at -the i.ancuMer 1'ress. Inc., Prince and Lemon Streets, 'Lancaster, J?-enn- svlvanici. .' Subscriptions and similar matter should be addressed to The Biological Bulletin, Marine Biological Laboratory, Woods Hole, Massachusetts. Agent for Great Britain : Wheldon and Wesley, Limited, 2, 3 and 4 Arthur Street, New Oxford Street, London, W. C. 2. Single numbers, $5.00. Subscription per volume (three issues). $14.00. Communications relative to manuscripts should be sent to Dr. W. D. Russell-Hunter, Marine Biological Laboratory, Woods Hole, Massachusetts 02543 between May 23 and September 1, and to Dr. W. D. Russell-Hunter, P.O. Box 103, University Station, Syracuse, New York 13210, during the remainder of the vear. Second-class postage paid at Lancaster, Pa. LANCASTER PRESS, INC., LANCASTER, PA. CONTENTS No. 1. FEBRUARY, 1971 RIDDIFORD, LYNN M. AND CARROLL M. WILLIAMS P\GE Role of the corpora carcliaca in the behavior of saturniid moths. 1. Re- lease of sex pheromone .......................................... 1 TRUMAN, JAMES W. AND LYNN M. RIDDIFORD Role of the corpora carcliaca in the behavior of saturniid moths. II. Ovi- position ........................................................ BAGUET, FERNAND AND JAMES CASE Luminescence control in PoricJithys (Teleostei) : excitation of isolated photophores .................................................... 15 FANKBONER, PETER V. The ciliary currents associated with feeding, digestion, and sediment re- moval in Adula (Botitla ) falcata Gould 1851 ........................ 28 FISHER. F. M., JR. AND C. P. READ Transport of sugars in the tapeworm Calli'obothrium verticillatum ...... 46 FRANZ, DAVID R. Population age structure, growth and longevity of the marine gastropod Urosalpinx cincrca Say .......................................... 63 MARKS, E. P. AND R. A. LEOPOLD Deposition of cuticular substances in vitro by leg regenerates from the cockroach, Lcncof>liaca uiadcrac ( V. ) .............................. 73 MORSE, M. PATRICIA Biology and life history of the nudibranch mollusc, Cor\pliclla stimpsoni (Verrill 1879) ...... '. ............................ " ............... 84 PEARSE, J. S. AND R. W. TIMM Juvenile nematodes (Echinocephalus pseud ouncinatus} in the gonads of sea urchins (Centrostephanus coronatits) and their effect on host gameto- genesis ......................................................... 95 ROBERTS, MORRIS H., JR. Larval development of Pat/itrns longicarpus Say reared in the laboratory. II. Effects of reduced salinity on larval development ................ 104 RODRIGUEZ, LEWIS V. AND REED A. FLICKINGER Bipolar head regeneration in planaria induced by chick embryo extracts . . 117 SCHWAB, DANIEL \Y. AND RICHARD E. SHORE Fine structure and composition of a siliceous sponge spicule .......... 125 STROSS. R. G. Photoperiod control of diapause in Ihiphnia. IV. Light and CO L ,-sensi- tive phases within the cycle of activation ............................ 137 WALKER, JOANNE G. Oxygen poisoning in the annelid Tnbifc.v tnbifc.r. II. Osmotic protection 156 YANG, WON TACK The larval postlarval development of Parthenopc scrrata reared in the laboratory and the systematic position of the Parthenopinae (Crustacea, Brachyura) .................................................... K>(> iv CONTENTS No. 2. APRIL, 1971 ERNST, CARL H. PAGE Sexual cycles and maturity of the turtle, Chryscinys picta 191 GWADZ, ROBERT W., GEORGE B. CRAIG, JR. AND WILLIAM A. HICKEY Female sexual behavior as the mechanism rendering Acdcs aegypti refrac- tory to insemination 201 MANGUM, CHARLOTTE P. AND CHARLES D. Cox Analysis of the feeding response in the onuphid polychaete Diopatra ciiprca (Hose) 215 NASH. DONALD J. Effects of prenatal X-irradiation on development and postnatal viability of inbred and hybrid mice 230 NORTH, BARBARA B., AND GROVER C. STEPHENS Uptake and assimilation of amino acids by Platynionas. II. Increased uptake in nitrogen-deficient cells 242 RUSHFORTH, NORMAN B. Behavioral and electrophysiological studies of hydra. I. An analysis of contraction pulse patterns 255 SASTRY, AKELLA N. AND NORMAN J. BLAKE Regulation of gonad development in the bay scallop, Aequipecten irradians Lamarck 274 SCHELTEMA, RUDOLF S. Larval dispersal as a means of genetic exchange between geographically separated populations of shallow-water benthic marine gastropods 284 SERAFY, D. KEITH Intraspecific variation in the brittle-star Ophiopholis aculeata (Linnaeus) in the Northwestern Atlantic ( Echinodermata ; Ophiuroidea ) 323 TURPEN, JAMES B. AND ROBERT \Y. ANGELL Aspects of molting and calcification in the ostracod Heterocypris 331 YUYAMA, SHUHEI Delay and quadripartition in sea urchin eggs induced by short exposure to 2-mercaptoethanol 339 No. 3. JUNE, 1971 CHENEY, DANIEL P. A summary of invertegrate leucocyte morphology with emphasis on blood elements of the Manila clam. Tapes scinideciissata 353 ENGSTROM, WILMA SILVA Removal of the fertilization membrane of fertilized eggs of Urcchis caitf>o and development of "membraneless" embryos 369 FlNGERMAN, MlLTON, CLELMER K. BARTELL AND ROBERT A. IvRASNOW Comparison of chromatophorotropins from the horseshoe crab Liinnliis polyphemus, and the fiddler crab. Vca pugilator 376 GOODWIN, M. H. AND M. TELFORD The nematocyst toxin of Metridium 389 CONTENTS V HAVEN, NORINE D. Temporal patterns of sexual and asexual reproduction in the colonial ascidian Metandrocarpa taylori Huntsman 400 KINDRED, JAMES E. An attempt at a subjective and objective classification of the acidophilic granulocytes of some marine fishes 416 LAHUE, R. AND \Y. C. CORNING Habituation in Lhniilus abdominal ganglia 427 LASSERRE, PIERRE The marine Enchytraeidae (Annelida, Oligochaeta ) of the eastern coast of North America with notes on their geographical distribution and habitat 440 XIMITZ, SISTER M. AQUINAS Histochemical study of gut nutrient reserves in relation to reproduction and nutrition in the sea stars, Pisastcr ochraccits and Potirio ininiata .... 461 RABINDRANATH, P. A new liljeborgiid amphipod (Crustacea) from Kerala, India 482 ROBERTS, MORRIS H., JR. Larval development of Paynrus longicarpus Say reared in the laboratory. III. Behavioral responses to salinity discontinuities 489 RUSHFORTH, NORMAN B. AND DONALD S. BURKE Behavioral and electrophysiological studies of hydra. II. Pacemaker activity of isolated tentacles 502 Vol. 140, No. 1 February, 1971 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY ROLE OF THE CORPORA CARDIACA IN THE BEHAVIOR OF .SATURNIID MOTHS. I. RELEASE OF SEX PHEROMONE LYNN M. RIDDIFORD AND CARROLL M. WILLIAMS The Biological Laboratories, Harvard University, Cambridge, Massachusetts 02138 Mating of Polyphemus moths under laboratory conditions requires the presence of a volatile emanation from oak leaves (Riddiford and Williams, 1967). The active material has been extracted from red oak leaves and shown to be trans-2- hexenal (Riddiford, 1967). Vapors of a dilute solution of this aldehyde were found to act upon the female antennae. The resulting nervous input to the brain provokes after a certain period of delay the "calling" behavior. The latter can be recognized by inspection in terms of the protrusion of the female genitalia thereby exposing the glands which emit the sex pheromone. In the case of virgin Cercropia females, calling behavior is elicited, not by a chemical, but by photoperiod. Thus, under both long- and short-day conditions, calling begins 1.5 to 2 hours before dawn and often continues for as long as 0.5 hour after lights-on. During this same pre-dawn period, male Cecropia moths become hyperactive even in the absence of females. The third silkmoth considered in the present study was Antheraea pernyi a semi-domesticated species which, like the completely domesticated Boinbyx mori, has for thousands of years been selected for ease of mating. Though virgin Pernyi females show no overt calling behavior under laboratory conditions, there is convincing evidence that the sex pheromone is continuously released to provoke mating at any time of day or night (Riddiford. 1970). In the present study carried out on these three species we have sought to determine whether the corpora cardiaca or corpora allata are involved in the con- trol of the release of sex pheromone. MATERIALS AND METHODS 1. Experimental animals Pupae of Antheraea polyphctnus and Hyalophora cecropia were purchased from dealers or reared outdoors on netted trees (Telfer, 1967). Cocoons of Antheraea pernyi were obtained from Japanese sources. The pupae were stored 1 Copyright 1971, by the Marine Biological Laboratory Library of Congress Card No. A38-518 LYNN M. RIDDIFORD AND CARROLL M. WILLIAMS at 5 C for at least 12 weeks; they were then returned to 25 C to provoke adult development. 2. Excision of corpora allata and/or corpora cardiaca These organs were removed from pupae by the technique described by Williams (1959). To check the completeness of extirpation, the excised organs were placed in a black dish containing Ringer's solution and examined under a dissecting microscope. If the excised organs were not self-evident, the extirpation was con- sidered incomplete and the animal was discarded. In certain individuals the glandular complexes were excised and three pairs of "loose" complexes reim- planted into the thoracic tergum. All individuals were placed at 25 C under controlled photoperiod (usually 17L:7D). Adult development was initiated after about 2 weeks and completed after an additional 3 weeks. 3. Behavioral assays Female Polyphemus moths were caged in a darkened room and exposed to the vapors of 0.05% aqueous trans-2-hexenal solution. The experiment was usually begun in the early evening. At half-hour intervals for at least the first 4 hours, the moths were inspected under dim red light for calling behavior ; they were again inspected the following morning. In the experiments performed on Cecropia the female moths were reared and caged in two constant-temperature rooms, one programmed for a short day (12L:12D) and the other for a long day (17L:7D). Under dim red light the moths were inspected for calling behavior at hourly intervals throughout the scotophase. In the experiments performed on A. pcrnyi, virgin females were caged with males and their mating behavior ascertained as described by Riddiford (1970). EXPERIMENTAL RESULTS 1. Delay in response of virgin female Polyphemus moths to vapors of trans-2- he. venal As described under Methods, 46 normal females were caged in a darkened room in the presence of the vapors of /raM^-2-hexenal. Observations under dim red light at 0.5 hour intervals indicated that at least one hour was required for the initiation of calling behavior and that 74 % of individuals were calling after a total of 4 hours. More detailed observations were carried out on a series of 9 virgin females which were placed, 1 or 2 at a time, in a 2-liter glass chamber in a darkened room. The chamber was ventilated by a gentle stream of air containing the vapors of a 0.05% aqueous solution of fraw^-2-hexenal. Observations of the moths were made at 15-minute intervals under dim red light. The results summarized in Figure 1 once again show that at least an hour elapses before the first individual initiates calling. Fifty per cent of individuals initiated calling within the first 2.25 hours and 100% within the first 4 hours. From these observations we learn that a latent period intervenes between the CORPORA CARDIACA AND "CALLING" BEHAVIOR "Calling" 20 30 60 90 120 150 180 210 Duration of exposure to 0.05% frans-2- hexenal (min) 240 FIGURE 1. The time required for virgin Polyphemus females to begin releasing sex pheromone ("calling") after exposure to vapors of an aqueous solution of 0.05% trans-2- hexenal in the apparatus described under Materials and Methods. A total of 9 moths was used in these determinations. presentation of the chemical stimulus and the initiation of the behavioral response. This delay was the first indication that a neuroendocrine relay mechanism might be involved. 2. Effects of allatectomy The corpora allata were excised from 20 female Polyphemus pupae without any damage to the nearby copora cardiaca. The allatectomized pupae were then placed at 25 C and allowed to develop into adult moths. The latter's response to fmw^-2-hexenal was then determined as described under Methods. The results summarized in Figure 2 are the same as seen for unoperated Polyphemus moths. Moreover, the allatectomized females mated when placed with males and ovi- posited a normal number of eggs which hatched as normal first-instar larvae. The experiment was repeated on 19 allatectomized Cecropia moths exposed to the short-day regimen of 12L:12D. Normal behavior was observed in terms of the presence of calling behavior during the final hour before lights-on. The results summarized in Figure 2 show that the absence of corpora allata in no way affected the response to photoperiod. 3. Effects of removal of the complex of corpora cardiaca and corpora allata The experiment described in the preceding section was repeated on 29 Poly- hemus and 38 Cecropia except that in this case the moths were derived from pupae lacking the entire complex of corpora allata and corpora cardiaca. The results summarized in Figure 2 show a great departure from normal behavior in LYNN M. RIDDIFORD AND CARROLL M. WILLIAMS that fewer than 20% of individuals showed calling behavior in response to the appropriate stimuli. At the conclusion of the experiment many of the moths, including all individuals which had shown a calling response, were sacrificed. The heads were excised, pinned under Ringer's solution, and carefully inspected for any trace of the excised glands. The several individuals which showed any such indications were eliminated from the experiment. In Figure 2 the 14 to 18% of individuals which displayed lOOr 80 z 60 < o 40 20 19 I I H. cecropia A. polyphemus CONTROL -CA -(CC + CA) FIGURE 2. The effect of allatectomy and allatectomy-cardiactomy on the "calling" response of Polyphemus and Cecropia females to vapors of 0.05% fron.y-2-hexenal and to photoperiod respectively. The numbers in parentheses above the bars indicate the number of females tested. the calling response showed no trace of corpora cardiaca. However, the dissection is a difficult one so there remains the possibility that the extirpation may have been incomplete in these individuals. 4. Re-implantation of the corpora allata-corpora cardiaca complexes The glandular complexes were excised from 11 female Polyphemus pupae and 2 female Cecropia pupae. Into the thoracic tergum of each individual were imme- CORPORA CARDIACA AND "CALLING" BEHAVIOR diately reimplanted 3 pairs of "loose" glandular complexes. The moths derived from these preparations were tested for calling in response to the appropriate stimuli. The results were as follows: only 2 (18%) of the Polyphemus moths showed calling behavior when exposed to trans-2-hexena\ vapors ; none of the Cecropia moths showed calling behavior in response to photoperiod. 5. Effects of denervatiny the corpora cardiaca In 5 female Polyphemus pupae the two pairs of nerves connecting the corpora cardiaca with the rear of the brain were severed. When the moths derived from these preparations were challenged with vapors of fraw^-2-hexenal, they showed no trace of the normal calling response. At the conclusion of the experiment autopsies performed on all 5 individuals showed no regeneration of the connections between brain and corpora cardiaca. 6. Experiment on female Pernyi moths As mentioned in the introduction, Antheraea pernyi is a semi-domesticated species which has been highly selected for ease of mating. In experiments re- ported by Earth ( 1965 ) the allatectomized female Pernyi moths were fully effec- tive in attracting and mating with males. The question therefore arises as to whether they can do so if the corpora cardiaca are also extirpated. To answer this question we excised the complex of corpora allata and corpora cardiaca from 16 female Pernyi pupae. The moths derived from these individuals were tested for the release of sex pheromone by caging them with normal males. Fifteen of the 16 females mated within 15 minutes, thereby documenting the continuous release of sex pheromone peculiar to the virgin females of this species. DISCUSSION 1. The role of the corpora cardiaca in the reproductive behavior of virgin female silkmoths In contrast to the continous and apparently spontaneous release of sex phero- mone by virgin females of semi-domesticated Pernyi silkmoths, the undomesticated Cecropia and Polyphemus silkmoths possess a neuro-endocrine mechanism for the control of pheromone release. Our experiments strongly argue that the corpora cardiaca, but not the corpora allata, are necessary for the calling behavior which accompanies the release of sex pheromone by the virgin female moths. That being so, the excision of the corpora cardiaca blocks the normal release of pheromone in response to environmental signals. This loss is not repaired by the reimplanta- tion of as many as three pairs of "loose" glandular complexes a finding which suggests that the nervous connections between the brain and corpora cardiaca are necessary for the behavioral response. When these nervous connections were selectively severed, calling behavior was blocked despite the continued presence of the denervated corpora cardiaca. Females lacking corpora cardiaca fail to mate because males are not attracted to them. However, when caged with males near a cage of normal calling females, they not only mate, but then go on to lay fertile eggs. The absence of corpora LYNN M. RIDDIFORD AND CARROLL M. WILLIAMS cardiaca therefore interferes with the attraction of males but not with the ability to mate and reproduce. 2. Dual role of llic corpora cardiaca The corpora cardiaca serve as neurohaemal organs in which the products of the brain's neurosecretory cells are released into the blood. In addition, the corpora cardiaca contain intrinsic neurosecretory cells and therefore qualify as genuine endocrine organs. When the corpora cardiaca are excised, their intrinsic cells are permanently lost, whereas the neurohaemal portion promptly regenerates from the cut ends of nervi corpora cardiaci I and II (Stumm-Zollinger, 1957). So, for our present purposes it appears that the intrinsic cells of the corpora cardiaca are the source of the hormone which triggers the calling behavior of virgin females. 3. The minimal circuitry In virgin Cecropia and Polyphemus moths the brain processes in-coming signals conveying specific environmental cues relating to the onset and timing of reproductive behavior. After this central integration, signals flow from the brain to the corpora cardiaca via the two pairs of nerves which interconnect them. On the basis of present knowledge we cannot say whether these signals are nerve impulses or neurosecretory agents. The signals in question converge on the intrinsic cells of the corpora cardiaca to provoke the release from these cells of a certain hormone. In the case of cockroaches, Milburn and Roeder (1962) have extracted from the corpora cardiaca a substance which causes rhythmic discharge in the phallic nerve when applied to the abdominal ganglia. Evidently, in the virgin moths an analogous factor is secreted by the intrinsic cells of the corpora cardiaca to promote the motor acts which comprise the calling behavior. A similar type of mechanism involved in oviposition behavior will be examined in the further detail in the following communication (Truman and Riddiford, 1971). Supported by NSF grants GB-7966 (LMR) and GB-7963 (CMW) and by the Rockefeller Foundation. We wish to thank Dr. James Truman for prepara- tion of the figures and for a critical reading of the manuscript. SUMMARY 1. In virgin female silkmoths the protrusion of the genitalia or "calling" behavior signals sex pheromone release. The wild Polyphemus and Cecropia silkmoths "call" in response to specific environmental cues which are chemical and photoperiodic respectively. The semi-domesticated Pernyi silkmoth is exceptional in that it shows no overt "calling" behavior and pheromone is apparently released continuously. 2. By appropriate experiments it was possible to show that the corpora cardiaca but not the corpora allata are prerequisite for the "calling" behavior. Thus, when the corpora allata are removed from female pupae, the behavior of the resulting CORPORA CARDIACA AND "CALLING" BEHAVIOR 7 moths is normal. By contrast, removal of the corpora allata-corpora cardiaca complex greatly reduces the number which "call." 3. In order to perform their function in the "calling" behavior the corpora cardiaca must have intact connections with the brain. Reimplantation of three pairs of corpora allata-corpora cardiaca complexes into animals lacking their own com- plexes fails to restore the ability to "call." "Calling" is also blocked when the nervous connections between the corpora cardiaca and the brain are severed. 4. Evidently, in response to the environmental signals the brain stimulates the release of a hormone from the intrinsic cells of the corpora cardiaca. This hormone then acts on the abdominal nervous system to provoke the protrusion of the female genitalia and the accompanying release of pheromone. LITERATURE CITED BARTH, R. H., JR., 1965. Insect mating behaviour: endocrine control of a chemical com- munication system. Science, 149 : 882-883. MILBURN, N., AND K. D. ROEDER, 1962. Control of efferent activity in the cockroach terminal abdominal ganglion by extracts of corpora cardiaca. /. Gen. ('<>//>. Phvsiol., 2: 70-76. RIDDIFORD, L. M., 1967. 7Ya.y-2-hexenal : mating stimulant for Polyphemus moths. Science, 158: 139-141. RIDDIFORD, L. M., 1970. Antennal proteins of saturniid moths: their possible role in olfaction. /. Insect Physiol, 16 : 653-660. RIDDIFORD, L. M., AND C. M. WILLIAMS, 1967. Volatile principle from oak leaves : role in sex life of the Polyphemus moth. Science, 155 : 589-590. STUMM-ZOLLINGER, E., 1957. Histological study of regenerative processes after transection of the nervi corporis cardiaci in transplanted brains of the Cecropia silkworm (Platysamia cccropia L.). /. Exp. Zool., 134: 315-326. TELFER, W., 1967. Cecropia. Pages 173-182 in F. H. Wilt and N. K. Wessells, Eds., Methods in Developmental Biology. Thomas Y. Crowell Co., New York. TRUMAN, J. W., AND L. M. RIDDIFORD, 1971. Role of the corpora cardiaca in the behavior of saturniid moths. II. Oviposition. Biol. Bull., 140: 8-14. WILLIAMS, C. M., 1959. The juvenile hormone. I. Endocrine activity of the corpora allata of the adult Cecropia silkworm. Biol. Bull., 116: 323-338. Reference : Biol Bull, 140: 8-14. (February, 1971) ROLE OF THE CORPORA CARDIACA IN THE BEHAVIOR OF SATURNIID MOTHS. II. OVIPOSITION JAMES W. TRUMAN AND LYNN M. RIDDIFORD The Biological Laboratories, Harvard University, Cambridge, Massachusetts 02138 Mating in the case of many insects serves as a pivotal event in the life of the adult female. This is especially true for the non-feeding Lepidoptera such as silktnoths of the family Saturniidae. In these insects one can define two conditions, the virgin state and the mated state, each with its own behavioral and physiological characteristics. A case-in-point is the silkmoth, Hyalophora cecropia. Here the virgin female assumes a "calling" posture for sex pheromone release at a specific time of the night (Riddiford and Williams, 1971). This behavior is terminated once mating takes place (Falls, 1933). Moreover, once mating has taken place oviposition is stimulated and longevity is reduced (Rau and Rau, 1914). Due to its quantitative nature we have chosen to study the rate of oviposition as an index of the change from the virgin to the mated state. MATERIALS AND METHODS 1. Experimental animals The experiments were performed on the silkmoth Hyalophora cecropia. Diapausing pupae were obtained from Nebraska and were chilled for at least 12 weeks at 5 C prior to use. 2. Excision of the corpora allata and corpora cardiaca The removal of the corpora allata or the corpora allata-corpora cardiaca com- plexes was performed as described by Williams (1959). 3. Castrations Pupae were anesthetized with CO 2 and placed dorsal-side up in a plasticine cradle. A square of cuticle was excised from the dorsal midline of the third abdominal segment, and the underlying heart was pushed to the side of the operating field. The pupal testes, which are situated in the dorso-lateral region of the fourth abdominal segment, were removed with Dumont #5 forceps. A few crystals of phenylthiourea and streptomycin (Williams, 1959) were placed in the wound, and the operated area was covered by a piece of plastic cover-slip which was sealed in place with melted wax. To prevent the pupa from dislodging the seal, the fourth abdominal segment was immobilized by placing melted wax between segments 3 and 4. After 4 to 5 weeks at 25 C the castrated pupae gave rise to apparently normal moths which mated and produced well-formed, albeit sterile spermatophores. 8 CORPORA CARDIACA AND OVIPOSITION 4. Matings of the moths The females were mated on the first or second day after their emergence. Unoperated females were placed with males in a 1 X 1 X 1 foot cage in the evening. The next morning non-mating animals were removed from the cage and stored elsewhere. The mating pairs were separated in the late afternoon. In order to assure that all operated females mated, these females were hand- mated by a technique developed by C. M. Williams (personal communication). The female was deeply anesthetized with CO 2 . An unanesthetized male moth was grasped by the wings and the genitalia were stimulated with a camel's hair brush to initiate the opening of the claspers. The tip of the female's abdomen was ex- posed by gentle pressure on the abdomen and brought into contact with the male genitalia. In most cases the males readily clasped the female and copula- tion ensued. The pair were separated 12 to 20 hours later. The success of a mating to an unoperated male was judged by the production of fertile eggs. A successful mating to a castrated male was indicated by the remains of a spermatophore in the female's bursa copulatrix. 5. Collection of eggs Each female was placed in a large paper bag. The number of eggs deposited each day was counted during each of the next six to seven days. At the end of this period, the females were sacrificed and dissected under Ringer's solution. The number of chorionated eggs remaining in the ovarioles was determined. This number was added to the total laid by each female and the latter expressed as a percentage of the total number of eggs that had matured. In the graphs of the oviposition patterns of mated moths, the data are referenced to the number of days after the termination of mating. In the case of unmated females, the end of day 1 was scored when the first 5 or more eggs had been oviposited. In most cases this occurred the third or fourth day after emergence. RESULTS 1. 1. The effect of mating on the pattern of oviposition Figure 1A summarizes the cumulative per cent of eggs oviposited by 12 virgin females over a period of 6 days. Oviposition occurred at essentially a constant rate of 7% per day. After 6 days an average of 40% had been deposited and 60% retained. To assess the effect of mating, 26 females were mated to unoperated males. Twenty-four of these crosses were fertile. The pattern of oviposition of these twenty-four is shown in Figure IB. On the first day after mating the percentage of eggs laid by mated females w-as nearly as great as the total laid by virgin females over the entire six-day period. After six days, most females retained fewer than a dozen eggs in their ovarioles. This pattern of oviposition is essentially identical to that which Taschenberg and Roelofs ( 1970) reported for mated Ceropia. The two unsuccessful matings were also of interest. Although the remains of a spermatophore were present in the bursa of both females, none of the eggs 10 JAMES W. TRUMAN AND LYNN M. RIDDIFORD 100 80 40 30 80 60 J L } DAYS 100 80 60 40 - B 5 6 I IOO D 60 40 J L - 6 i i -J 1 1 1 I L FIGURE 1. The oviposition patterns of Cecropia females: A., virgin females; B., females successfully mated to normal males; C, females successfully mated to castrated males; D., allatectomized females successfully mated to normal males. CORPORA CARDIACA AND OVIPOSITION 11 was fertile. Both individuals showed a pattern of oviposition similar to that of virgin females. 2. Matings to castrated males Fifteen females were mated to castrate males. As shown in Figure 1C, the pattern of oviposition was indistinguishable from that of virgin females. Upon autopsy, each female showed the remains of a spermatophore in the bursa copulatrix. Autopsy of the castrated males revealed normally developed accessory glands which were full of secretory products. Consequently, it was assumed that the spermato- phores were normal except for the absence of sperm. 3. Removal of the corpora allata The excision of the corpora allata does not interfere with oogenesis (Williams, 1959), or with the production and release of sex pheromone (Riddiford and Wil- liams, 1971) by female Cecropia moths. However, there remained the possibility that the corpora allata might be involved in the switchover from virgin to mated behavior in response to insemination. To test this possibility the corpora allata were removed from thirteen female pupae. The pupae were then placed at 25 C and allowed to develop into female moths. The females were hand-mated to normal moths. As seen in Figure ID, the removal of the corpora allata did not significantly interfere with the response of these females to insemination. 4. Removal of the complex of corpora allata and corpora cardiaca The corpora allata-corpora cardiaca complexes were removed from 21 female pupae. Into the thorax of four of these pupae were reimplanted 4 or 5 pairs of "loose" corpora allata-corpora cardiaca complexes. A total of 14 of the resulting female moths were successfully mated to normal males (three of these had received implants of the complexes). The removal of the complex of corpora allata and corpora cardiaca essentially abolished the response to insemination (Fig. 2A). As in the case of virgin females, the rate of oviposition was constant over a six-day period ; the average rate w r as about 10% per day as compared to 7 c /c for virgin females. The implantation of up to five pairs of complexes had no effect in restoring the mated response (Fig. 2A, open circles). Seven females lacking corpora allata and corpora cardiaca were mated either to castrated males or to normal males without successful insemination. All seven showed the same type of behavior as unoperated females under the same con- ditions i.e., the virgin pattern of oviposition (Fig. 2B). DISCUSSION 1. The stimulus for the change in female behavior In most insects mating stimulates oogenesis and provokes an abrupt change in the behavior of the female ( Norris, 1933). The specific stimuli which effect these changes are variable. In cockroaches the mechanical stimulus of the sperma- tophore in the bursa copulatrix brings about the mated response (Roth and Stay, 12 JAMES W. TRUMAN AND LYNN M. RIDDIFORD 1961). In Drosophila (Kummer, 1960) and mosquitoes (Leahy and Craig, 1965) the transfer to the female of a secretion from the male accessory gland is respon- sible for the change. In the Lepidoptera (Klatt, 1920; Norris, 1933) and Hemi- ptera (Davey, 1965) the presence of sperm in the female spermatheca is apparently the trigger. The present findings on Cecropia are in substantial agreement with those re- ported for other Lepidoptera. Matings with castrated males involved the passage of a sterile spermatophore but did not cause the switch in behavior. Consequently, the presence of sperm itself appears to be the stimulus which triggers the trans- formation of behavior to the mated pattern. 100 80 60 40 20 100 80 60 40 20 B DAYS FIGURE 2. The oviposition patterns of allatectomized-cardiectomized Cecropia females : A., females successfully mated to normal males the open dots refer to those which had 5 pairs of allata-cardiaca complexes reimplanted ; B, females mated to castrated males or unsuccessfully mated to normal males. 2. The endocrine relay: the corpora cardiaca The effect of mating on oogenesis is presumably mediated by the corpora allata (Engelmann, 1968). But in the case of the behavioral change, the involvement of the endocrine system has remained unexplored. The saturniids have proven to be convenient animals on which to study the change in oviposition behavior because in these insects egg maturation is independent of the corpora allata and the female moths emerge with essentially a full complement of eggs. Thus one circumvents any complications arising from a dependence of oviposition on oogenesis. As seen in the Results, the corpora cardiaca. but not the corpora allata are necessary for the behavioral change. After removal of the corpora allata-corpora CORPORA CARDIACA AND OVIPOSITION 13 cardiaca complexes, mating resulted in only a slight elevation in the oviposition rate over that of virgin females. This was in marked contrast to the results obtained by mating unoperated or allatectomized females to normal males. The fact that the implantation of up to five pairs of complexes into cardiacectomized females did not restore the oviposition response shows the necessity of intact connections between the brain and corpora cardiaca. Therefore, the relationship is identical to that controlling pheromone release by these moths (Riddiford and Williams, 1971). The intrinsic cells of the corpora cardiaca apparently produce an oviposition stimulating factor, and the release of this factor is controlled by the brain. It is important to note that the unsuccessful matings of allatectomized-cardia- cectomized females gave the same result as seen in virgin females or in females which were mated to castrate males. From this w r e learn that the removal of the corpora cardiaca does not interfere with the process of egg-laying itself, but only with the increase in oviposition rate which is provoked by insemination. 3. The role of the corpora cardiaca The corpora cardiaca appear to have a central role in controlling the behavior of the female moth. In the virgin condition the brain directs the release of a "calling" hormone from this organ (Riddiford and Williams, 1971). The results reported here indicate that after the reception of sperm by the female, the brain ceases to promote "calling" and now stimulates the release of an oviposition hor- mone from the corpora cardiaca. Consequently, the corpora cardiaca serve as the switch by which the brain changes the behavior of a female from the virgin to the mated condition. Supported by NSF grant GB-7966 (LMR), a NSF predoctoral fellowship (J.W.T.), and a grant from the Rockefeller Foundation. We w r ish to thank Prof. C. AI. Williams for helpful discussions during this investigation and for a critical reading of the manuscript. SUMMARY 1. In females of the Cecropia silkmoth, mating stimulates the rate of egg deposition. 2. When virgin Cecropia are mated to castrated males, they receive an apparently normal spermatophore except that it lacks sperm. Such females show an oviposition pattern corresponding to that of unmated females. Thus, the normal increase in oviposition rate is triggered by the reception of sperm by the female. 3. The corpora cardiaca are shown to be involved in this change in oviposition rate. After the corpora allata are removed from female pupae, the resulting moths respond normally to insemination. After removal of the corpora allata-corpora cardiaca complexes, the ovipositional response to mating is effectively abolished. Reimplantation of up to 5 pairs of corpora allata-corpora cardiaca complexes into pupae lacking their own complexes does not restore the normal oviposition pattern. 14 JAMES W. TRUMAN AND LYNN M. RIDDIFORD 4. It is concluded that the increase in oviposition rate is due to a hormone which is produced by the intrinsic cells of the corpora cardiaca. By an apparent reflex mechanism, the presence of sperm in the female's spermatheca causes the brain to trigger the release of the hormone in question from the corpora cardiaca. LITERATURE CITED DAVEY, K. G., 1965. Copulation and egg-production in Rhodnius prolixus: the role of the spermathecae. /. E.vp. Biol., 42: 373-378. ENGELMANN, F., 1968. Endocrine control of reproduction in insects. Ann. Rev. Entoinol., 13: 1-26. FALLS, O., 1933. Sex attraction in Samia cccropia. Trans. Kansas Acad. Sci.. 36: 215-217. KLATT, B., 1920. Beitrage zur Sexualphysiologie des Schwammspinners. Biolot/ischcs Zcn- tralblatt, 40: 539-558. KUMMER, H., 1960. Experimentelle Untersuchungen zur Wirkung von Fortpflanzungs-faktoren auf die Lebensdauer vor Drosophila melanogaster Weibchen. Z. Veryl. Ph\siol.. 43: 642-679. LEAHY, M. G., AND G. B. CRAIG, JR., 1965. Accessory gland substance as a stimulant for oviposition in Acdcs acc/yfti and A. allopictus. Mosquito News, 25: 448-452. NORRIS, M. J., 1933. Contributions towards the study of insect fertility. II. Experiments on the factors influencing fertility in Ephcstia kiihniclla Z. (Lepidoptera, Phy- citidae). Proc. Zool. Soc. London. 4: 903-934. RAU, P., AND N. RAU, 1914. Longevity in saturniid moths and its relation to the function of reproduction. Trans. Acad. Sci. St. Louis, 23: 1-78. RIDDIFORD, L. M., AND C. M. WILLIAMS, 1971. Role of the corpora cardiaca in the behavior of silkmoths. I. Release of sex pheromone. Biol. Bull.. 140: 1-7. ROTH, L. M., AND B. STAY, 1961. Oocyte development in Diploptera pinictata (Eschscholtz) (Blatteria). /. Insect Physio!.. 7: 186-202. TASCHENBERG, E F., AND W. L. ROELOFS, 1970. Large-scale rearing of Cecropia (Lepidoptera: Saturniidae). Ann. Entoinol. Soc. Aincr., 63: 107-111. WILLIAMS, C. M., 1959. The juvenile hormone. I. Endocrine activity of the corpora allata of the adult Cecropia silkworm. Biol. Bull., 116: 323-338. Reference : Biol. Bull., 140: 15-27. (February, 1971) LUMINESCENCE CONTROL IN PORICHTHYS (TELEOSTEI) : EXCITATION OF ISOLATED PHOTOPHORES 1 FERNAND BAGUET - AND JAMES CASE Unircrsitc dc Loircain, Lourain, Bcl: 10" 1 microwatt; at 20 C, 4.6 X 10"* microwatt. For a maximal stimulation at 20/sec and 20 C oxygen, the output is 1.4 X 10~ 3 microwatt. These computations are rendered highly approximate if only owing to the difficulty of ascertaining the geometry of light emission and the possi- bility of light absorption by associated tissues and especially chromatophores. We are most indebted to Mr. Jules Crane, Dr. Richard Ibara and Mr. M. S. Trinkle for assistance in obtaining Porichthys and for their most helpful advice and assistance. SUMMARY Luminescene of isolated photophores of midshipman fish, Porichthys myriaster and P. notatus, was studied as a function of electrical stimulus characteristics, temperature, oxygen concentration and pH. Single stimuli delivered into the PORICHTHYS PHOTOPHORE EXCITATION 27 photophore via microelectrode induced luminescence with a minimum latency of 100 msec. Externally applied stimuli were ineffective at rates of less than 2/sec and at 5/sec latency was between 5 and 19 seconds. These values suggest the possibility of both direct and indirect photophore excitability. Oxygen is essential for luminescence, but spontaneous occurrence of luminescence after electrical stim- ulation under nitrogen suggests that certain elements of the excitation-luminescence sequence are relatively insensitive to anaerobiosis. The temperature optimum is at about 20 C and the Q 10 is 4.5. Total light produced was greater at pH 7.4 than at 5.6. Maximal light production is estimated at 1.4 X 10~ 3 microwatt per photophore. LITERATURE CITED ARORA, H. F., 1948. Observations on the habits and early life of the batrachoid fish, Porichthys notatus Girard. Copcia. 1948: 89-93. CHASE, A. M., AND P. H. LORENZ, 1945. Kinetics of the luminescent and nonluminescent reac- tions of Cypridina luciferin at different temperatures. /. Cell. Comfy. Physiol., 25 : 53-63. COCHRAN, W. G., AND G. M. Cox, 1957. Experimental Designs, [2d. edition] J. Wiley, New York, 611 pp. CORMIER, M. J., J. M. CRANE AND Y. NAKANO, 1967. Evidence for the identity of the luminescent system of Porichthys porosissimus (fish) and Cypridina hilgcndorfii (crustacean). Biochcm. Biophys. Res. Commun., 29: 747-752. CRANE, J. M., 1965. Bioluminescent courtship display in the teleost Porichthys notatus. Copcia, 1965: 239-241. GREENE, C. W., 1899. The phosphorescent organs in the toad fish Porichth\s notatus Girard. /. Morphol, 15: 667-696. GREENE, C. W., AND H. H. GREENE, 1924. Phosphorescence of Porichthys notatus, the California singing fish. Aincr. J. Physiol., 70 : 500-507. HARVEY, E. N., 1931. Stimulation by adrenalin of the luminescence of deep-sea fish. Zoologica, 12: 67-69. NICOL, J. A. C., 1957. Observations on photophores and luminescence in the teleost Porichthys. Quart. J. Microscop. Sci., 98: 179-188. NICOL, J. A. C., 1967. The luminescence of fishes. Symp. Zool. Soc. London, 19: 27-55. OHSHIMA, H., 1911. Some observations on the luminous organs of fish. /. Collcqc Sci. (Tokyo), 27: 1-25. STRUM, J. M., 1969. Fine structure of the dermal luminescent organs, photophores, in the fish, Porichthys notatus. Anat. Rcc.. 164: 433-467. U&DENFRIEND, S., 1962. Fluorescence Assay in Biology and Medicine. Academic Press, London, 106 pp. WHITEAR, M., 1952. The innervation of the skin of teleost fishes. Quart. J. Microscop. Sci.. 93: 289-305. YOUNG, J. Z., 1933. The preparation of isotonic solutions for use in experiments with fish. Pubbl. Sta. Zool. Napoli, 12: 425. Reference : Biol .Bull.. 140: 28-45. (February, 1971) THE CILIARY CURRENTS ASSOCIATED WITH FEEDING, DIGESTION, AND SEDIMENT REMOVAL IN ADULA (BOTULA) FALCATA GOULD 1851 1 PETER V. FANKBONER Pacific Marine Station, Dillon Beach, California - Adula jalcata, the pea pod shell, is a common byssally attached rock borer in soft mudstone reefs at Bolinas and Moss Beach, California (Fig. 1). While it possesses the protective advantage of living cryptically, A. jalcata is subjected to an environmental stress not faced by most epifaunal mytilids. Namely, it must function within a burrow into which sediment is being continuously deposited from both particle-laden water passing over its burrow entrance and the mudstone by- products of its own mechanical boring. This report is a comparative study on the ciliary mechanisms of feeding, digestion, and sediment removal in A. jalcata. Sediment removal can hardly be considered separately from feeding and digestion, as it is during these later two processes that sediment is resolved from potential food material and extruded from the burrow. MATERIALS AND METHODS A clear plastic mold of the burrow was constructed for observing the method employed in removing the sediment produced during mechanical boring. A piece of mudstone containing a live specimen of A. jalcata within its burrow was cracked open and the boring mussel carefully removed. A soap model was made of the burrow and a transparent "Bioplastic" two-piece mold (dorsal and ventral pieces with respect to the orientation of the bivalve) was cast of the soap figure. The finished mold was washed in running seawater for several days to remove traces of water soluble chemicals present on the plastic's surface. The original living bivalve, was naturally positioned in the plastic burrow and the bivalve-burrow unit was lowered into running seawater. The bivalve was considered established in the plastic substrate after it had laid down byssal thread on the burrow floor and extended its siphonal process for feeding. The course of ciliary currents was determined by introducing carborundum #100, carborundum #600, carmine, or crushed mudstone and following the move- ments of these substances through a Wild M-5 dissecting microscope. The light source used was a 4000 candle power fiber optics light unit manufactured by Iota Cam Company. The corrosion-acetate technique (Fankboner, 1967), was used for casting the alimentary tract and the ducts of the digestive diverticula. 1 Portion of a thesis submitted to the faculty of the University of the Pacific, in partial fulfillment of the requirements for the M.S. degree. 2 Present address : Department of Biology, University of Victoria, British Columbia, Canada. 28 CILIARY CURRENTS OF ADULA FALCATA fV tSff)* ISO " V-* , t FIGURE 1. A lateral view of the burrow of A. falcata. The mudstone burrow has been cracked open to expose the left side of the enclosed bivalve. RESULTS The ciliary currents of feeding The mantle cavity and its organs are illustrated in Figure 2. A. falcata feeds in the usual bivalve manner by extending its siphonal process and filtering particle-laden water through its ctenidia. Of particular interest in the mantle cavity is a flap of tissue, the siphonal valve, extending ventrally from the anterior VM PA ROLP AN APR IN RILP 0.5CM FIGURE 2. The organs and ciliary currents of the mantle cavity of A. falcata after removal of the right shell valve, the right mantle lobe, and part of the musculature. Abbreviations used are : AA, anterior adductor ; AN, anus ; APR, anterior pedal retractor ; BY, byssus, CSM, cut surface of mantle lobe ; EX, exhalant siphon ; F, foot ; ID, inner demibranch ; IN, inhalant siphon; LI, ligament; OD, outer demibranch ; PA, posterior adduc- tor ; RILP, right inner labial palp ; ROLP, right outer labial palp ; SV, siphonal valve ; VM, visceral mass ; U, umbone. 30 PETER V. FANKBONER portion of the tissue septum separating the exhalant and inhalant siphons. The probable function of this structure will be discussed later. The ctenidia are heterorhabdic (possessing both principal and ordinary fila- ments) and eleutherorhabdic (filaments not united by organic interfiliamenter junctions). However, the filaments maintain a unit integrity as they are bound GC TC LFC FC I S FIGURE 3. A frontal view of the ventral portion of a ctenidial filament of A. falcata and its ciliary currents. The feathered arrows indicate the path followed by large particles or masses of mucus-bound material. The unfeathered arrows indicate the course taken by the finer, lighter particles. Abbreviations used are : CD, ciliary disc ; FC, frontal cilia ; GC, guard cilia; IS, interlamellar septum; LC, lateral cilia; LFC, lateral-frontal cilia; TC, terminal cilia. together in a uniform demibranch by the intermeshed cilia of the ciliary discs (Fig. 3). Each gill consists of two long, slender demibranchs. The inner is complete and lies slightly deeper than the anteriorly reduced outer demibranch (Figs. 2 and 3). The outer demibranchs are approximately ten filaments shorter at their anterior ends than the inner demibranchs. This condition, within the Family CILIARY CURRENTS OF ADULA FALCATA 31 Mytilidae, is not unique to A. falcata. Similar reductions of the outer demibranch are found in Aditla californiensis, Litlwphaga bisiilcata, Musculus (Modiolus) senhonsci, and Mytilus cditlis (personal observation). Dr. Charles R. Stasek (personal communication) of Florida State University at Tallahassee, has observed the same condition in Lithophaga plumula, Musculus (Modiolaria) laevigata, Mytilus californianus. and Septijer bijnrcatus. A functional advantage for this anatomical reduction is unclear but, for Petricola pholadiformis (a rock boring eulamellibranch with a reduced outer demibranch), much of the inner demibranch is exposed to the action of the sorting areas of the outer labial palp (Purchon, 1955a). FIGURE 4. The ventral aspect of the ctenidia-palp association in A. falcata. Abbreviations used are: DG, distal oral groove; ID, inner demibranch; LG, lateral oral groove; LILP, left inner labial palp ; LOLP, left outer labial palp ; M, mantle ; OD, outer demibranch ; PG, proximal oral groove. The most anterior filament on the outer demibranch is unusually large and its ventral tip has grown inward so that it conjoins its terminal food groove with that of the inner demibranch (Fig. 4). During feeding, this modified filament passes food strings to the food groove of the inner demibranch which in turn con- veys them to the palps for further sorting. The frontal ciliary currents and food grooves of the gills are identical to those of Atkin's (1937b) category B(l) for the Family Mytilidae. The first sites of particle sorting on the gills are the adjacent tracts of fine and coarse frontal cilia (Fig. 3). The fine frontal cilia select the smaller, less dense particles and convey this material ventrally to the food grooves (Fig. 4). Once in the food grooves, these particles are passed anteriorly to the distal oral groove from where they are later conveyed by cilia to the mouth and ingested. The longer, coarse frontal cilia carry the larger particles and mucus-bound masses ventrally to the tips of the filaments, where they are passed by the terminal cilia (Fig. 3) anteriorly to the palps for further sorting (Figs. 2 and 4). 32 PETER V. FANKBONER The total sorting efficiency of the outer ascending' portions of the ctenidial demibranchs is associated with those areas of the mantle wall with which the gills normally come in contact. The ciliary currents of the mantle are essentially rejectory and, because they are directed ventro-posteriorly, these currents pass at right angles to the ventro-anteriorly directed currents of the ascending portions of the gill's outer demibranchs (Fig. 2). This system of ciliary currents produces a ventrally directed resultant vector which insures that large mucus-bound masses are rejected as pseudofaeces faster and with more efficiency than by means of the demibranch or the mantle wall alone. DORSAL ABORAL ORAL VENTRAL O.IMM FIGURE 5. A section of the right inner labial palp of A. falcata, showing the palp folds and their ciliary currents. After food is collected by the gill's cilia it is kept segregated from unsorted and/or sediment material by guard cilia projecting from the ventrolateral walls of the food grooves (Fig. 3). This separation facilitates food particles proceeding uninterrupted to the folds of the labial palps for further sorting. In gross morphology, the paired labial palps of A. falcata are similar to those described by Kellogg (1915), for M. edulis (Figs. 2 and 4). The ciliary currents on the folds of the palps (Fig. 5) are as follows : ( 1 ) An orally-directed acceptance current passing over the crests of the palp folds; (2) A ventrally-directed rejection current moving down the crests of the palp folds; (3) An orally-directed acceptance current passing down the proximal slopes of the folds; (4) A ventrally-directed rejection current on the floor of the intrapalp fold groove; (5) An orally-directed acceptance current moving up the distal slope of the fold; (6) A dorsally directed resorting current on the upper portion of the distal slope of the fold; (7) A spiralling resorting current between the folds where they insert into the thick portion of the labial palp. The last mentioned current is probably an accumulative effect of fold currents three and five. CILIARY CURRENTS OF ADULA FALCATA 33 Particles which are accepted by the palps are carried orally over the crests of the palp folds and on to the mouth via the proximal oral groove (Fig. 4). Generally, rejected particles are passed direrctly to the ventral borders of the palps where they are collected by cilia lining the inner surface of the mantle lobes or the visceral mass (Figs. 2, 4, and 5). Subsequently, this rejected material is carried posteriorly to the embayment of the inhalant siphon (Fig. 6), and extruded from the burrow as pseudofaeces. On occasion, A. falcata draws the palps across the lips of its mouth in a wiping fashion, leaving particle-laden mucus- strings within the proximal oral groove (Fig. 4). This material is ingested by the mouth and conveyed to the stomach, where it undergoes further sorting and digestion. I.OCM FIGURE 6. The ventral aspect of the posterior portion of the mantle cavity in A. falcata. The valves have been expanded to demonstrate the ciliary currents (solid arrows) of the ctenidia and the inhalant siphon. The dashed arrows indicate the direction of the currents on the inside surface of the mantle-siphon lobes. The primary function of the labial palps of A. falcata is that of a sorting mechanism, but of almost equal importance is their role as a metering device, controlling the maximum volume of particles and mucus strings reaching the mouth. When the volume of food matter in the oral groove exceeds the ingestion rate of the mouth, mucus strings back up within the groove and are swept away by the rejectory cilia on the palp borders (Fig. 4). This rejected material either leaves the mantle cavity as pseudofaeces or, more rarely, is accepted by the cilia at the base of the inhalant siphon and returned to the ctenidia for resorting (Fig. 6). The ciliary currents of digestion The gut, associated organs, and ciliary currents of A. falcata are similar to those of Botula cinnamomea (Dinamani, 1967), M. edulis (Graham, 1949, and Reid, 1965), L. gracilis (Dinamani, 1967), L. nasuta (Purchon, 1957), and Perna viridis (Dinamani, 1967). However, because there are significant differences between A. falcata and these other mytilids, I will describe the stomach of 34 PETER V. FANKBONER A. jalcata in detail and bring out in the Discussion pertinent comparative structures and functions. The short, ciliated oesophagus leads from the mouth to the spindle shaped mid-gut and its digestive diverticula complex (Figs. 7 and 8). The style sac terminates just anterior to the posterior adductor, while the intestine forms a flattened loop originating at the posterior stomach wall and terminating within the exhalant siphon chamber as the anus (Fig. 2). The stomach is a complex sorting organ which houses the entrances to the caecum and the numerous openings to the dichotomies of the digestive diverticula (Fig. 7). There are sixteen duct openings to the "liver" which are equally SS HG FIGURE 7. A doral view of the stomach of A. jalcata, its structures, and its ciliary currents. Abbreviations used are : A, antero-dorsal tract ; C, caecum ; DD, ducts of digestive diverticula ; GS, gastric shield ; H, dorsal hood ; HG, hood groove ; I, intestinal groove ; IG, minor intestinal groove; L, left duct tract; LF, left fold of duct tract; O, oesophagus; R, right duct tract; RF, right fold of duct tract; S, shield tract; SS, style sac; TT, tongue of major typhlosole ; t, tongue of minor typhlosole ; X, appendix. divided into right and left duct pouches (Figs. 7 and 8). The ducts and tubules of the digestive diverticula are similar in morphology to those described by Owen (1955) for M. editlis. The right and left duct pouches form ciliated gutters along the floor of the stomach which veer to the left anteriorly, entering into the sorting caecum (Figs. 7 and 9). The gross ciliary currents of the pouch tracts are directed towards the caecum. The caecum is a short, finger-shaped pocket which opens into the stomach from the left anterior wall. The structure and the ciliary currents of this organ are illustrated in Figures 7, 8 and 9. The sorting mechanism of the caecum in A. jalcata is similar to Reid's (1965) type B classification with one exception. At the blind end of this finger-like pocket is a structure, not found in the caecum of M edulis, which I call the caecal pouch (Fig. 9). This pouch envelopes the terminal end of the caecum's sorting folds and appears to physically prevent the larger particles and mucus-bound mass from entering the hood groove. CILIARY CURRENTS OF ADI] LA FALCATA 35 There are four dominant ciliated folds in the stomach of A. falcata. The shortest of these is the minor typhlosole which enters the lumen of the stomach from the style sac and extends a short distance along the right wall of the stomach (Fig. 7). The main typhlosole also originates in the style sac and Figure 8. Drawing of the ventral aspect of a vinyl acetate cast of the alimentary tract of A. falcata. The duct tubules have been deleted. Abbreviations used are: C, caecum; L, left duct tract; MG, mid-gut; O, oesophagus; R, right duct tract; SS, style sac. continues as a meandering tongue along the floor of the stomach and disappears into the sorting caecum. Just prior to entering the caecum, it divides unequally into two folds. One of these folds is a continuation of the main typhlosole's tongue and the other becomes the left fold of the stomach (Figs. 7 and 9). The fourth fold of the stomach, the right fold, has its origin at the base of the right duct 36 PETER V. FANKBONER pouch and extends anteriorly along the floor of the stomach running parallel to the main typholosole and following it into the caecum. With the exception of the minor typholsole, all of these folds convey particles and mucus-bound materials anteriorly into the caecum for sorting. LF FIGURE 9. The sorting area of the caecum of A. falcata, its structure and ciliary currents. Abbreviations used are : CP, caecal pocket ; DD, duct of the digestive diverticula ; DT, duct typhlosole ; HG, hood groove ; I, intestinal groove ; L, left duct tract ; LF, left fold of duct tract; R, right duct tract; RF, right fold of duct tract; TT, tongue of major typhlosole. The intestinal groove has its source at the point where the main typhlosole sub-divides to form the left fold of the stomach (Fig. 7), and its functions in passing materials, which have been rejected by the sorting mechanisms of the stomach, posteriorly to the mid-gut as waste material. Forming a tributary to the intestinal groove is a crease in the right wall of the stomach, the minor intestinal groove. This groove has two functions. Its anterior half conveys particles and mucus-bound matter to ciliary currents leading CILIARY CURRENTS OF ADULA FALCATA 37 to the dorsal hood; its posterior portion is rejectory in nature and passes rejected materials to the intestinal groove (Fig. 7). Just before joining the intestinal groove, the minor intestinal groove forms a shallow pocket which, in M. edulis, is referred to as the "appendix" by Reid (1965). The function of the appendix in A. jalcata is to remove excess mucus-bound material from the food bolus spinning in the lumen of the stomach. When the bolus becomes too large from a high influx of food material from the oesophagus, it begins to press or extrude the excess into any grooves or pockets within the stomach wall. When such excess material is pushed into the appendix, ciliary currents convey it to the intestinal groove for expulsion from the stomach. There are two ridged ciliated tracts in the walls of the stomach. One, the shield tract, is found on the left wall of the stomach, just posterior to the gastric shield. The second, the antero-dorsal tract, is located in a position complementary to the shield tract, on the right wall of the stomach. The ciliary currents on both tracts are relatively \veak and, as far as I have been able to determine, are far less effectual than the other stomach sorting areas. A. jalcata is a relatively advanced mytilian bivalve (Yonge, 1955) and probably its shield and anterodorsal tracts have, through time, lost the sorting activity observed by Graham (1949) and Reid (1965) in the more primitive M. edulis. The translucent crystalline style projects from the style sac and its head is lodged against the lobes of the chitinous gastric shield (Fig. 7). The style's con- sistency is very soft and often possesses little more viscosity than a thick mucus. Occasionally, specimens of A. jalcata have styles which contain chunks of food material embedded within the center of the rod. The presence of embedded matter appears to be related to whether the animals are freshly broken out of their mudstone burrows and examined in the field or whether they are studied after living in a marine aquarium for several weeks. In the former situation, all of the styles are clear, but those which are kept submerged and actively feeding in the laboratory tend to develop dark corkscrew spirals of chunky food material down the style centers. The course of digestion begins following ingestion by the mouth of a particle- laden mucus string, which is moved by cilia down the oesophagus and into the stomach. Upon entering the stomach, there are several paths the mucus food string may take. The first of these paths, probably the most frequently used, is similar to that described by Reid (1965) for M. edulis. The food string is con- veyed down the anterior wall of the stomach and is carried in an oblique direction over the right stomach fold, the main typholosole, and the left fold. It then passes posteriorly over the left wall of the stomach until it is picked up by the mucus-bound mass enclosing the head of the crystalline style (Fig. 7). The style, which, when viewed from the anterior of the animal, is turning slowly in a clock- wise direction, begins to dissolve and forms a corkscrew in and around the food-mucus mass. During the spinning of the crystalline style which is coated with a "bolus" of potential food matter, small portions peel free from the turning mucus-bound mass. This is probably due to what Owen (1967) suggests is the effect of gastric fluids and abrasion by the stomach walls. Most of these food-mucus pieces drop into the right and left duct pouches, where they are either drawn into the main ducts of the digestive diverticula (Fig. 7) or are swept PETER V. FANKBONER anteriorly into the pouch ciliary tracts and folds for later sorting in the caecum (Fig. 9). Food entering the stomach at a rate faster than it can he sorted and utilized is bypassed to the mid-gut via the intestinal groove. When the mucus-food strings around the head of the crystalline style accumulate to excess, the material is usually scraped off into the appendix, where it is later passed to the intestinal groove and carried into the mid-gut as waste. In the duct pouches, fine particles are taken into the main ducts of the digestive diverticula by what Owen (1955) suggests in M. edulis is an inhalant counter current in the nonciliated portion of the main ducts, created by an exhalant current in the ciliated portion. Owen submits that in M. edulis the ciliary currents of the main ducts convey waste products from the diverticula to the stomach for re- moval. He proposes that the counter current functions in drawing food particles towards the diverticula, where they are later digested intracellularly. A. jalcata probably functions in a similar manner. Particles which are passed anteriorly in the pouch tracts or the folds lying on the floor of the stomach, undergo minor sorting before they enter the caecum. The smaller, more dense particles are swept off the main typhlosole into the major intestinal groove to be later passed into the mid-gut as waste material. The larger mucus-bound masses or strings of material and the smaller, lighter particles are conveyed by the main typhlosole, the pouch folds, and, in part, by the lower left wall of the stomach, into the caecum. Once in the caecum, potential food particles are sorted by the right and left folds and the main typhlosole (Fig. 9). The heaviest finest particles are rejected into the intestinal groove, while the lighter particles are carried into the caecal pocket where cilia sweep them into the hood groove. Particles which enter the hood groove are conveyed out of the caecum and to the dorsal hood. At the head of the dorsal hood, particles and mucus-bound masses accumulate until they spill over the gastric shield and are picked up by the head of the crystalline style. The larger food strings and mucus masses either remain in the caecum, where they are eventually broken down into finer sizes, or are passed out of the caecum via the ciliated surface of the caecum's posterior wall. This material is soon caught up in the currents of the stomach's left wall and carried to the head of the dorsal hood where it is wound around the head of the style. A second path a mucus-food string may take after entering the stomach lumen is to pass along the left anterior stomach wall into the sorting caecum. The subsequent treatment of the mucus food strings within the caecum is similar to that described in previous paragraphs. The final alternative route which may be taken by mucus-food strings entering the stomach is to be conveyed directly from the oesophagus to the minor intestinal groove. Within this groove, ciliary currents carry the food strings posteriorly where they enter the dorsal hood and are passed to the spinning mucus-food mass or bolus enveloping the anterior portion of the crystalline style. More rarely, food strings continue in a posterior direction along the minor intestinal groove and ultimately empty into the mid-gut. On occasions when the stomach is relatively empty, food strings fall into the right duct tract and pass on to the caecum for sorting. CILIARY CURRENTS OF ADULA FALCATA 39 General comments on criteria for acceptance or refection of particulate material by ciliary sorting mechanisms in A. falcata Particle selection or rejection that takes place in A. falcata, whether by the gills, palps or stomach, appears to be based upon the size, weight, concentration, and mucus-bound state of potential food material. In general, material small in size, light in weight, in low concentration, and free of mucus is selected over larger, heavier, highly concentrated, mucus-bound particles. I observed no evi- dence of a selective process other than a physical one. Sediment removal from the burrow The fate of sediment matter entering the mantle cavity during feeding has been, in part, previously discussed. Removal of the products of mechanical boring from the burrow is initially a different process, involving active participation by portions of the mantle folds and the foot. 0.5CM FIGURE 10. The ventral aspect of the anterior portion of A. falcata showing the folds of the mantle, the foot, and their ciliary currents. Yonge (1955) suggests that in A. falcata sediment material resulting from mechanical boring is removed from the burrow via the folds of the mantle and the inhalant siphon. My observations on a specimen living in a clear plastic burrow confirm this. Carborundum powder (#100 and #600) which I placed on the umbones of the specimen in the plastic burrow were deposited within a few minutes at the head of the burrow. Next, the deposited particles were contacted by the anterior, fused, inner folds of the mantle margin, conveyed by cilia ventrally (with respect to the bivalve) and passed into the lumen of the mantle cavity (Figs. 2 and 10). If the carborundum particles entering the mantle cavity were introduced in large quantities, they generally became embedded in the mucus produced by the mantle epithelium and were conveyed as stringy masses posteriorly by the cilia lining the walls of the mantle cavity. On reaching the tip the inhalant siphon, the mucus-bound particles were swept away by the currents of the exhalant siphon (Fig. 6). There is one obvious advantage to pseudofaeces leaving a bivalve in 40 PETER V. FANKBONER a mucus-bound slate. Wave action on the reef probably carries the mucus embedded material a considerable distance from the burrow opening before it breaks up. Hence, it is unlikely that the material would be introduced again in the feeding process. When small quantities of carborundum particles were taken into the mantle cavity, they were usually conveyed to the sorting folds of the labial palps where they were treated like incoming potential food material. Finer suspended par- ticles in the burrow are drawn into the mantle cavity through an iris-like opening formed by the anterior unfused portion of the ventral mantle marginal folds (Fig. 10). The water currents which draw in this suspended material are prob- ably created by the slight opening and closing of the shell valves because the fluids were taken into the mantle cavity in gasping draughts rather than the steady flow characteristic of ctenidial currents. After suspended material has entered the mantle cavity, it is collected by the filtering mechanism of the gills and processed as potential food material. The foot is long, slender, and serpentine in its movements and shape. Yonge (1955, page 387) submits: "Once the animal (A. falcata} is established in its boring the only, but all-important, function of the foot is to plant the byssus threads." My observations have revealed that the foot (Fig. 3) has the significant additional function of removing sediment from the walls of its mudstone burrow. The ciliary currents of the foot are shown in Figures 2 and 10. The extensible powers of the foot are unusual. I have, on occasion, observed it expanded to at least four-fifths the length of the shell valves. There does not appear to be any portion of the walls of the mudstone burrow or the shell valves which are inaccessible to the ciliated tip of the foot. Observations on A. falcata in the clear plastic burrow revealed that the foot occasionally extended and touched portions of the burrow walls in a browsing manner. Further, wherever the ciliated foot touched areas of the burrow walls which were coated with #600 mesh carborundum powder, the particles were removed and passed to the mantle cavity by the foot's ciliated surfaces. Mudstone particles and carmine were also introduced and were treated in a similar manner to carborundum according to particle size, particle density, and mucus-bound state. I was never able to place more participate ma- terial within the plastic burrow than could be handled by the mantle cavity's mechanisms for sediment removal. I was particularly impressed with the ease with which the ciliary currents of the inhalant siphon could move large, dense, mucus- bound masses (sometimes 0.5 cm in diameter) of #100 mesh carborundum powder in a vertical direction. In general, large particles were removed from the burrow as pseudofaeces via the inhalant siphon and smaller, suspended material entered the alimentary system and was later incorporated into faecal pellets. DISCUSSION In particle-laden waters, many eulamellibranchs rid their mantle cavities of clogging particles by first concentrating the undesirable material into mucus-bound masses at the base of the inhalant siphon and then directing the siphonal valves so that the incurrent stream of water washes the pseudofaeces from the mantle cavity ventrally through a gape or aperture between the mantle folds (Kellogg, 1915 ? and Yonge, 1948). Obviously a rock boring form such as A. falcata CILIARY CURRENTS OF ADULA FALCATA 41 cannot utilize this mechanism because it would result in effectively "fouling one's own nest." Instead, most waste material must be removed from the mantle cavity via ciliary currents of the inhalant siphon. The boring bivalves Hiatella and Borneo, have a slight variation on the theme for pseudofaecal cleansing. In Hiatella, pseudofaeces are first consolidated into tiny balls by ciliary vortices or swirls at the base of the inhalant siphon and then later washed out through the inhalant siphon by spasmodic contractions of the clam's two adductor muscles (Hunter, 1949). The genus Barnea has evolved a filmy flap, the collecting membrane, which extends from the posteroventral portion of the visceral mass deep into the inhalant siphonal embayment (Kellogg, 1915, and Purchon, 1955bj. Posteriorly directed ciliary currents of the collecting membrane deposit pseudo- faeces into the incurrent siphon, where it is removed from the bivalve via extrusion through the siphon's opening. Such a system bypasses the problem of the stream of incoming water washing away the pseudofaeces because it effectively shields the material until it is deposited near the point of extrusion. Yonge (1955) suggests that the siphonal valve (Fig. 2 and 6) of A. jalcata acts as a shield to protect pseudofaeces from the broad incoming stream of water from the inhalant siphon. I question this interpretation for two reasons. First, any material passing posteriorly through the inhalant siphon would obviously be subjected to the full force of the incoming water currents when it was carried by cilia over the siphonal membrane or the inner walls of the anterior portion of the mantle folds (Fig. 6). Second, A. falcata utilizes a gape in the ventral mantle folds just anterior to the siphonal process as a functional inhalant siphon during the removal of moderate to heavy loads of pseudofaeces. While this "functional" inhalant siphon is operative, the siphonal valve appears to work as a plug for the true siphon. Guard cilia generally occur in those bivalves which inhabit a substrate con- taining some muddy or silty material (Atkins, 1937a). The mudstone burrow of A. jalcata provides a similar habitat because sediment material from both mechanical boring within the burrow, and suspended silt from mudstone w r eathering outside the burrow, are deposited into the excavation. The deep marginal grooves and their guard cilia provide a mechanism for insuring that potential food material travels the full distance to the labial palps, while at the same time remaining segregated from the silt laden stream of water entering through the inhalant siphon. Stasek (1963), has determined, from fifty-five bivalve families, that a well- defined association is present between the ctenidium and the labial palps. Based on anatomical characters, he has segregated the members of these families into three major categories. As a mytilid, A. jalcata falls into the most primitive of these groups, Category I. Members of Category I are characterized by the following anatomical features : ". . . the ventral tips of at least the first few or, usually, of many filaments of the inner demibranch are inserted unjused into a distal oral groove (a designation originated by Kellogg, 1915)" (Stasek, 1963, page 91). In A. jalcata, and indeed in the six mytilian genera I have examined, the first few filaments of the inner demibranch are always inserted tmfused into a palp pocket lying within the lateral oral groove (LG) (Fig. 4). Kellogg's (1915, page 629) definition of the distal oral groove states (the italics are mine) : "The third is a groove in the mantle wall, close to and parallel with, the anterior edge of the inner demibranch, found 42 PETER V. FANKBONER in forms in which the outer demihranch does not extend so far forward as the inner the distal oral groove." Kellogg further submits that there is no distal oral groove in Mytilus, and he is correct on the basis of his purely anatomical definition. However, Stasek (personal communication) judges that, in the Mytilidae, a functional basis for a ctenidium-palp association is more valid than a purely anatomical one. While I agree with Stasek on this point, I think that his definition for Category I would be more convincing if it read as follows : ". . . the ventral tips of at least the first few or, usually, of many filaments of the inner demibranch are inserted unfused into an anatomical or functional distal oral groove. The stomach of A. falcata is well equipped to handle the fine sediment particles accompanying incoming food material. It differs from other mytilian rock borers, such as B. cinnamomea and L. gracilis (Dinamani, 1967), in that it possesses a well-developed sorting caecum. The folds and typhlosolar tongue of the caecum of A. falcata comprise the chief sorting mechanism of the stomach, as the antero- dorsal and shield tracts appear to be ineffectual. Sediment particles that get by the sorting mechanisms of the ctenidia and labial palps and are conveyed to the stomach of A. falcata are generally dealt with by the caecum and dumped into the intestinal groove to be carried out of the region of the stomach. The absence of a well-developed caecum in L. gracilis might be related to the chemical nature of its boring. Lithophaga spp. are usually found burrowed in a limestone substrate (Yonge, 1955) and during the boring process do not produce a particulate sedi- ment. Hence, they probably do not require the sorting mechanism found in the stomach caecum of A. falcata. Dinamani (1967) has suggested that in the Mytilidae the caecum functions more as a temporary reservoir of food than as a sorting mechanism. He observed that, in specimens of P. z'iridis starved for up to two days, the food matter within the caecum had disappeared. The caecum may have a purely storage function in mytilids where the sorting folds of the caecum are rudimentary, such as in B. cinnamomea, L. gracilis, and P. viridis, but in A. jalcata and M. cdulis, where the sorting mechanism is well-developed, the caecum is patently a device for resolving food matter from mucus and sediment particles. The complexity and functional significance of the appendix in the bivalve stomach ranges from a rudimentary, static groove in M. edulis (Reid, 1965) to the large food storage pouch found in the wood-boring Teredinidae (Purchon, 1968). Yonge (1949) submits that the appendix or antero-dorsal caecum in tellinids functions as both a storage cavity for sand grains to aid in trituration of food material, and a relief valve for excess food-mucus strings accumulating around the head of the crystalline style. The rejectory currents of the appendix of A. falcata precludes its functioning in a storage capacity. Moreover, I was never able to observe the overt abstraction of mudstone particles in any portion of the stomach. The appendix of A. falcata functions solely as a relief mechanism, and, instead of temporarily storing the bolus material pressed into it, the ciliary currents of the appendix convey this material to the intestinal groove, where it is subsequently removed from the stomach. The obvious advantage of the appendix to A. falcata is that the presence of a relief mechanism in the stomach prevents overeating and thus maintains the proper volume for optimum mixing of food with gastric juices. CILIARY CURRENTS OF ADULA FALCATA 43 The crystalline style from specimens of A. falcata maintained in aquaria often contained bits and pieces of food matter. Morton (1952) suggests that this is a common occurrence which is advantageous to the style bearer in that it permits recovery and digestion of food matter which would otherwise be lost as faeces. Nelson (1918) found similar structures in the styles of Anodonta grandis and Modiolns inodiolus. Nelson's explanation for this phenomenon was that, during the formation of the crystalline style, a clear mucus stream is secreted by the walls of the intestine and picked up by the ciliated surfaces of the typhlo- soles in the mid-gut. During the process of the mucus stream's being carried to the typhlosoles, it is twisted on itself and forms a corkscrew spiral. Frequently, odd particles of food are picked up by the twisted mucus stream and this accounts for the dark color of the spiral structure. Nelson further suggests that if an animal is starved, loses its style, and then suddenly starts feeding again, it usually generates a dark cored new style. In the case of A. falcata, continuous feeding with loss and subsequent regeneration of the style, as is probably the case in the subtidal zone as well as in aquaria, is the more common cause for the dark cored style being produced. Doubtless, as Morton (1952) has suggested, it is advantageous to the style bearer to recycle food material which would otherwise be lost, but on the other hand, the amounts of food involved in this process are probably too minute to be of any real importance to nutrition in the majority of the Bivalvia. Functionally, the crystalline style supplies most, but not all, of the digestive enzymes found in the lumen of the stomachs of style-bearing bivalves (Reid, 1968). According to Morton (1967), the style acts as a stirring rod and windlass as well. I found no evidence of the crystalline style of A. falcata functioning as a windlass for the string of mucus-bound food emerging from the posterior end of the oesophagus. The physics of Morton's observation suggest that there would have to exist some mechanism for increasing the rate of the food string moving down the oesophagus as the bolus of material around the style increased in diam- eter. Otherwise, as the first wrappings were going on the style, there would be a windlass mechanism, but as the bolus diameter increased, so would the rate of wrapping increase (if one can assume that the rate of the rod's turning is constant) and the food string would break from the strain. Another possibility is that, as the food bolus increases in diameter, increased friction with the walls of the stomach would act as a clutch, slowing the turning rate of the style and, thus, keeping the wrapping speed of food strings uniform. My observations on conditions associated with the acceptance or rejection of particulate material in A. falcata suggest that its sorting mechanisms are not highly efficient. However, permitting the passage of some sediment material into the stomach could be advantageous to the animal's well-being. The presence of sedi- ment matter in the gut may aid in trituration of food, as is the case in tellinids (Yonge, 1949). More likely, the consolidation of sediment particles into firm faecal pellets prevents reintroduction of that material into the mantle cavity or burrow. Reid and Reid (1969) report a sand-grain epiflora in the stomach of Macoma secta which comprises most of its nutrients. Since, volume for volume, fine sediment particles would possess a surface area proportionally higher than that of sand grains, it is possible that the small amount of sediment material in the stomach of A. falcata provides nutritional value as well. 44 PETER V. FANKBONER Adaptative radiation of the bivalve foot is illustrated most commonly in the function of locomotion. The creeping of the saddle oyster Enigmonia aenigmatica (Yonge, 1957), Lasaca rubra climbing a rock face via adhesion to a slime trail (Morton, 1960), and the leaping and digging-in of cockles are all executed by the bivalve's use of the foot. Within rock boring bivalves, if the foot is functional at all, it is utilized for stabilization of the visceral mass during boring, either by the laying-down of a byssus in the case of Adula, Lithophaga, and Tridacna (Yonge, 1955 and 1936) or in functioning as a sucker, gripping the head of the burrow, as reported by Purchon (1955b) for the Pholadidae. In A. falcata, the foot performs the additional function of cleansing its burrow walls of the participate mudstone produced during mechanical boring. I wish to express my gratitude to Dr. Charles R. Stasek, formerly of the California Academy of Sciences, for his advice and encouragement during the course of this study. Thanks are also due to Dr. E. H. Smith, Director of the Pacific Marine Station, for providing the facilities where much of this work was completed, and to Professor A. R. Fontaine, Professor G. O. Mackie, Dr. R. G. B. Reid, and Dr. R. A. Ring, of the University of Victoria, for comments and criticism of the manuscript. Excess figure charges were paid from a NRC operating grant A- 1427 to Dr. G. O. Mackie, Victoria, Canada. SUMMARY The mode of sediment removal by A. falcata,, and its relationship to the ciliary currents of feeding and digestion was determined. Where possible, morphological and functional comparisons were made with other rock-boring bivalves. 1. A. falcata moves sediment material from the walls of its burrow into its mantle cavity by means of the ciliated surfaces of its unusually extensible foot and the mantle folds. After entering the bivalve's mantle cavity, sediment particles are treated as potential food material by the ciliary sorting mechanisms associated with feeding and digestion. Most sediment matter is rejected from the burrow as pseudofaeces via extrusion through the inhalant siphon opening. However, small amounts of sediment particles find their way past rejectory mecha- nisms of the gills and labial palps, and pass into the stomach. Following sorting and subsequent rejection by the stomach, this material is incorporated into faecal pellets which are later swept from the area of the burrow by the strong currents of the exhalant siphon. 2. It is suggested that particle discrimination, during sorting by A. falcata, is dictated by the weight, size, and mucus-bound state of the material being sorted. LITERATURE CITED ATKINS, D., 1937a. On the ciliary mechanisms and inter-relationships of lamellibranchs. Part II : sorting devices on the gills. Quart. J. Microscop. Sci. 79 : 339-373. ATKINS, D., 1937b. On the ciliary mechanisms and inter-relationships of lamellibranchs. Part III : types of lamellibranch gills and their food currents. Quart. J. Microscop. Sci., 79: 375-421. DINAMANI, P., 1967. Variation in the structure of the stomach in the Bivalvia. Malacologia, 5(2): 225-256. CILIARY CURRENTS OF ADULA FALCATA 45 FANKBONER, P. V., 1967. The corrosion-vinyl acetate technique as an aid in the reconstruc- tion of the marine molluscan alimentary system. Veliger, 9(4) : 444-446. GRAHAM, A., 1949. The molluscan stomach. Trans. Roy. Soc. Edinburgh, 61(27) : 737-778. HUNTER, W. R., 1949. The structure and behavior of Hiatella gallic ana L. and H. artica L, with special reference to the boring habit. Proc. Roy. Soc. Edinburgh, 63B : 271-289. KELLOGG, J. L., 1915. Ciliary mechanisms of lamellibranchs with descriptions of anatomy. /. MorphoL, 26: 625-701. MORTON, J. E., 1952. The role of the crystalline style. Proc. Malacol. Soc. London, 29 : 85-92. MORTON, J. E., 1960. The responses and orientation of the bivalve Lasaca rnbra Montagu. /. Marine Biol. Ass. U.K., 39: 5-25. MORTON, J. E., 1967. Molluscs. Hutchinson University Library, London, 244 pp. NELSON, T. C, 1918. On the origin, nature, and function of the crystalline style of lamelli- branchs. /. MorphoL, 31(1): 53-111. OWEN, G., 1955. Observations on the stomach and the digestive diverticula of the Lamelli- branchia. Part I. The Anisomyaria and Eulamellibranchia. Quart. J. Microscop. Set., 96: 517-537. OWEN, G., 1967. Digestion. Pages 53-96 in K. M. Wilbur, C. M. Yonge, Eds., Physiology of Molhtsca, Volume II. Academic Press, New York. PURCHON, R. D., 1955a. The functional morphology of the rock-boring lamellibranch Petri- cola pholadiformis L. /. Marine Biol. Ass. U.K., 34: 257-278. PURCHON, R. D., 1955b. The functional morphology of the British Pholadidae (rock-boring Lamellibranchia). Proc. Zool. Soc. London, 124: 859-911. PURCHON, R. D., 1957. The stomach in the Filibranchia and Pseudo-lamellibranchia. Proc. Zool. Soc. London, 129: 27-60. PURCHON, R. D., 1968. The Biology of the Molhisca. Pergamon Press Ltd., London, 560 pp. REID, R. G. B., 1965. The structure and function of the stomach in bivalved molluscs. /. Zool., 147: 156-184. REID, R. G. B., 1968. The distribution of digestive tract enzymes in lamellibranch bivalves. Comp. Biochem. Physiol, 24: 727-744. REID, R. G. B., AND A. REID, 1969. Feeding processes of members of the genus Macoina (Mollusca: Bivalvia). Can. J. Zool., 47: 649-657. STASEK, C. R., 1963. Synopsis and discussion of the association of ctenidia and labial palps in the bivalved Mollusca. Veliger, 6(2) : 91-97. YONGE, C. M., 1936. Mode of life, feeding, digestion and symbiosis with zooxanthellae in the Tridacnidae. Sci. Repts. Great Barrier Reef Exped. (British Museum Nat. Hist.), 1: 283-321. YONGE, C. M., 1948. Cleansing mechanisms and the function of the fourth pallial aperture in Spisula subtruncata Da Costa and Lutraria lutraria L. /. Marine Biol. Ass. U.K., 27: 585-596. YONGE, C. M., 1949. On the structure and adaptations of the Tellinacea, deposit-feeding Eulamellibranchia. Phil. Trans. Roy. Soc. London Series B, 234: 29-76. YONGE, C. M., 1955. Adaptation to rock-boring in Botula and Lithophaga (Lamellibranchia, Mytilidae) with a discussion on the evolution of this habit. Quart. J. Microscop. Set., 96: 383-410. YONGE, C. M., 1957. Enigmonia aenigmatica Sowerby, a motile anomiid (saddle oyster). Nature, 180: 765-766. Reference : Biol Bull, 140 : 46-62. (February, 1071 ) TRANSPORT OF SUGARS IN THE TAPEWORM CALLIOBOTHRWM VERT1CILLATVM 1 F. M. FISHER, JR., AND C. P. READ Marine R'wloyical Laboratory, Woods Hole, Massachusetts, 02543, and Department of Biology, Rice University, Houston, Texas 77001 Since tapeworms have no digestive tract and no cavities in the body other than the so-called osmoregulatory canals and genital ducts, it has been assumed that nutrients enter the body through the outer surface. Further, several tape- worm species are known to require an external source of carbohydrate for growth and reproduction ; at least 14 species can metabolize no sugars other than glucose and galactose (Read and Simmons, 1963). Calliobothrium vcrticiUatum, a tetra- phyllidean cestode parasitizing the smooth dogfish, has been reported to absorb and metabolize glucose and galactose but not to utilize fructose, mannose, sucrose, lactose, trehalose, or maltose from the suspending medium (Read, 1957; Laurie, 1961). It seems highly probable that glucose, and perhaps galactose, is a required energy source for Calliobothrium. Although no data on the concentrations of free monosaccharides in the environment of the worm are available, it seemed desirable to study sugar absorption by the worm in vitro, with the view that sugar metabolism may limit growth and reproduction of Calliobothrium in its host. MATERIALS AND METHODS The hosts, Mustelus canis, were collected by a commercial fisherman and dis- tributed by the Supply Department of the Marine Biological Laboratory. The dogfish were maintained in large tanks with running sea water (18-22) until the time of an experiment. The holding period varied from one to three days. The hosts were killed by a blow on the chondrocranium and the spiral intestine quickly removed to a container immersed in ice. The intestinal valves were split longitudinally and the adult cestodes removed to a balanced salt solution also maintained at (250 mM NaCl, 4.4 mM KC1, 5.1 mM CaCL, 2.9 mM MgCl 2 , and 300 mM urea, buffered with 10 mM tris-maleate at pH 7.2, as described by Read, Simmons, Campbell and Rothman 1960). This solution is referred to herein as saline. The parasites were freed of adhering intestinal contents and mucus and sorted into groups of about eight to ten worms. Each sample was placed in 4 ml of saline in a 30 ml beaker. Beakers containing the parasites were preincubated in a shaking water bath at 20 for 30 minutes. Unless other- wise stated, all incubations were carried out at 20 C. At the end of an incubation, worms were rapidly removed from the incubation medium, quickly rinsed three times in saline, blotted on hard filter paper, and placed in 2 ml of 70% ethanol. Extraction of soluble materials was carried out at room temperature for at least 1 This work was supported by grants from the National Institutes of Health (GM 12263 and A I 01384). 46 SUGAR TRANSPORT IN TAPEWORM 47 18 hours with frequent agitation. Incubation times and other manipulations will be discussed in the context of the experiments. Wet weights of parasites were determined on a torsion balance. Dry weights and ethanol-extracted dry weights were obtained after drying the material to constant weight at 103 C in small tared aluminum cups. Estimation of parasite body water was made gravimetrically, i.e., wet weight minus dry weight. Isotopic methods Radiochemicals were obtained from New England Nuclear Corporation and Cambridge Nuclear Corporation. Radioactivity was determined on aliquots of ethanol extracts or incubation media using a low background gas flow counter (C 14 ) or a gamma ray spectrometer with a Nal crystal (K 42 and Na 24 ). Specific activities (c/m ^mole' 1 or c/m meq" 1 ) were determined by counting aliquots of various media after appropriate dilution with 70% ethanol. Chemical methods All unlabeled carbohydrates were purchased from commercial sources (Mann Research Laboratories, Pfanstiehl Chemical Co. and Sigma Chemical Co.). Polysaccharide and total ethanol-soluble carbohydrate were determined by the phenol-sulfuric acid method of Dubois, Gilles, Hamilton, Rebers and Smith (1956). Polysaccharide was insolated by digestion at 100 C in 30% KOH and precipitation with 1.2 volumes of 95% ethanol. For determination of the specific activity of the polysaccharide (yumoles 14 C-glucose incorporated//xmole glucose in polysaccharide), the precipitated material was washed five times with 70% ethanol containing 0.1% LiCl, by alternate dispersion and centrifugation. This was found to be necessary for the complete removal of ethanol-soluble radio- activity. The washed pellet was then dissolved in water and aliquots removed for the determination of radioactivity and total carbohydrate, using glucose or trehalose standards. Glucose was determined on aliquots of extracts or media by the glucose oxidase method ("Special Glucostat," Worthington Biochemical Corp.). The purified enzyme preparation was employed to minimize errors due to other carbohydrates in the experimental materials. Sodium and potassium were determined by flame photometry, using a Coleman Junior spectrophotometer with a flame attachment. Standard solutions contained equivalent amounts of each ion in the chloride form and ethanol at the same con- centration as samples. All inorganic reagents were obtained from Fisher Scientific Co. or J. T. Baker Chemical Co. RESULTS Effects of the gas phase Since carbon dioxide is known to affect carbohydrate metabolism in a number of animal parasites (von Brand, 1966; Prichard and Schofield, 1968; McDaniel, Read and Maclnnis, in press), it seemed advisable to determine whether the gas phase affected the absorption of glucose in Calliobothrwm. Data from such experiments are shown in Figures 1 and 2. When compared with the results 48 F. M. FISHER, JR., AND C. P. READ obtained in air, nitrogen-carbon dioxide had no significant effect on the absorption or accumulation of glucose, but had a dramatic effect on the incorporation of 14 C-glucose into glycogen. This effect is clearly due to the presence of carbon dioxide in the atmosphere since McDaniel et al. (in press) showed that under 30 60 minutes FIGURE 1. The absorption of "C-glucose by Calliobothrium under atmosphere of 95% Na-5% CO 2 . Bicarbonate was added to maintain pH at 7.2. Glucose in worms was determined by glucose oxidase () or by radioactivity (O). Glucose in the medium was determined by glucose oxidase (X). Inset's: Incorporation of 14 C-glucose into glycogen in same worm samples expressed as specific activity (micromoles I4 C-glucose per micromole of glycogen glucose). Each point in all curves is the average of worm samples. Compare with Figure 2. nitrogen or carbon dioxide-free air, incorporation of 14 C-glucose into the glycogen of Calliobothrium was markedly less than that observed when the gas phase contained carbon dioxide. Since absorption was not so affected, subsequent experi- ments were carried out in air. It may be pointed out that, under the conditions of the experiments shown in Figures 1 and 2, the kinetics of absorption were SUGAR TRANSPORT IN TAPEWORM 49 first order. Further, at concentrations which probably resemble those of the worm's environment, the worm clearly accumulates glucose against a concentra- tion difference (Table I). Later experiments were designed to determine whether or not this was a saturable system. 60 minutes FIGURE 2. The absorption of 14 C-glucose by Calliobothrinm under atmosphere of air Glucose in worms determined by glucose oxidase (O) or by radioactivity (). Similarly, glucose in the medium was determined by glucose oxidase (A) or by radioactivity (A). Inset is : Incorporation of I4 C-glucose into glycogen in same worm samples expressed as specific activity (micromoles I4 C-glucose per micromole of glycogen glucose). Each point in all curves is the average of two worm samples. Compare these data with Figure 1. Effect of temperature When worms were incubated with glucose at each of several temperatures for 60 minutes, sugar accumulation rose sharply with temperature to a maximum at about 20 C. Above 30 C, absorption sharply declined (Fig. 3). The data do not allow a very precise estimate of the optimum temperature for glucose accumulation, but the extent to which glucose is incorporated into glycogen appears to have a higher temperature optimum than that for glucose accumulation (Fig. 3). 50 F. M. FISHER, JR., AND C. P. READ TABLE I Effect of time on the accumulation of glucose by Calliobothrium. Incubation mixture contained 5 ml of 0.5 mM glucose in KRT. Results based on average of two determinations Incubation time (minutes) Glucose cone, inside (mM) Glucose cone, outside (mA/) Conc.i Conc.o 4.9 0.5 9.8 1 5.3 0.46 11.5 2 6.8 0.43 15.8 4 7.7 0.38 20.3 6 8.5 0.36 23.5 8 9.0 0.29 31.0 10 10.4 0.28 37.0 20 12.1 0.12 100.8 30 12.6 0.04 315.0 Effect oj pH The absorption of glucose was affected by hydrogen ion concentration; the highest rates were observed at about pH 8 to 9 (Fig. 4). There was a dramatic oo 50 FIGURE 3. Effect of temperature on the 60 minute accumulation of tissue glucose ( ) and on incorporation of 14 C-glucose into glycogen (O). Tissue glucose was determined by glucose oxidase. Specific activity is /mioles H C-glucose per /*mole of glycogen glucose. The medium contained 5 mM glucose at the beginning of the experiment. Each point is the average of two samples. SUGAR TRANSPORT IN TAPEWORM 51 pH 5.2 6.2 7-2 8.2 9.2 8 pH worm d i s t r. 15 10 anterior 4 6 Valve no. 8 post FIGURE 4. Effect of pH on rate of 14 C-glucose absorption in 2 min incubations (X) and the pH of intestinal contents in different segments of the dogfish spiral intestine (O). Worm samples were preincubated for 60 minutes in buffered saline at the appropriate pH and then incubated with 0.01 mM 14 C-glucose at the same pH as the preincubation medium. The V = icimoles per gram dry wt per hr. Normal worm distribution in the spiral intestine is indicated. TABLE II The effect of various compounds on the transport of glucose by Calliobothrium. Inhibitor concentration, 5 mM ; substrate concentration, 0.05 mM.: Incubation time 2 min Inhibitor % Inhib. No inhibitory effect Glucose 75.6 n-acetyl-glucosamine alanine Arbutin 73.2 D-glucosamine arginine Maltose 59.9 lactose aspartic acid a-Methylglucoside 58.6 fructose glycine Ga lactose 44.2 mannose leucine Cellobiose 25.2 3-0-methylglucoside methionine Salicin 23.2 rhamnose proline Ouabain* 24.5 sorbose ta urine Phlorizin** 96 sucrose trimethylamine valine * Present at 2 mM. ** Present at 0.01 mM. 52 F. M. FJSHER, JR., AND C. P. READ drop in the entry kinetics between 7.2 and 6.2. Determination of the pH in various segments of the spiral intestine of Miistclus revealed that there is a pH gradient. In the region of the first spiral, the pH is about 6 and rises to a pH of about 8 in the ninth spiral. These data are shown in Figure 4 and the observed distribution of Calliobothrium strobilae in the spiral intestine is shown J_ s FIGURE 5. Lineweaver-Burke plots of 14 C-glucose absorption in 2-min incubations without inhibitor (a) or with 0.01 mM phloretin (b) or 0.01 mM phlorizin (d). Samples of curve c were incubated for 60 minutes in 0.01 mM phlorizin, rinsed, and incubated for 2 min with 14 C-glucose. Each point is average of two samples. Abbreviations are : V = /orioles per gram per hour ; S = mM u C-glucose. for comparative purposes. The worm appears to inhabit those segments of the spiral intestine having a pH range which is highly favorable for the absorption of glucose by the parasite. It would be of interest to examine the effect of pH on the absorption of glucose by free segments of Calliobothrium, since these are normally found in the first two spirals where the pH is relatively low. Attempts to make such determinations have proven to be technically difficult. SUGAR TRANSPORT IN TAPEWORM 53 Effects of inhibitors In 2 min incubations, 5 mM 2,4-dinitrophenol had no effect on the rate of glucose transport. However, when the worms were preincubated for 10 min in 5 mJl/ dinitrophenol, glucose absorption was inhibited 30% in a subsequent 2 min incubation without dinitrophenol. Sodium iodoacetate at 0.5 mM produced a 30% inhibition of glucose transport in 2 min incubations. 80 40 600- 200 20 40 J_ S FIGURE 6. Lineweaver-Burke plots of I4 C-glucose absorption in 2-min incubations in the presence of galactose (A), maltose (D), a-methylglucoside (A), ouabain (O), or without inhibitor (). Inhibitors were present at 2 mM. Points for curves with galactose, maltose, and a-methylglucoside as inhibitors are averages of two determinations while those of the ouabain-inhibited and uninhibited glucose absorption curves are individual determinations. Inset is: Curves for absorption of glucose alone (a), glucose in the presence of 2 mM maltose (b), and glucose in the presence of 2 mM ouabain (c). Data points from the lower curves were used to construct inset curves. Abbreviations are : S = mM glucose ; V = ^uncles absorbed per gram dry weight per hour. Twenty-eight additional compounds were examined as inhibitors of glucose transport (Table II). Of those tested, eight produced significant inhibitions. When these same compounds were tested as stimulators of the efflux of pre- viously accumulated glucose, only those which inhibited glucose absorption caused an enhanced leakage of glucose. Further, sodium a-ketoglutarate, /3-glycerophos- phate, or dithiazanine had no effect on efflux of previously accumulated glucose. 54 F. M. FISHER, JR., AND C. P. READ r 60 I CuO 40 OuO E o 100 mM NaCI 200 FIGURE 7. Effect of Na + on the absorption of l4 C-glucose in 2-min incubation. Deleted NaCI was replaced by KC1. Glucose was present at 0.2 mM in all incubates and each point is average of two samples. The latter compound is an anthelminthic drug which blocks sugar transport in the isolated gut of Ascaris (Fisher, unpublished) but has no effect on glucose uptake or efflux in Calliobothrium. Phlorizin, which is known to be a strong inhibitor of glucose transport in other tapeworms (Phifer, 1960; Laurie, 1961), as well as various other cells and tissues, was the most powerful inhibitor of glucose transport in Calliobothrium. The inhibition appeared to be competitive in character in 2 min incubations (Fig. 5). However, when worms were incubated for 60 min in the presence of 0.01 mM phlorizin and rinsed in phlorizin-free saline, the inhibition of glucose uptake was only partially reversed (Fig. 5c). The effect of phlorizin is apparently dependent on the intact glycoside since the aglycone, phloretin, produced a negli- gible inhibition (Fig. 5b). Ouabain, a cardiac glycoside, produced a significant inhibition of glucose transport, but the glycone moiety of ouabain, rhamnose, was not inhibitory (Table II)., Two other glycosides, arbutin and salicin, had significant activity as inhibitors of glucose transport (Table II). The inhibitions produced by galactose, a-methylglucoside, maltose, and ouabain were examined in some detail. Lineweaver-Burke plots of the data indicated that the inhibitions are competitive in character (Fig. 6). TABLE III Effect of glucose entry on the ionic composition and wet weight of Calliobothrium Incubation time (minutes) K + meq. I" 1 Na + meq. I" 1 Wet weight increase (mgm%) 87 172 30 79 203 7.3 60 73 221 13.5 90 71 229 19.5 120 70 234 32.2 SUGAR TRANSPORT IN TAPEWORM 55 Na and K Effects Since sodium has been implicated in the transport of sugars and amino acids in other animal cells (Stein, 1967), it was deemed desirable to determine whether sodium is involved in glucose transport in Calliobothriiiin. \Yhen K + was sub- stituted for Na + in the saline, the rate of glucose transport was a linear function of Na concentration (Fig. 7). Further, it was found that tissue sodium increased when glucose was being absorbed, as might be expected in the co-transport of sodium and glucose. The increase in sodium was accompanied by less than compensatory decrease in tissue potassium and by an increase in tissue water (Table III). Control worms incubated in saline without glucose showed no significant changes in tissue sodium, potassium, or wet weight. 400 200 012345 m M ouabain FIGURE 8. The effect of various concentrations of ouabain on the absorption of 5 mM 14 C-glucose in 2-min incubations of Calliobothrinm. The V = jamoles absorbed per gram dry weight per hour. As previously noted, the glycoside ouabain inhibits glucose transport in Calliobothriiim. In 2 min incubations, the inhibition increases linearly between and 5 mM ouabain (Fig. 8). This inhibitor is known to act on sodium trans- port in a number of animal cells, blocking the sodium extrusion mechanism. In Calliobothrium, ouabain causes a marked net influx of sodium (Fig. 9). Exposure to ouabain causes the tissue sodium to rise from about 180 to 300 meq/1 in 30 minutes. This is accompanied by an enhanced net efflux of potassium which is not equivalent to sodium influx. Glucose partially alleviates the effect of ouabain on the fluxes of both sodium and potassium. It may also be noted in Figure 8 that ouabain inhibits the net accumulation of glucose by Calliobothrium, reducing it by about 63%. If ouabain acts on the system involved in maintaining sodium at internal concentrations which are lower than that of the ambient medium, it might be expected to affect efflux of sodium from the worm. To test this hypothesis, worms were equilibrated for 60 min in saline containing Na 24 , following which the worm samples were transferred repetitively at 5 min intervals to vessels 56 0> GO 30 20 10 300 200 o> 100 F. M. FISHER, JR., AND C. P. READ glucose - - -* glucose 4 ouabain ouabain lucose -i- ouabain glucose ouabain 40 minutes 120 FIGURE 9. Effect of ouabain on Na + and K + net fluxes and on the accumulation of glucose. Glucose and ouabain at concentration of 5 mM. All values are expressed in terms of tissue water and each point is average of two samples. Controls incubated without glucose and/or ouabain showed no net change in Na + or K + levels during the same time period. containing saline with or without 5 mM ouabain. The media were then assayed for Na 24 . Results of such an experiment are shown in Figure 10. The data indicate that sodium effluxes from at least two compartments. Ouabain appears to exert an effect on one compartment and not on the other. Efflux from the SUGAR TRANSPORT IN TAPEWORM 57 10 9 8 7 6 5 4 O. o 20 40 60 mi nutes FIGURE 10. Effect of ouabain on efflux of Na + . Worms were equilibrated for 60 min in saline containing Na 24 . Efflux was then measured for 60 min in the presence (O) or absence ( ) of 5 mM ouabain. See text for experimental details. "fast compartment" is clearly inhibited by ouabain, wbereas efflux rates from the other compartment appear to be the same for ouabain-treated and control samples. Accumulation of other hc.roses Since glucose was accumulated against a concentration difference, it was of interest to determine whether some other hexoses are accumulated by Calliobo- thriitm and to examine the relative rates at which these hexoses might be incorp- orated into parasite polysaccharide. Such data are presented in Table VI. Of those examined, galactose and glucose are accumulated against a concen- tration difference. Although a specific galactose oxidase was not employed to examine the levels of galactose in the worm, the amount of non-glucose, ethanol- soluble carbohydrate (Table IV, column d) is about four times that found in worms incubated with other sugars. It is concluded that a large proportion of this non-glucose carbohydrate is indeed galactose. Although galactose is trans- ported, very little of this sugar is incorporated into polysaccharide (Table IV, column g). Mannose and fructose are not transported and, as anticipated, neglig- ible amounts of label from these sugars are incorporated into polysaccharide. Additional experiments on the absorption of 3-0-methylglucose showed that the rate is a linear function of sugar concentration and that it is neither accumulated against a concentration difference nor metabolized by the worm. 58 F. M. FISHER, JR., AND C. P. READ The data indicate that 3-0-methylglucose enters the worm by diffusion and sub- stantiates the finding that this sugar does not react with the glucose transport sys- tem. DISCUSSION The outer surface of the syncytial tegument of Calliobothrium is elaborately differentiated for absorptive function, as is the case in all cestodes whose tegumentary ultrastructure has been studied (Lumsden, 1966). As we have shown, glucose is absorbed at high rates by the worm and, at concentrations which probably fall in the range normally encountered by the worm in its environment, Calliobothrimn accumulates glucose against a concentration differ- ence of up to 315 fold in a 30 minute period. The capacity of the worm to TABLE IV Accumulation of radioactive hexoses by Calliobothrium and incorporation into polysaccharide; incubation time: 60 min, incubation medium: 4.0 ml of 5.0 mM. substrate; average of 2 samples /hexose; gas phase: 95% N^-5% CO* Subst. C'< ("mM") (a) Glue. (mM) (b) EtOHsoi. CHO ("mM") (0 Non Glue. CHO ("mM") (d) Medium Glue. (mM) (e) Medium Total CHO ("mM") (f) Sp. act. Polysac. (g) Glue 46.78 31.23 38.60 7.37 3.84 3.94 0.1248 Gala 66.37 11.27 59.89 48.62 2.00 0.0011 Mann 2.16 12.94 24.67 11.73 5.01 0.0004 Lev 1.02 11.55 23.61 11.61 5.01 0.0001 8.87 23.51 14.64 (a) Based on specific activity of medium. (b) Determined by glucose oxidase. (c) Determined by phenol : H 2 SO4. (d) Determined by difference in, c b. (e) Determined by glucose oxidase. (f) Determined by phenol: H 2 SO 4 . (g) /iMoles C 14 substrate incorporated/jumole hexose in polysaccharide. accumulate glucose is not affected by the presence of carbon dioxide or bicarbonate in the medium, although carbon dioxide clearly affects the rate at which glucose is incorporated into glycogen. This is similar to results obtained with the cyclophyllidean tapeworm Hymenolepis dimimita (Fisher and Read, unpublished data). The temperature optimum for glucose absorption is perhaps slightly higher than that normally prevailing in the marine environment of the host. However, the body temperature of an active dogfish may be slightly higher than that of the surrounding sea water and it seems probable that the worm in its host would absorb glucose at something approaching the maximum rate at a given concentra- tion. There is a high correlation between the pH in that part of the spiral intestine inhabited by Calliobothrium and the pH values at which maximum rates of glucose absorption were observed. In the anterior part of the spiral intestine, SUGAR TRANSPORT IN TAPEWORM 59 not inhibited by the strobila of Calliobothriiim, the prevailing pH would reduce glucose absorption by 30 to 35%. This might be of significance in allowing normal function of the worm. On the other hand, the free segments of Callio- both riii in which are in varying stages of shelled egg development are typically found in the two anterior valves of the spiral intestine. The energy requirements of the free segments may be considerably less than the requirements of the strobila, since active tissue growth must be occurring at lower rates. The inhibition produced by dinitrophenol clearly does not occur at the surface since it produced no inhibition in 2 min incubations. It may produce its effect over longer periods of exposure by general interference with water and salt move- ments, lodoacetate, on the other hand, is a very non-specific poison and seems to produce an effect on the initial rates of glucose absorption probably combining with the carrier itself. Of other potential inhibitors tested, only a limited number of hexoses and glycosides were inhibitory. None of the amino acids and several sugars were not inhibitory. Of the inhibitory sugars, five were found to produce inhibition in a competitive manner. The others were not examined with respect to this point. However, the observed competitive inhibitions of glucose uptake by other sugars ; including the lack of inhibition by 3-0-methylglucoside and the inhibition by cellobiose and maltose suggest that the specificity of the site for glucose attachment to the transport mechanism in Celliobothrium differs from that in hamster mucosal cells (Crane, 1960) and in the tapeworm Hymenolepis diininuta (Read, 1961). Study of possible inhibitory effects of larger series of sugars, particularly hexoses, is required for definition of the points of glucose attachment in the transport mechanism. The lack of interaction by 3-0-methylglucose suggests that the third carbon on the hexose molecule is involved in glucose interaction. The finding that 3-0-methylglucose appears to enter the worm by diffusion, rather than by a mediated process, independently supports the concept that this sugar cannot react with the glucose transport system. The finding that maltose and cellobiose inhibit glucose uptake must be inter- preted with caution. It would be unwise to assume that the inhibitory effects are produced by the intact disaccharides. It is well known that disaccharidases are membrane-bound and act at the mucosal cell surface in the vertebrate intestine (Crane, 1968). The tapeworm surface has been described as a digestive-absorp- tive surface and, in the case of Hymenolepis diininuta, has been shown to release the products of hexose phosphate hydrolysis in the external medium (Arme and Read, 1970; Dike and Read, in press). Lumsden, Gonzalez, Mills and Viles (1968) also described an ATPase acting at the outer boundary of the tegument in H. diminuta. Further, the rabbit tapeworm Cittotaenia has a membrane-bound disaccharidase which hydrolyzes sucrose, liberating fructose into the external medium (Read and Rothman, 1958). Hence, there is a possibility that the apparent inhibition of glucose transport in CaUiobothrium by maltose and cello- biose might be due to effects of glucose liberated by the action of intrinsic membrane-bound disaccharidases situated in close proximity to the glucose transport system in the outer face of the CaUiobothrium tegument. The results of our experiments in which maltose in the external medium appeared to produce enhanced efflux of previously accumulated glucose might indicate the action of a 60 F. M. FISHER, JR., AND C. P. READ surface maltase. In studies to be published elsewhere, Fisher has shown that Calliobothrium does not secrete a maltase into the external medium. Clearly, the inhibition of glucose transport by maltose and cellobiose requires further study. Of the glycosides tested, phlorizin was the strongest inhibitor of glucose transport. Laurie (1961) reported that phloretin, the aglycone moiety of phlorizin, was a strong inhibitor of glucose absorption. We have not verified this. How- ever, Laurie's experiments were of 1 to 4 hour duration, while those of the present work were 2 min incubations. Phloretin may affect glucose accumulation rather than the initial rate of glucose transport in Calliobothrium. The changes in tissue sodium occurring with glucose absorption strongly suggest that glucose and sodium enter this worm by co-transport. Since sodium seems to activate glucose transport as an arithmetically linear function of sodium concentration, it may be concluded that reaction of only one sodium ion is required to activate the system for the transport of a glucose molecule. The data also suggest that a change in transmembrane electric potential should accompany glucose transport in Calliobothriinn, but this has not been measured. The effects of ouabain on sugar transport and sodium fluxes furnish additional evidence that sodium is involved in glucose transport and glucose accumulation. The results seem to be completely consistent with Crane's view that glucose accumula- tion in vertebrate cells is attributable to the maintenance of low intracellular sodium, with a resulting lower affinity of the membrane carrier for sugar on the intracellular aspect of the cell boundary (Crane, 1965). In the present study, ouabain has been shown to cause an increase in tissue sodium and a decrease in glucose accumulation. The effect on sodium levels seems to involve inhibition of the sodium extrusion process since sodium efflux is clearly inhibited. The partial reversal of the ouabain effect on the net influx of sodium cannot be readily ex- plained and would merit further investigation. While ouabain may inhibit glucose accumulation through effects on sodium extrusion, with a consequent lowering of the external/internal sodium ratio, there remains a problem in explaining the inhibition of glucose uptake by ouabain in 2 min incubations. This inhibition of the initial rate of glucose absorption seems to be competitive in nature, and it seems probable that ouabain interacts directly with the transport system. It might be suggested that ouabain might be hydrolyzed with the release of a competitively inhibitory sugar. However, we found the glycone moiety, rhamnose, to be without effect on glucose absorption. Thus, we may postulate that, in Calliobothrium, ouabain inhibits sugar transport directly by interacting with the membrane carrier and inhibits sugar accumulation indirectly by inhibiting the sodium extrusion mechanism. Further work is required to substantiate this interpretation. Sodium in the external medium is required for glucose absorption by both larvae and adults of the cyclophyllidean tapeworm, Taenia taeniaeformis. In the absence of sodium, no glucose transport occurs and, at 12 and 46 mM sodium, glucose absorption is reduced about 70 and 25%, respectively (von Brand and Gibbs, 1966). If the response to sodium concentration is linear, as is the case with Calliobothrium, about 60 mM sodium would be required to maintain glucose transport at its highest level in T. taeniaeformis. This has not been examined experimentally, although von Brand and Gibbs (1966) found that in media con- taining 115 mM sodium, glucose absorption equaled that observed in media con- SUGAR TRANSPORT IN TAPEWORM 61 taining 150 mM sodivun. Hymenolepis diminuta also requires sodium for glucose transport. This will be discussed elsewhere (Fisher, in preparation). The data on accumulation of various sugars by Calliobothrium verifies earlier reports (Read, 1957; Laurie, 1961) that this worm does not metabolize mannose or fructose. The worm is almost impermeable to these sugars. In the case of galactose, the worm seems to accumulate the sugar to higher levels than glucose can be accumulated. Calliobothrium has been reported to ferment galactose at rates almost equivalent to the rates of glucose fermentation (Read, 1957; Laurie, 1961 ). Thus, it may seem surprising that virtually no 14 C-galactose was incorporated into the polysaccharide of the worm. In a 60 min period, incorporation of 14 C-glucose into polysaccharide was 125-fold that of galactose. However, galactose is not glycogenic in the cyclophyllidean tapeworm Hymenolepis diniinuta. Although starved H. diminuta showed a dramatic net glycogenesis when furnished with glu- cose, no significant net glycogenesis occurred when galactose was available (Read, 1967). Like Calliobothrium, H. diminuta transports galactose into the tissues by a specific mediated system and ferments this sugar, but galactose is not glycogenic. A taxonomic note Euzet (1954) reported that he had examined Linton's (1891) tapeworm material collected from dogfish at Woods Hole. Linton had identified these worms as Calliobothrium eschrichtii Beneden, but Euzet concluded that Linton's material represented a new species which he named Calliobothrium lintoni. However, the animal studied in the present investigation (as well as in previous studies by Read, 1957; Read et al, 1960; Simmons, Read and Rothman, 1960; Laurie, 1961) is indeed Calliobothrium verticillatum (Rudolphi). We have found C. lintoni on several occasions in Mustelus canis. It is too small an organism for convenient physiological research. SUMMARY 1. Gas phase had no significant effect on glucose absorption or accumulation by Calliobothrium. 2. Optimum temperature for glucose accumulation is about 20 C. 3. Glucose transport was affected by pH, highest rates being observed at 8 to 9. This corresponds to the pH measured in that part of the spiral intestine inhabited by the strobilate worm. 4. Those sugars or glycosides which inhibited transport of glucose also stimulated efflux. Inhibitions produced by phlorizin, ouabain, galactose, maltose, and a-methylglucoside were found to be competitive in 2 min incubations. 5. The rate of glucose transport was a function of Na + in the external medium. Tissue Na + increased and K + decreased during glucose absorption. 6. Ouabain caused a net influx of Na + and this effect was attributed to the observed inhibition of Na + efflux. 7. Galactose was also accumulated by the worm but a negligible amount was incorporated in polysaccharide. Mannose, fructose, and 3 0-methylglucoside were not transported. 8. The data are discussed in terms of the specificity of the glucose transport mechanism. It is hypothesized that ouabain may inhibit two processes involved in glucose transport and accumulation. 62 F. M. FISHER, JR., AND C. P. READ LITERATURE CITKI) ARME, C., AND C. P. READ, 1970. A surface cn/yinc in 1 1 \ntcno1cpis diminuta (Cestoda). /. Parasitol, 56: 514-516. BRAND, T. VON, 1966. Biochemistry of Parasites. Academic Press, New York, 429 pp. BRAND, T. VON, AND E. GIBBS, 1966. Aerobic and anaerobic metabolism of larval and adult Tacnia tacniacjorinis. III. Influence of some cations on glucose uptake, glucose leakage, and tissue glucose. Proc. Helminthol. Soc. Wash., 33: 1-4. CRANE, R. K., 1960. Intestinal absorption of sugars. Physiol. Rev., 40: 789-825. CRANE, R. K., 1965. Na + -dependent transport in the intestine and other animal tissues. Fed. Proc. 24: 1000-1006. CRANE, R. K., 1968. A concept of the digestive absorptive surface of the small intestine. Pages 2535-2542 in C. F. Code, Ed., Handbook of Physiology. Alimentary Canal, Sect. 6, Vol. 5. American Physiological Society, Washington, D. C. DIKE, S. C., AND C. P. READ, 1970. Tegumentary phosphohydrolases of Hymenolepis diminuta. J. Parasitol., in press. DUBOIS, M., K. A., GILLES, J. K. HAMILTON, P. A. REBERS AND F. SMITH, 1956. Colori- metric method for determination of sugars and related substances. Anal. Chem., 28 : 350-356. EUZET, L., 1954. Quelques especes du genre Calliobothrium Van Beneden, 1850 (Cestoda: Tetraphyllidea). Bull. Soc. Ncuchatel. Sci. Natur., 77: 67-79. LAURIE, J. S., 1961. Carbohydrate absorption in cestodes from elasmobranch fishes. Comp. Biochem. Physiol., 4: 63-71. LINTON, E. (1891). Notes on entozoa of marine fishes of New England, with descriptions of several new species. Part II. U. S. Comm. Fish Fish., 1887: 719-895. LUMSDEN, R. D., 1966. Cytological studies on the absorptive surfaces of cestodes. Z. Para- sitcnk., 27: 355-382. LUMSDEN, R. D., G. GONZALES, R. R. MILLS AND J. M. VILES, 1968. Cytological studies on the absorptive surfaces of cestodes. III. Hydrolysis of phosphate esters. /. Para- sitol., 54: 524-535. MCDANIEL, J. S., C. P. READ AND A. J. MAC!NNIS, 1971. Some effects of carbon dioxide on glycogenesis in flatworms. /. Parasitol., in press. PHIFER, K. O., 1960. Permeation and membrane transport in animal parasites: On the mechanism of glucose uptake by Hymenolepis diminuta. J. Parasitol.. 46: 145-153. PRICHARD, R. K., AND P. J. SCHOFIELD, 1968. Phosphoenolypyruvate carboxykinase in the adult liver fluke, Fasciola hepatica. Comp. Biochem. Physiol., 24 : 773-785. READ, C. P., 1957. The role of carbohydrates in the biology of tapeworms. III. Studies on two species from dogfish. Exp. Parasitol., 6: 288-293. READ, C. P., 1961. Competitions between sugars in their absorption by tapeworms. /. Para- sitol., 47: 1015-1016. READ, C. P., 1967. Carbohydrate metabolism in Hymenolepis (Cestoda). /. Parasitol., 53: 1023-1029. READ, C. P., AND A. H. ROTHMAN, 1958. The role of carbohydrates in the biology of cestodes. VI. The carbohydrates metabolized in vitro by some cyclophyllidean species. Exp. Parasitol, 7: 217-223. READ, C. P., AND J. E. SIMMONS, JR., 1963. Biochemistry and physiology of tapeworms. Physiol. Rev., 43: 263-305. READ, C. P., J. E. SIMMONS, JR., J. W. CAMPBELL AND A. H. ROTHMAN, 1960. Permeation and membrane transport in animal parasites : Studies on a tapeworm-elasmobranch symbiosis. Biol. Bull, 119: 120-133. SIMMONS, J. E., JR., C. P. READ AND A. H. ROTHMAN, 1960. Permeation and membrane transport in animal parasites : Permeation of urea into cestodes from elasmobranchs. /. Parasitol., 46: 43-50. STEIN, W. D., 1967. The Movement of Molecules across Cell Membranes, Theoretical and Experimental Biology, Vol. 6. Academic Press, New York, 369 pp. Reference : Biol. Bull., 140 : 63-72. (February, 1971) POPULATION AGE STRUCTURE, GROWTH AND LONGEVITY OF THE MARINE GASTROPOD UROSALPINX CINEREA SAY * 2 DAVID R. FRANZ Biological Sciences Group, University of Connecticut, Starrs, Connecticut 06268 Urosalpinx cinerea Say is a West Atlantic temperate gastropod which is dis- tributed on the North American coast from Nova Scotia to Nassau Sound, Florida. The species also occurs in the British Isles where it was probably introduced with consignments of the American oyster, Crassostrea virginica Gmelin (Orton, 1927). U. cinerea preys on a wide variety of marine invertebrates (Carriker, 1955; Wood, 1968) and in localities where the oyster is cultured, destruction of young oysters may be a serious commercial problem. This species has been the subject of much research but information on its growth and longevity is incomplete, particularly for populations in North America. Information on the size of U. cinerea from different localities is available, notably the work of Cole (1942), Stauber (1943), Walter (1910), Federighi (1931a) and Myers (1965) and certain workers have attempted, unsuccessfully, to cor- relate size with temperature (Federighi, 1931b; Fraser, 1931). The problems in- volved in this type of correlation have been discussed by Carriker (1955) and Chestnut (1955). The only detailed studies of growth of U. cinerea are based on populations from English waters (Cole, 1942; Hancock, 1959). Both authors assessed growth by size frequency analysis and Cole made use of growth interruption marks on the siphonal canal in evaluating the individual components of his polymodal size frequency curves. Andrews (1955) and others have commented that no evidence was provided to support the assumption that the interruption marks on the siphonal canal were annual growth marks (annuli). Andrews also noted that Cole's analysis lacks data on young U. cinerea, up through two years, owing to the dif- ficulty in using an oyster dredge to collect small snails. Furthermore, Cole failed to provide evidence supporting his contention that the smooth curves fitted to the size frequency distributions represent year classes. Hancock's (1959) investi- gation included an analysis of size distribution of a population from the River Roach (Essex) as well as measurements of snails hatched and reared in the laboratory. By combining both sets of data, he was able to propose a cumulative growth curve for the population. This investigation was undertaken to provide information on the age struc- ture and growth of a North American population of U. cinerea (Southern New England) as a basis for future comparison with populations from other locations and ecological situations. 1 Contribution No. 69 of the Marine Research Laboratory, University of Connecticut, Noank, Connecticut. 2 This work was supported by the Research Foundation of the University of Connecticut, Grant No. RF-117. 63 64 DAVID R. FRANZ METHODS The subject of this study is a population of U. cincrea from the Mystic River, near Noank, Connecticut. The habitat is a rocky intertidal shore formed by a railroad embankment. The snails occupy approximately the lower third of the littoral zone where they occur on the sides and undersurfaces of rocks. Because of its location just within the mouth of the Mystic River, this shore habitat is well protected from wave action. The yearly salinity range is approximately 29- 31 ppt and the seasonal temperature range is from -1-25 C. Ice forms on the shore in late winter. Sampling was begun in June 1967 but for reasons noted below, most of this study is based on material collected in October, 1968, plus six samples from May through October, 1969. A total of approximately 2500 oyster drills was used in this analysis. The method followed in collection of the snails was as follows: collecting began about 30 minutes prior to LW and was continued for one hour. All snails were collected by hand by picking snails from individual rocks. Every rock selected was examined carefully to be certain that all snails were removed, thus reducing sampling bias in favor of large snails. The same stretch of shore and the same vertical level was sampled each time. Snails were taken to the laboratory for measurement and returned to the sampling location, usually the same day. In October, 1969, a supplementary sample of only larger drills was taken in order to increase the sample size of snails older than one season. Shell height (distance from the tip of the siphonal canal to the top of the spire) to the nearest 0.1 mm was measured with vernier calipers, but snails under 8 mm were measured with a calibrated stereomicroscope. In addition to shell height, the shell weights were determined for one sample (October 1969) as follows : the snails were treated with concentrated potassium hydroxide for one week, rinsed several times with tap water and flushed with a pasteur pipette, over dried overnight at 65 C and weighed to the nearest 0.01 g. The method of determining growth and longevity in this study is by size frequency analysis of both shell length and weight. An attempt has also been made to interpret and correlate growth interruption lines on the shell with the size frequency data. Difficulties encountered in this type of correlation in Mollusca have been discussed by Haskin (1954) and Wilbur and Owen (1964). A particularly vexatious problem in U. cinerea is caused by the extended reproduc- tive period and consequent large range in size of each year class. The extensive overlapping between year classes, in conjunction with the lower growth rates of older oyster drills, makes the identification of older year classes always difficult and sometimes impossible. In the present investigation the problem is reduced by the simultaneous analysis of both shell length and shell weight. RESULTS Shell length frequency graphs for eight sampling dates beginning September 1967 are shown in Figure 1. The September 1967 population is polymodal but the lower peak, below about 8 mm, represents the 1967 year class which can be traced from its first appearance in the sample taken in July, 1967 (not shown). Sometime between September 1967 and the following June, an event important to the subsequent success of this study occurred, namely, the apparent disappear- GROWTH OF UROSALPINX CINERI : .. I 65 ance of most of the 1967 year class. The cause remains unknown but the absence of this class is very evident in the June 1968 sample. The loss of most of this year class seems not to have seriously affected the population, as evidenced by the successful recruitment of the 1968 year class (Fig. 1, October 1968). The significance of the loss of the 1967 class becomes obvious in the May and June samples of 1969. In these, the total population is broadly bimodal. The lower Sept. 1967 10 20 Shell Length (mm) 25 FIGURE 1. Size distribution of Urosalpin.r from Noank, Connecticut, September 1967 through October, 1969. Numbers above the components of the distribution indicate the year of hatching. (left) component clearly represents the previous (1968) year class; the upper component, a mixture of year classes including 1965 and 1966. The largely vacant space between these two components would have been occupied by the 1967 year class. Actually, a remnant of the 1967 year class is present and as shown below, can be demonstrated by the analysis of shell weight. In any case, the absence of most of the 1967 year class allows a more precise characterization of the 1968 class than would otherwise have been possible. In the July sample, the 1969 year class makes its first appearance. By August, the entire population consists of three components, the new 1969 class at the left, the 1968 class occupying the center of the distribution, and a component of one 66 DAVID R. FRANZ or more year classes older than 1968 at the right. Note that beginning in May of 1969, the 1968 class is persistently polymodal although this was not evident the previous October. The reason for this is not known but since there can be no doubt that only a single year class is involved, this year class has been treated as a single unit. Upgrowth of the 1969 class is evident in the July through October samples. Growth of the 1968 class also continues so that by October 1969, the 1968 class merges with the remnants of the older classes. Combined Somple Oct. 1969 40- 30- o c CO 20^ ^~ ^ ^" ^~ ^~ - 5J- I s - ly 1 '' ro O") Cd CO m CO CO CO ro 1 ro i ro D ro i> cr> CO 1 1 IT) CO CO CO r- CO O ro ro ro to rO Shell Weight (Grams) FIGURE 3. Shell weight distribution of supplementary 1969 sample. Numbers above the modes indicate the probable years of hatching. and a smaller number of even larger snails which may or may not be survivors of an older year class. It is evident from Figures 2 and 3 that the analysis of shell weight permits the discrimination of several size components not evident from shell length frequency alone. Growth annuli To find supporting evidence for the growth data provided by the size analysis, the drills comprising the supplementary October 1969 sample were examined for the presence of annual growth interruption marks (annuli) on the shell, the underlying assumption being that snails examined in October at the end of their first growing season should lack annuli. Snails completing their second growth season should have a single annulus, after three seasons a second annulus should 68 DAVID R. FRANZ be present, and so on. The problem is to discriminate between a true annual growth interruption mark and a mark resulting from some other cause such as damaged outer lip. Consequently, the decision as to whether a mark is indeed a true annulus is highly subjective and in many cases impossible. Drills which were damaged or in which the cause of the interruption was in doubt were not included, so that data in Table I represent 168 out of an initial 224 snails. It is evident from the table that there is a close correlation between the mean weight of snails in each of the first four growth mark categories and the mid- points of weight components believed to correspond to the 1966 through 1969 year classes (Fig. 3). The correlation between the range in weight of each of the in E o i_ C9 I 4- 3' 2 I - 1 1 1 I I I I 10 15 20 30 40 Length (mm) FIGURE 4. Double logarithmic plot of shell length and shell weight, pooled October 1969 sample. Curve is fitted visually. growth mark categories and the range of their respective year class components is also close, with the exception of the group of snails with two growth marks. The excessive scatter in this group probably results from failure to discriminate between true and false annuli. This is particularly difficult in older drills because of erosion of the spire and partial masking by subsequent shell growth of a portion of the earlier annuli. Cumulative growth Figure 5 is a cumulative growth curve based on the analysis of both shell weight and length. When the cumulative shell length is plotted against age, the expected sigmoid growth curve is produced with a point of inflection near GROWTH OF UROSALPINX CINEREA 69 TABLE I Weight frequency data for U. cinerea possessing growth annuli Midpoint of weight interval (grams) (Modal groups are italicized) No. snails per growth interruption category "0 marks" "1 marks" "2 marks" "3 marks" "4 marks" 0.07 7 0.22 ' 21 3 0.37 3 3 1 0.52 7 0.67 2 21 3 0.82 28 3 0.97 1 16 7 1.12 4 1 1.27 5 5 1.42 1 1 1 1.57 2 2 1.72 1 1.87 1 1 2.02 2 2 1 2.17- 2 2.32 2 2 2.47 1 1 2.62 2 1 2.77 1 1 2.92 3.07 3.22 3.37 3.52 3.67 Mean weight (grains) 0.22 0.82 1.26 2.22 2.47 3-1 E o c52H o> O) 7. Introduction to Experimental Ecology. Academic Press, New York, 401 pp. MYERS, T. D., 1965. A comparative study of size variation in the Atlantic oyster drill Uro- salpinx cinerea (Say). Masters thesis. University North Carolina. Chapel Hill, North Carolina. ORTON, J. H., 1927. The habits and economic importance of the rough whelk tingle (Mnrex erimaceus). Nature, 120: 653-655. SHELDON, R. W., 1967. Relationship between shell-weight and age in certain molluscs. /. Fish. Res. Canada, 24(5): 1165-1171. STAUBER, L. A., 1943. Ecological studies on the oyster drill, Urosalpinx cinerea, in Delaware Bay, with notes on the associated drill En pleura cat/data, and with practical considera- tion of control methods. Unpnbl. Kept. Oyster Res. Lab., Rutgers University, New Brunswick, New Jersey, 180 pp. WALTER, H. E., 1910. Variations in Urosalpinx. Amer. Natur., 44: 577-594. WILBUR, K. M., AND G. OWEN. 1%4. Growth. Pages 211-242 in K. M. Wilbur and C. M. Yonge, Eds., Physiology of Mollnsca I. Academic Press, New York. WOOD, L., 1968. Physiological and ecological aspects of prey selection by the marine gastropod Urosalpinx cinerea ( Prosobranchia : Muricidae). Ifalacolopia, 6(3): 267-320. Reference: Biol. Bull.. 140: 73-83. (February, 1971) DEPOSITION OF ClTICrLAR SUBSTANCES IN VITRO BY LEG REGENERATES FROM THE COCKROACH. LEUCOPHAEA MADERAE (F.) E. P. MARKS AND R. A. LEOPOLD Metabolism and Radiation Research Laboratory, Agricultural Research Service, U. S. Department of Agriculture, l : aro<>. \ortli Itakota ^102 The appearance of cuticle-like membranes on the surface of insect tissues main- tained in vitro has been reported by several investigators. Denial (1956) found a thin membrane covering explants of leg imaginal discs from 3-hour-old Drosophila prepupae held in vitro. He also observed thickening of the cuticle and the formation of tar sal claws. Sengal and Mandaron (1969) explanted leg imaginal discs of late-instar Drosophila larvae with the cephalic complex (including the ring gland) ; when the cultures were treated with ecdysone, many leg discs showed morphogenic changes that included the appearance of setae, surface scultpturing, and claws. Marks and Reinecke (1964) reported the secretion of a gelatinous material on the epidermis of leg regenerates of the cockroach, Lcncophaea maderae (Fabr.), and Larsen (1966) cultured heart fragments from cockroach embryos and found a cuticle-like membrane covering vesicles that formed on these explants. Miciarelli Sbrenna, and Colombo (1967), working with explants of body wall from the abdomen of nymphs of the locust. Scliistoccrca gr eg aria (Forsk), found that the developmental stage of the donor insect greatly influenced the deposition of cuticle. Agui, Vagi, and Fukaya (1969) treated explants of epidermis from diapausing larvae of Chilo supprcssalis (Walker) by adding small amounts of ecdysterone to the culture medium, and within 48 hours, the explant shed its old cuticle and deposited a new one. Ritter and Bray ( 1968 ) cultured clots of blood from the cockroach. Gratn- phadorina portcntosa in a medium that was free of insect plasma. The subcultures developed strands, flakes, and platelets around and within the clot, and the investi- gators concluded that the deposits were made up of a protein-chitin complex and that the complex was deposited on the surface of the culture vessel. However, the electron micrographic studies of Locke (1964) indicated that at least in vivo, the precursors are secreted by the epidermal cells. Deposition of the cuticle in the regenerating leg of nymphs of the cockroach. L. jnadcrac. was studied in vivo by Leopold and Marks (unpublished observations), and they found that two sheaths formed during regeneration. One appeared early in the development of the regenerate and thickened to form a gelatinous envelope around the regenerate. This layer apparently consisted of a protein-carbohydrate complex. The second sheath contained chitin. appeared late in development, thickened, and became wrinkled and rugose. The initial sheath was sloughed before molting, and the underlying chitin-bearing layer became the functional cuticle. 73 74 K. P. MARKS AND R. A. LEOPOLD iV y.M..5-i.-' .^Y-.^v;: ' - FIGURES 1-6. FIGURE 1. A gelatinous sheath (a) laid down by the explain (b). The sharply denned outer border (c) is occasionally found in old, untreated cultures: ;';/ viva 35 days: in vitro S3 days. FIGURE 2. Gelatinous sheath (a) secreted by untreated explant (b) shows epithelial cells (c) extending into the sheath : in vivo 25 days ; in vitro 40 days. FIGURE 3. Detail of the surface of a mature leg regenerate that developed spontaneously in vitro. Seta (a) and surface sculpture (b) are characteristic: in vivo 35 days; in vitro 28 days. COCKROACH CUTICLE IN VITRO 75 In the present study, \ve have investigated some of the processes involved in cuticle secretion by combining the leg regenerate system \vith an in vitro methodol- ogy and histochemical analysis. MATERIALS AND METHODS Cockroach leg regenerates of different ages were prepared as described by Marks (1968). In the first tests, leg regenerates allowed to develop eight days after leg removal were dissected and placed under dialysis strips in Rose chambers with glass coverslips. The chemically defined M-7 medium was used. Six days later, they were examined with a phase contrast microscope, and the chambers were refilled with medium containing the test substance. After six more days, they were re-examined, and those showing cuticle-like deposits were scored as positive. In the second test, leg regenerates were allowed to develop from 10 to 35 days after leg removal and were then dissected and examined. Those showing evi- dence of a cuticle were discarded, and the remainder were placed in Rose chambers between the dialysis strip and a glass coverslip. A plastic coverslip was added, and the chambers were filled with M-18 culture medium to which 5% fetal calf serum was added. The test was arranged with pairs of chambers, each con- taining a pair of leg regenerates from opposite sides of a single cockroach. One chamber of each pair was treated with ecdysterone, and the other was used as the control. Some of the controls were completely untreated, and others were treated with the inactive analog 22-iso-ecdysone (kindly supplied by Dr. John Siddall, Zoecon Corp., Palo Alto, California). The pairs were examined daily with a phase contrast microscope, and differences between the treated and untreated chambers were recorded. If the control showed evidence of developing cuticle, the treated chamber paired with it was removed from consideration. After 15 days, pairs in which the test leg developed a cuticular deposit were separated, and the test leg was discarded. Then the control was given 5 to 10 /xg of ecdysterone to prove that its failure to produce cuticle had occurred because of the absence of the hormone and not from other causes. The presence of chitin in selected specimens was verified by using the fluores- cent enzyme technique of Benjaminson (1969). Frozen sections of the leg regenerate tissue were treated with chitinase conjugated with the fluorescent dye lissamine rhodamine B 200 chloride. They were examined by fluorescence micro- scopy, and fluorescence in cuticular structures was accepted as evidence of chitin. RESULTS AND DISCUSSION Deposition of the sheath Leg regenerates that were allowed to develop for eight days before explanation normally showed no cuticle-like deposits when they were explanted. However, FIGURE 4. Detail of the surface of a 10-day-old leg regenerate treated with 12 /ug/ml of ecdysterone. Setae are present but surface sculpture is absent: in T/TV 10 days; in vitro 18 days. FIGURE 5. Electron micrograph of cuticle from the leg of a newly molted cockroach. Epicuticle (a), sclerotized layer (b), and endocuticle (c) are clearly visible. Laminae (d) of the endocuticle are well formed and compact. FIGURE 6. Electron micrograph of cuticle formed by a 25-day-old leg regenerate in vitro shows well-formed epicuticle (a) but poorly formed endocuticle (b) : in rivo 25 days; in vitro 30 days. 76 !:. I'. M \RKS AXD K. A. LEOPOLD \b . 10 FIGURE 7. A section through cuticle from leg of freshly molted cockroach. This specimen is shown under phase contrast illumination. Cuticle (a), which is refractible, is underlain by the epidermis (b). FIGURE 8. Same specimen as figure 7 subjected to Benjaminson chitinase procedure and fluorescent microscopy. The chitinase-dye complex that has conjugated with the cuticle fluoresces. Note that the epidermis does not conjugate with the enzyme-dye complex and is no longer visible. COCKROACH CUTICLE IN VITRO 77 when they were held in vitro, 13 per cent produced deposits of gelatinous material within 14 days. The deposit appeared as a thin transparent layer that gradually thickened (Figs. 1 and 2) ; also, flakes and threads of darker material and occasional layering of this material appeared. As the cultures aged, the outer border of the layer often became sharply defined and yellowish (Fig. 1). Xo evidence of definitive cuticle appeared even though some cultures were held as much as 120 days. Fluorescent chitinase tests showed that the gelatinous layer, which probably represented the protective sheath seen in in vivo preparations, was devoid of chitin. The protective role of this sheath, which may be related to wound cuticle (Sannasi, 1968), is supported by the finding that methanol (0.5 //.I/ml of medium ) and incubates of muscle tissue were as effective in inducing sheath deposition as were ecdysterone (2.5 jug/ml of medium) and incubates of the prothoracic gland. This sheath material is probably the same as the gelatinous substance reported by Marks and Reinecke (1964) and may also be the same as the cuticular material reported by Larsen ( 1966) . Deposition of the cuticle The first evidence of the deposition of the cuticle in vitro was the appearance of a thin refractile layer between the epidermal cells and the protective sheath. As the deposition continued, the epidermal cells rounded up, contracted, and pro- duced a pebbled appearance. Parallel ridges appeared and increased in size and number so the entire surface of the explant had a rugose appearance. Setae formed from hair-like trichogen cells that protruded from the explant, and the tormagen cell became embedded in the surface and sclerotized (Figs. 3 and 4). The deposition of the cuticle was usually terminated by the withdrawal of the epidermal cells from the secreted structures and the development of a space between them. This process apparently represents the in vitro counterpart of apolysis. Examination of the ultrastructure showed that the cuticle regenerated in vivo possessed a thick endocuticle at the time of molting, and that the lamellae were compact and well formed (Fig. 5). Cuticle that formed in vitro (Fig. 6) showed a well developed epicuticle. but the endocuticle was only partially formed and not as compact as cuticle formed /// vivo; separation of the cuticle from the epidermis was apparent. FIGURE 9. Section through cuticle (a) formed by a 25-day-old leg regenerate in vitro seen with phase contrast illumination. Epidermis (b) has been largely stripped away in sectioning. The membrane is folded so the two layers are visible: in vivo 25 days; in vitro 20 days. FIGURE 10. The same specimen as in Figure 9 subjected to the Benjaminson chitinase procedure and viewed by fluorescent microscopy shows two layers of fluorescence. Xote fluorescence is not as intense as in Figure 8. FIGURE 11. A section through cuticle formed in vitro by a 10-day-old leg regenerate viewed under phase contrast illumination. Xote refractile droplets (a) appearing between the cuticle (b) and epidermis (c) : in vivo 10 days; in vitro 21 days. FIGURE 12. Same specimen as in Figure 11 subjected to the Benjaminson chitinase pro- cedure and viewed by fluorescent microscopy. Both droplets (a) and solid cuticle (b) fluoresce, indicating the presence of chitin. 78 E. P. MARKS AND R. A. LEOPOLD 18 FIGURES 13-18. FIGURE 13. A leg regenerate treated with 5 /*g/ml of ecdysterone shows seta formation. Base of seta lies in the socket (a) that protrudes from the surface of the cuticle (b). The unsclerotized tip of the trichogen cell (c) is visible: in riro 25 days; in i'itn> 25 days. FIGURE 14. Another area of the same leg regenerate shown in Figure 13 shows two setae with well-formed sockets (a) embedded in cuticle (b). Only the base of the seta is sclero- tized ; upper end remains membranous : in vivo 25 days ; in vitro 25 days. FIGURE 15. Reticular pattern (a) and abnormal setae (b) are present on the surface of cuticle developing in vitro. Epidermis (c) has withdrawn, leaving the cuticle behind: in vivo 25 days; in vitro 25 days. COCKROACH CUTICLE 7A T VITRO 7 ( ) The dense cuticle formed in vivo demonstrated a strong reaction for chitin ( Figs. 7 and 8) when it was tested by the fluorescent chitinase procedure. When the cuticle that formed in vitro was tested by this method, the reaction was weak but consistently positive ( Figs. 9 and 10). The lower intensity of the fluorescence probably resulted from the lower density of the endocuticle in the in vitro prepara- tions. In some specimens, particularly among the younger explants, droplets of hyaline material often accumulated between the developing cuticle and the surface of the epithelial cells. When one such explant was sectioned and tested for chitin, we found a pebbled inner surface that contained chitin in the form of droplets (Figs. 11 and 12). Such droplets appeared only after treatment with ecdysterone and were most commonly found when the dosage was very low. The development of setae in vitro occurred with greatest frequency in explants that were taken from older nymphs (Table I). In some cases, the socket formed on the surface of the cuticle (Fig. 13) rather than imbedded in it (Fig. 14). and generally the sclerotization was incomplete. We occasionally found cuticle forming over the surface of a vesicle, and when the vesicle later collapsed and withdrew the delicate epithelium, the remaining cuticle was left with a pattern representing the outline of the cells that deposited it (Figs. 15 and 16). The presence of the hyaline droplets often made it difficult to observe these patterns, but in those areas where they were absent, a pattern of ridges within the area laid down by a single cell was seen (Fig. 18). This apparently represented the imprint of the surface of the cell itself. When leg regenerates were dissected late in the molting cycle (35 days), a large number had already initiated cuticle development /;; vivo. Those that had not were placed in vitro, and more than half continued their development without stimulation (Table I). Therefore, meaningful studies of the induction of cuticle development could only be made by using paired chambers containing legs from opposite sides of the same insect. In an attempt to simplify our procedures and to increase our efficiency, we made a study of younger leg regenerates. Paired chambers were used, and leg regenerates from 5- to 30-day-old were treated with different doses of ecdysterone. The results indicated that while all doses above 2 jug/ml had about the same effect on cuticle production, the frequency with which the leg regenerates responded to stimulation increased with age (in vivo) until a 100 per cent response was reached at 20 days. There was also a corresponding increase in the frequency of seta formation with age. The increase in ability to respond to the hormone with increased age seemed to continue after the tissue was placed in vitro. Thus, 10-day-old legs treated FIGURE 16. Reticular pattern (a) is visible in this very thin cuticle. Droplets (b) prob- ably containing chitin (Figure 11) are present. Epidermis has withdrawn: in vivo 25 days; in vitro 10 days. FIGURE 17. Detail of the surface of a developing cuticle shows droplets accumulating between the cuticle and the epidermis. Droplets probably contain chitin ; in vivo 10 days ; in vitro 18 days. FIGURE 18. Surface detail of developing cuticle shows polygonal outlines of cells (a). Within the outlines is a pattern (b) that represents the negative image of the irregular surface of the cell. This may represent the first step in the formation of the surface sculpture : in vivo 25 days ; in vitro 10 days. 80 K. P. MARKS AND R. A. l.K< >!'< >U > four days after explanation gave a 20 per cent (2/10) response, but when the water controls for this series were treated with ecdysone 14 days later, there was a 66 per cent (4/6) response. A similar but less marked effect was obtained with 15-day-old leg regenerates. The untreated controls did not produce cuticle. A second test was set up to determine the minimum dose of ecdysterone that would produce a 100 per cent response in 25-day-old regenerates. Paired chambers were not used because regenerates of this age did not spontaneously produce cuticle under experimental conditions. In the 12 specimens treated with 2 jA of water and in the 10 specimens treated with 24 jug/ml of 22-iso- ecdysterone, there was no development of cuticle ; all specimens developed cuticle when they were later treated with 10 /^g/ml of ecdysterone for 25 hours. In the experimental chambers, doses of ecdysterone in water ranging from 2.5 /xg/ml of nutrient to 0.05 ju,g/ml were given; then after seven days, the chambers were emptied and refilled with fresh nutrient, and the date when cuticle deposition TABLE I The effect of age at lime of explanatation on development of cuticle in cockroach leg regenerates in vitro Number (and per cent) developing after Days after leg removal Number tested Per cent developing without treatment treatment with 2-10 /ig/ml ecdysterone Cuticle Setae 10 10 2 (20) 1 (10; 15 8 7 (88) 1 (12, 20 6 6 (100) 1 (33) 25 16 (1 16 (100) 10 (62; 30 4 (1 4 (100) 3 (75; 35 s (in* 54 5 (100) 3 (60,) * Eleven were set up, and six (54%) developed spontaneously. All five remaining regenerates developed cuticle when treated with ecdysterone. first appeared was recorded. A 100 per cent response was obtained with 2.5 jug/ml, a 90 per cent response was obtained with doses down to 0.5 jug/ml, a 37 per cent response was obtained with a dose of 0.2 jug/ml, and no response was obtained with a dose of 0.1 /xg/ml or less. All the controls that were later exposed to the hormone developed cuticle. The appearance of setae and surface sculpture and the presence of chitin indi- cated that the cuticular deposits formed in vitro in response to stimulation by ecdysterone are the same as those deposits that are present in vivo at the time of molting. Thus, it is likely that the structures reported by Denial (1956) and Sengel and Mandaron (1969) as appearing on imaginal leg discs taken from late-instar Drosophila larvae probably also contained chitin. In Demal's studies, development was spontaneous since stimulation occurred before the leg imaginal discs were explanted. but evaluation of the results of Sengel and Mandaron is more difficult because the brain and ring gland were present in the same cultures as the imaginal discs and because development occurred in some of the carrier- control cultures. Interaction between the glands and the ecdysone cannot be ruled out since Burdette, Hanley, and Grosch (1968) reported that the effect COCKROACH CUTICLE IN VITRO 81 of ecdysone on the ocular imaginal discs of Drosophila was enhanced by the presence of the cephalic lohes in the culture, and Williams (1952) showed that ecdysone can stimulate secretion by the prothoracic glands. However, the amount of ecdysone used by Sengel and Mandaron in the culture medium would prob- ably have been sufficient to induce cuticle secretion by the leg imaginal discs, even in the absence of the gland explants. The work of Agui et al. (1969) further demonstrated that the entire process of molting, including the shedding of the old cuticle and the deposition of the new, can be induced in short-term cultures by adding ecdysterone to the culture medium. In leg regenerates, the deposition of endocuticle in vitro is usually incomplete and varies considerably from one specimen to another. In general, the 35-day-<>l nun in length. FIGURE 3. Radula of a juvenile ( '. stiinpstnii at 7 rachidian tooth stage. Scale equals 0.01 mm. FIGURE 4. Juvenile at 8 rachidian tooth stage showing radula. eyes, cerebral ganglion and statocysts. Scale equals 0.03 mm. M. PATRICIA MORSK A 8 FIGURE 5. Two lateral radular teeth of an adult C. stimpsoni showing variations of the shape and presence of denticles. FIGURE 6. Portion of radular ribbon of C. stimpsoni showing the single row of rachidian teeth and laterals of one side. NUDIBKA.XCH MOLLUSC LIFE HISTOKV 87 strate in a probing manner and are utilized for holding the prey during feeding. The rhinophores are very contractile and can retract to one half their length upon being touched or when they come in contact with a foreign object such as a sea anemone. When contracted the rhinophores appear to be ringed but in extension during undisturbed activity this configuration is lacking. In the living forms the rhinophores are tinged a light orange and the distal third has a white external pigmentation. The foot is broad, well developed with two short anterior lateral extensions which add i to the total width of the foot at their point of origin. The width of the foot is relatively constant and the posterior end forms a rather abrupt point. During crawling activities the posterior portion of the foot does not project far from the body. The cerata have a central core of the digestive gland which i> red-brown in color. In each ceras this core terminates at the distal end just before an epidermal pigmentation of white dots appears. In some cerata these external dots are nu- merous and close together forming a ring and the tip beyond this ring is the translucent white color of the body. The cerata are round in cross section and taper evenly toward the distal end. In several specimens there were patches of two to five cerata which did not contain the central core of the digestive gland. The reproductive opening, anus and excretory pore all open on the right side of the animal between the mantle edge and the foot. The anus is located 7 posteriorly on the right side, the genital complex anteriorly on the same side and the smaller excretory pore in between nearer the genital openings. The area between foot margin and the first longitudinal row of cerata is pigmented with an epidermal white pigmentation of the same type found associated with the cerata and the rhinophores. The triseriate radula is variable. In the fifteen radulae examined, the num- ber of rows of teeth varied from 2-131. The denticulations on either side of the central tooth ranged from 8-13 and those of the lateral teeth varied from 9-11 fine denticles to no observable denticulations. On several radulae the younger lateral teeth of the ribbon appeared to be more slender and to have fewer denticles (Fig. 5 ). The central rachidian teeth measured 0.28-0. 38 mm in height (Fig. 0). The teeth are attached to the radular membrane by two square lateral projections which were described and figured by Krause (1892) and Knipowitsch (1902). The lateral teeth ranged between 0.17-0.22 mm in length and both slender and stout teeth as figured by Krause and Knipowitsch were found. Two well-developed jaws are associated with the buccal mass. The masticating edges of the jaws form an oval ring lining the oral opening into the buccal mass. Bergh (1885) illustrated a portion of the masticating edge of the jaw showing the denticles which agrees with my observations. Under oil immersion the denticles are found to be covered with tubercles which are most prominent on the reflected FIGURE 7. Side view of developing veliger of C . stinipsoiii showing the mantle fold and shell. FIGURE 8. Encapsulated veliger of C. stiiiipsuni with velum, foot, eyes, mantle fold and larval heart. FIGURE 9. Juvenile C. stimpsuni with four primary cerata with differentiated cnidn>;ir- Xote the beginning of the first pair of cerata in front of the anterior pair. M. PATRICIA MORSI-: outer edges of the masticating portion and considerably worn down toward the oral opening into the huccal mass. I -riding activity \Yhen numbers of C. stimpsoni were observed in Maine, the animals were found crawling on an extensive mud flat. Screening of the substrate was under- taken to determine a possible food source for the nudibranchs. The burrowing anemones, Edwardsia elegans Verrill and Halcampa duodecimcirrata (Sars), were found in the same habitat and later in the laboratory when with these organisms, C. stimpsoni exhibited a definite feeding response. The feeding animal projected the outer lip and mouth forward and lowered the oral tentacles to hold the prey. The epidermis of the anemone was rasped away by the radula and ingested. Speci- mens collected in 1969 were maintained in the laboratory using the hydroid Tnbularia sp. as their food organism. Reproduction and development Reproduction and development in Coryphella stimpsoni has been observed over a two-year period, and in each year the nudibranchs were collected in October, taken to the Marine Science Institute and held in the laboratory tanks. The animals copulated for several hours during which time the exchange of sperm occurred when the two individuals were aligned in opposite directions with their reproductive openings in communication. There was little movement during copulation. Egg masses, when viewed from above, were normally deposited in concentric whorls starting from the center and moving counterclockwise. The coil thus produced varied from H-2] whorls and adhered tightly to the surface on which it \vas extruded. In one case the individual whorl measured 2 mm in width and the entire egg mass measured 10 mm in diameter. The individual uncleaved eggs measured 0.25 mm in diameter and the egg capsule surrounding the single egg measured 0.41 mm in length by 0.30 mm in width. The egg masses were de- posited over a period of approximately 15-20 days during the month of December. The algal species Ascophyllum nodosinn and Chondrus crispits were introduced to the tanks and utilized by the nudibranchs for egg deposition. Although both these species are present on the intertidal area where the nudibranchs were col- lected and may serve as a substrate for egg deposition, C. stimpsoni also deposited eggs on the sides of the tanks and the undeside of several rocks in the tanks. Development time varied during the two years of observations and is influenced by temperature. The time of development from egg deposition until the juvenile crawled from the egg mass was 52 days at a temperature ranging from 4.0 C to 5.0 C. The time of development at a temperature ranging from 5.0 C to 8.5 C was from 25 to 34 days. This makes it difficult to predict development in nature and continual observations of development in the Maine habitat have not been made. The eggs cleave equally to the four-cell stage and a quartet of micromeres are given off clockwise at the eight-cell stage. A rounded ball of small cells results from repeated cleavages. Soon after this a cap-shaped gastrula is formed with a XUMHK AXCH MOLLUSC LIFE HISTORY S) ventral depression. The gastrula changes shape to a rounded form and the velum begins to differentiate. There is no observable internal differentiation, the embryo appears opaque due to the presence of a large amount of yolk and there is no movement within the egg capsule. Differentiation of the foot and velum of the nudibranchs continues with the development of two velar lobes and cilia. The foot, without an operculum, enlarges on the ventral side. At this stage a large yolk mass of the body begins to differentiate into three lobes. These lobes are partially united anteriorly, but posteriorly the mass on the left side is single and there are two smaller lobes on the right side. The two right lobes are united to one another and then together with the larger left yolk mass. The larvae begin to show some movement within the egg capsules. Further differentiation occurs with the formation of the mantle fold, the secretion of a thin shell surrounding the body of the veliger. and the development of a pair of eyes and statocysts. The veligers move within the egg capsule but not at any great speed and in many cases, they remain still with the velar cilia continuing to move randomly. The shell is very thin and is pres- ent for only a short period of time (Figs. 7 and 8). \Yhile still in the capsule larval characteristics begin to disappear and the veliger begins to resemble an adult. The mantle fold moves posteriorly forming the edge of the mantle. The four primordia of the cerata develop in front of the fold which subsequently becomes indistinct. However, it can still be seen just above the large digestive gland and posterior to the cerata primordia. The velum is reabsorbed and the velar cilia are reduced to patches near the eyes. .As the velum is reabsorbed the cilia on the muscular foot become active. Although the entire surface of the foot is covered, the cilia are numerous and close together along the anterior margin. These cilia cause the rotation of the embryo which is very minor at this stage. Small rudiments of the oral tentacles and rhinophores appear on the head and the dorsal portion of the visceral mass shows differentiation of four digestive gland primordia which develop simultaneously, in most cases while the nudibranchs are still within the egg capsule. The juveniles hatch from the egg mass from 25 to 52 days after egg deposition depending on the temperature of the surrounding water during development. Sev- eral days before hatching, the interior portion of the stroma surrounding the eggs becomes opaque and is invaded by microorganisms. Nematodes, protozoans and harpacticoid copepods can be seen moving through the coils. The outside of the egg string remains intact for about a week after the hatching process is com- plete. The individual egg capsules become slightly opaque just prior to their rupture and when the capsule wall collapses the animal crawls away from the flaccid structure. Eclosion is not clear. There is a great deal of activity of the foot prior to hatching and the cilia of the foot, especially the numerous anterior cilia, beat continually. Juveniles which escape from the egg capsule have been found, in some cases, to have from two to four radular teeth present which they may use to rasp through the capsule. Juveniles crawl directly from the egg capsule and take up an immediate benthic existence. They were positively phototactic and gathered on the sides of the vessels nearest the window if food was not present. However, if food was present, they crawled toward the hydroid to feed. The newly hatched juveniles appear '"> M. 1'ATKICIA MORSK opaque white with large granules of yolk in the body. The body is well differ- entiated from the foot. The surface epithelium is entirely ciliated with greater concentrations of cilia on the anterior portion of the foot and around the mouth area. Stiff sensory bristles occur on the oral tentacles, rhinophore primordia and the four primary cerata. Statocysts and eyes are present and the small juveniles continue to move over the surface of the substrate by both muscular action of the foot and by ciliation of the foot. In an experimental group, mixed cultures of the algae Dunaliella sp., Monochrysis lutheri and Isochrysis galbana were intro- duced into the bowls. The algae could be seen rotating in the stomach and gave the digestive system a green coloration. However, the animals were not able to survive on this diet and it is questionable as to whether they were able to derive any nutrition from the algae. The second time the life cycle was studied (1969-70), the hatched juveniles were immediately placed with the gymnoblast hydroid Sarsia uiirabilis which was collected from the intertidal rocks at Nahant. Feeding began immediately and, in the continual presence of the food organism, individuals reached a length of 6.6 mm at 60 days post-hatch with the development of approximately 60 cerata on the dorsal surface (Fig. 2). After ingestion there is an immediate coloration of the stomach as well as of the four extensions of the digestive gland in the four primary cerata. At this stage the left digestive gland supplies the anterior left and posterior pair of primary cerata. The smaller right digestive gland supplies the right anterior ceras. Ex- ternally the cerata appear bulky in feeding juveniles and at their tips the cnidosacs differentiate as internal opaque white structures. Squash preparations of living animals show an abundance of unexplored nematocysts stored in the cnidosacs. An animal that is disturbed with a probe under a dissecting microscope is seen to emit these unexploded nematocysts from the tip of the cerata. It would appear from the number of nematocysts present at any one time that they must be continually given off from the cerata in order to eliminate the structures from the digestive system. Food within the digestive gland extensions is continually flow r - ing to and from the stomach by contractions of the cerata and ciliary action of the cells lining the gut. Very soon after feeding begins fecal pellets accumulate. The food-organism was orangish-pink but the fecal pellets were red. The nudibranchs graze on the entire hydroid colony but appear to prefer the undifferentiated tips of growing coenosarc. The radula is used to graze on the hydroid and can be seen in a rasping motion during the feeding process. Squash preparations and whole mounts utilizing Turtox CMCS mounting media have shown that the newly hatched juveniles already have a radula with three to five rachidian teeth with well- developed denticles (Fig. 3). The radular teeth continue to increase in numbers during growth and in 1040 day post-hatch juveniles there are from 8-12 rachidian teeth on the radular ribbon (Fig. 4). Toward the posterior portion of the radular ribbon (at about rachidian tooth 7) a pair of slender lateral teeth can be seen on the radular ribbon. Juveniles that actively feed on the gymnoblasts begin to show formation of new cerata after about ten days of feeding. The first pair of new cerata form in front of the anterior pair of primary cerata (Fig. 9). The second pair arise in front of the second pair of primary cerata and the third single ceras differen- NUDIBRANCH MOLLUSC LIKK HISTORY FIGURE 10. Diagrammatic representation of the sequential development of the cerata from the primary cerata (AP, anterior primary ceras ; PP, posterior primary ceras). d, C 2 and C t represent the right cerata of pairs while Ca is a single posterior ceras. Other structures shown include the oral tentacle (T), anal opening (A), heart bulge (H) and rhinophore (R). tiates at the posterior extremity of the nudibranch body (Fig. 10). The foot extends beyond this portion of the body. The cnidosacs of these new cerata dif- ferentiate several days after the initial cerata formation. The fourth pair of cerata develop just behind the second pair of primary cerata. Cerata formation continues in pairs \vith a second lower group being established on either side of the middle two rows. At this stage the nudibranchs measure 2.0-4.5 mm in length. The eyes gradually sink to a position on either side of the cerebral ganglion and are relatively small at this later stage in comparison to earlier post-hatch stages. The cerata are not as bulky in the 4.5 mm length nudibranchs and the whole animal takes on a slender adult-like appearance. There are scattered spots of white epidermal pigmentation at the four-cerata stage and as the animal continues to differentiate they coalesce into rings around the cerata (Fig. 2). The adult heart is clearly visible and beating at a length of 4.6 mm (Fig. 11). FIGURE. 11. Juvenile Coryphella stimpsoni showing development of cerata at a length of 4.6 mm. ^2 M. PATH 1C I A MORSK DISCUSSION Yen-ill (18SO) indicated that Coryphella stimpsoni bore a close resemblance to Coryphella salmonacea Couthouy. Both species have evenly spaced cerata as opposed to those species of Coryphella with cerata in distinct groups. In addi- tion both species show a similar type of development in that the young crawl di- rectly from the egg mass when hatching (Morse 1969a, 1969b). The differences between adults of the two species are that C. stimpsoni has a distinct trefoil sur- rounding the mouth, deposits egg strings in December, usually occurs on mudflats and will feed on burrowing sea anemones, while Coryphella salmonacea is gen- erally larger, has only lateral bulges surrounding the circular mouth, deposits egg masses in March and April, and has only been found on rocky shores. The radula of C. stimpsoni is variable within one animal as well as among sev- eral nudibranchs. Variations include the dentition of the central and lateral teeth, shape of the lateral teeth and numbers of rows of teeth in the adults. This is in contrast to Odhner's (1939) opinion that in the genus Coryphella. the radula is relatively constant in each species. Schonenberger (1969) described the variation of the radula of Trinchesia granosa Schmekel throughout its life- history. A comparable variation was seen in C. stimpsoni and points out the need for further studies on large numbers of nudibranchs of a species to determine if variability exists. The development pattern of Coryphella stimpsoni is similar to development type 3 as described by Thompson (1967) ; that is the animal begins benthic life immediately after breaking out of the egg membrane. Two reports of develop- ment type 3 in eolid molluscs were listed by Thompson (1967). i.e., the report of Tardy (1962) on Eolidina aldcri and that of Roginskaya (1962a) on Cuthona pustitlata. In a second paper, Roginskaya (1962b) reported the development of C. piistnlata in more detail. Schonenberger (1969) reported a similar develop- ment in Trinchesia granosa. The term direct development implies a juvenile crawl- ing from the egg mass to a benthic existence, however E. alderi crawls out with the veliger shell still attached and C. pustulata loses the veliger shell after breaking out of the capsule but while still in the stroma of the egg mass. Both T. granosa and C. stimpsoni crawl out of the egg mass as juveniles with no shell. Rao (1961) described the differentiation of the eolid Cuthona adyarensis which assumed a benthic existence after a short free-swimming veliger phase. Rao (T961), Roginskaya (1962b) and Schonenberger (1969) all reported in detail the development of the oral tentacles, the rhinophores and the sequence of dif- ferentiation of the cerata. It appears that these processes are very similar although there is a noticeable difference in the sequence of cerata differentiation which may prove to be species specific. In Coryphella stimpsoni the oral tentacle rudiments, rhinophore rudiments and the large dorsal visceral mass are already present when the nudibranch breaks through the membrane. The visceral mass soon equally forms four primordia of the primary cerata. The yolk present in the juvenile is enough to nourish the juveniles during the period of differentiation, although with- out the natural food further development is halted. Thompson (1967) noted that four species of nudibranchs reported to have development type 3 have a boreo-arctic distribution which is also true of Cory- phella stimpsoni. One of the four species mentioned by Thompson, Cadlina laevis, XUDIHRAXCH MOLLUSC LIFE HISTORY ( >3 has been collected by the author within two miles of the locality from which the Coryphella stlmpsoni were taken. C. stimpsoni is similar to Cadlina laevis in that the veligers and hatching juveniles are filled with yolk granules. In both species the embryonic period is long, there is a reduction of the velum and shell, and there is an absence of an operculum. The term direct development is a bit misleading in the literature because of prior use, especially in relation to insect development. In the latter it generally refers to a series of embryonic stages where the animal's form closely resembles the adult. The juveniles do not have a distinct set of "larval characteristics." Among the molluscs the term has only been used for nudibranchs although many prosobranchs and pulmonates show a similar pattern of development. The im- portant factors remain that there are a group of opisthobranchs which crawl di- rectly from the enveloping capsule without a free-swimming veliger stage and in most cases when crawling-out show no external evidence of typical opisthobranch veliger morphology. Thompson (1967) has listed the characteristic reduction of veliger structures during early development which holds true of C. stimpsoni and which appears common to this form of development. It is suggested that his classification, i.e., "Development Type 3" be utilized in place of the often mis- leading term "direct development." SUMMARY 1. The morphology and life history of the nudibranch mollusc, Coryphella stimpsoni (Verrill 1879) is described. 2. Adults were collected in October and deposited egg strings in December. 3. Development corresponds to Thompson's (1967) Type 3, i.e., juveniles crawl directly from the egg mass and begin benthic life. 4. Development time from egg deposition to the benthic crawling stage was from 25 to 52 days. 5. Veligers are characterized by a large amount of yolk, a reduced velum, a thin shell and no operculum. 6. Sequential development of tentacles, rhinophores and cerata and the differen- tiation of other organs such as cnidosacs, heart and digestive structures in the juveniles are described. 7. This is the first report of an eolid nudibranch from New England which crawls directly from the egg mass as a juvenile to begin benthic life. LITERATURE CITED BERGH, R., 1885. Beitrage zur Kenntniss der Aeolidiaden VIII. Vcrh. Zoo!. Bot. Ges. U'cin, 35: 1-60. KNIPOWITSCH, N., 1902. Zoologische Ergebnisse der Russischen Expeditionen nach Spitz- bergen. Ann. Mils. Zool., 7: 355-459. KRAUSE, A., 1892. Mollusken von Ostspitzbergen. Zool. Jahrb. Abi. Syst. Oekol. Geogr. Tierre, 6: 339-376. MORSE, M. PATRICIA, 1969a. Contribution to the knowledge of New England Nudibranchs. Ann. Rep. Amcr. Malacol. Union, 1969 : 7-18. MORSE, M. PATRICIA, 1969b. Direct development in the nudibranch mollusc, Corvpliclla stimp- soni Verrill 1880. Amcr. Zool., 9(3) : 618. 94 M. PATRICIA MORSE ( IUHNER, NILS Hj, 1939. Opisthobrancliiate mollusca from the Western and Northern coasts of Norway. DC t. Kgl. Norske Vidensk. Sdsk. Skr. 1 : 1-92. RAO, K. Y., 1961. Development and life history of a nudibranchiate gastropod Cuthona adyarensis Rao. J. Mar. Biol. Ass. India, 3 : 186-197. ROGFIVSKAYA, I. S., 1962a. The egg masses of nudihranch molluscs of the White Sea. Biologiya Beloga Morya, Trudy Belomorskoy Biologisheskoy Stantsii Morsk. Go- snndarst Utiiv., 1: 201-214 (In Russian). ROGINSKAYA, I. S., 1962b. Biology, reproduction and development of Cuthona pustulata (Gastropoda, Nudibranchia). Dokl. Akad. Naiik SSSR, 146(2): 488-491 (In Rus- sian). SCHONENBERGER, NoRRKRT, 1969. Beitragc zur Entwickkmg und Morphologic von Trinclicsia granosa Schmekel (Gastropoda, Opisthobranchia). Pnbbl. Sta. Zool. Napoli, 37: 236-292. TARDY, I., 1962. Cycle biologique et metamorphose d'Eolidina alderi ( Gasteropode, Nudi- branche). C. R. Hebd. Seances Acad. Sci. Paris, 225: 3250-3252. THOMPSON, T. E., 1967. Direct development in a nudibranch, Cadlina lacvis, with a dis- cussion of developmental processes in opisthobranchia. /. Mar. Bin!. Ass. U. K., 47: 1-22. VERRILL, A. E., 1879. Notice of recent additions to the marine fauna of the eastern coast of North America. Aincr. J. Sci. f 17: 309-315. VERRILL, A. E., 1880. Notice of recent additions to the marine invertebrata of the north- eastern coast of America, with descriptions of new genera and species and crucial remarks on others. Proc. U. S. Nat. Mus.. 3: 356-405. Reference : Biol. Bull., 140: 05-103. ( February, 1971) JUVENILE NEMATODES ( ECHINOCEPHALUS PSEUDOUNCINATUS) IX THE GONADS OF SEA URCHINS (CEXTROSTEPHANTS CORON ATI'S} AND THEIR EFFECT ON HOST GAMETOGENESIS J. S. PEARSE AND R. W. T1MM II'. M. Keck Laboratories, California Institute of Technology, Pasadena, California ( >1}(> ( >, and Department of Nenuitology, University of California, Davis, California 95f>1'> Although sea urchins have been used as biological research material for many years, particularly as a source of gametes, there have been few reports of infection of these animals by nematodes. The nematode Echinomermella grayi (Gemmill and von Linstow, 1902) tentatively placed in the Mermithoidea by Chitwood (1933), is known from a single specimen found in the perivisceral coelom of Echinus esculentus off Britain, and perhaps a second specimen (Irving, 1910; Ritchie, 1910). Juvenile specimens of gnathostomatid nematodes have been re- ported twice from sea urchins : Echinocephalus uncinatus "perhaps accidentally" in tropical sea urchins (Shipley and Hornell, 1904) and a single specimen of E. pseudouncinatus in the gonads of Arbacia puiictulafa at Woods Hole, Massa- chusetts (Hopkins. 1935; Millemann, 1951). Nematodes have also been found in the gonads of the urchin Astropyga pulvinata off Acapulco, Mexico, but these have not been identified (]. S. Pearse, unpublished observations). Neither Hyman (1951, 1955), Johnson (1968), Johnson and Chapman (1969), nor Hol- land and Holland ( 1970) cite any other record of nematodes in echinoids. We report herein the regular occurrence of juvenile Echinocephalus pseudo- uncinatus in the gonads of Centrostephanus coronatus off Southern California. Between about 40 and 80% of the sea urchins contained the nematodes in their gonads. Moreover, gametogenesis in parts of the gonads was profoundly affected by the presence of the nematodes. MATERIALS AND METHODS Most specimens of Centrostephanus connuitus were collected from 3 to 10 m depth by scuba diving off the northeastern shore of Big Fisherman's Cove, Santa Catalina Island, near the Santa Catalina Marine Biological Laboratory. Other samples were taken from Pin Rock in Catalina Harbor on the opposite side of Santa Catalina I. -kind, and from Whistler's Reef off the mainland coast of Cali- fornia near Corona del Mar. These three areas are all in quite different waters: Big Fisherman's Cove is on the Santa Catalina Channel side of Santa Catalina Island; Catalina Harbor, although less than 3 km distant, is on the Pacific side of the island : and Whistler's Reef is about 25 km away on the opposite side of Santa Catalina Channel. Moreover, one sample of gonads from C. coronatus was taken on 26 November 1968 from Bahia Tortola (Turtle Bay) on the west-central coast of Baja California, Mexico. 95 W> J. S. PEARSE AND R. W. TIMM The sea urchins were lodged deep in crevices among the rocks during the day and were extracted with the aid of a bent metal rod. During the night, animals foraged in the open and were more easily collected. The specimens were dissected within a day after collection and all the gonads were carefully inspected for the presence of nematodes. Records were kept of the presence or absence nf worms in the gonads. with notation of their relative abundance. Gonads with nematodes were fixed in Bouin's solution or warm formalin- alcohol-acetic acid-water solution (FA A ; 1:5:1:4). Some of the fixed specimens were dehydrated in an acetone series, cleared in benzene, embedded in paraffin, sectioned at 10 /*, and stained in hematoxylin and eosin. Others were dissected whole from the gonads. TARLK I Incidence of specimens of Centrostephanus infected with Echinocephalus Number (.and per cent) Number of infected with Date Centrostephanus Echinocephalus Big Fisherman's Cove, Santa Catalina Island 11 Aim. ft> 38 2t> V 68.5) 20 Aug. 6<> 7 5 (71.5) 26 Aug. (>) 18 14 (78.0) 31 Aug. 69 17 14 (82.5) 15 Oct. 69 10 9 (90.0) 29 Nov. 69 20 16 (80.0) 2 Feb. 70 18 13 (65.0) 9 Mar. 70 25 21 (80.0) Catalina Harbor, Santa Catalina Island 28 Au S . f>) 22 >> 41.0) Whistler's Reef, Corona del Mar 10 Nov. 69 20 8 (40.0) 20 Nov. 69 ] 8 7 S9.0) OBSERVATIONS Frequency oj neinatode injection in Centrostephanus coronatus A total of 153 specimens of C. coronatus were collected from Big Fisherman's Cove, Santa Catalina Island, and examined for the presence of nematodes in their gonads. Of these, 118 (78%) were infected (Table I). The percentage infected in each sample ranged from 65 to 90%. All of the urchins in the August samples were sexed by microscopic examination of gonadal smears ; there were 46 males and 34 females, and 31 and 28 were infected, respectively. All the samples of C. coronatus from both Pin Rock in Catalina Harbor and Whistler's Reef off Corona del Alar had lower incidences of infection than those from Big Fisherman's Cove (Table I). Moreover, not only were a smaller percentage of animals infected, but those that were contained fewer nematodes. Most of the infected urchins in Big Fisherman's Cove had multiple infections, with several nematodes in each of the five gonads. In contrast, most of the XEMATODES IN SEA URCHIN GONADS () ' infected urchins from both Pin Rock and Whistler's Reef had only one or two nematodes in only one or two of the gonads, and detection of the infection re- quired careful dissection of all the gonads. Seventeen specimens of C. coronatiis were also collected and dissected on 26 November 1968 from Bahia Tortola (Turtle Bay) on the west-central coast of Baja California. At the time of the collection, the presence of nematodes was not known, but during the dissection, one gonad was found with an encysted nematode. The juvenile nematode The nematodes we found in the gonads of C. coronatiis are all juveniles. They do not differ in any respect from Millemann's (1951) description of Echino- cephalus pseudouncinatus. The most significant feature for identification is the structure of the head bulb (Fig. 1A), which bears 6 rows of about 40 hooks each, with a lateral separation of the hooks and smaller hooks on the dorsal and ventral areas. Figure IB shows the tail, which was not drawn by Millemann (1951). Echinocephalus pseudouncinatus was originally described from numerous ju- venile specimens found in the foot of pink abalones Haliotus corrugata from San Clemente Island off Southern California (Millemann. 1951). This is the only species of Echinocephalus presently known from Southern California. Juveniles of another species of Echinocephalus, E. uncinatus, are known to infect molluscs elsewhere (Shipley and Hornell, 1904; Baylis and Lane, 1920), and it is likely that E. pseudouncinatus juveniles regularly infect both molluscs and C. coronatiis in Southern California. Among the sea urchins of Southern California, however, juvenile E. pseudouncinatus seem to infect C. coronatus specifically. Hundreds of specimens of the sea urchins Strongylocentrotus purpuratus, S. jranciscanus, and Lytechinus anamesus from Southern California have been dissected during the past year and only a single specimen of E. pseudouncinatus has been found in the gonad of one S. purpuratus collected near San Diego. The single, remarkable record of what seems to be a juvenile specimen of E. pseudouncinatus in the gonad of a specimen of Arbacia pnnctnlata at Woods Hole, Massachusetts, should again be noted (Hopkins. 1935: Millemann. 1951 ). Possible adult hosts Adults of the genus Echinocephalus inhabit the intestine, usually the spiral valve, of elasmobranchs (Hyman, 1951; Yamaguti, 1961). The most conspicuous elasomobranch in the vicinity of Big Fisherman's Cove, Santa Catalina Island, is the California horned shark Heterodontus francisci. This species is known to feed on sea urchins and its teeth and bones often are colored purple, presumably from an accumulation of sea urchin naphthoquinone pigments (Leighton Taylor, Scripps Institution of Oceanography, personal communication). A similar purple colora- tion of teeth and bones occurs in the sea otter Enhydra lutris, which feeds on sea urchins (Fox, 1953). Other species of horned sharks are known to feed on sea urchins, including H. phillipi and H. galeatus off southeast Australia, and H. japonicus off Japan (Smith, 1942). Moreover, Saville-Kent (1897. page 192) noted the teeth of H. phillipi ". . . are not infrequently stained a deep purple, through constant indulgence in a dietary of the commoner purple urchin." J. S. I'KAKSK AND R. \V. TIMM Accordingly, three specimens of California horned sharks were collected in February 1070 from Rig Fisherman's Cove, Santa Catalina Island, and the in- testinal contents examined. ( )ne large female, measuring 76 cm total length, correspond to Figures 4 through 6; the scale line indicates 5 mm. Fir.ukK 4. Section through the normal-appearing aboral portion of the testis in Figure 3, showing the gonoduct (G) surrounded by perithemal (P) and hemal (H) spaces, and the testicular tubules full of spermatogenic cells (S). FIGUKK 5. Section through the nematode in the testis in Figure 3, showing a portion of the nematode (N) and extensive nutritive phagocytic tissue (NP) in a tubule surrounded by a fibrous cyst w^all CO, and adjacent tubules full of spermatogenic cells (S). FIGURE 6. Section through the shriveled oral portion of the testis in Figure 3 just below the nematode, showing the germinal epithelium full of degenerating cells (D) and large perihemal spaces (P). Figures 4-6 are from 10 ^ paraffin sections stained with hematoxylin and eosin ; all are at the same magnification and the scale line in Figure 4 indicates 100 /j.. 1<>2 J. S. I'EARSE AND R. W. TIMM active and normal in tin- aboral half of the testis, while the tubules were shrunken and rilled with degradation products orally (Figs. 3-0). The rinding that encysted juvenile worms can suppress urchin gametogenesis has important implications regarding gametogenic control. Suppression does not seem to be due to the release and diffusion of some substance by the worms because gametogenesis is suppressed only oral to the infections rather than all around them. Moreover, the worms are coiled and encased in cysts produced by the urchin gonads ; they are not likely to move through the gonads and disrupt gametogenesis. Rather, the encysted worms seem to block the passage of some material that is essential for gametogenesis. Such a gametogenic regulating sub- stance, although never directly demonstrated, is almost certainly present because gametogenesis occurs in synchrony among all five gonads in individual sea urchins ( Pearse, 1969). The encysted nematodes probably do not block transport of simple nutrients; radioactive tracing studies have shown that nutrient transfer in urchins occurs mainly through the perivisceral coelom ( Farmanfarmaian and Phillips, 1962). Instead, the apparent blockage of gametogenesis by the juvenile nematodes indicates the presence of a hormonal substance which regulates gameto- genesis in sea urchins and is transported within the gonad (perhaps through the perihemal or hemal spaces) rather than through the perivisceral coelomic fluid. We are indebted to 1 )r. Russel L. Zimmer, Resident Director, for providing aid and facilities for these studies at the Santa Catalina Marine Biological Labora- tory. Santa Catalina Island, and to Dr. Nicholas D. Holland, Dr. W. Duane Hope, Dr. Phyllis T. Johnson, and Mr. Leighton Taylor for advice and criticism. This study was supported in part by the Federal Water Quality Administration Grant No. 18050 DNV. The sample from Bahia Tortola, Baja California, was taken during Stanford Oceanographic Expedition, TE VEGA Cruise 20, supported by the National Science Foundation Grant Nos. GB6870 and GB6871. Part of the manuscript was prepared while the senior author was a visting faculty member of the University of California, Santa Cruz. SUMMARY AND CONCLUSIONS 1. The juvenile phase of the nematode Echinocephalus pseudouncinatus occurs commonly in the gonads of the sea urchin Centrostephanus coronatus off Southern California. The gonads of about 78% of the urchins in Big Fisherman's Cove, Santa Catalina Island, were heavily infected with the juveniles. About 40% of the urchins at Pin Rock in Catalina Harbor, Santa Catalina Island, and at Whistler's Reef off Corona del Mar on the mainland coast of California were in- fected, although the infection was usually not as severe as in Big Fisherman's Cove. A juvenile, probably of E. psuedouncinatus, also was found in gonads of C. coronatus collected from Bahia Tortola off west-central Baja California. Al- though juveniles of E. pseudouncinatus also occur commonly in the foot of pink abalones, they only rarely infect other species of sea urchins in Southern California. 2. The California horned shark Heterodontus francisci seems a likely host of the adult phase of E. pseudouncinatus. However, when specimens of the horned NEMATODES IN SEA URCHIN GONADS 103 shark were examined, adult specimens of Echinocephalus were found which do not seem to be the same species as our juvenile specimens. 3. The juvenile nematodes were encysted mainly in the spaces (perihemal or hemal) between the perivisceral coelomic epithelium and the germinal epithelium of the Ccntrostcphanus gonad. Large juveniles filled the host gonadal tubule and bulged into the perivisceral coelom. Host gametogenesis was suppressed in the infected gonadal tubule, especially in the oral parts of such tubules. Gametogene- sis in adjacent tubules did not seem affected. When the juveniles were in major tubules of the host gonad, severe suppression of host gametogenesis occurred in the oral parts of the gonad. It is suggested that encysted juveniles block the passage through the gonadal tubules of some hormonal substance that regulates urchin gametogenesis. LITERATURE CITED BAYLIS, H. A., AND C. LANE, 1920. A revision of the nematode family Gnathostomidae. Proc. Zool. Soc. London. 1920: 245-310. CHITWOOD, B. G., 1933. The systematic position of Echinoncma grayi Gemmill, 1901. J. Parasitol, 20: 104. FARMANFARMAIAN, A., AND J. H. PHILLIPS, 1962. Digestion, storage, and translocation of nutrients in the purple sea urchin (Strongylocentrotus purpuratus). Biol. Bull., 123: 105-120. Fox, D. L., 1953. Annual Biocliroities. Cambridge University Press, Cambridge, 379 pp. GEMMILL, J. F., AND O. VON LINSTOW, 1902. Icthyonema grayi. Arch. Naturgesch., 68 : 113-118. HOLLAND, N. D., AND L. Z. HOLLAND, 1970. A bibliography of echinoderm biology, continuing Hyman's 1955 bibliography through 1965. Pnbbl. Stas. Zool. Napoli, 37 : in press. HOPKINS, S. H., 1935. A larval Echinocephalus in a sea urchin. /. Parasitol., 21 : 314-315. HYMAN, L. H., 1951. The Invertebrates. Volume III. Acanthoccphela, Aschelminthes and Entroprocta. McGraw-Hill Book Co., New York, 572 pp. HYMAN, L. H., 1955. The Invertebrates. Volume IV. Echinodermata. McGraw-Hill Book Co., New York, 763 pp. IRVING, J., 1910. Nemertine within test of sea-urchin. The Naturalist (London), 1910: 6. JOHNSON, P. T., 1968. An Annotated Bibliography of Pathology in Invertebrates other than Insects. Burgess Publishing Co., Minneapolis, Minnesota, 322 pp. JOHNSON, P. T., AND F. A. CHAPMAN, 1969. An annotated bibliography of pathology in in- vertebrates other than insects supplement. Misc. Publ. No. 1, Center for Patho- biology, University of California, Irvine, 76 pp. MILLEMANN, R. E., 1951. Echinocephalus pseudouncinatus n. sp., a nematode parasite of the abalone. /. Parasitol., 37: 435-439. PEARSE, J. S., 1969. Reproductive periodicities of Indo-Pacific invertebrates in the Gulf of Suez. I. The echinoids Prionocidaris baculosa (Lamarck) and Lovcnia elongata (Gray). Bull. Mar. Sci., 19: 323-350. RITCHIE, J., 1910. Worm parasitic in sea-urchin. The Naturalist (London), 1910: 94. SAVILLE-KENT, W., 1897. The Naturalist in Australia. Chapman and Hall, London, 302 pp. SHIPLEY, A. E., AND J. HORNELL, 1904. The parasites of the pearl oyster. Kept. Govt. Ceylon Pearl Oyster Fish. Gulf Manaar (Herdmann, London), 2: 77-106. SMITH, B. G., 1942. The heterodontid sharks, their natural history and external development of Heterodontus (Ccstracion) japonicus based on notes and drawings by Basford Dean. Pages 647-770 in E. W. Gudger, Ed., The Basford Dean Memorial Volume, Archaic Fishes, Part II, Article VIII, American Museum of Natural History, New York. TAHARA, Y., AND M. OKADA, 1968. Normal development of secondary sexual characters in the sea urchin, Echinometra mathaci. Publ. Seto Mar. Biol. Lab., 16: 41-50. YAMAGUTI, S., 1961. Systcma Hehnintlium. Volume HI. The Nematodes of Vertebrates. Parts 1 and 2. Interscience Publishers, New York, 1261 pp. Reference : Biol. Bull., 140 : 104-116. (February, 1971 ) LARVAL DEVELOPMENT OF PAGURUS LONGICARPUS SAY REARED IN THE LABORATORY. II. EFFECTS OF REDUCED SALINITY ON LARVAL DEVELOPMENT x MORRIS H. ROBERTS, JR. 2 Virginia Institute of Marine Science, Gloucester Point, Virginia 23062 Temperature and salinity define a set of conditions within which planktonic organisms can survive and develop. Thorson (1946) described the restriction of some meroplankters to Kattegat water in the Oresund which was presumed to be based on either a temperature, or a salinity discontinuity or both. Bary (1963a, b, c), in an extensive study of North Atlantic plankton, clearly demonstrated a relationship between zooplankton distribution and temperature-salinity distribution. Banse (1956) observed the distribution of polychaete and echinoderm larvae with respect to various water masses in Kiel Bay. He concluded that these larvae were restricted to their "Gebirtswasser" by the temperature-salinity characteristics of these water masses. In a subsequent paper (1959) he described a similar situa- tion for copepods. Survival and rate of development of decapod larvae are temperature-dependent phenomena. It has been shown for a variety of species that there is some optimal temperature range above and below which larval mortality increases (Boyd and Johnson, 1963; Chamberlain, 1961, 1962; Costlow, 1967; Costlow and Bookhout, 1962, 1968; Costlow, Bookhout and Monroe, 1960, 1962, 1966; Coffin, 1958, I960). There is a unique range of temperature permitting survival for each decapod species so far studied. Further, it has been demonstrated that intermolt duration decreases with increasing temperature. While temperature affects animal distributions both in the open sea and coastal waters, salinity has the greatest influence in coastal waters and estuaries. The ability of larvae to survive reduced salinity is therefore of extreme interest in the case of estuarine species. Considerable research has been conducted to elucidate the effects of salinity on larvae of estuarine brachyurans (Chamberlain, 1961, 1962; Costlow, 1967; Cost- low and Bookhout, 1962, 1968; Costlow et al, 1960, 1962, 1966). Salinity strongly affects survival with tolerance ranges unique for each developmental stage and each species. There seems to be little effect on intermolt duration until the lethal salinity is neared, at which point a slight increase can be detected. Informa- tion is available for only three anomuran species, Pisidia longicornis (Lance, 1964), 1 Contribution Number 368 from the Virginia Institute of Marine Science, Gloucester Point, Virginia 23062. 2 Present address : Department of Biology, Providence College, Providence, Rhode Island 02918. This paper is part of a dissertation submitted to the School of Marine Science of The College of William and Mary in partial fulfillment of the requirements for the Doctor of Philosophy Degree. 104 SALINITY TOLERANCE, PAGURUS LARVAE 105 Pagurtis samuelis (Coffin, 1958) and Pagurus bernhardus (Bookhout, 1964). The general trends are the same as for brachyurans. In the present study the effects of reduced salinity on embryos and larvae of Pagurus longicarpus were determined. Efforts were made to detect changes in salinity tolerance with increasing developmental age. Despite the importance of temperature in larval development, this parameter was not included as a variable in these studies. Limitations of equipment and time made it impractical to conduct experiments at several temperatures. While there may well be important interaction effects of salinity and temperature, it did not seem reasonable to pursue this aspect of the problem at this time. The four zoeae and megalopa of Pagurus longicarpus were described in an earlier paper (Roberts, 1970). It was reported that these larvae did not corre- spond to those described by Thompson (1903) as P. longicarpus. It was sug- gested that Thompson had been working with P. annulipes. Nyblade (1970), based on his study of the larvae of the latter species, concluded that Thompson was indeed studying P. annulipes. MATERIALS AND METHODS Embryological development Information on embryological development was obtained incidentally to the culture of eggs for various experiments. Eggs removed from the pleopods of a single female were incubated in filtered water at several salinities following the procedures developed by Costlow and Bookhout (1960). Quantitative observa- tions, such as per cent hatch, w T ere not attempted, but it was noted if development and hatching occurred and if there were any delay in hatching. Larval development Larvae for these experiments were hatched from females maintained in 5-liter battery jars containing water of 20/cc salinity. The water was replaced daily. Preliminary studies showed that eggs did not develop at all salinities included in this study. Further, no consistent difference was detected between mortality of larvae hatched at salinities from 15 to 30% c versus larvae hatched at 20%o and subsequently transferred to water with salinities from 15 to 30%c. Larvae were reared in water of 20% c on a diet of nauplii of Artemia to the desired zoeal stage for each experiment. Tolerance tests were conducted in small finger bowls containing 200 ml water, 10 larvae per bowl. Fifty larvae were subjected to each test salinity in the range 10 to 30%c. All experiments were conducted at 20 2 C, a tempera- ture intermediate in the range tolerated by these larvae, though not necessarily opti- mal. This temperature is somewhat below the mean summer temperature in the York River estuary, but was convenient for the present study. Four experiments of the above design were run, each experiment beginning with a different developmental stage. The experiment number corresponded with the zoeal stage used at the start of the test; thus Experiment 1 began with Zoea I. Experiment 2 with Zoea II and so forth. It was hoped in this manner to detect any increase (or decrease) in ability to tolerate reduced salinity as a function of 106 MORRIS H. ROBERTS, JR. zoeal stage and to detect any effect of prior culture conditions at the test salinity on each stage. In the four experiments described above, tests were terminated after the molt to the megalopa since in most cases too few larvae reached this stage in healthy condition to produce meaningful results. An independent test, Experiment 5, was conducted with megalopae obtained from mass culture (100-200 larvae in 1000 ml, 20%c, 20 C). Since megalopae are aggressive and, in mass culture, frequently kill one another, compartmented boxes were used with one megalopa per com- partment (50 ml). The same series of salinities was used as for tests with zoeae. TABLE I Hatching time as a function of incubation salinity and degree of development at start of incubation Days to hatching Initial degree of development Temperature 20 to 30 % & Eyespot* 30 2 2 Eyestreak** 30 3 3 Eyestreakf 25 4 4 f yolk*** 25 5 6 "Early"**** 25 7 9 "Early" 25 7 8 * Eyespot refers to eggs with dark brown or black circular eyespots and a beating heart. ** Eyestreak refers to eggs in which the eye is represented by a brown crescent-shaped spot. *** ^ yolk refers to eggs with yolk occupying about | the egg capsule. **** "Early" refers to eggs with at most a small white germinal disc apparent. Further details concerning these stages are given in Coffin (1960). t Artificial sea water (Rila mix) used in this experiment. Water for all tests was prepared by diluting high salinity water collected at the branch laboratory at Wachapreague, Virigina. Water was collected from Finney Creek, filtered through a 1 ^ filter, passed beneath ultraviolet lights, and collected in carboys. After transit to the main laboratory at Gloucester Point, the water was stored in large darkened carboys until used. RESULTS Embryological development Development of embryos was followed at salinities of 10, 15, 20, 25 and 3Q% . The results are summarized in Table I. Embryonic development was success- fully completed at all salinities except W% at which development proceeded for one or two days after which the yolk broke down. The eggs were greatly swollen and deteriorated rapidly. The same result was obtained even with late- stage eyed embryos placed in water of W% salinity. Viable larvae hatched at I5%e but generally 24-48 hours later than those hatched at higher salinities. The delay in development was more pronounced if eggs in an early stage of de- velopment were used, No effect was detected at salinities from 20 to 30% c . SALINITY TOLERANCE, PAGURUS LARVAE 107 Larval development: Survival In all experiments, many deaths occurred just before, during, or immediately after a molt, especially the fourth molt. Larvae dying just prior to a molt showed anlagen of structures present in the succeeding stage beneath the old integument and separation of hypodermis from exuvium. These larvae apparently could not split the old integument, or could not remove themselves from it before they swelled. Larvae dying during the molt swelled while partly out of the exuvium. These larvae, with abdomen and part of the cephalothorax removed from the exuvium, were recorded as the stage following the molt. At the lowest salinities, the dorsum of the carapace was sometimes shed before the abdomen ; these larvae were classed as the stage prior to the molt. Those dying immediately after the molt usually appeared weak at the last observation before death, and may have succumbed from the effort of completing the molt. In Experiment 1, complete development to the megalopa was observed at all salinities from 15.5 to 30.5%c. No significant difference in survival was noted from 18.0 to 30.5%c with 25 to 60% of the larvae reaching the megalopa. At I5.5 c / f c, only 8% reached the megalopa. At 13.0%, about 70% molted to Zoea II and 10% to Zoea III, but none to Zoea IV. At W.5% only 2% survived to Zoea II and none to Zoea III (Fig. la). At salinities from 15.5 to 30.5% , the slope of the survivorship curve was relatively constant up to Zoea IV but showed a marked increase for Zoea IV. At 13.0%o there was a sharp increase in mortality for Zoea II. Mortality observed immediately following transfer to 10.5 and 13.0/c was due to the low salinity, not the salinity difference involved. Larvae transferred from 20.5 to 30.5% f , a salinity difference of equal magnitude, though in the oppo- site direction, did not exhibit significant mortality immediately after transfer. Further, 1 or 2 days after the larvae had been transferred to 10.5 and 13.0%o, mortality dropped to zero, and remained at zero until the next molt which was considerably delayed (see below). Observations of larval development at salinities of 18.0 and 13.0%o were omitted from the series in Experiment 2, but the results were essentially the same as in Experiment 1 with complete development to the megalopa observed from 15.5 to 30.5% . From 20.5 to 30.5%,, 35 to 55% survived to the megalopa; at 15.5%, only 4%. No larvae reached Zoea III at 10.5%. Most larvae in this latter salinity died shortly after transfer, the remainder just prior to the molt. The slopes of the survivorship curves are essentially constant up to Zoea IV at all salinities except 10.5% (Fig. Ib). Complete development occurred over the salinity range 13.0 to 30.5% in Experiment 3. From 18.0 to 30.5%, 60 to 83% reached the megalopa, at 15.5%, 40% and at 13.0%, 5% (Fig. Ic). Thus at 15.5% there was an apparent increase in survival to the megalopa from 4-8% to 40% while at 13.0%, the increase is from 0% to 5%. Megalopae were obtained in every salinity including 10.5% in Experiment 4. Better than 80% survived at all salinities from 13.0 to 30.5%, and 5.3% at 10.5%, (Fig. Id). While it might seem that there is a marked increase in salinity tolerance with increasing developmental age, a comparison of the family of survivorship curves 108 MORRIS H. ROBERTS, JR. ioo- 50- \ ^ \ '" x -," 30- \ '\ ""-. ''\ - \ \ 20- \ \ \ \ \ \ \ < \ \ \ \ \ > 10- \ \ : v \ o: ~ \ \ ~ \ \ 5- \ \ \ \ \ \ \ 3- \ \ * i \ \ \ 2- ^ V \ x v \ i >o x % i 11111 , i i i 100- p^-^^__ 1 __^ : ^~~~. + ~ \ "" x - - - -T^^^ \ - \ \ ~ N x^^ >> ^ \ 50- \ v ^" '\ N \ >i X \ \ 30- \ \ \ 20- I \ \ < \ \ > \ \ 30.5 \ >\c\ oc c \ i^^ \- 25 5 \ _ \ _ j , \ <-{J O \ w S 5 5 ~ \ \ \ - 18.0 \ 15.5 \ 13.0 3- \ \ 10.5 2- \ 5 ,b \ * > d o i i l i i i 1 1 i n in rz M m iz M STAGE FIGURE 1. Survivorship curves for the salinity tolerance experiments; (a) Experiment 1, (b) Experiment 2, (c) Experiment 3, (d) Experiment 4. In Experiment 4, the results were identical for 30.5 and 20.5 % and for 18.0 and 15.5 %. for the four experiments suggests that this is not the case. The curves for each experiment are exactly like those of Experiment 1 if the latter are displaced an appropriate number of developmental stages to the right. SALINITY TOLERANCE, PAGURUS LARVAE 109 In order to compare the mortality in each zoeal instar in the various experi- ments, the per cent mortality was calculated as follows : %M = ^^-Jb+1 X 100 Hi where Hi = number of Zoea i ;+i = number of Zoea fc + 1 These values are arrayed in Table II. Assuming that the per cent mortalities so obtained are binomially distributed, it is possible to estimate 95% confidence limits from tables for the appropriate sample size and point estimate (Diem, 1962). From a consideration of the 95% confidence limits and the point esti- mates for each cell in the array in Table II, I was led to the conclusion that three values were not adequate estimates of the mortality resulting from salinity. These values are enclosed in brackets. The initial zoeal stage used in each experiment had a lower per cent mortality than succeeding stages. Indeed, no difference could be detected in the per cent mortality of the initial stage of each experiment at any given salinity. This can be seen by comparing the per cent mortality for the initial stage used in each experi- ment at any given salinity in Table II. This results from the selection of only healthy larvae, i.e., those showing a strong positive light response, in setting up each experiment. The increase in mortality in succeeding stages, which usually was not very great except in Zoea IV, reflects at least in part the fact that some larvae remaining after the initial molt were not healthy. There was no significant difference in per cent mortality for any given zoeal stage over the salinity range 18.0 to 30.5% . This can be seen by comparing values for any given stage, e.g., Zoea III, over the entire range of salinity. Per cent mortality for the initial stage in each experiment at IS.S% was in the same range as that of 18.0 to 30.5%c, but many larvae were weakened as indicated by the significant increase in mortality in the succeeding stage. The same effect was noted at 13.0%o but at a significantly higher mortality level. If tolerance of reduced salinity had improved with increasing developmental age, one would expect a marked improvement in survival of later stages in \5.5%o in the later experiments. Per cent mortality for the megalopa (Experiment 5) was less than 6% from 18.0 to 30.5% c , which agrees with the results obtained for the zoeal stages. At lower salinities per cent mortality increased rapidly to 100% at 10.5%e. At lS.S%o the per cent mortality was significantly higher than at higher salinities with most deaths occurring near the time of molt to the juvenile. Deaths at 13.0%e also were associated with molting, while those at 10.5%e were associated with the stress produced by transfer to this salinity. This may represent a slightly lessened ability of the megalopa to tolerate reduced salinity. Larval development: Intermolt duration The mean time after hatching to each zoeal stage, in days, was calculated for all experiments. The number molting during each 24-hour period was asso- 110 MORRIS H. ROBERTS, JR. TABLE II Per cent mortality for the zoeal stages and megalopae at each salinity Salinity Experiment number Zoeal stages I II Ill IV Megalopa 30.5 1 2.0 8.2 15.5 31.4 / 2 [21.6] 12.8 50.0 / 3 0.0 22.5 / 4 2.6 / 5 5.6 25.5 1 [18.0] 9.8 [24.3] 51.8 / 2 10.0 6.7 57.2 / 3 . / 4 5.3 / 5 0.0 20.5 1 2.0 7.1 11.0 29.0 / 2 14.0 4.7 34.9 / 3 . 2.5 15.4 / 4 2.6 / 5 2.8 18.0 1 6.0 10.6 4.8 67.5 / 2 / 3 5.0 31.6 / 4 . 0.0 / 5 5.6 15.5 1 7.0 54.8 28.6 73.5 / 2 20.0 47.6 90.5 / 3 15.0 50.0 / 4 0.0 / 5 22.2 13.0 1 32.0 88.3 100.0 / / 2 / 3 35.0 92.0 / 4 15.8 / 5 66.7 10.5 1 98.0 100.0 / / / 2 100.0 / / / 3 97.6 100.0 / 4 94.7 / 5 100.0 ciated with the mean time after hatching for that period for the purposes of calcula- tion. This procedure assumes that the frequency of molting was normally dis- tributed through time. Mean post-hatching time, range of post-hatching time, and number of larvae involved for each zoeal stage are presented in Table III and Figure 2. Intermolt duration for each zoeal stage was then calculated as the difference between the SALINITY TOLERANCE, PAGURUS LARVAE 111 mean post-hatching times from Experiment 1, except in cases where no informa- tion was available. When information was available from the other experiments, it was used instead. Rigorous statistical comparison of mean post-hatching time for each molt was not possible because the observation intervals were very large relative to the interval between the first and last zoea completing a given molt. Such intensive grouping causes a gross underestimate of the true variance, thereby increasing the chance of rejecting the null hypothesis that the mean post-hatching times were equal, when it should be accepted. Intensive grouping also tends to bias the CO Q UJ 2 LEGEND EXPERIMENT NUMBERS 2 3 4 10 - 1 1 I 1 x- - j T T I 1 1 5 - i 1M i 1 I 1 T^ 1 1 10 - m 15 - 20- i 10- 1 ! i 11 1 T T ! ." + i i i ! * ' "Jl ' 7! it "i t T i 15 - - ' 1 '!--.! i , .. - : T i 1 5 - * el I 10 n l n 1 11 /-^ 1 12 305 25.5 205 180 155 130 10.5 SALINITY (% ) FIGURE 2. Range and mean post-hatch time for each zoeal stage. estimate of the mean in an indeterminate manner since there is no way of knowing in any given case whether the larvae molted at the beginning or end of an observation period. The intermolt duration over the salinity range 15.5 to 30.5% was 3.0 to 3.5 days for Zoea I, 2.7 to 3.8 days for Zoea II, 3.5 to 4.6 days for Zoea III, 4.6 to 6.4 days for Zoea IV. Intermolt duration increased slightly at 13.0 and 10.5% . Mean post-hatching times derived from Experiments 1 to 4 show very close agreement at each salinity except in some cases for the molt from Zoea IV to megalopa (Fig. 2). The poor agreement in this case results in part from the small numbers of larvae involved. Intermolt duration for the megalopa was obtained directly since the protocol for this experiment was slightly different. The intermolt duration was 6.5 to 7.5 days at salinities from 18.0 to 30.5/ , increasing to 8 and 9 days at 15.5 and 13.0% , respectively. The increased intermolt duration for Zoea IV and the 112 MORRIS H. ROBERTS, JR. TABLE III Mean post-hatching times to the end of each zoeal stage and intertnolt durations for each larval stage Develop- mental stage Experiment number Salinity. %o <(!.. 25.5 20.5 18.0 15.5 13.0 10.5 Zoea I 1 x 2.95 3.50 3.04 3.52 3.54 4.59 8.50 Interrnc r (2-4) n 49 (2-5) 41 (2-5) 98 (3-5) 47 (2-5) 93 (4-6) 34 (7-10) 2 It duration 3.0 3.5 3.1 3.5 3.5 4.6 8.5 Zoea II 1 x 5.88 6.23 5.82 6.60 7.31 8.75 r (3-8) n 45 (5-9) 37 (4-10) 91 (6-8) 42 (5-12) (8-10) 42 4 2 x 6.04 5.81 5.76 6.40 Intermc r (4-8) n 39 (4-7) 45 (4-8) 43 (5-9) 40 lt duration 2.9 2.7 2.7 3.1 3.8 4.2 Zoea III 1 x 9.50 10.81 9.94 10.95 10.80 . r (8-12) n 38 (8-15) 28 (7-14) 81 (9-14) 40 (9-16) 30 2 x 9 . 85 9.67 9.70 __ 10.83 r (8-13) n 34 (8-11) 42 (8-12) 41 (9-16) 21 3 x 10.40 . 9.81 9.71 10.24 10.81 1 1 . 50 Intermc r (9-12) n 40 (8-13) 39 (8-13) 38 (9-13) 34 (9-13) 26 (12) 1 )lt duration 3.6 4.6 4.1 4.4 3.5 (5.0) (5.7) Zoea IV 1 x 15.12 16.58 15.01 17.35 15.38 . - r (12-19) n 26 (12-22) 13 (11-23) 58 (15-19) 13 (14-18) 8 2 x 17.26 17.19 17.69 . . 19.50 r (13-21) n 17 (13-20) 24 (14-22) 27 (18-21) 2 3 x 16.40 . - 16.17 16.04 17.09 16.50 r (14-20) n 31 (14-20) 33 (14-20) 26 (15-20) 17 (15-18) 2 4 x 17.12 16.78 16.82 16.74 16.84 17.09 19.00 Intermc r (15-19) n 37 (15-18) 36 (15-19) 37 (15-21) 38 (15-18) 38 (15-20) 32 (16-21) 2 )lt duration 5.6 5.8 5.1 6.4 4.6 (5.7) (8.1) Megalopa* 5 x 7.35 r (6-8) n 34 6.47 (5-8) 36 6.55 (5-9) 35 7.04 (5-9) 34 8.18 9.08 (6-10) (7-11) 28 12 * Values for this stage are given as intermolt duration, in days, rather than post-hatching time, in days, as for all other stages. SALINITY TOLERANCE, PAGURUS LARVAE 113 megalopa compared to earlier zoeal stages has been noted before in Petrochims diogenes (Provenzano, 1968). DISCUSSION The notion that certain life history stages are more subject to limitation by abiotic environmental factors such as salinity, was first enunciated by Shelford (1915). Since then, considerable evidence has been amassed demonstrating that younger stages tend to be less tolerant than adults. Among decapod crusta- ceans, perhaps the best documented example is the blue crab, Callinectes sapidus, which can tolerate salinities from fresh to oceanic as an adult, but must return to water with a salinity in excess of I5%o for hatching of the eggs. Complete larval development occurs only at 20%c and above (Sandoz and Rogers, 1944; Costlow and Bookhout, 1959; Costlow, 1967). Other examples from various phyla may be found in recent reviews by Kinne (1964, 1966) and other papers in the literature. Salinity tolerance of decapod embryos has received only cursory attention. Broekhuysen (1936) cultured Carcimis maenas eggs at salinities from 10 to 5Q% C . At salinities from 20 to 40% , 16 C, and 25 to 40% , 10 C, complete embryonic development occurred, while at salinities above and below this range, development occurred to a degree, but hatching was not observed. At 10%c, no embryonic development was detected. Tolerance of each larval stage is unknown, but the adult tolerance is from 4 to 34% salinity (or above). Hatching of Hepatus epheliticus has been observed at all salinities from 20 to 40%c which is a greater range than tolerated by either larvae or adults (Costlow and Book- hout, 1962). In the present study, hatching of P. longicarpus was observed from 15 to 30%% c (Kinne. Shirley and Meen. 1963). I am indebted to Dr. Langley Wood, who served as chairman of my graduate committee, for his patience and helpful criticisms. I also appreciate the many fruitful discussions with Dr. Marvin L. Wass and Mr. Willard A. Van Engel and their critical reading of the manuscript. Dr. Morris L. Brehmer provided space in his laboratory and a critical review of the manuscript. Dr. Edwin Joseph provided a constant temperature water bath. I am very grateful to Mrs. Jane S. Davis and Mrs. K. Stubblefield for their patient aid in preparing the figures. I owe special thanks to my wife, Beverly Ann, for her constant encouragement and aid in the laboratory. During the course of this study I was the recipient of a National Science Foundation Graduate Fellowship. LITERATURE CITED BANSE, K., 1956. Uber den Transport von meroplanktischen Larven aus dem Kattegat in die Kieler Bucht. Ber. Dent. Wiss. Komm. Mccrcsjorsch., 14 : 147-164. BANSE, K., 1959. Die Vertikalverteilung planktischer Copepoden in der Kieler Bucht. Bcr. Deut. JViss. Komm. Meeresjorsch., 15: 357-388. BARY, B. M., 1963a. Temperature, salinity and plankton in the eastern North Atlantic and coastal waters of Britain, 1957. I. The characterization and distribution of surface waters. /. Fish. Res. Board, Can., 20: 789-826. BARY, B. M., 1963b. Temperature, salinity and plankton in the eastern North Atlantic and coastal waters of Britain, 1957. II. The relationships between species and water bodies. /. Fish. Res. Board, Can., 20: 1031-1065. BARY. B. M., 1963c. Temperature, salinity and plankton in the eastern North Atlantic and coastal waters of Britain, 1957. III. The distribution of zooplankton in relation to water bodies. /. Fish. Res. Board, Can.. 20: 1519-1548. BOOKHOUT, C. G., 1964. Salinity effects on the larval development of Pagurus bernhardiis (L.) reared in the laboratory. Ophelia, 1 : 275-294. BOUSFIELD, E. L., 1955. Ecological control of the occurrence of barnacles in the Miramichi Estuary. Bull. Nat. Mus. Canada, 137: 1-69. BOYD, C. M., AND M. W. JOHXSOX, 1963. Variations in the larval stages of a decapod crusta- cean, Pleuroncodcs planipes Stimpson (Galatheidae). Biul. Bull., 124: 141-152. BROEKHUYSEN, G. J., JR., 1936. On development, growth, and distribution of Carcinulcs maenas (L.). Arch. Kccrl. Zool, 2: 257-399. COFFIN. H. G., 1958. The laboratory culture of Pagunis saiiniclis Stimpson. Walla ll'alla Coll. Publ, 22: 1-5. 116 MORRIS H. ROBERTS, JR. COFFIN, H. G., 1960. The ovulation, embryology, and developmental stages of the hermit crab, Pagurus sainitclis Stimpson. Walla Walla Coll. Publ., 25 : 1-30. CHAMBERLAIN, N. A., 1961. Studies on the larval development of Neopanope texana sayi (Smith) and other crabs of the Zanthidae (Brachyura). Chesapeake Bay Institute, Johns Hopkins Univ., Tech Rep., 22 : 1-35. CHAMBERLAIN, N. A., 1962. Ecological studies of the larval development of Rhithropanopeus harrisii (Xanthidae, Brachyura). Chesapeake Bay Institute, Johns Hopkins Univ., Tech. Rep., 28: 1-47. COSTLOW, J. D., JR., 1967. The effect of salinity and temperature on survival and meta- morphosis of megalops of the blue crab, Callinectes sapidus Rathbun. Helgolaender Wiss. Mecresiintcrs., 15 : 84-97. COSTLOW, J. D., JR., AND C. G. BOOKHOUT, 1959. The larval development of Callinectes sapidus Rathbun reared in the laboratory. Biol. Bull., 116: 373-396. COSTLOW, J. D., JR., AND C. G. BOOKHOUT, 1960. A method for developing Brachyuran eggs in vitro. Litnnol. Oceanogr., 5 : 212-215. COSTLOW, J. D., JR., AND C. G. BOOKHOUT, 1962. The larval development of Hepatus epheliticus (L.) under laboratory conditions. /. Elisha Mitchell Sci. Soc., 78: 113- 125. COSTLOW, J. D., JR., AND C. G. BOOKHOUT, 1968. The effect of environmental factors on the development of the land-crab Cardisoma guanhumi Latreille. Amer. Zool., 8 : 397-410. COSTLOW, J. D., JR., C. G. BOOKHOUT AND R. MONROE, 1960. The effect of salinity and temperature on larval development of Sesarma cinereum (Bosc) reared in the labora- tory. Biol. Bull.. 118: 183-202. COSTLOW, J. D., JR., C. G. BOOKHOUT AND R. MONROE, 1962. Salinity-temperature effects on the larval development of the crab, Panopcus herbstii Milne-Edwards, reared in the laboratory. Phys'wl. Zool., 35: 79-93. COSTLOW, J. D., JR., C. G. BOOKHOUT AND R. MONROE, 1966. Studies on the larval develop- ment of the crab, Rhithropanopeus harrisii (Gould). I. The effect of salinity and temperature on larval development. Physiol. Zool., 39: 81-100. DIEM, K. (Ed.), 1962. Docuincnta Geigy, Scientific Tables. [6th Edition] Geigy Pharma- ceuticals, Ardsley, New York, 778 pp. KINNE, O., 1964. The effects of temperature and salinity on marine and brackish water animals. II. Salinity. Oceanogr. Mar. Biol., Ann. Rev., 2: 281-339. KINNE, O., 1966. Physiological aspects of animal life in estuaries with special reference to salinity. Netherlands J. Sea Res., 3: 222-244. KINNE, O., E. K. SHIRLEY AND H. E. MEEN, 1963. Osmotic responses of hermit crabs (Pagurus longicarpus Say) exposed to various constant temperatures and salinities. Crustaceana, 5: 317. LANCE, J., 1964. The salinity tolerances of some estuarine planktonic crustaceans. Biol. Bull, 127: 108-118. NYBLADE, C., 1970. Larval development of Pagurus ainnilipcs (Stimpson, 1862) and Pagurus pollicaris Say, 1817 reared in the laboratory. Biol. Bull., 139: 557-573. PROVENZANO, A. J., JR., 1968. The complete larval development of the West Indian hermit crab Pctrochinis diotjcnes (L.) (Decapoda, Diogenidae) reared in the laboratory. Bull. Mar. Sci., 18: 143-181. ROBERTS, M. H., JR., 1970. Larval development of Pagurus longicarpus Say reared in the laboratory, I. Description of larval instars. Biol. Bull., 139: 188-202. SANDOZ, M., AND R. ROGERS, 1944. The effect of environmental factors on hatching, moulting, and survival of the zoea larvae of the blue crab, Callinectes sapidus Rathbun. Ecology, 25: 216-228. SHELFORD, V. E., 1915. Principles and problems of ecology as illustrated by animals. /. Ecol, 3: 1-23. THOMPSON, M. T., 1903. The metamorphosis of the hermit crab. Proc. Boston Soc. Natur. Hist., 31 : 147-209. THORSON, G., 1946. Reproduction and larval development of Danish marine bottom inverte- brates. Mcdd. Dan. Fisk. Havundcrs, 4 : 1-523. Reference: Biol. Bull., 140: 117-124. (February, 1971) BIPOLAR HEAD REGENERATION IN PLANARIA INDUCED BY CHICK EMBRYO EXTRACTS * LEWIS V. RODRIGUEZ AND REED A. FLICKINGER Department of Biology, State University of New York at Buffalo, Buffalo. \ civ York When a planarian is transected the anterior cut end will regenerate a head while the posterior end will give rise to a tail. Child (1941), postulated in his axial gradient theory that quantitive axial gradients of physiological activity some- how account for axial patterns of differentiation. This theory predicts that the higher level of physiological activity at the anterior cut end leads to the formation of the head at that site. An alternative suggestion is that polarity is maintained by the presence of specific inhibitors which similarly follow an axial gradient pattern. When a planarian is decapitated, the brain forms before the eye spots or auricles. However, regeneration of the brain does not occur if another brain is present (Morgan, 1902; Rand and Ellis, 1926; and Miller, 1938). Such effects could be due to specific inhibitors or they could result from a metabolic competition. Decapitated planarians reared in water containing centrifugal supernatants of head homogenates had smaller brains, or the brains were absent, suggesting there are specific inhibitor substances (Lender, 1960). The aim of this investigation was to determine if polarity is determined by the presence of specific inhibitors or if it is the function of an axial gradient of physiological activity. Regeneration was allowed to occur in the presence of centrifugal supernatants of either heads or tails of planaria, Dugesia dorotocephala, to determine if these fractions possess inhibitory activity. In order to test the idea that competitive metabolic processes may play a role in determining polarity, small cut pieces of planaria were cultured in chick embryo extract and ultrafiltrate to see if these nutrients provided in these preparations could alter the normal re- generation polarity by inducing the formation of bipolar heads. MATERIALS AND METHODS Planarians, Dugesia dorotocephala, which were 1-2 centimeters in length were maintained in aerated tap water in covered containers containing 100 units of penicillin and streptomycin per milliliter and were starved for at least 5-7 days prior to use in an experiment. Before cutting, the worms were washed several times with boiled tap water, then while beneath a shallow layer of water they were exposed to bacteriocidal ultraviolet light for five minutes and then cut with sterile single edge razor blades. In obtaining the centrifugal supernatants for the inhibitor experiments the worms were cut into three equal parts, thus providing the head, mid-body and tail fractions. The middle piece always included the 1 This research was supported by grants from the National Science Foundation (GB-5500) and the National Institute of Health (GM-16236-01). 117 118 I.HYY1S V. RODRIGUEZ AND REED A. FLICKINGER pharynx. The protein concentrations of the centrifugal snpernatants were deter- mined by the Lowry method (Lowry, Rosebrough, Farr and Randall, 1951). The chick embryo extracts used in these experiments were obtained in the frozen state from Grand Island Biological Company. The effect of the chick embryo frac- tions upon DNA synthesis in pieces of planarians was tested by exposing pooled segments of planaria to 25 ^ c/ml of H 8 -thymidine in 2 ml of the chick embryo extract solutions and in a 1/10 dilution of a balanced salt medium (Niu-Twitty, 1953). At the conclusion of the incubation period the worms were washed five times with boiled tap water. The worms were then homogenized in cold 7% trichloroacetic acid (TCA) and the pellet washed three times by centrifugation. The residues were extracted twice with 1:1 ethanol-ether, twice with 1:4 ethanol- ether and once with ether alone to remove lipids ; the residues were then dried. . , The diphenylamine method (Dische, 1955) was used to determine DNA concen- tration" and levels of labeled DNA were estimated with a liquid scintillation counter. ^ RESULTS Experiments with centrifugal supernatants The head, tail and midbody centrifugal supernates used for the inhibitor experiments were prepared by homogenizing at 4 C in aerated tap water containing 50 units of penicillin and sterptomycin per milliliter. Homogenates were centri- fuged at 10,000 X g for 30 minutes at 4 C and the supernatants were collected. The supernatants were passed through sterile millipore filters of 0.45 and 0.22 p pore diameter before use. Heads were removed from a group of 25 worms by cutting just posterior to the auricles and the decapitated worms were exposed for 9 days at 20 C to supernatants of homogenized head, tail and midbody sections at concentrations between 2 and 100 pieces per two milliliters of homogenation medium. In terms of protein concentration, cut planaria were exposed to centri- fugal supernatants from head regions with a protein concentration of 6.05 ju.g/ml to 605 ftg/ml and to centrifugal supernatants from tail regions with protein con- centrations of 4.58 jug/ml to 458 /xg/ml. Fresh supernatants were prepared daily and the media changed each day of the nine day exposure period. Control and test worms were observed daily for the appearance of eyespots and auricles. After nine days of culture, test and control worms were fixed at which time both eye spots and auricles had appeared. They were then embedded in paraffin, 5 /j. serial sections made and sections were stained with 0.1% cresyl violet. All planaria exposed to head and tail concentrations of 2 through 87 pieces per two milliliters survived the nine days of incubation and regenerated normal eyespots and auricles. Tail and head concentrations of 100 pieces/2 ml killed test worms within 48 hours after first exposure. In control and test worms in- cubated in non-lethal supernatant concentrations, auricles appeared between 48- 72 hours, detectable eyespots between 60-96 hours, fully discernable eyespots be- tween 85-120 hours and fully regenerated heads by 6 days. There was no sig- nificant delay in time of appearance of the eyespots in the centrifugal supernatant fractions. Serial sections of control and test worms showed the presence of brains of equal length I 175 5 //, ) in both the control and test worms (Fig. 1). BIPOLAR PLANARIAX REGENERATION a 20/i FIGURE I. Transverse sections of heads that formed after 9 days of regeneration in (a) bubbled tap water and (b) a centrifugal supernate obtained by homogenizing 50 planarian heads in 2 ml of bubbled tap water. The more lightly stained tissues in the center of the sections are the brains. Effect of chick embryo c.vtract on polarity In a preliminary experiment, to determine what body levels and which nutrient medium produced bipolars, the planaria were transversely cut at 8 body levels (Fig. 2) and the cut segments of worms allowed to regenerate in chick embryo extract and chick embryo extract ultrafiltrate. Regenerating pieces from posterior levels 5 and 6 cultured in 1% chick embryo extract and \% chick embryo extract 120 LEWIS V. RODRIGUEZ AND REED A. FLICKINGER FIGURE 2. The levels at which worms were cut for the experiments with chick embryo extract. ultrafiltrate in 1/10 Niu-Twitty saline (Niu and T witty, 1953) were the only levels showing bipolar head regeneration, i.e., heads forming at each end of the cut segments of the worms. Segments of worms from regions 1 to 4 and 7 to 8 regenerated heads at the anterior end and tails at the posterior ends. No bipolar heads arose in pieces cultured in saline alone and these pieces formed heads at the anterior end surface and tails at the posterior end. In further experiments, twenty sections each of posterior body levels 5 and 6 from worms starved for 5, 10, 20 and 30 days were cultured in saline and nutrient media. The sections were exposed to test solutions for 4 days, after which regeneration was completed in 1/10 saline (Table I). No bipolar head regenera- tion occurred in the short posterior pieces from worms starved 5-10 days. However, 6 bipolar heads out of 26 cut pieces regenerated from cut posterior level 5 pieces and 8 bipolar heads out of 21 cut pieces were formed in posterior level 6 cut pieces in nutrient media from worms starved for 20 days. After 30 days of starvation, 9 of 34 level 5 pieces formed bipolar heads in nutrient media, while 7 of 25 level 6 pieces in nutrient media regenerated bipolar heads. None of the posterior level pieces produced bipolar heads when the medium was 1/10 Niu Twitty saline (Table I). A typical regenerate with a bipolar head is shown in Figure 3. BIPOLAR PLANARIAN REGENERATION 121 TABLE I Effect of nutrient media on incidence of bipolar head formation Number of bipolars Level Number of surviving sections Medium of worm Days of starvation 5 10 20 30 (1.) 1/10 Niu-Twitty saline + 1% chick embryo extract 5 0/20 = 0/19 = 4/10 = 40' , 4/16 = 25% (2.) 1/10 Niu-Twitty saline + 1 % chick embryo extract ultrafiltrate 5 0/19 = 0/19 = 2/16 = 12.5% 5/18 = 27.8% (3.) 1/10 Niu-Twitty saline 5 0/20 = 0/18 = 0/16 = 0/14 = (1.) 1/10 Niu-Twitty saline + 1% chick embryo extract 6 0/20 = 0/19 = 3/5 = 60% 3/12 =25% (2.) 1/10 Niu-Twitty saline + 1% chick embryo extract ultrafiltrate 6 0/20 = 0/19 = 5/16 = 31.25' , 4/13 = 30.7% (3.) 1/10 Niu-Twitty saline 6 0/20 = 0/17 = 0/19 = 0/15 =0 Isotopic experiments with worms exposed to chick embryo extract It is possible that the chick embryo extract causes bipolar head regeneration in short posterior pieces from starved planaria by stimulating cell division. In order to determine if the chick embryo extracts had stimulated DNA synthesis and cell division in the regenerating sections an experiment using H 8 -thymidine was performed. One hundred and fifty worms starved for 30 days were cut into lOO/i FIGURE 3. A posterior section of a worm which regenerated a head at each end after culture in 1% chick embryo extract. 122 LEWIS V. RODRIGUEZ AND REED A. FLICKINGER head (levels 1-4), midbody (between levels 4 and 5) and tail pieces (levels 5-8) (Fig. 2). Fifty such sections of each level were each incubated in two nutrient media and saline. To each of the media H 3 -thymidine was added to a final con- centration of 25 IL c/ml and cut worms were incubated for 24 hours. The results of this experiment clearly demonstrate a stimulation of DNA synthesis of all three levels of the pooled cut worms which had been cultured in nutrient medium (Table II). There was a greater stimulation of DNA synthesis at middle and posterior levels than at anterior levels. Those cut pieces giving rise to bipolar heads were found using posterior levels of cut worms (Table I). TABLE 1 1 Effect of nutrient media on DNA syntheses in plunaria starved for 30 days. Fifty cut pieces were incubated in 25 p. c/ml H"-thyniidine for 46 hours CPM MG DNA Level of worm % stimulation DNA synthesis compared to controls Medium Anterior Middle Posterior Anterior Middle Posterior (1.) 1/10 Niu-Twitty saline + 1' , chick embryo extract 21,400 16,756 14,872 19.4 37.9 32.1 (2.) 1/10 Niu-Twitty saline + 1' , chick embryo extract ultrafiltrate 18,460 15,539 12,600 13.7 33.1 19.8 (3.) 1/10 Niu-Twitty saline-control 17,238 10,391 10,100 DISCUSSION According to Lender (1960) decapitated planarians reared in water containing centrifugal supernatants of head homogenates either did not regenerate brains, or the brains were of reduced size. Lender felt that the diffusion of inhibitory sub- stances from the brain gave rise to an anterior-posterior gradient of inhibition. In recent work on inhibitors Ziller-Sengel (1967a, 1967b) has reported that the pharyngeal region of the planarian Dugesia lugitbris contains a specific inhibitor which delays the regeneration of the pharynx. The inhibitor is present in filtered extracts of the pharyngeal region and is species-specific since homogenates of the pharyngeal regions of Dugesia tigrina do not inhibit regeneration of the pharynx in Dugesia lugubris. The results of this investigation do not demonstrate the presence of a specific inhibitor of brain regeneration in 10,000 X g supernatants of heads of Dugesia dorotocephala. Centrifugal supernatants of homogenates from the three body levels at concentrations of 2 to 87 cut sections in 2 milliliters of homogenization medium (6-600 /xg protein /ml) do not exhibit inhibitory activity. Brondsted (1955) showed that a head regenerates in a window cut out of the fore part of the planarian, Bdellocephala, when the old head is present. This implies that inhibitory substances from a certain body structure do not travel through the body to inhibit the regeneration of identical structures. Brondsted's BIPOLAR 1'LAXARIAX REGENERATION 123 experiment, and the reports of Ziller-Sengel (1967 a, 1967b), showing a delay in the regeneration of the pharynx and not a loss or a reduction in size of the regener- ated organ, do not support the role of specific inhibitors in maintaining polarity. Exposure of posterior cut sections of planaria to nutrient media (chick embryo extract and chick embryo extract ultrafiltrate) after a sufficient time of starvation of the worms does alter the polarity of the regenerating pieces. It has been shown that cephalocaudal physiological gradients exist in planaria and that polarity of regenerates can be reversed by abolishing such gradients ( Flickinger, 1959; Flickinger and Coward, 1962; Coward, 1968). Nutrient media (Coward, Flickinger and Garen, 1964) accelerate both the rate and extent of regeneration in posterior regions of cut starved planarians. Although very short cut pieces often form heads at both ends (Janus heads), this did not occur in the control pieces in saline. The levels of the worm just posterior to the pharynx (levels 5 and 6) are in the fission zone of the planarian and thus may account for bipolar head formation in this region. It has been demonstrated that inhibition of DNA synthesis is necessary in order to produce bipolar head regeneration with cloramphenicol (Kohl and Flickinger, 1966). Therefore, it seemed important to learn if a stimulation of DXA synthesis, and presumably cell division, occurred when bipolar head forma- tion is induced by stimulation, rather than by inhibition. The results of the H 3 - thymidine incorporation experiments in which anterior, middle and posterior seg- ments were regenerating in saline show that after 30 days of starvation there still exists a cephalocaudal gradient of DNA synthesis as described by Coward and Flickinger (1965), and Kohl and Flickinger (1966). However, no difference in DXA synthesis existed between the pooled middle and posterior regions. Further- more, posterior body levels showed a greater increase in the percentage of stimula- tion of DNA synthesis with nutrient media than does the anterior level, indicating a greater stimulation of cell division in these levels than in the head level. The results of this investigation, and the reported work of previous investi- gators, strongly suggest the presence of a cephalocaudal metabolic gradient in planarians in which metabolic competition results in the dominance of anterior levels over each succeeding posterior level. Xo evidence for specific inhibitor substances was found in this study. SUMMARY 1. Decapitated planaria. Dugesia dorotocephala, exposed to head and tail centrifugal supernatants for nine days do not exhibit inhibition of brain regenera- tion. The experimental and control worms had brains of equal sizes. 2. A significant number of bipolar heads regenerated in small pieces from the posterior region of starved worms which were cultured in chick embryo extract and chick embryo extract ultrafiltrate. 3. Chick embryo extracts produce a stimulation of DNA synthesis in pieces of starved worms which were cut into anterior, middle and posterior sections. There is a greater stimulation at the posterior levels which can give rise to bipolar heads. This stimulation of DNA synthesis is thought to reflect an increase in cell division at these levels caused by the chick embryo extract. 124 LEWIS V. RODRIGUEZ AND REED A. FL1CKINGER LITERATURE CITED BRONDSTED, H. V., 1955. Planarian regeneration. Biol. Rev., 30 : 65-126. CHILD, C. M., 1941. Problems and Patterns of Development. University of Chicago Press, Chicago, 727 pp. COWARD, S. J., 1969. The relation of surface and volume to so-called physiological gradients in planaria. Develop. Biol., 18: 590-601. COWARD, S. J., AND R. A. FLICKINGER, 1965. Axial patterns of protein and nucleic acid synthesis in intact and regenerating planaria. Groivth, 29: 151-163. COWARD, S. J., R. A. FLICKINGER AND E. GAREN, 1964. The effect of nutrient media upon head frequency in regenerating planaria. Biol. Bull., 126 : 345-353. DISCHE, Z., 1955. Color reactions of the nucleic acid components. Pages 285-305 in E. Char- gaff and J. Davidson, Eds., The Nucleic Acids, Vol. 1. Academic Press, New York. FLICKINGER, R. A., 1959. A gradient of protein synthesis in planaria and reversal of axial polarity of regenerates. Groivth, 23: 251-271. FLICKINGER, R. A., AND S. J. COWARD, 1962. The induction of cephalic differentiation of regenerating Dugesia dorotocephala in the presence of the normal head and in un- wounded tails. Develop. Biol., 5: 179-204. KOHL, D. M., AND R. A. FLICKINGER, 1966. The role of DNA synthesis in the determination of axial polarity of regenerating planarians. Biol. Bull., 131 : 323-330. LENDER, T. H., 1960. L'inhibition specifique de la differentiation du cerveau des Planaires d'eau douce en regeneration. /. Embryo!. Exp. Morphol., 8: 291-301. LOWRY, O. H., N. J. RosEBRoroH, H. L. FARR AND R. J. RANDALL, 1951. Protein measure- ment with the Folin phenol reagent. /. Biol. Chein., 193 : 265-276. MILLER, J. A., 1938. Studies on heteroplastic transplantation in triclads. I. Cephalic grafts between Euplanaria dorotocephala and E. tigrina. Pliysiol. Zoo!., 11: 214-247. MORGAN, T. H., 1902. Growth and regeneration in Planaria htgiibrus. ll'ilhclm Roux Arch. Entwicklungsmech. Organismen, 13 : 179-212. Niu, M. C., AND V. C. TWITTY, 1953. The differentiation of gastrula ectoderm in medium conditioned by axial mesoderm. Proc. Nat. Acad. Sci., Washington, 39: 985-989. RAND, H. W., AND M. ELLIS, 1926. Inhibition of regeneration in two-headed or two-tailed planarians (maculata). Proc. Nat. Acad. Sci., Washington. 12: 570-574. ZILLER-SENGEL, C., 1967a. Recherches sur 1'inhibition de la regeneration du pharynx chez les planaires. I. Mise en evidence d'un facteur auto-inhibiteur de la regeneration du pharynx. /. Embryol. E.rp. Morphol.. 18: 91-105. ZILLER-SENGEL, C., 1967b. Recherches sur 1'inhibition de la regeneration du pharynx chez les planaires. II. Variations d'intensite du facteur inhibiteur suivant les especes et les phases de la regeneration. /. Embryol. Exp. Morphol.. 18: 107-119. Reference : Biol. Bull., 140: 125-136. (February, 1971 ) FIXE STRUCTURE AND COMPOSITION OF A SILICEOUS SPONGE SPICULE * DANIEL W. SCHWAB AND RICHARD E. SHORE Fundamental Research, Owens-Illinois, Inc., Toledo, Ohio 43601 and Department of Biology, University of Toledo, Toledo, Ohio 43606 \Yhen a siliceous skeletal spicule of a marine sponge is dissolved in hydro- fluoric acid (HF), a filament remains. We report here a new understanding of the structure and composition of the siliceous spicule and its axial filament. Biitschli (1901) showed that these filaments stain with several organic dyes and behave toward a variety of reagents as protein. He concluded that the HF-resis- tant residue, presumably all axial filament, contained a very small amount of organic material around which silica deposition is initiated. The view that the tc-/; ole axial filament, or even a major fraction of it, is "sans doute proteinique" (page 497 Levi. 1963) appears to be a misreading of this statement by Butschli. Biitschli further showed that spicules fractured in cross section and stained with various dyes exhibit a triangular core and several regularly spaced concentric rings. The laminar structure has been widely documented for large spicules. The early literature is exhaustively reviewed by Minchin (1909). Drum (1968^ applied rotary replication techniques in the study of the surface ultrastructure of siliceous spicules. In his electron micrographs he points out a filament that remains after HF removal of the spicule. From its binding of dyes and its loss on ultramicroincineration he concluded that the filament was an organic axial filament, primarily carbohydrate. He did not obtain enough material for direct chemical analysis. \Ve present three kinds of data on the structure and composition of the axial filament and the surrounding siliceous spicule. These are direct chemical analysis, phase-contrast light micrographs, and electron micrographs obtained by both direct (rotary shadow; and indirect (negative replica) procedures. Our major con- clusion is that there is sufficient carbon in the siliceous spicule to provide a major fraction of the visible axial filament as a protein or carbohydrate. Nevertheless, organic matter is only a small fraction of the HF-resistant residue. This con- clusion is not contrary to that of Butschli, but is contrary to that of more recent workers who have cited Butschli as their authority. The analytical methods used here were not available to Butschli and have not been combined in this manner by others. MATERIALS AND METHODS Grouih and maintenance of sponges Clumps of marine sponge were obtained through Pacific Biomarine from the coast of California near Los Angeles. These were held with sea urchins 1 This research was supported in part by a grant to R. E. Shore from the Department of Fundamental Research, Owens-Illinois Technical Center. 125 126 DAXIKI. \\'. SCHWAB AND RICHARD E. SHORE and other marine animals in a recirculating aquarium (Instant Ocean) at 13 in artificial sea water ( InMant Ocean) to which has been added sodium silicate to give 1 mM silicate, Growth of new tissue occurred in this aquarium. The sponge was identified as to species on the basis of morphology and spicule type. Acarnus crithacns has a peculiar acanthocladotyle as a minor macrosclere (DeLaubenfels, 1932). The spicules studied were the major macrosclere, a simple style. Preparation of spicules Spicules were freed from the surrounding sponge tissue and other contaminants in several ways: (a) nitric acid and density gradient centrifugation, (b) nitric acid and water washing, (c) crude enzymatic digestion and water washing. For light microscopy and some of the chemical analysis, spicules were freed from the sponge tissue by digestion in boiling concentrated nitric acid. The acid was washed off by repeated centrifugation through distilled water. The spicules were sepa- rated from organic debris and sand by isopycnic centrifugation on a density gradient of carbon tetrachloride and ethylene dibromide. They were washed in methanol and dried at 20. For electron microscopy spicules were cleaned in concentrated nitric acid at room temperature (4 hours) and washed in distilled water. Finally the styles were separated from the other spicules by differential settling in dis- tilled water and air dried at 60. Client leal analysis Acid-washed, density gradient spicules were freed from tiny flakes (presumably mica) of the same density (1.93 to 1.96) by gentle swirling in a shallow dish. The spicules were dried from methanol in a tared Teflon weigh boat. The spicule cake was flooded with 1 N HF and held at 55 on a water bath to dryness. This was repeated to constant weight of the residue at 55. Portions of the weighed, air-dried residue were subjected to three different analyses: (a) gas chromatography for C. II. and N, (b) emission spectrographv. and (c) atomic absorption. L 'n/li t microscopy Nitric acid-washed spicules were observed by phase-contrast light microscopy in a Wild A120 research microscope. The photographs in this paper were made with a 100 > oil immersion objective on Kodak Panatomic-X 35 mm film. Spicules were placed between a plastic cover glass and a plastic dish and indi- vidual spicules observed from the time the HF reached them. Timed observations thus represent time after contact of the observed spicule with the etching fluid. A spicule and its remnants were observed for as long as 2 hours. Electron, microscopy Drum (1968) developed a procedure using rotary shadow t<> make a replica from which the spicule was then removed by HF digestion. This three-dimen- sional replica allowed him to study the surface well, but the internal filament was SPONGE SPICULE FINE STRUCTURE 127 often displaced or obscured. The technique was modified in our laboratory to facilitate study of the filament. The spicule was partially removed by a brief etch in HF, exposing a portion of the axial filament. After rotary shadowing, the remaining spicule was removed by a second digestion in HF. The axial filament was held attached to the replica. Cleaned spicules were broken with the edge of a small spatula and sprinkled loosely onto a collodion-coated 200 mesh electron TABLE I Chemical analysis of spicules cleam-d in boiling HNO 3 Component Weight as % of spicule Weight as % of residue Atomic ratio to Si in residue (X100) Part A: By loss of weight in 1 N HF at 55* SiO 2 -r- H,O Residue 95.7 4.3 100.0 Part B: By atomic absorption, expressed as the element determined Si 17.5 \a 0.65 13.7 K 0.124 2.57 Al 0.154 3.07 Ca 0.014 0.29 Fe 0.007 0.14 100.0 94.0 10.5 18.2 1.16 0.4 I 'art C : By combustion and gas chromatography of the oxide, expressed as the element determined \ 0.073-0.57 1.46-11.5 H 0.022-0.08 0.43-1.6 C 0.016-0.042 0.32-0.83 Part D : By sem-quantitative emission spectrography, expressed as the oxide of the element named B Ba MR. Xi. Ti.Xr (each) Pb Sr 0.6-6.0 0.03-0.3 0.02-0.2 0.01-0.1 0.006-0.06 microscope grid. One drop of aqueous 5% HF was placed on the gird. After 2-4 to 3 minutes the drop was removed by touching the edge of the grid with filter paper. In a similar manner the grid was washed in three changes of dis- tilled water and dried at room temperature. A carbon film was evaporated in vacua onto the specimen during two to three complete revolutions of the sample. The remaining spicule was removed by floating the grid, spicule side down, on a drop of HF for 15 minutes. The grid was again washed three times in distilled water in the same manner. 128 DANIEL W. SCHWAB AND RICHARD E. SHORE FIGURES 1-7 FIGURES 1-4. Phase contrast photomicrographs of spicules and axial filaments in HF during digestion, all same magnification, scale line 20 /j.. FIGURE 1. Intact spicule at first contact with HF showing axial filament (arrow). FIGURE 2. Digestion in 1 N HF alone leaves the axial filament (arrow) protruding from partially digested siliceous cylinder. SPONGE SPICULE FINE STRUCTURE \ 1 ( > A technique was also devised for studying the ultrastructure of spicules in cross section. In general terms, lightly etched spicule sections were replicated with cellulose acetate. The replicas were removed and shadowed, and this nega- tive image examined. Cleaned spicules were embeddd in Epon 812 and poly- merized at 60 for 48 hours. The blocks were cut with a diamond saw so as to expose cross sections of a maximum number of spicules. The sawed surface was polished with Barnesite (Edmund Scientific Company) polishing compound and cleaned in an ultrasonic cleaner with distilled water. A negative replica of the polished surface was made by evaporating a drop of 8% cellulose acetate in acetone on the surface. After about 46 hours, the dry plastic negative replica was stripped off. The replica was shadowed with Pt-C (tan" 1 ^) and carbon coated. Spicule-containing areas were cut out, placed on 200 mesh electron microscope grids, and the plastic dissolved with acetone in a vapor condensation washer. The embedded spicules were then etched for 30 seconds in 5% HF and replicated again as before. Multiple replicas were made of the same surface. Specimens were studied in a Siemens Elmiskop I at 80 KV accelerating voltage. OBSERVATIONS Chemical analysis The results of the three types of chemical analyses are all summarized in Table I. The first column shows the composition for the whole spicules, the second for the non-volatile residue after 1 N HF digestion. The table compares the amounts of the various major elemental components on the basis of weight as a percentage of total spicule (column 1), weight as a percentage of the non- volatile residue after HF (column 2), and atoms per atom of silicon in the residue (column 3). The spicules would appear to be primarily silicic acid (or silica, since the H 2 O content was not directly determined). The residue may contain alkali silicates, but to have all the sodium and potassium present as silicate would require more silicon than is found. Were there very little alkali silicates in the residue there would be sufficient silicon for all the other cations to be present as metal silicates. The proportion C:H:N in the residue is between 1:50:6 and 1:100:260 as calculated from the range of values from the two portions of the inhomogeneous residue. In both cases the total carbon is in the range of \% of the residue; therefore, less than 2% of the residue could be organic matter of the 40% C composition expected of proteins and carbohydrates. Since the N:H ratio is so high it seems unlikely that the N is present as an amine, but more probably as a nitrate. FIGURE 3. Digestion in 1 N HF in the presence of citric acid removes the axial filament. The axial filament may be preferentially removed deep into the siliceous cylinder : arrow marks retreating tip of axial filament. Compare to Figure 1. FIGURE 4. After completed digestion in HF, axial filaments remain (3 shown). FIGURE 5. Electron micrograph of a spicule after partial removal in HF showing an axial filament protruding from the end, scale line = 10 /*. FIGURE 6. Same spicule as in Figure 5 after rotary replication and complete removal in HF. The axial filament can be seen extending from the tip of the conical etch pit, scale line 10 fj.. FIGURE 7. Spicule fragment partially removed in HF, rotary shadowed, and then com- pletely removed in HF. The etch pit is only partially replicated but the axial filament is evident, scale line = 10 n. 130 DANIEL W. SCHWAB AND RICHARD E. SHORE Light microscopy Measurements with an eyepiece micrometer agree with the published dimen- sions for the styles of Acarnus erithacus, namely 20/x X 340ju. An axial structure can be seen inside intact spicules at high magnification with phase optics. Figure 1 is a photomicrograph showing such a structure. It does not appear to he smooth but angular in profile. It does appear to be cylindri- cally symmetrical. During digestion in 1 N or 2 N HF, the siliceous outer portion of the spicule is removed and the axial filament remains (Figs. 2 and 4). The axial filament may be observed to kink, wave in the fluid, or lie on the substratum. Prolonged exposure to 2 N HF does not visibly alter the filament. By contrast to the persistence of the filament in HF alone, when citric acid is added to the HF (from 2 to 4.8 M citric) both the outer and axial portions of the spicule vanish within 15 minutes. The axial filament may even be de- stroyed prior to the siliceous cylinder. This is illustrated by a comparison of Figure 2 (HF alone) and Figure 3 (HF and citric). The siliceous outer portion of the spicules is at about the same stage of digestion. Electron microscopy of u'holc spicules With the rotary shadow technique it was possible to study the axial filament from a spicnle that had been partially removed, and then to remove the HF- sensitive material completely and study the same region again by examining the carbon replica made during the rotary shadowing. Figure 5 shows a partially removed spicule with a filament extending from the end. The eccentric position of the filament will be understood after comparison with Figure 6, which is a replica of the same spicule seen in Figure 5 after the spicule was completely re- moved. The internal relationship of the filament and spicule is revealed. The removal of siliceous material by HF occurs most rapidly in the interior of the spicule, producing a conical etch pit. The filament can be seen to lie along the face of this cone extending outside the spicule from the edge of the cone's base and extending inside the spicule from the cone's apex. The carbon of the replica apparently is attached to the filament holding it in position as it was prior to removal of the second portion of spicule. For deep etch pits at unfavor- able orientations to the carbon during replica formation, or for thin carbon coats, the filament may not be attached to the replica on the surface of the spicule. As Figure 7 shows, the interior portion of the filament is displaced slightly in such circumstances. Electron microscopy of spicule cross sections These figures are derived by shadowing a negative replica of the surface of the spicules so that protrusions on the spicule become pits on the replica, and pits on the spicule are protrusions on the replica. The conical etch pit already seen in the whole mount (positive shadow replica) is here a conical protrusion with a long shadow free of platinum. The etching time for the embedded spicules was necessarily shorter than for the partial removal of whole spicules. The result is much smaller etch pits and liner detail than was seen with the whole-spicule preparations. SPOXCil-: SI'ICl'LK K1XK STKUCTUkH 131 Figure 8 is a cross section of a style, the major macrosclere of Acarnus erithacus that was polished but not etched. Similar preparations after light etching with HF are shown in Figures 9 and 10. Occasionally, the axial filament FIG. 8-9 FIGURE 8. Electron micrograph of a spicule cross section after polishing but no etch ; C/Pt shadowed negative replica, scale line = 10 /*. FIGURE 9. Spicule cross section after polishing and a 30-second etch in 5% HF. The center etches very rapidly and forms a conical etch pit. Closely spaced concentric rings are also evident; C 7 Pt shadowed negative replica, scale line = 10 /u. 132 DANIEL W. SCHWAB AND RICHARD E. SHORE may be seen as a detail in the etch-pit cone (Fig. 10). Oblique sections through spicules provided supplementary evidence of the axial position of this filament, as shown in Figure 1 1 . The shape of the axial filament is not clear from these observations. It may be either triangular or circular in section. The filament appears to be about 0.4-0.6 ^ in diameter. The negative replicas exhibit circular surface protrusions (Figs. 9 and 10), indicating the existence of indentations produced by etching on the surface of the spicule. The more pronounced, that is, deeper, rings may correspond to those FIGURES 10-11 FIGURE 10. Replica uf an axial filament in an etch pit (arrows) ; C/Pt shadowed negative replica, scale line = 1 p.. FIGURE 11. Replica of an etch pit in oblique section showing the axial filament (arrows) ; C/Pt shadowed negative replica, scale line = 1 /J.. described by Butschli (1901) from light microscopy of stained fractured spicules. Several shallower grooves occur between and concentric with the deeper clefts. The rings are usually more pronounced closer to the center of the spicule. This is correlated with the more rapid removal of material in this region during etching that is expressed as the conical etch pit already mentioned above. The rings are somewhat irregularly spaced. The thinnest ones are about 0.2-0.3 ^ thick. DISCUSSION The pattern of chemical composition revealed by our analysis is in general agreement with the literature. The bulk of the siliceous spicule is cation-free SPONGE SPICULE FINE STRUCTURE 133 oxide of silicon, presumably (SiO 2 ) x H 2 O. The present data provide no value for the ratio of SiO 2 to water. Values have been reported in the literature ranging from 3 to 5. (Minchin reviews the literature before 1909; see also J0 r gensen, 1944). There is little agreement in the literature on the cation composition of siliceous spicules. presumably reflecting real differences among the forms examined (Minchin, 1909). To compare two marine Demospongiae, the present values are in agreement with Biitschli's for sodium and the traces of aluminum, though he found much more magnesium. From the composition and the density there is no reason to question the use of the term "spicopal" to refer to the siliceous portion of the spicule (Vosmaer and Wijsman. 1905), that is, to presume an amorphous silicic acid with traces of various cations with properties similar to the mineral opal. As to the nature of the axial filament, there can be little question now that the siliceous macroscleres have a component part that is both axial and HF- resistant. The filament can be directly observed before and during etching. The high magnification light microscopy leads us to conclude that the filament visible by phase microscopy is not derived from the outer surface of the spicule by collapse of a net or sheath. Nor can we see how it could be derived by the coalescence of isolated particles within the siliceous matrix that come together during dis- solution of the spicule. Organic matter on the native spicule surface would pre- sumably have been destroyed by the 4-hour digestion in boiling concentrated nitric acid. The direct observation during solution of the spicule in HF reveals an object protruding from the end of the retreating spicule that has the same dimensions as a line found within the spicule prior to etch. What portion of the axial filament may be protein or other organic material? The destruction of axial filament in mixed citric-hydrofluoric acid indicates only that a chelatable cation contributes to the stability of the filament. Other similar organic substances have been characterized ; in particular a presumably unrelated factor involved in cell aggregation has been purified from sponge (Moscona, 1968). Its insolubility depends on the presence of Ca ++ or other divalent cations. The most direct evidence of the possibility of a primarily organic filament comes from the CHN analysis. The carbon present (between 0.32 and 0.83% of the residue) would be sufficient to account for at least 0.8% of the residue as organic matter with 40% carbon, the range found in carbohydrates and proteins. As a fraction of the total spicule weight, this would be at least 4.0 X 10"*. The filament visible in electron micrographs appears to be about 0.5 /t in diameter and the spicules are about 20 ^ in diameter. The ratio of filament cross section to spicule cross section, that is, the ratio of radii squared, has a value of 6.3 X 10"*. Mass density of the organic material would be expected to be around 1.27 (gelatin) to 1.53 (starch) and of the spicopal about 1.96 (whole spicules) to 2.3 (opal). Thus a given mass of organic material would fill about 1.5 times the volume of the same mass of opal. "When the carbon found (expressed as the organic fraction of the total spicule), 4.0 X 10~ 4 . is multiplied by this mass density correlation factor, 1.5, the product is 6.0 X 10'*. That is, organic matter makes up by weight 0.4 parts per thousand of the whole spicule, and that much is expected to occupy 0.6 parts per thousand 134 DANIEL W. SCHWAB AND RICHARD E. SHORE by volume. We observe a structure that occupies about 0.63 parts per thousand by volume. We are led to conclude that there is sufficient carbon detected to provide a major fraction of the visible axial filament as organic matter with the carbon content of protein or carbohydrate. It is important to emphasize that this must be a minimum estimate since the harsh cleaning procedure, boiling concen- trated nitric acid, while removing the organic material from the native surface of the spicule might penetrate into the open tip of some spicules and remove some material. The data of Travis, Kranc^ois, Bonar, and Glimcher (1967) deserve comment at this point. As part of a broad comparative study of the organic material associated with the mineralized skeletons in animals, they examined the amino acid composition of an HF residue of siliceous spicules. They cleaned the spicules with distilled water only. Such a treatment is not likely to remove the collagen fibers that bind spicules in place, to say nothing of less specific cellular debris. Therefore, their data apply to a mixture of surface and axial material. It may well be that the surface material is more important in controlling the actual deposition of the siliceous body of the spicule. A second recent paper of importance is by Drum (1968). His conclusions on organic material rest primarily on incineration data. Unfortunately, organic material is not the only material volatilized by such a procedure. All three microscopy techniques used in this study indicate the existence of an HF-resistant filament that is axial in position. From the chemical analysis there would seem to be sufficient organic matter to account for the filament as carbo- hydrate or protein. Organic axial filaments have been demonstrated in calcareous spicules ( Minchin and Reid, 1908). Jones (1967) refutes much of that work, contending that the filaments are preparation artifacts and at best are impure calcite in the intact spicule. The present work does not appear to be subject to that kind of criticism. Concentric rings are revealed by gentle etching of spicule cross sections. From fracture studies the earlier workers, reviewed by Minchin (1909), had concluded that the spicule was composed of concentric lamellae. Schulze ( 1904) observed a similar ring pattern in Hexactinellids which he interpreted as lamellae of organic matter alternating with siliceous material. According to Biitschli, the concentric strata were layers of different retractility. The lowered refractility. he surmised, arose from a minutely alveolar structure. Our figures show that some of the rings are clefts and some are dikes, thus differing in etch sensitivity. It is not clear how such a pattern of concentric but irregularly spaced zones of etch-resistant and etch-sensitive material could be related to Schulze's pattern of regular rings, presumably of organic material. Biitschli's notion of differences in compactness appears to be clearly related to the kind of pattern we see. However, if microalveoli are present, they must be below the resolution of our replica technique. We must further emphasize the discrepancy in scale of the parts of the 20-/J, diameter spicule as reported here for Acarnus and as reported by Biitschli for Tethya and Geodea. His axial filament is about 2.4 /t in diameter, that is. a fraction 0.12 of spicule diameter. His concentric rings are about 1.8 p. thick, that is. 1 1 rings for the entire 20-//, diameter spicule. Our filament diameter is SPONGE SPICULE FINE STRUCTURE 135 about one-fourth of his, both in absolute terms and as a fraction of spicule diameter. Our rings are similarly smaller than his by a factor of 0.14. Rings of this small size are at the limit of resolution of the light microscope. It is easily understandable that Btitschli would not have detected these as individual rings. We cannot rule out the attractive possibility that subtle differences in composi- tion or physical structure of the spicopal give rise to both the differences in etch sensitivity that we report and the negative birefringence that others have re- ported (Minchin. 1909). Investigations are underway in this laboratory to determine the spatial distribu- tion of chemical elements within the spicule. \Ye gratefully acknowledge the expert assistance of several colleagues for pro- viding chemical analyses; W. H. Grieve for gas chromatography, R. H. Hall for emission spectrography and T. 1. Gomoll for atomic absorption. SUMMARY Siliceous spicules from Acarnus erithacus were studied using electron and light microscopy in conjunction with chemical analyses by gas chromatography. emission spectrography, and atomic absorption. Chemical data correlated with microscopy measurements indicate that there is sufficient carbon to provide a major fraction of the axial filaments as organic matter (assuming 40% carbon). The concentric ring structure reported earlier from light microscopy was studied on the electron microscopic level by utilizing carbon-platinum replicas of HF etched spicule cross sections. Ring spacings as small as 0.2-0.3 //, were detected using this technique. Correlation of chemical composition and fine structure is discussed with respect to the axial thread and the siliceous portions of the spicule. LITERATURE CITED Bi'TSCHLi, O., l c '(Jl. Einige Beobachtungen iiber Kiesel- und Kalknaddn von Spongien. Z. Wiss. Zool, 69: 235-280. DELAUBENFELS, M. W., l c >32. The marine and fresh-water sponges of California. Proc. U. S. Nat. Mas., 81: 1-140. DRUM, R. W., 1968. Electron microscopy of siliceous spicules from the freshwater sponge Hctcromycnia. J. I'ltrastntct. Res., 22: 12-21. GARRONE, R., 1969. Collagene, spongine et squelette mineral chez 1'eponge Ilaliclona rosea (O.S.) (Demosponge, Haploscleride). /. Microscop. 8: 581-598. JONES, W. C., 1967. Sheath and axial filament of calcareous sponge spicules. Nature, 214: 365-368. J0RGENSEN, C. B., 1 ( 44. On the spicule-formation of Spongilla lacnstris (L.) 1. The depend- ence of the spicule-formation on the content of dissolved and solid silicic acid of the milieu. Dei. Kgl. Danske 1'idenskabcrnes Sclskab Bioloyiske Mcddclclscr, 19(7) : 1-45. LEVI, C., 1963. Scleroblastes et spiculogenese chez une eponge siliceuse. C. R. Acad. Sci. Paris, 256 : 497-498. MFKCHIN, E. A.. 1909. Sponge-spicules. A summary of present knowledge. Ergebnisse und Fortschritte dcr Zooloqic, 2: 171-274. 136 DAXIEL W. SCHWAB AND RICHARD E. SHORE MINCHIN, E. A., AND D. J. REID, 1908. Observations on the minute structure of the spicules of calcareous sponges. Proc. Zool. Soc. London, 2 : 661-676. MOSCONA, A. A., 1968. Cell aggregation : Properties of specific cell-ligands and their role in the formation of multicellular systems. Develop. Biol., 18 : 250-277. SCHULZE, F. E., 1904. Hexactinellida. [Wissenschajtlichc Ergcbnissc der Deutschen Tiefsee- Expedition mij dcm Dampfer "Valdivia" 1898-1809, Vol. 4] G. Fischer, Jena, 266 pp. TRAVIS, D. F., C. J. FRANCOIS, L. C. BONAR AND M. J. GLIMCHER, 1967. Comparative studies of the organic matrices of invertebrate mineralized tissues. /. Ultrastntct. Res., 18: 519-550. VOSMAER, G. C. J., AND H. P. WijSMAN, 1905. On the structure of some siliceous spicules of sponges. I. The styli of Tcthya lyncurium. Acad. II' ct. Amsterdam, Proc., 8: 15-28. Reference: Kwl. Bull, 140: 137-155. (February, 1<>71 PHOTOPERIOD CONTROL OF DIAPAUSE IN DAPHNIA. IV. LIGHT AND OX-SENSITIVE PHASES WITHIN THE CYCLE OF ACTIVATION R. G. STROSS Department of Biological Sciences, State University of AY' York at Albany, Albany, New York 12203 The continual involvement of photoperiod in the life cycle of arthropods is demonstrated by the influence of photoperiod on both the initiation and termina- tion of diapause ( Adkisson, 1965). Photoperiod control of diapause in insects has been demonstrated for larval (Paris and Jenner, 1959; Wellso and Adkisson, 1964), pupal (Williams and Adkisson, 1964) and imaginal stages (deWilde, Duintjer and Mook, 1959). Embryonic diapause is also known to be initiated in response to photoperiodic induction. This has been demonstrated for a number of insects and for the Crustacean, Daphnia pitlc.r (Stress and Hill, 1968). In previous study with Daphnia the termination of diapause may require a light stimulus (Stress, 1966, 1969), thereby demonstrating the presence of a photoreceptor which may be absent in early embryos of some insects (Minis and Pittendrigh, 1968). Furthermore, the light stimulus may need to be "long-day" although many embryos are activated in "short-days" as well. Since photoperiod control of diapause initiation may be controlled by density of the experimental animals, i.e., facultative, the suggestion is that an unknown stimulus present in the water may interfere with the retention of diapause in short daylengths. The present study was undertaken to demonstrate photoperiod control of diapause termination in Daphnia. A second objective of the study was to test the applicability of two empirical models of photoperiodism, a test made possible by the remarkable feature of the diapaused embryo of Daplmia to respond to a single light period. Dunning (1959, 1964) described a physiological rhythm of maximum light sensitivity, the timing of which he suggested may be more influenced by the "dawn" than by the "sunset" of a natural or simulated day. A restatement of the model was made possible by a more explicit knowledge of the action of light on a circadian rhythm in an insect. The restated (coincidence) model (Pittendrigh and Minis, 1964) argues that the so-called time or phase of maximum sensitivity to light (photo-inducible phase) is coupled to the circadian cycle of the organism. Thus the light period of a daily cycle phases the circadian cycle, and when the light period is "long-day," it illuminates the organism at a time when it is photo-inducible. The dual effect of light is operationally significant when the experimental organism requires more than one inductive cycle. The coincidence model is satisfied when the long-day response is induced by a single pulse of light per cycle administered to an organism at the photo-inducible phase of its circadian cycle. The position of the inducible phase was first indicated to be near sunset (Pittendrigh and Minis, 1964) 137 138 R. G. STROSS and later generalized to include the possibility that it could be at dawn (Pitten- driidi. 1966). An alternative model of photoperiodism (Lees, 1966) suggests that time measurement is restricted to the dark period, and a long-day condition results (virginoparae) when the night is critically short (Lees, 1966). In an organism requiring but a single period of light for induction, two pulses of light suitably spaced with respect to one another but independent of the organism's circadian cycle would suffice to induce the long-day condition. This so-called "interval- timer" has been tested with consistent results in the aphid Megoura. A recent study (Hamner, 1969) with a moth (Carpocapsa) strongly indicates the presence of both a circadian and an interval-timer in the photoperiodic response. Previous study has shown that the diapaused embryo of Daphnia contains a minimum of two phases, a photo-refractory phase followed by a photo-sensitive phase (Stress, 1965). The photo-refractory phase in embryos collected from the wild is completed with exposure to low temperature ; 4 C was found adequate for embryos collected at temperate latitudes (Stress, 1966, 1969) but a lower temperature was necessary for embryos of D. middendorffiana collected from the arctic (unpublished). The duration of the photo-refractory phase may be dif- ferent in summer and winter diapausing strains, and much shorter in the former (Stress, 1969). The alternative condition of having the same or similar thermal optimum for both the active and diapause states has also been shown to exist in Supply House cultures of the species (Stross, 1966; Davison. 1969). The photo-sensitive phase in Daphnia pule.v may or may not require light for terminating the embryonic diapause. In a winter diapausing strain, a light require- ment is restricted to two situations : pre-mature termination in autumn before the diapause state intensifies, and in the spring when the embryos are incubated in a crowded and presumably oxygen deficient environment (Stross and Hill, 1968). In other strains including the ones lacking a low temperature optimum, light re- mains an absolute requirement, that is the diapause is maintained indefinitely in constant darkness. The suggestion that the process underlying activation may be basically photoperiodic has been apparent in both light-escaping and light- requiring strains (Stross and Hill, 1968; Stross, 1969). METHODS AND MATERIALS The egg pods (ephippia) containing usually a pair of diapaused embryos were removed from the bottom of the culture vessel and transferred to constant dark. They were mass incubated while either lying at the bottom of beakers (stagnant environment) covered with saran or suspended in nylon net within a flowing stream (Fig. 1). The embryos were mass produced in cultures of a strain of DapJnia pulex Leydig that came originally (1960) from a Biological Supply House. The strain is capable of completing the photo-refractory phase at 20 C (Stross, 1966), the incubation temperature in these experiments, although the environment may require other modification (see Results). The strain could be classed as the dicyclic type since at room temperature (21) the females readily reproduce exclusively the diapausing embryos when densely cultured in long day- lengths. The cultures were maintained in room light supplemented with fluor- escent lighting on an L18:D6 regimen. CO 2 AND PHOTOPERIODISM 139 At the time of light exposure embryos \vere transferred under safelight to individual vials (25 X 95 mm) containing 20 ml of medium which was either lake water or a synthetic substitute (Stress and Hill, 1968). Ten or 25 egg pods containing a determined number of embryos were placed into each vial. The vials were then covered with saran or, as in the gasing experiments, stoppered with a rubber stopper lined with saran. Vials were prepared by autoclaving in a strong bicarbonate solution followed by rinsing in distilled water ; this treatment was found to reduce the variance which can be a serious problem in hatching experiments. The embryos were exposed to "cool-white" fluorescent light and at intensities ranging from 1000 to 2000 lux, depending on the experiment, for intervals as indicated. FIGURE 1. A "flowing stream" for incubating diapaused embryos of Daplmia in constant darkness. The embryos are held in small cylinders of nylon screen which is inserted into the ball joints between the manifolds. In the experiments with modified atmosphere, the gas was introduced from pre- pared compressed sources (Linde-Union Carbide) with a sintered glass diffuser. The diffuser was introduced directly into the vial and the gas bubbled for 1.0 minute. Enriched CO 2 mixtures consisted of 10.0 and 49.0 per cent CCX, a 20 per cent (X concentration, and the balance N 2 . A 5.0 per cent CO, atmosphere was also used in which the CCX had been added to air (i.e., displacing both O 2 and N 2 ). Enrichment with CO 2 was carried out at the time of light exposure. Preliminary treatment with N 2 was carried out in the same manner, except that masses of embryos were treated in screw-top jars which were sealed following gasing. This procedure may not have removed all CX from the medium, although in the CCX enrichment the treatment was sufficient to bring the medium to a new pH equilibrium. 140 K. G. STROSS RESULTS Activation of the cliapaused embryo of Daphnia pulex Leydig may require light received as one long day or its "skeleton" (Fig. 2). In two experiments, light exposure was necessary at two separate intervals within a single light period. Embryos which had been incubated in constant dark and a natural thermocycle were first exposed to light at 0900 EST either for 2 or 16 hours. Those exposed for only two hours required a second exposure and the timing EXPOSURE (hours) HATCH 8 12 16 20 24 o: 0.0 0.0 0.0 72.0 63.6 g LLJ o.o 8 34.8 jg 32.4 i 54.0 55.2 19.4 FIGURE 2. Hatching of Daphnia embryos when given a single 16-hour light period or two pulses of light that form the skeleton of one 16-hour light period. The embryos were first exposed at 0900 EST after having been in constant darkness and near an open (September) window for 17 or 31 days. Prior to that time the embryos were in constant darkness and temperature (19 C) for four months. (See text for additional details.) of the second exposure was critical. When exposed to a second pulse from hours 14 to 16 following the start of the first pulse, activation (63.6 per cent), as measured by hatching, was equivalent to a continuous 16-hour light exposure (72.0 per cent). A second exposure from hours 10 to 12 was not effective (0.0 per cent hatch) as was a single exposure given from hours 14 to 16. Exposure to one 14-hour skeleton resulted in the hatching of i (34.8) the number activated by one 16-hour skeleton. This result suggested that the critical photoperiod for the embryos was approximately 14 hours of light. Further de- ductive support for 14 hours as the critical daylength was given by the number CO= AND PHOTOPERIODISM 141 of embryos activated with a light pulse of shorter duration than the standard two hours. A one-hour pulse from hours 13 to 14 activated as many embryos as the standard pulse given from hours 12 to 14, whereas a one-hour pulse given after hour 14 resulted in the activation of a larger number of embryos. Further reduction in the duration of the light pulse, as for example to one-half hour, reduced activation (Fig. 2) and verified results of preliminary experiments. It is to be emphasized that these results were obtained from exposure of the embryos to a single photoperiod or its skeleton. The requirement for two exposures appropriately spaced suggested that two photo-inducible phases exist and exposure to both may be a necessary part of long-day induction in Daplnlia embryos. Lees (1966) has also shown that two exposures per 24 hours were necessary to simulate long daylength in the aphid, Megoura, or more precisely, to keep the night interval critically short. The apparent similarity between the two evoked such questions as whether the day or night interval was being measured and whether the position of at least one of the photo-inducible phases recurred at circadian intervals. Further experimenta- tion proved impossible, however. The long day or two-pulse requirement proved transitory for embryos held in the dark in a stagnant environment. Subsequent tests with the same batch of embryos, but from different containers, revealed first the loss of the requirement for a second pulse followed by a de-synchronization of photo-sensitivity. In the third experiment, designed to locate the time of photo-sensitivity relative to the thermal dawn of the environment, the first two-hour pulse of light activated as many embryos as the skeleton of a 12 or 16-hour photoperiod. In this new state the embryos were not only responsive to a single pulse but showed a time dependent sensitivity to photo-activation. Three groups of embryos were exposed at the onset of thermal rise (0600) in an artificial thermoperiod, or before (0400) or after (1000) the thermal rise. Activation, as measured by hatching, was largest in the embryos exposed to light at 1000 and 83.0 per cent of the embryos hatched (Fig. 3A). Exposed six hours earlier, only 36.0 per cent hatched (P =0.01). This new condition followed only two weeks after the experiments that clearly showed a need for two pulses, and nine days after the embryos were placed in an artificial thermoperiod. In an effort to locate the "time" of maximum sensitivity to photo-activation, each of 12 groups were exposed to a different two-hour period within the 24-hour thermocycle. After only four additional thermocycles the success of activation had declined from a maximum of 83.0 per cent to approximately one half that number (Fig. 3A). Two groups of embryos showed maximum sensitivity, one exposed at 0800, a time which corresponded with the thermal rise, and a second exposed eight hours later at 1600. Oddly, the total activation of embryos for the two periods of maximum sensitivity was equal to the number of embryos that had been activated at 1000 in the preceding experiment performed four days earlier. Several interpretations are possible. The simplest seems to be that the embryos have retained only one phase of photosensitivity and that the rhythm of light sensitivity is responding to the thermocycle. One group of embryos seems to have retained a phase relation to the thermal rise, while a second group began 142 R. G. STROSS o m o < LU AGE OF EMBRYOS 6 1/4 MONTHS THERMOCYCLES 36 + 9 AGE OF EMBRYOS 6 1/2 MONTHS THERMOCYCLES 36+13 04 06 08 10 12 14 16 18 20 22 24 02 04 46 8 10 12 14 16 18 20 22 24 2 CLOCK HOUR | ,5.0 !_/ \- i i i i \ 04 06 08 10 12 14 16 18 20 22 24 02 04 HOUR OF DAY FIGURE 3. Hatching of embryos when given a single two-hour light pulse at various times in an artificial 24-hour thermal cycle. The embryos are from the same batch as shown in Figure 2 but are two weeks or more older. * S.E. is zero. a synchronous drift to a later time, and at the time of measurement occupied no known relationship to the thermal cycle. The positions of the photo-sensitive phase were also unstable. After an additional month of incubation in the dark, the embryos were activated when exposed at nearly all phases of the thermocycle. Disclosure of two distinct phases of light sensitivity will form the basis of later interpretation although the condition was unstable. Instability amounted to CO, AND PHOTOPERIODISM 143 the loss of one of two inductive phases requiring light for activation. This was accompanied by the development of an asynchronous and seemingly continual sensi- tivity to light. The presence of two photo-inductive phases in each thermocycle was also unstable in a second series of experiments with a new batch of diapaused embryos. In the second series the optimum interval between the two light exposures was clearly a function of when the embryos were first exposed to light pulses of two- hours duration. When exposed at the thermal rise, a 10-hour skeleton photo- period was more effective than a 16-hour skeleton photoperiod. The reverse was true when the embryos were first exposed 12 hours after the thermal rise. These results are readily interpretable if the embryos become entrained to the thermal oscillations in the environment but do not distinguish the thermal rise (dawn) from the decline (sunset). It would seem that the first pulse of light was exposing the embryos at their "sunset" and the second pulse at "dawn." A delay in the first exposure to light would then result in an initial exposure of the embryos at dawn. In other words the interpretation suggests that the embryos may respond when "seeing" light first at either sunset or at dawn. Substitution of potential synchronizers other than thermal oscillation was only partially successful. The simple transfer to fresh medium may have been effec- tive. When exposed to a 2-hour pulse at the time of transfer, few (3.3 per cent) embryos were activated. When exposed 24 hours later, 70.0 per cent were activated. Exposure at both times (hours to 2 and 24-26) was ineffective (9.3 per cent). Attempts to synchronize with light were disastrous. Single exposures of 15 minutes to embryos still in the original medium damaged the embryos. A small number (15.6 per cent) hatched but most of the affected (activated?) embryos disintegrated. Hourly exposures to the safelight (red light) at 24-hour intervals were only slightly successful. The foregoing results demonstrated that light may be required for activation. They clearly show that the Daphnia embryo may be discontinuously sensitive to light. Specific times of light sensitivity appear to bear a specific relationship to a natural or imposed thermocycle. Two conditions of photo-sensitivity were re- vealed. In one condition following the onset of the photo-sensitive phase of dia- pause, the embryos may require exposure to light at two distinct times within one thermocycle. One of these times may be interpreted as the dawn phase of the embryo's putative daily cycle, the second as the sunset phase. In a second condi- tion the embryo requires only a single exposure to light. Whether the single exposure serves as dawn or sunset cannot be interpreted, as a result of a seemingly ambivalent phase relationship of the embryo's endogenous cycle to the thermocycle of the environment. Photo-sensitising with elevated tensions of CO 2 Daphnia embryos held in darkness in a flowing stream (Fig. 1) for up to 14 months retain their viability but are photo-refractory. They may be made photo-sensitive with manipulation of the gaseous environment. Following incuba- tion in a medium bubbled with nitrogen, photo-activation is possible but only when the tension of CO., is elevated by bubbling the medium with air enriched with 5.0 per cent CO 2 ; detailed results are given below. The gas modification is a well 144 R. G. STROSS tested procedure for breaking the dormancy of many organisms including certain parasites (Fairbairn, 1961; Rogers and Sommerville, 1968) and seeds (Ballard, 1958, 1967; Ballard and Grant Lipp, 1969). Although some essential details are still lacking, the procedure may guarantee a uniform response from embryos incu- bated for six months to 14 months in the dark before exposure to light. The f unction (s) of C0 2 Elevated CCX tensions may perform two functions in the parasite (Sommerville, 1964). One is to break the dormancy and the second is to initiate a development leading to the molt of the diapausing instar. Sommerville (1964, 1966) showed TRMT HOUR 4 8 HATCH 16.0 0.0 70.0 64.0 27.0 13.0 a: CD o o GO ^ => FIGURE 4. Photo-activation in the presence of an elevated CO 2 tension. The medium has been enriched with a mixture of 5.0 per cent CO a in air. Embryos were in constant dark and temperature for ten months, six of which were in the "flowing stream" and four months prior to treatment in a medium treated initially with 100 per cent N 2 . Note some activation independent of CO 2 enrichment. the second function could be largely completed within the first 24 hours of incuba- tion. A fungal parasite of cucumbers needs both a light and a dark-requiring process in order to form spores (Barnett and Lilly, 1950. 1955). The dark-re- quiring process is suppressed by what has been identified as an elevated tension of CCX in the culture vessel. The two functions of CCX may not be separable in the activation of Daphnia embryos. It is clear that both light and elevated CCX tensions participate in the activation process. Moreover, the action (s) of CCX is effected in the dark. The exposure of embryos to a four-hour pulse of light either with or before elevation of CO 2 resulted in a high level of activation (Fig. 4). Conversely, pulsing the em- CO 2 AND PHOTOPERIODISM 145 bryos with CO 2 for four hours either with or before the light pulse had little or no effect. In the experiment the level of hatching in the CO 2 control was significantly greater than zero (16.0 per cent). This background level was elimi- nated in subsequent experiments by restricting the duration of incubation in nitrogen. Whatever the function of CO, in the dark, the presence of CO 2 restricts the light requirement to a maximum of four hours. A working hypothesis of activation is that the embryos may require stimula- tion at two separate inductive phases of a daily cycle as shown in earlier experi- ments (see above). One of the phases has an absolute requirement for light. The second phase might be inducible with light or a substitute such as an elevated tension of CO 2 . Results of the first experiment with CO 2 showed that CO 2 renders the embryo photo-sensitive. That CO 2 later substituted for a second pulse of light is an assumption consistent with the dual role of CO 2 as shown by Sommer- ville (1964) and consistent with the observation that CO 2 suppresses a dark (short-day) reaction (Barnett and Lilly, 1955). Thus it may be proposed that CO 2 may have two roles in the activation of the Daphnia embryo : to break photo- refractoriness and to substitute for light at one of two photo-inducible phases. Results supported the hypothesis that CO 2 substitutes for light at one of two inductive phases of a daily cycle, although in a way more striking than anticipated. Ten groups of embryos in triplicate were given an elevated level of CO 2 under safelight. The first group was exposed immediately to a two-hour pulse of light and the other nine were similarly exposed but at a later time in the following 32-hour interval (Fig. 5). Embryos were activated in all but the control groups and ranged from 44 to 100 per cent with no trend apparent. That is the embryos would appear to be photo-inductible at all times tested. The striking features of hatching were the synchrony and the length of the interval from light exposure to the time of hatching (Fig. 5). The interval assumed two basic patterns. When the embryos were exposed to a light pulse at any of four times in the first 12 hours, they hatched simultaneously and 83 hours after CO 2 elevation, that is they behaved as though CO 2 was the stimulus trigger- ing activation. Controls clearly showed that both light exposure and CO 2 eleva- tion were necessary for activation. In the subsequent 12-hour interval, that is from hours 12 to 24 following CO 2 elevation, the pattern of hatching suggested that light now provided the key stimulus. The situation reversed itself in the third 12-hours and once again the pattern of hatching suggested CO 2 was the key stimulus. The observed pattern of hatching strongly suggests that CO, triggers a dark- reaction. It may be supposed that light given during this phase initiates no immediate effect but that the light stimulus is "stored" until the beginning of a second phase at which time a second necessary process is initiated by light following which the embryo begins an irreversible development. If the embryos are not exposed to light until the second 12-hour period, the process of activa- tion that was initiated with CO 2 apparently stops and waits for the light to initiate the second (light requiring) process. Since photo-activation is here sug- gested to occur at dawn, the second 12-hour period might be equivalent to Running's photophase. Now the hatching response to light exposure in the inter- val from hours 24 to 34 is the exciting thing since it strongly suggests an endo- genous rhythm of recycling of the embryos to a state comparable to when the CO 2 146 R. G. STROSS tension was first elevated. Again CO, seems to activate a dark reaction and the embryo stores the light signal for dawn use. It is to be noted that the overall interval from light exposure to hatching is now shortened by approximately eight hours, almost as if in the first 24 hours some development had taken place in the presence of CO., but in the absence of light. One essential demand of this interpretation is that CO 2 initiates a dark reaction and that initiation is independent of the time at which the embryos are introduced to the CO L , elevation. To test this the same type of experiment as the preceding MEAN TIME HATCH 78 D ' I ' 67 D -A- I 98 D I 89 co 4 L i i i i i i i i i__i A - -I 60 oo I 56 D I 67 9 D I 53 g ' 69 -i A-B D I 44 C0 2 ONLY LIGHT ONLY 4 8 12 16 20 24 28 32 36 ^72 4 8 12 16 20 24 28 FIGURE 5. Mean hatching time of embryos exposed to a two-hour pulse of light at some time in a 34 hour interval following CO 2 elevation of the medium. Previously, embryos had spent three weeks in a Na environment following transfer from the "stream." "A" is the (variable) interval from onset of light to mean hatching time. "B" is the (fixed?) interval by which total development time (A) has been shortened in the second cycle following CO elevation. was repeated with four sets of embryos, with successive sets receiving the CO 2 elevation at hour 0, 6, 12 and 24. All sets were removed from the same container, which came from a stagnant medium, since the source in the flowing stream had been exhausted. With one important exception the embryos responded as predicted. Following gasing the embryos began development which advanced them a maximum of 13 hours toward hatching and there appeared to be no qualitative difference in any of the four groups (Fig. 6). Unlike the results of the preceding experiment, the embryos continued to develop for the first 20 hours following gasing instead of for the first 12 hours. Also unlike the preceding experiment, all but the group CO, AND PHOTOPERIODISM 147 started at hour 12 began a reversal at hour 20 following CO 2 elevation. At hour 32 the groups started at hours and 6 had reversed development to the point where hatching required the same duration as though light and the CO 2 elevation had been given simultaneously. Following hour 32 the two groups again began development toward hatching. Only the group started at hour 12 behaved similarly to the embryos in the preceding experiment in that development executed during the first 20 hours was retained throughout the subsequent 16 hours. \Yith signif- icant quantitative exceptions the experiment is believed to support the hypothesis that CO., triggers one of two essential processes necessary for termination of the diapause. START OF EXPERIMENT HOURS o 0+6 " - CHI2" D CM- 24" HOURS DELAY OF LIGHT EXPOSURE FIGURE 6. Mean hatching time of embryos given a two-hour light pulse simultaneous with COa enrichment or at some time within 34 hours following enrichment. Four sets oi embryos were treated with CCX at hour 0, or at 6, 12, or 24 hours later. Hatching time is expressed in hours from light exposure but relative to the hours required when both light and CO^ treatments were given simultaneously. Embryos previously incubated for one year in a stagnant environment in constant dark. The hypothesis also implies that the signal from the light pulse may be "stored" for initiation of the light requiring reaction following completion of the CO 2 - initiated reaction. Experimental evidence is supportive. When CO., is with- drawn at hour 10 following the elevation, activation is not significantly greater than zero despite a light pulse given with the CO., elevation (Fig. 7). However, when withdrawal is followed immediately by a second light pulse (hour 10 to 12), activation is achieved and the embryos hatch. If four hours are allowed to elapse between withdrawal of CO 2 and exposure (hours 14 to 16) to light, no embryos are activated indicating that a CO., controlled process is reversible. These results strongly support the demand for induction of two temporally separate processes. 148 R. G. STROSS one of which is satisfied by CO, and the second of which requires light. They also show the order in which the two inducible processes must be satisfied when CO 2 is employed to break the photo-refractory phase of diapause. Breaking diapause with modification of the gaseous environment When the embryos are incubated in a "stream," enrichment of CO, concentra- tion becomes a necessity for photo-sensitivity as shown above. CO 2 treatment was not sufficient by itself, however. Although the details are still being investigated, c C0 2 CONTROLS MEAN S.E. HATCH x /IOO EMBRYOS DARK CONTROL EXPERIMENTAL HOUR 10 14 L 43.0 670 7.2 71 9.5 7.1 2 8 10 12 14 16 18 20 22 HOUR MEAN TIME OF HATCHING RELATIVE TO CONTROL (HOURS) 4.7 - 61.0 2.3 -1.0 69.0 24.3 0.0 + 4.5 + 12.0 55.0 6.3 0.0 0.0 - - FIGURE 7. "Storage" of the light stimulus by COa as shown by the need for a second light exposure when CO 3 is withdrawn following ten hours after CO 2 elevation and the initial light exposure. the embryos require some preliminary exposure to a low O 2 environment. Follow- ing six months incubation in the "stream" a collection of embryos was divided into two groups. One group was subjected to a 100% N 2 environment for two weeks, the second to a 50% CO 2 , 20% O 2 environment for the same interval. Both groups of embryos were then exposed to light with or without CO 2 elevation (5% CO, in air). The groups treated with N, gave an 87.0 per cent hatch when exposed to both light and elevated CO.,, the CO 2 control, 0.0 per cent. The groups treated with 50% CO 2 and normal O, gave no hatching in light and 5 per cent CO 2 , although there was a 6.0 per cent hatch in the CO, controls. The low O 2 (No) environment may be considered a prerequisite to the effectiveness of CO 2 elevation in the presence of light. CO 2 AND PHOTOPERIODISM 149 The low O 2 and high CO 2 environment, which occurs normally with aerobic metabolism in a closed system may therefore act independently in breaking the photo-refractory phase of diapause in Daplinia. Presumably this is what happens in a stagnant medium and ultimately results in the disappearance of one of the two inducible phases within the daily cycle of the embryo. Ballarcl and Grant Lipp (1969) have shown that a low O 2 atmosphere achieves partially the effect of an elevated CO 2 atmosphere in breaking the dormancy of clover seeds. A preliminary exposure to low O 2 was not necessary for CO 2 effectiveness in their material, how- ever. ACTIVE DURATION A FUNCTION OF GENOTYPE TEMP. OTHER DURATION FIGURE 8. Schematic model of diapause interval in embryos of Daphnia pulex. Each of four states (A-D) are identified by the changing responsiveness of the embryos to specific stimuli in the environment, including temperature, O 3 and CO 2 tensions, and light. In a fifth state (E) the embryo is photosensitive. The light requirement for terminating the diapause may differ depending on the manner of entry to the photo-sensitive phase of diapause. (See text for details.) A further complexity is that the Daphnia embryos are not immediately sensitive to the low O 2 stimulus. Some interval of incubation in the stream is a necessary preliminary condition for response to low O 2 after only two weeks. Four weeks of exposure was indaquate when the embryos were removed from the culture vessel and placed directly in low O 2 . Preliminary experiments suggest that the embryos become primed for response to the low O 2 environment if the flow of medium in the stream is interrupted periodically. Photo-refractoriness in the diapaused embryo may represent a variety of inter- nal states, if responsiveness to the variety of stimuli employed in this and other studies may be used as criteria. A descriptive model (Fig. 8) of diapause sug- gests a minimum of five states, the first four (A-D) of which are photo-refractory. In the initial (A) state the diapause intensifies or deepens. During this state progressively fewer embryos are rendered photo-sensitive following such treat- ments as osmotic shock (Stress and Hill, 1968), decapsulation (Davison. 1969) etc. The following interval (B) is one of intense diapause, the duration of which 130 R. G. STROSS may be determined by temperature, genotype of tbe embryos, and, in the case of the strain used in ibis study, some unknown set of conditions. A third state ( C ) is marked bv tbe responsiveness of the embryo to low O 2 tensions, tbe effect of which is to induce a state (D) of sensitivity to a stimulus such as elevated CO.,. Tbe photo-refractory phase of diapause may be considered at an end when the "D" state has been satisfied. Since tbe light sensitizing action of CO 2 is reversible, tbe embryo may move reversibly from state C to state E, the final state, with the addition or withdrawal of CO 2 . In the final or E state the embryo is fully photosensitive and exposure of the embryo to light at the appropriate phase of an endogenous cycle will trigger activation. The final state may be forced with manipulation of the gaseous environ- ment, and it is suggested that only some strains of D. pnle.r are amenable, or it may result from incubation at the appropriate temperature. Alternative states of photo-sensitivity in the "E" phase of diapause development are described in Eigure 8. In one state (bottom line) the embryo is photo-inducible for only a part of a daily cycle w r hich may include two (e.g., Fig. 2) or one (e.g.. Fig. 3) phase of induction. In the alternative state (top line) the embryo may be continually sensitive to light during its "E" phase of diapause. Although sensitivity may be continuous, the rhythmic pattern of batching (Figs. 5 and 6) would suggest that photo-induction is also rhythmic, that is, the product of a photo- reaction is apparently "stored" until needed in an internally ordered activation process. DISCUSSION The diapaused embryo of Daphnia is potentially under the control of photo- period, as may be the entire life cycle. The requirement for light is readily demonstrable. However the requirement for one inductively long day may be obscured by the internal state of the embryo. In one state induction clearly requires one long day or two pulses of light that simulate one long day. The embryo may therefore have two, not one, photo-inducible phases in each endogenous daily cycle. Scant evidence suggests that either tbe dawn or sunset stimulus may be received first. In a second state only one light pulse of two hours is sufficient for inducing a return to active embryonic development. In this state the embryo is still manifesting time measurement since it is activated by the light pulse only at a certain time(s) or phase of a 24-hour thermocycle. The loss of one of two photo-inducible phases marks a second internal state and follows a first. Con- ceivably some arthropods have normally only one photo-inducible phase and its "loss" permits diapause termination in constant dark. There are many such examples of termination in constant dark and they include another strain of Daphnia fmle.r (Stross and Hill, 1968) as well as other arthropods, aquatic (Paris and Jenner, 1959) and terrestrial (Williams and Adkisson, 1964). That two photo-inducible phases may exist in arthropods other than Daphnia is demon- strated by the response of the aphid Mcyoitra to photoperiod (Lees, 1966). Other recent studies infer photo-induction of diapause termination at more than a single phase of a daily cycle. Hamner (1969), employing a regimen of CO, AND PHOTOPERIODISM 151 one short day followed by a cycle of complete darkness, shows that night inter- ruptions in the first night period are more effective in creating a long-day than are light interruptions at the appropriate time in the second cycle. A light-influ- enced preparative process, the product of which decays following the end of the main light period, was inferred. Saunclers (1970) using the same approach with a different insect described an opposite result in which interruptions during the second (and third) cycle were apparently more effective. Both studies suggest that two light controlled reactions may be involved in each inductive cycle. The reaction to light at supposedly two phases, one at sunset and the other at dawn, is consistent with the response to night interruptions when given in 24-hour cycles. When the main light period is suitably short (6 to 10 hours) there are normally two intervals in the night when a light interruption is inductive. Exposure to light early in the night may be inductive when it forms the sunset of a long day. Exposures in the late night are effective when they form the dawn of a long day (Pittendrigh and Alinis, 1964; Pittendrigh, 1966). Light exposure in the late-night is usually the more effective inducer of a long daylength (Adkisson, 1964, 1966; Saunders, 1968, 1970) presumably because the main light period now forms the sunset which may be interpreted as the phase of photo- inducibility. However, Pittendrigh (1966) strongly suggests the photo-inducible phase may be the late-night or dawn period, a view supported by the interpretation of the results in this paper. In that event the main light period would be in- volved with a second inductive phase which could be preparative for a second light reaction which may take place at dawn. The existence of a second inducible phase that occurs near (sunset) the end of the light period and which provides an essential product for a photo-induction that occurs at dawn may be supported by temperature manipulation. Experi- ments that combine a period of chilling with night interruptions have shown that chilling during the day may interfere with the preparation while those given at night seem to preserve the product of the preparative process. At least such an interpretation may be deduced from the results as described by Saunders (1968) and Saunders and Sutton (1969). The deliberate elevation of CO 2 tension, which may be necessary to break photo-refractoriness, introduces an additional complexity. The embryo now re- sponds to a light stimulus as though it were continuously photo-inducible. How- ever, the pattern of hatching suggests that the embryo may be photo-inducible only at certain phases of some internal cycle. To explain the continued sensitivity to light, it is suggested that some intermediate product is stored for a phase- specific process of activation. Storage apparently requires the presence of the elevated tension of CO 2 . Three potential functions of CO 2 in terminating the diapause of Daplinia are described in an extension of the so-called coincidence model (Pittendrigh and Minis, 1964) (Fig. 9). One function is to break the photo-refractory state of the embryo. The second is to store a light signal. The third function is to preserve for all, or a part, of each daily cycle a product of some dark reaction until the embryo is exposed to light. The three (x, y, z) functions of CO L , are deduced from the light requirements of the embryo in state one when there may be two photo-inductive phases each cycle. 152 R. G. STROSS The coincidence model describes photo-induction as a light activated enzyme converting a substrate when the concentration of the substrate achieves threshold concentration and that occurs once each circadian cycle. An extension (Fig. 9) describes two such substrate conversions, the products from each combining to form a final product. One intermediate product (Pii) reacts with a second inter- (A) LONG DAY 16 24 24 E . Ea Ei ^ t- SHORT DAY 12 24 24 . Ea A. (B) C0 2 ~ INDUCED TERMINATION JC02 Ei A A FIGURE 9. A model of photoperiodism describing the potential functions of CO 2 . The basic model is an extension of the coincidence model to include two "photo-inducible" phases in each daily cycle. Three functions of CO 3 are postulated to explain experimental observa- tions. They are the breaking of photo-refractoriness ("x"), substitution of one light stimulus ("y"), and "storage" of a light signal until the embryo becomes photo-inducible ("z"). See text for details. mediate (Pj 2 ) to form the final product (Pp), provided the concentration of Pi 2 is sufficient, as would be the case if the night were short. Viewed in these terms, the potential functions of CO 2 are easily described. The first function of CO 2 "unlocks" the machinery for making two intermediate products, shown in Figure 9B as the "x" function. This effect is analagous to excystment in parasites (Sommerville, 1964, 1966). The second function, "y," CO 2 AND PHOTOPERIODISM 153 is to preserve the product of the dark reaction which provides the second sub- strate (Pi,). This effect of CO 2 could be analagous to the preventative effect of CO, in short-day (long-night) induction (Barnett and Lilly, 1950, 1955) and possibly to the second function of stimulating molting in certain parasites. The third function, "z," is the presumed storage of the light signal until an essential substrate reaches critical concentration. Whatever the functions of CO 2 , it seems clear that termination of diapause in the Daphnia embryo may be controlled by daylength. Termination of the entire population in nature may therefore be synchronous. Synchrony is more likely to occur when the gaseous environment of the embryos is appropriately close to equilibrium with the atmosphere. Mr. Timothy Downing provided technical assistance. Dr. C. Edwards and Dr. J. Jacklet reviewed the manuscript and Dr. Jacket was instrumental in form- ulating the model of photoperiodism. Dr. J. Mackiewicz provided references to the literature on parasites. The author is grateful to these and others not specificially mentioned. Research supported in part from grants of the NSF and Joint Awards Council, State University of New York. SUMMARY The diapaused embryo of the cladoceran, Daphnia pulex, may require light for terminating the diapause, and a single photocycle may be adequate. In D. pulex the light-refractory phase of diapause may be broken or completed in a variety of environments. In the Supply House strain employed in these studies, the usual low temperature treatment is unnecessary when the embryos are placed in constant darkness in sealed containers. Alternatively, the refractory state in constant dark- ness may be broken with low O 2 and high CO 2 tensions. Both modifications were shown to be necessary and in that sequence. A single long-day light signal may terminate the diapause when the embryo passes from the photo-refractory to the photo-sensitive phase. Three kinds of light responses were observed, however. Each relates to the treatment given the embryos prior to light exposure. In one so called internal state the embryo requires one long day or two pulses of light. Scant evidence suggests that the first pulse may be interpreted as either dawn or sunset. In this state there are obviously two photo-inductive phases in each inductive cycle. The embryo may change, however, by "losing" one of the photo-inductive phases. In this condition a single two-hour pulse of light, if given at the appropriate time of a thermocycle, may terminate the diapause. A third internal state is introduced when CO 2 is employed to break the re- fractory phase of diapause. The diapause is terminated by a single two-hour pulse of light, and the embryo is continuously sensitive to the light stimulus. However, the pattern of hatching indicates that activation may occur at restricted phases within the induction cycle. Several roles of CO 2 may be deduced from the experimental results. An elevated CO 2 tension breaks photo-refractoriness. It also induces a "dark" 154 R. G. STROSS reaction which may need to he completed hefore the light may stimulate termina- tion of the diapause. Withdrawal of CO., after 10 hours prevents activation unless a second light pulse is given immediately thereafter. The apparent deferral of light stimulation suggests a third function of CCX. that of "storing" the light signal until the embryo becomes photo-inductive. The potential roles of CO., are described in a model of photoperiodism that requires photo-induction at two phases in each daily cycle. LITERATURE CITED ADKISSON, P. L., 1964. Action of the photoperiod in controlling insect diapause. Amer. Natnr., 98: 357-374. ADKISSON, P. L., 1965. Light-dark reactions involved in insect diapause. Pages 344-350 in J. Aschoff, Ed., Circadian Clocks. North Holland Publ. Co. Amsterdam. ADKISSON, P. L., 1966. Internal clocks and insect diapause. Science, 154: 234-241. BALLARD, L. A. T., 1958. Studies of dormancy in the seeds of subterranean clover (Trifoliniu subterranean L.) I. Breaking of dormancy by carbon dioxide and by activated carbon. Aust. J. Biol. ScL, 11: 246-260. BALLARD, L. A. T., 1967. Effect of carbon dioxide on the germination of leguminous seeds. Pages 209-219 in N. Borliss, Ed., Physiology, Ecology and Biochemistry of Germina- tion. Ernst-Moritz-Arndt-Universitat Press, Greifswald, East Germany. BALLARD, L. A. T., AND A. E. GRANT LIPP, 1969. Studies of dormancy in the seeds of sub- terranean clover (Trifolium subterranean L.) III. Dormancy breaking by low concentrations of oxygen. Aust. J. Biol. Sci., 22 : 279-288. BARNETT, H. L., AND V. G. LILLY, 1950. Nutritional and environmental factors influencing asexual sporulation of Choanephora cucurbitarum in culture. Phytopathology, 40: 80-89. BARNETT, H. L., AND V. G. LILLY, 1955. The effects of humidity, temperature and carbon dioxide on the sporulation of Choanephora cucurbitarum. Mycologia, 47: 26-29. BUNNING, E., 1959. Physiological mechanism and biological importance of the endogenous diurnal periodicity in plants and animals. Pages 507-530 in R. B. Withrow, Ed., Plwtoperiodism and Related Phenomena in Plants and Animals. American Associa- tion for the Advancement of Science Symposium #55, Washington, D. C. BUNNING, E., 1964. The Physiological Clock. Academic Press, New York, 145 pp. DEWiLDE, J., C. S. DUINTJER AND L. MOOR, 1959. Physiology of diapause in the adult Colorado beetle. I. The photoperiod as a controlling factor. /. Insect Phvsiol., 3: 75-85. DAVISON, J., 1969. Activation of the ephippial egg of Daphnia pulc.r. J. Gen. Phvsiol.. 53: 562-575. FAIRBAIRN, D., 1961. The in vitro hatching of Ascaris lumbricoidcs eggs. Can. J. Zoo/., 39: 153-162. HAMNER, W. M., 1969. Hour-glass dusk and rhythmic dawn timers control diapause in the codling moth. /. Insect Physio!.. 15: 1499-1504. LEES, A. D., 1966. Photoperiodic timing mechanisms in insects. Nature. 210 : 986-989. MINIS, D. H., AND C. S. PITTENDRIGH, 1968. Circadian oscillation controlling hatching: Its ontogeny during embryogenesis of a moth. Science, 159: 534-536. PARIS, O. H., AND C. E. JENNER, 1959. Photoperiodic control of diapause in the pitcher- plant midge. Metriocncmus knabi. Pages 601-624 in R. B. Withrow, Ed., Photo- periodism and Related Phenomena in Plants and Animals. American Association for the Advancement of Science Symposium #55, Washington, D. C. PITTENDRIGH, C. S., 1966. The circadian oscillation in Drosophila pscudoobscura pupae: a model for the photoperiodic clock. Z. Pflanzcnphysiol., 54 : 275-307. PITTENDRIGH, C. S., AND D. H. MINIS, 1964. The entrainment of circadian oscillations by light and their role as photoperiodic clocks. Aincr. Natur., 98: 261-294. ROGERS, W. P., AND R. I. SOMMERVILLE, 1968. The infectious process and its reaction to the development of early parasitic stages of nematodes. Pages 327-348 in B. Dawes, Ed., Advances in Parasitology. Academic Press, New York. CO 2 AND PHOTOPERIODISM 155 SOMMERVILLE, R. I., 1964. The effect of carbon dioxide on the development of third stage larvae of Haemonchus contortns in vitro. Nature, 202: 316-317. SOMMERVILLE, R. I., 1966. The development of Haemonchus contortns to the fourth stage in vitro. J. ParasitoL, 52: 127-136. SAUNDERS, D. S., 1968. Photoperiodism and time measurement in the parasitic wasp, Nason-ia ritripcnnis. J. Insect Physiol., 14: 433-450. SAUNDERS, D. S., 1970. Circadian clock in insect photoperiodism. Science, 168: 601-603. SAUNDERS, D. S., AND D. SUTTON, 1969. Circadian rhythms in the insect photoperiodic clock. Nature. 221 : 559-561. STROSS, R. G., 1965. Termination of summer and winter diapause in Daphnia. Ainer. Zool., 5: 701. STROSS, R. G., 1966. Light and temperature requirements for diapause development and re- lease in Daphnia. Ecology, 47 : 368-374. STROSS, R. G., 1969. Photoperiod control of diapause in Daphnia. III. Two-stimulus control of long-day, short-day induction. Biol. Bull., 137 : 359-374. STROSS, R. G., AND J. C. HILL, 1968. Photoperiod control of winter diapause in the fresh- water crustacean, Daphnia. Biol. Bull., 134: 176-198. WELLSO, S. G., AND P. L. ADKISSON, 1964. Photoperiod and moisture as factors involved in the termination of diapause in the pink bollworm, Pcctinophora gossypiella. Ann. Ent. Soc. Amcr.. 57: 170-173. WILLIAMS, C., AND P. L. ADKISSON, 1964. Physiology of insect diapause. XIV. An endocrine mechanism for the photoperiodic control of pupal diapause in the oak silk- worm Antheraca pcrnyi. Biol. Bull., 127: 511-525. Reference: Biol. Bull., 140: 156-165. (February, 1971) OXYGEN POISONING IN THE ANNELID TUBIFEX TUBIFEX. II. OSMOTIC PROTECTION 1 JOANNE G. WALKER 2 Department of Physiology and Biophysics, University of Illinois, Urbana, Illinois 61801 Walker (1970) studied the effects of hyperbaric oxygen on Tnbife.v titbifc.r, an Annelid normally living' in environments with extremely low oxygen tensions. These animals were killed by exposure to four atmospheres absolute oxygen pres- sure for 15 hours but recovered from exposures of up to eight hours or when longer exposures were interrupted by a sufficient interval at atmospheric conditions (Walker, 1970). In the present study the ability of various agents to modify the toxic effects of oxygen on T. tnbife.v has been investigated. MATERIALS AND METHODS T. tnbife.v was exposed to four atmospheres absolute oxygen pressure by the procedures described previously (Walker, 1970). The water used in this study both for worm exposures and for preparation of reagents was tap water which had passed through activated charcoal filters and to which 0.05 g disodium dihydrogen ethylenediamine tetraacetic acid (versene, Hach Chemical Co.) per liter was then added. The effects of various agents were determined by adding 0.5 ml of the test solution to each depression of the paraffin block which already contained 4.5 ml water and one worm. Addition of agents was made immediately before oxygen exposure except where otherwise noted. In some experiments the effect of com- plete removal of the environmental water after oxygen exposure and replacement with another medium was tested. When heat stress experiments were run worms were exposed to increased temperatures in a water bath but were returned to room temperature to score survival. Worm responses were judged by the following criteria : the presence or absence of movement and the presence of damage visible without a microscope. Worms having progressive damage and showing no movement for several observa- tion periods were scored as dead. Such worms either became a grey, opaque, motionless mass or disintegrated completely. An asterisk following data for per cent survival in the tables indicates that the value does not correspond to an actual observation time. Such data were obtained from a graph of the per cent survival versus time after oxygen exposure for the experiment in question. 1 This paper is a portion of a thesis submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy at the University of Illinois, Urbana, Illinois. 2 Present address : Department of Pharmacology, College of Medicine, Ohio State Uni- versity, 370 W. 9th Ave., Columbus, Ohio 43210. 156 OSMOTIC PROTECTION IN O 2 POISONING 157 RESULTS 1. Protection b\> solutes The possibility that glycolysis might he inhibited by hyperbaric oxygen in T. tiibife.r was first considered. Worms in either one per cent glucose or in water were exposed to oxygen in two doses interrupted by various intervals. These preliminary glucose experiments were a portion of the dosage fractionation experi- ments reported previously (Walker, 1970). In one experiment glucose was added during the interval following the first oxygen exposure. In another experiment worms were exposed to oxygen in water, in one per cent glucose, in 2,4-clinitro- phenol at a final concentration of 1 X 10" 4 M and in one per cent glucose plus 1 X 10" 4 M 2,4-dinitrophenol. The results of these experiments are pre- sented in Table I. TABLE I Survival of T. tubifex as affected by addition of glucose and of dinitrophenol during fractionated oxygen exposures (16 worms per group) Oxygen treatment Post-exposure survival (per cent) Worm group First dose hours Interrup- tion hours Second dose hours 48 hours 153 5 hours Water 8 2 8 24.1* 0.0 1 % Glucose (added at start of first exposure) 100.0 100.0 1% Glucose (added at end of first exposure) 51.7* 12.5 Water 10 4 8 12.5 9.0* 1 % Glucose (added at start of first exposure) 91.1* 87.5 Water 8 6 8 75.0 68.8 1 % Glucose (added at start of first exposure) 100.0 93.8 DNP 1 X H)- 4 M (added at start of first exposure) 97.7* 93.8 1% Glucose + DNP 1 X 10~ 4 M (added at start of first exposure) 93.8 93.8 * By interpolation. One per cent glucose had a striking protective effect on oxygen-treated T. tubifc.i'. Glucose had some protective effect even when it was added after the first eight-hour oxygen exposure. Control experiments showed that 2,4-dinitro- phenol at 1 X 10-* M was not lethal for unoxygenated T. tiibifex; this concentra- tion, reported to uncouple oxidative phosphorylation (Loomis and Lipmann, 1948), not only did not reverse the protective effect of glucose but appeared to have some protective effect of its own. T. tubifc.r were also exposed to oxygen in one per cent Proteose-Peptone (Difco) during the dosage fractionation experiments. This medium is very favor- able for the growth of microorganisms. Worm depressions containing Proteose- Peptone rapidly became contaminated and the worms had to be discarded. Proteose-Peptone nevertheless provided a protection comparable to that of glucose during the period before fouling. 158 JOANNE G. WALKER The protective effect of both glucose and Proteose-Peptone was immediately apparent from the behavior of oxygen-exposed worms when they were removed from the oxygen chamber. Worms exposed to oxygen in glucose or Proteose- Peptone were moving normally and were not contracted, while worms exposed in water were highly coiled and almost motionless. Since such different substances as glucose, Proteose-Peptone and 2,4-dinitro- phenol all protected T. hibife.r against the toxic effects of oxygen and particularly since a moderate NaCl concentration was produced in dissolving and adjusting the pH of the 2,4-dinitrophenol solution, the possible protective effect of NaCl was suggested. Three groups of worms were therefore exposed to oxygen for 16 hours TABLE II Survival of T. tubif ex following 18 hours of oxygen exposure in various concentrations of glucose or sodium chloride (16 worms per group) Post-exposure survival (per cent) Worm group 25 4 hours 53 1 hours 94 5 hours 144 =fc 3 hours Water 62.5 62.5 43.8 37.5 Glucose 0.1 M 87.5 87.5 87.5 75.0 Glucose 0.01 M 87.5 87.5 87.5 81.3 Glucose 0.001 M 62.5 50.0 43.8 31.3 Glucose 0.0001 M 56.3 56.3 37.5 18.8 Water 8.3 8.3 8.3 8.3 NaCl 0.1 M 100.0 31.3 31.3 31.3 NaCl 0.01 M 100.0 87.5 87.5 81.3 NaCl 0.001 M 18.8 18.8 18.8 18.8 NaCl 0.0001 M 46.6 46.6 46.6 46.6 Water 25.0 18.8 18.8 18.8 NaCl 0.05 M 100.0 100.0 100.0 100.0 NaCl 0.02 M 100.0 100.0 100.0 100.0 NaCl 0.005 M 68.8 55.5* 43.8 43.8 * By interpolation. in water, in one per cent glucose and in NaCl at a concentration osmotically equivalent to the one per cent glucose and their post-exposure survival scored. Survival of T. tnbijc.v at 219 hours after oxygen treatment was 25 per cent for worms exposed to oxygen in water, 100 per cent for worms exposed to oxygen in one per cent glucose and 87.5 per cent for worms exposed to oxygen in NaCl osmotically equivalent to the glucose. Again the protective effect was clear from the worms' behavior on removal from the oxygen chamber. Worms exposed to oxygen in either glucose or NaCl were more active and appeared more like normal worms than did the worms exposed to oxygen in water. To determine the concentration providing maximum protection, worms were exposed to an 18-hour oxygen dose in solutions of glucose or NaCl ranging in concentration from 0.1 to 0.0001 M. Results of these experiments are presented in Table II. OSMOTIC PROTECTION IN O 2 POISONING 159 It is clear that maximum protection resulted from treatment with glucose solutions between 0.1 M and 0.01 M. The NaCl solution giving the greatest protection had a concentration of 0.01 M. Sodium chloride at Q.I M offered equivalent protection during the first day after exposure, but its beneficial effect decreased markedly during the following day. It is of interest that this relatively high concentration of NaCl was beneficial even though one half of the control worms not exposed to hyperbaric oxygen died in this concentration of NaCl within two days. \Yhen the NaCl optimum was more sharply defined, optimum and equivalent protection occurred in 0.05 M and 0.02 M solutions. These results are presented graphically in Figure 1. It is interesting to note that one per cent glucose used in all other glucose experiments has a molarity of 0.056 ; NaCl which is osmotically equivalent to one per cent glucose has a molarity of 0.028. 100 75 o Glucose A NaCl 0.0001 0.001 0.01 O.I 1.0 OSMOLAR CONCENTRATION FIGURE 1. Survival of T. tnbifc.v following 18 hours of oxygen exposure in various osmolar concentrations of glucose or sodium chloride. Because of the marked protective effect of NaCl addition, it seemed essential to determine whether salts other than NaCl provided similar protection. Sodium nitrate and the chlorides of various monovalent and divalent cations were tested. All solutions were used at concentrations both above and below the osmotic equivalent of one per cent glucose. Complete dissociation of the salts at these concentrations was assumed. Results of these experiments are given in Table III. These experiments indicate that protection against oxygen damage may be nonspecific in that salts with qualitatively different ionic composition had similar protective capacities. Calcium chloride at both concentrations and MnCl 2 at the lower concentration tested provided protection against hyperbaric oxygen equi- valent to that produced by NaCl. Both concentrations of KG and the higher concentrations of CoCl, and MnCl 2 were extremely toxic to the unoxygenated control worms. Most of the unoxygenated worms in the higher concentrations of KC1 and AInCL survived for two days but all were dead in five days. In 0.02 M KC1 all the unoxygenated worms survived for six days but all were dead 160 JOANNE G. WALKER TABLE III Survival of T. tubifex following 18 hours of oxygen exposure in various salt solutions (16 worms per group) Post-exposure survival (per cent) Worm group 5 1 18 1 27 3 42 2 137 185 1 hours hours hours hours hours hours Water 81.3 62.5 62.5 62.5 62.5 62.5 NaCl 0.05 M 100.0 100.0 100.0 100.0 93.4 93.4 NaCl 0.02 M 100.0 100.0 100.0 100.0 87.5 87.5 KC1 0.05 M 93.8 87.5 87.5 43.8 0.0 0.0 KC1 0.02 M 100.0 100.0 93.8 93.8 25.0 0.0 NaNO 3 0.05 M 100.0 100.0 87.5 81.3 68.8 56.3 NaNO 3 0.02 M 93.8 [93.8 93.8 87.5 68.8 68.8 CaCU 0.02 M 100.0 100.0 100.0 100.0 100.0 100.0 CaCl 2 0.002 M 100.0 100.0 100.0 100.0 93.8 93.8 Water 100.0 25.0 25.0 18.8 18.8 18.8 CoCI 2 0.02 M 100.0 62.5 0.0 0.0 0.0 0.0 CoCl 2 0.002 M 100.0 43.8 6.3 6.3 6.3 6.3 MnCl 2 0.02 M 100.0 75.0 0.0 0.0 0.0 0.0 MnCU 0.002 M 100.0 87.5 81.3 81.3 81.3 81.3 at nine days ; in 0.02 M CoCl, half the unoxygenated worms were dead in six days. All of the unoxygenated control worms survived in all other salt solutions. It is all the more impressive, therefore, that even those salts which were evenutally toxic to unoxygenated control worms delayed the death of oxygen-exposed worms during the first day following oxygen exposure. Sodium nitrate solutions pro- duced a somewhat different effect. Protection by this salt was at first equivalent TABLE IV Survival of T. tubifex as affected by replacement of the medium following 18 hours of oxygen exposure (16 worms per group) Post-exposure survival (per cent) * ji nt gi imp 16 hours 48 hours 250 12 hours Water not replaced 0.0 0.0 0.0 Water replaced with water 25.0 25.0 25.0 Water replaced with 0.028 M NaCl 81.3 86.8* 62.5 0.028 M NaCl not replaced 100.0 100.0 100.0 0.028 M NaCl replaced with 0.028 M NaCl 100.0 100.0 100.0 0.028 M NaCl replaced with water 93.8 93.5* 81.3 Water not replaced 12.5 12.5 0.0 Water replaced with water 12.5 8.5* 6.3 Water replaced with 1% Glucose 37.5 33.6* 18.8 1% Glucose not replaced 100.0 96.0* 87.5 1% Glucose replaced with 1% Glucose 100.0 100.0 81.3 1 % Glucose replaced with water 100.0 100.0 43.8 By interpolation. OSMOTIC PROTECTION IN O 2 POISONING 161 to the protection given by NaCl. This protective effect gradually decreased until at the termination of the experiment the NaNO 3 -treated worms had a mortality equivalent to that of the controls. Results in Tables II, III and V indicate that some worms in water survived after uninterrupted oxygen exposures of 16 to 18 hours. A variation in sensitivity of T. tubijex to oxygen was observed throughout these experiments which could not be strictly correlated with increasing storage time in the laboratory but may be related to the adaptation previously reported (Walker, 1970). 2. Time factor in protection Exchange experiments were set up in an attempt to determine when osmotic protection from NaCl or glucose addition was most effective. Worms were given an 18-hour oxygen exposure in water which was withdrawn in a syringe imme- diately after oxygen exposure and replaced with 0.028 M NaCl. Other worms TABLE V Survival of T. tubttex following addition of glucose or of sodium chloride at various times after 18 hours of oxygen exposure (16 worms per group) Post-exposure survival (per cent) Worm group 12 hours 77 hours 205 hours Water addition time 60.0 53.4 33.7 1% Glucose addition time 87.5 62.5 50.0 0.028 M NaCl addition time 73.4 73.4 66.7 0.028 M NaCl addition f hour 100.0 86.7 86.7 0.028 JWNaCI addition 1 hour 100.0 86.7 60.0 0.028 If NaCl addition 2 hours 80.0 73.4 66.7 0.028 M NaCl addition 4 hours 87.5 75.0 68.8 0.028 If NaCl addition 8 hours 68.8 62.5 62.5 0.028 M NaCl addition 12 hours 68.8 68.8 68.8 were exposed to oxygen in 0.028 M NaCl which was replaced with water after oxygen exposure. A similar experiment was run using one per cent glucose. Results of these experiments are shown in Table IV. The presence of either saline or glucose during oxygen exposure resulted in a markedly increased worm survival. The continued presence of saline after removal from oxygen provided complete protection whether or not the saline was replaced by fresh saline solution. Even the replacement of saline with water after oxygen exposure did not greatly decrease worm survival. The presence of glucose after oxygen treatment greatly increased worm survival. Post-exposure replacement of glucose with water resulted in survival intermediate between the survival of worms treated continuously with water and worms in the continued presence of glucose. Addition of glucose after oxygen treatment may have a slight beneficial effect whereas addition of saline after oxygen exposure leads to markedly increased worm survival. Some worms which appeared to be moribund after exposure to oxygen in water resumed activity several hours after saline addition. Oxygen damage, however, was too great to be reversed completely and they eventually died. 162 JOANNE G. WALKER The protective effect of post -exposure glucose and saline addition was substantiated by additional experiments ; the data are presented in Table V. Since addition of saline immediately following oxygen exposure resulted in a marked increase in worm survival, the next experiment was designed to test whether saline addition later in the recovery period would provide any protection and to determine at what time during the recovery period the saline addition would become ineffective. One trial was also run to determine whether treatment of the worms with saline (0.05 M) during the two hours previous to oxygen exposure in water had any beneficial effect. Worms in 4.5 ml water were exposed to oxygen for 18 hours. A 0.5 ml volume of water or saline or glucose was added to each worm depression at the time of removal from oxygen or, in the case of saline, at various time intervals up to 12 hours after oxygen exposure. The final glucose concentration was one per cent ; the final saline concentration was osmotically equivalent to that of the glucose. The results, shown in Table V, indicate that saline addition as late as 12 hours after oxygen exposure enhanced survival. Glucose addition immediately after oxygen exposure resulted in survival intermediate between that of the controls exposed in water and that of worms treated with saline. In another experiment a two-hour saline pretreatment had no protective effect ; survival was indistinguishable from that of worms receiving no saline pretreatment. 3. Protection against other types of stress The foregoing experiments showed that there is little specificity regarding the ions involved in protection against oxygen poisoning. It appeared equally important to determine whether protection is specific for oxygen poisoning or whether increasing the osmotic concentration of the environment would reduce mortality following other types of stress as well. To test this possibility, heat and hydrogen peroxide were used as stressing agents. (a.) Heat stress. In preliminary tests worms heated for five minutes in a water bath at 45 C exhibited, within a few minutes, a typical pattern of response: marked hyperactivity followed by segmental constriction and segmental rupture which soon ended in worm disintegration. A five minute exposure at 38 C had no detectable effect other than to stimulate worm activity. Worms heated at 37 C for 60 to 75 minutes exhibited the constriction-rupture-disintegra- tion response sequence spread over several days. Thus, prolonged exposure to 37 C offered more accurate visualization of the stress effects and appraisal of possible protective capacities of saline addition. In subsequent experiments, each worm was placed in either 5 ml water or 5 ml 0.05 M NaCl in a test tube. The tubes were heated in a water bath at 37 C for 75 minutes and subsequent responses of the worms were observed. During the first day following the heat stress, worms in water without added salt exhibited a high mortality. Only 14.4 per cent of the heat-treated worms survived after 21 hours; none were alive at 103 hours. This response was not unlike the typical response following a fairly high dose of oxygen. Saline treat- ment resulted in a marked initial protection ; 92.9 per cent of the worms exposed to heat stress in a saline environment were alive after 21 hours. Most of the OSMOTIC PROTECTION IN O 3 POISONING 163 saline-protected worms eventually died, however, so it appears that heat stress is not strictly comparable to oxygen damage at these dosages. (b). Hydrogen peroxide stress. Hydrogen peroxide exposure was used as another type of stress. Preliminary experiments indicated that 20 /*g per ml was a suitable hydrogen peroxide dose. Twenty four worms were therefore treated with 20 fig per ml hydrogen peroxide in combination with 0.05 M NaCl and their responses were compared to those of worms in hydrogen peroxide alone. After 640 hours 29 per cent of the worms in H 2 O, and 78.3 per cent of the worms in H 2 O 2 in combination with NaCl were alive. These data suggest that saline addition in combination with hydrogen peroxide has a beneficial effect on worm survival. DISCUSSION It is difficult to account for the protective capacities of such a widely diversified collection of compounds as glucose, Proteose-Peptone, 2,4-dinitrophenol, NaCl, CaCln, KC1, MnCl 2 , CoCL and NaNO 3 . Various ions, particularly divalent cations, have previously been reported to have a protective effect in oxygen poisoning. Magnesium in vitro (Dickens, 1946a) and in vivo (Wittner, 1957), manganese in vitro (Dickens, 1946a; 1946b) and in vivo (Wittner, 1957; Gerschman. Gil- bert and Frost, 1958b) and cobalt in vitro (Dickens, 1946a; Horn, Williams, Haugaard and Haugaard, 1967) and in vivo (Wittner, 1957; Gerschman, Gilbert and Caccamise, 1958a; Gerschman ct al., 1958b) all were found to decrease the toxic effects of oxygen. Several reports of protection by salts against other agents similar to the protection found in the present study have appeared in the literature. Freese, Bautz-Freese and Bautz (1961) found that when increasing concentrations of NaCl were present inactivation of phage T 4 by the mutagenic agent hydroxylamine was inhibited. Saier and Giese (1967) reported that the changes occurring in Paramecium multimicro-mtcleatum caused by exposure to ultraviolet irradiation were inhibited or delayed as the salt content of the medium was increased. Some protection was obtained when glucose solutions osmotically equivalent to the salt solutions were tested. The medium providing the best protection was hypotonic to the organism but high in calcium and magnesium. T. tubife.r is a freshwater dweller ; the tap water used in the present study con- taining the disodium salt of versene at an effective osmotic concentration of 4.5 X 10~* M without other added solutes therefore corresponds closely to the worms' natural environment. Palmer (1968) reported that 10 per cent sea water (equiv- alent to 0.058 M NaCl; Prosser, 1961) was the highest salinity tolerated by T. tubifc.v without gradual acclimatization. Above this concentration worm deaths occurred in 24 hours. In the present study, NaCl at 0.1 M, the only concentra- tion above 0.05 M tested, caused deaths among unoxygenated control worms also. Thus the highest salinity tolerated by T. tubifex corresponds closely to the saline concentration giving maximum protection against hyperbaric oxygen and the other stressing agents employed in the present study. Palmer (1968) suggested that in higher salinities T. tubifc.v might be expected to require less energy expenditure to maintain itself hypertonic than in dilute environments causing continual tissue dilution. Such an argument could also be 164 JOANNE G. WALKER applied to explain the present findings of solute protection during oxygen poison- ing. Palmer's experiments (1968) however did not support her hypothesis; she found no change in the oxygen consumption of T. tubijex in environments from zero to 20 per cent sea water. The rather narrow concentration range for maximum protection indicates that the osmotic concentration of the environmental medium is an important factor in the response of T. tubijex to oxygen. The protective effect of added ions appears to have little specificity and to be independent of the ionic composition of the salt. The enhanced survival observed in these experiments following oxygen exposure of T. tubijex in the presence of ions or glucose indicates that almost complete protection can be obtained against these doses of oxygen in this organism. The protection provided by NaCl even when it is added some time after oxygen exposure indicates that in T. tubijex the effect of oxygen is at least partially re- versible. The enhanced survival of T. tubijex in 0.05 M NaCl after heating or exposure to hydrogen peroxide suggests that increased salinity may be effective in protecting this organism against other environmental stresses. I wish to express my deepest appreciation to Dr. Howard S. Ducoff for his advice and guidance throughout this study, to Dr. John D. Anderson for his continued interest and to the Department of Physiology and Biophysics for the use of their facilities. SUMMARY 1. One per cent glucose when present during oxygen exposure provided sig- nificant protection of T. tubijex from doses which resulted in high worm mortality. 2. Glucose protection was not reversed by 2,4-dinitrophenol ; 2,4-dinitrophenol was itself protective. 3. Sodium chloride at a concentration osmotically equivalent to one per cent glucose provided protection equal to or better than that of glucose. 4. Proteose-Peptone, CaCl 2 , KC1, CoCl 2 , MnCl 2 and NaNO 3 also provided various degrees of protection when they were present during oxygen exposure. 5. The solute concentration providing maximum protection against oxygen poisoning had an optimum at a concentration osmotically equivalent to one per cent glucose. 6. Sodium chloride increased the survival of oxygen-exposed T. tubijex even when it was added as long as 12 hours after oxygen treatment. 7. Sodium chloride provided at least partial protection against the stress of heating or hydrogen peroxide exposure. LITERATURE CITED DICKENS, F., 1946a. The toxic effects of oxygen on brain metabolism and on tissue enzymes. 1. Brain metabolism. Biochem. J., 40: 145-171. DICKENS, F., 1946b. The toxic effects of oxygen on brain metabolism and on tissue enzymes. 2. Tissue enzymes. Biochem. J., 40 : 177-187. FREESE, E., E. BAUTZ-FREESE AND E. BAUTZ, 1961. Hydroxylamine as a mutagenic and inactivating agent. /. Mol. Biol., 3 : 133-143. OSMOTIC PROTECTION IN O 2 POISONING 165 GERSCHMAN, R., D. L. GILBERT AND D. CACCAMISE, 1958a. Effect of various substances on survival times of mice exposed to different high oxygen tensions. Amer. J. Physiol., 192: 563-571. GERSCHMAN, R., D. L. GILBERT AND J. N. FROST, 1958b. Sensitivity of Paramecium candatiini to high oxygen tensions and its modification by cobalt and manganese ions. Amcr. J. Physio!., 192: 572-576. HORN, R. S., C. D. WILLIAMS, E. S. HAUGAARD AND N. HAUGAARD, 1967. Toxic effects of oxygen on carbohydrate metabolism. Fed. Proc., 26 : 709. LOOMIS, W. F., AND F. LIPMANN, 1948. Reversible inhibition of the coupling between phos- phorylation and oxidation. /. Biol. Chan., 173: 807-808. PALMER, M. F., 1968. Aspects of the respiratory physiology of Tubifc.v tnhifc.r in relation to its ecology. /. Zoo/. London, 154: 463-473. PROSSER, C. L., 1961. Water: Osmotic Balance. Pages 6-56 in C. L. Prosser and F. A. Brown Jr., Eds., Comparative Animal Physiology. [2nd. Edition] W. B. Saunders Co., Philadelphia. SAIER, F. L. AND A. C. GIESE, 1967. The effect of ultraviolet radiation upon osmoregulation in Paramecium. Photochcm. Photobiol., 6: 745-755. WALKER, J. G., 1970. Oxygen Poisoning in the Annelid Tubifc.v tubifc.v. I. Response to oxygen exposure. Biol. Bull., 138 : 235-244. WITTNER, M., 1957. Inhibition and reversal of oxygen poisoning in Paramecium. J. Proto- zool, 4: 25-29. Reference: BioL Bull., 140: 166-189. (February, 1971) THE LARVAL AND POSTLARVAL DEVELOPMENT OF PARTHEN- OPE SERRATA REARED IN THE LABORATORY AND THE SYSTEMATIC POSITION OF THE PARTHENOPINAE (CRUSTACEA, BRACHYURA) 1 WON TACK YANG Institute of Marine and Atmospheric Sciences, University of Miami, Miami, Florida 33149 According to Rathbun (1925), Parthcnope (Platylainbrus} serrata (H. Milne Edwards) has a range extending from the Bermudas, and Cape Hatteras, North Carolina, through the Gulf of Mexico and Bahama Islands, to Bahia, Brazil. The known substrate on which it occurs consists of sand, broken shells, gravel, corals or combinations of these. Holthuis (1959) found ovigerous females in Surinam from May to June. Williams (1965), citing the United States National Museum records, said ovigerous females have been collected during June in North Carolina, summer in Florida, and October in Cuba. The studies of larval development of parthenopid crabs have been limited to the first zoeal stage or planktonic materials, and have dealt mostly with Mediterranean species. Gourret (1884) first mentioned the zoeae of Parthenope massena (Roux) (which he indicated with the name Lambrus massena}. Cano (1893) described and illustrated three zoeal stages and Boraschi (1921) illustrated a telson of Parthenope species (as Lambnts sp. ). Bourdillon-Casanova (1960) described and illustrated the first two zoeal stages of P. massena (as L. massena). Heegaard (1963) hatched and described the first zoeal stages of P. massena and P. angulifrons Latreille (as L. massena and L. angulifrons] , presenting a color plate to show the chromatophore patterns. Aikawa (1937) described the first zoea of the Japanese species P. valida De Haan (as L. valid-its} reared from hatching. The present work is the first study of the complete larval and post- larval development of any species of the Parthenopidae, based on reared animals. Since Milne Edwards' (1834) definition of family Oxyrhinques (=Oxy- rhyncha), all major workers of Oxyrhyncha (e.g., Dana, 1851; Miers, 1879; Alcock, 1895; Rathbun, 1925; Sakai, 1938; Balss, 1957; Garth, 1958) have placed the Parthenopidae under Oxyrhyncha. MATERIALS AND METHODS Two ovigerous females used in the present work were collected by a Biscayne Bay shrimp trawler on 4 March, 1965. These were kept in a running sea-water aquarium with sand. When eye spots appeared in the eggs each ovigerous female was then placed into a 25-cm diameter finger bowl of clear plastic with filtered 1 Contribution No. 1291 from the Rosenstiel School of Marine and Atmospheric Sciences, University of Miami. This work was supported by Public Health Service research grant GM- 11244. 166 LARVAL DEVELOPMENT OF PARTHENOPE 167 Biscayne Bay water. The females were fed cut shrimp and the water was changed daily. Hatching occurred on 13 March and 4 April, 1965. After hatching, each larva was coded with a serial number and placed singly in a compartment of a plastic culture box. Each hatch consisted of 90 larvae reared at 25 C; 35 larvae from the first brood and 45 larvae from the second brood were reared at 20 C. Each compartment contained about 20 cc of filtered Biscayne Bay water. The first three zoeal stages were fed the smallest nauplii of just-hatched Artemia eggs, obtained by filtering through stainless steel screen (105 X 105 mesh, 0.0003 inch wire). The later zoeal and megalopal stages were fed unfiltered newly hatched Art cm in. Chopped fresh shrimp was fed to the early stage crabs. Animals were changed to compartments with clean seawater, fed fresh food daily and were checked for molted and dead individuals. Salinity range of the rearing seawater was 34-37/c. Exuviae and specimens of each developmental stage were preserved in 7% buffered formalin. These were stained with Mallory's acid fuchsin red or chlorazol black. The larval appendages for drawing were dissected in 85% lactic acid, while crab appendages were dissected in full strength ethylene glycol. The appendages after drawing were mounted permanently in Turtox CMCS, a stained water miscible mounting medium. Drawings were made with aid of a camera lucida. Details of the appendages were checked under high power (400 X or more). Four specimens of each stage were measured with a calibrated ocular micrometer. The carapace length of zoeae was measured in lateral profile from the anterior margin of the ocular peduncle to the extremity of the posterior margin of the carapace. Carapace length of megalopa and crab were measured dorsally from the tip of the rostral spine to the posterior margin of the carapace, and the width was measured across the widest part of the carapace. The size given for each stage is the arithmetic average of four specimens examined. The formulae for order of setation denote the number of setae from the proximal to the distal segment (or group) and was also based on four specimens. The female crabs are deposited in the museum of the Institute of Marine and Atmospheric Sciences of the University of Miami (UMML 32-3356 and UMML 32-3582). MORPHOLOGICAL RESULTS The number of zoeal stages through which Parthenope serrata passes to reach the megalopa is five or six. Most individuals surviving to the crab stage had passed through six zoeal stages. The text figures provided are as accurate as possible, hence the text is mainly restricted to comments on features not obvious from the figures or to emphasize any variations that were found. First zoea (Fig. 1) Carapace: Length approximately 0.37 mm. Dorsal, rostral and lateral spines as illustrated. Forehead rounded, with muscle bands plus a small protuberance. No anterior seta, nor infero-lateral setae present. A seta on each postero-lateral part of dorsal spine. Eye sessile with minute ocular papillae. Abdomen: Lateral knobs, postero-lateral process as shown. Tclson : Lunate, bearing a dorsal spine 168 WON TACK YANG A FIGURE 1. Parthcnope scrrata: First zoea ; A, lateral view; B, anterior view of carapace; C, abdomen; D, antenna 1; E, antenna 2; F, maxilla 2; G, maxilla 1; H, maxilliped 1; I, maxilliped 2 ; J, antero-ventral view of mandible ; K, postero-median view of mandible. Bar scales represent 0.2 mm. LARVAL DEVELOPMENT OF PARTHENOPE 169 Hl- DE- FG- FIGURE 2. Parthcnope serrata: Second zoea; A, lateral view, B, anterior view of carapace C, abdomen; D, antenna 1; E, antenna 2; F, maxilla 2; G, maxilla 1; H, maxilliped 1 I, maxilliped 2. Bar scales represent 0.2 mm. 170 WON TACK VANG on each fork. Occasionally a minute setnle present on lateral margin of telson halfway to fork, visible only under high magnification (400 X ). Mandible: Nine to ten teeth on antero-ventral margin, three teeth on right postero-medial margin (none on left). Maxilliped 1 : Endopodite, five-segmented, setae arranged (2, 2. 1, 2, 5). Color: Melanophores located on dorsal basal area of each lateral spine ; on posterior basal part of dorsal spine ; postero-ventral portion of carapace : on each mandible ; on each side of the ventral surface of abdominal somites 1-5 ; on basal part of antenna 2. Inconspicuous melanophore located on interorbital area of rostrum and at middle part of basipodite of maxilliped 1. Eye a light yellow-green. Pereiopods: Buds present, minute. Second zoea (Fig. 2) Carapace: Length approximately 0.48 mm. Eyes stalked. In subsequent stages dorsal spine continually shortens in relation to carapace length and anterior base thickens. Four setae added on forehead, plus five hairs along infero-lateral margin. Abdomen: Lateral process of somites 3-5 more elongated. Telson: Inner margin less lunate than zoea 1. Dorsal spine on fork now reduced. Maxilla 1: A plumose seta added to (proximal to) endopodite. Basal and coxal endites each with increased number of setae. Ma.villa 2: Scaphognathite setae increased to approximately eight ; apical process now divided into three. Maxilli- peds 1 and 2: Now with six natatory setae. Third zoea (Fig. 3) Carapace: Length approximately 0.59 mm. Forehead and anterior base of dorsal spine nearly a straight line. Additional setae added below forehead pro- tuberance. Groove forming in proximal portion of rostral spine between eyes. Abdomen: Somite 6 now separated from telson. A pair of mid-dorsal setae on somite 1 (one seta in some individuals). Antenna 1: One small aesthetasc added. Antenna 2: Small endopodite bud appears. Maxilla 1: Basal endite now with nine setae. Maxilla 2: Space between three processes of endopodite noticeably increased. Scaphognathite with approximately 10 setae in addition to apical processes. Maxillipeds 1 and 2: Eight natatory setae. Maxilliped 1: Setae now arranged (2, 2, 1, 2, 6). Fourth zoea (Fig. 4} Carapace: Length approximately 0.66 mm. Lateral spines further reduced. Rostral spine at interorbital area well grooved. Abdomen: Somite 1 usually with three mid-dorsal setae. Telson: Dorsal spine on fork now minute. Antenna 2: Endopodite bud elongated. Maxilla 2: Scaphognathite usually with 14 plumose setae plus three to five divided apical processes. Maxillipeds 1 and 2: Eight and nine natatory setae, respectively. Occasionally these numbers reversed ; or rarely eight on both. Pereiopods: Buds elongated and easily seen. Pleopods: Bud primordia recognizable. LARVAL DEVELOPMENT OF PARTHENOPE 171 FIGURE 3. Parthenope scrrata: Third zoea; A, lateral view, B, antero-lateral view of carapace; C, abdomen; D, antenna 1; E, antenna 2; F, maxilla 2; G, maxilla 1; H, maxill- iped 1 ; maxilliped 2. Bar scales represent 0.2 mm. 172 WON TACK YANG FIGURE 4. Parthenope scrrata: Fourth zoea; A, lateral view; B, antero-lateral view of carapace; C, abdomen; D, antenna 1; E, antenna 2; F, maxilla 2; G, maxilla 1; H, maxilliped 1 ; I, maxilliped 2. Bar scales represent 0.2 mm. LARVAL DEVELOPMENT OF PARTHENOPE 173 FIGURE 5. Parthenopc scrrata: Fifth zoea; A, lateral view; B, antero-lateral view of carapace ; C, abdomen ; D, antenna 1 ; E, antenna 2 ; F, maxilla 2 ; G, maxilla 1 ; H, maxilliped 1 ; I, maxilliped 2. Bar scales represent 0.2 mm. 174 Fifth zoea (Fig. 5) WON TACK YANG Carapace: Length approximately 0.79 mm. General form more depressed than previous stages along with more stout and shorter dorsal spine. Two small setae on infero-lateral margin ; a few minute setae posterior to these on future branchiostegite membrane. Abdomen: Somite 1 with approximately four FIGURE 6. Parthenopc scrrata; Sixth zoea; A, lateral view; B, antero-lateral view of carapace; C, abdomen; D, antenna 1; E, antenna 2; F, maxilla 2; G, maxilla 1; H, maxilliped 1 ; maxilliped 2. Bar scales represent 0.2 mm. LARVAL DEVELOPMENT OF PARTHENOPE 175 dorsal setae. Telson: Dorsal spine on fork now rudimentary. Antenna 1: Six terminal, two (rarely one) subterminal aesthetascs. Distal position anticipates segmentation. Inner flagellum hud appears on medial margin. Antenna 2: Endopodite bud now extends beyond exopodite. Maxttlipeds 1 and 2: Usually ten natatory setae each. Perewpods: Buds enlarged, chelation beginning. Pleo- pods: Buds much elongated. Gills: Buds now visible. Sixth zoea (Fig. 6) Carapace: Length approximately 1.01 mm. More depressed: dorsal spine stouter, inclined more posteriorly than previous stage. Lateral spine much shorter. Rostrum thickened, with median ridge on proximal portion plus two setae. Abdomen: \Yith approximately five mid-dorsal setae on somite 1. Antenna 1: Peduncle, inner flagellum, and dorsal flagellum differentiated. Partly segmented peduncle with a dorsal indentation. Dorsal flagellum with three groups of aesthe- tascs placed (5, 5, 2). Antenna, 2: Endopodite remarkably elongated plus two terminal setae; anticipating segmentation. Mandible: Anterior proximal portion with a palp bud. Maxilla 1: Form remains unchanged from second zoeal stage but setae increase in number. Max ill a 2: Scaphognathite with approximately 23 setae ; with branched apical processes. Maxilliped 1 : Ten natatory setae. Basi- podite setae increased (2, 2, 3, 3). Maxilliped 2: Eleven natatory setae. Perewpods: Buds well developed. Chelae almost formed. Pleopods: Elongated, with terminal setae. Gills: Buds continue forming in respective locations; in exuviae they appear to be laminated. Megdopa (Figs. 7, 8) Carapace: Length approximately 1.47 mm; width 0.98 mm. Medial portion of rostrum at interorbital region slightly depressed. Gastric region moderately inflated and without protruding ridges as in majiid megalopae. Hepatic lobe a hemisphere. Cardiac spine (dorsal spine in zoeal stages) with few minute setae. Abdomen: Six somites plus telson ; setation as shown. Antenna 1: Peduncle three-segmented ; dorsal portion of proximal segment concave, sup- porting proximal portion of eye peduncle with expanded lateral margin. Segmenta- tion between distal and penultimate segments of outer flagellum unclear. Antenna 2: No modifications on distal portion of basal article (such as in majiid megalopae). Mandible: Now with two-segmented palp. Maxilla 2: Apical processes of scaphognathite now bushy plumose setae, similar in form to marginal setae. A few setae on dorsal and ventral surface of scaphognathite. Endopodite with added lateral plumose setae. Maxilliped 1: Shape now radically changed; epi- podite now fringed with seven or more smooth hairs. Maxilliped 2: Epipodite bud with one or two terminal setae. Maxilliped 3: As illustrated. Pereiopods: Sparsely covered with setae. Right cheliped slightly larger than left. Dactyl of pereiopod 5 with long subterminal hair (^feeler) ; hair tip hook-shaped. Pico- pods: Present on somites 2-6, natatory setae arrangement on exopodite progressing distally as follows: 13, 14, 13, 10, 4. Numbers of natatory setae inconsistant, varying considerably between specimens, and on different sides of same individual. Appendix interna usually with three hooks. 176 WON TACK YANG FIGURE 7. Parthcnope scrrata : Megalopa A, lateral view ; A, dorsal view ; B, left cheliped ; C, right cheliped ; D, pereiopod 2 ; E, pereiopod 5 ; F, pleopod ; G, antenna 1 ; H, mandible ; I, maxilla 1. Bar scales represent 0.2 mm. LARVAL DEVELOPMENT OF PARTHENOPE 177 FIGURE 8. Parthenope serrata : Megalopa ; A, antenna 2 ; B, maxilla 2 ; C, maxilliped 1 ; D, maxilliped 2 ; E, maxilliped 3. Bar scale represents 0.2 mm. Gills: Development more advanced than in majiid megalopae. Maxilliped 2 with one elongate bud of future podobranch, incipient epipodite with two minute distal hairs. Maxilliped 3 with well laminated posterior arthrobranch, elongate 178 WON TACK YANG FIGURE 9. Parthenope serrata : First crab ; A, lateral view of carapace ; A', dorsal view ; B, ventral view of carapace ; C, ventral view of right cheliped ; D, ventral view of left cheliped ; E, pereiopod 5 ; F, antenna 1 ; G, mandible ; H, maxilla 1 ; I, maxilla 2. Bar scales represent 0.2 mm. LARVAL DEVELOPMENT OF PARTHENOPE 179 bud of anterior arthrobranch, and small bud of podobranch. Cheliped with two well laminated arthrobranchs. Pereiopods 2 and 3 each with a pleurobranch. Color: Carapace shaded pale yellow. Melanophores on carapace and append- ages placed as follows: stippled over dorsal surface of carapace; on proximal por- tion of rostrum at interorbital area ; on posterior base of cardiac spine ; on dorsal and inferior side of each branchial postero-lateral spine ; on peduncle of eye ; on first segment of peduncle of antenna 1 ; on coxopodite, meropodite and propodite of pereiopods 2-5 ; and on ischiopodite of maxilliped 3. Melanophores on cheliped as follows : two at base of hiatus, two at proximal portion of propodite, one at carpopodite, three small ones at meropodite ; paired on ventral side of abdominal somites 2-6. Basipodite of pleopods on somites 2-6 and telson each with a faint melanophore and red chromatophore. First Crab (Figs. 9, 10} Carapace: Length approximately 1.70 mm; width 1.48 mm. Depth of carapace at cardiac region slightly greater than at gastric region. Medial portion of rostrum depressed into groove. Abdomen: Seven-segmented. No marked locking mecha- nism between it and thoracic sternum. Antenna 1: Dorsal surface of peduncle cup-shaped, well calcified with several hairs, acting as ventral floor of eye stalk. Antenna 2: Flagellum less calcified than peduncle, with seven segments. Maxilla 1: Endopodite segmentation obscure. Plumose seta from second zoeal stage still present on lateral margin, proximal to endopodite. Maxilla 2: Scaphognathite now with approximately 52 plumose setae. Maxilliped 1: Endopodite hatchet-shaped, apparently one-segmented. Maxilliped 2: Epipodite still small with two terminal hairs. Maxilliped 3: Ischio-basipodite cleavage slightly marked, with serrated ischiopodite medial margin. Pereiopods: Lateral margins of cheliped strongly serrated as in adult. Right cheliped, as in megalopa, much larger than left. Pleopods: Buds of subsequent crab stage pleopods 2-5 apparently present, but hard to detect. Gills: Appear on appendages as follows : Maxilliped 2 with elongated bud of epipodite bearing one or two terminal hairs, and elongated podobranch bud. Bud of arthrobranch not apparent in this stage but is observed in second crab stage. Maxilliped 3 with well laminated posterior arthrobranch, and slightly laminated bud of podobranch and anterior arthrobranch. Cheliped with two well laminated arthrobranchs. Pereiopods 2 and 3 each with a pleurobranch. Color: Minute melanophores quite densely spotted on carapace, and on third maxilliped. Chelipeds, pereiopods 2-5, and proximal portion of antennae with small melanophores. Paired melanophores on external side of abdominal somites 2-6. REARING RESULTS Two broods from different ovigerous females collected at the same time and locality have been reared. Although the morphological study was based on material from the two broods, the present rearing results are of the brood which hatched on 4 April 1965. An ovigerous female (carapace length 17.2 mm; carapace width including lateral spines 24.5 mm) produced approximately 3900 zoeae in one brood. ISO WON TACK YANG Figure 11 illustrates survival and mortality of the animals reared at 25 C including the individuals which molted into megalopa after five zoeal stages. A little less than one half of the initial population molted into second zoea. After FIGURE 10. Parthcnope serrata : First crab ; A, antenna 2 ; B, maxilliped 1 ; C, maxilliped 2 ; D, maxilliped 3. Bar scale represents 0.2 mm. LARVAL DEVELOPMENT OF PARTHENOPE 181 O z < O o CO < N u I S1VWINV dO 38WnN OV30 FIGURE 11. Parthenope scrrata: rearing record at 25 C. The horizontal scale represents days after hatching. The vertical scale represents number of animals. The lower figure indicates number of deaths per day. The symbols 1 to 6, M, cl and c2 represent, respectively, zoea 1 to zoea 6, megalopa, crab 1 and crab 2 stages. 182 WON TACK YANG the second zoeal stage, the number of individuals surviving into succeeding stages gradually decreased. Approximately one-ninth of the initial population reached the megalopa stage, but mortality in the megalopa stage was high. In this rearing none of the megalopae which had passed through five zoeal stages molted into the first crab stage. Three individuals out of the initial zoeal population of 90 reached the first crab stage in about 30 days after hatching and two individuals developed into the second crab stage. Among the larvae of this brood reared in 20 C two individuals reached the first crab stage in about 45 days ; one individual reached the second crab stage but died immediately after molting. CRAB 1 MEGA- LOPA o l/l ZOEAI \ 34 5 6 7 ^ 9 10 11 MEAN DURATION IN DAYS FIGURE 12. Duration of the larval and post-larval stages of P. scrnita under two different temperatures. The mean duration of each stage at 20 and 25 C is shown in Figure 12. The data of individuals reared at 20 C are based on an initial zoeal population of 45 individuals. The duration of the first crab stage at 20 C rearing is based on a single individual. The first zoeal stage at both temperatures had a greater duration than the succeeding stage. The majority of individuals which had a very prolonged first zoeal stage failed to complete development although some reached the intermediate zoeal stages. The general pattern of mean duration is quite similar in both 20 and 25 C although mean duration of intermediate zoeal stages was longer at 20 C than at 25 C. A prolonged duration of the megalopa stage is well marked in both groups. Provenzano (1968) observed a similar pattern in a study of anomuran larvae, and suggested that the prolonga- tion of the metazoea and megalopa was probably due to reorganization of body structure. Robertson (1968) found a similar though less marked trend in his LARVAL DEVELOPMENT OF PARTHENOPE 183 TABLE I Comparison of characters of first zoeal stage of Partlienopidae of known parentage Species Parthenope serrala H. Milne Edwards P. valida De Haan P . massena (Roux) P. angulifrons Latreille Reference Present work Aikawa, 1937 Heegaard, 1963 Heegaard, 1963 Dorsal spine Much longer than carapace depth, tapers distally, bends pos- teriorly Almost equal to carapace depth, tapers distally, bends pos- teriorly Much longer than carapace depth, tapers distally, bends pos- teriorly Much longer than carapace depth, stout, and blunt, somewhat straight Rostral spine Long, pointed Long, pointed Long, pointed Long, pointed Lateral spine Fairly long Short (?) Fairly long Fairly long Postero-lateral processes on somites 3-5 Elongated pro- jection, pointed No projection, not pointed Elongated pro- jection, pointed No projection, not pointed Telson Lunate Lunate Lunate Slightly lunate Spine on telson fork 1 dorsal spine only 1 dorsal only 1 dorsal spine, 1 lateral seta 1 dorsal spine, 1 lateral seta Exopodite of antenna 2 1 -segmented, 2 terminal spines 1-segmented, 2 terminal spines 2-segmented, 2 terminal spines 1-segmented, tapers distally with 2 subter- minal spines Endopodite of maxilla 1 (seta dist.) (seta prox.) 2-segmented, (2-4) (1) 2-segmented (2-4) (1) 3-segmented (2-4) (0) 3-segmented (2-4) (0) Medial margin of endopodite of maxilla 2 (setae) 3 processes (2,2,3) 2 processes (?) (2,3 + 2) 3 processes (2,1,2) 2 processes (2,4) Endopodite maxilliped 2 3-segmented 3-segmented 1-segmented (?) 2-segmented (?) Locality Florida, U. S. A. Japan Mediterranean Mediterranean study of phyllosomas of the lobster, Scy Hants americanus (Smith). The marked prolongation of the megalopa stage has been seen in the larval development of majid crabs which have only two zoeal stages (Yang, 1967). DISCUSSION The number of zoeal stages in development of the Parthenopidae has been in question for a long period. After noting that Cano described three zoeal stages, Lebour (1928, page 555) predicted that "there probably are four or five." Bour- 184 WON TACK YANG dillon-Casanova (1960) also assumed that the Parthenopidae had more than two zoeal stages. Table I is constructed from the major anatomical distinctions and similarities among P. s errata and three species of parthenopid first zoeae of known parentage reared by Aikawa (1937) and Heegaard (1963). There are some apparent character differences among the four reared species of parthenopid zoeae. However, there are unique zoeal characters which allow the larvae to be identified as parthenopids. These are summarized from the four species plus Partlicnope (P.} agona and Solenolambrus typicus, also reared by the author (Yang, unpublished) : 1. The carapace has well developed dorsal, lateral and rostral spines. 2. The outline of the telson is lunate with a "dorsally located spine" on each fork. A minute hair-like lateral seta is sometimes present. 3. Lateral knobs are present on abdominal somites 2-3. 4. The forehead protuberance is well developed as in most majid zoeae and the postero-dorsal knob of the carapace also appears to be present in the majority of parthenopids studied. 5. The endopodite of maxilliped 2 appears to be three-segmented with a disto-medial seta on the two proximal segments. The carapace of the first zoeae is more or less spherical with long dorsal, rostral, and lateral spines. As zoeae molt to the subsequent stages, the distance be- tween the anterior part of the eye and the basal part of the dorsal spine increases. The length of the dorsal and lateral spines rapidly decreases in proportion to the carapace size as does the rostral spine, while carapace height is also reduced at each subsequent stage. The dorsal spine thickens basally and shifts posteriorly. Thus, the distance between the posterior base of the dorsal spine and the postero- dorsal margin of the carapace decreases. At the sixth (last) zoeal stage, the carapace is remarkably depressed with shorter dorsal and reduced lateral spines and broader rostral spines. Thus, the carapace form becomes similar to that of the megalopa. A Parthcnope megalopa retains the lateral spines of the zoeae on the branchial region. The so-called dorsal spine of the zoeal stage is a cardiac spine. This gradual change of shape is also seen in the zoeal stages and megalopa of Callinectes illustrated by Costlow and Bookhout (1959), and among other long-zoeae-stage crabs. Further, this shifting of the zoeal form into the megalopa is clearly observed in Stenorhynchus seticornis (Herbst) and Rochinia hystrix (Stimpson), both having but two zoeal stages (Yang, 1967). Costlow (1965), and Rice and Provenzano (1966) reviewed the variability in larval stages of decapod larvae. In the present study, in one rearing of P. serrata at 25 C, zoeae developed into the megalopa stage. Thirteen of these molted from the fifth zoeae to megalopa while eleven molted from the sixth zoeae. Of the former, ten individuals died during or immediately after molting and three died later. Of those produced after 6 zoeal stages, two megalopae subsequently de- veloped into crabs. In the fourth zoeae of the five-staged series there are nine and ten natatory setae on maxillipeds 1 and 2, respectively. The fifth zoea of these series each have ten natatory setae on maxillipeds 1 and 2. The five-staged zoeae have a well-developed endopodite on antenna 2 and pleopods buds at the fifth stage, and the length is similar to the normal six-staged zoeae, but the development of LARVAL DEVELOPMENT OF PARTHENOPE 185 antenna 1 appears to be poor without an inner-flagellum bud, and with only two rows of aesthetascs. In normal six-staged zoeal development, the inconsistent number of the plumose natatory setae was mentioned earlier. ' This variability of number of natatory setae is also reported by Yatsuzuka (1957) for Portitnits pclagicus (L., as Nep- tunus pelagicus} and five other species of crabs. I studied the larval and postlarval development of more than twelve species of Majidae from hatching, and determined that the zoeae of the Parthenopidae, which have Brachyrhynchan zoeal characters, are easily distinguished from the majid zoeae in the following respects : 1. Six zoeal stages are observed in P. serrata, whereas in the true majid crabs the number of zoeal stages is apparently fixed at two, reflecting precocious or abbreviated development. The early zoeae of P. serrata and of the other Parthenopidae discussed above are alike in their anatomy, which is of a type characteristic for crabs with multi-zoeal development. 2. The zoeae of the Parthenopinae have dorsal, rostral, and lateral spines on the carapace. In general, dorsal and rostral spines are seen in the true majid zoeae. In the zoeae of the Inachinae (Majidae) only a dorsal spine is present. 3. The telson of the zoeae of the Parthenopinae is lunate with a spine located dorsally on each fork. In the Majidae this spine is lateral. 4. The exopodite of antenna 2 differs from that of general majid zoeae in having 2 terminal instead of 2 subterminal spines, plus a well developed endopodite bud in the first zoeal stage. 5. The "anterior seta" (Bourdillon-Casanova, 1960), present in the majid zoeae, is wanting in the zoeae of Parthenopinae. Lebour (1928), discussing parthenopid zoeal characters, stated (page 555), "They are more like Cancridae, having all the carapace spines, antennae like Portunus (actually Macropipus), and only one lateral spine on the telson. If these larvae be correctly identified and representing the Parthenopidae, then this family does not agree with Majidae in any way, and is an exception among the Oxyrhyncha." Aikawa (1935) merely mentioned that the zoeae of the Partheno- pidae are remarkably close to those of the Cancridae, and later he placed the zoeae of the Parthenopidae into a group of the Cancridae (Aikawa, 1937). Although both parthenopid and cancroid zoeae share many zoeal features, the cancroid zoeae of known parentage described by Lebour (1928), Aikawa (1937), Fagetti (1960), Mir (1961), and Pool (1966) show a remarkably different type of telson compared to that of the Parthenopidae. The cancroid zoeae possess lateral-knobs only on abdominal somite 2. Lebour (1944) gave a brief description and illustration of a megalopa which she attributed to a species of Parthenope, based on the elongated chelipeds and lack of "feelers" on the dactyl of pereiopod 5. She mentioned that her megalopa also had lateral spines on abdominal somites 3-5, and the pleopod on somite 6 bore numerous spines. There are no such features in P. serrata. The elongated rostral spine and the posteriorly elongated cardiac spine of megalopa she examined are somewhat similar to P. serrata. Lebour (1928, page 490) stated, "The presence of feelers on the last joint of the last legs cuts off the Brachyrhyncha (with exception of Pinnotheridae) from the Oxyrhyncha and from Ebalia. . . ." Gamo (1958) said that the number 186 WON TACK YANG of feelers in megalopa of eight species of Grapsidae ranges from three or four. There is only one feeler on the megalopa of P. scrrata. Its tip is not serrated. The megalopa of P. scrrata can be distinguished from majid megalopae by the following : 1. In P. scrrata there is a smooth feeler on the dactyl of periopod 5; none in majids. 2. There are no knobs, projections or modifications on the distal portion of the basal article of antenna 2 in P. serrata. There are more flagellar segments than the four that occur in majid megalopae. 3. The first segment of the peduncle of antenna 1 in P. scrrata is inflated ; its dorsal surface is flattened and concaved. This serves as a partial support of the eyestalk. 4. The gills are much better developed, having the elongated bud of the future podobranch and a bud of the epipodite on maxilliped 2. This is char- acteristic of the megalopae of multizoeal crabs (Yang, 1967). The following characters distinguish the first crab stage of P. serrata from those of majid crabs : 1. In the first crab of P. scrrata, the shape of the carapace is quite similar to that of the adult, as in the first crabs of some majids (e.g., Stenorhynchns and Epialtus). However, the first crabs of other majids, e.g., Libinia, Microphrys and Macrocoeloma are very different from the adult. 2. The inflated and laterally expanded first segment of the peduncle of antenna 1 now serves as the ventral support of the eyestalk. In the early post- larval stages of majid crabs (Acanthonychinae, Pisinae, and Mithracinae), the knob on the disto-dorsal surface of the basal article of antenna 2 locks into the ventral socket of the rostrum. Thus, the basal article is the ventral support (floor) of the eyestalk in the majid crabs. 3. The podobranch of maxilliped 2 is well developed in the first, and second and the parent adult crabs. As in adult Cancer (cj. Pearson, 1908), it is located along the anterior margin of the branchial series. The developmental pattern, as well as the character of respective zoeal stages of the Parthenopinae, is completely different from those of the Majidae. The subfamily Parthenopinae in the Parthenopidae is much larger group (about 128 species) than the other subfamily Eumedoninae (about 25 species). The latter occur in the Indo-Pacific and the species are mostly commensal (Balss, 1957). The larvae of the Eumedoninae are not yet known and may differ from those of the Parthenopinae. The Hymenosomatidae, one of three families of Oxyrhyncha, has unique zoeal characters (lack of the dorsal and lateral spines, and peculiar telson) which differ from those of the Majidae and the Parthenopidae (Gurney, 1938). Boschi et a!. (1969) found three zoeal stages and the carapace with only a rostral spine in the hymenosomatid, Halicarcinus planatus (Fabricius) reared from hatching. Gurney (1942, page 282) suggested that the atypical zoeal characters of Hymenosomatidae, provide strong evidence against the inclusion of the Hymeno- somatidae in the Oxyrhyncha. Thus, it is now evident that in the Oxyrhyncha the zoeal characters of each family differ from those of the others, indicating that the Oxyrhyncha, like the Oxystomata, are a heterogeneous group. LARVAL DEVELOPMENT OF PARTHENOPE 187 In the adult classification, most revisers of Oxyrhyncha as mentioned in the previous section placed the Parthenopidae under Oxyrhyncha. Flipse (1930) also dealt with the Parthenopidae as Oxyrhyncha in his Siboga Expedition Report. However, Dana (1851, pages 426-427) considered the Parthenopidae as inter- mediate between the Maiinea and the cancroid crustacea because of the structure of the basal article of the antenna and the epistome, and Miers (1879, pages 635-636) stated that the Parthenopidae occupy a position almost intermediate be- tween the rest of the Oxyrhyncha and certain Cancroidea in respect to the struc- ture of the antennae. Miers (1879; page 641) also assumed the nearest affinities of the Parthenopidae in Oxyrhyncha are in the direction of Inachns through Inachoidcs. Cano (1893, page 580) provided a scheme, based on the adult characters, showing that the Parthenopidae diverged from the Inachidae-Majidae line. On the other hand, quite a few workers have not placed the Parthenopidae under Oxyrhyncha. Most of the authors did not give reasons for their classifica- tions. Strahl (1862a and b) separated the Parthenopidae from Oxyrhyncha and grouped it with Calappidae on the formation of the basal article of the antenna. The structure and formation of the rostrum, cephalothorax, antennae and orbits caused Ortmann (1893; pages 412-413) to place the parthenopid group under Cyclometopa. Guinot (1966 and 1967), seeking a "parthenoxystomienne" line, suggested the parthenopid crab (Acthra) and some oxystomous crabs (Osachila. Hcpatns, Hcpatella and Actaeomorpha) be united, because of their adult morpho- logical characters. Clearly, the systematic position of the Parthenopidae, based on adult char- acters, is still in question. The larval characters of the Parthenopinae are completely different from those of Majidae and Hymenosomatidae. Parthenopid larvae are different at least from the Majidae in the number of zoeal stages, the formation and pattern of antenna 2, and the gill formation. The larvae of the subfamily Eumedoninae are completely unkown. However, the larval characters of the Parthenopinae (Par- tJicnope) strongly suggest a relationship to the Brachyrhyncha rather than to the Oxyrhyncha. The author is grateful to Dr. Anthony J. Provenzano, Jr. for his encourage- ment during the course of the present study. Mr. Henry B. Roberts of the United States National Museum confirmed the identification of the female. Dr. L. B. Holthuis and Dr. Raymond B. Manning criticized the manuscript. Mr. Robert Gore kindly helped in final revision of the manuscript. Mrs. C. Edith Marks helped with rearing of specimens. SUMMARY 1. Larvae of an oxyrhynchous crab, Parthenopc (Platylambnis) serrata (H. Milne Edwards), were successfully reared in the laboratory from hatching to the second crab stage on a diet of Artcuiia nauplii. Six zoeal stages, one megalopa and the first crab stage are described and illustrated. 188 WON TACK YANG 2. Two series of larvae were reared at each of two temperatures, 20 and 25 C. The salinity ranged between 34 and 37%c. The mortality of the first zoeal stage reduced the initial population to less than half. The first crab stage was attained in approximately 30 days at 25 C, and in 45 days at 20 C, after hatching. The first zoeal stage and the megalopal stage showed more prolonged mean duration than intermediate stages and the pattern of mean duration was similar in both 20 and 25 C. 3. The major characteristics of four species of parthenopid first zoeae of known parentage are tabulated for comparison. The major distinctive characters of Parthenope larvae are presence of rostral, dorsal and lateral spines on the cara- pace, lunate telson with a dorsally located spine on each fork, normally six zoeal stages, and a smooth "feeler" at the tip of the fifth pereiopod in the megalopa. 4. The number of zoeal stages and the morphological characters are com- pared to those of oxyrhynchous crabs. The larval characters suggest that Parthe- nopinae should be removed from the superfamily Oxyrhyncha and placed in the Brachyrhyncha. LITERATURE CITED AIKAWA, H., 1935. Inachidae ka oyobi kinen-shu no zoea youchu no keitai [The zoeal morphology of the family Inachidae and allied species]. Zool. Mag. (Dobntsugaku Zasshi) Tokyo, 47 : 217-227. AIKAWA, H., 1937. Further notes on brachyuran larvae. Rcc. Occanoqr. Wks. Japan, 9: 87-162. ALCOCK, A. W., 1895. Materials for a carcinological fauna of India. No. 1. The Brachyura Oxyrhyncha. /. Asiat. Soc. Bcng., 64: 157-291. BALSS, H., 1957. Abteilung Brachyura, Borradaile, 1907. Pages 1505-1672 in H. G. Bronns, Ed., Klassen jind Ordmtngcn dcs Ticrreichs 5 (1) 7. BORASCHI, L., 1921. Osservazioni sulle larve dei Crostacei Decapodi, Brachiuri e Anomuri. Memorie R. Comitato Talassografico Italiano, 87 : 1-32. BOSCHI, E. E., M. A. SCELZO AND B. GOLDSTEIN, 1969. Desarrollo larval del cangrejo Halicarcimts planatus (Fabricius) en el laboratorio (Crustacea, Decapoda, Hymeno- somidae), con observaciones sobre la distribucion de la especie. Bull. Mar. Sci., 19: 225-242. BOURDILLON-CASANOVA, L., 1960. Le Meroplancton du Golfe de Marseille: Les larves de Crustaces Decapodes. Rec. Trav. Sfa. Mar. Endoitme Fac. Sci. Marseille, 30(18) : 1-286. CANO, D. G., 1893. Sviluppo e Morfologia degli Oxyrhynchi. Mitt. Zool. Sta. Ncapcl, 10: 527-583. COSTLOW, J. D., JR., 1965. Variability in larval stages of the blue crab, Callincctcs sapidns. Biol. Bull., 128: 58-66. COSTLOW, J. D., JR., AND C. G. BOOKHOUT, 1959. The larval development of Callinectcs sapidns Rathbun reared in the laboratory. Biol. Bull., 116: 373-396. DANA, J. D., 1851. On the classification of the Maioid Crustacea or Oxyrhyncha. Amer. J. Sci., Ser. 2,11: 425-434. FAGETTI, G. E., 1960. Primer estadio larval de cuatro Crustaceos Braquiuros de la Bahia de Valparaiso. Rev. Biol. Mar., 10: 143-154. FLIPSE, H. J., 1930. Die Decapoda Brachyura der Siboga-Expedition. VI. Oxyrhyncha : Parthenopidae. Siboga Exped., Monog., 39 : 1-96. GAMO, S., 1958. On the post larval stages of two species of crabs of the subfamily Varuninae, Grapsidae, Brachyuran Crustacea. Zool. Mag. (Dobutsiigaku Zasshi) Tokyo, 67 : 373-379. GARTH, J. S., 1958. Brachyura of the Pacific coast of America. Oxyrhyncha. Allan Han- cock Pacif. Exped., 21 : 1-854. LARVAL DEVELOPMENT OF PARTHENOPE 189 GOURRET, P., 1884. Considerations sur la faune pelagique du Golfe de Marseille. Ann. Mus. Hist. Natur. Marseille, Zool, 2(2): 1-175. GUINOT, D., 1966. Recherches preliminaires sur les groupements naturels chez les Crustaces Decapodes Brachyoures. I. Les affinites des genres Acthra, Osachila, Hcpatus, Hepatella et Actaeomorpha. Bull. Mas. Hist. Natur., Paris, 38: 744-762. GUINOT, D., 1967. Recherches preliminaires sur les groupements naturels chez les Crustaces Decapodes Brachyoures. I. Les affinites des genres AetJira, Osacliila, Hepatus, Hepatella et Actaemorpha (suite et fin). Bull. Mus. Hist. Natur., Paris, 38: 828-845. GURNEY, R., 1938. Notes on some decapod Crustacea from the Red Sea, VI : Some brachyuran larvae. Proc. Zool. Soc. London, 108 : 73-84. GURNEY, R., 1942. Larvae of Decapoda Crustacea. Ray Society, London, 306 pp. HEEGAARD, P., 1963. Decapod larvae from the Gulf of Napoli hatched in captivity. Vidensk. Medd. Naturhist Foren. Kjobenhavn, 125 : 449-493. HOLTHUIS, L. B., 1959. The Crustacea Decapoda of Suriname (Dutch Guiana). Zool. Verh., 44: 1-296. LEBOUR, M. V., 1928. The larval stages of the Plymouth Brachyura. Proc. Zool. Soc. Lon- don, 1928: 473-560. LEBOUR, M. V., 1944. Larval crabs from Bermuda. Zoologica, 29: 113-128. MIERS, E. J., 1870. On the classification of the maioid Crustacea or Oxyrhyncha, with a synopsis of the families, subfamilies, and genera. /. Linn. Soc. Zool., 14 : 634673. MILNE EDWARDS, H., 1834. Histoire naturelle des Crustaces, comprenant 1'anatomie, la physiologie et la classification de ces animaux. In Librairie Encylopedique de Roret Volume 1, Paris. MIR, R. D., 1961. The external morphology of the first zoeal stages of the crabs, Cancer magister Dana, Cancer antennarius Stimpson and Cancer anthonyi Rathbun. Calif. Fish Game, 47: 103-111. ORTMANN, A., 1893. Die Decapoden-Krebse des Strassburger Museums. VII. Theil Abtheilung: Brachyura (Brachyura genuina Boas) II. Unterabtheilung : Cancroidea, 2. Section: Cancrinea, 1. Gruppe: Cyclometopa. Zool. Jahrb. Abt. S\s. Ockol. Geogr. Tiere 7: 411-495. PEARSON, J., 1908. Cancer. In Liverpool Marine Biology Committee Memoirs XVI. Wil- liams and Norgate, London. POOL, R. L., 1966. A description of laboratory-reared zoeae of Cancer magister, and megalopa taken under natural conditions (Decapoda Brachyura). Crustaceana, 11: 83-97. PROVENZANO, A. J., 1968. The complete larval development of the West Indian hermit crab, Petrochirns diogencs (L) reared in the laboratory. Bull. Mar. Sci., 18: 143-181. RATHBUN, M. J., 1925. The spider crabs of America. Bull. U. S. Nat. Mus., 129: 1-693. RICE, A. L., AND A. J. PROVENZANO, 1966. The larval development of the West Indian sponge crab Drotnidia antillensis (Decapoda: Dromiidae). /. Zool., London, 149: 297-319. ROBERTSON, P. B., 1968. The complete larval development of the sand lobster Scyllarus americanus in the laboratory, with notes on larvae from plankton. Bull. Mar. Sci. 18: 294-342. SAKAI, T., 1938. Studies on the crabs of Japan. III. Brachygnatha, Oxyrhyncha. Yokendo Publishing Co., Tokyo. STRAHL, C, 1862a. Uber cine neue Species von Acanthocyclus Lucas, seine systematische Stellung und Allgemeines iiber das System der Dekapoden. Monatsbcricht Deutschen Akademie Wissenschaften Berlin, (1861), 12: 713-717. STRAHL, C., 1862b. On a species of Ruppcllia, Milne-Edwards, and the limit of the Brachyura. Ann. Natur. Hist., 9 : 299-303. WILLIAMS, A. B., 1965. Marine decapod Crustaceans of the Carolinas. Fish. Bull. Fish. Wildlife Serv. US., 65 : 1-298. YANG, W. T., 1967. A study of zoeal, megalopal, and early crab stages of some oxyrhynchous crabs (Crustacea: Decapoda). Unpublished Doctoral dissertation, University of Miami, 459 pp. YATSUZUKA, K., 1957. Study of brachyuran zoea (artificial rearing and development). Pages 571-590 in Suisan Gaku Shusei (Fisheries Compilation). Tokyo University Press, Tokyo. Vol. 140, No. 2 April,, 1971 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY SEXUAL CYCLES AND MATURITY OF THE TURTLE, CHRYSEMYS PICT A CARL H. ERXST Department of Bioloijy, Southwest Minnesota State College, Marshall, Minnesota 56258 The sexual cycles have been described for male Stcrnotherus odoratns (Risley, 1938), and both sexes of Terrapenc Carolina (Atland, 1951) and Terrapene ortiata (Legler, 1960). Sexual cycles of the painted turtle, Chryseinys picta have been previously described for Nova Scotia females by Powell (1967) and for both sexes of Michigan C. picta by Gibbons (1968). A similar study was conducted on Pennsylvania C. picta which revealed certain differences in the spermatogenetic and oogenetic cycles and the attainment of sexual maturity as reported by Powell (1967) and Gibbons (1968). The results of this study are presented here. METHODS This study was conducted at the \Yhite Oak Bird Sanctuary. 3 miles north of Manheim, Lancaster County, Pennsylvania. Turtles were captured by hand, with a dip net, or in conventional hoop-net traps. Routine measurements of each turtle included the maximum plastron length, the straight line carapace length, the total shell width at the bridge, and the total shell height at the bridge. The third- claw length of both the fore and hind foot, the preanal, and the postanal tail lengths were also measured to determine the characters of sexual dimorphism (Table I). Eighty adult specimens of C. picta (50 females, 30 males) were dissected and the reproductive tracts were removed to determine the male and female annual sexual cycles. Fresh ovaries were examined from 30 females having plastron lengths ranging from 106.3 to 136.8 mm, dissected during the periods from 5 August to 31 October, 1966, and from 11 March to 30 July, 1967. Previous examinations (August, 1965) of reproductive tracts of 20 other females had indicated that specimens of 110 mm plastron length were sexually mature. All ovaries were weighed before preservation. Follicles, corpora Intea, and corpora albicantia w y ere counted and measured after preservation. Follicles less than one millimeter in diameter were not counted. Follicles and oviducal eggs were recorded separate for the right and 191 Copyright 1971, by the Marine Biological Laboratory Library of Congress Card No. A38-518 192 CARL H. KKXST left sides. Corpora In tea and albicantia were studied under a binocular dissecting microscope. No histological studies were made of the female reproductive tract. Fresh testes were examined from 30 males having plastron lengths ranging from 87.8 to 112.9 mm, dissected during the periods from 23 August to 26 Sep- tember, 1966 and 5 March to 13 August, 1967. All testes were weighed, and the greater diameter was measured before preservation. The testes were fixed in 10 per cent neutral formalin and after two weeks transferred to 80 per cent ethyl alcohol for storage. Later the testes were embedded in paraffin, sectioned, and stained with hematoxylin and eosin. All measurements in the field and laboratory were made to the nearest tenth of a millimeter with dial calipers. A triple-beam balance was used to determine all weights. TABLE I .1 summary of sexual dimorphism in Chrysemys picta Character Range Male mean S.D. Range Female mean S.D. Plastron length 70.9-119.2 99.6 10.1 80.8-142.1 120.9 14.2 Shell width 67.0-99.5 81.0 6.6 78.5-114.3 94.4 8.6 Shell height 28.7-49.7 38.2 3.5 31.9-59.5 48.5 6.3 Third forefoot claw 5.1-13.0 10.8 1.4 3.7-9.2 6.6 0.7 Forefoot claw/hind foot claw 2.1-3.2 2.7 0.4 0.9-1.6 1.3 0.3 Preanal tail length 5.7-20.0 13.1 3.2 0.5-14.8 6.8 2.6 Postanal tail length 16.0-33.7 27.3 2.9 21.7-41.0 31.0 4.5 Total tail length 28.9-51.8 40.1 3.6 26.2-45.7 36.9 3.3 Tail length/plastron length 0.38-0.48 0.42 0.03 0.25-0.31 0.28 0.02 RESULTS I'cinalc sc.rnal cycle and maturity Ovaries weighed the most during April and May (Fig. 1). This corresponds to the mating season and many large ovocytes were found on the ovaries at this time. The ovaries weighed the least during July, August, and September, or after ovulation and nesting has occurred. No correlation was found between plas- tron length and ovarian weight, as smaller females often had heavier ovaries on the same date as larger females. For example, a female with a plastron length of 107.6 mm sacrificed on 23 August, 1966, had an ovarian weight of 12.6 grams while another female, 136.8 mm in length, sacrificed the same day, had an ovarian weight of only 9.4 grams. Enlarged ovocytes (15 to 25 mm) were present throughout the study periods, but were most numerous during April and May and least numerous during July and August. Cagle (1954) reported that female specimens of C. picta collected during May, June, and early July and containing oviducal eggs usually had large ovocytes still present. He assumed that the ovo- cytes represented second and third broods that would be deposited that same season. Although careful searches were made from April through September, no evidence was found at White Oak to support Cagle. Fresh nests and nesting CHRYSEMYS PICTA SEXUAL CYCLES 193 females were only found during June. Oviducal eggs were only found during late May, June, and early July. The oviducal eggs examined were pinkish white and slightly translucent, hut became white shortly after their removal. They ranged in length from 23.1 to 33.0 mm (mean 28.4) in diameter from 15.4 to 20.1 mm (mean 17.5) and in weight from 5.3 to 6.8 g (mean 5.9). 1 40 35 30 25 20 $ l> 15 10 5 I March April May June July August Sept October FIGURE 1. Seasonal fluctuations in ovarian weight in thirty specimens of Chryscmys pictu. The follicles found on the ovaries could be grouped by diameter as : large (greater than 15 mm), medium (11 to 15 mm), and small (6 to 10 mm) after the system proposed by Legler (I960). Those smaller than 6 mm were not consid- ered. Mature follicles had diameters of 20 to 25 mm. Both Powell (1967) and Gibbons (1968) reported 18 mm as the maximum size of the follicles they mea- sured. (Cagle (1944) considered female specimens of Chryscinys scripta to be mature if they contained follicles of 15 mm or greater. All female specimens of 194 f.AKI. II. KKNST C. picta witli the largest follicle 10 nun or more in diameter had two or more addi- tional groups of follicles present on the ovaries (6 to 9 mm in diameter and less than 5 mm). Apparently, these groups represent follicles which would mature and he ovulated during successive mating seasons. None were found to reach mature size before October, and therefore, second nestings as proposed by Cagle (1954), Powell (1967), and Gibbons (1968) would not occur in southeastern Pennsylvania. The ovarian cycle begins in July and August after ovulation and nesting have occurred. Many small follicles form on the germinal ridges of the ovaries. The ovocytes within the follicles increase in size and by late September and October there is little difference in size between these and the June size of the ovocytes that were not ovulated (about 15 mm). Those not ovulated in June grow and reach TABLE II Ovarian activity in ten specimens of Chrysemys picta Left ovary Right ovary Corpora lulea Follicles (diameter 15+ mm) Corpora lutea Follicles (diameter 15+ mm) 4 2 4 3 1 1 3 2 5 4 ( 2 2 3 1 3 1 1 2 2 3 4 2 5 1 3 1 2 5 4 2 I 3 4 2 2 3 mature size by early October (about 20 mm). This differs from Gibbons' (1968) and Powell's (1967) findings of little change in size during the summer. There is little change in size during the winter. Atland (1951) suggested that Terrapene Carolina follicles grow to nearly mature size in the season preceding ovulation and remain quiescent over winter. This appears also to be the case in C. picta. At- land also thought that some of the enlarged follicles were absorbed during hiberna- tion. The follicles of White Oak C. picta were of mature size in March and remained that way until the ovulation period (May and June). This differs from what Gibbons (1968) found in his Michigan C. picta. He reported that speci- mens taken in March contained yolked follicles 15 to 16 mm in diameter. These remained this size until May when they grew to mature size (about 18 mm). Ovarian activity alternates in C. picta. Counts of corpora lutea showed one ovary more active than the other in a given season, and higher counts of enlarged follicles (15 mm in diameter or larger) indicated that the opposite ovary would be more active in the following season (Table II). Legler (1960) reported this same condition in Terrapene ornata. The corpora lutea are cuplike in shape and 7.5 to 9.0 mm in diameter. The number of oviducal eggs always equalled the CHRYSEMYS PICT A SEXUAL CYCLES 195 number of corpora Iittca. Involution of the corpora littca occurs very rapidly and after a month they are barely visible on the ovarian surface and can be referred to as corpora albicantia. The corpora alhicantia disappear when a new clutch of eggs is ovulated. Extrauterine migration of ova has been reported in ('. picta, C. scripta, T. ornata, and Emydoidca blandhn/i, by Legler (1958) and in Sternotherus odoratus by Tinkle (1959). An examination of the female painted turtles from which oviducal eggs were removed revealed no evidence of ovular migration. All had equal numbers of corpora Iiitea and oviducal eggs on the same side. Legler (1960) thought such migration may serve to redistribute eggs to the oviducts when the ovaries are functioning at unequal rates. I.I 1.0 -X - 9 I - 8 "^ 0.7 2 0.6 1 - 5 -Q> "o 0.4 ."!> 0.3 ^ 0.2 O.I March April May June July August September FIGURE 2. Seasonal fluctuations in testes weight in thirty specimens of Chryscmys picta. Females were considered sexually mature if they : 1 ) contained follicles with diameters greater than 15 mm; 2) contained oviducal eggs; 3) were found mating; or 4) were found nesting. Only 7 females of 16 examined with plastron lengths below 100 mm contained follicles of 15 mm. The smallest had a plastron of 80.8 mm. Four had plastron lengths between 90 and 99 mm. All 7 showed 5 years growth annuli. All females examined over 100 mm plastron length contained some follicles of 15 mm. All of the females found containing oviducal eggs, mat- ing, or nesting, were over 110 mm in plastron length and at least 5 years old. White Oak females were mature at 110 mm plastron length after their fifth year. Possibly some mature at a smaller size (between 100 and 110 mm ). Cagle (1954) reported 20 Illinois females containing eggs were 126 to 160 m min plastron length, 14 Tennessee females 106 to 138 mm, and a single Michigan female was 152 mm. He assumed that a female having ovarian follicles 10 mm or greater in diameter was capable of depositing eggs; and found 41 Illinois females, plastron lengths 196 CARL H. ERNST 122 to 162 nun, and a single Louisiana female, length 125 mm, with follicles of this si/.e. 1 le stated that females became sexually mature when they reach a plastron length of 120 to 130 mm. Gibbons (1968) reported that sexual maturity in south- u csteri! Michigan was attained when females reached a plastron length of about 110 to 120 mm (at about 10 years of age). Male sc.nial cycle and maturity Testes weighed the most during March following emergence from hibernation (Fig. 2). The testes contained much sperm at this time and additional weight was possibly due to the proliferation of the sustentacular cytoplasm. Weight de- creased steadily during April and May as the sperm passed out of the testes, and into the epididymides. Weight increased in June as sperm maturation again began. The greatest diameter varied little, ranging from 6 to 8 mm during each month. Gibbons (1968) reported changes in the size of the testes. He reported the testes to be reduced in size from March to June, enlarged from July to September and to be small again in October. The changes in weight agree with those reported by Gibbons (1968) for Michigan C. [>icta. Spermatogenesis began in March when a few spermatogonia first appeared. At this time there were many Sertoli cells present and some cellular detritus was contained in the lumen of the seminiferous tubules. Sperm produced in the pre- vious cycle were also present in the lumen. As March progressed the clear cyto- plasm of the cells extended into the lumen of the turtle. During April there was much detritus in the lumen, and the sperm started to pass out of the tubules and into the epididymides. The cells bordering the tubule were no longer clear or extending into it. There were about an equal number of Sertoli cells as primary spermatocytes. As the month passed, a second layer of spermatocytes was formed. In May there were two or three rows of spermatocytes, both primary and secondary, surrounding the Sertoli cells which were decreasing in number. Sper- inatids were now present in small numbers. Detritus and some sperm still re- mained in the lumen. A few new sperm appeared around the border of the lumen in early June. Many of the lumens were now clear of detritus. Spermatids increased in numbers. Spermatocytes were present, but the Sertoli cells had practically disappeared. In late June many newly formed clumps of sperm could be seen around the lumen border. July and August were the most active periods of sperm production. By late July large numbers of sperm were found clumped in the lumens and by the end of August filled most of the lumen and lined the borders. During September the sperm moved to the center of the lumen so that few remained at the border at the end of the month. The cycle appeared complete by this time. Gibbons (1968) reported that the sperm passed into the epididymides during October. Figure 3 shows stages of the cycle. Risley (1938) found the testes of Stcrnotherus odoratus to be largest in August and smallest in early May. Recession of testes in spring coincided with the mating period and later increases in size with increasing spermatogenic activity CHRYSEMYS PICTA SEXUAL CYCLES %** &n* >" *visv. r? ; ^w rM,?^ ' s ' JW febrWrtljS^ |ii IP^I -; I /- T %, -^ '.*iP* v* ' .,'.'* /^ -*_-- FIGURE 3. Representative stages in the spermatogenic cycle of Chrysemys picta. Letters a to f, respectively, are sections of testes obtained on 14 March, 4 April, 11 May, 12 June, 12 July, and 23 August. and enlargement of the seminiferous tubules. Atland (1951) reported that the cycle of Terrapene Carolina was essentially like that of S\ odoratus. Legler (1960) showed the cycle of Terrapene ornata began in early May and ended in late October and except for a longer length was the same as that of S. odoratus and T. Carolina. The spermatogenic cycle of C. picta beginning in March and ending in September 198 CARL H. ERNST is even longer, but in main points does not differ from those species previously mentioned. Male specimens of C. picta were considered sexually mature if they contained mature sperm in their testes or epididymides. The smallest turtle containing mature sperm had a plastron length of 87.8 mm and showed 5 growth annuli. This turtle was sacrificed on 19 September, 1966, and possibly had mated the previous spring. Mature males have elongated forefoot claws (over twice as long as those of the hind feet), and elongated preanal tail lengths (Table I). Males first showed these characters at 70 mm plastron length in their fourth year. Apparently, males mature during their fourth year in Pennsylvania, but do not mate until the spring of their fifth year. Gibbons (1968) reported that male C. picta in southwestern Michigan mature when they reach 80 mm plastron length and showed some to be mature in their fourth year. Cagle (1954) reported male C. picta reached ma- turity at 90 mm in Michigan, 70 mm in Illinois, and reported Louisiana males sexually mature at 55 and 62 mm, but other males in the size range 50-60 mm were sexually inactive. Cagle (1954) stated that males of the southern popula- tions apparently may become sexually mature during the first year of life (in one complete growing season), and the northern males require at least two and pos- sibly three seasons to attain maturity. Cagle (1948) reported male Chrysemys scripta normally become mature at plastron lengths of 90 to 100 mm but occa- sionally individuals may become mature at a smaller size. Chrysemys scripta in their first mature season may be 2 to 5 years old. The attainment of maturity in the genus Chrysemys is apparently a factor of size rather than age. DISCUSSION The variations in the sexual cvcles and attainment of maturity between the specimens of Chrysemys picta in this study and those studied by Cagle (1954). Powell (1967), and Gibbons (1968) are of interest. Climatic conditions greatlv influence the physiological activity of turtles. Cold temperatures reduce the metabolic rate of Chrysemys picta (Rapatz and Musacchia, 1957) and probably also reduce the activity of the reproductive organs. White Oak painted turtles were active in all months except February (Ernst, 1969). and although Gibbons' (1968) females had similar ovulatory periods to those from White Oak, the annual activity period in southwestern Michigan was shorter (Gibbons, 1967). Sexton (1959) also reported a shorter annual activity period during his study of C. picta in southern Michigan. Although no figures are avail- able on the length of the annual activity period in Nova Scotia, it is probably also shorter than that in southeastern Pennsylvania. This is caused by the earlier onset of colder water temperatures and the delayed thawing of the ice cover in spring. According to Goode (1953) both southern Michigan and Nova Scotia have a continental forest climate with cool summers while southeastern Pennsylvania has a continental forest climate with warm summers. Nova Scotia has a surface tem- perature below 32 F in winter and from 50 to 68 F in summer. P)Oth southern Michigan and southeastern Pennsylvania have surface temperatures below 32 F in winter and above 68 F in summer. During January the normal temperatures of both southern Michigan and Nova Scotia range between 20 and 30 F, while CIIRYSEMYS PICT A SEXUAL CYCLES 199 those of southeastern Pennsylvania range between 30 and 40 F. The normal July temperatures of southern Michigan and southeastern Pennsylvania range between 70 and 90 F while those of Nova Scotia only range between 50 and 70 F. A shorter period of development caused by prolonged colder water temperatures could explain why the ovocytes of females from the more northern populations do not grow as large as do those from White Oak females. It would be interesting to compare data from Louisiana females. This could also explain the earlier maturity of both sexes at White Oak. Cagle (1954) has reported an inverse relationship between the attainment of sexual maturity and latitude in Chrysemys picta, as shown in the present study. Tinkle (1961) reported that in Sternothcrus odoratus both southern males and females reach sexual maturity at a smaller size than do the northern sexes. Inhibition by cooler temperatures may explain the lack of growth in ovocytes in Nova Scotia during the summer, but it is not clear why such a difference should exist between the Michigan and Pennsylvania populations which both have warm summers. Painted turtles from more southern localities are known to have longer repro- ductive periods. Cagle (1954) defined the nesting season of Chrysemys picta as that period in which females may contain oviducal eggs and reported a nesting season of early April to the last of July in Louisiana and from 12 May to 22 July in Illinois. Presumably this difference is caused by the warmer climate in Louisiana. Multiple nestings as supposed by Cagle (1954), Powell (1967), and Gibbons (1968) have never been proven. Since no female Chrysemys picta has been ob- served nesting more than once in a given season, this difference may not exist. If multiple nestings do occur they would be restricted to southern populations because of the short northern egg-laying season. The difference in the sexual cycles and attainment of maturity shown in this study point out that too often the results of a study of one population of a species are fallaciously taken for granted as being true for all such populations throughout the species' range. Gibbons and Tinkle (1969) have pointed out similar varia- tions in the clutch size of C. picta populations in a single geographic area. There is a critical need for more information on the factors affecting reproduction in turtle populations. I wish to thank J. Robert Heckman for preparing the histological sections of Chrysemys testes, Geoffrey Kampe for taking the photomicrographs, and Dr. Roger W. Barbour for his advice and criticisms. This paper is a portion of a dissertation presented to the faculty of the Depart- ment of Zoology, University of Kentucky, in partial fulfillment of the requirements for the degree of Doctor of Philosophy. SUMMARY 1. A mature male Chrysemys picta can be distinguished from a female by its long forefoot claws, thickened base of the tail, and short post-anal tail length. 2. Males become sexually mature when they reach a plastron length of 80 to 90 mm. They usually mature in their fourth year, but probably do not mate until the spring of the fifth year. 200 CARL H. ERNST 3. All sexually mature females contain follicles with diameters greater than 15 mm. All females containing oviducal eggs found mating or nesting were over 110 mm in plastron length and at least five years old. 4. Sexual maturity is apparently a factor of size rather than age. 5. Spermatogenesis begins in March with the most active period of sperm pro- duction occurring in July and August. The cycle is completed in September when the sperm start to pass into the epididymis where they are stored during the winter. 6. The ovarian cycle in southeastern Pennsylvania begins in July and August after ovulation and nesting, and is completed by the end of October. No change occurs during the winter. This differs slightly from the cycle of more northern females. 7. The ovarian activity alternates, and extrauterine migration of ova is known to occur. LITERATURE CITED ATLAND, P. D., 1951. Observations on the structure of the reproductive organs of the box turtle. J. Morplwl, 89: 599-621. CAGLE, F. R., 1944. Sexual maturity in the female of the turtle Pscudetnys scripta clcgans. Copeia, 1944: 149-152. CAGLE, F. R., 1948. Sexual maturity in the male turtle, Pscudom\s scripta troostii. Copeia, 1948: 108-111. CAGLE, F. R., 1954. Observations on the life cycles of painted turtles (genus Chrysemys). Amer. Midland Nattir., 52: 225-235. ERNST, C. H., 1969. Natural history and ecology of the painted turtle, Chryscinys picta (Schneider). Ph.D. dissertation, University of Kentucky, 207 pp. GIBBONS, J. W., 1967. Population dynamics and ecology of the painted turtle, Chrysemys picta. PhD. dissertation, MicJiigan State University, 111 pp. GIBBONS, J. W., 1968. Reproductive potential, activity, and cycles in the painted turtle, Chryse- mys picta. Ecology, 49: 399-409. GIBBONS, J. W., AND D. W. TINKLE, 1969. Reproductive variation between turtle populations in a single geographic area. Ecologv, 50: 340-341. GOODE, J. P., 1953. Goode's World Atlas. Rand McNally and Co., New York, 272 pp. LEGLER, J. M., 1958. Extra-uterine migration of ova in turtles. Hcrpetologica, 14: 49-52. LEGLER, J. M., 1960. Natural history of the ornate box turtle, Terrapene ornata ornata Agassiz. Univ. Kans. Pnbl. Mits. Natur. Hist., 11: 527-669. POWELL, C. B., 1967. Female sexual cycles of Chrysemys picta and Clemmys insculpta in Nova Scotia. Can. Field Nahtr., 81 : 134-140. RAPATZ, G. L., AND X. J. MUSACCHIA, 1957. Metabolism of Chrysemys picta during fasting and during cold torpor. Amcr. J. Physio!., 188: 456-460. RISLEY, P. L., 1938. Seasonal changes in the testes of the musk turtle, Sternotherus odoratns. J.Morphol.,63: 301-317. SEXTON, O. J., 1959. Spatial and temporal movements of a population of the painted turtle, Chrysemys picta marginata (Agassiz). Ecol. Monogr., 29: 113-140. TINKLE, D. W., 1959. Additional remarks on extra-uterine migration of ova in turtles. Her- petologica, 15: 161-162. TINKLE, D. W., 1961. Geographic variation in reproduction, size, sex ratio and maturity of Sternothaerus odoratus (Testudinata: Chelydridae). Ecology, 42: 68-76. Reference: Biol. Bull.. 140: 201-214. (April, 1971) FEMALE SEXUAL BEHAVIOR AS THE MECHANISM RENDERING AEDES AEGYPT1 REFRACTORY TO INSEMINATION 1 ROBERT W. GWADZ,- GEORGE B. CRAIG, JR., AND WILLIAM A. HICKEY > Vector Biology Laboratory, Department of Biolot/y. University of Notre Dame, Notre Dame, Illinois 46556 The yellow fever mosquito, Acdcs acijyf>ti (L.), is one of the most thoroughly studied insects from the standpoint of both basic and applied science. Because of its importance as a vector of human diseases, a great deal of literature has accumu- lated concerning aspects of its systematics, distribution, bionomics, disease relation- ships and control. In spite of this intensive study, little was known until recently about its mating behavior. Recent research on sexual behavior of mosquitoes has indicated that many widely accepted concepts were incorrect. Moreover, several points of controversy remain. One of the first serious investigations of sexual behavior in A. aegypti before and during coitus was undertaken by Roth (1948). In this paper he firmly estab- lished that flight sound of the female mosquito attracted males and initiated copula- tory behavior. Later. Spielman (1964) and Jones and Wheeler (1965) described the act of copulation in detail and attempted to explain the mechanism by which semen is transferred by the male to the female. j Early students of the behavior of A. aegypti observed that females begin to copulate soon after emergence and continue to do so throughout life. Most as- sumed that the terms "mating," "copulation" and "insemination" were synonymous, thus concluding that most copulations result in insemination. MacGregor (1915) stated that copulation takes place as soon as females are ready to fly. Roth (1948) noted that most females are mated within 105 to 145 minutes after emergence, and recorded a single female that underwent 50 successful copulations by 1 1 differ- ent males during a one hour period. Finally, Gillett (1955) observed that, in captivity, a single female of A. aegypti may mate as many as 40 times before her first blood meal. There has been considerable confusion and improper usage of terms affecting mosquito reproduction. The following is a partial list of terms related to behavior and reproduction as used herein : Mat in i/ - That portion of sexual reproduction which begins with the search for a sexual partner and ends with successful insemination. Copulation or coitus repre- sents only one portion of mating. Mating behavior includes swarming and other mate-finding mechanisms. Copulation - The act of sexual intercourse. Coitus, coition, sexual union. In mosquitoes, copulation is a rather mechanical process. It requires the bodies of 1 This investigation was supported in part by National Institutes of Health Research Grant No. AI-02753 and by National Institutes of Health Fellowship 1 F01 GM44588-01 from the National Institute of General Medical Sciences. 2 Present address : Department of Tropical Public Health, Harvard School of Public Health, 665 Huntington Avenue, Boston, Massachusetts. 3 Present address : Department of Biology, St. Mary's College, Notre Dame, Indiana 46556. 201 202 R. \V. (, \VADZ, G. B. CKAK';. JR. AM) \V. A. RICKEY both partners to he in the correct position and their genitalia to be at the proper degree of connection so th.it ejaculation and insemination can take place. Coupling - The events occurring when a sexnallv aroused male seizes a female and establishes genital contact. The male remains in contact and tries to achieve a position of coitus that will permit insemination. He may strive for the copula- tory position but not achieve it. A male encountering a sexually refractory female may couple, but cannot copulate. Insemination - The steps that occur between male ejaculation and deposition of sperm in the spermathecae. Sperm and seminal fluid are introduced into the bursa copulatrix of the female and, within a few minutes, sperm begin to migrate to the spermathecae where they are stored until the time of fertilization. A major revision in our understanding of the sexual behavior of A. acgypti occurred when it was demonstrated that females of this species are monogamous (Craig, 1967; Spielman, Leahy and Skaff, 1967). Copulation by a mature virgin female results in insemination. Although the female appears to copulate with other males, she is refractory to a second insemination for life. Subsequent in- semination is prevented through the action of a substance from the male accessory glands that is transferred to the female in the seminal fluid. This substance has been designated "matrone" (Fuchs, Craig and Hiss, 1968). Virgin females can be rendered refractory to insemination by implants of accessory glands from adult males (Craig, 1967; Spielman ct a!.. 1967), by injec- tion of an aqueous extract of glands (Craig, 1967), or by injection of material extracted from lyophilized whole bodies of males (Fuchs ct <;/., 1968). The par- tially purified substance, matrone, has been shown to be a protein ( Fuchs, Craig and Despommier, 1969). A second major revision of the life history of A. acgypti came with the descrip- tion of a post-emergence refractory period (Gwadz and Craig, 1968; Lea, 1968). During this refractory period, which may last 36 to 60 hours, females may couple repeatedly, but they are not inseminated. Lea ( 1968 ) showed that the onset of sexual receptivity could be prevented by removal of the corpora allata and restored by gland implants. He concluded that sexual receptivity was influenced by juvenile hormone, a product of the corpora allata. Later, Gwadz, Lounibos and Craig (1971) demonstrated that the post-emergence refractory period could be reduced by more than 70% by the topical application of a synthetic analogue of juvenile hormone to pupae or teneral adults. The mechanisms which permit a female of A. acgypti to couple without being inseminated have become a matter of controversy. Both young virgin females and previously inseminated females appear to copulate repeatedly but they are not inseminated. Craig (1967) felt that matrone might stimulate the inseminated female to hold her vaginal lips closed, thereby preventing the introduction of semen. Gwadz and Craig (1968) hypothesized that young virgin females are not insemi- nated because they withdraw their terminalia and prevent males from completing firm genital union. On the other hand, Spielman ct al. (1967, 1969) concluded that the explana- tions offered by Gwadz and Craig served only to describe some of the behavioral characteristics of females refractory to insemination. Spielman and co-workers MATING BEHAVIOR IN AEDES AEGYPTI 203 felt that these behavioral traits alone were not major barriers to insemination. It was their hypothesis that the failure of effective insemination was due in part to the physical condition of the bursa copulatrix of the female at the time of copula- tion. They attempted to show that young or previously inseminated females are unable to alter the consistency of the semen as it is introduced into the bursa. They felt that the seminal mass is expelled by the female or withdrawn by the male at the termination of coitus. They further contended that sperm and seminal fluid is transferred to females of all ages and sexual experience ; however, a female is capable of retaining semen from only one copulation during her lifetime. Observation on insemination capacity of males are relevant to this controversy. A male can inseminate 5 to 7 females but has little or no capability for sperm replenishment once he is depleted ( Gwadz and Craig, 1970 ; Hausertnan and Xijhout. in press). If, as Spielman contends, copulations with refractory females involve semen transfer and subsequent semen loss, this loss should be reflected as a reduction in the capacity of a male for further insemination. The experiment of Powell (Craig, 1967) shows that male capacity for insemination is not reduced after exposure to previously inseminated females. The present study was under- taken to amplify the preliminary results of Powell. Moreover, an attempt was made to analyze the nature of post-emergence and post-insemination refractory responses. MATERIALS AND METHODS All experiments were conducted with the ROCK strain of A. acyypti. ROCK is a relatively large, uniform, vigorous strain used in a number of laboratories and may be considered as characteristic of the type form of A. aeyypti aCf/yfyfi. This strain was originally obtained from D. W. Jenkins in 1959 and is the most com- monly used strain at the Vector Biology Laboratory of the University of Notre Dame. All rearing was conducted in accordance with the methods of Craig and Vande- Hey (1962). Larvae were reared in enamel pans in 2500 ml of tap water at a density of approximately 200 larvae per pan. Larvae were fed a suspension of liver powder (Nutritional Biochemicals Corporation, Cleveland, Ohio) at a con- centration of 10 g powder per 1000 ml of tap water. The feeding schedule con- sisted of 10 ml of suspension on the day of egg hatching (day 0), 10 ml on day 1, and 20 ml each on days, 2, 3 and 4. Pupation began late on day 4 and was completed by day 5. Care was taken to insure uniform growth and pupation, and pans were dis- carded if the larvae they contained failed to achieve greater than 95 r /c pupation by the middle of day 5. Sexes were separated by size and transferred to petri dishes ( 15 cm diameter) lined with moist paper towelling. These dishes were placed in gallon cardboard ice cream containers covered with bolting cloth. Adult emergence took place in these containers. All rearing and experimentation was performed at 27 1 C and 80 5% relative humidity. Day length was 16 hours. In experiments where the age of the female was of importance, emerging females were collected from the cages at hourly intervals. All females used in -204 R. W. GWADZ, G. B. CRAIG, JR. AND W. A. HICKEY these age-related studies were of known age 30 minutes. The ages of females used in all other studies were known with 12 hours. Newly emerged adults were transferred to gallon containers and maintained on apple slices as a source of food. In experiments where previously inseminated females were required, 2-3 day old females were placed with a surplus of males for two days. This treatment resulted in approximately 95% insemination of females. Determination of sc.vual receptivity Females of A. ac(>(>). Virgin females exposed to males of either strain showed a high count, indicating insemination. On the other hand, refrac- tory females gave counts very close to the average background count. In fact, the refractory females resembled the control, virgin females with no exposure to males. These females were refractory because they had previously been inseminated, and hence rendered monogamous by matrone. Clearly, the labelled males did not pass any seminal material to the refractory females. TABLE IV Radioactivity transferred to unlabelled ROCK females exposed for a 3-day period to virgin labelled males from, two different strains Isotope labelling of cf cf * Mean counts/minute/RocK 9 Material Cone, in MC/ml 9 virgin, no cf exposure (N = 10) 9 exposed to cf cf (N = 15) 9 virgin, mature 9 refractory** cf ROCK cf DISTORTER cf ROCK cf DISTORTER Isoleucine-C 14 Yeast Protein-C 14 0.05 0.10 24 25 62 53 83 52 28 28 26 28 * See Table I. ** 9 9 previously exposed to The results of the experiment with radioactive males (Table IV) fully confirm the earlier experiments (Tables II and III) which showed that prior exposure to refractory females did not diminish a male's capacity for insemination. Observations of refractory and receptive behavior Mating behavior of single females in suspended flight was observed with the aid of a stereoscopic microscope. Couplings resulting in insemination were mark- edly different from couplings by refractory females that did not result in insemina- tion. Virgin females, 4 to 6 days old, were exposed to suspended flight to free- flying males; of 41 females tested, all but one were inseminated after a single coupling (Table V). Observations of these matings indicated that each coupling involved firm genital union and corresponded to the descriptions of copulation by Spielman (1964) and Jones and Wheeler (1965). Figure 1 pictures a male and female of A. aeyypti engaged in copulation which resulted in insemination. The degree of genital contact of this successful copulation is illustrated in Figure 2. The mean duration of genital contact for the 40 successful inseminations was 12.04 seconds (Table V). This average copulation time for receptive virgin females in suspended flight closely corresponds to the 12.1 second copulation time for receptive females in free flight in a lantern chimney. It would appear that suspension of the female had no discernable effect either on the duration of coupling MATING BEHAVIOR IN AEDES AEGYPTI 209 or on the probability of insemination. Moreover, insemination of a receptive virgin female was normally achieved by the first male able to grasp her legs and bring his claspers into position for copulation. Indeed, if the terminalia of the male succeeded in touching the terminalia of the female, complete copulation and insemination was assured. "fc\ * FIGURE 1. Copulation by a sexually mature female of Acdcs acgypti. The arrow indi- cates the cerci of the female extending from beneath the claspers of the male. This copulation resulted in insemination of the female. The length of the male abdomen is approximately 3 mm. FIGURE 2. Illustration of the degree of genital contact characteristic of copulation by a sexually receptive virgin female. The male claspers (shaded) fully engage the female cerci (dark). FIGURE 3. Coupling without copulation by a sexually refractory female of Acdcs acgypti. The arrow indicates the cerci of the female free and well above the claspers of the male. This coupling did not result in sperm transfer or insemination. FIGURE 4. Illustration of the superficial genital contact typical of coupling by young refractory virgin females or previously inseminated females. The male claspers (shaded) do not touch the female cerci (dark). Copulation and insemination are impossible in this position. 210 R. W. GWADZ, G. B. CRAIG, JR. AND W. A. HICKEY Refractory females were exposed in suspended flight in the same manner as was employed with receptive females (Table V). Each refractory female was permitted a minimum of five couplings. None of 35 young tested actually copu- lated or were inseminated. Moreover, none of 25 matrone-injected females were inseminated. Obvious differences were noted in the behavior of refractory and receptive females. The duration of genital contact for couplings involving refrac- tory females was extremely variable, with a range from 1 second to over 7 minutes ; moreover, these copulations were never complete. Coupling by refractory females did not result in firm genital union (Figs. 3 and 4). The aedeagus of the male was so positioned that sperm transfer and insemination would be difficult if not impossible. Visual determination of the state of receptivity of individual females is not difficult. An observer is able to differentiate between receptive and refractory TABLE V Receptivity to insemination of various types of females of Aedes aegypti in suspended flight . Individual females were suspended by the mesonotmu and induced to fl\< in the presence of males; each coupling was observed and timed. No. 9 9 Female type No. couplings allowed per 9 Duration of genital contact in seconds Tested Inseminated Mature virgin (4-6 days old) 1 41 40* 12.04 3.30** Matrone-injected 4-5 25 <6*** Saline-injected 1 13 13 10.41 2.51** Young virgin (22-24 hrs.) 4-5 35 <6 *** * One unsuccessful coupling aborted at 6 seconds. ** Mean S.D. *** Most couplings lasted less than 6 seconds ; however, a few lasted over 60 seconds. Genital union was incomplete in all cases and no sperm was transferred. females by noting the coupling activities of males. Males respond to sexually recep- tive females by ready contact, firm genital union and copulation. Males coupling with refractory females may be able to grasp the terminal sternite of the female but fail to copulate ; they never achieve insertion of the aedeagus. These males may persist in their attempts to attain the proper coital position, but are never able to clasp the female cerci. The differences between the two types of coupling are so marked that females of unknown state can be determined as receptive or refractory on the basis of a single attempt if viewed with a stereoscope. However, to the unaided eye, both types of couplings appear to be normal copulations. Speed of onset of refractory behavior In order to determine the speed with which females change from receptive to refractory coupling responses, individual females were observed under conditions of suspended flight. Each female was permitted two couplings with free-flying males. Each coupling was observed and timed, and the interval between the first and the second coupling was noted. Eighteen mature virgin females were tested. MATING BEHAVIOR IX AEDES AEGYPT1 211 The first coupling of each female was characterized by a firmness of genital union typical of a successful copulation. None of the second couplings were successful. The behavior of females during the second coupling was comparable to the behavior of previously inseminated or matrone-injected females. Moreover, these females were refractory to repeated attempts by several males. The interval between the first and second couplings varied from 7 to 261 seconds with a mean of 93 seconds. This interval was not a function of female receptivity but was related to the speed with which the second male was attracted to the suspended female. On the basis of these observations it would appear that the switch from recep- tive to refractory behavior is an almost instantaneous result of insemination. The first coupling invariably resulted in a firm copulation. The second coupling was always superficial and typically refractory. Moreover, the switch is not the result of the act of copulation itself. Virgin females will copulate repeatedly with semen- depleted males. Obvously, no semen transfer can occur, yet these copulations always show firm genital union, however, these females remain sexually receptive and will immediately copulate with a fresh male. DISCUSSION Females of A. acgypti may couple repeatedly but they copulate only once, i.e., semen is transferred only once. Consequently, they are monogamous. A female utilizes gametes from the insemination of a single male throughout her reproductive life. Only mature virgin females will copulate and can be inseminated. Both young virgin females and previously inseminated females are refractory to insemi- nation ; they may couple but do not copulate. Behavioral responses of the refrac- tory female prevent firm genital union. Spielman et al. (1967, 1969) felt that insemination of refractory females could not be prevented exclusively by behavioral factors. They felt that the condition of the bursa copulatrix of the female determined whether or not a female could be effectively inseminated. They froze females during and after forced-copulation and concluded that semen was transferred to females regardless of age or previous mating status. Only mature virgin females would retain the semen ; refractory females ejected it from the bursa. Forcemated refractory females usually had internal masses of semen in the region of the genital atrium. Unquestionably, refractory females can be inseminated by forced copulation (Craig, 1967; Akey and Jones, 1968; Gwadz and Craig, 1968). However the forced-copulation technique overrides the behavioral responses which prevent in- semination under conditions of free flight. Therefore, data obtained using the forced-copulation technique are not relevant to normal insemination. We believe that our disagreement with the conclusions drawn by Spielman ct al. (1967, 1969) are due in large part to their reliance on the forced-copulation technique. Three lines of evidence from the present work bear out the contention that female behavior is adequate to prevent insemination of refractory females and that seminal expulsion is not involved. First, males exposed to refractory females showed no loss in insemination capacity. Males could inseminate 5-7 virgin females ; repeated prior coupling with refractory females did not change this figure. 212 K. \v. (.; \VAD2, G. B. CRAIG, JR. AND W. A. MICKEY Secondly, refractory females exposed to males do not show external sperm masses. Of 100 refractory females examined, not a single one had semen on the external genitalia. Thirdly, refractory females placed with radioactive males showed no increase in radioactivity. If semen had been expelled from the bursa onto the external surface, the total count per female should have been raised. From the standpoint of evolutionary parsimony, prevention of insemination of refractory females by seminal expulsion seems improbable. Considerable wastage of genetic material could occur. In our laboratory, Hausermann and Nijhout (in press) have shown that fecundity of male A. acgypti is strictly limited; a male cannot inseminate more than 5-7 females in a life-time. Yet, males of A. acgvpti show hyperactive sexual activity (Roth, 1948), which would quickly lead to sperm depletion in the presence of refractory females. On the other hand, prevention of insemination by a behavioral mechanism of refractory females seems far more likely. If the female fails to give the proper cue indicating sexual receptivity, the male will not ejaculate and sperm will be conserved. Spielman and his co-workers also argued that vacuolization of the bursal wall and spherule formation in the semen might be involved in the mechanism for semen retention. When seminal fluid comes in contact with a bursa in vivo or in vitro, physical changes do take place. However, there is no evidence to indicate that these events are in any way related to the retention of semen. Indeed, vacuoliza- tion and/or spherule formation are not always observed in females retaining semen (Spielman et a I., 1967, 1969). Spielman concluded that the bursa of the mature virgin female (but not the immature or non-virgin female ) has the capacity to "transform" seminal fluid and retain semen after copulation. However, Gwadz ( 1970 ) has shown that the copu- latory behavior of female A. acyypti is regulated by the caudal sympathetic nervous system and is not related to the state of the bursa. When the terminal abdominal ganglion is surgically removed from a normally refractory female, she can be inseminated and will retain semen after copulation in free-flight. Although the chemical nature of the bursa of the young or previously inseminated refractory female is unchanged by the operation, the behavioral responses are altered. Nor- mally refractory females that lack the terminal ganglion are not able to prevent copulation and insemination. Sexual behavior of females of A. aegypti is similar in many respects to that displayed by females of the house fly, Musca domestica. Adams and Hintz (1969) have demonstrated a post-eclosion refractory period controlled by the corpora allata. Allatectomized female flies are mounted by males, but no insemination takes place because of female behavior. Moreover, house flies are monogamous ; a component of the male seminal fluid renders females refractory to a second insemination for life (Riemann, Moen and Thorson, 1967). Sexually receptive females respond to courtship and mounting by extrusion of the ovipositor. This extrusion facilitates clasping and insemination by the male. Refractory females are mounted, but do not extrude the ovipositor ; males are unable to clasp properly and cannot in- seminate the female. There are still serious gaps in our knowledge of mating behavior of female mosquitoes. The mode of action and precise target organs for the various hor- mones involved in regulating sexual behavior still need to be explored. More- MATING BEHAVIOR IN AEDES AEGYPTI 213 over, almost nothing is known about female pheromones, sex attractants and copulation initiators. An even greater void exists in knowledge of the physiology of mating behavior of male mosquitoes. Because they are not disease vectors, males have reecived scant attention ; their sex-related physiology has been virtually ignored. Perhaps the greatest deficiency is in knowledge of mating behavior in the field. The where and when of mating, and the factors which bring the sexes together are questions of prime importance. The answers to these questions are especially relevant to those attempting programs of genetic control. These questions may prove to be among the most difficult to answer. We wish to thank Dr. Andrew Spielman of the Department of Tropical Public Health, Harvard School of Public Health for his critical reading of this manuscript. SUMMARY 1. Certain females of A. aegypti are refractory to insemination, either because they are too young or because of previous insemination. Earlier reports indicated that refractory females prevent insemination by expulsion of semen from the bursa copulatrix. The present work indicates that this is not the case. 2. Three lines of evidence indicate that refractory females prevent insemination by failing to give the cues that lead to male ejaculation. First, males exposed to refractory females show no loss in their ability to inseminate. Secondly, refrac- tory females exposed to males show no evidence of external sperm masses resulting from expulsion. Thirdly, refractory females placed with radioactive males show no increase in radioactivity, indicating no transfer of labelled semen. Since the forced-copulation technique overrides normal behavior of females, earlier data obtained using this technique are not relevant to normal insemination. 3. Copulation by a mature virgin female is characterized by firm genital union followed by insemination. A refractory female allows the male to couple but not to copulate. Coupling involves a superficial joining of the genitalia; no ejacula- tion or insemination occurs. LITERATURE CITED ADAMS, T. S., AND A. M. HINTZ, 1969. Relationship of age, ovarian development, and the corpus allatum to mating in the housefly, Musca domestica. J. Insect Physiol., 15: 201-215. AKEY, D. H., AND J .C. JONES, 1968. Sexual responses of adult male Aedes aegypti using the forced-copulation method. Biol. Bull., 135: 445-453. CRAIG, G. B., JR., AND R. C. VANDEHEY, 1962. Genetic variability in Aedes aegypti (Diptera : Culicidae). I. Mutations affecting color pattern. Ann. Entontol. Soc. Amer., 55: 47-58. FUCHS, M. S., G. B. CRAIG, JR. AND D. D. DESPOMMIER, 1969. The protein nature of the substance inducing female monogamy in Aedes aegvpti J. Insect. Plivsiol., 15: 701- 709. FUCHS, M. S., G. B. CRAIG, JR. AND E. A. Hiss, 1968. The biochemical basis of female monogamy in mosquitoes. I. Extraction of the active principle from Aedes aegypti. Life Sciences, 7 : 835-839. GILLETT, J. D., 1955. The inherited basis of variation in the hatching-response of Aedes eggs (Diptera: Culicidae). Bull. Entomol. Res., 46: 255-265. 214 K. W. GWADZ, G. B. CRAIG, JR. AND W. A. HICKEY GWADZ, R. W., 1970. The neuro-hormonal regulation of sexual receptivity in female Acdcs acgypti (Diptera: Culicidae). Ph.D. thesis, University of Notre Dame, 88 pp. GWADZ, R. W., AND G. B. CRAIG, JR., 1968. Sexual receptivity in female Acdcs acgvpti. Mosq. News, 28: 586-592. GWADZ, R. W., AND G. B. CRAIG, JR., 1970. Female polygamy due to inadequate semen trans- fer in Acdcs acgypti. Mosq. News, 30: 355-360. GWADZ, R. W., L. P. LOUNIBOS AND G. B. CRAIG, 1971. Precocious sexual receptivity induced by a juvenile hormone analogue in females of the yellow fever mosquito, Acdes aegypti. Gen. Comp. Endocrinol., in press. HAYES, R. O., 1953. Determination of a physiological saline solution for Aedcs acgypti (L.). /. Econ. Entomol., 46: 624-627. HICKEY, W. A., AND G. B. CRAIG, JR., 1966. Genetic distortion of sex ratio in a mosquito, Acdes acgypti. Genetics, 53: 1177-1196. JONES, J. C., AND R. E. WHEELER, 1965. An analytical study of coitus in Aedes aegypti (Lin- naeus). /. Morphol., 117: 401-424. HAUSERMANN, W., AND H. F. NIJHOUT, 1970. Loss of male fecundity following sperm deple- tion in Aedes acgypti. J. Mcd. Entomol., in press. LEA, A. O., 1968. Mating of virgin Aedes aegypti without insemination. /. Insect Physiol., 14: 305-308. MACGREGOR, M. E., 1915. Notes on the rearing of Stegomyia fasciata in London. /. Trap. Mcd.Hyg., 18: 193-196. RIEMANN, J. G., D. J. MOEN AND B. J. TnoRSON, 1967. Female monogamy and its control in houseflies. /. Insect Physiol., 13: 407-418. ROTH, L. M., 1948. A study of mosquito behavior. An experimental laboratory study of the sexual behavior of Aedes aegypti (Linnaeus). Amer. Midland Natur., 40: 265-352. SPIELMAN, A., 1964. The mechanics of copulation in Aedes aegypti. Biol. Bull., 127: 324-344. SPIELMAN, A., SR. M. G. LEAHY AND V. SKAFF, 1967. Seminal loss in repeatedly mated Aedcs aegypti. Biol. Bull, 132: 404-412. SPIELMAN, A., SR. M. G. LEAHY AND V. SKAFF, 1969. Failure of effective insemination of young female Aedes acgypti mosquitoes. /. Insect Physiol., 15: 1471-1479. Reference : Biol. Bull., 140 : 215-229. (April, 1971 ) ANALYSIS OF THE FEEDING RESPONSE IN THE ONUPHID POLYCHAETE DIOPATRA CUPREA (BOSC) x CHARLOTTE P. MANGUM AND CHARLES D. COX 2 Department of Biology, College of William and Mary, Williamsburg, Virginia 23185 Like other annelids, spontaneous behavior of the omiphid polychaete Diopatra citprca (Bosc) is dominated by rhythmic bursts of muscular activity that move large volumes of water through the tube. LTnlike other species, however, the frac- tion of oxygen withdrawn in respiration is quite small. Considered in conjunction with our observation of a feeding response to chemical stimulation (Mangum and Cox, 1966), the low rate of oxygen withdrawal suggests that tube irrigation in this species is primarily a means of testing the medium for the presence of food, and only secondarily serves to ventilate the organs of respiratory exchange. When chemical stimuli are present in supra-threshold concentrations, the animal gives a behavioral response which concentrates food particles and culminates in ingestion (Mangum, Santos and Rhodes, 1968). We have analyzed the response in terms of its separate behavioral components, and attempted to identify the substances in food which elicit the response. MATERIALS AND METHODS Behavioral observations On several occasions, the response to fortuitous food stimuli has been observed in nature. However, analysis of the response was accomplished in the laboratory in running seawater, with controlled stimulation. Worms were initially observed in their natural opaque tubes mounted in the normal vertical position. Later observations were made on worms placed horizontally in glass tubes. If observations are solely visual, some of the movements involved in the feeding response may be easily confused with other behavior patterns which generate water currents, viz. rejection of noxious chemical stimuli and ventilation movements. The similarity was also noted in Chaetopterus variopedatus by Wells and Dales (1951). Therefore, the nature of the response of D. cnprea to test substances was determined by visual observations in conjunction with quantitative measurements of the current (Fig. 1). The apparatus is similar to that used in investigations of ventilation rhythms (Dales, Mangum and Tichy, 1970; Mangum, 1971). The worm tube is attached to the recording chamber by the tail end, so that an upward deflection of the writing pen reflects current flow out of the tube. The direction and pattern of the current were determined by adding carmine particles to a test solution. Each test substance was presented to a minimum of ten worms. Providing that the precautions below are taken, inconsistency of the response is not a problem. Although there is some variation in the intensity of the response given to supra- threshold stimuli, it is not sufficient to be confusing and it is not clearly correlated 1 Supported by National Science Foundation research grants GB-5563 and GB-6884. 2 Current address : Department of Microbiology, University of Georgia, Athens, Georgia. 215 216 CHARLOTTE P. MANGUM AND CHARLES D. COX with intensity of stimulation. Therefore, we have treated the response as if it were all-or-none, which is very nearly correct. With the exception of the temperature dependence experiments, recalcitrance was rarely encountered within the period described helow and designation of an acceptable criterion was unnecessary. In no case did the worms show less than 90% uniformity to qualitative tests, and 80% to quantitative tests. D o **~r \ ALANINE \ M ARENARIA EXTRACT EGG ALBUMEN EGG ALBUMEN TIME (min) FIGURE 1. Currents generated by Diopatra cuprca. Feeding current given in response to clam extract and 10~ a M amino acid solutions. Rejection response given to egg albumen. Small deflections reflect ventilation movements. Each increment of ordinate is 1 nil ; each increment of abscissa is 1 min. All qualitative tests were made in running seawater, and ample time (ranging from 15 minutes to one hour, depending on the volume of the container) between tests was allowed to insure thorough renewal of the medium as well as subsidence of the response. It soon became apparent that the nutritional condition of the animal has a great influence on the response. In their natural habitat, worms have full guts during most months of the year (Mangum et al., 1968). Animals brought into the laboratory may not respond, or they may respond inconsistently for a short period. Our observations have been made only on worms which have been acclimated to laboratory conditions for a minimum of three clavs. Conversely, FEEDING RESPONSE IN DIOPATRA 217 very hungry worms respond to almost any stimulus ; animals kept in the laboratory without feeding for six weeks or longer, depending on the temperature, have been observed attempting to feed from the surface film of the water overhead. In this case, the response is given continually. Therefore, tests w T ere made only on ani- mals collected less than two weeks previously in warm weather, and three weeks in colder weather. The worms were maintained without feeding in the laboratory, either in running seawater at 16-23 C or in filtered, recirculating seawater at 16-20 C. Response to extracts of organisms Extracts of a large number of marine animals were made by simply cutting open the animal and collecting the liquid which flowed into a beaker. The extracts were centrifuged, filtered and immediately tested. In a few cases, vis. the sponges and ectoproct, external pressure was applied in order to express fluid. Extracts of several species of algae and one higher plant were prepared by homogenization. Subsequent to these initial tests, the response to clam extract was used as the criterion for a positive feeding response to chemical stimuli. Analyses of gut contents confirm the participation of at least one species of lamellibranch mollusc in the diet of D. cuprea (Mangum et al., 1968). Valves of the soft-shelled clam M\a arenaria were bent back 180, and the fluid (largely mantle fluid, muscle fluid and blood) collected. Some of the analyses were repeated on fluid from the clams Mercenaria mercenaria, Tagelus plebius and Spisula solidissima, as indi- cated below. Analysis of clam extract Centrifuged clam extract was dialyzed for 15 hours at 4 C against agitated seawater at a volume ratio of 1 : 1000, with a change of seawater every five hours. Dialysis was repeated on extracts of different individuals against running seawater at 21 C for 5 hours. Clam extract (32 ml) was precipitated with 28 ml cold 10% (w/v) trichloro- acetic acid, centrifuged and the supernatant fluid (pH initially 1.1) dialyzed against agitated standing seawater at 4 C until the pH reached that of seawater (12- 16 hr). Clam extract (25 ml) was precipitated first with an equal volume and then with H volumes of ethyl alcohol. The precipitate was washed three times with ethyl alcohol and then redissolved in a large volume of seawater. The solution was con- centrated to the original volume of 25 ml by adding dry Sephadex G-25, and then tested. A precipitate was also obtained by adding an equal volume of acetone, and then redissolved and tested. Additional acetone produced no precipitate. Clam extract was chromatographed at 4 C on columns of Cal Bio-Gel P10 (1.5 X 20 cm) and Sephadex G-75 (1.5 X 30 cm). The gels were swollen and eluted in filtered seawater, and the fractions were collected with a LKB Ultrorac collector and drop counter (15 drops/ml). Positions on the P10 column were marked with a mixture of \% (w/v) glycine and 0.1% Sephadex Blue dextran 2000. The glycine was detected with the ninhydrin reagent. The fractions were brought to room temperature and bioassayed. 218 CHARLOTTE P. MANGUM AND CHARLES D. COX The G-75 column (42 nil bed volume) was calibrated with 1.5% (w/v) re- crystallized ovalbumin (Fisher Co.) and 3 X 10"* M cytochrome c (Sigma Chemi- cal Co. ) . Cytochrome r was detected visually ; ovalbumin was detected by the Folin-Ciocalteau test (Clark, 1964). Blue dextran 2000 was used to determine the void volume. The pressure head was adjusted to yield a flow rate of 0.2 ml/ min, which remained constant after initial equilibration of the column for 48 hr. Molecular weights of the standards were assumed to be 13,000 for cytochrome c and 45,000 for ovalbumin. After initial separations on the G-75 column suggested that dilution of the sample on the column resulted in our failure to detect active substances by bio- assay, clam extract was concentrated prior to chromatography. Dry Sephadex G-25 (water regain 2.5 ml/g) was placed in the bottom of a 10 ml glass syringe filled with a small piece of nylon net at the neck. Eight ml of clam extract were then pipetted into the gel, gently stirred and allowed to stand at 5 C for 45 min. Unabsorbed fluid (1.5 ml) was then forced out of the syringe up to the point of compression of the gel, and chromatographed as above on Sephadex G-75. Alter- nate fractions were subjected to the Folin-Ciocalteau test and the Molisch test (Clark, 1964), or bioassayed. The amino acid content of extracts from M. arenaria and M. nierccnaria was determined according to the procedure described by DuPaul and Webb (1970), using a Technicon Auto-Analyzer. The analyses were performed by Dr. K. L. Webb and Mr. W. D. DuPaul on extracts from three individuals of M. mercenaria and six of M. arenaria, acclimated to 18-20%o salinity. Data from homogenates of tissue of 5. solidisshna have kindly been furnished by Mr. W. D. DuPaul. Response to commercially prepared chemicals A wide range of chemicals was tested at concentrations of 1% (w/v) in filtered seawater, unless otherwise specified. All of these substances were commercially prepared reagent grade chemicals, except for the protein globin, furnished by Dr. C. T. Grant. Attention was then focused on the amino acids (Sigma Chemical Co.), initially assayed at concentrations of 10~ 2 M. Fresh solutions were made daily. The assay procedure was altered to determine the threshold concentration of each of eight amino acids. The worm, in a tube fabricated in the laboratory from clean sand and hence undecorated, was placed in a 5 -inch finger bowl with a known quantity of filtered seawater. While constantly mixing with a magnetic stirrer, measured volumes of a 1 M solution of the amino acid were added quickly, to bring the total concentration of the medium to 10~ 7 M, and the response was noted. By adding further aliquots of the stock solution, the final concentration was raised in incre- ments of a factor 10 until the response was positive. The experiment was repeated on a minimum of five worms. Unfortunately, the recording apparatus is not suit- able for determinations of threshold concentrations, since it employs volumes of water in separate chambers which cannot be thoroughly mixed (Mangum. 1964). Therefore, the criterion for the feeding current component of the response was visual observation of anterior beating coupled with slow headward undulatory waves of muscular contraction. On several occasions we observed the initiation of beating unaccompanied by perceptible undulatory waves. We interpreted these FEEDING RESPONSE IN DIOPATRA 219 movements as the spontaneous initiation of a rhythmic burst of irrigation activity, and, consequently, discarded the results. A mixture of amino acids was prepared from the results of the analyses of clam extracts, and then hioassayed. FIGURE 2. Movements involved in generating the feeding current. RESULTS Description of the response When a small piece of clam meat is placed within a few centimeters of the anterior opening of the tube, the worm extends its anterior region and undergoes searching behavior, usually strongly oriented toward the food. We have never observed failure to locate and ingest the food except at temperatures a few degrees higher than those at which the response is blocked (Mangum and Cox. 1966). At no time have we observed a feeding response to an organism held in the vicinity 220 CHARLOTTE P. MANGUM AND CHARLES D. COX of the tube opening or to a member of the tube epibiota (Mangum ct a!., 1968), if we were certain that the organism was intact. However, the response is given immediately to an organism which is only slightly punctured. When a pipette of centrifuged, filtered clam extract is very gently emptied in the vicinity of the tube opening, the worm gives a charatceristic three-component response. It is this pattern of behavior which we have designated the feeding response and which has been the subject of our investigation. We have separated three components : the first rheotactic, the second chemotactic and the third both cliemo- and mechano-tactic. Initially, the antennae begin to rotate in their sockets, focused towards the source of the stimulus. Since antennal rotation can be elicited with many fluids, including seawater as well as clam extract, it cannot be considered a component of the chemically elicited response, but is instead a rheotactic component. The second component, which usually follows the first in time, consists of a set of movements which generate a current (Fig. 1) and an eddy at the tube aperture. The anterior end of the worm beats dorsoventrally against the sides of the tube, creating a centripetal eddy which concentrates particulate matter in the immediate vicinity. Simultaneously, an undulatory wave of contraction of the body wall muscles, also dorsoventral, passes very slowly in the anterior direction. Since the ventral beat of the anterior region is followed by a slight pause, there is a leak from the eddy in the dorsal direction, bringing concentrated particles into contact with the mouth region (Fig. 2). Because of the undulatory wave, the net flow of the feeding current is outwards from the tube (Fig. 1), and the tube is cleared of potential fouling material. We have been able to elicit this component of the response only with chemical stimuli. It is easily confused with normal tube irri- gation, the movements differing only in the much slower rate of passage of the undulatory wave and, consequently, the velocity of the current (Fig. 1). The feeding current movements also resemble those of the rejection response, which is accompanied by an extremely rapid headward undulation and, thus, a far greater volume moved (Fig. 1), as well as withdrawal to a deeper position in the tube. The third and final component, which may occur simultaneously with the second, consists of two sets of movements : ( 1 ) Parapodial clasping, in which the greatly enlarged parapodia of the anterior segments move ventrolaterally, with the result that the worm grasps large particles concentrated in the eddy. If an object is grasped, the movements cease; if not, they continue. (2) Jaw flexing and mouth opening, with the result that the worm tears off a piece of food held by the parapodia and ingests it. The two sets of movements are invariably coupled ; one does not occur without the other. This component can be elicited by either chemical or mechanical stimuli. It is given in response to a jet of clam extract, a piece of clam meat or a small piece of congealed rubber cement gently touching the tentacles or palps. Unlike clam meat, the worm grasps rubber cement, ingests it and soon regurgitates it. After a few minutes, the worm will take another piece of rubber cement, regurgitate it, and take another for at least eight trials. If alternate pieces of clam meat and rubber cement are fed, however, the worm no longer responds to rubber cement after five or six pieces of clam meat, whereas clam meat is still an effective stimulus. Although the state of satiety apparently FEEDING RESPONSE IN DIOPATRA 221 TABLE I Response to extracts of marine organisms* Species Response Porifera Microciona prolifera N.R. Halichondria sp. N.R. Coelenterata Metridium senile F.C. Diadumene leucolena N.R. Sipunculida Golfingia gouldi F.C. Annelida Amphitrite ornata F.C., M.O. Hydroides dianthus F.C. Nereis succinea F.C., M.O. Diopatra cuprea F.C., M.O. Arenicola marina F.C., M.O. Glycera dibranchiata F.C., M.O. Mollusca Ensis directus F.C. Aequipecten ir radians F.C. My a arenaria F.C., M.O. Mercenaria mercenaria F.C., M.O. Crassostrea virginica F.C., M.O. Tagelus plebeius F.C., M.O. Nassarius obsoletus F.C., M.O. Polinices duplicatus F.C. Spisula solidissima F.C., M.O. Arthropoda Homarus americana F.C. Carcinus meanas F.C. Artemia salina F.C., M.O. Ectoprocta Bugula turrita Rej. Echinodermata Asterias forbesi F.C. Arbacia punctulata F.C. Strongylocentrotus drobachiensis F.C. Echinarachinius parma F.C. Urochordata dona intestinalis Rej. Yertebrata Fundulus fundulus F.C. Chlorophyta C odium sp. F.C. Rhodophyta Porphyridium sp. N.R. Angiosperm Zostera sp. F.C. * N.R. = no response; F.C. = feeding current; M.O. = mouth opening; Rej. = rejection. 222 CHARLOTTE P. MANGUM AND CHARLES D. COX exerts a differential influence on the response to clam meat and rubber cement, the pattern of the response in a hungry worm is indistinguishable. The location of the receptors mediating chemically stimulated components of the response is uncertain. Although the worm is lavishly provided with anterior protrusions of various sorts, they seem to be unessential to the initiation of the response. After successive surgical removal of each of the five occipital antennae, including ceratophores, both sets of palps and several gills, the worm continues to give both components of the chemically stimulated response. Even after removal of the prostomium and several anterior setigers, the worm still generates a current of similar magnitude as the feeding current, if clam extract is applied. Indeed, a TABLE II Molar concentration of free amino acids in three species of lamellibranch molluscs Amino acid Mercenaria mercenaries extract (N = 3) jl/ya arenaria extract (N = 6) Spisida solidissirna (gill tissue) (N = 4) Aspartic acid Ta urine 7 X 10- 7 X 10-* 1 X 10-* 7 X 10- B 5 X 10- 5 2 X 10-* Glycine Histidine 2 X 10-" 2 X 10- 5 X lO-^ 4 X 10~ 5 2 X 10-* 1 X lO- 6 Threonine 4 X 10- 5 8 X lO- 5 3 X lO- 6 Alanine 1 X 10-" 3 X 10-* 5 X 10~ 5 Isoleucine 6 X 10-6 2 X 10- 6 1 X lO- 6 Leucine 1 X 10- 5 4 X 10- 6 1 X 10-6 Lysine Arginine Glutamic acid 5 X 10- 5 4 X 10- 5 1 X 10-7 5 X 10-5 4 X 10~ 5 7 X lO-s 4 X lO- 6 7 X lO- 6 3 X 10-5 Serine 5 X 10- s 3 X 10-* 7 X 10-6 Tyrosine Cysteine Tryptophan Ornithine 1 X 10~ 5 none none 1 X lO- 8 2 X 10- 5 X lO- 6 > 1 X 10-6 7 X 10- 1 X ID" 6 4 X 10- none 1 X lO- 6 Proline 2 X 10~ 6 1 X 10- 6 none Valine 1 X 10- 5 3 X 10~ 8 1 X 10-6 Methionine 6 X lO- 6 1 X lO-^ 3 X 10- 7 Phenylalanine 2 X 10-6 1 X 10~ 6 1 X lO- 6 localized muscular contraction of the body wall is elicited by directing a jet of clam extract at the dorsal surface of the isolated autotomized tail region. Finally, Sven Donaldson and James Case (personal communication from Dr. James F. Case, University of California, Santa Barbara) have recorded electrical potentials from the segmental nerves of several isolated postbranchial segments, when clam extract was applied. Temperature sensitivity of the response Animals collected at 3-8 C and acclimated in the laboratory to 5 for two weeks give the response to clam extract only rarely. Of 20 worms tested, 15 gave no response to either clam extract or to clam meat gently applied to the tentacles. Five worms produced a current somewhat in excess of normal spontaneous irriga- tion but none gave the mouth opening component. Only one in 20 ingested the FEEDING RESPONSE IN DIOPATRA 223 clam meat. In contrast, 18 of 20 animals acclimated to 10 C for two weeks gave the complete response to clam extract. Low temperature blockage of the response continues for many weeks; ulti- mately, however, inhibition is released. Of eight animals kept at 5 C in the lab- oratory without food for seven months, six ingested pieces of clam meat placed in the vicinity of the tube opening. Response to extracts of organisms The results of an assay of 32 marine organisms for the presence of substances eliciting the feeding response in D. cuprea are given in Table I. Many of the tests TABLE III Response of Diopatra cuprea to various substances* Substance Response Substance Response 0-D-Glucose, 1% a-D-Glucose, 1% D-Galactose, 1% a-Lactose, 1% /3-Lactose, 1% D-Arabinose, 1% Maltose, 1% L-Rhamnose, 1% D-Mannose, 1% D-Xylose, 1% D-Sorbitol, 1% Sucrose, 1% Glycerol, 1% Glycogen, 1% Inulin, 1% N-Acetyl glucosamine, 1% N-Acetyl galactosamine, 1% Hemoglobin (Fisher Co.), 1% Human globin, 10~ u M Arenicola marina blood, fresh A. marina blood, aged 24 hr at 5 C F.C. N.R. F.C. F.C. F.C. N.R. F.C., M.O. N.R. N.R. F.C. F.C. N.R. N.R. N.R. N.R. F.C., M.O. Ethyl alcohol Distilled water Vegetable oil emulsion Mineral oil emulsion Butyric acid, 1% Linoleic acid, 1% Trichloroacetic acid, 1% Reduced glutathione, 10~ B M F.C, F.C., M.O. M.O. F.C. F.C., M.O. Rej. Gelatin, 1% Casein, 1% Trypsin, 0.1% Trypsin, 1% Trypsin, 1%, boiled Egg yolk Egg albumen Clam extract, boiled Crab urine Urea, 1% Hydroxyurea, 1% Rej. Rej. F.C. F.C. Rej. N.R. Rej. N.R. N.R. N.R. F.C. Rej. Rej. F.C. Rej. Rej. Rej. F.C. F.C. F.C. = feeding current ; M.O. = mouth opening; Rej. = rejection; N.R. = no response. were made prior to analysis of the response into its separate components when the criterion was a recorded current of the correct magnitude together with visual observation of the movements. In all cases where mouth opening is not reported, we do not know whether it occurs. However, we have never been able to confirm the occurrence of the feeding current component without mouth opening in response to a freshly prepared natural extract. The condition of the extract certainly determines the nature of the response. While either freshly prepared or frozen and thawed clam extract causes the feeding response, boiled and cooled clam extract elicits a rejection response. Whole blood freshly collected from the polychaete Arenicola marina, causes the feeding response, 224 CHARLOTTE P. MANGUM AND CHARLES D. COX but 24 hour-aged blood which shows methemoglobin absorption bauds elicits a rejection response. D. cuprea responds to an extract of almost any organism, including itself. The absence of a response to the two sponges and to the red alga Porphyridium may reflect the difficulty of obtaining fluid rather than the absence of active substances. Lack of a response to extract of the anemone Diadumene leucolena cannot be similarly explained, however. Response to commercially prepared chemicals The response to commercially prepared chemicals is not highly correlated with molecular structure. For example, D. cuprea gives a feeding current response to TABLE IV Response of Diopatra cuprea to amino acids* Amino acid Response Amino acid Response Aspartic acid Rej. Alanine F.C. Glutamic acid Rej. (10- 6 M ) Threonine F.C. Taurine N.R. Lysine F.C. Histidine N.R. Cysteine F.C. (10- 6 M) Glycine N.R. Asparagine F.C. Cystine N.R. Methionine F.C. (10~ 6 M); M.O. (10~ 3 M ) Isoleucine F.C. Phenylalanine F.C. (10- 6 M); M.O. (1Q- 2 M) Leucine F.C. Proline F.C; M.O. Tryptophan F.C. (10-* M) Hydroxyproline F.C. (10- 7 M); M.O. Tyrosine F.C. (10~ 7 M) Valine F.C. (10-* M); M.O. (IQ~* M) Serine F.C. (10~ 6 M) Norvaline F.C. Arginine F.C. Norleucine F.C. Ornithine F.C.; M.O. * Rej. = rejection; N.R. = no response; F.C. = feeding current; M.O. = mouth opening; 10~ n M threshold concentration. a number of sugars and fails to do so to others (Table III). The activity of test substances is not related to chemical properties in any way which is obvious to us. Neither the optical activity of hydroxyl groups, nor the state of polymerization, nor the reducing properties of the molecules explain the distribution of activity. Although the worms do give the feeding current response to oil emulsions, which cannot be due to mechanical stimulation, lipids as a class would seem to be unlikely candidates for active substances in natural extracts. Several proteins elicit a response, although not all of those tested. The worms give at least the feeding current component to 20 of 25 amino acids tested, and the complete response to six (Table IV). Once again, there does not seem to be a common denominator in terms of chemical properties of the amino acids which do or do not give a response, or which elicit only the feeding current component. The threshold concentrations for the feeding current component seem rather low, and those for mouth opening moderately high. Most importantly, the threshold for the feeding current lies below the figure for the concentration of four amino acids in M. mercenaria extract (Table II), six in M. arenaria extract FEEDING RESPONSE IN DIOPATRA 225 (Table II), six in S. solidissima gill homogenates (Table II), and in S. solidissima muscle bomogenates (personal communication from W. D. DuPaul, Virginia Insti- tute of Marine Science). Tbe tbresbold for mouth opening is attained by only one amino acid, valine, in HI. arenaria muscle. When a mixture of 20 amino acids was prepared according to the recipe for M. arenaria extract (see below) given in Table II, 10 out of 10 worms tested gave a feeding component response to the solution. But 8 out of 10 failed to give the mouth opening component. None of the worms gave a rejection response, despite the supra-threshold concentration of glutamic acid. Therefore, synergistic effects are not completely absent. Analysis of clam extract When M. arenaria extract is diluted to 10~ 2 , the worm still gives all three com- ponents of the response. When it is further diluted to 10~ 3 , however, mouth open- 100 z o et 80 60 CYTOCHROME C OVALBUM I N BLUE DEXTRAN 1 2 345 MOLECULAR WEIGHT (x TO' 4 ) FIGURE 3. Calibration of Sephadex G-75 column. Closed circle ( ) indicates position of unknown active substance. ing no longer occurs. Dialysis of extracts of four species of lamellibranch molluscs yielded an interesting dichotomy : the non-dialyzable fraction from Mya arenaria, Spisula solidissima and Tagelus plebius elicited a complete feeding response, whereas that from Mercenaria mercenaria gave no response. A sample of each extract aged at the same temperature for the same length of time gave a complete response. The results were identical, regardless of the temperature or conditions of dialysis. Trichoroacetic acid precipitation of M. arenaria and T. plcbiits extracts fol- lowed by dialysis failed to alter the results. The non-dialyzable and non-precipit- able fractions elicited the feeding response. The redissolved ethyl alcohol precipitate from M. arenaria gave the feeding current component but not mouth opening ; the redissolved acetone precipitate gave no response. 226 CHARLOTTE P. MANGUM AND CHARLES D. COX When the extract of M. mercenaria was chromatographecl on Cal Bio-gel P10, only the fraction corresponding to the amino acid marker elicited a response, and only the feeding current component was given. Chromatography of M. arenaria extract yielded an active fraction corresponding to the amino acid marker, and also a second active fraction 1-3 ml behind the void volume marker. The worms gave a complete response to both fractions. Because the larger molecule was eluted so close to the void volume on Bio-gel P10 subsequent analyses were made on a more precisely calibrated column of Sephadex G-75 (Fig. 3). When Sephadex G-25 concentrated extract of M. arenaria was chromatographed on the G-75 column, the worms gave a complete response to fraction numbers 123-131, corresponding to the glycine marker, and also to numbers 57-58, suggesting a molecular weight of 45,000 if the substance behaves like the globular protein markers. However, the fractions gave a strong positive reaction to the Molisch test for carbohydrate as well as a weaker positive reaction to the Folin-Ciocalteau test (Clark, 1964). These results, in the light of elution of the second active substance behind the blue dextran marker on Bio-gel P10, suggest that the unknown substance is a glycoprotein. Glycoproteins do not behave on gel columns like ideal globular proteins (Whitaker, 1963 ; Andrews, 1965; Ward and Arnott, 1965). The results of Ward and Arnott (1965) suggest that the departure depends upon the ratio of carbohydrate to protein, which is not known for the glycoprotein in clam extract. However, the molecular weight of clam extract glycoprotein may be less than 20,000, since it was retarded by Bio-gel P10, and possibly considerably less since it was not precipitated by trichloroacetic acid. The manufacturers give the figure of 10,000 as the exclusion limit for Bio- gel P10, and 5000-17,000 as the effective operating range. It is possible that the failure to be precipitated by TCA is due to a high ratio of carbohydrate to protein. The concentration of amino acids in our extract of M. mercenaria, and M. are- naria, and in tissue homogenates of 6". solidissifiia is given in Table II. In M. mercenaria extract, the four amino acids serine, tyrosine, valine and phenylalanine occur in concentrations exceeding the threshold for the feeding current component. None of the amino acids occurs in concentrations exceeding the known threshold values for mouth opening. In our M. arenaria extract, six amino acids occur in concentrations above the threshold for the feeding current component, but none exceed that for mouth open- ing. The six are the same as those in M. merceniara, with the addition of cysteine and methionine. 5\ solidissima gill homogenates contain concentrations of the four amino acids tyrosine, cysteine, valine and phenylalanine equal to or in excess of threshold for the feeding current component of the response. Data from several other sources are pertinent. Carr's (1967b) determination of the amino acid content of extract from the shrimp Peanaeus duorarum predicts that it will elicite the feeding current component in D. cuprea, since valine and tyrosine occur in supra-threshold concentrations. Stephens and Preston (Univer- sity of California at Irvine, personal communication) measured amino acid content of coelomic fluid of the polychaete Glycera dibranchiata, which is also known to elicit the response (Table I). Based on their data, the feeding current component may be explained by supra-threshold concentrations of six amino acids, and mouth FEEDING RESPONSE IN DIOPATRA 227 opening to the presence of 2 X 1O 3 M concentrations of valine. Finally, the re- sults of DuPaul and Webb (1970) on M. arenaria (2Q% salinity) suggest that the feeding current response to muscle homogenates can be explained by supra- threshold concentrations of tyrosine, tryptophan, valine, phenylalanine, serine and cysteine. None of these occurs in threshold concentrations for mouth opening, and in fact the worm does not respond with this component to the preparation. Gill homogenates contain the first four amino acids in excess of threshold for the feeding current (personal communication from W. D. DuPaul and K. L. Webb, Virginia Institute of Marine Science). The concentration of valine (2-7 X 1O* M} explains mouth opening as well. DISCUSSION AND CONCLUSIONS Chemically mediated feeding responses by annelids to natural extracts were studied by Copeland and Wieman (1924), who reported that Nereis mrens gives a highly oriented response, culminating in ingestion, to extracts of crustaceans, lamellibranchs and gastropods. They also pointed out the importance of burrow irrigation in permitting the animal to test the medium for chemical stimuli. A response to crab extract by Onnphis creniita was noted by Tampi (1959). In most cases of chemically mediated feeding responses in aquatic invertebrates, it is not known whether the extract derives from an organism which comprises part of the natural diet. Nonetheless, these investigations are at least suggestive of the be- havioral mechanism which allows the animal to recognize its food. Identification of active substances has been attempted in the past two decades by analyzing responses of various animals to commercially prepared chemicals. Most investigators have assumed that good candidates for the role of active sub- stances are of small size, readily diffusible, low in density, and active in fairly low concentrations which may be expected to occur in nature. Amino acids and small peptides meet the physical requirements, and they are also believed to be wide- spread in animal fluids, although little quantitative information has been available until recently. Consequently, feeding responses of various animals to test solutions of amino acids and small peptides have been reported (Loomis, 1955 ; Fulton, 1963; Carr, 1967a, 1967b ; Crisp, 1967; Lindstedt, Muscatine and Lenhoff, 1968; Blanquet and Lenhoff, 1968). To evaluate the significance of these very inter- esting behavioral responses, one crucial question must be answered : does the threshold concentration of the particular substance actually occur in extracts of food which elicit a response? Unfortunately, there have been few reports of amino acid or peptide concentrations in test extracts. Suitability of the test extract has also been inadequately considered : does it originate from a source which is likely to resemble the animal's food ? In the present case, the response to 10~ 5 M methio- nine is of some interest to us, but there is no evidence to suggest that it is of any interest to Diopatra cuprea. To our knowledge, the only investigation in which extensive quantitative analy- sis of natural extracts has been undertaken in conjunction with the response to commercially prepared chemicals is that of Carr (1967a, 1967b), dealing with a graded response of the mud snail Nassarins obsoletns to shrimp extract. As a result of an elaborate analytical procedure, Carr (1967b) succeeded in preparing CHARLOTTE P. MANGUM AND CHARLES D. COX a quantitative solution of 12 compounds which duplicates 50% (probit scale) of the response to shrimp extract. The feeding response of D. cuprea seems to be far less exacting, a quality which is possibly of adaptive significance to a sedentary animal that obtains its food by a less direct mechanism. Absence of specificity to commercially prepared chemi- cals is certainly supported by evidence of a wide range of foods found in the gut (Mangum et at., 1968). Unfortunately, the organisms eaten by natural popula- tions of D. cnprca are too small to yield copious extracts for laboratory analysis. A truly satisfying account of a chemically mediated feeding response would involve the demonstration of supra-threshold concentrations of at least one active substance in each kind of food taken by the species in nature. We have certainly not achieved such a level of explanation. But we do know that one component of the response is given to substances which occur in supra-threshold concentrations in extracts of several lamellibranch molluscs. And that the lamellibranch Gemma gemma is found in the guts of worms taken in nature (Mangum ct al., 1968). The animal's food in nature may often be participate in form (Mangum et al., 1968). If the feeding current component is elicited by valine emanating from food particles, for example, then mouth opening and ingestion will be elicited by mechanical stimulation from the particles concentrated by the feeding current. Thus the process of feeding can be attributed to the presence of four amino acids in Al. mercenaria extract or 6 in Al. arenaria, with considerable margin to spare. Nonetheless, the observed response to chemical stimulation from M. arenaria extract includes the mouth opening component. The larger active molecule, apparently a glycoprotein, may be essential to understanding this component of the observed response. We are indebted to Mr. Robert A. Graham for his contributions to our under- standing of the response. LITERATURE CITED ANDREWS, P., 1965. The gel-filtration behaviour of proteins related to their molecular weights over a wide range. Biochcm. J '., 96: 595-606. BLANQUET, R. S., AND H. M. LENHOFF, 1968. Tyrosine enteroreceptor of Hydra: its function in eliciting a behavior modification. Science, 159: 633-634. CARR, W. E. S., 1967a. Chemoreception in the mud snail, Nassarius obsoletus. I. Properties of stimulatory substances extracted from shrimp. Biol. Bull., 133: 90-105. CARR, W. E. S., 1967b. Chemoreception in the mud snail, Nassarius obsoletus. II. Identifica- tion of stimulatory substances. Biol. Bull., 133: 106-127. CLARK, J. M., 1964. Experimental Biochemistry. W. H. Freeman and Co., San Francisco, California, 228 pp. COPELAND, M., AND H. L. WIEMAN, 1924. The chemical sense and feeding behavior of Nereis virens Sars. Biol. Bull, 47 : 231-238. CRISP, D. J., 1967. Chemoreception in cirripedes. Biol. Bull., 133: 128-140. DALES, R. P., C. P. MANGUM AND J. C. TICHY, 1970. Effect of changes in oxygen and carbon dioxide concentrations on ventilation rhythms in onuphid polychaetes. /. Mar. Biol. Ass. U.K., 50: 365-380. DuPAUL, W. D., AND K. L. WF.BB, 1970. The effect of temperature on salinity-induced changes in the free amino acid pool in M \a arenaria. Comp. Biochcm. Physiol., 32 : 785-801. FULTON, C., 1963. Proline control of the feeding reaction in Cordylophora. J. Gen. Physiol., 46: 823-837. FEEDING RESPONSE IN DIOPATRA 229 LINDSTEDT, K. J., L. MUSCATINE AND H. M. LENHOFF, 1968. Valine activation of feeding in the sea anemone Boloceroides. Comp. Biochcm. PhysioL, 26: 567-572. LOOMIS, W. F., 1955. Glutathione control of specific feeding reactions of hydra. Ann. New York A cad. Sci., 62: 209-228. MANGUM, C. P., 1964. Activity patterns in metabolism and ecology of polychaetes. Comp. Biochem. PhysioL, 11: 239-256. MANGUM, C. P., 1970. Respiratory physiology in annelids. Amcr. Sci., 58: 641-647. MANGUM, C. P., AND C. D. Cox, 1966. A feeding response to chemical stimuli in the onuphid polychaete Diopatra cuprea. Amcr. Zool., 6: 456-457. MANGUM, C. P., S. L. SANTOS AND W. R. RHODES, 1968. Distribution and feeding in the onuphid polychaete, Diopatra cuprea (Bosc). Mar. Biol., 2: 33-40. TAMPI, P. R. S., 1959. An ingenious method of collecting polychaete worms for fish bait. /. Mar. Biol. Ass. India, 1: 250-251. WARD, D. N., AND M. S. ARNOTT, 1965. Gel filtration of proteins with particular reference to the glycoprotein, luteinizing hormone. Anal. Biochem., 12: 296-302. WELLS, G. P., AND R. P. DALES, 1951. Spontaneous activity patterns in animal behaviour: the irrigation of the burrow in the polychaetes Chaetopterus variopedatus Renier and Nereis diver sicolor O. F. Muller. /. Mar. Biol. Ass. U. K., 29: 661-680. WHITAKER, J. R., 1963. Determination of molecular weights of proteins by gel filtration on Sephadex. Anal. Chem., 35: 1950-1953. Reference : Biol Bull, 140: 230-241. (April, 1971) EFFECTS OF PRENATAL X-IRRADIATION ON DEVELOPMENT AND POSTNATAL VIABILITY OF INBRED AND HYBRID MICE 1 DONALD J. NASH Department of Zoology, Colorado State University, Fort Collins, Colorado 80521 Studies of the effects of prenatal x-irradiation of mammals have emphasized in recent years long-term pathological and functional investigations. These inves- tigations have included studies on growth (Murphree and Pace, 1960; Nash and Gowen, 1962; Parish, Murphree and Hupp, 1962; Rugh, Duhamel, Chandler and Varma, 1964), reproduction (Russell, Badgett and Saylors, 1960; Rugh and Jack- son, 1958; Nash, 1969), and postnatal mortality and lifespans (Nash and Gowen, 1965; Rugh and Wohlfromm, 1965; Sikov, Resta and Lofstrom, 1969). In spite of the obvious significance to human populations, relatively little work has been done to explore the influence of genetic background in conditioning the response of embryos to x-irradiation. Most of the earlier studies either have utilized animals of unknown heterogeneous origin or single strains of uniform origin. A few studies have demonstrated the significance of genetic variability in embryonic response (Rugh, 1958; Callas, 1962; Nash and Gowen, 1962, 1965; Dagg, 1963; Nash, 1969). The recent review of Brill, Nance, Engel, Glasser, and Forgotson (1967) has pointed out the basic similarities between laboratory animal studies and long-term, postnatal effects observed in humans that have been exposed to prenatal irradia- tion. The studies reported in the present paper were designed to examine effects of prenatal x-irradiation upon the responses of inbred and hybrid mouse embryos as measured by morphological changes and changes in postnatal viability. MATERIALS AND METHODS The mice that were used in these studies were obtained from first pregnancies of matings within and among three inbred strains : BALB/Gw, K/Gw and S/Gw. Descriptions of the characteristics of the strains are provided in previous papers (Nash and Gowen, 1962, 1965). The matings produced progeny which were either inbred or hybrid, and provided nine different inheritance types. Mated females were examined daily for the presence or absence of a vaginal plug, the plug being the sole criterion used to time the period of gestation and the approximate age of embryos at irradiation. The afternoon of the day on which the plug was observed was considered to be day ^ of gestation. Females were irradiated at 4 PM, and the gestational ages of the embryos at the time of treatment were considered to be approximately 6i, 10i, 14^ or \7\ days. Developmentally, these ages correspond to a time shortly after implantation, a period of major organogenesis, a period of minor organogenesis, and a time of rapid fetal growth, respectively. However, there are a number of factors which !This work has received assistance from Contract AT dl-1) 107 from the Atomic En- ergy Commission and U. S. Public Health Service Grant HD 01898 from the National Insti- tute of Child Health and Human Development. 230 PRENATAL X-IRRADIATION AND VIABILITY 231 are operative which make even the ahove approximate estimates of developmental stages subject to error. There is considerable variation not only between litters, but also among individuals of the same litter, as to fertilization time, implantation time, and subsequent developmental time. Allen and MacDowell (1940), for example, observed that embryos of the same chronological age may differ as much as 24 hours in developmental age. The source of irradiation was a General Electric Maxitron which was operated at 250 pkv, 30 ma with 0.25 mm Cu + 1 mm Al filtration at a distance of 50 cm from anode to mid-mouse. The dose rate was approximately 133r/minute. Pregnant mice were exposed to single doses of whole-body irradiation of 0, 20, 80, 160, or 320 roentgens. TABLE I Percentage of developmental abnormalities induced by irradiation at 10\ days gestation in inbred and hybrid mice Morphological characters observed at birth 80r 160r 320r Inbred Hybrid Inbred Hybrid Inbred Hybrid Number of animals observed 72 76 22 48 17 33 Abnormal 7 3 64 29 100 100 Eyelids open Cranial blister 1 1 9 14 2 2 6 100 12 85 Vaulted cranium 2 24 6 Abnormal tail 6 1 41 10 100 100 Overgrowth of feet Forefeet 5 Hind feet 23 10 12 15 Reduction of feet Forefeet 5 4 65 70 Hind feet 2 59 67 Mice were examined twice daily so that newborn litters usually were found within 12 hours of birth. Newborns were given a gross, external morphological examination. Progeny were checked daily for deaths and all mice were maintained at least through 75 days of age. Litters remained with their treated dams through the suckling period. All mice were weaned at 30 days, and the males and females were separated at that time. Additional data from these experiments describing effects on postnatal growth, reproductive performance, and lifespans are provided in several previous papers (Nash and Gowen, 1961, 1962, 1963a, 1963b, 1965; Nash, 1969). RESULTS Morphology The only three treatments which yielded abnormalities in more than singular occurrence were doses of 80, 160, or 320r at 10| days (Table I). An obvious 232 DONALD J. NASH increase in the percentage of abnormal mice is noted with increasing levels of irradiation, a dose of 320r causing all progeny to he abnormal to some degree. It was observed also that whereas none of the mice treated with 80r had more than one abnormality, and only about 15 per cent treated with 160r had more than one abnormality, all of the progeny exposed to 320r had multiple malformations. Digital abnormalities were restricted to 160 and 320r. The overall incidence of overgrowth remained approximately the same with both doses. However, a distinct difference in response between the forefeet and hindfeet was shown. Only a single case of overgrowth was found in the forefeet, while all other cases of over- growth were observed in hindfeet. TABLE II Means and standard errors of litter sizes at birth after in utero radiation Dose Embryological age Number litters Number young Mean standard error Or 13 101 7.8 .6 20r 6} days 12 92 7.7 .4 10} days 14 106 7.6 .7 14} days 13 116 8.9 .7 17$ days 16 134 8.4 .5 80r 6} days 14 114 8.1 .3 10} days 18 148 8.2 .5 145 days 14 122 8.7 .5 17} days 11 88 8.0 .7 160r 63 days 14 97 6.9 .6 10} days 10 71 7.1 1.0 14} days 20 167 8.4 .5 17} days 14 113 8.1 .6 320r 10} days 13 60 4.6 .5 14} days 9 71 7.9 .5 17} days 24 197 8.2 .4 The incidence of reduction was slight after a dose of 160r, but increased to 66 per cent after 320r. The difference in response between fore- and hindfeet was not significant. It was of primary interest to compare the response of the two broad types of genetic constitutions used in these studies : inbreds and hybrids. The comparison of the responses to the induction of developmental abnormalities is most revealing with the dose of 160r, since with 320r all progeny were affected, and with 80r, the percentage of abnormal mice was too small to detect a significant difference. Fol- lowing a prenatal exposure to 160r, 64 per cent of the inbred progeny showed some malformation, but only 29 per cent of the hybrid progeny did (X 2 = 6.10; 0.01 60- cr "40- 5? 20- or "40- S5 20- i i \ I \ 4 12 26 40 60 75 PAYS AFTFR BIRTH CONTROL 4 12 26 40 60 75 20 R 80 R I60R 100- 80- 14 DAYS x. 100- \c: _ "~ so- 17 DAYS |S^rr^_ Z "-^ z \ 2 60-. ', N ^ ... > SO- 1 "~"~---~ . "~v i co 40- ^ 20- S 20- . , , . . . n 26 40 60 DAYS AFTER BIRTH 75 12 26 40 60 DAYS AFTER BIRTH 75 FIGURE 2. Effects of prenatal irradiation on postnatal survival of mice. The percentage of progeny that were stillborn or died within 12 hours of birth reached a maximum of 100 per cent on day 10^ after a dose of 320r (Figure 1). The embryological age of 10i days appears to be most sensitive to neonatal mor- tality. The LD 50 at birth is somewhere between 80 and 160r, the former causing only 5 per cent neonatal deaths, the latter 76 per cent. The only other treatment that increased neonatal mortality appreciably was 320r at 144 days, which caused 23 per cent of the progeny to be stillborn. Doses of 20r or 80r appeared to be ineffective in increasing neonatal mortality when given at any gestational age. A differential response of inbred and hybrid progeny was evident following a dose of 160r at 10i days. Mortality was 100 per cent in the inbreds, but only PRENATAL X-IRRADIATION AND VIABILITY 235 TABLE III Effects of sex upon postnatal viability following prenatal irradiation (per cent alive) Days after birth Embryonic age Irradiation at treatment dose 4 12 26 40 60 Untreated Or Male 98 98 93 93 91 Female 98 95 87 78 76 6% days 20r Male 95 90 88 88 85 Female 80 70 70 67 63* 80r Male 95 82 74 68 67 Female 89 82 76 76 76 160r Male 93 86 86 86 84 Female 81 72 67* 63* 60** lOidays 20r Male 98 91 89 82 76 Female 89* 80* 75* 73 73 80r Male 87 79 74 66 66 Female 75 62 56 52 52 14 J days 20r Male 90 90 86 84 80 Female 84 84 79 74 69 80r Male 87 83 67 65 65 Female 78 70 63 63 63 16()r Male 83 72 64 63 61 Female 60* 49** 44** 44** 43** 17| days 20r Male 92 86 77 74 71 Female 91 82 72 72 71 80r Male 85 83 79 79 77 Female 85 83 83 83 83 160r Male 98 95 89 86 84 Female 88 84 81 76 74 320r Male 84 66 47 43 40 Female 74* 53** 36* 35 28* * P < 0.05. ** P < 0.01. 64 per cent in the hybrids (X 2 = 8.83; 0.001 < P < 0.01). The difference be- tween inbreds and hybrids in all other treatments was not significant. However, after 320r at 144 days, which did cause an increase in mortality, the data support the contention that the inbreds were more sensitive than the hybrids (30 per cent and 20 per cent mortality, respectively). 236 DONALD J. NASH TABLE IV Comparison of postnatal viability of inbred and hybrid following prenatal irradiation (per cent alive) mice Days after birth Embryonic age Irradiation at treatment dose 4 12 26 40 60 Untreated Or Inbred 87 87 83 83 83 Hybrid 99* 96 90 83 81 6^ days 20r Inbred 52 52 52 52 45 Hybrid 98** 86** 85** 83** 81** 80r I nbred 84 84 77 74 74 Hybrid 88 77 70 68 67 160r Inbred 84 78 75 75 72 Hybrid 75 68 66 63 62 10! days 20r I nbred 74 72 72 67 62 Hybrid 96** 87 79 75 73 80r I nbred 60 49 49 40 39 Hybrid 93** 84** 74** 71** 71** 14! days 20r Inbred 76 76 66 66 63 Hybrid 85 85 83* 78 73 80r Inbred 60 51 38 38 38 Hybrid 94** 91 ** 83** 81** 81** 160r I nbred 49 40 31 30 30 Hybrid 88** 76** 73** 73** 70** 17^ days 20r Inbred 85 70 66 64 61 Hybrid 93 92** 79* 78** 77* 80r I nbred 92 92 92 92 88 Hybrid 81 78 75 75 75 160r Inbred 79 79 74 72 72 Hybrid 85 81 77 73 70 320r I nbred 66 49 27 19 19 Hybrid 78 61 47* 45** 37 Significant difference between inbred and hybrid mice. * P < 0.05. **P < 0.01. Postnatal viability Effects of prenatal irradiation on postnatal survival through 60 days of age are shown in Figure 2. The data include only those mice that survived the first 12 hours after birth. Due to prenatal and neonatal mortality, all levels of irradia- PRENATAL X-IRRADIATION AND VIABILITY 237 tion are not represented at all embryological ages. The data are presented within each of the four embryological ages pooled by sex and genotype. There appears to be some effect on postnatal viability following prenatal irradiation at any of the four gestational ages. Effects are minimum following irradiation at 6i days, reach their maximum following exposure at 10^ days, and then decline at 144 and 17^ days. After a dose of 80r at 1(H days only 55 per cent of the mice that were born 100- 80- > _j < 60- o cc 40- 20- HYBRIDS HYBRIDS INBREDS INBREDS 10 1/2 DAYS 14 1/2 DAYS 14 1/2 DAYS 10 1/2 DAYS 12 26 40 DAYS AFTER BIRTH 60 FIGURE 3. Effects of prenatal irradiation on postnatal survival of inbred and hybrid mice following irradiation at lOi or 14* days of gestation. survived until 60 days of age. Sixty-two per cent survived the same dose at 144 days and 78 per cent at 174 days. Table III gives the data on the postnatal survival of mice according to sex. The data were analyzed by means of variance analyses within each of the embryo- logical age irradiation dose treatment combinations. The mean squares of within sexes variances were tested against the mean squares of between sexes variances, and significant F values are indicated in Table III. In all treatments in which 238 DONALD J. NASH there is a significant difference in postnatal viability between the sexes, more female progeny die than male progeny. Among the untreated, control mice, although the difference does not reach significance, more males survive than females. There does not appear to be any consistent pattern between survival differences between the sexes and level of irradiation and gestational age at treatment. Prenatal irra- diation does not appear to affect the basic sex difference in postnatal viability that was observed in these data. The genetic data of the present investigations were analyzed by comparing the response of inbred and hybrid progeny within each of the treatments (Table IV). Among the control mice inbred progeny suffered a greater mortality than the hybrids the first few days after birth, but had comparable survival thereafter. In all of the prenatally-irradiated groups in which there was a significant difference in postnatal viability between inbred and hybrid mice, it was the inbreds that had the greater mortality. In those treatments which produced significant differences in postnatal viability, the major portion of the mortality had taken place by 12 days after birth. The greatest differences in mortality between inbreds and hybrids are exhibited following irradiation at 10^ days or 14^ days. Results of the overall response of the two types of genotypes pooled by level of irradiation are shown in Figure 2. DISCUSSION Results of the present study, as well as the previous reports describing other postnatal responses of the same group of experimental mice (Nash and Gowen, 1962, 1965; Nash, 1969), have demonstrated that prenatal irradiation may have pronounced, long-term effects on postnatal morphology and function. Such effects may develop in animals that appear normal at birth. The viability findings re- ported in the present paper are in basic agreement with those obtained by Rugh and Wohlfromm (1965) in the mouse and Sikov, Resta, and Lofstrom (1969), in the rat. Rugh and Wohlfromm's comprehensive study indicated that the post- natal LD 50 /, values following prenatal irradiation between days 6 and 18 of gesta- tion fluctuated with respect to certain developmental stages. They found that the most radiosensitive stages were on days 8 to 10 of gestation, a time of active differ- entiation of the nervous system. The methodology employed in the present studies differed from that used by Rugh and Wohlfromm (1965) in two major aspects. First, their studies utilized just one inbred strain of mice (CF X ), whereas the present studies used mice of nine different inheritance types. Secondly, in their studies all newborn litters were given to foster mothers that had not been irradi- ated. In the present studies litters were maintained with their natural mothers who also had been exposed to the same amount of whole-body irradiation as had their litters. Results thus reflect a combination of direct effects of radiation on the progeny and indirect effects that may have been mediated through the somatic effects of irradiation on the mother. In a previous paper (Nash and Gowen, 1962), it was pointed out that there was some indirect evidence that effects of prenatal irradiation on postnatal growth were due to direct effects on the irradiated embryos and not due to effects on the lactating ability or maternal behavior of the mother. Similar postnatal growth patterns were obtained following a dose of PRENATAL X-IRRADIATION AND VIABILITY 239 320r at \7\ days gestation to fetuses and mother and a dose of 320r on the day of birth to the litter only. In addition, many of the effects on postnatal growth did not become evident until sometime after weaning. It is possible that some specific dose-embryological age combination might have an effect on the maternal organism which would influence postnatal viability of the progeny with or without a corresponding effect on postnatal growth. With the above limitations on consideration of results of both investigations in mind, some striking quantitative differences in response to prenatal irradiation of different genotypes become evident. In general the LD 50 / : , values obtained by Rugh and \Yohlfromm from the CF, strain are higher than those of mice in the present studies, both inbreds and hybrids. The hybrids did tend to exhibit a greater radioresistance than the inbreds as measured by effects on development and postnatal viability. Heterosis in radiation response to prenatal exposure also has been noted by Rugh (1958) and Nash (1969). The present experiments do not yield any direct information on the underlying physiological mechanism of the greater radiation resistance of the hybrids. It is possible that some of the differ- ences in response are due, in part, to differences in developmental rates which may be under genotypic control. The same chronological age of gestation need not reflect necessarily the same developmental stage of the embryos. Painter (1928) and Venge (1950) demonstrated the influence of genetically determined variation in the timing of embryological development upon subsequent growth and development. Examination of differences in size of two breeds of rabbits revealed that the differences could be traced back to early embryological stages and differ- ences in cleavage rates of fertilized ova between large- and small-sized breeds. Additional experiments which would allow for differences in developmental rate of different genotypes are necessary. To date, this approach apparently has been used only by Dagg (1963). Differences in embryonic response also may reflect inherent differences in susceptibilities to radiation-induced effects on prenatal development and postnatal growth and viability. SUMMARY 1. Developmental malformations and effects on postnatal viability were mea- sured in inbred and hybrid mice that had been irradiated at different stages of embryological development. Three genetically differentiated inbred strains of mice, BALE, K, and S, and their six possible types of hybrids, including reciprocals, were used. The design for these studies was factorial. There were five levels of irradiation: 0, 20, 80, 160, and 320r ; five embryological stages: untreated, 6.1, 10i, 144, and \7\ days of gestation; two sexes; and mice of nine different inheritance types. Pregnant females were exposed to single whole-body 250 pkv x-rays on a specified day of gestation as timed from the appearance of a vaginal plug. Mal- formations were scored at birth and postnatal mortality was followed through 75 days of age. 2. Morphological anomalies observable at birth were found only after irradia- tion at 10i days gestation. Anomalies were observed after doses of 80r or more. Neonatal mortality also was highest following irradiation at 10| days. The LD no at birth was found to be between 80 and 160r, with a dose of 320r causing all progeny to be stillborn. 240 DONALD J. NASH 3. A differential response both in the induction of malformations and in the incidence of neonatal deaths was found between inbred and hybrid genotypes. Following treatment with 160r at 1(H days there was 64 per cent abnormal and 100 per cent stillborn progeny among the inbreds compared to 29 per cent abnor- mal and 64 per cent stillborn among the hybrids. 4. Effects of prenatal irradiation on postnatal survival to 60 days were at a minimum after treatment at 6i days gestation, reached a maximum following irradiation at 10^ days, and then declined at 14^ or 17i days. No consistent pat- terns of survival differences between males and females and level of irradiation and embryological stage at the time of treatment was observed. 5. Among the control or non-irradiated mice, inbred progeny suffered a higher mortality than the hybrids especially during the first few days after birth. In those treatments which produced a significant increase in postnatal mortality, inbred genotypes responded to a greater degree than did hybrid genotypes. LITERATURE CITED ALLEN", E., AND E. C. MACDOWELL, 1940. Variation in mouse embryos of eight days gestation. Anat. Rcc.,77: 165-173. BAGG, H. J., 1922. Disturbances in mammalian development produced by radium emanation. Amer.J. Anat., 30: 133-161. BRILL, A. B., W. E. NANCE, E. ENGEL, S. R. GLASSER AND E. H. FORGOTSON, 1967. Radiation induced abnormalities in human development. Pages 69-87 in S. L. Marcus and C. C. Marcus, Eds., Advances in Obstetrics and Gynecology, Volume 1. Williams and Wil- kins, Baltimore, Maryland. CALLAS, G., 1962. Embryology of X-ray induced cleft palate in mice. Anat. Rcc.. 142: 336. DAGG, C. P., 1963. Some effects of X-irradiation on the development of inbred and hybrid mouse embryos. Pages 91-102 in W. D. Carlson and F. X. Gassner, Eds., Effects of Ionizing Radiation on the Reproductive System. Pergamon Press, New York. GOWEN, J. W., Ed., 1952. Heterosis. Iowa State University Press, Ames, Iowa, 552 pp. GOWEN, J. W., AND J. STADLER, 1956. Life spans of different strains of mice as affected by acute irradiation with 100 pkv X-rays. /. Exp. Zoo!., 132: 133-156. GRAHN, D., 1954a. Genetic variation in the response of mice to total body X-irradiation. I. Body weight response of six inbred strains. /. Exp. Zoo!., 125: 39-61. GRAHN", D., 1954b. Genetic variation in the response of mice to total body X-irradiation. II. Organ weight response of six inbred strains. /. E.vp. Zoo/., 125: 63-83. IOB, T. T .,G. J. LEIBOLD AND H. A. FITZMAVRICE, 1935. Biological effects of roentgen rays. Development of critical periods in mammalian development with X-rays. Amer. J. Anat., 56: 97-117. MURPHREE, R. L., AND H. B. PACE, 1960. The effects of prenatal radiation on postnatal devel- opment in rats. Radiat. Res.. 12: 495-504. NASH, D. J., 1969. Lifetime reproductive performance of mice exposed as embryos to X- irradiation. Biol. Bull.. 137: 189-201. NASH, D. J., AND J. W. GOWEN, 1961. Reproductive performance of mice exposed to em- bryonic X-irradiation. Amer. Zoo/., 1: 377. NASH, D. J., AND J. W. GOWEN, 1962. Effects of X-irradiation on postnatal growth in the mouse. Biol. Bull., 122: 115-136. NASH, D. J., AND J. W. GOWEN, 1963a. Lifespans of mice exposed to embryonic X-irradiation. Rec. Genet. Soc. Amer., 48: 903-904. NASH, D. J., AND J. W. GOWEN, 1963b. Heterosis in the response of mouse embryos to irradia- tion. Proc. A7 Int. Congr. Genet.. (The Netherlands}, 1: 78-79. NASH, D. J., AND J. W. GOWEN, 1965. Effects of X-irradiation of mice exposed in utero during different stages of embryological development on duration of mature life. Bid. Bull., 128:425-458. PRENATAL X-IRRADIATION AND VIABILITY 241 O'BRIEN, J. P., 1956. Vertebrate radiobiology : embryology. Annu. Rev. Nuclear Sci., 6: 423-489. PAINTER, T. S., 1928. Cell size and body size in rabbits. /. Exp. Zool., 50: 441-453. PARISH, N. R., R. L. MURPHREE AND E. W. HUPP, 1962. Growth and sexual development of prenatally irradiated cattle. /. Anim. Sci., 21: 473^78. RUGH, R., 1953. Vertebrate radiobiology: embryology. Annu. Rev. Nuclear Sci., 3: 271-302. RUGH, R., 1958. Increased radioresistance through heterosis. Science, 127: 144-145. RUGH, R., 1959. Vertebrate radiobiology: embryology. Annu. Rev. Nuclear Sci., 9: 493-522. RUGH, R., 1968. The Mouse: Its Reproduction and Development. Burgess Publishing Com- pany, Minneapolis, Minnesota, 430 pp. RUGH, R., AND S. JACKSON, 1958. Effect of fetal X-irradiation upon subsequent fertility of the offspring. /. Exp. Zool., 138: 209-221. RUGH, R., AND M. WOHLFROMM, 1965. Prenatal X-irradiation and postnatal mortality. Radial. Res., 26: 493-506. RUGH, R., A. DUHAMEL, A. CHANDLER AND A. VARMA, 1964. Cataract development after em- bryonic and fetal X-irradiation. Radiat. Res., 22: 519-534. RUSSELL, L. B., 1950. X-ray induced developmental abnormalities in the mouse and their use in the analysis of embryological patterns. I. External and gross visceral changes. /. Exp. Zool., 114: 545-601. RUSSELL, L. B., S. K. BADGETT AND C. L. SAYLORS, 1960. Comparison of the effects of acute, continuous, and fractionated irradiation during embryonic development. Pages 343-359 in A. A. Buzzati Traverso, Ed., In Immediate and Lou' Level Effects of Ionizing Radiation. Taylor and Francis Ltd., London. SIKOV, M. R., C. F. RESTA AND J. E. LOFSTROM, 1969. The effects of prenatal X-irradiation of the rat on postnatal growth and mortality. Radiat. Res., 40: 133-148. STADLER, J., AND J. W. GOWEN, 1957a. Contributions to survival made by body cells of geneti- cally differentiated strains of mice following X-irradiations. Biol. Bull., 112: 400-421. STADLER, J., AND J. W. GOWEN, 1957b. Radiological effects on resistance mechanisms of geneti- cally differentiated strains of mice exposed to Salmonella typhimurium, /. Infec. Dis., 100: 284-299. VENGE, O., 1950. Studies of the maternal influence on the birth weight in rabbits. Acta. Zool., 31: 1-148. Reference : Biol. Bull., 140 : 242-254. ( April, 1971 ) UPTAKE AND ASSIMILATION OF AMINO ACIDS BY PLATYMONAS. II. INCREASED UPTAKE IN NITROGEN-DEFICIENT CELLS * BARBARA B. NORTH AND GROVER C. STEPHENS Department of Developmental and Cell Biology, University of California, Irvine, California 92664 Nitrate, ammonia, and atmospheric nitrogen are usually considered the only significant nitrogen sources for marine phytoplankters (Dugdale and Goering, 1967). However, organic nitrogen as free amino acid is present in nearshore and estuarine waters at concentrations comparable to those of nitrate and ammonia. Five X 10 7 to 5 X 10~ 6 moles/liter of total dissolved amino acids have been re- ported by several investigators (Chau and Riley, 1966; M. E. Clark, California State College at San Diego, personal communication ; Degens, 1968 ; Hobbie, Crawford and Webb, 1968; Webb and Wood, 1967). North and Stephens (1967, 1969) have studied the uptake and assimilation of glycine by Platymonas, a rock pool phytoplankter which is occasionally abundant in inshore locations. This organism grows rapidly in batch culture with glycine, nitrate, or ammonia as its sole nitrogen source (supplied at 2 X 10~ 3 moles/liter). Cells harvested from such cultures are capable of rapid and continuous uptake of glycine-C 14 present at low, natural concentrations (1 X 1O 6 moles/liter). Uptake rates are proportional to concentration in this concentration range. The labelled amino acid in the cells enters oxidative and synthetic pathways. The contribution of amino acid nitrogen accumulated from low ambient concentration to the nitrogen requirement of the cells can be evaluated from measurements of growth rate, amino acid uptake velocity, and cell nitrogen. This report describes the stimulation of amino acid uptake in Platymonas grown on restricted nitrogen supply. The possibility of interactions between amino acid uptake and the acquisition of other nitrogenous compounds in the environment is also discussed. For example, nitrate reductase formation is suppressed by ambient ammonia in some algae (Epply, Coatsworth, and Solorzano, 1969). Con- versely, amino acids are known to suppress and desuppress nitrate reductase syn- thesis in some higher plants (Filner, 1966). Finally, data concerning uptake of dicarboxylic and polybasic amino acids and their suitability as a nitrogen source are presented. MATERIALS AND METHODS The original culture of Plat\monas, the organism employed, was obtained from H. A. Lowenstam (California Institute of Technology). The culture medium contained 2 X 10 3 or 2 X 10~ 4 g-at N/liter (as nitrate, glycine, glutamate, or arginine), 2 X 10" 4 moles/liter K 2 HPO 4 , 1Q- 5 moles/liter FeCl 3 , and 0.001% EDTA in artificial seawater (Instant Ocean, Aquarium Systems Inc., Ohio). Sterile liquid cultures were inoculated from axenic agar slants. Cultures were maintained at 20 C with cool-white fluorescent lights (10 4 ergs/cm 2 /sec). Con- 1 This work was supported by ONR Grant N00014-66 C0326 and NSF Grant GB 17263. 242 AMINO ACID UPTAKE IN PLATYMONAS 243 tinuous culture apparatus as described in North and Stephens (1969) was also used for one series of experiments. Cells grown on 2 X 10~ 3 g-at N/liter ("high" nitrogen cultures), were har- vested in late log phase for uptake experiments. Cells supplied with 2 X 10~ 4 g-at X/liter ("low" nitrogen cultures) were harvested in late log or early sta- tionary phase of the growth curve. It should be emphasized that the difference 500- LU t 100- U >- 50- \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ N \ \ 600 700 800 900 CELL NITROGEN FIGURE 1. Glycine uptake vs. cell nitrogen content for Platymonas from continuous culture. Nitrogen content varied in response to changes in dilution rate of the culture medium ; abscissa : /ug N/10 8 cells; ordinate : moles glycine X 10 9 /hr/10 8 cells; (O) are cells from steady-state growth conditions. The vertical bar delineates 2 standard errors on each side of the regres- sion line. between high and low nitrogen cultures is qualitative in this work, and the initial concentration of nitrogen in both cultures is considerably higher than levels in the sea. For amino acid uptake measurements, cells were harvested, washed, and resus- pended at 5 X 10 5 cells/ml in Millipore-filtered artificial seawater. C 14 amino acids were added to give 1 /*c/20 ml of cell suspension. For glycine, arginine, and glutamate, this produced substrate concentrations of 2 to 4 X 10~ 7 moles/liter. The amino acid analogues a-amino isobutyric acid (AIB), 1 -amino- 1-carboxy-cyclo- 244 BARBARA B. NORTH AND GROVER C. STEPHENS pentane (cycloleucine), and the amino acid taurine, had lower specific activities, and minimum substrate concentrations were about 10~ B moles/liter (cj. Table II). Higher concentrations were achieved with addition of C 12 amino acids. To determine uptake rates, 1 ml aliquots were taken periodically. The cells in the aliquot were collected on filters (0.45 /A, EHWP, Millipore), washed with 10 ml artificial seawater, and the radioactivity in the cells counted. Counts/ minute (cpm) in the first aliquot and the time of sampling (1 or 2 minutes) were treated as blank values and subtracted from all subsequent samples. Uptake rates were also measured by following the disappearance of C 14 from the seawater. Radioactivity in 1 ml aliquots of seawater was counted after removing cells by filtration. All samples were counted in a Beckman CPM- 100 liquid scintillation system, corrected for quenching and background, and cpm converted to moles. Amino acid uptake rates were obtained from short term observations during which rates remained linear with time. Uptake velocities were calculated by the weighting method of Wilkinson (1961) : for Y = Vx (Y = moles taken up at time x at velocity F), the standard error (S.E.) of Y was assumed to be proportional to Y because of volumetric sampling errors. That is, S.E. (Y ) = KY where K is a constant per cent error in sam- pling. This provides a weighting factor of 1/F 2 for calculation of the velocity V and the S.E. of V. Uptake velocities were determined at several substrate concentrations. From such data, North and Stephens (1967) have shown that glycine accumulation in- volves a saturable step and can be adequately described using the Michaelis-Menton equation, V V m& x(S/S + K s ) where V uptake velocity, F max = maximum up- take velocity, S = substrate concentration, K s = substrate concentration at which the velocity is half-maximal. An unweighted V vs. V /S linear transformation of the Michaelis-Menten equation was used to evaluate the kinetic parameters K s and F m ax (cf- Figure 2). This transformation was employed in preference to the more commonly used Lineweaver-Burk plot (1/F versus l/S) because it provides more accurate estimates of K s and F max (Dowd and Riggs, 1965). Cells were extracted with cold 5% trichloroacetic acid (TCA) to provide a mea- sure of C 14 assimilation. This solubilizes amino acids and various small organic compounds but does not extract nucleic acids, lipids, polypeptides, or protein. In experiments with C 14 amino acid analogues, cells were extracted with cold TCA, 80% ethanol, and 0.66 N perchloric acid. Insoluble and total radioactivity were counted directly, and soluble activity calculated by difference. Cell nitrogen con- tent was measured with a semi-micro Kjeldahl technique (Steyermark, 1961) in continuous culture experiments and with a Coleman nitrogen analyzer for other experiments. RESULTS Amino acid uptake velocities increased when nitrogen supply in the culture medium was lowered. This stimulation occurred in cells harvested from both batch and continuous culture. A logarithmic increase in rate of glycine uptake was correlated with a decrease in nitrogen content in populations from continuous culture (Fig. 1). A similar rate increase occurred in batch culture cells (Table I). AMINO ACID UPTAKE IN PLATYMONAS 245 K H a s s I -^ ^l ^i s o s So Pu >> -a . S umulation j O 1 3 > s O iO rt 1 O X 00 ^^ ^^ ^^ 4_) p 5 u O O u CS C 11 O z s^ -v u OO O Is <-M O ~) 00 \Q t~** oO ^5 LO GO fO OO ssgsssss^sssss 4-) u ^ u H / "" 1 " S -s ' "3 o -t o ^ ^ -f -f ^ f--l O "0 Ol vO > i OO OO ^1 1^- -f je x '/i *7 5 s. o L-i ^ r^i O ^H O rvlO^O^ 'lOO^OlOOOOO OJ 2 x s- -s en "3 o OJ 3 3 5 * - rt Q 3 a 'o a"o PS s_ CO o -X IA> ro f OO ON OO rf ', OO ^ -^ t O 3 a! O C "e "2.2 'o'-y 0) i i i i O O 7 7 7 7 7 7 7 7^ 7 7 7 7 7 7 o be tmino ncentr; 0) "3 g X XX X XXXXXXXXXXXXXX ^ T~I s r^ r/5 ro ro a~ 2 'o a! O 4> OJ OJ 0) ^ U *!>, E O O O O >,>,>>>,>, >,'b/.'be'be'be =233 zr s 1 a be be be be te be be b/. be be rs rt cs rs be be be be 'i '-^ /* V- U3 O 1 --\ O ~l OO rr, l-~ OO <^1 O OO rf ~. l-~ OO <^l >**. t~~ O OO A u d O o E _x ^H ^ t "' ^ "* ^ "' "' * rv, C II r^ "^^ *^" ; conditio * c.. l : 5 _> s j -C , JS S be ^ b ^ tuo ^ t/: -s br,. ;J b/. is bp & b/) b/j J r~ Q .ti i> " 3 "3 o LJ J 1 ^L 03 O. 2^1- O 4-> Q, lirl -O C3 Nitrogen s 1 iidd f H f *. __ fc_ f tm f m I* . .-sOO^N^' UW N- ^^J = 2 X 10-" g-. :e rate at a su centration. at ion time for a W - ^ ' ^ ^J r ^ __ ^J 4_l <^J i3 -* (0 O 246 BARBARA B. NORTH AND GROVER C. STEPHENS The low nitrogen cells (i.e., cells grown on 2 X 10" 4 g-at N/literj contained less cell nitrogen and acquired animo acids from solution more rapidly than high nitrogen cells (grown on 2 X 10" 3 g-at N/liter). Stimulation also ap- 2- 20 25 V/S FIGURE 2. The relation between glycinc uptake rates and glycine concentration for cells grown on 2 X 1(T 3 and 2 X 10~ 4 M NO 3 . The plot is a linear transformation of the Michaelis- Menten equation: V -K,(V/S) + V m ^; abscissa: V/S where V moles glycine X 10 8 /hr/10 8 cells and S = moles glycine X 10 T /liter ; ordinate : V = moles glycine X 10 8 /hr/10 s cells. The solid lines were determined by a least squares analysis. The dotted line encloses the 95% confidence band. peared to be independent of the form in which the nitrogen was supplied during growth. Four different nitrogen sources were employed (nitrate, arginine, gluta- mate, and glycine) each at high and low concentrations. Division rates in log phase were comparable on all substrates and at both nitrogen concentrations. AMIXO ACID UPTAKE IN PLATYMONAS 247 Cells from each culture were checked for their ability to acquire arginine, glycine, and glutamate from dilute solution. The resulting uptake patterns were comparable in each high and low nitrogen pair (Table I). In every case, rates from low nitrogen cultures were substantially higher than corresponding high nitrogen cells. In addition, arginine uptake always exceeded glycine, and glutamate rates were always lowest. This pattern was independent of the nitrogen source on which the cells had been grown. Even though low nitrogen conditions consistently increased amino acid uptake rates and decreased cell nitrogen, the overall correla- tion between nitrogen content and uptake rate found in continuous culture (Figure 1 ) was not found in these batch culture experiments. In the first experiment summarized in Table I, cells were harvested in log phase TABLE II I'ptake of amino acids and analogues by Platymonas grown on two concentrations of nitrate Culture conditions Analogue uptake Accumulation ratios ** [N0 3 ] (moles/liter) [Cell] (cells/ml) C 14 analogue Concentration (moles/liter) Uptake rate ( moles Uptake* rate x ion (moles liter inside \ moles/liter outside/ V hr 108 C ell s ) 2 X 10-" 5.9 X 10 s GLYCINE 4.7 X 10- 7 194 4128 2 X 10-* 5.9 X 10 5 A1B 114 X 10~ 7 . 372 326 254 2 X 10- 1 5.9 X 10 5 CYCLOLEUCINE 127 X 10~ 7 112 88 65 2 X 10-* 5.9 X 10 s TAURINE 218 X 10~ 7 2 X 10-3 16.8 X 10 5 GLYCINE 4.7 X 10- 7 25 532 2 X 10-3 16.8 X 10 5 AIB 114 X 10~ 7 26 23 18 2 X 10-3 16.8 X 10 5 CYCLOLEUCINE 127 X 10~ 7 2 X 1C- 3 16.8 X 10 5 TAURINE 218 X 10" 7 * Uptake rate at a substrate concentration of 10 5 moles/liter, calculated from measured rates assuming that rate is proportional to substrate concentration. ** Inside concentration calculated using packed cell volume (70 X 10" 14 liters H 2 O cell). growth when cell densities and growth rates were comparable in both high and low nitrogen cultures. Therefore, nitrogen was not growth-limiting in this ex- periment. In the second experiment, the low nitrogen cultures were in the early stationary phase, were less dense than the high nitrogen cultures, and presumably had exhausted the available nitrogen. Nonetheless, amino acid uptake was stimu- lated in both experiments. The maximum velocity of glycine uptake (F ma x) and the concentration at which the system mediating uptake was half-saturated (K s ) were evaluated for high and low nitrogen cells. Glycine uptake rates were measured at ambient concentrations ranging from 2 X 10~ 6 to 5 X 10~ 5 moles/liter for low and high nitrogen cultures (Figure 2). K s and F ma x were evaluated from the slope and intercept of the line. In cells from low nitrogen cultures, /" ma x = - 5181 2677 ([moles glycine X lO"/hr 10 s cells] 95% confidence interval). K,-- 19 15 ([moles glycine X 10 /liter] 95% confidence interval). In high nitrogen cul- tures, Fnmx 644 381, Kg-- 14 14. A portion of the increase in r max may be due to the larger size of cells from the low nitrogen culture. These cells may 248 BARBARA B. NORTH AND GROVER C. STEPHENS have 50% more surface area than high nitrogen cells. However, this explanation accounts for only a fraction of the difference observed in F max . Thus, amino acids were acquired from the medium more rapidly by cells grown on restricted nitrogen, whether in continuous culture or in fixed volume cultures. It was possible that stimulation of uptake resulted from more rapid utilization of amino acid within the cells (i.e., assimilation), particularly since glycine is rapidly converted into an array of alcohol-soluble compounds (North and Stephens, 1967). This could stimulate uptake rates without enhancing the ability of the cell to accu- mulate amino acids against an internal concentration gradient. Accumulation of 60- 40- 0. u 20- TOTAL CPM 10 20 30 40 MINUTES 50 X TCA INSOL TCA SOL. . TOTAL CPM~ X TCA INSOL TCA SOL. 60 10" 4 M NO, 10" 3 M FIGURE 3. Incorporation of glycine-C 1 * into TCA-soluble and TCA-insoluble fractions of Platymonas grown on 2 X 10~ 3 and 2 X 10~ 4 M NO 3 ; ( ) represents total CPM in cells; (X X) CPM in TCA-insoluble fraction; (O O) CPM in TCA-soluble fraction. C 14 taurine and C 14 amino acid analogues by cells grown in high nitrogen and low nitrogen batch cultures was compared. Taurine was not acquired by the cell under any of the conditions employed. Alpha-aminoisobutyric acid (AIB) and 1 -amino- 1-carboxy-cyclopentane (cycloleucine), however, are accumulated much more rapidly by nitrogen deficient cells (Table II). AIB-C 14 and cycloleucine-C 14 in the cells were readily extracted with 80% ethanol, verifying that these analogues were not metabolized into insoluble compounds. In another experiment, AIB was extracted quantitatively from Platymonas and identified by autoradiography of a thin layer chromatogram of an extract (perchloric acid). The ratio of internal to external concentration of these analogues was high at the end of the experiment (Table II) and indicated accumulation against a substantial concentration differ- AMINO ACID UPTAKE IN PLATYMONAS 249 ence. Larger ratios would have been achieved had the experiment continued since uptake was still linear at the end of the observations. Thus, uptake stimulation is accompanied by increased rates of accumulation into the cell. The rate at which glycine radioactivity was assimilated into the cold TCA- insoluble fratcion of cells from low and high nitrogen cultures was also investigated. The amino acid C 14 obtained from the medium entered synthetic pathways more rapidly in the low nitrogen cells (Fig. 3). No conclusions can be drawn about absolute rates of amino acid assimilation without further information about the size and specific activities of amino acid pools. However, the data indicate that ambient amino acids were incorporated into cells of either history. The rates of arginine uptake were unaffected by the presence of nitrate or ammonia at levels 20 to 1000 times more concentrated than the amino acids (Table III). A similar result was obtained with glycine-C 14 . Thus, there is no short term interaction at levels of inorganic nitrogen as high as 10~ 4 moles/liter. TABLE III Arginine uptake by Platymonas in the presence of nitrate and ammonia. Cells were incubated in 1.5 X 10~~' M arginine- C 14 for 10 minutes % radioactivity remaining in medium at 10 minutes High nitrogen culture Low nitrogen culture control 75.2 28.6 10- 5 M NO 3 76.5 28.7 10-* M NO 3 75.5 29.4 10~ 5 MNH 3 75.9 28.1 10^ M NHs 78.2 28.7 Of course, our earlier results indicate that supplying nitrate at levels around 2 X 10~ 3 moles/liter would eventually depress amino acid uptake rates, but the time course of this effect has not been investigated. However, such levels far exceed any natural concentrations that might be encountered by Platymonas. DISCUSSION We are investigating the significance of natural dissolved amino acids to the nutrition of Platymonas. The fact that arginine, glutamate, and glycine can pro- vide nitrogen for growth in batch culture is neither necessary nor sufficient to argue for a normal nutritional role of dissolved amino acids. An amino acid or other organic compound in the environment may supplement other nitrogen sources or compensate for a particular synthetic deficiency without being a complete nitro- gen source. In addition, an amino acid that supports growth at high concentra- tions may have no importance in the sea. Nonetheless, neutral, dicarboxylic, and polybasic amino acids are all complete nitrogen sources for Platymonas. In gen- eral, utilization of amino acids supplied at the high concentrations characteristic of batch cultures appears to be widespread among phytoplankters (Danforth, 1962}. A more cogent approach to the nutritional role of dissolved amino acids is to study the utilization of these compounds when supplied at concentrations reported 250 BARBARA B. NORTH AND GROVER C. STEPHENS from natural waters. The work reported here and in earlier contrihutions from our laboratory supports the following statements: (1.) Amino acids are accumu- lated by Platyuionas from very dilute solutions, concentrations of which are in the range reported for natural waters. (2.) Accumulated amino acids enter respira- tory and synthetic pathways in the cell. (3.) Uptake continues for long periods (demonstrated for glycine). (4.) The uptake rate depends on the amino acid and the nutritional history of the cells. Rates are higher in cells which have been grown under conditions where nitrogen is restricted though not to the extent that it limits the growth rate. (5.) When cells are stimulated to take up amino acids more rapidly, the pattern of uptake is constant regardless of the nitrogen source on which they were grown. The uptake rate of arginine is always greater than glycine, and glutamate is consistently lowest. (6.) The stimulation of amino acid uptake is accompanied by increased rates of accumulation, as shown by accelerated uptake of amino acid analogues. (7.) Cells grown at reduced nitrogen concen- trations show an increase in the maximum velocity (F ma x) of amino acid uptake. No change in K s is apparent. (8.) Amino acid uptake proceeds in the presence of nitrate and ammonia. The conditions that increase amino acid uptake in the laboratory might be expected to occur under natural conditions. We have previously shown ( North and Stephens, 1969) that restriction of nutrient levels in continuous culture pro- duced a marked increase in rates of amino acid uptake. In the present report, increased rates were also obtained when nitrogen supply was reduced in batch cultures. North, Stephens and North (in press) have also demonstrated com- parable increases in uptake rates for various amino acids in several other marine phytoplankters. There are other reports of amino acid uptake by phytoplankton. A light- dependent glycine transport system has been reported for the freshwater alga, Sccncdesnins (Cseh and Szabo, 1965). Hellebust and Guillard (1967) report rapid uptake of glutamate by the diatom, Melosira. Glutamate uptake is also reported in a study of four phytoplankters by Sloan and Strickland (1965) though they conclude that the observed rates are too low to have any nutritional sig- nificance. In Platynwnas, glutamate uptake is also quite slow compared to glycine and arginine. Even though growth on restricted nitrogen increases glutamate uptake ten-fold, rates are too low to supply a significant amount of nitrogen (see below ) . Culture conditions may be responsible for some of the differences found in the literature. Other nitrogen-sensitive transport systems have been reported in algae. Yaden ( 1965 ) discusses increases in glucose and amino acid uptake in Eui/lcna. Uptake was stimulated by heterotrophic growth in the dark, starvation in the dark, or streptomycin bleaching. Nitrate (Eppley and Thomas, 1969) and ammonia (Fitz- gerald, 1968; Syrett, 1962) uptake systems may also respond to nitrogen depriva- tion. The former authors report an increase in F max for nitrate uptake in nitrogen- deficient cells. Although our data justify the conclusion that /' n mx increases in batch culture cells grown under low nutrient conditions, the limitations of the batch culture technique are severe. We can neither control nor predict the growth con- ditions which will give reproducible values for a particular kinetic parameter in such cultures. Thus, there is no overall correlation between uptake rates and any AMINO ACID UPTAKE IN PLATYMONAS 251 of the usual culture parameters such as growth rate, nitrogen content, cell concen- tration, etc. This is not surprising since the conditions in hatch cultures are changing continuously and rapidly. Even though some data reported here suffer the limitations of hatch culture technique, the contribution of amino acid uptake to the nitrogen economy of the cells was evaluated. Nitrogen uptake velocities (from a solution of 5 X 10~ 7 moles/liter) were calculated for glutamate, glycine, and arginine for cells from high and low nitrogen cultures. Uptake rates, when divided by cell nitrogen content, are equivalent to growth constants in terms of nitrogen ( Dugdale and Goering, 1967) ; that is, the constants represent growth rates or generation times (generation time == In 2/growth rate) assuming that growth is limited by amino acid nitrogen. The calculated constants (Table I) indicate that glycine might supply sufficient nitrogen to support growth at generations as short as 160 hours. Arginine uptake is consistently faster than glycine, and the greater nitrogen con- tent of the molecule could support a generation as short as 17 hours. These calcu- lated generation times cannot be compared directly with generation times under natural conditions, since the latter have not been measured. One can speculate that natural doubling times might well be ten times longer than generation times observed in batch culture. However, if calculated generation times exceed batch culture growth by fifty-fold, the substrate concerned probably cannot contribute significantly to the economy of the organism under natural conditions. Since the shortest generation time observed in batch culture of Platymonas \vas 22 hours, we conclude that glycine and arginine may be important nitrogen sources in the sea. Uptake of glutamate was slower than both arginine and glycine and could support a generation time of 886 hours. This does not represent a significant nitrogen contribution under any of our conditions (although glutamate was a com- petent nitrogen source in batch culture). Again the limitations of batch culture should be emphasized. Caperon (1969) suggests that nitrogen limited cell growth should be described by a response function which incorporates both the nutritional history of the cell and the current growth conditions. Such a response function is undoubtedly required to describe behavior of the cell populations in batch culture. Until this is done, kinetic parameters obtained from such experiments should probably not be used to predict behavior of natural phytoplankton populations. This difficulty can be eliminated if cells are grown in continuous culture and allowed to reach a steady state. Under steady-state conditions, the cells have a constant and repro- ducible nutritional history, and well defined growth parameters. When generation times calculated from amino acid uptake velocities and nitro- gen content are compared with measured generation times of the steady-state cell populations, a more defensible estimate of the nutritional significance of dissolved amino acids is achieved (cf. Figure 1). At moderate growth rates (generation time = 51 hours) and natural glycine concentrations (5 X 10~ 7 moles/liter), the uptake system can supply sufficient nitrogen to support a calculated doubling time of 860 hours. However, at the slower growth rates achieved by reducing the nutrient supply (generation time =- 166 hours), the glycine contribution becomes more important. Under these conditions, uptake rate increases and cell nitrogen decreases as in low nitrogen batch cultures. Glycine uptake could support a gen- 252 BARBARA B. NORTH AND GROVER C. STEPHENS eration of 95 hours. This assumes that amino acid uptake continues unabated. Sustained uptake has been demonstrated for high nitrogen cells (North and Stephens, 1969), but similar experiments have not been performed on nitrogen- deficient cells. These calculations do not correct for leakage of nitrogen-containing compounds. However, such losses are probably small. Hellebust (1965) reports that Tetra- selmis (Platymonas) leaks 1-3% of its photo-assimilated carbon, depending on light conditions. If the excreted product is entirely protein, this leakage rate adds a maximum of 5% to the cell nitrogen requirement. The amino acid molecule also provides reduced carbon to the cells. The accu- mulated amino acid can be expressed as a fraction of respiratory rate by calculat- ing the oxygen required for complete oxidation of accumulated amino acid and comparing this figure to oxygen consumption in the dark. Dark respiration of Platymonas, as determined in a Gilson respirometer, ranges from 1.3 to 3.5 /Ainoles O,/hr/10 8 cells. Populations from continuous cultures can obtain 0.5 //.moles glycine/hr/10 8 cells. Complete oxidation of this amount of glycine requires about 0.9 jwmoles of O 2 , or 25-67% of the total respiration. This is not to say that Platymonas can grow exclusively on amino acid carbon in the dark, even though amino acid uptake is not light-dependent (North and Stephens, unpublished results). On the contrary, Platymonas appears to be an obligate phototroph (Gibor, 1956). Nonetheless, the utilization of amino acids or other organic substrates reduces the photosynthetic carbon requirements and may be important to algae in certain habitats (Bunt, 1969). Though there is no short-term effect on amino acid uptake by high concentra- tions of nitrate or ammonia, we have not examined the possibility of a reciprocal situation, i.e., the affect of amino acids on the utilization of other nitrogen sources, particularly nitrate. Such investigations may be important in view of recent eco- logical emphasis on the characteristics of nitrate uptake. Dugdale and Goering (1967) suggest using uptake velocities for nitrate and ammonia to measure nitro- gen flow in primary productivity studies in situ. K K and F m ax have been used to analyze and predict interspecies competition and succession (Dugdale, 1967; Eppley, Rogers and McCarthy, 1969). Finally, a nitrate reductase assay in the field has been proposed as a sensitive measure of rates of nitrate assimilation and the presence of nitrate (Eppley, Coatsworth, and Solorzano, 1969). Interactions between nitrite, nitrate, and ammonia assimilation were examined and taken into account in the above studies. The effect of amino acids on these systems is un- known, however, and these authors have emphasized the need for such investiga- tions. We propose to study the effect of low concentrations of amino acids on nitrate reductase content of steady-state populations. SUMMARY The marine phytoplankter, Platymonas, increases rates of amino acid uptake when grown on a restricted nitrogen supply. Uptake of glycine, arginine, and glutamate increases as much as 10-fold when cells are deprived of nitrogen during growth. In a given culture, arginine uptake is always faster than glycine, and AMINO ACID UPTAKE IN PLATYMONAS glutamate accumulation is always slowest. Nitrogen-deprivation also stimulates accumulation of non-metabolizable amino acid analogues. Kinetic studies suggest that [ 7 max increases and 7v s . remains unchanged in N-deprived cells. When low concentrations of glycine or arginine are supplied to cells grown on restricted nitrogen, uptake rates are sufficient to fulfill the nitrogen requirement in Platy- monas. High concentrations of nitrate or ammonia do not interfere with amino acid uptake. Thus amino acids in the ocean may be an important nitrogen source for some phytoplankters. LITERATURE CITED BUNT, J. S., 1969. Observations on photoheterotrophy in a marine diatom. /. Phycol., 5: " 37-42. CAPEROX, J., 1969. Time lag in population growth response of Isochrysis galbana to a variable nitrate environment. Ecology, 50: 188-192. CHAU, Y. K., AND J. P. RILEY, 1966. The determination of amino-acids in seawater. Deep Sea Res., 13: 1115-1124. CSEH, E., AND E. SZABO, 1965. The glycine uptake of Scenedesnnts obtnsiuscnlus. Annal. Biol. Tihany. Hinuiaria. 32: 247-253. DANFORTH, W. F., 1962. Substrate assimilation and hu'.erotrophy. Pages 99-123 in R. A. Lewin, Ed., Pliysiolof/y and Biochemistry of Algae. Academic Press, New York. DEGENS, E. T., 1968. Molecular composition of nitrogenous compounds in seawater. Woods Hole Occanogr. hist. Ref. No. 6-52: 1-27. DOWD, J. E., AND D. S. RIGGS, 1965. A comparison of estimates of Michaelis-Menten kinetic constants from various linear transformations. /. Biol. Chan., 240: 863-869. DUGDALE, R. C., 1967. Nutrient limitation in the sea : dynamics, identification, and significance. Limnol. Occanogr., 12: 685-695. DUGDALE, R. C., AND J. J. GOERING, 1967. Uptake of new and regenerated forms of nitrogen in primary productivity. Limnol. Occanogr., 12: 196-206. EPPLEY, R. W., J. L. COATSWORTH AND L. SOLORZANO, 1969. Studies of nitrate reductase in marine phytoplankton. Li unto I. Occanogr.. 14: 194-205. EPPLEV, R. W., J. N. ROGERS AND J. L. MCCARTHY, 1969. Half-saturation constants for uptake of nitrate and ammonium by marine phytoplankton. Limnol. Occanogr., 14: 912-920. EPPLEY, R. W., AND W. H. THOMAS, 1969. Comparison of half-saturation constants for growth and nitrate uptake of marine phytoplankton. /. PliycoL, 5: 375-379. FILNER, P., 1966. Regulation of nitrate reductase in cultured tobacco cells. Biochim. Biophys. Ada. 118: 299-310. FITZGERALD, G. P., 1968. Detection of limiting or surplus nitrogen in algae and aquatic weeds. /. Phycol.. 4: 121-126. GIBOR, A., 1956. The culture of brine algae. Biol. Bull.. Ill: 223-229. HELLEBUST, J. A., 1965. Excretion of some organic compounds by marine phytoplankton. Limnol. Occanogr., 10: 192-206. HELLEBUST, J. A., AXD R. R. L GUILLARD, 1967. Uptake specificity for organic substrates by the marine diatom, Melosira nummuloidcs. J. Phycol., 3: 132-136. HOBBIE, J. E., C. CRAWFORD AXD K. L. WEBB, 19h8. Amino acid flux in an estuary. Science, 159: 1463-1464. NORTH, B. B., AXD G. C. STEPHENS, 1967. Uptake and assimilation of amino acids by Platy- moiias. Biol. Bull., 133: 391-400. NORTH, B. B., AND G. C. STEPHENS, 1969. Amino acids and Platymonas nutrition. Pages 263- 273 in R. Margalef, Ed., Proceedings of Vlth Int. Scazveed Symposium. Subsecretaria Mercante Marina, Madrid. NORTH, W. J., G. C. STEPHENS AND B. B. NORTH, in press. Marine algae and their relations to pollution problems. Proc. FAO Tech. Conj. on Marine Pollution, Rome, 1970, in press. 254 BARBARA B. NORTH AND GROVER C. STEPHENS SLOAN, P. K., AND J. I). H. STRICKLAND, 1966. Heterotrophy of four marine phytoplankters at lo\v substrate concentrations. /. Pliycol., 2: 29-32. STEYERMARK. A., 1961. Nitrogen: micro-kjeldahl method. Pages 188-220 in Quantitative Or- ganic Microdiuilysis. Academic Press, New .York. SYRETT, P. J., 1962. Nitrogen assimilation. Pages 171-188 in R. A. Lewin, Ed., Physiology and Biochemistry of Algae. Academic Press, New York. WEBB, K. L., AND L. WOOD, 1967. Improved techniques for analysis of free amino acids in sea- water. Pages 440-444 in N. B. Slova, Ed., Automation in .-innlytii' Chemistry. Tcch- nicon Symposium. Mcdiad, New York. WILKINSON, G. X., 1961. Statistical estimations in enzyme kinetics. Biochcm. J., 80: 324-332. YADEN, S., 1965. Uptake of organic compounds by Euglcna gracilis. Ph.D. thesis, Univer- sity o\ Minnesota, 33 pp. Reference : Biol. Bull, 140: 255-273. (April, 1971 ) BEHAVIORAL AND ELECTROPHVSIOLOGICAL STUDIES OF HYDRA. I. ANALYSIS OF CONTRACTION PULSE PATTERNS. 1 NORMAX B. RUSHFORTH Department of Biology and Department of Biometry, Case Western Reserve University . Cleveland, Ohio 44106 A characteristic feature of the behavior of some hydroid polyps is that their spontaneous activity consists of rhythmically re-occuring behavioral events. Ex- amples are: the concerted oral movements of the tentacles followed by peristalsis of the proboscis in Tubularia (Josephson and Mackie, 1965) ; hydranth contraction of Obclia (Morin, 1969) ; stalk contraction in isolated polyps of Corymorpha (Rushforth, unpublished observations) ; and contraction of the body column and tentacles in Hydra (Passano and McCullough, 1963). In these four hydroids electrical potentials are associated with the periodic movements. The potentials are relatively large, on the order of several millivolts when externally recorded. Large spontaneous potentials and associated periodic movements, however, are not universal in hydroids. Neither rhythmic behavior nor endogenous events have been observed in Pcnnaria (Fred Delcomyn, University of Oregon, personal communication) and Hydractinia (Darrell Stokes, University of Hawaii, personal communication). In Cordylophora, although very slow rhythmic behavior occurs (Fulton, 1963), the animal is electrically silent unless fed or stimulated (Mackie. 1968). At the present time, no hypotheses have been proposed to explain the existence of these divergent classes of hydroid behavior. The behavior and electrical activity in Hydra Passano and McCullough (1962) were the first investigators to record elec- trical correlates of spontaneous behavior in Hydra. In a series of papers they described the properties of two endogenously active coordinating systems in this animal (Passano and McCullough, 1964, 1965; McCullough, 1965). One system termed the Rhythmic Potential System, produces pulses (RP's) rhythmically with frequencies ranging from 1 to 10 per minute. These pulses are small, relatively rapid, compound potentials on the order of tens of microvolts when recorded externally. While potential RP pacemakers are located throughout the whole column, RP's usually begin in the basal region of the animal. The pulses are conducted in the column at about 4 cm per sec. Such rhythmic potentials trigger gastrodermal muscle contraction, causing elongation of the body column (Shibley, 1969). A second coordinating system produces large, slow potentials when recorded with external electrodes. The pulses are of up to 500 msecs in duration and have amplitudes up to several tens of millivolts. The system was designated the Contraction Burst System by Passano and McCullough (1964), since the pulses 1 Supported in part by grants MH-10734 and GM-12302 from the National Institutes of Health. 255 256 NORMAN B. RUSHFORTH were associated with contraction of the epidermal musculature and occurred usually in hursts. On the basis of long-term recordings (lasting 10 to 12 hours) from five species of hydra, Rushforth (1966) found that in many animals the electrical pulses and body contractions occur primarily as single, widely spaced events rather than in bursts. In view of this finding the potentials were termed Contraction Pulses (CP's) rather than Contraction Burst Potentials. This term has been used subsequently by Rushforth (1967a) and others (Josephson, 1967, Josephson and Macklin, 1967, 1969). In retrospect, it would have been more appropriate to have called these potentials Column Contraction Pulses to dis- tinguish them from contraction pulses recorded from the tentacles of the animal (Rushforth, 1967b). However, in order not to add another term to the rapidly proliferating nomenclature of pulses recorded from hydroids, the term Contrac- tion Pulses (CP's) will be retained in the present study. The pulses are equiva- lent to those designated as Contraction Burst Potentials by Passano and his co-workers. The present study is a quantitative description of the temporal patterns of Contraction Pulses in two species of Hydra: the large Belgian species H. pirardi Brien and the North American variety H. pseudoligactis. It is likely that H. pseudoligactis is synonymous with H. canadensis which deserves priority ( Adshead, Mackie and Paetkau, 1963). However, the animals have been cultured for several years as H. pseudoligactis and this name will be continued to be used until it can be shown that the species is the same as that Rowan described as H. canadensis (Rowan, 1930). These species were chosen since they represent the extremes on a continum of behavior and electrical activity of ten species of hydra thus far investigated. The contraction pulse activity of H. f>irardi consists predominately of bursts of potentials while that of H. pseudoligactis consists almost exclusively of single events. Under conditions of external stimulation, however, the electrical activity of the latter species may be changed to give a temporal pattern of bursts of potentials similar to that of H. [>irardi. This investigation is a first step in an effort to construct testable models for the production of potentials by pacemaker systems in hydroids. The current study was undertaken to quantify the characteristics of inter-pulse intervals in two selected species of Hydra. The characteristics of the interval measurements will be compared in subsequent studies with those observed for other hydroids, and with those of intervals generated using theoretical models. MATERIALS AND METHODS The conditions employed for culturing the hydra were similar to those de- scribed by Loomis and Lenhoff (1956). Animals were raised in a medium con- sisting of 'glass distilled water to which 1.5 X 10' 5 M Ca C1 2 , 1.2 X 10~ 3 M NaHCO,, and 1.2 X 10~ 4 M Na 4 EDTA had been added. The hydra were fed daily with a plentiful supply of freshly hatched Artemia salina nauplii. The temperature was held at 21 1 C and the animals were reared in constant illumination, approxi- mately 30 foot candles. The electrical recordings were from hydra randomly selected from the culture dishes 24 hours after feeding. Continuous recordings of the spontaneous electrical CONTRACTION PULSE PATTERNS IN HYDRA 257 activity of individual hydra were made using a glass suction electrode (Josephson, 1967). The individual recording periods were 10 to 12 hours in duration. The electrode has a bell-shaped opening 0.4-0.6 mm in internal diameter about 1 mm from the open end. The other end of the electrode is attached to a syringe by means of flexible tygon tubing and is held in a micromanipulator. The hydra is lightly sucked into the open end of the electrode, and sits with its base in the bell-shaped opening (Fig. 1). The electrode is filled with culture solution and contains a silver wire inserted through the tubing to increase the conductivity. Electrical contact in this system is made between the tissue of the hydra base, the culture in a con- FIGURE 1. Hydra placed in the recording electrode: upper left: H. firardi , ^.,- tracted state ; upper right : II. [>inirdi in an elongated position ; Icnver : //. t>scudt>Ii<~/actis in a normal relaxed position. medium and the silver wire. A reference electrode consists of a coil of silver wire approximately 5 cm from the recording electrode. Electrical potentials from the preparation are amplified using capacito.r-coupled amplifiers having long time constants, with the final display on an oscilloscope and penwriter. Both electrodes are placed in 200 ml of culture solution in a finger bowl. The recording electrode is positioned so that the hydra touches neither the bottom of the dish nor the upper surface film with its tentacles. It was found that increased suction on the base of the hydra in the electrode affected both the frequency and temporal patterning of pacemaker activity. There- fore, during the recording periods the suction was kept as low as seemed feasible. Occasionally because of the light suction on the electrode, the hydra would pull 25.x NORMAN B. RUSHFORTH its base out of the electrode during the recording period. It was found possible to feed hydra attached to the electrode and have them produce buds over several days while still recording their pacemaker activities. Visual observations using a dissecting microscope were made of the overt be- havior of the hydra while simultaneously monitoring its electrical activity. Col- umn contractions were manually recorded during phases of the electrical record- ings using a telegraph key which activated a channel of the penwriter. Such visual observations were made at the start and termination of a 12 hour record of 40 uv 600 uv IOSECS / 1 < k i n / v r- 1 | \ I \ 1 1 \ \ \ II -- p \ \ v*"^ \ \ \ \ \ 1 l_ I T II I I 1 1 si - p 1 ' SI t 1 LI TR T LI SI SI R SI I 1 FIGURE 2. Electrical records of hydra. Upper two records : //. pseudoligactis, lower two records : //. Pirardi, for definition of the symbols see the text. The positive polarity is in the upward direction in this and other figures of electrical recordings. electrical activity to ensure correspondence between the recorded pulses designated as CP's and column contractions of the animal. In order to distinguish between the behavioral event of column contraction and the associated electrical event the following terms are used : ( 1 ) a column con- traction designates a single contraction of the body column which may be visually observed, or recorded using a transducer as described by Josephson (1967); (2) a contraction burst is a behavioral event consisting of a series of individual column contractions, each contraction in the series separated by less than 20 seconds from a subsequent or preceding column contraction in the burst. At the termina- CONTRACTION PULSE PATTERNS IN HYDRA 259 tion of a contraction burst the hydra appears as a tight ball with its column fully contracted; and (3) a Column Contraction Pulse (designated here as a Con- traction Pulse, or CP ) is a single electrical potential preceding a column contrac- tion by 0.1 to 0.2 seconds. The electrical correlation of a contraction burst is a burst of Contraction Pulses (a CP burst). Measurements were made from the penwriter records of the intercontraction pulse interval lengths to an accuracy of 0.4 seconds. Both inter-pulse interval histograms and joint interval scatter diagrams were used to characterize the dis- tributions of the inter-pulse intervals. These methods are described later in the results section. I- lJ O o: LJ 50 100 150 200 250 300 INTERVAL LENGTH (SECS.) 350 FIGURE 3. Inter-pulse interval histogram for //. piranii. The intervals are from a 10 hour record from a single animal. \Yith H. pseudoligactis, the CP system was sometimes induced to fire pre- dominately in bursts rather than as single CP's by: (1) increasing the electrode suction; (2) giving intermittant light stimulation; or (3) exposing the animal to live Artemia salina nauplii. The CP pattern was determined for three different levels of suction : 0, 10. 20 cm of water, as measured by a manometer attached to the tygon tubing of the recording electrode. Intermittant light stimulation con- sisted of successive exposures of the hydra to two minutes of "strong light" and two minute periods of ambient light. The ''strong light" was supplied by an American Optical Lamp, Model 353 which housed a G. E. bulb, model 1493, operated at 6 volts. The bulb was placed 15 cm from the preparation and light 260 NORMAN B. RUSHFORTI1 \vus directed through a heat filter and the culture solution onto the whole hydra. The CP activity was monitored during H hour period of interniittant light stimulation and compared with that for a similar period following the termination of such stimulation. Contraction I'ulses were recorded from unstimulated hydra and their temporal patterns compared with those attained after the animal had ted to satiation on artemia. The nauplii were introduced one at a time into the medium ahove the hydra, creating minimal mechanical disturbance to the preparation. CO o LJ CO O o: LJ O I- 0_ 50 CO 40 30 20 50 100 150 200 250 INTERVAL LENGTH (SECS.) 300 350 FlGUKK 4. Inter-pulse interval are troni a 10 histogram tor //. pscudoligactis. hour record Irom a single animal. 400 450 The intervals RESULTS Temporal patterns of spontaneous Contraction Pulses Inter-species differences. The temporal patterns of spontaneous Contraction I'ulses were markedly different in the two species. Figure 2 shows a 140 second portion of a continuous record from a single //. pseudoligactis (tipper pair o! records), and a //. pirardi (lower pair of records). The upper channel of both pairs of records has greater sensitivity and was vised to record both the CT'> which are the large potentials, as well as Rhythmic Potentials (RP's) and Tentacle Contraction Pulses (TCP's) which are much smaller potentials. These smaller pulses are clearly evident in the record of H. pirardi. At the sensitivity used to record RP's and TCP's considerable truncation of the CP amplitude occurs in the penwriter records. Thus, the second channel was used at a lower sensitivity to record undistorted CP's unaccompanied by the smaller potentials. CONTRACTION PULSE PATTERNS IN HYDRA 261 H. pseudoligactis gave CP's predominantly as single widely spaced events. While //. pirardi gave both single pulses and hursts of potentials. When the interval lengths between successive CP's for all intervals in a 10 hour record were plotted in the form of a percentage histogram, the resulting distribution was clearly bimodal in the case of H. pirardi (Fig. 3). The lower modal class, consisting of approximately 40 per cent of the intervals for this preparation. 400 350 300 250 M O LJ CO 200 150 100 50 I 50 100 150 200 250 SEC I (I) 300 350 400 FIGURE 5. First order joint interval scatter diagram for H. pirardi. For further details see text. The preparation is the same as that for Figure 3. represents short intervals occurring within bursts, while the upper modal class (about 13 per cent of the intervals) represents longer intervals not in the bursts. The trough between the two modes was used as a criterion to classify intervals into long and short intervals in the electrical recordings. On the basis of such histograms it was found that 20 seconds was a suitable cut-off value to char- acterize intervals as belonging to the long or short class. A burst was denned as consisting of two or more successive pulses separated by short intervals (SI). An inter-burst interval (IBI) was denned as the sum of all long inter- 262 NORMAN B. RUSHFORTH v;ils (LI) between two adjacent hursts. The segment of the //. pintrdi record depicted in Figure 2 consists of seven inter-pulse intervals, five of which are short intervals while the remaining two are long intervals. The record contains two bursts and one inter-burst interval. 400 x x x 350 x x X X X X x * 300 X X 250 X X x ^_^ X X ~ X X X x v X X x x i x x x x x i i x x x x ,_,200 X* X \ X X O /x x x x x \ A Y UJ 150 x \ x X X X X "HV' , ; M 5,' J . x .1 * \ **vX)tK 100 Xx** * A v wXw XX J( v "Y * x $ x x Jx*x x x x x x x X x^3rx&^ X x X x X x x X)(X x X X x * jobix ^ r ^r ^ x x XX X$TMT> ^ i X ^t X 50 x x x JK*1 * ^ x x x x x x x x x X XX l^ x x x X ^ * x x x x X 50 100 150 200 250 300 350 400 SEC I (I) FIGURE 6. First order joint interval scatter diagram for H. pscudoligactis. For further details see text. The preparation is the same as that for Figure 4. In contrast to the histograms for H, pirardi which contained a large percentage of short intervals (about 50 per cent for the preparation shown in Figure 3), those of H. pscudoligactis did not show bimodality and had few short intervals. The inter-pulse histogram for all intervals in a 10 hour recording period for the H. pscudoligactis preparation previously described, is plotted in Figure 4. In this case only 9 per cent of the intervals are short intervals (i.e., less than 20 sees) indicating that there are few bursts for this animal. CONTRACTION PULSE PATTERNS IN HYDRA 263 The temporal patterns of CP's for the two species were compared using a second graphical procedure, a joint interval scatter diagram. This method has been used to quantitatively describe the spontaneous activity of a single neuron (Rodiech, Kiang and Gerstein, 1962). In Figure 5 the first order joint interval scatter diagram is plotted for the H. pirardi preparation. This is a graph of each successive inter-CP interval plotted as the ordinate measure against the inter-CP interval of the preceding interval as the abscissa. Thus the i + 1th interval is plotted against the ith interval, giving a graph of n -- 1 points from a record of n inter-CP intervals. As is seen for this preparation, records containing a large number of bursts give characteristic L-shaped plots. Points close to the origin constitute intervals between pulses within a burst. Those near the abscissa and parallel to it are intervals which precede bursts ; those near to the ordinate depict intervals following a burst. In contrast to this graph, the first order joint TABLE I Measurements of intervals between Contraction Pulses for H. pirardi and, H. pseudoligactis. The data are from 10-12 hour recordings for 10 hydra of each species Mean (Standard error) Measurement H. pirardi H. pseudoligactis Number of intervals/12 hr 1167 (137) 397 (30) P < 0.001 Short interval (sec) 9.0 (0.9) 6.3 (0.1) P < 0.05 Long interval (sec) 94 (14) 128 (8) P < 0.05 Per cent of intervals classified as long intervals 42 (4) 89 (3) P < 0.001 Number of bursts/12 hr 279 (29) 26 (6) P < 0.001 Number of intervals/burst 2.5 (0.3) 1.7 (0.1) P < 0.05 Mean interval/burst (sec) 9.3 (0.9) 7.0 (1.0) P > 0.10 Burst lengths (sec) 20.7 (1.4) 10.8 (1.1) P < 0.001 Inter-burst interval (sec) 151 (20) 1746 (336) P < 0.001 interval scatter diagrams for the H. pseudoligactis shows points widely dispersed and roughly symmetrically distributed about 45 line through the origin (Fig. 6). The graphs for all ten preparations of H. pseudoligactis were quite similar to that shown here, and were quite different from the L-shaped plots characterizing the burst patterns of the ten specimens of H. pirardi. The temporal patterns of the contraction potential records from the two species were further compared with measurements utilizing long and short intervals and the derived quantities of bursts and inter-bursts. A summary of these measure- ments from the ten hydra of the two species is given in Table I. Mean values together with their standard errors are presented for nine parameters of Contrac- tion Pulse activity. It is evident from this table that the overall frequency of CP's, as measured by the number of intervals/12 hours is much greater for H. pirardi. This is a result of a greater number of bursts and therefore a smaller inter-burst interval length than for PI. pseudoligactis. When the burst characteristics of the two species are compared it is noted that the burst lengths of H. pirardi are longer. This results from both a larger 264 NORMAN B. RUSHFORTH number of intervals per burst and a greater mean interval per burst for this species. Differences in the distributions for burst and inter-burst characteristics were investigated in order to determine statistical measures which most clearly distinguished between the two species. It is seen from Figure 7 that two measure- ments, the number of bursts/12 hours and the burst length may be used to separate animals of the two species into non-overlapping sets. Thus, we see that for these two species the temporal characteristics of the column contraction pace- maker potentials might be used as criteria for taxonomic classification. 500 CO cr. x 400 C\J \ CO H CO 300 cr 00 o: UJ CD 200 100 OH.PSEUDOLIGACTIS x H. PIRARDI X X 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 BURST LENGTH (SECS.) FIGURE 7. Graph of the number of brusts/12 hours and the burst length for ten prepara- tions of H. pseudoligactis (open circles) and H. pirardi (crosses). Characteristics of bursts of contraction pulses: (a) Amplitudes oj CP's ztnthin bursts. Measurements were made of some additional characteristics of the bursts of H. pirardi. There were insufficient bursts in the records of H. pseudoligactis to apply statistical analyses. However, the general relationships observed for H. pirardi, appear to apply to the records of spontaneously produced CP bursts in H. pseudoligactis, and are quite evident when animals of this species are stimulated to give bursts of potentials. An easily discernible feature of bursts of potentials is the increase in the ampli- tudes of successive CP's in a burst. Both Passano and McCullough (1964) and CONTRACTION PULSE PATTERNS IX HYDRA 265 Josephson (1967) noted that CP's are compound potentials consisting of both positive and negative phases when recorded externally. In the present study the pulses contain both positive and negative phases (and also an after potential due to capacitor discharge in the capacitor-coupled recording system). The initial positive and negative components were measured in the long-term records. In preparations of H. pirardi the amplitudes of both the negative and positive components of the potentials were found to increase for successive pulses in a burst. Except for one of the ten preparations studied, in which the amplitudes of the first two pulses in a burst were approximately the same, a successive increase in the amplitude of CP's in a burst was a constant feature of the records. In Table II, the relationship of the CP amplitudes and the position of the pulse TABLE II Amplitude of a CP and position of the CP in a group of CP's for H. pirardi. The data are from a 12 hour recording from a single animal Number of pulses per group Mean amplitude (mv) : K- !T^ ^i 3 ^ Number of groups Position of pulse in the group 1 2 3 4 1 1.93 1.44 96 2 1.77 1.46 2.49 1.80 87 3 1.81 1.42 2.53 1.76 3.18 2.44 49 4 1.88 0.82 2.53 2.06 3.17 2.90 3.64 3.53 27 in a group of CP's is given for the H. pirardi preparation previously described. Analysis of variance tests indicated that the increase in both phases of CP's within a burst is statistically significant (P < 0.05). Characteristics of bursts of contraction pulses: (b} Lengths of short intervals within a burst and long intervals within an inter-burst. Passano and McCullough (1964) observed that spontaneous CP bursts in H. pirardi lasted less than 30 seconds and consisted of from 5 to 8 pulses. They noted that the inter-CP interval lengths in a burst were shorter at the mid-point of the burst than either at the start or end of the burst. Analyses were made of the inter-pulse intervals in recordings of H. pirardi, to determine both the interval lengths of short intervals within a burst, and the lengths of long intervals within an inter-burst interval. Data were pooled for 10 animals and the mean lengths were determined for intervals at various positions in a burst and in an inter-burst. The relationship between the interval length and the position of the interval in a burst is given for bursts of various sizes 266 NORMAN B. RUSHFORTH TABLE III Interval length and position of interval in a burst for H. pirardi. The data are based on 10-12 hour recordings of each of 10 individual animals Mean interval (sec) Number of short intervals per burst Position of interval in the burst Number of bursts 1 2 3 4 5 1 10.4 860 2 10.1 9.8 789 3 9.5 7.4 8.5 597 4 8.5 6.0 5.7 6.9 392 5 8.7 5.7 5.2 5.5 6.9 302 in Table III. Although in onr recordings of H. pirardi the burst lengths were smaller (about 20 seconds, see Table I ) and consisted of fewer pulses (2 to 5 usually), the characteristic pattern of interval lengths within a burst observed by Passano and McCullough (1964) was evident. Interval lengths were shortest in mid-burst than in the early stages or at the termination of a burst. The length of the first short interval was significantly greater (P < 0.001) than the lengths of other intervals in the burst. Table IV reveals the relationship between the length of a long interval and its position in the inter-burst interval. The first long interval is significantly longer (P < 0.001) than the subsequent long intervals in the inter-burst interval. However, there is a systematic decrease in the lengths of progressive intervals in the inter-burst interval. Thus, the temporal pattern of short intervals within a burst is quite different from that of long intervals within an inter-burst interval. Modifications of CP patterns in H. pseudoligactis induced by external stimulation The previous results indicate that in H. pirardi, there are normally many spon- taneous bursts of CP's contrasted with H. pseudoligactis in which bursts of CP's are rare events in unstimulated preparations. Any models devised for pacemaker TABLE IV Interval length and position of interval in an inter-burst interval for H. pirardi. The data are from 10-12 hour recordings for each of 10 individual animals Number of long intervals per inter- burst interval IBI (sec) Position of interval in the inter-burst interval Number of inter- burst intervals 1 2 3 4 5 1 106 106 1595 2 145 93 52 983 3 205 86 66 53 246 4 244 78 59 62 45 77 5 356 88 74 70 66 58 27 CONTRACTION PULSE PATTERNS IN HYDRA 267 activitv in II\dra should include provisions for both tvpes of pulse patterns. //. pseudoligactis will produce hursts when appropriately stimulated. Effective stimuli include: (1) increased suction of the hydra in the recording electrode; (2) intermittant light stimulation of the whole preparation; and (3) exposure to stimuli inducing a feeding reaction in the animal. When H. pseudoligactis is induced to give bursts of CP's, their temporal patterns are similar to those pre- viously noted for preparations of H. pirardi. ( 1 ) The effect of increased electrode suction. In the course of long-term recordings of the spontaneous activity of several species of Hydra, it was observed that animals damaged during placement in the electrode produced many RP's. McCullough (1965) also noted an increased RP frequency in hydra placed in stressful recording situations. For animals such as H. pirardi, which normally produce many bursts of CP's, the initial effect is an elongation of the inter-CP burst interval accompanied by more pulses per burst and longer burst lengths. On the other hand, for hydra usually giving single CP's, such as H. pseudoligactis, the effect is to change the CP pattern to one primarly of burst of potentials. f " ' fl (l \ L 1 1 i \l [ L I ' ' ' 1 \5OO>uv it Tttr nrr IBI- IBI- r IB, TIT Imv IBI- IOSECS. FIGURE 8. Electrical record of H. pseudoligactis under high electrode suction (20 cm of water). In order to study the induction of bursts in H. pseudoligactis more syste- matically preparations were exposed to three different levels of suction in the recording electrode (0, 10, 20 cm of water). It was found that either inter- mediate (10 cm) or high (20 cm) pressure induced bursts of CP's accompanied by enhanced RP frequencies during inter-CP burst intervals. An example of an electrical record of a preparation at the higher pressure is given in Figure 8. Both the CP bursts and single RP's are easily discernible in the upper channel recording, and contrast markedly with the records characteristic of unstimulated H. pseudoligactis (see upper records of Figure 2). In Table V, nine measurements of inter-CP intervals are compared for the three pressure regimes. As the suction on the base of the hydra is increased there is a significant increase in the frequency of CP bursts (P < 0.01), and therefore a marked reduction in the inter-burst interval length. There is sig- nificant decrease in the per cent of intervals which are long intervals, and also a reduction in the long interval length (P < 0.01). However, while increased pressure results in more short intervals within a burst (P < 0.05), the mean inter- val per burst is not affected significantly (P > 0.10). CP bursts in H. pseudoligactis induced by increased electrode suction have the same characteristics as those of unstimulated H. pirardi: (1) both positive and 268 NORMAN B. RUSHFORTH negative phases of the CP amplitudes increase progressively with successive pulses in a burst; (2) intervals are shorter in mid-burst than at the start or end of the burst; and (3) long intervals significantly decrease in length with successive positions in the inter-burst interval. (2) The effect of intermit tant light stimulation. Several studies have shown that the effect of light on Hydra is complex, giving rise to both inhibition and excitation of pacemaker systems (Passano and McCullough, 1962, 1964, 1965; Rushforth, 1967a). When H. pirardi is stimulated with light for two minutes (about 600 ft-c), followed by two minute periods of ambient light (about 30 ft-c), the CP pattern is altered. The normal burst pattern is changed to one in which bursts occur exclusively during the two minute periods of direct illumination (Rushforth, 1967a). In contrast when preparations of H. pseudoligactis are ex- posed to this regime, CP's are inhibited during light stimulation. On direct illumination of this species, the frequencies of both RP's and TCP's are increased. TABLE V Effects of pressure on CP interval measurements in H. pseudoligactis. The data are based on a one hour recording at each pressure for each of eight animals Mean (Standard error) Measurement Low pressure Intermediate High pressure (0 cm) pressure (10 cm) (20 cm) Number of intervals/12 hr 260 (37) 438 (59) 930 (376) Short interval lengths (sec) 9.4 (1.2) 9.7 (5.8) 11.0 (0.7) Long interval lengths (sec) 201 (28) 160 (27) 130 (29) Per cent of intervals classified as long intervals 94 (8) 61 (7) 36 (10) Number of bursts/12 hr 10 (4) 62 (7) 165 (82) Number of intervals/burst 1.2 (0.2) 2.0 (0.5) 1.8 (0.4) Mean interval length/burst (sec) 10.4 (1.3) 10.8 (0.8) 11.3 (0.7) Burst lengths (sec) 11.7 (2.4) 18.1 (3.0) 20.5 (5.3) Inter-burst interval lengths (sec) 4653 (1095) 655 (97) 272 (99) During the interposed periods of ambient light, however, the CP System is in- duced to fire in bursts. For a sample of 80 periods of illumination, the CP frequency was significantly smaller (0.15 0.04 versus 5.45 0.17, P < .01) and the RP frequency significantly greater (7.15 0.19 versus 1.34 0.14, P < 0.01) compared with control periods of ambient light. An analysis of the stimulated bursts in the ambient light period showed the short interval lengths depended on their positions in the burst in a similar manner to spontaneously produced bursts in H. pirardi, or bursts in H. pseudoligactis stimulated by increased electrode suction. The intervals are shorter in mid-burst. (3) Effects of feeding. In previous studies of the contraction behavior of various species of Hydra, it was observed that feeding stimuli inhibited spontaneous contractions and those induced by external stimuli (Rushforth, 1967a). Subse- quent experiments in which both the contraction behavior and correlated electrical activity were simultaneously monitored showed that following CP inhibition during CONTRACTION PULSE PATTERNS IX HYDRA 269 feeding of an Artemia or exposure to the tripeptide reduced glutathione the CP frequency was increased significantly before returning to prestimulus levels. Dur- ing a period of approximately 1 hour of increased CP activity, the temporal pat- tern of electrical pulses of H. psendoligactis were quantified by comparing inter- pulse measurements taken before and after artemia ingestion. After feeding there are relatively fewer long intervals and considerably more bursts than before exposure to Artemia, for comparable time periods. There is a significant reduction both in the long and short interval lengths (P < 0.01), while the number of pulses per burst is significantly increased after ingestion of the artemia (P < 0.01 ). The bursts induced by feeding have the same general features as those elicited by light or by increased electrode suction, or as those spontaneously produced by H. pirardi. Short intervals are shortest in mid-burst, long intervals decrease systematically within an inter-burst, and the amplitudes of pulses increase in size either throughout the burst or in the initial part of the burst levelling off to a constant value in the late stages. DISCUSSION Passano and McCullough (1964) described several characteristics of the pacemaker system of coordinating periodic contractions of the body column in Hydra in some detail. In observations of three species (H. littoralis, H. pirardi, and H. carnca ) they emphasized that electrical output of this system consisted primarily of bursts of potentials. However, they noted that single contractions, causing only a slight shortening of the animal, comprise approximately 10 per cent of the hydra's contraction behavior. In the current study the temporal pattern of burst of potentials for undisturbed species of H. pirardi has been contrasted with the pattern found in H. psendoligactis. In this latter species unstimulated hydra give primarily single, widely spaced pulses and only occasionally produce bursts of potentials. However, when these animals are stimulated, the temporal pattern is changed to one largely consisting of bursts. Three features of the interval patterns were characteristic for unstimulated H. pirardi and for H. psendoligactis, induced to produce bursts: (1) amplitudes of pulses increased with successive positions in a burst; (2) intervals were shorter in mid-burst than at the start or termination of the burst; and (3) intervals decreased in length with successive positions in the inter-burst interval. These three features have been observed in other species of Hydra (Passano and McCullough, 1964, Rushforth, unpublished ) and appear inherent properties of the temporal CP pattern in this coelenterate. In studies of the electrical properties of the body wall of H '. o!i(/actis, Josephson and Macklin (1967, 1969) indicate that CP's are generated by the epithelial cells. They suggest that the inner membranes of the ectodermal epitheliomuscular cells are sites of CP production. Strumwasser (1967) suggested that the process gene- rating periodic bursts of spikes in a neuron could be explained on the basis of an internal oscillator. "The frequency of his internal oscillator determines the burst frequency, whereas the interaction between the amplitude of the oscillator and a fixed discharge threshold determines the length of the burst and the number of spikes in it" (Strumwasser, 1967. page 306). Thus long intervals (between 270 NORMAN B. RUSH FORT 11 bursts) and short intervals (within hursts) in Aplysia are thought to result from a single generating mechanism. In Hydra, the interval lengths of long intervals were found to have quite a different functional relationship to their position in an inter-hurst interval from that of short intervals within hursts. Both in unstimulated preparations of H. pirardi, and in H. pseudoligactis induced to give bursts of potentials by in- creased electrode suction the intervals decrease successively in length throughout the inter-burst interval. The interval immediately following a burst was signif- icantly longer than the subsequent intervals prior to the next burst. The different patterns suggest that different mechanisms seem to be involved in the production of short and long intervals. MULTIPLE EVENT GENERATOR SINGLE EVENT GENERATOR w CO o o OL O FIGURE 9. Schematic model of the CP system of Hydra. An arrow indicates excitatory input, a reverse arrow indicates inhibitory input. The following model is offered as one which could account for the observed CP patterns. The proposed model is shown schematically in Figure 9. There are two elements, one producing single events and another producing multiple events. Both elements trigger the CP conducting system and thus initiate CP's. The multiple event generator has short term excitation to itself, such excitation leads to firing in bursts, and is also responsible for the decrease of interval lengths in the initial phases of the burst. There is also inhibitory feedback to the multiple event generator, probably mediated via the CP conducting system. This inhibitory feedback is slow rising, and the inhibitory effects of several CP's in succession are accumulative. The growing inhibition within a burst leads to increased intervals toward the end of the burst and eventually the termination of the burst. The joint effect of the excitatory and inhibitory components gives rise to the CONTRACTION PULSE PATTERNS IN HYDRA 271 observed pattern of intervals within ;i burst, i.e. intervals are shorter in mid- burst than at the start or end of the burst. The inhibitory effect of single CP's upon the multiple event generator can be directly demonstrated by driving the CP System with single stimuli. Exciting the CP System once each 15 to 30 seconds with single electric shocks or pulses of mechanical stimuli inhibit CP bursts ( Passano and McCullough, 1964, Rush- forth and Burke, in preparation). Indirect evidence for the inhibition of CP bursts by CP singles may be derived from an analysis of inter-CP intervals in long-term records consisting of both single pulses and bursts. The inter-burst interval is significantly longer (P < 0.01 ) if it includes a CP single than if no CP is present in the interval. \Yith increasing numbers of single CP's the inter-CP burst interval is correspondingly increased in length (see Table IV). It is proposed that there is inhibitory feedback to the single event generator mediated by the CP conducting system. Such inhibition results in the decrease in the intervals between single CP's following a burst of CP's throughout the inter-burst interval. This suggests gradual escape of the single event generator from long lasting inhibition following a CP burst. The model proposed for the CP System is incomplete. Modifications of the model are necessary to take into account interactions of the CP System with both the Rhythmic Pulse System and the Tentacle Pulse System. The model is likely to be further modified to take into account the effects of external stimuli on the pacemaker system. Photic, mechanical and chemical stimuli are likely to have complex influences on the pacemaker systems, possibly affecting more than one component. A major problem, as with all other hydroids, is the identification of the cellular elements involved in the pacemaker and conduction activities. Passano and Mc- Cullough (1965) proposed that the pacemaker system for CP's was the ectodermal nerve net. The size of the CP's, however, seemed too large to be exclusively the result of activity from the small dispersed neurons in the nerve net. Josephson and Macklin (1967, 1969) demonstrated that Hydra has a transepithelial potential. The fluid in the body cavity is typically 15 to 40 mv positive with respect to the outer bathing solution. Superimposed on this transepithelial potential are the negative going (depolarizing) CP's. It was hoped initially that direct measure- ment of the transepithelial potential and possible triggering of CP's in response to depolarizing current pulses might provide direct measurements for the construc- tion of models for the production of CP's. However, CP's were found to have several surprising features in relation to the transepithelial potential which made such efforts unfeasible. Josephson and Macklin (1969) found: (1) the amplitude and frequency of CP's are essentially independent of the transepithelial potential when the latter is altered by imposed current; (2) there is little change in the impedance of the hydra column during CP firing; (3) when the transepithelial potential is clamped at zero, CP's continue to appear spontaneously as current spikes. There are four membranes between the inside of the body cavity and the outside solution, only one of which is likely to be the CP-generating membrane. Josephson and Macklin have shown that the CP-generating membrane forms only a small fraction of the total transverse impedance of the column. This unfortu- nately means that the CP-generating membrane cannot adequately be studied by 272 NORMAN B. RUSHFORTH measuring potentials across the body wall, or passing current through it, since the presence of large series impedance elements masks the properties of the CP- generating membrane. The mechanisms by which CP's are conducted along the hydra column are unknown. Activity in the nerve net may trigger epithelial cells to produce these large electrical potentials. However, nerve free epithelia capable of producing propagated electrical pulses have been found in hydrozoans (Mackie, 1965, Mackie and Passano, 1968), so that it is possible that CP conduction as well as initiation is due to properties of the epithelial cells. In the present study the same charac- teristics of contraction potentials have been investigated in two species of Hydra. These characteristics impose boundary conditions on the parameters for models of pulse production. In subsequent papers we will investigate the properties of iso- lated hydroid pacemaker systems and the interactions of such pacemakers in the intact animal. I wish to acknowledge with gratitude the assistance of Mr. Paul Bakunas, Mr. Otto Morgenstern, Mrs. Nancy Rushforth, Mr. Richard Collins, and Mr. Howard Nearman. I wish to thank Dr. R. K. Josephson for his helpful criticism of the manuscript. SUMMARY 1. The temporal patterns of spontaneously produced Contraction Pulses (CP's) were investigated in two species of Hydra. It was found that H. pseudoligactis frequently give CP's as single, widely spaced events, whereas, H. pirardi produce many CP bursts. Inter-pulse interval histograms and joint order scatter diagrams, constructed on the basis of 10 to 12 hour recordings from 10 preparations of the two species, were employed to depict these contrasting temporal patterns of pulse production. The burst frequency and the burst length for CP's were found to be so different for the two species that these measurements alone allowed a unique designation of the animals into the two species classes. 2. CP bursts from unstimulated H. pirardi had the following characteristic features: (a) both positive and negative phases of the CP amplitudes increased progressively with successive pulses in a burst; (b) intervals were shorter in mid- burst than at the start or end of the burst ; and (c) intervals progressively decrease in length with successive position in an inter-burst interval. 3. The CP pattern of H. pseudoligactis can be changed from one primarily of single CP's to one consisting predominantly of bursts of pulses by the following external stimulation: (a) increased electrode suction; (b) intermittant light stimu- lation ; and (c) exposure to feeding stimuli. The three features of CP bursts found in unstimulated H. pirardi, are exhibited by the CP burst pattern induced by such stimuli in H. pseudoligactis. LITERATURE CITED ADSHEAD, P. C., G. O. MACKIE AND P. PAETKAU, 1963. On the Hydras of Alberta and the Northwest Territories. Bull. Nat. Mus. Canada, 199: 1-13. DAVIS, W. J., AND R. K. MURPHY, 1969. Discharge patterns of swimmeret motoneurones in the lobster, simulated with a digital computer. /. Exp. Biol., 50: 119-128. FULTON, C., 1963. Rhythmic movements in Cordylophora. J. Cell. Com/>. Physio!., 61 : 39-52. CONTRACTION PULSE PATTERNS IN HYDRA 273 IKEDA, K., AND C. A. G. WIERSMA, 1964. Autogenic rhythmicity in the abdominal ganglia of the crayfish: the control of swimmeret movements. Cmnp. Biochcm. Phvsiol., 12: 107-115. JOSEPH SON, R. K., 1967. Conduction and contraction in the column of hydra. /. Exp. Biol., 47: 179-190. JOSEPH SON, R. K., AND G. O. MACKIE, 1965. Multiple pacemakers and the behavior of the hydroid Titbularia. J. Exp. Biol., 43: 293-332. JOSEPHSON, R. K., AND M. MACKLIN, 1967. Traiisepithelial potentials in Hvdra. Science, 156: 1629-1631. JOSEPHSON, R. K., AND M. MACKLIN, 1969. Electrical properties of the body wall of Hydra. J. Gen. Physiol, 53: 638-665. LOOMIS, W. F., AND H. M. LENHOFF, 1956. Growth and sexual differentiation of hydra in mass culture. /. Exp. Zool., 132: 555-574. MACKIE, G. O., 1965. Conduction in the nerve-free epithelia of siphonophores. Amer. Zool., 5: 439-453. MACKIE, G. O., 1968. Electrical activity in the hydroid Cordylophora. J. Exp. Biol., 49: 387-400. MACKIE, G. O., AND L. M. PASSANO, 1968. Epithelial conduction in hydramedusae. /. Gen. Physiol., 52: 600-621. McCuLLOUGH, C. B., 1965. Pacemaker interaction in Hydra. Amer. Zool., 5: 499-504. MORIN, J. G., 1969. Mechanisms controlling behavior and bioluminescence in the colonial hydroid Obclia. Ph.D. thesis, Harvard University, Cambridge, Massachusetts, 156 pp. PASSANO, L. M., AND C. B. McCuLLOUGH, 1962. The light response and the rhythmic poten- tials of hydra. Proc. Nat. Acad. Sci. U. S. A., 48: 1376-1382. PASSANO, L. M., AND C. B. McCuLLOUGH, 1963. Pacemaker hierarchies controlling the be- havior of hydras. Nature, 199: 1174-1175. PASSANO, L. M., AND C. B. MCCULLOUGH, 1964. Coordinating systems of behavior in Hydra. I. Pacemaker systems of the periodic contractions. /. Exp. Biol., 41: 643-664. PASSANO, L. M., AND C. B. McCuLLOUGH, 1965. Coordinating systems and behavior in Hydra. II. The rhythmic potential system. /. Exp. Biol., 42: 205-231. RODIECK, R. W., N. Y. S. KIANG AND G. L. GERSTEIN, 1962. Some quantitative methods for the study of spontaneous activity of single neurons. Biophys., 2: 351-368. ROWAN, W., 1930. On a new hydra from Alberta. Trans. Roy. Soc. Canada, 24: 165. RUSHFORTH, N. B., 1966. An analysis of spontaneous contraction pulse patterns in Hydra. Amer. Zool., 6: 524. RUSHFORTH, N. B., 1967a. Chemical and physical factors affecting behavior in Hydra: inter- actions among factors affecting behavior in Hydra. Pages 369-390 in W. C. Corning and S. C. Ratner, Eds., Chemistry of Learning. Plenum Press, New York. RUSHFORTH, N. B., 1967b. Behavioral responses of isolated tentacles of Hydra. Biol. Bull., 133: 449. SHIBLEY, G. A., 1969. Gastrodermal contractions correlated with rhythmic potentials and pre- locomotor bursts in Hydra. Amer. Zool., 9: 586. STRUMWASSER, F., 1965. The demonstration and manipulation of a circadian rhythm in a single neuron. Pages 442-462 in J. Aschoff, Ed., Circadian Clocks. North Holland Publish- ing Co., Amsterdam. STRUMWASSER, F., 1967. Types of information stored in a single neuron. Pages 291-319 in C. A. G. Wiersma, Eds., Invertebrate Nervous Systems. University of Chicago Press, Chicago, Illinois. Reference : Biol. Bull., 140 : 274-283. (April, 1971 ) REGULATION OF GONAD DEVELOPMENT IN THE BAY SCALLOP, AEQUIPECTEN IRRADIANS LAMARCK 1 AKELLA N. SASTRY AND NORMAN J. BLAKE Graduate School of Occonography, Unircrsity of Rhode Island, Kingston, Rhode Island 02881 Gonad development in the annual reproductive cycle of temperate climatic zone marine invertebrates is correlated with seasonal changes in the environment. Gonad development has heen induced in molluscs (Loosanoff and Davis, 1950; Sastry, 1963) and barnacles (Crisp, 1957; Patel and Crisp, 1960) outside the normal breeding period by maintaining the animals at the appropriate tempera- tures and providing food. The effect of photoperiod on maturation and breeding of Balanus has been reported by Barnes (1963). The influence of food on gonad development in sea urchins and scallops has been reported by Giese (1959) and Sastry (1966). Barnes (1967) found that interruption of food supply leads to regression of ovarian tissue in barnacles, especially in the early stages of develop- ment. Although environmental influence on gonad growth and gametogenesis is apparent, the mechanism of control is still unclear. Sastry (1968) reported that gonad growth and gametogenesis are initiated in bay scallops, Aequipecten irradians Lamarck, when the animals are exposed to a minimum threshold temperature with food. In the absence of food or when the temperature remains below the threshold level, the oogonia develop but oocyte growth does not occur. This suggests that nutrient reserves from the ingested food are utilized during gonad growth and gametogenesis. The nutrient reserves are apparently transferred from the storage organ, the digestive gland, to the gonad and utilized by the developing gametes for synthesis of various biochemical con- stituents. The present paper considers the influence of temperature on the transfer of nutrient reserves. MATERIALS AND METHODS Source and maintenance of animals Bay scallops collected during the last week of November from Buzzards Bay, Massachusetts (10 C and 32%o), were obtained from the Supply Department, Marine Biological Laboratory, Woods Hole. The scallops were maintained in the laboratory at 5 C and 15 C in recirculating sea water aquaria (Day no Aqua- Labs, Inc., Lynn, Massachusetts) with a photoperiod of 12 hours light and 12 hours darkness. Scallops were fed daily on naked flagellates, Monochrysis lutheri Droop cultured in the laboratory on an enriched sea water medium described by Davis and Guillard (1958). Three liters of food organisms (2-3 million cells/ml) were provided daily to a group of 80 scallops held at each experimental temperature. 1 This study was supported by grants from the National Science Foundation, Sea Grant College Program, University of Rhode Island, Kingston, Rhode Island 02881. 274 GONAD DEVELOPMENT IN BAY SCALLOPS 275 Assay of reproductive activity At the beginning of the experiments, ten scallops were sacrificed, and the gonad index (gonad weight/body weight X 100), digestive gland index (digestive gland weight/body weight X 100), and the stage in gametogenesis were determined (Sastry, 1966). Five animals were removed at intervals from each experimental temperature and the same measurements made. The oocyte diameter in each ani- mal from each sample was measured with an ocular micrometer. Radiotracer experiments Radiotracer experiments were performed by injecting uniformly labelled "C- leucine (New England Nuclear Corp., Boston, Massachusetts) into the digestive gland of the scallops. Radioactive leucine (specific activity of 240 MC/rmV/ dis- 20 X LJ Q Q 10 O O I 5C 5C 30 DAYS 45 60 FIGURE 1. Gonad growth response of scallops at 5 C and 15 C. solved in 0.5 ml. 0.01 N HC1 was added to 0.5 ml of sea water (32%o) to make an approximately isotonic solution. Ten /xl (0.5 /xC) of solution per animal was injected into scallops held at 5 C and 15 C on the 15th. 30th and 45th day after the beginning of the experiment. The scallop injected with the labelled amino acid were placed in battery jars containing sea water of equilibrated temperature and kept in contsant temperature incubators. Each scallop was fed 100 ml of food organisms a day. Four or five animals were sacrificed after one day and again after one week to determine the gonad and digestive gland indexes and the stage in gametogenesis. Pieces of body tissues weighing 15-20 mg taken from each animal were dissolved in 1 ml of Soluene-100 (Packard Instrument Co.) in low potassium vials. The vials were kept for at least 18 hours at 35 C to aid in dis- solving the tissue. The vials were then filled with Bray's solution (Bray, 1960). The scintillation mixture contained 60 g naphthalene, 4 g PPO, 500 ml POPOP, 100 ml methanol, 20 ml of ethylene glycol in 880 ml of p-dioxine to make up to 276 AKELLA N. SASTRY AND NORMAN J. BLAKE a total volume of 1 liter. Radioactivity was measured with a Packard Tri-Carb liquid scintillation counter (model 3380) attached to a computer print out. Ap- propriate corrections were made for background and quenching. The total disintegrations per minute (DPM) for organs of each animal were calculated by the following method : Total DPM/organ -- DPM/mg tissue X total weight of organ (mg). The DPM for all body organs were added to obtain the total DPM per animal. Incorporation of labelled amino acid into the organs is calculated by the following method : Radioactivity index- (Total DPM/organ X 100/Total DPM/animal) Statistical analysis of the data was made with the aid of IBM-360 computer. RESULTS Gonad grou'tli response of scallops at 5 C and 15 C The mean gonad index values of scallops held at 5 C and 15 C are shown in Figure 1. The gonad index remained approximately the same for all samples from 5 C The scallops at 15 C showed an increase of gonad index following day 30. The differences between the gonad indexes of samples kept at the two temperatures became significant by day 45. The gonad and digestive gland indexes of scallops show a reciprocal relation- ship during the reproductive cycle (Sastry, 1970a). The digestive gland index, greater than the gonad index during the vegetative and resting periods, decreases with the gonad growth. This decrease has been attributed to the transfer of nutrients to the gonads (Sastry. 1966, 1968). The relationship between the two organ indexes of scallops exposed to 5 C and 15 C is shown in Figure 2 and Table I. The digestive gland index remained near a constant level above the gonad index for scallops at 5 C. At 15 C, however, the digestive gland index fluctuated, and showed an increase with the growth of gonads. Apparently the clear reciprocal relationship observed for field population does not apply to animals held in the laboratory with abundant food. Gametogenesis in scallops at 5 C and 15 C The scallops examined at the beginning of the experimental period had only primary germ cells. The oocyte growth in scallops held at 5 C and 15 C is shown in Figure 3. Oogonia developed at 5 C but there was no oocyte develop- ment. The scallops held at 15 C developed oocytes in the cytoplasmic growth phase by clay 21 and the vitellogenesis growth phase by day 36. Most oocytes were in the vitellogenesis growth phase on day 51. Spermatocytes and spermatids were observed in the scallops at the beginning of the experiment. In the scallops held at 5 C spermatocytes were the dominant stage throughout the experimental period, but a few spermatozoa also were ob- served. At 15 C spermatozoa were predominant, but spermatocytes and sperma- tids were present. GONAD DEVELOPMENT IN BAY SCALLOPS 277 Incorporation of 1 *C-leucine into the digestive gland and gonad of scallops at 5' and 15 C, 24 hours after injection C The mean radioactivity index for 14 C-leucine incorporation into the digestive gland and the gonad of scallops at 5 C and 15 C is given in Figure 4 and Table II. The animals injected with the labelled amino acid on day 15 and 30 at 5 C showed a wide variation of incorporation into the digestive gland. The mean radioactivity index for labelled amino acid incorporation was slightly higher in the I 5 to Q CO UJ o 5 l5 Q 10 5C DIGESTIVE GLAND INDEX GONAD INDEX I 5C 15 30 DAYS 45 60 FIGURE 2. Mean gonad index and digestive gland index of scallops exposed to temperatures of 5 C and 15 C. sample injected on day 30. Incorporation of "C-leucine into gonads was much less than into the digestive gland in all the samples. The gonads of samples in- jected on days 30 and 45 took up slightly more than those injected on day 15. In the scallops held at 15 C the incorporation of 14 C-leucine into the digestive gland of all samples was about the same except for day 30, when it was slightly higher. Again the gonads took up much less amino acid than the digestive gland. The mean radioactivity index for incorporation of 14 C-leucine into the gonads was about the same in samples injected at different time intervals. Incorporation of 14 C-leucine into both digestive gland and gonad of scallops held at 15 C was higher than in the animals at 5 C (Table II). 278 AKELLA N. SASTRY AND NORMAN J. BLAKE Incorporation of l *C-leucine into digestive gland and gonad of scallops at 5 C and 15 C, one week after injection The incorporation of 14 C-leucine into the digestive gland and gonad of scallops at 5 C and 15 C is shown in Figure 4 and Table III. Uptake of labelled amino acid by the digestive gland was about the same on day 15 and day 45 for scallops held at 5 C. The incorporation of 14 C-leucine into the gonad, while much lower than that for the digestive gland, was slightly higher on day 45 than on day 15. In the scallops held at 15 C, the mean radioactivity index for incorporation of labelled amino acid into the digestive gland was about the same in all the sam- TABLE I Mean gonad and digestive gland indexes of scallops at 5 C and 15 C Day Organ 5C 15 C Significance (t-test) t G D 8.7 1.4 11.6 J.6 8.7 1.4 11.6 1.6 15 G D 7.3 1.3 9.6 0.8 8.4 0.8 10.4 0.7 1.58 1.74 21 G D 8.0 1.1 10.7 0.3 9.8 1.8 9.0 0.8 1.87 4.41* 30 G D 7.2 0.6 9.6 0.8 7.7 0.9 10.9 1.4 1.08 1.08 36 G D 7.4 0.9 10.1 0.8 8.6 1.1 9.0 0.1 2.10 1.95 45 G D 7.9 1.0 10.3 0.4 12.2 1.6 13.4 1.4 5.05* 4.64* 51 G D 8.3 1.1 9.4 1.1 14.7 1.4 12.1 1.2 8.13* 3.91* G, Gonad ; D, Digestive gland ; * P < 0.05. pies. At this temperature, however, the gonads incorporated more 14 C-leucine than the digestive gland on days 30 and 45. The difference in the uptake of 14 C-leucine into the gonads of scallops held at 5 C and 15 C is statistically significant (Table III). Incorporation of 14 C- leucine into the digestive gland of scallops held at 15 C was higher than in the animals held at 5 C. The difference is only significant in the samples injected on day 45 (P < 0.05). DISCUSSION The influence of environmental factors on the gonad development of marine invertebrates has been demonstrated, but the mechanisms controlling gonad growth in the reproductive cycle are not yet clear. Breeding condition has been induced in GONAD DEVELOPMENT IN BAY SCALLOPS 279 80 60 =2, cr LU h- LJ UJ h- >- O O o < LJ 40 20 . . |5C - 5C 30 DAYS 45 60 FIGURE 3. Oocyte growth in scallops at 5 C and 15 C. TABLE II Incorporation of u C-leucine into gonad and digestive gland of scallops at 5 C and 15 C, 24 hours after injection Day of injection Organ 5C Radioactivity index 15 C Radioactivity index Significance (t-test) t 15 G D 3.8 2.3 13.6 10.4 11.1 7.8 18.9 4.4 1.99 0.96 30 G D 6.8 3.8 20.7 17.7 7.3 3.4 26.7 10.0 0.23 0.66 45 G D 7.0 2.0 14.0 2.5 9.8 1.4 19.1 2.5 2.51* 3.00* G, Gonad; D, digestive gland; * P < 0.05. 280 AKELLA N. SASTRY AND NORMAN J. BLAKE bivalve molluscs and barnacles by maintaining tbe animals at appropriate tempera- tures in the winter and supplying food. This suggests a temperature and nutri- tional influence on gonad development (Loosanoff and Davis, 1950; Patel and Crisp, 1960; Sastry, 1963). Crisp (1957) reported that Balanus required main- tenance at colder temperatures for a prolonged period before they could be induced to breed at warm temperatures. However, exposure to warm temperatures alone failed to induce breeding condition in Balanus. Starvation of a number of species X UJ Q 30 >- 25 UJ 2 O UJ _l o 10 30 < UJ 20 10 o 5 C 15 C ---DIGESTIVE GLAND GONAD After one week After 24 hours 15 30 45 14 DAY OF C-LEUCINE INJECTION FIGURE 4. Incorporation of 14 C-leucine into gonad and digestive gland of scallops, 24 hours after injection and one week after injection. at the beginning of the reproductive period was reported to prevent gonad growth which suggests that food influences gonad development (Giese, 1959; Sastry, 1966; Barnes, 1967). Apparently, the nutrient reserves from the ingested food must be transferred to gonads for growth to occur and these materials are utilized there in the biochemical synthesis of the developing gametes (Giese, 1959; Barnes, Barnes and Finlayson, 1963). It has been suggested that nutrient reserves are transferred from the storage organ in barnacles and starfish or directly from gut to perivisceral fluid and then to gonads in sea urchins without significant storage (Giese, 1959; GONAD DEVELOPMENT IN BAY SCALLOPS 281 Farmanfarmaian and Phillips, 1962 ; Barnes, Barnes and Finlayson, 1963 ; Mauzey, 1966). The transfer of nutrient reserves to the gonads is thought to occur with the initiation of gonad development by establishment of a gradient between source and utilization (Barnes, Barnes and Finlayson, 1963). However, there are only a few studies examining the distribution and utilization of nutrients during repro- duction in marine invertebrates. Allen (1962) found that Mya arenaria fed with 3L 'P labelled Pheodactylum assimilates greatest concentrations of 32 P in the digestive gland with little evidence of rapid build up in the ovary. In Calanus, how r ever, Marshall and Orr (1955, 1961) found that 70% of the 32 P assimilated by the female copepod passes rapidly into the ova. The results of the present study show that transfer of nutrient reserves to gonads occurs in scallops with the initiation of gonad development. Gonad development is initiated in fed animals maintained within a range of temperatures (Sastry, 1968). Under these conditions, the nutrient reserves are apparently transferred to the gonads. Temperature below the TABLE III Incorporation of l *C-leucine into gonad and digestive gland of scallops at 5 C and 15 C one week after injection Day of injection Organ 5C Radioactivity index 15 C Radioactivity index Significance (t-test) t 15 G D 4.2 1.1 13.2 12.9 15.7 7.8 20.6 11.8 3.12* 0.96 30 G D 20.0 12.0 18.2 1.8 45 G D 8.5 2.2 12.4 3.4 20.4 3.4 19.8 3.0 6.54* 3.48* G, Gonad; D, Digestive gland; * P < 0.05. required minimum supresses gonad growth and the nutrient reserves are accumu- lated in the digestive gland. The nutrient reserves stored in the digestive gland may supplement the nutritional intake needed for gonad growth. However, the animals maintained at temperatures exceeding a certain maximum also fail to develop gonads. Perhaps this is due to increased metabolic utilization or failure to regulate the synthetic processes of the developing gametes. Gametogenesis occurs simultaneously with gonad growth in scallops. Oocytes enter the cytoplasmic growth phase when the animals are exposed to the required minimum temperature with food. Either subthreshold temperature or starvation will prevent the oocytes from developing (Sastry, 1968). Apparently initiation of the cytoplasmic growth phase is regulated by the transfer of nutrients to the gonads, this transfer, of course, being dependent on the temperature. These nutrients are obviously utilized by the oocytes during their growth to mature eggs. The nutrient transfer to the gonads is therefore not only regulated by environmental factors, but also by the nutritional needs of the developing oocytes. The significant increase in the incorporation of 14 C-leucine into gonads of scallops with oocytes in the cytoplasmic growth phase at 15 C further suggests AKELLA N. SASTRY AND NORMAN J. BLAKE that the rate of nutrient transport to the gonads is regulated by the temperature and the stage of oocyte development (Figs. 3 and 4). However, the uptake of labelled amino acid into gonads of scallops at 5 C and 15 C measured 24 hours after injection is not significantly different. The nutrient transfer may occur slowly at a rate regulated by the developing oocytes. The difference in the incorporation of 14 C-leucine into the gonads of scallops at 5 C and 15 C could have been influenced by the temperature, or the stage of oocyte development, or both. Incorporation of 14 C-leucine into gonads of scallops with oogonia at 5 C is much less than those with oocytes in the cytoplasmic growth phase at 15 C, especially one week after injection. Incorporation of 14 C- leucine has been shown to be involved in protein synthesis (Raven, 1961 ; Williams, 1965 ). It has also been reported that protein synthesis occurs in growing medium- sized oocytes but is absent in very young or mature unfertilized eggs of the sea urchin. It is possible that oocytes entering the growth phase stimulate gonad development, with the rate of gonad development depending on the rate at which nutrient reserves are utilized by the growing oocytes for synthesis of various bio- chemical constituents. Oocytes entering the cytoplasmic growth phase may begin protein synthesis by utilizing the nutrients taken up from the digestive gland. The cytoplasmic growth phase of oocytes may be triggered either by direct action of temperature or by neurosecretion or both (Lubet, 1966; Sastry, 1970b). It is a pleasure to thank Dr. Eugene Jackim and Dr. Earl Davey, National Marine Water Quality Laboratory, West Kingston, for the use of facilities for radioactivity measurements and for assistance during this study. Thanks are also due to Miss Katherine Saunders for technical assistance. SUMMARY 1. Gonad growth and oocyte cytoplasmic growth were induced in scallops ex- posed to 15 C with food. At 5 C only oogonia developed. 2. Scallops injected with 14 C-leucine and measured 24 hours after injection showed greater incorporation of labelled amino acid into both digestive gland and gonad at 15 C than at 5 C, but the difference in uptake between the two groups was not significant. 3. Incorporation of 14 C-leucine into gonads of scallops measured one week after injection was significantly higher at 15 C than at 5 C. 4. The gonad growth and oocyte growth of scallops is controlled by food and temperature. Both external environmental factors and the stage of oocyte devel- opment seem to regulate nutrient transfer to gonads and therefore control growth activity. LITERATURE CITED ALLEN, J. A., 1962. Preliminary experiments on the feeding and excretion of bivalves using Phaeodactylum labelled with 32 P. /. Mar. Biol. Ass. U. K., 42: 609-623. BARNES, H., 1963. Light, temperature and breeding of Balainis balanoides. J. Mar. Biol. Ass. U.K., 43: 717-728. BARNES, H., 1967. The effect of starvation and feeding on the time of production of egg masses in the boreo-arctic Cirripede, Balanus balanoides (L.). /. Exp. Mar. Biol. Ecol., 1: 1-6. GONAD DEVELOPMENT IN BAY SCALLOPS BARNES, H., M. BARNES AND M. FINLAYSON, 1963. The seasonal changes in body weight, biochemical composition and oxygen uptake of two common boreo-arctic barnacles, Balanus balanoidcs and B. halanus. J. Mar. Biol. Ass. U. K.. 43: 185-212. BRAY, GEORGE A., 1960. A simple efficient liquid scintillator for counting aqueous solutions in a liquid scintillation counter. Anal. Biochcm., 1 : 279-288. CRISP, D. J., 1957. Effect of low temperature on the breeding of marine animals. Nature, 179: 1138-1139. DAVIS, H. C., AND R. R. GUILLARD, 1958. Relative value of ten genera of microorganisms as food for oysters and clam larvae. Fish. Bull. U. S. Fish Wildlife Scrv., 58: 293-304. FARMANFARMAIAN, A., AND J. H. PHILLIPS, 1962. Digestion, storage, and translocation of nutrients in the purple sea urchin (Strongvlocentrotus purpuratus). Biol. Bull., 123: 105-120. GIESE, A. C., 1959. Comparative physiology : Annual reproductive cycle of marine inverte- brates. Anmt. Rev. Physio!., 21: 547-576. LOOSANOFF, V. L., AND H. C. DAVIS, 1950. Conditioning of V . mercenaria for spawning in winter and breeding its larvae in the laboratory. Biol. Bull., 98: 60-65. LUBET, P., 1966. Essai d'analyse experimentale des perturbations produites par les abaletions des ganglions nerveux Chez Mytilits edulis L and Mytilns galloprovincialis LMK (Mollusques Lamellibranches). Ann. Endoc. 27: 353-365. MAUZEY, K. P., 1966. Feeding behavior and reproductive cycles in Pisastcr ochraccus. Biol. Bull., 131: 127-144. MARSHALL, S. M., AND A. P. ORR, 1955. On the biology of Calanus fitnnarcliicits. VII. Food uptake, assimilation and excretion in adult and stage V. Calanus. J. Mar. Biol. Ass. U.K., 34: 495-529. MARSHALL, S. M., AND A. P. ORR, 1961. Studies on the biology of Calanus finnwrchiciis. XII. The phosphorous cycle : excretion, egg production and autolysis. /. Alar. Biol. Ass. U. K., 41 : 463-488. PATEL, B., AND D. J. CRISP, 1960. The influence of temperature on the breeding and moulting activities of some warm water operculate barnacles. /. Mar. Biol. Ass. U. K., 39: 667-680. RAVEN, CHR. P., 1961. Oogcncsis: The Storage of Development Information. Pergamon Press, New York, 274 pp. SASTRY, A. N., 1963. Reproduction of the bay scallop, Acquipcctcn irradians Lamarck. Influ- ence of temperature on maturation and spawning. Biol. Bull., 125: 146-153. SASTRY, A. N., 1966. Temperature effects in reproduction of the bay scallop, Aequipectcn irradians Lamarck. Biol. Bull., 130: 118-134. SASTRY, A. N., 1968. The relationships among food, temperature and gonad development of the bay scallop, Aequipectcn irradians Lamarck. Physiol. Zool., 41: 4453. SASTRY, A. N., 1970a. Reproductive physiological variation in latitudinally separated popula- tions of the bay scallop, Aequipectcn irradians Lamarck. Biol. Bull., 138: 56-65. SASTRY, A. N., 1970b. Environmental regulation of oocyte growth in the bay scallop, Aequi- pecten irradians Lamarck. E.vperentia, in press. WILLIAMS, J., 1965. Chemical constitution and metabolic activities of animal eggs. Pages 13-71 in Rudolf Weber, Ed., The Biochemistry of Animal Development, Descriptive Biochemistry of Animal Development, Volume 1. Academic Press, New York. Reference : Biol. Bull., 140: 284-322. (April, 1971) LARVAL DISPERSAL AS A MEANS OF GENETIC EXCHANGE BETWEEN GEOGRAPHICALLY SEPARATED POPULATIONS OF SHALLOW-WATER BENTHIC MARINE GASTROPODS 1 RUDOLF S. SCHELTEMA Woods Hole Oceanographic Institution, ll'oods Hole, Massachusetts 02543 If the larvae of marine gastropods are capable of frequent long-distance dis- persal, then they can he expected to serve, not only as a means of colonizing or re-colonizing new regions, but also as agents of gene flow between widely separated or disjunct populations. Populations that become completely isolated will tend to evolve in different directions. On the other hand, frequent genetic exchange between populations of the same species will counteract the divergent evolution between them, particularly if the immigrant veligers introduce especially advantage- ous genes upon which selection may act. To support the hypothesis that larvae are important in maintaining genetic continuity between shoal-water populations of gastropods with amphi-Atlantic distributions, it is necessary first to give evidence that larvae are indeed dispersed over long distances and second to show that the frequency of larval dispersal is related to converging or diverging evolution of isolated adult populations of gastropod species. Two kinds of evidence are useful in demonstrating long-distance dispersal : ( 1 ) the relationship between the circulation and the geographical distribution of larvae and ( 2 ) the duration of pelagic larval life and its relationship to the velocity of trans-Atlantic currents. Thus, larvae should be present along the entire length of the surface currents required to transport them between the geographically dis- junct adult populations, and the length of pelagic larval life must be long enough so that currents will carry veligers the required distance within the time allotted them. To understand the frequency of larval dispersal it is necessary to know: (1) the probability that larvae will survive and be transported by off-shore currents to far-distant populations and (2) the approximate number of larvae produced by the originating or parent population. In this account I will attempt to show that larvae are dispersed over long dis- tances, that the frequency of dispersal is much higher than might superficially be supposed, and finally, in a preliminary way, to show that the relative frequency of larval dispersal does indeed seem to be related to morphological similarities or differences found between spatially separated populations. METHODS The conclusions presented here are based on 857 plankton tows taken through- out the tropical and warm-temperate North Atlantic Ocean (Fig. 4). The samples were obtained from the research vessels ATLANTIS II and CHAIN by making oblique tows with a three-quarter-meter plankton net having a mesh of 223 or 316 microns. In most instances a net with the smaller mesh was used. The samples were taken 1 Contribution No. 2565 from the Woods Hole Oceanographic Institution. 284 DISPERSAL AND GENETIC EXCHANGE by paving out 200 meters of wire at approximately 20 meters per minute and then hauling' the net inboard at a similar rate. The duration of a tow was 20 minutes and the ship moved at approximately 5.5 km/hr (3 knots). The maximum depth to which the net descended depended on current and weather conditions, but was usually about 100 meters. Preliminary sorting was done at sea where living larvae were separated from the sample. Observations of live veligers proved to be quite valuable for later identification. Certain species were selected for rearing to settlement ; these were maintained in one-liter polyethylene boxes and fed from phytoplankton cultures maintained at sea. In addition to tows made by me from the research vessels of the Woods Hole Oceanographic Institution, I have also had access to a number of other plankton samples loaned to me by various colleagues. Very few previous taxonomic investigations have been made on tropical gastropod larvae (vide Simroth, 1895; Thorson, 1940; Dawydoff, 1940; Lebour, 1945), so that the adult affinities of each veliger form must be determined. Identifications can be made either by (1) comparing the larval shell with the protoconch of an adult and (2) rearing the larvae in the laboratory through settle- ment and post-larval development. In practice, at least with tropical species, the latter method presents consider- able difficulties. Though I have held some species of veliger larvae to settlement in the laboratory (viz., Cyinatiitiu nicobariciim, Thais haeniastoma, and Philippia krebsii), generally it is not readily possible to maintain the post-larvae to a stage at which identification becomes possible, because the feeding requirements of most tropical gastropods are still quite unknown. Identifications of the gastropod veligers considered here were made by careful comparison between larval shells and protoconchs of young adult specimens. I am indebted to the curators of molluscs, Drs. J. Rosewater, U. S. National Museum, R. T. Abbott and R. Robertson, Academy of Natural Sciences of Phila- dephia, and W. J. Clench and K. J. Boss, Museum of Comparative Zoology, Harvard University, for allowing me the use of collections under their custody. Mr. H. Lewis of the Philadelphia Shell Club kindly made specimens of Cymatiidae available to me. Usable specimens having protoconchs in good condition are relatively rare for most species. In identifying any species of larva, I have looked at the protoconch of specimens of all other closely related species that occur in the tropical and warm-temperate Atlantic if the material were available. RESULTS AND DISCUSSION Evidence for long-distance larval dispersal The relationship between circulation and geographical distribution of larvae The importance of larval dispersal was already recognized by Alfred Russel Wallace in his work on The Geographical Distribution of Animals (1876). Wal- lace knew that the univalve and bivalve Mollusca have free-swimming larval stages and recognized that "they thus have a powerful means of dispersal, and are carried by tides and currents so as ultimately to spread over every shore and shoal that offers conditions favorable for development." (Vol. I, page 30). Simroth (1895) 286 RUDOLF S. SCHELTEMA described and figured pelagic Mollusca, including gastropod veliger larvae, from samples taken on the I'lanklan l : ,.\'[>cditnn\, though he did not establish the identity of most of the forms he described. Thorson ( l ( )6l, page 466) recognized that certain genera of gastropods, principally tropical in their geographical distribu- tion (e.y., Lamcllaria, Tonna, Cassis, Charonia, Cyiiiti/imn, and Bnrsa], have larvae that "have been found midway across the oceans." Recently, Robertson (1964) and Scheltema (1964, 1966a) have described the dispersal of two species of gastropods into the Sargasso Sea and Gulf Stream, while Mileikovsky (1966) has reviewed existing knowledge on the dispersal of molluscan larvae, particularly the dispersal of prosobranch veligers in the Norwegian Sea. The gastropod families that are now known to have representatives with long pelagic developments include the families Neritidae and Naticidae (first mentioned here), and the Triphoridae, Lamellariidae, Architectonicidae, Cymatiidae, Cassidae, Tonnidae, Muricidae, Bursidae, Coralliophilidae, Ovulidae, and Cypraeidae (Scheltema, 1966b). While scattered records have already shown the existence of larvae in the open sea, no systematic attempt has yet been made to determine their occurrence over large areas of the ocean and in particular their presence throughout the entire length of major current systems of the equatorial and north temperate Atlantic Ocean. These data are the most important in demonstrating the dispersal of larvae over long distances. The evidence from plankton collections I have made shows that dispersal of gastropod larvae into the open sea is not a rare event but a commonplace occur- rence (Fig. 4). Taking into account all collections examined, including samples from all major currents of warm waters of the North Atlantic Ocean, 70 per cent contain gastropod larvae from the shoal-water benthos. Many species are found during all months of the year. Because of their common occurrence and in order to avoid the cumbersome expression "long-distance larvae from shoal-water bottom- dwelling organisms of the shelf," I have applied to these open-sea forms the term teleplanic larvae or teleplanos (from the Greek telep'lanos, meaning far-wandering) (Scheltema, in press). The teleplanos may be denned as larvae that (T) originate from shoal-water, continental-shelf benthos, (2) are regularly found in the open sea, (3) have a pelagic development of long duration, and (4) serve as a means for dispersal over long distances. The term is not restricted taxonomically to any phylum, but specifically excludes larvae of the holoplankton. Many teleplanic gastropod veligers have morphological adaptations which enhance their survival on the open ocean, such as long periostracal spines, the reduction or complete lack of shell calcification, and an increase in length of the velar lobes used for swimming and feeding. The North Atlantic circulation. The direction and route of larval dispersal are determined by the principal ocean currents. The circulation of the North Atlantic Ocean may be regarded as an enormous anticyclonic gyre. On the west is the Gulf Stream moving northeastwardly ; on the north is the North Atlantic Drift moving toward the east and dividing into a northern and southern arm ; bounding the eastern portion of the gyre is the Canary Current moving to the southwest ; and finally on the south is the North Equatorial Current moving west- wardly across the tropical North Atlantic toward the West Indies (Fig. 4). De- tails of this circulation are obviouslv much more complex. Large temporary eddies DISPERSAL AND GENETIC EXCHANGE 287 are known to occur in the Gulf Stream (Fuglister, 1963) and the North Atlantic Drift is grossly over-simplified by this description (vide Worthington, 1962; Mann, 1967). Notwithstanding, this generalized conception allows prediction of the transport expected from an object floating on or near the ocean surface. In the Equatorial Atlantic Ocean the North Equatorial Current is separated from the westwardly flowing South Equatorial Current by a weakly developed Equa- torial Countercurrent flowing toward West Africa. The Countercurrent moves only a relatively small volume of water and is seasonal in occurrence (Schumacher, 1940, 1943; U' S. Hydrographic Office Monthly Pilot Charts). However, directly beneath the westwardly flowing South Equatorial Current a strong eastwardly moving Equatorial Undercurrent flows from South America toward the island of Sao Tome off West Africa (Voorhis, 1961 ; Voigt, 1961 ; Metcalf, Voorhis and Stalcup, 1962; Khanaychenko, Khlystov and Zhidov, 1965; Rinkel, Sund and Neuman, 1966; Strum and Voigt, 1966). This warm, shallow current (20 to 27 C), whose core extends between 50 and 100 meters depth, provides a means whereby tropical larvae may be dispersed from west to east across the tropical Atlantic (Chesher, 1966, page 210; Scheltema, 1968). Description of larvae and their geographical distribution in the North and Equatorial Atlantic Ocean. To illustrate specific instances of teleplanic gastropod larvae in temperate and tropical Atlantic waters, I have selected to consider in detail, from among many species, ten representative forms taken from the families Cymatiidae, Tonnidae, Cassidae, Muricidae, Architectonicidae, and Neritidae. Since the larvae of these ten species are either completely unknown or not pre- viously well described, each is illustrated and briefly described here. Family Cymatiidae The family Cymatiidae or hairy tritons is a group of carnivorous gastropods which are very well represented in the tropical Atlantic. The three species con- sidered here, C \nnatlnm parthenopeum (von Salis), Cymatium nicobaricnni (Roding), and Charonia variegata (Lamarck), all have very wide geographical distributions. Cymatium parthenopeum is known throughout most of the warm and tropical North and South Atlantic and as Monople.r australasiae Perry in the Indo-Pacific (Clench and Turner, 1957). Cymatium nicobaricnui is common in the western Atlantic and Indo-Pacific but has also been reported from a few records in the eastern Atlantic (Dautzenberg, 1890; Odhner, 1931; Clench and turner, 1957). The egg masses of several Cymatiidae are known from recent studies of Laxton (1969). The larva of Cymatium partJienopenm has an uncalcified, resilient, translucent, amber-colored, conical, multispiral shell which becomes somewhat darker in color when nearing completion of planktonic development (Fig. 1, d, e, f). The first whorl following the embryonic shell is cancellate ; the remaining whorls are glossy and smooth (Fig. 1, j). The shell in early stages is holostomatous, becoming siphonostomatous with a short siphonal canal when reaching the intermediate stage (Fig. 1, d; see also Scheltema, 1966a, Figs. 4a and 4b). The outer lip has a varix when larval growth is completed (Fig. 1, e, f). The slightly impressed suture of the body whorl is mostly obscured by the smooth, overlying _>S,K RUDOLF S. SCHELTEMA m FIGURE 1. DISPERSAL AND GENETIC EXCHANGE 289 periostracum. The operculum shows practically no markings ; a very faint eccen- tric nucleus is sometimes visihle. In the living larva the body is light cream- colored, while the foot is gray to black. The larva has two tentacles and four long velar lobes lacking pigmentation (ride Scheltema, 1966a, Fig. 2). When growth is completed the larval shell is about 3.5 mm in length. The unknown larval shell figured by Simroth (1895, Table V, Fig. 16) is probably this species. The teleplanic larvae of Cymatium parthenopeum are found throughout the North Atlantic gyre, in the Caribbean, and from two locations in the western end of the south Equatorial Current (Fig. 5). The larva of Cymatium nicobaricum was described by Lebour (1945, pages 477-478, Fig. 24, as Cymatium chlorostomum) from the waters near Bermuda, and the species was also figured but unidentified by Simroth (1895, PI. V, Figs. 10 and 11). It is very easily recognized because it is the only Cymatium larva in the Atlantic having a coeloconoid shell (Fig. 1, k, 1). The first H whorls following the embryonic shell are cancellate. On the completely developed larva the embryonic and sometimes a portion of the cancellate whorls are already worn. The early and intermediate stage larvae have long, slightly curved, periostracal spines (Fig. 1, h, 1) which in older larvae may no longer be present (Fig. 1, k). Minute axial lines are sometimes barely visible in the periostracum. The oper- culum (Fig. 1, m) differs from most Cymatiidae in that its nucleus is only very slightly eccentric. Only Charonia varicgata has a similar operculum (Fig. 1, i), but the larvae of Cymatium nicobaricum cannot be confused with this species. In the living larva of Cymatium nicobaricum the body pigmentation is pink with darker reddish patches. There are four long velar lobes which completely lack pigmentation. The length of the fully grown larval shell is about 4.6 mm. It is very often encrusted with fouling organisms including Folliculinidae, green algae, and encrusting bryozoa. Cymatium nicobaricum larvae are restricted in the samples to the western Atlantic and North Atlantic Drift (Fig. 5) but were never found in the eastern Atlantic. Charonia variegata has the largest known larvae of any of the Cymatiidae in the Atlantic, the shell exceeding 5 mm in length when fully developed. Its ovate, FIGURE 1. Larval shells, opercula and protoconchs of three species of Cymatiidae: (a) larval shell of Charonia varicgata. Northwestern Atlantic, 3927'N, 5017'W; (b) protoconch of Charonia raricgata. Western Atlantic, Tobago Island, B.W.I. (ANSP 209812) ; (c) Cha- ronia raricgata larval shell, North Equatorial Current, 1911'N, 4724'W ; (d) intermediate stage veliger larva of Cymatium parthenopeum. Gulf Stream, 3843'N, 7046'W; (e) proto- conch of Cymatium parthenopeum. collection of H. Lewis (No. 2493) ; (f) late veliger larva of Cymatium parthenopeum at final stage before settling, off the Azores, 3632'N, 3014'W; (g) larva of Charonia variegata. Eastern Atlantic off the West Coast of Africa, Atlantide Station No. 122, 0129'S, 0850'E; (h) apical view of intermediate stage larva of Cyinatium nicobari- cum, Gulf Stream, 3544'N, 6653'W; (i) operculum from larva of Charonia raricgata, internal view, specimen from same station as (a) above; (j) embryonic shell and first cancellate whorl of the pelagic larva of Cymatium parthenopeum, Gulf Stream, same station as (d) above; (k) late veliger larva of Cymatium nicobaricum at final stage before settlement, Western Atlantic near Bermuda, 3245'N, 6436'W; (1) intermediate stage larva of Cymatium nicobaricum. Western Atlantic, Gulf Stream, same station as (h) above; (m) operculum from larva of Cymathtm nicobaricum. internal view, specimen from same station as (h) above. Scale = 1 mm in all figures except i, j, and m where it is equivalent to 0.5 mm. Numbers prefixed by "ANSP" are lot numbers in the Academy of Natural Sciences of Philadelphia. 290 RUDOLF S. SCHELTEMA FIGURE 2. DISPERSAL AND GENETIC EXCHANGE 291 hairy shell is dark umber-hrown. barely translucent, and nncalcified in the early stages. The protoconch found on museum specimens of this species is smooth and secondarily calcified from within, giving a rather massive appearance (Fig. 1, b). Sometimes there is evidence of periostracal fragments adhering to the proto- conch, remains of the elaborate pattern of short, soft spines which give the larvae a hirsute appearance (Fig. 1, a, c, g). The whorls directly following the embryonic shell are cancellate but usually already worn in late larval development ( Fig. 1, a). The last three whorls are shouldered. The sutures are normally obscured by the periostracum, which is continuously laid down with the growth of each succeeding whorl. In early larvae the periostracal spines on the first two or three whorls directly following the cancellate whorl are more or less axially arranged (Fig. 1, a). In succeeding whorls the periostracal spines are spirally arranged (Fig. 1, c). With growth of the larva the periostracum of the adapical whorls is overlain with new periostracum upon which the spines are also spirally arranged. On older and larger larval shells the outer parietal lips are glossy and semitransparent in transmitted light but appear very dark brown in reflected light. The operculum of the larva is characteristic, having the nucleus almost central as it also is in the adult (Fig. 1, i; vide Clench and Turner, 1957 for illustration of adult operculum). This is probably the species figured but unidentified by Simroth (1895; PL VI. Figs. 15). Since the protoconch of the eastern Atlantic species Cliaronia nodifera (Lamarck) has not been seen, it is not certain that the larval shell of this species can be separated from Charonia variegata. Variations in shell morphology are not consistently related to the geographical distribution of the larvae, and therefore two species of veligers are not distinguishable. The dispersal of Cliaronia variegata larvae by ocean currents is very evident from their widespread geographical dis- tribution throughout the North Atlantic gyre, in the South Equatorial Current, and from scattered records in the Caribbean Sea (Fig. 6). FIGURE 2. Larval shells, opercula and protoconchs of the Tonnidae and Cassidae : (a) larval shell of the early intermediate stage veliger of Tonna galca. Sargasso Sea east of Ber- muda, 3205'N, 5910'W; (b) intermediate stage larva of Tonna galca. North Equatorial Cur- rent, 2434'N, 4720'W ; (c) intermediate stage larva of Tonna galca, same specimen as (b") above; (d) shell of fully developed veliger larva of Tonna c/alca, Western Atlantic, Gulf Stream, 3622'N, 6753'W; (e) protoconch of Tonna galca, apical angle 105, Western At- lantic, S. Inlet, Lake Worth, Boynton, Florida (ANSP 181564); (f) apical view of fully grown larva of Tonna galca, same specimen as (d) above; (g) larval shell of fully grown veliger, Tonna maculosa. Western Atlantic, Sargasso Sea, 2505'N, 5748'W; (h) larval operculum of intermediate stage veliger of Tonna maculosa, viewed from inside, specimen from Gulf Stream between Cape Cod and Bermuda; (i) larval operculum of intermediate stage veliger of Tonna galca, viewed from inside, specimen from Gulf Stream between Cape Cod and Bermuda; (j) protoconch of Tonna maculosa, apical angle 94, Western Atlantic, Key West, Florida (Scheltema coll.) ; (k) larval shell of fully grown veliger, Tonna maculosa, same speci- men as (g) above; (1) two views of the protoconch of Phaliuni granulatum, Western Atlantic, Matthew Town, Great Inagua, Bahamas (ANSP 173078); (m) operculum from larva of Phalium granulatum, viewed from inside, specimen from North Atlantic Drift, 4411'N, 4411'W; (n) three views of fully developed larva of Phaliuni granulatum, off the West Afri- can coast, 2032'N, 2234'W; Co) three views of intermediate stage larva, Phaliuni granulatum. Western Atlantic, Gulf Stream, 3817'6936'W; (p) three early stage larvae of Phalium granulatum, Western Atlantic, northern end of Gulf Stream, 4114'N, 5728'W. Scale = 1 mm in all figures except m, where it is equivalent to 0.5 mm. Numbers prefixed with ANSP refer to museum lot numbers in the collection of the Academy of Natural Sciences of Phila- delphia. 292 RUDOLF S. SCHELTEMA m FIGURE 3. DISPERSAL AND GENETIC EXCHANGE 293 Family Tonnidae The Tonnidae or tun shells are large gastropods found in tropical and warm temperate waters throughout the world. Members of this family are restricted to the continental shelf and occur mostly on sandy bottoms. Little is known re- garding the life history of most species. Thorson (1940, pages 192-193) has figured the egg mass and early larva of Tonna (?) maculosa while Knudsen (1950, pages 97-98, Fig. 8) shows an egg mass that he attributes to Tonna costatum (Menke). These flat gelatinous masses with "egg spaces" containing about 100 eggs each may have from 350-660 thousand embryos. The eggs are relatively large, 300 microns in diameter, and the early larvae of Tonna ( ?) maculosa illustrated by Thorson (1940, page 193, Fig. 15) is about 380 microns in longest dimension. Measurement of the embryonic shell on Tonna galca closely corresponds to this, being about 400 microns in length. The two equatorial Atlantic species are Tonna galca (Linne) and Tonna maculosa (Dillwyn). Larvae of both these species are encountered in the Gulf Stream. The early intermediate stage larva of Tonna galca has long, straight, fairly stout, spirally arranged, periostracal spines. This stage appears to be similar to that illustrated by Lebour (1945, page 475, Fig. 21) as Tonna perdi.v (Linne). In the later intermediate stage larva, as a new whorl is added, additional spines are also formed (Fig. 2, b, c). The globose shell is smooth, white, and glossy. When completely developed the larva has usually lost all its periostracal spines and then measures about 3.5 mm in length (Fig. 2, d, f). The typical long, drawn-out basal periostracal spine seen in the earlier stages is also frequently lost. At this late stage of development the larva of Tonna galca may be confused with that of Tonna maculosa. The two differ, however, in the shape of the aperture FIGURE 3. Larval shells, opercula and protoconchs of some species belonging to the fami- lies Muricidae, Ovulidae, Architectonicidae, and Neritidae : (a) Thais (?) rustica veliger larva, Eastern Atlantic from off West Africa, 2259'N, 2030'W ; (b) protoconch of Thais rustica, Western Atlantic, Port Royal, Jamaica (USNM 442303) ; (c) protoconch of Pedicularia sicula Jccussata, off coast of Georgia, Western Atlantic, depth 792 m, U. S. Fish. Comm. Sta. No. 2415 (USNM 108408) ; (d) Thais hacmastoina larval shell, Western Atlantic east of Bahamas, 2506'N, 6742'W; (e) protoconch of Thais haemastoma, Western Atlantic, Corpus Christi Bay, Texas (USNM 125561); (f) larval shell of Pedicularia sicula. Western Atlantic, Gulf Stream, 3627'N, 6801'W; (g) larval shell of Pedicularia sicula, Eastern Atlantic, off Azores, 3714'N, 2744'W; (h) larval shell of Thais hacmastoma. Western Atlantic east of Bahamas, same station as (d) above; (i) larval shell of Thais hacmastoina, Mid-equatorial Atlantic, South Equatorial Current, 0234'N, 2403'W; (j) larval shell of Thais hacmastoina, Eastern Atlantic off West Africa, 0800'N, 1515'W; (k) Philif>fia krcbsii larval shell, fully devel- oped, Western Atlantic, North Atlantic Drift, 3926'N, 4405'W ; (1) PhiUppia krcbsii larval shell, same as (k) above; (m) operculum of Philip f>ia krcbsii, inner view and side view show- ing apophysis; (n) three larval shells of Smaragdia riridis. Western Atlantic, Gulf Stream, 3040'N, 7340'W; (o) operculum taken from larva of Smaragdia riridis, northern end of Gulf Stream, Western Atlantic, 4126'N, 5545'W. Arrow shows position of minute reinforcing bar; (p) two larval shells of Smaragdia riridis. Eastern Atlantic off West Africa, 0720'S, 0723'E ; (q) protoconch of Smaragdia viridis I'iridcmaris showing larval shell slightly im- mersed, Western Atlantic, Castle Harbor, Bermuda, depth 7-9 m (ANSP 267608). Scale = 1 mm for all figures. Numbers prefixed by "ANSP" refer to lot numbers in the collection of the Academy of Natural Sciences of Philadelphia. Numbers prefixed by "USNM" are lot numbers from the collection of the United States National Museum, Washington, D. C. 294 RUDOLF S. SCHELTEMA (compare Fig. 2, <1 with 2, g) ; the length, number, and angle of the spines (if still present, Fig. 2, k ) ; the apical angle of the spire, usually greater than 100 in Tonna galea (Fig. 2, d, e) and less than 100 in Tonna maculosa (Fig. 2, g, j) ; and the shape of the reinforcing bar on the operculum whereby it is attached to the foot musculature (Fig. 2, h, i), being curved in Tonna maculosa but straight in Tonna (jalca. Simroth (1895, PI. IV) under the larval name Macgillivrayia has figured Tonna larvae. The large operculum (PI. IV, Fig. 7) is that of Tonna galea, but Figs. 1-3 on the same plate are probably Tonna maculosa. Larvae of Tonna galea were found throughout the North Atlantic gyre, in the South Equatorial Current, off the west coast of Africa, in the mid tropical Atlantic, off the northeast coast of Brazil, and in the Caribbean Sea, whereas those of Tonna maculosa were found only in the western Atlantic (Fig. 7). Family Cassidae The family Cassidae, known in the vernacular as helmet or bonnet shells, is closely related to the two preceding families (i.e., Tonnidae and Cymatiidae) and like them has a worldwide distribution in tropical and warm temperate waters. The adults live on sandy bottoms or in Thalassia beds and feed upon echinoids. The egg mass of P ha Hum granulatum is made up of capsules built into tower-like structures (vide Abbott, 1968, PI. I), and has up to 300 eggs in each capsule. The larvae most frequently encountered in the North Atlantic that belong to the family Cassidae are those of Phalium granulatum (Born). This larva is smaller than any of those previously considered here, the shell being somewhat less than 2.5 mm at settlement. The shell is globose (Fig. 2, n, o), though with a somewhat flattened spire, particularly in earlier stages (Fig. 2, p). The sutures are ad- pressed; the first H whorls are amber-brown while the remainder of the shell is smooth, glossy, and white. The operculum (Fig. 2, m) is distinctive with its two strengthening supports. Two other species of PJialhtm occur in the tropical North Atlantic, (1) Phalium coronadoi (Crosse), a rarely collected species known only from the coast of North Carolina and Cuba, and (2) Phalium saburon (Bruguiere), a species whose distribution overlaps that of Phalium granulatum, being found in the Azores, Bay of Biscay, and southward along the west coast of Africa to Ghana. Proto- conchs of both these species have been examined and compared to Phalium gran- ulatum (Fig. 2, 1) ; Phalium coronadoi and Phalium saburon both have smaller larvae with wider apical angles. The geographical distribution of Phalium granulatum veliger larvae includes the Gulf Stream, the North Atlantic Drift, the Canary Current off the west coast of Africa, and the South Equatorial Current (Fig. 8). Family Muricidae The Muricidae are a family of carnivorous gastropods which feed on bivalves, barnacles, and other gastropods. The genus Thais (sometimes relegated to a sepa- rate family Thaididae) is represented in the tropical Atlantic by at least four very common species which definitely are known to have pelagic phytoplankto- trophic larvae. The egg capsules of the various species of Thais are laid singly DISPERSAL AND GENETIC EXCHANGE 295 rather than in distinctive masses as in some of the Cymatiidae and in Phalium granulatum and in most species are usually between 5 and 8 mm in height. The larval shell of Thais haemastoma (Linne) was illustrated by Craven in 1877 as the species Sinusigera colbeauiana. Later in 1883 he figured an early post-larval shell which this time he correctly identified as Thais haemastoma. Dautzenberg (1889, PI. II, Fig. 5) has figured an early post-larval shell from the Azores showing details of the protoconch of Thais haemastoma. Recently Moore (1961) accurately described the shell of the fully developed larva, but his FIGURE 4. Geographical distribution of plankton tows upon which the present study is based. The darkened circles indicate stations where teleplanic gastropod veligers were found ; the open circles denote stations where no such gastropod larvae were taken. In instances when plankton tows were less than one degree apart, points shown on the above and subsequent charts may represent more than one tow. Some stations were repeated at the same location at different times of the year. Locations included in the figure are only those for which entire plankton samples have been examined. Stations appearing on sub- sequent charts but not on this figure represent positive records from loaned plankton tows. They were not used for negative records because material had been previously sorted from them or preservation was poor, and because they had been collected differently. The dashed lines indicate the 21 C summer and winter isotherms. Major surface currents in the North and tropical Atlantic are designated by arrows and may be identified by reference to letters : A = Gulf Stream, B = North Atlantic Drift, C = Canary Current, D = North Equa- torial Current, E = Equatorial Countercurrent, and F = South Equatorial Current. 296 RUDOLF S. SCHELTEMA accompanying illustration (page 26) is confusing and lacks necessary details of shell ornamentation. The larval shell of Thais haemastoma at the completion of growth is about 1.5 mm in length (Fig. 3. d, h, i, j). There are five whorls and an apical embryonic shell. The base of the columella has a cancellate pattern formed from seven or more axial and about an equal number of spiral cords. Below the sutures there are minute axial riblets. The third whorl has a single ribbed spiral cord which continues on the succeeding whorls. The larva is a typical Simtsigera with a long projection separating the adapical from the basal portion of the outer lip. A varix, frequently with a serrate adapertural edge, is formed in the adapical portion of the outer lip. The body whorl may have minute punctae extending back about ] whorl from the edge of the outer lip. The body whorl is bounded adapically by a spiral band which is continuous with the varix. Certain minute features are lost with preservation and the shell tends to become "chalky" with time. There are closely related species in the Thais haemastoma complex (e.g., Thais rustica [Lamarck], vide Clench, 1947) which seem to have larvae quite similar to Thais haemastoma scnso stricto (compare Fig. 3, b and 3, e ; the latter protoconch is worn). Certain larvae which were taken off the west coast of Africa and which otherwise appeared identical, differed in having a less calcified shell, being slightly larger, having a somewhat greater apical angle, and having a narrower varix lacking a serrate edge (Fig. 3, a, off west Africa should be compared with Fig. 3, d, h, i, j). These somewhat different west African larvae most closely resemble the protoconch of Thais rustica, though this species in the eastern Atlantic has been reported only from St. Helena. The geographical distribution of TJiais haemastoma larvae includes the Gulf Stream, the North Altantic Drift, the Canary Current, and the South Equatorial Current (Fig. 9). Family Architectonicidae Architectonicidae or sundial shells are a family of subtidal gastropods repre- sented in tropical waters throughout the world. Some subtropical and warm temperate forms are also found. The adults of most known species are restricted to depths of a few meters to about 200 meters near the edge of the continental shelf. Their food is apparently confined to coelenterates, particularly the zoanthid genera Palythoa and Zoanthus (Robertson, 1967) and also the polyps of scleractinians and actiniarians (Robertson, Scheltema and Adams, 1970). Many Atlantic species are now known to have teleplanic larvae. Moreover, of the ten species found by Marche-Marchad (1969) along the coast of west Africa, seven are also known from the West Indies. The larvae of at least some of these species are euryther- mal. PhUippia krebsii has been found in temperatures from 18.9 to 28.9 C. Other architectonicid larvae have been found at temperatures as low as 13.5 C. The larval shell of PhUippia krebsii (Morch) is illustrated by Robertson (1964, Fig. 4 and Figs. 7-10) and Robertson et al. (1970, Figs. 3 and 6). It is hyperstrophic, smooth, glossy, and more or less transparent (Fig. 3, k, 1). The operculum is circular, externally concave, with a central internal apophysis that extends into the foot musculature (Fig. 3, m). The apophysis is constricted DISPERSAL AND GENETIC EXCHANGE 297 at the base as is also found in the adults belonging' to the subgenus Psila.vis (Rob- ertson, 1970a, Fig. 10, page 74). The living larva has four velar lobes with pale orange pigmentation along their borders. The anterior of the foot is darkly pig- mented except for a conspicuous white food-rejection tract passing from beneath 60 30 FIGURE 5. Geographical distribution of the larvae of Cymatium parthenopeum (von Salis) and Cymatium nicobaricuui (Roding) : 1 = stations with Cymatium parthenopeum, 2 stations with Cymatium nicobaricum, 3 = stations where both species were taken. Small open circles are stations where neither species was encountered in tows. Stippled areas show the approximate geographical regions where adult Cymatium parthenopeum is to be found to depths of 63 meters (Clench and Turner, 1957). The geographical distribution of post-larval Cymatium nicobaricum roughly corresponds to that of Cymatium parthenopeum in the west- ern Atlantic, but in the eastern Atlantic Cymatium nicobaricum is known only from the Ma- deira and Canary Islands (Clench and Turner, 1957; Odhner, 1931; Dautzenberg, 1890). Arrows show direction of surface currents. the mouth to the anterior base of the foot. Philippia krebsii is readily distinguished from other species of Architectonicidae in the Atlantic by its large size (1.5 mm along the longest axis) and by the presence of a prominent anal keel (vide Robert- son, 1964, page 6, Table I). The geographical distribution of Philippia krebsii larvae essentially includes all the major surface currents of the tropical and tem- perate North Atlantic Ocean (Fig. 10). 298 RUDOLF S. SCHELTEMA Family Neritidae The larval shell of the neritid Smaragdia viridis is smooth, glossy white with adpressed sutures (Fig. 3, n, p, q). The embryonic shell is only 120 microns longest dimension, but the fully developed larva is 650 microns along this axis. The living larva has a simple bilobed velum and two small tentacles. The diges- tive gland has a green, almost chartreuse, cast which is retained for several months after preservation. The small operculum is paucispiral and has a minute reinforc- ing bar next to the nucleus (Fig. 3, o, see arrow). In later life the adult shell is cryptically colored and is frequently found associated with Thalassia in the west- ern Atlantic and Posidonia in the Mediterranean. Possibly it may be dispersed on drifting turtle grass (vide Menzies, Zaneveld and Pratt, 1967). The geographical distribution of Smaragdia viridis veliger larvae includes the Gulf Stream, the western portion of the North Atlantic Drift, the Canary Current, the South Equatorial Current, and Caribbean Sea ( Fig. 11). Family Ovulidae The veliger larva of the ovulid Pedicularia sicula Swainson is an ovate, amber- brown Sinusigera. The embryonic shell, somewhat lighter in color and approxi- mately 125 microns in length, is followed by two spiral whorls with axial costae and a single spiral cord. The two remaining whorls have oblique reticulate ornamen- tation (Fig. 3, f, g). Sometimes this ornamentation becomes interrupted and confused. The outer lip of the fully developed larval shell has a thickened varix and the typical beak, a projection found on all Sinusig era-type veligers. The fully developed larval shell is between 0.9 and 1.0 mm in length. The protoconch of the post-larva, as is common to members of the Cypraeidae and Ovulidae, becomes completely hidden in the adult by succeeding whorls. It is therefore necessary to have relatively young specimens for comparison with the larva (Fig. 3, c). A number of names have been applied to the species Pedicularia sicula (vide Allen, 1956). The limited amount of material examined reveals no good reason for separating the eastern and western Atlantic forms of Pedicularia into separate species, and the western Atlantic form, usually known as Pedicularia, decussata (Gould), should probably be considered no more than a subspecies of the eastern Atlantic and Mediterranean Pedicularia sicula. Specimens of North Atlantic spe- cies are rare in museum collections so the degree of variability is not well under- stood. The protoconch of a Pacific species of Pedicularia was first figured by John Dennis MacDonald (1858), assistant-surgeon on H. M. S. HERALD in a paper entitled, "On the probable metamorphosis of Pedicularia and other forms; affording presumptive evidence that the pelagic Gastropoda, so called, are not adult forms, but, as it were, the larvae of well-known genera. . . ." Dautzenberg (1889) has very well illustrated the protoconch of an Azorean specimen of Pedicularia sicula (PI. IV, Fig. 2a) and the larva described as an unknown ovulid by Lebour (1945, pages 474-475, Fig. 21) off Bermuda is probably Pedicularia sicula (s.L). Hedley (1903, Fig. 69, page 342) also figured the protoconch of a Pedicularia species from off New South Wales. Other larvae from the family Ovulidae which have DISPERSAL AND GENETIC EXCHANGE 299 been previously described are Sinniia patula by Lebour (1932), Simnia spelta and (?) Ovula ovum by Thiriot-Quievreux (1967). All tbese forms, though having certain similarities, can be distinguished from Pedicularia sicnla. Likewise, the protoconchs of Cypraeidae in the Museum National d'Histoire Naturelle in Paris figured by Ranson (1967) cannot be confused with Pedicularia because of differ- ences in the shell ornamentation. However, because of the scarcity of suitable mate- so FIGURE 6. Geographical distribution of Charoma variegata (Lamarck) larvae: filled black circles = Charonia variegata larvae ; small open circles = stations where Charonia variegata larvae did not occur. The stippling shows approximate geographical regions from which post-larval Charonia variegata (s.s.) are known (Clench and Turner, 1957) and does not include distributional data of Charonia nodifera (Larmarck). Arrows show direction of surface currents. rial with protoconchs, it has not been possible to make extensive comparisons with other forms of Cypraeidae and Ovulidae in the tropical North Atlantic. The larvae of Pedicularia sicnla (s.l.) were found in the Gulf Stream, the North Atlantic Drift, the Mediterranean Sea, and the North and South Equatorial Currents (Fig. 12). Relationship between the North and Equatorial Atlantic circulation and dispersal of teleplanic gastropod veligcrs. Two basic patterns of geographical distribution are 300 RUDOLF S. SCHELTEMA distinguishable among teleplanic gastropod veligers. The first of these is that found in such forms as Charonia varici/ata, Tonna galca, Philippia krcbsii, Pedicularia sicula (s.L), and Cymatium parthenopeum. Larvae of these species were found in all three trans-Atlantic surface currents, viz., the North Atlantic Drift and the North and South Equatorial Currents. Some species such as Phalium granulatum, 30 FIGURE 7. Geographical distribution of the larvae of Tonna galca (Linne) (= 1) and Tonna maculosa (Dillwyn) (=2). Small open circles denote stations where no species of Tonna veligers were found. Stippled areas show geographical regions where post-larval Tonna galea are found to the edge of the continental shelf (Turner, 1946). Tonna maculosa post-larvae are known in the eastern Atlantic only from scattered records, specifically in the Canaries (see: Fischer-Piette and Nickles [1946, pages 45-48] for relevant literature.) Arrows indicate direction of surface currents. Thais haemastoma, and Smaragdia virldis were missing in the North Equatorial Current but moved in a westerly direction on the South Equatorial Current only. The second basic distributional pattern is that in which larvae were missing in the North Atlantic Drift. Such a distribution is not represented among the ten species considered here, but examples may be found in the gastropod genera Bursa and Cypraea. These genera evidently have larvae which are dispersed eastwardly on the Equatorial Undercurrent, which provides them with a means for larval trans- DISPERSAL AND GENETIC EXCHANGE 301 port from the South American to the West African coast (vide Scheltema, 1968). The first distributional pattern is restricted to forms which have eurythermal larvae ; otherwise they would not survive the low temperatures of the North At- lantic Drift in which they regularly occur (see 21 C isotherm. Fig. 4). Veligers along the entire periphery of the North Atlantic gyre theoretically can be carried between any two points along its edge. The only limitation to the distance a 30 FIGURE 8. Geographical distribution of the veliger larvae of Phalium granulatum (Born) : filled black circles = larvae of Plniliiiiii granulatum; small open circles = plankton stations where Phalium granulatum was missing from tows. Stippled areas denote geographical regions where Phalium granulatum post-larvae have been found. The eastern and western Atlantic forms are designated subspecifically as Phalium granulatum nndulatiim and Phalium granulatum granulatum, respectively (Abbott, 1968). Arrows show direction of surface currents. particular larva may be carried is the length of pelagic life in relation to current velocity. The second distributional pattern appears to be characteristic of stenothermal tropical forms, excluded from the colder northern route across the Atlantic but effectively carried eastward in the warm Equatorial Undercurrent. There may of course also be some species in which some factors other than temperature limit the geographical distribution of the larvae. 302 RUDOLF S. SCHELTEMA Duration of pelagic larval life and its relationship to the -velocity of trans-Atlantic Ocean currents Ekman (1953) in his volume Zoogeograpliy of the Sea concluded that trans- oceanic drift of larvae is unlikely because "on the whole the planktonic stage is too J o. ' T FIGURE 9. Geographical distribution of Thais haemastoma larvae. Thais haemastoma is known as post-larvae in the eastern Atlantic from two subspecies, Thais haemastoma haema- stoma (=1) and Thais haemastoma jorbcsi (2); and in the western Atlantic (in addition to the subspecies Thais haemastoma haemastoma), from two additional "subspecies," Thais haemastoma floridana (=3) and Thais haemastoma haysae (=4). The adult distribution of these various "subspecies" is taken from Clench (1947) and Knudsen (1956): 5 = larvae of Thais haemastoma; 6 = larvae belonging to the Thais liacmastoma complex but whose spe- cific identity is uncertain (see Figure 3, a). Small open circles are stations where Thais haemastoma was not found. Arrow show direction of surface currents. short to account for dispersal over thousands of kilometers," but he added that "our knowledge of the duration of larval stages in most animals is very meagre" (page 21). Thorson (1961) in a stimulating theoretical paper has summarized data on the length of pelagic larval life of 195 species of bottom invertebrates and concluded that in most instances the period of larval development was too short to account for larval transport across major ocean basins. Among the 18 proso- DISPERSAL AND GENETIC EXCHANGE 303 branch species that were considered, not one had a larva with a development of over 9 weeks. Thorson specifically excluded members of tropical gastropod genera from consideration, because he lacked data for the duration of their pelagic larval life. He acknowledged, however, that "true long-distance larvae . . . seem mainly to be associated with tropical seas and to be rare or lacking in temperate and cold regions" (Thorson, 1961, page 473). Mileikovsky (1966) also believes "that the FIGURE 10. Geographical distribution of Philippia krebsii (Morch) larvae: rilled black circles = Philippia krebsii larvae ; small open circles = plankton stations where Philippia krebsii did not occur. The stippling indicates approximate geographical distribution of post-larvae of Philippia krebsii (Robertson, 1964; Robertson et al., 1970), which are found to 200 meters depth. The species is presently known in the eastern Atlantic only from the Canary and Cape Verde Islands. Arrows show direction of surface currents. larval drift of all benthic invertebrates in warmer waters of currents in the tropical regions is more prolonged and lengthy than in their counterparts in waters of temperate and high latitude currents" (page 400, in translation). Since Thorson in his summary article was considering only cold-temperate and boreal species, this information on the length of larval development should not be extrapolated to make conclusions on the dispersal of tropical gastropods in ocean currents. Lacking experimental data, an approximate measure of the minimum length of pelagic larval development must be obtained indirectly. There are very few 304 RUDOLF S. SCHELTEMA data on the growth of any gastropod veliger larvae. However, the growth rate of the prosobranch Nassarius obsolchts (Say) is well known under a variety of laboratory conditions (Scheltema, 1%7). Let us assume as an approximation that the average shell growth of Nassarius obsolcfus between 20 and 25 is repre- sentative of all phytoplanktotrophic gastropod veliger larvae. If we use this rate and apply it to the measurement of Cymatium parthenopeum larvae an estimated value for the minimum time of pelagic larval development may be obtained. Thus in the species Cyniatiuni parthenopeum the total pelagic shell growth is represented by the difference in size between the small embryonic shell on top of the cancellate whorl (Fig. 1, j) and that of the fully developed, ready-to-settle larva, almost 2 mm long (Fig. 1, f ). This method has been applied to the other species considered (Table I, left-hand column). Admittedly the estimates are only rough, and the TABLE I Estimated duration of the pelagic stage in ten gastropod species having teleplanic larvae Estimated avg. days Species required for growth ("developmental Days held in culture ("delay period") Total days pelagic period")* Cymatium nicobaricum 207 113** 320 Cymatium parthenopeum 155 138 293 Charonia variegata 219 57 276 Tonna galea 148 94 242 Tonna maculosa 198 (198) Phalium granulatuni 107 (107) Tluiis haemastoma 62 28** 90 Philippia krebsii 67 ca. 7** 74 Smaragdia viridis 25 30 55 Pedicularia sicula 42 (42) * Based on average growth rate of Nassarius obsoletus in laboratory culture (vide Scheltema, 1967, Table I, page 261). ** These species were successfully held to settlement in the laboratory. accuracy of the method depends partly upon the shell geometry of the species being studied and upon the assumption that early growth is linear. However, in the only case where other data are available, the estimate obtained by this method is in fairly close agreement. Hence, Butler (1954) estimated from field data a mini- mum pelagic life of about 50 days for Thais haemastoma, not far off from my estimate of 62 days. The length of pelagic larval development determined by the method just out- lined will most certainly tend to underestimate. Larvae for which the proper algal food is known usually grow faster under ideal laboratory conditions than in their natural environment. Moreover, it is well known that many gastropod species can delay settlement. For example, certain predaceous nudibranchs will delay metamorphosis in the absence of their normal coelenterate prey (Thompson, 1958, 1962; Hadfield and Karlson, 1969), and Fretter and Graham (1962, page 448) report that metamorphosis of Odostomia (Pyramidellidae) is accelerated by the presence of the polychaete worm Pomatoceros and delayed by its absence. Other DISPERSAL AND GENETIC EXCHANGE 305 gastropod larvae have a more general response. Thus the intertidal snail Nassarius obsoletus, primarily a deposit-feeder, delays settlement in the absence of an appro- priate sediment (Scheltema, 1956, 1961), and Kiseleva (1966, 1967) has shown that the prosobranchs Rissoa splendida and Bittiitin reticulatum settle preferentially on the alga Cystoseira. In some species settlement may be simply a tactile response to a surface, apparently the case with the architectonicid Philippia krcbsii. * >: y^4 ' / *. ,' ; r ^' "** " * ' . o 6Q FIGURE 11. Geographical distribution of the larvae of Smaragdia riridis (Linne) : filled black circles = larvae of Smaragdia riridis; small open circles = plankton stations where Smaragdia riridis was not found. Stippled areas show approximate regions where post-larval Smaragdia riridis is found ; eastern and western Atlantic subspecies are respectively designated Smaragdia riridis riridis (Linne) and Smaragdia viridis riridcmaris (Alaury). Distributional data are compiled from the collections of the Museum of Comparative Zoology and the Phila- delphia Academy of Natural Sciences. Arrows show direction of surface currents. The larval development of many phytoplanktotrophic gastropod veligers can be divided for convenience into two periods (Scheltema, 1967). The first of these one of rapid growth and morphological development may be termed the "develop- mental period." Its end is marked by the inflection point on the cumulative growth curve and the completion of external morphological development (Schel- tema, 1967, page 260, Fig. 4). Larvae are ready to settle at this time. Estimates of the duration of pelagic larval development based on tin- differences in size be- 306 RUDOLF S. SCHELTEMA tween the embryonic and fully developed larval shell are roughly equivalent to the length of the developmental period. The second period, termed the "delay period," is terminated by a settlement response. In species with very specific post-larval requirements, such a response can be very precise. Very little further growth occurs during the delay period. Many veliger larvae captured in the open sea have completed the developmental period and reached the delay period. Indeed, it was the constancy in size that confused Craven (1883) into believing that the veliger larvae of the Sinusigera- type (Fig. 3, a-j) were fully developed holoplanktonic species. He writes, "The two principal arguments in favor of this view were the great distances from land at which they were often found and the constant dimensions of each species" (page 141). It is the delay period which gives to many species the greatest flexibility in the length of their pelagic larval life. TABLE II Velocity of the eastwardly and westwardly flowing currents of the north and tropical Atlantic Ocean Estimated number of Estimated current days required for Region velocity km/hr trans-Atlantic drift North Atlantic Drift (Bahamas to Azores) 0.5-1.3 128-300 (Bahamas to N. W. Africa) 0.9 400 North Equatorial Current (N. W. Africa to West Indies) 0.9-1.2 128-171 South Equatorial Current (Gulf of Guinea to Brazil) 1.0-3.2 60-154 Equatorial Undercurrent (Brazil to Gulf of Guinea) 2.0 96 Data from (1) Pilot Charts, U. S. Hydrographic Office; (2) direct measurements: Dickson and Evans, 1956, North Equatorial Current; Stalcupand Metcalf, 1966, Equatorial Undercurrent; and (3) drift bottle data : Guppy, 1917 ; Bumpus and Lauzier, 1965 ; and W. H.O.I, files courtesy of Mr. Dean F. Bumpus. I have attempted to obtain some measure of the potential length of the delay period experimentally by culttiring larvae caught at sea in plankton nets. The endurance of some veliger species under laboratory conditions is truly remarkable (Table I, column designated "delay period"). Thus Cymatium nicobaricum was held for 113 days and Cymatium partlicnopeum for 138 days. Not all species were successfully held in culture, and some species were not caught during the time that culture observations were made. Addition of the estimated time of the developmental period to that of the delay period probably gives a better estimate of the maximum pelagic life of a species of gastropod than the estimated developmental period alone (last column, Table I). One may compare these data for the duration of pelagic life with the current velocities of the major eastwardly and westwardly flowing currents of the North Atlantic Ocean (Table II). Under favorable hydrographic conditions the duration of pelagic larval life of Cymatium nicobaricum, Cyiiiatiitiu parthenopeum, Charonia variegata, Tonna galea, Tonna macitlosa, and probably Phaliirm c/raniilatnm are all DISPERSAL AND GENETIC EXCHANGE 307 sufficiently long to account for their trans-Atlantic dispersal over any of the four major routes. The remaining four species, Thais haemastoiiia. I'liilippia. krcbsii, Smaraydia viridis, and Pedicularia sicula seem to have but little opportunity for trans-Atlantic transport either by way of the North Atlantic Drift or the North Equatorial Cur- rent, and only under the most favorable conditions could cross in the South Equa- 60 FIGURE 12. Geographical distribution of the larvae of Pedicularia sicnla Swainson : filled black circles = larvae of Pedicularia sicula ; small open circles = stations at which larvae of Pedicularia sicula were not taken. Stippled areas designate geographical regions from which the post-larvae are known (Allen, 1956). The western Atlantic form Pedicularia decussata Gould should probably be regarded as no more than a subspecies of Pedicularia sicula (s.s.). Arrows show direction of surface currents. torial Current and the Equatorial Undercurrent. Yet three of the four remaining species under consideration have larvae throughout the entire North Atlantic Drift and two are found also along the North Equatorial Current (see Figs. 9, 10, 11, and 12). The answer to this paradox must be that the duration of pelagic larval devel- opment among these species is longer than estimated. Three possibilities suggest themselves. The first is that the developmental period is much extended under conditions of the open sea. It is known, for example, from observations in the 308 RUDOLF S. SCHELTEMA laboratory that when food is less than optimal, length of the developmental period can be very greatly extended, in some species evidently for months (Millar and Scott, 1967; Turner and Johnson, 1968; personal unpublished observations). It is of course also well known that the amount of phytoplankton in the open sea is very substantially less than in neritic waters. Laboratory experiments show that a reduction in water temperature may extend the length of the developmental period of gastropod veligers up to fifty per cent (Scheltema, 1967, in Nassarins obsoletus). This observation is particularly relevant to the duration of pelagic larval develop- ment in the North Atlantic Drift as during some months of the year the surface- water temperature there can be ten degrees lower than in the tropical Atlantic. The second possibility is that the potential length of the delay period may extend over a longer time than has been estimated. Species which as adults have very particular habitats or which prey on specific organisms as food may be suspected of having marked settlement responses. Thus Pedicularia sicula, as an adult found only on stylasterine hydrocorals (Robertson, 1970b), may be expected to have a highly developed settlement response. A delay in settlement until a suitable envi- ronment is encountered can greatly increase the length of the delay period and the total duration of pelagic existence. The third possibility is that certain species having relatively small egg capsules may attach these to floating objects, and that the larvae found in the open sea are released subsequently from these capsules. This is clearly not a possibility for the first six species in Table I as their egg masses are quite large and are never attached to floating objects. The capsules of Thais haemastoma are relatively small and conceivably could be attached to drifting materials. It is possible that occa- sionally the adults of certain small species such as Smaragdia inridis are mechani- cally transported, in this instance on turtle grass with which Smaragdia iriridis is normally associated. Clench (1947, page 76) cites the case of a specimen of Thais Iiaenwstoina taken from a floating log 300 miles northwest of the Madeira Islands. However, floating objects are seldom seen in the open sea, even in sea-lanes, and in any case could not possibly account for the large numbers of larvae found there. It is now possible to summarize more specifically the case of the four species in question. Thais haemastoma is the only species on the list in Table I which shows marked seasonally in its reproduction. Its larva has been found in the North Atlantic Drift only in the autumn, when the falling sea-surface temperature may extend the duration of its larval development (sea-surface temperature is about 21 C during October). Philip pia krebsii is probably able to extend its larval development beyond that suggested by the data in Table I (see also Robertson, 1964, for estimate of length of larval development). In the absence of a tactile stimulus, it can probably greatly extend the duration of its pelagic existence (Robertson et a!.. 1970). Probably the larvae of Smaragdia viridis are infrequently dispersed across the North Atlantic Drift because of their relatively short planktonic development. The alternative of the adults' drifting on turtle grass is, however, quite possible. Finally, the delay period of Pedicularia sicula larvae, for which there are no data, probably can be extended over long periods. This seems the most reason- DISPERSAL AND GENETIC EXCHANGE 309 able hypothesis to explain its widespread occurrence in all the North Atlantic sur- face currents (Fig. 12). Probability and frequency of long-distance dispersal of vcliger larvae Estimates of the probability and frequency of dispersal from drift-bottle data and survival rates Larval transport between continents has often been likened to a sweepstakes, because it is believed that the odds that any specific larva will drift across the Atlantic Basin are very small. It is generally concluded from such a quasi-quanti- tative statement that only rarely can a larva be dispersed in either direction across the Atlantic. The probability of successful trans-Atlantic larval dispersal, />. depends upon the drift coefficient, />,/, and the survival coefficient, /\, such that: P == PdPs 0) TABLE 1 1 1 Percentage drift-bottles released along the North American Coast and subsequently recovered in the Eastern Atlantic Ocean (based on 156,276 bottles)* Percentage of total bottles Region of recovery released Northern Europe (Great Britain, Ireland, Denmark, Norway) 0.10 Azores 0.05 Southern Europe (France, Spain, Portugal) 0.04 Northwest Africa (Including Canary Islands) 0.01 Total percentage recovery in Eastern Atlantic 0.20 * Computed from data in files of Woods Hole Oceanographic Institution and made available through the courtesy of Mr. Dean F. Bumpus. Specifically, the drift coefficient can be defined as the likelihood that larvae will enter and remain within a current that will transport them across the Atlantic Ocean Basin rather than back onto their own shore. The value of />,/ depends upon (1) the distance between the parent population and the ultimate destination of the larvae, that is, the greater the distance the less likelihood for success, and (2) the size of the target area, or, the amount of coastline which the larvae are likely to encounter. An approximate value for the drift coefficient may be derived from a knowledge of drift-bottle and drift-card recoveries. It is assumed that if a larva remains within the surface currents, it will be carried in the same manner as the drift bottles or cards and that consequently the likelihood of its transport across the Atlantic Ocean, other factors being equal, will be similar to that of such bottles or cards. The per cent recovery in both directions, west to east along the North Atlantic Drift and east to west along the North Equatorial Current, is re- markably similar (Tables III and IV), even though the data are from entirely different sources and the sample sizes differ by two orders of magnitude. From 310 RUDOLF S. SCHELTEMA these recovery data an average value of 1 .9 X 10" '' may be assigned to p tl . As probaMy not all bottles that actually make the crossing are eventually recovered, the value of />,/ is doubtlessly conservative. The survival coefficient /\, is the probability that a pelagic phytoplanktotropic larva u'ill survive to settlement without regard to where it may be dispersed. It must include possibility of mortality by predation, disease, or starvation. Because such information is not known for any tropical gastropod veliger, the value for larval survival, />.., must be derived by an indirect method. Over the long course of time it is evident that each member of a species must reproduce itself at least once during its lifetime ; otherwise the species will decrease in numbers to eventual extinction. Therefore the minimum survival of any viable species must on the average be such that for each member there is produced during TABLE IV Percentage drift-cards released along the North-West African Coast and subsequently recovered in the Western Atlantic Ocean (based on 1,X()0 cards)* Percentage Region of recovery of total Anguilla (British West Indies) 0.06 East Coast of Florida (Cocoa Beach) 0.06 Outer Banks (North Carolina) 0.06 Total percentage recovery in Western Atlantic 0.1 8 * Data computed from Stander, Shannon and Campbell, 1969, Fig. 1, page 296; Table I, page 294-295. its lifetime one larva that ultimately survives to sexual maturity. Thus in a dioecious species : P. = 7 P) tf where t f is the fecundity or total number of eggs produced per female. The numerator "2" in the expression on the right-hand side of the equation assumes an equal number of males and females in the population. The equation further assumes that chances for survival are equally good in coastal and off-shore waters. The expression probably tends somewhat to underestimate larval survival because it includes post-larval mortality. However, the latter is regarded as being very small compared to the mortality of larvae. Selected examples can now be considered. Members of the genus Tonna are known to produce broad, flat, ribbon-like egg masses with ca. 300,000 to 600,000 eggs. Thorson (1940, page 192) estimated that Tonna (?) macnlosa in the Iranian Gulf produced 350,000 eggs and Knudsen (1950. page 98) describes an egg mass of a Tonna from off West Africa with 6(>0, 000 eggs. It is not known how many egg masses a female produces, but for ease in calculation let us assume that each spawns five per year over a two-year period and every mass contains DISPERSAL AND GENETIC EXCHANGE 311 500,000 eggs. The probability that a particular larva from an egg mass will sur- vive and ultimately reach the opposite coast of the Atlantic Basin will then be : p ---- Pd-- = [1.9 X 10-' ! ][4 X 10- 7 ] == 7.6 X 10- 10 (3) or approximately one in eighty billion. Butler (1954) reports that each female Thais haemastoma produces 500,000 eggs per season, and a result similar to that of Tonna will then be obtained. Thus the probability that any specific larva will be transported across the Atlantic Basin is very small. However, what concerns us here is not, as in a sweepstakes, the likelihood that a particular horse will finish, but rather the total number that will complete the race. It does not only matter what the odds for dispersal of a particular larva may be; what is also important is the total number of larvae that are dispersed (vide Simpson, 1952, pages 167-168). If enough larvae are prodviced, the chances may become very good that at least some veligers will successfully drift across the Atlantic. Therefore the frequency of trans-Atlantic dispersal is dependent, not only on the probability of such an event, but also on the number of larvae produced by the parent population. Thus, d ---- pt (4) where d is the frequency of dispersal, and t is the number of larvae produced by a population. The value of Ms a function of the population size, t v , and the num- ber of eggs produced per female, //. In a dioecious species, assuming that half of the population is female, the total number of larvae produced is t p /2-t f . As- suming once more that the survival coefficient is 2/t f and substituting in (4) above, d = which is simply to write that the frequency of dispersal depends upon (1) the probability that larvae will be carried across the Atlantic by the major ocean cur- rents and (2) the size of the parent population from which the larvae originate. Since />rf). This value is approximate for pa, in the North and Equatorial Atlantic Ocean. The curve shows the ex- pected frequency of larval dispersal (d) computed from various assumed population sizes when only one larva survives for each adult member of the parent population. Fluctuations in larval survival for most benthic species are expected to lie within the stippled region (vide Fig. 11; Thorson, 1946, page 439, Fig. 199; Coe, 1956; Loosanoff, 1964, 1966). years it fell between 10 and 100 organisms ; for nine years it was between 100 and 1000 organisms, and finally during two years it fell between 1000 and 10,000 organisms (Fig. 14). From these and other available data (Thorson, 1946; Coe, 1956) it appears that in most instances the fluctuation in successful settlement will lie within four and certainly five orders of magnitude. These data allow one to predict that the variation in the frequency of long-distance dispersal will probably also fall within similar limits (Fig. 13). In considering the frequency of long-distance dispersal it has been necessary to assume that the duration of larval development is sufficient to allow transoceanic DISPERSAL AND GENETIC EXCHANGE 313 dispersal, because it is not possible to arrive at a coefficient which uniquely repre- sents the length of larval life and yet is biologically unrelated to some character defined by the survival coefficient, p s . The theoretical model shows that, over long periods of time (decades or more), the frequency of dispersal is largely dependent upon the parent population size, but that over shorter periods it may be more closely related to survival of the larvae. Estimates oj dispersal frequency from occurrence of larvae in the plankton A further consideration of evidence from the plankton tows and its relationship to the preceding theoretical model can now be made. It is first necessary to determine the limitations of the collection method. If a plankton net of three- quarter-meter diameter is towed at approximately 5.5 km/hr (3 knots) for twenty minutes and the efficiency of the net is assumed to be ninety per cent, then the TABU; V Number of gastropod veligers present in each of the major east-west currents of the North Atlantic Ocean when homogeneous distribution and minimal detectable concentrations of larvae by the present, method of sampling are assumed (Minimum detectable concentration = 1.35 X 10~* larvae/ m?) Volume of water* (m") Total number larvae at cone, of one/ tow North Atlantic Drift North Equatorial Current South Equatorial Current Equatorial Undercurrent 2.45 X 10 5 3.43 X 10 5 2.26 X 10-> 0.98 X 10 5 3.31 X 10 11 4.63 X 10" 3.05 X 10" 1.32 X 10" * In computing the sea-surface area of each major ocean current the following co-ordinates when connected by straight lines were arbitrarily considered to enclose the currents: North Atlantic Drift 45 N-60W; 45N-20W; 35N-10W; 35N-72W; North Equatorial Cur- rent 25N-65YY; 25N-20\V; 10N-25\V; 10N-55\V; South Equatorial Current 5N- 50\V; 5 N-20YY; 5S-7E; 5S-35W; Equatorial Undercurrent 2N-35\V; 2N-5E; 2S-5E; 2S-35YY. capture of a single organism represents 1.35 X 10~ 3 organisms per cubic meter. This figure represents the lower limits of sensitivity obtained by the method of capture. When larvae occur at even lower concentrations, they will not be cap- tured in every plankton tow taken. When their concentration is greater than 1.35 X 10~ 3 /m 3 they will be expected to occur more than once in each sample. The numbers of larvae actually taken varied with the species and the distance from the point of their origin. In my samples the numbers ranged from one to over 500 per tow ; however, it seldom occurred that more than ten gastropod veliger larvae of any particular species were taken in a mid-ocean sample. Thus the con- centration of larvae in most tows, by comparison with the more common holo- planktonic species, was very low. Even though the concentration of larvae per cubic meter may be small the volume of water in which the veligers are found is very large. If one arbitrarily delimits the surface area of each of the major east- west currents of the North and Equatorial Atlantic and considers only the upper 50 m (or in the case of the 314 RUDOLF S. SCHELTEMA Equatorial Undercurrent a vertical distance between 50 and 100 m depth) one can compute the volume of water in which most of the larvae are to be found (Table V). Assuming the lowest measurable concentration of one larva per tow, and considering the case in which a single veliger occurs in each sample of a trans- oceanic section, one can now roughly calculate the number of larvae that ought to 10 1 10 10' FIGURE 14. Intensity of larval settlement of the sea-star Asteria forbcsi (stippled bars) and the oyster Crassostrca rirginica (solid bars) over a period of twenty-five years. The ordinate gives the percentage of years during which settlement occurred at a density value (number/0.25 nr ) shown on the abscissa. The success of settlement is regarded as being directly related to larval survival and, at least in the species represented here, the median density of settlement is approximately equal to the survival when p s = 2/t t ( i.e., when each individual in the population gives rise to one offspring). There are other species where the latter relationship is probably not true, specifically, when the frequency-density curve is skewed (data taken from Loosanoff, 1964, 1966). be present throughout the length of any such current at one instant in time. Thus if the concentration of gastropod veligers is 1.35 X 10~ 3 /m 3 , the number of larvae in the North Equatorial Current will be | 3.43 X 10 s km 3 ] 1 1.35 X 10 C larvae/knr] := 4.63 X 10 11 or about 460 billion veligers. East-west sections made across the Atlantic in the North Equatorial Current during the months of January and June show that, in DISPERSAL AND GENETIC EXCHANGE 315 half of the species considered, larvae occurred during both the \vinter and summer months. The implication is that the total larvae present in the current may ap- proximately assume a steady state and that as larvae are removed from the western extremity of the current, new ones are added at the eastern end. Assuming' a travel time for the larvae of about 170 days (Table II ), the number of larvae trans- ported across the Atlantic Basin on the North Equatorial Current each year will be somewhat over two times the total number of larvae present at any one instant. Perhaps these computations are somewhat fanciful. Actually the larvae are not known to be uniformly distributed, although uniform distribution is of course the underlying assumption in the foregoing calculations. Indeed, none of the species are found in every sample, whereas, on the other hand, many frequently occurred in numbers greater than one per tow. What the computations do clearly indicate is that large numbers of larvae may be dispersed and yet be completely undetected TABLL. \'f Relative frequency at which ten species of gastropod veliger larvae occur in plankton samples from the nii-taii pcratc and tropical surface -waters of North Atlantic Ocean (based on 65S samples) Species X umber of stations found Percentage of stations found Cymatintn par then open m 177 26.9 f'hilippia krebsii 136 20.7 Pedicularia sicula 114 17.3 Smaragdia viridis 95 14.4 Charonia variegatu 91 13.8 Tonna galea 89 13.5 Phalium granulalum 33 5.0 Thais haenmstoma 32 4.9 Cymatium nicobaricu in 28 4.3 Tonna niaculosa 4 0.6 by the present method of sampling. If the larvae of a species can be found in mid- ocean with consistent regularity, it is likely that it frequently will be dispersed across the Atlantic. The fate of the larvae after they reach their destination is yet another matter. The number of larvae required to have any genetic impact on the recipient popula- tion must depend upon the selective advantage that the characters carried by the larvae from their parent population may have in the new environment. Relative rates of genetic exchange and the apparent effect on geographically sepa- rated populations of marine gastropods The frequency with which a species of larva is found in a series of tows along a section across the Atlantic Basin can give some indication of the rate of trans- oceanic exchange relative to other veliger species taken from the same samples. As a first approximation the data from all stations in the North and Equatorial Atlantic can be pooled (Table VI). If the frequency of occurrence is expressed as the per cent of all stations at which a species was found and the ten species con- 316 RUDOLF S. SCHELTEMA sidered are accordingly ranked, thru Cymatium parthenopeum will be at the top of the list, appearing in 26.9 per cent of the samples, whereas Tonna maculosa will he at the bottom, being found in only 0.6 per cent of the samples. Cyinatinui parthenopeum is a species that as an adult is commonly found in the shelf fauna throughout the warm and tropical North Atlantic Ocean (Clench and Turner, 1957), whereas Tonna niacnlosa is principally found in the western Atlantic, being known only from a few scattered records in the eastern Atlantic ( Fischer- Piette and Nickles, 1946). One can perhaps conclude that Tonna niacn- losa is occasionally re-introduced into the eastern Atlantic and has never become established, or alternately, that it has been consistently overlooked in eastern Atlantic collections. Notwithstanding its long pelagic larval development ( Table I ) , Cymatium nicobaricum lias an adult distribution somewhat similar to that of Tonna uiaculosa. In the eastern Atlantic it has been found only on the Madeira and Canary Islands (Clench and Turner, 1957; Odhner, 1931; Dautzenberg, 1890). Regarding the remaining gastropod species one can ask whether the relative amount of isolation between their eastern and western Atlantic populations is reflected in the degree of morphological difference between them. The assumption is that the degree of difference is inversely related to the amount of genetic exchange between them and that the rate of gene-flow is related to the frequency with which larvae of a particular species are found relative to other species of veligers in the plankton. Hence, the fewer samples in which a veliger species is found, the greater the probability that populations on either side of the Atlantic wall be represented by distinct subspecies. The case of Philippia krcbsii is problematic since the species is not easily col- lected as an adult and was not even known to occur in the eastern Atlantic until quite recently (Robertson, 1964; Robertson ct a/., 1970). The genus Pedicularia in the North Atlantic is often considered to include two species, one on either side of the Atlantic. These two forms are, however, probably best regarded as a single species, perhaps composed of two geographical subspecies. In any case not enough adult specimens are presently available for determining morphological variability of the eastern and western Atlantic forms. Smaragdia riridis has two subspecies, viridis in the eastern Atlantic and vindc- inaris in the western Atlantic, although the frequency of its occurrence appears high on the list. However, 75 per cent of all specimens found in east-west trans- Atlantic currents occurred at the western end of the North Atlantic Drift (Fig. 11 ). Moreover, the length of its larval development probably limits the frequency of its dispersal (Table I). Both Charonia varicyata and Tonna yalca are highly variable species. Neither are regarded as having geographic subspecies (Clench and Turner, 1957; Turner, 1946). The percentage of stations at which Phalium granulatum is found is consider- ably less than the preceding two species. Phalium granulatum was recently re- viewed by Abbott (1968), who considers it to have three subspecies, namely, nndulatum in the eastern Atlantic and Mediterranean, gramtlatinn in the western Atlantic including the West Indies, Central America, and Northern Brazil, and centriqtiadratum in the eastern Pacific from Southern California to Peru. DISPERSAL AND GENETIC EXCHANGE 317 Near the bottom of the list and perhaps one of the most interesting examples is Thais haemastoma. This highly variable species is usually considered to have four subspecies: haemastoma and forbesi in the eastern Atlantic and haemastoma, floridana, and liaysac in the western Atlantic (Fig. 9). The dispersal in the easterly direction is in the North Atlantic Drift, where it has been found only in the late summer and early fall. Butler (1954) observed that the species breeds in the Gulf of Mexico during June, July, and August. Hence, it is one of the few examples in which there appears to be some seasonably in the occurrence of larvae. Dispersal in a westerly direction is evidently restricted to the South Equatorial Current. Here the larvae have also been found in winter months. The data on the frequency of larval occurrence may now be briefly summarized. The veliger larvae of Cyuiatiitin parthcnopcinu, Charonia variegata, and Tonna . - -j 1.5 4) C "o. M 1.O Maine Newfoundland Greenland 5 10 Disk Diameter 15 FIGURE 1. Linear regression lines for arm spine length as a function of disk diameter in the three populations of O. aculeata. All measurements are in millimeters. VARIATION IN OPHIOPHOLIS ACULEATA 327 foundland and Greenland. The frequency of specimens with thorny spine tips decreases clinally from 57-15 per cent, from Maine to Greenland. The abonil surface oj the disk The mean length of the longest aboral disk spines (Table I) does not sig- nificantly differ between the three samples. An increase in spine length is correlated with an increase in disk diameter, but the correlation coefficients of the linear regression lines are low, ranging from 0.63-0.76. The morphology of the aboral disk spine tip (Table II ) does not significantly differ between the three samples. The frequency of specimens with thorny aboral spine tips ranges from 4-8 per cent for the three samples. The number of naked radial shields is not significantly different in the three samples (Table II). Specimens with covered radial shields comprise 62.5 per cent of the combined samples, those with naked shields comprise 29.0 per cent and those with partially covered radial shields comprise 8.5 per cent of the com- bined samples. The mean number of naked aboral disk plates (Tables I and III) increases along a cline from 32 in the Maine sample to 47 in the Greenland sample. This number is not correlated with disk diameter. The arms The mean number of aboral arm platelets (Tables I and III) differs signif- icantly between the three samples, but does not vary clinally. The Maine sample has the largest mean of 16.3 and the Newfoundland sample has the smallest mean of 14.7. The Greenland sample is intermediate with a mean of 15.5 platelets. There is no significant difference between the samples for the presence of aboral arm platelet spines (Table II). Platelet spines are present only on small specimens (Table IV) with a disk diameter of 8 mm or less. The spines do not occur on all small specimens since individuals as small as 2 mm were without platelet spines. The frequencies for the number of arm spines (Table II) are essentially the same in all the samples. The number of arm spines increases from 47 with an increase in mean disk diameter from 2.6-13.8 mm (Table IV). The length of the arm spines increases linearly with an increase in disk- diameter (Fig. 1). The arm spines increase clinally in relative length from the Maine to the Greenland sample. The slope for the Greenland sample is signif- icantly greater than the slopes of the more southern samples. Even though the slopes for the Maine and Newfoundland samples are not significantly different, the y-intercept is significantly greater in the latter, resulting in proportionately shorter arm spines in the Maine sample. The regression equations are: Maine y == 0.90.r + 0.44, r -- 0.86, S.E. = 0.13 mm Newfoundland y == 0.96.r + 0.50. r = 0.86, S.E. = 0.13 mm Greenland y =: 1.20.r + 0.41, r := 0.84, S.E. = 0.19 mm y = arm spine length (mm) .r = disk diameter (nun) 328 D. KEITH SERAFY The width of the arm spines increases linearly with an increase in disk diameter. The slopes and v-intercepts of the regression line do not differ significantly between the three samples. The regression equation for the combined sample is: y == O.Sl.r -- 0.01, r -- 0.93, S.E. = 0.07 mm y -- arm spine width (mm) x disk diameter (mm) The width of the arm spines increases linearly with an increase in arm spine length (Fig. 2). As shown above, arm spine length increases clinally with increase in latitude, but the width remains constant. This results in the arm spines being 1.5 I 1 ' 0) c 0.5 Maine Newfoundland Greenland 0.5 1.0 1.5 2.0 Arm Spine Length FIGURE 2. Linear regression lines for arm spine width as a function of arm spine length for the three populations of O. aculcata. All measurements are in millimeters. more acute in the more northern latitudes. Slopes for the linear regression lines differ significantly between the three samples. The regression equations are : Maine V = 0.72.r -- 0.21, r = 0.89, S.E. = 0.11 mm Newfoundland ~v = 0.63.r -- 0.16, r = 0.80, S.E. = 0.21 mm Greenland y == 0.49.r -- 0.09, r -- = 0.83, S.E. = 0.19 mm y arm spine width (mm) x = arm spine length (mm) DISCUSSION Characters that change clinally include ( 1 ) number of naked aboral disk plates. (2) length of the arm spines, and (3) length-to-width ratio of the arm spines. Characters that are essentially the same in Maine and Newfoundland, but are distinctly different in Greenland include ( 1 ) length-to-width ratio of the oral shields, and (2) number of oral papillae. VARIATION IN OPH1OPHOLIS ACULEATA 329 Characters that are essentially the same in Newfoundland and Greenland, but are distinctly different in Maine include ( 1 ) length of the oral disk spines, and (2) morphology of the oral disk spine tip. A character that does not change clinally but is distinctly different in all populations is the number of aboral arm platelets. Characters that are essentially the same in all populations include ( 1 ) morphol- ogy of the aboral disk spine tip, (2) number of naked radial shields, (3) presence of aboral arm platelet spines, (4) length of the aboral disk spines, (5) width of the arm spines, and (6) number of arm spines. None of the above characters are sufficiently different between samples that their coefficient of difference approaches the 1.28 value suggested by Mayr (1969) to be a basis for subspecific distinction. The magnitude of the differences is large enough to indicate the samples came from three phenotypically distinct populations. The uniformity of some characters in all the populations may indicate the expres- sion of these characters is not significantly altered by environmental factors, such as temperature or salinity. Those characters that vary clinally between popula- tions are probably under the influence of some environmental parameter. Tw'o meristic ranges given for 0. aculeata by previous workers, are too small. Mortensen (1927) and Dyakonov (1954) state there are 3 (rarely 2) oral papillae in 0. aculeata. The present study shows there are 2-6 oral papillae with most specimens having 31. The number of arm spines ranges from 6-7, according to Mortensen (1927], and 5-7 according to Dyakonov (1954). The present study shows a range of 4-8. None of the specimens used in this study had more than 7 arm spines ; however, some very large specimens, 22 mm in disk diameter or greater, had 8 arm spines on the sixth arm segment. Because the number of oral papillae and number of arm spines is size dependent, expansions in these ranges are expected when examining extremely small or large specimens. Grieg (1893), cited in Mortensen (1933), states that Greenland specimens of 0. aculeata are more spiny on the disk than ones from the Norwegian coast. According to Mortensen (1933), a large percentage of the Greenland specimens collected during the Danish Ingolf-Expedition had long spines on the disk. These spines were most prominent along the borders of the radial shields. He noted that they were particularly evident on specimens from the more northern localities with negative bottom temperatures. Clark (1911) notes that Alaskan specimens of 0. aculeata have a more spinous disk than ones from Maine. This study indicates the oral disk spines of the Maine sample are significantly longer than those of the Newfoundland or Greenland samples and the aboral spine length is essentially the same in all three samples. A few specimens in the United States National Mu- seum, Washington, D. C. and the Museum of Comparative Zoology, Harvard Uni- versity, collected in the high Arctic have very long aboral disk spines, particularly around the radial shields. If the aboral disk spines are significantly longer in the specimens from the high Arctic, then the increase in length must occur somewhere north of the northernmost sampling area used in this study (approximate Lati- tude, 67N). Dyakonov (1954) described 0. pilosa from the Sea of Okhotsk. This species is supposed to differ from 0. aculeata in the following ways: 1) Short spinelets occur along the edge of the disk and terminate in a sharp thorn or crown of thorns ; 330 D. KEITH SERAFY 2) The aboral arm platelets are numerous, small, rounded, and usually possess a spine; 3) The radial shields are naked. All these characters fall well within the range of variation displayed by O. aculeata. This makes the validity of O. pilosa questionable. A reevaluation of O. pilosa will probably indicate it is a junior svnonvm of O. aculeata. * I thank Prof. H. B. Fell of the Museum of Comparative Zoology, Harvard Uni- versity and Miss M. E. Downey, United States National Museum for loans of specimens from their respective museums. Drs. J. H. Dearborn, H. H. DeWitt and Mr. F. J. Fell critically read the manuscript. I am grateful to Dr. D. Dean of the Ira C. Darling Center, University of Maine and to Drs. F. A. Aldrich and I. Emerson of the Memorial University, Newfoundland, for assistance in obtain- ing material. SUMMARY 1. Fourteen taxonomic characters were analyzed separately for Maine, New- foundland and Greenland populations of O. aculeata. 2. The number of naked aboral disk plates, length of the arm spines and length- to-width ratio of the arm spines are characters that vary clinally. 3. The length-to-width ratio of the oral shields and the number of oral papillae are essentially the same in the Maine and Newfoundland populations but are sig- nificantly different in the Greenland population. 4. The length of the oral disk spines and morphology of the oral disk spine tip are essentially the same in the Newfoundland and Greenland populations but are significantly different in the Maine population. 5. The number of aboral arm platelets does not change clinally but is distinctly different in all populations. 6. The morphology of the oral disk spine tip, number of naked radial shields, presence of aboral arm platelet spines, length of the aboral disk spines, width of the arm spines and number of arm spines are characters essentially the same in all populations. 7. There are 2-6 oral papillae and 4-8 proximal arm spines present on 0. aculeata. 8. The diagnostic characters of 0. pilosa Dyakonov, 1954, fall within the range of variation displayed by 0. aculeata. LITERATURE CITED CLARK, H. L., 1911. North Pacific ophiurans in the collection of the United States National Museum. U. S. Nat. Mas. Bull,, 75: 1-302. DYAKONOV, A. M., 1954. Ophiitroids of the USSR Seas. Izdatel'stvo Akademii Nauk SSSR Moskva-Leningrad, 131 p. [Translated from Russian, 1967, by the Israel Program for Scientific Translations, Jerusalem, TT 67-51490.] GRIEG, J. A., 1893. Ophiuroidea. Norske Nordhavs Expedition 1876-1878, Zool. (Christiania), 23: 1-41. MAYR, E., 1969. Principals of Systematic Zoology. McGraw-Hill Book Company, New York, 428 pp. MORTENSEN, T., 1927. Handbook of the Echinodenns of the British Isles. Oxford University Press, New York, 421 pp. MORTENSEN, T., 1933. Ophiuroidea. Danish Ing olf -Expedition, 4(8) : 1-121. Reference : Biol. Bull., 140 : 331-338. (April, 1971) ASPECTS OF MOLTING AND CALCIFICATION IN THE OSTRACOD HETEROCYPRIS JAMES B. TURPEN 1 AND ROBERT W. ANGELL Department of Biological Sciences, University of Denver, Denver, Colorado 80209 The ostracod is a marine or freshwater crustacean with a bivalved carapace that encloses the rest of the body. The valves composing the carapace are heavily calci- fied and, being an integral part of the exoskeleton, they are shed during each molt. The valves of the new exoskeleton are then calcified while the rest of the exo- skeleton remains unmineralized. The calcification process associated with molting in ostracods is not well known. X-ray diffraction studies by Kesling (1951) on Cypridopsis vidiia determined that the calcium is deposited within the ostracod valve as calcite. Fassbinder (1912), experimenting with the same species, concluded that the immediate source of the calcium used in calcification was from the body of the ostracod rather than from the aqueous environment. This was determined by growing the ostracods in cal- cium-free water and feeding them calcium-free food, and finding that they could still calcify the new valves after a molt. Van Morkhoven (1962) reported that the animal's diet will affect the calcification of the valves, with rich calcium diets pro- ducing secondary thickening and numerous protuberances. Finally, Kesling (1951) proposed that C. vidna has no special cells which secrete the shell material ; it is formed rather by all the epidermal cells within the duplicature. He believed the ostracod calcification process was different from that of other crustaceans that have been studied. The objective of this study was to gain some information about this calcifica- tion process. Experiments using Ca 45 as a tracer were conducted to determine whether any calcium is reabsorbed from the valves prior to molting and whether calcium is stored for use in later valve calcification. Autoradiographs were pre- pared to determine which cells were active in calcifying the new valves. MATERIALS AND METHODS Cultures of Heterocypris were started from individuals collected at ponds in Cherry Hills Village, Colorado. These cultures were maintained in "Chlorofree" (Lambert-Kay Inc., Los Angeles) treated tap water at a pH of 7.4. Parboiled rice grains and hay were added as food. Isolated individuals were maintained in small culture dishes containing 22 ml treated tap water at pH 7.4, a one-quarter- inch piece of hay, one-quarter of a rice grain, one drop of detritus from the stock- cultures, and some ground calcite to provide a rough substrate. Cultures were maintained at approximately 23 C. In all the tracer studies the dosage of Ca 45 (Nuclear Chicago) was one micro- curie of the isotope per twenty-two ml of treated tap water. Determination of radioactivity was made by counting the organisms in metal planchets using a 1 Present address : Department of Biology, Tulane University, New Orleans, Louisiana 79112. 331 332 JAMES B. TURPEN AND ROBERT W. ANGELL Nuclear Chicago model 186 decade sealer and a Geiger-Muller tube with a window thickness of 1.4 mg/cm 2 . Cytological and autoradiographic material was prepared by fixing in 5% for- malin adjusted to pH 8.0 with NaOH, dehydrating with graded ethyl alcohols, and embedding in Epon 812. The embedded organisms were decalcified by nick- ing the anterior end of the Epon block to expose the valves, placing the blocks in 1 N HC1 for one hour, and re-embedding in Epon 812 in a vacuum oven. This technique protected the soft tissues from acid degradation and prevented the loss of cellular calcium. The material was sectioned at two microns on a Sorvall Mt-1 ultramicrotome. Sections were collected on distilled water at pH 8.0 and mounted on glass slides without any subbing. Sections were stained with either standard hematoxylin 5 minutes at 60 C and Saffranin-0 0.1% for 3 minutes at 60 C, or fast green FCF 0.1 %, pH 2.0 at 60 C for five minutes, or toluidine blue (Trump, Smuckler and Benditt, 1961) modified by staining in Columbia staining jars for thirty minutes at 60 C. Sections of specimens for autoradiography were fixed, embedded, and sectioned as above. Two-micron sections were mounted on slides, dipped in Kodak NTB-2 Nuclear Track Emulsion, and exposed for fourteen days in light-tight boxes. After development they were stained with toluidine blue, dehydrated, and mounted under coverslips. The mineralogical composition of the valves was determined by X-ray diffrac- tion of powdered valves mounted in a Debye-Scherrer powder camera. EXPERIMENTAL PROCEDURES Determination of instar times Three groups of fifteen animals each were isolated and observed hourly from the time of hatching until the adult stage was reached. The occurrence of each molt and the general condition of each animal was recorded. Determination of calcium reabsorption Twenty cultures were established, each containing 22 ml treated tap water and one microcurie Ca 45 . Ten cultures contained ten animals each ; ten cultures con- tained one animal each. The animals were allowed to molt in the radioactive media. After molting and calcification of the new valves, the animals were rinsed three times in treated tap water and transferred to identical cultures without Ca 45 . The animals were then allowed to molt in the non-radioactive media, and the discarded valves and the animals were collected, washed, and dried for radioactivity assay. Controls were prepared by placing ten dead ostracods killed by drying in an identical culture of Ca 45 for the duration of the experiment. Additional controls were prepared by collecting the discarded non-radioactive valves from the first molt and exposing them to cultures containing Ca 45 for the time period of the experiment. Determination of calcium storage prior to molt Thirty animals in the seventh instar were isolated in individual cultures con- taining one microcurie of Ca 45 . The animals were allowed to undergo one com- OSTRACOD MOLTING AND CALCIFICATION 333 plete molting cycle in this medium. At this stage of development the next molt normally occurs at 72 hours ; therefore ten animals were removed 24 hours after molting in Ca 4r ', ten after 48 hours, and ten after 72 hours. Each group of or- ganisms was dried on metal planchets after rinsing them three times in treated tap water. Controls were established by exposing dead animals to the same concentration of Ca 45 for 24, 48, and 72 hour intervals. OBSERVATIONS AND RESULTS As observed in our cultures, an ostracod beginning to molt can be distinguished by the development of clear areas along the ventral margins of the valves where they are starting to separate. During this time there is also a slight ballooning of TABLE I Itistar time periods under laboratory culture conditions. Molting occurs every one to three days depending on the stage of development attained by the individual Time in hours between molts Instar No. of animals observed Shortest Longest Mean 1st to 2nd 45 24 24 24 2nd to 3rd 45 24 48 26 3rd to 4th 45 24 48 29 4th to 5th 39 24 48 43 5th to 6th 36 24 72 53 6th to 7th 34 48 72 62 7th to 8th 34 48 96 74 8th to 9th 31 72 130 77 the animal that causes it to lose its oval shape and to appear more rounded. As this swelling increases, swimming movements become erratic and the edges of the valves are rubbed against the bottom. This activity is interrupted by frequent periods of motionlessness. The actual discarding of the exoskeleton begins with the scraping of the inner surface of the valves by the thoracic legs followed by the withdrawal of the two pairs of antennae from the old exoskeleton and then the withdrawal of the remain- ing appendages. Once all the appendages are free, there is a further swelling of the ostracod that causes the old exoskeleton to separate completely from the body. The newly emerged individual appears soft and swollen. The transparent, uncal- cified valves are gradually calcified over a period of several hours becoming opaque as calcification is completed. Determination of instar times Under laboratory culture conditions molting occurs every one to three days depending on the stage of development of the individual. As may be seen in Table I the younger instars molt more rapidly than the more mature animals and there is considerable individual variation in instar duration. 334 JAMES B. TURPEN AND ROBERT W. ANGELL Determination of calcium reabsorption Ostracods placed in water containing Ca 4r> do not incorporate Ca 4r> into the valves until the next molt and the calcification of the new valves. If the ostracods are then transferred to non-radioactive water, during the following molt the radio- active valves are shed, and none of the Ca 45 in them is reahsorbed for use in calci- fying the new valves. All the animals (110 individuals) that molted in Ca 4r '- containing water incorporated Ca 45 into the valves as was shown hy radioactive assay of the valves after they had been discarded during the following molt. The paired valves registered approximately 400 to 500 cpm. The ostracods and their new valves (108 individuals), when assayed by drying the whole animal on a FIGURE 1. A simplified diagram of a section through the ventral margin of the duplicature : A, Outer chitinous layer ; B, Calcified valve ; C, Inner chitinous layer ; D, Cell of outer lamella ; E, Interlamellar space; F, Sub-dermal cell; G, Cell of inner lamella. planchet and then counting, contained no Ca 45 . Thus, there was no reabsorption of Ca 45 from the old valves prior to molting that was retained and used for calci- fying the new valves. There was no uptake of Ca 45 in either the dead control animals or in the valves discarded during the first molt in Ca 4 -"' and left in the containers for the duration of the experiment. Determination of calcium storage The ostracods used in this experiment were allowed to molt in a medium con- taining Ca 45 . Therefore, the valves calcified during this molt contained Ca 45 and had counts (400-500 cpm) equivalent to the radioactive valves recovered in the calcium reabsorption experiment above. Any significant storage of calcium in the tissues of the ostracod should be reflected in a higher cpm as the time of the next OSTRACOD MOLTING AND CALCIFICATION 335 molt approached. The ostracods were kept in the Ca 45 medium for 24, 48, and 72 hours after molting to determine whether there would be an additional Ca 45 buildup that could represent calcium stored for the following molt and subsequent valve calcification. In the ten ostracods exposed to Ca 45 for 24 hours after molting there was a range from 403 to 536 cpm with a mean of 482 cpm. In the 10 exposed for 48 hours after molting there was a range from 437 to 534 cpm with a mean of 486 cpm and in the 10 exposed for 72 hours after molting there was a range from 44S to 515 cpm with a mean of 485 cpm. The controls only registered background. \Yith the exception of 4 counts all the numbers lie within the 93% probability level for the deviation of the mean, indicating that there is no significant variation in TABLE 1 1 Results of autoradiograph v showing time of calcification during molting process. During specimen preparation the valves were decalcified. Thus, the autoradiographs show only the Ca K present in the cells active in calcification (the cells of the outer lamella) and not the increasing accumulation of Ca 45 in the valves. At the end of seven hours the valves are calcified and the amount of Ca 45 detectable in the cells of the outer lamella drops to zero. Number of sections Number of section- showing incorporation Time after molting observed (animals) of Ca 45 5 minutes 120 (3) 30 minutes 160 (4) 160 1 hour 80 (2) 80 2 hours 120 (3) 80 3 hours 200 (5) 110 6 hours 160 (4) 90 7 hours 160 (4) 90 8 hours 80 (2) 1 1 hours 80 (2) 14 hours 80 (2) 18 hours 80 (2) the observed counts. Therefore, these results indicate that there is no buildup or storage of environmental calcium within the body of the animal prior to the molt. Histology The general histological features of the duplicature of Heterocypris as deter- mined by examination of sections stained with either haematoxvlin or fast green are presented in Figure 1 . The cells of the outer lamella are large, irregular in shape, and have processes extending outward to the valve and inward into the space between the outer and inner lamellae. These cells appear to be amoeboid and, during calcification, their positions within the lamella appear to change so they are found concentrated in the area being calcified. In the early stages of calcifica- tion they are concentrated at the margins of the valve, and, as these portions are calcified, they move in toward the mid-region which is the last part of the valve to be calcified. The cells of the inner lamella are smaller in size and more regular in shape. Instead of being a loose aggregation of cells, they are organized into a more closely 336 JAMES B. TURPEN AND ROBERT W. ANGELL packed epithelium. Although the cells have small processes extending into the interlamellar space, they do not have an amoeboid appearance. The cells have fewer nucleoli than those of the outer lamella, and they stain less intensely both with fast green used as a general protein stain and with toluidine blue. Autoradiography of sectioned material Calcification of the valves starts within thirty minutes after the molt, and the cellular activity associated with calcification is completed within seven hours (Table II). With the cessation of this activity there is no further change in the appear- ance of the valves, and calcification has evidently been completed. The autoradio- graphs show a localization of the isotope only in the cells of the outer lamella. They also show that calcification starts at the margins of the valves and gradually proceeds toward the mid-region where calcification is finally completed. Mineral composition of values The valves are mineralized exclusively with calcite. DISCUSSION Several conclusions may be drawn from the results presented here. The first is that calcium dissolved in the water is the source of calcium for calcifying the valves of Hcterocypris, although the incorporation pathways are unknown. Second, the animal does not reabsorb calcium from the old valves prior to the molt, and it does not build up a major store of calcium from other sources preparatory to molt- ing. Third, calcification of the new valves is completed within seven hours after the molt and follows a pattern from the dorsal and ventral edges medially and from the anterior to the posterior of the animal. Finally, the cells of the outer lamellae are responsible for the deposition of calcite in the valves. The Ca* 5 added to the culture water was incorporated into the new ostracod valves during calcification following the next molt. Whether the Ca 45 was first incorporated into the food by action of the bacterial and fungal microfauna asso- ciated with the rice grains and was then ingested during feeding or whether it was directly absorbed by the cells of the body is not positively known. However, we favor the latter because: (1.) with no discernible method of calcium storage the animal would have to eat and digest a tremendous amount of food at the proper time to provide the large amounts of calcium necessary for the valves, (2.) in the many autoradiographs of Heterocypris used in this study there w T as no localization of Ca 4 -" in the gut region or in the food it contained and (3.) the only cellular concentration of Ca 45 found during the entire calcification process was in the cells of the outer lamella. Van Morkhoven (1962) has reported that specimens of Cypris pttbera, when fed on a diet of crushed snails rich in calcium, developed secondary thickening of the valves. It would appear from this that in C. pub era there is an assimilation of calcium from the food source. There could also have been an enrichment of the water in the culture by the dissolution of some of the crushed snail shell. Our attempts to duplicate this experiment with Heterocypris failed because the ostracods OSTRACOD MOLTING AND CALCIFICATION 337 would not eat crushed snails taken from their native pond nor would they eat any other kind of meat offered to them. Perhaps a difference in dietary requirements reflects a difference in calcification mechanisms. Our finding that calcium is neither stored nor reahsorbed by Heterocypris is also different from Fassbinder's (1912) report that in C. vidita the immediate source of calcium for valve calcification is calcium stored in the body at least five days prior to molting. This was shown by culturing the ostracods for that period of time in calcium-free water with calcium-free food and finding they could still calcify their new valves after molting. Although \ve believe that modern tracer methods are more reliable than the methods available to Fassbinder and that this difference could account for our divergent results, there may be alternative explana- tions. The possibility exists that under normal conditions of adequate available environmental calcium an ostracod does not store calcium in its tissues or reabsorb calcium from the old valves prior to molting. However, under conditions of cal- cium stress when environmental calcium is not available, calcium may be with- drawn from the old valves and stored for calcification of the new ones. If the calcium of the valves is not reused, and it was not in our experiments, the problem becomes one of calcium storage. In some highly ornamented benthic species the valves must weigh several times as much as the weight of the organic material of the body. Where and in what form could this amount of calcium be stored? The presence of gastroliths, such as are found in the decapod crustaceans, is doubt- ful, and considering the small size of the digestive tract, it is doubtful that if they were present, they could store the amount of calcium needed for valve calcifica- tion. For these reasons it is our opinion that Heterocypris and probably most ostracods have not evolved the calcium storage mechanisms of the decapod Crustacea, but instead have evolved cellular mechanisms for the rapid concentration and deposition of this element. We believe that the cells of the outer lamella are responsible for calcifying the valves because they are in contact with the calcified valves by means of their numerous processes, because they are concentrated in the region of active calcifica- tion, and because they are the only cells to show Ca 4n incorporation in autoradio- graphs. The lack of Ca 45 in any other cells of the body during the entire calcifica- tion process leads to the tentative conclusion that the calcium is not transported through the body of the animal but is directly concentrated from the water by the cells of the outer lamella and transported intracellularly through the cell processes to the site of deposition in the valve. In his study of C. vidua Kesling found no morphological differences between the cells of the inner and outer lamella although he postulated a functional differ- ence for the two. If he had prepared a sequence of sections of animals at various periods of the molting cycle, perhaps morphological differences would have become apparent. In Heterocypris sp. the differences are marked during this time, and need further study. There appears to be an increase in the number of cells of both lamellae as the time of molting approaches, and during the calcification phase a movement of cells of the outer lamella occurs as explained above. Some of the changes are related to the production of mucopolysaccharides for the new exoskeleton and probably for the elaboration of the enzymes utilized in the molt. We have stained sections taken from animals at various time intervals during the 338 JAMES B. TURPEN AND ROBERT W. ANGELL molting cycle with the PAS reaction and have noted a marked increase in the amount of poly sacchar ides in the duplicature just before molting, hut this needs more study before the complexities of the process are understood. SUMMARY Observations on living ostracods (Heterocypris) show that instars one to four require one to two days between molts, and instars five to eight require two to three days between molts. Tracer experiments with Ca 45 show that during the molting cycle the animal neither reabsorbs Ca from the old valves prior to molt nor stores Ca preparatory to molt. Autoradiographs indicate that calcification of the valves occurs within seven hours after the molt and proceeds from the dorsal and ventral edges towards the valve center. The autoradiographs and other stained preparations indicate that the amoeboid-like cells of the outer lamella elaborate the calcite portion of the exoskeleton. LITERATURE CITED FASSBINDER, K., 1912. Bietrage zur Kenntnis der Susswasserostracoden. Zool. Jahrb. Abt. Anat. Out. Ticrc, 32: 533-576. KESLING, R. V., 1951. The morphology of ostracod molt stages. Illinois Biol. Monogr., 21(1-3) : 1-126. TRUMP, B. F., E. A. SMUCKLER, AND E. P. BENDITT, 1961. A method for staining epoxy sections for light microscopy. /. Ultrastruct. Res., 5 : 343-348. VAN MORKHOVEN, F. P. C. M., 1962. Post Paleozoic Ostracoda: Their Morphology, Taxonomy, and Economic Use, Volume 1. Elsevier Publishing Co., Amsterdam, London, New York, 204 pp. Reference : Biol. Bull, 140 : 339-351. (April, 1971 ) DELAY AND QUADRIPARTITION IN SEA URCHIN EGGS INDUCED BY SHORT EXPOSURE TO 2-MERCAPTOETHANOL SHUHEI YUYAMA ' Department of Radiology and Department of Biology, Case Western Reserve University, Cleveland. Ohio and Marine Biological Laboratory, Woods Hole, Massachusetts The effects of 2-mercaptoethanol (ME), an inhibitor of cell division in a va- riety of cells (Mazia, 1958; Limbosch-Rolin and Brachet, 1961; Fautrez and Fautrez-Firlefyn, 1963; Fotakis, 1963; Limbosch-Rolin. 1963; Meeker, 1964; Mazia and Zeuthen, 1966), have been studied extensively by Mazia and his co- workers using echinoderm eggs. They reported that ME blocks cleavage reversibly if treatment is initiated before the "point of no return" which corresponds approxi- mately to metaphase, and that the delay in cleavage is equal to the time spent in ME (Mazia, 1958; Mazia and Zimmerman, 1958). In addition, they found that DNA synthesis continues (Bucher and Mazia, 1960) and that the level of respiration is high enough to support cleavage (Mazia and Zimmerman, 1958) during the ME treatment. The mitotic apparatus (MA) can be isolated during the ME treatment, although the fibrous structures of MA seem disrupted with phase contrast microscopy (Mazia and Zimmerman, 1958). Pressure studies of Zimmerman ( 1964 ) revealed that ME also weakens cortical gel strength. It was concluded that ME arrests cleavage at the time of exposure to ME, because the delay is equal to the duration of the treatment, and that this arrest is somehow linked to the abnormality in structure of the isolated MA. In addition to the synthesis of DNA and ATP, the first cleavage of echinoderm eggs appears to require protein synthesis (Hultin, 1961 ; Zotin, Milman and Faustov, 1965; Young, Hendler and Karnofsky, 1969). In order to further elucidate the mechanisms of ME action on cleavage, it is necessary to obtain more detailed information on the response of cells to ME during the cleavage cycles and on the relationship between protein synthesis required for division and the effects of ME. These problems have been investigated in the present study. The phenomenon of quadripartition (QP), a direct division of one cell into four, occurs during recovery from the prolonged blockage of cleavage by ME. QP involves the generation of a tetrapolar mitotic apparatus from a bipolar one (Mazia and Zimmerman, 1958). Mazia, Harris and Bibring (1960) established that each mitotic center is duplex in nature, i.e., there are four potential centers at metaphase, and that duplication of centers occurs at about telophase. They further suggested that ME is capable of blocking duplication of the centers while permitting the "moving-apart" of existing centers according to the regular mitotic schedule. These conclusions were based on several observations. First, the quadripartitioned cells skipped the next cleavage although other mitotic events including the formation of a monopolar mitotic spindle occurred. Secondly, QP could not be induced by the ME treatment unless a certain time had elapsed after fertilization and unless exposure to ME was of such duration that it extended past the division of controls and the ME treated cells divided after the second divi- 1 Present address : Department of Zoology, University of Toronto, Toronto, Ontario, Canada. 339 340 SHUHEI YUYAMA sion of controls. The inhibition of duplication of mitotic centers by ME was veri- fied by Went (1966) using a more unambiguous agent (benzimidazole) for assay- ing the number of mitotic centers. In the present study, new observations have been made that may modify the present hypothesis regarding OP. The capacity of androgenic merogones to undergo QP was also examined. MATERIALS AND METHODS Arbacia pitnctnlata eggs, obtained by electrical stimulation, were shed directly into a beaker filled with filtered sea water (Harvey, 1952). Lytcchinus variegatits eggs were shed by placing a few drops of 0.55 M KG into the test (Palmer, 1937). All eggs were washed twice. The following procedures were used for the short exposure experiments. Eggs were inseminated and washed about a minute later to remove excess sperm. At 3-5 minute intervals after insemination, 1 ml of egg suspension, containing 400- 500 eggs, was mixed with 1 ml of 0.2 M ME dissolved in sea water. After the desired period of exposure (5-80 minutes depending upon the experiments), eggs were washed by centrifugation or by dilution (a dilution factor of 1:10,000). Both centrifugation and dilution methods gave essentially the same results. The controls were treated similarly. The time when approximately 50% of the eggs had cleavage furrows was estimated visually. In order to determine the per- centage of quadripartition (QP), 200 cells were examined before the onset of the second cleavage. Data obtained from asynchronously dividing cells were discarded. After several preliminary experiments, O.I M ME was chosen because this was the lowest concentration that gave consistent results. For example, 0.7 M did not block cleavage consistently in both Lytechinus and Arbacia eggs. In some experiments ME-treated eggs were further treated with puromycin, a specific inhibitor of protein synthesis (Nathans, 1964). Each aliquot of eggs was quickly washed twice by dilution (for 1 minute) and 0.2 ml of egg suspension containing about 200 eggs was mixed with 0.2 ml of 1 X 10~ 3 M puromycin on a depression slide. The slide was placed in a petri dish with filter papers moistened with sea water to prevent evaporation. In experiments where eggs were first pulsed with puromycin and then treated with ME, 1 ml of 1 X 10~ 3 M puromycin was mixed with 1 ml of fertilized egg suspension, washed three times after a 15 minute period of incubation and then subsequently, at different times, exposed to ME for 20 minutes. Androgenic merogones were obtained by inseminating the enculeate fragments of eggs obtained by cutting (for detailed method see Rustad, Yuyama and Rustad, 1970). Briefly, unfertilized eggs were cut with a thin glass needle into approxi- mately equal halves in a Syracuse dish coated with 0.2 % agar. Enucleate and nucleate halves were sorted immediately after each cut, since the half containing a female pronucleus can be determined during the time of cutting. Approximately 30 minutes were required to cut 60 eggs. Extreme caution was taken to prevent polyspermy in fertilizing the cut eggs. Determination of the incorporation of 3 H-leucine was performed in the following manner. Sea urchin eggs were washed twice with sea water filtered through Millipore filters, inseminated, washed and transferred into 10 ml of sea water MERCAPTOETHANOL INDUCED DELAY 341 containing either ME or puromycin. After 5 minutes, 1 /zc/ml 3 H-leucine (5.0 c/mmole, New England Nuclear) was added. The reaction was terminated by adding an equal volume of \0 r /c trichloroacetic acid (TCA) and the samples 70 50 (D C 30 10 10 30 50 TAI in minutes FIGURE 1. Delay-ME exposure diagram of Lytechinus eggs. Eggs were fertilized, and treated by ME for 15 minutes, initiated at different times after insemination. Lengths of each vertical line indicate the first cleavage period of Lytechinus eggs. Duration of the ME treatment is indicated by the solid line. Numbers associated with each vertical line indicate the recovery period. The mitotic stages of the control eggs are indicated on the right side of the graph. Abbreviations used are: TAI = time after insemination; m = multiple division in some eggs ; d = time of control cleavage ; mp = metaphase ; nm = disappearance of the nuclear membrane ; dc = centriole separation ; fp = fusion of pronuclei ; excess delay = cleav- age period of ME treated eggs cleavage period of control egg duration of the ME treatment. were heated for 15 minutes at 90 C. An excessive amount of non-radioactive leucine was added and the samples were kept overnight in the cold room. The samples were filtered through glass fiber filters, washed with 70% ethanol, dried, placed in vials containing toluene and proper fluors and the radioactivity was counted using a Nuclear Chicago Scintillation Counter. 342 SHUHEI YUYAMA RESULTS Cleavage delay In the first series of experiments, Lytechinus eggs were inseminated and ex- posed to 0.1 M ME for periods of 15 minutes, starting at various times after fertilization. Typical results are shown in Figure 1. During the first cleavage m in. LU o 10 CO CO LU U X LU -10 1 div. 2 div. T A I , mm. mm, o LU A X ^ i 20 RECOVERS D i i i i i 1 1 20 40 60 TIME in ME , m i n. FIGURE 2. Delay-ME exposure diagram of Arbacia eggs, (a) The experimental design is similar to that of Figure 1, although eggs are treated for 20 minutes in this experiment. Different lines indicate separate experiments. ( i) ) Dependence of the recovery period upon the duration of the ME treatment, the treatment was initiated either at the beginning of the streak stage (line 1) or several minutes before metaphase (line 2). cycle, there was a change in delay, although the exposure period was kept constant. With batches of cells whose first division cycle was 45 minutes, the delay decreased when the treatments were initiated up until 13 minutes after insemination, increased linearly up until 37 minutes and decreased rapidly after this time. During the period when delay increased linearly ( the linear delay phase ) , the interval between the end of exposure and the time of cleavage was constant (Fig. 1). Although the period of the first cleavage cycle varied among different MERCAPTOETHANOL INDUCED DELAY 343 batches of eggs (45-60 minutes) and although the amount of delay varied with different experiments, the essential pattern of delay was the same in six different experiments. The start of the linear delay phase superficially corresponds to the beginning of the streak stage; division (separation) of the mitotic centers appears ro occur about this time (Wilson and Mathews, 1895). The end of the linear delay phase corresponds approximately to metaphase after which sensitivity of cells to ME is rapidly lost. The above experiments established that the delay induced by ME treatment is not constant but fluctuates during the cleavage cycle. In order to examine whether or not this fluctuation of delay would occur in another genus of sea urchin eggs, the same short exposure experiments were per- formed with Arbacia eggs. The delay pattern of Arbacia eggs was essentially the same as that of Lytechinus (Fig. 2a). In addition, the linear delay phase was also observed in the second division cycle. Since the linear delay phase reflects the constant period between the end of exposure and th time of cleavage (see Fig 1), it suggests that ME damages a certain event related to cytokinesis and that recovery from this damage takes a constant period of time. In order to ex- amine whether this ME sensitive event is present from the beginning of the streak stage or develops at later stages, the dependence of the recovery period upon the exposure time to ME was studied. Fertilized eggs were exposed to ME for various periods of time starting at two different stages (beginning of the streak stage and immediately before the end of the linear delay phase; ca. metaphase). Typical results are shown in Figure 2b. When the ME treatments were initiated at the beginning of the streak stage, the recovery period gradually decreased and then became constant after periods of exposure for about 20 minutes or longer. When the exposure was initiated immediately before the end of the linear delay phase, the recovery period did not depend upon the period of exposure between 5 and 80 minutes. These observations suggest that the ME sensitive event develops about 20 minutes after the beginning of the streak stage. This time is several minutes before metaphase. At about metaphase, eggs lose sensitivity to the ME treatment. In all experiments the recovery period (an interval between the end of exposure to ME and the time of cleavage) is 20-25 minutes and is longer than the interval between metaphase and cleavage (8-12 minutes). In the next series of experiments, the effects of temperature on the recovery period were investigated. An Arrenius plot of the data indicated that the energy of activation for the recovery period is 16,400 calories. In the third series of experiments, the relationship between protein synthesis required for division and the blockage of cleavage by ME was investigated in Arbacia eggs. Eggs were exposed to puromycin (5 X 10~ 4 .17 ) immediately after fertilization for a period of 15 minutes. This treatment resulted in approximately 30 minute delays in cleavage ; the synchrony of cleavage was not appreciably changed by the treatment. When these puromycin-treated cells were exposed briefly to ME at various times during the cycle, the appearance of the linear delay phase was delayed for approximately 30 minutes (Figure 3). In four sepa- rate experiments, there was a precise correlation between the delay of cleavage due to the puromycin treatment and the delay of the appearance of the linear delay 344 SHUHEI YUYAMA phase. These results indicate that the ME sensitive event arises only after the protein synthesis required for division. Arbacia eggs were continuously treated with puromycin. When the treatments were initiated at different times after fertilization, cleavage was blocked up until about 10 minutes after insemination. After this period, cleavage was delayed up until the beginning of the streak stage (delay decreased rapidly during this period) and then the first cleavage was no longer affected by the puromycin treatment. These findings confirmed the observation of Rustad and Burchill (1966). When the cells were treated with ME for 20 minutes during the puromycin sensitive period and then incubated immediately with puromycin (5 X 10~ 4 M), it was found that sensitivity to puromycin is markedly reduced after the ME treatment. This < 20 j ULJ s? p / / Q ^ // '' '' 10 y/ ' "*>'' y CO / / ' ^;^- ^ / 00 // s^ / LLJ U n '^^^ f O X cd P d ID 01 LU -10 puramycn i 1 20 40 60 80 TAI in minutes FIGURE 3. Delay-ME exposure diagram of the puromycin treated cells. The delay pat- terns of Arbacia eggs treated by ME alone (C) are compared with those pulsed with puromycin for 15 minutes before the ME treatment (P). Lines with the corresponding num- bers represent experiments performed with the same batches of eggs. Abbreviations used are : cd = cleavage time of controls ; cp = cleavage time of puromycin treated controls. suggests that protein synthesis required for division continues during the ME treatment. In order to verify this point, eggs were incubated with 3 H-leucine (1 /tc/ml) for a period of 30 minutes in the presence of ME (0.1 M) or puromycin (5 X 10** M). The hot TCA insoluble radioactivity was determined. The average counts per minute (from two separate experiments) were: Control 8495; ME treated 6/14; puromycin treated 1317. During the linear delay phase, no additional delay was observed with eggs which were treated with puromycin after the ME treatment, suggesting that recovery from the ME treatment does not require protein synthesis. Went (1962) reported that dithiodiglycol (DTDG), an oxidized product of ME, solated the mitotic apparatus and inhibited cleavage. Since the effective con- centration of DTDG (1 X 10-"' M) was much lower than that of ME (1 X 10' 1 M}, there was a possibility that the effects of ME were due to DTDG contamination in the ME solution, as contamination by \% DTDG in ME may produce the effect. The effects of DTDG (1 X 10~ 3 M) on the cleavage of Arbacia eggs MERCAPTOETHANOL INDUCED DELAY 345 \vere studied. \Yith DTDG the sharj) transition of sensitivity which was the characteristic of the ME effect, was not observed during the first cleavage cycle. Eggs exposed to DTDG at later stage of the cycle showed a characteristic abnor- mality of furrow formation (wrinkled all over the egg surface). After short ex- posures to DTDG (20 or 30 minutes), eggs divided so asynchronously that it was impossible to estimate the 50% division time precisely. These observations rule out the possibility that the effects of ME observed in the present study were due to contamination of the ME solution. TOO r so - OL o 60 - * 30 "J L X ^c. '| 20 [ vO /' ^^ CN r - ^ /. m 20 V * 10 Q % / ^^ \ / 1 1 7~ v w\ 1 1 1 '. 1 . J/ 20 40 60 TAI in m nutes FIGURE 4. Early quadripartition in Lytcchinns eggs. Experimental design is similar to that of Figure 1, although eggs \vere treated by ME for 20 minutes. Triangles = cleavage time of bipartitioned cells ; circles = cleavage time of quadripartitioned eggs. Solid bars indicate the percentage of QP; arrows = cleavage time of controls. Quadripartition Frequencies of quadripartition (QP) varied widely in different experiments. Typically, a majority of cells underwent OP only after prolonged exposure to ME. Sometimes, however, high frequencies of QP were observed in Lytechinus eggs after a 15-20 minute exposure to ME. The result of one of these observations is shown in Figure 4. In this experiment, fertilized eggs were exposed to ME for a period of 20 minutes starting at different times after insemination. With a 58 minute cell cycle, frequencies of QP increased markedly (from 4% to 39%), starting at 32 minutes after insemination, reached a peak (77%) at 37 minutes and decreased (42%) at 42 minutes. At 47 minutes, some cells were no longer blocked and no significant percentage of QP was observed. In the case where 39% showed QP, cells were removed from ME 6 minutes before the onset of the first cleavage (not a single cell in the control started to divide at this time), and they divided 18 minutes before the second cleavage. In the case where 77% showed QP, cells were removed 1 minute before the first cleavage and they divided 346 SIU'IIKI Yl'YAMA 15 minutes before the second cleavage. This phenomenon will he referred to as "earlv quadripartition." Tliese observations demonstrate that it is not necessary to treat cells for a prolonged period of time to induce quadripartition. The ma- jority of cells undergoing QP divided about 5 minutes earlier than hipartitioned cells. The phenomenon of "early QP" was not observed in Arbacia eggs, however. In the next series of experiments, the role of the female pronucleus on QP was investigated. Whole eggs, encleate halves and nucleate halves of Arbacia eggs were inseminated and treated with ME for 30 minutes starting 30 minutes after the insemination. QP occurred in androgenic merogones in two separate experi- m ME so r nnctulata and Lytechinus varicgatus were treated with 0.1 M 2-mercaptoethanol (ME) for short periods initiated at different times after fertilization. In both the first and second cleavage cycles, there was a phase during which cleavage delay increased linearly. This linear delay phase reflects the con- stant period of recovery from damage caused by ME treatment. The ME sensi- tive event is estimated to be present only briefly (about 5 minutes) immediately prior to the metaphase and appears to be associated with processes that determine the cleavage furrow. This ME sensitive event must develop only after completion of the puromycin sensitive event (protein synthesis required for division), because there is an exact correlation between the puromycin induced delay and delay in the appearance of the linear delay phase. Lytechimis eggs treated briefly with ME before the first cleavage underwent quadripartition well before the second cleavage of controls. ME, therefore, ap- pears to cause a premature splitting of centrioles prior to their duplication. Quadri- partition can also be induced in androgenic merogones, verifying the old concept concerning the role of sperm centrioles in normal cleavage. LITERATURE CITED BIBRING, T., 1962. An analysis of experimentally induced four-cell division in developing echinoderm eggs. Ph.D. thesis, University of Calif., Berkeley, California. BOVERI, T., 1900. Zellen-Studien IV. t' T her die Natur des Centrosomen. Jena. Z. Natnrw., 35: 1-220. 350 SHUHEI YUYAMA BUCHEK, N. L. R., AND 1). MAX i A, 1%(). Deoxyribonucleic acid synthesis in dividing eggs of tlie sea urchin, Strongylocentrotus purpiiraliis. J . Biophvs. Biocheni. Cvtol., 7 : 651-655. COSTELLO, D. P., 1961. On the orientation of centrioles in dividing cells, and its significance: a new contribution to spindle mechanises. Biol. Bull.. 120: 285-312. DE HARVEN, E., AND W. BERNHARD, 1956. Etude au microscope electronique de 1'ultrastructure de centriole chez les vertebres. Z. Zcllforsch.. 45: 378-398. DIRKSEN, E. R., 1961. The presence of centrioles in artificially activated sea urchin eggs. /. Biophys. Biochein. CytoL, 11: 244-247. EPHRUSSI, B., 1933. Contribution a 1'analyse des premiers stades du developpement de 1'oeuf. Action de la temperature. Arch. Biol., 44: 1-147. FAUTREZ, J., AND N. FAUTREZ-FIRLEFYN, 1963. A propos de la localisation du noyau dans les blastomeres de 1'oeuf d'Arteinia saliiia. L'influence du ^-mercaptoethanol et due dithiodoglycol sur cette localisation. Develop. Biol., 6: 250-261. FOTAKIS, N. S., 1963. Die Wirkung von /3-Merkaptoathanol und a-Piponsaure auf die HeLa- Zellkultur. E.rp. Cell Res.. 30 : 600-602. HARRIS, P. G., 1961. Electron microscope study of mitosis in sea urchin blastomeres. /. Biophys. Biochem. CytoL, 11: 419-431. HARRIS, P. G., 1962. The effect of mercaptoethanol on the mitotic spindle. Volume II NN-1 in S. S. Breeze, Jr., Ed., International Congress Electron Microscopy. Academic Press, New York. HARVEY, E. B., 1952. Electrical method of sexing Arbacia and obtaining small quantities of eggs. Biol. Bull., 103: 285. HENSHAW, P. S., 1940. Further studies on the action of roentgen rays on the gametes of Arbacia punctitluta. II. Modification of the mitotic time schedule in the eggs by exposure of the gametes to roentgen rays. Amer. J. Roentgenol. Rod. Thcr., 43: 913-916. HIRAMOTO, Y., 1956. Cell division without mitotic apparatus in sea urchin eggs. E.rp. Cell Res., 11: 630-636. HULTIN, T., 1961. The effect of puromycin on protein metabolism and cell division in fertil- ized sea urchin eggs. Experientia, 17: 410-411. KAWAMURA, K., 1960. Studies on cytokinesis in neuroblasts of the grasshopper, Clwrtophaga viridijasciata (de Geer). E.rp. Cell Res., 21: 1-18. KUNO, M., 1954. Comparative studies on experimental formation of multinucleated eggs of sea urchins by means of various agents. Embryologia, 2 : 43-49. KUNO-KOJIMA, M., 1960. The effect of DNP and NaN 3 on fertilized eggs of the sea urchin with special reference to the induction of the abnormal cleavage. Einbrvoloqia, 5 : 71-77. LiMBOSCH-RoLiN, S., 1963. Les effets du |3-mercaptoethanol et du dithiodiglycol sur la crois- sance de Eschericliia coli et du Saccharomyces cerevisiac. E.rp. Cell Res., 29: 61-72. LiMBOSCH-RoLiN, S., AND J. BRACKET, 1961. Etude des effets du ^-mercaptoethanol et du dithiodiglycol sur la division cellulaire chez les oeufs d'Amphibiens. E.rp. Cell Res., 24: 120-130. MAZIA, D., 1958. SH compound in mitosis. I. The action of mercaptoethanol on the eggs of the sand dollar Dcndraster c.rcentriciis. E.rp. Cell Res., 14: 486-494. MAZIA, D., 1961. Mitosis and the physiology of cell division. Pages 77-412 in J. Brachet and A. E. Mirsky, Eds., The Cell, I 7 ol. 3. Academic Press, New York. MAZIA, D., AND E. ZEUTHEN", 1966. Blockage and delay of cell division in synchronized population of Tctnihvinoni by mercaptoethanol (monothioethylene glycol). C. R. Trar. Lab. Carl she re/'. 35: 341-361. MAZIA, D., AND A. M. ZIMMERMAN, 1958. SH compound in mitosis. II. The effect of /3-mercaptoethanol on the structure of the mitotic apparatus in sea urchin eggs. E.rp. Cell Res., 15 : 138-153. MAZIA, D., P. G. HARRIS AND T. BIRRING, I960. The multiplicity of the mitotic centers and the time-course of their duplication and separation. /. Biophvs. Biochein. C\tol., 7: 1-20. MEEKER, G. L., 1964. The effect of /3-mercaptoethanol on synchronously dividing populations of Astasia longa. Exp. Cell Res., 35 : 1-8. MERCAPTOETHANOL INDUCED DELAY 351 NATHANS, I)., 1964. Puromycin inhibition of protein synthesis: Incorporation of puromycin into peptide chains. Proc. Nat. Acad. Sci., 51 : 585-592. PALMKK, L., 1937. The shedding reaction in Arbacia punctulata. 1'liysiol. /mil., 10: 352-367. RAPKIJST, L., 1931. Sur les processus chimiques au cours de la division cellulaire. Ann. Physinl. Physlcocliiin. Hint., 7: 382-418. RAPPAPORT, R., 1965. Geometrical relations of the cleavage stimulus in invertebrate eggs. /. rheoret. BioL, 9: 51-66. ROBBIXS, E., AND G. JENTZSCH, 1969. Ultrastructural changes in the mitotic apparatus at the metaphase-to-anaphase transition. /. Cell BioL, 40: 678-691. RUSTAD, R. C., AND B. R. BuRCHiLL, 1966. Radiation-induced mitotic delay in sea urchin eggs treated with puromycin and actinomycin D. Rod. Res., 29 : 203-210. RUSTAD, R. C., S. YUYAMA AND L. C. RUSTAD, 1970. Nuclear cytoplasmic relations in the mitosis of sea urchin eggs. II. The division times of whole eggs, and haploid and diploid eggs. BioL Bull., 138: 184-193. SAKAI, H., 1963. Studies on sulfhydryl groups during cell division of sea urchin egg. VI. Behavior of -SH groups of cortices of eggs treated with ether-sea water. Exp. Cell Res., 32: 391-393. SAKAI, H. AND K. DAN, 1959. Studies on sulfhydryl groups during cell division of sea urchin egg. I. Glutathione. Exp. Cell Res., 16: 24-41. SCOTT, A. C., 1960. Furrowing in flattened sea urchin eggs. Biol. Bull., 119: 260-272. SCOTT, A. C., 1965. A new concept of the mechanism of cytokinesis. BioL Bull., 127 : 389. SWANN, M. M., 1954. The mechanism of cell division: experiments with ether on the sea urchin egg. Exp. Cell Res., 7: 505-517. SWANK, M. M., AND J. M. MITCHISON, 1953. Cleavage of sea urchin eggs in colchicine. /. Exp. BioL, 30: 506-514. WENT, H. A., 1962. Structural modification of the mitotic apparatus during the early cleavages in sand dollar eggs. Chromosome-, 13 : 219-242. WENT, H. A., 1966. An indirect method to assay for mitotic centers in sand dollar (Den- raster c.rcentricus') eggs. /. Cell BioL. 30: 555-562. WILSON, E. B., 1902. Experimental studies in cytology. II. Some phenomena of fertiliza- tion and cell division in etherized eggs. III. The effect on cleavage of artificial obliteration of the first cleavage furrow. Arch. Entivicklunamcch. Oryamsmcn H'il- hclm Roux 13: 353-395. WILSON, E. B., AND A. P. MATHEWS, 1895. Maturation, fertilization and polarity in the echinoderm egg. New light on the "quadrille of the centers." J. MorphoL, 10: 319-342. YOUNG, C. W., F. J. HENDLER AND D. A. KARNOFSKY, 1969. Synthesis of protein for DNA replication and cleavage events in the sand dollar embryo. Exp. Cell Res., 58: 15-26. ZIMMERMAN, A. M., 1964. Effects of mercaptoethanol on the furrowing capacity of Arbacia eggs. BioL Bull., 127: 345-352. ZIMMERMAN, S. B., T. H. MURAKAMI AND A. M. ZIMMERMAN, 1968. The effects of selected chemical agents on furrow induction in the eggs of Arbacia punctulata. Biol. Bull., 134: 356-366. ZOTIN, A. I., L. S. MILMAN AND V. S. FAUSTOv, 1965. ATP level and cleavage of sea urchin eggs Strongvlocentrotus drobachiensis ( O. F. Muller). Exp. Cell Res., 39: 567-576. Vol. 140, No. 3 June, 1971 THE BIOLOGICAL BULLETIN PUBLISHED BY THE MARINE BIOLOGICAL LABORATORY A SUMMARY OF INVERTEBRATE LEUCOCYTE MORPHOLOGY WITH EMPHASIS ON BLOOD ELEMENTS OF THE MANILA CLAM, TAPES SEMIDECUSSATA l DANIEL P. CHENEY - College of Fisheries, 204 Fisheries, University of H'ashiut/tini, Seattle. ]\'ashin-T3 c 3 >n 3 3 o a cu Cd Oj >, >- J-'i E tj w SH ij C 3 C E c -^ t/i Q cu cd Si c ^ ^" ii * fj c3 bd g.2 cu U tr. ** j^ MU *- u cd cd.2 . cu cd i S~ cu 4 ~ *-^ -^J So.2 g ~ UM U, fcu \J %, & COJ= =. u o 3 '*-- y a_cd 'x cic'bjb ct 2 _-" E S|.o| O oc "cd O "cd _ L- l_ CU Secrete s 1 Stem cell encaps >> *cu O cu o u .-g M C v> _g ajt cu Cu tyj c/} CUOi Oxygen t Seasonal elemen a c "o "o 00 o > o Cardium, Cardium, Precurso: a. ^ 3 a 1 a. 1 tN N o cd o cd cs cd *-H i S 00 1 i i -3 I -3 i -3 X X 00 00 00 Morphology g /) O o >i CJ 3> ~7i 2 "o 3 C cu c u cd iriable, cytoplasm with or without granules Archeocyte ranched pseudopodia milar to those in circulatory svstem 6 a"3.c j cu O O M ^_c"5 -2 E " a || cj' i- 1 " CJ [2 C;5 philic granules in the cytoplasm yaline cytoplasm = vert, lymphocyte. Nuc. i to f dia. of cell. Chromatin clumped uc. \ size of cell, eccentric & ovoid ; nucleolus often visible. Chromatin lightly clumped uc. j to 5 size of cell. Chromatin heavily clumped : Crustacean forms : Mammalian forms irefrigent granules, colored or colorless, depending on species irefrigent granules, colored or colorless, depending on species > " ffli/5 & ffi 2 Z; II II 03 CQ o cu i CJ CJ Cu cu CO en Where obs Mesoglea Mesoglea Mesoglea Mesoglea Mesoderm Pseudocoel Mesenchyme E E QJ OJ w w 1-. k-. C u Circ. system Hemolymph Hemolymph Hemolymph Hemolymph (smears) Hemolymph (smears) Hemolymph (smears) Hemolymph Hemolymph Hemolymph (smears & tions) Hemolymph (smears & tions en 5 jj cu in > Published nam< and synonyms Scleroblast (1) Archeocyte (35) Amoebocyte (1) Interstitial cell Neoblast Ameboid cells ( Lymphocyte Hyaline amoebocyte Granular amoebocyte "erythrocyte" Leucocyte (3) Leucoblast Macronucleocyt Granular eosino philic amoe- bocyte Granular baso- philic amoe- bocyte (7) Granulocyte (4^ Lymphocyte (4 Coarse grained amoebocyte ( Fine grained amoebocyte ( cu cu cu 0) CU cu cu ft t^ * >t >. ^ ^ i**! CJ CJ CJ Cj' CJ -1-* o o o o o ^ 1 CJ CJ O CJ cu cu y cu u cu cu cu ^ 3 *-> 3 *J -t-J QJ cu cu cu cu cu > >, CU M cu i _ H >,V >: 'S-, CJ cj ~~ cd cj' CJ -W CJ u *rt cu cu >. cu cu o _C5 cu .3 "3 _o >. o _O i~ e _c o S S "3 ^-S "o "3 c a o ,c ft "3 o^ "3 C ~a "c3 3 P !" cd ^J ' CJ cd "3 CO cod cd 3 cd c cd K B CU >t > J ffiK O WK ,J O 'o cd CQ ^-< o ^ O ^J 2 \c~ cd o ^^, CO ^r- ^ J3 ON cd J cu ft ^_^ . U - 1 " 2 8 3 00 T3 ^^ ^ o cd "? ^ cu * a O !S *o cd ro S cu^ ^* XI CJ ^ fc cd c '5'-- o a ^ cd | .S a cd'*' o CJ cd cd . cd "S j- CJ - VI CN cu OJ "^ OJ OJ *>> '^* C o cu - o o _S "3 "o ~- 5 |^ CM u PL^ ^ Z S 2 j MANILA CLAM LEUCOCYTES 355 S -a '3 '" s o 7 C i O ._' rt CO Ofl 3 2b XI n "5 f ^ 3 'D X 0.2 C k. rt c; art? Numeral o .2 2 x 4J ^ ,-1 CO 5 =|t ^t* ,-, O O "^ ^ - = s u cjfa.2 S E nj u agocytosis and libit amoeboid i filiform and Ii . Busycon - " O ^- ? /: = 5 u 3 _g f3 O -i ? 5.'? ' = - S cjv? g ^-3 uc^ x E 1^0 "3 "3 "o i - p, cj QJ ~* *- 1 x-.S a C. x "g f. ii ^ o g 'x r^s ? 3 7 ?"" = ~ 00 c3 t: a y X O Is? r E "1 | 1 ||| ^Ti 'Si O Modified a but slow Function ii pinocyto Viviparus T -i" O.STJ -j Z- a ^ .5 o j > cs - 3 s.yi E n Z E 3= > S U Lymnaea \ Lvmnaea s Lymnaea s a. a. a a. a. 1 a. OJ N 2 S x . Tf r^ CJ v 1 ^ a u-. cn X X X *~\ C"0 1 S " i 00 " IT; v S 2 - f. CJ .. CJ u O M CJ tn A_ ~3 s a,_j ,- g .0 d >/; C 71 O 3 ^ c 3 -3 -a =5 o2 "cj ~ >_K Morphology _rt i >. o 14H O i| H u a 3 >. > C o O 2 c o so * *-* ~ l X 2ip ||E| C (S -T ^. >. _j 2 E P ^ ^g M '' X 3 o a'H. voi^ Lit r^i lll-sa^l |-|^lf^ 7 s s 2 ? .5" 2 ? o "> Z "ES" ^-i * ** y i-i *C iU C-^'S S I^T" _OO o S ~ . OJ -^ x >, = j^ GO f % S HI?? II ill 2.2 o "^_ ~^ X x > H I 'Z C I?? =. i 1 T " S~ x-'"f *= --=.= s S~'S u ^ t- ^j " rt f Bill O J^ = .2 2 5"x E - S. 2 = i '- '., % P-cudopoilia. Ma artifact and arise 1 eakup of normal ce t 2 CS c cd "2"H.i b fe s c tfl~ c rt g & g 5;^} J ^gg.iZ *^ X |E M=> x-E 3. S Z J ? -S 3 3 - o cd i; u a> 4J C) 3J IH flj OJ -^ E ^ r E^E /, "C X X X .= # /-* ~; J^ _ _= ^Z -C _ , o U E E tit i? >. g >, ^J -c E E I! Ii ~ ^ r- TO 5 * >* E 3 E E i 2 = ^ If ll i 4J _M ^ ,_ 2 = 2 "5 r ~ r '^"M) C c 2 E " 2 E C X l| U 111 c"o c .2 5.2 F '"^ ^ aj CJ i- u ^ ' ' 1, " ' 1. ' CJ ' ' t/ <- CJ "^ i, J I S ~ r - < sc w " _ _- n: s Published names and synonyms Basophilic (5) amoebocyte (5) "erythrocyte" Coarsely granular amoebocyte (6) Finely granular eosinophil (6) Basophilic corpuscle (6) Brown cell (8) Granular amoe- bocyte (<)) Hyaline amoebocyte (9) Hyaline amoe- bocyte (2) Lymphoid cells (10) Granular macro- phage (10) Eosinophilic granular amoe- bocyte "normal cells" (ID Leucocyte (11) "'X *a3 U "rf c f. o rv. CJ 2 o CJ CJ OJ ^J y *^ . ^ C2 ^ CJ V V >. CJ CJ CJ 4-J '2. ^ o 5 c o ^ CQ U O CJ 5 T3 1 ! If 3 5 CJ o CJ 3 g o CJ 3 .x CJ o CJ u 'cj u CJ t- CJ Cj o J a < a O O J 1-1 o < w ^ _M o" u * a fN c - ni ^"V E a 2 a o -^^ _J """ PH ca^ tt < H 356 I > \\IKI. P. CHENEY c CJ ~ '" /' ' ./ "5 3 i cu ^ E ^-^^ _rt OJ T3 ~~* T"' u 5E - C _^ C (U cu s g o E ca o >- CJ / ,- _'5 s ? O '"' >,w M G .. no -/. cu ^cu OJ 1-1 ^ fe ^I^S . _j o ca '/' Ctf O _- -/. ^ c'H 3 o O ^" ^ -jj ^- Ct GJ ^ CJ -^ CO OJ 3 t x< ^1 *> "H >; 2 ""^ rt ^ ^ 03 W3 U 2 S o '" T3 ^2 to ^j O ^ y. Q "~* "C S E ,_0 - P 1 E.2 O -^ J 1 ! l| D- ? o QJ .t- -tJ t^ Cu "^ >i fc; a; . >> 'f> rf) . 3 W) Q Q o '"* ~" ^r . ,-H *-* rt o.ri ^^ Q C C 7* .2T) = ^ T) cj u S 2 3 c ci S "5 o o u _ c ^ ^ o ri"c rt *J _i_> y O U ^ ^ M-i X 3 < ^ ^ z O ^ K) U a. a. a. .2 .2 a. a. a. cu T3 a. 3_ N X O rO CN-S c 7 ' a. a. i 5 i"-5 .2 15 .2 So .2 IT) IT) _ "" ~ 00 o if. a * H ^* ~" 3 yj >, 'X c Morphology and granulated-pos- y phagocytic "normal s" e and contain granule derable variation in si i form. Nuc. vesicular nd, oval or lentiform. o d /* 2 p cu a u u O 3 CO yj ^ o o!^ cj ^ ,' " > zj -^ 3 "3 5 "5S.a __ ^ . ajC*-- S-=S omatin ne cytoplasm, slight rgin around darkly ined nuclei dant cytoplasm othe e similar to above 1 nucleus surrounded t ronglv basophilic oplasm ical or kidney-shaped :. Weakly basophilic oiilasm y basophilic cytoplasr j3.S , 03 !|5|||aMs^lS ** ! ' M-S"3 b Vj u CO *-> S 111 l S=a| u >, c a S c.i ^?* < *~- c5 rt j_i *- (J >- 2 U -^; 3 [fl u ^ cu fc D ^6 T) V o -a -a -a CU 3 n CO rf cd cd n cu -c "2 "3 "H CO X! ^ V ^* j3 j~; 3 3 '53 O cu a E a E CO "5 S E u s C/) CO tfi US CJ c " C cu "o E cu "o E 3 tfi CO X CO CO E "o "o cu "o E K "Z '~ O '"0 3 - 3 - C - O cj O cj O cj "3 cu *a1 cu "3 cu '" m '" T' E 1-1 J? (d J-l ffi K o O O U CU cu 4-1 1 U ^^ CJ U 0) cu Published nam and synonym Wandering cells (11) Small rounded amoebocytes (ID Amoebocyte (1 Small amoeboc Type A (14) Large amoeboc Type B (14) Spherical leuco cvte (13) Leucocyte (13) Amoebocyte (1 Basophilic amo bocyte (15) "a ~o"^ "a o^ 'C o 3 ^3 ^_ "a ~ T; M" E u ^ a cu ; cu OJ ; c S CO >. I o' o O > " o CJ 3 cu S cu .- General Granulo Leucocy granul 11 13 E o> 0) BO CO Leucocy Leucocy Leucocy o M CO 33 a "a "E O 00 S "2 "S 'u '0 'o co o) JO U s 5 3. ^" E CO __ 3 "8 a ca <-" "O (X j| "flj a ca G C C/3 ill Lepidoptera. Similar usu. smaller than body ;. Imp. in trephocytosis ;i'J). Prob. differentiate asmatocytes (18) 'hilia u PL, Granules Cell- ai leucoc\ S X > ' c u t, Probably T .- > _^ o o ' ^ a; 3 x" 1 ' u C x X" O i>.2 o,-i ^ - P x C'>^D Q u i- S- ^ o x ^ ^* ^~ U * c S l i* As above intense portant nnitntv ~- Z '~ ^^ a. D .2 a. a. a a a. a. 3. N; "o O !S -V] fV r3 V c c - 55 a. IO *-*-< 1 1 'N ^1 a. * NO cs i T o O O \r. 00 Q 00 u-. -" *~ X 41 j~ , 75 "SS^ gill H'-f OJ x CJ j~; _^ o , 3 3 >.-^i > _o i U. ^ "^ >, M _o O O rt "Slil 8^1 ill! n (5 E x "3'^ ~ L ~ :T >. 7 "M "l^ E HI 3 '3 o 3 <33 So Z am. ~.= i^3 '="3 2 > S 2 * t ^ Ilssf : ill| ! ^ ^ ^ ^ x'H ^-2 c 7 *- rt co o 3 .*; a i"- r S " "3 .5 S-" B S .y IH O S |II ^ C x 5 x ** Ells Is o -f _, i_ ^, t- O sc"S o ~ '-> Sill / -3 >'~Z. " ^i- 9C 3 Bl M rt gS D^ - '-'"3^-a o 2 - x 13 i C c c r- a; rt <^ ^ --^H^ 5 ^ "^ III extensions ave many dist round, colorle philic granule S 5"S'x"g.S X -j'5^ 3 -^-_ o ? >,"rt O S *~ 5 5 c E E Z S & a S .-^ o ^ ^ ^ rt J rt . rt " M'C 55 3 rt"aj ^ o >- u ^ 1! u H "^ '^ ^ ^ x " "~ Jirfijg cu o u S *1 O H C/3 22 _ > OS > a" T3 , SJ 2 rt rt rt a! a) X .2 ~ 333 3 3 5 _^ _-; ^ S* _-; __ O o = C > >; > _>i > t r, u-2 P E S a E E "3 "3 E S "7, o '" o r o '' 0x0 r ' O U u OJ OJ u 0) OJ O O U O O u = K I I S i X o "^ ^^ 11 ~ 5- c 4J 6 x "f '1 ? i, ~i e e c ^ O f QJ ^ +_> f 1 o 41 ^ X- '-' ? ^2 - 2 41 C a; ^ O ' o 2 & >. ' o ^~ o' C - C* S & ^ .S >. >, o Z r- ^ CC ' ' S>i O ~, ' 1, ^ ^Tj r/j X rt """"'' 5 y "3 >. OJ c rt ^^ *-^ QJ ^j -^ ^ 3 / v "* ^ ^-. O i- .'rt 11 CL, .2? C nog E 1 2 J S O 5 = 5 .S >-3 J U P QJ ^ E = X - O "x a ~ >. c OO w" c ~^"s. 5 ~" d' < a. J OJ __ (U : u a; Cl i> V +J OJ > 5J MJ f. 1) ^ 5!, 4J i . i^-- 4-> 4-> f*-* a 4_1 o U o o' _o o o 1-. O CJ 00 = O lasmatocytes < fused for podocytes (1Q transform into podocyt Very rare (in 2 species) In Drosofhilia Immature granulocyte (2 Many of living cells ex] pseudopodia 0} > o E - y u -u v 2 2 > ^ >- z 3 ~ >- ^ .ti u o /. ctf i- 3 11 llollg s" | 6 '^i1 2S r S|?-3ST MSS- ! 3 ^^ '-cac>. a -aj.. llim^^ SSIliiltil ua;o^j-i-'^^Pajro position ucleus central to eccentr Granules vary from amp ~ =^0 30 || ||J f|| |i| "Or-Cj^^'Cdr-'-^rt >,"w '-^ "^ rr, 2 -*"" T? r- 1 ^ ,r-CC -^0^^ 3^i oo -2SI..J T-, 0. -w CJ ^ "3 ^ s|g"5 g s , 3 > a . tu u nj . u 3i S F-o; 5" yj 3 *^ 4 ^- y? OJ y OJ GO i- ed o '^-c Ua^ls-Su y^y-.SSSSjS P > < z Z z & (S) 06 O a 0) aj yj ^ r* i , r- r 1 " ^ O a ~tr "tr " p, o, t: -x o aj E C c E E E 5 a; '*> J^ ** ^- ^> '^- -*' ^*' ' "o "o 'o 'o o o "3 o > E E E E E S. E E OJ a; OJ B E E I I a a a a yj ^ S OJ e E is i- PN rsj o' o o !Q * s n! _rt ^ l^ _o jD CJ -^ ' ^ rvi CJ ^ c 3 a^ O 3 3 i- ^ >. ' R ' ' a g ^^ C 7^ >. "S i cr, >."" M >. ^ "'I -3 u, M ^-v tn oc -^ e ^ J3 ^ """ ,-x c Is 1 11 *^ O Q^" QJ . % "3 B o -A 10 u l/^ ^^ --H l/^ 1 _^ ^ ^ - -rt jD O ^ tj^ o E 13-3.3 13 ^i Oflrs g'N'rt^N "3 J- C^ Cd -2 *- J^ i- 3 ^ OJ " ' Hi o rt >> x " E ^ rt ' ' "S ^* a, o 2 IX J EW GO j j a GO ao U (U 0) 2 i! S 2 ^1 ^* CJ ,^ o cj* c- 1 y 4-> V o ofa O ^3 o o o 2 l~l 4-> O X "5 -^- *3 - 3 "3 3 So S"-- 2 JW = r rt o ns (3 c cs 13 o^ C > a o - x X x 3 cu c Cv C X ^ -c c S'-5 rt n -- i S5'~ i 2.-- z. '0 T 'o CJ Cu p. J:^o C o "^ O O *-! 3 "" 1 t! si "cu O cu cu.s CJ a I x cd c ^ ^ . '- ' ~- _o 5 C ^*-i ^ O *" o iJ x - ~ o O " CJ "^ 3 ^ O " &J o 3E o 2 _CO CU C CJ III ^ s ~ a'~ 1 ?s nil 1 IH CU ^15_ S<2"c3 o en E S c cu tu 'J) 2 H 0~ x i- >: < < ^1 O oJ Cu a. a. a. a. a. a. a. 4. cu a. O oj CN M ^- 00 i/~- o O W t'-S T-5 rs.2 1 T3 f '-5 7 xx H 1 "" t vO 00 >c "" 00 O *"" J3 CO o fill u 1 .' rt ^- x. ~ ^ j2"5 2 cj .r: co 3JC ^ E Cfi nJ 1 1 !l|l laS-g o -^ ^C ^7 ^ u' V 5:-- w . J;- i > a ~ - ^ - '' y r- "t" " X* ^5 *"" '"" 2^ ' '*' n, i"* 1 CO ^ x . >. M ^ x '-C C rf *J r- 3 >. > - '> 3 cui cu . - CO F ^ "^ " O 11^^ 2 3.5 & ^.'""3^ c c _O " O "^ '-J c5* _ 5 +_> -3 4_> C *~ < CO >> co '" - .5 '"^ /' *3 ^ J ~ C 3 3 H n "o o ^3 "^ *" I*-! f~" ^ Cd o w 'C o ^ cu w O 3"^ CJ x - a.'"" j ^' >, 3 'S "* ^ *" ^ c . r- ^ 3 O x " !/) CJ 4_ - rt cd ^ cu a. ^_, ^i -t-> ^ ^, ^ ^ Tf ? -f 'i "^ r^ ? . JO CO X **" ct x ^ x x ^ '> "C -^ 5 S ^* i ^ X i/~ 3 X (U E e - 5 a x ^- "* -f 3 ^ *o IH o T K, 3 - "r* cd ^U - g x rt ^.S " ^~"S ""c 3 '_S 3 t ^ . j-r X'5'c'n' 'o s u "o 5 S^ 5 o of |j -S^l i"s >^ ^ * o (U "o'|-" '-"= S^^'i | 2 Ji; 1 3 5 P S S 3 X _2 '-^ I/ "~L "" u'cc-^-^ 5 2" 2 j-j rz^ >, -S x 'p * 5 " S 5 n "3 s "x r" "g j?e ^s.s"i5 M^:^~ Ucj^Zcj *o "M ^ rt ll llll I'C o " c5 u s ^n3"o^^o-s^ * S g >. " 1 2 1 ll II ^ j? II |2 > ( J? " II -d OJ l_l -w - - c = - - -r ~ c fv _, '3 '^ '3 3 3 3 3 3 3 5 o, re CC c C 1- ^c QD ^e i- ^c OJ ^CJ O o CJ ^0 ^0 CJ u ^ E E E g E g o Q _c _o c o c o c o Q ^ E ^ "cu "c^ ^ Cu *7* "T* *flj CU c o o O O O o o o K o O U O U U U as g^ C8 - ^-y y 1 % j, 11 ~ Cu s S "3 ^_. ^-* fM 5. ^ g cu ; o2 cu C .= >> X W 3J ^"^ "7' *U r-^ CJ *" 2 s! ^^ |-|E 4J -*-> u >. o r2 cu 1 ^ 1 2 cu f o!? 3"^>. o^"i II O x o S p rr 5 J Ja"rt - o| f'SIl s & i'S rt-2 2^- 2'S w i-l S o cTTfi. E > E cu PL, ( >i C cu cu CU o CJ CU ^ - cj r> - 15 C = cu _o "3 _0 >> "3 S o o u cu :: cu ly 3 13 "3 5 c'5, o co ^r 3 g "3 "3 O K K w u o 1- CU J u a j. ^o CO VJ Tf c 1 ^" QO"C_X T3 rt^^ C nj ^ cd ^~ co ST3". c. ilS --" * ^ 3 u 3 " T" 3J.C - T3 ^ ^ T3 -JOO OH J2 ^ ' '^ w w s "s K < 360 DANIEL P. CHENEY cu IT Sg- o | Properties and special characteristics 'efined for Arbacia sp. where th account for 42% of all cellular elements in coelomic fluid (32. or Arbacia. Red form (eleocyt is rich in glvcogen and contair echinochrome (28). CDC-Rei 48% and green 10% (32) ound only in the irregular urch so far ound in all regular echinoids. Cells do not contain hemoglob but do disintegrate rapidly up contact with injured tissue (2* 'as shown to have a blade-like flagella. May be a precursor phagocytes (32) bundant in coelomic fluid ound in small numbers annot yet be correlated with o echinoderm leucocytes a to DH l r I > < U, O a. a. 3. a. a. cd cu 2 d ^: 3 c 35 i -5 X 7 '-57 a. X o t ^ Cf~' & - m - 2 c i E -C x OJ 0; t. g i g C i r M 3 o M "o p] Ul luroiorm. Vary pherical to ovoid mroid form. Gree and red spherules teen seen. Spheru form sometimes r than in other for Eg E^p- >rt li; l'gjl-3 sSlts si^5 fi j C >T3 3 S c 25og S 'o 3 J 'o -^ T. o a . cu ~ E o v a '3 Sa'o |SSJ| E I* 'o l_ - "^ o a c ca X 3 CU CJ often contains gra Short pseudopodia tive phagocytes and usu. vermifoi les and rods have ; , for saffraniri. Ni 2 a (C |E O o Ki ~ &--0.2 O O Cu u --- ~ '- "- 2 I x2.c c ^_j 1/3S - 1 - CU ~2 CU Cu ^- c J2 >, o n o g -g u 5 >. g ?-jg cat. o _O "o ffi c - ~ llll c 3 ffi 3 ca ll 5 ^ 3 Z a a e-g ee?i rt^oO ri y) 3 U "o II II 33 P-, II II u u II UJ o cu IH o; a ~ -a _. a a pQ '3 3 3 '3 '^ '^ 3 '2 O TH C qr; * ^ tC ^ q: 12 u CJ CJ CJ CJ o o u CJ u CU g 'E 1 g 5 E E E 'g " _C O O _O o _o j> "3 *~ "^ ~ cu *cu *CU "3 OJ o O O O o o o U U U o U u O cu , , ca Published name; and synonyms Colorless morula cells (29) White mulljerry cells (28) Colorless trepho- cyte (32) Colorless spherul cell (34) Colored morula cell (29) Red mulberry cell (28) Eleocyte (29, 30] Colored or red CJ ^ ^ _ ^v| QJ s ^ ^ ^* O '', "*, . . > r^ . v_x "w ' X _ J^ ' ' ~J "^ 3J ^^ if ^^ *" *~* ""^ ^ ^- U J M IX Cu J ca 00 Cu ^ 00 CJ o -r. 3 U, (28, 29) Bladder filaform phagocyte (29, 30) Colorless morula cell (29. 30) Colored and colorless morul cell (5, 29) Phagocyte (29) Cfl 11 2 * ~ 5 cu cu cu i > CJ o u' 4-1 o O cu o o cu u CJ OJ ^ 3 3 ^ 3 o EU cu >, >! CJ CJ r^-. f^- cj U *3 ^O CU CU OJ OJ ~3 3 _3 -H 3 3 ^ r^ C r- ^H """ M a ^ u ca ca ? ? ca cd S O O K O IM c K t/j c c/j rt O LJ "* c 0/1 "7 V O 1 " o Os" o C (N G fN r^; "^ i/T Cd -ga _ E ^, _3 OJ > ca ^ rt 2a "tn cu r- "S c 0, i- a; 4_i 'o C- 1 x >1 ~, < O MA. VILA CLAM LEUCOCYTES 361 "2 8 K *s* W^ 23 i^'f cn r ^t .2 t^ Q ^ ^j "o -S H -r-, P rt ., ^ IT I/ C CT r^ 3 ". r "" C g"^ o i C i_ ca aj P ? a %i || = 1 c ' o - - '.n >- u ca Cj ^~! s. o o Sf OJ O"Q 2t || 11 -S 513 S 5 u. C 5 ^ 0, t; M . o 2 S *o o ?^ cu E|E|. -5 ?f 111 pg Jj^ = JH J^ I" 5 6 a. CU 17* s; 1 w -"" dj u^ 1 ul'3 fit _o - r, i-> E III i *" C "3 O O o c a; O C '/, o J5 o, IH o g 6 - 2 , il lE.I |||^ s "2 "s 5 3 Is >. 15 >.- "~ u M'-W g .^ .2 ;> ^: 'o 0" { aJ 1 |5|| "3 "o ^J3 M j'S g_o > U N vT^ ^ r 3 a a Q o c> c II S II II II _)' o E cu > IH 0] .| o c 1 i -* O E C5 o o o O) o o< o UU u O C/j ^ ^ ^ la yS ^ s sc i -r ^ SS = 2 u ^ c , , o tj ^ _ >, 1 cu C A > ^- >-, r f. *^- c ** - "c ^~^ ^ ^ t w, '- 3"H S "^ = -** C O 1* 2:1; "~ ^ c >, U - H; ^ . U, t C s o,K u 11" < cu U OJ 0) a T* r* -J-* >> o o o ^ O o dr " 00 u o u a; cu tu cu > 11 >. s ^-i ^-. o ~ >H V CU OJ _O 3 .1 C C. "H ^ 5 ^"j c " cu o o o tn C/7 ca c*3 C a" t a r^l rt ^^ c ^ o CD ^; E CO jf 2 "2 cs.2 JS u rtja eu 3 O U C u O K .- .i: Ci rt a c CO -x s o'd H S H ^-^; ffl 366 DANIEL P. CHENEY occur in contrast to the sharp differences noted between large and medium types. The cells with small nuclei are round or extremely elongate. Small elongate, irregular nuclei were usually euchromatic; however, heterochromatic or pycnotic round nuclei (usually 1.5-3^) were seen in varying proportions depending on the sample time and staining intensity. This heterochromasia was usually but not always associated with large cells (small nuclei) containing many granules. In bright field, with unstained preparations, this granularity presented a yellow to yellowish-brown coloration, the latter denned as a pigment cell character. The nuclei ot all cells seemed to become more eccentric as granularity increased and the shape remained round except when compressed against the cell mem- brane. Except for some large leucocytes, nucleoli were not visible. TABLE II Differential counts of ronncctirc tissue leucocytes in the Manila dam, Tapes semidecussata. Group I (n == 14) was held for 1 to 30 days in naming scau'iiler. Group II (n == 14) was held for 30 to 50 days in the same svslein. 1011 cells were counted per individual Group Nucleus dia ( Cell dia (n) Mean % S.D. % Subgroup total % Group Nucleus dia (ji) Cell dia (M) Mean ', S.D. % Subgroup total % I 3-6 11.3 5.9 II 3-6 14.0 7.1 1.5-3 6 12 19.5 5.5 40.2 1.5-3 6 12 32.1 12.8 59.0 12-16 9.4 3.5 12-16 12.9 4.7 3-6 31.6 12.3 3-6 21.1 9.8 3-6 6-12 18.4 6.2 53.9 3-6 6-12 12.9 7.3 37.8 12-16 3.9 2.2 12-16 3.8 2.9 3-6 1.2 1.5 3-6 1.0 1.3 6-8 6-12 2.9 1.7 5.9 6-8 6-12 1.5 1.9 3.2 12-16 1.8 1.3 12-16 0.7 1.3 1.5-3* All 24.5 6.6 1 .5-3* All 40.0 13.2 * Heterochromatic nuclei. Changes in the proportions of leucocytes possibly 7 due to prolonged storage of clams under artificial conditions are shown in Table II. A comparison of differ- ences in relative numbers between cell sizes and groups held for 1 to 30 days and 30 to 50 days suggest an increase in small-heterochromatic forms in the latter group. Tests employing the single classification analysis of variance were con- ducted to determine the significance of differences between groups for each nuclear size classification shown in Table II. The significance levels for acceptance of the hypothesis that the means of both groups were similar (M, =: M 2 ) were as follows: all small forms, P < 0.001 (F : : 18.74); heterochromatic small forms, P < 0.001 (F = : 19.68) ; medium forms, P < 0.005 (F = : 10.57) ; and large forms, not significant (F 2.31). The heterochromasia exhibited by many small cells probably indicates some loss of cell function or degeneration (Robbins, 1967). The gross and histologic appearance of both groups was similar. MANILA CLAM LEUCOCYTES 367 SUMMARY The morphology and known functions of leucocytes have been tabulated and compared for most invertebrate phyla. The leucocytes of the Manila clam are similar in form to molluscan leucocytes described by other investigators. Two general types were seen. One was a hyaline leucocyte ("leucocyte") and the other a granular leucocyte ("granulocyte" ). Cells in these categories showed extreme variations in size and staining characteristics. A decrease in nuclear size often was correlated with an increase in nuclear heterochromasia and eccentricity and cytoplasmic granularity. The average cell count for all circulating leucocytes was K,50 180 per mm 3 . LITERATURE CITED ABRAHAM, M., 1963. Etude morphologigae sur les coelomocytes des Echinides reguliers ( A morphological study on the coelomocytes of regular echinoid.es). Israel, J. Zeal., 12(1): 101-116. ANDREW, W., 1962. Cells of the blood and coelomic fluids of tunicates and echinoderms. Amer. Zoo/., 2 (3) : 285-297. ANDREW, W., 1965. Comparative Hcmatology. Grune and Stratton, New York, 188 pp. ARVY, L., AND M. GABE, 1949. Contribution a 1'etude morphologique du sang des Polyplaco- phores. Bull. Soc. Zoo/. France, 74: 173-179. ARVY, L., AND M. GABE, 1951. Donnees morphologiques sur le sang du Dentale. Bull. Lab. Maritime Dinard., 35: 15-22. BANG, F. B., AND B. G. BANG, 1962. Studies on sipunculid blood: immunologic properties of coelomic fluid and morphology of "urn cells." Call. Biol. Mar., 3 : 363-374. BETANCES, L. M., 1922. Les cellules du sang de quelques Lamellibranches. Arch. D'Anal. Microscop., 18: 309-327. BLOOM, W., AND D. W. FAWCETT, 1962. A Textbook of Histology. Saunders, Philadelphia, 720 pp. BOOLOOTIAN, R. A., 1962. The perivisceral elements of echinoderm body fluids. Amcr. Zoo/., 2: 275-284. CAHN, A. R., 1951. Clam culture in Japan. U. S. Fish. U'ildl. Sen: Fish. Leafl., 399: 1-80. CAMERON, G. R., 1932. Inflammation in earthworms. /. Pathol. Bacterial., 35: 933-972. CHENG, T. C., H. W. F. YEE, E. RIFKIN AND M. D. KRAMER, 1968. Studies on the internal defense mechanisms of sponges. III. Cellular reactions in Terpios ceteki to implanted heterologous biological materials. /. Invert. Pathol., 12: 29-35. O'ENOT, L., 1891. Etudes sur les sang et les glandes lymphatiques dans la serie animale (2 e partie: Invertebres). Arch. Zoo/. Exp. Gen. Serie 2, 9: 13-90, 365-475, 593-670. CUENOT, L., 1897. Les globules sanguins et les organes lymphoides des invertebres. Arch. . hint. Microscop., 1: 153-192. DREW, G. H., AND B. A. CANTAB, 1910. Some points in the physiology of lamellibranch blood corpuscles. Quart. J. Microscop. Sci., 54(4) : 605-622. DUNDEE, D. S., 1953. Formed elements of the blood of certain freshwater mussels. Trans. Amer. Microscop. Soc., 72(3) : 254-264. ENDEAN, R., 1966. The coelomocytes and coelomic fluids. Pages 301-328 in R. A. Boolootian, Ed.. Physiology of Echinodcnnata. Interscience, New York. FLORKIN, M., 1960. Blood chemistry. Pages 141-160 in T. H. Waterman, Ed., The Physiol- oo v of Crustacea. J'ol. I. Academic Press, New York. GEORGE, W. C., AND I. H. FERGTSOX, 1950. The blood of gastropod molluscs. /. Morphol., 86: 315-327. GEORGE, W. C., AND J. NICHOLS, 1948. A >tudy of the- blood of some Cru>tacea. /. Morphol., 83(3): 425-443. GRASSE, P., Ed., I960. Traife de /.oolauie. Vol. I'. Masson, Paris. LIBRARY 368 DANIEL P. CHENEY HAIGER, S. A., 1964. A histological and cytological study of the "brown cells" found in the "auricular pericardial gland" and other tissues of the oyster, Crassostrea viri/inica (Gmelin). M.S. thesis, University of Dchiwirc, 51 pp. HAIV.IITOX, I., 1934. Amoebocytes and allied cells in Invertehrata. /. Ru\. Microscop. Soc., 54:246-262. HYMAX, L. II., 1 ( >(>7. The Invertebrates. I'ol. VI, Molhiscn I. McGraw-Hill, New York 7')2 pp. JiHixsoN, P. T., 1969. The coelonu'c elements of sea urchins (Strongylocentrotus) 1. The normal coelomocytes ; their morphology and dynamics in hanging drops. /. Invert. Pathol, 13: 25-41. JONES, J. C., 1959. A phase contrast study of the blood cells of Prodcnia larvae. Quart. J . Microscop. Set., 100(1) : 17-24. JONES, J. C., 1962. Current concepts concerning insect hemocytes. Amcr. ZooL, 2(2): 209- 246. JOXES, J. C., 1964. The circulatory system of insects. Pages 2-107 in M. Rockstein, Ed., The Physiology of Insects, I 7 ol. Ill . Academic Press, New York. KINDRED, J. E., 1929. The leucocytes and leucocytopoietic organs of an Oligochaete Plicretiuni indica (Horst). /. Morphol, 47(2) : 435-467. LIEBMAN, E., 1950. The leucocytes of Arbacia piinetu/ata. Biol. Bull.. 98: 4n-59. MAUPIN, B., 1969. International scries of nmnoi/raplts in pure anil applied hiolouy. Division: Zoology. Vol. 41. Blood platelets in nnin and animals. I'ol. 1 and J. Pergamon Press, New York. METCHNIKOFF, E., 1968. Lecture on the comparative pathology of inflammation, delivered at the Pasteur Institute in 1S Q 1. Dover, New York. MULLER, G., 1956. Morphologic, Lebensablauf, and Bildungsort der Blutzellen von Lyiniitien stai/nulis L. (Course of life and site of formation of the blood cells of L. stai/nalis I,.). Z. Zelliorschung, 44: 519-550. OWEN", G., 1966. Feeding. Pages 1-52 in K. M. Wilbur and C. M. Yonge, Ed., Physiology of Mollusca, Vol. II. Academic Press, New York. PAN, C., 1958. The general histology and topographic microanatomy of Aitstralorhis gla- hnittts. Bull. Miis. Comp. ZooL. Harvard Univ., 119: 237-290. RIZKI, M. T. M., 1957. Alteration in the haemocyte population of Drosophila melanogaster. J. Morphol., 100(3) : 437-458. RIZKI, M. T. M., 1962. Experimental analysis of hemocyte morphology in insects. Amcr. ZooL, 2(2) : 247-256. ROBBINS, S. L., 1967. Pathology. Saunders, Philadelphia, 1433 pp. ROSE, C., AND S. SHOSTAK, 1968. The transformation of gastrodermal cells to neoblasts in regeneration Phagocata gracilis (Leidy). E.rp. Cell. Res., 50: 553-561. TAKATSUKI, S., 1934. On the nature and function of the amebocytes of Ostrea edulis. (Juart. J. Microscop. Sci., 76: 379-431. WAGGE, L. E., 1955. Amoebocytes. Pages 31-78 in G. H. Bourne, and J. F. Danielli, Eds., International rcvieiv of cytology, Vol. IV. Academic Press, New York. WIGGLESWORTH, V. B., 1956. The haemocytes and connective tissue formation in the insect, Rhodnins proli.nis (Hemiptera). Quart. J. Microscop. Sci., 97: 89-98. WOOD, P. J., AND L. P. VISENTIN, 1967. Histological and histochemical observations of the hemolymph cells in the crayfish, Orconcctcs virilis. J. Morphol., 132(4): 559-567. YONGE, C. M., 1923. The mechanism of feeding, digestion and assimilation in M\a arenaria. J.E.i-p. Biol.. 1(1) : 15-63. Reterence : Biol. Bull., 140: 369-375. (June, 1971) REMOVAL OF THE FERTILIZATION MEMBRANE OF FERTILIZED EGGS OF URECHIS CAUPO AND DEVELOPMENT OF "MEMBRANELESS" EMBRYOS l WII.MA SILVA KXGSTKOM - Department <>/ y.olof California, AYr/,v/<'v, Cdlifumia "I72n Embryos of i'rcchis cuitpo are invested by a fertilization membrane, and unlike many other invertebrates, embryos of [Credits do not lose this membrane during "hatching" of the blastulae. The removal of the fertilization membrane is a pre- requisite to embryo dissociation; embryo dissociation has been employed recently to investigate the role of cell interaction in the regulation of biosyntheses (Giudice, Mutulo and Moscona, 1967; Pfohl and Giudice, 1967; Shiokawa and Yamana. 1967; Sconzo, Pirrone, Mutolo and Giudice, 1970; Engstrom, in preparation). This paper describes a method to remove the fertilization membrane and the sub- sequent development of the "membraneless" embryos. MATERIALS AND METHODS Collection and culture of //ainctcs Specimens of I/rechis cauf>o collected from the mudflats of Bodega Bay, Cali- fornia, were housed in aquaria containing aerated recirculating sea water at 8-12 C. The collection and filtering of eggs has been described (Gould, 1967). Unless otherwise stated, all incubations of eggs and embryos were at 15 C. A one per cent suspension of eggs (vol/vol) was inseminated with a fresh suspension of sperm. Five minutes later, those eggs destined for demembranization were washed and resuspended in millipore-filtered sea water (FSW). Control eggs (with membranes) were similarly washed, resuspended at 2 C in FSW contain- ing 0.1 mg each of penicillin and streptomycin/ml (PSSW) and then transferred to a 15 C water bath for subsequent development. Embryos were cultured in silicone-treated glass petri dishes in volumes not exceeding 5 mm in height in order to allow good aeration and development of the embryos. Untreated glass, agar- coated glass and plastic were not suitable containers for culturing membraneless embryos due to adherence of denuded embryos to these substrates. Removal of the jcrtilization membrane \Yhen the fertilization membrane (FM) commenced to rise (10-15 minutes post fertilization), eggs were centrifuged, resuspended in calcium-free sea water (28.4 g NaCl. 5.4 g MgCl 2 -6H 2 O, 0.78 g KC1. 7.12 g MgSO 4 -7H 2 O, 0.2 g N T aHCO 3 per liter, pH 7.8 all reagents were analytical grade from Mallinckrodt) containing 1 inM disodium ethylenediamine tetraacetate, pH 7.8-7.9 (CFSW- EDTA), centrifuged and finally resuspended at 2 C in 20-25 volumes of a sucrose-EDTA solution (SE) 0.87 M sucrose, 5 rml/ DTA, pH 7.8-7.9. The 1 Work performed under USPHS Predoctoral Fellowship 1-F1-GM-33, 889-01 and sup- ported in part by a grant to Dr. Fred H. Wilt from USPHS (GM 13882). 2 Present address : Department of Biology, University of Oregon, Eugene, Oregon. 369 370 WILMA SILVA ENGSTROM egg suspension was transferred to a 15 C water bath and periodically inverted very gently to mix the suspension. Fertilization membranes became progressively thinner until they were no longer visible by light microscopy ; this typically re- quired 10-15 minutes. At the first sign of complete dissolution of the FM, the suspension was gently diluted with 30 volumes of FSW. The above treatments always resulted in the lysis of some eggs (5-20%). The final suspension was centrifuged at 40 X y for .] minute in a Sorvall GSA rotor. These demembranated eggs were resuspended at a concentration of \ f /o in I'SS\V (assuming no loss of eggs during SE treatment ) and cultured under the same conditions as the controls. When necessary, the cultures were gently rotated to resuspend any embryos adher- ing to the siliconized glass. Mechanical treatment, thiol reagents, proteolytic enzymes, nonelectrolytes and combinations thereof were tried without success in removing membranes. Al- though prolonged exposure to urea or pronase did remove the FM, further develop- ment was inhibited. See Berg (1967) for a review of methods of FM removal. TABLE I Removal of the fertilization membrane. Fertilized eggs were treated with the reagents listed under Treatment; the duration of exposure is noted under Time; the Results indicate whether the FM remained ( + ) or was removed ( ) following treatment Treatment Time (post fertilization) Results Sucrose (0.87 M), 15C 8 minutes-24 hr + Sucrose (0.87 M) + EDTA (1 HIM), 15 C 7 minutes-lf hr Sucrose (0.87 M) + EDTA (10 HIM), 15 C 10-20 minutes (no division bv 7 hr) Sucrose (0.87 M) -f- EDTA (5 HIM), 2 C 10 minutes-25 hr + Sucrose (0.87 M) + EDTA (5 mM), 15 C 10-20 minutes - (good development) Sucrose (p.87 M) + EDTA (5 mM), 15 C un- 10-40 minutes + fertilized eggs in FSW at 2 C for 6 hr. prior to fertilization OBSERVATIONS Removal oj the FM A non-electrolyte (0.87 M sucrose) in combination with a chelating reagent (5 mM EDTA) proved to be effective both in removing the FM of freshly fer- tilized Urechis eggs (10-25 minutes postfertilization ) or blastulae, and in allow- ing development to ensue. As can be seen from Table I, too high a concentration of EDTA (10 mM) led to FM~ eggs which did not continue development, while too low a concentration (1 mM EDTA) removed the FM of only a small propor- tion of the eggs. The FM appeared to be solubilized at about the same rate over its entire surface. The temperature of incubation in SE proved to be important, 15 C being effective while 2 C was not. In addition, if unfertilized eggs were held at 2 C for 6 hours, though they fertilized and developed well, their FMs were resistant to dissolution. FERTILIZATION MEMBRANE REMOVAL 371 Development of embryos lacking FM Through the 4-cell stage, FM~ embryos lagged by 5-10 minutes, thereafter the embryological development of membraneless and control embryos seemed equiva- lent as far as rate of development and morphology of developmental stages (Figs, la and b). Cleavage planes were the same in control and experimental embryos for the period studied, i.e., through the 16-cell stage. FM~ embryos were able to form respectable blastulae which began swimming shortly after 9 hours. Culture dishes of demembranated embryos always contained a small number of isolated cells. Gastrulation began at 17 to 17] hours in control and experimental embryos. Both control and FM" embryos exhibited a concentration of red pigment granules (hematin (Baumberger and Michaelis, 1931)) on the surface around the blasto- pore. Generally, the diameter of the blastopore of an FM~ early gastrula was 1.5 times that of the control. By 21 hours, the embryos were well-defined gas- trulae. At this time a small population of green pigment granules was present on the surface below the prototroch over the region of the future intestine. Newby (1940) states that Urcchis does not possess green pigments though two other spe- cies of echiuroids do (Tlialusscnia and Echhirus). By 25 hours, both experimental and control embryos possessed a tuft of long apical cilia and the appropriate divi- sions of the gut. At this time, one difference in these embryos had become quite apparent, the membraneless embryos were more elongate (Fig. Ic and d). This difference was more strikingly expressed as development proceeded. At the level of the prototroch in the trochophore larva (43 hours ) , the width of the experi- mental embryo was typically only about 0.7 while its longitudinal axis was 1.3 times that of the control embryo ( Fig. le and f ). By two days, FM" and control em- bryos were feeding and able to contract along their longitudinal axis, notably changing their shapes. Newby (1940) observed such contraction at four days of development. An unexpected difference between FM" and control embryos was that at all swimming stages, the membraneless embryos were less motile than the corresponding control embryos. Unless noted, the embryology described above agrees with that presented in Newby 's monograph on 1'rcchis development (Newby. 1940). Lastly, cultures of FM~ embryos always contained some double-embryos whereas the controls did not. Sometimes these embryos were merely "joined" at some point, usually at the pretrochal level; more often, their apical ends were "fused," there being a digestive system at each end (Fig. Ig). It may be in- ferred that one function of the FM is, so to speak, to keep an embryo unto itself. This same phenomenon of double embryos has been observed in the sea urchin (Vacquier and Mazia, 1968b). From the above observations, it was concluded that removal of the FM does not interfere with normal embryogenesis in terms of such events as blastulation. gastrulation. ciliation or pigmentation, but does have some influence upon the motility, shape and "oneness" of these embryos. Biochemical data suggest that FAT embryos are normal in their time of initiation of synthesis of rRNA (Eng- strom, in preparation). 372 WILMA STI.YA ENGSTROM a g FJGTKK 1. I'rci'his embryos as viewed by light microscopy: (a.) 6 : ; hour FM', apical view; (b.) 71 hour FM~, a])ical view; (c.) 24 hour FM + , latero-posterior view; (d.) 24 hour FM", latero-anterior view; (e.) 46 hour FM + , lateral view r ; (f.) 46 hour FM~. lateral view; (g.) "Double" embryos (FM") at 4 days as viewed by light microscopy; lateral view. Bar on figure represents 50 UL. FERTILIZATION MEMBRANE REMOVAL 373 DISCUSSION Mechanism of membrane r In hindsight, it was not surprising that a nonelectrolyte along with a chelating agent would remove the FM from a marine embryo which appears to retain its FM into late development. The necessity to invoke tin- use of a chelating agent in addition to sucrose suggests that ions, especially divalent cations, play a role in the integrity of the FM of Credits eggs. These findings are in line with Temin's ( 1956) proposal that the FM of the Urechis egg is composed of two layers, an outer protein layer, soluble in nonelectrolyte, and an inner calcium-proteinate layer. In addition, the removal of the FM from unfertilized eggs of 5. f>nrf>uratus with non-electrolyte can he prevented by the addition of ions, particularly divalent cations (Dean and Moore, 1949; Moore, 1932). The sensitivity of membrane removal to temperature may mean that by 2\ hours EDTA had not yet had access to the Ca-proteinate layer. It is possible, though not to be expected, that Ca ++ ion deposition in the membrane was greater or the bonding more tenacious at lower temperature. Unfertilized eggs seem to "age" and render the FM more resistant to the action of SE. It is difficult to account for this finding but, since the eggs were not observed for a long period of time, it is possible the FM would have eventually dissolved. Resistance to treatment cannot have been due to the usual "hardening" process since the FM can be removed at least as late as the blastula stage. The unexpected difference in the degree of motility between the membraneless and control embryos might be explained by supposing that the FM has an energy- providing role. Since the cilia pass through the FM this seems a reasonable hypothesis. A second alternative is that some component (s) of the cilia may have been solubilized or inactivated during exposure of the embryo to SE. It has been shown that some components of isolated cilia of Tctraliyincna are solubilized in low ionic strength-EDTA buffer ( Renaud, Rowe and Gibbons, 1968). Since SE treatment of Urcdiis embryos preceded the appearance of cilia by at least 7 hours, the above speculation presumes: (1) the existence of a pool of ciliary proteins and (2) that the effect of this nonelectrolyte-chelating solution (SE) would be at the level of ciliary protein units or subunits. Evidence for a pool of ciliary proteins has been reported by Auclair and Siegel (1966). Lastly, the present observations cannot exclude the possibility that FM" embryos simply have fewer cilia and so are less motile than FM + embryos. I >cTclopnic)it of membraneless embryos I'p through the trochophore stage, the membraneless embryos appeared normal in their emhryological development though abnormal in shape. This difference was first detected at the beginning of gastrulation. Recall that the blastopore of the FM~ embryo was larger than that of the normal embryo and that older experi- mental embryos were longer and narrower than controls. The explanation for the earlier observation is probably that during invagination. the FM being relatively inelastic, acts as a restraint against outward directed cell movements. One would expect this lack of restraint to be reflected in the overall diameter of the blastula, especially in the vegetal region. At about 48 hours of development, the archenteron of the control embryo has undergone distent ion and stretching. By this time, 374 WILMA SILVA ENGSTROM cuticle formation has also begun (New by, 1940). Without the restraining influ- ence of the FM, this distention of the archenteron might cause the membraneless embryo to become more elongate. ( )ne can conclude that there was no cuticle formation to prevent this shape change in FM~ embryos. The importance of the FM for the shape of the embryos suggests that the I'M is retained into later devel- opment. Support for this point comes from studies on worms from another phylum. Sipunculoidea. First it should be recalled that unlike sea urchins, Urechis hlastulae do not shed the FM when they "hatch." rather the cilia pass through the I'M. This same behavior is seen in three different genera of sipunculids; further, the FM persists in the sipunculid trochophore and is transformed into the cuticle of the metamorphosing larva (Rice, 1967). Further, either a hyaline layer is normally absent, removed by nonelectrolyte, or if present, is unimportant in influencing the shape of the embryo. Circum- stantial evidence suggests that there is no hyaline layer in FM~ embryos. Moore (1930) found that pretreatment of unfertilized .S'. f>iirf>iinitiis eggs with nonelec- trolyte prevented the formation of a FM or hyaline layer upon fertilization. A more direct kind of evidence is that FM~ Urcchis embryos were able to form double embryos. This has also been found with sea urchin embryos which are lacking a hyaline layer (Vacquier and Mazia, 1968b). Assuming that no hyaline layer is present in FM" embryos, one can infer that intercellular bonding between blastomeres is important in the maintenance of structure of the embryos. This is similar to the case in the sand dollar, Dendraster c.rccntncits. in which cell-hyaline layer interaction is weak with intercellular bonding assuming the dominant role in binding blastomeres together (Vacquier and Mazia, 1968a). In summary, it may be that the FM and intercellular bonding are important in maintaining the shape of an Urechis embryo. SUMMARY It has been shown that the fertilization membrane of the I'rcclns embryo can be removed by treatment with a sucrose- F I )T A solution. The resulting mem- braneless embryos appear to undergo nearly normal embryogenesis, differing from the controls only in shape, degree of motility and ability to form double-embryos. LITERATURE CITED AUCLAIR, W., AND B. SIEGEL, 1966. Cilia regeneration in the sea urchin embryo: Evidence for a pool of ciliary proteins. Science, 154: 913-915. BAUMBERGER, J. P., AND L. MICHAELIS, 1931. The blood pigments of U. canpo. Biol. Bull., 61: 417-421. BERG, W., 1967. Some experimental techniques for eggs and embryos of marine invertebrates. Pages 767-776 in F. H. Wilt and N. K. Wessels, Eds., Methods in Derelopiiicntnl Biolof/y. Crowell, New York. DEAN, R. B., AND A. R. MOORE, 1949. The protective action of cations on the membrane forming capacity of echinoderm eggs. Biodynainics, 6: 213-216. GIUDICE, G., V. MTTOLO AND A. MoscoNA, 1967. Synthesis of ribosomal RNA during sea urchin development. Biocliiiii. Biopliys. Acta, 138: 276-285. < .oi'Li>, M., 1%7. Echiuroid worms: Urechis. Pages 163-171 in F. II. Wilt and N T . K. Wessels, Eds., Met/axis in Developmental Biolnfiy. Crowell, New York. MOORE, A. R., 1930. Fertilization and development without membrane formation in the egg of the sea urchin. Strongylocentrotus pitrpiiratns. Protoplasma, 9: ( '-17. FERTILIZATION MKMBRANE REMOVAL 375 MOOKK, A. R., 1932. Tlie dependence of cytoplasmic .structures in tlie egg of the sea urcliin on the ionic balance of the environment. ./. Cell. Comp. I'liysiol., 2: 41-51. NEWBY, W., 1940. The embryology of the echiuroid worm, Creeliis etinpo. Memoirs, Ameri- can Philosophical Society, Philadelphia, 16: 1-219. PFOHL, R. J., AND G. GIUDICE, 1967. Tlie role of cell interaction in the control of enzyme activity during embryogenesis. Bioehiin. Biophys. Acttt, 142: 263-266. KKXAUD, F. L., A. J. RO\VE AXD I. R. GIBBONS, 1968. Some properties of the protein forming th'e outer fibers of cilia. /. Cell Bin!., 36: 79-90. RICE, M., 1%7. A comparative >tud\ of tlie development of Phascolosoi/iu aiid.^si.cii. (iolfnit/iu pugettensis and Themise pyrnides with a discussion of developmental patterns in sipuncula. Ophelia. 4: 143-171. SIONZO, G., A. M. PIKROXE, V. Mr-rout AMI G. Ginm K, 1 ( '7(). Synthesis of ribosomal RNA in disaggregated cells of sea urchin embryos. Biaeliiin. /-Hupliys. Acta, 199: 441-446. SHIOKAWA, K., AND K. YAMANA, 1967. Pattern of RNA synthesis in isolated cells of Xenupus lacvis embryos. Develop. BioL, 16: 368-388. TEMIK, H., 1956. Studies of activation in eggs of Urechis caiipo, Xereis limbata and Astcrias jorbcsi. Biol. Bull., Ill: 313. \'.\CQUIER, V., AXD D. MAZIA, 1968a. Twinning of sand dollar embryos by means of dithio- threitol (Structural basis of blastomere interactions). Exp. Cell Res., 52: 209-221. VACOI/IKR, V., AND I). MAZIA. 1968b. Twinning of sea urchin embryos by treatment with dithiothreitol. Exp. Cell Res., 52: 459-468. Reference : Biol. Bull., 140: 370-388. (June, 1971) COMPARISON OF CHROMATOPHOROTROPINS FROM THE HORSESHOE CRAB LIMULUS POLYPHEMUS, AND THE FIDDLER CRAB, UCA PUGILATOR^ MILTON FINGERMAN, CLELMER K. BAUTKLL AND ROBERT A. KKASXOW I'epartinent of Biology. Tnlane University, New Orleans, Louisiana 70118, Department of Bioloay, Louisiana State University, New Orleans. Louisiana 70122, and Marine Bioloi/ienl Lal>oratory, ll'ooils 1 1 ale, Massaehusctts 02543 Chromatophore responses of fiddler crabs have been utilized to study the chro- matophorotropins that these crabs produce themselves and, in addition, to assay for chromatophorotropic substances from other species of invertebrates. Brown and Cunningham ( 1 l >41 > found that extracts of the central nervous system (CNS) of the horseshoe crab, Linniliis polypJiennts, cause dispersion of the pigment in the melanophores of the fiddler crab, Uca piKjna.r, and concentration of the pigment in its leucophores. The melanin-dispersing and white pigment-concentrating re- sponses evoked by the seven portions of the CNS tested paralleled each other so closely that Brown and Cunningham suggested both effects may have been caused by a single substance. More recently, Rao and Fingerman (19/Oa'i investigated the chromatophorotropic activity of extracts of radial nerves from the sea star, Astcrias annircusis, and found that after gel filtration certain fractions which caused melanin dispersion in the tiddler crab, I 'en pugilator, also evoked pigment concen- tration in the leucophores and erythrophores. Extracts of horseshoe crab CNS have not been tested on erythrophores of fiddler crabs. On the other hand, ex- tracts of the eyestalks of the fiddler crab, Uca pugilator, have been clearly shown to contain substances that will cause dispersion of the pigment in its melanophores, leucophores and erythrophores and concentration of the pigment in its leucophores and erythrophores (Carlson, 1935; Sandeen, 1950; Brown, 1950; Rao, Fingerman and Bartell, 1967). Differential solubility analyses and gel filtration studies of chromatophorotropins from the eyestalks of the fiddler crab, Uca pni/ilator. indicate that each chromatophorotropin has either a pigment-dispersing or pigment-concen- trating action but not both (Rao, Fingerman and Bartell, 1967; Fingerman and Couch, 1967; Fingerman and Rao, 1969a, 1969b). The present investigation was undertaken (a) to determine whether the active material in the CXS of the horse- shoe crab has the same elution pattern from Bio-Gel P-6 and evokes the same chromatophoric responses in the fiddler crab as did that from the radial nerves of the sea star and (b) to compare some of the properties of chromatophorotropins from the CNS of the horseshoe crab and the eyestalks of the fiddler crab. MATERIALS AND METHODS The horseshoe crabs. Limn/us polyphemus, were collected in the vicinity of Woods Hole, Massachusetts, by members of the Supply Department of the Marine Biological Laboratory. \Ye wish to express our appreciation for their efforts. The fiddler crabs, Uca pugilator, were obtained from Panacea, Florida. The total 1 This investigation was supported by Grant GB-7505X from tbe National Science Foundation. 376 CHROMATOPHOROTROPIC SUBSTANCES 377 length of the horseshoe crabs used in this investigation \vas 105-125 mm; carapace width of the fiddler crabs was 16-19 mm. The extracts were assayed on the melanophores, erythrophores and leucophores of the fiddler crab. The melanin and red pigment of eyestalkless fiddler crabs was maximally concentrated ; the red pigment of intact fiddler crabs in black con- tainers is maximally dispersed ; and the white pigment of intact Panacea fiddler crabs in a black container is maximally concentrated. The white pigment of eye- stalkless Panacea fiddler crabs is found in all stages ranging from maximal concen- tration to maximal dispersion. Eyestalkless individuals whose white pigment was maximally dispersed were selected for use in these experiments. Therefore, by making use of intact and eyestalkless fiddler crabs assays could be made for melanin- dispersing, red pigment-dispersing, white pigment-dispersing, red pigment-concen- trating and white pigment-concentrating substances. The Hogben and Slome (1931) method was used to stage the chromatophores on the anteroventral surface of the second or third walking leg. According to this system stage 1 represents maximal concentration of the pigment, stage 5 maximal dispersion and stages 2, 3 and 4 the intermediate conditions. These chromatophore stages were used to calculate Standard Integrated Responses (SIR) to extracts as defined by Finger- man, Rao and Bartell (1967). The SIR is a measure of both the amplitude and duration of the response. Extracts of the tissues to be assayed, the CXS of the horseshoe crab and the eyestalks of the fiddler crab, were prepared in several different ways depending upon the aim of the experiment. The particular method used will be presented in the appropriate place below as the experiments are described. In those gel filtration experiments where distilled water was used to elute the active material from the column, each fraction was mixed with an appropriate volume of 400% physiological saline (Pantin, 1934) to provide an extract that was isosmotic with the blood of the fiddler crab. Otherwise 100% saline was used. The injected dose was 0.05 ml per crab. Control crabs received 100% saline alone. Each fraction obtained from the gel chromatography experiments was assayed on three crabs ; all other extracts were assayed on five crabs. EXPERIMENTS AND RESULTS Comparison of the chromatophorotropic activity of the ethanol-sohtble fraction, the 95% methanol: chloroform (2 \1) -soluble fraction and a saline extract of the central nervous system of the horseshoe crab The first objective of this series of experiments was to assay extracts of the horseshoe crab CNS prepared directly in saline on the melanophores and leuco- phores of a fiddler crab as did Brown and Cunningham (1941 ). but Uca pinjilator would be used instead of the species, Uca f>uc/na.r, they used. In addition the effect of such an extract on the erythrophores of a fiddler crab would also be deter- mined for the first time. Secondly, it has been demonstrated that the ethanol- M>luble fraction (Rao, Bartell and Fingerman, 1967) and the methanol: chloroform- soluble fraction (Bartell, Rao and Fingerman, 1971) of the eyestalks of the fiddler crab, Uca pin/ilator. prepared in concentrations higher than 0.3 eyestalk equivalent per dose evoke much >tnmger melanin-dispersing responses than do saline extract > 378 M. FINGERMAN, C K. I',ARTl-;i.I. AM) K. A. KK ASXOVY hut below this concentration tin- differences among the res])oiises to such extracts are slight. However, a high activity ethanol-soluhle fraction is not obtained when freeze-dried eyestalks are u.sed ( l\ao, Kartell and Fingerman, 1968). For com- parative purposes melanin-dispersing and white pigment-concentrating responses 80 - 60 a: - 40 c/) 20 B C D FIGURE 1. The Standard Integrated Responses (SIR) of eyestalkless fiddler crabs to ex- tracts of the central nervous system of the horseshoe crab; solid bars, melanin-dispersing re- sponses; empty bars, white pigment-concentrating responses. A, B, responses to extract pre- pared by triturating the tissue directly in saline; C, I), responses to the ethanol-soluble frac- tion; E, F, responses to the 95% methanol: chloroform (2:l)-soluble fraction. evoked by the ethanol-soluble and methanol : chloroform-soluble fractions of the horseshoe crab CNS would be compared with the responses to the saline extract to learn whether the chromatophorotropic material in the horseshoe crab CNS behaves in a similar manner to the melanin-dispersing material in the eyestalk of the fiddler crab. The extracts used in this particular set of experiments were prepared in a final concentration of 0.1 equivalent of the CNS of the horseshoe crab per dose CHROMATOFHOROTROPIC SUBSTANCES 379 of 0.05 ml. The first extract was made by triturating three freshly dissected CNS in 1.5 ml saline which provided the desired final concentration. Before injection the extract was centrifuged for three minutes at 1815 X oly- plictuus, were assayed for chromatophorotropic activity on the fiddler crab, Uca [>ii(/ilator. The extracts caused pigment dispersion in the melanophores and pig- ment concentration in the leucophores but had no effect on the erythrophores. 2. The ethanol-soluble and 95% methanol: chloroform (2: 1 (-soluble fractions of the central nervous system from the horseshoe crab evoked melanin-dispersing and white pigment-concentrating responses which had larger amplitudes and longer durations than did the responses caused by extracts prepared directly in saline. 3. Neither gel filtration nor acetone fractionation was effective in separating the melanin-dispersing activity from the white pigment-concentrating activity in extracts of the central nervous system of the horseshoe crab. These responses appear to be caused by either the same molecule or by different substances having similar elution patterns from Bio-Gel P-6. 4. Extracts of the eyestalk of the fiddler crab were fractionated on Bio-Gel P-6. The melanin-dispersing, white pigment-dispersing and red pigment-dispersing ac- CHROMATOPHOROTROPIC SUBSTANCES 387 tivities were eluted \villi an R f of 0.57, the same value as that of the major peaks of melanin-dispersing- and white pigment-concentrating activities from the central nervous system of the horseshoe crah. In contrast, the white pigment-concentrat- ing and red pigment-concentrating activities of the fiddler crah separated from the pigment-dispersing activities, having been eluted later from the column of Bio-Gel P-6 with an R f of 0.28. 5. The chromatophorotropic material in the central nervous system of the horseshoe crah is more closely related to the melanin-dispersing material of the fiddler crah than to the white pigment-concentrating hormone of the fiddler crah. 6. The chromatophorotropins in the central nervous system of the horseshoe crab and the radial nerves of the sea star differ from each other in their elution patterns from a column of Bio-Gel P-6 and in the chromatophore responses they evoke in the fiddler crah. LITERATURE CITED ABRAMOWITZ, A. A., AXD R. K. ABRAMOWITZ, 1938. On the specificity and related properties of the crustacean chromatophorotropic hormone. Biol. Bull., 74: 278-296. BAGNARA, J. T., 1964. Stimulation of melanophores and guanophores by melanophore-stimulat- ing hormone peptides. Gen. Conip. Endocrinol., 4: 290-294. BARTELL, C. K., K. R. RAO AXD M. FIXGERMAN, 1971. Comparison of the melanin-dispersing fractions in extracts prepared initially in ethanol, saline, or distilled water from eye- stalks of the fiddler crab, Uca pni/ilator. Camp. Biachcm. Physial.. 38A: 17-36. BROWX, F. A., JR., 1950. Studies on the physiology of Uca red chromatophores. Biol. Bull., 98: 218-226. BROWX, F. A., JR., AXD O. CUNNINGHAM, 1941. Upon the presence and distribution of a chromatophorotropic principle in the central nervous system of Liiinilns. Biol. Bull., 81 : 80-95. CARLSON, S. P., 1935. The color changes in Uca piiuilafor. Proc. Nat. Acad. Sci.. 21: 549- 551. FINGERMAX, M., AND E. F. COUCH, 1967. Differentiation of chromatophorotropins from the prawn, Palaemonetes -cuhiaris. and the fiddler crab, Uca fnijilator. /. E.vp. Zool., 165: 183-194. FIXGERMAX, M., AXD K. R. RAO, 1969a. A comparative study of leucophore-activating sub- stances from the eyestalks of two crustaceans, Palaemonetes rnli/aris, and Uca pitt/ila- tor. Biol. Bull.. 136: 200-215. FINGERMAX, M., AXD K. R. RAO, 1969b. Physiology of the brown-black chromatophores of the stomatopod crustacean Squilla einpusa. Pliysiol. Zoo!., 42: 138-147. FINGERMAN, M., K. R. RAO AND C. K. BARTELL, 1967. A proposed uniform method of report- ing response values for crustacean chromatophorotropins : the Standard Integrated Response. Experientia, 23: 962. HOGBEX, L., AXD D. SLOME, 1931. The pigmentary effector system. VI. The dual character of endocrine co-ordination in amphibian colour change. Proc. Ro\. Soc. London, Scries B, 108: 10-53. KLEIXHOLZ, L. H., 1970. A progress report on the separation and purification of crustacean neurosecretory pigmentary-effector hormones. Gen. Comp. Endocrinol.. 14: 578-588. PANTIX, C. F. A., The excitation of crustacean muscle. /. E.r/>. Biol.. 11: 11-27. RAO, K. R., AXD M. FINGERMAX, 1970a. Chromatophorotropic activity of extracts of radial nerves from the starfish, Asterias amurensis. Z. Vcrgl. Physiol.. 67: 133-139. RAO, K. R., AND M. FIXGEKMAX, 1970b. Action of biogenic amines on crustacean chromato- phores: I. Differential effect of certain indolealkylamines on the melanophores of the crabs Uca pinjihitor and Carcinns niaenas. Experientia, 26: 383-384. 388 M. FINGERMAN, C. K. BARTELL AND R. A. KRASNOW RAO, K. R., AND M. FINGERMAN, 1970c. Action of hiogenic amines on crustacean chromato- phores, II. Analysis of the response of erythrophores in the fiddler crab, Uca pit<]ilator, to indolealkylamines and an eyestalk hormone. Comp. Gen. Phaniiacol., 1: 117-126. RAO, K. R., C. K. BARTELL AND M. FINGERMAN, 1967. Relationship between the response of melanophores in the fiddler crab, Uca pugilator, and the concentration of eyestalk extract. Z. Vcrgl PhysioL. 56: 232-236. RAO, K. R., C. K. BARTELL AND M. FINGERMAN, 1968. Solubility and stability properties of the melanin-dispersing substances in the eyestalks of the fiddler crab, Uca pugilator. Z. Vergl. PhysioL, 60: 1-13. RAO, K. R., M. FINGERMAN AND C. K. BARTELL, 1967. Physiology of the white chromato- phores in the fiddler crab, Uca puro[>ini]ttus Girard ( 1852), and Proeoinharns eeonfinae Hobbs (1942). 398 M. H. GOODWIN AM) M. TELFORD 2. The toxin appears to be composed of a proteinaceous, undialyzable fraction, and a dialyzable fraction containing two aromatic amines. 3. The amines are probably fused or polysubstituted ring compounds, one of which is carboxylated. 4. The crude toxin has varying potency, depending upon the species of assay animal used. 5. With the possible exception of histamine, none of the compounds previously suggested as possible components of the nematocyst toxin were found. LITERATURE CITED ACKERMAN, P. G., AND G. TOKO, 1963. Blood lipids. Pages 249-272 in S. Frankel, S. Reit- man, and A. C. Sonnenwirth, Eds., Clinical Laboratory Methods and Diagnosis. C. V. Mosby Co., St. Louis. BARNES, J. H., 1966. Studies on three venomous Cubomedusae. Pages 307-332 in W. J. Ross, Ed., The Cnidaria and tlieir Evolution. Academic Press, Aberdeen. BLANQUET, R., 1968. Properties and composition of the nematocyst toxin of the sea anemone Aiptasia pallida. Cninp. Biochcm. Physiol.. 25: 893-902. BLISS, C. I., 1952. 7V; r Statistics of Bioassay. Academic Press, New York, 183 pp. BREGOFF, H. M., E. ROBERTS AND C. C. DEL\VICHE, 1953. Paper chromatography of quater- nary ammonium bases and related compounds. /. Biol. Clicin., 205: 565-574. DAWSON, R. M. C., D. C. ELLIOT, W. H. ELLIOT AND K. M. JONES, 1959. Data for Biochemi- cal Research. Clarendon Press, Oxford, 299 pp. DIMITROV, G. D., AND R. C. KANKONKAR, 1968. Fractionation of npera ntselli venom by gel filtration. II. Comparative study of yellow and white venoms of Vipcra russelli with special reference to the local necrotizing and lethal actions. To.ncon. 5: 283-288. HAND, C., 1954. The sea anemones of central California. U'asmann J. Biol., 12: 3-251. HASTINGS, S. G., J. B. LARSEN AND C. E. LANE, 1967. Effects of nematocyst toxin of Physulia pliysalis on the canine cardiovascular system. Proc. Soc. E.rp. Biol. Med., 125: 41-45. KLINE, E. S., AND V. S. WARAVDEKAR, 1960. Inhibitor of succinoxidase activity from Hvdnt littoralis. J. Biol. Chan., 235 : 1803-1808. LANE, C. E., 1961. Physulia nematocysts and their toxin. Pages 169-178 in H. M. Lenhoff and W. F. Loomis, Eds., The Biology of Hydra. University of Miami Press, Coral Gables, Florida. LANE, C. E., AND E. DODGE, 1958. The toxicity of Phy satin nematocysts. Biol. Bull., 115: 219-226. LENTZ, T. L., AND J. G. WOOD, 1964. Amines in the nervous system of coelenterates. /. Histochem. Cytochcin.. 12: 37. LOWRY, O. H., N. J. ROSEBROUGH, A. L. PARK AND R. RANDALL, 1951. Protein measurement with the Folin phenol reagent. /. Biol. Chein., 193 : 265-275. MARTIN, E. J., 1966. The macromolecular toxin of sea anemones. Pages 136-138 in R. I. Bowman, Ed., Proceedings of the Symposia of the Galapagos International Scientific Project, 7-18 Jan., 1 ( >64, Berkeley. University of California Press, Berkeley. MATHIAS, A. P., D. M. Ross AND M. SCHACTER, 1960. The distribution of 5-hydroxytrypta- mine, tetramethylammonium, homarine, and other substances in sea anemones. /. Physio!., 151: 296-331. MEITES, S., AND W. R. FAULKNER, 1962. Manual of Practical Micro and General Procedures in Clinical Chemistry. Charles C Thomas, Springfield. PHILLIPS, J. H., 1956. Isolation of active nematocytes of Metridnim senile and their chemical composition. Nature, 178: 932. PHILLIPS, J. H., AND D. P. ABBOTT, 1958. Isolation and assay of the nematocyst toxin of Mctridiiim senile fiinhriatnm. Biol. Bull., 113: 296-301. PICKEN, L. E. R., AND R. J. SKAER, 1966. A review of researches on nematocysts. Pages 19-50 in W. J. Ross, Ed., The Cnidaria and their Ei'olution. Academic Press, Aber- deen. XEMATOCYST TOXIX OF METRIDIl'M 399 RICHET, C., AND P. PORTIEK, 1936. Rccherches sur la toxine des coelenteres et las phenomenes d'anaphylaxie. Resultats dcs campaniles scientifiques Monaco. 95: 3-24. SNEDECOR, G. W., AND W. G. COCHRAN, 1967. Statistical Methods. Iowa State University Press, Ames, Iowa, 593 pp. UDENFRIEND, S., C. T. CLARK AND A. XEISSBACH, 1955. The estimation of 5-hydroxytrypta- mine in biological tissues. /. Biol. Chcin., 215: 337-344. WELSH, I. H., 1956. On the nature and action of coelenterate toxins. />tv/> Sea Res. $"/> />.. 3: 287-297. WELSH, J. H., 1960. 5-hydroxytryptamine in coelenterates. Nature. 186: 811-812. WELSH, J. H., 1961. Compounds of pharmacological interest in coelenterates. Pages 179-186 in H. M. Lenhoff and W. F. Loomis, Eds., The Biolouy of Hydra. University of Miami Press, Coral Gables, Florida. WELSH, J. H., AND P. B. BROCK, 1958. Quaternary ammonium bases in the coelenterates. Biol. Bull.. 115: 551-561. WEST, E. S., AND W. R. TODD, 1962. Textbook of Biochemistry. Macmillan Company, New York, 1356 pp. YANAGITA, T. M., 1959. Physiological mechanism of nematocyst responses in sea anemone. /. /:./>. Hiol.,36: 478-494. Reference: ttil. />'////.. 140: 400-415. (June, 1'Vl ) TEMPORAL PATTERNS OF SEXUAL AND ASEXUAL REPRODUCTION IN THE COLONIAL ASCIDIAN METANDROCARPA TAYLORI HUNTSMAN 1 NORINE D. HAVEN - Hopkins Marine Station of Stanford University, California Metandrocarpa taylori is a colonial ascidian common on rocks in the low inter- tidal zone along California shores. Reddish-orange colonies spread out in flat sheets, often covering large expanses of granite rocks, particularly on their lower surfaces. Individuals, which resemble small simple ascidians, are joined to one another basally by a thin layer of test containing blood vessels. Previous studies on M. taylori have been concerned largely with hud develop- ment (Abbott, 1953), analysis of the factors involved in hud initiation (Newberry, 1965), and larval structure and behaviour (Abbott, 1955). The major objective of the present work is the comparison of sexual and asexual reproductive activities in M. taylori, including consideration of how these two processes interact in colonies throughout the year. There is also a brief considera- tion of the seasonal aspects of reproduction in relation to water temperature and phytoplankton. A study of the gonads was also necessary in order to determine criteria for sexual reproductive activity. Except for the observation of Abbott (1955) that larvae settled in the summer months, nothing has been recorded as to the extent of either the sexual or asexual reproductive cycle. Likewise, there is no adequate description of the gonads in the literature; Huntsman (1912) and Ritter and Forsyth (1917) described populations of M. taylori that differed somewhat from each other in the number and arrangement of the gonads. MATERIALS AND METHODS Specimens of Metandrocarpa ta\lori used in the study of sexual reproduction were collected at Pescadero Point, Monterey County, California (3631'47"N.) at approximately two week intervals from June 1, 1961 to June 2, 1962. Colonies were sampled by using a chisel to remove clusters containing about a dozen indi- viduals each. Approximately ten different colonies were sampled on each collect- ing date. The specimens were relaxed for 1224 hours in a 7% MgSO 4 -seawater solution, then fixed in Bourn's fluid and stored in 70% EtOH. Adults in each cluster, connected basally by a common tunic, are readily separated from one an- other intact. Gonads are easily exposed by removing the upper half of the zooid by a horizontal cut, then carefully lifting out the remaining portion of the branchial basket along with the stomach and intestine. Since M. taylori usually occurs on irregular surfaces in situations where growth is difficult to monitor, colonies used in the study of asexual reproduction were estab- lished on glazed tiles and maintained in racks attached in the intertidal zone adja- 1 This investigation was supported in part by Public Health Service Fellowship No. GPM- 12, 688-C1 from the Division of General Medical Sciences, and in part by Public Health Service Training Grant GM-647 from the National Institute of General Medical Sciences. - Present Address : Simon Fraser University, Burnaby 2, British Columbia, Canada. 400 REPRODUCTION IN METANDROCARPA 401 cent to Hopkins Marine Station. Three clusters of adult zooicls from separate colonies were collected each month at Pescadero Point, cleaned of debris, and tied with thread to the center of six-inch square glazed ceramic tiles. Tiles were kept in running seawater for a month in the lab. to allow the zooids to attach and begin to bud. The tiles were then placed in slotted redwood boxes bolted to intertidal rocks protected from heavy surf approximately 0.2 feet above mean lower low water. Once a month, all of the tiles bearing cultures were transferred to laboratory aquaria, observed, photographed with a Polaroid camera, and returned to the sea boxes within 24 hours. Individual colonies were maintained until they grew to the edges of a tile. While there \vas some damage and loss of colonies due to storms, most colonies expanded to cover one side of the tile by the end of four months in the intertidal zone. The gonads of twenty adult zooids collected biweekly from different colonies were measured to provide data on sexual reproduction. Female gonads were as- signed to the following size classes: small (ovaries less than 0.1 mm in diameter) ; medium (ovaries 0.1 mm or more in diameter, eggs still encased in ovarian epi- thelium) ; and developing (mature eggs and various stages of embryos, both lying free in the mantle cavity ) . Size classes for testes were small ( less than 0.2 mm in greatest diameter) ; medium (0.2-0.4 mm), and large (over 0.4 mm ). RESULTS Gonads In Metandrocarpa tavlori the separate testes and ovaries are partially embedded in the ventral floor of the mantle cavity. The ovaries extend in a horseshoe shape, back and outward on both sides of the medial endostylar ridge. The testes extend posteriorly from these on each side, in a line on the right but bunched on the left due to presence of the stomach. The gonads are illustrated in Haven (1965). The average number of ovaries was seven (range 0-13) and the average number of testes was nine (range 0-17). The number and the arrangement of gonads in Metandrocarpa tavlori are in general accord with the reproductive system described by Ritter and Forsyth (1917) for M. inicliaclscni, a species placed in synonomy with M. taylori by Abbott ( 1953) . In animals not actively engaged in sexual reproduction the ovaries are small (less than 0.1 mm in diameter) and contain developing eggs in early stages. Three or four enlarging eggs lie in each ovary. As the eggs enlarge and become yolkv. they stretch the overlying epithelium. Xo more than one or two eggs are fully mature at the same time in an individual ovary and it appears that usually only one egg is released at a time from an ovary. When the eggs reach a diameter of approximately 0.3 mm they are released from the ovary, apparently by rupture of the gonad epithelium since an oviduct is lacking, and come to lie free in the mantle cavity. Fertilization probably occurs at this time. The larvae are usually retained in the adult until they are capable of swimming away to settle (Abbott. 1955). Small testes (less than 0.2 mm in diameter) have a rounded base and a vas deferens protrudes from the center of each. Sections of testes of this size show 402 NOR1NK D. HAVEN densely packed cells with no lumen evident. Testes of intermediate sizes (0.2 mm- 0.4 nun in diameter) are pear-shaped. Smears of these testes show many nurse cells with attached spermatids, together with immature sperm. With further enlargement (over 0.4 mm diameter) the testes develop tabulations, and smears contain a good proportion of highly motile sperm. Collections frequently con- tained a number of adult zooids with no visible testes, suggesting the possibility of complete regression of the testes during periods of inactivity. Sexual reproduction in the population The annual pattern of sexual reproductive activity, based on the most advanced stages of gonads present in individuals, is shown in Figure 1. Sexual reproduc- IOCH I'/ I medium ova but no embryos or larvae present developing embryos and larvae present 13 26 II 26 II 24 8 2! 9 Aug Aug Sep Sep Oct Oct Nov Nov Dec I 3 19 2 17 3 16 I 18 29 19 2 Jan Jan Feb Feb Mar Mar Apr Apr Apr May Jun 1962 100 J HI medium testes but no large testes present large testes FIGURE 1. Sexual reproductive activity of Metandrocarpa tuylori in the collections of 1961-1962. Zooids are classed according to the most advanced condition of reproductive activ- ity of the male and female systems present. Dates of collection are shown on the horizontal scale. tion continues throughout the year. At least 10% of the population always showed some developing embryos and larvae, and these stages w r ere present in a maximum of 70% of the sample in the collection of July 27, 1961. In addition, medium sized eggs were present in other individuals from 5%> to 50% of the time. At no time was less than 30%> of the population engaged in sexual reproduction, as indicated bv the presence of medium sized eggs and/or developing embryos and larvae. Summer was the period of greatest sexual activity. In the three month period from June 12. 1961 to September 11. 1961, an average of 79% of the population contained medium sized eggs or developing larvae or both and the number of indi- viduals containing developing embryos and larvae was 30/ higher than in the remainder of the year. REPRODUCTION IN METANDROCARPA 403 The lower half of Figure 1 shows the condition of the testes in each collection. During the summer, when embryo production was increased, there was no corre- sponding increase in the proportion of individuals with large testes. but increased summer activity is indicated when the classes with medium sized and large sized testes are considered together. Throughout the year a rise or fall in the incidence of medium and large testes usually corresponded to a similar change in the incidence of medium sized and developing eggs ; the correlation coefficient for this was highly significant ( r = 0.784 ; P < 0.0005 ). Collections of M . taylori taken by Dr. Donald P. Abbott of Stanford Univer- sity from April 1949 to March 1950 were similarly analysed and also showed greater sexual reproduction during the summer months and early fall. TABLE I Summary of egg and testes stages present simultaneously in individual zoo ids Individual Zooids with: Collections of 1949-50 period (X = 440) Collection? of 1961-62 period (N = 540) No ovary visible Small eggs only Small + medium sized eggs Small, medium sized and developing eggs Medium sized -f developing eggs Developing eggs only Small + developing eggs Medium sized eggs only Total No testes visible Small testes only Small + medium testes Medium testes only Medium + large testes Large testes only Total 13% 43 11 9 15 4 3 2 100% 10% 22 10 30 11 17 100% 1% 45 20