Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Regulation of spawning behaviour in the female goldfish, Carassius Auratus Stacey, Norman Edward 1977

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


831-UBC_1977_A1 S83.pdf [ 9.05MB ]
JSON: 831-1.0094202.json
JSON-LD: 831-1.0094202-ld.json
RDF/XML (Pretty): 831-1.0094202-rdf.xml
RDF/JSON: 831-1.0094202-rdf.json
Turtle: 831-1.0094202-turtle.txt
N-Triples: 831-1.0094202-rdf-ntriples.txt
Original Record: 831-1.0094202-source.json
Full Text

Full Text

THE REGULATION OF SPAWNING BEHAVIOUR IN THE FEMALE GOLDFISH, CARASSIUS AURATUS by NORMAN EDWARD STACEY B.Sc., University of Brit i s h Columbia, 1970 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY We accept this thesis as conforming to the required standard The University of British Columbia i n the Department of Zoology 1977 Norman Edward Stacey, 1977 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of Brit ish Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the Head of my Department or by his representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of The U n i v e r s i t y of B r i t i s h Columbia 2075 Wesbrook'Place Vancouver, Canada V6T 1W5 Date 2 7 / f ABSTRACT This study of the regulation of spawning behaviour in the female goldfish (Carassius auratus) identified four endogenous factors believed to play major roles i n spawning behaviour: (i) stimuli from ovulated eggs, (i i ) ovarian steroids, ( i i i ) pituitary hormones, and (iv) prostaglandins. Spawning behaviour is synchronized with ovulation by the stimulus of an intraovarian mass of ovulated eggs. Normally, female goldfish begin to spawn on the morning of the day of ovulation and perform as many as several hundred spawning acts over a period of several hours; spawning behaviour ceases when a l l ovulated eggs have been shed. The duration of spawning be-haviour was extended i f oviposition (the release of eggs through the ovipore) was prevented by placing a plug in the ovipore. Spawning behaviour was ter-minated when ovulated eggs were removed by hand-stripping and restored when ovulated eggs were injected through the ovipore and into the ovarian lumen. This effect of eggs on spawning behaviour was not restricted to the day of ovulation but was seen in a l l fish with ovaries in any stage of vitellogene-si s . Injection of several substitutes for ovulated eggs induced low levels of spawning behaviour. Injection of ovulated eggs failed to induce spawning behaviour in fe-male goldfish with regressed, nonvitellogenic ovaries. Pretreatment with a variety of gonadal steroids restored the spawning response to egg injection in these intact, regressed fi s h . Hypophysectomized fish did not perform spawning behaviour when injected with ovulated eggs. Pretreatment of hypophysectomized fish with homogenized goldfish pituitaries or partially purified salmon gonadotropin (SG-G100) restored the response to egg injection. Aminoglutethimide, an inhibitor of steroid synthesis, blocked the effect of SG-G100 on spawning behaviour, i i i suggesting gonadotropin may exert i t s e f f e c t on behaviour by stimulating steroidogenesis. However, s t e r o i d treatments were t o t a l l y i n e f f e c t i v e i n res t o r i n g the response to egg i n j e c t i o n i n hypophysectomized f i s h . Prostaglandin (PG) appears to be involved i n mediating the behavioural response to ovulated eggs. Indomethacin (IM), an i n h i b i t o r of PG synthesis, blocked the onset of spawning behaviour following egg i n j e c t i o n . I n j e c t i o n of PGF„ restored spawning behaviour i n egg-injected, IM-treated f i s h ; PGE was l e s s e f f e c t i v e and PGE^ was without e f f e c t . I n j e c t i o n of PGF^.induced normal spawning behaviour i n f i s h which had not been i n j e c t e d with ovulated eggs, suggesting that ovulated eggs induce spawning behaviour by stimulating synthesis of PG. The e f f e c t s of PG on spawning behaviour of hypophysectomized f i s h treated with SG^GIOO or steroids p a r a l l e l e d the e f f e c t s of egg i n j e c -t i o n on f i s h r e c e i v i n g s i m i l a r treatments; SG-G100 restored the spawning re-sponse to i n j e c t i o n of PG, while s t e r o i d treatments were without e f f e c t . Mechanisms by which ovarian and p i t u i t a r y hormones and prostaglandins may influence spawning behaviour are discussed and a model of the regulation of spawning behaviour i s proposed. In addition, an attempt i s made to pro-vide a t h e o r e t i c a l basis for comparing the regulation of sexual behaviour i n female vertebrates. iv TABLE OF CONTENTS Page CHAPTER I - GENERAL INTRODUCTION 1 CHAPTER II - MATERIALS AND METHODS 4 A. General Maintenance 5 B. Hypophysectomy 6 C. Histology of regressed ovaries 9 D. Behavioural testing procedures 13 E. Description of spawning behaviour 16 CHAPTER III - THE ROLE OF OVULATED EGGS IN THE SPAWNING BEHAVIOUR OF FEMALE GOLDFISH 21 A. Experiment 1. Effect of Removal and Replacement of Ovulated Eggs 22 1. Introduction 22 2. Materials and Methods 22 3. Results and Discussion 25 B. Experiment 2. Induction of Spawning Behaviour by Injection of Egg Substitutes 29 1. Introduction 29 2. Methods, Results and Discussion 30 C. Summary of Chapter III 34 CHAPTER IV - THE EFFECT OF STEROIDS ON SPAWNING BEHAVIOUR OF FEMALE GOLDFISH 35 A. Experiment 3. Steroid-Induced Spawning Behaviour in Intact, Regressed Female Goldfish 36 1. Introduction 36 2. Materials and Methods 38 3. Results 39 4. Discussion 44 B. Experiment 4. Ineffectiveness of Steroids in Hypophysectomized Female Goldfish 46 1. Introduction 46 2. Materials and Methods 46 3. Results 47 4. Discussion 47 C. Experiment 5. Ineffectiveness of Long-Term Steroid Treatment in Hypophysectomized Fish 48 1. Introduction 48 2. Materials and Methods 48 3. Results and Discussion 49 D. Summary of Chapter IV 5 0 V. Table of Contents (cont'd) CHAPTER V - ROLE OF THE PITUITARY IN THE SPAWNING . BEHAVIOUR OF FEMALE GOLDFISH 5 1 A. Experiment 6. Effect of Hypophysectomy and Pituitary Replacement on Spawning Behaviour 52 1. Introduction 52 2. Materials and Methods 52 3. Results 53 4. Discussion 56 B. Experiment 7. Effect of Salmon Gonadotropin and Aminoglutethimide on Spawning of Hypophy-sectomized Female Goldfish ...... 57 1. Introduction 57 2. Materials and Methods 58 3. Results ...... 61 4. Discussion 66 C. Summary of Chapter V 67 CHAPTER VI. THE ROLE OF "PROSTAGLANDINS IN SPAWNING BEHAVIOUR OF FEMALE GOLDFISH 68 A. Introduction 69 B. Experiment 8. Inhibition by Indomethacin of Spawning Induced by Ovulated Eggs 74 1. Introduction 74 2. Materials and Methods ...... 74 3. Results and Discussion 77 C. Experiment 9. Effect of Prostaglandins on Spawning of Indomethacin-Treated Fish 78 1. Introduction 78 2. Materials and Methods 78 3. Results and Discussion 78 'D. Experiment 10. Effect of Prostaglandins on Spawning Behaviour 82 1. Introduction 82 2. Methods, Results and Discussion 82 E. Experiment 11. Effect of Steroids and Salmon Gonadotropin on Prostaglandin-Induced Spawning Behaviour in Hypophysectomized Fish 86 1. Introduction 86 2. Materials and Methods 87 3. Results 87 4. Discussion 88 F. Summary of Chapter VI 92 CHAPTER VII - GENERAL DISCUSSION ' 93 A. Introduction 94 B. The Role of the Pituitary 95 C. The role of Steroids 99 1. Introduction 99 2. Effects of steroids i n intact female goldfish 100 Table of Contents (cont'd) 3. Possible mechanisms of steroid action D. The Role of Ovulated Eggs E. The Role of Prostaglandins F. The Regulation of Spawning Behaviour G. A Comparative Approach to the Study of Female Sexual Behaviour BIBLIOGRAPHY APPENDIX - Some Aspects of the Histology of Regressed Ovaries - Appendix Figures v i i . LIST OF TABLES Table P a g e I Effect of removal and replacement of ovulated eggs on the spawning behaviour of ovulated female goldfish 27 II Effect of injection of egg substitutes on spawning behaviour of female goldfish 32 III Effect of steroids bn spawning behaviour of intact, regressed female goldfish 42 IV Effect of injection of goldfish pituitary homogen-ate on the spawning behaviour of hypophysectomized female goldfish. 55 V Vehicle volumes injected in Experiment 7 60 VI Effect of salmon gonadotropin (SG-G100) and aminoglutethimide on the spawning behaviour of hypophysectomized female goldfish 64 VII Effect of incubation time on latency to the f i r s t spawning act in recently ovulated female goldfish... 71 VIII Effect of indomethacin on spawning in response to injection of ovulated eggs 76 IX Effect of prostaglandins on the spawning response to injection of ovulated eggs in female goldfish treated with indomethacin 80 X Effect of prostaglandins on the spawning behaviour of female goldfish without eggs in the ovarian lumen 4^ XI Effect of salmon gonadotropin (SG-GL00) and steroids on prostaglandin-induced spawning behaviour i n hypophysectomized female goldfish 90 v i i i . LIST OF FIGURES Figure Page 1. Proposed model of the regulation of spawning behaviour in the female goldfish 118 2. Early stage of degenerating early yolk vesicle or previ-tellogenic oocyte 148 3. Follicular hypertrophy in degenerating oocyte 148 4. Advanced degenerating early yolk vesicle oocyte 148 5. Advanced degenerating early yolk vesicle oocyte 150 6. Advanced degenerating previtellogenie oocyte from hypo-physectomized fish 150 7. Yolk nucleus in oocyte of hypophysectomized fish 150 8. Nuclear migration stage i n degenerating oocyte 150 9. Yolk vesicles in normal oocyte 152 10. Yolk vesicles in normal oocyte 152 11. Yolk vesicles associated with membrane of presumptive pre-atretic oocyte 152 12. Yolk vesicles associated with membrane of presumptive pre-atretic oocyte 152 13. Yolk vesicles and presumptive nucleoli in hypertrophied f o l l i c u l a r layer from hypophysectomized fish 154 14. Pre-nuclear breakdown stage oocytes in ovary of hypo-physectomized fish 154 15. Degenerating nucleus at early stage of atresia 154 16. Nucleus with spherical body at periphery of degenerat-ing oocyte ...... 15^ 17. Degenerating early yolk vesicle oocyte from hypophysec-tomized fish 156 18. Advanced degenerating early yolk vesicle oocyte with nucleus and spherical body ....... 156 ix. L i s t of Figures (cont'd) Figure Page 19. Nucleoli and spherical body in cytoplasm of advanced degenerating previtellogenic oocyte 156 20. Clumped nucleoli in atretic oocyte from hypophysec-tomized fish 156 21. M-l macrophage and presumptive nucleoli i n atretic oocyte of hypophysectomized fi s h 158 22. M-2 macrophage in atretic oocyte of intact, regressed fish 158 23. Macrophage aggregation adjacent to early atretic 158 oocyte 24. Macrophages entering early atretic oocyte of intact regressed fish 158 25. Advanced degenerating previtellogenic oocytes i n ovary of hypophysectomized fish 160 26. Spaces in ovary of hypophysectomized fish apparently caused by atresia of previtellogenic oocytes 160 27. Spaces in ovary of hypophysectomized fish apparently caused by atresia of previtellogenic oocytes 160 28. Injected ovulated eggs in ovary of hypophysectomized fish 162 29. Macrophage aggregation adjacent to injected ovulated oocyte 162 X. ACKNOWLEDGEMENTS I am very grateful to Dr. N.R. Liley, my supervisor, for suggesting the problem, for guidance and support during the investigation, and for encouragement, constructive criticism, and patience during the preparation of the manuscript. I would also like to thank my research committee, Dr. W.S. Hoar, Dr. A.M. Perks, arid Dr; E.M. Donaldson, for their helpful comments and discussion. I am grateful to Dr. D.L. Kramer and Dr. K.H. Khoo for their suggest-ions and enthusiasm, to Dr. E.M. Donaldson for the generous g i f t of salmon gonadotropin, to Dr. J. Pike (Upjohn Company) for his g i f t of prostaglandins, and especially to Ms. M.E. Hurlburt. This study was supported by a National Research Council operating grant to Dr. N.R. Liley and by a University of British Columbia graduate fellowship to myself. 1 CHAPTER I GENERAL INTRODUCTION 2 GENERAL INTRODUCTION Research into many aspects of vertebrate reproduction has expanded 'explosively' in recent years. However, considering the numbers of ethol-ogists, animal psychologists, neurophysiologists, and biochemists whose c r i t i c a l attentions now focus on the problems of hormonal control of female reproductive behaviour in many species of mammals , the research effort concerned with these phenomena in lower vertebrates is at best modest. Undoubtedly, this inequitable distribution of s c i e n t i f i c labor stems from differences in the degree of relevance to problems of medicine and f e r t i l i t y in man and domestic animals. As greater emphasis is placed on the culture of edible fish species, greater emphasis also w i l l be placed on the study of reproductive physiology and i t s relation to sexual behaviour in fish. Remarkably l i t t l e information concerning the hormonal regulation of sexual behaviour in female fish i s available (Hoar, 1965; Liley, 1969) and much of this pertains to a reproductively specialized teleost, the viviparous guppy (Pdecllla reticulata). In this species, the correlation of receptivity with gonadal state and the results of hypophysectomy, ovariectomy, and replacement therapies suggest that, as in higher verte-brates, ovarian and gonadotropic hormones participate in the regulation of sexual behaviour (Liley, 1968, 1969, 1972; Liley and Donaldson, 1969; Liley and Wishlow, 1974). Apart from an early study (Noble and Kumpf, 1936; discussed in Liley, 1969) which reported partial restoration of sexual behaviour following treatment of female Hemichromis bimaculatus with an un-identified ovarian extract, replacement therapies (steroid hormones or ovarian extracts) have been ineffective in restoring spawning behaviour 3 of ovariectomized oviparous fish (Liley, 1969). Though the results of many studies acknowledge the 'close temporal relationship between ovul-ation and the spawning act' in female fish (Liley, 1969), the possible significance of this sequence in the regulation of spawning behaviour . seems generally to have been overlooked. However, the observation that removal of ovulated eggs terminated the spawning behaviour of female goldfish led Yamazaki (1965) to hypothesize that 'ripe eggs i n the ovarian lumen stimulate the spawning behaviour of females via some pathway'. Yamazaki's results further suggested that a role for ovulated eggs in spawning behaviour could explain why the technique of ovariectomy and steroid replacement therapy had yielded negative results in earlier studies of egg-laying fish (Liley, 1969) . In this thesis I have tested and found much support for Yamazaki's hypothesis concerning the role of ovulated eggs in spawning behaviour. In subsequent experiments, evidence i s presented which indicates that pituitary and ovarian hormones influence the tendency of female goldfish to perform spawning behaviour when ovulated eggs are i n the ovarian lumen. The mechanism by which eggs evoke spawning behaviour has been examined and i t i s suggested that protaglandins are involved. Emphasis has been placed on correlating sexual receptivity with ovarian histology in both treated and untreated fish. In summarizing the results of this study, comparative aspects of hormonal control of reproductive behaviour are emphasized and a basis for comparing reproductive behaviour in female vertebrates i s proposed. CHAPTER II MATERIALS AND METHODS General Maintenance A l l fish used in this study were purchased from Hartz Mountain Pet Supplies Limited (Richmond, B.C.). To ensure an adequate supply of female goldfish, most shipments of fish were sexed at the ware-house; the remainder were shipped unsexed to the lab at the University of British Columbia. Common, comet, and intermediate types were used. Colors ranged from deep orange to silver and included many variations in mottled patterns, black tipped and white tipped fins, etc. Individuals with the wild, olive coloration were not used. The fish weighed from 15 to 60 g but the majority were in the range of 20 to 30 g. To 'regulate' the reproductive state of both males and females, fish were maintained under two environmental regimes; cold water, long photoperiod (1.2 C; 16L:8D) and warm water, long photoperiod (20 C; 16L:8D). As reported by Yamamoto et al (1966), goldfish kept at 14 C or less develop to a prespawning stage but neither ovulate nor spermiate. When raised to 20 C, fish in this condition ovulate or spermiate within several days. Usually, fish newly acquired from the warehouse (where they had been kept in running, dechlorinated, cold tapwater) were placed in large stock tanks of various sizes supplied with running dechlorinated tapwater thermostatically regulated at 10-13 C and illuminated for 16 hours per day withh20 - 40 watt fluorescent lights. In some cases, cold water stocks were kept outdoors in six-feet diameter fibreglass tanks under natural photoperiod but with some continuous low-level illumination. These tanks were not heated and sometimes were as cold as 5 C. To avoid introduction of disease, new shipments of cold water stock fish were kept separate from established stocks for at least several months. Cold water stock tanks contained no vegetation 6 or loose substrate. In some cases, cold water stocks were separated by sex. Warm water stocks were kept 3 - 6 per 40 1 tank in a fish room with air temperature control (water temperature 20 - 22 C) on a 16L:8D photoperiod. Each tank was supplied with 1" - 2" of quartz sand as bottom substrate, a floating mat of water sprite (Ceratopterus thalictroides) , and an air stone. F i l t r a t i o n was used at the outset but quickly abandoned: outside box f i l t e r s because of the maintenance involved, subgravel f i l t e r s because of the acid water conditions produced when used i n conjunction with Ceratopterus. A rapid f a l l in pH was by far the greatest cause of mortality i n the warm water stock fish. In fish exposed to this condition, skin (and presumably g i l l ) mucous apparently i s denatured and becomes opaque; moribund individuals transferred to neutral water nearly always re-covered within a few hours. Fortunately, an easily observable i n -dicator, an increase in water cl a r i t y , always preceded the development of c r i t i c a l l y low pH by at least a day. Entrapment of much suspended detritus i n the fine roots of Ceratopterus aids greatly in maintenance of water cl a r i t y , although this was l i k e l y of more importance to the experimenter than to the subjects. A l l stocks were fed frozen brine shrimp ad libitum at least every other day. In addition, cold water stocks were fed Clark's New Age Trout Feed (Moore-Clark Company, Salt Lake City, Utah) on an irregular basis. B. Hypophysectomy Hypophysectomized fish were used in a number of experiments re-ported in this thesis. Mortalities directly due to the operation were low; however, disease and a relatively high incidence of incomplete 7 removal of the pituitary usually resulted in less than half the operated fish being used for experiments. Hypophysectomy was performed as demonstrated by Yamazaki (personal communication). Fish which had been kept at 20 C for at least several days were chilled for 30 to 60 minutes in ice water to which had been added a small amount of tricaine methanesulphonate (MS-222, 0.01 %) . Fish used directly from cold water stock tanks often failed to remain anaesthetized for the duration of the operation; addition of more MS-222 in these cases usually resulted in higher mortality. Anaesthetized fish were wrapped in moist paper towelling, positioned belly-up on a grooved wet sponge, and covered with crushed ice. The gular membrane on the l e f t side was s l i t and a retractor inserted between the second and third g i l l arch to expose the roof of the buccal cavity. The dorsal buccal epithelium was then cut and folded back and the underlying parasphenoid bone dr i l l e d away with a dental burr. With practice, the location of the pituitary could be accurately deter-mined by reference to conspicuous nerves on the surface of the bone; these nerves were cut i n a l l hypophysectomies and sham operations. When the pituitary could be seen through the bone, probe and forceps were used to expose the pituitary which then was removed by aspiration. The cut edges of the buccal epithelium were repositioned but not sutured. Following removal of the pituitary, fish were placed in 40 1 aquaria f i l l e d with 25% sea-water (7?>/'pp)at 5 C and allowed to warm slowly to room temperature. Each tank had a subgravel f i l t e r and a i r -stone but, i n order to maintain salinity and avoid pH problems, no plants were added. No food was given un t i l 3 or 4 days after the oper-ation. A l l fish were kept i n the dilute seawater throughout post-operative recovery, ovarian regression, hormone treatment, and be-havioural testing. 8 A disease (or diseases) which virtually every hypophysectomized fish in my lab contracted was not identified. However, i t developed with a readily identifiable set of symptoms and could be cured easily with low mortality provided treatment was begun early. The f i r s t i n d i -cations of infection were a lowering of the dorsal f i n , folding of the paired fins, slight loss of balance, and accelerated swimming near the surface. Within a day, areas of the body surface became reddened, apparently by enlargement and rupture of the blood vessels in the skin, and the i n i t i a l symptoms were more pronounced. Treatment consists of the addition of 0.5 to 0.75 g K^C^O^ to a 40 1 tank. Fish undergoing treatment do not feed, apparently an effect of the drug, as healthy treated fish behave similarly. Several days following the disappear-ance of the reddening and the resumption of normal swimming, 50% of the aquarium water is replaced with 25% sea-water, and several days later the water is changed completely. Preliminary tests showed that fish which had been completely hypo-physectomized for more than one month (1) lost a l l body color, becoming a pale white, (2) had highly regressed ovaries, and (3) could not be induced to spawn in standard testing procedures. Prior to assignment to experimental groups, prospective test females could be rejected for incomplete hypophysectomy on the basis of body coloration. This method is effective but not foolproof, pre-sumably as pigmentation and reproduction are regulated by different pop-ulations of pituitary cells which may be removed separately in an i n -complete ablation. Following behavioural testing, heads were checked for pituitary remnants under a dissecting microsope; heads were not ex-amined histologically due to the time involved and a lack of confidence in the technique. Various aspects of ovarian histology were the f i n a l 9 c r i t e r i a for accepting or rejecting behavioural data from fish judged on the basis of body coloration to be completely hypophysectomized. This procedure is discussed i n the following section. C. Histology of Regressed Ovaries in Intact and Hypophysectomized Fish In interpreting the effects of various hormonal manipulations on the spawning behaviour of female goldfish, I have placed considerable emphasis on the ovarian histology of treated fi s h . The basic assump-tions involved in this approach derive from the results of a number of studies of female teleost reproductive endocrinology which indicate that pituitary hormones stimulate growth of ovarian f o l l i c l e s and de-position of yolk. Correlations observed i n the present study between the state of ovarian development and sexual receptivity have been used to develop histological c r i t e r i a by which spawning behaviour data were accepted or discarded. The application of this procedure to various types of experiment is discussed below. A more detailed account of some aspects of the histology of regressed ovaries of intact and hypo-physectomized fish i s given in the Appendix. In previous studies, the growth of goldfish oocytes has been divided into two basic stages. (Yamazaki, 1961), a f i r s t growth phase in which the oogonia develop into primary yolkless oocytes about 150y ,in diameter, and a second growth phase characterized by the for-mation of two types of cytoplasmic inclusions, yolk vesicles (cortical granules) and yolk granules (proteinaceous yolk). Yolk vesicles f i r s t appear in oocytes about 150\L in diameter. The yolk granule stage, which does not begin un t i l the oocyte is about 300u. in diameter, is divided into primary, secondary, and tertiary stages depending on the cytoplasmic pattern of granule distribution. Following hypophysectomy of female goldfish, yolk-laden or second growth phase oocytes become atretic (Yamazaki, 1961, 1965). This 10 sensitivity of the yolky oocytes i s related to stage of development. Oocytes i n late yolk vesicle, primary and secondary yolk granule stages are the most sensitive and begin to degenerate within a few days of pituitary removal. Next to degenerate are the tertiary yolk stage oocytes. Least sensitive are the early yolk vesicle stage oocytes which often are present and apparently healthy several months after the operation. There i s some evidence from this study that distinct changes in the distribution of the yolk vesicles precedes atresia in these oocytes. Eventually, a l l oocytes with yolk degener-ate in the absence of the pituitary, no new yolk formation occurs, and the ovary remains i n a regressed condition composed of various stages of f i r s t growth phase oocytes (Yamazaki, 1965). Published accounts suggest that in goldfish, as in other teleosts, previtello-genic oocytes are independent of the pituitary, at least to the extent that they do not degenerate following hypophysectomy. How-ever, in the present study atresia of previtellogenic oocytes was observed many times in inteact and i n hypophysectomized fish and their presence served as a valuable indicator of advanced ovarian regression. Details of previtellogenic atresia are discussed i n the Appendix. Treatment of hypophysectomized fish with gonadotropin prepar-ations induces growth of oocytes and deposition of yolk vesicles and yolk granules (Yamazaki, 1965; Yamazaki and Donaldson, 1968). It appears that the mechanism by which the pituitary stimulates formation of yolk granules i n fish i s similar to that in other vertebrates (Chester Jones et a l . , 1972; Gallien, 1975). At the level of the ovarian f o l l i c l e , gonadotropin stimulates both the production of estrogens (which induce the formation and mobilization of hepatic yolk proteins) and the uptake of yolk proteins from the blood 11 (Campbell and Idler, 1976; Emmersen and Petersen, 1976). Except for a study by Khoo (1974) in which estradiol, estrone, and es t r i o l were reported to induce yolk vesicle formation in hypophysectomized female goldfish, there is no information concerning the effect of steroids on :formation of this type of yolk. Under the experimental conditions of this study, spawning be-haviour could be induced without hormonal pretreatment in intact fish with ovaries in any stage of vitellogenesis: fish in which the ovaries contained only f i r s t growth phase oocytes were not receptive. For this reason, spawning behaviour data from a l l fish with ovaries containing any yolk are discarded from the results of experiments designed to test the abi l i t y of exogenous hormones to induce receptivity in intact, fish. In this study, ovaries are described as nonvitellogenic i f the oocytes contain neither yolk vesicles nor yolk granules. In experiments involving hypophysectomized f i s h , the problem of distinguishing acceptable behavioural data was more d i f f i c u l t . Basic-ally , data from fish judged on the basis of body coloration to be com-pletely hypophysectomized were considered acceptable i f there was histological evidence of previous and ongoing atresia. However, as treatment of hypophysectomized fish either with pituitary homogenates or with gonadotropin preparations arrested atresia and induced yolk formation, completeness of hypophysectomy in these cases was determined on the basis of body coloration alone. Unlike the situation in intact f i s h , the decision to accept or reject behavioural data was based not only on the presence or absence of yolk vesicles, which may persist for months after hypophysectomy, but also on the presence or absence of degenerating previtellogenic and vitellogenic oocytes. This was complicated by the fact that ovarian histology in hypophysectomized fish i s influenced by the state of ovarian 12 development at the time of the operation. For example, ovaries from two females sampled one month after complete hypophysectomy may show very different patterns of degeneration. The ovaries of a female spontaneously regressed at the time of hypophysectomy w i l l show no signs of second growth atresia but likely w i l l contain various stages of degenerating previtellogenic oocytes. In contrast, a female hypo-physectomized when the ovaries were in the early stages of yolk granule formation w i l l have extensive second growth phase atresia, but normal yolk vesicle stage oocytes w i l l probably be present and the degener-ation of previtellogenic oocytes w i l l not have commenced. In general, information from previous studies (Yamazaki, 1961, 1965; Khoo, 1974) on the sequence of events and time course of atresia following hypophysectomy has aided in distinguishing ongoing but incom-plete ovarian regression from: incomplete hypophysectomy. However, the finding (Khoo, 1974) that estrogens induce yolk vesicle formation i n hypophysectomized female goldfish suggested that i n hypophysectomized fish receiving steroid treatment, completeness of pituitary removal could not be accurately determined on the basis of ovarian histology. Ih the present study, the results of Experiment 5, equivalent in design to that carried out by Khoo, appeared to show that estradiol and -dihy-drotestosterone induced formation of yolk vesicles. However, as de-generation of yolk vesicle stage oocytes may not be complete by the sixth week after hypophysectomy (Yamazaki, 1965; personal observation) and start-of-treatment controls were not included in Khoo's nor in the present study, i t i s equally reasonable that these steroids may simply inhibit the degeneration of early yolk vesicle stage oocytes. In fact, Khoo's statement that i n some oocytes the 'induced' vesicles were scattered randomly in the cytoplasm, suggests that these oocytes were in the 13 early stages of atresia (see Appendix). In Experiment 11 of the present study, where steroid treatment was not begun until the third or fourth month after hypophysectomy, at which time degeneration of early yolk vesicle stage oocytes i s complete, the ovaries of fish re-ceiving estradiol showed no evidence of yolk vesicle formation. D. Behavioural Testing Procedures A l l behavioural tests were carried out in 60 1 aquaria supplied with an undergravel f i l t e r , 2" of quartz sand, and a floating layer of Ceratopterus. Development of low pH from the use of undergravel f i l t e r s was not a problem as these tanks were set up for only short periods. A l l experiments involved the induction and maintenance of sexual activity in male goldfish and most experiments also involved the i n -duction of ovulation in females. Both techniques are simple and dependable. For the induction of ovulation, gravid female goldfish which had a distended abdomen (preferably soft and asymetrical) and expanded ovipore were selected from the cold water stock tanks about noon on day 0 and warmed to approximately 20 C over a period of several hours. Females treated in this way were never observed to ovulate on day 1, while a highly variable proportion (presumably dependent on the state of ovarian maturity) ovulated spontaneously the morning of day 2. Intraperitoneal (i.p.) injection of approximately 3 IU /g HCG(0.6% NaCl vehicle) on the afternoon of day 1 induced ovulation by 0600 h on day 2 in virtually a l l cases and this technique was used routinely (for a more thorough discussion of the timing of events in goldfish ovulation, see Stacey and Pandey ['1975]). In goldfish, the paired ovaries are enclosed dorsally by ovisacs 14 which, are separate anteriorly and which join at the posterior of the ovaries to form a short oviduct. Following ovulation, many released oocytes move dorsally between the ovarian lamellae and f i l l the ovi-duct and ovisacs. The remainder of the ovulated eggs remain within the ovary until those in the ovisacs and oviduct are gradually re-leased through the ovipore during oviposition. Ovulation is detected by applying slight pressure to the abdomen to release a stream of eggs from the ovipore. The technique for inducing sexual activity in.male goldfish is similar to that for the induction of ovulation in females. Stock males were usually separated from females at the time of purchase and kept under standard cold water stock conditions. Two or three days before they are required for behavioural testing, males which have well developed tubercles on the opercula or on the leading rays of the pectoral fins are injected i.p. with 3 IU/g HCG and then warmed to 20 C over several hours. Males in prespawning condition spermiate on the day after being warmed and injected and usually display sufficient sexual activity the following day. Most males remain highly active for at least a week; supplementary HCG injections or exposure to spawning females usually extend this period. In earlier experiments, only one female at a time was tested in each observation tank as i t was f e l t that the spawning behaviour of one female might influence the behaviour of another and that, as males given a choice of females usually show marked preference for certain individuals,unequal stimulation of the females would result. Observa-tions-confirmed neither of these suspicions. To increase the number of females which could be observed at one time, and in some cases to re-duce the amount of courtship directed toward each female, as many as 5 females were tested simultaneously in a 60 1 tank. 15 The i n i t i a l method of behavioural testing was to place the female in the observation tank shortly before testing was to begin. Obser-vation commenced with the introduction of an active male which had previously been kept in an all-male tank. It soon became evident that the sexual activity of males was much greater i f the female was introduced to the male's tank and i f more than one male was present. Thus, in later experiments, from 3 to 5 active males were kept in each observation tank from the time they had been warmed and injected with HCG. The general level ,of sexual activity in each tank was assessed on the morning of testing by placing an ovulated or gravid female into each observation tank and waiting un t i l chasing had commenced. In tanks where male activity was only moderate, the 3 most active males were used for courting test females and the remaining males were removed for the duration of the test period. Only two males were used when male activity was high (the usual case), and only one male when the activity was extremely high. In a small aquarium, excessive chasing and butting can k i l l a female goldfish which is unable to escape. Behavioural testing sessions were usually of 3 h duration. Most responding females commenced spawning within the f i r s t hour of testing and stopped spawning by the end of the third hour. It was f e l t that extending the test period would increase the incidence of egg-binding (see experiment 2). Spawning activity was recorded on grid charts as numbers of spawning acts in each 5 minute interval of the test period. Following most tests, females were k i l l e d by over-anaesthetization in MS-222, and the ovaries fixed in Bouins Fluid, embedded in paraffin, sectioned and stained with Mallory connective tissue stain. 16 E. Description of Spawning Behaviour When an ovulated female and a spermiated male g o l d f i s h are placed together i n an aquarium provided with green vegetation, the usual course of events i s as follows. The male approaches the female and may make contact ( e s p e c i a l l y i n the region of the ovipore) or may turn away and approach several times before making contact. Usually the female does not swim quickly at t h i s stage and often appears to present the ovipore region to the male. This occurs as a turning away from the male, sometimes with the t a i l higher than the head. E s p e c i a l l y when the male f a i l s to investigate a receptive female, i t may be observed that the female w i l l approach the male, sometimes making contact, but usually swim-ming past the head of the male. .-Normally, males take-the i n i t i a t i v e i n i n v e s t i g a t i n g the female, and the female appears to play a passive r o l e . The male soon begins to butt the female (pushing the female with his snout) i n the region of the b e l l y , ovipore and caudal peduncle i f the female i s swimming away, and i n v i r t u a l l y any part of the body i f the female remains stationary. A chase develops, the female swimming at a moderate speed, the male either swimming very close behind, or swimming beside and maintaining contact with the female. If green vegetation i s not present, the female does not perform o v i p o s i t i o n behaviour though some eggs may be dropped, apparently inadvertently due to the pressure of the male. The female i n i t i a t e s spawning behaviour by performing a ' r i s e ' . In t h i s , she swims up toward f l o a t i n g vegetation, pushes:.the snout i n (the head may or may not be covered), and occasionally mouthes the vegetation. T y p i c a l l y , r i s i n g i s followed by a spawning a c t ( o v i -p o s i t i o n ,),. The female enters the vegetation and turns on her side, as the male p a r a l l e l s t h i s motion from s l i g h t l y behind and below. The p a i r 17 (female above, male below) then quickly swims up, breaking the water surface, and then down, leaving the vegetation. It is during this rapid, synchronized rising and arching, in which the male appears to attempt to push the female out of the water, that the eggs and sperm are released. Whether the female exercises any control over egg release during the spawning act is not known; however, even slight pressure against the belly such as occurs during prespawning chasing is sometimes sufficient to release some eggs. Netting ovulated females invariably causes egg release. Following emergence from the vegetation, the female may resume swimming at a moderate speed, followed closely by the male. More often, the female w i l l return to the vegetation and complete one or several spawning acts in quick succession. This 'clumping' of spawning acts interspersed with periods of chasing and butting agrees with Yamazaki's (1965) description. Rising does not always lead to completion of a spawning act. Most females which have not begun to spawn w i l l perform a number of rises before completing the f i r s t spawning act, after which the ratio of rises to spawning acts decreases. Occasionally, the male is responsible for the failure of rising to lead to spawning, either because the speed of his approach startles the female, or because of incorrect orientation to the female. In the cases where rising does not lead to spawning, the female either swims quickly away from the vegetation or remains motion-less in the vegetation for as long as several minutes, not responding to the actions of the male. Though female goldfish w i l l sometimes spawn oh submerged vege-tation, and a minority of individuals seem to prefer this site (B.Partridge, personal communication), the majority of ovipositions occur on floating 18 vegetation both in the laboratory and under natural conditions (Innes, 1949) . Due to the problem of distinguishing a spawning act on submerged vegetation from an attempt by the female to avoid the male, and because only floating vegetation was provided in warm water stock tanks, only floating vegetation was used in experimental situations. Spawning may continue for several hours and involve several hundred spawning acts. Generally, termination of spawning is coincident with the shedding of a l l ovulated eggs, though on occasion small numbers of eggs have been removed from females which had ceased spawning behaviour (Yamazaki, 1965; personal observation). Though this was observed on only a few occasions and was not investigated further, i t appeared that females which had finished spawning (i.e. had shed a l l ovulated eggs) inhibited the male's chasing and butting by increasing the swimming speed and becoming very effectively evasive for a short period of time. Observations by myself }and by others (Innes, 1949) under more natural conditions, suggest that the courtship of male goldfish is a highly competitive, vigorous, and lengthy event. In experimental situations where .3 or .4 males were placed with an equivalent number of females in 60 1 aquaria, one male invariably initiated courtship. Usually, this quickly stimulated courtship of the same female by one or more of the remaining males. In fact, the com-petitive stimulation of courtship between males was so reliable, that adding a new,^active male to an experimental tank to stimulate courtship in inactive resident males became standard experimental procedure. There i s evidence (Partridge, Liley, and Stacey, 1976) that male goldfish are attracted to and stimulated to court female goldfish by a pheromone, probably of ovarian origin. This description of events in a normal spawning situation has ignored the question of whether var i a b i l i t y in the stimulus provided by the male is a significant factor in the responsiveness of the female. This problem was not 19 investigated experimentally as several observations indicated that provided the male i s sufficiently aroused to complete the spawning act, the intensity of male courtship has no obvious effect on the spawning behaviour of the female. For example, on many occasions in which receptive females (either natur-ally ovulated or with experimentally induced receptivity) had been placed with relatively inactive males which had not begun to court, the female was seen to rise (a behaviour which usually e l i c i t s rising even in inactive males) and, after allowing the male to position, perform a normal spawning act without any of the usual preliminary investigation, butting, or chasing on the part of the male. This i s not evidence that prespawning courtship has no stimulating effect on the female, but i t does demonstrate that i t is not prerequisite. Furthermore, in the multi-male, multi-female testing pro-cedure used in many of my experiments, the males invariably showed a prefer-ence for certain of the females. While most i n i t i a l l y 'unattractive' females became and remained attractive after they had begun to spawn, the usual s i t -uation was that certain females consistently e l i c i t e d more courtship through-out the test, regardless of their spawning behaviour. In extreme cases, one or more females which never spawned throughout a test received almost constant courtship, while other spawning females were attended only at the times they entered the floating vegetation. These observations have been interpreted as indicating that the occurrence and rate of spawning are functions of the physiology of the female goldfish. Obviously, the presence of the male is required at the time of the oviposition behaviour, but provided the male i s sexually active enough to respond to the stimulus of a rising female and complete the spawning act, the intensity of the male's other courtship activities do not affect the spawning behaviour of the female. This view is at odds with that of Yamazaki (1965) 20 who suggests, without providing data, that prespawning courtship or chasing by males stimulates both ovulation and oviposition. An important problem in the recording of female goldfish spawning be-haviour is the choice of an appropriate parameter with which to measure sexual receptivity (the tendency to perform spawning behaviour). The apparent lack of female prespawning behaviours leave only two obvious indices of receptivity, rising behaviour and the spawning act. Though rising behaviour might at f i r s t appear to be an acceptable measure of sexual receptivity, i t has several serious drawbacks. The most important is that behaviour similar to rising occurs commonly in a nonsexual context in both males and females. Second, though each spawning act i s necessarily pre-ceded by a rise, the tendency for rising to be disrupted prior to the com-pletion of a spawning act usually results in the performance of many more rises than spawning acts. This is particularly evident when spawning be-haviour is beginning or is proceding at a low frequency. As noted previously (page 17), aspects of male spawning behaviour may alter the relation between the numbers of rises and spawning acts. The spawning act is a distinctive, stereotyped, and easily quantified behaviour which i s restricted to a sexual context; i t i s the only measure of sexual receptivity used in this study. The term receptivity is used s t r i c t l y in a behavioural-senseuandniimplies nothing' ahoutuunderlyingrphysiological mechanisms. CHAPTER III THE ROLE OF OVULATED EGGS IN THE SPAWNING BEHAVIOUR OF FEMALE GOLDFISH 22 A. Experiment 1. Effect of Removal and Replacement of Ovulated Eggs on the Spawning Behaviour of Ovulated Goldfish 1. Introduction That ovulated eggs within the ovary are involved in the induction of spawning behaviour in female goldfish was suggested by Yamazaki (1965) whose observations on the effects of egg removal led him to speculate that 'ripe eggs in the ovarian lumen stimulate the spawning behaviour of females via some pathway'. In the present study, preliminary examination of the role of ovula-ted eggs i n stimulating spawning behaviour utilized several simple tech-niques. Yamazaki's results were repeated. Spawning ceased i f a l l ovula-ted eggs were removed by hand-stripping (gently squeezing the area between the pectorals and the ovipore between moistened fingers). As stripping only a portion of the eggs did not terminate spawning, i t was concluded that the effect of complete egg removal was not due to disturbance or i n -jury, but rather to the absence of ovulated eggs, as Yamazaki had suggested. It was found that the use of an 'oviduct plug' to prevent the release of ovulated eggs greatly extended the duration of spawning in ovulated fish. Whereas the normal duration of spawning is less than two hours, 'plugged' females would spawn throughout the day of ovulation and sometimes continue the following day. The following experiment was carried out to determine whether place-ment of ovulated eggs in the ovarian lumen would restore rather than simply o prolong spawning. 2. Materials and Methods' Females which had ovulated spontaneously (in response to warming but 23 without HCG injection) the morning of the test day were taken from holding tanks and placed singly in 60 1 observation tanks. A spermiated male was introduced into each tank after 20 to 30 minutes. The latency to the f i r s t spawning act and the number of spawning acts in the following 20 minutes provided a record of the normal spawning activity of each fi s h . Immediately following this i n i t i a l 20 minute spawning period, each fish was removed, anaesthetized in MS-222, and the eggs removed and stored in egg-injection syringes. For egg removal, females were dried gently with paper towelling and hand-held such that the ovipore was over (and the anal f i n outside) a small plastic cup of about 5 ml capacity. It is essential to use plastic apparatus when handling ovulated eggs, as eggs w i l l eventually adhere to glass. Furthermore, eggs must not be allowed to come into contact with water as this causes adhesion and hardening. Eggs are released into the cup.; by gentle pressure against the belly and drawn up into a plastic 1 ml syringe fit t e d with bevelled PE tubing. Preliminary tests had shown that a single sequence of stripping re-moved only a portion of the ovulated eggs even i f considerable pressure was exerted on the belly; when stripping was repeated in 10 or 15 minutes, additional eggs could usually be expressed with only slight pressure. It i s believed that i t is the eggs in the oviduct and in the posterior por-tions of the ovisacs which are removed during each stripping. Following removal of these accessible eggs, additional ovulated eggs move dorsally along the channels between the ovarian lamellae and into the ovisacs and oviduct, from which they can be removed by further stripping. Following egg removal the fish were revived and returned to the obser-vation tanks where the behaviour of male and female was recorded for one 24 hour. During the f i r s t 30 minutes, the females were removed at 10 minute intervals and squeezed to remove remaining ovulated eggs; for the second 30 minutes, the pairs were l e f t undisturbed. The purpose of this one hour observation period was to ensure that egg removal was complete and that spawning had terminated. In preliminary tests i t had been found that low levels of spawning activity occurred i f even a small number of ovulated eggs remained in the oviduct. Following the observation period, the females were removed from the observation tanks, anaesthetized in MS-222, and given one of the following treatments: Group I - a no-treatment handling control in which each fish was simply revived and returned to the observation tank, Group II - an ovipore plug control in which each fish was f i t t e d with an ovipore plug, revived, and returned to the ob-servation tank with the male, Group III - an egg-injection treatment in which each female was injected with i t s own ovulated eggs, fitt e d with an ovipore plug, revived, and returned to the spawning tank with the male. Oviduct plugs consisted of PE tubing (various diameters) with a f i r e -polished glass plug i n the proximal end and recurved barbs cut in the dis-ta l end (total length 4-8 mm). In inserting the plugs, i t was often help-f u l to f i r s t insert a fine glass probe into the oviduct and stretch the anterior margin (the junction with the anus) forward. This precaution re-duces the chance of placing the plug in the rectum. No attempt was made to replace a l l the eggs removed from each Group III female as preliminary tests had shown that this usually resulted in the rupture of the oviduct 25 and failure to induce spawning behaviour. Each female was injected with approximately 0.025 ml eggs per gram body weight. Though the time between treatment of each fish and i t s return to the observation tank was not constant (as more than one fish was treated at a time), effort was made to minimize this interval. Thus, of the 33 fi s h tested, a l l but one were returned to the observation tanks within 10 min-utes of treatment. Time of the f i r s t spawning act and a l l subsequent spawn-ing activities were recorded for one hour, 3. Results and Discussion There were no differences between the three treatment groups in the latency to the f i r s t pretreatment spawning act following i n i t i a l introduc-tion to the male, nor in the number of spawning acts in the 20 minute pre-treatment period (Table I). Removal of ovulated eggs was highly effective in terminating spawning behaviour; only one of 11 control fish (Group I) spawned during the posttreatment period. The injection of ovulated eggs (Group III) restored spawning behaviour. This is obvious whether the number of fish responding or the spawning rate of responding fi s h i s considered. Presence of an ovipore plug may have contributed to the spawning behaviour of Group III f i s h , as more fish spawned following treatment with an ovipore plug (Group II) than following the control treatment (Group I); however, the posttreatment responses of Group I and Group II were not significantly--different (Mann-Whitney U-test). A number of informal tests conducted concurrently with the present experiment showed that the abi l i t y of egg injection to induce spawning be-havior i s not restricted to the day of ovulation. Injection of eggs from 26 TABLE I EFFECT OF REMOVAL AND REPLACEMENT OF OVULATED EGGS ON THE SPAWNING BEHAVIOUR OF OVULATED FEMALE GOLDFISH 27 Pre-Treatment Spawning Post-Treatment Spawning Group Treatment Latency (min) No. Spawning Acts/20 min. No. Spawning % Acts/20 min. Restored I No Treatment 6 34 0 0 6 60 0 0 (n=ll) 5 21 0 0 8 17 0 0 6 25 0 0 7 18 0 0 50 41 7 17 200 14 0 0 6 56 0 0 41 31 0 0 4 33 0 0 mean = 14.4 mean =31.8 (1/11) II Ovipore Plug 19 44 0 0 15 20 0 0 20 26 0 0 (n=ll) 2 55 0 0 3 38 0 0 10 54 10 18 4 40 1 2 5 25 0 0 15 38 4 10 2 40 11 27 2 56 0 0 mean = 18.8 mean= 39.6 (4/11) III Ovulated Eggs 20 12 10 83 and 6 57 37 65 Ovipore Plug 4 60 57 95 (n=ll) 6 20 20 100 10 29 20 69 5 30 7 23 8 59 56 94 4 26 46 177 5 30 31 103 3 40 8 20 22 9 15 167 mean = 8.4 mean= 33.8 (11/11) * = No. spawning acts in the most active 20 minute period of spawning in the test hour. No difference between groups in pretreatment spawning behaviour (Mann-Whitney U test). 28 ovulated donor females consistently induced normal spawning behaviour both i n females which had ovulated as much as a month or more before being tested and in fish which had failed to ovulate after transfer to warm water. These results both support Yamazaki's' hypothesis concerning the s t i -mulatory effect of ovulated eggs on spawning behaviour and demonstrate that the normally sequential processes of ovulation and oviposition can be easily dissociated. The simple technique of egg injection thus allows the endocrine regulation of sexual behaviour to be examined independently of the control of ovulation. This is c r i t i c a l when i t i s realized that the traditional approach of ovariectomy and replacement therapy is rather inappropriate in this case due to the dependence of the female's spawning behaviour on internal cues provided by eggs in the ovarian lumen. 29 B. Experiment 2. Induction of Spawning Behaviour by Injection of Egg Substitutes 1. Introduction Although the technique of egg injection is a simple and reliable method for inducing spawning behaviour, i t is associated with three obvious problems: (i) the necessity, and occasional d i f f i c u l t y , of inducing ovulation in egg-donor females, ( i i ) the possibility of differences in stimulus quality of eggs from donor females, and ( i i i ) the tendency for eggs to become adhesive (bound) on contact with water and thus to lose the a b i l i t y stimulate spawning. Egg binding was a common problem and reduced fi n a l sample sizes in a l l experiments requiring egg injection. Binding sufficient to inhibit spawn-ing usually occurs as hardened eggs attached to the withdrawan ovipore plug or as a small cluster of eggs which may be expressed with some d i f f i c u l t y from the ovipore by pressure on the belly. In some cases, binding may i n -volve the hardening and adhesion of the entire mass of injected eggs. In cases where fish with bound eggs were not inhibited from spawning, i t was assumed the binding had occurred after the onset of spawning behaviour. Data for a l l non-responding fish for which there was any suggestion of egg binding were discarded. Finding an injectable egg substitute for the induction of spawning would eliminate the d i f f i c u l t i e s of using eggs and greatly expedite experi-ments involving the a r t i f i c i a l induction of spawning behaviour (i.e., spawn-ing without ovulation). To this end, three potential egg substitutes 30 have been tested; petroleum j e l l y , gelatin, and Dow Corning 200 Silicone Fluid. 2. Methods, Results, and Discussion The discouraging results of the preliminary tests did not justify a full-scale experiment to test egg substitutes. Test conditions were not standardized and varied in such aspects as duration of the test period, the time elapsed since the previous ovulation of the test female, and whe-ther or not an ovipore plug was used (Table II). Petroleum j e l l y and Dow Corning 200 Silicone Fluid were simply drawn up i n an egg injection syringe and injected. Gelatin was dissolved in warm dechlorinated tapwater at two concentrations, 0.3 g/ml and 0.6 g/ml, and stored overnight at 4 C. Injection was as for petroleum j e l l y and the Silicone Fluid. As seen i n Table II, both gelatin and petroleum j e l l y induced spawning behaviour in some fish , although the response to these substances was much less than that obtained with egg injection. These limited results demonstrate that the ab i l i t y to induce spawning behaviour i s not a property unique to ovulated eggs and that in some fish physical cues alone may be sufficient to induce the response. The failure of the injected materials to duplicate the response to ovulated eggs i s be-lieved to result from their i n a b i l i t y to duplicate the physical stimulus provided by eggs. The ina b i l i t y of physically altered (adhesive and hardened) bound eggs to induce spawning supports this contention. An alternative explanation for the low stimulus quality of the tested egg substitutes is that they lack some chemical stimulant normally associa-ted with ovulated eggs or f l u i d . The fact that binding of only a few eggs 31 TABLE II EFFECT OF 'EGG SUBSTITUTES' (PETROLEUM JELLY, GELATIN, AND SILICONE FLUID) ON SPAWNING BEHAVIOUR OF RECEPTIVE FEMALE GOLDFISH 32 Treatment Ovipore Condition Duration of No. of Plug of Observation Spawning Used F i s h Period A c t s (hours) * Ovulated + 1 week POV 3 260 Eggs II 1.5 65 Petroleum + II 2 0 J e l l y + it 2 0 - II 3 0 + II 3 0 + II 3 3 - it 3 22 + II 3 108 Gela t i n (low — II 2 0 concentration) - II 2 0 - II 2 0 Gelatin (high — M 2 35 concentration) - II 2 0 - II 2 0 — preovulatory 2 0 S i l i c o n e _ 1 week POV 1.5 0 F l u i d - II 1.5 0 * Tested 1 week a f t e r ovulation 33 near the ovipore w i l l often completely i n h i b i t spawning even though the remainder of the injected eggs appears normal, would seem to argue against this alternative. C. Summary of Chapter I I I 1. The duration of spawning behaviour of ovulated f i s h i s extended by plugging the ovipore to prevent egg release. 2. In ovulated f i s h , spawning behaviour i s terminated by removing ovulated eggs. Spawning behaviour i s restored by i n j e c t i n g ovulated eggs through the ovipore and into the ovarian lumen. 3. Injection of ovulated eggs into preovulatory and postovulatory f i s h induces spawning behaviour within several hours. 4. On contact with water, ovulated eggs become hard and adhesive and lose the a b i l i t y to induce spawning behaviour. 5. Several substitutes for ovulated eggs (gelatin, petroleum j e l l y ) induce spawning behaviour when injected into the ovarian lumen; these substances are less effective than ovulated eggs. CHAPTER IV THE EFFECT OF STEROIDS ON THE SPAWNING BEHAVIOUR OF FEMALE GOLDFISH 36 A. Experiment 3, Steroid-Induced Spawning Behaviour i n Intact, Regressed Female Goldfish 1. Introduction Normally, gonadal maturation and ovulation in goldfish are accelera-ted by raising the water temperature to 20 C. However, the ovaries of spent or mature fish kept at 20 C for an extended period (2 to 3 months is usually sufficient) become regressed (Khoo, 1974); vi r t u a l l y a l l oocytes are in the previtellogenic stage although some may contain small yolk vesir-cles (Yamazaki, 1961, 1965). Since vitellogenesis resumes i f regressed fe-males are returned to cold water (12 C) soon after spawning, temperature regression appears not to result from an incapacity of the post-spawning reproductive system, but rather from an inhibition of the system by pro-longed high temperature. The mechanism by which high temperature prevents ovarian recrudescence i s not known; however, there i s some evidence (Scruggs, 1951; Nagahama, 1973) that gonadotropin production is low in the post-spawn-ing period. Examination of a similar phenomenon in another teleost, Gillichthys mirabilis, suggests the mechanism may operate by a decrease both i n gonadotropin secretion and in the response of the gonad to gonado-tropin (De Vlaming, 1972). Presumably, the functional significance of this temperature-induced regression i s that the stimulatory effect of elevated temperature on gonadal maturation is confined to the early part of the warm season when rearing conditions for the fry are optimal. Injection of ovulated eggs into temperature-regressed fish invariably failed to induce spawning behaviour. Khoo (personal communication) found that the activity of steroid dehydrogenases i s very low in intact, regressed female goldfish and Liley (1972) has shown that estradiol is effective in 37 restoring the sexual receptivity of both hypophysectomized and ovariec-tomized female guppies. On the basis of these findings, i t seemed a reasonable hypothesis that the lack of a spawning response to egg injection was due to reduced steroid, in particular to reduced estradiol, production. To test this hypothesis, an experiment was carried out which showed that injection of estradiol restored the spawning of temperature-regressed fish i n response to egg injection. As there was no reason to believe that estradiol i s the only steroid capable of inducing spawning behaviour, the effects of other steroids were also examined; these additional experiments revealed an unexpected lack of specificity in the behavioural response to steroid treatment. Ten steroids known to be synthesized by teleost ovaries (Eckstein, 1970; Eckstein and Katz, 1971; Lambert et_ al_. , 1971; Colombo and Belvedere, 1976; OZon, 1972) were tested; cholesterol, pregnenolone, 17'1* -OHr-pregneno-lone, progesterone, 17? -OH-progesterone, dehydroepiandrosterone, androstene-dione, testosterone, 11-ketotestosterone, and 17,6 -estradiol. The effects of estrone, e s t r i o l , 5? -dihydrotestosterone, and androsterone were also ex-amined. Dehydroepiandrosterone, androstenedione, testosterone, e s t r i o l , estrone, and estradiol were chosen as they induce sexual behaviour in fe-male mammals (Beyer et_ al., 1970, 1971) and because the latter three also stimulate receptive behaviour in female guppies (Liley, 1972). Cholesterol, pregnenolone, 17°^  -OH-pregnenolone, and 17°?. -OH-progesterone do not stimu-late sexual behaviour in female mammals and were not expected to induce spawning behaviour. Similarly, i t was thought that treatment with proges-terone would not stimulate spawning behaviour. Facilitation by progesterone of sexual behaviour in female rodents requires estrogen pretreatment (je.g., 38 Joslyn e_t a l . , 1971) and in ovariectomized female guppies progesterone f a i l s to restore receptivity (Liley, 1972). The 5 a -reduced androgens, androsterone and dihydrotestosterone, were tested to examine the possibility that androgens stimulate spawning behaviour following aromatization to estrogens; there is evidence that in mammals 5^ -reduced steroids are not aromatized (Thompson et aT., 1971). B. Materials and Methods Although this study is presented as a single experiment, i t i s in fact a series of small experiments carried out over a period of a year and a half. Tank f a c i l i t i e s were such that i t was not possible to obtain re-gressed fish and preovulatory egg-donor females in sufficient numbers to perform the entire experiment at one time. Each small experiment tested the effects of several steroids and a saline control injection. Test fish had been kept at 20 C for 3 to 9 months and were assumed to be regressed. Assignment to test groups was not random, but was arranged so that the 4-6 fish in each 40 1 aquarium could be identified individually by morphological characters or color patterns. Only one treatment group was kept in each tank. Steroids (Sigma) were ground to a fine powder in a Misco homogenizer, suspended at a concentration of 2.5y g/pl in 0.6% NaCl containing 4 drops Tween 80/100 ml and injected at a dosage of 20 Ug/g. Suspensions were made up 10 ml at a time and kept i n sealed centrifuge tubes at 4 C. They were discarded when depleted to 4-5 ml as a precaution against degradation and changes in concentration. Fish were ligh t l y anaesthetized in MS^ -222 prior to. injection and lai d on.a wet paper towel. The steroid suspension was agitated immediately prior 39 to injection, which was always intraperitoneal through the right body wall, slightly above and behind the base of the pelvic fins... The needle (25-30 gauge) was inserted at least 1/2" through the body wall to minimize the efflux of injected vehicle. Following injection, f i s h were immediate-ly returned to the holding tank to recover. Each fish received 5 injec-tions on alternate days over a 9 day period. On the morning of the day following the fi n a l injectionX/ occasionally the following day i f ovulated eggs were not available) fish were anesthetized in MS-222, injected with ovulated eggs (approximately 0.025 ml/g), fitted with an ovipore plug, and l e f t to recover i n 2 1 beakers for 30 to 60 minutes. Fish were then placed in 60 1 observation aquaria containing actively courting males and spawning behaviour was recorded for 3 hours. In some of the later tests in which more females than males were placed in each observation tank and the chance of injected eggs being expressed during chasing was low, ovipore plugs were not used. Following the 3 hour test period, females were removed from the obser-vation tanks and anaesthetized and the ovipore plugs were carefully re-moved. The fish were then checked for binding of the injected eggs by gently squeezing the belly near the ovipore with a moistened finger, and egg-bound, non-responding fish discarded from the sample. A l l fish were sacrificed and the ovaries fixed and prepared for histological examination. C. Results Of 191 'regressed' fi s h which received steroid or saline injections in this experiment, only 41 individuals achieved the requirements of successful egg injection without binding, and complete absence of yolky 40 oocytes in the ovaries. The spawning data for these fi s h are presented in columns I and II of Table III. Spawning scores in columns III, IV, and V are from incompletely regressed fish with ovaries in the..early stages of yolk vesicle formation. Ovaries of fish in column III contained some yolk vesicle stage oocytes; as they also contained early stages of degener-ating yolk vesicle oocytes, i t appeared that temperature regression had not yet been completed. In contrast, as the ovaries of the few fish i n column IV contained some oocytes in the yolk vesicle stage, but had only advanced atretic yolk vesicle oocytes, i t appeared either that temperature regression had been arrested prior to completion, or that vitellogenesis had recommenced. The fi s h in column V showed no signs of atresia and were in the early stages of yolk vesicle formation. The histological differences between vitellogenic (stages III, IV, and V) and nonvitellogenic ovaries (stages I and II) were much more obvious than were the differences between stages I and II or among stages III, IV, and V. In vitellogenic ovaries, oocytes containing yolk vesicles were usually quite numerous. On the other hand, the incidence of degenerating oocytes was nearly always low and i t is quite l i k e l y that i f larger portions of the ovary had been examined, some stage IV and V ovaries would have been reclassified as stage III. Data from egg-bound fish and from fish with advanced yolk formation are not i n -cluded i n Table III. None of the lOgregressed control females injected with saline (columns I and II) performed any spawning acts during the 3 hour test period. Of the 5 saline-treated fish (column III) which had some early yolk vesicle stage oocytes plus signs of ongoing regression, only one individual performed one spawning act. In contrast, 4 of the 5 females in columns IV and V performed high levels of spawning during the test period. TABLE III EFFECT OF STEROID TREATMENTS ON SPAWNING BEHAVIOUR OF FEMALE GOLDFISH WITH TEMPERATURE-REGRESSED OVARIES 42 No. of spawning acts of No. of spawning acts of Treatment fish with ovaries Treatment fi s h with ovaries in stage in stage I II III IV V I II III IV V 0 0 1 19 16 Saline 0 .0 0 0 0 0 0 32 0 0 0 95 0 0 0 0 0 Choles- 0 0 terol 0 46 29 0 Proges- 0 terone 0 0 17oc._oH- 0 0 0 0 Proges-terone 0 • Pregnen1^ 17 12 33 58 olone 17<*-0H-Pregnen-10 56 19 83 101 olone 1 Dehydro- 1 epiandro- 209 sterone 23 Androstene- - 70 109 77 0 dione 77 44 Testos- 36 80 54 terone 7 176 314 4 100 33 14 86 Estra- 119 0 38 40 diol 3 43 131 6 14 56 38 0 0 E s t r i o l 6 14 0 0 0 Estrone 33 32 23 8 24 Dihydro- 32 16 testos- 40 terone 36 11-Keto- 86 77 48 1 s testos-terone 32 106 21 45 Andro- 7 52 17 82 sterone 59 28 0 18 83 55 Stage I (no yolk vesicles; degenerating primary oocytes) and Stage II (no yolk vesicles; no degenerating primary oocytes) ovaries are completely regressed. Stage III '(yolk vesicles; degenerating secondary oocytes), Stage IV (yolk vesicles; advanced degenerating secondary oocytes), and Stage V (yolk vesicles; no degenerating oocytes) ovaries are not completely regressed. Spawning scores are total number of spawning acts per 3 hour test period. Data from fish with advanced stages of yolk formation not included in table. On the basis pf the above histological interpretation, and of the relation between spawning behaviour and ovarian histology in saline-treated fi s h , i t is suggested that only data from fish with nonvitello-genic ovaries (stages I and II) be accepted as evidence for the action of exogenous steroids on spawning behaviour. Many fish with ovaries in stages III, IV, and V probably were receptive prior to receiving steroid treat-ments . In contrast to the lack of response of the control group, regressed individuals in groups treated with ly^-OH-pregnenolone, androstenedione, testosterone, estradiol, dihydrotestosterone, 11-ketotestosterone, and androsterone performed high levels of spawning following egg injection. The regressed fish treated with pregnenolone responded to egg injection, sug-gesting that this steroid also affects spawning behaviour. Es t r i o l and estrone appear to be ineffective, although the sample sizes are small. IT*-OH-progesterone does not restore receptivity in regressed f i s h ; none of 5 individuals tested showed any spawning response. There is no i n -formation as to the effectiveness of progesterone in restoring receptivity as none of the group receiving progesterone were completely regressed. How-ever, i t is of interest that none of these four marginally vitellogenic fi s h displayed any spawning behaviour following egg injection. This result may simply indicate a period of poor egg injection technique. Alternatively, i t suggests an inhibitory effect of progesterone on responsiveness to egg injection in fish commencing ovarian maturation. Two other progesterone-treated fish with much more advanced yolk formation (not shown in Table III) spawned normally following egg injection. No fish in the groups injected with cholesterol or dehydroepiandroster-one had completely regressed ovaries. 44 D. Discussion The results of this experiment demonstrate that a variety of steroids restores the responsiveness of temperature-regressed fish to injection of ovulated eggs. The mechanism of action of the effective steroids is not known. They may stimulate central nervous structures regulating sex-ual behaviour, or may have some peripheral effect in sensitizing the fish to the stimulus of ovulated eggs in the oviduct. The only conclusion which can be drawn regarding the diversity of effective steroids is that under the conditions of this experiment the mechanism controlling sexual behaviour does not appear to be highly steroid specific. There are several obvious explanations which could account for this. One explanation may simply be that the dosage of steroid used was pharmacologically high and thus specificity was masked. Certainly the dosage used in this study is much higher than those found effective in i n -ducing receptivity in female mammals. However, in a study (Liley, 1972) which employed essentially the same dosage as in the present work, estradiol restored sexual behaviour of ovariectomized female guppies, and testoster-one was ineffective. Another plausible explanation is that the ovary of the temperature-regressed goldfish possesses an active steroidogenic pathway capable of converting exogenous steroids to one or more behaviourally active forms. Such an interpretation is not at variance with the reported steroidogenic potential of the previtellogenic ovary of the eel, Angullla (Colombo and Belvedere, 1976). The results of this experiment also emphasize that fish with ovaries in advanced stages of regression are responsive to the injection of eggs, 45 provided that the ovaries contain at least some yolk v e s i c l e deposition. Thus i t i s evident that the mechanism regulating spawning behaviour i s functional under endocrine conditions which l i k e l y are very different from those present at the time of ovulation. The present experiment provides no evidence as to whether the exogen-ous steroids act d i r e c t l y to influence behaviour, or exert some action on or with the p i t u i t a r y . The following experiments, using hypophysectomized f i s h , attempt to answer this question. 46 B. Experiment 4. Ineffectiveness of steroids on spawning behaviour of hypophysectomized female goldfish 1. Introduction In Experiment 3 i t was shown that a number of steroids restore spawn-ing in response to injection of ovulated eggs in intact, nonreceptive, tem-perature-regressed female goldfish. Liley (1972) has shown that pituitary hormones are not necessary for the expression of sexual behaviour i n the female guppy; estradiol injections alone restore the receptivity of hypo-physectomized fish. To determine whether the pituitary is involved i n the action of steroids on female goldfish spawning behaviour, an experiment similar to Experiment 3 was performed with hypophysectomized fi s h . Seven steroids were tested. Estradiol and testosterone were used as they were known to be effective i n intact fish. Although some evidence suggests they are ineffective in intact female goldfish, estrone and e s t r i o l were used as they are effective in the female guppy (Liley, 1972). As deoxycorticoster-one has been shown to induce ovulation and oviposition i n hypophysectomized female Heteropneustes (Sundararaj and Goswami, 1966), this steroid was also tested. Progesterone was also tested and cholesterol was used as a control. 2. Materials and Methods Females which had been kept at 20 C for 4 to 6 months and therefore were assumed to have regressed ovaries, were hypophysectomized and kept in 25% seawater for 4 to 9 weeks before beginning steroid treatment. Prepara-tion, dosage, and frequency of injection of steroids was as described i n Experiment 3. However, due to d i f f i c u l t i e s i n inducing ovulation in donor females, the duration of treatment varied from 7 days (4 injections) to 19 days (10 injections). Collection and injection of ovulated eggs were as 47 described in Experiment 1 and testing procedures were as described i n Experiment 3. 3. Results As this experiment was carried out before discovery of the treatment for 'red disease', mortalities were high and the sample sizes are small. Of 24 f i s h surviving the treatment, the data from six were discarded as the fish had bound eggs. None of the remaining 18 (estradiol - 7; e s t r i o l - 3; estrone - 2; testosterone - 1; progesterone - 3; cholesterol -2) displayed any spawning activity during the three hour test period. No fish treated with deoxycorticosterone survived to be tested. 4. Discussion It appears that hypophysectomy interferes i n some way with the a b i l i t y of exogenous steroids to potentiate the spawning response to. injected eggs. This may indicate either that the effect of steroids on spawning behaviour of intact fish is mediated by the pituitary, or that a pituitary factor i s required in addition to the steroid. It i s also possible that the mechanism regulating spawning behaviour requires stimulation by a steroid or steroids other than those used in this experiment. In the female rat, there is evidence that the lordosis response to exogenous steroids is decreased i f the hormonal treatment is preceded by a prolonged period of hormonal deprivation (Damassa and Davidson, 1973; Beach and Orndoff, 1974) . Thus the ineffectiyeness of steroids in this experiment may have been due to a decline in sensitivity to steroids in the 4 to 9 week interval between hypophysectomy and the i n i t i a t i o n of treatment (the period of hormone deprivation may have been longer as the fish most l i k e l y were regressed prior to hypophysectomy). The following experiment explores this latter possibility. 48 C. Experiment 5. Ineffectiveness of Long-Term Steroid Treatment on Spawning Behaviour of Hypophysectomized Female Gold-fis h 1. Introduction In Experiment 4 i t was suggested that the ineffectiveness of steroids in inducing sexual behaviour of hypophysectomized female goldfish might be due to a decrease in responsiveness to steroids following hypophysectomy. A similar phenomenon is reported to occur in female rats following ovariec-tomy (Beach and Orndoff, 1974; Damassa and Davidson, 1973). In goldfish, where ovarian development is normally very prolonged, hormonal priming of sexual behaviour i s l i k e l y to occur over an extended period. In addition, i f the hypothesis that unreceptive, temperature-regressed fish are capable of in vivo steroid conversion is correct, i t is li k e l y that the hypophyseal-gonadal axis i s active, although operating at a level insufficient to induce.oocyte growth, vitellogeriesis, and sexual re-ceptivity. Thus, the degree of hormonal deprivation and the time required to rest-ore receptivity by steroid treatment may be greater i n hypophysectom-ized than in temperature-regressed fi s h . To examine this possibility, I carried out an experiment similar to Experiment 4, except that the interval between hypophysectomy and i n i t i a t i o n of steroid treatment was shorter and the duration of steroid treatment was longer. Two steroids (estradiol and dihydrotestosterone) were tested, each known to be effective i n restoring receptivity in intact, temperature-regressed fish (Experiment 3). 2. Materials and Methods Female goldfish which had been warmed to 20 C several weeks earlier were hypophysectomized and kept in 25% seawater for a further 3 to 4 weeks 49 before being assigned to the steroid or saline treatment groups. Prepara-tion and injection of steroids and procedures for handling and testing fish were as described in Experiment 4, However, the period over which fish received injections was increased to 40 days; injections were given every fourth day for the f i r s t 20 days and on alternate days for the next 20 days. 3. Results and Discussion Of 21 fish which survived the injection schedule, one was too sick to test and 5 had bound eggs. None of the remaining 15 fi s h (saline - 5; es-tradiol - 5; dihydrotestosterone - 5) showed any spawning activity during the three hour observation period. It has been postulated (Experiment 4) that the prolonged steroid depri-vation following hypophysectomy decreased the sensitivity of goldfish target tissues to exogenous steroids; a decrease i n the concentration of steroid receptor proteins may account for this proposed insensitivity. There is evidence that the concentration of estrogen receptor i n the rodent uterus is influenced by estrogen (see review by Milgrom et a l . , 1973). However, i t is doubtful whether the lack of effect of the 40-day treatment i n the present experiment is the result of insufficient steroid priming, as the binding capacity of mammalian tissue increases within hours of exposure to estrogen (Milgrom et al_. * 1973) . It is emphasized, however, that i n the present experiment treatment was not begun until as much as a month after hypophysectomy; the possibility that responsiveness to egg injection might be maintained or restored by earlier treatment has not been examined. The failure of exogenous steroids to restore sexual behaviour of hypo-physectomized female goldfish differs from the effects of similar treatments 50 in female rats (Pfaff, 1970) and guppies (Liley, 1972). As steroids i n -duce receptivity in intact but not i n hypophysectomized goldfish, i t is suggested that some pituitary factor(s) may be involved in the regulation of spawning behaviour. The experiments described in Chapter V were car-ried out to explore this possibility. D. Summary of Chapter IV 1. In female goldfish with 'regressed' (nonvitellogenic) ovaries, injection of ovulated eggs does not induce spawning behaviour. 2. A number of steroids restore responsiveness to egg injection in regressed fish. 3. Hypophysectomy abolishes the spawning response to injection of ovulated eggs. 4. Steroid treatments do not restore responsiveness to egg injection in hypophysectomized fish. CHAPTER, V ROLE OF THE PITUITARY IN THE SPAWNING BEHAVIOUR OF FEMALE GOLDFISH A. Experiment 6. Effect of Pituitary Replacement on the Spawning Behaviour of Hypophysectomized Female Goldfish 1. Introduction In previous experiments, steroid injections restored the sexual be-haviour of intact, temperature-regressed fish whereas similar treatment of hypophysectomized fish was without effect. These results suggested the involvement of the pituitary in the control of responsiveness to egg injec-tion. To examine this possibility, hypophysectomized fish were injected either with estradiol or estradiol plus homogenized goldfish pituitaries, and tested for the abi l i t y to spawn following the injection of ovulated eggs. 2. Materials and Methods Female goldfish were kept at 20 C for 1 to 2 weeks, hypophysectomized (or sham operated), kept a further 2 weeks in 25% seawater, and assigned to one of three treatment groups: (1) pituitary replacement: Females in this group were injected i.p. with fresh macerated goldfish pituitaries (2/day) on alternate days over a 20 day period, during the last 8 of which they also received estradiol (20 ug/g i.p. on alternate days). Preparation of estradiol was as in previous experiments. Injected pituitaries were dissected from freshly k i l l e d donors (males and females; 10-15 g), ground i n a small mortar and pestle, and mixed with a small volume of cold 0.6% NaCl to give an injection volume of 0:2-0.3 ml. As Experiments 4 and 5 showed that estradiol alone does not restore responsiveness to egg injection in hypophysectomized f i s h , a l l fish were primed with estradiol to eliminate the possibility that lack of responsiveness could be due to insufficient estradiol, rather than to the absence of some pituitary factor. (2) saline control: Treatment was as for group (1) except that roughly equivalent volumes of 0.6% NaCl were injected in place of macerated pituitaries. (3) sham control: Females in this group were sham hypophysectomized and treated as in group (2). On the morning of the day following the last injection, each fish was anaesthetized in MS-222, injected with ovulated eggs (0.025 ml/g), f i t t e d with an ovipore plug, and l e f t in a 2 1 beaker of 25% seawater for 40 to 80 minutes before being placed with one or two active males in a 60 1 obser-vation tank. After a three hour observation period the fish were anaesthe-tized, checked for bound eggs, and checked by dissection for pituitary rem-nants. The ovaries were fixed in Bouin's Fluid and prepared for histological examination. 3. Results Sham hypophysectomy had no obvious effect on female goldfish spawning behaviour. Five of seven sham fish spawned at a frequency of normal intact fish (Table IV). SH-3, which spawned only 3 times during the test period, had slightly bound injected eggs and regressed ovaries with very l i t t l e yolk vesicle formation. Even though estradiol was injected, hypophysectomy essentially abo-lished the response to injected eggs. Only 2 of the 8 females tested per-formed any spawning activity and in both fish the response was minimal. Injection of a homogenate of goldfish pituitary restored the spawning behaviour of most of the test females. Furthermore, this treatment induced 54 TABLE TV EFFECT OF INJECTION OF GOLDFISH PITUITARY HOMOGENATE AND ESTRADIOL ON THE SPAWNING RESPONSE TO INJECTION OF OVULATED EGGS IN HYPOPHYSECTOMIZED FEMALE GOLDFISH 55 Group II III Treatment Hypophysectomy 20 day pituitary treatment 8 day estradiol treatment Hypophysectomy 20 day saline treatment 8 day estradiol treatment Sham hypophysectomy 20 day saline treatment 8 day estradiol treatment No. Spawning Acts/3 hour Test Period 56 2 25 2 0 155 72 0 3 57 0 67 19 0 25 0 0 0 12 1 137 2 0 11 111 123 (mean = 42.3) (mean = 58.5) No significant difference between spawning behaviour of pituitary-injected (Group I) and sham hypophysectomized (Group III) fish (Mann-Whitney U-test). 56 yolk deposition ranging from early yolk vesicles to early yolk granules. 4. Discussion The results of this experiment demonstrate a role for the pituitary in the induction of spawning in response to injected eggs, and support YamazakiTs (1965) finding concerning both the time course of second growth stage atresia following hypophysectomy and the induction of yolk formation following pituitary replacement. In the present experiment, 5 of the 8 hypophysectomized controls con-tained no yolk in the ovaries. The slight yolk vesicle deposition seen in the other three fish (one of which spawned twice during the test period) is similar to that of some of Yamazaki's 9-week post-operative fi s h . As in my experiment, Yamazaki apparently did not carry out histological checks for completeness of hypophysectomy and thus there i s a possibility that hypophy-sectomized fish with such minimal yolk deposition may have retained p i t u i -tary fragments. However, considering the high incidence of atresia of second growth phase oocytes in these ovaries and the post-hvpophysectomv persistence of yolk vesicles seen in other experiments, i t seems more li k e l y that post-hypophysectomy regression in these fi s h was incomplete. 57 B. Experiment 7. Effect of Salmon Gonadotropin and Aminoglutethimlde on Spawning Behaviour of Hypophysectomized Female Goldfish 1. Introduction In the preceding experiment, injection of homogenized goldfish p i t u i -tary material into hypophysectomized fish demonstrated a role for the pituitary in spawning behaviour, but provided no information as to the na-ture or the mode of action of the pituitary factor(s) involved, In the pre-sent experiment an attempt was made to test the hypothesis that i t is gonadotropin which i s involved i n the regulation of spawning behaviour by treating hypophysectomized females with a partially purified spring salmon (Oncorhynchus tshawytscha) gonadotropin preparationSG-G100 (gift of Dr. E.M. Donaldson). To examine the possibility that gonadotropin might exert i t s effect through stimulation of steroidogenesis, some gonadotropin-treat-ed fi s h also were injected with the steroid enzyme inhibitor, aminoglutethi-mide (AG, Ciba). As AG inhibits side-chain cleavage of cholesterol (Gaunt et a l . , 1968; Gower, 1974), injection of this chemical should restrict or prevent entry of precursors into the ovarian steroidogenic pathway. If gonadotropin induces spawning behaviour simply by stimulating ovari-an steroid production, then concomitant treatment with AG should block or at least reduce the response; but i f gonadotropin influences behaviour by some mechanism not requiring steroids, then AG should have no inhibitory effect. However, i t is possible that the mechanism regulating spawning be-haviour requires both gonadotropin and steroid. In this case, treatment with AG would also block spawning. To control for the possibility that spawning behaviour i s induced by a combination of gonadotropin and estrogen, 58 a l l fish receiving gonadotropin and AG also were injected with estradiol, which previously had been found not to induce receptivity in hypophysec-tomized fish (Experiments 4, 5, and 6). The major fault in the design of this experiment is that i f only gona-dotropin and estradiol are required to restore responsiveness to egg injec-tion, treatment with AG would appear to have noe effect. If however, a steroid other than estradiol i s required, treatment with AG should block the effect of gonadotropin even though estradiol is present. 2. Materials and Methods' Females were transferred from 12 C to 20 C for several days prior to hypophysectomy, after which they were kept in 25% seawater for 23 to 27 days before beginning treatment. Six treatment groups were used (see Table V): (i) high dose gonadotropin (SG-G100; 15 yg/g) plus estradiol (E 2; 20 yg/g) (i i ) low dose SG-G100 (3 yg/g) plus E 2 ( i i i ) high dose SG-G100 plus AG (100 yg/g) plus E 2 (iv) low dose SG-G100 plus AG plus E 2 (v) control saline (0.6% NaCl) plus E 2 (vi) control E 2 A l l groups received injections (either SG, AG, saline, or E 2) on a l -ternate days over a 20 day period. Tn addition, groups 1 to v were given E 2 on alternate days for the last 9 days of treatment. SG-G100 was administered in two dosages to establish a dose-response relationship for groups i and i i ; i t was thought that this dose effect might also be evidence in groups i i i and iv. Group v served as control for the E„ injections given to groups i to iv and group v i provided an additional TABLE V VEHICLE VOLUMES (PER INJECTION DAY AND TOTAL) INJECTED IN EXPERIMENT 7 Volumes of vehicles (ml) injected per 20 g fi s h on each of f i r s t 5 injection davs second 5 injection days Treatment Group SG AG SAL E 2 Total SG AG SAL E 2 Total Over 10 injection days i , i i .10 .10 .10 .08 .18 1.40 i i i , iv .10 .06 .16 .10 .06 .08 .24 2.00 V .10 .10 .10 .08 .18 1.40 v i .16 .16 .16 .16 1.60 SG = salmon gonadotropin (SG-G100) AG = aminoglutethimide Sal = saline E„ = estradiol o 61 control for the extended E^ exposure assumed to result from SG-G100 stimu-lation of steroidogenesis in groups i and i i . Preparation and injection of was as described previously (Experi-ment 3). SG-G100 was prepared immediately before injection by dissolving the lyophilized hormone in cold 0.6% NaCl to a concentration of 0.6 ug/ul (low dose) or 3.0 yg/ul (high dose). AG was ground to a fine powder, sus-pended at a concentration 30 ug/ul in 0.6% NaCl containing 4 drops Tween 80/100 ml and injected i.p. at a dosage of 100 yg/g. To reduce handling to a minimum, no attempt was made to give equal numbers of injections to a l l groups. However, concentrations of E^ suspen-sions were adjusted so that groups receiving three chemicals at a time (groups i i i and iv) received only 50% more vehicle than group v i which was given only single injections. As indicated in Table V, the differences in total vehicle volumes injected over the duration of the experiment were even smaller. On day 20, each fish was anaesthetized in MS-222, injected with ovula-ted eggs (0.02 ml/g), fitt e d with an ovipore plug, and l e f t in a 2 1 beaker of 25% seawater for 60-75 minutes before being placed with one or two ac-tive males (one female per 60 1 observation tank). The fis h were observed continuously for three hours,, anaesthetized, checked for bound eggs, and sacrificed. The heads were checked by dissection for completeness of-hypo^-physectomy, and ovaries were fixed and prepared for histological examina-tion. 3. Results As a result of mortality and binding of injected eggs, the fin a l 62 sample sizes in this experiment were small. The data are presented in Table VI. As in previous experiments, estradiol failed to restore receptivity in hypophysectomized fish (Group v); one of four fish showed a low level of response. It is not clear whether the response of this fish was due to residual receptivity or to the estrogen treatment. No completely hypophy-sectomized group v i (estradiol-treated) fish survived to be tested. At both the high and low dosages (Group 1 and i i ) , salmon gonadotropin was effective in restoring the response to injected eggs, a l l 7 fi s h re-sponding in the behavioural tests. In addition, the gonadotropin had a marked effect on vitellogenesis, ranging from the induction of the early yolk vesicle stage in most fis h , to the early yolk granule stage in one female. These effects on oocyte development are consistent with those re-ported by Yamazaki and Donaldson (1968) ; the greater degree of ovarian re-sponse i n their study l i k e l y was due to higher gonadotropin dosage and healthier fish. Except for one degenerating oocyte in one f i s h , gonadotro-pin treatment also prevented the degeneration of f i r s t and second growth stage oocytes. This is in contrast to the results of estradiol treatment (groups v and vi) where 8 of 10 f i s h had desperating oocytes. Administration of aminoglutethimide was highly effective in inhibiting spawning i n response to egg injection in hypophysectomized, gonadotropin' treated fish (Group i i i and i v ) ; only one of six fish responded i n the be-havioural tests. However, the ovarian histology of fish treated with gona-dotropin and aminoglutethimide was similar to that of the fish receiving gonadotropin alone; yolk vesicles were present in a l l ovaries and atresia of oocytes was inhibited completely. 63 TABLE VI EFFECT OF SALMON GONADOTROPIN (SG-G100) AND AMINOGLUTETH1MIDE ON THE SPAWNING RESPONSE TO EGG INJECTION IN HYPOPHYSECTOMIZED FEMALE GOLDFISH Group i i i M i iv v v i Treatment SG - 15 yg/g First 10 days Second 10 days SG - 15 yg/g E - 20 yg/g SG - 3 yg/g SG - 3 yg/g E 2 - 20 yg/g SG - 15 yg/g AG - 100 yg/g SG - 15 yg/g AG - 100 yg/g E - 20 yg/g SG - 3 yg/g AG - 100 yg/g SG - 3 yg/g AG - 100 yg/g E - 20 yg/g saline E 2 - 20 yg/g saline E - 20 yg/g E 2 - 20yg/g Z No• spawning acts/3 hour test period 116 61 6 38 72 36 18 0 0 0 44 0 0 0 0 5 no fish tested SG = salmon gonadotrpin (SG-G100) AG = aminoglutethimide E 2 = estradiol 4> - vn -I LEAF 65 OMITTED IN PAGE NUMBERING. 66 4. Discussion The results of this experiment demonstrate that in hypophysectomized female goldfish, treatment with a partially purified salmon gonadotropin preparation (SG-G100) combined with estradiol (I^) is effective in restor-ing spawning in response to the injection of ovulated eggs. As in previous experiments, treatment with alone was ineffective. These findings sug-gest that gonadotropin is the pituitary factor responsible for restoring receptivity in the hypophysectomized fish injected with goldfish pituitary homogenate in Experiment 6. However, as the gonadotropin preparation used in this experiment l i k e l y contained some thyrotropic activity (Donaldson and McBride, 1974), i t is possible that i t was thyrotropin rather than, or i n addition to, gonadotropin which restored receptivity. When hypophysectomized fish receiving SG-G100 and E^ also were injected with the steroid enzyme inhibitor, aminoglutethimide (AG), the response to injections of eggs was strongly inhibited. This inhibition of spawning be-haviour by AG may simply have been pharmacological and independent of any effects on steroidogenesis; however, neither the non-spawning behaviours nor the general condition of AG-treated fish were obviously different from those of the other groups. If AG affected spawning by inhibiting steroidogenesis, then i t appears the effect was due to inhibition of the production of an essential steroid (or steroids) other than estradiol. Whether or not AG was administered, SG-G100 stimulated yolk production. This is to be expected, as gonadotropin and estradiol were present in a l l cases. C. Summary of Chapter V 1. Injection of homogenized goldfish pituitaries or partially puri-fied salmon gonadotropin into hypophysectomized female goldfish restores responsiveness to injection of ovulated eggs. 2. In hypophysectomized fish treated with salmon gonadotropin and estradiol, injection of the steroid enzyme inhibitor, aminoglutethimide, inhibits responsiveness to injection of ovulated eggs, suggesting a steroid other than estradiol may play a role in spawning behaviour. CHAPTER VI THE ROLE OF PROSTAGLANDINS IN THE SPAWNING BEHAVIOUR OF FEMALE GOLDFISH 69 A. Introduction The majority of the preceding experiments were designed to examine the effects of various hormones on the spawning responses of female gold-fish to the injection of ovulated eggs. These studies have been concerned with relatively long-term effects of hormonal deprivation and replacement over periods of days or weeks. In this section, the short-term regulation of sexual behaviour is examined in experiments involving the mechanism by which egg injection triggers spawning. In his study of the spawning behaviour of goldfish, Yamazaki (1965) noted that even though a female had ovulated, spawning would not occur un-less both aquatic vegetation and a sexually active male were present. Re-moval of the vegetation inhibited spawning immediately, while replacement restored i t within minutes. My observations also show that i f an ovulated female, isolated from other fish for days, is placed with a sexually active male, spawning can occur within a minute. However, a consistent feature of spawning induced by egg injection is that there is a variable latent period (as much as several hours) between the time eggs are injected and the time spawning begins. Although this latent period may be nothing more than an inhibition induced by anaesthetization and handling during egg i n -jection, some preliminary data suggest this i s not so. Seven highly receptive fish with second growth phase oocytes were anaesthetized and injected with ovulated eggs: two females (#1,2) were placed with active males immediately, four (#3-6) were placed with the males after 200 minutes 'incubation' in isolation from other f i s h , and one (#7) was placed with males following a second anaesithetization at the end of the 200 minute incubation (Table VII). In the fish placed with males 70 TABLE VII EFFECT OF INCUBATION TIME (DELAY BETWEEN EGG INJECTION AND PLACEMENT OF FEMALE WITH MALE) ON LATENCY TO FIRST SPAWNING ACT IN RECENTLY OVULATED (3-to 7-DAY BOSTOVULATORY) FEMALES. 71 Fish Incubation Latency to No. spawning acts (min) f i r s t in f i r s t 40 min spawning act of spawning 1 - POV-7 10 37 52 2 - P0V-4 4 55 25 3 - POV-7 200 3 74 4 - POV-7 200 9 19 5 - P0V-3 200 2 118 6 - POV-4 200 9 52 7 - P0V-3 200* 12 66 *Fish #7 received a second anaesthetization of 200 minutes. 72 immediately after recovery from anaesthetic, the latency to the f i r s t spawning act was considerably greater than that in fi s h incubated for 200 minutes. This was true even when an incubated fish was anaesthetized a second time. These limited results suggest that eggs in the ovarian lumen stimulate spawning by a different mechanism than do the presence of male goldfish or aquatic vegetation. Whereas presentation of the latter stimuli e l i c i t s spawning within a few minutes in fish which nay'e had ovulated eggs in the ovarian lumen for some time, injection of ovulated eggs is followed by a latent period which apparently is not due simply to anaesthetization. Of course, the second anaesthetization at 200 minutes i s a poor control as i t indicates only that MS-222 has l i t t l e effect on the behaviour of a fish which is already receptive. It was thought that the latent period might represent the time re-quired for oviduct stimulation to effect changes in central neural struc-tures controlling sexual behaviour. As a number of workers have reported that injection of neurohypophyseal hormones induces a 'spawning reflex', oviposition, or parturition i n a variety of teleost species (Liley, 1969; Macey et a l , , 1974), i t seemed possible that these hormones might be i n -volved i n mediating the effects of egg injection on the spawning behaviour of female goldfish. No formal experiments were carried out to test this hypothesis as i n a l l preliminary tests oxytocin was completely ineffective in inducing spawning behaviour in female goldfish with vitellogenic oocytes or in increasing the rate of spawning in ovulated f i s h . Similar results have apparently been obtained by Ptckford (unpublished results cited in Macey e_t a l . , 1974) . 73 In female mammals, physical stimulation (stretch) of the reproduc-tive tract and other smooth muscles (situations roughly comparable to ovulation o,r egg injection i n goldfish) is associated with the release of prostaglandins (Poyser et a l . , 1971; Piper and Vane, 1971; Csapo, 1973). Prostaglandins (PG), the 'intrinsic uterine stimulant' (Csapo, 1973), have been implicated in the oviposition of birds (Hertelendy, 1972, 1973; Heftelendy et a l . , 1974, 1975; Wechsung and Houvenaghel, 1976), in the parturition of mammals (Liggins et a l . , 1973; La'bhsetwar, 1974; Aiken, 1974; Flint et a l . , 1974; Currie, 1975; Umo et a l . , 1976), and in the mechani sm of action of some IUDs (Spilman and Duby, 1972; Chaudhuri, 1975). The findings that steroids may alter the release of PG from the genital tract (Roberts et al_., 1975) , that the response o.f. the genital tract to PG may be modified by steroids (Csapo, 1973; Spilman, 1974), and that PG may play a role in the action of LH on ovarian tissue (Kuehl et a l . , 1970; Marsh et a l . , 1974) a l l suggested that investigation of PG function in the female goldfish might shed light on the problem of the induction of spawn-ing behaviour by egg injection. The following experiments present the re-sults of these investigations. 74 B. Experiment 8. Inhibition by Indomethacin of Spawning Behaviour Induced by Injection of Ovulated Eggs 1. Int ro due tion It was postulated that i f PGs were involved in goldfish spawning behaviour, their synthesis or release may be increased by the stimulus provided by egg injection. In female mammals, endogenous PG production can be inhibited by injection of indomethacin (IM), a potent inhibitor of PG synthesis which apparently affects a complex of enzymes referred to as prostaglandin synthetase (Vane, 1971). As a f i r s t step i n demon-strating a role for PG in goldfish spawning behaviour, IM injection was used to inhibit endogenous PG production. To examine the speed and dura-tion of action of IM, injections were given at various'times in relation to the injection of eggs and the onset of spawning behaviour. 2. Materials and Methods A l l female goldfish used in this experiment had ovaries i n various stages of vitellogenesis and therefore could have been expected to spawn when injected with ovulated eggs. Four treatment groups were used (see Table VIII). Indomethacin (IM, Sigma) was injected i.p. as a saline sus-pension (0.6% NaCl; 4 drops Tween 80/100 ml) at a dosage of 10 ug/g (5yl/g). This dosage had been effective in a study of goldfish ovulation (Stacey and Pandey, 1975). At 1000 h on the test day, fish were anaesthetized in MS-222, weighed and injected with ovulated eggs (0.02 ml/g). They were then l e f t for 1 hour, placed i n 60 1 observation tanks with actively courting males, and observed continuously for three hours. Females in Group i had received TABLE VIII EFFECT OF INDOMETHACIN ON SPAWNING IN RESPONSE TO INJECTION OF OVULATED EGGS 76 Group Treatment N No. Mean • No. Spawning Respond- Acts* (range) ing i IM 10 h before eggs i i IM with eggs i i i saline with eggs IM after 20 minutes spawning iv saline with eggs 7 5 Q 9 0 0 9 1 8 0+ 0+ 6-H- (2-11) (!)+++ 5-H- (1-9) saline after 20 minutes spawning 19+4- (5-37) * Based on responding individuals. + Mean no. per 3 h test period ++ Mean no. per 20 min +++ Mean no. after 2nd injection IM = indomethacin 77 IM at 2400 h the previous evening, 10 hours before egg injection, while those in Group i i had received IM while anaesthetized for egg injection. Groups i i i and iv received saline injections coincident with egg injec-tions as a control for the IM injection given to Group i i . In addition, Group i i i received an IM injection and Group i v a second saline injection 20 minutes after spawning had commenced. For this second injection fish were not anaesthetized, but simply netted from the observation tank, i n -jected while hand held, and replaced immediately. 3. Results and Discussion As shown in Table VIIT, none of the females injected with IM 10 hours priord to, or coincident with, injection of ovulated eggs spawned during the 3 hour test period; in contrast, a l l females injected with saline at the time of egg injection (Groups i i i and iv) spawned in the test period. Following IM injection, one Group i i i female performed one spawning act (5 minutes post-injection), whereas a l l group iv females continued to spawn following the second control saline injection. Injection of an IM suspension is thus seen to have a rapid and rela-tively long-lasting inhibitory effect on the spawning response to injected eggs. (In contrast, courting males, which were observed for up to three hours post-injection, showed no obvious behavioural response to the injec-tion of IM; however, no quantitative measures were taken. It i s not known whether treatment with IM inhibits the release of sperm). 78 C. Experiment 9. Effect of Prostaglandins on the Spawning Behaviour of Indomethacin-Treated Female Goldfish 1. Introduction This experiment tested the a b i l i t i e s of injected prostaglandins (PGE^, P^E2' ^^2^^ t 0 o v e r c o m e t^ i e indomethacin-induced inhibition of spawning following egg injection. These three PGs were chosen as they were found to be effective in overcoming the indomethacin blockade of ovu-lation in female goldfish (Stacey and Pandey, 1975). 2. Materials and Methods Five treatment groups were used and the i n i t i a l treatment for a l l was as for Group i i in the preceding experiment. Fish were anaesthetized, weighed, and injected with ovulated eggs (0.02 ml/g) and IM (10 yg/g) be-tween 900 and 1000 h on the test day. A l l fi s h were then l e f t for one hour to recover, netted, injected i.p. with the saline vehicle (Group i) or one of the PGs (Groups i i - v ) , placed in the observation tanks with ac-tive males, and observed continuously for 2 hours. PGE^ (Group i i ) and PGE2 (Group i i i ) were prepared by dissolving in 95% ethanol (1 mg/0.1 ml) and diluting 9X with buffered saline (20 mg Na2C03/100 ml 0.6% NaCl). PGF^ (Group iv) was simply dissolved in buffered saline (1 mg/ml). Group v received PGF,^ prepared in the same manner as PGE^ and PGE2 as a control for possible behavioural effects of the ethanol solvent. A l l PGs (Upjohn) were injected i.p. at a dose of 5 yg/g (5 yl/g). 3. Results and Discussion None of the Group i fish treated with the saline vehicle spawned during the 2 hour test period (Table IX). Though the sample size was 79 TABLE IX EFFECT OF PROSTAGLANDINS ON THE SPAWNING RESPONSE TO INJECTION OF OVULATED EGGS.IN FEMALE GOLDFISH TREATED WITH INDOMETHACIN 80 Group Treatment N No. r e - Mean No. Spawning sponding Acts* (range) i IM, eggs and s a l i n e 6 0 0 i i IM, eggs and PGE 1 5 0 0 i i i IM, eggs and PGE 2 7 2 5 ( 4 , 6 ) i v IM, eggs and PGF^ 7 5 27 (7-62) V IM, eggs and PGF-(alcohol v e h i c l e ; 4 3 52 (5-115) * Based on i n d i v i d u a l s responding i n 2 h test period. Indpmetnaeini;,(IMi)i and eggs i n j e c t e d 1 h before t e s t , PGs injec t e d at s t a r t of te s t 81 small, PGE^ (5 fish) had no effect i n overcoming the indomethacin block-ade, and PGE2 was only slightly active, inducing low levels of spawning in 2 of the 7 fish tested. In contrast, PGF^x was quite effective i n inducing spawning behaviour in IM-treated fish ; presence of ethanol in the vehicle does not seem to account for the ineffectiveness of PGE^ and PGE^, as 3 of the 4 Group V fish spawned normally. The latencies from the injection of PG to the onset of spawning were quite variable (5-100 minutes; mean 40 minutes). There were no s i g n i f i -cant differences among the latencies of the responding fi s h in groups i i i , i v , and v. Of the 8 fish which spawned following PGF20Q injection, a l l but one continued to spawn for the remainder of the observation period. 82 D. Experiment 10. Effect of Prostaglandins Alone (Without Egg Injec-tion) on the Spawning Behaviour of Female Goldfish 1. Introduction Based on the findings that indomethacin (IM) blocked the spawning re-sponse to injected eggs, and that PGs (especially PGF2a:) were effective in overcoming this blockade, this experiment was conducted to determine whether PG injection alone (i.e., without prior treatment with ovulated eggs) was sufficient to induce spawning behaviour, and whether the order of PG potenciess was as found in Experiment 9 (PGF^ > PGE2 > PGE^ . 2. Methods, Results, and Discussion Female goldfish were injected i.p. (no anaesthetization) with either PGF„ , PGE.. , or PGE_ at 5 yg/g, placed immediately with active males, and 1 2 observed continuously.for 3 hours. No control group was used. As shown in Table X, PGF2co was highly effective in inducing spawning behaviour even though ovulated eggs were not present in the ovarian lumen. Several of the fish i n this group were injected with IM 20 minutes after spawning had commenced; there was no apparent effect on spawning behaviour, PGE2 was marginally effective in inducing spawning behaviour (2 o.f 10 fish responding) and PGE^ was without effect. As in Experiment 9, the latencies to the onset of spawning were quite variable (5-135 minutes; mean 45 minutes). Also, the duration of spawning in the fish responding to PGF. was considerable, several individuals re-sponding regularly for longer than 2 hours. These results suggest that the effect which ovulated eggs in the ovari-an lumen have "an spawning behaviour is mediated through the release of TABLE X EFFECT OF PROSTAGLANDINS ON THE SPAWNING BEHAVIOUR OF FEMALE GOLDFISH WITHOUT OVULATED EGGS TN THE OVARIAN LUMEN 84 Group Treatment N No. Re- Mean TCHQ. Spawning sponding Acts* (Range) i PGE^ 10 0 0 i i PGE2 10 2 11 (7,15) i i i PGP. 13 13 59 (11-208) Based on individuals responding i n 3 h test period. PGs injected at start of test. 85 prostaglandins, and that i t i s some action of prostaglandins, independent of the presence of eggs, which induces spawning i n response to male courtship. 86 E. Experiment 11. Effect of Hypophysectomy and Gonadotropin Replacement on Prostaglandin-lnduced Spawning Behaviour in Female Goldfish 1. Introduction In experiments presented in Chapters IV and V, treatments with p i -tuitary and ovarian hormones were found to affect the responsiveness of female goldfish to egg injection. A variety of steroid hormones were effective in inducing receptivity in intact, temperature-regressed fi s h , whereas in hypophysectomized fish, a l l steroids tested were found to be totally ineffective. Treatment with goldfish pituitary homogenate or a part i a l l y purified salmon gonadotropin preparation (SG-G100) was highly effective in restoring spawning behaviour in hypophysectomized fish. As the injection of the steroid enzyme inhibitor, aminoglutethimide, inhibited the response to SG-G100, i t was suggested that the mode of action of gona-dotropin on spawning behaviour may be through the stimulation of ovarian steroidogenesis. A number of interpretations of these results is possible. In many mammalian studies, prostaglandin release has been shown to be correlated with reproductive cycles and to be influenced by gonadotropin and steroids (see references i n section A, this chapter). As PG is ap-parently essential for spawning behaviour to occur, i t seemed a plausible explanation that the in a b i l i t y of steroids to restore receptivity in hypo-physectomized fish is due to the absence of PG synthesis following hypophy-sectomy. The following experiment examined this hypothesis by comparing the response to PGF^ injection i n hypophysectomized fish treated either with steroids, salmon gonadotropin, or a saline vehicle. 87 2. Materials and Methods Females which had been hypophysectomized 3 to 4 months earlier were injected i.p. (no anaesthetization) as follows: Group i - saline vehicle (0.6% NaCl; 4 drops Tween 80/100 ml), Group i i - 10 yg/g spring salmon (Oncorhynchus tshawytscha) gonadotropin (SG-G100, lot #BCR-3) dissolved i n saline, Group i i i ^ 20 yg/g 17$-estradiol suspended in saline. A l l treatment groups received injections at 5 yl/g on alternate days for a 15 day period. On the morning of the day following the last injection, fish were injected i.p. with 5 yg/g PGF^, placed immediately with actively courting males, and observed continuously for 3 hours. Fish in groups i and i i were then sacrificed and the heads examined by dissection for p i t u i -tary remnants. Ovaries were fixed and prepared for histological examina-tion . To examine the possibility that gonadotropin-induced receptivity was due to steroids other than estradiol, Group i i i was divided into two groups; Group i i i A received estradiol injections on alternate days for a further 15 day period, while Group i i i B received injections of a mixture of ster-oids (estradiol, 20 yg/g; testosterone, S^dihydrotestosterone, 11-keto-testosterone, progesterone, and deoxycorticosterone a l l at 5 yg/g) for the same period of time. Testing and related procedures following this second injection period were the same as those following the f i r s t . 3. Results Injection of PGF^ failed to induce spawning behaviour in female goldfish which had been hypophysectomized for 3 to 4 months; a 2 week 88 pretreatment with salmon gonadotropin restored PG-induced spawning be-haviour i n 8 of 13 hypophysectomized fish tested (Table XI). Treatment .with estradiol or a combination of steroids was without effect in restor-ing the responsiveness to PG, only 1 of 12 fish showing a low level of spawning in each testing session. As in the two preceding experiments, the latency to spawning was variable (20-160 minutes; mean 60 minutes) and the duration of spawning was often considerable, 4 of the 8 responding fish in Group i i spawning for longer than 2 hours. Histological examination of the ovaries revealed no yolk deposition in any Group i or Group i i i fish. The ovaries of the 6 most active spawn-ers in Group i i contained small numbers of oocytes in the early yolk vesi-cle stage. Of the 5 non-responding Group i i f i s h , one had ovaries with early yolk vesicles, three had no yolk vesicle stage oocytes, and the ovaries of one contained only a^-stage corpora lutea (Khoo, 1975). No oocytes containing yolk granules were observed. The greater ovarian re-sponse to SG-G100 reported for hypophysectomized goldfish by Yamazaki and Donaldson (1968) may have been due either to a longer injection schedule (3 weeks rather than 2) or a shorter period following hypophysectomy (2 months rather than 3 or 4). SG-G100 was also effective in preventing oocyte degeneration: where-as 13 of 16 Group I and 8 of 12 Group i i i fish had ovaries containing de-generating previtellogenic oocytes, these were found in only 3 of 13 Group i i fish. 4. Discussion In hypophysectomized fish, salmon gonadotropin restores the spawning TABLE XI EFFECT OF SALMON GONADOTROPIN (SG-G100) AND STEROIDS ON SPAWNING BEHAVIOUR IN RESPONSE TO INJECTION OF PROSTAGLANDIN F ^ IN HYPOPHYSECTOMIZED FEMALE GOLDFISH 90 Group Treatment N No. re- Mean i.oNo. Spawning sponding Acts* (range) i saline 16 9 , 0 i i salmon gonado-tropin 13 8 26 (5-53) i n estradiol 12 1 (3) i i i A estradiol 6 1 (5) i i i B steroid mixture 6 0 0 Based on individuals responding in 3 h test period PGs injected at start of test 91 response to prostaglandin (PGF^) i n j e c t i o n , while treatment with e s t r a d i o l or a mixture of steroids has no e f f e c t . The lack of e f f e c t of e s t r a d i o l i s not l i k e l y to be due to an i n h i b i t o r y e f f e c t of high s t e r o i d dosage, as the same dosage given to i n t a c t regressed f i s h (Ex-periment 3) and with salmon gonadotropin to hypophysectomized f i s h (Ex-periment 7) d i d not i n h i b i t spawning. The r e s u l t s of this experiment demonstrate that responsiveness to PG i s modulated by gonadotropin; however, they do not support the hypothe-s i s that the ineffectiveness of steroids on the responsiveness of hypo-physectomized f i s h to egg i n j e c t i o n i s due to the absence of PG. In hypophysectomized f i s h treated with salmon gonadotropin or s t e r o i d s , the e f f e c t s of PG on spawning behaviour are s i m i l a r to the e f f e c t s of egg i n -j e c t i o n , suggesting that f i s h which f a i l to respond, to eggs simply may not respond to endogenous PG, 92 F. Summary of Chapter VI 1. Injection of indomethacin, an inhibitor of prostaglandin synthe-sis, blocks the effect of ovulated eggs on spawning behaviour. 2. Injection of prostaglandins restores spawning behaviour i n egg-injected fish treated with indomethacin and induces spawning i n fish which have not been injected with ovulated eggs. 3. The spawning response to prostaglandin injection i s abolished by hypophysectomy and restored in hypophysectomized fish by treatment with salmon gonadotropin. Steroid treatment i s ineffective in restoring the response to prostaglandin injection in hypophysectomized fish. CHAPTER VII GENERAL DISCUSSION 94 A. Introduction This study not only raises a number of questions as to the nature of the roles of the hormones examined; i t also suggests fascinating problems concerning the evolution of the regulation of sexual behaviour in female vertebrates in general. There is much scope for speculation on the mechanism which regulates spawning behaviour in female goldfish. However, i t must be acknowledged that a l l conclusions derived from this study are biased to the extent that they are based on results obtained in an experimental situation de-signed to measure only oviposition behaviour. As detailed by Beach (1976), hormones may influence three basic aspects of female reproductive be-haviour; (i) attractivity - measured in terms of the appetitive sexual res-ponses evoked in conspecific males, ( i i ) proceptivity - appetitive sexual behaviours evoked in females by males, and ( i i i ) receptivity -the consummatory phase of the mating sequence. As mentioned above (page 16 ), in situations where the male goldfish is relatively sexually inactive, females may exhibit proceptive behaviour including components which would f a l l into the categories Beach (1976) refers to as a f f i l i a t i v e , s o l i c i t a t i o n a l , approach-withdrawal, and contact responses. One aspect of non-behavioural stimuli contributing to the attractivity of female gold-fish consists of unidentified substances which apparently are released by the ovaries of preovulatory and ovulated fish and function as olfactory stimulants of appetitive sexual responses in males (Partridge et al.,1976). The present study has dealt solely with the endocrine control of the spawning act, the consummatory phase of female goldfish sexual behaviour; 95 i t is possible that had data been gathered on the effect of hormones on the stimulus quality of females or on the tendency of females to approach males or engage i n other types of proceptive behaviour, the results and conclusions might have been different. The results of experiments presented in this thesis identify four endogenous factors believed to play major roles in spawning behaviour; (i) pituitary hormones (apparently gonadotropin), ( i i ) ovarian steroids, ( i i i ) stimuli from ovulated eggs in the oviduct, and (iv) prostaglandins. As so l i t t l e i s known of the endocrine regulation of female reproductive behaviour in other teleostean species, speculation concerning the function of these four factors in the control of spawning behaviour in goldfish is based largely on information from studies of reproduction and sexual behaviour in higher vertebrates, particularly mammals. B. The Role of the Pituitary The results of hypophysectomy and pituitary replacement therapy demon-strate that the pituitary plays at least an indirect role in regulating spawning behaviour in female goldfish. That pituitary gonadotropin i s involved in this regulation is suggested by the effectiveness of salmon gonadotropin in restoring receptivity in hypophysectomized fi s h . It is possible that the effect of salmon gonadotropin on spawning behaviour was due to, or enhanced by, contamination of the gonadotropin preparation with thyrotropin. Donaldson and McBride (1974) found evidence that salmon gonadotropin injection in adult salmon led to increased thyroid activity, an effect which they attributed to thyrotropic activity in the gonadotropin preparation. It is not clear how thyrotropin or thyroxin might be involved in stimulating receptivity. If these hormones were 96 required to mediate the e f f e c t of steroids on behaviour, then the lack of e f f e c t of steroids i n hypophysectomized f i s h would be understandable. However, there apparently i s no evidence that thyrotropin or thyroxin en-hances the action of steroids on target issues. It i s not known how gonadotropin restores responsiveness to egg i n j e c t i o n i n hypophysectomized f i s h . However, i t i s suggested that e i t h e r or both of two basic mechanisms may be involved: 1. gonadotropin may d i r e c t l y a f f e c t central neural structures c o n t r o l l i n g sexual behaviour, or 2. gonadotropin may exert an i n d i r e c t control over behaviour by stimulating the formation of s t e r o i d hormones which i n turn act d i r e c t l y on the mechanism regulating sexual responsiveness. There i s no d i r e c t evidence from t h i s or other studies of teleosts that gonadotropin exerts a d i r e c t control over female sexual behaviour. However, i t has been suggested that i n male Gasterosteus aculeatus (Hoar, 1962; Baggerman, 1966) and Cymatogaster aggregata (Wiebe, 1967) gonadotropin may a f f e c t behaviour associated with reproduction through a mechanism independent of the gonads. Gonadotropin has been suggested to exert a d i r e c t control over sexual behaviour i n the female guppy ( L i l e y , 1968); more recent evidence ( L i l e y and Donaldson, 1969; L i l e y , 1972) f a i l s to provide support for t h i s hypothesis. There appears to be no evidence that gonadotropins enhance sexual behaviour i n female mammals except through stimulation of steroidogenesis. In ovariectomized r a t s , hypophysectomy does not a f f e c t l o r d o s i s induced by high estrogen dosage ( P f a f f , 1970) and f a c i l i t a t e s female sexual behaviour when threshold dosages of estrogen are employed (Crowley et al.,1976). 97 As injection of luteinizing hormone supresses lordosis responses in hypophysectomized-ovariectomized individuals, Crowley e_t al.(1976) suggest •'that this gonadotropin may affect sexual behaviour by inhibiting the release of luteinizing hormone-releasing hormone (LH-RH). LH-RH stimulates lordosis in the female rat (Pfaff, 1973; Moss and Foreman,1976). There is some evidence that gonadotropin induces receptivity in female goldfish by stimulating steroidogenesis. As estrogen enhances yolk formation in other teleosts (Campbell and Idler, 1976; Emmersen and Petersen, 1976), i t is likely that salmon gonadotropin, which induces yolk formation i n hypophysectomized goldfish (Yamazaki and Donaldson, 1968) , also stimulates the production of estrogen and other steroids. Thus, the ab i l i t y of aminoglutethimide to inhibit spawning behaviour of salmon gonadotropin-treated goldfish suggests that gonadotropin influences recept-i v i t y by stimulating steroidogenesis. A similar mechanism has been proposed to explain the effect of salmon gonadotropin on sexual behaviour in the female guppy (Liley and Donaldson, 1969). In addition to inhibiting side-chain cleavage of cholesterol (Gaunt et a l . , 1968; Gower, 1974), aminoglutethimide has been shown to inhibit aromatization (Thompson and S i i t e r i , 1973) and to block the stim-ulatory effect of testosterone on sexual behaviour in the male rat (Beyer e_t a l . , 1976). It is not li k e l y that aminoglutethimide suppressed spawning behaviour by inhibiting aromatization, as a l l fish receiving gonadotropin and aminoglutethimide were also injected with estradiol. The fact that estradiol did not overcome the inhibitory effect of aminoglute-thimide suggests that spawning behaviour may be regulated by steroids other than (or in addition to) estradiol. 98 The possibility that the effect of aminoglutethimide on spawning behaviour is pharmacological, and not related to an effect on steroido-genesis, cannot be excluded. In mammals, aminoglutethimide affects the synthesis of thyroid (Rallison e_t a l . , 1967) and adrenocortical hormones (Philbert et a l . , 1968) and depresses brain activity (Elazar and Blum, 1971). However, neither i n the present study nor in that of Beyer et a l . (1976) did aminoglutethimide treatment produce any obvious changes in non-reproductive behaviour. Salmon gonadotropin restores receptivity in hypophysectomized female guppies but not in fish which have also been ovariectomized (Liley and Donaldson, 1969) , suggesting that the hormone exerts i t s effect on behaviour by stimulating ovarian steroidogenesis. The subsequent finding that estrogen treatment induces receptivity in hypophysectomized female guppies (Liley, 1972) demonstrates that in this teleost gonadotropin is not essential for the expression of sexual behaviour. As i n the guppy, steroid hormone restores receptivity in the hypophysectomized female rat (Pfaff, 1970) ; in fact, hypophysectomy has been found to enhance behavioural res-ponsiveness to estrogen treatment in the rat (Crowley et a l . , 1976), possibly by removing short-loop feedback inhibition of LH-RH release (see Kuhl and Taubert, 1975). In female goldfish, the results of steroid treatment of intact, re-gressed fish and of aminoglutethimide injections in hypophysectomized, gonadotropin treated individuals indicate that steroids may be involved in regulating spawning behaviour. However, the failure of steroid treatments alone to restore receptivity of hypophysectomized fish suggests that, in contrast to the situation in rats and guppies, gonadotropin may play an 99 indispensable role in the sexual behaviour of female goldfish. Whether this apparent requirement for gonadotropin is of physiological s i g n i f i -cance, or simply results from inappropriate steroid therapy applied to hypophysectomized fish, cannot be determined without further study. C. The Role of Steroids 1. Introduction This study provides the f i r s t demonstration that steroids stimulate spawning behaviour i n an oviparous female fish* In intact goldfish with regressed, nonvitellogenic ovaries, a variety of steroids restores spawning behaviour in response to injection of ovulated eggs; hypo-physectomy abolishes this effect of steroids on behaviour. These findings suggest that the pituitary i s involved i n the effects of steroids in intact f i s h , but do not clarify the relative contribution of steroid and pituitary hormones to the regulation of spawning behaviour. On the basis of the results presented in this study, i t is not possible to determine whether the induction of spawning behaviour results from the combined action of pituitary and ovarian hormones, or requires endocrine input from only one of these organs. In the following discussion, speculations concerning the role of steroids in goldfish spawning behaviour are based on the effects of steroids on reproductive physiology and behaviour of other female verte-brates. Therefore, i t is emphasized at the outset that the relationship between ovarian development and the onset of sexual receptivity in goldfish appears to be unique. The onset of estrous behaviour in mammals has been attributed to 100 periovulatory fluctuations in steroid hormones (Davidson and Levine, 1972); also in the guppy, female sexual behaviour normally occurs when the ovary is in an advanced stage of development (Liley, 1968). In female goldfish, however, spawning behaviour may be induced in the absence of endocrine stimuli associated with the later stages of ovarian development; fi s h with partially regressed ovaries (containing minimal yolk vesicle formation) perform normal spawning behaviour when injected with ovulated eggs. It could be argued that this finding in goldfish i s analogous to the results of experiments i n female rats in which high levels of sensory stimulation (manual palpation of flanks and perineum) induce lordosis responses i n the absence of or with low doses of estrogen (see discussion by Crowley et al_. , 1976). However, the fact that normal sensory stimuli (presence of ovulated eggs in the genital tract, courtship of male goldfish) induce spawning behaviour at any time during an extended period of receptivity (when yolk vesicles or granules are present i n the oocytes) raises the possibility that, i f steroids are involved in regulating spawning behaviour, the role of these hormones may be fundamentally different from the role they play i n reproductive behaviour of other female vertebrates. 2. Effects of Steroids in Intact Female Goldfish Two obvious explanations could account for the diversity of steroids which induce spawning behaviour i n intact, regressed goldfish. The mechanism controlling ;receptivity may respond to a variety of steroids. Alternatively, steroid conversion may occur in regressed fish ; some of the steroids shown to restore responsiveness to egg injection may be metabolized to one or more active forms which influence behaviour. 101 .There is no published information to indicate whether mechanisms con-tr o l l i n g female sexual behaviour in other species of oviparous fish are sensitive only to certain steroids. However, in the viviparous female guppy, only estrogens have been shown to induce receptivity (Liley, 1972). In this species, treatment with estradiol, estrone, e s t r i o l , or diethylstilboestrol restores sexual behaviour while C o r t i s o l , corticosterone and progesterone are without effect; females treated with testosterone exhibit male sexual behaviour. The results of mammalian studies indicate that, where conversion of androgens to estrogens can be ruled out, sexual behaviour i n females is induced only by estrogens (progesterone is also required in some species to f a c i l i t a t e the priming effects of estrogen:,- Ciaccio and Lisk, 1967; Joslyn et a l . , 1971 ). Dihydrotestosterone, which cannol be aromatized, f a i l s to induce estrous behaviour i n female rabbits (Beyer e_t al_. , 1970) or to increase sexual motivation in female rats (McDonald and Meyerson, 1973). On the other hand, a number of estrogens (estradiol, estrone, estriol) are capable o-f'inducing receptivity in rats (Beyer et a l . , 1971) and guinea pigs (Feder and Silver, 1974). As es t r i o l apparently is not converted to estradiol (Ruh et a l . , 1973), i t appears that more than one naturally occurring estrogen may directly influence sexual behaviour in female mammals. Sexual behaviour i n female mammals and i n the female guppy appear to be induced specifically by estrogens. Therefore i t is unlikely that the effectiveness of the wide variety of steroids (estradiol, androgens, 17*?; -OHr-pregnenolone, and possibly pregnenolone) which restores responsive-ness to egg injection i n female goldfish i s due to a lack of specificity in the response'to steroids. Rather, the effects of steroids in goldfish are 102 more consistent with the concept that many of the exogenous steroids are converted in vivo prior to affecting behaviour. In some mammalian studies (e.g. Beyer e_t al_. , 1970; Whalen et a l . , 1972), the effects of non-estrogen steroids on female sexual behaviour have been attributed to in vivo conversion of the exogenous steroids to estrogens. The capacity for extragonadal conversion of behaviourally active exogenous steroids has been demonstrated in several species (review by Ryan et a l . , 1972) and much recent work has attempted to correlate the metabolism of androgens in central neural tissue with the effects of androgens and estrogens on sexual behaviour (Perez-Palacios et^ al_. , 1975; Naftolin and Ryan, 1975). Recent demonstrations (Christensen and Clemens, 1976; Beyer et a l . , 1976) that inhibition of aromatizing enzymes blocks the stimulatory effect of testosterone on sexual behaviour of male rats further support the concept that the effects of some steroid treatments on behaviour may be mediated by in vivo steroid conversion. Although there is no evidence that i n vivo steroid conversion occurred under the conditions of my experiments, there is evidence that the capacity for conversion is present in goldfish (Khoo, 1974, 1975) and other teleosts (Ozon, 1972). Of relevance is a study by Colombo and Belvedere (1976) who examined ovarian steroid synthesis in sexually immature Anguilla in which the oocytes were about to commence vitellogenesis. In vitro incubation of the previtellogenic ovarian tissue with pregnenolone or progesterone yielded testosterone plus several intermediate metabolites. If the steroidogenic properties of the immature Anguilla ovary are similar to those of the temperature-regressed goldfish, then the effects of pregnenolone and 103 17<x -OH-pregnenolone on spawning behaviour could be explained as the r e -s u l t of conversion of these steroids to androgens or estrogens. D i f f i c u l t i e s a r i s e when the concept of i n vivo conversion i s used to i n t e r p r e t the behavioural e f f e c t s of androgens and e s t r a d i o l . For example, i f i t i s assumed that e s t r a d i o l stimulates the mechanism con-t r o l l i n g spawning behaviour, then the e f f e c t s of androstenedione and testosterone could be explained as the r e s u l t of aromatization. However, i f the g o l d f i s h i s s i m i l a r to mammals i n being unable to aromatize 5 a - r e -duced steroids (Thompson et^ al_. , 1971), the e f f e c t s of androsterone and dihydrotestosterone could not be explained as the r e s u l t of conversion of these steroids to estrogen ( i t i s not known whether 11-ketotestosterone can be converted to estrogen). On the other hand, i f i t i s assumed that androgens regulate r e c e p t i v i t y i n the female g o l d f i s h , then the r e s u l t s of aridrogenstreatments could be interpreted as evidence that the mechanism c o n t r o l l i n g spawning behaviour responds to a v a r i e t y of androgens; i n female guppies ( L i l e y , 1972) and mammals (Beyer et a l . , 1971; Feder and S i l v e r , 1974), sexual r e c e i p t i v i t y i s induced by more than one estrogen. This l a t t e r i n t e r p r e t a t i o n f a i l s to account for the behavioural a c t i v i t y of e s t r a d i o l . That estrogen i s l i k e l y to be d i r e c t l y involved i n stimulating spawn-ing behaviour i s suggested by the e f f e c t s of estrogen on sexual behaviour of female guppies ( L i l e y , 1972), Mzardsm,(.Crews, 1975), b i r d s (Noble, 1972), and many mammals (Young, 1961; Beach, 1964; Davidson and Levine, 1972). However, androgen appears to control r e c e p t i v i t y i n the female rhesus monkey ( E v e r i t t and Herbert, 1975), Therefore, considering both the 104 behavioural effects of the androgens used i n this study and the fact that plasma androgens increase prior to spawning in female goldfish (Schreck and Hopwood, 1974) and in other female teleosts (Schmidt and Idler, 1962; Schreck et a l . , 1972; Katz and Eckstein, 1974; Campbell et a l . , 1976), the possibility that androgens may regulate spawning behaviour should not be ignored. As female sexual behaviour in the guppy and in many mammals'is i n -duced only by estrogens (and androgens which can be metabolized to estro-gens) i t is unlikely that both estrogens and non-aromatizable androgens act directly to stimulate receptivity i n female goldfish. A more reason-able explanation of the effects of estrogens and androgens on spawning be-haviour of intact, regressed goldfish i s that either estrogen or androgen (perhaps both) acts indirectly by stimulating gonadotropin release. This possibility i s considered in the following section of the discussion. In summary, no single explanation is li k e l y to account for the divers-ity of steroids which induce responsiveness to egg injection in intact, regressed goldfish. Pregnenolone and 17* -OH-pregnenolone probably are converted i n vivo to behaviourally active metabolites. Similarly, andros-tenedione and testosterone may be metabolized prior to affecting behaviour. However, the effectiveness of estradiol, and of androsterone and dihydro-testosterone (which l i k e l y cannot be aromatized), indicates that both androgen and estrogen restore receptivity in intact f i s h . These steroids may act directly on the mechanism controlling spawning behaviour. Alter-natively, androgen and / or estrogen may induce receptivity indirectly by stimulating gonadotropin release. 105 3. Possible Mechanisms of Steroid Action Exogenous steroids may restore responsiveness to egg injection i n -directly, by stimulating gonadotropin secretion, or directly, by acting on central neural or peripheral structures controlling spawning behaviour. Implicit i n the hypothesis that steroids induce receptivity by stim-ulating gonadotropin release is the assumption that in gegressed, un-receptive fish plasma gonadotropin is absent, or present in low concen-trations. Plasma gonadotropin has been measured in preovulatory (Breton et a l , , 1972) but not in regressed goldfish; however, Nagahama (1973) des-cribed signs of nuclear and cytoplasmic degeneration in pituitary gona-dotrophs from female goldfish with regressed ovaries. Also in Gillichthys  mirabilis, which undergoes a temperature-induced gonadal regression simi-lar to that of goldfish (De Vlaming, 1972), ultrastructural examination of gonadotrophs indicates that secretory activity of these cells i s depressed in regressed fish (Zambrano, 1972). In female brook trout and sockeye salmon with ovaries i n early stages of development, plasma gonadotropin levels are low or undetectable (Grim e_t al_. , 1975) . As partially regressed goldfish with minimal yolk vesicle deposition perform spawning behaviour when injected with ovulated eggs, low plasma gonadotropin titres may be sufficient to induce sexual receptivity. Thus, injection of steroids may restore responsiveness in regressed fish by stimulating only small increases in gonadotropin release. In female rats, steroids have been shown both to inhibit and to f a c i l i -tate release of gonadotropin (Davidson; 1969; Everett, 1969). Although some of these effects are exerted at the level of the hypothalamus, estrogen (Cooper and McCann, 1975) and testosterone (Perez-Palacios ejt a l . , 1976) 106 also increase the response of the pituitary to luteinizing hormone-releas-ing hormone. Steroids may act on the teleost hypothalamus and pituitary to regulate gonadotropin secretion; in the male sunfish (Lepomis cyanellus), Pfaff e_t al_. , (unpublished results cited i n Morrell et. a l . , 1975) demon-strated retention of labelled testosterone in the anterior pituitary and nucleus lateralis tuberis (an infundibular region suggested to regulate gonadotropin secretion i n goldfish [Peter, 1970] ). Generally, exogenous steroids exert inhibitory effects on the ovaries of intact fish (see review by Pickford and Atz, 1957). In the catfish, Heteropneustes f o s s i l i s , estradiol or testosterone treatments reduce both the size and number of pituitary basophils, suggesting that ovarian atresia following administration of these steroids results from inhibition of gona-dotropin secretion (Sundararaj and Goswami, 1968). Similar indirect evi-dence indicates that exogenous steroids inhibit gonadotropin secretion in female goldfish; injection of estrogens or testosterone induces atresia of vitellogenic oocytes (Khoo, 1974) and estrogen treatment produces degener-ative changes in gonadotrophs (Nagahama, 1973). The results of these studies suggest that, in fish with maturing ovaries, estrogen and testos-terone lower gonadotropin output to levels insufficient for maintenance of yolky oocytes; however, these findings do not eliminate thetpossibility that in regressed fish these steroids may stimulate a low rate of gonado-tropin release. Thus whether such a mechanism can account for the effects of steroids on spawning behaviour of regressed fish remains an open question. As noted above, i f steroids are directly involved in the regulation of spawning behaviour, they may act on central neural structures or may affect the responsiveness of the genital tract to the presence of ovulated eggs. 107 Estrogen stimulates sexual behaviour in female mammals by actions on specific areas of the brain, in particular the region of the preoptic nucleus - anterior hypothalamus. There is evidence that the preoptic area is involved i n the control of reproductive behaviour in teleosts. Ele c t r i c a l stimulation of the preoptic region evokes courtship behaviour in male b l u e g i l l sunfish, Lepomis macrochirus (Demski and Knigge, 1971), re-lease of milt i n male sunfish and goldfish, and release of eggs in ovulated female goldfish (Demski et a l . , 1975). Furthermore, electrolytic lesions of the preoptic area i n Fundulus abolish the spawning reflex response to injection of neurohypophyseal hormones (Macey et a l . , 1974). There is no evidence in teleosts that steroids are involved in the function of the pre-optic area or of other brain areas which may regulate reproductive behaviour. Although Pfaff et a l . (unpublished results cited in Morrell et al_. 1975) demonstrated binding of testosterone in the nucleus later a l i s tuberis of the male sunfish, Lepomis cyanellus, the functional significance of this finding i s not known. Steroids may influence spawning behaviour by an action on the genital tract. In the female rat, physical stimulation of the perigenital area plays a role in sexual behaviour (Pfaff et a l . , 1973; Kow and Pfaff, 1976) and estrogen increases the sensitivity of this region (Kow and Pfaff, 1973-4). Similarly, steroids may affect spawning behaviour in goldfish by sensitiz-ing the oviduct to the stimulus provided by ovulated eggs. In discussion to follow i t is suggested that ovulated oocytes are transported to the ovipore by c i l i a r y action and that i t is the portion of the oviduct near the ovipore which is sensitive to ovulated eggs. Steroids may regulate this proposed c i l i a r y ova transport; treatment of goldfish fry with ethinylestradiol or 108 methyltestosterone induces hypertrophy and extensive c i l i a t i o n of ovi-duct epithelium, processes which are normally associated with vitellogenesis (Takahashi and Takano, 1971). In female mammals, the effects of estradiol on proliferation of c i l i a in the genital tract are well known (More and Masterton, 1976). In summary, evidence that exogenous steroids inhibit gonadotropin secretion in goldfish and other teleosts suggests that steroids stimulate spawning behaviour in intact, regressed female goldfish by an action on the genital tract or on the central nervous system. Steroids may influence spawning behaviour solely through an action on the genital tract, activat-ing a mechanism which detects ovulation and relays this information to the central nervous system. It is suggested that i f steroids influence spawning behaviour by an action on the central nervous system, the mechanism involved may be different from that in female mammals. For example, in female mammals, elevated plasma estrogen levels induce a brief period of receptivity near the time of ovulation; endocrine stimuli associated with ovulation are not necessary to induce receptivity in goldfish, as females are responsive to injection of ovulated eggs at any time during v i t e l l o -genesis. Thus, steroids in mammals function as chemical messengers from the ovaries to the brain, signalling that the ovaries are prepared for ovu-lation. In female goldfish, stimuli generated by intraovarian ovulated eggs synchronize spawning behaviour with ovulation; steroids ( i f they are involved) appear only to prime the mechanism regulating receptivity, and not to transmit specific information regarding the state of ovarian development. D. The Role of Ovulated Eggs. Spawning behaviour of female goldfish i s temporally linked with ovulation J 109 by the stimulus of ovulated eggs in the genital tract. The ab i l i t y of ovulated eggs to induce spawning behaviour i s not restricted to the day of ovulation but i s seen i n a l l fish with ovaries in any stage of v i t e l l o -genesis. As discussed i n detail above, both pituitary and ovarian hormones may affect responsiveness to ovulated eggs. The site of action of ovulated eggs i n inducing spawning behaviour i s not known; however, several observations suggest that the terminal portion of the oviduct may be involved. For example, whereas stripping of a l l easily removed eggs (believed to be those i n the oviduct and posterior ovisac) usually terminates spawning for 10 to 15 minutes, removing only a portion of these eggs, and thus leaving additional eggs at the ovipore, has l i t t l e effect on spawning behaviour. It was also observed that in fe-males from which a l l accessible eggs had been removed, resumption of spawning generally was restricted to individuals which, on further stripping, were found to have eggs at the ovipore. Furthermore, in some cases where spawning behaviour following egg injection i s inhibited by egg binding (hardening and adhesion of eggs in contact with water),, only a small number of eggs at the ovipore are bound and the remainder of the injected eggs in the oviduct and ovisacs are apparently normal. It i s not known how ovulated eggs are transported to the ovipore. Eggs may be moved by the activity of smooth muscles or, more l i k e l y , by coordin-ated c i l i a r y motion as has been suggested to occur in ovum transport in the rabbit oviduct (Halbert et a l . , 1976) and in the frog coelom (Suvarnalatha and Sarkar, 1972). In goldfish, both the ovarian lamellae and the ovisacs possess conspicuously ciliated epithelium. The effect of ovulated eggs on behaviour is thought to be mediated at 110 least partially by physical cues, as substitutes for eggs (gelatin, petroleum jelly)are marginally effective. The decreased a b i l i t y of bound eggs to induce spawning may be due either to inappropriate physical characteristics or to localized 16ss of, aachemisal - stimuMnt. There-is^evidence that stimulation of the genital tract influences female sexual behaviour not only in teleosts but also i n other vertebrate classes. The close temporal correlation between ovulation and spawning (Liley, 1969) suggests that a mechanism whereby ovulated eggs stimulates spawning behaviour may be widespread in this group. In previous studies of oviparous fish (see review by Liley, 1969),failure to induce female sexual behaviour by steroid therapy may have been due to the absence of the requisite stimuli from internal sexual structures. Stimuli from ovulated eggs may function in the reproductive behaviour of some female Anura. Intraperitoneal saline injections (10-30 ml) induce oviposition behaviour in preovulatory and recently spent Rana pipiens (Noble and Aronson, 1942) indicating that abdominal distension may serve a function similar to oviduct distension in goldfish. Noble and Aronson (1942) suggest that, as saline injection f a i l s to induce oviposition behaviour in females treated more than thirteen days after ovulation, the behavioural response to abdominal distension may be influenced by hormones. Although physical s t i -mulation of the vagina and cervix f a c i l i t a t e s lordosis responses in the female rat (Rodriguez-Sierra e_t a l . , 1975), this phenomenon differs in several aspects from the effects of ovulated eggs on spawning behaviour i n goldfish. For example, the facilitatory effect of cervical probing occurs within a minute, persists for several hours after withdrawal of the stimu-lus, and is neither facilitated by estrogen treatment nor reduced by hypo-physectomy. I l l E. The Role of Prostaglandins Prostaglandins (PGs) appear to play a role in the spawning behaviour of female goldfish. The results of the present study suggest that stimu-lation of the oviduct following ovulation 'or injection of ovulated eggs or egg substitutes causes the release of PGs (most l i k e l y PGF2a ••) which then act directly or indirectly to induce spawning behaviour. The fact that indomethacin (an inhibitor of PG synthesis) eliminates a l l spawning within minutes suggests that endogenous PG is utilized rapidly and that spawning behaviour following ovulation o-t egg injection results from PG released over extended periods. The often considerable duration of spawning following i.p. PG injection indicates a slow uptake from the peritoneal cavity. The vari a b i l i t y in the latency to the onset of the behavioural response probably reflects the imprecision of this route of administration. The fact that some fish begin to spawn within a few minutes of receiving PG injection suggests that the latent period preceding spawning behaviour induced by egg injection may be the time required for the stimulus of eggs in the oviduct to elevate PG to effective levels. The capacity for PG biosynthesis has been demonstrated in several tissues of the carp (Christ and Van Dorp, 1972) and PGE and PGF have been isolated from the testes of several teleost species (Nomura et al_., 1973) . The source of endogenous PG which induces spawning behaviour in goldfish i s not known. PG may be released from the oviduct following physical stim-ulation, as has been shown to occur in mammalian uteri (Poyser et a l . , 1971), or i t may be released in the brain in response to afferent signals generated in the oviduct. It i s also possible that PG is released by ovarian macrophages in 112 response to egg injection. Macrophages have been implicated in the action of IUDs (Myatt et a l . , 1975) and i t has been suggested (Higgs and Youlten, 1975) that PG production by leukocytes might regulate the emigration of leukocytes from blood vessels during acute inflammation. If similar mechanisms were activated egg injection in goldfish, the latency from egg injection to onset of spawning might represent the time required for sufficient numbers of cells to aggregate at the injected eggs. Macrophage aggregations in the vic i n i t y of injected eggs are observed frequently i n histological preparations (Fig.29 in the Appendix). The fact that,! there is much evidence that the number and phagocytic activities of macrophages are greatly increased by estrogen and fluctuate with the female reproduct-ive cycle (Vernon-Roberts, 1969) suggests that the possibility of macro-phage involvement in prostaglandin production and spawning behaviour merits investigation. Recent unpublisheddstudies of female spawning behaviour in two other oviparous teleosts, (Jordanella floridae (Crawford, 1975) and Gasterosteus (Lam, Chan and Pandrey,1976) , support the present findings that PGs are i n -volved i n oviposition behaviour. Indomethacin injection abolished the re-sponses of ovulated female sticklebacks; following PGF^ injection, these responses were at least partially restored and in some cases oviposition occurred. Though injection of PGF„ into female Jordanella failed to i n -duce oviposition or even behavioural coordination with courting males, spawning reflex responses similar to those seen in the spawning behaviour of this and related cyprinodonts were usually e l i c i t e d . Both in Fundulus (Pickford, 1952; Wilhelmi e_t a l . , 1955)and in female Jordanella (Crawford, 1975), injections of neurohypophyseal hormones induce 113 a spawning reflex response similar to that observed in female Jordanella following administration of PGF„ . In contrast, treatment with neuro-hypophyseal hormones f a i l s to stimulate spawning behaviour in goldfish (present study; Pickford, unpublished results cited in Macey et a l . , 1974) and in several other teleosts (see Macey e_t a l . , 1974). On the basis of the earlier work on Fundulus (Pickford, 1952; Wilhelmi e_t a l . , 1955), which demonstrated that pharmacologically high doses of neurohypophyseal hormones (injected intraperitoneally) are required to induce the spawning reflex response and that these hormone treatments are equally effective i n intact and gonadectomized fis h , Macey et_ al_. (1974) proposed that neuro-hypophyseal hormones may act on some brain centre to e l i c i t the behavioural response; the finding (Macey et _al., 1974) that electrolytic lesions of the nucleus preopticus impaired or abolished the spawning reflex response to hormone injection provided support for this hypothesis. However, i t has recently been shown (Peter and Knight, unpublished results cited in Peter, 1976) that, even when neurohypophyseal hormones are administered by intra-ventricular injection, the dosage required to e l i c i t the spawning reflex response is similar to that required intraperitoneally. Peter (1976) i n -terprets these results as evidence of a peripheral rather than a central effect of these hormones and further suggests that a peripheral action of 'neurohypophyseal hormones is probably not a part of the normal mechanism for triggering spawning behaviour in teleosts'. The response of Fundulus to neurohypophyseal hormones is similar to that of female Jordanella injected with either oxytocin of YGF^^ ; in neither species do treated f i s h show behavioural coordination with the opposite sex. Thus i t is questionable whether the behavioural effect of 114 PGF2 A C i n female Jordanella i s physiological. Furthermore, i t i s stressed that the effect of neurohypophyseal hormones i n Eundulus, where both gonadectomized (Wilhelmi e_t a l . , 1955) and hypophysectomized fish (Pickford, 1952) show a spawning response without coordination of the sexes, contrasts sharply with the a b i l i t y of "PGF^^ t o induce a l l aspects of spawning behaviour i n female goldfish, but only in the normal spawning environment (in the presence of males and aquatic vegetation) and under pituitary stimulation. The mode of action of PG in inducing spawning behaviour in female goldfish i s not known. Hall et a l . (1975)showed that injection of PGE2 into the third ventricle of ovariectomized, estrogen-primed female rats induced receptivity similar to that following treatment with estradiol and progesterone. As LH-RH, shown to induce or increase lordosis in hypo-physectomized, estradiol-primed female rats (Pfaff, 1973), may be released under the stimulation of PGE2 (Eskay et a l . , 1975; Ojeda et a l . , 1975), Hall e_t al_. suggest that the mechanism by which PGE2 increases sexual be-haviour in female rats may involve the release of LH-RH. Hypothalamic implants of LH-RH have been shown by Dyer and Dyball (1974) to affect electrical activity of hypothalamic neurons, and these workers stress the possibility that i f this hypothalamic peptide functions both as a neuro-transmitter and as a releasing factor, a single neuron or group of neurons could influence both behaviour and pituitary function. There i s evidence that the goldfish hypothalamus contains the teleost equivalent to mammalian LH-RH (Peter, 1973; Crim et a l . , 1976); release of this hypothalamic factor may be involved in the action of PG on goldfish spawning behaviour. In the only report of an effect of PG on the sexual behaviour of male 115 mammals, twice-daily injections of PGF^la: (PGE^ was without effect) i n -duced at.70% increase i n mean number of ejaculations of male rabbits after 9 days of treatment (Agmo, 1975). The long latency of this response suggests a less direct effect on behaviour than i s l i k e l y to occur in the female goldfish. PG (and, indirectly, ovulation and egg injection) may exert some effect on ovarian or oviductal smooth muscle which i s monitored centrally through afferent stimulation. Portions of the female goldfish reproductive tract may be similar to mammalian uteri which, as well as responding to PG (Aiken, 1974), w i l l also release PG when physically stimulated by stretch (Poyser et a l , , 1971) or by insertion of a foreign object (Spilman and Duby, 1972) . The finding that PGP 2 * * s c o n s W e r a b l y more effective than PGE^ .^. ± r\ • stimulating spawning behaviour may be due to differential activity at a single site of action. Alternatively, VGF^ a n ^ P G E2cx m ay a c t a t different sites, the lower potency of PGE2 resulting from insufficient concentration of this prostaglandin i n i t s target tissue following intraperitoneal i n -jection. The effects of PG injection on hypophysectomized female goldfish which have been treated with salmon gonadotropin or steroids parallel the effect of egg injection on fish receiving similar treatment. Salmon gonadotropin replacement therapy restores the spawning response to injection of either eggs or PG, while steroid treatments are without effect. If egg injection inducessspawning behaviour by stimulating PG synthesis and release, then the failure of egg injection to evoke a response in hypophysectomized fish may simply reflect thetinability of hypophysectomized f i s h to respond to PG. 116 Alternatively, pituitary removal may disrupt many steps in a chain of events leading from the distension of the oviduct, by ovulation or egg i n -jection, to the onset of spawning behaviour. F. The Regulation of Spawning Behaviour In the foregoing discussion1,' more than one mechanism has been suggested to explain the function of each of the endogenous components which i n f l u -ence spawning behaviour. In the present section, the control of spawning behaviour is described by a model (Fig.l ) based on what is believed to be the most lik e l y set of mechanisms involved. Implicit i n the construction of the proposed model i s the assumption that there are fundamental simi-l a r i t i e s in the endocrine control of sexual behaviour i n female goldfish and mammals; without this constraint, the regulation of spawning behaviour could be described by other models consistent with the data presented in this study. As indicated in Figure 1, gonadotropin stimulates oocyte~development and synthesis of steroids which both maintain and sensitize the oviduct and prime hypothalamic (and/or preoptic) areas controlling spawning behaviour. Following ovulation, signals generated by stimulation of the oviduct induce changes in the steroid-primed hypothalamus; this mechanism may involve pro-staglandin (PG) at the level of the oviduct and the hypothalamus. Spawn-ing occurs when specific external cues (sexually active male; substrate for egg deposition) are present. The model attributes the lack of effect of steroids in hypophysectomized fish to the failure of exogenous steroids to duplicate the action of endo-genous steroid production and predicts that, in the absence of gonadotropin, 117 Figure 1. Model of the regulation of spawning behaviour in female goldfish (a) gonadotropin stimulation of oocyte growth and steroidogenesis (b) steroid stimulation of ovisac and oviduct (c) steroid action on sensitivity of oviduct (d) steroid priming of spawning centre GRH - gonadotropin releasing hormone, OE -ovulated egg, P - pituitary, PG - prostaglandin, SC - spawning centre 118 external stimuli gonadotropi n MODEL OF THE REGULATION OF SPAWNING BEHAVIOUR IN THE FEMALE GOLDFISH 119 some exogenous steroid treatment of hypophysectomized fi s h should be cap-able of restoring responsiveness to Injection of eggs or PG. This model also attributes the behavioural action of gonadotropin solely to an effect on steroidogenesis. If i t could be shown that the a b i l i t y of PG to induce spawning behaviour persists after ovariectomy, then PG treatment of gona-dectomized fi s h could provide valuable information regarding the roles of gonadotropin and steroids on spawning behaviour. The model proposes that PG acts both on the oviduct and the hypothala-mus to stimulate spawning behaviour. Inhibition of PG^-induced spawning ,by surgical removal of the oviduct would provide evidence that PG acts at this peripheral site. Both induction of spawning by intraventricular i n -jection of PG and inhibition of egg injections-induced spawning by intra-ventricular injection of indomethacin would provide support for a central action of PG on spawning behaviour. As indicated i n the proposed model, gonadotropin releasing hormone (GRH) is suggested to play a role in spawning behaviour; stimulation of the oviduct following ovulation generates afferent input to the hypothala-mus, inducing the release of GRH which then acts on the central neural centre(s) controlling spawning behaviour. The speculation that GRH is i n -volved i n spawning behaviour is based mainly on the results of mammalian studies. In the rat, vaginal stimulation induces changes in neuronal activ-i t y in brain centres controlling sex behaviour and LH-RH release (Blake and Sawyer, 1972), changes in hypothalamic LH-RH content (Takahashi et a l . , 1975), and persistent lordosis to manual stimulation (Rodriguez-Sierra et a l . , 1975); there i s much evidence that mechanisms by which genital stimuli induce gonadotropin (and presumably LHr-RH) release are widespread 120 in mammals (reviews by Jochle, 1973, 1975). Furthermore, LH-RH has been shown to induce lordosis in female rats (Pfaff, 1973; Moss and -Foreman, 1976) and the effect of on lordosis has been suggested (Hall et a l . , 1975) to be due to release of LH-RH (see Eskay et a l . . 1975, and Ojeda et a l . , 1975, for effects of PG on LH-RH release). These findings, which indicate that genital stimuli may enhance sexual behaviour in female mammals by releasing LH-RH, raise the possibility that ovulated eggs may induce spawning behaviour by a similar mechanism. There i s no published information on the effect of GRH on spawning behaviour of teleosts. However, the findings that plasma gonadotropin titres are greatly elevated in ovulated females of several salmonid species and in ovulating and partially spawned female sockeye salmon (Crim ejt al_., 1975) suggest not only that GRH release may increase during ovulation but also that high rates of release may persist i n ovulated f i s h , perhaps due to the stimulus provided by ovulated eggs. In goldfish, levels of plasma gonadotropin are high on the day of ovulation and return to pre-ovulatory values the following day (Breton et_ al_., 1972). If the pres-ence of ovulated eggs stimulates GRH and gonadotropin release in goldfish, the relatively transient increase in plasma gonadotropin observed by Breton et a l . (1972) may have resulted from a decrease in stimulus quality of the ovulated eggs (my observations show that in most cases where ovu-lated females are not placed with males'until the day after ovulation, at least slight egg binding occurs and few or no spawning acts are performed). The hypothesis that GRH plays a role in spawning behaviour could be tested indirectly by determining changes in plasma gonadotropin levels following injection of ovulated eggs. Behavioural observation following 121 intraventricular injection of mammalian LH-RH would provide a more direct approach; this technique has been used in examining the effects of neuro-hypophyseal hormones on the spawning reflex response of Fundulus (Peter, 1976). However, as there i s evidence that mammalian LH-RH and teleost GRH may be different (Deery, 1974), failure to induce spawning behaviour with mammalian LH-RH would not eliminate the possibility that endogenous GRH i s involved. G. A Comparative Approach to the Study of Female Sexual Behaviour In the preceding discussion, which considered the regulation of spawning behaviour i n terms of mechanisms proposed to regulate sexual behaviour in female mammals, i t has been assumed that similarities in the endocrine control of reproduction i n mammals and teleosts may be para-l l e l e d by similarities in the endocrine control of reproductive behaviour. Although this approach has been useful in interpreting experimental re-sults, i t affords only a limited view of the relationship between mechan-isms controlling reproductive behaviour in goldfish and mammals. A more comprehensive view of this relationship may result from considering these behaviours i n an evolutionary context. In the course of vertebrate evolution, reproductive strategies have changed greatly, the presumed ancestral mode of reproduction, external f e r t i l i z a t i o n , giving rise to internal f e r t i l i z a t i o n associated with ovi-parity, viviparity, etc. Regardless of the strategy employed, however, the formation and fate of the oocytes involves a similar series of stages: pituitary-stimulated development, ovulation, passage of ova or embryos through the coelomic cavity (gymnovarian teleosts) or reproductive tract (cystovarian teleosts, mammals) where they may be held for varying periods, 122 expulsion of ova or foetuses. What has changed drastically i s the stage at which the male intervenes and sexual behaviour occurs. For example, i n female goldfish and frogs, where f e r t i l i z a t i o n is external, sexual be-haviour and the release of ova are synchronous. In oviparous lizards and birds, and in viviparous guppies and mammals, the advent of internal f e r t i l i z a t i o n has been accompanied by an advancement i n the timing of sexual behaviour i n relation to oviposition and parturition. Thus, as the role of the female in reproductive behaviour changed from donator of ova (external fertilization) to recipient of sperm (internal f e r t i l i z -ation) , the occurrence of sexual behaviour became temporally dissociated from the expulsion of sexual products; in contrast, the persistent role of the male as gamete donator necessitated a close temporal association between sexual behaviour and gamete release in a l l male vertebrates. It i s proposed that in the ancestral, externally f e r t i l i z i n g female vertebrate (i.e. goldfish) a single mechanism, primed by endocrine factors associated with ovarian development and activated by physical stimulation of the genital tract following ovulation, performed the dual function of regulating sexual behaviour and releasing gametes. As the timing of sexual behaviour shifted with the evolution of internal f e r t i l i z a t i o n , the mechanism controlling female sexual behaviour came to be influenced by hormonal factors associated with earlier stages i n oocyte development. For example, stimuli from ovulated eggs are required to induce spawning behaviour in female goldfish, whereas estrogen alone i s sufficient to stimulate receptivity i n female mammals in which sexual behaviour precedes ovulation. However, as stimulation of the genital tract affects sexual behaviour in female mammals, the ab i l i t y of female mammals to respond to 123 steroids alone may be due to elaboration of a basic mechanism regu-l a t i n g sexual behaviour i n a l l female vertebrates. These highly speculative proposals, which greatly s i m p l i f y com-plex processes, are offered i n the hope of providing a conceptual basis f o r comparing the regulation of sexual behaviour and rel a t e d phenomena i n female vertebrates. 124 BIBLIOGRAPHY 125 Agmo, A. 1975. Effects of prostaglandin E^ and F on serum luteinizing hormone concentration and on some sexual functions in male rabbits. Prostaglandins, 9_: 451-457. Aiken, J.W. 1974. Prostaglandins and prostaglandins synthetase inhibi-tors: studies on uterine motility a n ^ function. In: Prostaglandin Synthetase Inhibitors, H.J. Robinson and J.R. Vane (Eds.)., Raven Press, New York. Baggerman, B. 1966. On the endocrine control of reproductive behaviour in the male three-spined stickleback (Gasterosteus aculeatus L.). Symp. Soc. Exptl. Biol. 20: 427-456. Beach, A.W. 1959. Seasonal changes in the cytology of the ovary and of the pituitary gland of the goldfish. Canadian Journal of Zoology, 37: 615-625. Beach, F.A. 1964. Biological basis for reproductive behaviour. In: Social Behaviour and Organization Among Vertebrates. W. Etkin (Ed.), pp. 117-142, University of Chicago Press, Chicago. Beach, F.A. 1976. Sexual attractivity, proceptivity, and receptivity in female mammals. Hormones and Behaviour, ]_: 105-138. Beach, F.A., and R.K. Orndoff. 1974. Variation in the responsiveness of female rats, to ovarian hormones as a function of preceding hormonal deprivation. Hormones and Behaviour, 5_: 201-205. Beyer, C , N. Vidal, and A. Mijares. 1970. Probable role of aromatization in the induction of estrous behaviour by androgens in the ovariec-tomized rabbit. Endocrinology, 87: 1386-1389. Beyer, C., G. Morall, and R. Vargas. 1971. Effect of diverse estrogens on estrous. behaviour and genital tract development in ovariectomized rats. Hormones and Behaviour, _2: 273-277. Beyer, C., G. Morali, F. Naftolin, K. Larsson, and G. Perez-Palacios. 1976. Effect of some ftntiestrogens and aromatase inhibitors on androgen-induced sexual behaviour in castrated male rats. Hormones and Beha- vior, 7_ 353-363. Blake, C.A., and C.H. Sawyer. 1972. Effects of vaginal stimulation on hypothalamic multi-unit activity and pituitary LH release in the rat. Neuroendocrinology, 10: 358-370. Breton, B., R. Bi l l a r d , B. Jalabert and G. Kann. 1972. Dosage radioimmuno-logique des gonadotropins plasmatiques chez Carassius aiiratus, au cours du inycthemere et. pendant 1'ovulation. General and Comparative  Endocrinology, 18: 463-68. 126 Campbell, CM., J.M. Walsh and D.R, Idler. 1976. Steroids i n the plasma of the winter flounder (Pseudbpleuroriectes americanus Walbaum). A seasonal study and investigation of steroid involvement in oocyte maturation. General and Comparative Endocrinology 29: 14-20. Campbell, C.M., and D.R. Idler. 1976. Hormonal control of vitellogenesis in hypophysectomized winter flounder (Pseudopleuronectes americanus Walbaum). General and Comparative Endocrinology, 28: 143-150. Chaudhuri, G. 1975. Inhibition by aspirin and ;iridomethacin of uterine hypertrophy induced by an IUD. J. Reproduction and F e r t i l i t y , 43: 77-81. Chester Jones, I., D. Bellamy, D.K.O. Chan, B.K. Follet, I.W. Henderson, J.G. P h i l l i p s , and R.S. Snart. 1972. Biological actions of steroid hormones in nonmammalian vertebrates. in:Steroids in Nonmammalian  Vertebrates, (Ed.) D.R. Idler, pp. 414-480. Academic Press, New York. Christ, E.J., and D.A. Van Dorp. 1972. Comparative aspects of prostaglan-din biosynthesis in animal tissues. Biochimdca et Biophysica Acta, 270: 537-545. Christensen, L.W., and L,G. Clemens. 1976. Blockage of testosterone-in-duced mounting behaviour in male rats with intra-cranial application of the aromatization inhibitor, androsta-1,4,6-triene-3,17-dione. Endocrinology, 97: 1545-1551. Ciaccio, L.A., and R.D. Lisk, 1967. Facilitation and inhibition of es-trous behaviour i n the spayed female golden hamster, (Mesocricetus  auratus). Endocrinology, 88: 810-815. Colombo, L., and P.C. Belvedere. 1976. Steroid biosynthesis by the ovary of the European eel, Anguilla anguilla L., at the silver stage. General and Comparative Endocrinology, 28: 371-85. Cooper, K.J., and McCann. 1975. Influence of ovarian steroid on pituitary responsiveness to LH-releasing hormone (LH-RH) in the rat. In  Hypothalamic Hormones. M. Motta, P.G. Crosignani and L. Martini (Eds.) pp. 161-168, Academic Press. Crawford, S. 1975. Analysis of Jordanella floridae reproductive beha-viour. B.Sc. Thesis, U.B.C. Crews, D. 1975. Psychobiology of reptilian reproduction. Science, 189: 1059-1065. Crim, L.W., E.G. Watts, and D.M. Evans. 1975. The plasma gonadotropin profile during sexual matur-ation. in a variety of salmonid fishes. General and Comparative Endocrinology, 27: 62-70. 127 Crlm, L.W., R.E. Peter and R. Bil l a r d . 1976. Stimulation of gonadotro-pin secretion by intraventricular injection of hypothalamic extracts in the goldfish Carassius auratus. General and Comparative Endocrino- logy, 30: 77-82. Crowley, W.R., J.F. Rodriguez-Sierra and B.R. Komisaruk. 1976. Hypophy-sectomy facilitates sexual behaviour in female rats. Neuroendocrino- logy, 20: 328-338. Csapo, A.I. 1973. The prospects of PGs in postconceptional therapy. Prostaglandins, ,3_: 245-289. Currie, W.B. 1975. Secretion rate of prostaglandin F during induced labor in goats. Prostaglandins. 9_: 867-879. Damassa, D., and J.M. Davidson. 1973. Effects of ovariectomy and constant light on responsiveness to estrogen in the rat. Hormones and  Behaviour, 4_: 269-279. Davidson, J.M. 1969. Feedback control of gonadotropin secretion. In: Frontiers in Neuroendocrinology. W.F. Ganong and L. Martini (Eds.) pp. 343-388, Oxford University Press, New York. Davidson, J.M., and S, Levine. 1972, Endocrine regulation of behaviour. Annual Review of Physiology, 34: 375-408. Deery, D.J. 1974. Determination by radioimmunoassay of the luteinizing hormone-releasing hormone (LH-RH) content of the hypothalamus of the rat and some lower vertebrates. General arid Comparative Endocrinology 24: 280-285. Demski, L.S., and K.M. IRnigge, 1971. The telencephalon and hypothala-mus of the bl u e g i l l (Lepomis macrochirus): evoked feeding, aggressive and reproductive behaviour with representative frontal sections. Journal of Comparative Neurology, 143: 1-16. Demski, L.S., D.H. Bauer, and J.W. Gerald. 1975. Sperm release evoked by electrical stimulation of the fish brain: a functional-anatomical study. Journal of Experimental Zoology, 191: 215-232. De"Vlaming,V.L. 1972. The role of the endocrine system in temperature-controlled reproductive cycling in the estuarine gobiid f i s h , Gillichthys mirabilis. Comparative BjOchemistry7and • 3?hysiolagy, 41(A): 679-713. Dodd, J.M. T972. The endocrine regulation of gametogenesis and gonad maturation i n fishes. General and Comparative Endocrinology Supple-ment, 2 : 675-687. 128 Donaldson, E.M., and J.R,. McBride, 1974. Effect of ACTH and salmon gonadotropin on interrenal and thyroid activity of gonadectomized adult sockeye salmon (Oncorhynchus nerka). Journal of the Fisheries  Research Board of Canada, 31: 1211-1214. Dyer, R.G,, and R.E.J. Dyball. 1974. Evidence for a direct effect of LRF and TRF on single unitd activity in the rostral hypothalamus. Nature, 252: 486-488. Eckstein, B, 1970. Metabolic pathways of steroid biosynthesis in the ovarian tissue of a teleost, Tilapia aureaa. General arid Comparative  Endocrinology, 14: 303-312. Eckstein, B., and Y. Katz. 1971. Steroidogenesis in post- and pre-spawned ovaries of a cichlid f i s h , Tilapia aurea. Comparative Biochemistry and  Physiology, 38A: 329-338. Elazar, Z., and B. Blum. 1971. Effect of drugs on i n t e r i c t a l spikes and afterdischarges i n experimental epilepsy. Archives of International  Pharmacodyn. 189: 310-318. Emmersen, B.K., and T.M. Petersen. 1976. Natural occurrence, and experi-mental induction by estradiol-17-B, of a lipophosphoprotein ( v i t e l l o -genin) in flounder (Platichtys flesus, L.). Comparative 'Biochemistry and .-Physiology, 54B: 443-446. Eskay, R.L., J. Warberg, R.S. Mical and J.C. Porter. 1975. Prostaglandin E„-induced release of LH-RH into hypophyseal portal blood. Endocrino- logy, 97: 816-824. Everett, J.W.. 1969. Neuroendocrine aspects of mammalian reproduction. Ann. Rev. Physiol. 31: 383-416. Everitt, B.J., and J. Herbert. 1975. The effects of implanting testos-terone propionate into the central nervous system on the sexual beha-viour of adrenalectomized female rhesus monkeys. Brain Research, 86: 109-120. Feder, H.H., and R. Silver. 1974. Activation of lordosis in ovariectom-ized guinea pigs by free and -esterified forms of estrone, estradiol-17B and e s t r i o l . Physiology and Behaviour, 13: 251-255. F l i n t , A.P.F., A.B.M. Anderson, T,P. Patten, and A.C. Turnbull. 1974. Control of utero-ovarian prostaglandin F during labor in the sheep: acute effects of vaginal and cervical stimulation. Journal of  Endocrinology, 63: 67-87. 129 Gallien, L. 1975. Sequential endocrine activities controlling oogene-si s : a general survey. American Zoologist, 15: Suppl. 1, Chapter 12, 197-214. Gaunt, R., B,G. Steinetz, and J.J. Chart. 1968. Pharmacologic altera-tion of steroid hormone functions. C l i n i c a l Pharmacology and Thera-peutics, 9_: 657-81. Gower, D.B. 1974. Modifiers of steroid hormone metabolism: a review of their chemistry, biochemistry, and c l i n i c a l applications. J^ . Steroid Biochem. _5: 501-523. Gurr, E. 1962. Staining animal tissues. A practical theory. Leonard H i l l , London. Halbert, S.A., P.Y. Tam, and R.J. Blandau. 1976. Egg transport i n the rabbit oviducts: the roles of c i l i a and muscles. Science, 191: 1052-1053. Hall, N.R., W.G. Luttge, and R.B. Berry. 1975. Intracerebral prostaglan-din E 2: effects upon sexual behaviour, open f i e l d activity and body temperature in ovariectomized female rats. Prostaglandins, 10: 877-888. Hertelendy,, F. 1972. Prostaglandin-induced premature oviposition in the coturnix quail. Prostaglandins, _2: 269-279. Hertelendy, F. 1973. Block of oxytocin-induced parturition and oviposi-tion by prostaglandin inhibitors. Life Sciences, 13: 1581-1589. Hertelendy, F., M. Yeh, and H.V. B i e l l i e r , 1974. Induction of oviposition in the domestic hen by prostaglandins. General and Comparative  Endocrinology, 22: 529-531. Hertelendy, F., H.V. B i e l l i e r , and H. Todd. 1975. Effects of the egg cycle and route of administration on prostaglandin-induced oviposi-tion of hens and quail. J. Reprod. and F e r t i l . , 44: 579-582. Higgs, G.A., and L.J.F. Youlten. 1972. Prostaglandin production by rab-bit peritoneal polymorphonuclear leukocytes in vitro. British J.  Pharmacology, 44: 330 P. Hoar, W.S. 1962. Hormones and the reproductive behaviour of the male three-spined stickleback (Gasterosteus aculeatus). Animal Behaviour, 10: 247-266. Hoar, W.S. 1965. Comparative physiology: hormones and reproduction i n fishes. Annual Review of Physiology, 27: 51-70. 130 Innes, W.T. 1949. Goldfish v a r i e t i e s and water gardens. Innes Publish. Co., Philadelphia. Jochle, W. 1973. Coitus-induced ovulation. Contraception. _7_: 523-564. Jochle, W. 1975. Current research i n coitus-induced ovulation: A review. J. Reproduction and F e r t i l i t y (Suppl. 22), 165-207. Joslyn, W.D., H.H. Feder, and R.W. Goy. 1971. Estrogen conditioning and progesterone f a c i l i t a t i o n of lo r d o s i s i n the guinea p i g . Physiology  and Behaviour, _7: 477-482. Katz, Y., and B. Eckstein. 1974. Changes i n s t e r o i d concentration i n blood of female T i l a p i a aurea ( T e l e o s t e i , Cichlidae) during i n i t i a -t i o n of spawning. Endocrinology 95: 963-967. Khoo, K.H. 1974. Steroidogenesis and the ro l e of steroids i n the endo-crine c o n t r o l of oogenesis and v i t e l l o g e n e s i s i n the g o l d f i s h , Carassius auratus. Ph.D. Thesis, University of B.C. Khoo, K.H. 1975. The corpus luteum of go l d f i s h (Carassius auratus L.) and i t s functions. Canadian Journal of Zoology, 53: 1306-1323. Kow, L.-M., and D.W. P f a f f . 1973-74. E f f e c t s of estrogen treatment on the s i z e of receptive f i e l d and response threshold of pudendal nerve i n the female r a t . Neuroendocrinology, 13: 299-313. Kow, L.-M., and D.W. P f a f f . 1976. Sensory requirements f o r the l o r d o s i s r e f l e x i n female r a t s . Brain Research, 101: 47-66. Kuehl, F.A. J r . , J.L. Humes, J. Tarnoff, V.J. C i r i l l o , and E.A. Ham. 1970. Prostaglandin receptor s i t e : evidence f o r an e s s e n t i a l role i n the action of l u t e i n i z i n g hormone. Science, 169: 883-884. Kuhl, H., and H.-D. Taubert. 1975. Short-loop feedback mechanism of l u t e i n i z i n g hormone: LH stimulates hypothalamic L-cystine arylamidase to i n a c t i v a t e LH-RH i n the rat hypothalamus. Acta Endocrlnologica, 78: 649-663. Labhsetwar, A.P. 1974. Prostaglandins and the reproductive cycle. Fed. Proc., 33: 61-77. Lambert, J.G.D., J.H.H. T^ i j s s e n and V. Eylath. 1971. Conversion of H-androstenedione and H-testosterone to H-oestradiol-178 i n v i t r o by the ovary of the guppy, P o e c i l i a r e t i c u l a t a . Kon. Nederl. Akad. West. (Amsterdam) 74: 52-59. 131 Liggins, G.C., R.J. Fairclough, S.A. Grieves, J.Z. Kendall and'B.S. Knox. 1973, The mechanism of i n i t i a t i o n of parturition in the ewe. Recent Brogress In Hormone Research, 29: 111-159. Liley, N.R. 1968. The endocrine control of reproductive behaviour i n the female guppy, Poecilia reticulata Peters. Animal Behaviour, 16: 318-331. Liley, N.R, 1969. Hormones and reproductive behaviour i n ,;;f.ishes. In: Fish Physiology. W.S. Hoar and D.J. Randall (Eds.)'., Vol. 3, pp. 73-116, Academic Press, New York. Liley, N.R, 1972. The effects of estrogens and other steroids on the sexual behaviour of the female guppy, Poecilia reticulata. General  and Comparative Endocrinology, Suppl. 3: 542-552. Liley, N.R., and E.M. Donaldson. 1969. The effects of salmon pituitary gonadotropin on the ovary and the sexual behaviour of the female guppy, Poecilia reticulata. Canadian Journal of Zoology, 47: 569-573. Liley, N.R., and W. Wishlow. 1974. The interaction of endocrine and experiential factors in the regulation of sexual behaviour in the female guppy, Poecilia reticulata. Behaviour, 48: 185-214. McDonald, P.G., and B.J. Meyerson. 1973. The effect of oestradiol, tes-tosterone and dihydrotestosterone on sexual motivation in the ovari-ectomized female rat. Physiology and Behaviour, 11: 515-520. MacKay, N.J. 1973. The effects of methallibure (I.C.I.-33,828) and thiourea on gametogenesis in the f i r e t a i l gudgeon, Hypseleotris g a l l i . General and Comparative Endocrinology 20: 221-235. Macey, M.J., G.E. Pickford and R.E. Peter. 1974. Forebrain localization of the spawning reflex response to exogenous neurohypophysial hor-mones in the k i l l i f i s h , Fundulus heteroclitus. J. Experimental  Zoology, 190: 269-280. Marsh, J.M., N.S.T. Yang, and W.J. LeMaire. 1974. Prostaglandin synthe-sis in rabbit graffian f o l l i c l e s in vitro. Effect of luteinizing hormone and cyclic AMP. Prostaglandins, 7: 269-283. Milgrom, E., M. Luu Thi and E.E. Baulieu. 1973. Control mechanisms of steroid hormone receptors in the reproductive tract. 6th Karolinska Symposium on Research Methods in Reproductive Endocrinology. Protein Synthesis in Reproductive Tissue. A. Diczfalusky (Ed.) , Geneva, Switzerland, May, 1973. Karolinska Institutat, Stockholm, Sweden, pp. 380-403. 132 More, I.A.R., and R.G. Masterton. 1976. The role of oestrogen in the control of ciliated cells of the human endometrium. J. Reproduction  and F e r t i l i t y , 47: 19-24. Morrell, J.I., D.B. Kelley, and D.W. Pfaff. 1975. Sex steroid binding in the brains of vertebrates: studies with light microscope auto-radiography. In: The Ventricular System in Neuroendocrine Mechanisms, Proceedings of the second brain-endocrine interaction symposium. K.M. Knigge, D.E. Scott, M. Kobayashi, and S. I s h i i (Eds.)., Karger, Basel, pp. 230-256. Moss, R.L., and M.M, Foreman, 1976. Potentiation of lordosis behaviour by intrahypothalamic infusion of synthetic luteinizing hormone-releasing hormone. Neuroendocrlnology, 20: 176-181. Myatt, L., M.A. Bray, D. Gordon, and J. Morley. 1975. Macrophages on intrauterine contraceptive devices produce prostaglandins. Nature, 257: 227-228. Naftolin, F., and K.J. Ryan. 1975. The metabolism of androgens in central neuroendocrine tissues. J. of Steroid Biochemistry, 6_: 993-997. Nagahama, Y. 1973. Histo-physiological studies on the pituitary gland of some teleost fishes, with special reference to the hormone-pro-ducing cells in the adenohypophysis. Memoires of the Faculty of Fisheries, Hokkaido, University 21(1) : 1-63. Noble, G.K., and K.F. Kumpf. 1936. The sexual behaviour and secondary sex characters of gonadectomized fi s h . Anatomical Record, 67: Suppl. 113. Noble, G.K., and L.R. Aronson. 1942. The sexual behaviour of Anura. I. The normal mating patterns of Rana pipiens. Bull. Amer. Mus. Natural  History, 80: 127-142. Noble, R. 1972. The effects of estrogen and progesterone on copulation in female quail (Coturnix coturhix japonica) housed in continuous dark. Hormones and Behaviour, _3: 199-204. Nomura, T., H. Ogata, and M. Ito. 1973. Occurrence of prostaglandins in fish testis. Tohoku J. of Agricultural Research, 21: 138-144. Ojeda, S.R., J.E. Wheaton, and S.M. McCann. 1975. Prostaglandin E2~in-duced release of luteinizing hormone-releasing factor (LRF). Neuroendocrinology, 17: 283-287. Ozon, R. 1972. Estrogens in fishes, amphibians, reptiles, and birds. In: Steroids in Nonmammalian Vertebrates. D.R. Idler (Ed.)., Academic Press, N.Y. pp. 390-413. 133 Partridge, B.L., N.R. Liley, and N.E. Stacey. 1976. The role of phero-mones in the sexual behaviour of the goldfish. Animal Behaviour, 24: 291-299. Perez-Palacids, G., K. Larsson, and C, Beyer. 1975. Biological s i g n i f i -cance of the metabolism of androgens in the central nervous system. J. of Steroid Biochemistry, 6^: 999-1006. Peter, R.E. 1970. Hypothalamic control of thyroid gland activity and gonadal activity i n the goldfish, Carassius auratus. General and  Comparative Endocrinology, 14: 334-356. Peter, R.E. 1973. Neuroendocrinelogy of teleosts. American Zoologist, 13: 743-755. Peter, R.E. 1976. The preoptic nucleus in fishes: A comparative discus-sion of function-activity relationships. (In preparation). Pfaff, D.W. 1970. Mating behaviour of hypophysectomized rats. J_. Comparative and Physiological Psychology, 72.: 45-50. Pfaff, D.W. 1973. Luteinizing hormone-releasing factor potentiates lordosis behaviour in hypophysectomized ovariectomized female rats. Science, 182: 1148-1149. Pfaff, D.W., C. Lewis, C. Diakow, and M. Keiner. 1973. Neurophysiological analysis of mating behaviour as hormone-sensitive reflexes. In  Progress i n Physiological Psychology. E. Stellar and J.M. Sprague (eds.), Vol. 5, pp. 253-297. Academic Press, New York. Philbert, M., J.P. Luton, B. Mathieu de Fossey and H. Bricaire. 1968. Les inhibiteurs de la cortico steroidogenese II. Nouvel inhibiteur de l a cortico-surrenole: 1'aminoglutethimide (16'038 Ba). La Presse  Medicale 76: 1107-1110. Pickford, G.E. 1952. Induction of a spawning reflex in hypophysectomized k i l l i f i s h . Nature, 170: 807-808. Pickford, G.E., and J.W. Atz. 1957. The physiology of the pituitary gland of fishes. New York Zoological Society, New York. Piper, P., and J. Vane. 1971. The release of prostaglandins from lung and other tissue. Annals of the New York Academy of Science, 180: 363-385. Poyser, N.L., E.W. Horton, C.J. Thompson and M. Loss. 1971. Identifica-tion of prostaglandin F^ released by distension of guinea-pig uterus in vitro. Nature, 230: 526-528. 134 Rallison, J.L., L.F. Kumagai and F.M. Tyler. 1967. Goitrous hypothy-roidisms induced by aminoglutethimide, anticonvulsant drug. J^ . of Cl i n i c a l Endocrinology, 27: 265-272. Roberts, J.S., B. Barcikowski, R. Wilson, R.C. Skarnes, and J.A. McCracken. 1975. Hormonal and related factors affecting the release of prosta-glandin from the uterus. J. Steroid Biochemistry, (i: 1091-1097. Rodriguez-Sierra, J.F., W.R. Crowley and B.R. Komisaruk. 1975. Vaginal ^stimulation in rats induces prolonged lordosis responsiveness and sexual receptivity. J. of Comparative and Physiological Psychology, 89: 79-85. Ruh, T.S., B.S. Katzellenbogen, J.A. Katzellenbogen and J. Gorski. 1973. Estrone interaction with the rat uterus: in vitro response and nu-clear uptake. Endocrinology, 92: 125-134. Ryan, K.J., F. Naftolin, V. Reddy, F. Flores, and Z. Petro. 1972. Estro-gen formation in the brain. American J. of Obstetrics and Gynecology, 114: 454-462. Schmidt, P.J., and D.R. Idler. 1962. Steroid hormones in the plasma of salmon at various stages of maturation. General and Comparative  Endocrinology, 2_: 204-214. Schreck, C.B., R.T. Lackey, and M.L. Hopwood. 1972. Evaluation of diel variation in androgen levels of rainbow trout (Salmo gairdneri). Copeia, _4: 865-868. Schreck, C.B., and M.L. Hopwood. 1974. Seasonal androgen and estrogen patterns in the goldfish, Carassius auratus. Transactions of the  American Fisheries Society: 2} 345-348. Scruggs, W.M. 1951. The epithelial components and their seasonal changes in the pituitary of the carp (Cyprinus carpio L.) and goldfish (Carassius auratus L.). J. Morphology, 88: 441-469. Spilman, CH. 1974. Oviduct response to prostaglandins: influence of estradiol and progesterone. Prostaglandins, ]_: 465-471. Spilman, CH., and R.T. Duby. 1972. Prostaglandin mediated luteolytic effect of an intrauterine device in sheep. Prostaglandins, 2_: 159-168. Stacey, N.E., and S. Pandey. 1975. Effects of indomethacin and prosta-glandins on ovulation of goldfish. Prostaglandins, 9_: 597-607. Sundararaj, B.I., and S.V. Goswami. 1966. Effects of mammalian hypophyseal hormones, placental gonadotropins, gonadal hormones, and adrenal cor-ticosteroids on ovulation and spawning in hypophysectomized catfish, Heteropneustes f o s s i l i s (Bloch). J. Experimental Zoology, 161: 287-296. 135 Sundararaj, B.I., and S.V. Goswami. 1968. E f f e c t s of estrogen, progester-one, and testosterone on the p i t u i t a r y and ovary of c a t f i s h , Heteropneustes f o s s i l i s (Bloch). J . Exper. Zool. 169: 211-228. Suvarnalatha, M., and H.B. Devaraj Sarkar. 1972. Egg transport i n the peritoneal cavity of the skipper frog, Rana cyanophlyctis (SCHN). B i o l . Rep. 6_: 234-237. Takahashi, H. and K. Takano. 1971. Sex hormone-induced precocious hy-pertrophy and c i l i a t i o n of e p i t h e l i a l c e l l s i n the ovarian lumen of the g o l d f i s h . Annot. Zool. Japan, 44: 32-41. Takahashi, M,, J . J . Ford, K. Yoshinaga and R.O, Greep. 1975. E f f e c t s of c e r v i c a l stimulation and anti-LH r e l e a s i n g hormone serum on LH-relea s i n g hormone content i n the hypothalamus. Endocrinology, 96: 453-457. Thompson, E.A.,- S.B. Bolton, and P.K. S i i t e r i . 1971. K i n e t i c studies of placental s t e r o i d aromatase. Fedn, Proc. Fedn. Am. Acts. Exp. B i o l . , 30: 1160. Thompson, E.Q., and P.K. S i i t e r i . 1973. Studies on the aromatization of C-19 androgens. Annals New York Academy of Science, 212: 378-388. Umo, I., R.J. F i t z p a t r i c k , and W.R. Ward, 1976. P a r t u r i t i o n i n the goat: plasma concentrations of prostaglandin F and s t e r o i d hormones and uterine a c t i v i t y during l a t e pregnancy and p a r t u r i t i o n . J . Endocrino- logy, 6J8: 383-389. Vane, J.R. 1971. I n h i b i t i o n of prostaglandin synthesis as a mechanism of action of a s p i r i n - l i k e drugs. Nature New Biology, 231: 232-235. Vernon-Roberts, B. 1969. The e f f e c t s of s t e r o i d hormones on macrophage a c t i v i t y ! ; I nternational Review of Cytology, pp. 131-159. G.H. Bourne and J.F. D a n i e l l i (Eds.), Academic Press. Wechsung, E., and A. Houvenaghel. 1976. A possible r o l e of prostaglandins i n the regulation of ovum transport and o v i p o s i t i o n i n the domestic hen. Prostaglandins, 12: 599-609. Whalen, R.E. , C. B a t t i e , and W.G. Luttge. 1972. Anti-estrogen i n h i b i t i o n of androgen induced sexual r e c e p t i v i t y i n r a t s . Behavioural Biology, 7: 311-320. Wiebe, J.P. 1967. The reproductive physiology of the viviparous seaperch, Cymatogaister aggregata CGihbb'ns. Ph.D. Thesis, U.B.C. 136 Wilhelmi, A.E., G.E. Pickford, and W.H. Sawyer. 1955. Initiation of the spawning reflex responses in Fundulus by the administration of-fish and mammalian neurohypophyseal preparations and synthetic oxytocin. Endocrinology, 57: 243-252. Yamamoto, K., Y. Nagahama, and F. Yamazaki. 1966. A method to; induce a r t i f i c i a l spawning of goldfish a l l through the year. Bull. Jap. Soc. Sci. Fisheries, 32: 977-983. Yamazaki, F. 1961. The effects of hypophysectomy on the ovary of the goldfish, Carassius auratus, Bull. Fac. Fish. Hokkaido Univ., 12: 167-180. Yamazaki, F. 1965. Endocrinological studies on the reproduction of the female goldfish, Carassius auratus L., with special reference to the function of the pituitary glandr:. Mem. Faculty of Fisheries, Hokkaido University, 13: 1-64. Yamazaki, F., and E.M. Donaldson. 1968. The effects of partially puri-fied salmon pituitary gonadotropin on spermatogenesis, vitellogenesis, and ovulation in hypophysectomized goldfish (Carassius auratus). General and Comparative Endocrinology, 11: 292-299. Young, W.C. 1961. The hormones and mating behaviour.ialn: Sex and Internal Secretions. W.C. Young (Ed.). 3rd Edition, Vol. 2, pp. 1173-1239. Williams and Wilkins, Baltimore. Zamb.rano;} j D. 1971. The nucleus latercalis tuberis system of the gofrlid f i s h Gillichthys mirabilis - III. Functional modifications of the neurons and gonadotropic c e l l s . General and Comparative Endocrino- logy, 17_: 168-182. 137 APPENDIX 138 SOME ASPECTS OF THE HISTOLOGY OF REGRESSED OVARIES In this study, the ovarian histology of experimental f i s h has been used both i n assessing the e f f e c t s of various injected hormones and i n determining the completeness of hypophysectomy and ovarian regression. The process of s e l e c t i n g ovarian c h a r a c t e r i s t i c s which could be corre-l a t e d with various endocrine states revealed several aspects of ovarian histology which apparently have not been reported previously. Degenera-t i o n of p r e v i t e l l o g e n i c ( f i r s t growth phase) oocytes was observed many times i n i n t a c t and hypophysectomized f i s h ; a s e r i e s of stages of degener-ation i s described below and some information i s provided concerning the e f f e c t s of steroids and gonadotropin. C y t o l o g i c a l features of p r e a t r e t i c early yolk v e s i c l e stage oocytes also are discussed. As the existence of degenerating p r e v i t e l l o g e n i c oocytes (DPVOs) was not discovered u n t i l completion of a l l experiments, no s p e c i a l techniques were employed i n preparing ovaries for h i s t o l o g i c a l examin-a t i o n . A l l ovaries with DPVOs were r o u t i n e l y f i x e d i n Bouin's F l u i d and wax embedded (Paraplast). Blocks were cut at 5 u and stained with e i t h e r haematoxylin and eosin, Mallory trichrome connective t i s s u e s t a i n (Gurr, 1962), or PAS and l i g h t green. As tissues were sampled from various areas of the ovary and as the cutt i n g o r i e n t a t i o n was not standardized, no quantitative assessment of ovarian histology was undertaken. DPVOs were found i n the ovaries of i n t a c t , temperature-regressed females and i n hypophysectomized f i s h , though the incidence was much higher i n the l a t t e r group. In the early stages of a t r e s i a (Fig.2), DPVOs were very d i f f i c u l t 139 to distinguish from degenerating early yolk vesicle stage oocytes (DEYVOs). However, as atresia progresses, the f o l l i c l e s of the DEYVOs develop into corpora atretica, passing through stages 1 to { as described by Khoo (1975) for more advanced oocytes: i.e. the f o l l i c u l a r layer hypertrophies, the oocyte contents are taken up by the granulosa and invading macrophages (Figs. 3 and 4), the hypertrophied f o l l i c l e f i r s t surrounds an empty cavity (Figs. 4 and 5), and finally collapses into an irregular mass of cells. On the other hand, DPVOs do not form corpora atretica, there is very l i t t l e or no f o l l i c u l a r hypertrophy, and although macrophages invade and remove the contents of the oocyte, there is no collapse of the f o l l i c l e , which eventually surrounds a space approxi-mately the size of the evacuated oocyte (Fig. 6). It i s d i f f i c u l t to be certain that a l l small degenerating f o l l i c l e s that form corpora atretica also contain yolk vesicles, as the amount of yolk vesicle material may be very small (one vesicle/5y section). Stain-ing with PAS and light green has proven effective in detecting small amounts of yolk vesicles, but this method is complicated by the fact that the yolk nucleus (an accumulation of mitochrondria) not only i s taken up by the hypertrophied f o l l i c l e cells (as are the yolk vesicles), but, as with the yolk vesicles, is strongly PAS-positive (Fig. 7). The f i r s t recognizable stage of atresia in previtellogenic oocytes involves migration of the nucleus toward the egg membrane, breakdown of the nucleus, disappearance of any difference in staining properties of nucleoplasm and cytoplasm, and a general decrease In the basophilia of the two (Fig. 8). The early stages of the loss of staining may be due simply to the dilution of the cytoplasm with the weakly staining nucleoplasm; 140 in later stages, the weaker staining i s l i k e l y due to changes in cyto-plasmic constituents as the cytoplasm takes on a more granular appear-ance and eventually i s composed of aggregated debris. Whether other f o l l i c u l a r events precede these early changes is not known. However, a common finding i n ovaries that contain DEYVOs is that the yolk vesicles in the non-degenerating oocytes no longer maintain a normal cytoplasmic distribution (Figs. 9 and 10), but are very close to, i f not touching, the oocyte membrane (Figs. 11 and 12). In a l l other aspects the oocytes appear normal. As the yolk vesicles in oocytes which have just become atretic are not visible in the cytoplasm but can be found in the hyper-trophied f o l l i c u l a r layer (Fig,13), i t appears that yolk vesicles migrate to the extreme periphery of the cytoplasm prior to the onset of atresia in early yolk vesicle stage oocytes. No other cytoplasmic or nuclear abnormalities are evident at this time. Migration of the nucleus toward the periphery is seen commonly in DPVOs and DEYVOs and has been reported by Yamazaki '(1961) and observed in the present study to occur i n oocytes with more advanced yolk vesicle formation (Fig.14). In the early vitellogenic and previtellogenic oocytes, the nucleus moves as a unit toward the egg membrane, at which point some or a l l of the nucleoli are released. The f o l l i c u l a r layer may (Fig.15) or may not (Fig.16) be hypertrophied. In many i f not a l l instances, the nuclear membrane, often retaining a large (10 - 12u) spherical structure with highly variable staining properties (Figs. 17 and 18), returns to the centre of the oocyte, where i t may persist after f o l l i c u l a r hyper-trophy has begun. It is believed that i t is this stage which has been interpreted by Khoo (1974) as evidence of pregnenolone-induced yolk 141 granule formation. It i s not known whether nuclear migration occurs in DPVOs which have no detectable f o l l i c u l a r development. However, in oocytes degener-ating at this stage, the nucleoli do not remain at the periphery of the c e l l but are found in the general cytoplasm after nuclear breakdown. In oocytes which have become atretic at a slightly more advanced stage i n which the f o l l i c l e i s partially developed, small numbers of nucleoli can be found within the scattered hypertrophied f o l l i c u l a r c e l l s , the rest remaining i n the cytoplasm (Fig.19). As degenerating oocytes are found in which hypertrophy of f o l l i c u l a r cells and uptake of nucleoli occur and yet in which no yolk vesicles can be detected, i t appears that f o l l -icular envelopment of the oocyte is completed prior to yolk vesicle formation. It seems that f o l l i c u l a r hypertrophy in the partially envel-oped previtellogenic oocyte is transient and that no corpus atreticum i s formed. In the early atretic stages of oocytes with developed f o l l i c l e s , nucleoli recently deposited at the periphery are usually, clumped in a small area; the degenerating nucleus often is seen close to these nucleolar concentrations (Figs. 8 and 15). Eventually the nucleoli are distributed f a i r l y evenly within the hypertrophied f o l l i c u l a r cells (Fig.3); i t i s not known how this i s accomplished. Unlike the situation in degenerating oocytes without developed f o l l i c l e s , nucleoli are not found free in the cytoplasm. In DPVOs and DEYVOs, the stage of nuclear migration is also characterized by invasion of macrophages which i n i t i a l l y tend to accumu-late in and around the degenerating nucleus. In some instances, 142 the nucleoli at thisvstage are found in clusters (Fig.20). Two readily distinguishable types of macrophage are found both i n DPVOs and in DEYVOs. The most common, which w i l l be referred to as M-l type macrophages, are small, usually spherical but occasionally ellipsoidal cells with clear cytoplasm and small central nuclei (Figs. 6, 20, and 21). The second type (M-2) is a larger, roughly spherical c e l l with granular cyto-plasm and a large, eccentric, bilobed nucleus (Fig,s.6 and 22). As cells similar to both ,M-l and M-2 macrophages are usually very numerous in the channels between, the ovarian lamellae and often are found ad-jacent degenerating oocytes (Figs. 23 and 24), i t appears that the free macrophages of DPVOs and DEYVOs are of extrafollicular origin. Although this was not quantified, i t appeared that the less common M-2 macro-phage occurred most frequently in the less developed atretic oocytes. Macrophages are present i n DPVOs from the time of nuclear degenera-tion u n t i l a l l oocyte constituents have been removed. When this has occurred, the non-hypertrophied f o l l i c u l a r envelope may remain intact around the cavity (Fig.25); however, many of these envelopes apparently disintegrate. Thus, ih many long-term hypophysectomized , and in several intact, regressed fi s h , i t appeared that extensive atresia of p r e v i t e l l -ogenic oocytes had created large open spaces in the ovary (Figs. 26 and 27). This was not simply the result of atresia of large yolky oocytes and shrinkage of the resulting corpora atretica, as in that process the entire ovary decreases in volume while maintaining close association of normal and atretic oocytes (Fig.28). In addition, corpora atretica are persistent structures which are evident for at least several months:,after their formation. 143 After cytoplasmic debris has been removed, DEYVOs resemble the Y - stage atretic f o l l i c l e s described by Khoo (1975), except that they are smaller (usually less than 50y in cross section). Eventually these small structures collapse to form persistent 6- stage corpora atretica. No experiments were conducted to investiage the endocrine basis of degeneration of early yolk vesicle or previtellogenic oocytes. However, some incidental observations are worth noting. In none of the experiments reported i n this study was there any indication that treatment with exogenous steroids induced the formation of yolk vesicles. There, was, however, some indication that injection of estradiol or 5°= - dihydrotestosterone delayed onset of atresia in early yolk vesicle oocytes following hypophysectomy (Experiment 5). For example, whereas 5 of 8 fish receiving estradiol and 4 of 8 fis h receiving 503 - dihydrotestosterone had small numbers of normal yolk vesicles in their oocytes, there were no normal yolk vesicles in any of 6 saline-treated fish. That this indicated inhibition of atresia rather than induction of yolk vesicle formation i s suggested by the fact that whereas a l l 6 saline-injected females had many DEYVOs, these were found in low numbers in only 4 of the estradiol and 4 of the dihydrotestoster-one groups. In addition, yolk vesicles associated with the egg membrane were abundant in 6 of the fish receiving estradiol and in 5 of the fish receiving dihydrotestosterone, whereas these structures were seen in only 3 of the fish receiving saline, and then only i n low numbers. This i s interpreted to mean that in fish treated with saline, the proposed transition of normal early yolk vesicle oocytes to oocytes with mem-brane vesicles and fin a l l y to DEYVOs has proceeded more rapidly than i t 144 has in fish receiving estradiol or dihydrotestosterone. In a l l three experiments involving hypophysectomy and injection of pituitary material, there was good evidence that both homogenized goldfish pituitaries and partially purified salmon gonadotropin (SG-G100) inhibited atresia following hypophysectomy. In the two experiments in which replace-ment therapy began within one month of hypophysectomy, none of J.11 fish receiving homogenized pituitaries (Experiment 6) and only 1 of 10 fish re-ceiving SG-G100 (Experiment 7) had DPVOs in their ovaries. In contrast, following estradiol treatment, 7 of 8 fish i n Experiment 6 and 8 of 10 fis h in Experiment 7 had DPVOs. SG-G100 appeared to be less effective in inhibiting atresia in long-term hypophysectomized goldfish, as the ovaries of 4 of 15 females treated with SG-G100 in Experiment 11 contained DPVOs; however, this was s t i l l a lower incidence of atresia than that found in fish treated with estrogen (8 of 12 fish) or saline (13 or 16 fish). As females receiving SG-G100 also had more yolk vesicles associated with the egg membrane, i t is possible that some of the atresia involved oocytes in which yolk vesicles were induced by the gonadotropin. Most studies of oocyte atresia in teleost ovaries have dealt with the degeneration of second growth stage (vitellogenic) oocytes (references cited by Khoo, 1975) which are generally accepted to be dependent on the pituitary (review by Dodd, 1972). Previtellogenic oocytes have been sug-gested to be independent of pituitary influence (Dodd, 1972) and atresia of these oocytes has received scant attention i n the literature. Beach (1959) described corpora atretica in goldfish which he believed were de-rived from previtellogenic oocytes; however, as these structures had a hypertrophied granulosa layer, they may have developed from early yolk 145 ve s i c l e stage oocytes. MacKay (1973) reported 'small corpora a t r e t i c a apparently derived from degeneration of previtellogenic oocytes' i n ovar-ies of f i r e t a i l gudgeons (Hypseleotris g a l I i i ) receiving methallibure treatment. As these structures, found only i n females given long-term (2 month) methallibure treatment, appeared to have no granulosa layer, i t i s l i k e l y they were DPVOs. The results of the present study indicate that the Correlation be-tween the post-hypophysectomy degeneration of teleost oocytes and the presence of yolk deposition i s not as simple as has been suggested i n e a r l i e r work. Although these findings are only suggestive, they point to a greater p i t u i t a r y influence on previtellogenic oocytes than has been recognized previously. APPENDIX FIGURES 147 Figure 2: Early stage of degenerating early yolk vesicle (DEYVO) or previtellogenic (DPVO) oocyte. A few red nucleoli (N) are seen at the periphery. Mallory trichrome. Figure 3: Follicular hypertrophy more advanced than in Figure 2. Mallory trichrome. Figure 4: Advanced DEYVO from intact regressed fish. Hyper^ trophied f o l l i c l e surrounds empty cavity. NSte that presumptive nucleoli (N) are s t i l l numerous. Mallory trichrome. 149 Figure 5: Advanced DEYVO from female goldfish hypophysectomized for 2 months. PAS-positive yolk vesicles (V) and green nucleoli (N) are s t i l l obvious. PAS-light green. Figure 6: Advanced DPVO from fish hypophysectomized for 2 months. Most oocyte contents have been removed and several M-l (1) and M-2 (2) macrophages remain. Note absence of f o l l i c u l a r hypertrophy. Mallory trichrome. Figure 7: Yolk nucleus in oocyte of fis h hypophysectomized for 4 months, PAS-light green. Figure 8: Nuclear migration stage in oocyte from fi s h hypophysec-tomized for 4 months. Note breakdown of nuclear structure and loss of cytoplasmic basophilia. Mallory trlchrome. 151 Figure 9; Normal configuration of yolk vesicles (V). Note vesicles tend to form a ring and do not contact the oocyte membrane. Mallory trichrome. Figure 10\ As i n Figure 9 but stained with PAS-light green. Figure 11; Yolk vesicles (V) associated with membrane of presumptive pre-atretic oocyte. From fish hypophysectomized for 2 months. Mallory trichrome. Figure 12: As in Figure 11 but stained with PAS-light green. 153 Figure 13: Yolk vesicles (V) and presumptive nucleoli (N) in hypertrophied f o l l i c u l a r layer from fish hypophysectomized for 2 months. PAS-light green. Figure 14: Presumptive pre-nuclear breakdown stage oocytes (0) in ovary of fish hypophysectom-ized for 2 months. Note abnormal yolk vesicle distribution. Mallory trichrome. Figure 15: Degenerating nucleus at early stage of atresia. Spherical body (S) appears outside nuclear membrane and is very lightly stained. Presumptive nucleoli (N) are seen both within the nuclear membrane and within the presumpt-ive granulosa cells (G) in the slightly hyper-trophied f o l l i c l e . From intact regressed female. Mallory trichrome. Figure 16: Peripheral nucleus with spherical body (S) and several nucleoli (N). F o l l i c l e is not hypertrophied. Mallory trichrome. Presumptive DEYVO from fish hypophysectomized for 2 months. Nucleus i s unusual as i t stains differently than the cytoplasm and appears to retain some nucleoli. Mallory trichrome. Advanced DEYVO from intact, regressed f i s h . Degenerated nucleus with spherical body (S) is present. Hypertrophied granulosa cells (G) and presumptive nucleoli (N) can also be seen. Mallory trichrome. Advanced DPVO from intact, regressed f i s h . Presumptive nucleoli (N) are seen both in the slightly hypertrophied f o l l i c l e and in the cytoplasm. Spherical body (S) and M-l macro-phages (1) also are present. Mallory trichrome Clumped nucleoli (N) from degenerating nucleus atretic oocyte of fish hypophysectomized for 4 months-M-l macrophages (1) can be seen. Mallory trichrome. 157 Figure 21: M-l macrophage (1) and presumptive nucleoli (N) i n nuclear debris of oocyte of fish hypophysectomized for 4 months. PAS-light green. Figure 22: M-2 macrophage (2) i n atretic oocyte of intact, regressed fish. Mallory t r i -chrome . Figure 23: Macrophage aggregation adjacent to early atretic oocyte. Mallory trichrome. Figure 24: Macrophages entering early atretic oocyte in ovary of intact regressed fish. Mallory trichrome. 159 Figure 25: Ovary of female goldfish hypophysectomized for 4 months. Two nearly evacuated DPVOs (E) can be seen. Older presumptive DPVOs (D) are apparently disintegrating. Mallory trichrome. Figure 26: Ovary of fish hypophysectomized for 4 months showing open spaces apparently caused by extensive atresia of previtellogenic oocytes. Mallory trichrome. Figure 27: As in Figure 26 but at a lower magnifi-cation. Mallory trichrome. 161 Figure 28: Ovary of female g o l d f i s h hypophysectomized for 2 months showing close association of normal (N) and a t r e t i c (A) oocytes. The two large oocytes (0) are ovulated oocytes (from another female) which were in j e c t e d through the ovipore. Mallory trichrome. Figure 29: Macrophage aggregation adjacent to i n j e c t e d ovulated oocyte. D e t a i l from Figure 28. Mallory trichrome. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items