Reproductive patterns in three species of large coral reef sponges

The type and frequency of reproduction of three common sponge species was monitored histologically and by direct field observations of spawning events over 2 successive years at the reefs of Curaçao. Ircinia strobilina showed year round reproductive activity by the production of spermatic cysts in varying intensities in a major part of the population. Production of oocytes and larvae were only observed in the period from September through April, indicating an actual breeding season of 8 months a year. The sexes in this viviparous sponge seem to be separate, but protandry cannot be ruled out. Less than 10% of the sponge tissue was found to consist of reproductive elements. Neofibularia nolitangere is a gonochoristic oviparous sponge with a short annual breeding season of 2 months, in which two successive, highly predictable, short spawning periods occurred synchronously for the whole population. Up to half of the sponge tissue was transformed to reproductive elements. Agelas clathrodes showed low reproductive activity. Field observation suggests that this sponge is an oviparous hermaphrodite, probably annually releasing male and female gametes synchronously in a varying number of individuals of its population during a short breeding period.The variation in reproductive patterns in this small group of large massive coral reef sponges spans a range similar to that found in the morphologically highly variable group of stony corals and illustrates their ecological differentiation on the reef.     

Aspects of the spawning of western Atlantic butterflyfishes (Pisces: Chaetodontidae)

Synopsis The status of knowledge of spawning among the five shallow waterChaetodon species in the western Atlantic is reviewed. Spawning has been observed for three species in Puerto Rico, St. Croix and the Bahamas, with possible courtship in a fourth.Chaetodon aculeatus spawned near the time of sunset over objects on the reef as single female/male pairs or as two females and one male, with pair spawning in rapid succession. Spawning occurred during much of the lunar month from February to April and it is uncertain whether any lunar periodicity to spawning exists. Male-male aggression was noted. Spawning sites (coral heads) were alternated daily and it is likely that females spawn only once every two days. A single female produced as many as 2090 eggs in a single spawning.Chaetodon capistratus spawned during much of the lunar month from February to April. It spawned about 5 min afterC. aculeatus, occasionally using the same sites, and alternated sites daily. A female produced as many as 3710 eggs in one spawning.Chaetodon striatus spawned from February to April but it is unknown if it has any lunar spawning cycle. No predation attempts by piscivores on spawning adults were seen. Predation byMelichthys niger on eggs ofC. striatus occurred. No egg predation was observed forC. aculeatus andC. capistratus. With an assumed four month reproductive season, alternate day spawning and observed egg production values,C. aculeatus andC. capistratus produce respectively about 100 000 and 200 000 eggs per large female per year. The reproductive strategy of smaller species may be to produce moderate numbers of eggs per day over a spawning season of at least a few months while larger species may produce more eggs per day for a shorter period.     

Isotope analysis reveals proportional change and site-selection variation of river- and lake-produced eggs of a landlocked migratory fish

Changes in the proportions of river- and lake-produced eggs of a landlocked amphidromous fish, ayu (Plecoglossus altivelis altivelis) in the Lake Biwa water system, Japan, were monitored by stable isotope analysis, based on different δ¹⁵N and δ¹³C values of prey organisms between the lake and its tributaries. During the 3 month reproduction season, the δ¹⁵N values of spawned eggs decreased with time. This result implies that there was a shift from lake-produced eggs to river-produced eggs within a reproductive season, based on the observation that adult fish in the lake had previously been shown to have eggs with distinctly higher δ¹⁵N values in their ovaries than those in the tributaries. This explanation was also supported by the change in δ¹³C values of the spawned eggs. Furthermore, eggs with lower δ¹⁵N and higher δ¹³C values tended to be spawned at less variable depths, suggesting that females spawning river-produced eggs selected the spawning sites from a narrower range. We conclude that stable isotope ratios of spawned eggs can be indicators of the relative contributions of different food chains and can enable comparisons of reproductive characteristics between types of egg.     

Reproduction in puyen, Galaxias maculatus (Pisces: Galaxiidae), in the southernmost extreme of distribution

Reproductive aspects of a peripheral population of Galaxias maculatus are described and the relationship with the physiology and handling of energetic reserves under marginal environmental conditions is investigated. The G. maculatus population of Tierra del Fuego has an extended reproductive season, with differences in timing and duration compared to other populations of continental Patagonia, New Zealand and Australia. Elevated gonadosomatic indexes (I_G) were observed during this period in both sexes (maximum I_G = 33.49% males; 35.94% females). The high abundance of mature males (with high I_G values) on the spawning grounds during the reproductive season suggests that they were waiting for the return of the mature females. Larger females reached total maturation at the beginning of the reproductive season, whereas the size of maturing females diminished toward the end (mean TL = 96 mm, October; 70 mm, February). Both sexes showed an extremely high investment in reproduction, reaching a maximum I_G of about 35% for both sexes. At the beginning of the reproductive season females reached the maximum median I_H (3.37%) and males the minimum (0.96%), suggesting differences in the role of the liver in the management of energetic reserves during sexual maturation. The variation in the fat index (I_F) suggests that fat reserves were used to survive winter (maximum median I_F > 1%, autumn; minimum about 0.2%, spring).     

Life history characteristics of the protogynous parrotfish Calotomus japonicus from northwest Kyushu, Japan

In this study, we examined the life history characteristics of the parrotfish Calotomus japonicus, using individuals collected between May 2003 and May 2008 off the Nagasaki Peninsula in northwest Kyushu, Japan. Age determinations were performed using scales. Marginal increment analysis revealed that growth rings were formed annually around July. Growth in both sexes was fitted to the von Bertalanffy growth function (L _∞ = 513, k = 0.28, t ₀ = 0.03, where L _∞ is the theoretical asymptotic total length in mm, k is the growth rate coefficient and t ₀ is the theoretical time at zero length). Observed maximum age for both sexes was 8 years. We also characterized the reproductive biology of this species based on a gonadosomatic index and histological examinations of the gonads. The spawning season extends from July to October, with peak spawning activity occurring during July and August. Fish reach sexual maturity by the second year of life. Females are assumed to be multiple spawners, since we observed specimens with postovulatory follicles in ovaries containing either yolk globule oocytes or migratory nucleus oocytes. All males had secondary testes, which were characterized by the presence of an ovarian lumen structure and sperm sinuses in the gonadal wall. This indicates that all males, irrespective of whether they were initial or terminal phase males, had undergone a sexual transition. Sex change appears to occur during the spawning season, and thereafter sex-changed males are able to fertilize female eggs throughout the remainder of the current spawning season.     

Growth,reproduction, and feeding oF Fundulus heteroclitus (L.) on a North Carolina salt marsh

Fundulus heteroclitus (L.) from a North Carolina Spartina marsh teed largely on small crustaceans (amphipods, tanaids and copepods) and polychaetes. Fish longer than 30 mm standard length also ingested considerable amounts of living plant material. Smaller individuals were distinctly carnivorous. Recognizable particles of Spartina detritus occurred in less than 15% of the guts examined. The relationship of weight to length changed significantly during the year. Although females were larger than males of the same age. males were heavier than females of the same length, except for a brief period at the peak of the spawning season in the early spring. An average second season fish may lay up to 512 eggs from March through August, but first season fish did not reach reproductive size by the end of the spawning season. Growth of first season fish in mid-summer averaged 5% of their total weight per day. These significant seasonal changes in ecological properties of killifish populations are important in any estimates of growth, reproduction, and production.     

Non-copulatory spawning and female participation during early egg care in a marine sculpinHemilepidotus gilberti

The reproductive behavior of a marine sculpin,Hemilepidotus gilberti, was observed in the aquarium. The behavioral sequence comprised fight for terriotory, courtship and fight for females, spawning and caring for the egg mass. Females very slowly deposited eggs enclosed with viscous ovarian fluid onto the substrate. Almost all the eggs contained in the ovary were released in a single spawning, its completion requiring 21–22 min. Territorial males emitted of spawning. Sneaking by neighboring males followed subsequently. This was the first observation of non-copulatory spawning among marine sculpins. Females cared for the eggs for severl days after spawning, and then male alone care continued until hatching (47–49 days after spawning). Participation by females in egg caring is likely to be caused by their spawning only once during the spawning season and the consequent lack of urgency for post-spawning feeding.     

The choice of oviposition site and the effects of density and oviposition timing on survivorship in<Emphasis Type="Italic"> Rana japonica</Emphasis>

In species that lack parental care and whose larvae are restricted to a given space throughout development, choice of egg-laying or reproductive sites often greatly influences fitness. Furthermore, reproductive success of individuals may be affected not only by conditions of the breeding area but also by the time when they start to breed, because intraspecific variation in the timing of breeding may result in asymmetric competition between larvae. We addressed these issues in an experimental and field study of the Japanese brown frog, Rana japonica. We tested whether the survival success of larvae decreased with an increase in the number of deposited egg masses and with a greater delay in oviposition. We found that the number of egg masses and timing of oviposition together significantly predicted larval survival per clutch. Moreover, we observed the natural oviposition of R. japonica to examine the prediction that if the density of larvae and the timing of oviposition affect survivorship of larvae, R. japonica will avoid depositing eggs in pools in which egg masses have already been deposited. We found that for small pools only, R. japonica tended to favor oviposition at unoccupied pools (those lacking egg masses and larvae) in the spawning season. For large pools, however, adults favored those already occupied by egg masses in the spawning period. The density of egg masses and tadpoles or the timing of oviposition may have less effect on spawning or the survival of tadpoles in large pools than in small pools.     

Reproduction and distribution of two species of Goniastrea (Scleractinia) from the Great Barrier Reef province

The reproductive biology of Goniastrea aspera at Magnetic Island was compared with that of a very similar sympatric species, G. favulus as reported by Kojis and Quinn at Heron Island. The development of gametes was similar in both species, but there was no evidence for an adolescent protandrous period of development in G. aspera such as that recorded for G. favulus at Heron Island. Other reproductive differences between the two species were found in egg size and the mode of spawning. The eggs of G. aspera are smaller and more numerous than those of G. favulus. Goniastrea aspera expelled buoyant packets of eggs and sperm, while G. favulus had sticky sinking eggs which were released separately from the sperm. The spatial pattern of the two species was examined on the reef flat at Magnetic Island to determine whether the observed differences in spawning behaviour and egg buoyancy might have an effect on egg retention and the distribution of adult colonies. The results of this comparison failed to detect any difference in the degree of aggregation of the two species. This is not the result which would be expected if sticky sinking eggs helped retain developing larvae in the vicinity of the adult. These results, together with evidence from a wide range of coelenterates and observations on the larvae of G. aspera point to post spawning larval behaviour as the most likely factor in determining where these species will settle.     

Reproductive patterns of the Pacific oyster Crassostrea gigas in France

Summary

In France, national management programs focus research on understanding reproductive factors in Crassostrea gigas to confront problems of the oyster industry. However, little information has been documented in which reproductive patterns include sexual changes. The reproductive cycle of oysters at three sites of the Atlantic coast of France was examined from 1996 to 1998, and the seasonal variations in oocyte size-frequencies, and sex ratio were described. The results showed a synchronism within the population concerning reproductive behavior. Young oocytes are generated after spawning and show no apparent changes during winter. Growth of oocytes begins in spring and cells reach maturity in April-May and are ready for a single spawning season in June-July. Oocytes that were not released during spawning are reabsorbed within the gonad. The significant difference between sites is that spawning occurred 1 month later in the southern area. A modal analysis showed that oocyte populations in the sample individuals are primordially bimodal, but with polymodal occurrences in June-July, in some cases. Irregular alternative sexuality was detected at all sites, and hermaphrodites appear to be a transition phase that allows changes from male to female during early spring. Previous observations, together with the study of the development of oocyte cohorts over time, permit a hypothetical model concerning the kinetics of gametogenesis in C. gigas. The model suggests that primary oocytes are generated from energy supplied from degenerating, as well as young oocytes that do not reach the mature stage within the gonad during autumn-winter. It seems that, during vitellogenesis, there is disintegration of smaller cells coupled with transfer of energy to the larger oocytes, which continue to grow and mature.     

Spatial variation in the effects of size and age on reproductive dynamics of common coral trout Plectropomus leopardus

The effects of size and age on reproductive dynamics of common coral trout Plectropomus leopardus populations were compared between coral reefs open or closed (no‐take marine reserves) to fishing and among four geographic regions of the Great Barrier Reef (GBR), Australia. The specific reproductive metrics investigated were the sex ratio, the proportion of vitellogenic females and the spawning fraction of local populations. Sex ratios became increasingly male biased with length and age, as expected for a protogyne, but were more male biased in southern regions of the GBR (Mackay and Storm Cay) than in northern regions (Lizard Island and Townsville) across all lengths and ages. The proportion of vitellogenic females also increased with length and age. Female P. leopardus were capable of daily spawning during the spawning season, but on average spawned every 4·3 days. Mature females spawned most frequently on Townsville reserve reefs (every 2·3 days) and Lizard Island fished reefs (every 3·2 days). Females on Mackay reefs open to fishing showed no evidence of spawning over 4 years of sampling, while females on reserve reefs spawned only once every 2–3 months. No effect of length on spawning frequency was detected. Spawning frequency increased with age on Lizard Island fished reefs, declined with age on Storm Cay fished reefs, and declined with age on reserve reefs in all regions. It is hypothesized that the variation in P. leopardus sex ratios and spawning frequency among GBR regions is primarily driven by water temperature, while no‐take management zones influence spawning frequency depending on the region in which the reserve is located. Male bias and lack of spawning activity on southern GBR, where densities of adult P. leopardus are highest, suggest that recruits may be supplied from central or northern GBR. Significant regional variation in reproductive traits suggests that a regional approach to management of P. leopardus is appropriate and highlights the need for considering spatial variation in reproduction where reserves are used as fishery or conservation management tools.     

Reproductive behaviour of the filefishRudarius ercodes: male spawning parades and female choice

Synopsis The reproductive behaviour ofRudarius ercodes is described from undersea observations in Aburatsubo Bay, Japan. Reproductive behaviour can be separated into four parts: (1) Prespawning Search = searching for spawning sites by females and searching for gravid females by males, (2) Spawning Parade = males follow a gravid female in a line, competing with each other to reach the head of the queue, (3) Spawning = the female takes the spawning position, males rush to the side of the female, and mating occurs between one female and several anterior males of the spawning parade, (4) Parental Care = females attach adhesive eggs to seaweed with the mouth and guard them until embryos hatch. There is no male parental care. The reproductive season ranges from May to October and spawning occurs early in the morning every day. Females begin feeding early in the morning, but males feed little at this time.R. ercodes shows neither territorial behaviour nor fixed-pair spawning. One male might spawn several times in one morning. One female spawns at most once every 5 days. The mating system of this species is promiscuous. The probable function of the spawning parade as a style allowing female choice is discussed.     

Assessment of Maturity,Reproduction and Reproductive Potentials of Torpedo Scad, <Emphasis Type="Italic">Megalaspis cordyla</Emphasis> (Linnaeus, 1758) from Northern Arabian Sea Coast of Pakistan

Port landing specimens of torpedo scad, Megalaspis cordyla were collected at monthly intervals from August 2013 to June 2014 from the northern Arabian Sea coast of Pakistan. The life history characteristics revealed that torpedo scad has a prolonged breeding season which occurred with two peaks: from May to June and from November to February. The spawning seasonality was also validated with microscopic gonad examination.The highest GSI value was noted in May and January whereas the lowest in November was recorded for both sexes. The age at which individuals of both sexes become 50% mature (t₅₀) was estimated as 2.44 year. Life span (t_max) calculated as 3.96 years for male and females. The size of 270 and 265 mm at which 50% individuals attained sexual maturity (L₅₀) was determined for males and females respectively. The reproductive potential (fecundity)was made from ten ripe females ranged from 29–43.8 cm TL and weighing 220–540 g of IV–V developmental stage was enumerated as 32306–236677 eggs (CV = 2.135) with an average number of eggs 86544 counted per ripe female/season. This study contains fundamental data of an important fish species in the northern Arabian Sea coast of Pakistan. It provides vital information to the fisheries managers for implementation on sustainable use of the resource during spawning season.     

Two spawning seasons and mating system of the bastard halibut, Tarphops oligolepis

We investigated the breeding habits of the bastard halibut, Tarphops oligolepis, in the southwest of Kyushu Island, Japan. This fish was found to have two spawning seasons in a year (around July and November); individual fish spawned over the two spawning seasons. During the spawning seasons, males established territories. Home ranges of females overlapped with those of other females and with territories of plural males. The courtship partner changed during a day, and multiple matings among both sexes were observed during a spawning season, suggesting that T. oligolepis bred promiscuously. This is the first report of the mating system among Paralichthyidae. Received: October 18, 2000 / Revised: May 7, 2001 / Accepted: July 9, 2001     

Gonad maturation and spawning of cobia,Rachycentron canadum (Linnaeus, 1766) off the Dungun coast,Malaysia

The reproductive biology of cobia, Rachycentron canadum, from the coastal waters of Dungun, Malaysia was studied from June 2014 to May 2015. From commercial trawls, a total of 201 samples (combined sexes) were collected (fork lengths [FL] 37.5–124.0 cm; body weights 0.5–20.4 g). The overall sex ratio of females to males was 1:0.9, which was not significantly different (χ² = 2.12, df = 1; p < .05). Estimations of length at 50% maturity (L₅₀) showed that both sexes matured at approximately 75 cm FL; estimated spawning frequency was 6 days. Mean batch fecundity (BF) ranged from 0.55 to 4.32 million eggs. The average number of eggs per gram of ovary was from 2,100 to 5,400 eggs, and relative fecundity 147 eggs/g. There was a weak positive correlation (r² = .48) between BF and female FL as well as BF with an ovary‐free body weight (r² = .56), possibly due to females being in a continuous spawning condition and some possibly half‐spent, based on the histological examination of the female gonads. Despite cobia being asynchronous spawners, the gonadosomatic index in both males and females showed peaks in June, November, and particularly March. Based on histological examination, spawning‐capable males were encountered throughout the study period, whereas spawning‐capable females in the late developing subphase were found mostly in March and April. This is the first study on the reproductive aspects of cobia in Malaysian waters.     

Reproductive biology of brown trout, Salmo trutta macrostigma Dumeril 1858, in a tributary of the Ceyhan River which flows into the eastern Mediterranean Sea

The study describes some key elements of the reproductive biology, including spawning season, age at sexual maturity, fecundity and egg diameter of the native brown trout, Salmo trutta macrostigma, in a tributary of the Ceyhan River. A total of 197 brown trout (118 females and 79 males) were captured in 2000–2001 by electric fishing. In observations on monthly changes, the gonadosomatic index (GSI) and the monthly frequency distribution of egg diameter confirmed that spawning lasted from November to January. Some 27.7% of the females and 62.5% of the males attained sexual maturity in their second year. The smallest fork length (FL) of brown trout attaining sexual maturity was 17.4 cm for males and 17.8 cm for females. Mean fecundity in age groups II, III, IV and V were 360, 452, 693 and 1283 eggs per female, respectively. One 9‐year‐old female had a unique 3232 egg count. The mean fecundity of the sampled population was 554 eggs per fish, positively correlated with the FL (mm) (R = 0.8227 ) and body weight (R = 0.8130). The diameter of mature eggs in the spawning season ranged from 3.250 to 5.930 mm, with a 4.146 mm average. Mean egg diameter in age groups II, III, IV and V in the spawning season were 0.813, 3.799, 4.663 and 5.243 mm, respectively. Fecundity, egg weight and diameter were statistically different in all age groups.     

Life history of the endangered Japanese striped loach,Cobitis kaibarai (Cypriniformes: Cobitidae), with special reference to its reproductive ecology and the influence of creek reshapings on its population density

In recent years, the biodiversity of freshwater fishes has been markedly decreasing worldwide because of anthropogenic activities. The Japanese striped loach, Cobitis kaibarai (Cypriniformes: Cobitidae), is a primary freshwater fish endemic to northern Kyushu, Japan. This species is designated as endangered IB class in the Red List by the Japan Ministry of the Environment. Its population is decreasing, possibly because of habitat loss and degradation. To conserve C. kaibarai populations, information on its basic ecology is necessary; nonetheless, its detailed life history and reproductive ecology have yet to be clarified. In this study, the authors conducted monthly capture–mark–recapture surveys and periodical observations to investigate the life history, spawning sites and season of C. kaibarai. They also evaluated the influence of creek reshaping (concrete revetment) on the C. kaibarai population in Saga Plain, northern Kyushu. Between 2015 and 2018, more individuals were captured during winter than summer. The average body width of females peaked in early June and small immatures were confirmed from July. Some individuals were captured across 15 or more months after their initial marking. In the survey of reproductive sites, eggs and larvae of C. kaibarai were found in shallow areas in mid-June; these were temporarily submerged following the increase in water level from early June. Therefore, C. kaibarai spawns in shallow areas during this season. Based on the capture–mark–recapture surveys, the estimated population density of C. kaibarai significantly decreased in a census site that had undergone creek reshaping, which contrasted with the results in a control site, where no significant difference was detected. The standard length of C. kaibarai increased following creek reshaping, suggesting that the proportion of C. kaibarai postponing spawning had increased, possibly because of degradation of spawning environments. The results of this study provide important ecological knowledge for the conservation of C. kaibarai and emphasize the importance of shallow waters for floodplain spawners.     

Patterns of migration between feeding and spawning sites in a coral reef surgeonfish

Many coral reef fishes exhibit regular localised migrations between feeding and spawning areas, but the factors affecting these migration patterns, such as the distance, frequency and spawning site fidelity are poorly understood. The aim of this study was to investigate the patterns of migration to spawning sites of the surgeonfish, Ctenochaetus striatus (Acanthuridae). We explored relationships amongst an individual’s size and sex, the distance and frequency it migrated from its feeding area to spawning sites, fidelity to particular spawning sites and the number of individuals that aggregated to spawn. In order to achieve this, 406 C. striatus were captured and tagged on inshore reefs in Kimbe Bay (5°30′S 150°6′E), New Britain, Papua New Guinea. Tagged individuals were consistently observed within spatially discrete but overlapping feeding areas (maximum diameter averaging <13 m). The mean distance migrated was 58 m (ranging from 2 to 291 m). No tagged individuals were witnessed spawning at more than one site. Whilst most individuals (n = 88) migrated to the spawning site that was closest to their feeding areas, those that migrated to sites further away (n = 9) always spawned at sites where the number of conspecifics aggregating was larger. Neither the size nor the sex of individuals limited migration distance. However, males migrated significantly more frequently than females (on average once every 2 days vs. once every 3 days), and migration frequency was positively correlated with size in females. Migration distance did not affect the frequency with which individuals spawned. Whether patterns of migration are determined by cost-benefit optimisation, tradition, or an alternative mechanism is unknown.     

Sexual reproduction of the brooding sponge <Emphasis Type="Italic">Rhopaloeides odorabile</Emphasis>

Sexual reproduction of Rhopaloeides odorabile, a common Great Barrier Reef sponge, was quantified using histological sections from samples collected over two reproductive seasons. Rhopaloeides odorabile is viviparous and gonochoristic with a male biased sex ratio (2.5:1). Commencement and cessation of gametogenesis coincides with rising and falling sea surface temperatures (≈24–29°C). Spermatogenesis occurs from October until January. Females initiate oogenesis in September with the asynchronous development of oocytes, embryos and larvae occurring within the brood chambers. A larval release period of 5–6 weeks occurs during January and February. The minimum size of reproductive individuals was 176 cm³ for females and 192 cm³ for males. The total reproductive output for both sexes shows a positive correlation with size. A relative reproductive output index quantified the maximum reproductive investment at approximately 1 and 3% of the total choanoderm for females and males, respectively, which represents a lower range of reproduction in comparison to other viviparous sponges. Low reproductive output in R. odorabile may be offset by an extended spawning period, alleviating the risk of releasing larvae in potentially adverse conditions via one synchronised spawning event and increasing overall larval survival.     

Experimental test of the effect of maternal hormones on larval quality of a coral reef fish

Maternal hormones can play an important role in the development of fish larvae. Levels of the stress hormone, cortisol, in females are elevated by social interactions and transferred directly to the yolk of eggs, where they may influence developmental rates. In some vertebrates, prenatal exposure to high levels of testosterone determine early growth rates, social status and reproductive success. The present study examined whether post-fertilization exposure of eggs of the tropical damselfish, Pomacentrus amboinensis (Pomacentridae), to natural levels of cortisol or testosterone directly affects larval morphology at hatching. Maternal and egg levels of cortisol and testosterone varied widely among clutches of eggs from local populations around Lizard Island on the Great Barrier Reef. The morphology of larvae produced by these local fish populations also varied widely and differed significantly among sites (e.g., standard length: 2.6–3.4 mm; yolk sac area: 0.01–0.13 × 10⁻² mm²). Laboratory experiments showed that elevated cortisol levels in the egg reduced larval length at hatching, while slight elevations in testosterone increased yolk sac size. The influence of testosterone, and to a smaller extent cortisol, on larval morphology differed among egg clutches. These differences were partly explained by differences in initial egg hormone levels. Morphological changes induced by experimental hormonal regimes encompassed the entire range of variability in body attributes found in field populations. It is unclear whether cortisol influences growth alone or development rate or both. Testosterone appears to influence yolk utilization rates, and has no significant effect on growth, in contrast to its role in later developmental stages. Maternally derived cortisol and testosterone are important in regulating growth, development, and nutritive reserves of the embryo and larvae of this fish species. Factors that influence the maternal levels of cortisol and testosterone may have a major impact on larval mortality schedules and, therefore, on which breeding individuals contribute to the next generation. Received: 19 August 1998 / Accepted: 16 November 1998