The evolutionary advantage for hermaphroditism in reef fish is a much debated topic, with numerous hypotheses. An animal which, as it grew, assumed the sex advantageous to its current size would thereby increase its reproductive potential. Avisea and J. Mankb is a wonderful place to start, and is referenced heavily here. Back to topic, sequential hermaphroditism occurs in three primary forms.
Protogyny , in which a species begins life as a female and is capable of changing into a male at a later stage; Protandry , the opposite of protogyny, in which a species begins life as a male and may later switch to a female; and serial bi-directional sex change , where a species is capable of switching back and forth between male and female. The first form of hermaphroditism, called protogynous proto -first, gyno -female hermaprodites, is the commonest and most prevalent of the three types.
It is widely known to occur in Labridae, as well as many other reef fishes including Pomacanthidae and Serranidae. In protogynous hermaphrodites, species are capable of sex change from females to males, and this is predicted to be evolutionary favoured when reproductive outputs of males outweighs that of being a female. Simply put, being a male rocks and he gets to have more sex, and therefore this theory supports the evolution of protogynous hermaphrodites.
As such, protogynous hermpahroditism is often seen in fish who display a haremic lifestyle, where most of the breeding is dictated and controlled by a large dominant male. Using Cirrhilabrus as an example, females and initial phased IP males are often smaller, drab and less showy than large terminal phased TP males.
IP males are primary, or satellite males that are born as males, while TP males are secondary males which arise from sex changed females. In a harem of Cirrhilabrus , the TP males are the brightest, most colourful males that monopolise most of the breeding rights. Take note that while Cirrhilabrus is an example of protogynous hermaphrodites, it is a diandric genus where males are not exclusively derived from females.
In monandric genera such as Centropyge and Genicanthus , males are exclusively derived from females. All fishes in monandric genera are born as females, and when the need arises, are capable of changing into males.
Regardless of monandric or diandric protogynous hermaphrodites, the basis of sex change is the same. With the removal of the terminal and dominant male from the harem, the next most aggressive female in line will sex change to take its place.
All female protogynous hermaphrodite species possess germ cells for both sex organs, and when the social situation calls for a change in sex, are capable of suppressing the female gonads in lieu of the development for male ones.
Testosterone production suppresses the female gonads and fuels the development of male sex organs and secondary characteristic. In sexually dimorphic and dichromatic species, this change is evident externally, with increased development of secondary male linked characteristics such as colours, finnage and other traits not found in females.
Above is an example of Genicanthus semifasciatus undergoing a sex change. All Genicanthus are sexually dichromatic, and so the process of sex change is easily documented based on physical attributes. The females are mainly unmarked sans a black gill cover, mask and tail lobes. The males are striped vertically halfway down the body, and in exchange for the black facial and tail markings, gains instead a yellow mask that runs along the limits of the vertical stripes.
During the changing process, a sequence of simultaneous colour changes occurs whereby the fishes loses female traits and gains male patterns at the same time. Internally, the male sex organs are also developing. In genera like Centropyge where sexual dichromatism is not so obvious, witnessing the transformation and therefore sexing the species is a little bit more tricky.
This change can take as little as weeks, and within that time, the changing female continues to express her dominance until the change is ready. By then, the now fully functional male can assume breeding privileges. The two large fish are the only sexually mature fish and are a male and female breeding pair. All of the smaller fish are male. If the large breeding female is removed, her male mate changes sex to female and the next largest fish in the group rapidly increases in size and takes over the role as the sexually mature male.
Can you imagine how different the movie Finding Nemo would have been had an ichthyologist IK-thee-ALL-uh-jist , a fish expert been asked for advice? A good example of a protogynous fish is the Indo-Pacific cleaner wrasse. If the male is removed from the harem, the largest female begins courting the other fish and develops male organs within two weeks. If you look at the different groups that have evolved hermaphroditism on the fish phylogeny, you can see that fish engaging in this is sort of gender-bending are scattered across the phylogeny.
This suggests that hermaphroditism has evolved independently in fish many times. Helfman, G. The Diversity of Fishes. Blackwell Publishing, Malden, MA. In addition, the reported high GSI values in this study were derived from only eight protandrous species 14 , three of which were in fact those of individuals reproducing as female rather than male, or, unlike for other species in the dataset, do not correspond to the maximum recorded male GSI value for the species and so are not comparable.
Finally, recent phylogenetic trees, which are essential for all comparative studies are better resolved and more comprehensive than those employed by earlier studies e. Here we have compiled the largest dataset to date of sexual systems, spawning modes and body size in the family Sparidae. Using modern phylogenetic comparative approaches, we have investigated the evolutionary history of sexual systems and tested the predictions of the SAM, considering spawning mode and mating systems, that protogynous and protandrous species should exhibit lower GSI values due to expected lower sperm competition Table 1 than gonochoristic species, while accounting for body size.
Our study therefore combines sexual systems, mating systems, sperm competition and the principles of the SAM in several sparid species, while accounting for the potential confounding effects of allometry and phylogeny.
We used FishBase www. We verified, and if necessary corrected, the sexual system reported in FishBase for each individual species used in this study against the primary literature 14 , 25 , 54 , 63 , We carefully revised previous assignments of sexual systems in four species, which were mainly based on the gonadal morphology of individuals collected at single or different ages Sometimes this approach cannot distinguish functional active hermaphrodites from non-functional hermaphrodites i.
The assignment of the correct sex is further complicated in non-reproducing juveniles, which can present a bisexual gonadal stage. Altogether, these peculiarities can make diagnosis of sexual system extremely challenging 14 , 54 , We resolved any discrepancies from previous studies using newly published data in which care was taken not to incur in the above problems Table S1.
Out of a total of recognized sparid species, we could assign the sexual system in 68 species Table S1. When several GSI values were reported for a given species e. We used two molecular phylogenetic trees with time-calibrated branch lengths, an essential step for robust analyses in a phylogenetic comparative framework Specifically, we used a phylogeny of Actinopterygians 67 based on a genes 6 mitochondrial and 21 nuclear genes and a phylogeny for the family Sparidae 68 , based on three mitochondrial and two nuclear genes.
These trees included 58 and 55 species, respectively, out of the 68 species with sexual system information in our dataset. The ancestral state reconstruction infers the evolutionary history of a trait along a phylogeny given the character states of species in the tree and provides estimates of the probable character state of each node in the phylogeny. This approach is based on a Markov model of evolution for discrete traits We reconstructed the ancestral character states of sexual system using maximum likelihood ML , setting all transition rates between G, PA and PG free to vary, i.
We ran these analyses on both phylogenetic trees using the R package ape However, we can only report the results of the ancestral state reconstruction using the phylogenetic tree of Rabosky et al. We used phylogenetic generalized least square models PGLS 61 , 71 , 72 to test the predictions of the SAM using the R package caper 73 and ML estimation, with both phylogenetic trees 67 , Possible allometric effects on the GSI were thus accounted for using either body length or weight as additional independent variable.
Continuous variables were log 10 -transformed to meet assumptions of normality, with the exception of GSI values. Results for GSI were qualitatively similar whether this variable was log 10 -transformed, transformed with logit function or left untransformed, thus we report the results of GSI in percentage, not transformed. All model residuals were normally distributed in all analyses. We also tested the SAM prediction within the genus Diplodus.
This was the only genus that provided limited but sufficient data to consider at least two different sexual systems G vs. PA for statistical analysis of the relationship between sexual system and male GSI values in very closely related species. Importantly, Diplodus species have a narrow range of body sizes and thus there is less variability and potential confounding effects of allometry.
However, the analysis could not be carried out in a phylogenetic context in this genus because too many species were missing from both phylogenies.
This represents a substantial increase in the number of species previously investigated Reconstruction of the ancestral character state in a phylogenetic context showed that gonochorism was only slightly more likely to be the ancestral sexual system in the Sparidae family likelihood at the root While both forms of sequential hermaphroditism, especially protogyny, evolved rapidly to gonochorism PA to G: 0.
Finally, the transitions between the two forms of sequential hermaphroditism were both estimated to be zero PA to PG: 0. Ancestral state reconstruction in the Sparidae using the phylogenetic tree by Rabosky et al.
Sexual system is coded as gonochorism grey , protandry blue and protogyny red. The pie area indicates the likelihood of character state at each node for the three states. Transitions rates between sexual systems in the Sparidae derived from the ancestral state reconstruction in maximum likelihood.
G: gonochorism; PA: protandry; PG: protogyny. We did not find any significant difference in total male body length Fig. For the 46 sparid species in the tree where male GSI values were available Table S3 , GSI values of protandrous and protogynous species were higher and lower respectively than that of gonochoristic species, with statistically significant differences between protogyny and the other two sexual systems Fig.
Results were qualitatively similar regardless the phylogenetic tree used Table S4 and were not influenced by allometric effects, tested using either length or weight as a covariate in the model Table S5. These low numbers did not allow testing predictions for mating systems formally. However, the data appear to suggest that protandrous sparids have higher GSI values than gonochoristic species regardless of whether they spawn in groups or pairs Fig. The black dots indicate individual values in species with information for the three variables: sexual system, spawning mode and male GSI.
We found no significant differences in weight t 2. However, the latter had a significantly shorter length than the former t 4. Importantly, despite being smaller in size, protandric Diplodus species had a significantly higher GSI t 4. The median is indicated by solid black horizontal line. The black dots indicate individual values. With a larger dataset of sexual systems in the family Sparidae than previously used, this study reveals that protandry and protogyny can evolve from gonochorism, although the ancestral state remains still uncertain in this family.
Importantly, we find that transitions between the two forms of sequential hermaphroditism are unlikely, if ever, to occur.
We find strong support for the SAM predictions, incorporating mating system and sperm competition, that protogynous species should exhibit lower levels of sperm competition relative to gonochoristic species Table 1 , as quantified by their low GSI values, consistent with their mating systems that allow large males to monopolize access to fertile females.
However, we find no evidence in support of similar predictions in protandrous sparids, i. Unexpectedly, protandrous species have similar GSI values to those of gonochoristic species and higher GSI than protogynous species, regardless of their mating system. Below we propose how a compensatory mechanism, together with mating system and spawning mode, may explain this unexpected finding.
Using twice as many species relative to an earlier study, recent molecular, time-calibrated phylogenies and modern phylogenetic comparative approaches, our study shows that gonochorism is only marginally more likely to be the ancestral state in this family. We find that gonochorism can evolve into both protogyny and protandry. However, sequential hermaphroditism is an evolutionary unstable state as it reverts quickly back to gonochorism, suggesting that both types of sequential hermaphroditism are costlier to sustain than gonochorism.
These results may explain why, despite hermaphroditism being anatomically and physiologically possible in fish in contrast to other vertebrates 75 , gonochorism predominates among fishes Thus, it is perhaps not surprising that our analysis reveals that the evolutionary transition rate from gonochorism to protandry is very low and that transitions from protogyny to protandry and vice versa are unlikely to occur; once canalized towards initial development as a female, it may be too costly to switch the developmental pathway to male-first sex changer and vice versa.
We tested whether the sparids conform to SAM predictions when incorporating sperm competition, mating system and spawning mode as previously done in protogynous and gonochoristic epinephelids Specifically, we tested whether gonochoristic species, which often spawn in large groups or aggregations and are typically characterized by intense sperm competition, have a higher GSI than species with either types of sequential hermaphroditism.
Sperm competition is indeed less intense in haremic species often protogynous , where the presence of few dominant large males drastically reduce the interaction between sperm of different males 24 , 35 , It is also generally accepted that protandrous hermaphrodites normally reproduce in small, random mating groups no size-assortative or in strictly monogamous pairs 23 , 24 , 77 as, for example, in most clownfish family Pomacentridae of the genus Amphiprion , such as A. In both mating systems, this would imply a low degree of sperm competition and thus low values of GSI, as predicted by the SAM 22 , 23 , These results, therefore, support the SAM prediction, incorporating sperm competition, in protogynous species Table 1.
However, our study also reveals that protandrous sparids have, on average, the highest male average GSI value 3. Thus, even when mating in pairs or small groups, protandrous males invest heavily in the gonads, indeed even more than gonochoristic species that mate in large aggregations with intense sperm competition.
Further, we demonstrate that these results are not determined by differences in body size, as we find no allometric effects on GSI values, neither across all species, nor within one genus Diplodus containing closely related species of similar size with different mating systems.
Altogether, our study unambiguously demonstrates that, while gonochoristic and protogynous sparids conform to a broader version of the SAM including sperm competition, spawning mode and mating system, protandrous sparids do not. We suggest that these results may be explained by both spawning mode and a compensatory mechanism determined by high sexual size dimorphism SSD in protandrous species. In protandrous species, sexual size dimorphism has been confirmed in several species such as Diplodus annularis and Lithognathus mormyrus 81 , However, more data to would be needed to confirm this observation.
Some protandrous sparid species like Acanthopagrus berda, Sarpa salpa 83 , Diplodus capensis 84 and D. Indeed, we often tend to oversimplify the complexity of sequential hermaphroditism: not all protogynous species are haremic, as not all protandrous species mate in pairs.
A recent study 28 , 85 has revealed a broader variation in effective population size in protogynous species that differs from the more limited expectations obtained when all protogynous species were considered haremic by default. Instead, some protogynous species were found to be group spawners, altering the expectations that all protogynous species should have low effective population size due to their supposed haremic system.
Similarly, not all protandrous species are monogamous or mate in pairs and small groups. Furthermore, many of the protandrous sparids that engage mostly or exclusively in pair mating, including Sparus aurata 86 and Rhabdosargus sarba 58 , exhibit a surprisingly high GSI of 4.
Therefore, while spawning in aggregations can explain the high GSI of many gonochoristic and protogynous species, mating system and spawning mode alone cannot explain the high GSI consistently found in protandrous sparids. This is not unique to the sparids. For example, the majority of damselfishes family Pomacentridae reproduce in pairs and, as predicted by SAM 14 , exhibit lower GSI values max.
This corroborates the suggestion that protandrous species can exhibit high GSI values, regardless of mating system. Here we propose that it is precisely the nature of protandry what explains the high GSI in protandrous sparids with pair mating.
Specifically, protandrous males first sex are always smaller than females. Thus, given that fecundity increases with size in females 90 , small protandrous males need to produce large amounts of sperm to effectively fertilize highly fecund females much larger than themselves, even when mating in pairs and under conditions of low levels of sperm competition. To do so successfully, they need to invest disproportionally in the gonads. For small-sized protandrous males, there might be a body size threshold below which the GSI has to increase to ensure enough sperm production to fertilize the eggs produced by the larger females.
Consistently, evidence of sperm production adjustment in relation to the amount of eggs to be fertilized has been reported in a coral reef fish, Thalassoma bifasciatum Taken together, this evidence supports our suggestion that protandrous males have higher GSI than expected because they need to ensure the fertilization of eggs produced by much larger females than themselves.
Briefly, Bateman principles state that, due to the smaller cost of producing sperm when compared to eggs: i male reproductive success RS increases with mate number whereas female RS does not; ii males have greater variance in RS than females, and iii the sex with the greater variance in RS undergoes stronger sexual selection 92 , In fact, there is evidence that sperm depletion is a problem for many males across several taxa reviewed in For example, in the simultaneous hermaphrodite polychaete Ophryotrocha diadema , small protandrous males can have difficulties fertilizing a full clutch of eggs Sperm depletion has been also documented in fish 98 , Moreover, in external spawners as most fishes are, mating rates how many females can be fertilized and ejaculation rates how many sperm should be released in the water should also be considered potential causes of higher GSI Thus, a higher than expected GSI may be related not only to sperm competition, which would relate to strong sexual selection in males in protandrous species spawning in groups, but also to a physiological compensatory mechanism that allows males to fertilize the many eggs that large females produce.
To test these ideas further, field studies are needed to corroborate whether the protandrous sparids with the highest GSI values are those that spawn in large groups or aggregations rather than in random matings or pairs. Furthermore, laboratory experiments aimed at determining the actual fertilization capacity of small-sized protandrous sparids, specifically fertilization rates with different amounts of sperm, will be key to determine whether there is indeed a size threshold below which the GSI needs to increase in order to ensure fertilization of the eggs released by the larger females.
This evidence would advance substantially our understanding of the relationship between sexual systems and mating systems in this diverse family in particular and in teleosts in general. To conclude, this study provides the most updated analysis of the distribution and incidence of different sexual systems in the family Sparidae, an ideal model taxon in which to study the evolution of sexual systems and SAM predictions.
We have found that both protandry and protogyny can evolve from gonochorism and back, but evolutionary transitions between the two types of sequential hermaphroditism are unlikely if ever to occur. We show that the predictions of the SAM incorporating mating system, spawning mode and sperm competition, hold well for protogynous and gonochoristic species.
In contrast, protandrous species do not conform to theoretical expectations. The high GSI values of some protandrous sparids suggest that males compete to fertilize the eggs of females while others mate in pairs in the absence of male-male competition but still invest greatly in gonad tissue.
We suggest that this is due to a compensatory mechanism that is intrinsic to protandry: boosting male investment in the gonads to ensure successful fertilization of the considerable number of eggs released by highly fecund females that are much larger than protandrous males. Pandian, T. Sexuality in fishes. CRC Press Smith, C. The remarkable reproductive diversity of teleost fishes. Article Google Scholar. Leonard, J. The Evolution of Sexual Systems in Animals.
In Transitions Between Sexual Systems. Springer, Cham Chapter Google Scholar. Mackiewicz, M. A mixed-mating strategy in a hermaphroditic vertebrate.
Costa, W. Androdioecy in Kryptolebias killifish and the evolution of self-fertilizing hermaphroditism. Charnov, E. The theory of sex allocation. Princeton University Press, New Yersey Maynard Smith, J. The evolution of sex. Cambridge University Press, Cambridge Neiman, M. Why sex? A pluralist approach revisited. Ghiselin, M. The evolution of hermaphroditism among animals. Policansky, D.
Sex change in plants and animals.
0コメント