Despite the widespread recognition of adaptive radiation as a driver of speciation, the mechanisms by which natural selection generates new species are incompletely understood. The evolutionary radiation of endemic East Asian cyprinids has been proposed as evolving through a change in spawning habits, involving a transition from semibuoyant eggs to adhesive eggs in response to crosslinked river-lake system formation. Here, we investigated the molecular mechanisms that underpin this radiation, associated with egg hydration and adhesiveness. We demonstrated that semibuoyant eggs enhance hydration by increasing the degradation of yolk protein and accumulation of Ca2+ and Mg2+ ions, while adhesive eggs improve adhesiveness and hardness of the egg envelope by producing an adhesive layer and a unique 4th layer to the egg envelope. Based on multiomics analyses and verification tests, we showed that during the process of adaptive radiation, adhesive eggs downregulated the “vitellogenin degradation pathway,” “zinc metalloprotease pathway,” and “ubiquitin-proteasome pathway” and the pathways of Ca2+ and Mg2+ active transport to reduce their hydration. At the same time, adhesive eggs upregulated the crosslinks of microfilament-associated proteins and adhesive-related proteins, the hardening-related proteins of the egg envelope, and the biosynthesis of glycosaminoglycan in the ovary to generate adhesiveness. These findings illustrate the novel molecular mechanisms associated with hydration and adhesiveness of freshwater fish eggs and identify critical molecular mechanisms involved in the adaptive radiation of endemic East Asian cyprinids. We propose that these key egg attributes may function as “magic traits” in this adaptive radiation.

Among ray-finned fishes that provide parental care, many spawn in constructed nests, ranging from bowls, burrows and ridges to nests made of algae or bubbles. Because a nest by definition is a construction that enhances the nest-builder's fitness by helping it meet the needs of the developing offspring, nest-building behaviour is naturally selected, as is a preference for spawning with mates that provide well-built nests. However, nest-building behaviour can also be sexually selected, when nest traits increase mating success, protect against sperm competition or nest take-overs by conspecifics. Here, we offer a systematic review, with examples of how competition for sites and location of fish nests relates to sexual selection. We examine direct and indirect benefits of mate choice linked to nest traits, and different types of nests, from a sexual selection perspective. Nest-related behaviours are often under both natural and sexual selection, and we disentangle examples where that is the case, with special attention to females. We highlight some taxa in which nest building is likely to be sexually selected, but lack of research has left them uninvestigated. Some of them are established aquarium species, making them particularly amenable for future research. Finally, we compare with arthropods, amphibians and birds. This article is part of the theme issue ‘The evolutionary ecology of nests: a cross-taxon approach’. pmc Philos Trans R Soc Lond B Biol Sci Philos Trans R Soc Lond B Biol Sci 136 transb 7503623 RSTB Philosophical Transactions of the Royal Society B: Biological Sciences 0962-8436 1471-2970 The Royal Society PMC10331916 PMC10331916.1 10331916 10331916 37427477 10.1098/rstb.2022.0139 rstb20220139 1 1001 70 14 60 Articles Review Articles How sexual and natural selection interact and shape the In ray-finned fishes (Actinopterygii), it is mostly the male, alone or together with the female, that builds nests and care for the offspring [ 3 , 4 ]. Despite that pattern, males typically compete for mating opportunities, and females are choosy [ 5 ]. However, when preferred nests get filled with eggs, females may compete intra-sexually over those nests. This can generate both natural and sexual selection on female competitive ability and choosiness. We return to this topic in §7. Fish nests (and bowers) have been thoroughly reviewed in an evolutionary ecology context previously [ 3 , 6 ]. Here, we build on this work and focus on aspects of nests that may be sexually selected and examine to what extent nest-holders and mate-choosers are under natural versus sexual selection. We have searched the literature within ray-finned fishes and the presented examples are meant to be systematically broad, illustrate points we want to make, and indicate areas where more research is needed. Box 1 describes our use of important definitions related to nests and eggs. The methods and the list of the reviewed literature are found in the electronic supplementary material 1. Box 1. Definitions related to nests and eggs. Similarly, the line between different broods will be obscured when sequential and overlapping broods are cared for in the same nest. Here we will show that (1) it is important to tease apart how natural and sexual selection act on nests and their builders, (2) different types of nests differ in how natural and sexual selection are likely to act on them, and (3) sexual selection can act directly on nest-related traits through various mechanisms before, during and after spawning. 2 . Nest sites (a) Location and shape of nest sites The location of a nest site is important for offspring survival and development and hence choice of site is naturally selected in both sexes. However, at another site with only small shells, shell size was a better predictor for reproductive success than male size, and there was low variance in reproductive success, since most males got their shells filled with eggs [ 136 ]. Thus, size of the breeding sites can clearly affect the opportunity for sexual selection among males (cf. J. transcriptus , §2a). (f) Self-produced nesting material: glue, mucus and bubbles Some fishes produce nesting material themselves, and again, there is evidence of both natural and sexual selection behind these adaptations. But how should we label the selection on female choosiness or specific preference that arises from these benefits? Most nest traits that benefit the offspring directly (e.g. through improved survival or development) generate natural selection on female mate-choice (§3a). Direct benefits related to nest building that benefit the choosy female herself (e.g. a safe spawning site) also fall under natural selection. Because nests are constructions, they are extended phenotypes of the nest builder [ 63 , 64 ]. If the nest indicates parental abilities, then female choosiness and preference for nest traits will again be naturally selected, and females may compete for access to such males. (d) Sexual selection in nest-building frogs, arthropods and birds The overlap between natural and sexual selection in connection to nest-building behaviour reviewed in this paper is by no means unique to ray-finned fishes. Some frogs build foam nests. Their foam nests resemble the foam nests of gouramis and Callichthyinae catfishes (i.e. bubble nests, §4f) [ 111 , 142 , 143 ]. A visible frog foam nest increases a male's chances of mating [ 177 ] and nest building during mating arguably allows females to choose direct benefits through protection of the offspring from predation and desiccation [ 178 ]. As far as we know, this has never been done, but should be on everyone's to-do-list. To conclude, natural selection on offspring often (via mate choice) leads to sexual selection on parents, and there are plenty of behaviours related to nest building in fishes that are likely to be under sexual selection, in addition to the natural selection inherent in any construction called a nest.

Over the past 2 decades, biologists have come to appreciate that hybridization, or genetic exchange between distinct lineages, is remarkably common—not just in particular lineages but in taxonomic groups across the tree of life. As a result, the genomes of many modern species harbor regions inherited from related species. This observation has raised fundamental questions about the degree to which the genomic outcomes of hybridization are repeatable and the degree to which natural selection drives such repeatability. However, a lack of appropriate systems to answer these questions has limited empirical progress in this area. Here, we leverage independently formed hybrid populations between the swordtail fish Xiphophorus birchmanni and X. cortezi to address this fundamental question. We find that local ancestry in one hybrid population is remarkably predictive of local ancestry in another, demographically independent hybrid population. Applying newly developed methods, we can attribute much of this repeatability to strong selection in the earliest generations after initial hybridization. We complement these analyses with time-series data that demonstrates that ancestry at regions under selection has remained stable over the past approximately 40 generations of evolution. Finally, we compare our results to the well-studied X. birchmanni × X. malinche hybrid populations and conclude that deeper evolutionary divergence has resulted in stronger selection and higher repeatability in patterns of local ancestry in hybrids between X. birchmanni and X. cortezi. 22 8 26 8 2024 e3002742 e3002742 PLoS Biol 10.1371/journal.pbio.3002742 39186811 Over the past 2 decades, biologists have come to appreciate that hybridization, or genetic exchange between distinct lineages, is remarkably common—not just in particular lineages but in taxonomic groups across the tree of life. As a result, the genomes of many modern species harbor regions inherited from related species. This observation has raised fundamental questions about the degree to which the genomic outcomes of hybridization are repeatable and the degree to which natural selection drives such repeatability. We complement these analyses with time-series data that demonstrates that ancestry at regions under selection has remained stable over the past approximately 40 generations of evolution. Finally, we compare our results to the well-studied X . birchmanni × X . malinche hybrid populations and conclude that deeper evolutionary divergence has resulted in stronger selection and higher repeatability in patterns of local ancestry in hybrids between X . birchmanni and X . cortezi . How repeatable are the genomic outcomes of hybridization, and how much does natural selection drive such repeatability? In replicate natural populations of hybrid ants that have evolved independently for tens of generations, researchers found remarkably high repeatability in local ancestry patterns across 3 hybrid populations, driven in part by selection against deleterious load inherited from one of the parental species [ 29 ]. Past work from our group asked about repeatability in patterns of minor parent ancestry in naturally occurring Xiphophorus birchmanni × X . malinche populations that formed independently in different river systems [ 57 ]. We found moderate predictability in local ancestry patterns between replicate X . birchmanni × X . malinche populations [ 17 , 58 ]. Since we had access to time-series data for both the Santa Cruz and Chapulhuacanito populations, we were interested in evaluating how ancestry at minor parent deserts and islands has changed over the last 40 generations. Given that both hybrid populations are estimated to be over 100 generations old, we would expect that loci under strong or moderate selection would be fixed even at the earliest time points in our data set. Indeed, we find that regions that fall into shared ancestry deserts tend to have low minor parent ancestry in 2003 and maintain low ancestry through time ( Fig 4B and 4D ). The results of wavelet analyses indicate that broad-scale changes in ancestry along the genome in one hybrid population (at the scale of >8 Mb) predict a remarkable ~90% of the variance in the other hybrid population. We found that this cross-population repeatability was robust to iterations of the analysis controlling for potential technical confounders (see Methods ; S10 Table ). Since shared patterns of ancestry deviations are not predicted under neutrality, these results demonstrate that the correlations we observe are attributable to natural selection driving parallel changes in minor parent ancestry in the 2 hybrid populations, presumably due to selection on the same loci. Since these correlations are strongest at the broadest spatial scales in the genome, this indicates that natural selection acting shortly after hybridization was important in establishing them. The degree of cross-population repeatability we observe here exceeds that reported in other studies that have found evidence for such patterns [ 29 , 38 , 56 , 57 ]. What mechanisms could drive such high repeatability in minor parent ancestry across independently formed hybrid populations? Given the frequency of hybrid incompatibilities in Xiphophorus [ 17 , 35 , 58 ] and the fact that neither X . birchmanni or X . Moreover, in cases where deserts are not replicated across populations, minor parent ancestry still tends to be low in the second population (on average falling in the lowest quartile of minor parent ancestry; e.g., Fig 3A ). Consistent with our findings that selection acted early after hybridization, we find that minor parent deserts are typically large (on average 1.8 Mb). These regions are exciting candidates to pursue as we begin to map hybrid incompatibilities between X . birchmanni and X . cortezi in natural populations and in the laboratory [ 67 ]. Beyond these genome-wide patterns, we know the precise locations of 2 loci that

The study of natural selection and local adaptation is a thriving field of research. Local adaptation is driven by environment components and results in locally adapted phenotypes with higher fitness relative to other phenotypes from other locations in the species range. Tests of local adaptations have traditionally been done using transplant experiments, but the advent of next-generation sequencing methods have allowed the study of local adaptation to move from a phenotypic to a genomic approach. By using genome scans and state-of-the-art statistical tests, researchers can identify genes putatively under selection and study the genomic architecture of local adaptation, which often includes the observation of clustering of adaptive genes concentrated in fewer genomic regions known as “genomic islands of divergence”. The two species of North Atlantic eels, the European and the American eel, are excellent species for studying selection since they are panmictic and present large population sizes, show a wide distribution range across extremely heterogenous environments, and are subject to high mortalities. We reviewed studies of natural selection and local adaptation in American eel, European eel, between life cycle stages, between European and American eel. Finally, we discussed genome architecture in relation to local adaptation in eels and the role of both genetic (i.e., local adaptation) and non-genetic (i.e., phenotypic plasticity) in the survival of eels across their distribution range.

Gene action in fish of tetraploid origin. II. Cellular and biochemical parameters in clupeoid and salmonoid fish. By use of cell size, protein and hemoglobin content, and enzyme activities as markers, it becomes apparent that in the course of evolution the gene expression of anciently tetraploid fish of the order Ostariophysi was diploidized, but no such regulatory mechanism has evolved in the phylogenetically tetraploid species of the order Isospondyli. This finding is discussed in terms of possible selective neutrality of tetraploid expression and the phylogenetic age of Isospondyli. Published in Biochemical genetics (1975)

It seems likely that rod photoreceptors may have evolved in conjunction with the change from larval to juvenile stage through metamorphosis in indirect developing fishes. During evolution, the contraction and/or loss of the larval stage has resulted in earlier appearance of rod photoreceptors during development although they always arise later than cone photoreceptors. This ontogenetic developmental sequence supports Walls's (1942) proposal that cones are phylogenetically older than rods and suggests that rods may have evolved several times. Published in Journal of neurobiology (1990)