Introduction
Studies of sex-specific differences in ecology, physiology and behaviour enhance our understanding of processes driving evolution and maintenance of biodiversity, and assist in developing conservation management (Amos et al., 2014; Pavlova et al., 2013). However, lack of sex-specific genetic markers hampers such studies in many monomorphic fish species. Sex-determining systems in teleost fish are highly variable, including sequential hermaphroditism, genotypic sex determination (with X0, XY or complex XY being more prevalent than Z0, ZW or complex ZW), environmental dependency and environmental sex determination (Bachtrog et al., 2014; Mank, Promislow, & Avise, 2006). In fishes with genotypic sex determination, sex chromosomes are often homomorphic, where X/Y or Z/W pairs are similar or nearly identical in gene content and size (Bachtrog, 2013; Bachtrog et al., 2014). Various genes (includingamhy , amhr2 , bcar1 , dmrt1/dmY/dm-W ,gsdf, sdY and members of the SOX: SRY -likeHMG -box-containing gene family), single nucleotide polymorphisms (SNPs), inversions, or multiple loci, can all contribute to sex determination in fish (Bao et al., 2019; Graves & Peichel, 2010; Martínez et al., 2014; Natri, Merilä, & Shikano, 2019). Environmental influence on genotypic sex determination in fish (Devlin & Nagahama, 2002; Penman & Piferrer, 2008) results in occurrence of environmental sex reversal, which can trigger transitions between genotypic and environmental sex and facilitate evolution of new genotypic sex-determination systems (Holleley et al., 2015; Schwanz, Georges, Holleley, & Sarre, 2020). Due to rapid and frequent sex chromosome turnover in fish, high variation in sex-determination genes and systems is observed among closely-related species and across populations of the same species (Darolti et al., 2019; Kottler et al., 2020; Natri et al., 2019; Takehana et al., 2014). Thus, identification of reliable, generally applicable sex-linked markers in fish can be challenging.
When present, homomorphic sex chromosomes occur where gametologs have recombined relatively recently, and may result from recent switches in the chromosome pair used for sex determination or transition from temperature-dependent to genotypic sex determination (Bachtrog et al., 2014). In contrast, heteromorphic sex chromosomes, where X and Y—or Z and W—are diverged and highly distinct, evolve from homomorphic sex chromosomes through suppression of recombination between sex chromosomes and subsequent degradation of Y or W, which accumulate mutations and repeats and lose genes that similarly affect the sexes (Charlesworth, Charlesworth, & Marais, 2005). This inhibition of recombination starts from a sex-determining locus and occurs progressively and stepwise along chromosomes, resulting in genomic regions with different levels of sequence divergence between homologs (i.e. evolutionary strata), corresponding to different times since cessation of recombination (Charlesworth et al., 2005).
Several genomic approaches can be used to detect sex chromosomes (reviewed in Palmer, Rogers, Dean, & Wright, 2019). Three are particularly applicable to wildlife, because they use DNA sequence obtainable from non-lethally-collected samples and do not require captive breeding to build linkage maps. The first approach uses different ploidy of sex chromosomes diverged through Y or W degeneration: the homogametic sex (e.g. XX females or ZZ males) will have two copies of the same sex chromosome, whereas heterogametic sex (e.g. XY males or ZW females) will have one copy of each sex chromosome. This pattern will be reflected in read-depth coverage: for example, old strata in which recombination between homologs has long ceased will show half as many reads on X or Z in the heterogametic sex as the homogametic one, and Y- or W- loci will be absent in the homogametic sex (Darolti et al., 2019; Gan et al., 2019; Vicoso, Emerson, Zektser, Mahajan, & Bachtrog, 2013). But in young strata or homomorphic sex chromosomes, having high similarity between X and Y or Z and W, the read depth will be similar between sexes and to that of autosomal regions. For these younger regions of sex chromosomes, an approach based on differences insex-specific SNP density across genomic regions is more appropriate (Darolti et al., 2019; Wright et al., 2017). With Y or W accumulating mutations faster than X or Z due to reduced recombination and weaker purifying selection, higher SNP density in young strata is expected in the heterogametic than homogametic sex. In contrast, in older strata with substantial Y or W degeneration, X- and Z-linked loci will be effectively hemizygous in the heterogametic sex (halved read depth), and higher SNP density is expected in the homogametic sex. Finally, older strata of Y or W chromosome could be detected using a k-mer approach, by matching short reads of the homogametic sex to the genome of the heterogametic sex (Carvalho & Clark, 2013), or usingin-silico whole genome subtraction (Dissanayake et al., 2020). A combination of different approaches might be needed to detect sex-linked loci when the age of sex chromosomes is unknown.
Sex determination is not well understood in fish of the family Percichthyidae. Species from this family dominate the Australian freshwater fish fauna, and three genera occur in eastern Asia (Coreoperca and Siniperca ), and South America (Percichthys ). Many Australian species are threatened, including the endangered Macquarie perch Macquaria australasica and trout cod Maccullochella macquariensis . Although sex of an adult can be ascribed when it produces gametes, lack of sexual dimorphism and ways to determine the sex of individuals non-invasively outside breeding seasons hinders better understanding of species biology and more efficient conservation. Shams et al. (2019) examined karyotypes for two percichthyids, Murray cod Maccullochella peelii and golden perchMacquaria ambigua , and for both reported male heterogametic sex-chromosome systems (XX females/XY males) with diploid chromosome number 2n = 48. Heteromorphism was detected in sex chromosomes of Murray cod, but homomorphism in golden perch (Shams et al., 2019). In Macquarie perch, the sister-species of golden perch (Lavoué et al., 2014), a set of >1200 genome-wide SNPs did not reveal markers consistent with Y-linkage (i.e. present in males only) or strict X/Y homology, i.e. always heterozygous in males and always homozygous in females (Lutz et al., 2021), suggesting that it may also have young sex chromosomes. Infrequent (0.17-3.7%) synchronous hermaphroditism was reported in captive and wild Murray cod (Gooley, Anderson, & Appleford, 1995; Ingram, Ho, Turchini, & Holland, 2012) but not in Macquarie perch (Appleford, Anderson, & Gooley, 1998). There are no reports of sex change in Percichthyidae. Hatchery work suggests that environmental conditions may influence sex determination at an early stage of development in some percichthyids (Ingram et al., 2012). Lyon et al. (2012) reported 2.5 times as many females as males in a population of trout cod stocked from hatchery-bred fish. Because sex-ratio biases in stocked fish have implications for recovery of populations through stocking, being able to determine the genetic sex of fish, and the extent to which it can be overridden by environmental variation, would benefit conservation management of threatened percichthyid species.
With the overarching goal of uncovering the genetic basis of sex determination in two percichthyid perches, we aimed to (i) sequence and assemble the genomes of Macquarie perch and golden perch, (ii) annotate both genomes using respective transcriptome sequences, (iii) identify sex-linked loci using reduced-representation sequencing (DArTseq) data for both species and whole-genome resequencing (WGS) data for 100 known-sex Macquarie perch, (iv) identify candidate sex-determining genes using annotations, (v) develop a non-lethal, affordable and rapid molecular assay to determine the genetic sex of Macquarie perch and test its validity across four percichthyids (Macquarie and golden perches, Murray and trout cods), and (vi) test whether the candidate sex-determining region of Macquarie perch contains sex-specific variation across the four percichthyids using amplicon sequencing. Whereas the reduced-representation approach yielded few sex-linked loci, WGS data for Macquarie perch data revealed a small genomic region inherited in a predominantly XY fashion on a SOX1b -containing scaffold. This suggests a XY sex-determining system in Macquarie perch, with SOX1b as a novel candidate sex-determining gene, and the sex-linked region as its candidate regulatory locus. A test of a molecular sexing assay targeting a SNP with a male-specific allele in the sex-linked region, and amplicon sequencing data for four percichthyid species indicated that the Macquarie perch sexing region is species-specific and either specific to populations related to those in which it was detected or can be influenced by environment. The resources developed here will facilitate evolutionary research on sex chromosome turnover in fish, as well as conservation and ecological research in Percichthyidae, a dominant component of the freshwater Australian fish fauna. The workflow described here could be used for developing molecular tests to determine genetic sex of other fish species with monomorphic sex chromosomes.