4.1 Comparison between diploids and triploids in the native range
Our analyses of genetic diversity and structure between diploid and triploid C. auratus showed that overall nucleotide diversity of diploids was comparable to that of triploids based on both mitogenome and nuclear genome (Table 3). This is inconsistent with results of previous studies showing a slightly higher genetic diversity in triploids than diploids (Liu et al., 2017b; Luo et al., 2014). This may reflect the fact that our sampling work mainly focused on the native range of source populations for introduction into Tibet, with a higher percentage of diploids than triploids, while triploids actually have wider geographic distribution than diploids (Liu et al., 2017b; Liu et al., 2017c). By using the same procedures of variant calling, total number of detected SNPs in triploids was slightly higher than that in diploids, which is consistent with a previous study based on SNPs from transcriptome data (Ren et al., 2018). Furthermore, at the individual level, although the mean observed heterozygosity of diploids was significantly higher than that of triploids when considering ploidy level (Figure 5), the mean frequency of heterozygous loci across genomes in diploids was significantly lower than that in triploids. Similarly in other unisexual organisms that are derived from auto-polyploidization (uneven ploidy in general), they show levels of genetic diversity close to that of their sexual diploid counterparts (Joly & Bruneau, 2004; Lee et al., 2016; Liu et al., 2015b). However, unisexual organisms with interspecific hybrid origins, such as vertebrate P. formosa andP. eos-neogaeus and some plants, usually have a much higher genetic diversity than their sexual parental species (Angers & Schlosser, 2007; Robertson et al., 2010; Vallejo-Marin & Lye, 2013; Warren et al., 2018). Hence, genetic diversity of unisexual organisms depends on their specific origin relative to related sexual species.
Previous studies revealed that triploid C. auratus have undergone multiple independent polyploidy origins from sympatric diploids (Liu et al., 2017b; Luo et al., 2014). Our results also supported this point by showing that native diploids were genetically clustered with triploids within some lineages. Based on mitogenomes, diploids and triploids coexist in almost each lineages, indicating the same origin within the species. However, by using genomic SNPs, diploids were clustered with triploids only within two clusters (Figure 3), and triploids shaped three more clusters. Given that the genetic variation of triploids remains almost frozen across generations because of unisexual gynogenesis (Wang et al., 2022), we assumed that the two shared clusters reflected recent triploidization events and the three triploid-specific clusters originated from more ancient triploidization events. Genetic structure of contemporary diploids linked to a geographical separation pattern should be shaped by local adaptation or other factors. However, the genetic structure of triploids is accumulated by previous triploidization from diploids. Based on results of this study, the triploid C. auratus has at least five polyploidy origins and might possess more origins when investigating more samples from a broader range of locations.