Cryptic species, biogeography, and patterns of introgression in the fish genus Mogurnda (Eleotridae) from the Australian wet tropics: A purple patch for purple-spots

Accurately delimiting species is an essential first step towards understanding the true biodiversity of an ecosystem and any subsequent efforts to identify and protect taxa at risk of extinction. Current molecular evidence suggests that purple-spotted gudgeons (genus Mogurnda) harbour high levels of cryptic biodiversity across their broad distributional range. The present study uses a large single nucleotide polymorphism (SNP) dataset plus a companion allozyme dataset to clarify taxonomic uncertainty, patterns of introgression, and biogeographic relationships among Mogurnda populations within the Queensland Wet Tropics (QWT), a known biodiversity hotspot. Both datasets were strongly concordant in identifying a total of seven taxa split among distinct southern, northern, and lowlands groups. No two taxa were found in strict sympatry, but many appear to be parapatric and occur within the same drainage basin. Although clear evidence of introgression was only evident at six sites (∼4%), the genomic signature of modest historic admixture between proximally-distributed taxa was detected at multiple other sites. Nevertheless, all primary genetic and phylogenetic analyses strongly supported the integrity and diagnosability of these seven taxa. We therefore nominate these as novel candidate species for what appears to be yet another hyper-cryptic complex within the Australian freshwater ichthyological fauna. These results offer up intriguing ecological scenarios and conservation implications for multiple candidate species with narrow ranges in specialised habitat. We conclude by exploring the major biogeographic patterns displayed by QWT Mogurnda.

Dispersal inference from population genetic variation using a convolutional neural network

The geographic nature of biological dispersal shapes patterns of genetic variation over landscapes, making it possible to infer properties of dispersal from genetic variation data. Here, we present an inference tool that uses geographically distributed genotype data in combination with a convolutional neural network to estimate a critical population parameter: the mean per-generation dispersal distance. Using extensive simulation, we show that our deep learning approach is competitive with or outperforms state-of-the-art methods, particularly at small sample sizes. In addition, we evaluate varying nuisance parameters during training-including population density, demographic history, habitat size, and sampling area-and show that this strategy is effective for estimating dispersal distance when other model parameters are unknown. Whereas competing methods depend on information about local population density or accurate inference of identity-by-descent tracts, our method uses only single-nucleotide-polymorphism data and the spatial scale of sampling as input. Strikingly, and unlike other methods, our method does not use the geographic coordinates of the genotyped individuals. These features make our method, which we call "disperseNN," a potentially valuable new tool for estimating dispersal distance in nonmodel systems with whole genome data or reduced representation data. We apply disperseNN to 12 different species with publicly available data, yielding reasonable estimates for most species. Importantly, our method estimated consistently larger dispersal distances than mark-recapture calculations in the same species, which may be due to the limited geographic sampling area covered by some mark-recapture studies. Thus genetic tools like ours complement direct methods for improving our understanding of dispersal.