UC Berkeley

A major goal in conservation is to understand and halt the increasing rate of extinction that many natural populations face due to threats such as habitat loss and fragmentation, climate change, and disease. Extinction rates are particularly high among freshwater fishes, due to high levels of species endemism and frequent habitat loss. Declining populations are expected to face reduced genetic variation, increased inbreeding, and increased genetic load, which can reduce adaptive capacity and fitness. Together, these factors can eventually lead to an “extinction vortex”, a self-reinforcing process in which small populations are increasingly likely to decline towards extinction. However, it is unclear why some populations go extinct in response to small population size and isolation while others do not. For my dissertation, I have developed genomic resources for the critically endangered Devils Hole pupfish (Cyprinodon diabolis) as a model system for understanding the genomic consequences of small population size and isolation and the resulting impacts on fitness and future viability. This species lives in the smallest known range of any vertebrate and has experienced severe recent population decline since the mid 1990s, down to as low as 35 individuals in 2013. Through my dissertation, I aim to not only inform management of Devils Hole pupfish, but also more broadly explore how population genomics across space and time can be leveraged to inform conservation of endangered species. In Chapter 1, I leveraged genomics to investigate how varying levels of isolation and population size influence population structure, inbreeding, and mutation load in Death Valley pupfishes (Tian et al. 2022 PRSB). In Death Valley, there are two other species of pupfish, Cyprinodon nevadensis and Cyprinodon salinus. Towards the end of the Pleistocene (10-15 kya), Death Valley was filled with pluvial lakes but as the climate became more arid, pupfishes became isolated. Through PCA and admixture analyses, I found strong population structure between the three species within Death Valley. In addition, I demonstrated that Devils Hole pupfish are extremely inbred, with an average inbreeding coefficient of 0.54, and that there is elevated mutation load in this species relative to other desert pupfishes. Finally, I identified unique, putative, loss-of-function alleles and deletions in genes associated with sperm motility and hypoxia to inform management of this conservation icon.In Chapter 2, I generated chromosome-scale de novo reference genomes of C. diabolis and C. nevadensis mionectes, and 180 resequenced genomes of Death Valley and Ash Meadows desert pupfishes, including genomes from 1940s formalin-fixed historical specimens and embryonic lethal C. diabolis individuals that have a heart defect observed at 5 dpf that is occurring at 25% frequency in the population to assess how demographic history and genetic variation interact to shape the genomic content and future viability of Devils Hole pupfish. I found evidence of strong population structure and that effective population size has declined across Death Valley pupfishes over the last ten thousand years, coinciding with the drying of pluvial lakes during the late Pleistocene. Genetic diversity is low across desert pupfishes, but up to an order of magnitude lower in wild C. diabolis (π = 0.000021), with the captive population having 43% less genetic diversity than in the wild, despite consistent introductions of wild genetic diversity. Despite population decline, C. diabolis genetic diversity has not significantly decreased over time, likely due to historical levels already being very low. In addition, I found strong evidence of inbreeding depression in C. diabolis; a 10% increase in inbreeding is associated with a 75% decrease in the odds of survival. Furthermore, I found evidence of purging of strongly deleterious alleles in Devils Hole pupfish relative to C. nev. amargosae, and an overall increase in moderately deleterious alleles. Finally, I used genome scans and association mapping to investigate the genetic basis of the embryonic lethal heart defect. The genomic resources that I generated across space and time lay the groundwork for continued conservation and management of this critically endangered species.In Chapter 3, I estimated a mutation rate for Devils Hole pupfish. Mutation is the ultimate source of genetic variation and across taxa, mutation rates vary considerably. One theory for explaining this variation is the drift-barrier hypothesis, which assumes that since most mutations are either deleterious or neutral, it should be selectively beneficial for organisms to reduce the mutation rate and maintain genome integrity. However, in populations with small effective population size and thus reduced efficacy of selection, the mutation rate may drift upwards due to mutator alleles persisting. Despite the lack of a pedigree, I estimated a mutation rate by identifying autozygous regions, counting heterozygous SNPs within them, and analyzing the distribution of autozygous fragment sizes to infer a per-generation mutation rate. My estimate of 1.16 x 10-8 bp-1 generation indicates that the Devils Hole pupfish mutation rate is 2X higher than that of the average for fishes and consistent with the drift barrier hypothesis. In addition, I found unique mutational spectra difference between embryonic lethals and adults. Furthermore, I was able to infer updated historical effective population sizes and divergence times. This chapter established a key parameter for evolutionary genomics research and conservation of the Devils Hole pupfish.