Populations of fishes are increasingly threatened by over‐exploitation, pollution, habitat destruction, and climate change. In order to better understand the factors that can explain the amount of genetic diversity in wild populations of fishes, we collected estimates of genetic diversity (mean heterozygosity and mean rarefied number of alleles per locus) along with habitat associations, conservation status, and life‐history information for 463 fish species. We ran a series of phylogenetic generalized least squares models to determine which factors influence genetic diversity in fishes after accounting for shared evolutionary history among related taxa. We found that marine fishes had significantly higher genetic diversity than freshwater fishes with marine fishes averaging 11.3 more alleles per locus than their freshwater counterparts. However, contrary to our expectations, genetic diversity was not found to be lower in threatened versus not‐threatened fishes. Finally, we found that both age at maturity and fecundity were negatively related to genetic variation in both marine and freshwater fishes. Our results demonstrate that both life‐history characteristics and habitat play a role in shaping patterns of genetic diversity in fishes and should be considered when prioritizing species for conservation.

Captive breeding programs are often initiated to prevent species extinction until reintroduction into the wild can occur. However, the evolution of captive populations via inbreeding, drift, and selection can impair fitness, compromising reintroduction programs. To better understand the evolutionary response of species bred in captivity, we used nearly 5500 single nucleotide polymorphisms (SNPs) in populations of white-footed mice (Peromyscus leucopus) to measure the impact of breeding regimes on genomic diversity. We bred mice in captivity for 20 generations using two replicates of three protocols: random mating (RAN), selection for docile behaviors (DOC), and minimizing mean kinship (MK). The MK protocol most effectively retained genomic diversity and reduced the effects of selection. Additionally, genomic diversity was significantly related to fitness, as assessed with pedigrees and SNPs supported with genomic sequence data. Because captive-born individuals are often less fit in wild settings compared to wild-born individuals, captive-estimated fitness correlations likely underestimate the effects in wild populations. Therefore, minimizing inbreeding and selection in captive populations is critical to increasing the probability of releasing fit individuals into the wild.