Drivers of genetic diversity across the marine tree of life

Abstract

Why do some species have more genetic diversity than others? This question is one of the greatest remaining mysteries in evolutionary biology, and is particularly urgent in marine species, which are experiencing catastrophic anthropogenic impacts. We address this critical gap by estimating genetic diversity for 93 marine species sampled over 9,000 localities. For each species, we aggregate biotic traits and abiotic geographic data for their ranges. We show that diversity increases with range extent and planktonic dispersal. We hypothesize that these traits increase a species' ability to avoid or recover from bottlenecks, thereby maintaining diversity. Our findings provide insights into the factors interacting to shape genomic variation in the ocean, and offer a predictive framework for understanding marine biodiversity in the face of global change.

Figure 1:. Species-level genetic diversity varies by orders of magnitude in the world’s oceans.

(a) Time-calibrated phylogeny of all species (N = 93) and map of all samples (N = 9,430) included in this study (b) Set of discrete (gray-scale) and continuous (blue-scale) biotic and abiotic predictors collated for each species. Colors represent the trait state or value for each species. (c) Species-level genetic diversity estimated using a spatially explicit model and genomic sequence data.

Figure 2:. The estimated effects of biotic and abiotic predictors on species’ genetic diversity vary in their magnitude and direction.

Estimated effect sizes (presented via their marginal posterior distributions) are standardized by the standard deviation of their predictor and are ordered from top to bottom by mean effect on species-level genetic diversity. Significant predictors – those for which the 95% equal-tailed credible interval of the marginal distribution of effect sizes did not overlap zero (gray, vertical dashed line) - are denoted with an asterisk.

Figure 3:. Species-level genetic diversity is most significantly predicted by species range extent and the presence of a planktonic life stage.

(a) Species with larger range extents have higher genetic diversity. Shaded polygons represent 50%, 80%, and 95% model prediction credible intervals. (b) Species that have at least one planktonic life stage have higher genetic diversity than species that are never planktonic. Each point represents a unique species (N = 93) and is colored by taxonomic clade (as in Figure 1).