Revisiting an old riddle: what determines genetic diversity levels within species?

Abstract

Understanding why some species have more genetic diversity than others is central to the study of ecology and evolution, and carries potentially important implications for conservation biology. Yet not only does this question remain unresolved, it has largely fallen into disregard. With the rapid decrease in sequencing costs, we argue that it is time to revive it.

Figure 1. Autosomal nucleotide diversity levels across species.

Autosomal genetic diversity is given as the average number of pairwise differences per base pair, in percent, and is shown on a log10 scale. Each estimate represents the mean of at least three loci and in most cases is based on only non-coding or synonymous sites. The estimates are ordered by diversity level, labeled by species name, and colored by the phylum to which each species belongs. The number of species in each phylum is given in parentheses in the legend.

Figure 2. Autosomal nucleotide diversity levels across species, grouped by phylum.

Diversity estimates for each species are the same as in Figure 1; here they are ordered within phylum, and phyla are presented in order of their median diversity levels. Within Chordata, open circles indicate mammals, and within Arthropoda, they denote Drosophila species. We note that the three most diverse chordates are all invertebrate sea squirts. In panel (A), estimates are colored by the phylum to which each species belongs and horizontal bars mark the median estimate for each phylum; for Magnoliophyta, a dashed line marks the median for selfing species (open circles) and a solid line marks the median for outcrossing species. (We do not provide p values for comparisons because of the lack of phylogenetic independence.) Crosses denote estimates for individual populations and are shown when population structure was reported in the original study. In panel (B), estimates are colored according to whether the species lives in a terrestrial, freshwater, or marine environment (not all species are categorized). Horizontal bars indicate the median for each category within each phylum (only shown when more than two species fall in the category). The number of species in each habitat is given in parentheses in the legend.

Figure 3. Drosophila species phylogeny (top), autosomal nucleotide diversity estimates (bottom), and geographic ranges.

Diversity levels are significantly correlated with the range category a priori ordered as island, narrow endemic, broad endemic, cosmopolitan using a generalized least squares method, and controlling for the phylogeny (F = 21.49, df = 20, p = 0.0002). Names along the branches of the phylogeny identify the Drosophila subgroup to which the species below belong(s); branch lengths displayed are arbitrary. For a definition of the four range categories, see Text S1. Horizontal lines mark the median diversity of species within each range category. We note that the estimates for Drosophila buzzatii and Drosophila subobscura include loci within polymorphic inversions and represent the average diversity within a chromosomal arrangement.

Figure 4. Comparison of autosome and sex chromosome nucleotide diversity.

The ratio of sex chromosome to autosome diversity is plotted for the 29 species in which both estimates were available from the same population(s). Colors indicate the phylum to which the species belong. Within Chordata, open circles denote mammals and open triangles birds; within Arthropoda, open circles denote Drosophila species. The number of species in each group is given in parentheses in the legend. Within species, crosses represent the ratio estimated from different populations, with the median of the estimates shown as a triangle (birds) or circle (all other species). Solid horizontal lines indicate the median sex chromosome to autosome ratio for arthropods and chordates, colored as in the key. The black dashed line indicates where sex chromosome diversity equals three-fourths of autosomal diversity.