The determinants of genetic diversity in butterflies – Lewontin’s paradox revisited

Abstract

Under the neutral theory genetic diversity is expected to be a simple function of population size. However, comparative studies have consistently failed to find any strong correlation between measures of census population size and genetic diversity. Instead, a recent comparative study across several animal phyla identified propagule size as the strongest predictor of genetic diversity, suggesting that r-strategists that produce many offspring but invest little in each, have greater long-term effective population sizes. We present a comparison of genome-wide levels of genetic diversity across 38 species of European butterflies (Papilionoidea). We show that across butterflies, genetic diversity varies over an order of magnitude and that this variation cannot be explained by differences in abundance, fecundity, host plant use or geographic range. Instead, we find that genetic diversity is negatively correlated with body size and positively with the length of the genetic map. This suggests that variation in genetic diversity is determined both by fluctuation in N e and the effect of selection on linked neutral sites.

Migratory behavior is positively associated with genetic diversity in butterflies

“…Our results suggest that migratory butterflies tend to have comparatively large N e . For example, in migratory monarch populations, N e is inferred to be in the order of 10 7 (Zhan et al, 2014; Talla et al, 2020), compared to N e of 10 5 – 10 6 in sedentary populations (Zhan et al, 2014), the latter falling within a general range estimated for butterflies (Mackintosh et al, 2019). Our heterozygosity estimates for the monarch butterfly also followed a pattern of nearly twice the difference between behavioral categories: ca.…”

“…A positive association between genetic diversity at neutral sites and N e is expected, since larger populations have a higher input of novel mutations and reduced genetic drift (Wright, 1931; Kimura, 1983; Nevo, 1978; Leffler et al, 2012; Mackintosh et al, 2019). Our results suggest that migratory butterflies tend to have comparatively large N e .…”

“…Indeed, there seem to be a few outliers that are not consistent with a correlation between genetic diversity and long-term population size. In Mackintosh et al, 2019, estimates of neutral genetic diversity for some migratory butterflies were among the bottom 15 % of European species, such as Vanessa atalanta, Pontia daplidice or Pieris brassicae . In our dataset, an exception is Danaus erippus (1.16 %), which appears to have a heterozygosity level below average.…”

“…A deviation from a strict linear relationship between N e and heterozygosity is not unexpected given a set of possible confounding factors. For example, recent geographic expansions and corresponding increases in population size, could lead to populations that have not yet reached mutation-drift equilibrium (Buffalo, 2021) – this has been suggested for example in P. brassicae and P. rapae (Mackintosh et al, 2019; Ryan et al, 2019). Another potential factor concerns the role that natural selection can play in decreasing neutral variation at physically linked sites (Ellegren & Galtier, 2016).…”

“…Other life history traits could shape species' heterozygosity levels and therefore act as confounding factors. As genetic diversity has been found to correlate with body size in butterflies (Mackintosh et al, 2019, Dapporto et al, 2019 and body size may correlate with population size (White et al, 2007), we included wingspan estimates obtained from the literature (Table S5) as a covariate in the pGLS test. In addition, because hybridization can elevate heterozygosity estimates, we excluded four potentially hybridizing taxa from our dataset (Table S1; Dupuis & Sperling, 2015, 2016Herrell & Hazel, 1995;Mavárez et al, 2006).…”