Evolutionary shifts to self‐fertilisation restricted to geographic range margins in North American Arabidopsis lyrata

Abstract

Cross-fertilisation predominates in eukaryotes, but shifts to self-fertilisation are common and ecologically and evolutionarily important. Reproductive assurance under outcross gamete limitation is one eco-evolutionary process held responsible for the shift to selfing. Although small effective population size is a situation where selfing plants could theoretically benefit from reproductive assurance, empirical tests of the role of population size are rare. Here, we show that selfing evolved repeatedly at range margins, where historical demographic processes produced low effective population sizes. Outcrossing populations of North American Arabidopsis lyrata have low genetic diversity at geographic margins, with a signature of post-glacial range expansion in the north and rear-edge isolation in the south. Selfing populations occur at the margins of two genetic groups and never in their interior. These results corroborate small effective population size as the promoter of self-fertilisation and have important implications for our understanding of species turnover, range limits and range dynamics.

Quantitative Genetic Architecture at Latitudinal Range Boundaries: Reduced Variation but Higher Trait Independence

“…In many taxa, range-edge populations are relatively small, isolated from gene flow, and exposed to elevated rates of genetic drift (Eckert et al 2008). This seems to be true in A. lyrata as well, because variation at microsatellite markers is eroded near the distribution boundaries (Griffin and Willi 2014). Specifically, for the nine populations studied here, marker data show reduced expected heterozygosity in the south and north (see "Methods").…”

“…First, expected heterozygosity (H E ), a proxy for long-term population size in the absence of significant gene flow, is significantly lower at the latitudinal extremes for the nine populations studied here (mean: 0.53; quadratic term 5 standard error: 20:0045 AE 0:0018; P p :043; data from Griffin and Willi 2014). This implies decreased effective population size at the trailing southern and leading northern edge.…”

“…This implies decreased effective population size at the trailing southern and leading northern edge. Second, the nine populations have a low fixation index (F IS 0:07) with no significant latitudinal pattern, indicating that they are all predominantly outcrossing (Willi and Määttänen 2011;Griffin and Willi 2014). Third, a large-scale progeny-array analysis on 13 other outcrossing populations of A. lyrata (multilocus outcrossing rate t m 1 0:8) revealed that neither the rate of outcrossing nor the fraction of full-sibs (rather than halfsibs) among the outcrossed offspring were linearly or quadratically related to latitude (all P 1 :09; data from Willi and Määttänen 2010).…”

“…The eastern cline included populations from the Appalachians and the East Coast: North Carolina, Virginia, western Maryland, New Jersey, and upstate New York. The two clines fall within separate genetic clusters, based on analysis of microsatellite markers (Willi and Määttänen 2010;Griffin and Willi 2014). From each population, we collected mature fruits from 30 plants The geographic distribution of A. lyrata ( fig.…”

“…There is much evidence that populations at range margins have reduced neutral genetic marker diversity (Eckert et al 2008;Griffin and Willi 2014). However, few studies have mapped the distribution of quantitative genetic variation and genetic integration for ecologically relevant traits to address the causes of distribution limits.…”