Connectivity rescues genetic diversity after a demographic bottleneck in a butterfly population network

Abstract

Demographic bottlenecks that occur when populations fluctuate in size erode genetic diversity, but that diversity can be recovered through immigration. Connectivity among populations and habitat patches in the landscape enhances immigration and should in turn facilitate recovery of genetic diversity after a sudden reduction in population size. For the conservation of genetic diversity, it may therefore be particularly important to maintain connectivity in the face of factors that increase demographic instability, such as climate change. However, a direct link between connectivity and recovery of genetic diversity after a demographic bottleneck has not been clearly demonstrated in an empirical system. Here, we show that connectivity of habitat patches in the landscape contributes to the maintenance of genetic diversity after a demographic bottleneck. We were able to monitor genetic diversity in a network of populations of the alpine butterfly, Parnassius smintheus, before, during, and after a severe reduction in population size that lasted two generations. We found that allelic diversity in the network declined after the demographic bottleneck but that less allelic diversity was lost from populations occupying habitat patches with higher connectivity. Furthermore, the effect of connectivity on allelic diversity was important during the demographic recovery phase. Our results demonstrate directly the ability of connectivity to mediate the rescue of genetic diversity in a natural system.

Keywords: alpine butterfly; connectivity; demographic bottleneck; genetic diversity; recovery.

Fig. 1.

Map of the study patches occupied by a network of P. smintheus populations in Kananaskis, Alberta, Canada. Ridge-top meadow habitat is in white. Locations of 11 patches included in this study are shown as red circles with black letter labels. Map data: Google, DigitalGlobe 2015.

Fig. 2.

Changes in P. smintheus population size and allelic richness (A_R) in the network over different years. (A) Mean A_R over seven microsatellite loci in 2008, 2010, 2011, and 2013 (with rarefaction to four genes). Significant predictors of A_R change were 2010 population size during phase I, none during phase II, and 2012 connectivity during phase III. (B) Boxplots of yearly P. smintheus abundance estimates for all populations, showing interquartile range (IQR; boxes), maximum and minimum estimates up to 1.5 × IQR (whiskers), and outliers beyond 1.5 × IQR (open circles).

Fig. 3.

Relationship between the proportional loss of allelic richness (A_R) across a demographic bottleneck and habitat patch connectivity in P. smintheus. Proportional loss of allelic richness was measured from 2008 and 2013 and was best explained by connectivity in 2012. Solid dots indicate individual patches. Least-square line of best fit is shown.