The impact of long-term artificial removal experiment on the genetics of an alpine butterfly

Background/Question/Methods

In the current age of widespread anthropogenic effects across the globe, endemic organisms are vulnerable to habitat fragmentation and localized decline in population more than ever before. As such, insight into metapopulation-level dynamics is key to understanding the regional persistence of a species in the face of a localized extinction or population bottleneck. However, long-term field research addressing the effects of catastrophic declines in the context of a population network is scarce, due to difficulties in detecting such events with observational data, and the need for persistent sampling of populations over a prolonged period. In this study I investigated the genetic effects of an extended, localized population bottleneck using a long-term population experiment conducted within a well-studied metapopulation system of an alpine butterfly species (Parnassius smintheus), in which all adult butterflies observed in 2 local patches were removed for 8 consecutive years. I genotyped 473 historical samples collected throughout the removal experiment at 197 SNPs (single nucleotide polymorphisms), containing both putatively neutral and adaptive loci. I tracked the impact of long-term population bottleneck on the genetics of the local population and used population assignment tests to assess immigration from the surrounding population network.

Results/Conclusions

Despite the continuous removal, genetic diversity in the local populations was resilient, most likely due to immigration from surrounding populations. Genetic diversity measures (mean allelic richness and expected heterozygosity), while showing no significant difference between consecutive pairs of years in most cases, showed a significant, positive relationship between genetic diversity and time (LMM, p=0.02), indicating gradual recovery in genetic diversity throughout the experiment. Yearly proportion of loci detected to be out of Hardy-Weinberg equilibrium was positively correlated with time (LMM=0.0006), while no pair-wise comparisons of yearly measures in genetic differentiation was significant, suggesting a constant influx and intermixing of migrants within the focal patches. Additionally, there was an interesting observation in the population assignment analyses in which the most densely populated source population decreased in migrant contribution despite increasing in population size, indicating potential inverse density-dependent dispersal. Overall, my study confirmed that gene flow by between-patch immigration is the key mechanism conferring resilience against localized, long-term population disturbances, highlighting the importance of population-level dispersal dynamics in the persistence of a metapopulation.