Meeting Abstract

P1.169  Friday, Jan. 4  Evaluating methods of demographic inference and testing for balancing selection using genomic data from the checkerspot butterfly Euphydryas gillettii MCCOY, R.C.*; BOGGS, C.B.; PETROV, D.A.; Stanford University; Stanford University; Stanford University rmccoy@stanford.edu

Motivated by recent theoretical work, our study refocuses attention on the relative importance of balancing selection versus other evolutionary forces in maintaining consequential genetic variation in natural populations. As many tests for selection are sensitive to demography, our system is unique in allowing us to explicitly incorporate detailed knowledge of demographic history into our tests. We investigate the interaction of selection and demography in Euphydryas gillettii, a univoltine checkerspot butterfly that was intentionally introduced to Gothic, Colorado in 1977. The population established and subsequently experienced severe fluctuations, including extreme bottlenecks of fewer than 25 adult individuals as estimated by annual mark-release-recapture experiments. We prepared and sequenced barcoded cDNA libraries from 8 whole larvae from the introduced Colorado population and 8 whole larvae from its ancestral Wyoming population on the Illumina HiSeq2000 platform. These data were used to assemble the E. gillettii transcriptome de novo using Trinity and discover expressed SNPs using GATK. After filtering, we obtained a SNP set to which we applied several common methods of demographic inference and allele frequency spectrum-based tests for selection. This isolated population with well known demographic history allows us to compare the ability of methods of demographic inference to estimate the timing and strength of the bottleneck based on genomic data. Deviations from expected neutral patterns of genomic variation given the known demography suggest natural selection. In particular, our data reveal the action of purifying selection in spite of strong drift during recurrent bottlenecks.