Ancestral population reconstitution from isofemale lines as a tool for experimental evolution

Experimental evolution is a powerful tool to study adaptation under controlled conditions. Laboratory natural selection experiments mimic adaptation in the wild with better-adapted genotypes having more offspring. Since the selected traits are frequently not known, adaptation is typically measured as fitness increase by comparing evolved populations against an unselected reference population maintained in a laboratory environment. With adaptation to the laboratory conditions and genetic drift, however, it is not clear to what extent such comparisons provide unbiased estimates of adaptation. Alternatively, ancestral variation could be preserved in isofemale lines that can be combined to reconstitute the ancestral population. Here we assess the impact of selection on alleles segregating in newly established Drosophila isofemale lines. We reconstituted two populations from isofemale lines and compared them to two original ancestral populations founded from the same lines shortly after collection. No significant allele frequency changes could be detected between both ancestral populations and simulations showed that drift had a low impact compared to Pool-Seq-associated sampling effects. We conclude that laboratory selection on segregating variation in isofemale lines is too weak to have detectable effects, which validates ancestral population reconstitution from isofemale lines as an unbiased approach for measuring adaptation in evolved populations.