Rapid divergence and convergence of life‐history in experimentally evolved Drosophila melanogaster

Abstract

Laboratory selection experiments are alluring in their simplicity, power, and ability to inform us about how evolution works. A longstanding challenge facing evolution experiments with metazoans is that significant generational turnover takes a long time. In this work, we present data from a unique system of experimentally evolved laboratory populations of Drosophila melanogaster that have experienced three distinct life-history selection regimes. The goal of our study was to determine how quickly populations of a certain selection regime diverge phenotypically from their ancestors, and how quickly they converge with independently derived populations that share a selection regime. Our results indicate that phenotypic divergence from an ancestral population occurs rapidly, within dozens of generations, regardless of that population's evolutionary history. Similarly, populations sharing a selection treatment converge on common phenotypes in this same time frame, regardless of selection pressures those populations may have experienced in the past. These patterns of convergence and divergence emerged much faster than expected, suggesting that intermediate evolutionary history has transient effects in this system. The results we draw from this system are applicable to other experimental evolution projects, and suggest that many relevant questions can be sufficiently tested on shorter timescales than previously thought.

Genomics of Parallel Experimental Evolution in Drosophila

“…This finding supports the contention that convergent, or “parallel” depending on favored usage, evolution can result from polymorphisms that were shared in an ancestral population (cf. Stern 2013), as all selected lines in this study were derived from a common ancestral population (Rose et al 2004; Burke et al 2016). …”

“…Rose 1984) that were cultured from eggs laid by females 9–10 weeks old: 5 “AO” populations given 146 generations of A selection; 5 “BO” given 121 generations of B selection; and 5 “ n CO” given 37 generations of C selection. Their derivation is described in detail in Burke et al 2016. These generation numbers for the three selection regimes are all calibrated to the generation at which flies were sampled for pooled DNA sequencing.…”

“…In this usage, our experimental system is well qualified to examine such “parallelism,” as all 30 study populations share a moderately distant ancestor by the standards of experimental evolution: The Ives population (vid. Ives 1970; Rose et al 2004; Burke et al 2016). …”

“…Here we employ Drosophila stocks that feature 5-fold replicated groups of outbred populations that have been subjected to parallel sequences of selection regimes for decades (Rose et al 2004; Burke et al 2016). Three selection regimes were repeatedly imposed on these five-population groups, here called “A-type”, “B-type”, and “C-type” (fig.…”

How phenotypic convergence arises in experimental evolution

“…; Burke et al. ; Rebolleda‐Gómez and Travisano ). Nevertheless, divergent evolutionary responses have also been observed (e.g., Cohan ; Cohan and Hoffmann ; Melnyk and Kassen ).…”