An Experiment Testing Two Hypotheses of Speciation

Publisher Website
Google Scholar
Abstract
We initiated 17 laboratory populations of Drosophila simulans: a heterogeneous, chromosomally monomorphic base population (BASE) from which the other 16 were derived; four pairs of selection lines, each pair of which was selected for three different traits in three successive years; and eight genetic-drift lines, each of which experienced six flush-crash cycles after an initial founder event. Every 6 mo. for 3 yr we performed tests for reproductive isolation among the selection lines and BASE, and among the drift lines and BASE, with the following results. 1. Mean hybrid fitness was reduced in crosses between BASE and the drift lines. In the last test, fitness was 28% lower in hybrids than in parental lines. 2. This reduction in fitness became greater over the course of the experiment, principally in four of the eight lines. 3. The opposite, hybrid vigor, was observed consistently between drift lines after the second flush-crash cycle. 4. Mean hybrid fitness was reduced in crosses between BASE and the selection lines. 5. The reduction in fitness between BASE and the selection lines did not increase linearly over time, but varied erratically. 6. Weak sexual isolation was observed between BASE and the drift lines, for the experiment as a whole. The isolation was mainly caused by the behavior of three drift lines. 7. Overall, sexual isolation was not observed between BASE and the selection lines. The results of this experiment and previous, related ones support the view that either genetic drift or selection is by itself sufficient to produce reproductive isolation as a pleiotropic by-product. In the case of sexual isolation, not only can it be established as a by-product of microevolutionary forces, but there is a bias toward the development of positive assortative mating and against the development of negative assortative mating. The stronger and more rapid establishment of reproductive isolation by the drift lines than by the selection lines conforms to the predictions of Carson (1970) and Templeton (1980a,b).