Data from: Early warning indicators of population collapse in a seasonal environment

Bi-seasonal (breeding, <b>b</b>, and <b>non-breeding</b>, <b>nb</b>) abundances (<b>count</b>) from experimental population of <i>Drosophila melanogaster</i> exposed to chronic season-specific habitat loss. Populations were exposed to habitat loss in either the breeding or non-breeding period (<b>seasonT</b>), and were subject to one of two rates of loss: 10% or 20% per generation (<b>lossT</b>). In addition, a subset of individuals were measured from half of the replicate populations (<b>rep</b>; <b>treat_rep</b>). From each individual, measures of locomotion (<b>activity</b>; <b>crosses</b>) were collected using a Drosophila activity monitor, in which the movements of individual flies are autonomously recorded using infrared light beams. These individuals were then frozen, dried, and weighed (<b>weight</b>). <br><b>See readme.txt for definitions of variables in both datasets.<br></b><br>The abundance data has been used to determine whether populations losing breeding and non-breeding habitat decline in different ways and whether the season of decline is detectable in simple vital rates. By combining the abundance and phenotype data, we explored whether the production of trait-based early warning indicators of population collapse differs between populations losing breeding and non-breeding habitat.