Population Responses to Habitat Fragmentation: Statistical Power and the Random Sample Hypothesis

摘要: Oikos 76: 235-242. The random sample hypothesis states that small habitat fragments are samples from large ones. Therefore, the effect of fragmentation on population size is only related to pure loss, area and isolation do not influence in landscape. An alternative probability occupation a fragment fragment. Simulation models included effects were used investigate survival landscapes with different proportions suitable habitat. Four combinations sensitivity dispersal ability used. These estimate statistical power reject hypothesis. detect differences, between model based which survival, was low high proportion habitat, but increased as decreased. results tests support or might be due differences In where there no functionally still continuous two hypotheses will give very similar predictions about fragmentation. threshold value landscape true (threshold for fragmentation) vary across species landscapes. H.