| General opinion in the contemporary literature is that disturbance strongly influences patterns of species diversity, and that diversity is greatest at intermediate levels of disturbance. However, empirical evidence is equivocal, and quantitative predictions about the strength of diversity-disturbance relationships are lacking. Using Markov models of dynamics of real communities, I derived predicted changes in diversity when communities are subjected to quantified disturbance gradients. I also derived predictions regarding effects of sampling intensity and species selectivity of disturbance on diversity-disturbance relationships. My models predict peaked relationships should be relatively rare, variation in diversity over disturbance gradients typically should be low, and relationship shape varies with sampling intensity. These results are broadly consistent with a review of published diversity-disturbance relationships. A meta-analysis of 197 published diversity-disturbance relationships was performed to determine how frequently observed relationships are peaked; how strong, in general, relationships are; and whether various attributes of a study influence the observed shape and strength of a relationship. Non-significant relationships were the most common, and peaked responses were reported in only 16% of cases. Variation in diversity explained by disturbance was variable, but averaged ∼50%. Results suggest that strong and/or peaked relationships may arise from procedural artifacts. I conclude that there is little evidence to support the belief that disturbance should have a consistently important role in determining patterns of diversity, or that diversity-disturbance relationships are typically peaked.; Anthropogenic habitat loss is often cited as the most important cause of recent species' extinctions. Recent studies identifying hot spots of imperiled species have suggested that habitat loss is the primary factor threatening the survival of imperiled species. However, interpretation of such hot spots is not practical without knowledge of where imperiled species have been lost. I determined distributions of richness and losses of imperiled species in Canada, and statistically examined the relationship between species' losses and various landuse variables. Several hot spots of losses were identified in southern regions of Canada. The combination of frequent and intensive insecticide applications, routine herbicide use, and habitat loss due to agricultural development appears to be the most important threat to imperiled species. |
