The use of taxonomic distinctness to assess environmental disturbance of insect communities from running water

Summary 1. Taxonomic distinctness (or average taxonomic breadth) of a sample is a measure of diversity that has the great advantage of not being sensitive to variations in sampling effort, unlike measures of species richness. It can also be tested for departures from expectation and can thus be used to determine whether a site has suffered from a genuine loss of diversity. 2. Taxonomic distinctness or delta+ was measured using all insect species from three datasets obtained from running water habitats: reference or least disturbed sites from channel and bank habitats on rivers from the Australian state of Victoria; test sites distributed along a water quality gradient in Victoria and sites immediately below a series of dams in Victoria and New South Wales. The last two datasets comprised sites subject to a range of obvious disturbances. 3. At reference sites delta+ was generally within the range of expected values (95% confidence limits) for both habitats as would be anticipated for these essentially undisturbed sites. However, 13–14% of sites fell below the lower confidence limits. Sites in river basins with less cover of natural habitat had slightly but significantly lower delta+ values than those with more cover and these sites contributed to this anomaly. 4. Delta+ was inversely correlated with water quality (r = −0.54, P = 0.002) as measured by the first axis of a PCA of water quality variables. It was also inversely correlated with individual variables, such as turbidity and total phosphorous. 5. Delta+ was below expectation for 18 of the 19 dam sites and at 13 of these sites delta+ values were below the lower confidence limit. 6. Delta+ was a sensitive index of diversity that generally distinguished sites suffering from environmental disturbance. Only a few previous studies of river ecosystems have used delta+, which did not respond consistently to disturbance because either too few taxa at species or higher taxonomic levels were available or because longitudinal changes in species composition swamped the signal from delta+. Neither of these problems occurred with the current datasets. 7. Delta+ responded more readily to disturbance than number of species, perhaps because taxonomic distinctness reflects some of the multivariate nature of community samples, whereas number of species is blind to this aspect.