RAPID ASSESSMENT OF AUSTRALIAN RIVERS USING MACROINVERTEBRATES - COSTAND EFFICIENCY OF 6 METHODS OF SAMPLE PROCESSING

The cost and efficiency of different techniques for processing rapid-a ssessment samples of lotic macroinvertebrates were evaluated. Subsampl es of 50, 100, and 150 selectively picked individuals were compared to determine the optimal subsample size for selective picks. Selective s ubsampling of a set number of animals was then compared with subsampli ng by picking for a set time (30 min), exhaustive picking of a sample, and randomly subsampling 100 animals. Identification was to family le vel. Efficiency was assessed by comparing the ability of data generate d by each technique to discriminate between unpolluted reference sites and sites with mild to severe pollution by municipal effluent and urb an stormwater runoff. This discrimination was tested using both univar iate metrics (SIGNAL biotic index, family richness, and EPT [Ephemerop tera, Plecoptera, Trichoptera] family richness) and multivariate tests (analysis of similarity-ANOSIM). Cost was assessed as the amount of t ime each method required for picking and identification. The selective 100-animal subsample is recommended as the most cost-effective method for assessing sites affected by municipal sewage-treatment-plant disc harges and urban stormwater runoff. SIGNAL was the best of the 3 metri cs, having the most highly significant differences between reference a nd polluted sites and being most robust to variations in processing me thod. ANOSIM could distinguish between the reference and polluted site s using any of the sample processing methods, but the degree of the di stinction varied with sample processing method.