A MULTIVARIATE METHOD FOR DETERMINING OPTIMAL SUBSAMPLE SIZE IN THE ANALYSIS OF MACROINVERTEBRATE SAMPLES

The depauperate macroinvertebrate communities of two disturbed urban s treams were used to demonstrate a technique for determining optimal su bsample size for the multivariate representation of community structur e. Although multivariate analyses have commonly been applied to subsam pled data, the effect of subsampling on multivariate patterns has not previously been investigated in detail. The minimum requirement for ap plying the method is eight fully processed community samples distribut ed across two treatments, with n = 4 (being a representative subset of a larger study to which subsampling will be applied). The data from e ach sample are subsampled repeatedly with the aid of a simulation prog ram. Variation in community structure between treatments is tested by analysis of similarity for the complete data and for multiple sets of subsampled data. The optimal subsampling strategy is defined as the mi nimum effort required to achieve a median R-statistic no less than the value derived for the complete data. Subsampling to a fixed proportio n is a less efficient strategy than subsampling to a fixed number of i ndividuals. For the urban stream communities, subsampling to 300 indiv iduals was adequate for four out of six comparisons and was adequate f or all comparisons when a minimum proportion of 10% was applied. Sever al non-abundant taxa were found to be important contributors to differ ences between treatments. Subsamples need to be large enough to adequa tely estimate the abundance of such taxa. Further investigation is req uired to assess the relationship between optimal subsample size for mu ltivariate analysis and the shape of cumulative species-abundance curv es.