Abstract
A bottom-seated artificial substrate periphyton sampler was designed and evaluated for use in shallow, swift rivers. The sampler was found to be robust, but 19% of samples were lost owing to flooding and 17% because of vandalism. Both glass and perspex are usable as substrates providing the surfaces are well etched. Vertically oriented substrates gave the highest sample precision and are recommended for routine monitoring. Variability within samplers was generally greater than that between samplers. Mean within-sampler coefficients of variations were 54% for algal density, 12% for algal diversity, 43% for chlorophyll .alpha., and 36% for ash-free dry weight. The numbers of replicates required to obtain given percentage errors for various determinands are presented as nomographs. One sampler per station enabled most moderate or large changes in the structure of communities to be detected.