Arscott, D.B., S.T. Larned, M. Scarsbrook, and P. Lambert. 2010. Journal of the North American Benthological Society 29(2):530–545.
Changes in community structure and life-history traits of benthic invertebrates were examined along a longitudinal intermittence gradient in an alluvial river. The gradient was characterized with modeled and measured hydrologic, chemical, and physical environmental variables. The invertebrates were collected in the Selwyn River, southeastern New Zealand, at multiple sites in each of 4 river sections with distinct hydrological conditions (perennial-losing, ephemeral, intermittent, perennial-gaining). Values of hydrological metrics for each site were generated with an empirical model developed for the Selwyn River. The metrics included 4 that characterized intermittent flow (flow permanence, flow duration, drying frequency, distance to nearest perennial site). Most invertebrate richness and density metrics were significantly higher in the perennial-losing and perennial-gaining sections than in the ephemeral and intermittent sections. A principle components analysis (PCA) separated invertebrate samples from the 4 sections along 2 primary factors. Nine of 13 hydrological metrics, including the 4 intermittence metrics, were correlated with the PCA site scores. Linear regressions indicated that most taxon-richness metrics and some density metrics were related to flow permanence, flow duration, or both. Based on the regression analysis, we predicted that 1.9 taxa/m2 are added with each 10% increase in flow permanence, and 0.5 taxa/m2 are added with each 10-d increase in flow duration. Results from a nestedness analysis indicated that communities at ephemeral and intermittent sites were nested subsets of the communities at perennial sites, and the nesting order of sites was related to both flow permanence and flow duration. Assemblages of taxa with particular life-history traits (life span, fecundity, maximum size, and voltinism) varied linearly with flow permanence and flow duration. The variation in invertebrate communities along the Selwyn River was primarily the result of progressive removal of desiccation-sensitive taxa with increasing intermittence, not to selection for desiccation-resistant specialists. Quantitative intermittence–ecology relationships are needed to predict the consequences of future changes in flow intermittence, but such relationships are rare. The univariate relationships reported in our study contribute to a small but growing array of intermittence–ecology relationships.