Interactive effects of urban stormwater drainage, land clearance, and flow regime on stream macroinvertebrate assemblages across a large metropolitan region

Abstract

Urban impacts on streams are mediated by alteration of land cover and subsurface drainage. Understanding interactions of these 2 effects is critical for urban land and water management. We used boosted regression trees to analyze relationships between environmental predictors, including land-cover measures that indicate urban impacts, and Stream Invertebrate Grade Number Average Level (SIGNAL) or the occurrence of 60 macroinvertebrate families across the Melbourne (Australia) region. SIGNAL and 51 families were negatively correlated with attenuated imperviousness (AI). Twenty-four families exhibited a steep decline in probability of occurrence between 0 and ≤3% AI, consistent across the range of natural discharge (Q). Responses to attenuated forest cover (AF) were more variable among families (39 positive, 15 negative). AF had a less positive influence on SIGNAL at high than at low AI because the many (high SIGNAL-grade) families that were strongly negatively correlated with AI were likely to be absent at high AI. Interactions that suggested a modification of response to AI or AF as a result of the influence of another variable were restricted to 19 families. For 4 families with occurrences that were positively correlated with AF, the negative effect of AI was reduced in streams with high AF. For 14 families, responses of occurrence to AI or AF varied with Q or antecedent discharge in multiple ways. A reduced effect of AI was observed in low-Q streams or after very dry antecedent conditions for 10 families. Despite strong effects of the regional gradient in rainfall and vegetation across our study area, catchment-scale urban stormwater drainage was consistently associated with reduced occurrence of most families. The relative number of families predicted to be lost at low levels of AI underestimated relative loss of genera or species. Urban stormwater runoff probably is a strong driver of species loss in streams, but riparian forest can increase the probability of occurrence of a small number of tolerant families.

Note: Journal articles and conference papers (and links where available) are available under open access arrangements where possible. Otherwise please contact your institution’s library, the authors, or publishers to organise full access.