Predicting the constraint effect of environmental characteristics on macroinvertebrate density and diversity using quantile regression mixed model

Abstract

Various factors, such as habitat availability, competition for space, predation, temperature, nutrient supplies, presence of waterfalls, flow variability and water quality, control the abundance, distribution and productivity of stream-dwelling organisms. Each of these factors can influence the response of the density of organisms to a specific environmental gradient, inflating variability and making difficult to understand the possible causal relationship. In our study, we used quantile regression mixed models and Akaike’s information criterion as an indicator of goodness to examine two different datasets, one belonging to Italy and one belonging to Finland, and to detect the limiting action of selected environmental variables. In the Italian dataset, we studied the relationships among five macroinvertebrate families and three physical habitat characteristics (water velocity, depth and substratum size); in the Finnish dataset the relationships between taxa richness and 16 environmental characteristics (chemical and physical). We found limiting relationships in both datasets and validated all of them on different datasets. These relationships are quantitative and can be used to predict the range of macroinvertebrate densities or taxa richness as a function of environmental characteristics. They can be a tool for management purposes, providing the basis for habitat-based models and for the development of ecological indices.