Determinants of invertebrate community structure in glacial‐melt streams of southeast Tibet

Abstract

A widely examined predictive model of invertebrate community dynamics in glacial‐melt streams describes longitudinal changes in community structure with changing water temperature and channel stability with increasing distance from glaciers. Previous studies conducted in Europe, Greenland, New Zealand, and South America have supported the predictions of the invertebrate model and contributed to its refinement. However, none has evaluated if the model fits invertebrate community dynamics over a full range of distances from the glacier and water temperature conditions within glacial‐melt streams in southeast Tibet. We sampled invertebrates and measured water temperature, specific conductivity, turbidity, and associated glacier‐related variables within 14 sites in three subalpine glacial‐melt catchments in southeastern Tibet's Three Parallel Rivers region during 2010, 2011, 2013, and 2015. Our sites encompassed a temperature gradient from the upstream metakryal sites (maximum summer water temperature <2°C) to the furthest downstream site (maximum summer temperature >10°C) near the Mekong River. We evaluated the relationships of invertebrate community structure with in situ water temperature and channel stability which are the focal habitat variables in the invertebrate model. The additional habitat variables of distance from the glacier, glacier size, conductivity, and turbidity were evaluated to see if these were more important determinants of community structure than in situ water temperature and channel stability. Minimum and in situ water temperatures were positively correlated with distance from the glacier but Pfankuch Channel Stability Index bottom scores were not. Thus, the physical template within our study area differed from the expected template of the invertebrate model. Similar to the invertebrate model, in situ water temperature by itself or combined with Pfankuch index best explained five invertebrate response variables. In contrast with the invertebrate model, conductivity and turbidity best explained invertebrate taxa richness, density, and the site scores of the first and second detrended correspondence analysis axes of relative abundance. The invertebrate model predicts that only Diamesinae will occur in metakryal sites. However, in our metakryal sites we frequently captured 13 taxa (two Nemouridae morphotypes, Diamesinae, Orthocladiinae, Rhyacophila, Epeorus, Taeniopterygidae, Baetis, Capnia, Simuliidae, Limnephilidae, Himalopsyche, and Collembola). Invertebrate‐habitat relationships and taxa occurrence trends in glacial‐melt catchments in southeast Tibet differed from the invertebrate model predictions. Our findings highlight the need to develop a regional version of the invertebrate model applicable to Asian glacial‐melt streams with unstable stream channels throughout their catchments and that do not freeze in the winter.