Lake macroinvertebrate assemblages and relationship with natural environment and tourism stress in Jiuzhaigou Natural Reserve, China

•Macroinvertebrate assemblages in high-elevation lakes in Jiuzhaigou Natural Reserve, China were investigated to assessing the effect of tourism.•Light traps added many EPT taxa to benthic samples.•Variation of taxa and EPT richness calculated for the composite samples (benthic + light trap) were well predicted by GLM with elevation, tourism index, and total-N.•\Both metrics for benthic samples were poorly explained.•GLM based on natural environmental factors can be used to set biological expectations under the least disturbances in assessing tourism impacts.

With increasing human population and urbanization, tourism in natural reserves and other protected lands is growing. It is critical to monitor and assess the impacts of tourism on ecosystem health. However, there is a general lack of information on biological communities in natural reserves of developing countries and of tools for assessing human impacts. In the present study, we investigated macroinvertebrate assemblages in nine lakes in Jiuzhaigou Natural Reserve of China. Both benthic (20 dips of D-net) and light-trap samples (2 h) were collected at each lake and all benthic specimens and adults of Ephemeroptera, Plecoptera, and Trichoptera (EPT) were identified and counted. Water temperature and water quality variables were measured on site or in the Lab. Seventy taxa were recorded and dominated by dipterans and caddisflies. Light traps contributed 47% of taxa richness and 66% of EPT richness at the lakes. Detrended Correspondence Analysis showed that water temperature and tourism index were strongly associated with the changes of assemblage composition. Taxa richness and EPT richness calculated for the composite samples (benthic + light trap) were well fit with Poisson generalized linear model (adjusted R2 = 0.83 and 0.85, respectively), generally decreasing with increasing elevation, tourism index, and total-N. Tourism index was ranked as the top predictor for EPT richness based on multiple model weights, and elevation for taxa richness. In comparison, when based on benthic samples, neither of the metrics could be fitted with the seven environmental variables selected. These findings highlight the benefit of combined use of the sampling methods for lake monitoring and offered an analytical guide to developing biological indicators of lake ecosystem health in protected areas.