Evaluation of precipitation impacts on benthic macroinvertebrate communities at three different stream types

We investigated the impacts of two natural disturbances (i.e., heavy rain and dry events) on benthic macroinvertebrates at three different stream types: perennial, intermittent and ephemeral streams. Co-occurrence networks displayed the differences of macroinvertebrate community stability among three different stream types. Cluster analysis and non-metric multidimensional scaling demonstrated distinct impacts of heavy rain and drought on macroinvertebrate communities. The samples from the heavy rain and dry events were clearly classified under different clusters than other samples regardless of the stream type. Lotic preferred species such as Kiotina decorata and Sweltsa nikkoensis were selected as an indicator species in the perennial stream, whereas lentic preferred species such as Ephemera strigata and Davidius lunatus were chosen in the intermittent stream. In the ephemeral stream, oligochaetes such as Dorydrilus sp. and Fridericia sp., which are readily found in unstable habitats with their short-life cycle, were selected as indicator species. Based on the Random Forest model, the amount of precipitation was the most influential factor, mainly for determining the distribution of benthic macroinvertebrate species, followed by electric conductivity and velocity. The Threshold Indicator Taxa Analysis (TITAN) detected 32 (17.6%) of 182 taxa in total as robust 42 response, and most of the taxa (26 taxa) were sensitive to the increase in precipitation, whereas 6 taxa were tolerant. The community change points for the amount of precipitation were 121.2 mm/month for tolerant species and 36.6 mm/month for sensitive species. Finally, our results revealed that the occurrences of two extreme hydrological disturbances (i.e., heavy rain and dry events) were primary factors for determining the spatial and temporal extent of habitat diversity and stability in the Korean peninsula streams under the least disturbance conditions.