Environmental DNA metabarcoding reveals the impact of different land use on multitrophic biodiversity in riverine systems.

Human-induced changes in land use drive an alarming decline in river biodiversity and related ecosystem services worldwide. However, how different land use shapes aquatic multitrophic communities is still not well understood. Here, we used the biodiversity dataset from bacteria to fish captured by the environmental DNA (eDNA) approach in the four riverine systems with spatially different land use (i.e., Slightly disturbed group, Upstream disturbed group, Downstream disturbed group, and Strongly disturbed group) to reveal the changes in multitrophic biodiversity in relation to human land use. Firstly, our data showed that spatially different land use determined the pollutant loads of the riverine systems, most pollutants (e.g., TN and NH3-N) had significant differences among the four riverine systems. Secondly, taxonomic α diversity across multitrophic levels did not necessarily change significantly, yet the change in community structure can be considered as a more sensitive indicator to reflect different land use, because different land use shaped the unique structure of multitrophic communities, and the dissimilarity of community structure was closely associated with land use gradient (e.g., positive relationships in the Slightly disturbed group, negative relationships in the Strongly disturbed group). Thirdly, different land use induced the shifts of key taxa, resulting in the variation of community structure and the change of co-occurrence network. Overall, these findings suggest that spatially different land use plays a critical role in shaping aquatic multitrophic communities, and an in-depth understanding of the interdependences between biodiversity and land use is a critical prerequisite for formulating river management strategies.