Comparisons of eDNA and traditional sampling methods for enhanced assessment of benthic macroinvertebrate diversity in lotic environments with spatially varying land use

Abstract

Environmental DNA (eDNA) metabarcoding is emerging as a beneficial tool for assessing benthic macroinvertebrate diversity due to its speed and simplicity of sampling. This study evaluates the effectiveness of eDNA, qualitative (kick or scoop net), and quantitative (Surber net) sampling methods in characterizing benthic macroinvertebrate diversity in lotic ecosystems. The samplings were conducted across eleven unique river sites, each showing diverse types of riparian land cover including forest, agricultural, and urban areas. We identified 114, 79, and 110 macroinvertebrate species, respectively, via qualitative, quantitative, and eDNA-based sampling methods. Venn diagram analysis revealed the complementary nature of these three methods, each detecting a unique array of taxa. Especially, eDNA sampling could detect rare species and identify the species which were difficult to classify into the lowest level based on eye-observation. Non-metric multidimensional scaling (NMDS) analysis showed differences in longitudinal gradients across survey sites, regardless of the sampling method employed. Notably, spatial autocorrelation analysis demonstrated that capture-based sampling methods (qualitative and quantitative) revealed both positive autocorrelation at shorter distances and negative autocorrelation at more distant sites, suggesting species gradient. However, eDNA sampling only showed a positive autocorrelation at shorter distances. Our findings validate the synergistic use of eDNA-based and traditional sampling methods, illustrating how their combined application provides a deeper and more nuanced understanding of benthic macroinvertebrate diversity in freshwater ecosystems. This comprehensive approach not only reveals hidden species and complex ecological patterns missed by single-method studies but also enhances spatial analysis capabilities, as evidenced by our spatial autocorrelation findings. These insights are crucial for developing effective biodiversity conservation strategies and ensuring sustainable management of aquatic resources.