eDNA metabarcoding: a non-invasive method to track temporal community dynamics in temporary rivers

Resum

Temporary rivers (TRs) are dynamic ecosystems that typically shift between hydrological phases (i.e. flowing, disconnected pools, and dry). Disconnected pools have high conservation value as refugia for aquatic species during the dry season. However, these pools have been neglected by bioassessment programs. To assess the biological quality of these habitats, traditional methods (i.e., morphology-based) can be invasive and eliminate a large proportion of the community, diminishing their function as refugia. Within this context, environmental DNA (eDNA) metabarcoding offers a promising sampling alternative, since it is a minimally invasive method to compile community information from sediment samples. We tested the effectiveness of sediment eDNA alongside bulk DNA to characterize the macroinvertebrate communities and assess the biological quality of disconnected pools in TRs, comparing them with traditional methods. Additionally, we assessed the ability of sediment eDNA to reflect the components of beta diversity—nestedness and turnover—along the disconnection process. The biological quality was determined through macroinvertebrate indices widely used in Spain (i.e., IBMWP, family richness, and IASPT). Sediment eDNA was sampled every 15 days from three TRs in Catalonia, NE Spain. Macroinvertebrates were sampled only three times (connected, disconnecting, and disconnected pools). Macroinvertebrate samples were used to produce bulk DNA but also sorted and identified using morphology. Sediment eDNA and bulk DNA samples were processed via metabarcoding analysis of the mitochondrial COI gene. Although sediment eDNA was less effective in representing indices (i.e., IBMWP and family richness), it provided valuable insights into the temporal changes in community composition. Species turnover contributed more than the nestedness component in beta-diversity patterns, as stated in previous studies. Given the high community turnover detected in disconnected pools in TRs, we show that eDNA is a valuable tool for bioassessment because frequent monitoring of these ecosystems is needed to capture the transitions between hydrological phases.