eDNA provides accurate and precise estimates of abundance by integrating bioenergetics and particle mass-balance

Abstract

Anthropogenic activities have led to an unprecedented crisis in freshwater biodiversity loss. The capacity to monitor the abundance of wild populations is critical to conserving biodiversity, but conventional physical specimen collection methods are invasive, costly, and labour-intensive. Environmental DNA (eDNA) offers a promising alternative, being easy to sample, with studies under controlled laboratory conditions showing consistent correlations between eDNA concentration and abundance. However, applying eDNA to monitor abundance remains contentious, as eDNA particle dynamics and the ecology of eDNA production can decouple this relationship in natural ecosystems. To address this, we integrated bioenergetics and mass-balance frameworks to relate eDNA concentrations to freshwater fish populationabundanceestimated through conventional mark-recapture in Brook Trout (Salvelinus fontinalis) across nine Rocky Mountains lakes, five of which underwent size-selectively harvest over two years. Our integrated framework improved the variance explained in eDNA concentrations from 24% to 71%. The integrated model accurately distinguished most abundance estimates across populations and sampling periods, detecting both natural and harvest-induced reductions in abundance within several populations. This study is the first to empirically integrate the DNA production ecology and particle dynamics, demonstrating that properly accounting for ‘the ecology of eDNA’ enables rapid, accurate abundance quantification. We also discuss future research and applications of eDNA in light of these results.Significance statementAddressing the unprecedented freshwater biodiversity crisis requires cost- and time-effective methods for estimating abundance and biomass. Our study empirically demonstrates the potential of eDNA for rapid and accurate assessment of abundance in natural populations by integrating hydrology, chemistry and bioenergetics into a mass-balance modelling framework. Through a replicated whole-ecosystem experiment, we show that eDNA not only detects differences in fish abundance and biomass across populations, but also captures natural and experimentally induced abundance changes within populations. Our modeling approach enables eDNA to generate abundance and biomass estimates comparable in precision and accuracy to high-quality mark-recapture methods.