Estimating Organism Abundance Using Within-Sample Haplotype Frequencies of eDNA Data

Environmental DNA (eDNA) provides powerful insights into species presence and community composition but remains limited in its capacity to infer species abundance or population structure. Here, we show that the deviation between within-sample haplotype frequencies and the overall population-level haplotype frequencies can be used to estimate the number of individual contributors to a given sample. We first establish the theoretical framework for approximating population haplotype frequencies directly from eDNA data, enabling application even in the absence of tissue-derived references. Building on this foundation, we introduce a maximum likelihood estimator to infer the number of contributors and assess its performance through simulations spanning a range of haplotype frequency distributions and noise scenarios. These approaches assume that all samples are drawn from a single, panmictic population. We find that accurate estimates are attainable when haplotypes are sufficiently variable, population frequencies are well-characterised, and samples are large enough to capture frequency deviations. By bridging population genetic theory and eDNA, our method complements existing molecular approaches and offers a novel path towards quantifying abundance from eDNA metabarcoding data.