A framework for inferring biological communities from environmental DNA

Environmental DNA (eDNA), genetic material recovered from an environmental medium such as soil, water, or feces, reflects the membership of the ecological community present in the sampled environment. As such, eDNA is a potentially rich source of data for basic ecology, conservation, and management, because it offers the prospect of quantitatively reconstructing whole ecological communities from easily obtained samples. However, like all sampling methods, eDNA sequencing is subject to methodological limitations that can generate biased descriptions of ecological communities. Here, we demonstrate parallels between eDNA sampling and traditional sampling techniques, and use these parallels to offer a statistical structure for framing the challenges faced by eDNA and for illuminating the gaps in our current knowledge. Although the current state of knowledge on some of these steps precludes a full estimate of biomass for each taxon in a sampled eDNA community, we provide a map that illustrates potential methods for bridging these gaps. Additionally, we use an original data set to estimate the relative abundances of taxon-specific template DNA prior to PCR, given the abundance of DNA sequences recovered post-PCR-and-sequencing, a critical step in the chain of eDNA inference. While we focus on the use of eDNA samples to determine the relative abundance of taxa within a community, our approach also applies to single-taxon applications (including applications using qPCR), studies of diversity, and studies focused on occurrence. By grounding inferences about eDNA community composition in a rigorous statistical framework, and by making these inferences explicit, we hope to improve the inferential potential for the emerging field of community-level eDNA analysis.