Using environmental DNA to improve species distribution models for freshwater invaders

Abstract

Species Distribution Models (SDMs) have been reported as a useful tool for the risk assessment and modeling of the pathways of dispersal of freshwater invasive alien species (IAS). Environmental DNA (eDNA) is a novel tool that can help detect IAS at their early stage of introduction and additionally improve the data available for a more efficient management. SDMs rely on presence and absence of the species in the study area to infer the predictors affecting species distributions. Presence is verified once a species is detected, but confirmation of absence can be problematic because this depends both on the detectability of the species and the sampling strategy. eDNA is a technique that presents higher detectability and accuracy in comparison to conventional sampling techniques, and can effectively differentiate between presence or absence of specific species or entire communities by using a barcoding or metabarcoding approach. However, a number of potential bias can be introduced during (i) sampling, (ii) amplification, (iii) sequencing, or (iv) through the usage of bioinformatics pipelines. Therefore, it is important to report and conduct the field and laboratory procedures in a consistent way, by (i) introducing eDNA independent observations, (ii) amplifying and sequencing control samples, (iii) achieving quality sequence reads by appropriate clean-up steps, (iv) controlling primer amplification preferences, (v) introducing PCR-free sequence capturing, (vi) estimating primer detection capabilities through controlled experiments and/or (vii) post-hoc introduction of "site occupancy-detection models." With eDNA methodology becoming increasingly routine, its use is strongly recommended to retrieve species distributional data for SDMs.