How to design optimal eDNA sampling strategies for biomonitoring in river networks

Abstract

The current biodiversity crisis calls for appropriate methods for assessing biodiversity. In this respect, environmental DNA (eDNA) holds great promise, especially for aquatic ecosystems. While initial eDNA studies assessed biodiversity at single sites, technology now allows analyzing samples from many points simultaneously. However, the selection of these sites has been mostly motivated on an ad‐hoc basis. To this end, hydrology‐based models might offer a unique guidance on where to sample eDNA to most effectively reconstruct spatial patterns of biodiversity. Here, we performed computer simulations to identify best‐practice criteria for the choice of positioning of eDNA sampling sites in river networks. To do so, we combined a hydrology‐based eDNA transport model with a virtual river network reproducing the scaling features of real rivers. In particular, we conducted simulations investigating scenarios of different number and location of eDNA sampling sites in a riverine network, different spatial taxon distributions, and different eDNA measurement errors. We found that, due to hydrological controls, non‐uniform patterns of eDNA concentration arise even if the taxon distribution is uniform and decay is neglected. Best practices for sampling site selection depend on the taxon's spatial distribution: when taxa are concentrated in some hotspots and only few sampling sites can be placed, it is better to preferentially locate them in the downstream part of the catchment; when taxa are more evenly distributed, and/or many sites can be placed, these should be preferentially located upstream. We also found that uncertainties in eDNA concentration estimates do not necessarily hamper model predictions. Knowledge of eDNA decay rates improves model predictions, highlighting the need for empirical estimates of these rates under relevant environmental conditions. Our simulations help define strategies for designing eDNA sampling campaigns in river networks and can guide the sampling effort of field ecologists and environmental authorities.

Integrating physiology and environmental dynamics to operationalize environmental DNA (eDNA) as a means to monitor freshwater macro‐organism abundance

“…Mass-balance frameworks, similar to lentic systems, may be a particularly promising approach. A number of recent studies have integrated hydrology and geomorphology at the river-network scale to model both the distribution of eDNA and likely distribution of progenitor organisms using a mass-balance framework (Carraro et al, 2018(Carraro et al, , 2021. Such models assume that eDNA in river-networks is subject to transport processes that can be approximated based on well-studied hydraulic properties of river networks and eDNA degradation/deposition processes (Carraro et al, 2021), and are a promising approach to integrating the complex environmental dynamics of lotic systems to infer abundance from observed eDNA patterns.…”

“…A number of recent studies have integrated hydrology and geomorphology at the river-network scale to model both the distribution of eDNA and likely distribution of progenitor organisms using a mass-balance framework (Carraro et al, 2018(Carraro et al, , 2021. Such models assume that eDNA in river-networks is subject to transport processes that can be approximated based on well-studied hydraulic properties of river networks and eDNA degradation/deposition processes (Carraro et al, 2021), and are a promising approach to integrating the complex environmental dynamics of lotic systems to infer abundance from observed eDNA patterns. An additional advantage is that such frameworks can also be built to accommodate known effects of environmental variables (e.g., temperature) and organism physiology (e.g., body size) on eDNA production rates (Carraro et al, 2018).…”

Environmental DNA from a small sample of reservoir water can tell volumes about its biodiversity

“…Twelve were assigned to higher levels: 3 to order (Hemiptera, Neuroptera and Ploimida), 3 to class (Ostracoda, Bangiophyceae and Florideophyceae), 1 to subclade (Copepoda), 1 to clade (stramenopiles), and 3 to phylum (Chlorophyta, Tubulinea and Discosea), as shown in Figure 2A. The frequency of DNA fragments in the heatmap represents a rough estimation of the abundance of the species, even though this DNA abundance should be interpreted with caution as it may reflect body size of individual species [35] or drift [36]. Taxa richness was similar at each of the 3 depths (Poisson Generalized Linear Model β depth = -4.334e-17, z = 0, p = 1), , with each community presenting 35 taxa and according to the.…”

Extracting abundance information from DNA-based data

“…The accurate extraction of species-abundance information from DNA-based data could contribute usefully to the reconstruction of diets and quantitative food webs, the inference of species interactions, the modelling of population dynamics and species distributions, the biomonitoring of environmental state and change, and more prosaically, the inference of false positives and negatives (Thomas et al 2016; Deagle et al 2019; Peel et al 2019; Carraro et al 2020, 2021; Abrego et al 2021; Rojahn et al 2021). Here we use the term abundance to mean any estimate of biomass or count of individuals .…”