Comparison of qPCR and metabarcoding for environmental DNA surveillance of a freshwater parasite

Abstract

Analysis of environmental DNA (eDNA) has been successfully used across freshwater ecological parasitology to inform management of ecologically and economically important species. However, most studies have used species-specific quantitative polymerase chain reaction (qPCR) assays to detect target taxa. While generally effective, this approach limits the amount of community and management-supporting data that can be obtained from eDNA samples. If eDNA metabarcoding could be conducted with the same accuracy of a single species approach, researchers could simultaneously detect a target species while obtaining vast community data from eDNA samples. We sampled 38 freshwater sites on Fort McCoy, Wisconsin and compared qPCR to metabarcoding for eDNA detection of the ectoparasitic gill louse Salmincola edwardsii, an obligate parasite of Salvelinus fishes (chars). We found no evidence to suggest S. edwardsii occupancy or detection probabilities differed between qPCR and metabarcoding. Further, we found that the number of S. edwardsii reads from metabarcoding were negatively predictive of C _T values from qPCR (C _T value indicates cycle a significant amount of target eDNA is detected, with lower C _Ts indicative of more DNA), demonstrating that our metabarcoding reads positively predicted qPCR DNA quantities. However, the number of reads was not predictive of overall qPCR score (number of positive qPCR replicates). In addition to S. edwardsii, metabarcoding led to the detection of a vast community of over 2600 invertebrate taxa. We underscore the necessity for conducting similar analyses across environments and target species, as the ecology of eDNA will vary on a per-study basis. Our results suggest that eDNA metabarcoding provides a highly sensitive and accurate method for detecting parasitic gill lice while also illuminating the broader biological community and co-occurrence of species in the environment.

Keywords: Salmincola edwardsii; eDNA; metabarcoding; occupancy modeling; qPCR.

FIGURE 1

A Salmincola edwardsii infestation on a brook trout (Salvelinus fontinalis) at Fort McCoy, Wisconsin, USA. Photo Credit: John Noble.

FIGURE 2

Sites on Fort McCoy, Wisconsin, USA where stream water was collected to test for the presence of Salmincola edwardsii environmental DNA. Colored triangles indicate whether gill lice eDNA was detected, detected only with qPCR, detected only with metabarcoding, or detected with both methods.

FIGURE 3

Occupancy and detection probability estimates with 89% Bayesian credible intervals for gill lice (Salmincola edwardsii) environmental DNA from streams on Fort McCoy, Wisconsin, USA based on molecular assay (qPCR or metabarcoding).

FIGURE 4

Predicted C _T values from qPCR as a function of reads from metabarcoding for detection of gill lice (Salmincola edwardsii) environmental DNA from streams on Fort McCoy, Wisconsin, USA. Note that as C _T values decrease, the number of reads increases.