Including environmental covariates clarifies the relationship between endangered Atlantic salmon (Salmo salar) abundance and environmental DNA

Collecting environmental DNA (eDNA) as a nonlethal sampling approach has
been valuable in detecting the presence/absence of many imperiled taxa;
however, its application to indicate species abundance poses many
challenges. A deeper understanding of eDNA dynamics in aquatic systems is
required to better interpret the substantial variability often associated
with eDNA samples. Our sampling design took advantage of natural variation
in juvenile Atlantic salmon (Salmo salar) distribution and abundance along
9 km of a single river in the Province of New Brunswick (Canada), covering
different spatial and temporal scales to address the unknown seasonal
impacts of environmental variables on the quantitative relationship
between eDNA concentration and species abundance. First, we asked whether
accounting for environmental variables strengthened the relationship
between eDNA and salmon abundance by sampling eDNA during their spring
seaward migration. Second, we asked how environmental variables affected
eDNA dynamics during the summer as the parr abundance remained relatively
constant. Spring eDNA samples were collected over a 6‐week period (12
times) near a rotary screw trap that captured approximately 18.6% of
migrating smolts, whereas summer sampling occurred (i) at three distinct
salmon habitats (9 times) and (ii) along the full 9 km (3 times). We
modeled eDNA concentration as a product of fish abundance and
environmental variables, demonstrating that (1) with inclusion of
abundance and environmental covariates, eDNA was highly correlated with
spring smolt abundance and (2) the relationships among environmental
covariates and eDNA were affected by seasonal variation with relatively
constant parr abundance in summer. Our findings underscore that with
appropriate study design that accounts for seasonal environmental
variation and life history phenology, eDNA salmon population assessments
may have the potential to evaluate abundance fluctuations in spring and
summer.