Detecting spawning of threatened chum salmon Oncorhynchus keta over a large spatial extent using eDNA sampling: Opportunities and considerations for monitoring recovery

Abstract

AbstractOver the last century, anthropogenic activities have caused substantial declines in the abundance of chum salmon (Oncorhynchus keta) throughout their range in western North America. As a result, chum salmon were listed as threatened under the U.S. Endangered Species Act in 1999. Recovery strategies have been developed, but limited baseline data on distribution impedes implementation of these strategies. Traditional methods for identifying spawning distribution (e.g., spawning ground surveys) may be inadequate for rare species like chum salmon because of a low detection probability. In contrast, environmental DNA (eDNA) sampling is extremely sensitive to species presence and has the potential to supplement or replace existing methods for monitoring listed species. However, legal and administrative issues associated with accurately describing distribution of a listed species put a premium on understanding factors that influence detection of a species by eDNA sampling. In this study, we (a) developed and tested a quantitative PCR‐based eDNA assay for chum salmon, (b) collected eDNA samples to describe the spawning distribution of this species in Columbia River tributaries between Bonneville and The Dalles Dams, and (c) tested whether spawn surveyors could inadvertently transport chum salmon DNA between streams on their boots and waders. The newly developed assay was specific and sensitive to chum salmon DNA in both tissue and eDNA samples. Chum salmon DNA was detected in the positive control site and in four streams, which increased known contemporary spawning locations in the study area. In the contamination trial, surveyors successfully introduced chum salmon DNA into the study area but eDNA was only detected intermittently and locally over the next 11 days. The accuracy of eDNA sampling, combined with success in detecting a rare species at the population scale, makes this technique well‐suited for monitoring recolonization of chum salmon during the recovery process.