All species inevitably leave genetic traces in their environments, and the resulting environmental DNA (eDNA) reflects the species present in a given habitat. It remains unclear whether eDNA signals can provide quantitative metrics of abundance on which human livelihoods or conservation successes depend. Here, we report the results of a large eDNA ocean survey (spanning 86,000 km² to depths of 500m) to understand the abundance and distribution of Pacific hake Merluccius productus, the target of the largest finfish fishery along the west coast of the United States. We sampled eDNA in parallel with a traditional acoustic-trawl survey to assess the value of eDNA surveys at a scale relevant to fisheries management. Despite local differences, the two methods yield comparable information about the broad-scale spatial distribution and abundance. Furthermore, we find depth and spatial patterns of eDNA closely correspond to acoustic-trawl estimates for hake. We demonstrate the power and efficacy of eDNA sampling for estimating abundance and distribution and move the analysis eDNA data beyond sample-to-sample comparisons to management relevant scales. We posit that eDNA methods are capable of providing general quantitative applications that will prove especially valuable in data- or resource-limited contexts.

There are two main datasets for this paper. Both were collected during the 2019 U.S.-Canada Integrated Ecosystem & Acoustic-Trawl Survey for Pacific hake aboard the NOAA Ship Bell M. Shimada from July 2 to August 19. The first is data associated hake DNA concentrations as identified in water samples collected and analyzed using qPCR analysis. The second dataset is estimates of hake biomass derived from analysis of the acoustic-trawl information. Refer to the primary manuscript for detailed information on data collection and processing protocols.

We provide scripts for reading in and manipulating data as well as estimating statistical models and visualizing result from the raw data. Please refer to scripts in the repository for mainpulations of the raw data.