Environmental DNA analysis of river herring in Chesapeake Bay: A powerful tool for monitoring threatened keystone species

Abstract

Environmental DNA (eDNA) sampling has emerged as a powerful tool to detect and quantify species abundance in aquatic environments. However, relatively few studies have compared the performance of eDNA-based abundance estimates to traditional catch or survey approaches in the field. Here, we have developed and field-tested a qPCR assay to detect eDNA from alewife and blueback herring (collectively known as 'river herring'), comparing eDNA-based presence and abundance data to traditional methods of quantification (ichthyoplankton sampling and adult observations). Overall, the qPCR assay showed very high target specificity in lab trials, and was successful in detecting river herring for 11/12 Chesapeake Bay tributaries in spring 2015 and 2016, with 106 out of 445 samples exhibiting positive eDNA hits. We found a strong correlation between eDNA abundance and ichthyoplankton count data (Spearman's Rho = 0.52), and Phi-tests (correlation of presence/absence data) showed higher correlation between eDNA and ichthyoplankton data (Phi = 0.45) than adult data (Phi = 0.35). Detection probability was significantly lower on western vs. eastern shore tributaries of Chesapeake Bay, and blueback herring and alewife were more likely detected on the western and eastern shores, respectively. Temporal patterns of eDNA abundance over the spring spawning season revealed that alewife were present in high abundances weeks ahead of blueback herring, which aligns with known differences in spawning behavior of the species. In summary, the eDNA abundance data corresponded well to other field methods and has great potential to assist future monitoring efforts of river herring abundance and habitat use.

Fig 1. Map of eDNA and field sampling locations in Chesapeake Bay.

Inset shows the focus region (Chesapeake Bay in Red) within the United States of America.

Fig 2. qPCR results for lab trials of target and non-target fish.

300K and 3K copy samples are the PCR results for respective synthesized standards on the standard curve, ‘GzShad’ is gizzard shad, ‘Men’ is menhaden, ‘Hick’ is hickory shad, ‘Amer’ is American shad, and ‘Awife’ and ‘BBack’ are alewife and blueback herring, respectively. BL is a no template control (blank).

Fig 3. Histogram of positive eDNA detection copy numbers from 2015–2016.

eDNA abundance (copy number) is plotted on a log scale from 1 copy up to 1 million (x-axis), with the count of occurrences (number of samples) within each histogram bin on the y-axis. eDNA copy numbers for each sample reflect the initial number of mtDNA copies per 40 mL of water filtered.

Fig 4. Map of positive river herring eDNA detections across Chesapeake Bay.

Size of data points (positive detections) proportional to the magnitude of amplification (mean mtDNA copies) and are colored by tributary. eDNA copy numbers for each sample reflect the initial number of mtDNA copies per 40 mL water filtered.

Fig 5. Map of positive eDNA detections broken down by species (alewife or blueback herring).

Size of the point is proportional to the magnitude of amplification (mean mtDNA copies). eDNA copy numbers for each sample reflect the initial number of mtDNA copies per 40 mL of water filtered.

Fig 6. Temporal distribution of eDNA copy number (abundance) by species, during spawning.

Data for years 2015 and 2016 combined. Multiple data points (filled circles) at the same date (i.e. same x-axis coordinate) reflect samples taken either on the same day at different sites (same year) or the same date across years.

Fig 7. Correlation results (Phi coefficients) among the abundance metrics.

Panel A shows the summary tables for each pairwise comparison (0 = no detection, 1 = positive detection); panel B shows a multiple correlation plot displaying Phi-coefficient values, with filled circles proportional to the magnitude of the Phi coefficient.