Although environmental DNA shed from an organism is now widely used for species detection in a wide variety of contexts, mobilizing environmental DNA for management requires estimation of population size and trends in addition to assessing presence or absence. However, the efficacy of environmental‐DNA‐based indices of abundance for long‐term population monitoring have not yet been assessed. Here we report on the relationship between six years of mark‐recapture population estimates for eulachon (Thaleichthys pacificus) and “eDNA rates” which are calculated from the product of stream flow and DNA concentration. Eulachon are a culturally and biologically important anadromous fish that have significantly declined in the southern part of their range but were historically rendered into oil and traded. Both the peak eDNA rate and the area under the curve of the daily eDNA rate were highly predictive of the mark‐recapture population estimate, explaining 84.96% and 92.53% of the deviance, respectively. Even in the absence of flow correction, the peak of the daily eDNA concentration explained an astonishing 89.53% while the area under the curve explained 90.74% of the deviance. These results support the use of eDNA to monitor eulachon population trends and represent a >80% cost savings over mark‐recapture, which could be further increased with automated water sampling, reduced replication, and focused temporal sampling. Due to its logistical ease and affordability, eDNA sampling can facilitate monitoring a larger number of rivers and in remote locations where mark‐recapture is infeasible.

The spread of nonindigenous species by shipping is a large and growing global problem that harms coastal ecosystems and economies and may blur coastal biogeographical patterns. This study coupled eukaryotic environmental DNA (eDNA) metabarcoding with dissimilarity regression to test the hypothesis that ship‐borne species spread homogenizes port communities. We first collected and metabarcoded water samples from ports in Europe, Asia, Australia and the Americas. We then calculated community dissimilarities between port pairs and tested for effects of environmental dissimilarity, biogeographical region and four alternative measures of ship‐borne species transport risk. We predicted that higher shipping between ports would decrease community dissimilarity, that the effect of shipping would be small compared to that of environment dissimilarity and shared biogeography, and that more complex shipping risk metrics (which account for ballast water and stepping‐stone spread) would perform better. Consistent with our hypotheses, community dissimilarities increased significantly with environmental dissimilarity and, to a lesser extent, decreased with ship‐borne species transport risks, particularly if the ports had similar environments and stepping‐stone risks were considered. Unexpectedly, we found no clear effect of shared biogeography, and that risk metrics incorporating estimates of ballast discharge did not offer more explanatory power than simpler traffic‐based risks. Overall, we found that shipping homogenizes eukaryotic communities between ports in predictable ways, which could inform improvements in invasive species policy and management. We demonstrated the usefulness of eDNA metabarcoding and dissimilarity regression for disentangling the drivers of large‐scale biodiversity patterns. We conclude by outlining logistical considerations and recommendations for future studies using this approach.

Running head: indigenous eDNA monitoring of eulachon 21 22 23Although environmental DNA shed from an organism is now widely used for species detection 24 in a wide variety of contexts, mobilizing environmental DNA for management requires 25 estimation of population size and trends rather than simply assessing presence or absence. 26However, the efficacy of environmental-DNA-based indices of abundance for long-term 27 population monitoring have not yet been assessed. Here we report on the relationship between 28 six years of mark-recapture population estimates for eulachon (Thaleichthys pacificus) and 29 'eDNA rates,' which are calculated from the product of stream flow and DNA concentration. 30Eulachon are a culturally and biologically important anadromous fish that have significantly 31 declined in the southern part of their range but were historically rendered into oil and traded. 32Both the peak eDNA rate and the area under the curve of the daily eDNA rate were highly 33 predictive of the mark-recapture population estimate, explaining 84.96% and 92.53% of the 34 deviance respectively. Even in the absence of flow correction, the peak of the daily eDNA 35 concentration explained an astonishing 89.53% while the area under the curve explained 90.74% 36 of the deviance. These results support the use of eDNA to monitor eulachon population trends 37 and represent a >80% cost savings over mark-recapture, which could be further increased with 38 automated water sampling, reduced replication, and focused temporal sampling. Due to its 39 logistical ease and affordability, eDNA sampling can facilitate monitoring a larger number of 40 rivers and in remote locations where mark-recapture is infeasible. 41 42

Spread of nonindigenous organisms by shipping is one of the largest threats to coastal ecosystems. Limited monitoring and understanding of this phenomenon currently hinder development of effective prevention policies. Surveying ports in North America, South America, Europe, Southeast Asia, and Australia we explored environmental DNA community profiles evident of ship-born species spread. We found that community similarities between ports increased with the number of ship voyages, particularly if the ports had similar environments, and when indirect stepping-stone connections were considered. We also found 57 known non-indigenous taxa, some in hitherto unreported locations. We demonstrate the usefulness of eDNA-based tools for global biodiversity surveys, and highlight that shipping homogenizes biodiversity in predictable that could inform policy and management.