The unintentional transport of invasive species through the global shipping network causes substantial losses to social and economic welfare. Addressing this global challenge requires a non-indigenous species (NIS) risk assessment tool for identifying species of concern and the likely sources of these species. In this research we developed and employed novel methods in network analysis, data fusion, and environmental DNA (eDNA) analysis of water from commercial ports to forecast NIS risk under future scenarios. These methods integrated interdisciplinary models of navigation infrastructure, global trade, ecoregions, and future climate change into a network of ship-borne species flow. To test these predictions, we first evaluated the ability different eDNA metabarcoding primers (COI and 18S) to estimate taxonomic similarity of metazoans (multicellular animals) between ports, and then correlated these estimates with between-port shipping risk predicted from the models.
Results/Conclusions
Network modeling results suggest several large-scale species flow patterns, including clusters of ports that experience high ship traffic among themselves, as well as important nodes that connect clusters. This led to the identification of a relatively few number of ports which contribute much of the species flow around the world (e.g. Singapore alone contributes 26% of total predicted between-cluster flow from/to the Pacific cluster which contains 818 ports). To test model predictions, we first evaluated the use of two primers, COI and 18S, in metabarcoding of commercial port eDNA samples. Preliminary data suggest that eDNA metabarcoding can both detect known species as well as unique, but unidentified, operational taxonomic units (OTUs). However, we found that detection of known species and unknown OTUs varied by primer (COI and 18S), a finding which has implications for using eDNA metabarcoding for NIS early detection. Taxonomic similarities between ports estimated from these eDNA metabarcoding data are currently being used to test model predictions of past and future ship-borne NIS spread. Through iterative engagement with decision-makers, our work informs global and US policies and management practices that improve the sustainability of coastal ecosystem services.