Early Detection of Invasive Species in Vulnerable Harbors and Embayments of the Great Lakes
Jack Kelly, Greg Peterson, Corlis West, Joel Hoffman and Anett Trebitz
U.S. EPA, Mid-Continent Ecology Division, NHEERL/ORD, Duluth, Minnesota
Our goal is to define a model, cost-effective approach for detecting invasive species in vulnerable coastal systems. A case study of the Duluth-Superior Harbor/St. Louis River on Lake Superior has provided a number of “first records” of invasive species, including the New Zealand Mud snail and the Quagga mussel. The study has confirmed the presence of nearly two dozen invaders whose first detection was within the last ~20 years. The distribution of fish, benthic invertebrates, and zooplankton within this system (>48 km 2) have been determined using a spatially-balanced probability (EMAP) design and by directed sampling. A limited set of directed sampling zones were chosen by GIS spatial analyses of potential vectors of introduction, habitat attributes, and an invasive “vulnerability” index. We are using our data to compare different spatial designs and sampling styles/gear in order to develop “efficient and best” strategies for monitoring. Detecting non-indigenous species at an early stage of invasion is akin to searching for rare species — this is inherently an inefficient process, although it can be successful with massive effort. Species-area (or species-effort) curves are one analytical tool that we are using to assess detection efficiency. Results show substantial differences in the rates of species detection among fish sampling gears. Preliminary results suggest that directed sampling may be more effective than a probabilistic design, given equal sampling effort. This abstract does not necessarily reflect EPA policy.