You are here:
Can we say: There is a <5% chance a new fish has invaded the St. Louis River? Evolving aquatic invasive species early detection
Citation:
Hoffman, J., A. Trebitz, Greg Peterson, J. Schloesser, H. Quinlan, AND J. Kelly. Can we say: There is a <5% chance a new fish has invaded the St. Louis River? Evolving aquatic invasive species early detection. Presented at Minnesota Chapter of the American Fisheries Society, Mankato, MN, March 03 - 05, 2014.
Impact/Purpose:
The Great Lakes Water Quality Agreement, Annex 6 calls for a U.S.-Canada, basin-wide aquatic invasive species early detection network by 2015. The objective of our research is to explore survey design strategies that can improve detection efficiency, and to develop performance metrics to evaluate survey success...The results indicate a sustained, long-term program is necessary to effectively interpret and report monitoring results.
Description:
The Great Lakes Water Quality Agreement, Annex 6 calls for a U.S.-Canada, basin-wide aquatic invasive species early detection network by 2015. The objective of our research is to explore survey design strategies that can improve detection efficiency, and to develop performance metrics to evaluate survey success. Since 2005, we have been testing elements of the network in the Duluth-Superior harbor, a Great Lakes non-native species introduction hotspot and a potential beachhead for invasion into interior lakes and rivers. Our program has detected 22 non-native aquatic invertebrate species and 10 non-native fish species in the harbor. We have used randomization approaches and rarefaction theory to estimate the effort required to detect 100% of the predicted species pool. Based on the randomization approach, there is a >95% probability of detecting the most rare non-native fish detected in about 100 samples. Established non-native species, however, are more common than those at an early stage of invasion. Based on rarefaction, to sample >95% of the species pool would require >110 samples and as many as 400 samples. In an examination of the 5-yr monitoring time-series, we conclude that the cumulative effort offers the potential to detect rare, newly introduced non-native fishes; however, the detection power is relatively low for any given year and annual data tend to present artificially high confidence compared to the longer time-series. The results indicate a sustained, long-term program is necessary to effectively interpret and report monitoring results.