Looking for a Needle in a Haystack : Sampling Effort to Detect and Remove Asian Carps During Response Activities in the Laurentian Great Lakes Basin

Abstract

Early detection and response programs can prevent or delay species invasions. The potential to detect and remove species during response activities can be informed by the relationships between species abundance, sampling scheme, the probability of capture, and response effort. These relationships were investigated for Asian carps, which have been the subject of recent response efforts in the Great Lakes basin. Simulation models were used to examine the response effort required to detect and remove Asian carps in a fixed 75 ha area. The mean relative effort to detect a single fish ranged from 0.07 to 13.48 passes of the response area given: abundances from 1 to 25 fish (0.01 to 0.33 fish/ha), probabilities of capture from 0.05 to 0.70, and an assumed systematic sampling design. Response effort required for detection decreased in a nonlinear manner as fish abundance or the probability of capture increased. The mean relative effort required for local removal was 0.72 to 69.55 passes of the response area, and a similar nonlinear relationship was observed where small increases in the probability of capture above 0.05 led to substantial decreases in effort required for local removal. Completing five passes of the response area with non-detections (and an assumed low probability of capture, i.e., < 0.15) resulted in a moderate probability (p > 0.45) that fish remained within the response area. In addition, the probability of capture needed to be > 0.25 for low abundances (i.e., 1 fish) and > 0.55 for high abundances (i.e., 20 fish) to result in a high probability of local removal (p > 0.80) with five passes of the response area. Factors including fish aggregation behaviour, the size of the response area, sampling scheme (i.e., systematic, random, repeat, and informed sampling), and fish avoidance were also examined to determine their influence on effort required for detection and local removal. Improved knowledge of occupied fish habitat and resulting informed sampling had the greatest influence on effort required for detection and local removal under most probabilities of capture and fish abundances. These results provide a preliminary assessment of the potential effort required to successfully detect and remove Asian carps during response activities in the Great Lakes basin.