Effort and potential efficiencies for aquatic non-native species early detection

Abstract

Our objective was to determine the effort required for high-probability early detection of non-native zooplankton, benthic invertebrates, and fish using Duluth–Superior Harbor — a Great Lakes port under intense non-native species introduction pressure — as a case study. Initially, we allocated samples using a spatially balanced random design. We then resampled the harbor, but allocated samples to a few targeted areas. We detected 21 non-native invertebrate and 10 non-native fish species; however, many rare zooplankton and benthic invertebrates were likely missed. The two designs did not have significantly different species accumulation curves, but the targeted area design samples had higher species richness and detected non-native species with a significantly higher probability. It was possible to reduce the effort required to detect established non-native species. In contrast, the effort required to detect an ultra-rare, newly arrived species remained large. Based on statistical estimation theory, the effort required to detect 95% or more of species present could exceed enumerating 750 zooplankton samples (~500 000 individuals, ~90 species), 150 benthic invertebrate samples (~100 000 individuals, ~250 species), and 100 fish samples (~75 000 individuals, ~40 species).

Comparison of fish detections, community diversity, and relative abundance using environmental DNA metabarcoding and traditional gears

“…Similar studies have found that, depending on the body of water, 60 to >200 samples from a single site (i.e., lake or embayment) during a sampling event with an optimized mixture of fisheries gears would be needed to sample most (>90%) of a fish community (Hoffman et al, 2011;Hoffman et al, 2015;Trebitz et al, 2009). We consistently found that the multilocus eDNA metabarcoding assays could detect 50% or 95% of a fish community in fewer samples, as compared to a traditional gear approach.…”

Sampling Design for Early Detection of Aquatic Invasive Species in Great Lakes Ports

“…We estimated average species-effort curves and their associated 95% confidence intervals and S asym values using the EstimateS v8.20 software (Colwell 2006). Whereas a survey encounters species in a sequence that depends on which sites are sampled first, the average increase in species with each additional sample taken can be obtained by randomizing site order (see Hoffman et al 2011 for technical details). We used sample-based abundance data to maintain the observed patchiness in the catch (Gotelli and Colwell 2001), combining the 2010-2012 data for each port to standardize effort but also combining data from all three ports to gain insight regarding how the different ports scale as a possible monitoring "network."…”

“…We began AIS surveillance at Duluth-Superior in 2006 (details in Trebitz et al 2009;Hoffman et al 2011;). We used a random stratified survey design to provide an unbiased selection of sites dispersed throughout the harbor (Stevens and Olsen 2004) with two depth strata based on the depth limit of aquatic vegetation establishment (2.1 m; Angradi et al 2013).…”

“…Recognizing that a substantial search effort is needed to find a rare species (e.g., at the early stage of introduction; Hoffman et al 2011) and that monitoring resources are generally constrained, improving efficiency (i.e., number of species detected for a given sampling effort) for early detection is paramount. Species detection (i.e., likelihood of encountering a species) can be increased through additional sampling effort, whether in a short-term intensive phase or spread over multiple years.…”

How Quickly Do Fish Communities Recover From Boat Electrofishing in Large Lakes?

“…Researchers have widely applied this for general fisheries assessments, but it may not adequately monitor trends at specific locations that might serve as habitat for rare or invasive species. Consequently, some monitoring programs have expanded their efforts to include both probabilistic (random) and nonprobabilistic (targeted) site sampling (e.g., Gutreuter 1993;Hoffman et al 2010).…”