Author ORCID Identifier
Invasive species, Early detection, Risk assessment, Taxonomy, FlowCAM, Great Lakes, Hamilton Harbour
Many zooplankton surveys underestimate species richness owing to difficulties in detecting rare species. This problem is particularly acute for studies designed to detect non-indigenous species (NIS) when their abundance is low. Our goal was to test the difference in detection efficiency between traditional microscopy and image analysis (i.e., FlowCAM). We hypothesized that detection of rare species should become easier as they become abundant in a sample, if they are morphologically distinct, or if counting effort increased. We spiked different densities of Cladocera into zooplankton samples from Lake Ontario to simulate rarity, and assessed detection rate. Our results indicated that there was a positive relationship between the probability of finding at least one spiked NIS and its abundance, distinctiveness, and counting effort employed. FlowCAM processed more subsamples, though morphologically similar taxa were distinguished more readily with microscopy. The expected probability for detecting one individual spiked into a sample with ~ 8000 individuals (300 counted) was 3.60%, though observed values were considerably lower using both classical microscopy (4.58 × 10−3 to 1.00%) and FlowCAM (0.10 to 3.00%). Our experiments highlight that many plankton ecologists use subsample counts too low to detect rare native species and NIS, resulting in low species richness estimates and false negatives.
Stanislawczyk, Keara; Johansson, Mattias L.; and MacIsaac, Hugh J., "Microscopy versus automated imaging flow cytometry for detecting and identifying rare zooplankton" (2018). Hydrobiologia, 807, 1, 53-65.
Available for download on Sunday, January 01, 2119