Comparison of Three Benthic Macroinvertebrate Passive Sampling Devices for Non-Wadeable Streams

Rivers and streams provide essential ecosystem services to the degree that the monitoring and maintenance of these systems becomes imperative. Biomonitoring provides managers and policymakers with the tools to make informed decisions, and macroinvertebrates are often the object of biomonitoring because they are ubiquitous in most systems and are known to be good indicators of water quality. However, methods for sampling macroinvertebrates in non-wadeable streams (i.e., large rivers) have not been standardized across states and regions and an established method for macroinvertebrate biomonitoring in large rivers of Georgia is not currently available. My study compared macroinvertebrates collected with three types of passive sampling devices to assess their suitability for sampling non-wadeable systems. Hester-Dendy samplers, mesh packs filled with swamp laurel oak (Quercus laurifolia) leaves (leaf samplers), and mesh packs filled with laurel oak sticks and twigs (wood samplers) were deployed at three sites on the Savannah River and three sites on the Ogeechee River for approximately 30 days during the fall of 2014. I examined mean, standard deviation, and variance components from 53 common bioassessment metrics and 2 multi-metric indices to identify differences in colonizing macroinvertebrates between the sampling devices. I estimated variance components using 2-way ANOVA to determine sources of variation (e.g., sites or devices). I further compared assemblages colonizing sampling devices using Permutational Multivariate Analysis (PERMANOVA) followed by Similarity Percentages (SIMPER) analysis. The abundance of true flies (Order: Diptera), the abundance of midges (Family: Chironomidae) and 9 of the 53 metrics (i.e., Diptera taxa, % Amphipoda, % Gastropoda, % Oligochaeta, % Dominant individuals, Dominant individuals, Collector taxa, % Predator, and % Burrower) were determined as significantly different between sampling devices based on a 2-way ANOVA. Macroinvertebrate assemblages colonizing the three sampling devices differed (PERMANOVA; F14,37=1.6078, P=0.001), and SIMPER results showed these differences were driven by the proportions of taxa collected by each device. Estimates of variance components attributed large percentage (i.e., \u3e20%) of variability to sites, rather than devices, with the exception of % Predator and Diptera taxa. My study suggests all three sampler types are suitable for collecting macroinvertebrate from non-wadeable systems and determining the precision and overall efficiency of sampling devices is an important step towards developing standard operating procedures for the bioassessment of large rivers