An evaluation of five aquatic sampling methods useable in an Appalachian mountain stream, Great Smoky Mountains National Park, Sevier County, Tennessee

Five aquatic sampling methods were examined and evaluated for potential use in a future quantitative study in Mannis Branch, Great Smoky Mountains National Park, Sevier County, Tennessee. The methods examined were rock, kick, artificial substrate, leaf pack, and drift sampling. A primary focus of the study was to examine the variability among samples and sample units and to explore analytical and data manipulation tools for preparing sample data for analyses. Total counts and in many cases counts of functional feeding guilds (e.g., scrapers, collectors, predators, and shredders) were analyzed. Samples were tested for normality and equal variances prior to statistical significance tests and appropriate transformations applied to non-normal and heteroscedastic sample counts. Two transformations were routinely used and compared where sample data required manipulation, log (x + 1) and Box-Cox power transformation. Primarily parametric (e.g., one-way ANOVA) and, secondarily, nonparametric analyses (e.g., Kruskal-Wallis rank sums test) were performed to provide preliminary assessment of the variability and differences among sampling sites. Where one-way ANOVA and Kruskal-Wallis tests detected significant differences, multiple comparisons tests typically detected one or more sites downstream of a natural barrier (i.e., a waterfall) as having higher benthic macroinvertebrate counts relative to other sampling sites. An assessment is made for each method's continued use for quantitative purposes in Mannis Branch and is based on required sample sizes, time and effort involved in collecting and processing samples, and effectiveness for sampling the benthic fauna.