An evaluation of non-parametric species richness estimators for mussel assemblages in the Green River, Kentucky

Species richness is a primary concern in most ecological studies. Non-parametric estimators have been found effective at estimating species richness for many faunal groups. Freshwater mussel assemblages in the Green River have characteristics that make the ability to estimate the number of species present at a given location desirable. Some species once common in the Green River are now very rare and infrequently detected. These species may be on the brink of extirpation or even extinction which makes it vitally important to focus sampling on areas where they are most likely to occur. Mussel beds where sampled using a variety of sampling methods including random tossing of quadrats, systematic sampling with 3 random starts and qualitative sampling for up to 8 years at 13 sites in the Green River. Data from these events were analyzed using 7 non-parametric estimators to estimate species richness. The estimates produced were ranked by accuracy. A total of 45 species were detected within the upper Green River over the course of this study. The number of species detected per site ranged from 16 to 30. All estimators performed well, but overall, the second order jackknife was most accurate, and the bootstrap was least accurate. At almost all sites some species were detected in only one sampling event or exclusively in incidental collections. Additional species may be present at even lower numbers but remain undetected. Greater sampling effort than was used in this study may be necessary to detect all species present at a site. Discretion should be used before large sampling efforts using excavation are undertaken to reduce impacts to habitat and avoid stress to mussel. Some mussel species seem to be restricted to specific areas of the Green River between Lock and Dam 6 to the Green River Dam. The absence of these species in some areas probably leads to the total number of species found in this reach of river being higher than the number of species present at any one site. The second order jackknife provides a useful tool for targeting sites where undetected mussel species are most likely to be found. The ability to better target these sites can lead to a reduction in the amount of sampling effort needed to closely approximate the species richness of additional sites in the Green River and possibly other rivers as well.