Laserfiche WebLink
ISMP backwater sampling evaluation <br />A main premise of making comparisons between density estimates was that the high- <br />effort DMR sampling should provide relatively accurate and precise estimates relative to the <br />low-effort ISMP estimates. For the sake of estimating the bias of the ISMP sampling technique, <br />we consider the estimates derived from DMR sampling to be the "true" estimates (unbiased, no <br />variance) of presence and abundance of fishes in backwaters. We acknowledge that there is error <br />in such abundance estimates and quantify that variation with profile likelihood confidence limits, <br />but assume that the error was small relative to that for the two-seine haul ISMP technique. <br />Comparison of the proportion of ISMP density estimates that fell within the 95 % confidence <br />intervals of DMR estimates would thus provide a conservative measure of the bias of the ISMP <br />technique relative to DMR. Specific goals of comparisons between estimates of detection rates <br />and density derived from ISMP and DMR data were: <br />1) quantify the bias of the ISMP technique to detect presence of green sunfish, <br />largemouth bass, and three abundant non-native cyprinid species (fathead minnow <br />Pimephales promelas, sand shiner Notropis stramineus, red shiner Cyprinella <br />lutrensis) in each backwater, and <br />2) quantify the bias of the ISMP technique to estimate the density of two centrarchid <br />and three cyprinid species in each backwater. <br />If data from DMR sampling in individual backwaters was too sparse to compute abundance <br />estimates in CAPTURE, the total number of fish captured was used as an abundance estimate. <br />Thus, no confidence limits were calculated for those estimates. <br />-7-