9436 - Laserfiche WebLink

Home Browse Search

Investigators began at the top of each subreach and electrofished downstream at a constant rate with electrical current applied constantly to the water column. Investigators did not make judgements about locations that might hold Colorado pikeminnow and attempt to `sneak up' on fish. Backwaters, tributary mouths, and other habitat features were sampled once. When a rare fish was captured, sampling was stopped and the fish processed and released as near to the capture site as possible. Sampling began again and proceeded downstream until another rare fish was captured or the end of the subsection was reached. At the top of the next subreach, data sheets were completed and the elapsed time clock (recording shocking time in seconds) was reset before sampling began again. All Colorado pikeminnow encountered were captured when possible. Colorado pikeminnow that were positively identified (i.e. brought to the surface by the anode) but could not be captured were recorded as observed. All captured fish were checked for a PIT (passive integrated transponder) tag (Carlin tags prior to 1991), measured (total length, mm), and weighed (g). Unmarked fish were tagged before release. Data Analysis Capture data were analyzed as number of Colorado pikeminnow (CPM) collected per unit, of electrofishing effort (CPE). Initial analyses compared CPE calculated using both sampling time (elapsed time on the VVP converted to hours [CPM/hr]) and total miles electrofished (CPM/mi) to determine if the resulting trends differed. Initial analyses also compared number of Colorado pikeminnow captured (CPM) with number of Colorado pikeminnow collected + number observed (CPM+obs) to determine if any differences occurred (e.g. efficiency of experienced vs inexperienced sampling crews). One sample consisted of the electrofishing done along one shoreline of one subreach. CPE was calculated for each sample, and then mean CPE was calculated for each reach, river, and all rivers combined for every year of sampling. River-wide calculations (i.e. all sampling reaches combined) used only data from the original I3 sampling reaches to ensure consistency among years. Mean values calculated using the four values described above were plotted and compared using linear regression to verify that observed trends were consistent. River-wide CPE values were compared using two-factor, repeated measures analysis of variance (ANOVA) with reach as the repeated measure (subject factor), year the within factor, and river the between factor (Zar 1984). When differences were identified with ANOVA, pairwise comparisons were made using two different tests -Fisher's Least Significant Difference (LSD) Multiple- Comparison Test (considered a liberal test) and Tukey-Kramer Multiple-Comparison Test (conservative). River-wide trends in relative abundance were assessed with linear regression by comparing mean CPE over time (years). Annual mean CPE was compared to population estimates for the four different rivers with linear regression when estimates were available 5