Laserfiche WebLink
(Osmundson ,and Burnham 1998; Nesler 2000; Osmundson 2002). Analyses were- done with <br />Number Cruncher software (Hintze 2401) with significance accepted at P <_ 0.05. Data from <br />the Colorado and Green rivers were also broken into lower and upper river segments <br />(Colorado River - upper =reach 6, lower =reaches 7 and 8; Green River -upper =reaches <br />1, 2, and 3, lower =reaches 4 and 5); differences among years were assessed using the same <br />analyses. <br />Size structure was described using length-frequency distributions based on 1-cm length <br />groups. Annual length-frequency distributions were calculated for each river and lower and <br />upper segments of the Green and Colorado Rivers. An average length-frequency distribution <br />was also calculated for each river to describe major differences in population structure among <br />rivers. Fish collected from all reaches (including alternating reaches) were included in length- <br />frequency distributions. Fish recaptured within the same year were used only once in <br />analyses. <br />RESULTS <br />Total number of Colorado pikeminnow captured during ISMP sampling was variable, but <br />the numbers increased substantially in later years (Table 2). Total fish captured in the 13 <br />original reaches ranged from 49 in 1988 to 314 in 2000. These totals include a few fish that <br />were handled twice in the same year. Three alternating reaches were sampled in 2000 which <br />increased the total number of fish captured in that year to 429. Over the 15-year study, the <br />largest number of fish was captured in the Green River, followed by the White, Colorado, and <br />Yampa rivers, respectively. <br />Calculation of CPE <br />Catch per effort was calculated four different ways for each river and sampling reach to <br />ensure that trends were consistent among the different parameters: 1) CPM/hr, 2) <br />CPM+obs/hr, 3) CPM/mi, and 4) CPM+obs/mi (Tables A-1-A-7). Regression ofriver-wide <br />CPE as a function of year resulted in significant positive relationships (P<_0.05) for all four <br />parameters in most cases (lower Colorado River and lower Green River were exceptions; <br />Table A-8). Analysis of variance for all values also produced significant differences in most <br />eases (upper Colorado River, lower Green River, and Yampa River were exceptions; Table A- <br />8). Most importantly, all CPE parameters were significantly related to each other in all four <br />rivers (Figure 2). Because the different CPE values all showed positive trends in each river <br />and CPM/hr is the most commonly used measure of CPE for electrofishing samples, it was <br />selected for the remaining analyses. <br />6 <br />