Laserfiche WebLink
weakly positive. Thus there was no evidence of increasing numbers of one of these three species <br />depressing numbers of Colorado squawfish within backwaters as would be expected if competition or <br />predation were occurring. The correlations probably reflect similar habitat use rather than lack of <br />competition or predation. Haines and Tyus (1990) also found no evidence that abundance of one <br />species in a backwater was associated with reduced numbers of another. <br />Recent laboratory studies have addressed the issues of competition and predation. Karp and Tyus <br />(1990a) found that red shiner and fathead minnow were more aggressive than small Colorado <br />squawfish in laboratory studies and postulated these species could adversely affect growth or survival <br />of YOY Colorado squawfish. Beyers et al. (in review) documented competition between Colorado <br />squawfish and fathead minnow in aquazia when limited food resources were present. Predation was <br />also demonstrated in the laboratory with controlled studies (Muth and Beyers in review). Rupert et <br />al. (1993) collected adult red shiners from the lower Yampa River that contained fish larvae in their <br />stomachs. Although they did not identify larval Colorado squawfish in the samples, the high <br />abundance of red shiner in the basin, sympatry with YOY Colorado squawfish, and their consumption <br />of fish larvae is cause for concern. Although negative effects aze not easily documented with ISMP <br />or similar field data, these laboratory studies suggest that introduced fishes are increasing mortality of <br />YOY Colorado squawfish. <br />Potential Modifications to ISMP <br />In 1989, investigators became concerned that mean CPE ,alone was not providing an accurate <br />index of relative abundance of YOY Colorado squawfish every fall. They felt estimates of relative <br />abundance could be improved by measuring the number and surface azea of backwaters in every <br />reach. Therefore aerial video was incorporated into ISMP on a trial basis to examine its potential for <br />measuring number and surface area of backwaters and improving the annual estimate of relative <br />abundance. Participants felt that a gross estimate of YOY Colorado squawfish numbers could be <br />obtained by multiplying mean CPE by total azea of backwaters. <br />McAda (1993) evaluated 3 yeazs of ISMP data collected in conjunction with aerial video <br />enumeration of backwater number and azea. He concluded that aerial video did not significantly <br />improve the annual estimate of relative abundance and did not justify the additional cost. He <br />recommended that collection of aerial video as part of YOY Colorado squawfish monitoring be <br />discontinued; therefore the video data aze not presented again. However, McAda (1993) also <br />suggested aerial video of short river reaches in conjunction with intensive ground truthing may still be <br />appropriate for specific studies-such as the nursery habitat investigation being conducted by UDWR. <br />However, there may be other ways that ISMP could be improved. ISMP data are highly vaziable, <br />which makes detecting important differences among yeazs and reaches difficult (McAda 1989a). <br />Unfortunately, increasing sample effort does relatively little to improve statistical reliability of data <br />analyses. McAda (1989a) concluded that the extra cost of collecting additional samples was not <br />justified by the small increase in the power of any statistical analyses that might be done. ISMP <br />already samples two backwaters per 5-mile subreach which is often the maximum number located in <br />many subreaches every yeaz and increasing the number of sampling sites might not be feasible <br />anyway. However, some backwaters aze quite lazge and deep. It might be appropriate to increase <br />the number of seine hauls taken in the lazge backwaters to increase the certainty of getting a reliable <br />sample. ISMP guidelines originally suggested that investigators sample at least 25% of the surface <br />area of all backwaters sampled. However, that requirement was dropped when participants felt it <br />difficult to accomplish (McAda 1989c). <br />23 <br />