7177 - Laserfiche WebLink

Home Browse Search

The distance moved during 15-minute intervals was analyzed with respect to flow and temperature. Colorado squawfish movement data were divided into two groups; one exhibiting movement and one in which no movement occurred. When these two groups were subjected to an Analysis of Variance (ANOVA), there was a significant difference (P <.Ol) between them with respect to temperature and flow (daily discharge). When subjected to regression analysis (linear) there was a weak positive correlation (r=0.1) between distance moved and flow (linear model valid at p <.01). The regression of distance and temperature also indicated a weak positive correlation (r=.006, linear model valid at p <.05). Although, we did not have curvelinear multiple regression capability, a multivariate linear regression of flow and temperature on depth produced a positiive correlation (r=0.17, linear model valid at p <.01). Colorado squawfish were observed engaging in various types of local movements during the study. These movements are best exemplified by fish No. 165 (Fig. 3). This fish selected two sites separated by about 3 miles in May and early June, 1980. The fish was found either at a large eddy produced by the junction of the Green and Duchesne rivers or at a river bend located 3 miles up the Duchesne River. It is assumed that the fish selected these two locations as feeding stations. Many other Colroado squawfish displayed similar movements to a variety of habitat types. Habitat Preferences General Colorado squawfish were observed predominantely in shoreline habitat and associated with sand substrates, although use of other habitats (eddy, run, backwater, pool) and substrates (silt, boulder, rubble, gravel) were also observed, depending upon river system and time of year. Because some investigators have noted that behavior is altered with the attachment or implantation of radiotransmitters (Stasko and Pincock, 1977) the habitat and substrate selected by radiotelemetered Colorado squawfish was compared with those recorded for Colorado squaw- fish collected by conventional gear (Figs. 12, 13). Differences in habitat and substrate use recorded by the two different techniques in the Green River and in the combined Green, Yampa and White rivers were slight. The radiotelemetry data is probably more representative since the location of radiotagged fish is more precise and 24-hour data can be obtained. Also, the types of habitat represented by broad shallow sand substrates in mid-river are difficult to sample by conventional gear since larger fish escape and it is believed the radiotelemetry data is more representative for these cases. Water depths and velocities recorded at the observed locations of radiotelemetered Colorado squawfish (Table 5) indicated fish selected depths and velocities averaging 1.4 m and 0.2 m/s in the Green River, 0.9 m and 0.1 m/s in the Yampa River, and 0.7 m and 0.5 m/s in the White River. When these data were analyzed by ANOVA, there was a significant 23