Laserfiche WebLink
classified sample sites were migrating sand waves. The second most common ba <br />was secondary (scour) channels. These accounted for 49% (329 of 668) of class <br />sampled. The remaining types of backwaters occurred as follows: shoreline edd <br />vortex, 3%; isolated pool, 1%; and low velocity flow, 0.5%. <br />Four flow parameters (peak flow {cfs}, date {Julian} of peak flow, flow <br />at 50% and 75% of peak) were regressed against backwater numbers, volume (Figu <br />(Figure 4). The number of backwaters present in any given season was not rela <br />parameters from the previous spring. Significant negative linear relations di <br />between at least one flow parameter and mean backwater area and total backwate <br />each season. Additional negative relations (p<0.10) existed for mean backwate <br />spring and summer and for total backwater area in spring. Of 48 regression fu <br />significant at p<0.05 and four wexe significant at p<0.10 (Table 3). All rema <br />explained from 66% (R2=0.66) to 91% (R2=0.91) of the variability between years <br />regression relation for mean backwater area in spring and fall could be furthe <br />adding additional variables to the model. A three variable model of mean back <br />spring increased R2 to 0.99. Similarly, R2 increased to 0.97 for a two varia <br />backwater area in the fall (Table 3). <br />Colorado 5auawfish Habitat Use <br />Colorado squawfish were captured in 308 (29%) sampled backwaters. Tncid <br />Colorado squawfish presence in samples was greatest for summer and fall sampli <br />