Laserfiche WebLink
These correlations held for September samples, also. However, no correlation was evident for <br />~~ir"igrating sand waves when samples from June and July 1994 were excluded from analysis <br />~~ <br />(because they covered the entire 133.5 kilometer river reach) and only samples from the four, <br />~; <br />five-mile study sections were analyzed. Scour (secondary channel) backwaters were negatively <br />associated with duration of flows above 75% of the peak (r=-0.70). Although shoreline eddy <br />habitats were not significantly correlated with peak flow events, there was a highly significant <br />negative correlation (r'=--0.79) with flows during sampling periods. <br />Habitat Use <br />Chubs -- Chi-square analyses of chub use of backwater habitats by formation category <br />were not significant for any sampling period (Table 3). Chubs did not use any specific type of <br />t <br />backwater more or less than expected. Further analysis to determine if the distribution of <br />backwater types between the chub use and non-use classes differed, also failed to show <br />significance for any sampling period (Table 4). <br />Wilcoxon signed-rank analyses were conducted on individual habitat variables measured at <br />each backwater site. When pooled across sampling periods, significant differences (p<0.05) <br />existed between chub use and non-use backwaters for 23 individual variables. Seventeen of these <br />variables were associated with backwater depth. On average, chub use backwaters were <br />longer, wider, and deeper (greater area and volume) than unused backwaters (Table 5). <br />Differences did exist for some water temperature measurements, but levels of significance varied <br />greatly. Mean water temperature was higher in chub use backwaters, but this difference was not <br />13 <br />