Laserfiche WebLink
<br />species (using presence/absence), abundance was negatively <br />correlated. Abundance of the two juvenile catostomid species <br />showed the highest number of positive correlations with <br />nonnatives. This included positive correlations between adult <br /> <br />and juvenile E. promelas and juvenile Q. lutrensis. Juvenile Q. <br /> <br />discobolus also had a significant positive correlation with adult <br /> <br />and larval C. lutrensis. Juvenile C. latipinnis abundance was <br /> <br />negatively correlated with E. zebrinus as was the abundance of <br /> <br />juvenile R. osculus with Q. affinis and larval C. lutrensis. The <br /> <br />only significant correlation of an adult/sub-adult native species <br /> <br />with a nonnative was between Q. latipinnis and juvenile Q. <br /> <br />lutrensis. <br /> <br />DISCUSSION <br />Recent attempts by Power et al. (1995) to model the food-web <br />dynamics of large rivers recognized the need to investigate the <br />role of off-river water bodies as rearing areas for juveniles. <br />They also indicated the need to examine the relative importance <br />of seasonal life-history bottlenecks. In this study we examined <br />both factors. Secondary channels were shown to provide habitats <br />used by many larvae and juveniles of both native and nonnative <br />fish. There is, however, potentia1ly a juvenile competitive <br />bottleneck (Persson 1986, Persson and Greenberg 1990), where <br />resources for adult size classes may not be limited but <br />interspecific competition during the juvenile stage may limit <br />recruitment. This would seem to have the greatest effect during <br /> <br />14 <br />