8029 - Laserfiche WebLink

Home Browse Search

6 alternate prey. However, mean predation rates in the presence of alternate prey (5-43 %) were markedly lower than those in trial 4 (75-98%), which offered predators no alternate prey (Table 1). In trial six, three tanks were randomly assigned to each of four treatments (three more tanks were used than in trials one-five): high water with cover, high water without cover, low water with cover, and low water without cover. Water depth in 'high water tanks was 27 era, and 15 crn in low water' tanks. Cover provided by emergent vegetation in natural shoreline habitats was simulated in `with covet' treatments. Scirpus stems were collected from Upper Klamath Lake, cut to approximately 60 cm, and wedged into holes in 1/4" hardware cloth to achieve a stem density of approximately 565 stems/m2 (stems were spaced approximately 4 cm apart). These assemblies were lowered into the appropriate tanks with the Scirpus stems hanging down from the hardware cloth so that half of a tank bad Scirpus stems in it and the other half was open water. Scirpus stems were held in place on the tank bottom by sheets of black, 1/2" mesh plastic netting. Each tank was stocked with 20 randomly selected shortnose sucker larvae (14-21 mm standard length), followed by 10 randomly selected fathead minnows without regard to gape width. Trial six ran for 40 hours. A severe outlier (only 2 larvae eaten in a tank in the high water/cover absent treatment) prompted use of a nonparametric two factor ANOVA, performed on ranked percentages of larvae eaten per tank. Null hypotheses of no effect of water depth and cover presence/absence were not rejected (water depth P>0.75; cover 0.10<P<0.25); however, a significant interaction of water depth with cover presence/absence was detected (0.05<P<0.10). Subsequent nonparametric multiple comparisons showed significant differences (P<0.10) in predation rates between low water/cover present and low water/cover absent treatments, while neither of these differed significantly from the high water treatments (Table 2). While power was not estimated for these tests, it was undoubtedly low. As a supplement to the nonparametric analysis, the severe outlier (10% mortality in the high water/cover absent factor combination) was excluded, and a two factor, parametric ANOVA was performed. Mean predation rates still did not differ significantly relative to water depth in tanks (P>0.15), but were significantly lower when cover was present (P<0.06). In addition, there was no significant interaction between water depth and cover (P>0.22).