9703 - Laserfiche WebLink

Home Browse Search

razorback suckers were added, and observation resumed. This experiment ran for 13 days. The third trial was similar to the previous test; the same two flathead catfish were used and 10 new razorback suckers were added. However, the sanctuary zone was reduced to one-fourth of the tank's area (fig. 5) For four days, we periodically recorded where fish were observed in the tank. Figure 5. Tank configuration lone-fourth sanctuary) used for trial 3. The divider was placed along the center of the tank, dividing it into equal halves for trials 1 and 2. Statistical Analysis Razorback Sucker Predation Trials Fish treatment varied between years: in 2006, treated fish were exercised only; in 2007, treated fish were exercised and exposed to a predation event. We were interested if treated fish were preyed upon less than control fish (Ha' predation was random and mortality of treatment and control fish was equal; HA, predation was not random and mortality of treatment fish was less than mortality of c011lrol fish). The number of predation trials in both years had relatively small sample sizes and values were often small and discrete (for example, number of fish eaten was <5 in many trials). Therefore, we used nonparametric tests to compare mortality of razorback suckers: either the Wilcoxon Matched Pairs Ranks test or Multi-Response Permutation Procedure (MRPP) Matched Pairs test. Significance was determined at the 0.05 level for all tests on predation. Bonytail Predation Trials We examined predation of bony tail in small arena experiments using groups of two to three bony tail per treatment. A trial ended as soon as a predation event was observed, and we deterrnJl1cJ II the consumed fish was a treatment or control. We examined the binomial proportion of trial success using a Goodness of Fit test (HO, predation was random and ratio of trials was I: 1 ; HA, predation was not random and treatment fish succeeded in more trials than control fish). 8