8029 - Laserfiche WebLink

Home Browse Search

4 Methods and Results All experimental trials were conducted in a recirculating water system at the Klamath Tribe Native Fish Hatchery in Chiloquin, Oregon. Experimental fish tanks were light blue, rectangular, 68 L Rubbermaid tubs (45 cm long, 40 cm wide, and 41 cm tall). Water flowed into each tank at about 2 L f min, and flowed out via a centrally located standpipe which was screened to prevent escape of larval fish. Water temperature remained a fairly constant 17'C, which is within the range of ambient temperatures reported in shoreline areas of Upper Klamath Lake by Buettner and Scoppettone (1990). Trials one through five each ran for 14-18 hours, while trial six ran for 40 hours. Water depths were maintained at about 27 cm except in trial six when it was decreased to 15 cm in low water treatments. Sucker larvae used in the experiments were spawned from adults captured in the Sprague and Williamson Rivers in May, 1992. Fathead minnows were beach seined from Upper Klamath Lake near Modoc Point in early June, 1992. Fathead minnows were treated for a columnaris infection with 20 ppm nitrofurazone upon arrival at the hatchery. Many fish died during the seining process in the lake, indicating that the infection was well advanced. The nitrofurazone treatment successfully controlled the infection and all remaining fathead minnows were healthy when used in the experiments two weeks later. Trials one through five focused in part on assessing the influence of fathead minnow gape width (i.e. maximum mouth width at full gape) on their ability to prey on various sizes of sucker larvae. I measured gape widths to the nearest 0.1 mm on 50 randomly selected, live fathead minnows (48-68 mm TL) using a hand ruler and magnifying glass. Using the resulting frequency distribution, I stratified gape widths into three groups: small (2.0-2.3 mm), medium (2.5-2.8 mm), and large (3.1-3.8 mm). Fathead minnow gape width classes were then used as the three treatments in trials one through five. Tanks (experimental units) were randomly assigned to treatments (fathead minnow gape width class in trials one-five) until there were three replicate tanks per treatment. The percentage of larvae eaten in each tank was transformed to normality (aresine transformation) and used in one-way ANOVA tests to assess differences among treatment means. Statistical tests were performed using SYSTAT 5.0 (WdEnson 1990). Statistical power was determined post hoc when the null hypothesis was not rejected following methods of Cohen (1988). A severe outlier in trial 6 prompted use of a nonparametric two-way ANOVA (Zar 1984) followed by multiple comparisons (Conover 1980). 1 followed the admonitions and suggestions of Yoccoz (1991) concerning