7646 - Laserfiche WebLink

Home Browse Search

r. , salmonids from hatcheries, barges, and bypass fa- cilities. Methods Fish collection and maintenance.-Northern squawfish (mean fork length, FL, 432.6 mm; SO, 36.1; N = 70) were collected from the Columbia River by boat electroshocking and acclimated for at least 3 weeks in indoor 1,400-L circular tanks at temperatures approximating those in the river (l5-170C). They were fed a maintenance diet of live juvenile coho salmon Oncorhynchus kisutch (mean FL, 103.8 mm; SO, 7.2; N = 50) during acclimation and between experiments. The north- ern squawfish were starved for approximately 2 d prior to an experiment. Juvenile chinook salmon O. tshawytscha (mean FL, 90.6 mm; SO, 11.8; N = 100) were used as prey for experiments. Coho and chinook salmon were obtained from the Little White Salmon Na- tional Fish Hatchery complex and held in out- door, 1,400-L circular banks. Tanks received flow- through well water at 6-8"C; juvenile salmonids were not maintained for long periods at higher experimental temperatures because their health deteriorated. Most salmon were acclimated to ex- perimental temperatures for 3 d before a trial be- gan. Due to space constraints, prey were accli- mated to higher temperatures for I h prior to the start of one set of trials (circular tanks with 50% dead prey). Prey to be offered dead were killed by blows to the head shortly before an experiment. Dead fish were given adipose fin clips to distin- guish them from live prey that might die during a trial. Experimental systems. -Experiments were conducted in two indoor fiberglass tank systems. The first system was a series of 1,400-L circular tanks (1.5 m in diameter, 0.8 m deep). A recir- culating-water system supplied each tank with 20 L of filtered well water per minute (Lucchetti and Gray 1988). Temperature was maintained at 15- 170C. A natural photoperiod (15 h light: 9 h dark- ness) was simulated; light was provided by a series of fluorescent and incandescent lamps. We also conducted experiments in an 11,300- L, flowing-water raceway (7.6 m long x 1.2 m wide x 1.2 m deep). Temperature was maintained at 16-180C. Five incandescent lights simulated ambient photoperiod (IS h light: 9 h darkness). A centrifugal pump (1,700 Urn in) recirculated water. Because northern squawfish in the field are typically found in areas with flows less than 70 cm/s (Faler et a!. 1988), flow in the raceway was NOTES 681 kept at a low velocity. Water velocity varied spa- tially, ranging from about 73 cm/s in midwater upstream to 3 cm/s near the bottom downstream. Proportions of available dead prey. - Three northern squawfish per tank were used for each trial in the circular tanks. Sixty chinook salmon (live plus dead fish) were placed into each tank in the morning; 24 h later, uneaten prey were col- lected and counted. Eighteen trials were conduct- ed with 50% dead juvenile chinook salmon (30 dead and 30 live) and 12 trials with 20% dead chinook salmon (12 dead and 48 live). Prey num- ber was based on maximum daily consumption rates presented by Vigg and Burley (1991). Ten northern squawfish were used for each race- way trial after a minimum IO-d acclimation pe- riod. Trials were conducted during 3-h evening intervals (2000-2300 hours), the first 0.5 h of which was lighted and the remainder of which was in darkness. At the beginning ofa trial, 100 live and dead chinook salmon were placed simultaneously into the raceway at the upstream end. Uneaten prey were removed and counted at the end of 3 h. Three trials were conducted for each percentage of dead prey (50% and 20%). Predation during light and dark periods. - Ex- periments were conducted in circular tanks to de- termine if predation by northern squawfish dif- fered between light (215-270 lux measured 10 cm below the water's surface) and dark (<0.0 I lux) periods. Three northern squawfish per tank were used for each trial. Twenty-two lighted trials were conducted at 0900-1300 hours, and 27 dark trials were conducted at 1800-2200 hours. Three dead (20%) and 12 live juvenile chinook salmon were placed into each tank. Low prey numbers allowed us to visually count live and dead salmon re- maining in a tank without removing them; this was done at 0.5 hand 2 h after the start of II light trials to determine predation rate. Data analysis. - Trials in which more than 50% of the prey were consumed were not included in analyses because if one prey type is selected, the relative abundance ofthe preferred prey decreases as the experiment progresses (Coutant 1973). Each trial was considered a replicate within each of the six experimental conditions (Table I). Ifadequate numbers of prey were consumed per trial, chi- square analysis was used to determine if replicates were homogeneous (Sokal and Rohlf 1981; Zar 1984). Ifper-trial consumption was too low for an unbiased chi-square analysis (Zar 1984), trial re- sults for an experimental type were pooled for analysis. Expected ratios of dead and live prey