Consequences to Juvenile Sockeye Salmon and the Zooplankton Community Resulting from Intense Predation

Abstract

Stocking of juvenile sockeye salmon Oncorhynchus nerka into nonanadromous lakes at levels comparable with juvenile densities in anadromous lakes supporting natural sockeye salmon populations effected up to a 90% reduction in zooplankton biomass. The zooplankton communities subsequently became resistant to predation as the vulnerable Daphnia, Diaptomus, and ovigerous Cyclops were virtually eliminated and the more agile (nonovigerous) Cyclops and smaller Bosmina persevered and became predominant. Relying on a standing crop of zooplankton that was severely depressed by intense grazing the previous season, juvenile sockeye salmon experienced at least a 3-fold reduction in numbers and biomass between fry and smolt stages from the previous year. Our experimental results confirmed our empirical findings that, in rearing-limited lakes, smolt biomass production becomes a function of zooplankton biomass. Once restructured by excessive predation, some zooplankton communities were unresponsive or slow to respond to either reduced grazing pressure and/or to subsequent nutrient treatment. This delayed recovery of overgrazed zooplankton populations reduced growth and survival (rearing efficiency) for ensuing cohorts. In response, Frazer Lake sockeye salmon populations fell far below replacement, the dominant-year run segment collapsed, and the amplitude of high versus low return per spawner ratios increased. In contrast, less damaged zooplankter populations responded to nutrient treatments, leading to a 3-fold increase in fry-to-smolt survival and a 6to 20-fold increase in smolt biomass. The degree to which juvenile sockeye salmon foraging decreases biomass levels and changes the species composition of the zooplankton community ultimately determines the duration of zooplankton recovery and lowered sockeye salmon production. Authors: J. P. KOENINGS is a special assistant to the commissioner of the Alaska Department of Fish and Game, P.O. Box 25526, Juneau, AK 99802-5526. G. B. KYLE was a limnologist (now retired) with the Alaska Department of Fish and Game, Commercial Fisheries Management and Development Division, 34828 Kalifornsky Beach Road, Suite B, Soldotna, AK 99669-8367. Acknowledgments: Many fishery technicians and biologists — conducted field activities and collected data. The Limnology Laboratory in Soldotna — analyzed water and zooplankton samples. Sponsorship: These lake investigations were funded by the Kodiak and Cook Inlet Aquaculture Associations, Lower Cook Inlet Seiners Association, and State of Alaska. INTRODUCTION Spawner-recruit relationships (Ricker 1954) and cyclical dominance (Welch and Noakes 1990) affect cohort abundance of sockeye salmon Oncorhynchus nerka, but both processes assume zooplankton populations recover immediately from predation and for the most part are density-independent. Unfortunately, a vast majority of curves fit to spawner-recruit data are not statistically significant (Dahlberg 1973; Geiger and Koenings 1991). One reason for this might be that escapement/cohort effects on subsequent zooplankton populations are not independent but may carryover to subsequent years. These effects are included in some simulation models that use brood year interactions to forecast cyclic changes in sockeye salmon run strength in Alaska (Eggers and Rogers 1987) and Canada (Walters and Staley 1987), but for the most part, these effects have not been considered. Even though freshwater density-dependent effects have been clearly demonstrated in Alaskan (Koenings and Burkett 1987b; Kyle et al. 1988), Russian (Bugaev 1989), and Canadian (McDonald et al. 1987) sockeye salmon populations, biological mechanisms responsible for the negative interaction between cohorts are not well defined. Koenings and Burkett (1987b) proposed that a lake’s sockeye salmon rearing capacity is the fry density that produces smolts of minimal or threshold size (about 60 mm or 2 g). Loading densities above this capacity are excessive (Geiger and Koenings 1991) and may invoke changes (e.g., overgrazing) in rearing capacities for subsequent brood years. In fact, successive escapements 2–3 times above the rearing capacity in Frazer Lake (Kyle et al. 1988) caused the collapse of a dominant-year run, and subsequent brood year return per spawner ratios fell below replacement levels. Such top-down effects, if related to overgrazing 121 Predation Consequences for Juvenile Sockeye Salmon and Zooplankton • Koenings and Kyle the forage base, are reversible by nutrient treatment (Hyatt and Stockner 1985; Koenings and Burkett 1987b; Kyle 1994; Kyle et al. 1997). Top-down control by rearing sockeye salmon reduces the size of prey items, lowers zooplankton fecundity and density, displaces vulnerable prey species, and thereby restructures the zooplankton into a predator-resistant community (Koenings and Burkett 1987b; Kyle et al. 1988). Once established, such an assemblage may resist immediate reversal to bottom-up (producer) control, either through decreased predation or increased primary production. Based on observations from whole-lake manipulation experiments (lake stocking and nutrient treatment) on 4 lakes, we examine whether intense predation by high sockeye fry densities are independent and reversible. This information is useful to identify bottlenecks for enhancement (Koenings and Burkett 1987b) and has significant implications to management (Koenings and Burkett 1987a; Geiger and Koenings 1991).