D. Mcdermot et Ka. Rose, An individual-based model of lake fish communities: application to piscivore stocking in Lake Mendota, ECOL MODEL, 125(1), 2000, pp. 67-102
A general individual-based fish community simulator is presented. The model
tracks the daily feeding, growth, movement, reproduction, and mortality of
individuals of up to six species for multiple generations in up to three s
patial boxes. The version presented in this paper has been configured and c
alibrated for Lake Mendota, Wisconsin. Six species are followed: yellow per
ch, bluegill, white bass, cisco, walleye, and northern pike. The environmen
t consists of daily temperature, dissolved oxygen, and prey densities in ea
ch of the epilimnion, hypolimnion, and littoral zone spatial boxes. Feeding
parameters and larval mortality rates were calibrated until all species pe
rsisted at reasonable biomasses, with realistic mean lengths at age and die
ts by life stage. Two alternate baseline calibrations are presented which d
iffer in their degree of interannual variability. Correlation analysis of s
urvival and growth rates showed that larvae were influenced by competition,
yearlings by predation, and young of the year (YOY) juveniles by both. Adu
lt growth was density-dependent for planktivores and positively related to
forage biomass for piscivores. Cisco dynamics were effectively independent
of the other species. The calibrated model was used to compare the effects
of a biomanipulation experiment (piscivore enhancement) versus a coincident
fish die-off event, and to evaluate alternative stocking regimes for their
ability to sustain improved water duality. Predicted total zooplankton con
sumption was used to indicate effects on algae and water quality. Cisco die
-off produced similar short-term but much larger long-term reductions in zo
oplankton consumption than piscivore enhancement. Delayed changes in YOY ju
venile and larval survival illustrated complex indirect food web responses
to piscivore enhancement. None of the three alternative stocking regimes an
alyzed yielded ideal management results. Stocking either had little effect
on zooplankton consumption, or resulted in significant changes in piscivore
mean lengths or in the composition of the fish community. (C) 2000 Elsevie
r Science B.V. All rights reserved.