THE AUTHORS 



DANNY C. LEE is a research biologist with the Inter- 

 mountain Research Station, Forestry Sciences Labora- 

 tory, Boise, ID. He received masters' degrees in 

 ecology from the University of Tennessee and applied 

 statistics from Louisiana State University, and a Ph.D. 

 degree in wildlife and fisheries sciences from Texas 

 A&M University. He joined the Forest Service in 1 991 , 

 and has been modeling the population dynamics of 

 salmonids since 1983. His current research focuses 

 on the dynamics of small, fragmented populations of 

 fishes in the Intermountain West. 



JEFFREY B. HYMAN is a fellow in the Quality of the 

 Environment Division at Resources for the Future, 

 Washington, DC. He received an M.S. degree in 

 statistics and a Ph.D. degree in ecology from the 

 University of Tennessee and an M.S. degree in 

 zoology from Washington State University. His re- 

 search interests are in the dynamics and conservation 

 of animal populations and in the effects of environmen- 

 tal heterogeneity. He has been involved in ecological 

 modeling research in the Columbia River Basin since 

 1990. 



RESEARCH SUMMARY 



The Stochastic Life Cycle Model (SLCM) simulates 

 the life cycle of anadromous salmonids and is designed 

 to mimic the basic mechanisms regulating populations 

 of Pacific salmon, while capturing some of the intra- 

 annual and interannual variation inherent in these 

 populations. This model was designed for population 

 viability assessments combining advanced modeling 

 techniques with concepts from the field of conservation 

 biology. 



While the basic structure of the SLCM is similar to 

 other life-cycle models, it differs in several ways. First, 

 the SLCM incorporates stochastic or probabilistic 

 processes at each step in the life cycle. The binomial 

 distribution is used extensively to introduce demo- 

 graphic stochasticity in survival; the beta distribution 

 is used to introduce environmental stochasticity. 

 Because of its stochastic nature, the model's predic- 

 tions must be expressed in probabilistic terms. Multiple 

 games are run using a Monte Carlo approach to 



generate probability distributions for future outcomes. 

 Second, the model is designed to use inputs from 

 more detailed models for specific life stages, in com- 

 bination with a minimum number of empirically based 

 parameters. SLCM users can choose among alterna- 

 tive models for the more contentious life stages, such 

 as juvenile migration and adult harvest, incorporating 

 the results of their preferred models. An ancillary 

 calibration model has been developed that allows the 

 SLCM to be fitted to a historical time trace of popula- 

 tion estimates, constraining expectations of survival 

 and their variances to historical levels. 



The model also allows considerable flexibility in 

 describing the dynamics of juvenile production. Users 

 can choose among three density-dependent relation- 

 ships to describe egg-to-smolt survival, including the 

 Beverton-Holt, the Ricker, and a logistic response 

 function, or use empirically based conditional probabili- 

 ties. A variety of scenarios involving hatchery and 

 natural production are possible, ranging from natural 

 production only, to a combination of hatchery and 

 natural production involving supplementation of adults, 

 fry, or smolts. Allocation of naturally produced and 

 hatchery-produced adults among terminal harvest and 

 hatchery and natural spawning follows a set of adjust- 

 able rules that affords priority to natural escapement 

 and hatchery broodstock needs. 



The model is written in the SAS® programming 

 language, which allows the model to operate on a 

 variety of computing systems and provides enhanced 

 flexibility in the analysis of model output. Users of the 

 model need not be proficient in SAS. Simple input 

 forms and ancillary programs to analyze model results 

 allow users to run the model with a minimum of prior 

 instruction. 



ACKNOWLEDGMENTS 



Funding for this research was provided by the U.S. 

 Department of Agriculture, Forest Service, Intermoun- 

 tain Research Station, and the U.S. Department of 

 Energy, Bonneville Power Administration, Division of 

 Fish and Wildlife. We thank John D. Mclntyre, Charles 

 Paulsen, and Daniel Goodman for their contributions 

 to the manuscript. 



The use of trade or firm names in this publication is for reader information and does not 

 imply endorsement by the U.S. Department of Agriculture of any product or service 



Intermountain Research Station 

 324 25th Street 

 Ogden, UT 84401 



