Incorporating Uncertainty into Demographic Modeling: Application to Shark Populations and Their Conservation

Abstract: I explored the effect of uncertainty in demographic traits on demographic analyses of sharks, an approach not used before for this taxon. I used age-structured life tables and Leslie matrices based on a prebreeding survey and a yearly time step applied only to females to model the demography of 41 populations from 38 species of sharks representing four orders and nine families. I used Monte Carlo simulation to reflect uncertainty in the estimates of demographic traits and to calculate population statistics and elasticities for these populations; I used correlation analysis to identify the demographic traits that explained most of the variation in population growth rates ( λ ). The populations I examined fell along a continuum of life-history characteristics that can be linked to elasticity patterns. Sharks characterized by early age at maturity, short lifespan, and large litter size had high λ values and short generation times, whereas sharks that mature late and have long lifespans and small litters have low λ values and long generation times. Sharks at the “fast” end of the spectrum tended to have comparable adult and juvenile survival elasticities, whereas sharks at the “slow” end of the continuum had high juvenile survival elasticity and low age–zero survival ( or fertility ) elasticity. Ratios of adult survival to fertility elasticities and juvenile survival to fertility elasticities suggest that many of the populations studied do not possess the biological attributes necessary to restore λ to its original level after moderate levels of exploitation. Elasticity analysis suggests that changes in juvenile survival would have the greatest effect on λ, and correlation analysis indicates that variation in juvenile survival, age at maturity, and reproduction account for most of the variation in λ. In general, combined results from elasticity and correlation analyses suggest that research, conservation, and management efforts should focus on these demographic traits.