Recruitment Distributions Revisited

Abstract

Recruitment variability is a striking feature of the dynamics of many marine fish populations. Shelton (1992. Can J. Fish. Aquat. Sci. 49: 1754–1761) suggested that consideration of the survival probabilities of individuals or groups of individuals is the most appropriate strategy for modeling recruitment variability. A fundamental distinction can be made between environmental and demographic stochasticity. Here, environmental stochasticity refers to time-dependent variation in vital rates caused by environmental fluctuations. Demographic stochasticity is due to chance variation in the integer number of births and deaths in a population with time-invariant vital rates. The underlying basis for the simulations conducted by Shelton is consistent with the concept of demographic stochasticity. For large effective cohort size, these results converge to the deterministic case. Shelton's basic approach can be formalized using well-known results for a pure-death stochastic process. This derivation clarifies the importance of discrete processes for this class of models. Models of demographic stochasticity may be most useful for certain elasmobranch and marine mammal populations, while those incorporating environmental stochasticity may be more appropriate for many marine teleost and invertebrate populations. The critical distinction is between time-varying and time-invariant processes.