LAUSR

Abstract In fisheries ecology and stock assessment, recruitment signifies the transition from early stages of the life cycle which are characterized by environmentally-driven variability and density-dependence in mortality rates, to a recruited phase when natural mortality is largely stable and density-independent. Age or size at recruitment is an important structural element of age-structured fisheries models and stock assessment methods, yet the choice is often somewhat arbitrary and little guidance exists for when, biologically, a fish can be considered recruited. Such models may be mis-specified when recruitment is assumed to occur too early, i.e. when ‘recruited’ fish are still subject to density-dependent mortality even though the model specifies that they are not. We synthesize empirical data on density-dependence in mortality at different life stages within wild, stocked, cultured and experimentally manipulated fish populations. Results suggest the existence of a broad, approximately invariant pattern of density dependence across a wide range of fish life histories. Density-dependent mortality is most likely to occur, and potentially strongest, while fish are smaller than 10% of population asymptotic lengthy L∞. Both the likelihood of encountering density-dependent mortality and its potential strength are far more moderate for fish between 10 and 20% of L∞, and diminish in fish larger than 20% of L∞. This result provides the first empirically-based guidance for determining an appropriate size or age at recruitment for stock assessments. The result also provides information relevant to other fisheries management issues, such as the design and evaluation of fish stocking programs.