Does spatial variation in environmental conditions affect recruitment? A study using a 3-D model of Peruvian anchovy

Abstract

We used a 3-dimensional individual-based model (3-D IBM) of Peruvian anchovy to examine how spatial variation in environmental conditions affects larval and juvenile growth and survival, and recruitment. Temperature, velocity, and phytoplankton and zooplankton concentrations generated from a coupled hydrodynamic Nutrients–Phytoplankton–Zooplankton–Detritus (NPZD) model, mapped to a three dimensional rectangular grid, were used to simulate anchovy populations. The IBM simulated individuals as they progressed from eggs to recruitment at 10cm. Eggs and yolk-sac larvae were followed hourly through the processes of development, mortality, and movement (advection), and larvae and juveniles were followed daily through the processes of growth, mortality, and movement (advection plus behavior). A bioenergetics model was used to grow larvae and juveniles. The NPZD model provided prey fields which influence both food consumption rate as well as behavior mediated movement with individuals going to grids cells having optimal growth conditions. We compared predicted recruitment for monthly cohorts for 1990 through 2004 between the full 3-D IBM and a point (0-D) model that used spatially-averaged environmental conditions. The 3-D and 0-D versions generated similar interannual patterns in monthly recruitment for 1991–2004, with the 3-D results yielding consistently higher survivorship. Both versions successfully captured the very poor recruitment during the 1997–1998 El Niño event. Higher recruitment in the 3-D simulations was due to higher survival during the larval stage resulting from individuals searching for more favorable temperatures that lead to faster growth rates. The strong effect of temperature was because both model versions provided saturating food conditions for larval and juvenile anchovies. We conclude with a discussion of how explicit treatment of spatial variation affected simulated recruitment, other examples of fisheries modeling analyses that have used a similar approach to assess the influence of spatial variation, and areas for further model development.