Environmental influences on the interannual variation and spatial distribution of Peruvian anchovy (Engraulis ringens) population dynamics from 1991 to 2007: A three-dimensional modeling study

Abstract

A hydrodynamic-biogeochemical model was coupled to an individual-based fish model (IBM) to study the influence of physical and biological processes on Peruvian anchovy recruitment. Temperature and the concentrations of four plankton groups from a 1991 to 2007 simulation of a Pacific basin-scaled Regional Ocean Model System (ROMS) coupled with the Carbon Silicate Nitrogen Ecosystem (CoSiNE) model were used as inputs to the anchovy IBM. The anchovy IBM domain covered the upwelling area (0–20°S and 70–85°W) from 0 to 100m depth, and was 166 by 120 cells with 10 vertical. A cohort of eggs was started each month from 1991 to 2007, and individuals within each cohort followed through daily development, growth, mortality, and movement for one year. Growth was represented with a bioenergetic equation that used temperature and plankton concentrations from the ROMS–CoSiNE simulation as input. Mortality rate was stage-dependent and length-dependent. Movement of eggs and larvae was based on passive transport, and movement of juveniles and adults was a combination of passive transport and behavioral movement. Average number of days required to reach 5cm and the number surviving to 5cm were used as measures of recruitment. Averaged temperature and plankton concentrations within the IBM spatial domain showed strong interannual variation, and spatial and temporal patterns typical of the Peruvian upwelling system. Modeled anchovy growth and survival also showed strong interannual variation that resulted in large fluctuations in recruitment. Growth in a normal year resulted in anchovy requiring about 60–80 days to reach 5cm and the number of recruits was around 1010. Averaged anchovy length and number of survivors after 6 months for all monthly cohorts combined showed that good growth conditions and high survival occurred in the region from 5°S to 17°S, within 200km offshore and in the upper 100m. Vertically, centers of population distribution occurred at depths of 10m and between 50 and 70m. During the 1997–1998 El Niño, anchovy growth rate decreased so that it took fish 150–270 days to reach 5cm, and the numbers of survivors was about five orders of magnitude lower. Model results showed anchovy during El Niño conditions were located closer to shore, farther south, and at deeper depths than in normal years. Model results of interannual variation were insensitive to the length of anchovy used to define recruitment, how passive transport and behavioral movement were combined within different life stages, and to the starting locations of the monthly egg cohorts within the IBM domain. We discuss how our results can be used to infer anchovy recruitment under future climate change.