Abstract
The impact of environmental variability, including ocean currents, temperature and food, on larval dispersal and growth of the Peruvian anchovy (Engraulis ringens) was investigated using an individual-based modelling approach combined with a realistic 3-D regional ocean model. This approach employed several different criteria in order to define larval recruitment numerically. The first recruitment criterion employed was a threshold age of 30 days, which corresponds to the approximate duration of the passive pelagic larval phase of the species. The objective of this first approach was to assess larval recruitment in relation to ocean currents through larval retention on the continental shelf off Peru. The simulated recruitment patterns obtained were predominantly influenced by spawning depth and month. Subsequently, a second recruitment criterion was employed with a threshold on larval length (2 cm), for which the modelling of larval dispersal was complemented with a larval growth model. This model simulating larval growth as a function of temperature and food employed two different versions of the Dynamic Energy Budget (DEB) bioenergetic theory. The initial version (DEBstd) focused exclusively on the larval phase and assumed isomorphic growth. The simulated larval recruitment patterns were found to be in line with those obtained when using the age criterion, yet it offered a more comprehensive insight into the recruitment mechanism in relation to the environment, particularly in terms of its capacity to investigate the diversity of simulated ages at recruitment. Moreover, this version of the model was also capable of reproducing the negative impact of El Niño on E. ringens recruitment. The second version of the growth model (DEBabj) was a re-estimation of the DEBstd growth parameters, but assuming non-isomorphic growth. This development was done with the aim of creating a more realistic model, as well as a full life cycle model that could be extended to other studies investigating the ecology of the species in the future. Here, the challenge was to reliably express the physical length during the larval period. This issue was due to the hypothesis of acceleration in the early life stages and the influence of the shape coefficient (δ), an auxiliary parameter of the DEB model, on this physical length. Further exploration is required to obtain a scheme that correctly reflects the changes in shape and size during the larval period. To address this challenge, we put forth the use of two alternative criteria for recruitment that are independent of the shape coefficient: maturity, and volumetric structure. Using these criteria lead to some changes in the simulated recruitment pattern, which require further investigation. Building on the ichthyoplankton dispersal modelling tool Ichthyop, we have developed Ichthyop-DEB, a software that can simulate simultaneously larval dispersal and growth of E. ringens within the Peruvian upwelling ecosystem. This new tool integrates physical and biological processes within a threedimensional oceanic environment and is a step towards developing a numeric avatar of the ocean. Ichthyop-DEB can now be easily applied to other species whose life cycle includes a larval pelagic phase.