Abstract
We developed and evaluated a physics-forced marine ecosystem modelling framework for reconstructing spatial and seasonal lower-trophic dynamics over multi-decadal time horizons. The framework links the spatiotemporal capabilities of Ecopath with Ecosim (EwE) to physical-oceanographic and atmospheric reanalyses and was implemented for the pelagic zone of the Strait of Georgia, Canada, as the Strait of Georgia Ecosystem Model of Lower Trophic Levels (SOGEM-LTL). The model was run as a four-decade hindcast from 1980 to 2018 and represents 22 functional ecological groups spanning primary producers, mixotrophs, crustacean and gelatinous zooplankton, ichthyoplankton, nekton, bacteria, and detritus. We evaluated the hindcast against observational benchmarks for spring-bloom timing, dissolved inorganic nitrogen inventories, phytoplankton composition and variability, and zooplankton biomass and anomalies. SOGEM-LTL reproduced broad interannual variation in spring bloom timing, the first-order seasonal cycle of nutrient depletion and recovery, and spring crustacean zooplankton anomalies. Agreement depended strongly on the temporal and spatial support of the comparison: seasonal and phenological indicators were generally more skillfully reconstructed than local point-to-cell comparisons or annual summaries. The resulting outputs provide a spatially resolved four-decade reconstruction of seasonal lower-trophic fields for the Strait of Georgia. The study demonstrates that an ecologically detailed, relatively low-compute, E2E-capable framework can complement more process-rich biogeochemical models for investigating bottom-up forcing and match–mismatch dynamics.

This publication has 87 references indexed in Scilit: