1. Subseasonal to decadal ocean forecasting can make significant contributions to achieving effective management of living marine resources in a changing ocean. Most applications rely on indirect proxies, however, often measured at the ocean surface and lacking a direct mechanistic link to the dynamics of marine populations.

2. Here we take advantage of three-dimensional, dynamical reconstructions and forecasts of ocean biogeochemistry based on a global Earth System Model to hindcast and assess the capacity to anticipate fluctuations in the dynamics of bigeye tuna (Thunnus obesus Lowe) in the Pacific Ocean during the last six decades. We reconstructed spatial patterns in catch per unit effort (CPUE) through the combination of physiological indices capturing both habitat preferences and physiological tolerance limits in bigeye tuna.

3. Our analyses revealed a sequence of four distinct regimes characterized by changes in the zonal distribution and average CPUE of bigeye tuna in the Pacific Ocean. Habitat models accounting for basin-wide fluctuations in the thermal structure and oxygen concentration throughout the water column captured interannual fluctuations in CPUE and regime switches that models based solely on surface information were unable to reproduce. Decade-long forecast experiments further suggested that forecasts of three-dimensional biogeochemical information might enable anticipation of fluctuations in bigeye tuna several years ahead.

4. Synthesis and applications. Together, our results reveal the impact of variability of biogeochemical conditions in the ocean interior on the dynamics of bigeye tuna in the Pacific Ocean, raising concerns about the future impact of ocean warming and deoxygenation. The results also lend support to incorporating subsurface biogeochemical information into ecological forecasts to implement efficient dynamic management strategies and promote the sustainable use of marine living resources.

Reconstructions and forecasts of bigeye tuna CPUE in the Pacific Ocean. The data set  also includes environmental covariates and habitat quality indices described in the text and derived from BGC fields. Ocean reanalysis and decade-long forecast experiments provided by the Ensemble Coupled Data Assimilation system integrating the Carbon, Ocean Biogeochemistry and Lower Trophics biogeochemical model [ECDA-COBALT; see Park et al 2018 and Park et al 2019].

Please check the file `readme.qmd` for further details.

Dataset includes both R rds files that can be explored using R function readRDS, and NetCDF files that, among many other alternatives (see e.g. the software list at Unidata), can be explored using software packages like Panoply or ncview.