The potential yield of oceans has been a research topic since the early 60s. Direct extrapolation of catch data has led to regional and global models. These models attempt to capture and integrate ecological, economical and social aspects, or so it seems. Looking into the parameterization of such models is paramount to understand uncertainties in current estimates of global fisheries yield.
A review of the literature from the past 50 years reveals a range of estimates varying between a predicted collapse and over 2 billion tonnes of seafood harvestable per year. These estimates are, however, the direct results of various global models, which still show a strong rift between biomass-based and socioeconomic approaches. The definition of what is “harvestable” still depends on the perspective - ecological or human. Although bioeconomics and end-to-end models aim at bridging this rift, their focus still reflects the disciplinary biases of their components, and they have fallen in the fallacy of complexity. Truly integrated socio-ecological models are rare at global levels, although expanding on the range of regional ones available. The challenge is not expanding regional models globally, but developing models complex enough to represent the feedback effect between biological, economical and human components, while simple enough to reach a broad audience.
Even so, many points of focus in these components require attention. Landings (reported catch) still is used as a proxy for mortality, ignoring illegal fishing or discards, which can account for over a third. Drivers of the global fleet remain vastly unknown, leading to estimates of future catch still extrapolated from past trends. Future socio-ecological model of global fisheries should aim at addressing these uncertainties and reach a balance between parsimony, scope and fit.