Discussion
Our study supports the hypothesis that temperature is a main driver of large-scale latitudinal variation in fish community biomass. This result is likely driven by a reduced trophic transfer efficiency and a faster turnover rate of fish biomass in warmer waters. As expected, demersal community biomass is negatively related to fishing exploitation and positively related to zooplankton prey production. All these findings are consistently observed across the different spatial scales studied. We find no evidence that temperature fluctuations and recent warming have impacted demersal community biomass. Even though we found no effect of recent warming, our study provides an empirical basis for long-term climate predictions and suggests a set of explanatory variables that are most important.
The lack of a relationship between demersal fish biomass and the detrital bottom flux and the positive but weak relationship between zooplankton prey production and demersal fish biomass in the SEM were unexpected, as prey production should ultimately constrain the energy available to fish. From a trophodynamic perspective, the weak relationship between prey production and biomass and the strong relationship with temperature suggests that temperature-modulated impacts on fish turnover rates and/or trophic transfer efficiencies are more important than the baseline prey resources in determining demersal fish biomass, at least for the range of systems and scales considered here. This finding is supported by the trophodynamic model, which required a strong negative relationship between the transfer efficiency and temperature to obtain skillful demersal fish biomass predictions.
In retrospect, the weak relationship with prey production is not too surprising as the data compilation covers a considerable thermal range (-1 to 27 °C), while the studied shelf systems have moderate to high productivity and productivity varies less than a factor 4 (Figure 1d). It is thus expected that prey production could become more important for predicting changes in demersal fish biomass in both the SEM and the trophodynamic model along a broader productivity gradient. For example, a gradient from the shelf to the deep ocean that covers larger differences in benthic prey production (Wei et al. 2011).
It is important to note that the negative relationship between demersal fish biomass and temperature does not necessarily imply that the potential sustainable fishing catch will be lower in warmer shelf systems. Catch is a flux (biomass removed per unit time) similar to production and is differently affected by temperature compared to biomass. Increased biomass turnover times at higher temperatures, for example, decreases the biomass associated with a given production after warming, e.g., du Pontavice et al . (2021). In contrast with the demersal fish biomass results herein, estimates of plankton food web production available to fish can provide moderately skillful fisheries catch predictions at a global scale (Friedland et al. 2012; Stocket al. 2017). Similar to the demersal fish biomass results herein, a strong negative dependence between the transfer efficiency and temperature significantly improved fisheries catch estimates (Stocket al. 2017).