Evaluating drivers of spatiotemporal variability in individual condition of a bottom-associated marine fish, Atlantic cod (Gadus morhua)

An organism's body condition describes its mass given its length and is often positively associated with fitness. The condition of Atlantic cod (Gadus morhua) in the Baltic Sea has declined dramatically since the early 1990s, possibly due to increased competition for food and hypoxia. However, the effects of biotic and abiotic variables on body condition have not been evaluated at local scales, which is important given spatial heterogeneity. We evaluate changes in distribution, experienced environmental conditions, and individual-level condition of cod in relation to covariates at different spatial scales using geostatistical models with spatial and spatiotemporal random effects. Sprat, Saduria entomon, temperature and oxygen were positively associated with condition, and depth was negatively associated. However, the effects of explanatory variables were small-spatial and spatiotemporal latent variables explained 5.7 times more variation than all covariates together (year excluded). Weighting environmental oxygen with local biomass densities revealed steeper declining trends compared to the unweighted oxygen in the environment, while the effect of weighting was less clear for condition. Understanding the drivers of spatiotemporal variation in body condition is critical for predicting responses to environmental change and to effective fishery management; yet low explanatory power of covariates on individual condition constitutes a major challenge.

Automated summary

An organism's body condition is positively associated with its fitness, but the condition of Atlantic cod in the Baltic Sea has declined due to competition for food and hypoxia. The effects of biotic and abiotic variables on body condition at local scales have not been evaluated, which is important considering spatial heterogeneity. Geostatistical models were used to assess changes in cod distribution, environmental conditions, and individual-level condition, revealing that temperature, oxygen, sprat, and depth are associated with cod condition. However, the explanatory power of these variables is low compared to spatial and spatiotemporal latent variables.