Homogeneous spatial structure provides bet-hedging capacity for populations to spread risks in fluctuating environments. Theories have suggested that population spatial variability is influenced by age structure, abundance, and the environment. However, clear empirical evidence for such causal links is lacking. Here, using a 25-year survey data from the North Sea, we quantified the above-mentioned causal effects by empirical dynamical modeling that determined causality based on the concept of state-space reconstruction rather than correlation. We showed that all variables causally affected population spatial variability. Specifically, truncated age structure elevated population spatial variability. Warming and spatially heterogeneous temperature were likely to enhance population spatial variability. Abundance effect tended to be regulated by aggregation tendency of a species. Our findings highlighted a potential double jeopardy of fishing on increasing vulnerability of fishes to environmental variability via age truncation and abundance reduction, providing a rationale to incorporate spatial dynamics into fisheries management.