Abstract
Aim Global warming increasingly threatens marine fish, but environmental drivers of coastal fish biodiversity and fish trait responses to these drivers remain unclear, hindering understanding of biodiversity–environment relationships. This study disentangled direct/indirect effects of environmental factors on coastal China Seas fish communities and identified trait responses to these drivers. Location China Seas. Methods Piecewise structural equation modelling (pSEM) quantified direct/indirect environmental effects on fish communities; fourth‐corner analysis combined with RLQ identified trait responses to drivers. Results The direct and indirect effects of temperature on fish species richness were contrasting. The direct effect was negative (standardized coefficient: −0.29), which was associated with the adverse temporal impacts of global warming on biodiversity; the indirect effect was positive (value: 0.72), which explained the spatially higher biodiversity observed in warmer seas. Salinity, water depth and seafloor substrate were confirmed as important drivers of fish communities in marine systems, with effects linked to fish physiology (salinity), habitat availability (water depth, seafloor substrate) and feeding activities (seafloor substrate). Four key traits were identified as responsive to the environmental drivers: generation time, caudal peduncle throttling, pectoral fin size and food consumption to biomass ratio. The Bohai Sea and Yellow Sea were associated with fish having longer generation time, smaller caudal peduncle throttling, smaller pectoral fin size and lower food consumption to biomass ratio; the East China Sea showed intermediate trait characteristics; and the South China Sea exhibited trait patterns opposite to those of the Bohai Sea and Yellow Sea. Main Conclusions Contrasting temperature effects explain spatiotemporal fish biodiversity patterns. Northern coastal fish consistently have longer generation time, lower swimming capability and lower consumption ratio than southern fish. This work clarifies environment‐fish community causality and advances understanding of trait–environment associations.