Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
Climate warming and anthropogenic nitrogen enrichment are concurrent global‐change drivers that may alter soil biological and biogeochemical functions, yet whether their combined effects represent true non‐additive interactions remains unresolved. We conducted a factorial meta‐analysis of experiments containing complete control, warming treatment (W), nitrogen addition treatment (N), and combined warming plus nitrogen treatment (WN). The primary dataset included 26 independent experimental platforms and 1774 effect sizes. Individual treatment effects, the combined treatment effect, and the true W × N interaction were estimated separately; cluster‐robust variance estimation used the CR2 small‐sample correction (CR2), and response variables were classified into six soil functional modules. In the primary dataset, W (+2.15%), N (+3.44%), and WN (+5.84%) produced weak positive pooled responses, whereas the combined treatment became significant only in the expanded dataset (+7.15%, p = 0.020). The overall W × N interaction remained near zero across the strict, primary, and expanded datasets; in the primary dataset it was −0.54% (95% confidence interval, −4.51% to 3.59%; p = 0.794), indicating that the stronger combined response was largely consistent with additivity on the natural‐log response ratio (lnRR) scale. Module‐level responses nevertheless diverged. Microbial attributes showed a consistent negative interaction (−11.01%, p = 0.001; CR2 p = 0.045), driven mainly by microbial biomass. Carbon‐pool variables showed a recurrent but uncertain positive tendency (+7.81%, p = 0.043) because heterogeneity was high, an Egger‐type test indicated a potential small‐study effect, and the estimate remained positive but non‐significant after excluding methodologically complex variables (+6.93%, p = 0.134). Enzyme‐related variables tended positive, whereas flux and transformation processes tended negative. Excluding soil pH, stoichiometric ratios, and percentage‐based indices shifted the overall interaction from −0.54% to −1.87% (a 1.33‐percentage‐point change) while preserving the near‐zero inference and principal module directions. Overall additivity therefore masked process‐dependent responses among soil functional components.