Fig. 3: Global-scale correlations and structural equation model showing the relative importance of climate, soil properties, agriculture management, and socio-economic on nitrogen (NUE) and phosphorus use efficiency (PUE) of four major crops (rice, wheat, maize, soybean).

a Spearman’s rank-order correlation between the NUE (PUE) of four major crops and environmental variables. All tests were two-sided. To control the false discovery rate in multiple comparisons, the resulting p-values were adjusted using the Benjamini-Hochberg method. Both the correlation coefficients and their corresponding adjusted exact p-values are annotated in the figure3a. *: p < 0.05; **: p < 0.01; ***: p < 0.001; ****: p < 0.0001. b Structural equation model between the NUE (PUE) of four major crops and environmental variables. Data are presented as path coefficients derived from bootstrap resampling (n = 100 iterations). The exact p-values are reported above the paths. The overall predictive performance of the model is evaluated using the Goodness-of-Fit (GOF) statistic, with higher values indicating better prediction. The amount of variance explained by the model (R²) is also shown for each response variable. All abbreviations are as follows: MAT, mean annual temperature; MAP, mean annual precipitation; ET, evapotranspiration; AI, aridity index; IND, inorganic nitrogen deposition; OND, organic nitrogen deposition; Sand, sand content; Silt, silt content; Clay, clay content; BD, bulk density; TC, total carbon; TN, total nitrogen; TP, total phosphorus; AP, available phosphorus; Tillage, no-till or not; During, experimental period; Input N, nitrogen fertilizer input; Input P, phosphate fertilizer input; Input K, potash fertilizer input; FAT, fertilizer application types; FAF, fertilizer application frequency; FAP, fertilizer application placement; IM, irrigation method; GDPPC, gross domestic product per capita; PD, population density; UR, urbanization rate; MYS, mean years of schooling.