Global patterns reveal a non-linear relationship between soil pH and salinity

Soil pH is widely used as an indicator of soil health (Lebron et al., 2025) because it regulates various chemicals, physical, and biological processes. Soil electrical conductivity (EC), a proxy for soil salinity (Hassani et al., 2024; Shokri et al., 2025), is often analysed alongside soil pH to characterise soil chemical conditions. Although studies suggest a strong correlation between soil pH and EC, their relationship has rarely been quantified at continental or global scales. In this study, we applied a Generalized Additive Model (GAM) to globally investigate the relationship between soil pH, soil EC, and a set of environmental covariates representing soil texture, climate, terrain, and vegetation. Soil pH observations were obtained from the WoSIS and LUCAS databases. Our results suggest that soil EC is the most influential predictor (16.00%) for soil pH predictions, followed by soil water balance (14.00%), NDVI (12.00%), bulk density (10.00%), and minimum temperature (10.00%). Our analysis also shows that the relationship between soil pH and EC was distinctly non-linear. Soil pH increased from 6.59 to 7.29 as EC rose from 0 to 0.6 dS/m, then declined gradually until EC reached approximately 32 dS/m, beyond which pH stabilised near 6.71. Overall, the identified non-linear pH and salinity relationship provides new insight into the chemical constraints shaping soil conditions at continental to global scales.