Shifts in driver dominance shape divergent ARG prevalence patterns and nonlinear responses across the Yellow River Basin.

The impact of environment and biology on antibiotic resistance genes (ARGs) has been widely documented, but the interaction pathways underlying shifts in ARG prevalence patterns remain unclear. This study employed metagenomic analysis to investigate how environmental and biological factors are associated with divergent spatial prevalence patterns of ARGs in the Yellow River. Then, Copula function, Monte Carlo simulation and generalized additive models (GAM) were further integrated to decipher nonlinear responses in divergent prevalence patterns of ARGs. The results revealed prevalence-dependent associations underlying ARG prevalence patterns: MGE-associated processes showed the strongest association with core ARG abundance, whereas environmental factors were indirectly associated with stochastic ARGs through microbial diversity, with some level-I risk ARGs showing localized accumulation. Nonlinear response analysis further revealed prevalence-dependent associations of ARG abundance with elevation and tnpA gradients. This study reveals previously unrecognized shifts in driver dominance of environmental and biological contributions to ARG prevalence patterns.