Nonylphenol stress intensifies phosphorus limitation and restructures microbial co-occurrence networks in soil

Nonylphenol (NP) is a globally concerned endocrine disruptor, yet its ecological risks in soil are often evaluated solely by residual concentrations or degradation rate, overlooking the more subtle functional and structural disturbances to soil microbiomes. Here, we conducted a 90-day concentration-gradient microcosm experiment to integrate NP degradation dynamics with extracellular enzyme stoichiometry, vector-based microbial nutrient limitation, community succession, and co-occurrence network reorganization. We found that NP was efficiently degraded (>75% even at 120 mg kg⁻¹ within 90 days), but rapid degradation did not prevent profound functional shifts. Medium-to-high NP concentrations (≥30 mg kg⁻¹) significantly suppressed β-glucosidase (max inhibition 68.4%) and, more importantly, altered enzyme stoichiometric ratios (Carbon (C): nitrogen (N) and C: phosphorus (P) decreased linearly with NP concentration). Vector analysis revealed a critical transition: Medium-to-high NP exposure shifted microbial metabolic indicators from C-P co-limitation toward stronger P limitation. Concurrently, NP exerted strong selective pressure, reducing α-diversity but enriching NP-tolerant and putative degradative taxa (Proteobacteria, Lysobacter, Pseudomonas). This compositional restructuring drove microbial co-occurrence networks toward a more connected yet topologically reorganized "NP-adapted" configuration, with keystone taxa shifting from conventional nutrient cyclers (Massilia, Nitrospira) to stress-tolerant degrader genera (Truepera, Pseudonocardia). Collectively, we demonstrate that NP's ecotoxicological fingerprint lies not in its persistence but in its ability to decouple C-N-P acquisition strategies and force a network‑level adaptive reorganization - even under substantial degradation. Thus, risk assessment for NP‑contaminated soils must move beyond degradation data alone to include enzymatic stoichiometric imbalances, microbial P‑limitation status, and co‑occurrence network topology.