Microbial network complexity supersedes diversity in predicting soil multifunctionality during dryland grassland restoration: Implications for degraded calcareous ecosystems

Zang, Zhenfeng , Li, Yingxue , Tang, Fuhao , Zhao, Wei

2026-11-01 AGRICULTURE ECOSYSTEMS & ENVIRONMENT 2026   411(卷), null(期), (null页)

查看原文

Grassland restoration is critical for combating land degradation affecting 30% of global drylands, yet mechanisms associated with functional recovery remain poorly understood. A persistent challenge in microbial ecology is the inconsistent relationship between microbial diversity and soil ecosystem functioning, which may partly arise because bulk soil assessments overlook the strong micro-heterogeneity of soil habitats. Here, we tested whether soil aggregate size mediates diversity-multifunctionality relationships through microbial network complexity rather than taxonomic richness. Along a 45-year grassland restoration chronosequence in calcareous soils, we fractionated aggregates into three size classes (>2 mm, 0.25-2 mm, <0.25 mm) and quantified 14 ecosystem functions spanning carbon, nitrogen, and phosphorus cycling. Combining high-throughput sequencing with co-occurrence network analysis showed that bacterial diversity was negatively associated with multifunctionality (P < 0.001), while network complexity accounted for 37% of the variation in multifunctionality, exceeding all diversity metrics. Small macroaggregates (0.25-2 mm) maintained 25% higher multifunctionality than microaggregates (<0.25 mm), and this advantage was associated with 42-89% greater network complexity rather than species richness. Structural equation modeling further indicated that phosphorus limitation enhanced multifunctionality indirectly through shifts in network architecture, with greater network robustness under P-stressed conditions, suggesting that resource stress may promote tighter microbial cooperation. Together, these findings show that aggregate-scale heterogeneity helps resolve the diversity-function paradox that is often obscured in bulk soil studies. Our results highlight soil structure and microbial interaction networks as key determinants of ecosystem recovery in degraded calcareous grasslands.