Abstract

Soil health is fundamental to sustainable agriculture and ecosystem resilience, with microbial communities playing a pivotal role in nutrient cycling, organic matter decomposition, and soil structure maintenance. This study investigates the relationship between microbial co-occurrence networks and soil health indices in a 200-hectare agricultural landscape in Saskatchewan, Canada. Using high-throughput 16S rRNA sequencing, we characterized bacterial and fungal communities across 120 soil samples. Network analysis revealed that soils with higher soil health indices (e.g., soil organic carbon [SOC], aggregate stability) exhibited denser microbial networks with keystone taxa such as Bacillus and Mortierella. Random Forest models linked network metrics (e.g., node degree, clustering coefficient) to soil health indices, achieving an R² of 0.88 for SOC prediction. Key findings show that microbial network complexity correlates positively with soil health, offering insights for microbial-based soil management strategies.