Soil quality shapes the composition of microbial community stress response and core cell metabolism functional genes

The capacity for soils to perform vital functions, such as agricultural production, is dependent on numerous properties. Their simultaneous effect on soil biota is of interest to assess impacts of agricultural management. Using a soil quality index (SQI) based on a priori assumptions of eight soil physico-chemical properties that promote plant growth and microbial biomass, we sought to: 1) investigate the effect of land use on SQI; and 2) test a relationship between SQI and the composition of microbial functional genes. In 29 soils under four distinct land uses (cotton, wheat, pasture and native vegetation) gene composition was most distinct in cotton. The SQI followed the gradient cotton < wheat < native vegetation < pasture, with pasture significantly greater than other land uses. Of 67 functional gene markers, gradient boosted machine learning identified five genes that correlated strongly with SQI. These were stress response (oxyR and dnaJ), core carbon, nitrogen and sulfur metabolism (PTS-Glc-EIIA, glnH and sat, respectively). Nutrient cycling functional genes did not correlate with SQI. A structural equation model of the relationship between soil properties, SQI, and the aforementioned genes was used to visualise these interactions (root mean squared error of 0.1, R² of 0.64 and p < 0.001). We conclude that certain land use practices improve or degrade soil quality relative to native vegetation, and SQI primarily correlates with microbial stress response and core metabolism. This demonstrates capacity for microbial communities to adapt to environmental stress while certain functions remain relatively resilient.