Predicting the impact of climate warming on soil quality using bacteria and machine learning.
Journal:
Journal of environmental management
Published Date:
Jul 5, 2025
Abstract
In the context of global warming, a substantial portion of global soil is in a state of degradation, which poses a significant threat to biodiversity and food production worldwide. Moreover, monitoring soil quality typically requires measuring numerous physical, chemical, and biological indicators, resulting in high costs. In this study, 286 soil samples were obtained from the climate-sensitive Tibetan Plateau and subjected to 16S rRNA amplicon sequencing to reveal the relationships between soil quality, soil bacteria, and climate warming. The results indicated that climate-sensitive bacteria could effectively predict soil quality indices through machine learning (R > 0.76). This suggests that 16S rRNA sequencing can replace numerous soil indicators, providing comprehensive information on soil quality and reducing the costs associated with soil quality monitoring. Additionally, model predictions demonstrated a slight increase in soil quality when only the average annual temperature increased by 1.5 °C. However, when other climatic factors (precipitation and temperature during specific periods) also changed (future climate scenarios for 2021-2040 and 2080-2100), which is more realistic than only increasing the average annual temperature, soil quality declined and the greater the increase in temperature, the more severe the decline in soil quality. These findings provide valuable insights for soil management in the face of increasingly severe climate warming conditions. In summary, our research offers new perspectives for soil quality monitoring from a microbiome standpoint and indicates that future climate warming could pose a threat to soil quality.
Authors
Keywords
No keywords available for this article.
