Bista et al. (2026) Understanding soil moisture dynamics and key drivers for single and mixed cover cropping systems using machine learning

Identification

• Journal: Agricultural Water Management
• Year: 2026
• Date: 2026-09-17
• Authors: Prakriti Bista, Olufemi Adebayo, Huichao Yin, Kenneth C. Carroll, Rajan Ghimire
• DOI: 10.1016/j.agwat.2026.110777

Research Groups

• Department of Plant and Environmental Sciences, New Mexico State University, Las Cruces, NM, USA
• Agricultural Science Center, Clovis, NM, USA

Short Summary

This study investigates soil moisture dynamics in single-species and mixed cover cropping systems using machine learning. The results show that cover crop mixtures can improve water capture and utilization, and agronomy-guided machine learning can capture complex soil-plant-atmospheric interactions.

Objective

• Evaluate the effect of single (grass and legume individually) and cover mixture (grass and legume) on soil moisture dynamics, precipitation use efficiency, and crop water productivity compared to no-cover crop/control or fallow-based cropping system.
• Identify key environmental and management drivers regulating soil moisture at the surface (5 cm) and subsurface (30 cm) depths using machine learning model with SHAP analysis.

Study Configuration

• Spatial Scale: Local scale, New Mexico State University Agricultural Science Center near Clovis, NM (34°35′ N, 103°12′ W; elevation 1368 m).
• Temporal Scale: Two years, from March 2023 to June 2024.

Methodology and Data

• Models used: Extreme Gradient Boosting (XGBoost) machine learning model.
• Data sources: Field measurements of soil water content, weather data from a local weather station, irrigation data, and soil temperature data.

Main Results

• Cover crop treatments consistently retained greater soil water content than no cover crop control or fallow in 5 cm depth across all phases.
• The highest VWC was observed under peas, followed by oats and the PO mixture, while NCC had the lowest VWC.
• At 30 cm, VWC exhibited lower temporal variability and generally higher moisture than at 5 cm.
• Soil water storage during the no-crop phase was positive and highest in the PO mixture (+8.15 mm), whereas all other treatments lost water.

Contributions

• This study provides critical insights into cover crop selection and management to improve soil water conservation and agricultural sustainability in semi-arid soils.
• The results highlight the potential of cover crop mixtures to improve water capture and utilization, and agronomy-guided machine learning can capture complex soil-plant-atmospheric interactions.

Funding

• This research was funded by [project/program name and reference code].

Citation

@article{Bista2026Understanding,
  author = {Bista, Prakriti and Adebayo, Olufemi and Yin, Huichao and Carroll, Kenneth C. and Ghimire, Rajan},
  title = {Understanding soil moisture dynamics and key drivers for single and mixed cover cropping systems using machine learning},
  journal = {Agricultural Water Management},
  year = {2026},
  doi = {10.1016/j.agwat.2026.110777},
  url = {https://doi.org/10.1016/j.agwat.2026.110777}
}