Work

Long-term soil nitrogen and phosphorus dynamics under contrasting input systems and environmental drivers in semi-arid Canadian Prairies: Implications for nutrient management

Year: 2026

Type: article

Abstract: This study provides rare 18-year evidence on how climate and terrain interact with input systems to shape soil nitrate (NO₃⁻–N) and available phosphorus (available P) dynamics in semi-arid Canadian Prairies. A long-term field experiment was conducted from 1994 to 2013 at Scott, Saskatchewan, Canada, to assess three input systems (organic (ORG), reduced (RED), and high (HI) and three cropping diversity levels under variable growing season precipitation (GSP), temperature, and terrain attributes. Machine learning and mixed models revealed that environmental factors explained more variation in soil nutrient status than management practices: monthly GSP accounted for up to 30% of NO₃⁻–N variation, GDD for 33%, and terrain attributes for up to 50% of available P variation. Inputs explained 6–12% of variation. NO₃⁻–N declined over time, while available P accumulated in reduced-input systems, exceeding high-input systems by the end of the study. We concluded that environmental factors were the key drivers of long-term soil nutrient dynamics in this study. This work is among the first to integrate machine learning with long-term field data to quantify climate–terrain–management interactions for nutrient cycling, offering actionable guidance for precision agronomy. Incorporating environmental factors alongside fixed treatment effects in analyses of crop inputs and rotational diversity is needed to optimize fertilizer management strategies.

Source: European Journal of Agronomy

DOI: 10.1016/j.eja.2026.128224