A robust DayCent model calibration to assess the potential impact of integrated soil fertility management on maize yields, soil carbon stocks and greenhouse gas emissions in Kenya

Summary

This study presents a robust Bayesian calibration of the DayCent biogeochemical model to simulate maize productivity and soil carbon dynamics under integrated soil fertility management in Kenya. Using data from four long-term field experiments involving varying organic resource additions with and without mineral nitrogen fertiliser, the researchers significantly improved model performance for yield simulation (EF from 0.33 to 0.51) and SOC change prediction (EF from −1.3 to 0.54) after calibration. The calibrated model demonstrates spatial transferability through leave-one-site-out cross-validation, suggesting potential utility for regional or national-scale upscaling of ISFM impact assessments.

Regional applicability

Whilst the study is specific to Kenyan agroecosystems and ISFM contexts relevant to sub-Saharan African yield gaps, the methodological approach to model calibration and validation may have limited direct applicability to UK temperate cereal production, which typically operates under different soil, climate and management regimes. However, the framework for robust model parameterisation could inform development of decision-support tools for UK sustainable intensification scenarios.

Key measures

Maize yield, soil organic carbon (SOC) changes, global warming potential, Nash-Sutcliffe modeling efficiency (EF), organic and mineral fertiliser treatments

Outcomes reported

The study calibrated and validated the DayCent biogeochemical model to simulate maize productivity, soil organic carbon changes, and greenhouse gas emissions under integrated soil fertility management (ISFM) practices across four long-term field experiments in Kenya.

Topic tags