Characterisation of smallholder farming systems and greenhouse gas emissions simulation from maize cropping system in tharaka-nithi county, kenya

Abstract

The influence of soil fertility management technologies on crop production has widely been researched in Tharaka-Nithi County. However, data on their contribution towards national greenhouse gas budget is scanty. This study aimed at characterising smallholder farming systems and simulating greenhouse gas emissions, maize yields, yield scaled nitrous oxide (N2O) emissions and N2O emission factors from different soil fertility management technologies in Tharaka-Nithi County. Three hundred households were interviewed to obtain data for farming systems characterisation and evaluation of socio-economic factors influencing the diversity of farm typologies. Interview schedules were administered using open data kit collect mobile App. Multivariate analysis was done to characterise smallholder farming systems. To evaluate socio-economic factors influencing farm diversity, Chi-square, t-test, and multinomial regression analysis were carried out using the Statistical Package for Social Sciences (SPSS version 23). For calibration and validation of the DeNitrification DeComposition (DNDC) model, a one-year soil greenhouse gas quantification experiment data were used. The data were obtained from a field experiment conducted in Kigogo primary school. It was laid out in randomised complete block design under four soil fertility treatments as control (no external inputs), inorganic fertiliser (NP, 23.23, 120 kg N ha-1 yr-1), animal manure (goat manure, 120 kg N ha-1 yr-1) and animal manure + inorganic fertiliser (120 kg N ha-1 yr-1) replicated thrice. Climate, soil properties, N2O fluxes, maize yields and farm management data were used. The model was evaluated using modelling efficiency, mean error, coefficient of determination, mean absolute error, and root mean square error (RMSE). The experimental data were subjected to Analysis of Variance in SAS 9.4 software and mean separation done using least significance difference at p = 0.05. The results showed six farm types: Type 1, comprising cash crop and hybrid cattle farmers; Type 2, involving food crop farmers; Type 3, composed of coffee-maize farmers; Type 4, consisting of millet-livestock farmers; Type 5, comprising highly diversified farmers, and Type 6, had tobacco farmers. Land size, total tropical livestock unit, the proportion of land and amount of nitrogen applied to different cropping systems were significant in the construction of farm typologies. The DNDC model was fair in simulating daily N2O fluxes (54% ≤ normalized RMSE (nRMSE) ≤ 68% and 0.26 ≤ modelling efficiency (MEi) ≤ 0.49) and good to excellent performance in simulating cumulative annual soil N2O fluxes (6.16 ≤ nRMSE ≤ 12.86 and 0.63 ≤ MEi ≤ 0.86) across soil fertility treatments. The cumulative observed and simulated annual soil N2O fluxes ranged between 0.21±0.01 and 0.38±0.02 kg N2O-N ha-1 yr-1 and 0.20 kg N2O-N ha-1 yr-1 (control) to 0.38 (fertiliser) kg N2O-N ha-1 yr-1. The simulated N2O yield scaled emissions, and emission factors ranged from 0.022 to 0.029 g N Kg -1 grain yield and 0.03 % to 0.14% under manure and fertiliser treatments, respectively. Based on the low observed and simulated emission factors, using the Intergovernmental Panel on Climate Change (IPCC) Tier 1 default factor of 1% overestimates agricultural soils GHG emissions in the Central Highlands of Kenya. Manure and fertiliser combination should be promoted to enhance the three pillars of climate-smart-agriculture (CSA) as food security, climate change mitigation and adaptation.