Comparative analysis of on-farm greenhouse gas emissions from smallholder crop-livestock farming systems in Sub-Saharan Africa – A case study in Kenya

Abstract

This thesis addresses the contribution of the smallholder mixed crop-livestock systems to greenhouse gas emissions. The first objective of the thesis was to quantify region-specific enteric methane emission factors for cattle in two counties (Bomet and Nandi) in Kenya. The calculated Tier 2 EFs were compared to the Intergovernmental Panel on Climate Change (IPCC) Tier 1 values for the Africa region and the emission factors were 25% to 45% lower than the IPCC default values for the county of Bomet in Kenya, a substantial difference. These differences were likely caused by the differences in feed digestibility values which were 13% higher than the IPCC values (62.2% vs 55%). Tier 2 emission factor values from a neighboring region in Western Kenya - Nandi County (published results) located close to Bomet county had similar agro-ecological zones and the difference ranged between 4% lower to 4% higher than the emission factors from the Nandi region, a much smaller difference. The smaller difference was likely caused by the similar livestock keeping and climatic conditions in the two regions. In addition, Nandi Tier 2 emission factors were derived using the same approach as used for Bomet County region-specific activity data. The second objective was to conduct a partial life cycle assessment (LCA) to quantify the carbon footprint of smallholder crop-livestock farming systems. The emissions intensities of the farms were highly variable and ranged from emissions intensities as found in developed countries to very high emissions intensities. This result was caused by the differences in farm outputs and herd structures. Farms in the highlands (Nandi and Bomet counties) showed lower emissions intensities while farms in semi-arid regions (Nyando county) showed much larger emissions intensities. Conducting an LCA is data demanding while emissions reporting especially for the livestock sector in Africa requires locally specific data. As such simpler methodologies for acquiring such location-specific (and thus country-specific) data on the performance of the livestock systems and feed characteristics is needed. The third objective developed a simple, cost-effective protocol to derive locally specific data for quantifying enteric methane emission factors. Use of heart girth measurements, spot sampling of milk yield and use of default milk energy content showed the least bias compared with the actual measurements of either using specialized equipment or daily recording. However, the results were sensitive to feed quality values used by the model because the values from the literature may not be representative for specific regions and therefore if not actively collected may result in large bias and uncertainty.