Influence of Smallholder Farmers’ Management Practices on Soil Organic Carbon in Western Kenya

Abstract

Declining soil organic carbon (SOC) content among smallholder farmers is mainly attributed to different agricultural management practices. A plethora of literature is available on research that has been done on the effect of agricultural management practices on SOC, however, there is limited information on site-specific management practices that improve and sustain carbon content. This study’s specific objectives were (i) to assess the predominant agronomic management practices among smallholder farmers and their influence on SOC and potential for sequestration, (ii) to assess the effect of soil physicochemical properties on soil organic carbon and (iii) to simulate SOC dynamics under different management practices using CropSyst model. For the first objective, using secondary data on household surveys from an earlier on-farm survey, a list of farms was generated. Farms with odd numbers and those that had been practicing cultivation for more than 10 years were chosen at a random. A total of 60 farms selected were selected, with 30 farmers selected from Shikomoli (Vihiga) and 30 farms from Mukuyu (Kakamega) villages. Data on land use history, inputs, crop residue retention and soil and water conservation strategies were collected. In addition, soil samples were obtained at consequent depths of 0-20 cm, 20-40 cm and 40-60 cm and analysed for SOC content. Total nitrogen, total organic carbon, silt, clay, sand, pH, available phosphorus, bulk density, exchangeable aluminium, zinc, copper, iron, manganese, and aggregate stability were all analysed in the soil samples for the second objective. For the third objective, the CropSyst model was calibrated and validated with data from farm’ fields in Western Kenya and later used to assess soil organic carbon (SOC) sequestration under business-as-usual compared to conservation agriculture. From the 60 farmers identified in the first objective, six farms representing different farmers under maize cropping systems were selected. The main criteria for selection was farms with contrasting soil texture and management practices representing the two sites. The model was calibrated using average farmer reported yields and measured SOC. The calibrated model was then used to project long-term (50 years) SOC trends assuming farm would continue to practice current management practices or, alternatively, conservation agriculture. The results indicated that users of manure in combination with chemical fertilizer were 57% in Shikomoli and 40% in Mukuyu, users of manure combined with crop residue retention and chemical fertilizer were 26% in Shikomoli and 10% in Mukuyu, users of chemical fertilizer alone were 17% in Shikomoli and 33% in Mukuyu and users of chemical fertilizer combined with crop residue retention were 17% in Mukuyu alone. xiv Hence, in both sites the predominant management practice was manure combined with chemical fertilizer. The key soil and water management techniques used in both sites were trenches and grass strips, with 27 percent in Shikomoli and 64 percent in Mukuyu. Under the combined application of manure, fertilizer, and crop residues, the SOC stock (Mg/ha) content analysis indicated significantly higher carbon stocks in Mukuyu (40.4 Mg/ha) and Shikomoli (29 Mg/ha). The lowest carbon stock resulted from combined application of fertilizer and crop residue in Mukuyu (of 32 Mg/ha) and combined manure and fertilizer application in Shikomoli (22.2 Mg/ha). SOC stocks reduced down the depth in Shikomoli at 0-20 cm depth (26.91 Mg/ha), 20-40 cm (24.17 Mg/ha) and 40-60 cm depth (22.56 Mg/ha) while in Mukuyu at 0-20 cm (40.01 Mg/ha), 20-40 cm (33.83 Mg/ha) and 40-60 cm (30.11 Mg/ha). In Shikomoli and Mukuyu, the 0-20 cm soil depth had the highest soil organic carbon (SOC) stocks (26.9 Mg/ha, 40.01 Mg/ha) while the lowest SOC stocks (22.56 Mg/ha, 30.11 Mg/ha) were observed at 40-60 cm soil depth, respectively. SOC stocks were positively correlated with total nitrogen, carbon to nitrogen ratio, silt, clay, and manganese, while sand and bulk density were negatively correlated with SOC stocks. Under the conventional farmer management practices, SOC stocks decreased for all the farm while under conservation agriculture, SOC stocks increased. Differences in simulated dynamics could be attributed to climatic conditions, soil texture, yield variability, initial carbon content, crop water stress, nitrogen stress and biomass yields. For current farmer management practice, the average annual loss of SOC stocks in the upper 20 cm ranged from 0.03 to 0.19 t C/ha/yr. SOC stocks increased significantly to 0.28 t C/ha/yr after switching to conservation agriculture. The results from this study indicated that loss of SOC stocks among smallholder farming systems could be reversed by adopting management practices that minimized soil disturbance and increase residue retention. Keywords: soil organic carbon, agricultural management practices, smallholder farmers, CropSyst model, business-as-usual scenario