A road to food? : efficacy of nutrient management options targeted to heterogeneous soilscapes in the Teso farming system, Uganda

Key words: Land use change; Heterogeneity in soil fertility; Targeting; Integrated soil fertility management; Nutrient use efficiencies; Rehabilitation of degraded fields; Fertiliser requirements, Finger millet; QUEFTS model; Smallholder systems; sub-Saharan Africa.

Poor soil fertility in smallholder farming systems in sub-Saharan Africa is recognised as a major factor responsible for low per capita food production and escalating food insecurity. Increasing food production in most smallholder farming systems requires intensification with nutrient inputs. Targeting nutrient management interventions to heterogeneity can greatly enhance the use efficiency of the scarce nutrient inputs and can help in identification of ‘best fits’ (most suitable options for niches within the systems). This thesis aimed at contributing to understanding how to target nutrient management options to heterogeneity for improved crop production in the Teso farming system in eastern Uganda.
Land use change analysis between 1960 and 2001 showed that 48-78% more land was brought into cultivation and disappearance of communal grazing lands. Productivity of the farming system is also low. Population growth, political-instability-mediated collapse of institutions that supported production and marketing of cotton, and cattle rustling account for the changes in land use and productivity of the system. Balances of N, P and K were positive on larger farms (LF) and negative on the medium farms (MF), small farms with cattle (SF1) and without cattle (SF2), but were negative at the crop scale on all the farm types. Livestock, crop yield, labour availability and access to off farm income are the sustainability indicators in the system.
There were no topographic-gradients in soil pH, SOC, total N, Exch. Mg, Exch. Ca, Exch. K, CEC, sand and clay in the two villages with different geo-morphological features characterised except for extractable P which was 3 - 5 times higher in the in the top soils of the profiles in the valley bottoms than those in the upper landscape position of the toposequences. Soil organic carbon (SOC) concentrations significantly differed (P<0.05) in surface soil properties between landscape positions and even significantly much larger (P<0.001) between field types. Fields classified as of good, medium and poor soil fertility by farmers had average SOC concentrations of respectively 9.3-15 g kg–1, 6.6-11 g kg–1, 5.5-7.0 g kg–1. In contrast with other studies in smallholder farming systems in sub-Saharan Africa, spatial analysis did not reveal a particular generalized pattern in variability in soil fertility across farms. Within-farms, larger contents of SOC were associated with larger amounts of silt + clay and on locations of former kraals. The field scale, which is easily recognised by farmers, is an important entry point for targeting soil fertility management technologies.
Heterogeneity in soil fertility affected performance of legumes established with and without P and their residual effect on subsequent finger millet crops. Legume biomass and N accumulation differed significantly (P<0.001) between villages, landscape position, field type and P application rate. Mucuna accumulated the most biomass (4.8-10.9 Mg ha–1) and groundnut the least (1.0-3.4 Mg ha–1) on both good and poor fields in the upper and middle landscape positions. N accumulation and amounts of N2-fixed by the legumes followed a similar trend as biomass, and was increased significantly by application of P. Grain yields of finger millet were significantly (P<0.001) higher in the first season after incorporation of legume biomass than in the second season after incorporation. Finger millet also produced significantly more grain yield in good fields (0.62-2.15 Mg ha–1) compared with poor fields (0.29-1.49 Mg ha–1). Farmers preferred growing groundnut and were not interested in growing pigeonpea and mucuna. They preferentially targeted grain legumes to good fields except for mucuna and pigeonpea to poor fields. Benefit-cost ratios indicated that legume-millet rotations without P application were only profitable on good fields. Green grams, cowpea and soyabean without P can be targeted to good fields on both upper and middle landscape positions in both villages but mucuna without P to poor fields on the middle landscape position in Chelekura village and cowpea without P to poor fields on the upper landscape position in Onamudian village.
Application of N, P fertilisers alone (0, 30, 60, 90 kg ha–1), N+P at equal rates of single application, and manure (3 t ha–1) supplemented with N (0, 30, 60 and 90 kg ha–1) to degraded fields closed the within farm yield gap in finger millet by only 24%-43 %. The inability of the options to close the yield differences was because of poor nutrient use efficiencies (<25%) and other nutrient limitations (S and K) and physical limitations due to surface crusting. With large heterogeneity in soil fertility within smallholder farming systems, blanket recommendations are of limited value.
Using the Quantitative Evaluation of Fertility of Tropical Soils (QUEFTS) model calibrated for finger millet, balanced fertiliser requirements for a target millet yield of 2000 kg ha–1 was estimated at 83 kg N ha–1 and 52 kg P ha–1 and 56 kg K ha–1 for the sandy loam soils of Chelekura village and 64 kg N ha–1 and 31 kg P ha–1 and 40 kg K ha–1 for the sandy clay loam soils in Onamudian village. Targeting nutrient management options can result in larger benefits from nutrient management interventions and specific attention can be afforded to specific constraints to avoid wastage of resources. Combining organic resources and mineral fertilisers is needed for higher crop yields and nutrient use efficiencies. However, the SOC thresholds for higher mineral fertiliser use efficiencies need to be determined for different soil types (silt + clay) and crops as well as making farm/ system scale reconfigurations of cropping systems that will enhance efficiency in resource use. Supportive policy frameworks should be put in place to enhance investment in soil fertility management and thus increase food production.