Live fences – a hidden resource of soil fertility in West Kenya

In the highlands of Western Kenya, intensified land use combined with low use of mineral fertilizers at farm level is driving soil nutrient depletion and declining yield levels. All farms and individual field crop plots are surrounded by life fences and the area under such hedge structure is estimated to cover 3–5% of the total farmland area. The land below the hedges is not tilled and may receive nutrient and carbon inputs by the occasional addition of field crop residues and litter fall. We hypothesized that the areas covered by live fences represent not only an important land resource, but are also largely untapped sites characterized by high soil fertility in an otherwise largely degraded environment. We characterized physico‐chemical attributes of top soils collected on farmland and from adjacent live fences composed of five different fencing species in representative environments of Kakamega district (sandy Acrisol and bimodal rainfall distribution vs. clay Ferralsol and monomodal rainfall distribution). In addition, the maize production potential of these soils was assessed in a supplementary pot experiment. Concentrations and total amounts of soil C, N, N supplying capacity, exchangeable K, as well as aggregate stability tended to be higher in hedge structures than in field crop soil. The amount of labile (permanganate‐oxidizable) carbon and the carbon management index in fence lines on Acrisol were similar to those of the adjacent Kakamega rain forest reserve, while they were reduced by > 50% in crop fields. These trends were reflected in the biomass and N and K uptake by maize in potted soil. Effects were generally larger in Acrisol than in Ferralsol and were most pronounced with Tithonia diversifolia. We conclude that live fences are soil fertility hotspots that may be valorised in the future by replacing the generally unproductive fencing species with economical perennial crops.