Influence of nutrient management strategies on variability of soil fertility, crop yields and nutrient balances on smallholder farms in Zimbabwe

An improved understanding of soil fertility variability and farmers' resource use strategies is required for targeting soil fertility improving technologies to different niches within farms. We measured the variability of soil fertility with distance from homesteads on smallholder farms of different socio-economic groups on two soil types, a granite sand and a red clay, in Murewa, northeast Zimbabwe. Soil organic matter, available P and CEC decreased with distance from homestead on most farms. Soil available P was particularly responsive to management, irrespective of soil type, as it was more concentrated on the plots closest to homesteads on wealthy farms (813 mg kg(-1)), compared with plots further from homesteads and all plots on poor farms (2-6 mg kg(-1)). There was a large gap in amounts of mineral fertilizers used by the wealthiest farmers (> 100 kg N and > 15 kg P per farm; 39 kg N ha(-1) and 7 kg P ha(-1)) and the poorest farmers (< 20 kg N and < 10 kg P per farm; 19 kg N ha(-1) and 4 kg P ha(-1)). The wealthy farmers who owned cattle also used large amounts of manure, which provided at least 90 kg N and 25 kg P per farm per year (36 kg N ha(-1) and 10 kg P ha(-1)). The poor farmers used little or no organic sources of nutrients. The wealthiest farmers distributed mineral fertilizers evenly across their farms, but preferentially targeted manure to the plots closest to the homesteads, which received about 70 kg N and 18 kg P per plot (76 kg N ha(-1) and 21 kg P ha(-1)) from manure compared with 23 kg N and 9 kg P per plot on the mid-fields (26 kg N ha(-1) and 10 kg P ha(-1)), and 10 kg N and 1 kg P per plot (and ha-1) on the outfields. Crop allocation on the homefields was most diversified on the wealthiest farms where maize was allocated 41% of the area followed by grain legumes (24%) and paprika (21%). Maize was allocated at least 83% of the homefields on farms with less access to resources. All the farmers invariably applied nutrients to maize but little to groundnut. Maize grain yields were largest on the homefields on the wealthy farms (2.7-5.0 t ha(-1)), but poor across all fields on the poor farms (0.3-1.9 t ha(-1)). Groundnut grain yields showed little difference between farms and plots. N and P partial balances were largest on the wealthy farms, although these fluctuated from season to season (-20 to +80 kg N per farm and 15-30 kg P per farm; average 21 kg N ha(-1) and 8 kg P ha(-1)). The partial balances on the wealthy farms were largest on the homefield (20-30 kg N and 13 kg P per plot; > 26 kg N ha(-1) and > 13 kg P ha(-1)),but decreased to 10-20 N and 6-9 kg P per plot (< 20 kg N ha(-1) and 13 kg P ha(-1)) in mid-fields and -7 to +10 kg N and -1 to +1 kg P per plot (< 10 kg N ha(-1) and < 2 kg P ha(-1)) in the outfields. N and P balances differed little across plots on the poor farms (-2 to +4 kg per plot; -5 to +4 kg ha(-1)) due to limited nutrients applied and small off-take from small harvests. This study highlights the need to consider soil fertility gradients and the crop and nutrient management patterns creating them when designing options to improve resource use efficiency on smallholder farms. (c) 2006 Elsevier B.V. All rights reserved.