Ways Forward? Technical Choices, Intervention Strategies and Policy Options 359these are tightly constrained by limited access to markets, poor veterinary services
and the high cost of mineral fertilizer. In Zimbabwe, the two locations demon-
strated marked differences in terms of their productivity and market access. Man-
gwende is in a relatively high-potential area, with reasonable rainfall although poor
soils, and is well connected to markets in Harare and Chitungwiza. By contrast,
Chivi is in a semi-arid zone where frequent droughts reduce productive potential.
Although there are good road connections to Masvingo to the north and South
Africa to the south, there are fewer market opportunities.
The research on differences in soil-fertility management between socioeco-
nomic groups has helped identify households for which soil-nutrient decline con-
stitutes a serious problem. Poorer households in all sites faced particular difficulties
in maintaining the productivity of their farmland, due to very limited supplies of
livestock dung and poor access to credit and inorganic fertilizer. Such difficulties
are a reflection of their limited assets and capacity to mobilize resources more gen-
erally. Poor farmers in Dilaba, Mali and Chivi, Zimbabwe were additionally con-
strained by the disappearance of fallow land and lack of access to credit. This
diversity has major implications for design of technical options, extension
approaches and policy frameworks for improving soil-fertility management.
Farmers’ Concern for Soil-Fertility Management: The
Broader Context
As the case studies have shown, the capacity of different farmers to invest in
improving soil-fertility management depends on a range of factors which operate
at different levels. For the household, these factors include access to labour, live-
stock, land, capital and cash. In all sites, better-off farmers were much better able
to invest in larger-scale use of organic and inorganic sources of nutrients. For
example, in some sites, having some means of transport, such as a cart, is also
critically important in allowing large quantities of biomass to be moved to and
from homestead, kraal and crop lands. But there was also great diversity in how
such inputs are used, since farmers do not spread them universally across all fields
and crops. Instead, nutrients tend to be focused on smaller plots of land where
higher-value crops (such as vegetables) are grown. In addition, rotation of land
spreads fertility inputs over a series of years. For cotton farmers in southern Mali,
nutrient inputs were positive for cotton in the year in which cotton is cultivated,
but if account is taken of the subsequent two years of rotation with maize, millet, and
sorghum, then over three years there is a net outflow of nutrients. Nutrient losses are
much more significant on fields where lesser-value crops are grown and which are
not a high priority for farmers. However, even here, farmers often make very effective
use of the very limited quantities of nutrients available through careful placing and
timing of such inputs in relation to rainfall and crop development. In all sites, farm-
ers were experimenting with very small additions of mineral fertilizer at sowing or