No abstract

No abstract

Farmers' perceptions of soil fertility change were compared with observations on soil quality changes and crop performance in soils from a chronosequence representing a range of soil ages since conversion from forest to cropland (0 to 57 years). A majority of the farmers, 92 per cent, had observed a decline in soil fertility on their land. Farmers use crop yield, indicator plants, soil softness and soil colour to judge soil fertility. They identified 11 plants that they used to indicate high soil fertility and four plants that they used to indicate low soil fertility. There was a strong correlation (r = 0·96) between soil organic matter content (loss on ignition) and farmers' ranking of soil fertility based on colour and softness of soil samples from the chronosequence. The biotest experiment with maize showed an exponential decline in biomass production along the chronosequence, confirming the results of farmers' soil fertility ranking. In the biotest, total soil N predicted produced biomass well (r 2 = 0·95), whereas the relationship with soil available P (Olsen) was less obvious. Among the eight analysed plant nutrients in the maize leaves, N content was found to correlate best with biomass production (r 2 = 0·94). We conclude (i) that there is good agreement between farmers' knowledge and scientific indicators of soil fertility and (ii) that the major reason for declining soil fertility in Beseku is the decrease in N mineralization over time. Interventions should focus on supporting farmers to implement a diversified nutrient management strategy that can maintain or increase long-term productivity of the soil.

No abstract

No abstract