A participatory on-farm fertiliser trial was carried out in 1999 and 2000 to determine area
specific fertiliser rates for different red soils of poor inherent fertility classified by
farmers as ‘kossi’ and ‘dimile’ for wheat production. ‘Kossi’, mostly located around
homesteads are the most fertile soil due to the accumulation of organic matter, but
represents the smallest share of the total cultivated land. It is cropped mostly without the
application of inorganic fertiliser. ‘Dimile’ soils representing the largest share of the
cultivated area are characterised by low soil fertility as a result of continuous cultivation
and soil erosion. It is a worn out soil whose productive capacity is almost always hinged
to adequate application of organic and inorganic fertilisers. For the reasons mentioned
above, farmers demanded the blanket recommendation, 60-60 N-P2O5 kg ha-1
, be
modified in tune with the different soil types as classified by the farmers.A factorial
combination of two wheat varieties (HAR 604 and local) and four selected rates (0-0, 9-23,
18-46 and 41-46) of NP fertiliser for 'kossi' and 5 levels (0-23, 18-46, 41-46, 64-46 and 73-
69 kg N-P2O5 ha-1) for 'dimile' soils were laid out in RCBD with two replications. Wheat
grain yields were significantly (P<0.05) affected by the application of NP fertiliser in 1999
and in 2000 (P<0.01) on ‘dimile’ soils but significantly (P<0.05) only in 1999 on ‘kossi’
soils. The application of 18-46 and 41-46 kg N-P2O5 ha-1 on ‘kossi’ soils increased wheat
grain yield by about 24 % and 34%, respectively compared with the untreated plots. While
on ‘dimile’ soils mean grain yields of wheat were increased by about 37 %, 76 % and 77 %
with the respective application of 41-46, 64-46 and 73-69 kg N-P2O5 ha-1 compared with the
lowest NP rate (9-23 kg N-P2O5 ha-1). Similarly, plant heights, spike size, number of grains
per spike and total biological yield were positively influenced due to NP fertiliser
application on dimile soils. In addition, a partial budget analysis conducted using the
1999/2000 and 2000/2001 annual average wheat grain prices of 1.95ETB/kg and 1.32
ETB/kg showed that the use of 64-46 kg N-P2O5 ha-1 gave a marginal rate of return of 331
% and 179 %, respectively. This is well above the minimum acceptable marginal rate of
return of 100 %. Hence, the application of 64-46 kg N-P2O5 ha-1 as urea and DAP for
bread wheat production by smallholders could be recommended on dimile soils. But on
kossi soils a lower rate (9-23 kg N-P2O5 ha-1) or even without fertiliser would be
profitable, as the soil is rich in nutrients.