The concentration of native available N in tropical soils fluctuates c
onsiderably in response to seasonal changes in soil water potential. S
uch fluctuation reflects the net effect of inputs of N from mineraliza
tion, fertilizers and the atmosphere, and removal by plant uptake, imm
obilization, leaching and gaseous losses. The greatest concentrations
normally occur during the transition between the dry and wet seasons.
In East-Africa, up to 184 kg mineral N ha(-1) has been measured in the
0-40 cm soil layer and in Trinidad, 143 kg ha(-1) was found in the 0-
10 cm layer. Release and accumulation of mineral N occur as a result o
f the influence of soil water potential on microbial activity. This is
due to changes in microbial motility, solute diffusion, microbial sur
vival and the release of protected organic matter. A quantitative unde
rstanding of these processes should increase the efficiency of use of
this valuable N resource by crops. Current methods of forecasting mine
ralization under field conditions include measurement of the soil mine
ralization potential, the release of N from seasonal inputs of litter
and model predictions. Litter quality is important. Its composition, i
n particular its nitrogen, lignin and soluble polyphenol content has a
major impact on its N mineralization rate. Crop uptake, gaseous and l
eaching losses decrease the concentration of soil mineral N during the
wet season. These losses are important under moist tropical condition
s. For example, at Port Harcourt and Ibadan in Nigeria, leaching losse
s were large in spite of NO3- adsorption which decreased the depth of
NO3- leaching relative to through-flow. To minimise these losses, it i
s essential to synchronise plant nutrient demand with supply by minera
lisation. This is particularly important at the start of the tropical
rainy season when high rates of mineralisation often in excess of the
relatively low levels of crop demand, are observed. Fertilizer recomme
ndation, the time table for cropping and the farming system used there
fore need to take into account the seasonal availability of N. The CER
ES model simulates crop growth and development and the N-cycle. As dev
elopment and validation continue, such models should provide a strong
basis for better soil, crop and fertilizer management practices. A bet
ter understanding of the processes should provide a strong basis for f
uther development of such models.