Mining of Nutrients in African Soils Due to Agricultural Intensification - Principles of Sustainable Soil Management in Agroecosystems - page 406

inputs, where low income leads to low inputs, low yields, and so on. While the num-
ber and distribution of African subsistence farmers caught in the poverty trap is not
well known, the widespread incidence of the problem in various countries can be
surmised from the calculation of soil nutrient balances using models and available
national level data as well as data on other scales.

15.2 RATES OF NUTRIENT DEPLETION

15.2.1 M ethodology
Nutrient balance calculations can be made at a range of scales from plot to farm to
region to nation and beyond. At the global level, Grote et al. (2005) and Craswell et
al. (2003) assessed the net flows of nutrients in commodities transported internation-
ally in trade. The total Nitrogen, Phosphorus and Potassium (NPK) in the different
food commodities can be calculated from the amount of the commodities multiplied
by average values for their N, P, and K contents. One problem is that these studies
have not been able to source and include data on plantation crops, which commonly
receive much of the fertilizer used in many African countries. Furthermore, this
approach does not provide information about within-country nutrient flows, which
in an increasingly urbanized Africa may largely be in the direction from rural to
urban areas (Craswell et al. 2003; Drechsel and Zimmermann 2005). Nevertheless,
whether derived from nationally or regionally aggregated data, the international
trade balance provides an indication of the massive scale of NPK movement in food
commodities globally.
At the national and farm scale, NPK mining can be calculated from the following
(Henao and Baanante 1999):

Rn t = ∑ t ( AP t + AR Δ t - RM Δ t - L Δ t ) (15.1)

where Rn t is the amount of nutrient coming from organic and inorganic sources
remaining available in the soil after a period of time t ; AP t is the inherent soil nutri-
ent available at time t ; AR Δ t is the total of the nutrient in mineral and organic com-
pounds added or returned to the soil during time interval Δ t ; RM Δ t is an estimate
of the plant nutrients removed in crop harvest and residues during time interval Δ t ;
and L Δ t is the amount of nutrient lost via various pathways during time interval Δ t .
The nutrient balance calculation generally is based on estimates of the extent of
five inputs—mineral fertilizers, organic inputs, BNF, wet/dry deposition, and sedi-
mentation—and of five losses, such as crop product, crop residues, gaseous losses,
leaching, and soil erosion. Partial nutrient balances can be estimated from data
on fertilizer use and crop yields, utilizing databases such as FAOSTAT. Sheldrick
et al. (2002) combined FAOSTAT data with nutrient efficiency ratios to develop
nutrient audits without resorting to more detailed modeling. More detailed anal-
ysis tools such as the International Fertilizer Development Center (IFDC) model
and NUTMON (Henao and Baanante 1999; Stoorvogel et al. 1993) applied at the
farm or supranational level ( Figure 15.3 ) require estimates of transfer functions for
soil erosion and other components of nutrient losses and gains. Of the three major