This paper examines a number of agronomic field experiments in different regions of sub-Saharan Africa to assess the associated variability in the efficiencies with which applied and available nutrients are taken up by crops under a wide range of management and environmental conditions. We consider N and P capture efficiencies (NCE and PCE, kg uptake kg(-1) nutrient availability), and N and P recovery efficiencies (NRE and PRE, kg uptake kg(-1) nutrient added). The analyzed cropping systems employed different soil fertility management practices that included (1) N and P mineral fertilizers (as sole or their combinations) (2) cattle manure composted then applied or applied directly to fields through animal corralling, and legume based systems separated into (3) improved fallows/cover crops-cereal sequences, and (4) grain legume-cereal rotations. Crop responses to added nutrients varied widely, which is a logical consequence of the wide diversity in the balance of production resources across regions from arid through wet tropics, coupled with an equally large array of management practices and inter-season variability. The NCE ranged from 0.05 to 0.98 kg kg(-1) for the different systems (NP fertilizers, 0.16-0.98; fallow/cover crops, 0.05-0.75; animal manure, 0.10-0.74 kg kg(-1)), while PCE ranged from 0.09 to 0.71 kg kg(-1), depending on soil conditions. The respective NREs averaged 0.38, 0.23 and 0.25 kg kg(-1). Cases were found where NREs were > 1 for mineral fertilizers or negative when poor quality manure immobilized soil N, while response to P was in many cases poor due to P fixation by soils. Other than good agronomy, it was apparent that flexible systems of fertilization that vary N input according to the current seasonal rainfall pattern offer opportunities for high resource capture and recovery efficiencies in semi-arid areas. We suggest the use of cropping systems modeling approaches to hasten the understanding of Africa's complex cropping systems.
