Agro-climatic classification systems for estimating the global distribution of livestock numbers and commodities

Investment in agricultural research in developing countries is being increa singly targeted at those agro-climatic zones and issues where the economic and environmental benefits may be expected to be greatest. This first requi res that the zones themselves be defined, along with information on domesti c livestock numbers and commodity output within agro-climatic zones in diff erent countries. Different methods for classifying agro-climatic zones were compared. These included methods based on estimated length of growing peri od (LGP) using rainfall and temperature data, the ratio of precipitation to potential evapotranspiration (PET), and on more detailed agronomic models, remote sensing data and land use information. Zonation based on LGP has al ready been linked to existing national livestock data. By defining agro-cli matic zones and relating concentrations of livestock populations to those o f humans, it is possible to make realistic estimates of livestock populatio ns and the production of livestock commodities for most developing countrie s. Detailed agro-climatic analyses. of Mainland East Asia and Sri Lanka hav e recently been undertaken using the GROWEST agronomic model. Using this mo del as the basis of agro-climatic classification appears to be significantl y superior, particularly in temperate environments, to approaches based sol ely on LGP. Different ways of subdividing countries and continents into agr o-climatic or agro-ecological zones (AEZs) are reviewed in this paper. In a ddition, we show how the numbers of production and commodities from domesti c livestock can be allocated to such zones. We also indicate how some of th is information can be applied. (C) 2001 Elsevier Science Ltd. All rights re served.