AgMIP, Opportunities and Challenges of a Global Framework in Regional Southern Africa.

World population is expected to increase, particularly fast over the African continent. Increasing food production is thus vital in order to match the future food demand and work towards a self-sustained Africa. The Agricultural Model Intercomparison and improvement Project (AgMIP - www.agmip.org) is proposing an innovative integrated assessment of agricultural systems subject to future conditions. The project is designed to create interactions between future climate projections, crop models and socio-economic situations so as to provide more holistic impact assessment, and explore representative agricultural pathways for systems to adapt in the future. The global project supported 2 regional projects in southern Africa; one focusing on Botswana, Namibia, Lesotho, South Africa and Swaziland, and the other focusing on Malawi, Mozambique, South Africa and Zimbabwe. We would like to present and discuss the AgMIP first phase achievements: a solid methodological framework and preliminary results, and giving special emphasise to challenges and opportunities that offer the AgMIP global initiative for regional southern Africa. Southern Africa is predominantly a semi-arid region with high rainfall variability, characterized by frequent droughts and floods. The climate is highly seasonal with hot, wet summers and cool, dry winters (except for south western South Africa). The region is highly diverse in terms of cropping and socio-economic systems. The agricultural system is 'dual', with well-developed commercial farming systems on large farms, and subsistence-based production systems on small farms. The yields obtained by the traditional smallholder farmers are very small in comparison to the world's standards or the commercial farmer production levels in the region. AgMIP framework provides a consistent approach to connect climates, crop models and socio-economic conditions, hence enabling national, regional and global impact assessments, comparisons and improvements. The best available historical records of 1980-2010 were gathered and used in combination with gridded data where needed. The future climate projections reflect 20 existing GCMs, 5 have been used in southern Africa at this stage. AgMIP has linked with 19 crop models (number varying depending on the crop considered), the DSSAT and APSIM crop models have been used at this stage. Maize and sugarcane were modelled in South Africa, mixed crop-livestock systems in Zimbabwe, and subsets of those in the other countries. The agricultural systems are simulated under different adaptation conditions in order to assess the independent and combined socio-economic impacts. At this stage 3 questions were addressed: (i) what is the sensitivity of current agricultural systems to climate change, (ii) what is the impact of future agricultural systems to climate change and (iii) what benefits adaptation could bring? The project results concur with IPCC headlines and scale of changes. Temperatures over southern Africa are increasing with high confidence, while rainfall projections are uncertain. Despite progress made to face climate variability, those future conditions are expected to stress water availability. In lesser extent, the project revealed a potential delay in on-set of the rainy season, which could lead into shorter rainy season and further stress water availability. Appropriate measures exist and have a beneficial impact to mitigate and adapt agricultural systems to climate change stress. How to go now from those recommendations to finding the measures that are both applicable at small farms scale and steer sufficient governmental support?