From Food Production to Food Security - Global Environmental ...

levels, but as they stress the influence of weather and climate, and their basis in observedrelationships, large-area crop models do not currently simulate the non-climatic determinantsof crop yield. These non-climatic stresses contribute to the yield gap (Challinor et al., 2009),i.e. the gap between potential and actual yields. However, as such models do not encompasschanges in the proportion of land under cultivation, it is not possible to estimate how regionalproduction will actually change. Certainly, reliable information is needed on plot-levelresponses to environmental stresses that can be scaled up geographically, but not in isolationfrom the other major regional drivers of food systems. Coupling models at different spatiallevels from plot to region allows the study of interactions and feedbacks among biophysicaland social components at different levels. There is therefore a need to design interdisciplinaryresearch that starts with GEC objectives at a regional level, and to build systems that facilitatebetter understanding of these interactions and feedbacks. The suit of ‘point’ (or plot-level)crop models now available (e.g. DSSAT, APSIM, SUCROS) provide a valuable foundationfor such work. Regional-level studies can be greatly facilitated, and very useful informationprovided to social and economic models, when the point models are integrated withdownscaled climate model results.Other modelling studies at regional level address how ‘mega environments’ for major cropswill change (e.g. for wheat, Ortiz et al., 2008b); and how the biogeography of major andlocally important crops, and crops’ wild relatives will be affected (e.g. Jarvis et al., 2008).Cross-scale and cross-level interactions are not, however, generally included in modellingstudies, other than where a spatial scale issue has direct relevance, as is increasingly the casefor multi-scale scenario studies (Ingram and Izac, 2010: Paper 6).In addition to considering ‘up-scaling’ research on food production, there is a need to alsoconsider research at more integrated levels for other aspects of the food system. Food storageis another key determinant of food security, and is especially important during times of stress.It is, however, a complex issue, crossing a number of levels on spatial, temporal andjurisdictional (and possibly other) scales. While research has addressed the issue of strategicfood reserves at village level (e.g. Mararike, 2001) and national level (e.g. Olajide andOyelade, 2002), there is insufficient research into how best to establish long-term foodreserves at regional level. These could be a highly effective means of coping with impacts ofmajor droughts or other stresses that manifest at the regional level, but the issues are oftenhighly charged politically and progress can be slow. For instance, since the 1980s, SADC hasconsidered the establishment of a strategic food reserve to deal with the growing frequency ofnatural disasters. Early proposals were based on considerations of enough physical maizestock for 12 months’ consumption, but the SADC Council of Ministers have only recentlyagreed that the food reserve proposal should be revisited and should include consideration ofboth a physical reserve and a financial facility, supporting the notion of enhancedintraregional trade (Drimie et al., 2011).Other food system activities such as food distribution and logistics and consumption patternsalso warrant further analysis at regional level. An example of an initial analysis of currentknowledge and future research needs of all the major activities of the European food system61