Spatial variation of crop yield response to climate change in East Africa

https://doi.org/10.1016/j.gloenvcha.2008.08.005Get rights and content

Abstract

There is general consensus that the impacts of climate change on agriculture will add significantly to the development challenges of ensuring food security and reducing poverty, particularly in Africa. While these changes will influence agriculture at a broad scale, regional or country-level assessments can miss critical detail. We use high-resolution methods to generate characteristic daily weather data for a combination of different future emission scenarios and climate models to drive detailed simulation models of the maize and bean crops. For the East African region, there is considerable spatial and temporal variation in this crop response. We evaluate the response of maize and beans to a changing climate, as a prelude to detailed targeting of options that can help smallholder households adapt. The results argue strongly against the idea of large, spatially contiguous development domains for identifying and implementing adaptation options, particularly in regions with large variations in topography and current average temperatures. Rather, they underline the importance of localised, community-based efforts to increase local adaptive capacity, take advantage of changes that may lead to increased crop and livestock productivity where this is possible, and to buffer the situations where increased stresses are likely.

Introduction

Climate change poses a serious and continuing threat to development. Scholes and Biggs (2004), referring to Sub-Saharan Africa as the food crisis epicentre of the world, conclude that projected climate change during the first half of the twenty-first century will make this situation worse. Climate change will add burdens to those who are already poor and vulnerable (IPCC, 2007). At the same time, agriculture in Sub-Saharan Africa will continue to play a crucial role through its direct and indirect impacts on poverty, as well as in providing an indispensable platform for wider economic growth that reduces poverty far beyond the rural and agricultural sectors (DFID, 2005).

Overall, crop yields in Africa may fall by 10–20% to 2050 because of warming and drying, but there are places where yield losses may be much more severe, as well as areas where crop yields may increase (Jones and Thornton, 2003). Many developing countries in Africa are seen as being highly vulnerable to climate variability and change (Slingo et al., 2005), in part because they have only a limited capacity to adapt to changing circumstances (Thomas and Twyman, 2005). A high reliance on natural resources, limited ability to adapt financially and institutionally, low per capita Gross Domestic Product (GDP) and high poverty, and a lack of safety nets mean that the challenges for development are considerable (Thomas and Twyman, 2005).

Despite this, there is considerable and increasing activity on the part of development agencies and governments to come to grips with these challenges, including the development of appropriate adaptation strategies. Given the scale of the problems involved, development agencies could greatly benefit from information that quantifies the impacts that may arise, so that development assistance can be targeted in appropriate places, depending on the development objectives that are being pursued. There are, however, considerable knowledge gaps concerning the interacting and multiple stresses on the vulnerability of the poor in Africa. There is a critical need to undertake analytical assessments of vulnerability to increased climatic variability and climate change, to better understand the implications for poverty reduction as well as to be able to assess adaptation initiatives (Huq and Reid, 2005, Nyong, 2005).

Some vulnerability mapping has been carried out for the continent, including a preliminary attempt to help locate critical research activities and identify areas that may be severely affected by climate change and where agricultural populations are already vulnerable, environmentally and socially (Thornton et al., 2006). These “hotspots” include the mixed arid–semiarid systems in the Sahel, arid–semiarid rangeland systems in parts of eastern Africa, the coastal regions of eastern Africa, and many of the drier zones of southern Africa, for example. However, there is a need for better and multi-level vulnerability analyses to help target adaptation work, because there is considerable spatial heterogeneity of not only the impacts of climate change but also households’ access to resources, poverty levels, and ability to cope. There is also a critical need for better understanding of the information needs of decision-makers concerning climate change and variability, and tools need to be designed and implemented that can help meet these needs. While coping with climate change and variability is not a new challenge for African farmers, existing coping mechanisms may not be up to the challenges posed by the changes projected. The situation is made more complex by the fact that while we know something about the changes possible in climate in future years, we know much less about likely changes in climate variability and the probabilities of extreme events. Particularly for vulnerable people who are highly dependent on natural resources for their livelihoods, the impacts of extreme events, particularly the lower tails of probability distributions, may have a social and economic importance that far outweighs their apparent probability of occurrence. Increases in mean precipitation are likely to be associated with increases in variability (IPCC, 2007). In several regions, including parts of Africa, inter-annual climatic variability is strongly related to El Niño-Southern Oscillation (ENSO) events, and thus will be affected by changes in ENSO behaviour (Conway et al., 2007). More work is needed on the issue of changing weather variability in the future and what its impacts may be.

Even without considering changes in weather variability as a result of climate change, the patterns of crop yield impacts may be highly heterogeneous. Much of the agricultural impacts work to date has been carried out at relatively low spatial resolution, often at the scale of the globe, region, or country (for example, Parry et al., 2004, Cline, 2007, Lobell et al., 2008). Particularly for organizations that work with a “pro-poor” mandate in developing countries, in addition to the relatively broad-brush information that such studies provide, there is a need for more detailed information on the impacts of climate change on agricultural systems, so that effective adaptation options can be appropriately targeted. In this paper, we build on previous work (Jones and Thornton, 2003) that identified possible country-level changes in maize production to the middle of this century. Here we investigate in more detail the different types of crop response to climate change as represented by a combination of two climate models and two contrasting greenhouse-gas emission scenarios. For the East Africa region, we analyse the spatial differences in simulated main-season maize and secondary-season Phaseolus bean yields to 2050, and attempt some simple characterisation of crop response. The object of doing this is to assess the possibility of using such information for preliminary targeting of adaptation options at relatively high resolution. The next section of the paper outlines the methods used. Section 3 presents some simulation results, and the discussion in Section 4 sets out what we see as the major implications of the results for climate change impact assessment and adaptation targeting work in an African context.

Section snippets

Methods

A block diagram of the methods used is shown in Fig. 1, made up of strands dealing with climate and weather data, soils data, crop and crop management data, and the crop models used. These elements are described below.

Results

Fig. 2A shows average simulated maize yields when grown in the primary season under current climatic conditions, together with the coefficient of variation of this simulated yield (30 replicates). Similarly, Fig. 2B shows the average simulated bean yields in the secondary season where this is feasible and their coefficient of variation. As noted above, these maps do not show where maize and beans are currently grown; they merely show simulated yields in areas where they could be grown.

Discussion

The results above show that crop yield responses to the changing rainfall amounts and patterns and the generally increasing temperatures projected by GCMs are heterogeneous. They may vary by crop type, by location, and through time. Results also indicate that under the four GCM–scenario combinations considered, the aggregate production decreases are projected to be rather modest to 2050. These aggregate production changes, however, hide a large amount of variability, and under the higher

Conclusions

Information on the nature, extent and location of the impacts of climate change on households in Africa that are particularly dependent on natural resources for their livelihoods is crucial if appropriate adaptation options are to be designed and implemented to deal with changing crop and livestock production potential. These impacts may be highly variable across space and through time, as a result of the interactions between temperature increases and shifts in rainfall patterns and amount. In

Acknowledgments

We are grateful to Simon Anderson, Russ Kruska and Tom Owiyo for inputs. We thank two anonymous referees for helpful comments on an earlier version of the paper. We acknowledge funding from the US National Science Foundation under NSF awards numbers 0119821, ‘BE/CNH: Climate and Land Use Change Processes in East Africa’, and 0308420, ‘BE/CNH: An Integrated Analysis of Regional Land–Climate Interactions’, and the Michigan State University Foundation. Views expressed here are the authors’ own, as

References (55)

  • B. Wafula

    Applications of crop simulation in agricultural extension and research in Kenya

    Agricultural Systems

    (1995)
  • ASARECA, 2005. Fighting poverty, reducing hunger and enhancing resources through regional collective action in...
  • N.H. Batjes et al.

    Potential emissions of radiatively active trace gases from soil to atmosphere with special reference to methane: development of a global database (WISE)

    Journal of Geophysical Research

    (1994)
  • K.J. Boote et al.

    The CROPGRO model for grain legumes

  • D. Conway et al.

    GCM simulations of the Indian Ocean dipole influence on East African rainfall: Present and future

    Geophysical Research Letters

    (2007)
  • W.R. Cline

    Global Warming and Agriculture: Impact Estimates by Country

    (2007)
  • Department for International Development (DFID), 2005. DFID’s Draft Strategy for Research on Sustainable Agriculture...
  • Donatelli, M., Campbell, G.S., 1997. A simple model to estimate global solar radiation. PANDA Project, Subproject 1,...
  • FAO (Food and Agriculture Organization of the United Nations), 1978. Report on the AgroEcological Zones Project, vol....
  • FAO (Food and Agriculture Organization of the United Nations), 1995. Digital Soil Map of the World and Derived Soil...
  • R.A. Fisher et al.

    Statistical Tables for Biological, Agricultural and Medical Research

    (1967)
  • R.J. Hijmans et al.

    Very high resolution interpolated climate surfaces for global land areas

    International Journal of Climatology

    (2005)
  • G. Hoogenboom et al.

    BEANGRO, a process oriented dry bean model with a versatile user interface

    Agronomy Journal

    (1994)
  • Huq, S., Reid, H., 2005. Climate change and development: consultation on key researchable issues. IIED, London, UK....
  • Hutchinson, M.F., 1989. A new objective method for spatial interpolation of meteorological variables from irregular...
  • ICASA, 2007. The International Consortium for Agricultural Systems Applications website. Online at...
  • IPCC (Intergovernmental Panel on Climate Change), 2000. Emission scenarios, summary for policy makers. Online at...
  • Cited by (310)

    View all citing articles on Scopus
    View full text