With that in mind, a team from Stanford University’s Food Security and Environment Program, Lawrence Livermore National Laboratory and the US National Center for Atmospheric Research has used statistical models of climate and crop response to quantify which regions are likely to suffer the most. They reported on their work in Science.
"The majority of the world’s one billion poor depend on agriculture for their livelihoods," said David Lobell of Stanford University. "Unfortunately agriculture is also the human enterprise most vulnerable to changes in climate. Understanding where these climate threats will be greatest, for what crops and on what time scales, will be central to our efforts at fighting hunger and poverty over the coming decades."
Lobell’s colleague Rosamond Naylor believes that the study is particularly timely as the international donor community is starting to invest in agricultural productivity in the developing world again. "Our study will help show where these investments might be the most worthwhile," she said. "We know we can’t do everything right away, but this helps us know where to start."
Lobell, Naylor and colleagues found that southern Africa and southern Asia are likely to be "hunger hotspots". Southern Africa could lose more than 30% of its sweetcorn production – the main crop in the region – by 2030, while southern Asia could see losses of 10% or more in its staple crops of millet, sweetcorn and rice.
Model outcomes
To perform their study, the researchers used statistical crop models along with climate projections for 2030 obtained from 20 general circulation climate change models. The crop models helped predict the response of the plants to changes in temperature and rainfall. The team analysed 12 food-insecure regions, including much of Asia, sub-Saharan Africa, the Caribbean and Central and Latin America, choosing 2030 as a timeframe as largescale investment in agriculture typically takes 15 to 30 years to realize full returns.
"To identify which crops in which regions are most under threat by 2030, we combined projections of climate change with data on what poor people eat, as well as past relationships between crop harvests and climate variability," said Lobell.
Each of the 20 climate models, which were used in the World Climate Research Programme’s Coupled Model Intercomparison Project phase 3 (WRCP CMIP3), gave a different result. All the models predicted a rise in temperature, with the median projection of average temperature change being around 1°C in most regions. Few models projected less than 0.5°C warming while some predicted as much as 2°C.
The results for precipitation were more mixed – all regions had at least one model that projected an increase in precipitation and at least one projecting a decrease. The median projections ranged from about –10% to +5%. General circulation models typically indicate that precipitation from December to February will decrease in South Asia and Central America and increase in East Africa, while precipitation from June to August will decrease in southern Africa, Central America and Brazil.
Uncertain times
This range of projections meant that the researchers had to include uncertainty in their predicted changes in food production. The uncertainty varied depending on the region and on the type of crop – due to different plant sensitivities to rainfall levels.
"Areas of West Africa and the Sahel stand out as regions with very high rates of food insecurity and with a very high dependence on agriculture, but also with a fair amount of uncertainty regarding climate change impacts," said Marshall Burke of Stanford. "For these regions, you get half of the climate models telling you it’s going to get wetter and the other half giving you the opposite. As a result, our study raises the potential for very bad impacts in these regions but with much less certainty than in other regions."
That raises the issue of whether an organization looking to invest in improving agriculture in developing countries should plump for the safe bets such as maize in southern Africa and rice in Southeast Asia, where all models agree that impacts will be negative, or should try to safeguard crops such as sorghum in the Sahel or millets in Central Africa, which could well suffer badly but the case is less clear-cut.
And how should that investment best be spent? The opportunities for adapting agriculture to climate change range from the relatively cheap and straightforward – such as delaying or advancing planting dates, switching crop variety or even crop type – to the more expensive, including expanding irrigation and developing new crop varieties.
"Although we do not attempt to identify the particular adaptation strategies that should be pursued, we note that, in some regions, switching from highly impacted to less impacted crops may be one viable adaptation option," say the researchers in their paper.
They also stress the need for more research to investigate the need for adaptation, especially in regions such as Central Africa where current data sets don’t provide adequate information about climate-yield relationships, as well as calling for research on a finer spatial scale to resolve local hunger hot spots within a region. Work is also ongoing at Stanford’s Food Security and Environment Programme to investigate the impact of biofuels expansion on climate change and the world’s poor.
While the crop yield outcomes for particular regions may as yet be uncertain, it’s clear that planning and adaptation will be vital in ensuring that people have enough to eat as the century progresses, particularly if populations grow as predicted.