The legacy of drought on food security in Mali’s Sahel

Low food production plus a lack of livelihood diversification has left some communities more vulnerable than others to climate variability. They face recurring periods of malnutrition and food insecurity.

By Elisabeth Gawthrop

Farmers in Molobala Village, near Sikasso, Mali. Sikasso is in the south of the country, in the “maize/cotton/fruit” region as identified in the map below. The relationship between climate and food security is not just the direct climatic impact on production. While rainfall in Sikasso, for example, is generally more consistent, the region is not food secure and is known as a “food security paradox” because the source of insecurity is not entirely clear. Francesco Fiondella/IRI

Food production in Mali has increased dramatically since the droughts and famine of the 1970s, but year-to-year fluctuations in rainfall still affect food security in the country (see graph below). IRI scientists and colleagues at the World Food Programme recently authored a study combining climate information with household survey data that identifies regions of the country most vulnerable to food insecurity. The authors examined rainfall data along with survey results about farmers’ sources of livelihood and access to markets, as well as how much of the food farmers consumed was from their own production.

Climate and agricultural production in Mali. Time series of annual production of the main staple crops, in the solid line, and May–October seasonal precipitation, in the dashed line. National agricultural statistics are from the UN Food and Agriculture Organization. Rainfall data from GPCC is averaged over 12°N–16°N, 15°W–5°E.

In the most vulnerable region they identified — the livelihood zone labeled “millet/transhumant herding” in the Sahelian center of the country — food availability is problematic because of insufficient local production. This region, in light blue in the map below, faced one of the most dramatic changes in climate in the last few decades of the 20th century. From 1941–1970, there was rainfall sufficient for growing crops in two out of every three years, on average. From 1971–2000, however, there was only sufficient rainfall in one out of every three years. The first two maps in the three-image panel below show this shift with contour lines.

Livelihood zones in Mali as defined by the Famine Early Warning Systems Network in 2005, in color, overlaid by the 1979–2007 average yearly precipitation from GPCP with contour every 200 mm/yr.

“All of Mali felt the drought, but in this region you’re at the margin of where agriculture is sustainable, so the theory is that recurrence of drought put farmers here over the tipping point,” said Alessandra Giannini, the study’s lead author.

Although climate conditions have since improved, the farmers haven’t recovered as quickly. The authors present the argument that the persistent crisis conditions over many years, which were especially significant in the millet/transhumant herding area, eroded the wealth that is usually accumulated in livestock. In the survey, around 42% of those engaged in agriculture in this region were not involved in livestock raising. In other regions, this figure ranged from 6% to 37% (see third panel in below maps).

On top of low local food production, many farmers in this zone supplement their income in other ways that are still vulnerable to climate, such as fishing, production of garden vegetables (distinguished from production of staple and cash crops) and sale of livestock. Thus, their ability to buy food in times of drought is further limited. In contrast, farmers in other areas in the country have livelihoods that are more diversified, and thus less reliant on a good rainfall season.

These results are consistent, the authors say, with other studies identifying this region as “the geographical front-line of child malnutrition” and with the “location of stressed food security conditions in 2012.”

“The interesting thing about this study was that we could go beyond the typical work a climate scientist does when looking at food security,” said Giannini. “Usually, we just look at correlations between climate and productivity. But it isn’t just the physical climate that influences food security, and there are also climate effects at many timescales.”

For example, while climate can cause a “shock event” via severe drought over a few seasons or years, the way that longer-term, persistent drought shapes livelihoods would not show up in a simple correlation.

One of the study’s co-authors, Richard Choularton, worked for the World Food Programme during the time of the study and is now a senior associate at Tetra Tech. “I think the most important finding in our study was that the poorest people in the zones we identified as having the greatest impact from the 1970’s climate shift appear to be those who were least able to adapt,” Choularton said.

Better-off households were more diversified to include sources of food and income that weren’t as sensitive to fluctuations in climate, while the most food-insecure households depended almost entirely on rain-fed crop and livestock production.

“These households could not afford to adapt,” Choularton said. They faced what’s referred to as an adaptation deficit — the inability to manage current levels of climate risk to the extent that investing in adaptation measures to climate change becomes impossible.

“In contrast, in pastoralism livelihood zone in the drier and harsher northern Sahel, the poorest and most food insecure people have long had to diversify their sources of food and income to survive in their harsh climate,” Choularton said. While by some measures they may be more food insecure overall, in some ways they have more tools to deal with a changing climate and may be more resilient, he said.

The study also demonstrates the flexibility of IRI’s Data Library in being used for overlaying data characterized by different spatiotemporal dependence.

The shading in panels (a) and (b) is the percentage of households that produced millet at the time of the household survey. The shading in panel (c) indicates the proportion of households by livelihood zone that engage in agriculture, but not in livestock raising. (Here and in all similar plots, the northernmost “desert” livelihood zone is masked.) The contours in (a) and (b) show the number of years that it rained at least 450 mm during the May–October season (a) during the wet 1941–1970 period and (b) during the dry 1971–2000 period, contour is every 10. Note the location of the 20-year line in (a) is essentially replaced by the 10-year line in (b).

“What I liked about this work from a data point of view is that we could use the Data Library to easily put together diverse datasets, including the more standard climate data, which is at a coarse resolution and defined by grid boxes, as well as national-level crop production data and household surveys that were represented by individual points,” said Rémi Cousin.

The survey data is meant to be a snapshot from a relatively normal year, Cousin said. But because it is from just one year (2005 in this case), the study had to rely more on spatial analysis than on observations over time. This is where the Data Library’s power became especially useful. Depending on the need, Cousin said, they could crunch the data to the geographical level of interest: from grids to points to geometries like districts and livelihood zones.

Ultimately, the lesson learned from this study is that policies should promote the diversification of livelihoods into non-climate sensitive activities, the authors write. They argue that this can help counteract the “deepening of adaptation deficits among the poorest — too poor to diversify, hence adapt.”

“As the climate in the Sahel and around the world changes,” Choularton added, “this study shows that we need to pay particular attention to vulnerable households who are highly climate sensitive and do not have the means or resources to adapt to their changing climate.”

People with this kind of vulnerability might appear to be able to make ends meet, he said, but they are on the margin being eroded by climate change. “Without significant investments in livelihoods, social protection, climate risk management, resilience and adaptation, the processes of destitution, deepening of poverty, and increase in humanitarian crises we have seen in parts of the Sahel will only get worse.”

Farmers in Molobala Village, near Sikasso, Mali. Sikasso is in the south of the country, in the “maize/cotton/fruit” region as identified in the map below. The relationship between climate and food security is not just the direct climatic impact on production. While rainfall in Sikasso, for example, is generally more consistent, the region is not food secure and is known as a “food security paradox” because the source of insecurity is not entirely clear. Francesco Fiondella/IRI