Effects of climate variability and climate change on crop production in southern Mali

• The number of dry days/season is the key determinant for cotton yield.
• On average 1 mm of rain is converted into 2 kg of cotton.
• Rainfall is the key limiting factor of maize yield in the driest districts.
• Daily minimum air temperature increase of 0.06 °C per year during the period from 1965 till 2005.
• An increase of 0.08 °C of maximum temperature leads to a cotton yield loss of 24 kg ha−1.

In West Africa predictions of future changes in climate and especially rainfall are highly uncertain, and up to now no long-term analyses are available of the effects of climate on crop production. This study analyses long-term trends in climate variability at N’Tarla and Sikasso in southern Mali using a weather dataset from 1965 to 2005. Climatic variables and crop productivity were analysed using data from an experiment conducted from 1965 to 1993 at N’Tarla and from a crop yield database from ten cotton growing districts of southern Mali. Minimum daily air temperature increased on average by 0.05 °C per year during the period from 1965 to 2005 while maximum daily air temperature remained constant. Seasonal rainfall showed large inter-annual variability with no significant change over the 1965–2005 period. However, the total number of dry days within the growing season increased significantly at N’Tarla, indicating a change in rainfall distribution. Yields of cotton, sorghum and groundnut at the N’Tarla experiment varied (30%) without any clear trend over the years. There was a negative effect of maximum temperature, number of dry days and total seasonal rainfall on cotton yield. The variation in cotton yields was related to the rainfall distribution within the rainfall season, with dry spells and seasonal dry days being key determinants of crop yield. In the driest districts, maize yields were positively correlated with rainfall. Our study shows that cotton production in southern Mali is affected by climate change, in particular through changes in the rainfall distribution.