Separate and combined effects of temperature and precipitation change on maize yields in sub-Saharan Africa for mid- to late-21st century

https://doi.org/10.1016/j.gloplacha.2013.02.009Get rights and content

Highlights

  • General circulation models project different changes in the wet season precipitation.

  • We identify the general circulation models with largest changes in each grid cell.

  • We separate the effects of changes in temperature and precipitation on maize yield.

  • We identify the limiting factor for maize growth and development.

  • The importance of temperature and precipitation for maize growth varies in space.

Abstract

Maize (Zea mays L.) is one of the most important food crops and very common in all parts of sub-Saharan Africa. In 2010 53 million tons of maize were produced in sub-Saharan Africa on about one third of the total harvested cropland area (~ 33 million ha). Our aim is to identify the limiting agroclimatic variable for maize growth and development in sub-Saharan Africa by analyzing the separated and combined effects of temperature and precipitation. Under changing climate, both climate variables are projected to change severely, and their impacts on crop yields are frequently assessed using process-based crop models. However it is often unclear which agroclimatic variable will have the strongest influence on crop growth and development under climate change and previous studies disagree over this question.

We create synthetic climate data in order to study the effect of large changes in the length of the wet season and the amount of precipitation during the wet season both separately and in combination with changes in temperature. The dynamic global vegetation model for managed land LPJmL is used to simulate maize yields under current and future climatic conditions for the two 10-year periods 2056–2065 and 2081–2090 for three climate scenarios for the A1b emission scenario but without considering the beneficial CO2 fertilization effect.

The importance of temperature and precipitation effects on maize yields varies spatially and we identify four groups of crop yield changes: regions with strong negative effects resulting from climate change (<− 33% yield change), regions with moderate (− 33% to − 10% yield change) or slight negative effects (− 10% to + 6% yield change), and regions with positive effects arising from climate change mainly in currently temperature-limited high altitudes (>+6% yield change). In the first three groups temperature increases lead to maize yield reductions of 3 to 20%, with the exception of mountainous and thus cooler regions in South and East Africa. A reduction of the wet season precipitation causes decreases in maize yield of at least 30% and prevails over the effect of increased temperatures in southern parts of Mozambique and Zambia, the Sahel and parts of eastern Africa in the two projection periods. This knowledge about the limiting abiotic stress factor in each region will help to prioritize future research needs in modeling of agricultural systems as well as in drought and heat stress breeding programs and to identify adaption options in agricultural development projects. On the other hand the study enhances the understanding of temperature and water stress effects on crop yields in a global vegetation model in order to identify future research and model development needs.

Introduction

Process-based crop growth models are frequently used to simulate climate change impacts on agricultural crops in sub-Saharan Africa and many studies can be found in the literature for either the whole region (Jones and Thornton, 2003, Liu et al., 2008, Thornton et al., 2011, Folberth et al., 2012) or for individual African countries (Adejuwon, 2006, Thornton et al., 2009, Laux et al., 2010). These models compute important biophysical and biochemical processes, like photosynthesis, respiration and transpiration or the dynamics of carbon and water at the leaf-level (Tubiello and Ewert, 2002, Bondeau et al., 2007) and are therefore able to simulate the effect of increasing temperatures, changing precipitation and elevated atmospheric CO2 concentrations on crop development and yields. Climate projections from general circulation models (GCMs) on air temperature, precipitation and annual atmospheric CO2 concentrations are typically used as input for these models.

For sub-Saharan Africa GCM projections agree well in the level of average temperature increases between 3 to 4 K in the 2090s compared to the 1990s in the A1b projections, with deviations between seasons and regions (Christensen et al., 2007). The likelihood that the summer average temperature will exceed the current highest summer temperature on record is greater than 90% in West and East Africa in the 2050s and in nearly all parts of sub-Saharan Africa in the 2090s (Battisti and Naylor, 2009). In contrast GCM projections of changes in precipitation are more diverse and agreement on the direction of change is high only for eastern and southern Africa. Changes in precipitation in the Sahel and along the Guinean Coast are highly uncertain. Rainfall is likely to increase over eastern Africa and likely to decrease in southern Africa during winter and in western parts of southern Africa (Christensen et al., 2007). Analysing an ensemble of 14 GCM projections and three emission scenarios shows that the length of the growing season in the 2090s will be reduced by 5 to 20% relative to current conditions in most parts of Africa and by more than 20% in the Sahel and southern Africa (Thornton et al., 2011). As a result arid areas with a growing season length of less than 120 days are likely to expand by 5–8% in the 2080s for two emission scenarios (B2, A2) (Fischer et al., 2002). Additionally an increase in the number of extremely wet seasons in West Africa and East Africa by 20% and an increase of extremely dry seasons by 20% in southern Africa combined with an increase in the rainfall intensity is expected (Christensen et al., 2007).

Temperature and precipitation changes might limit crop growth and development to a different extent depending on the current growing conditions and the magnitude of climate change. Published studies either analyze the combined effect of precipitation and temperature changes on crops in a climate change impact study or highlight only the importance of one climate variable for crops. To our knowledge there is only one study separating the effects of temperature and precipitation on crop yields in Africa. In this statistical analysis for individual countries in sub-Saharan Africa Schlenker and Lobell (2010) show that impacts on aggregated crop yields due to temperature changes are much stronger (− 38% to + 12%) than impacts due to precipitation changes (− 3% to + 3%) for five different crops. Consequently they doubt that shifts in the distribution of growing season rainfall will outweigh temperature effects on yield. This is contradictory to findings from studies on the effect of rainfall variability on crops which highlight the importance of variable wet season starts and the occurrence of dry spells for crop yields (Barron et al., 2003, Sultan et al., 2005). Dry spells in the flowering phase at two semi-arid locations in East Africa are estimated to reduce potential maize yields by 15–75% depending on the soil water-holding capacity (Barron et al., 2003). Long periods of drought in low-rainfall years have already seriously affected Africa's agriculture and economy in the past (Sivakumar et al., 2005) and will remain a danger in water-limited environments. A recent survey among crop modeling experts suggests that in a crop model which investigates crop response to climate variability, precipitation variation has the greatest influence on crop yields (Rivington and Koo, 2011).

We test the hypothesis that both, changes in temperature and precipitation will have an important influence on crop yields in sub-Saharan Africa depending on the location, the current climatic conditions and the projected climate change. To analyze the effects of precipitation and temperature changes separately and in combination is important for understanding and modeling climate change impacts on agriculture. We also investigate the effects of changing precipitation variability on crop yields by varying the mean daily precipitation, the total wet season precipitation and the number of small and large precipitation events in our simulations. On the one hand this analysis helps to identify the limiting factors for agricultural production in different environments and prioritize adaptation strategies to climate change. The success of breeding programs and farmers in selecting drought- or heat-tolerant crop cultivars will depend on their knowledge about changing growing conditions and the severity of different types of abiotic stresses. On the other hand it enhances the understanding of temperature and water stress effects in the vegetation model in order to identify future research and model development needs. Comparing the separated effects of changing temperature and precipitation with the combined effect will also reveal if a combination of drought and heat stress would have an even more significant effect on maize yields as known from several studies with maize, sorghum, barley and various grasses (Barnabas et al., 2008). We choose maize (Zea mays L.) as an example crop as it is the most important food crop in sub-Saharan Africa in terms of harvested area.

Section snippets

Climate data

The study area comprises all land area in Africa from 40° N to 40° S and from 20° W to 60° E. Daily precipitation data for the baseline climate 1991–2000, b-1995 hereafter, were taken from the WATCH Forcing Data (WFD) (Weedon et al., 2011). This data set combines monthly precipitation totals from the Global Precipitation Climatology Center (GPCCv4) (Rudolf et al., 2005, Schneider et al., 2008, Fuchs, 2009) and reanalysis data on day to day variability from the European Centre for Medium-Range

Results

We focus on results for grid cells with unimodal rainfall distributions, and if at least 0.1% of the grid cell area is used for maize production. We show changes in mean annual temperature, length of the wet season and precipitation during the wet season used as input data for the global dynamic vegetation model and impacts of these changes on maize yield in p-2060 and p-2085 relative to the baseline period b-1995. As crop yield changes are simulated using stylized climate scenarios for each

Understanding crop yield changes

The crop yield changes presented in Fig. 4 result from stylized climate experiments with rather large changes in the wet season precipitation and the wet season length and show the climate change effect on maize yields in each grid cell. They differ a lot between regions reflecting the differences in initial climate conditions and the corresponding crop's growing conditions as well as the magnitude of climate change. A cluster analysis allows for identifying groups of grid cells with similar

Summary and conclusions

We show that the importance of the agro-climatic variables temperature, wet season precipitation and wet season length for maize yields, varies in space depending on the initial climate limitations and the magnitude of climate change. Crop yields change considerably in regions with unsuitable or extreme growing conditions where even slight climate change results in strong relative effects on crop yields and in regions which are exposed to strong temperature and precipitation changes. The

Authors' contribution

The contribution of the different authors was as follows: K.W. and C.M. conceived the original idea of studying the effects of changing precipitation variability on crops. K.W. expanded this into a study on separating the effects of temperature and precipitation on crops. C.M. prepared the daily precipitation files from 18 GCMs, which was the basis for analysing changes in the wet season length and wet season precipitation. All authors were involved in developing and discussing the methodology.

Acknowledgments

We acknowledge the modeling groups, the Program for Climate Model Diagnosis and Intercomparison (PCMDI) and the WCRP's Working Group on Coupled Modelling (WGCM) for their roles in making available the WCRP CMIP3 multi-model dataset. Support of this dataset is provided by the Office of Science, U.S. Department of Energy. We thank Sebastian Ostberg and Jens Heinke for preparing the LPJmL input data on temperature and cloudiness and Jens Heinke for providing the WATCH forcing data. CM acknowledges

References (51)

  • P.K. Thornton et al.

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

    Global Environmental Change

    (2009)
  • F.N. Tubiello et al.

    Simulating the effects of elevated CO2 on crops: approaches and applications for climate change

    European Journal of Agronomy

    (2002)
  • K. Waha et al.

    Adaptation to climate change through the choice of cropping system and sowing date in sub-Saharan Africa

    Global Environmental Change

    (2013)
  • J.O. Adejuwon

    Food crop production in Nigeria. II. Potential effects of climate change

    Climate Research

    (2006)
  • E.A. Ainsworth et al.

    What have we learned from 15 years of free-air CO2 enrichment (FACE)? A meta-analytic review of the responses of photosynthesis, canopy properties and plant production to rising CO2

    New Phytologist

    (2005)
  • B. Barnabas et al.

    The effect of drought and heat stress on reproductive processes in cereals

    Plant, Cell & Environment

    (2008)
  • D.S. Battisti et al.

    Historical warnings of future food insecurity with unprecedented seasonal heat

    Science

    (2009)
  • A. Bondeau et al.

    Modelling the role of agriculture for the 20th century global terrestrial carbon balance

    Global Change Biology

    (2007)
  • T. Butt et al.

    The economic and food security implications of climate change in Mali

    Climatic Change

    (2005)
  • J.H. Christensen et al.

    Regional climate projections

  • D.P. Dee et al.

    The ERA-Interim reanalysis: configuration and performance of the data assimilation system

    Quarterly Journal of the Royal Meteorological Society

    (2011)
  • N.S. Diffenbaugh et al.

    Observational and model evidence of global emergence of permanent, unprecedented heat in the 20th and 21st centuries

    Climatic Change

    (2011)
  • D.E.S. Dodd et al.

    Early detection of the start of the wet season in semiarid tropical climates of western Africa

    International Journal of Climatology

    (2001)
  • K. Ebi et al.

    Smallholders adaptation to climate change in Mali

    Climatic Change

    (2011)
  • FAO

    The State of Food Insecurity in the World 2011. How Does International Price Volatility Affect Domestic Economies and Food Security?

    (2011)
  • Cited by (60)

    View all citing articles on Scopus
    View full text