Modelling and forecasting roots & tubers losses and resulting water losses in sub-Saharan Africa considering climate variables

https://doi.org/10.1016/j.pce.2020.102952Get rights and content

Abstract

The implications of climate change coupled with anthropogenic activity on water resources have caused great concern, particularly in areas vulnerable to water stress such as sub-Saharan Africa. We focused on the future magnitude of food loss (FL) in African regions, using an ARIMAX model to fit and forecast roots & tubers (R&T) losses of five major crops cultivated in Africa regions, including cassava, potato, sweet potato, yam, and “other” roots & tubers. The forecast was done up to 2025 under the influence of five exogenous variables, namely, gross domestic product, harvested area, precipitation, temperature, and food production. In addition, the future crop water requirement (CWR) of production under climatic variables, and the associated water loss embodied in FL were quantified by means of CROPWAT 8.0. Our findings showed that in 2025 the magnitude of FL is expected to increase by 19.06%, 104.78%, and 27.72% at 2013 levels for East Africa, Middle, and West Africa, respectively. Under future climate the CWR of the selected crops is expected to be higher in West Africa (1790.24 m3/tonne), than in East (989.03 m3/tonne), and Middle Africa (903.64 m3/tonne). The future water loss embodied in FL is expected to be 114.37, 112.80, and 12.06 m3/cap/yr for the West, Middle, and East Africa regions, respectively. Our results show that measures aimed at preventing FL will also alleviate pressure on available water resources.

Introduction

The implications of climate change on the environment and anthropogenic activities is profound. It may be linked to agricultural practices, crop production, water resources, and so on (Bocchiola et al., 2013; Kusangaya et al., 2014; Mimikou et al., 2000; Shrestha et al., 2017; Stancalie et al., 2010; Tingem et al., 2008). Agricultural production is vulnerable to climate change, particularly in low income countries including in sub-Saharan African (SSAn) regions (Maddison et al., 2007; Muller et al., 2011). The low resilience of African agricultural practices to climate change is likely to generate greater food loss (FL) impacting on food security (Muller et al., 2011).

FL can be considered as the decrease in edible food mass throughout the part of the supply chain that specifically leads to edible food for human consumption (Gustavsson et al., 2011). FL is more related to the production, postharvest, and processing stages of the food supply chain (Kummu et al., 2012). FL is historically a major concern in Africa (Feukam Nzudie et al., 2020). The continent has a high prevalence of undernourished people, estimated at 20.4% of its population in 2017 (FAO, 2018). Preventing FL becomes even more difficult when coupling this with the impacts of climate change. For example, in most SSAn regions, FL in fresh vegetables has a tendency to increase due to hot climates (Gustavsson et al., 2011). Furthermore, weather conditions more generally impact on FL (Huang et al., 2017; Kaminski and Christiaensen, 2014). A crop which requires a warm dry climate to attain a specific dryness can be spoilt when exposed to a humid climate (Abass et al., 2014). Climate may also promote the development and spread of pests and plant diseases on/off-farm which indirectly act on FL (Manjula et al., 2009). Other than climate, there are also other factors which can influence FL. The economic situation of a nation may also impact on FL (Aulakh and Regmi, 2013; Gustavsson et al., 2011). For example, less developed (lower income) countries are more likely to generate FL due to lower levels of technological advancement, such as crop storage and transport facilities, and food processing. Other factors affecting FL include the magnitude of food production and the harvested area. This may be explained by the fact that a relatively high production (in the case where demand is lower) is likely to lead to loss of some of the excess food (The Economist Intelligence Unit, 2014). Since FL can be expressed per unit of harvested area, the latter is also an important influencing factor. Overall, FL is being driven and aggravated by the aforementioned factors in SSAn regions.

FL impacts on food security, economic prosperity, and natural resources such as water (Rezaei and Liu, 2017). Numerous studies have been conducted into the impacts of FL on water resources. Ridoutt et al. (2010) found that mango loss was responsible for 16.6 GL of blue water consumption in Australia. Kummu et al. (2012) used a process analysis to find that FL was responsible for 24% of total water resources globally, which translated into 27 m3/cap/yr. Liu et al. (2013) found that total water embodied in FL in China was equivalent to 13.5 × 104 GL in 2010. By means of process analysis, Le Roux et al. (2018) found that 4 GL/yr water was lost from the Steenkoppies Aquifer in South Africa due to FL in vegetables. However, these studies stop short of considering the implications of climate change on FL and associated water resource. Furthermore, FL reduction in the long term could potentially alleviate pressure on water resource. As such, and to inform integrated water management and planning practices, quantitative assessment of the implications of climate change on FL and associated water resources is required.

Roots & tubers (R&T) play an important role in feeding the world and tackling food insecurity (Scott, 2000). In SSAn regions ca. 20% of food energy consumption is derived from R&T (Scott et al., 2000). However, large amounts of R&T is lost before reaching final consumption. Most R&T FL occurs post-harvest than at other stages of the food supply chain in many SSAn countries (Gustavsson et al., 2011). Aforementioned, FL induces water loss which is in turn linked to the crop water consumption. As known the water consumption in growth stage is highly depend on the crop itself. As an example, taking potato as a crop belonging to R&T, its water consumption is about 500–700 mm/growing period (FAO, 1986). While the range of water consumption for other crops is from 300 to 2500 mm/growing period for bean and sugarcane respectively (FAO, 1986). Although potato has a relatively lower water consumption during its growing stage compared to the previous range, the point is how significant a large consumption of R&T could impact on total water consumption. In this study, we use an ARIMAX model to forecast the magnitude of R&T associated FL in East, Middle, and West SSAn. The major R&T crops cultivated in SSAn are classified as cassava, potatoes, sweet potatoes, yams, and “other R&T”. Our forecast has been undertaken to 2025, considering five independent variables, namely, Gross domestic product (GDP), harvested area (Har), precipitation (Pre), temperature (Temp), and production (Pro). The ARIMAX model was selected since it can integrate both a dependent or response variable (in this case FL) and several independent or exogenous variables (Anggraeni et al., 2017). In addition, we investigate the implications of future climate change on crop water requirement (CWR) in order to assess the water embodied in the expected FL. Note, that due to insufficient data in other African regions, only East, Middle, and West Africa were considered.

Section snippets

Methodology

Fig. 1 presents a framework describing the relationship between climate change, FL, and water resource as developed for the current study. The first step was to investigate the potential implications of climate and exogenous variables on human activity by means of the ARIMAX model, forecast until 2025. The second step was to quantify future CWR of production under climatic variables, and the third step was to investigate the implications of climate and human activity through the quantification

Model development

The results of the Granger-causality test between the dependent variable and the independent variables are summarized in Table 2. Considering East Africa, all five independent variables showed no Granger causality with the dependent variable (see the Conclusion column for the East Africa case); their p-values were all greater than 0.1 (10% considered for this study). This infers that the relevant independent variables cannot be used to improve or give much information in model development for

Conclusions

Climate change variables such as temperature (Temp), precipitation (Pre), and wind speed can potentially affect food loss (FL) and associated water resource loss. Here, we used five exogenous variables including two climatic variables (Pre and Temp), two food production variables (harvest area and food production), and one economic variable (gross domestic product, GDP) to forecast the magnitude of R&T losses for cassava, potato, sweet potato, yams, and “other” roots & tubers. By means of the

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (No. 72074136; 72033005; 72022009). The authors are grateful to Albert Ling Sheng Chang for his valuable assistance.

References (67)

  • S. Shrestha et al.

    Quantifying the impact of climate change on crop yield and water footprint of rice in the Nam Oon Irrigation Project, Thailand

    Sci. Total Environ.

    (2017)
  • G. Stancalie et al.

    Using earth observation data and CROPWAT model to estimate the actual crop evapotranspiration

    Phys. Chem. Earth, Parts A/B/C

    (2010)
  • X. Zhao et al.

    Index decomposition analysis of urban crop water footprint

    Ecol. Model.

    (2017)
  • W.A. Abia et al.

    Agriculture in Cameroon: proposed strategies to sustain productivity

    Int. J. Res. Agric. Res.

    (2016)
  • W.M. Adams et al.

    The Physical Geography of Africa

    (1996)
  • M.M. Aldaya et al.

    The Water Footprint Assessment Manual: Setting the Global Standard

    (2012)
  • S.R. Alimi et al.

    Toda-Yamamoto causality test between money market interest rate and expected inflation: the Fisher hypothesis revisited

    Eur. Sci.

    (2013)
  • R.G. Allen et al.

    FAO Irrigation and drainage paper No. 56. Rome

    Food Agric. Organ Unit. Nation.

    (1998)
  • B.H. Andrews et al.

    Building ARIMA and ARIMAX models for predicting long-term disability benefit application rates in the public/private sectors

    (2013)
  • W. Anggraeni et al.

    The performance of ARIMAX method in forecasting number of tuberculosis patients in Malang regency, Indonesia

    Int. J. Appl. Eng. Res.

    (2017)
  • AQUASTAT

    Distribution of Physical Water Scarcity by Major Hydrological Basin (Global)

    (2020)
  • J. Aulakh et al.

    Post-harvest Food Losses Estimation-Development of Consistent Methodology

    (2013)
  • C. Brouwer et al.

    Irrigation Water Management: Irrigation Water Needs

    (1986)
  • C. Chatfield

    Time-series Forecasting

    (2000)
  • N.D. Chauvin et al.

    Food Production and Consumption Trends in Sub-saharan Africa: Prospects for the Transformation of the Agricultural Sector

    (2012)
  • J. Cilliers et al.

    African Futures 2050-the Next Forty Years

    (2011)
  • CRUTS v3.24.01 Data Variables

    (2017)
  • P. Döll

    Vulnerability to the impact of climate change on renewable groundwater resources: a global-scale assessment

    Environ. Res. Lett.

    (2009)
  • J. Doorenbos et al.

    Calculation of Crop Water Requirements

    (1992)
  • N. Dussadee et al.

    Effect of plant shading and water consumption on heat reduction of ambient air

    Chiang Mai J. Sci.

    (2018)
  • S.H. Ewaid et al.

    Crop water requirements and irrigation schedules for some major crops in southern Iraq

    Water

    (2019)
  • Irrigation Water Management: Irrigation Water Needs

    (1986)
  • The State of Food Security and Nutrition in the World 2018. Building Climate Resilience for Food Security and Nutrition

    (2018)
  • Cited by (2)

    View full text