Sustaining food self-sufficiency of a nation: The case of Sri Lankan rice production and related water and fertilizer demands (2024)

Associated Data

Supplementary Materials

Abstract

Keywords: Self-sufficiency, Food security, Agricultural intensification, Water footprint, Nitrogen runoff, Water resources

Introduction

Global food trade plays an integral role in redistributing agricultural production. Because the resources required for food production are heterogeneously distributed across the planet, the import of food allows a country’s population to exceed the number of people that the locally available resources could otherwise support. Trade dynamics, however, are susceptible to long-term demographic changes as well as short-term fluctuations in political and environmental conditions. This fact is noteworthy given that 78% of countries are not self-sufficient in terms of domestic calorie production (Davis et al. 2014). During times of scarcity (e.g., 2007/2008 food crisis and 2010 droughts/wildfires in Russia), countries that are heavily dependent on food imports are at a distinct disadvantage in terms of food security as compared to nations who can rely on local crop production to meet dietary demands. Moreover, countries in surplus may need to reduce their food exports in the coming decades as a result of increasing domestic demand (Suweis et al. 2013). Indeed, recent work has shown that, as the global food system has become more dependent on trade, it has lost resilience and grown more susceptible to shocks and crises (Suweis et al. 2015). To minimize the vulnerability associated with dependence on food imports, national self-sufficiency policies have been implemented around the world, including in Qatar, Japan, and India (Fader et al. 2013).

With rice comprising approximately 40% its total crop production (FAO 2014a), Sri Lanka provides an interesting example of a country with a long-standing, national self-sufficiency policy. Since gaining independence in 1948, agricultural policies in Sri Lanka have generally been aimed at improving its self-sufficiency in all food crops. Particularly, the target for rice production (the staple food of the country) was established at 100% of domestic demand (Imbulana et al. 2006). Due to a combination of high yielding varieties, paddy expansion, and increased use of irrigation and fertilizer, rice production has steadily risen to meet this target. As a result, Sri Lanka has been almost entirely reliant on its own rice production since 2005 (DCS 2014; FAO 2014a). While the country is currently able to meet domestic demand for rice, it is unclear whether the country can continue to do so under projected population growth.

Sri Lanka’s population is projected to increase by approximately 18% (from 20.3 to 23.9 million people) before plateauing by mid-century due to increased economic development and decreasing fertility (DCS, 2012; UN 2012). Current rice yields are also already approaching the maximum attainable yield (i.e., the highest crop yield that a farmer can attain in a given climate using conventional technologies such as irrigation and fertilizers), leaving limited room for improving crop productivity (Mueller et al. 2012). In view of Sri Lanka’s continuing efforts to maintain self-sufficiency in terms of rice production, it is important for the country to take stock of what the projected demand may be in the coming decades and whether rice production can respond to meet this demand. In addition, decision-makers must be aware of what the consequences for water resources may be, as further intensification of rice production will likely necessitate the use of greater amounts of nitrogen fertilizer and irrigation water and may contribute to increased nutrient runoff and reduced environmental flows.

With these various considerations in mind, here we examine three possible scenarios of intensification—improving rice yields, maximized crop harvest frequency, and optimized water use efficiency—to determine how many people might be fed and what the impact on water resources may be. In doing so, this study offers a simple assessment of projected rice demand in Sri Lanka, potential avenues for increasing supply to meet self-sufficiency goals, and what these options may mean for future water use and quality. Intended as an example of how to evaluate future food supply, human demand, and resource needs simultaneously, our work here focuses on changes in water demand and pollution from fertilizers. We therefore acknowledge that other factors, both biophysical (e.g., climate change, greenhouse gas emissions, and soil loss) and socioeconomic (e.g., job creation and commodity markets), may also be important to decision makers and that their inclusion in a similar analysis can help provide a more holistic perspective of the impacts and benefits of intensification. Our analysis provides a framework by which other countries may gain insight into their prospects for self-sufficiency and, in turn, identify what policy changes may be needed to improve their outlook.

Materials and methods

To assess Sri Lanka’s ability to meet future national demand for rice, rice production statistics were collected and analyzed for trends. Literature values were then used to determine the maximum harvest frequency and attainable rice yields. Using diet projections for South Asia, the number of people that Sri Lanka can support with the increased production is determined and compared to the country’s population growth estimates. We then identified the irrigation water and nitrogen fertilizer inputs necessary to increase the yield from current levels to close the yield gap (calculated as the difference between the maximum attainable yield and the current yield) by 50, 75, and 90%. The water and nitrogen inputs were then used to compute the blue and gray water footprints for each of these yield gap closures (YGC), where blue water represents irrigation water derived from surface and groundwater sources and gray water represents changes in water quality from fertilizer runoff. The blue and gray water footprint calculations are based on the water footprint network methods and allow quantification of changes in water use and quality (Hoekstra et al. 2011). These concepts of blue and gray water—as well as the calculations used to quantify them—are described in detail below. Finally, improvements in water resource use efficiency were compared to total and renewable freshwater resources (RWR) to determine absolute changes in water use.

Site description

Due to local agro-ecologies, two distinct rice production zones can be delineated in Sri Lanka: the wet zone (WZ) and the dry zone (DZ) (Fernando et al. 2010). The WZ receives approximately 2500mm of rain annually while the DZ receives approximately 1550mm of rain annually. To maximize rice production in the two zones, different cultivation practices and seed varieties (such as a preference for longer duration varieties in the WZ where water is not as much of a constraint) are employed (Weerakoon et al. 2011). Nubin (2002)’s delineation of the two zones was used to determine which rice districts fall within the wet and dry zones (Fig.1). Rice is cultivated during two seasons in both zones, with Maha being the major season and Yala the minor.

Open in a separate window

Fig.1

Wet and dry zones of Sri Lanka

Results

Rice yields have steadily increased in Sri Lanka at both the national and sub-national levels, showing some fluctuation due to drought (e.g., in 2010) but no signs of yield plateau or stagnation (Fig.2). Rice yields, rice production, and harvested area are all higher in DZ than WZ for both Maha (major) and Yala (minor) growing seasons for all years considered in this study (with the single exception of rice yield in the 1984 Maha (major) season). Overall when considering historical yields and rice land expansion, we find that 48% of the increased production in Maha (major) and 50% of the increased rice production in Yala (minor) is attributable to higher yields. In much of the country, rice yields are already approaching Mueller et al. (2012)’s estimated maximum attainable yield of 5 tons ha⁻¹. The frequency with which rice has historically been harvested in Sri Lanka has remained stable at~1.1 harvests year⁻¹ over the past 30years, a value well below the maximum possible harvest frequency of 2 harvests year⁻¹ estimated by Ray and Foley (2013).

Open in a separate window

Fig.2

Rice yields during the Maha (major) and Yala (minor) cultivation seasons in the two zones. In 2013, 68% of rice produced in the WZ and 61% rice produced in the DZ came from the Maha (major) season (DCS 2014). Rice yields have steadily increased in Sri Lanka at both the national and sub-national levels. Maha-WZ (a), Maha-DZ (b), Yala-WZ (c), and Yala-DZ (d) yields are best fit with an increasing linear function. For the Yala-DZ yield, an exponential fit had a lower rmse, but rmse was not significantly different from a linear fit. See Table S1 for more details on these statistics. Maha-DZ rice yield data for 1986 were not available

Based on linear extrapolations of historical rice yield data, we estimate that rice yields in WZ and DZ will reach the maximum attainable yield in the years 2053 and 2025, respectively. Because rice yields in DZ may plateau soon, this region will make up a decreasing amount of the country’s rice production: 75.5% in 2000, 74.5% in 2030 and 71.8% in 2050. Despite the limited improvement possible for rice yields, 24.7 million people can be fed in 2030 and 25.3 million people in 2050 if rice yields grow at current rates until reaching the maximum attainable yield (Fig.3). As a result, domestic rice production in Sri Lanka would exceed the demands of the projected populations of 23.3 million in 2030 and 23.8 million in 2050 (UN 2012). Under current irrigation technology, increasing current rice yield to 90% of the attainable yield would increase rice production by 0.89 million tons and would increase freshwater use to nearly 9km³ H₂O year⁻¹ (Table2). Alternatively, maintaining current yields but increasing the frequency of harvests from the year 2000 rate (1.1 harvests per year) to the maximum (2 harvests per year) would increase rice production by 2.22 million tons. With this increase in harvest frequency, Sri Lanka could feed 45.1 and 46 million people in 2030 and 2050, respectively (Fig.3). In comparison, we estimate that the additional rice production required to meet domestic demand by the year 2050 is 0.62 million tons.

Open in a separate window

Fig.3

Water use and human demand for rice production. Per ton (a) and national (b) water footprints of rice production are shown for different yield gap closures (YGCs). We include all three footprints together (green, blue and gray), but it is important to note that only blue and green water represent consumptive uses of water. c Current and projected populations (orange) are estimations from the UN (2012). These estimates of demand are compared to the number of people able to be fed under three rice supply scenarios—‘Yield,’ ‘Harvest,’ and Efficiency’—described in detail in “Materials and methods” section. The ‘Efficiency’ scenario utilized the water use efficiency value at 50% YGC. Values are presented in Table S3

The 50% YGC has the lowest water footprint (green+blue+gray) for rice production but, if maintained, would be insufficient to feed future populations (Fig.3c). This lower WF for 50% YGC reflects decreased values in both the blue and gray WFs. The per ton blue WF associated with YGC followed a U-shaped curve with 50% yield gap having the smallest blue WF and the 90% yield having by far the largest blue WF (Fig. S1). The per ton gray WFs showed a similar U-shaped behavior with 50 and 75% YGC showing substantial reductions in nutrient loading per ton of rice and a return to current levels with further increases in yield (Fig. S2).

Current consumptive water use for rice production (i.e., blue water+green water) constitutes 39.5% of all annual water withdrawals (agricultural or otherwise) (Table2) and 9.7% of Sri Lanka’s total RWR (FAO 2014b). With current irrigation technology, increasing current rice yields to 90% results in an increase to 68.9% of total freshwater withdrawals for rice production. Conversely, maintaining current yields but increasing the frequency of harvests represents a similar increase in the share of total freshwater resource use from 40 to 72% (Table S2). From current levels, the total gray WF increases by 33% (to 1.15km³ H₂O year⁻¹) for 90% YGC and increases 82% (to 1.57km³ H₂O year⁻¹) for maximizing harvest frequency alone (Table S2).

Discussion

Sri Lanka is marginally self-sufficient in terms of rice production (Fig.3). The prospects for the long-term rice self-sufficiency of the country are promising not only because the Sri Lankan population is expected to plateau (UN 2012) before rice yields are maximized but also because per capita dietary demand for rice is expected to largely remain constant (Alexandratos and Bruinsma 2012). This potential to increase rice production through improving yields is encouraging from a food security perspective. However, pursuing this type of intensification—as will likely be the case in Sri Lanka and in all of South Asia (Bruinsma 2009)—may bear substantial impacts with as much as a 250% increase in the required volume of irrigation (blue) water (from current yield to 90% YGC; Table2) and a 23% increase in the amount of applied nitrogen under current (ca. 2000) resource use efficiency. Further, simply increasing rice yields will be a far less efficient use of freshwater resources than will closing the harvest gap. For a given increase in freshwater consumption, closing the harvest gap can increase rice production by 2.5 times more than closing yield gaps alone. This would seem to indicate that closing the harvest gap is the preferred method of rice intensification, especially considering that Sri Lankan rice yields will likely plateau in the coming decades. Nevertheless, confounding socioeconomic (e.g., infrastructural maintenance and labor costs) and biophysical (e.g., climate extremes) factors likely make this option less viable (Ray and Foley 2013).

Currently, 90% of the total water withdrawals in Sri Lanka are used for agricultural purposes (FAO 2014b) and rice comprises one-third of Sri Lankan crop production (FAO 2014a). Achieving increased rice production may have serious consequences for the country’s freshwater resources with overall consumptive water use under the most intensive production scenario representing 68.9% of national freshwater withdrawals. In addition, the most efficient levels of YGC (i.e., ‘Efficiency’ scenario; Figs.3c, S1, and S2) are likely not adequate to meet future demand while buffering against climatic variability. Therefore, decision-makers may need to empower extension service agents to educate farmers on the value of improving rice water and nutrient use efficiencies. Indeed, one study found that irrigated rice production in Sri Lanka has the potential to reduce its water demand by~50% through efficiency improvements in irrigation infrastructure (Amarasinghe et al. 1999). Nitrogen fertilizer use efficiency can also be increased while still improving yields (Mueller et al. 2014). Further, alternative methods of cultivation are a possibility and can significantly reduce agriculture’s impact if implemented on a large enough scale. One such alternative method is the system of rice intensification (SRI). Although Sri Lankan farmers cited greater labor requirements as compared to conventional rice farming methods, a 3-year study reported that SRI helped improve yields by 44% while markedly reducing the use of synthetic fertilizers (Namara et al. 2004). In highlighting irrigation and fertilizer use in this study, we also note that additional measures and considerations related to agricultural intensification can also be incorporated to look at carbon and nitrogen footprints, soil loss, the influence of climate, and even social and economic impacts in a country.

Achieving greater rice production sustainably is already a formidable challenge because of the greater resource requirements. This task is further complicated by the various potential impacts on rice production associated with climate change. Climate change is particularly relevant to Sri Lanka’s self-sufficiency because the country does not produce a large rice surplus and the vast majority of the country’s rice supply is produced in DZ. Shifting precipitation patterns, elevated temperature, and more frequent extreme weather events may contribute to crop stress while longer growing seasons and the CO₂ effect may benefit rice production. Thus it remains unclear if the potential impacts of climate change on yields may offset any yield gains arising from improved technology (e.g., irrigation and fertilizers). While Lobell et al. (2011) found a detectable impact on historical rice yields in nearby India and Bangladesh, a study by Knox et al. (2012) predicted that South Asian rice yields are not expected to significantly decrease despite the possible increases in rice blue water demand due to climate change. This increase in rice irrigation water requirement due to climate change has been studied by De Silva et al. (2007) and so was not quantified here. In their study, the authors showed that increases in average Sri Lankan paddy rice irrigation water requirement will be modest (+13–+23% by 2050) under climate change but that certain areas of DZ—Batticaloa and Ampara districts in particular—will be more affected (De Silva et al. 2007). Preventing crop stress during (potentially more frequent) times of relative water scarcity may therefore mean that environmental uses of water receive a lower priority and that aquatic ecosystems will likely receive greater fertilizer pollution as a result (Meyer et al. 1999; Kundzewicz et al. 2008). Based on UN and International Water Management Institute indicators, several rice-producing areas relying on irrigation are expected to experience severe water scarcity in the next 10years and will need to increase water withdrawals in turn (Amarasinghe et al. 1999). It may therefore be necessary for legislators in Sri Lanka to define and allocate adequate environmental flows in order to protect biodiversity and important ecosystem services.

Decisions made now regarding the intensification of rice production will have important implications for food security and the environment in Sri Lanka. Should economic growth in the agricultural sector become the primary focus for Sri Lanka, the country holds a large potential to increase and export rice production through increased cropping intensity. With many Asian countries (e.g., India, Pakistan, and Indonesia) expected to see rapid population growth in the coming decades (UN 2012) and with climate change contributing to elevated food prices and increased food price volatility (IPCC 2014), Sri Lanka could become a regional exporter of rice (Fig.3; Table2). While maximized rice production can increase economic development and help alleviate poverty, it also means that Sri Lanka must be willing to use more natural resources with potentially greater impact on the environment. If, on the other hand, Sri Lanka simply places emphasis on domestic food security (as has been the case with its self-sufficiency policies), increased rice production will ensure that the country is not import-reliant and is therefore less vulnerable to spikes in food prices. While maintaining the option of food imports can buffer Sri Lanka against local climatic stress in the future, this may not be a reliable option under certain climatic conditions. One recent study found that as Sri Lanka may have limited access to the global rice market due to its limited financial influence in the region—if, for instance, a large drought event impacted rice production in Asia (Puma et al. 2015). Sri Lanka’s reduced dependence on food imports is also important given recent evidence that the global food system has become more vulnerable to shocks as reliance on food trade has increased (Suweis et al. 2015) and that exporting countries may reduce exports in times of food scarcity (Puma et al. 2015). Indeed, food price shocks may become more frequent, as another recent study found that global rice prices have taken on a nonlinear (and more volatile) behavior (Elser et al. 2014). However, the same departure from linear behavior is also true for fertilizer prices (Elser et al. 2014) and may affect the economics of domestic rice production in Sri Lanka given its growing reliance on fertilizer imports (FAO 2014a). How exactly Sri Lanka may react to a high price regime—both for food and agricultural inputs—is unclear and will depend on how the country adopts new technologies for crop nutrient use efficiency and nutrient reclamation (Rittman et al. 2011). Over the past several decades, the Sri Lankan government has attempted (with mixed results) to distribute technologies to smallholder farms and rural people. In particular, Sri Lankan governments have invested heavily in irrigation projects, which helped to improve crop yields, enable rural access to water and, in turn, reduce poverty in these areas (Amarasinghe et al. 2005). However, the maintenance of this infrastructure as well as the monitoring of water use has frequently been lacking (Godaliyadda et al. 1999). Given that paddy rice cultivation employs nearly 2 million farmers, the government is under significant political pressure to continue providing a fertilizer subsidy (Weerahewa et al. 2010). Though this subsidy has helped to increase land productivity and rice yields, it continues to place a heavy burden on the Sri Lankan GDP while encouraging the over-application of fertilizers. It may therefore be necessary for decision-makers to re-evaluate such policies in the context of water footprint and resource use efficiencies in order to more holistically determine the best path forward for the country.

Conclusion

The case presented here is one example of how a country can examine its potential for increasing domestic crop supply, how this supply relates to projected human demand, and what the associated water and fertilizer demands may be. These analyses can provide a first insight into what the food security path of a country may be if it continues as it has and, if this is not desirable, help inform what changes may need to occur in the country’s food production system in order to ensure self-sufficiency (or whatever the targeted goal may be). It can also allow decisions-makers to weigh the costs and benefits of agricultural alternatives (in this case, increasing crop harvest frequency). Sri Lanka can anticipate continued independence from rice imports, but further efforts should be made to increase the efficient use of water and nutrients and to ensure that added rice production occurs as sustainably as possible while buffering against future climatic uncertainty and stress. In all, an assessment of demand, the potential to increase supply, and possible impacts on natural resources and the environment will leave decision-makers much better equipped to address the complex issue of food security and to establish informed and concrete targets to achieve it.

Electronic supplementary material

Acknowledgments

We thank Paolo D’Odorico, George M. Hornberger, and Michael L. Pace for their useful insights during the preparation of this manuscript. We also thank Nathaniel Mueller for kindly providing data on irrigated area and nitrogen application rates. This study was funded by the National Science Foundation (Grant no. DGE-00809128, DGE-0909667, and EAR-1204685).

Biographies

Contributor Information

Kyle Frankel Davis, Phone: 434-924-7761, Email: ude.ainigriv@sz5dfk.

Jessica A. Gephart, Email: ude.ainigriv@trahpeg.

Thushara Gunda, Email: ude.tlibrednav@adnug.arahsuht.

References