report

Food Systems At Risk

Transformative Adaptation for Long-Term Food Security

Download PDF
Chapter 2

Incremental Adaptation Alone Will Leave the World Hungry

This section examines available evidence to establish why transformative approaches to adaptation are needed to avert or minimize looming food security challenges and explores a range of issues that are preventing transformative adaptation from being more widely implemented. It assesses how mounting ecosystem degradation will undermine the ability of farmers, fishers, and herders to rely on traditional ways of managing climate variability and other risks.

Ashwini Chaudhary/Unsplash

2.1 Shortfalls in Staple Crops Projected

Recent analyses expose both the value of incremental adaptation measures in protecting the global food supply from climate change impacts, and also worrisome gaps between their likely effectiveness and the projected impacts of climate change on yields of staple crops. Figure 2 illustrates this point for the key global staple crops of wheat, rice, and maize. Based on a meta-analysis model of ~27,000 data points from studies published over the last four decades, Aggarwal et al. (2019) calculated variance around an ensemble mean of multiple studies of each particular crop, country, and time slice to illustrate projected percent changes in yield relative to a baseline of 1960–90 for 2020, 2050, and 2080. The orange and blue bands indicate a 95 percent confidence interval based on a thousand replications of the model. The orange bands illustrate the reference case of average modeled climate change impacts on these crops globally without adaptation. The blue bands represent average modeled impacts of climate change globally on these three crops over the coming decade with incremental adaptation measures. Each blue dot represents the blue band disaggregated to show individual countries.

Although the blue bands in the 2020 graphs are near the 1960–90 baseline for wheat and slightly under for rice and maize, the model predicts that yields of rice and maize could fall over 30 percent over the coming decades without adaptation (orange bands), and by a global average of up to 10 percent with adaptation (blue bands). However, while overall global declines are expected to be fairly minor with adaptation, some individual countries (depicted by the blue dots) fall well below the baseline and can expect to experience significantly declining yields of these staples even as their populations expand. These are countries where transformative adaptation is likely to be needed, while incremental adaptation may be sufficient in those with less dramatic declines. The graphs include latitude along the x-axis, making it clear that wealthy countries further from the equator will fare better than developing countries in the tropics, which are more likely to reach the limits of incremental adaptation sooner.

Figure 2 | Average Impacts of Climate Change on Crop Yields, with and without Incremental Adaptation Measures, 2020, 2050, and 2080

Note: Based on a meta-analysis model of ~27,000 data points from studies published over the last four decades, Aggarwal et al. (2019) calculated variance around an ensemble mean of multiple studies of each particular crop, country, and time slice to illustrate projected percent changes in yield relative to a baseline of 1960–90 for 2020, 2050, and 2080. The orange and blue bands indicate a 95 percent confidence interval based on a thousand replications of the model. The orange bands illustrate the reference case of average modeled climate change impacts on these crops globally without adaptation. The blue bands represent average modeled impacts of climate change globally on these three crops over the coming decade with incremental adaptation measures. Each blue dot represents the blue band disaggregated to show individual countries.

Source: Reprinted from Aggarwal et al. (2019).

Despite this and other emerging evidence regarding the limitations of incremental adaptation measures, the vast majority of agricultural adaptation—including climate-smart agriculture (CSA) as it is commonly practiced—focuses on such measures. The intention of such efforts is to preserve existing food systems by building resilience to climate change impacts, rather than recognizing that more fundamental changes to what can be produced, where, and how will increasingly be needed. CSA projects rarely explore what will happen when incremental measures become insufficient to fully manage increasing climate risks. There is relatively little research available on how to respond when crops and livestock reach their physiological limits of how much additional heat or drought they can tolerate, sources of irrigation water are reduced by permanent drying trends or salinization from sea level rise, or marine species cannot be bred to handle dramatically increased ocean acidity. This is despite a growing body of research that indicates such limits are already being reached in some locations and contexts.

The evolving field of agroecology “seeks to optimize the interactions between plants, animals, humans and the environment while taking into consideration the social aspects that need to be addressed for a sustainable and fair food system” (FAO n.d.). Agroecology and other types of nature-based solutions show great promise for advancing adaptation while reducing further losses in biodiversity and the unsustainable use of natural resources. However, relying exclusively on such approaches will grow increasingly risky as climate change impacts intensify. Projected shifts in global ecosystems will undermine a key assumption of agroecology: that ecosystems are stationary and stable and can thus be counted on to continue providing the same range of ecosystem services. When, for example, rainforests shift to grasslands, and grasslands shift to deserts, the amount of watershed regulation they provide will change, as will the interactions between wild pollinators and cultivated crops—both of which could undermine agricultural production. In addition, as Searchinger et al. (2018) suggest, there may be limited environmental contexts in which agroecology can contribute efficiently to meeting the concurrent goals of limiting global temperature increases and feeding a growing global population.

For similar reasons, although critically important, it should not be assumed that local knowledge and traditional solutions alone will be adequate to manage increasing climate-related agricultural risks. Such place-based expertise often evolved within fairly stable ranges of climate variability over generations. When those ranges shift beyond what traditional coping strategies can handle to unprecedented flooding or heatwaves, entirely new pests and diseases, or other novel challenges, traditional knowledge alone may not always suffice. Archaeological evidence suggests that climatic shifts contributed to the downfalls of the Maya civilization and those of the U.S. Southwest, to name just a few. Adaptation measures based on traditional knowledge should be recognized, valued, and considered along with less context-specific solutions—but not treated as silver bullets that can solve all climate-related challenges.

Local economies and markets will also have to respond to unprecedented circumstances. And while farmers are indeed often the best agents of change to influence other farmers—for example, the pioneer farmers referred to below when discussing autonomous transformations—they will need enhanced access to new types of crops and livestock and guidance on how to raise and market them.

As Figure 2 indicates, the limits to adaptation for agricultural crops will not be uniform across the globe. Looking more closely at maize, the third most important crop on the basis of harvested area, Ramirez-Cabral et al. (2017) found that under an A2 emissions scenario (i.e., at the higher end of emissions scenarios defined by the IPCC, but not the highest; see Nakicenovic and Swart 2000) for 2050 and 2100, tropical areas will experience the highest loss of climatic suitability, while regions closer to the poles will become more suitable. South America will have the greatest loss of climatic suitability, followed by Africa and Oceania, with large areas that are currently suitable for maize becoming limited by heat and dryness. On the other hand, Asia, Europe, and North America will become more suitable.

Figure 3 maps out hotspots where strong impacts of climate change are projected to lead to large gaps in wheat, maize, and rice production. The projections are based on assessments of impacts with adaptation on crop yield at the country scale for the 2050s and the food production gap (the difference between 2050 food demand and current food supply). Countries with high food gaps and high impacts of climate change are most vulnerable.

Figure 3 | Climate Change Hotspots

Source: Reprinted from Aggarwal et al. (2019).

A potential upside of this analysis may be that even countries projected to experience severe deficits in one or two of these staple crops—such as India, where both wheat and maize productivity are expected to decline—will often be able to continue producing similar amounts of other crops. For example, climate change impacts on rice in India are projected to be low, and it may be possible to grow more rice per hectare with improved varieties, inputs, and cultivation techniques, though this will be true only in locations where water supplies are adequate.

But a range of factors will impede farmers’ ability to switch between crops. Some of these—e.g., soil type, seasonal microclimatic conditions, absence or presence of pests and pollinators—will be difficult to overcome. Others—such as access to information on how to grow new varieties; required inputs such as seeds, credit, processing facilities, and markets; and sociocultural factors such as consumer willingness to consume less traditional foods—become more possible to develop when the need for them is recognized in advance so they can be planned for, financed, and implemented; that is, where transformative approaches to adaptation can be applied.

2.2 Traditional and Cash Crop Yields Vital to Food Security Expected to Decline

In addition to these impacts on staple crops, those that are less important in global markets but are essential to food security are also likely to be severely affected; for example, a 43 percent decline in plantain yields in Central Africa is expected over the next 20 years (Fuller et al. 2018), while bananas and beans are also in jeopardy (Rippke et al. 2016). Traditional and wild crops hold potential for filling food security gaps, although they have been largely overlooked in climate change research (Niles et al. 2020). Wild crops include any seeds, roots, or leaves that can contribute to people’s diets and are collected from uncultivated areas. Such crops are a significant portion of production in many low-income countries compared with the world’s major staples. For example, in Niger, traditional crops (e.g., millet, cow pea, sorghum) are produced at a ratio of 46 to 1 (production, tonnes) compared with major world crops (e.g., maize, wheat, rice) (FAO 2020b; Varshney et al. 2012). These crops (e.g., bambara) are particularly important to women, who grow the majority of them, and smallholder farmers, who rely on them for food security (Oyugi et al. 2015). Other examples of traditional crops include amaranth, jute mallow, desert date, and Shona cabbage.

Furthermore, cash crops such as coffee are also at risk; declines in the production of cash crops will weaken the ability of those who depend on them to purchase food and thereby further undermine food security. More than 120 million smallholders rely on coffee for their livelihoods, but by 2050 climate change will threaten 50 percent of the area currently suitable for its production (Climate Institute 2016), meaning that the livelihoods of coffee farmers are in jeopardy. Although incremental adaptation measures such as improved varieties and better water and shade management can help, the crop is likely to meet the physiological limits of its heat tolerance as temperatures rise (Kath et al. 2020). Without transformative action to adapt in areas where the crop is losing viability, such as by moving coffee production upslope to cooler areas and substituting crops that can thrive in warmer conditions, as explored in Section 3 of this report, the ramifications will be enormous.

2.3 Livestock Systems Are Also at Risk

As noted in the TACR paper on transformative adaptation for livestock production (Salman et al. 2019), climate change is also affecting the rapidly expanding livestock production sector, upon which an estimated one billion people living in poverty depend for food and income. An estimated 180 million people in developing countries depend on livestock grazing on drylands for their livelihoods (Thornton et al. 2008; Salman et al. 2019). Livestock production is particularly important in the semi-arid agroecological zones most at risk from climate change, some of which are growing too hot for current livestock breeds and facing desertification, which means that they will no longer be able to support current levels of grazing even if incremental solutions such as additional watering holes and improved forage are provided. As some of these hotspot areas lose viability for livestock production, herders’ livelihoods will be threatened. The IPCC has linked climate change to lower animal growth rates in Africa (IPCC 2019), while the FAO projects global demand for livestock products to increase by 70 percent to feed a population estimated to reach 9.7 billion by 2050 (FAO 2019).

Livestock production systems around the world are already changing in response to demographics, markets, and economic development—but these transitions rarely consider long-term climate risks. Without the proper information and resources, livestock systems may not withstand intensifying direct and indirect climate impacts such as changing disease dynamics. Technical and financial support are needed to avoid excluding or disadvantaging those living in poverty and other vulnerable groups (Salman et al. 2019).

2.4 Some Regions, Landscapes, and Human Systems Are at Risk of Nearing Tipping Points

In a growing number of locations, current agricultural livelihoods may soon no longer be possible. These include the Mekong Delta, where salinization due to sea level rise threatens rice production; chronic drought in California, which threatens fruit and vegetable cultivation; and the creation of a dustbowl in East Anglia, England, due to drought and land degradation, which threatens homegrown vegetable production (Benton et al. 2017).

Particular regions and types of landscapes are likely to reach their adaptive limits sooner than others, especially over longer timeframes as the effects of slow-onset climate change impacts emerge. Figure 4 identifies types of ecosystems that are most vulnerable to water stress and other slow-onset events based on a review of available literature.

Figure 4 | Ecosystems Most Vulnerable to Water Stress and Other Slow-Onset Climate Change Impacts

Source: Authors.

National-level data can mask important limitations, such as the difficulty of shifting between crops in places with unsuitable precipitation patterns, inadequate water supplies, or inappropriate soil—not to mention the cultural and behavioral challenges of producing food that satisfies global markets or meets local preferences. The most damaging impacts of warming on rainfed maize, wheat, and rice have already been substantially moderated by shifting the locations where they are cultivated over time, along with expanding irrigation (Sloat et al. 2020). However, continued crop migration will be limited by socioeconomic and political factors, as well as land suitability and water resources, and care must be taken to prevent substantial environmental costs by pushing agriculture into uncultivated areas. Food production must be increased through sustainable intensification (i.e., higher yields per unit of land) rather than expanding crop or grazing land and converting forests, savannas, and peatlands to farmland. This will often require stronger legal protection for natural areas (Searchinger et al. 2018).

Such broad-scale projections may also gloss over the true impacts on social groups expected to be hardest hit, such as female-headed households, the poorest farmers, landless tenant farmers, and day laborers (Niles and Brown 2017; Niles and Salerno 2018). 

According to the IPCC Global Warming of 1.5°C special report, in a growing number of places in the world, climate change is pushing systems—including ecological, but also agricultural, hydrological, economic, and others—toward severe and widespread risks (IPCC 2019). Figure 5, derived from that report, depicts the risks to natural, managed, and human systems around the world as temperatures rise.

As shown in Figure 5, crop yields, a central focus of this paper, are expected to experience moderate to high climate change impacts. Other primary systems required to maintain global agricultural productivity that are at risk as temperatures rise include terrestrial ecosystems and coastal and fluvial flooding; climate change impacts are projected to range from moderate to very high in these systems. The graphic also includes moderate to high impacts of rising temperatures on heat-related morbidity and mortality, which is critical to agricultural productivity; farmers will not be able to maintain productivity when more days become simply too hot for them to work in their fields or herd their animals. Note, however, that Figure 5 includes temperature increases only up to 2.5°C. A more recent review of all available evidence (Sherwood et al. 2020) finds that the odds of a temperature shift below 2 degrees is less than 5 percent under a high-emissions scenario (the world’s current trajectory), while there is a 6 to 18 percent chance that temperatures will shift more than 4.5°C, or 8.1 degrees Fahrenheit. This estimate closely matches cumulative emissions over the past 15 years (Berwyn 2020). This would push critical systems toward tipping points beyond which changes would be irreversible and the limits of adaptation would be reached.

Other systems that humans depend on for food security, such as warm-water coral reefs and tropical freshwater fisheries, as well as coastal areas that are home to 10 percent of the world’s population, are also subject to tipping points where severe, irreversible climate change impacts occur as temperatures rise and the limitations of adaptation are reached (IPCC 2019). Beyond these tipping points, these systems will no longer exist in the form they do today—including the food systems of an increasing number of places. This signals that fundamental, systemic transformations are needed.

Figure 5 | Impacts and Risks for Selected Natural, Managed, and Human Systems

Note: Crop yields, a central focus of this paper, are expected to experience moderate to high climate change impacts. Other primary systems required to maintain global agricultural productivity that are at risk as temperatures rise include terrestrial ecosystems and coastal and fluvial flooding; climate change impacts are projected to range from moderate to very high in these systems.

Source: Reprinted from IPCC (2019).

2.5 Entire Ecosystems Are Shifting; Food Systems Must Shift Along with Them

The projected impacts of climate change will lead to dramatic, ongoing changes in the plants and animals found in a particular area, comprising shifts of entire ecozones. For example, some recent climate change–linked fires in the Amazon rainforest are expected to result in permanent conversion to grasslands (Cooper et al. 2020), semi-arid areas are becoming deserts in regions like the Sahel (Huang et al. 2016), sea level rise is eating away at coastlines in West Africa (Croitoru et al. 2019), and groundwater aquifers are becoming too salty to be used for irrigation or drinking water in the Nile Delta (Abd-Elhamid et al. 2016).

Figure 6 depicts the magnitude to which global terrestrial ecosystems are expected to change as planetary warming continues. It illustrates that, while severe ecosystem changes (in red) are largely limited to Arctic areas with 2°C warming, they will become far more widespread and affect more populous areas with 3.5°C warming. Most of the planet’s ecosystems will be severely affected and no longer function as they currently do if 5°C of warming is reached.

Figure 6 | Likelihood of a Severe Change in Ecosystems

Notes: Abbreviation: °C: degrees Celsius. While severe ecosystem changes (shown in red) are largely limited to Arctic areas with 2°C warming, they will become far more widespread and affect more populous areas with 3.5°C warming. Most of the planet’s ecosystems will be severely affected and no longer function as they currently do if 5°C of warming is reached.

Source: Reprinted from Gerten et al. (2013).

Recent analysis (IE&P 2020) finds that approximately one billion people live in countries that lack the resilience to manage the ecological changes they are expected to face between now and 2050. Of the 157 countries covered by this analysis’s Ecological Threat Register, 22 percent will face catastrophic food insecurity—that is, they are likely to experience a substantial increase in undernourishment—by 2050. Without swift and substantial action, this could also result in unprecedented displacement of people.

Regions that experience a mean annual temperature of over 29°C are projected to expand from 0.8 percent of the world’s surface to 19 percent. Such areas are currently largely concentrated in the Sahara and affect relatively small numbers of people, but are expected to grow to encompass areas inhabited by one-third of the global population. This is in contrast to a projected shrinking of zones with mean annual temperatures of approximately 11–15°C, where most current production of crops and livestock largely takes place (Xu et al. 2020) and could therefore be considered the most suitable temperature range. Under a business-as-usual climate change scenario, the geographical position of this temperature niche is projected to shift more over the coming 50 years than it has since 6000 BP (before present). The regions that will be most affected are among the poorest in the world—and thus our efforts at finding and applying new approaches to adaptation will be most needed in these areas.

As climate change risks affect many types of ecosystems simultaneously, essential connections between agricultural production and natural systems will be impacted in ways that are likely to undermine the ecosystem services farmers depend on, such as regulating water supplies or supporting pollinators. For example, pollinators such as birds, bees, and butterflies are essential for the production of 35 percent of the world’s crops (Abrol 2012). Yet the variety of wild pollinators essential to crop production is falling due to climate change impacts (Giannini et al. 2017). New pests and diseases are already spreading into areas unaccustomed to coping with them. This may overwhelm local knowledge and capacity to deal with outbreaks by using traditional, non-chemical measures, as was the case with the recent locust plague in East Africa. Such changes may also affect rural households by jeopardizing the forest products they depend on, and increase the risks of wildfires, landslides, and other impacts.

Food system transformations should occur in alignment with ecosystem shifts to improve resilience and make more possible the sustainable use of water, land, and energy. While it may be possible, for example, to grow lettuce in the desert Southwest of the United States and tomatoes in greenhouses in arid northern Mexico, this can be accomplished only with depletion of scarce ground water for irrigation and heavy infusions of energy to cool greenhouses. This is an example of an unintended consequence of increasing the risk of maladaptation, which is addressed in greater detail in Section 4.2.

Start reading