Main

Food production, storage, processing, distribution, retail and consumption are all exposed to wide-ranging forms of environmental change, including slow-moving changes in average conditions (for example, climate, nutrient and water cycling), smaller-magnitude variations around those means and larger, anomalous disruptions1,2. Environmental variability and disruptions are manifested not only as floods, droughts and extreme heat (that is, climate variability) but also as natural hazards, pests, disease, algal blooms, coral bleaching and aflatoxins3,4,5.

Environmental variability includes fluctuations within the historical range but increasingly consists of extreme—and at times unprecedented—events driven by climate change and other forms of human-environmental change. While expected changes in climatic and environmental conditions over the coming decades will certainly have important implications for future food security, nutrition, and human and planetary health1,2,4,6,7, environmental variability and extreme events are already leading to harvest losses and distribution disruptions each year8. In many cases, these short-term impacts (that is, those occurring within the span of a production/growing season) have been shown to be stronger than those caused by slow-moving changes in climate and the environment7. Such developments point to a growing need to protect the stability of not only food production but also other steps in food supply chains in the face of rising environmental variability and risk9.

Environmental variability can affect food supply chains at local to global scales. In regions with limited global market access and where food production and consumption are tightly coupled (that is, short supply chains common in subsistence-based food systems), environmental variability can more readily translate into local shortages of specific foods10, creating food insecurity depending on available substitutes, the impacts on livelihoods11, and how household diets and nutrient intake are modified12. Increasingly, the impacts of environmental variability are not limited to local producers but spread through longer supply chains. A prominent example was seen in 2008, when a drought in key grain-producing regions, combined with rising biofuel demand, high oil prices, decreasing grain stocks and the depreciation of the US dollar, led to a spike in global grain prices13. That set off a series of rice export bans, furthering shortages and ultimately driving more than 130 million people into poverty and an additional 75 million people into malnourishment13. Although only partially an environmental shock, this global-scale disruption illuminated the potential for shocks to cascade through the food trade network and to impact geographically distant places and people.

As food supply chains grow increasingly globalized, it is important to understand the potential novel risks of how environmental variability is propagated or attenuated within food systems. Ongoing globalization and the rise of multinational food corporations have enhanced the efficiency and complexity with which producers and consumers are linked. Research is only beginning to understand how these dynamics influence the propagation of the effects of environmental variability through global and local food supply chains, how large-scale events (for example, blockades, recessions and pandemics)14,15,16 may compound these effects, and what they ultimately mean for the stability and affordability of nutritious diets. Identifying the key processes and actors in food supply chains—and understanding their interactions with, and exposure to, environmental variability and the economy—is central to reducing food and nutrition security risks associated with periodic food shortages and to improving resilience within food systems.

Here we perform a scoping review (Extended data Fig. 1) to synthesize existing knowledge on the impacts of environmental variability on food supply chains and the transfer of such impacts along local and global food supply chains. We first highlight the entry points for environmental variability along each step in the supply chain, drawing on examples from the scoping review. Then, we present the quantitative review results related to which food items, supply chain steps and types of environmental variability have been the focus of research to date. Finally, we summarize the qualitative review results on the response option space for actors along the supply chain.

On the basis of our findings, we contend that a new wave of food systems research focused on the full food basket, diverse forms of environmental variability and local-to-global interactions across the supply chain is required to understand the diverse impacts of environmental variability on the spectrum of food systems. This new knowledge can inform a range of policy options and business decisions across the supply chain that provide flexibility in responding to variability in food supplies and that enhance the resilience and adaptability of food systems to known and novel shocks.

Results

Entry points for environmental variability

Environmental variability can affect the quantity, quality and timing of food as it travels through the supply chain (Fig. 1). Most entry points for environmental variability to impact food supply chains occur at production. Exposure to variability and shocks is particularly pronounced for rainfed agriculture and smallholders (that is, households with <5 ha), who account for the majority of the world’s food production (57% (ref. 17) and 55% (ref. 18), respectively) and rely largely on locally available natural resources. A large body of literature documents the impacts of floods, droughts, natural disasters and other extreme events on food production19,20,21. In some instances (for example, disease and pests), individual food items may be impacted, while other phenomena (for example, large-scale drought and algal blooms) may induce declines across entire portions or regions of the food production basket. Environmental variability also presents challenges related to the quality and safety of food items, possibly leading to changes in the nutrient content of some species as well as increased food safety risks. Increasing variations in rainfall have been linked to greater variability in the protein content of wheat22 while higher temperatures have been associated with changes in the fatty acid content of some fish4,23 and declining milk quality in dairy cows24. Temperature, humidity and precipitation variability, along with increased pests, are associated with increased aflatoxin levels during production, while algal blooms can produce toxins that enhance the risk of shellfish poisoning25,26. Environmental variability also affects the transmission of animal diseases (which can necessitate mass animal culling or fishery closures), foodborne illness and zoonosis (that is, animal-to-human transmission).

Fig. 1: Entry points for environmental variability in food supply chains.
figure 1

1) Environmental variability can enter at each step of the food supply chain. 2) Economic conditions and infrastructure quality can dampen or enhance these environmental factors. A variety of geopolitical and other economic factors influencing the interactions of environmental variability and food supply chains are also discussed in the text. 3) The impacts of this variability on food quantity and quality can also be passed between steps in the food supply chain. 4) The cascading effects represent the main outcomes that could occur without intervention from supply chain actors, and the direction of the arrow next to each effect indicates whether there would be an expected increase or decrease. Depending on the specific context, only certain cascading effects may emerge in response to environmental variability. The asterisks indicate a type of environmental variability, economic influence or cascading effect that was identified within the scoping review. Temp., temperature; conn., connectivity; econ., economic.

Full size image

Specifically concerning storage, subsistence farmers who have limited growing seasons and depend on stored crops are particularly sensitive to extreme rainfall events affecting food quantity and quality27,28. Aflatoxin levels in crops increase if storage conditions are not sufficiently dry, while bacterial contamination and foodborne disease are common in the absence of cold storage. Broadly, higher temperatures and changes in rainfall patterns can increase the occurrence of bacteria, viruses, parasites and fungi, and their associated foodborne diseases in stored foods25,29,30.

Distribution and market access can also be substantially impacted by environmental variability. Floods can disrupt road networks and food trade, typically leading to losses of more perishable or time-sensitive food items31,32. Compounded by changes in the affordability and diversity of foods available in markets, physical and economic barriers can impact food and nutrition security33. Flooding and rising sea levels can also increase water runoff and contamination, leading to increased waterborne disease and chemicals in the food supply25. Further, extreme weather events disrupt food distribution by damaging existing infrastructure or slowing food shipments, which can increase risks of food damage, spoilage or contamination34. Compared to fresh perishable foods, processed foods may be better able to withstand the impacts of environmental variability related to spoilage, creating a trade-off between the durability, safety and convenience of processed food and their often lower nutritional quality35.

Consumption may also be impacted by environmental disruptions through shifts in food availability, affordability and acceptability. For subsistence farmers, such shifts may directly influence the quantity and quality of foods available for consumption within the household. Reduced incomes and higher food prices can have important implications for nutrition, limiting food purchasing power and leading to shifts in dietary intakes, often towards cheaper highly processed foods36,37,38. Environmental disruptions can also limit the ability to prepare food or to fully absorb nutrients from ingested food, often leading to compromised nutritional outcomes (for example, childhood stunting39). For instance, in Bangladesh, cooking stoves are often located outside on the ground and can be submerged by flood waters35. In Australia, extreme high temperatures were associated with increased risk of intestinal infections and nutritional and metabolic issues among young children40. In India, children living in communities exposed to recurring floods were at least 2.3 times more likely to be severely wasted, stunted or underweight41,42.

Propagation of environmental variability

Understanding the impacts of environmental shocks on food security and nutrition requires connecting production to consumption through all supply chain steps. We analysed peer-reviewed literature to investigate which food items, environmental disruptions and supply chain steps have been the focus of research to date. The scientific evidence for the effects of environmental shocks on food systems is highly concentrated on the main staple crops (maize, rice and wheat) and relatively few types of shock (Figs. 2 and 3). Of the 325 papers examined (Supplementary Table 1), nearly half looked at the impact of shocks on maize, wheat and rice (37%) and/or other cereals (10%). Another 18% did not focus on any specific food item (Fig. 2a). The types of food and shock studied are probably interrelated, as there was an emphasis on the temperature (22%) and precipitation (41%) extremes that threaten cereal crops (Fig. 2b). Other shock types, such as cyclones, harmful algal blooms and disease events, were roughly evenly represented in the studies (Fig. 3).

Fig. 2: Environmental disruptions of food supply chain steps.
figure 2

a,b, Literature review results (n = 325 studies) for each supply chain step disaggregated by food group (a) and shock type (b). Studies have concentrated on environmental variability related to the production of maize, wheat and rice and have focused on extreme precipitation and temperature. Certain studies examined more than one food item; as such, the number of observations in the figure exceeds the number of studies.

Full size image
Fig. 3: Environmental disruptions and food groups.
figure 3

Literature review results disaggregated by shock type and food group (n = 325 studies). As a result of the emphasis on staple grain production, studies also heavily focused on related disruptions (that is, extreme temperature and precipitation). Certain studies examined more than one food item; as such, the number of observations in the figure exceeds the number of studies.

Full size image

Across the reviewed studies, 42% focused on a subnational scale (that is, a region within a single country), 35% on a national scale (that is, a single country) and 15% on an international scale (that is, multiple countries or global), with the remainder being unspecific about the scale. Of the studies set in a specific country or region, over a quarter were set in Asia and nearly a fifth in Africa. The subnational focus, particularly in regions where subsistence food production makes up a large portion of the food supply, is highly valuable for understanding the impacts of environmental variability on local food security; yet, as a quarter of all food calories are traded internationally43, our understanding of the novel food security risks arising from increasingly global food supply chains is limited.

Further, we found that studies on environmental shocks are biased towards production; 89% included the production stage and 73% focused on production alone (Fig. 2a). Consumption, here broadly defined as including outcomes related to food expenditure, dietary intake, food security and nutrition outcomes, was included in just 15% of studies. A similar percentage (16%) looked at more than one supply chain stage and only 6% explicitly looked at shock propagation between supply chain steps. Given the heterogeneity of the methods and disciplines that cover the current literature on the propagation of environmental shocks through food supply chains, it was not possible to assess the representativeness or prevalence of specific linkages between environmental variability and food supply chain shock propagation.

In all, our findings indicate that while each step in a food supply chain may be independently affected by environmental variability in various ways (Fig. 1), upstream shocks can be propagated to subsequent supply chain steps under certain conditions—and this variability can be transferred both locally (Box 1) and globally, as exemplified by some studies44,45,46,47,48,49,50. For more local systems, as in the case of subsistence households, some of these steps are absent or diminished (Box 2), and environmental variability is more likely to affect their diets depending on the availability of substitutes, the effects on household incomes and the quality of the foods consumed. Smallholder farmers who are unable to sell their produce owing to suboptimal quality (or questionable safety) tend to consume it themselves, possibly compromising the nutritional value and diversity of their diets25. Low-income populations in urban settings may face a similar trade-off and purchase less expensive foods despite lower quality35,36 increasing their exposure to aflatoxin and other foodborne illnesses and disease. For more global food supply chains, a range of recent network-based modelling studies have attempted to evaluate the distant impacts of food supply shocks, finding that highly globalized countries44, net importers and low-GDP countries45,46,47, and countries with low strategic reserves48 are most exposed to external shock events. Other agent-based modelling work has found that shocks travel more easily across networks when the perturbed product is directly traded51 and that the indirect economic effects of supply chain shocks (for example, firm-to-firm disruptions that a natural hazard causes) can exceed direct impacts (for example, physical damage from the natural hazard)52.

One of the primary ways by which shock propagation along supply chains has been examined in the literature is through the effects of disruptions on prices. Price transmission across supply chains informs economic decisions at various stages, but volatile prices can create uncertainty for downstream supply chain actors53. The effect of transmission along a supply chain depends on the structure of the sector and the policy context. Foods with multiple sourcing options, substitute products and longer shelf lives can better buffer supply and dampen variability54,55. Concentrated market power also restricts price transmission when retailers exercise their power to keep consumer prices stable during volatile periods53. Conversely, foods with few substitutes, foods requiring longer grow-out periods, sectors with fewer industry actors and regions with limited infrastructure tend to have lower capacity to dampen supply or price variability, resulting in greater propagation along the supply chain. In cases where variability reaches consumers (often in the form of price fluctuations), physical and economic access dictates that these shortages are not evenly distributed between countries or between individuals. Thus, it is essential to understand not only how environmental variability can lead to instability in food supply but also how economic mechanisms may be used to modulate these fluctuations across the population.

Supply chain actors and their response option space

Approximately two-thirds of the coping strategies identified through our review focused on the production level. However, there are multiple points across the supply chain where governments, the private sector, intergovernmental, civil society and non-governmental organizations, and individual farmers can act to reduce the negative impact of environmental variability on food security7. Indeed, given the interconnectedness of supply chain steps, the full suite of food system actors will need to intervene—from production to consumption—to counteract the many ways in which environmental variability and extreme events manifest and propagate (Fig. 1). Opportunities to address this variability depend on the actor, the step in the supply chain, the timescale of implementation and the type of environmental variability (Table 1).

Table 1 Potential responses to environmental variability across the supply chain
Full size table

Responses at the production level focused on improving farmers’ resilience in terms of producing sufficient quantity and quality of food in the presence of environmental variability. While government institutions (including agricultural ministries, national agricultural research centres, and extension and agricultural policymakers) are the key actors involved in these responses, the private sector, intergovernmental organizations (for example, the World Bank) and other organizations such as CGIAR are also important for enabling stable production56. One important way to increase the resilience of agricultural production is by choosing crops, cultivars, species and breeds that are not only less susceptible to stresses such as drought, pests and salinity, but also offer other benefits for livelihoods or nutrition17,57. For instance, subsidy programmes in India aimed at promoting traditional climate-resilient cereals can increase iron and zinc supply, lower farmer input requirements and reduce water, energy and emission footprints58. The adoption of biofortified crops can potentially offer similar win-win situations; in sub-Saharan Africa, for instance, orange-fleshed sweet potato is more drought tolerant than other staple crops while also improving vitamin A intake59. Overall, programmes to introduce traditional or biofortified foods must consider cultural acceptability and may require promotional efforts to increase the willingness of consumers to shift their tastes to new varieties60,61. Other strategies at the production level include: shifting the timing and patterns of planting62; adopting agroforestry practices63; diversifying farms64; increasing the adoption of technology and on-farm infrastructure, agrochemicals and irrigation65; water and soil conservation practices66; and shifting or expanding growing areas67. However, strong extension services are needed to ensure that farmers have the knowledge and skills to change production practices. It is important to note that large multinational companies are often vertically integrated and have increased global connectivity, being more resilient to certain shocks owing to their ability to reallocate production and sourcing in response to climate shocks68,69. In other cases, globalized supply chains with specialized foreign markets are at a greater risk to supply chain disruptions70.

Investment by governments and other actors, including the private sector, can also increase resilience to environmental variability across supply chains. For example, increasing access to irrigation (where sustainable)71,72 and promoting water-efficient management practices73 can stabilize production during times of drought74, while improvements in storage and road infrastructure can help protect harvested crops and ensure market access under extreme weather events75. Crop and index insurance can protect farmer incomes against environmental risk, particularly in low- and middle-income countries, and in some cases enable them to access credit for investing in inputs, higher-value crop varieties76 and new technologies77. Besides protecting rural livelihoods, such programmes may be crafted to incentivize more sustainable and resilient on-farm choices.

Food processing strategies can help increase the shelf life of perishable foods (for example, fruits and vegetables), while fortification and product reformulation can increase the nutritional quality of foods—both of which can prevent food insecurity in times of limited food diversity and shortages associated with environmental variability. Given that processed foods may be more resilient to price volatility, ensuring that their consumption does not lead to nutritional compromises is essential. Policy options and outreach efforts at both the retail and consumer level can help incentivize consumption shifts, such as subsidies for more, and taxes for less nutritious foods (for example, ultraprocessed foods)78,79,80. This is particularly important given that food prices are often the primary driver of food choice and shifts in dietary intake when consumers experience shocks81. Social safety nets (for example, provision of cash transfers and food supplementation) can help mitigate the effects of environmental shocks on both food quantity and quality82. Also, strengthening food safety standards across supply chains can help ensure greater accountability among actors, including governmental and international bodies such as CODEX Alimentarius (that is, Food Code). Lastly, our review elicited several individual-level responses to environmental variability, such as suspending cultivation, migrating, seeking off-farm employment and selling assets (Table 1).

Increasing cohesion among supply chain actors will be necessary to support responses to growing environmental variability. Actors’ different goals can create barriers to such responses, which is especially apparent for trade policy. Given the interconnectedness of global food systems, a government’s policies to support (or thwart) trade can have important implications for their ability—and that of others—to respond to environmental variability83. The private sector responds to variability within the constraints of the policy environment set forth by governments, and the extent to which a company is vertically integrated across multiple steps in a food supply chain also determines its suite of options in responding to environmental variability (Box 2). How governments and the World Trade Organization shape trade policies can restrict certain policy responses. For example, shocks are more readily transmitted to importing countries (and disproportionately so to low-income countries47) owing to an intensification of international trade and changes in the distribution of cereal reserves48. Countries at risk of experiencing these shocks can potentially respond by increasing their domestic reserves of cereal commodities or importing from a diverse set of suppliers. For exporting countries, embargoes can help maintain adequate domestic food supplies during times of limited production; yet, hoarding and export bans may affect each country’s resilience to environmental variability either positively or negatively, depending on the specific actions taken and the commodities targeted13. The Bali Package represents one recent effort to address this trade-distorting behaviour by which the World Trade Organization seeks to protect the policy space of governments to quickly enact trade strategies to better ensure domestic food security84. It is possible that a similar approach can be taken to allow greater flexibility in responses to environmental variability as well.

Moving beyond food production shocks

Our current understanding of the various ways by which environmental variability affects (and propagates through) food systems is insufficient. Expanding our knowledge beyond a focus on production shocks for staple crops will be essential for enhancing the resilience and sustainability of food supply chains. By considering the various steps and interactions of food supply chains, different actors can more effectively develop targeted interventions to cope with environmental variability and avoid unintended consequences within food systems. In particular, adaptive strategies can be improved by understanding the shape and evolution of food supply chains, the potential for surprises and spillovers, and the risk of simultaneous and unprecedented shocks.

The shape and evolution of food supply chains have gradually evolved away from strong local interactions between producers and consumers towards greater complexity and distance. Over the past half-century, globalization and market consolidation have meant that power over global food system decisions has become concentrated43,85. These trends of vertical integration and homogenization have had profound upstream and downstream effects that span supply chains (Box 2)—ranging from ecosystem impacts in areas of production68,86 to modified options for consumers87,88—while the growing reliance of many countries on food imports has also left them increasingly vulnerable to shocks that occur beyond their borders45,46,89,90. The increase in processed or multi-ingredient foods has also led to converging and diverging product streams that make it more difficult to link the sources of raw materials to an end product91. Taken together, such changes have also meant that an individual consumer interested in sustainable purchasing is able to exercise less control over the sourcing of their food and the exposure of their diet’s production to environmental risk (Box 2)85. Being able to trace food’s passage through a supply chain not only ensures quality and sustainability standards but also allows governments to steer companies towards reliable and less vulnerable food supply decisions beyond economic incentives to do so. Substantial progress has been made to promote traceability in some countries (for example, the UK, the Netherlands and Japan) and for selected products (for example, the livestock sector in Uruguay and dairy products in France and the UK); however, for most countries and food items, understanding the steps connecting producers and consumers remains a critical obstacle to enhancing the resilience and adaptability of food supply chains to short-term environmental variability92. To address this, a host of innovative approaches offer opportunities to enhance information-sharing capabilities93, facilitate real-time responses to supply chain disruptions, couple environmental externalities with production94, account for the challenges of monitoring and reporting across countries, and recognize the differences in supply chains between countries and across food items. Key corporate actors that are increasingly influential in the dynamics of food systems95 can mobilize independent of governments to address food sustainability and resilience challenges and ultimately benefit from the outcomes96,97,98.

Surprises and spillovers can result from environmental variability due to direct and indirect interactions throughout a supply chain, among different foods and across spatial and temporal scales8. Such unexpected events are typical of complex systems, and models fail to capture the range of possible behaviours or the emergent properties of many actors interacting. Actors’ behavioural responses may differ within each specific disturbance context, including the prices of substitute foods, the broader economy and responses of other actors99,100. Price spikes of an input to food production due to environmental variability can lead actors to substitute resource inputs or consumers to substitute food products, creating spillovers into other supply chains (Fig. 1).

Substitution depends on the availability and affordability of suitable substitutes and can affect production of foods with seemingly little connection. For example, fishmeal and fish oil are important ingredients for some farmed fish and livestock feeds. The vast majority of fishmeal and fish oil is produced along the west coast of South America, and catch is heavily influenced by El Niño conditions, leading to price fluctuations. While alternative feed ingredients are easily available for poultry feeds, there are few nutritionally equivalent alternatives available for some farmed fish, so fluctuations in the fishmeal-oriented fisheries have a greater spillover effect into aquaculture than poultry. While identifying past surprises and spillovers is an important step towards anticipating potential food system responses to environmental variability, such events are difficult to predict. Fostering general resilience through enhancing production and trade diversity, maintaining food reserves and production buffers, managing feedbacks within the food system and supporting leadership and trust at multiple levels101,102 offers a path towards maintaining food system function in the face of surprises.

Simultaneous and unprecedented shocks can pose novel risks to global food systems. Agricultural trade agreements have been based on the premise that a disruption in supply can be compensated by increased sourcing from non-impacted areas (and to a limited extent by strategic reserves for certain commodities103,104). This paradigm, widely adopted by countries and corporations alike, assumes that shocks in one region occur independent of those in other places and will not be large enough to overwhelm the available pool of food supply and storage48. Yet, evidence suggests that large-scale climate teleconnections105,106,107 as well as a changing climate and increasing variability108 can lead to simultaneous and/or compounded109,110 impacts on crop production. For instance, rising temperatures are expected to increase the likelihood of simultaneous maize production losses in major production areas111. Global climate oscillations significantly influence productivity in two-thirds of the world’s croplands112. Synchronized climate anomalies associated with the trans-Pacific El Niño/Southern Oscillation (ENSO) have led to concurrent influences on major crop yields in geographically distant regions113. These studies suggest that maintaining asynchrony and diversity in areas of food production and sourcing—combined with food storage during bumper years—can limit instability in food supplies under increasing climatic and environmental variability114. These global teleconnections point to the possibility of coordinated risks that have yet to be adequately incorporated into climate-smart food security strategies. Examining the probability of synchronized extreme climate events and their linkages with agricultural output anomalies can better inform adaptive strategies of food supply chain actors115. More broadly, this scoping review points to a lack of investigation of numerous combinations of co-occurring environmental disruptions and the need to understand their effects on (and propagation through) food supply chains.

Conclusions

Our analysis shows that most research to date has focused on the associations between environmental variability and production. Although there is a growing body of evidence on the associations between environmental variability and nutrition outcomes, little research exists on the exact pathways by which these shocks influence nutrition. In some cases, we know that agriculture can influence nutrition outcomes through food production, income and women’s empowerment pathways116, but these have not been adequately examined in the context of environmental variability. Moreover, it is unclear the extent to which environmental variability affects nutrition outcomes due to shifts towards poorer-quality diets and/or increased disease risk leading to disruptions in the absorption of nutrients in different contexts. Without this evidence, little can be done to identify the most appropriate responses to environmental variability.

Also, little research has examined the propagation of shocks through food supply chains and their subsequent effect on consumption. This limits our understanding of the true effects of environmental shocks on the food system as a whole and how local shocks can have global impacts (and vice versa). Given the interconnectedness of food systems, a shock that influences one step of the food supply chain will undoubtedly impact subsequent steps. Likewise, a shock in one country or region can have serious implications for another. As our global food systems become more integrated, understanding the dynamics of propagation across supply chains through empirical research is critical to inform adequate interventions.

Our assessment of coping strategies identified a suite of actions that various supply chain actors can adopt in preparation for or in response to environment variability. Our findings indicate that multiple strategies across the food system will be needed in response to environmental variability, but that our ability to identify the most appropriate package of strategies to adopt remains limited. Understanding the specific interventions that can be taken within a variety of contexts to dampen the effects and propagation of environmental variability along food supply chains is critical to link food systems research with actionable outcomes. Moreover, additional research is needed that examines the interaction between people and their food system in response to shocks; these interactions are probably multifaceted and can have important implications for food security, diets and nutritional outcomes.

Finally, increasingly complex and globalized supply chains are exposed to environmental variability and extreme events at multiple points, leading to the possibility of shock propagation, spillovers and simultaneous shock events. The literature on these types of event is relatively scarce, but insights from resilience literature point towards the importance of production and sourcing diversification and strengthening internal feedbacks for enhancing the resilience of the global food production system117. Although not an environmental shock, the COVID-19 pandemic has revealed how a shock in one country can have a ripple effect across the globe and how a single event can simultaneously impact multiple geographies and points along the supply chain. Such shocks can produce differential effects on food availability (for example, unharvested fields due to reduced on-farm labour), physical (for example, restricted movement and transportation) and economic (for example, loss of livelihoods) access, and intake (for example, dietary shifts towards highly processed foods with an extended shelf life)15,16,118. This underscores the urgent need to expand our understanding of shock events beyond their impacts on production, to minimize systemic risks and enhance the capacity to cope with future disruptions.

Methods

We conducted a systematic scoping review to examine the propagation of environmental shocks throughout food supply chains (Extended data Fig. 1). Scoping reviews apply a systematic approach to reviewing the literature to map the evidence on a given topic and identify the main concepts, theories and knowledge gaps119. Given the heterogeneity of the methods and disciplines that cover the literature on the propagation of environmental shocks through food supply chains, a scoping review was chosen as the most appropriate approach to summarize findings from the existing literature119,120,121. We used the PRISMA checklist for scoping reviews to guide our methodological approach (Extended Data Table 1).

The scoping review was conducted in June 2019, with guidance from a medical librarian, using the Web of Science database and the following search terms: environmental variability (or climate variability, shock, propagation, weather, extreme event, disaster); crop (or food, livestock, fish); and production (or nutrition, food security, diet, access, availability, system, supply, trade, price transmission, price volatility). Our inclusion criteria included studies that: were published in English in 2008 or later (to capture articles published after the 2008 global food crisis); examined supply chains in the 1950s or later; measured an outcome that is directly part of the food system (for example, yield, catch, price, production, infrastructure, food quality and safety, and diets (nutrition through the diet pathway)); and focused specifically on an aspect of environmental variability (for example, disasters, temperature/precipitation/humidity extremes, disease or pest outbreaks, algal blooms and ENSO). We also included studies examining price shocks given that many of these shocks are related to environmental variability. We excluded books and book chapters as well as studies that solely assessed climate risks or vulnerability as well as those that examined the impacts of slow-moving climate on food supply chains. To ensure that our search was sufficiently sensitive to identify articles that met our inclusion criteria, we employed an iterative approach to refine the final search strategy. Specifically, we performed a first-iteration identification utilizing a set of initial search terms and examined whether known articles5,8,11,24,44,45,46,47,48,49,89,90,105,111 that met the inclusion criteria were retrieved. We then repeated the refinement of search terms until the search strategy captured all of the known articles that met our inclusion criteria. The exact search terms were coded in Web of Science as: (“climate varia*” OR “environment* varia*” OR “shock*” OR “propagat*” OR “weather” OR “extreme event*” OR “disaster*”) AND (“crop*” OR “livestock” OR “fish*” OR “food*”) AND (“production” OR “nutrition*” OR “security” OR “diet*” OR “access” OR “availability” OR “system” OR “supply” OR “trade” OR “price transmission” OR “price volatility”) NOT (“food web*”).

The records identified through our search (n = 15,271) were imported into R and any duplicates were removed. All three authors screened the titles and abstracts to assess whether they met our inclusion criteria. Two authors then independently reviewed all abstracts that met our inclusion criteria to verify the screening process. In the event that there was disagreement among the two authors, the third author was consulted and a decision on its inclusion was reached by consensus. We extracted data from articles in two stages: at the abstract and at the full-text review stage. We examined the abstracts that met our inclusion criteria to identify whether the study examined the food supply chain beyond production (that is, yields, productivity and production quality). If the study focused only on production, we extracted data related to the food/crops examined as well as the type of shock. At this stage, we also assessed whether the study solely focused on coping strategies. If the study focused on short-timescale (for example, within a growing season) or intermediate-timescale (for example, switching crops for the following season) coping strategies, we extracted data related to the coping strategies used at this stage as well. The studies that focused solely on coping strategies were then excluded from the full-text review, given that the focus of this paper is on the propagation of shocks across supply chains.

In the second stage of data extraction, we reviewed the full texts of the remaining articles that met our inclusion criteria (n = 325). If there was uncertainty regarding whether the study met our inclusion criteria on the basis of the full-text review, the authors jointly discussed the article and came to a consensus on its inclusion. Of the studies that met our inclusion criteria after the full-text review, we extracted information regarding the food/crops examined, the steps of the supply chain considered in the analysis, the type and year of the shock, whether there was propagation and information related to the study methods including the primary outcomes assessed and the study setting. The information extracted from the included studies allowed us to summarize the existing literature and identify gaps in the knowledge base. More specifically, we used the extracted data to quantitatively summarize the types of crop and environmental variability, the supply chain step examined and evidence of propagation along the supply chain. Given the heterogeneity of the included studies, we complemented the quantitative data collection with examples from the included studies to contextualize our findings. All extracted data were recorded into an Excel spreadsheet. We first piloted the data extraction spreadsheet with a small number of articles and made necessary adjustments before completing the data extraction process. Once the extraction was completed, we reviewed the data to ensure that there was consistency in the coding of output variables (for example, crop name and type of environmental variability).

Reporting Summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.