The role of resilience in food system studies in low- and middle-income countries

(1)

Contents lists available atScienceDirect

Global Food Security

journal homepage:www.elsevier.com/locate/gfs

The role of resilience in food system studies in low- and middle-income

countries

Markus A. Meyer

Department of Natural Resources, Faculty of Geo-Information and Earth Observation (ITC), University of Twente, Hengelosestraat 99, NL, 7514AE, Enschede, the Netherlands A R T I C L E I N F O Keywords: Resilience Food Value chain Supply chain Sustainability Agriculture A B S T R A C T

Current changes in food systems in low- and middle-income countries are not only associated with increasing demand for food driven by population growth. Shifts in diets with diﬀerent production and processing structures likely have a more considerable impact in future: higher land requirements (e.g., more animal-based proteins) or changing production regions and structures (e.g., higher share of (imported) processed food). Whether the ex-pected transition in low- and middle-income countries that is considerably larger than in high income countries is resilient, is questionable. Beyond, the resilience of food systems is not consistently assessed and hardly syn-thesized for low- and middle-income countries. A systematic review based on a keyword search was therefore conducted to identify research gaps on food system resilience for low- and middle-income countries.

This review identiﬁed three clusters of studies: studies in cluster 1 (global models and primary production) mainly model global (change) impacts on crop production, mostly disregard the downstream of the value chain and use globally accessible crop data (e.g., for cereals). Studies in cluster 2 (farmers' behavior under global change) mainly analyze with surveys and workshop formats the resilience of farmers and farming households under global change phenomena such as climate change and equally explore especially farmers’ adaptative capacity. Studies in cluster 3 (food systems and value chains) take a more holistic perspective and mostly qualitatively analyze the resilience of food systems based on existing studies. Quantitatively, most studies model the robustness of food systems and not their transformability.

Future research should quantitatively assess the adaptiveness and transformability of food systems and study the resilience of food systems beyond primary production. Studies should quantitatively model the impact be-tween the food system and associated environmental and social systems and along the food value chain to generate relevant insights for food system governance.

1. Introduction

Increasing and reliable food supply is needed to meet the expected population growth by 26 percent to 9.7 billion people in 2050. This increase varies for sub-regions between two percent (Europe and North America) and 99 percent in Sub-Saharan Africa (UN, 2019). Beyond population growth, changing diets (especially in low- and middle in-come countries) will equally aﬀect land requirements (Röös et al., 2017). Expected shifts in diets to more animal-based proteins combined with population growth in low- and middle income countries will likely become a driver of land-use change. Population dynamics will not be the main driver as in previous periods (Alexander et al., 2015; Springmann et al., 2018). A transition towards western diets with, e.g., more animal-based proteins and processed foods as described inPopkin (2017)will have a considerable impact on land use and food security.

Röös et al. (2017)show that current cropland is insuﬃcient for protein production of projected diets in 2050.

Diet shifts in low- and middle-income countries will be unlikely resilient and sustainable due to, e.g., increasing cropland scarcity. Resilience as a concept is increasingly used to address food security challenges and to conceptualize the complexity of interactions between multiple actors (e.g., value chains) and the impact of random external shocks (e.g., natural hazards) or changes in food systems (Tendall et al., 2015). Resilience is also a concept to integrate diﬀerent disciplines (Béné et al., 2016) and to link and balance diﬀerent targets, e.g., food security and sustainability (Prosperi et al., 2016). Quantifying the re-silience of food systems is crucial as the expected food system trans-formation can lead to unsustainable outcomes and food insecurity (Prosperi et al., 2016).Schipanski et al. (2016)also expect that a better understanding of resilience can reduce the vulnerability of food

https://doi.org/10.1016/j.gfs.2020.100356

Received 14 October 2019; Received in revised form 9 January 2020; Accepted 29 January 2020 E-mail addresses:m.a.meyer@utwente.nl,markus.a.meyer@gmail.com.

(2)

systems, i.e., improve food security. A better understanding may equally improve the governance of food system transformation in low-and middle income countries to prevent instead of remediate food system failures (Béné et al., 2016). However, resilience is less fre-quently considered and synthesized when analyzing food systems. Re-cent reviews in developing countries take a rather overarching per-spective, e.g., at the interface of food systems, urbanization, and agriculture (Abu Hatab et al., 2019a,2019b) and do not investigate the conceptualization of resilience in food system studies. Beyond, farming systems and transformation pathways strongly diﬀer between low- and middle and high-income countries (Reardon et al., 2019). These dif-ferences may lead to deviating resilience outcomes and may require additional research and diﬀerent governance approaches for low- and middle-income countries.

We deﬁne resilience as the ability of a system to maintain its societal functions through the system's robustness, adaptiveness, and trans-formability in response to internal and external stress (Folke et al., 2010;Meuwissen et al., 2019). The resilience of food systems extends beyond the resilience of farms and farmers to the entire value in con-trast to other studies (Hooks et al., 2017;Meuwissen et al., 2019). We also see the links of the agriculture-based food system to the social and environmental systems as crucial (partly in line with Tendall et al. (2015)) (e.g., drinking water quality, rural demographics). However, less studies address the resilience of food systems and associated socio-ecological systems.Prosperi et al. (2016)synthesize research on resi-lience in social-ecological systems andBehzadi et al. (2018)for supply chain management. We think that it is necessary to integrate and analyze the resilience of value chains and associated socio-ecological systems simultaneously.

To capture the domains dealing with food systems, three major re-search strands need to be considered (Bolwig et al., 2010; Vroegindewey and Hodbod, 2018): value-chain analyses with a social and an economic emphasis, social-ecological systems' research with a strong emphasis on resilience, and supply-chain analyses originating from operations research considering mainly the economic network's resilience. In this context,Vroegindewey and Hodbod (2018)provide a conceptual framework for supply chains with a focus on developing countries. Beyond, governance research on food security and food systems equally requires attention regarding the resilience of food systems as postulated byO'Connor et al. (2017).

This review aims (i) to identify major applications, literature strands, and conceptualizations of resilience for food systems in low-and middle-income countries. (ii) We review how diﬀerent studies jointly analyze the resilience of the agriculture-based food system and associated social-ecological systems. (iii) We identify the current use and the suitability of resilience in the context of food system and food security research as well as research gaps and future applications. 2. Methods

2.1. Systematic literature search

We conducted a systematic literature review based on a ISI Web of Science (WoS) keyword search (TOPIC) (September 04, 2019) for journal articles for the following search terms (((agricultur* AND (value chain* OR supply chain*) OR ((value chain* OR supply chain*) OR (value-chain* OR supply-chain*))) OR“food system"*) AND resilien*). It resulted in 367 publications. For the systematic review, we only used journal articles and screened them following the scheme inFig. 1. We used the review and opinion papers if applicable to relate theirﬁndings to this study and to identify the added value of this review.

2.2. Categories and criteria of the systematic review

Regarding resilience, we reviewed existing studies for the addressed resilience issue (i.e., the deﬁnition and conceptualization of resilience in

the reviewed studies), the resilience focus (i.e., the focus or challenge of the resilience assessment), the resilience item (i.e., the element assessed within the food or the associated system), which is partly in line with Meuwissen et al. (2019), but who focus on the farming system. In that respect, we analyzed, whether resilience was qualitatively or quanti-tatively analyzed and which properties of resilience were assessed: ro-bustness, adaptiveness or transformability. However, we did not only include studies emphasizing on the resilience of farming systems as Meuwissen et al. (2019)or the food system as a whole with food se-curity as resilience item or overall target (Tendall et al., 2015). We rather expected that the resilience of embedding or related systems (e.g., the environment or society) were equally crucial and may provide relevant resilience items with respect to food and farming systems. This review aimed to explore not only the conceptualization of resilience for food or farming systems but also their impact on the resilience of re-lated systems. We followedTendall et al. (2015)andMeuwissen et al. (2019)and defined robustness as a systems' ability to withstand shocks and to move to the previous state. We deviate fromMeuwissen et al. (2019) regarding our definition of adaptiveness, who define it as a system's carrying capacity to withstand shock due to changing inputs and external parameters. We rather see this as part of the robustness of a system. We see adaptiveness as a system's ability to provide a desired level of a resilience item (e.g., food, farm income) under changing input levels and external parameters (e.g., precipitation, market prices) which is in line withFolke et al. (2010)andWalker et al. (2004). Our definition is more comparable toTendall et al. (2015), who define re-sourcefulness and adaptability as resilience properties that explain them as the recovered amount of food security after a shock. In contrast toMeuwissen et al. (2019)andTendall et al. (2015), we do not narrow it to the farming and food system but look at the food and associated environmental and social systems. We also define transformability, originally defined by Folke et al. (2010), not as a system's carrying capacity (Meuwissen et al., 2019) but as a system's ability to provide a desired level or outcome of a resilience item under changed system structures and functioning. The term carrying capacity provides a strong notion from socio-ecological system's research as inFolke et al. (2010)which considers how much disturbance a system can stand. We Fig. 1. Search strategy used to identify relevant papers for the scope of this review.

(3)

emphasize more on the outcome in line withTendall et al. (2015), who see food security as an outcome dimension.

Regarding the emphasis on the value chain, we followedTendall et al. (2015) and reviewed, which elements of the value chain were assessed (producer, processor, logistics, sales, consumer). Equally, we re-viewed the environmental, social, economic, and institutional challenges. We thereby followed the elaboration byMeuwissen et al. (2019) on dimensions that cover the impacts of food systems on public and private goods. We included these sustainability challenges as in other reviews (Meyer and Leckert, 2018) since we aimed at investigating how resi-lience and sustainability are linked or overlap as done with a diﬀerent perspective inProsperi et al. (2016). The value-chain products were structured following the groups of the FAO commodity list (FAO, 2016). This categorization was extended if applicable. We also reviewed major contextual factors (“context”) and literature strands (i.e., classiﬁed the research domains targeted in the reviewed studies such as agricultural economics or rural development) and assessed the spatial scale and the countries, where a study was conducted.

2.3. Data analysis

To identify major groups of studies on the resilience of food systems, we followed three steps. First, we condensed the descriptions of qua-litative review criteria. The criteria resilience issue, focus, and item, as well as the context, and the literature strand were reduced to the main keywords with the package tm in R (R Development Core Team, 2018). We automatically cleaned the data for stop words, removed sparse terms (threshold: 0.9), and condensed the terms to the word stem. We calculated the term frequency – inverse document frequency for the remaining major terms. This indicator combines the term frequency for each paper and the relative importance in the entire set of reviewed studies per criterion (Ferreira et al., 2013). The list of studies with the cleaned input data is available in the supporting information.

Second, we followedMeyer and Leckert (2018) and conducted a multiple factor analysis to reduce dimensionality in the data with the package FactorMineR in R (R Development Core Team, 2018). We

deﬁned for the diﬀerent analysis criteria and categories the following groups: scale, value-chain element (producer, processor, logistics, sales, consumer), sustainability challenges (environmental, social, economic, institutional), product, and the results from the text mining: resilience issue, focus, and item, the context, and the literature strand. Based on the factor analysis, we conducted a hierarchical cluster analysis (Sébastien et al., 2008). To identify the optimal number of clusters, we cut the tree where the branches were visually longest and the relative gain of within cluster inertia highest. To interpret the clusters, we calculated the ar-ithmetic mean per cluster for all criteria.

3. Results 3.1. Overview

The main literature strands of the identiﬁed studies are agricultural economics and agribusiness (32), as well as rural (14) and sustainable development (12). Most studies only assessed one value-chain step (25 studies). Only few studies assessed all value-chain steps (7). Qualitatively, primary production was assessed most (65) and logistics (17) were assessed least. Most studies assessed all sustainability di-mensions and challenges. Economics were most prominent (61) and institutions least frequent (40). The spatial scale of most studies is re-gional (30) and rarely global and supra-national (9).

The mainly analyzed resilience items were the food system in-cluding supply and production (22), farmers and farming households (22), as well as crop production and varieties (9). Value-chain actors beyond farmers, if not covered within the term food system, are hardly in focus. The methods used were mostly qualitative (23) and quanti-tative surveys (22), document analysis and other desk research (13), as well as focus group discussions (12). For the product, most studies as-sessed either the food system without clear focus crops (23) or dealt with cereals (18). Less frequently, studies focused on oil-bearing crops (8), or fruits and vegetables (6 each).Fig. 2shows the spatial hot spots of the reviewed studies, which are especially Brazil and India for the given focus on low- and middle-income countries, but also Southeast Fig. 2. Number of studies on food systems and value chains per country (low- and middle-income countries).

(4)

Asian as well as South and Central American countries. 3.2. Conceptualizing resilience

In 52 of the reviewed studies resilience was explicitly qualitatively and quantitatively analyzed. The 28 qualitative studies address ro-bustness (21), adaptiveness (18), and transformability (13). The 24 quantitative studies include robustness (23), adaptiveness (7), and transformability (1). This reveals a major gap in quantifying adaptive-ness and especially transformability of food systems and value chains in low- and middle-income countries.

The 23 quantitative robustness studies were split into four major conceptual groups. Eight studies used time-series analysis and tried to identify patterns between target indicators (e.g., crop yield, food supply or prices) and stressors (mostly climate variability, but also other fac-tors such as population dynamics). Six studies assessed the robustness through composite indices and expert judgment at country level and four surveyed farmers' and other value-chain actors on factors aﬀecting their robustness. Five studies empirically assessed the robustness before and after or with and without shocks such as extreme weather events, disease outbreaks, or political crises and implemented robustness through multiple model runs or surveys before and after the shock. Four studies modeled the abstract or theoretical robustness towards external (e.g., market prices) and internal shocks (mostly supply chain conﬁg-urations) through simulation models. Major resilience items were crop yield (seven studies) and food availability (four studies). Five studies put the country including the food sector as resilience item. The re-maining studies used farm-related resilience items (livestock produc-tion and farm income), social factors (value-chain actors’ or population health), and market parameters (food demand and market prices) as resilience items.

The eight quantitative adaptiveness studies took diﬀerent perspec-tives. Four studies modeled the behavior of food-system-related resi-lience items (e.g., crop yield, market prices, demand, caloric intake) under external stressors and modeled the response of adapted systems with environmental (e.g., GHG emissions), social (e.g., caloric intake) or economic (e.g., farm income) indicators. Two studies used surveys to determine value-chain actors' adaptive capacity to external stresses through surveying demographic or economic factors. Two studies used ratings of countries’ readiness for climate change (Notre Dame Global Adaptation Initiative: assessment of economic, social and governance factors).

The two quantitative transformability studies modeled scenario-based impacts of food and related waste systems under future system configurations (scenarios). They quantified the impact of future sys-temic changes such as diet shifts or major policy changes such GHG taxes on the environment and society. They hardly quantified the transformation process.

3.3. Main groups of studies

We obtained three clusters for the reviewed studies from a hier-archical cluster analysis: cluster 1 (n: 5), cluster 2 (n:35), and cluster 3 (n: 30). For the resilience issue, item and focus, the clusters aligned comparably well and had three diﬀerent foci: crop production (cluster 1), farm households and farmers (cluster 2), and food systems (cluster 3). Cluster 2 more strongly included major resilience issues and ex-plicitly included resilience as central study focus in contrast to the other clusters. Cluster 3 saw not only the food system but also climate change as major resilience issue (Fig. 3).

Value-chain steps beyond the producer were mostly under-represented. Cluster 1 was stronger for logistics, cluster 2 for sales, and clusters 2 and 3 for consumers. Studies in all clusters consistently ad-dressed economic challenges. Cluster 1 mostly emphasized on en-vironmental challenges, and disregarded institutions and social aspects. Cluster 3 addressed the challenges in a balanced manner and

Fig. 3. Resilience issue, focus, and item (arithmetic mean of the term frequency per cluster; scale (0–1)); terms with high ratings per cluster (similar colors) should be jointly interpreted. (For interpretation of the references to color in thisﬁgure legend, the reader is referred to the Web version of this article.)

(5)

Fig. 4. Value-chain elements, sustainability challenges, and products (arithmetic mean of the term frequency per cluster; scale (0–1)); terms with high ratings per cluster (similar colors) should be jointly interpreted. (For interpretation of the references to color in thisﬁgure legend, the reader is referred to the Web version of this

(6)

emphasized social aspects. Regarding the assessed products, cluster 1 mostly emphasized cereals and cereal products, and oil bearing crops. Cluster 2 equally focused on cereals, but also on stimulant crops and fruits. Cluster 3 focused on food in general and not on speciﬁc crops (Fig. 4).

Regarding the chosen methods, cluster 1 strongly focused on mod-eling, whereas cluster 2 mostly used quantitative and qualitative sur-veys, focus group discussions, and workshops, as well as models. Cluster 3 considerably relied on document analysis compared with the other clusters. Regarding the literature strand, cluster 1 focused on global change research. Cluster 2 on agricultural economics and rural devel-opment, and cluster 3 on sustainable development research. With re-spect to the context, cluster 1 considered global impacts, and cluster 2 climate change and market-related parameters. Cluster 3 considered major context factors mostly equally. Regarding the spatial scale, most studies in cluster 1 were at the global scale and in clusters 2 and 3 at the regional scale (Fig. 5).

The identiﬁed clusters showed three major types of studies mainly diﬀerentiated by the type of external shocks and drivers, the assessment methods, the resilience focus, and the sustainability challenges (seeFig. 6): studies in cluster 1 mainly analyzed global (change) impacts on crop or primary production. Downstream value-chain steps were hardly con-sidered. Cluster 1 equally used rather easily accessible and globally relevant crops such as cereals. Thereby, the chosen tools were mainly related to crop modeling and (spatial) statistical analyses. Studies in cluster 2 mainly analyzed the resilience of farmers and farming households under global change phenomena such as climate change and equally explored farmers’ adaptative capacity with respect to chosen crops (e.g., produced amount and sales). They mostly relied on social science methods such as surveys or workshop formats. Studies in cluster 3 took a more holistic perspective and saw the resilience of the entire food system as important. They emphasized on social challenges, but also included environmental and economic challenges. Equally, these studies covered more often several steps of the value chain. The chosen methods were mostly document analysis and social science methods.

4. Future research needs

Generally, we discussed how the reviewed studies conceptualize food system resilience (section4.1). Central element in the food system is the value chain (see alsoTendall et al. (2015)) and its relevance in the reviewed studies is discussed (section4.2). Then, we investigate how the food system resilience links to the embedding environmental (sec-tion4.3) and social systems (section4.4) and how the interaction (e.g., feedbacks) can be assessed (section4.5). Beyond the conceptualization, food system resilience is addressed regarding quantiﬁcation (section 4.6) and governance (section4.7).

4.1. Conceptualization of resilience in food system studies

The definition of resilience in the reviewed studies was often very generic aiming at food system resilience in general without a clear definition of resilience properties. A clear specification of essential re-silience properties for food systems and associated environmental and social systems comparable to the one given for farming systems by Meuwissen et al. (2019)is needed. We developed such definition in section 2.2. We reviewed the resilience properties robustness, trans-formability, and adaptiveness. About one third of the studies quantified resilience and emphasized robustness. Few quantified transformability (two studies) and adaptiveness (eight studies). Robustness was assessed for food availability (especially crop yields) or market prices at country or global level given available data for time-series analyses. Other re-silience items such as value-chain actors' health, or farm income were either assessed before and after a shock or along time series. Others used composite indices and expert judgment to quantify country-level Fig. 5. Method, literature strand, and context (arithmetic mean of the term

frequency per cluster; scale (0–1)); terms with high ratings per cluster (similar colors) should be jointly interpreted. (For interpretation of the references to color in thisﬁgure legend, the reader is referred to the Web version of this article.)

(7)

robustness. Few studies dynamically quantiﬁed adaptiveness by mod-eling the behavior of resilience items (e.g., crop yield) under changing climate. Composite indices were used to rate countries' adaptive capa-city and surveys to determine value-chain actors’ adaptive capacity to shocks such as extreme weather events. Two reviewed studies only quantiﬁed transformability and investigated the outcome rather than the transformation process.

In that respect, future research should emphasis especially on modeling and quantifying the transformability and the adaptiveness of food systems’ resilience. Beyond, the current focus on rather easily accessible resilience items (e.g., crop yields) should be broadened and other relevant resilience items addressed in the following sections. 4.2. Value-chain perspective

The reviewed studies mostly miss a holistic perspective on the re-silience of food systems and do not consider value-chain steps in a balanced manner.Tendall et al. (2015)stress the need to take diﬀerent perspectives: the food system, the value chain, and individual value-chain actors. Improved linkages in food systems (e.g., regionally in-tegrated value chains) may foster resilience (Abu Hatab et al., 2019b). Nevertheless, existing studies mostly emphasize individual value-chain actors (mostly farmers) and less frequently consider other steps of the value chain. This is equally stated for reviews with emphasis on food security (Reardon et al., 2019) and sustainability transitions and food security (El Bilali, 2019). One reason for this focus on farmers might be that farmers are often seen as most vulnerable value chain actor from an economic point of view (e.g., size-related low bargaining power compared with processors or retailers) (Trienekens, 2011). A con-siderable rise of consolidating logistics, processing, and sales entities is a challenge for smallholders. Farmers in low- and middle-income countries miss the assets for modernization (Reardon et al., 2009), which may hinder their transformation and adaptation. In addition, research on food security most traditionally focusses on agriculture and rural development (Ericksen et al., 2009). In that respect, the resilience downstream the value chain is hardly modeled. A small share of studies from operations research (logistics and processing) (e.g.,Aboah et al.

(2019)) or market research (sales and consumer studies) took the entire value chain into account, but miss the embedding social and environ-mental systems as discussed in the following sections.

4.3. The environmental perspective on food system resilience

Future research needs to broaden the perspective and to assess how (i) food system transformation impacts environmental resilience; (ii) studies need to determine the impact of environmental drivers (i.e., others than climate change) on food systems' resilience.

(i) The reviewed studies mostly emphasize food systems or farmers and assess their resilience under external shocks which are mostly economic or environmental. The studies (clusters 2 and 3) often assess the adaptation of value-chain actors, especially farmers, under climate change. Studies rather disregard the impact of transforming food systems and of changes in farmers' behavior on environmental resilience. Environmental or biophysical capacity underpins food systems' resilience (e.g., through water or land availability) (Seekell et al., 2017) and should be considered. For example, changes from drought adapted crops such as Teﬀ to Maize in Ethiopia likely aﬀect water availability.

(ii) Regarding environmental drivers, most studies focus on the impact of climate change on farming households or on farmers and on the food system. The adaptation of farmers to climate change is in focus (cluster 2) and also considerably assessed in theﬁeld of risk-related research on farmers' behavior under climate change (Eitzinger et al., 2018). Beyond climate change and GHG emissions and other global impacts (cluster 1), the regional scale should be emphasized (Prosperi et al., 2016) and other relevant environmental impacts assessed: soil quality, water quality and availability, and biodi-versity (Meyer and Leckert, 2018).

4.4. The social perspective on food system resilience

Societal changes and demographic trends such as changes in con-sumption patterns, in income or age are hardly considered in the Fig. 6. Thematic clusters of existing studies on food system resilience grouped by external shocks, assessment methods, focus and sustainability challenges; the resilience focus is drawn in color (Cluster 1: global models and primary production; cluster 2: farmers' behavior under global change; cluster 3: food systems and value chains); icon source:https://icons8.com/. (For interpretation of the references to color in thisﬁgure legend, the reader is referred to the Web version of this article.)

(8)

reviewed studies. This research gap is equally highlighted for food system research in general (Ruben et al., 2019) and was also earlier stated (Ericksen et al., 2009) but not sufficiently addressed in the re-viewed studies.Springmann et al. (2018)analyze the impact of dietary changes and waste reduction on planetary boundaries at the global scale. However, the drawn assumptions on food system transformations and consumer behavior are hardly suitable for governance and man-agement at lower spatial levels (e.g., regions). The impacts on food system configuration and resilience likely differ between regions. Therefore, studies at the regional level are needed to provide sufficient spatial resolution and differentiation (e.g., to draw conclusions for land-use systems and regional or national diet shifts).

4.5. Feedbacks: food system impacts on society and environment and vice versa

Environmental impacts are not only the outcome of food system resilience, but also drivers of food system resilience in contrast to Tendall et al. (2015). This bidirectionality (food system↔ environ-ment) is not suﬃciently addressed in the reviewed studies. The re-viewed quantitative studies currently still follow the input-output per-spective. Environmental parameters are either taken as drivers or inﬂuencing factor of food system resilience (clusters 1 to 3) and in fewer cases also as outcome (cluster 1: global studies).

Linkages between environmental and social systems should be more strongly considered. Also feedbacks that may lead to leakage eﬀects should be considered. A prominent example is indirect land-use change due to bioenergy production (Searchinger et al., 2008). The narrow perspective on the bioenergy system allowed to overlook this leakage eﬀect on food production and the environment (especially higher GHG emissions instead of savings due to cropland displacing forest driven by energy crops on previous cropland).

4.6. Quantiﬁcation of impacts

4.6.1. Spatial and temporal quantiﬁcation of resilience

Most assessments in the reviewed studies are either qualitative at the regional scale or quantitative at the global scale. The regional scale assessments if focusing on food systems rather capture the state of the regional food system and the adaptation behavior of farmers or con-sumers, mostly based on surveys or focus groups discussions. The re-levant quantification of the impact of food system transformation on the resilience of food systems is mostly missing at the regional scale. There is mainly quantification, partly spatially explicit, at the global scale (e.g., impacts on virtual water (Zhao et al., 2019) or the global production of major crops under different environmental or economic shocks (Mehrabi and Ramankutty, 2019)). However, most spatially explicit studies such asMehrabi and Ramankutty (2019) only model crop production and the farming system. The spatial modeling of the impact of food systems and value chains is missing and hardly done at the regional scale. The current separation of the regional and global perspective might be due to low data availability and due to the di ffi-culty to align different preferences and priorities for resilience between world regions (e.g., food security in Sub-Saharan Africa vs. sustainable production in Western Europe) (Challies et al., 2014). Future studies should also quantify the impact of changing consumption patterns and model its pathways through value chains to grasp the impact on the resilience of land systems and rural societies. Here, agent-based land-use modeling could be an option. It allows to depict complex decision making along the value chain, but is hardly done (Utomo et al., 2018). Equally, it allows to show interactions with land-use systems (e.g., for cropping or pastoral systems (Schulze et al., 2016;Thober et al., 2018)). For the temporal dynamics, the reviewed studies mostly assess the current status of farmers' resilience (especially cluster 2) and in fewer cases the resilience of food systems (cluster 3) (e.g., Jacobi et al. (2018)). Current studies rather qualitatively describe spatial dynamics,

e.g.,Padulosi et al. (2015). Future research should quantify temporal dynamics of food systems' resilience to identify more or less resilient spatial conﬁgurations. In addition, the future resilience of food systems, e.g., could be quantiﬁed through modeling approaches such as scenario techniques to explore and to assess the future development of food systems (Mason-D'Croz et al., 2016).

4.6.2. Quantiﬁcation of drivers and internal dynamics within the value chain

The impact of societal preferences and trends (e.g., consumption) and environmental impacts (e.g., climate change) on food system resi-lience is mostly missing as previously noted byEricksen et al. (2009) andVroegindewey and Hodbod (2018). Only approaches from opera-tions research on optimizing and modeling food system networks exist (push perspective) and were recently reviewed (Aboah et al., 2019). Another review focusses on risks in supply chains but not on resilience or the systemic behavior of associated environmental or social systems (Behzadi et al., 2018). To close this gap, it is necessary to integrate diﬀerent frameworks such as food security and socio-ecological systems research (Prosperi et al., 2016) and to include the pull perspective on value chains (e.g., through modeling consumer demand and diet shifts with agent-based models).

4.7. Governance research and resilience

It is questionable whether governance should foster resilience as it does not necessarily improve the situation for all individuals in society or in the environment (Béné et al., 2016). Others see resilience as a tool to overcome or counteract market asymmetries or weak governance such as the terms of trade or neoliberal market structures (O'Connor et al., 2017). However, this requires studies at the appropriate scale and resolution to see whether more resilient food systems are less prone to weak governance. The reviewed studies hardly quantify the impact of governance on resilience. There is a need for indicators on the impact of governance on the distribution and consumption of food, to quantify the agency of actors, and to assess scale-overarching impacts of gov-ernance. The agency of actors is qualitatively assessed in the reviewed studies. This is partly in line with Delaney et al. (2018). Scale-over-arching factors such as market failures (e.g., negative environmental and social externalities) require an integration of global and regional perspectives. The impact assessment of global governance on regional resilience is mostly missing as equally stated bySeekell et al. (2017). Therefore, future research should analyze the impact of policies on food system resilience in a joint manner: policy issuers and aﬀected regions (e.g., the Common Agricultural Policy of the European Union on pro-ducing regions in low- and middle-income countries) as done for the environmental impacts of EU bioenergy imports (Meyer et al., 2016). 5. Conclusions

Future research needs to emphasize on the link between the food system and associated environmental and social systems to a larger extent. The resilience of food systems should not be seen in an isolated manner and major linkages to the embedding environmental (e.g., water quality and availability) and social systems (e.g., labor markets) require consideration. On this path, future research should quantify feedbacks between economic entities, society, and the environment. Within the food system, future research should assess the resilience of value chains and go upstream the value chain beyond primary pro-duction and also consider the impact of consumption on food system resilience (pull perspective (e.g., assessing the impact of diet shifts on food system resilience)). The quantiﬁcation of resilience at the sub-national level was hardly done and requires quantitative modeling (e.g., agent-based models) with increased temporal resolution and spatial explicitness. Conceptually, future studies should not only analyze ro-bustness but also quantify transformability and adaptiveness as equally

(9)

relevant resilience properties. Declaration of competing interest

The authors declare that they have no known competingﬁnancial interests or personal relationships that could have appeared to in ﬂu-ence the work reported in this paper.

Acknowledgments

The author would like to thank Andy Nelson and Rick Hogeboom for fruitful discussions on the topic of this paper. Equally, the author likes to thank two anonymous reviewers for helpful comments to im-prove this paper. This work was funded by the 4TU Centre for Resilience Engineering (DeSIRE).

Appendix A. Supplementary data

Supplementary data to this article can be found online athttps:// doi.org/10.1016/j.gfs.2020.100356.

References

Aboah, J., Wilson, M.M.J., Rich, K.M., Lyne, M.C., 2019. Operationalising resilience in tropical agricultural value chains. Supply Chain Manag.. Int. J. 24 (2), 271–300.

https://doi.org/10.1108/SCM-05-2018-0204.

Abu Hatab, A., Cavinato, M.E.R., Lagerkvist, C.J., 2019a. Urbanization, livestock systems and food security in developing countries: a systematic review of the literature. Food Sec 11 (2), 279–299.https://doi.org/10.1007/s12571-019-00906-1.

Abu Hatab, A., Cavinato, M.E.R., Lindemer, A., Lagerkvist, C.-J., 2019b. Urban sprawl, food security and agricultural systems in developing countries: a systematic review of the literature. Cities 94, 129–142.https://doi.org/10.1016/j.cities.2019.06.001. Alexander, P., Rounsevell, M.D.A., Dislich, C., Dodson, J.R., Engström, K., Moran, D.,

2015. Drivers for global agricultural land use change: the nexus of diet, population, yield and bioenergy. Glob. Environ. Change-Hum. Policy Dimens. 35, 138–147.

https://doi.org/10.1016/j.gloenvcha.2015.08.011.

Behzadi, G., O'Sullivan, M.J., Olsen, T.L., Zhang, A., 2018. Agribusiness supply chain risk management: a review of quantitative decision models. Omega-Int. J. Manag. Sci. 79, 21–42.https://doi.org/10.1016/j.omega.2017.07.005.

Béné, C., Headey, D., Haddad, L., Grebmer, K. von, 2016. Is resilience a useful concept in the context of food security and nutrition programmes?: some conceptual and prac-tical considerations. Food Sec 8 (1), 123–138. https://doi.org/10.1007/s12571-015-0526-x.

Bolwig, S., Ponte, S., Du Toit, A., Riisgaard, L., Halberg, N., 2010. Integrating poverty and environmental concerns into value-chain analysis: a conceptual framework. Dev. Pol. Rev. 28 (2), 173–194.https://doi.org/10.1111/j.1467-7679.2010.00480.x. Challies, E., Newig, J., Lenschow, A., 2014. What role for social-ecological systems

re-search in governing global teleconnections? Glob. Environ. Change-Hum. Policy Dimens. 27, 32–40.https://doi.org/10.1016/j.gloenvcha.2014.04.015. Delaney, A., Evans, T., McGreevy, J., Blekking, J., Schlachter, T., Korhonen-Kurki, K.,

Tamas, P.A., Crane, T.A., Eakin, H., Forch, W., Jones, L., Nelson, D.R., Oberlack, C., Purdon, M., Rist, S., 2018. Governance of food systems across scales in times of social-ecological change: a review of indicators. Food Security 10 (2), 287–310.https://doi. org/10.1007/s12571-018-0770-y.

Eitzinger, A., Binder, C.R., Meyer, M.A., 2018. Risk perception and decision-making: do farmers consider risks from climate change? Climatic Change 151 (3–4), 507–524.

https://doi.org/10.1007/s10584-018-2320-1.

El Bilali, H., 2019. Research on agro-food sustainability transitions: where are food se-curity and nutrition? Food Sese-curity 11 (3), 559–577.https://doi.org/10.1007/ s12571-019-00922-1.

Ericksen, P.J., Ingram, J.S.I., Liverman, D.M., 2009. Food security and global environ-mental change: emerging challenges. Environ. Sci. Pol. 12 (4), 373–377.https://doi. org/10.1016/j.envsci.2009.04.007.

FAO, 2016. FAOSTAT Commodity List.

Ferreira, R., Souza Cabral, L. de, Lins, R.D., Pereira e Silva, G., Freitas, F., Cavalcanti, G.D.C., Lima, R., Simske, S.J., Favaro, L., 2013. Assessing sentence scoring techniques for extractive text summarization. Expert Syst. Appl. 40 (14), 5755–5764.https:// doi.org/10.1016/j.eswa.2013.04.023.

Folke, C., Walker, B., Scheﬀer, M., Chapin, T., Rockström, J., 2010. Resilience thinking: integrating resilience, adaptability and transformability. E&S 15 (4), 20. Hooks, T., Macken-Walsh, A., McCarthy, O., Power, C., 2017. Farm-level viability,

sus-tainability and resilience: a focus on co-operative action and values-based supply chains. Studies in Agricultural Economics 119 (3), 123–129.https://doi.org/10. 7896/j.1718.

Jacobi, J., Mukhovi, S., Llanque, A., Augstburger, H., Kaeser, F., Pozo, C., Peter, M.N., Delgado, J.M.F., Kiteme, B.P., Rist, S., Speranza, C.I., 2018. Operationalizing food system resilience: an indicator-based assessment in agroindustrial, smallholder farming, and agroecological contexts in Bolivia and Kenya. Land Use Pol. 79, 433–446.https://doi.org/10.1016/j.landusepol.2018.08.044.

Mason-D'Croz, D., Vervoort, J., Palazzo, A., Islam, S., Lord, S., Helfgott, A., Havlík, P., Peou, R., Sassen, M., Veeger, M., van Soesbergen, A., Arnell, A.P., Stuch, B., Arslan,

A., Lipper, L., 2016. Multi-factor, multi-state, multi-model scenarios: exploring food and climate futures for Southeast Asia. Environ. Model. Software 83, 255–270.

https://doi.org/10.1016/j.envsoft.2016.05.008.

Mehrabi, Z., Ramankutty, N., 2019. Synchronized failure of global crop production. Nat. Ecol. Evol. 3 (5), 780–786.https://doi.org/10.1038/s41559-019-0862-x. Meuwissen, M.P.M., Feindt, P.H., Spiegel, A., Termeer, C.J.A.M., Mathijs, E., Mey, Y.d.,

Finger, R., Balmann, A., Wauters, E., Urquhart, J., Vigani, M., Zawalińska, K., Herrera, H., Nicholas-Davies, P., Hansson, H., Paas, W., Slijper, T., Coopmans, I., Vroege, W., Ciechomska, A., Accatino, F., Kopainsky, B., Poortvliet, P.M., Candel, J.J.L., Maye, D., Severini, S., Senni, S., Soriano, B., Lagerkvist, C.-J., Peneva, M., Gavrilescu, C., Reidsma, P., 2019. A framework to assess the resilience of farming systems. Agric. Syst. 176https://doi.org/10.1016/j.agsy.2019.102656.102656. Meyer, M.A., Leckert, F.S., 2018. A systematic review of the conceptual diﬀerences of

environmental assessment and ecosystem service studies of biofuel and bioenergy production. Biomass Bioenergy 114, 8–17.https://doi.org/10.1016/j.biombioe. 2017.05.003.

Meyer, M.A., Seppelt, R., Witing, F., Priess, J.A., 2016. Making environmental assess-ments of biomass production systems comparable worldwide. Environ. Res. Lett. 11 (3).https://doi.org/10.1088/1748-9326/11/3/034005.34005.

O'Connor, D., Boyle, P., Ilcan, S., Oliver, M., 2017. Living with insecurity: food security, resilience, and the world food programme (WFP). Global Soc. Pol. 17 (1), 3–20.

https://doi.org/10.1177/1468018116658776.

Padulosi, S., Mal, B., King, O.I., Gotor, E., 2015. Minor millets as a central element for sustainably enhanced incomes, empowerment, and nutrition in rural India. Sustainability 7 (7), 8904–8933.https://doi.org/10.3390/su7078904.

Popkin, B.M., 2017. Relationship between shifts in food system dynamics and accelera-tion of the global nutriaccelera-tion transiaccelera-tion. Nutr. Rev. 75 (2), 73–82.https://doi.org/10. 1093/nutrit/nuw064.

Prosperi, P., Allen, T., Cogill, B., Padilla, M., Peri, I., 2016. Towards metrics of sustainable food systems: a review of the resilience and vulnerability literature. Environ Syst Decis 36 (1), 3–19.https://doi.org/10.1007/s10669-016-9584-7.

R Development Core Team, 2018. A language and environment for statistical computing.

https://www.R-project.org.

Reardon, T., Barrett, C.B., Berdegué, J.A., Swinnen, J.F.M., 2009. Agrifood industry transformation and small farmers in developing countries. World Dev. 37 (11), 1717–1727.https://doi.org/10.1016/j.worlddev.2008.08.023.

Reardon, T., Echeverria, R., Berdegué, J., Minten, B., Liverpool-Tasie, S., Tschirley, D., Zilberman, D., 2019. Rapid transformation of food systems in developing regions: highlighting the role of agricultural research & innovations. Agric. Syst. 172, 47–59.

https://doi.org/10.1016/j.agsy.2018.01.022.

Röös, E., Bajželj, B., Smith, P., Patel, M., Little, D., Garnett, T., 2017. Greedy or needy?: land use and climate impacts of food in 2050 under diﬀerent livestock futures. Global Environ. Change 47, 1–12.https://doi.org/10.1016/j.gloenvcha.2017.09.001. Ruben, R., Verhagen, J., Plaisier, C., 2019. The challenge of food systems research: what

diﬀerence does it make? Sustainability 11 (1).https://doi.org/10.3390/su11010171. Schipanski, M.E., MacDonald, G.K., Rosenzweig, S., Chappell, M.J., Bennett, E.M., Kerr, R.B., Blesh, J., Crews, T., Drinkwater, L., Lundgren, J.G., Schnarr, C., 2016. Realizing resilient food systems. Bioscience 66 (7), 600–610.https://doi.org/10.1093/biosci/ biw052.

Schulze, J., Frank, K., Priess, J.A., Meyer, M.A., 2016. Assessing regional-scale impacts of short rotation coppices on ecosystem services by modeling land-use decisions. PloS One 11 (4), e0153862.https://doi.org/10.1371/journal.pone.0153862.

Searchinger, T., Heimlich, R., Houghton, R.A., Dong, F., Elobeid, A., Fabiosa, J., Tokgoz, S., Hayes, D., Yu, T.-H., 2008. Use of U.S. croplands for biofuels increases greenhouse gases through emissions from land-use change. Science 319 (5867), 1238–1240.

https://doi.org/10.1126/science.1151861.

Sébastien, Lê, Josse, Julie, Husson, François, 2008. FactoMineR: an R package for mul-tivariate analysis. J. Stat. Software 25 (1), 1–18.https://doi.org/10.18637/jss.v025. i01.

Seekell, D., Carr, J., Dell'Angelo, J., D'Odorico, P., Fader, M., Gephart, J., Kummu, M., Magliocca, N., Porkka, M., Puma, M., Ratajczak, Z., Rulli, M.C., Suweis, S., Tavoni, A., 2017. Resilience in the global food system. Environ. Res. Lett. 12 (2).https://doi. org/10.1088/1748-9326/aa5730.

Springmann, M., Clark, M., Mason-D'Croz, D., Wiebe, K., Bodirsky, B.L., Lassaletta, L., Vries, W. de, Vermeulen, S.J., Herrero, M., Carlson, K.M., Jonell, M., Troell, M., DeClerck, F., Gordon, L.J., Zurayk, R., Scarborough, P., Rayner, M., Loken, B., Fanzo, J., Godfray, H.C.J., Tilman, D., Rockström, J., Willett, W., 2018. Options for keeping the food system within environmental limits. Nature 562 (7728), 519–525.https:// doi.org/10.1038/s41586-018-0594-0.

Tendall, D.M., Joerin, J., Kopainsky, B., Edwards, P., Shreck, A., Le, Q.B., Kruetli, P., Grant, M., Six, J., 2015. Food system resilience: deﬁning the concept. Glob. Food Sec. Agric. Policy Econ. Environ. 6, 17–23.https://doi.org/10.1016/j.gfs.2015.08.001. Thober, J., Schwarz, N., Hermans, K., 2018. Agent-based modeling of

environment-mi-gration linkages: a review. E&S 23 (2).https://doi.org/10.5751/ES-10200-230241.

Trienekens, J.H., 2011. Agricultural value chains in developing countries: a framework for analysis. Int. Food Agribus. Manag. Rev. 14 (2), 51–82.

UN, 2019. World Population Prospects 2019. Ten Key Findings, New York City. Utomo, D.S., Onggo, B.S., Eldridge, S., 2018. Applications of agent-based modelling and

simulation in the agri-food supply chains. Eur. J. Oper. Res. 269 (3), 794–805.

https://doi.org/10.1016/j.ejor.2017.10.041.

Vroegindewey, R., Hodbod, J., 2018. Resilience of agricultural value chains in developing country contexts: a framework and assessment approach. Sustainability 10 (4).

https://doi.org/10.3390/su10040916.

Walker, B., Holling, C.S., Carpenter, S.R., Kinzig, A., 2004. Resilience, adaptability and transformability in social–ecological systems. E&S 9 (2), 5.

Zhao, H., Qu, S., Guo, S., Zhao, H., Liang, S., Xu, M., 2019. Virtual water scarcity risk to global trade under climate change. J. Clean. Prod. 230, 1013–1026.https://doi.org/ 10.1016/j.jclepro.2019.05.114.