Tradeoffs and synergies between food security and forest cover in Brazilian drylands

Global food demand is expected to increase in the next decades, pushing agricultural expansion and deforestation. However, food production in agricultural lands is just one dimension of food security, to which forest goods and services also contribute. In this paper, we aimed to explore the relationship between forest cover and food security. We hypothesized that food security is improved by both human‐made and green infrastructure combined. To test this relationship, we explore the relationships between forest cover and a multidimensional index of food security that included both socioeconomic and natural variables taken from Brazilian official databases for 1141 municipalities of the Brazilian Caatinga (a seasonally tropical dry forest). The index was formed by 12 principal components axes (12 PCs) and we found that financial poverty (PC 1) and economic inequality (PC 2) were the main determinants of food insecurity in Caatinga. We found that lowest food security values were found in two contrasting contexts: one is represented by poor and unequal municipalities with high forest cover while the other refers to poor and less unequal municipalities but with little forest cover. Municipalities with intermediate levels of forest cover had slightly higher food security, a consistent pattern across time (2006–2017). Win–win scenarios where both forest cover and food security increased with time were almost as common as lose–lose situations (25% and 22% respectively). This suggests a sort of balance between forests and human‐made land uses and reinforces that natural capital contributes to food security. Zero‐hunger is a main issue for sustainable development goals, and our results adds to the notion that both sustainable use of forests and socioeconomic improvements must coexist rather than being treated as antagonistic policies.

More food production, however, does not guarantee access, and conversion of natural habitats into agricultural fields has limited impact on food security due to access limitations (FAO, 2020).On the other hand, natural habitats can play a major role on food security for rural populations as sources of plants, fisheries, wild meat and insects, for example (Baudron et al., 2019).Access to forests can alleviate poverty by allowing many kinds of traditional management practices such as slash-and-burn agriculture, extensive pastoralism, and diverse types of extractivism (Jagger et al., 2022).Understanding the relationships between food security and forests is crucial to achieve biodiversityfriendly schemes of food production such as: sustainable intensification of agricultural production (Pretty & Bharucha, 2014) and crop yield improvement (Schütz et al., 2018) that must reduce demand for new lands and, therefore, halt deforestation (Lambin et al., 2018).
Fighting climate change while the population grows demands conciliating food security and protection of forests (Melo et al., 2021).For this, we need a paradigm shift that places food security in all its dimensions, including not only the access to the products of agricultural fields but the access to forest goods and services.
According to the Food and Agriculture Organisation of the United Nations (FAO, 2013), there are four main dimensions of food security.
First, availability, refers to the amount of food available in the system, considering, among others, food production and population size (Burchi & De Muro, 2016).Then, utilization refers to how people cook, process, and store the food available in the system, but this dimension is also related to water and sanitation issues, both affecting food utilization and health (Ericksen, 2008;FAO, 2020).A third aspect, stability, deals with the capacity of the system to guarantee food supply in the face of several types of disturbances, such as climate, market, or political instabilities (Kah, 2017;Tendall et al., 2015).Finally, access is the capacity of people to access the food produced in the system, either buying it or being capable of producing it themselves (FAO, 2013;Klassen & Murphy, 2020).The last is thought to be the main cause of food insecurity because the amount of food needed to meet current basic world population demands is actually produced (Barrett, 2010).
Economic poverty precludes access to the food market while utilization of natural resources can alleviate food insecurity of rural poor (Miller & Hajjar, 2020).Therefore, addressing food insecurity in a broad way demands understanding the nature of the relationship between extent of natural habitats and the several dimensions of food security.
Although the scientific literature highlights the tradeoff between the establishment of new crop fields and the increasing deforestation (Meyfroidt, 2018), forest can help to improve food security (Miller et al., 2020;Rasmussen et al., 2020).Agroforestry systems, for example, can contribute to improve availability, access, and stability of food systems at regional scales (Rosenstock et al., 2019).In many poor regions of tropical countries, forests are used as grazing fields for extensive pastoralism practices while helping to maintain forest cover (Alencar et al., 2022;Baudron et al., 2019).Despite this important linkages between forests and food security have been recently recognized (Bahar et al., 2020), the forest-food nexus needs to be better explored (Melo et al., 2021).Important knowledge gaps on the role of natural habitats for poverty alleviation and food security still persist and limit the quantification of forests to food security.Because a diverse set of natural and socioeconomic drivers can affect food security, we need to test and re-create ways of measuring the determinants of access to food.
Human-made or "grey" infrastructure is surely required to improve food security (Devereux, 2016).For example, roads help to guarantee access to markets and water dams to irrigation schemes (de Fraiture & Wichelns, 2010;Khan et al., 2009).The same is true for the "green" infrastructure (sensu da Silva & Barbosa, 2017) if we consider natural habitats as complementary sources of food items.However, current development models usually replace "green" by "grey" infrastructure and threaten landscapes of crossing a tipping point that compromises the ability of natural habitats to provide the services and goods that may improve food security (Swift & Hannon, 2010).Little grey infrastructure is a signal of underdevelopment that might reduce food availability due to limited access to food markets (Khan et al., 2009).On the other hand, little remaining natural habitats may represent a lack of complementary source of food for people, thus also reducing food security (Vysochyna et al., 2020).If it is true, it is reasonable to expect that a combination of better social indicators and enough natural habitats must provide better food security.At least theoretically, this is in accordance with the concept of "optimal landscapes" that both preserves natural habitats and produce food in a landscape structure that allows the high levels of food production, ecosystem services, and biodiversity conservation (Arroyo-Rodríguez et al., 2020).
The Brazilian seasonally dry tropical forest, also known as Caatinga, constitutes an opportunity for assessing the tradeoffs and synergies between food security and forests.This region is characterized by high levels of social vulnerability (Hummell et al., 2016) and low food security when compared with the other regions of Brazil (Gubert et al., 2017).Around 60% of its forest cover is preserved, though severely degraded and fragmented (Antongiovanni et al., 2018).People in this region depend largely on small-scale agriculture and extensive pastoralism that are periodically affected by seasonal droughts, thus reducing food availability (Melo, 2017).Natural resources are therefore an important asset for the 28 million people living in this dry forest and are likely to provide both goods and services that contribute to food security.The objective of this work is to understand the nature of the relationship between food security and forests at the regional scale.For this we focused on: (1) identifying the main socioeconomic indicators of food insecurity in the Caatinga through a multidimensional index of food security and; (2) understanding the spatial configuration of food insecurity and deforestation.Our results offer important evidence on the contribution of forests to food security that should be useful to regional landscape management and challenge the notion that current models of development based on landuse change can alleviate poverty and food insecurity.

| The Caatinga socioecological system
The Caatinga dry forest is characterized by an extended dry period in which rainfall is scarce, high evaporation rates, and great inter-and intra-annual rainfall variation (de Andrade et al., 2017).The rainfall amount goes from 400 to 1500 mm per year across the biome, being considered the wettest seasonally dry forest in the world (de Andrade et al., 2017).The vegetation is dominated by small-leaved, thorny trees, and several species of succulents (Queiroz et al., 2017).Water for irrigation comes mostly from rivers and wells, but as almost all rivers are intermittent, the largest share of the agriculture is solely rainfed or relies on rainfall accumulated in the wetter valleys (Sampaio et al., 2017).Water for human consumption comes from diverse sources (e.g., wells, cisterns, water tank trucks), but they are not necessarily of adequate quality, which can impact local food security (Sena et al., 2018).The Caatinga has very diverse soil conditions, but most of the areas present low fertility which reduces the agricultural aptitude (Sampaio et al., 2017).Combined, those environmental characteristics contribute to the low productivity of the agricultural system and to poor socioeconomic conditions of most farmers (Tabarelli et al., 2017).
The set of municipalities comprising the Caatinga have the lowest levels of human development in Brazil (da Silva et al., 2017).Most of the municipalities are considered rural (IBGE, 2017), where most farmers produce maize and beans associated with small-scale animal production for their subsistence (Sampaio et al., 2017).Goat production is the main strategy that poor rural farmers use worldwide in arid and semiarid regions (Caatinga included) especially during long droughts because of their tolerance to such climatic conditions and their ability to feed on natural vegetation (Devendra, 1999).Bovine production is mostly developed in a pasture with exotic grasses; however, only capitalized farmers normally engage in cattle production (Sampaio et al., 2017).Furthermore, the Caatinga occurs in one of the most densely populated semi-arid regions in the world (da Silva et al., 2017) with millions of people living in rural communities highly dependent on forest products (de Albuquerque et al., 2017).Although the Caatinga still has intermediate levels of forest vegetation cover, the forest fragments are not evenly distributed across the biome (Antongiovanni et al., 2018) and have high levels of chronic anthropogenic disturbance linked to the historical use of native vegetation by local people (Antongiovanni et al., 2020).

| Food security index and forest cover change
We developed an index of food security for the Caatinga using the four dimensions described in FAO (2013) to select the variables to represent each dimension.All variables were collected from the Brazilian official sources (e.g., IBGE, Ministry of Health) using 1141 municipalities of the Caatinga as sample units (Figure 1).We selected 38 variables that are summarized in Table 1.The respective premise and sources of each variable are in Table S1.It is important to notice that some variables seem to be redundant or misleading.For example, it may look redundant to include variables on both high and low Body Mass Index, but actually they capture two extremes of a healthy weight for adults which has different causes and consequences.Overweight is now considered one of the triple burdens of food insecurity (i.e., undernutrition, hidden hunger, and overweight) and is related to a diet rich in ultra-processed food, causing many health problems in adulthood (UNICEF, 2019).The lack of nutritious food, on the other hand, could cause undernourishment and low weight, especially if the lack of food occurs during early childhood, which could have negative health consequences for lifetime (FAO, 2013).Another example is the variable related to the social organization of farmers.It is still debated in the literature if cooperatives always bring benefits to small-farmers, however, in general, they do contribute to rural development and to reduce poverty among small-holder farmers (Johnson & Shaw, 2014).Some studies argue the issue of inclusiveness to be a key point in determining the real contribution of cooperativism to rural development (Bijman & Wijers, 2019).In Brazil, cooperativism is considered an important instrument of rural development and tends to contribute positively to small farmers, despite some farmers are indeed excluded (Stattman & Mol, 2014).Therefore, even if this contribution is not for all farmers or not evenly distributed among them, we believe that the organization of producers should have a positive effect on food systems and assume a positive impact on food security.
We built a multidimensional food security index for two timepoints (2006-2017) to understand how it changed over time and how it is related to forest cover change.We built one index using a Principal Component Analysis for each year to reduce a large number of variables into a few dimensions following a method proposed by Hummell et al. (2016).We selected only the PCs with an eigenvalue greater than 1 (Cutter et al., 2003) and then we changed the cardinality of each PC depending on whether the variables that compose the PC contribute positively or negatively to food security (Table S2).
The final food security index for each year was calculated as the sum of the individual scores of each PC for each municipality based on Hummell et al. (2016).We estimated the food security change in the municipalities by calculating the difference between the final index score between 2006 and 2017.
We gathered the data of native vegetation cover from the Map-Biomas project (Souza et al., 2020).We considered all categories of forest (forest, savanna, mangrove, and forest plantations) and nonforested native vegetation (wetland, grassland, salt flats, rocky outcrops, and other non-forest formations) to calculate the forest cover percentage for each municipality.We grouped all types of native vegetation into one class (i.e., native vegetation cover) because people in the Caatinga derive many uses related to food security from all types of native vegetation and not only from forests (de Albuquerque et al., 2017).Then, we identified and spatialized the municipalities with synergies (win-win and lose-lose) and tradeoffs (win-lose and lose-win) between forest cover and food security change, respectively.We considered all municipalities that gained food security from 2006 to 2017 as a "win" group, irrespective of the scores' size, as well as "lose" if the food security score was below zero.

| Statistical analyses
We first used a Principal Components Analysis to reduce dimensionality of our Multidimensional Food Security Index (MFSI).We used z-transformation to standardize scores of the principal components and calculate the MSFI.Because we changed the cardinality of the dimensions to always increase MFSI, the index is the sum of both positive (increasing security) and negative values (decreasing security).
Because the dimensions were formed by different variables across years we provided a table to help readers interpret what drives food insecurity (Table S1).The absolute changes of values of MFSI and forest cover were used to create maps that help to understand spatial distribution of both changes in forest cover and food security.To understand the relationship between forest cover and MFSI and its two most important principal components in 2006 and 2017, we built spatial regression models with a quadratic term of forest cover when testing the effect of forest cover on MFSI and its first principal component (PC-1) and without quadratic term for the second PC.In all models, spatial error was included to check for nonlinear relationships.We used the errorsalm function from the package "spdep" in the R environment.Moran I test, using the function moran.test was then used to test whether after accounting for spatial error, there was still spatial autocorrelation of the residuals.

| Multidimensional food security index
Our proposed Multidimensional Food Security Index (MFSI) was composed of 12 dimensions (principal components) that cover very different aspects of food security and explained up to 70% of the whole variance contained within the 38 variables for both years (2006 and 2017, Tables S1 and S2).Economic poverty was by far the most important dimension followed by socioeconomic inequality (see Tables 2, S1 and S2).More than half of the variance of the MFSI is attributed to the remaining 10 dimensions but with little contribution of each one, individually.There was a shift in the rank of importance of dimensions and small changes on the components, or variables, of each dimension between years.For example, the third most important (an availability variable) had a more important role in this dimension.
Environmental protection was placed at 12th dimension in 2006, increased its importance in 2017, and was ranked as the 4th most important dimension.Many other changes alike happened between years, suggesting that food security can be secondarily influenced by many different drivers across time (Table 2).
T A B L E 1 Variables used to build a food security index for the Caatinga municipalities and the respective measure and brief definition.(Figure 2a).Regarding food security, 561 municipalities registered a decrease in food security while the remaining 580 improved this measure from 2006 to 2017.We found a large variability in food security change, which goes from small changes to three orders of magnitude (>32,000%) of net gain or net loss (Figure 2b).Overall, we found that 29% of the municipalities lost forest and gained food security (Figure 3) between 2006 and 2017.These were followed by the municipalities that gained forest while losing food security (24%).
The worst scenario was registered for 25% of the municipalities that lost both forest and food security in the time period.However, the optimal combination of gaining both forest cover and food security was registered in almost 22% of the municipalities (Figure 3).
Also, the correlation between forest cover and food security was complex and spatially structured (all models performed better with spatial error) and, as expected, presented a boom-bust pattern described by a weak but significant quadratic function for both years (for 2006; z = À2.40,p = 0.016 and for 2017; z = À3.13,p = 0.001).
Briefly, food security tends to increase in mid-deforested and reaches its peak around 50% of deforestation when it comes to drop again to similar levels of food security found in highly forested municipalities.
A deeper analysis suggests that this pattern is mostly driven by economic poverty that presents a u-shaped curve in response to forest cover for both years (Figure 4) However, the second most important dimension (PC 2) of food security presented a direct relationship with forest cover, suggesting that inequality increases in more forested areas.As a result, low levels of food security can be grouped in two types: (1) poor people living in forested areas with social inequality and; (2) poor people living in deforested areas with smaller social inequality.In our study, the tradeoffs (gain-lose or lose-gain) group the poorest municipalities, but the ones losing forest and gaining food security are more unequal (in terms of income) while the ones gaining forest and losing food security are less unequal.

| DISCUSSION
The relationship between our multidimensional index of food security  Clearly, the access variables were the most important for our MFSI and are mostly related to socioeconomic conditions of the population.
Monetary poverty is a historical and widespread problem for the Caatinga population and as well as in other parts of the world, it is among the most important drivers of food insecurity (Fan & Brzeska, 2016;Nahid et al., 2021).On the other hand, the role of socioeconomic inequality is not clear but wealth concentration can also mean generalized low access to food (Klassen & Murphy, 2020).The interplay of the other dimensions and their variation in the degree of importance between years precludes the search for general patterns and lead to the notion that a great number of factors can shift in importance to determine food security over time (FAO, 2022).One important lesson is the prevalence of access variables within the dimension of the index we created.Stability of food production also had some influence and highlighted the importance of collective association of farmers and the long-term protection of water resources (Parraguez-Vergara et al., 2018).Surprisingly, neither utilization nor availability variables contributed significantly to our index, reinforcing the rationale that food security is less influenced by food production or capacity to process food (Barrett, 2010).Yet, our results concur that access is far more important than the other dimensions to determine a population's food security (Barrett, 2010;FAO, 2020) even in high-income countries, such as the U.S. (O'Hara & Toussaint, 2021).
The evolution of the index, however, reveals that despite the numerous national policies aimed at combating hunger and poverty, such as the Fome Zero (Zero Hunger) program (da Silva et al., 2013), nearly half of the assessed municipalities experienced a decline in food security from 2006 to 2017.This is congruent with recent findings showing that these programs failed to systematically reduce poverty and inequality in the region (Dyngeland et al., 2020), the two most relevant dimensions of food security in this context.Yet, the largest variations in food security were mainly positive, suggesting that some municipalities did benefit from these policies.Enhancing food security for people in the Caatinga and other dry forests is a complex problem and may require additional political solutions addressing other dimensions of food security yet to be considered by governments, such as the agency of people and communities and sustainability of food systems (Clapp et al., 2022).

| Food security and forest cover in caatinga
We found a nuanced scenario with both synergies (win-win and loselose) and tradeoffs (win-lose and lose-win) between food security and forest cover occurring in nearly similar proportions in the Caatinga.
The combination represented by the negative synergy (lose-lose) between forest cover and food security is thought to be the worst scenario where an increasing food insecurity can be worsened by loss of natural resources (Meyfroidt, 2018).Many of these municipalities, but also the lose-gain municipalities, are located in a northeastern-to southwestern axis which is experiencing an expansion of a commodity-driven economy (fruit plantation and soybean) that exerts a pressure of deforestation and concentrates wealth (Weinhold et al., 2013).Double positive scenarios where both forest cover and food security increases may represent development moments when poverty alleviation and politics for reducing social inequality are decoupled from deforestation and must probably rely on services and industry rather than agricultural expansion, combined with forest protection policies (Liu et al., 2017).Municipalities that lost forest and gained food security are probably experiencing the initial phases of commodity-driven development when a rapid increase in socioeconomic indicators derive from the establishment of new agricultural frontiers as already shown for the Amazon deforestation frontier (Rodrigues et al., 2009).The other side of this tradeoff where municipalities present net gains of forest cover but food security decreases may represent the "bust" phase of commodity-driven economy when land abandonment occurs due to the displacement of agricultural frontiers to cheaper lands leaving behind a poor and unequal society (Barbier & Hochard, 2018).Both types of tradeoffs resemble the initial and final phases of "boom-and-bust" development, respectively (Barbier, 2020).

| The role of forests for food security
Although our results do not address the role of forest to food security directly, there are some indirect connections that can be made based on the established literature.Forest goods and services are considered an important source of poverty alleviation as they provide vegetables, bushmeat, and work as rangelands for husbandry and provide nutrient cycling needed for shifting agriculture (Baudron et al., 2019).Our index of food security could not grasp such services because this kind of data are hardly available on the scale of our analyses.However, the fact that the highest values of food security were registered in intermediate levels of forest cover suggests that a combination of basic "grey" and enough "green" infrastructure offers more opportunities for poverty alleviation and increases access to natural goods and services that ultimately reduce food insecurity.Our results indicate that improving socioeconomic conditions (i.e., reducing poverty and inequality) can be more effective to increase food security than stimulating food production through agricultural land expansion.Positive associations between forest cover and several nutrition indicators have been reported elsewhere in the literature (Ickowitz et al., 2014;Johnson et al., 2013;Powell et al., 2011) whereas a significant inverted-U relationship between tree cover and fruit and vegetable consumption was registered in a study that compiled data from 21 African countries (Ickowitz et al., 2014).Forests can play an important role as a safety net for low-income people, particularly by providing family income complement (Miller & Hajjar, 2020).In rural Malawi, for example, forests play an important role for poor rural farmers while coping with food shortages during climate shocks, either by providing a direct source of food or through the selling of forest goods (Fisher et al., 2010).In the northern part of Caatinga, forest resources are used by 49% of the rural population to supplement their income, contributing by 10%-50% in most families and representing more than 50% of total income for some households (MMA, 2016).Also, the silvopastoral system in the region is known to be a traditional land use system, contributing largely to food security in the region (Pinheiro & Nair, 2018).It is crucial that further studies assess the degree to which local and regional food systems depend on forest goods and services, both within and without the food market.

| Future direction
Fifteen percent of the Caatinga municipalities (n = 210) had forest cover lower than 20%.Forest restoration is a legal obligation in rural properties in Caatinga when native vegetation covers less than 20% of the property area, or when the buffer areas of forest around water resources (i.e., Areas of Permanent Preservation-APP) are degraded (Brasil, 2012).The legal obligation can be seen as an opportunity to promote co-benefits between restoration and food security by the generation of jobs and income related to forest restoration (Mesquita et al., 2010).Forest restoration should directly support food security since it can promote stability for food production through water and soil protection (Soares-Filho et al., 2014) or improve the availability of food resources through agroforestry systems (Chamberlain et al., 2020).Variables related to protection of forests appear as one of the dimensions of our index (PC 10 in 2006 and PC 3 and 4 in 2017 -Table 2 and Tables S2 and S3) while variables related to the lack of protection contributed negatively to food security (PC 12 in 2006 andPC 9 in 2017).Thus, forest restoration in municipalities can help to increase food security mainly if productive restoration such as agroforests are implemented (Yang et al., 2020).Then, it is possible to restore many deforested areas without competing with food production and also creating opportunities to improve access and stability of food systems.The priority of restoration should be given to lowincome farmers and to municipalities with high levels of poverty and deforestation, since these are the places where it should have the greatest positive impact in increasing food security and also as a matter of environmental justice (Cousins, 2021;Reij & Garrity, 2016).
On the other hand, promoting restoration in municipalities with high forest cover could lead to more tradeoffs with food security than co-benefits.In this situation, policies aiming to increase food security should attack social inequality (Misselhorn et al., 2012) and take advantage of large tracts of native vegetation to maintain it under legal protection, whereas protected areas can help to improve people's well-being and food security (Naidoo et al., 2019).

| CONCLUSION
We highlighted that native vegetation should be taken into account when thinking about food security and sustainable food systems.Forests play an important role in maintaining the stability and productivity of the local food system (Chamberlain et al., 2020;Melo et al., 2020), but are not usually addressed in food security studies (FAO, 2013;Gubert et al., 2017;Ozturk, 2015;but see Vysochyna et al., 2020).Our approach shows that there are no intrinsic, unavoidable tradeoffs between forest cover increase and food security.In fact, as poverty and inequality were the main source of food insecurity in Caatinga, well-designed ecological restoration programs can help to alleviate poverty by creating jobs and promoting more income to rural families.Although restoration is an important ally to reduce poverty, this is not a solution for food insecurity in Caatinga and other drylands.There is still an urgent need for social policies that directly aim to reduce poverty in all its dimensions (UNDP & OPHI, 2020) which will greatly improve food security.Those policies should promote ways out of poverty that do not compromise the already overpressured natural systems (see Chamberlain et al., 2020;Cousins, 2021 e.g.), they should be focused in the most socially vulnerable and environmentally degraded municipalities, such the ones with low forest cover and low food security or the ones that had negative synergies (lose-lose) between forest and food.Otherwise, those policies might not reduce poverty and food insecurity where it is most needed.
dimension in 2006 was made by variables linked to the financial stability in food production alone; however, in 2017 gender inequality F I G U R E 1 Map of Brazil identifying the Caatinga biome from where the data of 1141 municipalities were used to understand the relationship between forest and food security.[Colour figure can be viewed at wileyonlinelibrary.com]

F
I G U R E 3 Geographical distribution of municipalities that experienced combinations of either gains or losses of forest cover and food security, between the years of 2006 and 2017.In the legend levels, the first word always refers to forest cover and the second word, after hyphen, refers to food security (either gain or lose).[Colour figure can be viewed at wileyonlinelibrary.com]F I G U R E 2 Map of the 1114 municipalities of the Caatinga region showing variation in forest cover (a) and food security index (b) between the years of 2006 and 2017.The food security index was built through a Principal Component Analysis using several variables related to access, availability, stability and utilization of food.The value for each municipality was calculated as the sum of the individual scores of each PC (with eigenvalue greater than 1) for each municipality, and then we calculated the difference from 2006 to 2017.Therefore, positive values of change means an increase in food security and negative values, a decrease from one year to another.[Colour figure can be viewed at wileyonlinelibrary.com] 4.1 | Multidimensional food security index (MFSI) Relationships between the percentage of forest cover per municipality and the values of food security index (a, b) and its main components (c-f) for the years of 2006 (left column) and 2017 (right column).[Colour figure can be viewed at wileyonlinelibrary.com] Principal components of the food security index for 2006 and 2017 and the percentage of explained variance.Numbers refer to the order of importance of each principal component and the names are the author's interpretation of the subset of variables loaded for each principal component.
social evensess.The second one is a "grey food insecurity" formed by poor municipalities with low forest cover and more social inequality.In between these extremes, there is a zone of relatively high food security and intermediate forest cover with less poor people and intermediate levels of social inequality.T A B L E 2