Ecosystem approaches and health in Latin America

Important environmental changes that have become increasingly pronounced in the last two centuries and that are seriously affecting human health require the development of integrated and participatory scientific approaches that can result in proposals for institutional and public policy changes. The purpose of this article is to offer some elements that can contribute to a line of reflection based on studies with ecosystem approaches in the Latin America context. The authors begin with a brief description of current scientific literature in public health that links ecosystems and human health in Latin America; next, they describe and compare the two prevailing trends that form the basis for the theoretical and methodological debates on ecosystem approaches; they also review the empirical research in Latin America or concerning Latin American countries in which an ecosystem approach has been adopted. The results point to limited scientific output on the interface between ecosystems and human health; aspects involving public participation and implementation of institutional changes and public policies are still in a rather incipient stage. Ecosystem; Environmental Health; Public Health Introduction The scale, magnitude, and uncertainties permeating the current environmental crisis reveal how human activities have produced drastic environmental changes at the local and global levels, resulting in numerous serious health problems. Due to their very complexity, these problems require a search for alternative approaches that combine socioeconomic and biophysical aspects for better understanding and solutions. The need to develop these alternative approaches has intensified at the transition from the 20th to the 21st century. The report of the World Resources Institute (WRI) 1 for 2000-2001 illustrates this process, pointing to the need to adopt an ecosystem approach premised on the capacity to contribute to: (1) the combination of diverse types of information that allows a careful weighing of the trade offs among various ecosystem goods and services and among environmental, political, social, and economic goals; (2) developing wiser policies and more effective institutions to improve ecosystem management; and (3) public participation (particularly that of local communities) in ecosystem management. Another major initiative is the Millennium Ecosystem Assessment program 2, conducted by the United Nations, launched in 2001 and with results published in 2005. The program was designed to respond to needs for information on ARTIGO ARTICLE Freitas CM et al. 284 Cad. Saúde Pública, Rio de Janeiro, 23(2):283-296, fev, 2007 how ecosystem changes can affect human wellbeing, with answers that can be obtained at different levels (local, national, and global) to improve ecosystems management and thus help improve human well-being. Ecosystems changes in these systems can have present and future consequences, resulting in socio-ecological problems that must be included on the global scientific and political agendas. The approaches developed to understand and solve them should allow dealing with the dual relationship between scientific knowledge and process decision-making. The current article aims to contribute to the development of these approaches in the Latin American context. We begin by briefly describing the current scientific literature in the public health field concerning ecosystems and human health in Latin America. Next we describe and compare the two approaches at the base of the theoretical and methodological debate on ecosystem and health. Finally, we analyze the empirical studies with ecosystem approaches by Latin American researchers or on Latin American countries based on the two approaches: (1) situating their studies in relation to the two approaches and (2) identifying the presence and limits of applying the three premises identified in the WRI report 1. Ecosystems and ecosystem approaches in Latin American public health journals In the early 21st century in Latin America, and particularly in Brazil, discussions and studies have emerged in the public health field based on these approaches. In 2001, the journal Cadernos de Saúde Pública/Reports in Public Health published one single article 3 and a special issue devoted to the theme 4. These articles as a whole dealt with such issues as application of these approaches for an understanding of tropical diseases; integrated management of diseases and natural resources; community mobilization and participation; and surveillance and monitoring strategies for environmental and health problems. In 2002, other types of publications (theoretical and conceptual) like a book chapter on health and sustainable environment 5 and a document 6 published by the Regional Office for Latin America and the Caribbean (in Uruguay) of the International Development Research Centre (Canada) have emerged as important contributions. A survey of Latin American public health journals available on SciELO (http://wwww.scielosp.org) up to 2004, using “ecossistema” and “ecosystem” as descriptors, identified 23 articles. Of these, four adopt a vector ecology approach 7,8,9,10; two study the interface between vector ecology and the ecosystem as a modified place 11,12; eight deal with the ecosystem as a place that has been modified and become disease-prone 13,14,15, in which viruses are present 16,17, where diseases occur 18,19, and as the place in which intervention should occur 20; and nine adopt ecosystem approaches which in some way include the premises laid out in the WRI report 1. Only one of the nine studies was published by a Brazilian researcher 3. The others are distributed geographically as follows: four by Latin America researchers, namely from Peru 21, Colombia 22,23, Paraguay 24, and Argentina 25; two by Canadian researchers 26,27; one by a Swedish researcher 28; and one involving cooperation between researchers from Kenya, Sweden, and Italy 29. Only recently has the Latin American public health literature incorporated the words “ecossistema” or “ecosystem” (the oldest articles only date back to 2000). This recent incorporation has meant that in most of the articles, the ecosystem is not treated from a systemic approach, but as the place for the vector (vector ecology), virus, disease, or intervention. The two ecosystem approaches We identified two current approaches at the base of studies. The first focuses on the identification and measurement of signs and symptoms of changes in ecosystems and their current or future potential to affect human health, so as to back decision-making and management with scientific information 30,31,32. The second values the development of contextualized and participatory approaches for understanding and searching for strategies to manage ecosystem changes in given places (villages, hamlets, and small towns, for example) and their impacts on the health of local communities 33,34,35. Ecosystem health approach The ecosystem health approach aims to be a science that integrates the natural sciences (biophysical dimension), social sciences (socioeconomic dimension), and health sciences (human health sciences). Using the metaphor of the ecosystem as a patient, it proposes to: (1) diagnose ecosystem dysfunctions by monitoring signs and indicators, distinguishing between “healthy” (desirable) ecosystems and “unhealthy” (undesirable) ones; (2) offer ecosystem management ECOSYSTEM APPROACHES AND HEALTH 285 Cad. Saúde Pública, Rio de Janeiro, 23(2):283-296, fev, 2007 options that reduce the costs of post-damage interventions, losses of economic opportunities, risks to human health, and social disruption due to environmental degradation 36,37. As an integrating science, the ecosystem health approach seeks to transcend the limits of the dominant approaches, both economic (centered on market prices that reflect the current scarcity of natural resources and failing to consider the consequences for future generations), ecological (a tendency to set society and economic activities aside, treating them as “external” forces), and engineering (the search for targeted solutions based on command and control strategies) 36,37. The approach considers four dimensions and eight attributes. The first dimension is biophysical, assessing the ecosystems’ structures and functions (nutrient cycles, energy flows, and diversity of species and habitats, among others). The socioeconomic dimension emphasizes differences in the ecosystems’ productive capacity and the valorization of ecosystems’ services by populations and their repercussions on economic policies. The human health dimension seeks to establish a causal nexus between the imbalance in the ecosystems’ health status and diseases and human health risk. The spatial-temporal dimension considers the different responses to the multiple forms of environmental stress that produce complex changes with a cumulative and/or synergistic effect that can threaten the very viability of the ecosystems at the local and/or global level 36,37. As for attributes, the approach suggests eight criteria or indicators for ecosystem health that are applicable to the integration of the abovementioned dimensions. The first three criteria (vigor, resilience, and organization) are considered prime components of ecosystem health (structure and functions) and are characterized by their predominantly biological origin. The other criteria represent the management, planning, and sustainability capacity of the mitigation and compensation measures adopted by society for situations involving environmental harm 36,37. Table 1 describes these eight attributes. Ecosystem approach to health The point of departure for the ecosystem approach to health is that disease and health manifestations occur in complex socio-ecological contexts, characterizing ecosystems as self-organizing holarchic open systems (SOHO). The objective is to identify connections between human health an


Introduction
The scale, magnitude, and uncertainties permeating the current environmental crisis reveal how human activities have produced drastic environmental changes at the local and global levels, resulting in numerous serious health problems.Due to their very complexity, these problems require a search for alternative approaches that combine socioeconomic and biophysical aspects for better understanding and solutions.
The need to develop these alternative approaches has intensified at the transition from the 20 th to the 21 st century.The report of the World Resources Institute (WRI) 1 for 2000-2001 illustrates this process, pointing to the need to adopt an ecosystem approach premised on the capacity to contribute to: (1) the combination of diverse types of information that allows a careful weighing of the trade offs among various ecosystem goods and services and among environmental, political, social, and economic goals; (2) developing wiser policies and more effective institutions to improve ecosystem management; and (3) public participation (particularly that of local communities) in ecosystem management.
Another major initiative is the Millennium Ecosystem Assessment program 2 , conducted by the United Nations, launched in 2001 and with results published in 2005.The program was designed to respond to needs for information on ARTIGO ARTICLE Cad.Saúde Pública, Rio de Janeiro, 23(2):283-296, fev, 2007 how ecosystem changes can affect human wellbeing, with answers that can be obtained at different levels (local, national, and global) to improve ecosystems management and thus help improve human well-being.
Ecosystems changes in these systems can have present and future consequences, resulting in socio-ecological problems that must be included on the global scientific and political agendas.The approaches developed to understand and solve them should allow dealing with the dual relationship between scientific knowledge and process decision-making.
The current article aims to contribute to the development of these approaches in the Latin American context.We begin by briefly describing the current scientific literature in the public health field concerning ecosystems and human health in Latin America.Next we describe and compare the two approaches at the base of the theoretical and methodological debate on ecosystem and health.Finally, we analyze the empirical studies with ecosystem approaches by Latin American researchers or on Latin American countries based on the two approaches: (1)  situating their studies in relation to the two approaches and (2) identifying the presence and limits of applying the three premises identified in the WRI report 1 .

Ecosystems and ecosystem approaches in Latin American public health journals
In the early 21 st century in Latin America, and particularly in Brazil, discussions and studies have emerged in the public health field based on these approaches.In 2001, the journal Cadernos de Saúde Pública/Reports in Public Health published one single article 3 and a special issue devoted to the theme 4 .These articles as a whole dealt with such issues as application of these approaches for an understanding of tropical diseases; integrated management of diseases and natural resources; community mobilization and participation; and surveillance and monitoring strategies for environmental and health problems.In 2002, other types of publications (theoretical and conceptual) like a book chapter on health and sustainable environment 5 and a document 6 published by the Regional Office for Latin America and the Caribbean (in Uruguay) of the International Development Research Centre (Canada) have emerged as important contributions.
A survey of Latin American public health journals available on SciELO (http://wwww.scielosp.org)up to 2004, using "ecossistema" and "ecosystem" as descriptors, identified 23 articles.Of these, four adopt a vector ecology approach 7,8,9,10 ; two study the interface between vector ecology and the ecosystem as a modified place 11,12 ; eight deal with the ecosystem as a place that has been modified and become disease-prone 13,14,15 , in which viruses are present 16,17 , where diseases occur 18,19 , and as the place in which intervention should occur 20 ; and nine adopt ecosystem approaches which in some way include the premises laid out in the WRI report 1 .Only one of the nine studies was published by a Brazilian researcher 3 .The others are distributed geographically as follows: four by Latin America researchers, namely from Peru 21 , Colombia 22,23 , Paraguay 24 , and Argentina 25 ; two by Canadian researchers 26,27 ; one by a Swedish researcher 28 ; and one involving cooperation between researchers from Kenya, Sweden, and Italy 29 .
Only recently has the Latin American public health literature incorporated the words "ecossistema" or "ecosystem" (the oldest articles only date back to 2000).This recent incorporation has meant that in most of the articles, the ecosystem is not treated from a systemic approach, but as the place for the vector (vector ecology), virus, disease, or intervention.

The two ecosystem approaches
We identified two current approaches at the base of studies.The first focuses on the identification and measurement of signs and symptoms of changes in ecosystems and their current or future potential to affect human health, so as to back decision-making and management with scientific information 30,31,32 .The second values the development of contextualized and participatory approaches for understanding and searching for strategies to manage ecosystem changes in given places (villages, hamlets, and small towns, for example) and their impacts on the health of local communities 33,34,35 .

Ecosystem health approach
The ecosystem health approach aims to be a science that integrates the natural sciences (biophysical dimension), social sciences (socioeconomic dimension), and health sciences (human health sciences).Using the metaphor of the ecosystem as a patient, it proposes to: (1) diagnose ecosystem dysfunctions by monitoring signs and indicators, distinguishing between "healthy" (desirable) ecosystems and "unhealthy" (undesirable) ones; (2) offer ecosystem management options that reduce the costs of post-damage interventions, losses of economic opportunities, risks to human health, and social disruption due to environmental degradation 36,37 .
As an integrating science, the ecosystem health approach seeks to transcend the limits of the dominant approaches, both economic (centered on market prices that reflect the current scarcity of natural resources and failing to consider the consequences for future generations), ecological (a tendency to set society and economic activities aside, treating them as "external" forces), and engineering (the search for targeted solutions based on command and control strategies) 36,37 .
The approach considers four dimensions and eight attributes.The first dimension is biophysical, assessing the ecosystems' structures and functions (nutrient cycles, energy flows, and diversity of species and habitats, among others).The socioeconomic dimension emphasizes differences in the ecosystems' productive capacity and the valorization of ecosystems' services by populations and their repercussions on economic policies.The human health dimension seeks to establish a causal nexus between the imbalance in the ecosystems' health status and diseases and human health risk.The spatial-temporal dimension considers the different responses to the multiple forms of environmental stress that produce complex changes with a cumulative and/or synergistic effect that can threaten the very viability of the ecosystems at the local and/or global level 36,37 .
As for attributes, the approach suggests eight criteria or indicators for ecosystem health that are applicable to the integration of the abovementioned dimensions.The first three criteria (vigor, resilience, and organization) are considered prime components of ecosystem health (structure and functions) and are characterized by their predominantly biological origin.The other criteria represent the management, planning, and sustainability capacity of the mitigation and compensation measures adopted by society for situations involving environmental harm 36,37 .Table 1 describes these eight attributes.

Ecosystem approach to health
The point of departure for the ecosystem approach to health is that disease and health manifestations occur in complex socio-ecological contexts, characterizing ecosystems as self-organizing holarchic open systems (SOHO).The objective is to identify connections between human health and activities or events that disturb the ecosystem's status and function 27,33,34 .
Under the theoretical framework developed by the ecosystem approach to health, complexity and uncertainty are inherent to socio-ecological systems, which involve a set of hierarchical groupings on multiple scales (spatial and temporal) that tend to organize in social and ecological feedback cycles, which in turn can contribute complexly to establishing critical instability points that sometimes result in the emergence of new structures and organizational forms.These alterations, characterized by different levels of uncertainties, can result in abrupt changes in the socio-ecological systems, ranging from minor alterations to environmental tragedies involving the emergence of pests or epidemics 27,33,34 .
The methodology is characterized by: (1) targeting local and regional ecosystem and health problems and (2) pluralism as the basic research strategy, incorporating multiple methods and forms of participation by local social actors.The approach is oriented by valorization of social collaborative learning processes involving specialists and local social actors, without separating problem understanding (research and analysis) from proposals for management and public policy strategies 33,34 .
The perspective is that pluralism, participation, and social and collaborative learning processes can lead to a type of adaptive management which emerges as an alternative, as well as complementing traditional anticipatory management.In adaptive management, the differences between how the future is anticipated and how it actually plays out are viewed as learning opportunities.The adaptive focus of the ecosystem approach to health presumes that decisions concerning environmental issues involve the mapping and construction of scenarios for how the territories or social and ecological systems should co-evolve as a self-organized entity.This path should also allow identifying the social stakeholders and interests, life histories, concerns, and future perspectives 34,35 .
Like the ecosystem health approach, the ecosystem approach to health also considers attributes or indicators that allow identifying whether an ecosystem is healthy or not, but its methodology focuses on the process and considers two fundamental aspects: (1) the frontiers of an ecosystem and/or environmental problem are established through negotiation between the various social actors and (2) the roles and responsibilities of the different social actors are defined at each step.These two aspects require that those adopting the approach set clear rules for negotiation, ways of involving different actors with opposing interests and resolving social at appropriate moments, and strategies to maintain the actors' par-ticipation throughout the process, reaching the formulation of management and public policy strategies 34,35 .

Ecosystem approaches and public health in Latin America
To analyze this item, we contend that the premises in the report People and Ecosystem: The Fraying Web of Life 1 are all necessary in order to scientifically recognize the "systems" in ecosystems holistically rather than sectorially, contributing directly to decision-making.These premises have also appeared in recent books published on the theme in Canada 34,35 , whose focus has influenced the research produced in Latin America 21,22,23,24,25 .Thus, it is no coincidence that the five selected articles have their research results published in the special issue of Cadernos de Saúde Pública/Reports in Public Health entitled An Eco-logical Approach to Human Health: Emerging and Communicable Diseases, which resulted from an event organized and financed by the Canadian International Development Research Centre (IDRC) at the Brazilian National School of Public Health in November 1999.After demarcating the field of analysis, we seek to situate these articles, which simultaneously reflect on a moment of transition and more systematic introduction of the ecosystem approach to health in Latin America, in relation to their proximity to the dominant watersheds, and to analyze the presence and limits of applying the three premises identified in the WRI report 1 .

Combination of diverse information
According to this first premise, an ecosystem approach necessarily involves a combination of diverse information that allows demonstrating the interfaces between goods and services from various ecosystems, which should be balanced with the environmental, political, social, and economic goals 1 .Of the five articles selected for a more detailed analysis, the study by Sosa-Estani et al. 25 on hantavirus pulmonary syndrome in Argentina was the only one with combined information on different spatial scales and diverse environmental/ecological variables for the problem at issue.The authors worked with a set of information that is relatively common to studies in the health field (serology; death records for infected individuals; identification of viral genotypes and phenotypes; identification of the rodent species involved).From the perspective of the ecosystem health approach, of the eight attributes that should be considered to assess an ecosystem's health, only the eighth (the effects on human health) was examined through the set of information that was collected.From the perspective of the ecosystem approach to health, the problem's analysis only involved a brief description of the socio-ecological changes in the three regions, but without involving the local communities in this description.
The other four articles 21,22,23,24 involve assembling pictures that allowed combining a larger set of information divided into two major groups, systematized in Tables 2, 3, 4, and 5.The first set dealt with spatial scales, collecting information at the individual/family/household/residential (Table 2), neighborhood/village/community (Table 3), landscape/municipal (Table 4), and regional/national/global levels (Table 5).At each level, the second set, shown in Tables 2, 3, 4, and 5, dealt with the different dimensions, combining information grouped at the environmental/ecological levels, related to the ecosystem, economic, social, cultural diseases dimensions, and those related to the impacts of the interventions in the ecosystem context.
Both the ecosystem health approach and the ecosystem approach to health deal with the scales analyzed in the articles.However, the ecosystem approach to health prioritizes the local scales (individual/family/household/residential and neighborhood/village/community), although it refers to the others.In the selected studies, there are clear influences by the ecosystem approach to health, due partially to the influence of the International Development Research Centre through funding and events.However, as observed in the Millennium Ecosystem Assessment 2 , a complete evaluation of the interaction between humans and ecosystems requires a multi-scale approach so as to allow the analysis of exogenous forces on a location or region to permit assessing the differential impact of changes in ecosystems on human well-being and health and point to different and combined answers on the different scales.Thus, the fact that the target studies focused on the local scale still proved to limit the search for solutions to the problems.
In these four articles, of the eight attributes considered in the ecosystem health approach, only the effects on human health were effectively considered, and the point of departure for most of the studies was a specific disease (Chagas disease, malaria, leishmaniasis).Although referenced on the ecosystem health approach, the attributes considered primary components of ecosystem health (vigor, resilience, and organization) and others such as management options, subsidies, and damage to neighboring systems were not discussed in any of these articles.The only attribute considered (and even then only in limited fashion) was the maintenance of ecosystem services, which has emerged as a key criterion for evaluating ecosystem health 2 .
As an attribute, ecosystems services refer to the functions that benefit human communities and that programs like the Millennium Ecosystem Assessment have viewed as key element for considering aspects pertaining to human health and well-being.According to the Millennium Ecosystem Assessment, ecosystems services include: (1) provisioning (food, fresh water, fuel, fibers, biochemical compounds, genetic resources); ( 2) regulation (clima regulation, disease regulation, flood regulation, detoxification); (3) supporting (services necessaries for production of other ecosystem services as soil formation, nutrient cycling primary production); (4) cultural services (non-material benefits obtained from ecosystems, such as leisure and tourism, spiritual, religious, aesthetic, and educational values, cultural heritage, and feeling of belonging to a place).
In relation to provisioning services and the environmental/ecological variables, the only reference we found was in the article by Murray & Sánchez-Choy 21 , focused on the local scales (Tables 2 and 3).The information discussed by these authors was quite general and dealt with the presence of (and access to) natural resources for agriculture, fishing, hunting, gathering, and water consumption.This same article also includes some information on supporting services, focused on the ecosystem's productivity as measured by soil fertility.For regulation services, the articles by Carrasquilla 22 and Rojas 23 discuss the environmental/ecological dimensions (Tables 2 and 3) at the local levels, with general information on the alteration in the landscape due to the forms of its use and occupation as a process that disturbs the relationship between humans and the ecosystem, facilitating disease through the intensification of contact with vec-

Impacts of interventions related to the ecosystem context
Hygiene practices associated with environmental health 21 Access to quality water for human consumption 21 Vector-favorable household microclimatic conditions 24 Construction materials 23,24 Forms and places in landscape occupation (whether near the forest or periphery) 22,23 Location and type of building materials (closer to the forest and more isolated from the village) 23 Socioeconomic status and food expenses 21 Income by source 21 Productivity in the economic activity 21 Health education as an element for reducing care for malaria cases 22 Availability of financial resources limiting the use of health services and preventive measures 22 Fish tanks as sources of both family income and potential vector foci 22 Impact of morbidity on the disease influencing productivity 24 Subsistence economy 24 Land tenure, family farming, and family income 24 Access to foodstuffs with quality diet 21 Family-level decisions (mother) on use of preventive measures or health services 22 Women making decisions on the family's health and home improvements 22 Morbidity and deterioration of living conditions 24 Schooling 24 Social stigma of persons with diseases 24 Jobs for housing improvements 24 Community participation in control and surveillance strategies (including schools) 24 Knowledge, beliefs, and practices leading to actions favoring or preventing disease 22 Practices allowing greater exposure to vector in the household or peri-domicile 22 Interruption of work activities to provide care for sick individuals 22 Perception of the disease 24 Incorporation of rural culture in the urban environment 24 Use of housing 24 Home ownership 24 Sustainable conditions for vector and species domiciliation 24 Family growth, facilitating vector feeding 24 Domestic animals incorporated into the household, serving as feeding sources for synanthropic animals 24 Peri-domiciliary characteristics 24 Location of dwellings (outskirts of village or close to forest 23 Type of local materials used in homebuilding, increasing odds of exposure to vector 23 Elimination of vector's microhabitat from the household by eliminating cracks through home improvement 24 Vector elimination using chemical products 24 Replacement of building materials and housing improvements in order to decrease vector exposure 23 Resettlement of scattered families 23 tors.Also in relation to regulation services, climatic aspects (rainfall, humidity, temperature, global warming, El Niño, La Niña) appear in the articles by Carrasquilla 22 , Rojas 23 , and Rojasde-Arias 24 , and as environmental/ecological dimensions, but situated on the regional or global scale (Table 5).
The dimensions that the authors referred to as ecosystem-related focused on ecosystem regula-tion services and dealt with climatic aspects 23,24 , ranging from the residential level (as a function of the building materials and their geographic location near forests) to the regional and global levels (as a function of alteration in the landscape and global climate changes).At the residential level (Table 2), the attempt was to indicate how such factors favor the presence and increase in the vector population at the regional and glob-

Impacts of interventions related to the ecosystem context
Number of houses favoring or preventing the problem 22 Location of houses favoring or preventing contact with vectors 22 Homebuilding rate in recently occupied areas 22 Forms and distribution of landscape occupation 22 Vector-favorable organization of peri-domiciliary space 24 Community's degree of dependence on the ecosystem 24 Macroeconomic projects 24 Residential grouping 23 Level of community organization 23 Number of inhabitants 23 Disease prevalence 23 Geographic location 23 Time since village was settled 23 Level of grouping (nuclear versus scattered villages) points to better or worse community participation and lower disease prevalence 23 Access to foods given their prices 21 Increase in the number of fish tanks as sources of family income 22 Groups for production and distribution of mosquito nets to poorest families 22 Forms of organization for marketing the community's produce 24 Improvements in facilities (railroads) for marketing the community's produce 24 Social practices (games, conversation) at times of greatest mosquito attack 22 Access to education 21 Access to health services 21 Community organization 21 Community resources and installations 21 Community history 21 Prioritization of municipalities in control and education related to the disease 24 Formation of farmers' cooperatives 24 Community participation through groups to eliminate mosquito attack sites 22 Lack of community organizations 24 Lack of economic culture 24 Dispersion of dwellings forces residents to cross the forest more often for visits or participation in community activities 23 Domiciliary and peri-domiciliary vector species, woodpiles, synanthropic reservoirs 24 Dispersion of dwellings 24 Peri-domiciliary characteristics 24 Resettlement of scattered families and promotion of nuclear villages 23 Vector elimination through mass interventions involving housing improvements and use of chemical products 24 al levels (Table 5) and how changes in regional climate patterns affected agricultural practices and individual behavior as a function of global climate changes, altering rainfall and flood patterns, for example.The environmental/ecological and ecosystem-related dimensions comprise a set related to ecosystem changes.A second set of dimensions, discussed next, relates to economic, social, and cultural aspects.
Economic dimensions 22,23,24 can be classified in three major groups, related to: (1) the negative impact of morbidity and mortality resulting from disease (Tables 2 and 3); (2) the positive impact of intervention through reduction of morbidity and mortality, highlighting the need for the government to take responsibility for these expenditures (Tables 2 and 3); and (3) organization of the local economy (land tenure, productivity, yields, financing, marketing of crops, corruption, un- Ecosystem productivity (soil fertility) 21 Ecological diversity 21 Presence of natural resources 21 Use of natural resources (agriculture, fishing, hunting, gathering, etc.) 21 Economic activities 22 Forms of work and their relationship to the landscape 22 Municipal government assumes costs of continuing the program 22 Increase in unemployment rate 22 Decrease in town's economic capacity due to public debt, corruption, and low income level 22 Contribution by municipal government to control programs (staff, transportation, vehicles) 24 Lack of commitment to disease control by local authorities 22 Reorganization of health services with new responsibility for municipal government 22 Politics plays central role in the health sector 22 Lack of people's trust in politicians and lack of belief in social and economic leaders 22 Lack of inter-sector approach to social development and problem-solving activities 22 Lack of medical care for persons with the disease 24 Lack of prenatal follow-up of the disease in endemic areas 24 Whether community organization is dependent on external agents 22 Community participation in prevention and health promotion activities is dependent on health authorities, without considering the community's preferences and priorities 22 Large number of traditional healers, as the first choice of care before the hospital or health post 22 Lack of municipal funding for health; health system centralized 24 employment, access to and utilization of natural resources, migratory flows resulting from impoverishment in given regions, etc.) (Tables 4 e 5).The social dimensions 22,23,24 were a broad group and in some cases were close to the economic ones.They related to family and community social organization; disease-related social stigma; disease-related decision-making in the family by women; centralization by the Ministry of Health of decisions related to the local level; community organization to prevent disease; participation in decision-making and formation of cooperatives; access to education, health services, and housing; migratory movements; omission by the public sector; and loss of community trust in decision-makers, among others (Tables 2, 3, 4, and 5).
The cultural dimensions 22,24 related to a wide range of aspects: perception of disease; cultural changes and shocks resulting from migratory movements; knowledge, beliefs, and practices actions that increased or decreased the disease; uses of housing; dependence on external agents for community organization to combat the disease; and differences in cultural norms in different ethnic groups (Tables 2, 3, 4, and 5).
The dimensions related to the disease in the ecosystem context are mixed with all of the above, since they related both to ecosystem changes as well as economic, social, and cultural ones.They involve forms of social and cultural organization at the household, peri-domiciliary, and community levels, favoring the vector's presence and proliferation (individual and residential levels); characteristics resulting from ecological changes in the location and region, favoring the vector's presence and proliferation (from the local to the regional level); socioeconomic characteristics favoring poverty and combined in turn with ecological changes, contributing to increase the disease, more specifically in Tables 2 and 3.
The last group of dimensions was limited to the impacts of interventions related to the ecosystem context 23,24 and can be classified in two groups: (1) interventions aimed at reducing or eliminating conditions favoring the vector in the community, involving the use of chemical products along with programs to control the disease and plans to improve rural housing and in some cases to group the housing in defined areas (Tables 2 and 3) and ( 2 Climate (rainfall, temperature, humidity) 22,23 Degree of ecosystem conservation 24 Ecological balance/imbalance 24 Macroeconomic projects 24 Temporary crops 24 Makeshift buildings in grain areas 24 Vector population increase during hot seasons 24 Quality of health system 24 Seasonal variations 24 Predominance of subtropical climate 24 Environmental changes 24 Rainy and flood seasons 24 Economic interests in macro-projects 24 Pattern of vegetation 23 Presence of transmission foci 23 Study site's geographic location and climate 23 Ethnic and cultural characteristics of local population 23 Climate changes (El Nino, La Niña) 23 Global warming 23 Global economic crisis 23 Changes in the region's climate patterns, affecting agricultural practices and people's behavior 23 Increased poverty in the developing countries 23 Decreased financing of the health sector with reduction in disease control activities 23 Lumbering by multinational companies 22 Population expelled from certain areas, resulting in migration and new settlements 24 External contribution to program development 24 El Niño 22 Low crop prices 24 Flaws in alternative programs (starting with production) and prices, contributing to major migratory movements to urban areas 24 National Agricultural Revitalization Plan 24 Migration from rural areas 22 Changes in health services financing 22 Situation of municipalities (prevalence of infestation and population serology) 24 Health priorities 24 Disease control as a decision by the Ministry of Health 24 Improvements in the National Housing Plan 24 Origin of local population (African-descendent, indigenous, etc.) 22 Decentralization of vertical programs 22 Health sector reform 22 Differences in cultural norms between different ethnic groups 24 Peasant farming culture still exists in large areas of the country 24 Primary agricultural activity with low yields and poorly marketed products 24 Migration to new settlements with endemic expansion 24 Period of acute cases of the disease 24 High temperatures contributing to increased vector production in the year 24 Region is appropriate for coexistence of vector and wild animals (reservoirs), making the individuals at greatest risk have to work in or cross the forest for some reason 23 Changes in rainfall patterns affect vector behavior and disease transmission 23 The disease primarily affects people in the poorest countries 23 Poorest countries lack resources to invest in disease prevention 23 (continues)

Dimension
Scales of integration Regional/National/Global

Impacts of interventions related to the ecosystem context
National plan to improving rural housing 24 National program to control diseases 24 Development of regional development poles 24 Environmental management 24 Improved quality of roads 24 Prioritization of programs for access to water for human consumption, gas, and electricity 24 Generation of work alternatives to avoid people's exposure 23 Stop global warming 23 Redefine government health care policies 23 Recruit more aid from the international community 23 by creating regional development poles (including access to services like fresh water, electricity, and gas) and generation of work alternatives to avoid greater exposure to vectors (Table 5).At the global level, the interventions involved the redefinition of government health care policies, reduction of global warming, or increased aid by the international community (Table 5).
The articles, particularly those focused on the ecosystem approach to health, although limited in terms of their treatment of information pertaining to the attributes described in the ecosystem health approach (even those relating only to ecosystems services), proved capable of combining a set of diverse information and to a certain extent managed to demonstrate the interfaces between goods and services in the various ecosystems and their interfaces with economic, social, and cultural aspects.The ecosystem health approach helped contextualize these attributes.The main limiting factor was that although the articles cited other scales, they failed to work with them in a linked way and were focused on the local scales.

Formulation of institutional and policy changes
According to this second premise, an ecosystem approach necessarily involves the formulation of broad public policies and more effective institutions to implement them.The basic idea is that once the problem is known by collecting various types of information, this knowledge should be connected to the relevant action.This involves formulating a set of policies or measures -legal, economic, financial, institutional, and social interventions -that reduce or eliminate the direct and indirect impacts on the ecosystems and that directly and indirectly affect human health and well-being 1,2 .This basic idea is present in the two focuses discussed in this article, the ecosystem health approach and the ecosystem approach to health.
Of the five articles analyzed, those by Rojas 23 and Sosa-Estani et al. 25 focus on the diagnosis of environmental problems, without proposing institutional changes or even formulating public policies to prevent or control the problem.
Due to their richness in connecting the diagnosis to proposals for institutional changes and public policy-making and the proposals themselves, the other articles will be analyzed in greater detail, since they provide material and inspiration for other studies with ecosystem approaches.
The paper by Rojas-de-Arias 24 was the most complete in this sense, as part of a project aimed at analyzing three strategies for Chagas disease prevention (insecticides; housing improvement; and a program combining improved housing and insecticides) and using the results to foster better public policies to control the disease.The strategies were proposed on the following scales in the ecosystem: (1) residential -elimination of vectors using insecticides and the vector's microhabitat through housing improvements; (2) community -elimination of the vector in the community by massive use of insecticides in the peri-domicile and domicile and improved housing; (3) regional -national housing improvement plan and national Chagas disease control program; and (4) global -improved environmental management of farm produce access and marketing routes and electricity, water, and gas supply programs.The article considered strategies both at the various levels and in different dimensions (cultural, social, and economic).For example: (1) weaknesses at the municipal level involving lack of local health institutions and the limitations of a centralized health system, seen as a cultural dimension); (2) lack of medical care at the municipal level for Chagas patients and lack of prenatal care for pregnant women with Chagas disease in some areas, considered a social dimension; (3) the need for Chagas disease control to be a Ministry of Health decision and for housing improvements to be incorporated into the housing plans, considered a social dimension.
The paper by Carrasquilla 22 analyzed malaria control in the urban context, highlighting public policy-making for expanded control and prevention, considering the capacity of institutions to lead them.Primary health care was the underlying strategy adopted by the proposal, with actions focused at the local community level.The strategy discussed: (1) the precarious nature of local health services for providing diagnosis and treatment, proposing the reorganization of these services with new responsibilities for malaria diagnosis and treatment as a local government task; (2) the need for greater coordination of control activities between government agencies (at the national, state, and municipal levels), nongovernment organizations, and communities; (3) the need for an inter-sector approach to social development, crucial for reducing malaria exposure and incidence.
In the study by Murray & Sánchez-Choy 21 , the objectives and methods were designed to create a participatory research environment in order to ensure that the results would guide the community's action plans and the development of public agricultural and health policies.In the eight communities, community action plans were developed on the basis of the research results and local health indicators, identifying ecosystems management strategies allowing diversification of the local diet, increased food security, and relief of health problems.

Public participation in management
The WRI report 1 sees public participation in ecosystem management, especially by local communities, as an essential element.In the ecosystem health approach, public participation is considered important, but is not developed as an integral element of the methodology.The researcher is the key figure formulating hypotheses and analyzing the results.Meanwhile, in the ecosystem approach to health, public participation is considered an integral element.The researcher seeks to involve the various stakeholders in formulating hypotheses and analyzing the results, from the perspective of social and collaborative learning processes between experts and local social actors leading to adaptive management of problems.
The article by Sosa-Estani et al. 25 did not involve any direct participation by the public or local community in the research.In three other articles, local community participation took place during the development of targeted local prevention strategies focused on education.In Rojas-de-Arias 24 , participation took place at the family level, when at least one member of the family agreed to participate in the training for housing improvements needed to prevent Chagas disease.In this article, community participation occurred in the villages by organizing meetings to encourage local involvement by training volunteers for educational work and involving local health institutions in the project.In Carrasquilla 22 , community participation took place during the elaboration of educational materials and the production and distribution of mosquito nets to prevent malaria.At the end of the paper, the author recognizes that the research team was not successful in involving local actors, especially the local politicians and authorities, despite recognizing the importance of such participation.
The article by Murray & Sánchez-Choy 21 is the only one reporting on participation by community members as inherent to the methodology.In this study, community members participated actively from the data collection phase onward.This broad participation beginning in the initial stages, closer to the ecosystem approach to health, served as the basis for promoting a broad dialogue among community members and launched the establishment of an environment in which the community can organize and plan an action platform for future problem-solving.

Final remarks
As stated in the introduction, the principal objective of the current article was to contribute to the development of ecosystem approaches in the Latin American context.We have thus sought to raise some issues and reflections that contribute to this process.
It became evident in the first section that there is still a limited scientific output in the form of articles attempting to deal with the interface between ecosystems and human health, with even fewer attempting to integrate this interface through the development of ecosystem approaches to health.
Second, the overall term "ecosystem approaches" entails important differences.The ecosystem health approach is efficient in its communicational thrust by using the metaphor of ecosystem as a patient.It allows raising the awareness of a large segment of the population concerning the interrelations between health and the environment.However, this same metaphor tends to limit the understanding of health to the predominantly biomedical aspects.In ad-dition, although the ecosystem health approach is concerned with defining four dimensions, including the socioeconomic, it does not clearly demonstrate how to deal with them in an integrated way, thus maintaining the society/nature dichotomy.However, it does make strides in proposing attributes to be investigated in order to evaluate the health of ecosystems, by defining the eight fundamental criteria.These criteria have great potential to become ecosystem indicators for evaluations, diagnoses, and monitoring strategies.
However, in order to further develop these criteria as indicators, there are still barriers that need to be overcome in the Latin American countries both in terms of the amount and quality of available data, which end up limiting the potential to gather various types of information, as stated by the WRI 1 and verified in our analysis of the articles, demonstrated in Tables 2, 3, 4, and 5.In our assessment, the Latin American countries face four major obstacles for gathering various types of ecosystem information: (1) the limited availability of ecosystem data as compared to social or economic data; (2) data and measurements that are ideally specific for a given ecosystem display limitations for extrapolating to other scales (bio-regions, eco-districts) and do not fit easily into the political and administrative boundaries of municipalities or states that underlie the vast majority of available data and information 38 ; (3) institutional weakness (absent or precarious human, technical, and financial resources), resulting in absent or discontinuous ecosystem monitoring programs, with low quality of available data; and (4) difficulty in determining ecosystem health values or states as benchmarks, since this often involves subjective judgments about what should be considered "normal" or "acceptable".
Other relevant differences between the ecosystem health approach and the ecosystem approach to health include the formulation of political and institutional changes, as well as public participation.These two aspects are more present in the ecosystem approach to health, which has had more influence on the research in Latin America, as discussed above.
Interestingly, the articles analyzed in the third section of the current article, which proposed changes and formulated policies, were nearly all focused on the residential or community scale, reaching the municipal scale at most.Even articles that included the global scale 21 did not move forward on issues acknowledged since 1986 in the Brundtland Report 39 and that were present in the United Nations Conference on Environment and Development (Rio 92) 40 , like proposals on the need for changes in institutions operating at the regional or global scale (United Nations Environment Program, World Health Organization, International Bank for Reconstruction and Development, World Bank, International Monetary Fund, World Trade Organization, etc.) and whose decisions affect the local level, as well as interventions for reorienting global policy and the global economy (which are known to contribute to deepening poverty and generating global problems, like global warming and loss of biodiversity).
In relation to public participation, of the five articles analyzed in the third section, participation was quite limited in three 22,23,24 .The study by Murray & Sánchez-Choy 21 was the only one reporting on public participation as inherent to the research methodology.
Although public participation is crucial to implementing ecosystem approaches, it becomes quite complex to the extent that it involves great diversity and value conflicts.Much more than the intellectual competence of the social actors for the normative implementation of the technical options, it is the compatibility of the value system that allows the formulation and definition of the problems and the subsequent adherence to rules and procedures to be implemented in the responses to such problems.Still, since a community's value system depends heavily on the local social context, it is difficult to generalize and manage, especially in the contexts of poverty and social inequality characterizing Latin America, where survival issues often override those pertaining to improved environmental quality or even the integrity of ecosystems.
Thus far, we see a trend for ecosystem approaches to health to be more prevalent in Latin America, although most such approaches show limited development in their formulation management and public policy strategies, especially in terms of public participation.Progress in studies with this approach requires further development of these aspects, in addition to a better definition and formulation of the variables and a better link with the regional and global scales.In addition, the series of reports that are now becoming available through the United Nations in the global program Millennium Ecosystem Assessment 2 since 2005, including regional evaluations and studies on the relationship between ecosystems and health, should tend to strengthen perspectives in Latin America that identify more with the ecosystem health approach, since they heavily prioritize the evaluation of ecosystem attributes, especially changes in ecosystem services, to the detriment of social participation processes ranging from problem understanding to the search for solutions.

Resumo
As grandes mudanças ambientais que vêm se acentuando nos últimos dois séculos e que afetam a saúde humana exigem o desenvolvimento de abordagens científicas integradas, participativas e que resultem em proposições de mudanças institucionais e nas políticas públicas.O objetivo deste artigo é oferecer elementos para uma reflexão sobre uma linha destas abordagens integradas, os enfoques (eco)sistêmicos, na realidade latino-americana.Para alcançar este objetivo, realizamos breve descrição do quadro atual de produção científica no âmbito da Saúde Pública sobre ecossistemas e saúde humana na América Latina; descrevemos e comparamos as duas vertentes que se encontram na base do debate teórico e metodológico sobre os enfoques ecossistêmicos; analisamos os trabalhos empíricos que adotaram enfoques ecossistêmicos produzidos por pesquisadores da América Latina ou relacionados a países do continente.Os resultados apontam para uma pequena produção científica que considera a interface ecossistemas e saúde humana, sendo ainda bastante incipiente no que se refere aos aspectos referentes à participação e à formulação e implementação de mudanças institucionais e nas políticas públicas.Ecossistema; Saúde Ambiental; Saúde Pública Contributors C. M. Freitas planned and organized the literature review and systematization and analysis of the available material, besides heading all the stages in drafting the article.S. G. Oliveira and G. E. Schütz participated in the planning and systematization of the literature related to the ecosystem health approach and ecosystem approaches to health in Latin America.M. B. Freitas and M. P. G. Camponovo contributed to the systematization of the literature related to the ecosystem health approach and ecosystem approaches to health in Latin America.

Table 2
Aggregation: individual/family/household/residential -dimensions/scales and variables present in a study adopting ecosystem approaches.

Table 3
Aggregation: neighborhood/village/community -dimensions/scales and variables present in a study adopting ecosystem approaches.

Table 4
Aggregation: Municipality/landscape -dimensions/scales and variables present in studies adopting ecosystem approaches.

Table 5
Aggregation: regional/national/global -dimensions/scales and variables present in studies adopting ecosystem approaches.