(2016) Une Histoire des Efforts de Gestion des Terres au Niveau du Paysage en Haïti : Leçons Apprises d'Études de Cas Couvrant Huit Décennies

A History of Landscape-level Land Management Efforts in Haiti Lessons Learned from Case Studies Spanning Eight Decades September, 2016 Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized A History of Landscape-level Land Management Efforts in Haiti Lessons Learned from Case Studies Spanning Eight Decades September, 2016 Andrew Tarter, Katie Kennedy Freeman, Klas Sander Acknowledgements The report A History of Landscape-level Land Management Efforts in Haiti is a product of the Agriculture Global Practice of the World Bank. The initial draft of the report was prepared under an activity led by Pierre Olivier Colleye and was intended as background information for government consultations conducted in May, 2015. The authors would like to thank the participants of this workshop, including the Ministry of Agriculture (MARNDR), for their valuable inputs. The final report considers feedback from this workshop, as well as comments from multiple experts across various institutions. The authors would like to thank the peer Reviewers - Peter Jipp, Julian Lee, Ademola Braimoh, Gerhard Dieterle, and Holger Kray - for their review and feedback. In addition, the team would like to thank the Haiti Country Management Unit – Mary Barton-Dock, Pierre Xavier Bonneau and Michelle Keane – for their support and Practice Manager Laurent Msellati for his guidance. The team acknowledges and thanks the Program on Forests (PROFOR) for supporting the publication of this report. Overall there is immense ecological diversity in Haiti, which results from the peculiarities of a landscape where most of the land area is in mountain. Agriculture is affected by the extreme variations in rainfall, variable soil types, differences in temperature according to altitude, and a mix of rugged highlands, fertile river bottom-lands and coastal pains. When the effects of these ecological variations are mediated through the prevailing social structure, they have a bearing on development potential and an understanding of production relations in Haitian economy. -Anthropologist Glenn Smucker, 1982. A watershed is formed by all those lands that shed water into a stream; it can be a few acres that shed water into a ravine or creek; or it can be millions of acres that drain into rivers, such as the Nile and Mississippi. Watershed management is the management of the lands, animals and people of a watershed so as to achieve the maximum benefits to man, while conserving the soil, forests, range and the water itself. -Département de l’Agriculture, des Resources Naturelles et du Développement Rural (DARNDR), Gouvernement de la République d’Haïti, 1962. 5 Table of contents LIST OF PROJECT ACRONYMS 8 EXECUTIVE SUMMARY 9 BACKGROUND 13 Ecological Vulnerabilities 13 Human Vulnerabilities 13 BOX 1. Conceptual Definitions of ‘Landscape-level Land Management’ and ‘Watershed Management’ 15 Outline of the Report 15 SECTION I: THE HISTORICAL AND CONTEMPORARY ENVIRONMENT OF HAITI 16 Looking Back to Move Forward 17 The Historical Deforestation of Haiti 17 The Colonial Period 17 The Post-Independence Period 18 Deforestation in the 20th Century 19 The Current Condition of the Rural Environment of Haiti 21 FIGURE 1. The Current Extent of Tree and Shrub Cover in Haiti in 2010 25 FIGURE 2. The Current Extent of Tree and Shrub Cover on the Island of La Gonâve, Haiti: 1990 and 2010 25 Land Inheritance, Land Tenure and Land Security 26 Land Inheritance 26 Land Tenure 27 Land Security 27 Soil Conditions in Haiti 28 Soil and the Relationship to Ecological Zones of Haiti 28 Livestock in Rural Haiti 29 TABLE 1. Principal Watersheds of Haiti 30 Hydrology in Haiti 30 FIGURE 3. Principal and Secondary Rivers, Haiti 31 SECTION II: LANDSCAPE-LEVEL MANAGEMENT PROJECTS IN HAITI 33 The Logic of Watersheds as a Landscape of Intervention 34 The Geospatial Distribution of Rivers in Haiti 34 FIGURE 4. An Early Delineation of Haitian Watersheds, Aggregated into 30 Areas 35 A History of Landscape-level Land Management Efforts in Haiti 6 FIGURE 5. An Early Grouping of Watersheds Adapted by Arrondissements and Divided into Seven Principal Areas 36 FIGURE 6. Watershed Prioritization Map for Project Interventions, Haiti 37 FIGURE 7. Watersheds of Haiti, According to the Ministère de l’Environnement 38 FIGURE 8. Current MdE Map (Right) Ostensibly Based on 1972 OAS Grouping (Left) 39 Case Studies on Landscape-level Land Management Projects in Haiti 40 Case 1: The Haiti Pilot Project (HPP) 40 Case 2: Organisation de Développement de la Vallée de l’Artibonite (ODVA) 41 Case 3: Développement Rural Intégré Petit-Goâve-Petit Trou de Nippes (DRIPP) 42 TABLE 2. Case Studies of Landscape Projects in Haiti, 1948-2014 43 Case 4: L’Organisme de développement du Nord (ODN) 44 Case 5: The Agroforestry Outreach Project (AOP) 45 Case 6: Secretariat Technique á l’Aménagemen des Bassins Versants (STABV) 46 Case 7: Agroforestry II (AFII) 46 Case 8: Productive Land Use Systems (PLUS) 47 Case 9: The Targeted Watershed Management Project (TWMP) 48 Case 10: Forest and Parks Protection Technical Assistance Project (FPPTP) 48 Case 11: Agriculturally Sustainable Systems and Environmental Transformation (ASSET) 49 Case 12: Hillside Agricultural Program (HAP) 49 Case 13: Le Développement Economique pour un Environnement Durable/ The Durable Economic and Environmental Development (DEED) 50 Case 14: Watershed Initiatives for Natural Environmental Resources (WINNER) 50 Case 15: Appui à la Valorisation du potentiel Agricole du Nord, pour la Sécurité Économique et environnementale (AVANSE) / Feed the Future (FTF) 51 Summary of Section II 51 SECTION III: CHANGES TO MODELS OF LANDSCAPE-LEVEL LAND MANAGEMENT IN HAITI, LESSONS LEARNED, AND CONCLUSIONS 52 Changes to Models of Landscape-Level Land Management in Haiti 53 FIGURE 9. Differing Approaches to Watershed Management 54 ‘Micro-level’ or ‘Plot-based’ Models 55 ‘Macro-level’ or Entire Watershed Models 56 Haitian Government Models 56 Haitian Government Policies on Watershed Management 57 Lessons Learned 57 Issues of Appropriate Administrative Oversight and Perimeters of Project Delineation 57 Ecological (protective) versus Social (productive) Approaches 57 Participatory Approaches versus Command-and-Control Approaches 58 Macro-level Approaches versus Plot-based Approaches 58 Land Inheritance, Land Tenure, and Land Security 59 Conclusion 60 7 FIGURE 10. Strengths, Weaknesses, Opportunities, and Threats (SWOT) Gap Analysis 60 Strengths 61 Weaknesses 61 Opportunities 62 Threats 63 Recommendations 63 SECTION IV: SOURCES CONSULTED 65 SECTION V: APPENDICES 72 APPENDIX A: Challenges to this Literature Review 73 APPENDIX B: Landscape Level Management Schemes in Haiti 73 APPENDIX C: Watershed Management Principals of the Ministries of Environment and Agriculture 75 Ministry of Environment 75 Ministry of Agriculture 76 A History of Landscape-level Land Management Efforts in Haiti 8 List of Project Acronyms Agroforestry II Agroforestry Outreach Project Agriculturally Sustainable Systems and Environmental Transformation The Artibonite Project Appui à la Valorisation du potentiel Agricole du Nord, pour la Sécurité Économique et environnementale/ Feed the Future Le Développement Economique pour un Environnement Durable/The Durable Economic and Environmental Development Développement Rural Intégré Petit-Goâve-Petit Trou de Nippes Forest and Parks Protection Technical Assistance Project Hillside Agricultural Program The Haiti Pilot Project Macaya Biosphere Reserve L’Organisme de Développement du Nord Organisation de Développement de la Vallée de l’Artibonite Productive Land Use Systems Pwojè Sove Tè Secretariat Technique á l’Aménagemen des Bassins Versants Targeted Watershed Management Project Watershed Initiatives for Natural Environmental Resources AFII AOP ASSET ATP AVANSE/FTF DEED DRIPP FPPTP HAP HPP MBR ODN ODVA PLUS PST STABV TWMP WINNER 9 Executive Summary A variety of public and private institutions have initiated landscape-level land management 1 projects in Haiti since the first multilateral development project of the United Nations was implemented there in 1948. The rubrics guiding these projects have been diverse: sustainable rural development; integrated landscape management; agro-ecosystem management; agroforestry; ecosystem stewardship; and watershed management—to name just a few. Despite a vast nomenclature that reflects real differences in landscape-level land management approaches and policies, most such projects implemented in Haiti have moved along a temporal trajectory toward the following set of shared, complementary objectives: i) Natural resource conservation, management, and/or restoration; ii) Improving and/or increasing agricultural productivity; and iii) Improving the lives of project beneficiaries by increasing economic activities. The success or failure of landscape-level land management projects in Haiti has not been determined by these complementary objectives, but by the differential emphasis of overall project approaches and policies toward achieving these objectives. The present report opens with a brief history of ecological degradation in Haiti—a necessary exercise for two reasons: (1) understanding how ecological degradation has proceeded historically in Haiti is required for mitigating future degradation; and (2) as the rural landscape has been altered significantly since the earliest landscape-level projects were implemented, a grasp of current ecological conditions in Haiti is also necessary. 1 Here and throughout this report, landscape-level land management is conceptualized as an overarching category that refers to varied large-scale efforts at land management, including but not limited to ‘watershed management’. After providing a brief ecological history, the report then shifts to examine fifteen different landscape-level land management projects executed in Haiti between 1948 and 2014, detailed in the form of brief but summative case studies that highlight salient project features, and differences in approaches and policies. The primary findings presented below emerge from these fifteen case studies, the preceding ecological history, and a review of many documents presented in the bibliography of this report. The bibliography serves as a point of reference for further inquiries, guided by citations throughout this report. The primary findings from the review of literature and case studies suggest: • There is wide consensus on the need for landscape-level land management in Haiti, and increasing consensus on watersheds 2 as the appropriate location for effective project implementation. The majority of Haiti’s population is embedded in an agrarian landscape that relies on agricultural production as their primary livelihood strategy. Agricultural possibilities will continue to decline without effective landscape-level land management strategies. Centuries of deforestation and climatic drying have made access to water a major determiner of on- going agricultural production. Since rainfall is increasingly variable and unreliable in many of the denuded locations in Haiti, watersheds represent one of the most effective locations for landscape improvement efforts. • There have been two common approaches or models to landscape-level land management projects that have been enacted in Haiti. Both require trade-offs. The ‘macro-level’ approach favors complete 2 Watershed management indicates efforts at landscape- level land management that are delineated by varied conceptual and operational definitions of a watershed. A History of Landscape-level Land Management Efforts in Haiti 10 landscape coverage at the expense of social and ecological heterogeneity. The ‘plot-based’ approach favors highly adaptive, plot-specific policies at the expense of full landscape coverage. The former has tended to emphasize ecological outputs through command-and- control models, while the latter has tended to emphasize both social and ecological outputs, developed through beneficiary participation in all levels of project identification, formation, implementation, and ownership. • ‘Macro-level’ approaches to landscape-level land management are frequently marked by the following features: ——Macro-level approaches tend to target entire watersheds, but still often fail to achieve 100% coverage; ——Macro-level approaches tend to come with a priori assumptions and solutions developed by experts, which frequently rely on a complete restructuring of local agricultural or land management practices, rather than building on existing structures. This trend frequently results in less-than-enthusiastic participation or in resistance to participation from project beneficiaries; ——‘Participation,’ from the perspective of macro-model project implementers, is frequently viewed as beneficiaries’ initial adoption of technologies or practices, and the contribution of time or labor toward this initial adoption; ——One of major reasons for the failed efforts of many macro-level landscape management projects in Haiti has been the adoption of a laundry list of project objectives meant to address the complexities of the problem— an emphasis on breadth over depth, but an emphasis applied indiscriminatingly across all lands and inhabitants within a landscape. The attempt to implement wide-ranging objectives through a top-down manner frequently results in project funds and personnel stretched thin, with fewer project objectives ever fully realized. • ‘Plot-based’ approaches to landscape-level land management are frequently marked by the following features: ——Plot-based approaches tend to account for the inter- and intra-heterogeneity of land, society, markets, roads, etc., of landscapes; ——Plot-based models frequently rely on prior and ongoing research to assess and address these levels of heterogeneity; ——Plot-based approaches have tended to adapt project objectives in an iterative, farmer- driven process; ——Plot-based approaches necessarily involve local participation at all stages of project identification, implementation, ownership, management, and continuity; ——Plot-based approaches rarely achieve 100% project coverage within a given watershed; and ——Plot-based approaches have proved the most successful in terms of project adoption by beneficiaries and continuity after the end of the project cycle. • Case studies show a clear, diachronic trend away from ‘macro-level’ approaches, toward ‘plot-based’ approaches. The 15 case studies presented in this report, and other cases reported on in the literature review, indicate that ‘plot-based’ approaches tend not only to be more effective, but are increasingly embraced in Haiti over ‘macro- level’ models. This is especially true of arable land holdings, which are highly fragmented across most landscapes in Haiti. In certain cases, such as instances of large tracts of government-owned land, uninhabitable areas, or large plots of collectively-owned land held by networks of kin, the case for ‘macro-level’ approaches may still be made. However, the long-term success of landscape- level land management approaches with these former two land categories—characterized by a lack of local land ownership and thus low investment—has historically proved untenable. 11 • Despite a historical trend away from ‘macro- level’ models toward ‘plot-based’ models, both approaches have failed to achieve their objectives at different times for some of the following reasons: ——One historical determinant of landscape- level land management project success or failure has been the differential emphasis placed on varied project objectives. Project objectives selected for emphasis often appear to be driven by the a priori demands of donors or implementing agencies, often emphasizing a particular technology, approach, or model that ultimately proves incompatible in the Haiti context. The results have been low adoption rates or low continuity after project timeframes expire; ——Unsuccessful projects were frequently weighted down by multiple competing objectives that aimed to recognize the interrelated complexity of landscape-level land management, but simultaneously stretched project resources thin in an effort to address this complexity; ——Successful programs limited the scope of what might be accomplished within the confines of the project timeframe, and focused on a few clearly defined policies tied to even fewer and more-clearly defined objectives; ——Successful programs frequently took a substantial period of preparation to align the visions and goals of donors, varied project partners, stakeholders, and beneficiaries; ——Successful programs involved local participation at all levels, including project identification, preparation, and implementation; and ——Allowing processes of stakeholder alignment and multi-tiered collective participation permitted successful programs to identify meta-objectives and policies that addressed a multitude of concerns important to all project stakeholders, leading to higher levels of project adoption, participation, ownership, and achievement of objectives. • Land tenure does not appear to be an impediment to beneficiary participation. Haitians utilize both informal and formal systems of land tenure, which permit the inheritance, management, leasing, sale, and purchase of land. While these tenure systems are complex, and thusly can appear insecure, experience demonstrates that they are highly functional and commonly do not represent an impediment to the implementation of landscape-level land management projects in rural Haiti. As the leasing of land is a common, established practice in rural Haiti, access to land appears to be a larger determiner of investment incentives to farmers than informal or formal tenure status. • Land security is a potentially volatile issue that deserves reflection. Considering historical and contemporaneous land grabs and forced displacements, land insecurity from external forces represents a real concern, particularly with projects that target large parcels of arable land. However, the vast majority of arable plots in rural Haiti have been divided into fragments that are discontinuously dispersed across landscapes, serving as a de facto protective measure against external threats to land security. Larger plots of land are collectively owned by kin groups, but typically agriculturally unproductive, which may also serve to protect them against external forces. • Participatory models have demonstrated more success than other more command-and- control approaches. The historical command-and-control approaches to landscape-level land management, including watershed management, have been met with skepticism and suspicion on the part of local project beneficiaries in Haiti. Within these projects, levels of participation have been low, as have levels of adoption. Experience suggests that participatory approaches are crucial to long-term project success. Since local farmers occupy watersheds and are the direct project beneficiaries, their inclusion in all aspects of project identification, preparation, A History of Landscape-level Land Management Efforts in Haiti 12 and implementation is crucial to project success and sustainability. Successful project approaches included multiple stakeholders, such as members of local government, and local governing groups, community groups, church groups, and other solidarity structures. • Frequently effective government participation is limited by overlap between project and political administrative delineations and ecological zones. Many watershed management projects in Haiti wrestle with the fact that rivers and their tributaries do not adhere to either the political boundaries of Haiti’s departments or subsequent administrative units, or to the distribution of ecological zones in the country (Delatour et al. 1984). The differences between ecological zones have been greatly diminished (Ehrlich 1985), adding doubt to models that delineate watersheds and policies based exclusively on ecological life zones. • Project continuity As of yet, all documented examples of landscape-level land management schemes in Haiti have suffered from issues of funding and project sustainability after donor cycles have concluded. In Haiti, virtually all support has been through non-governmental donors and it is unclear in most cases if the internal rate of return of project investments make them viable without sustained donor support. Nearly seventy years after the first landscape- level land management project of the United Nations was executed in Haiti, similar projects continue to be implemented, and many continue to suffer avoidable historical shortcomings. The broader development community and project beneficiaries in Haiti have much to learn from the past in order to most effectively design and implement successful landscape-level land management projects in the future. This meta- analysis has been crafted for this very audience, as a point-of-departure for future efforts at landscape management in Haiti. The key findings listed above summarize a review of the literature and the fifteen case studies subsequently presented in this report. The reader interested in a more nuanced, in-depth understanding of how these individual cases unfolded in Haiti is encouraged to consult the extensive bibliography of sources at the end of this report. 13 Centuries of deforestation in Haiti have played out on an overwhelmingly mountainous terrain that receives seasonal rains and torrential downpours from frequent tropical storms and hurricanes. The standard of living for many rural Haitians has been negatively impacted as a result of deforestation and removal of the vegetative cover. Some of these impacts, such as the loss of valuable topsoil, have been directly observable, but other impacts of deforestation in Haiti have been less visable. Ecological Vulnerabilities From a strictly ecologically standpoint, deforestation in Haiti has caused or contributed to: • Damage to Haiti’s riparian systems and the soil-silting of lakes and unique coral reefs; The lack of a protective vegetative cover has led to an increase in the rapidity of surface water runoff out of watersheds, reducing the ability of rainwater to recharge the aquifers. Since Haiti’s rivers are aquifer-fed, river water levels have been lowered, and in many cases permanent rivers have become seasonal. Soil- silting of lakes is raising water levels, changing water salination levels, and destroying or threatening fresh-water, brackish water, and ocean fish habitats (Hotz and Christian 2015). • A decline and loss of Haiti’s diversity of endemic flora and fauna; Haiti is home to some of the highest levels of biological diversity in the Caribbean (Swartly and Touissant 2006). The disappearance of potentially rare, endemic flora and fauna represents a loss of unknown proportions. The erosion of specific soil profiles results in a decreased ability of native flora to regenerate, paving the way for exotic or invasive species that lower overall biological diversity due to an increasingly homogenized habitat for both flora and fauna. • Increased climatic drying; The decline of a protective vegetative layer subjects Haiti’s land to increased evapotranspiration—the drying of soil through direct exposure to sun and wind, which work together to wick away surface moisture. The heat rising from arid land drives away higher moisture air, in many cases reducing the likelihood of rainfall, and causing the pseudo- drought phenomena experienced throughout much of rural Haiti for the last several decades. The social science literature is replete with well-documented testimonies of Haitian farmers who contribute drought and pseudo- drought conditions directly to the over-cutting of trees. • Increased terrestrial vulnerability to the effects of tropical storms and hurricanes; Without complex root systems to hold Haiti’s mountainous terrain in place, entire mountainsides frequently wash out during tropical storms and hurricanes. Streams and rivers swell from immense groundwater runoff, causing flooding, increased erosion and widening of riparian systems, and the soil silting of lakes and ocean habitats. Individual trees and remaining tree stands are subject to higher exposure from wind than trees aggregated in forests, resulting in the widespread downing of remaining arboreal stands. The effects of tropical storms and hurricanes are not only exacerbated due to Haiti’s deteriorating ecosystems, they also contribute to that deterioration in an iterative, degenerative terrestrial cycle. Human Vulnerabilities Since most rural Haitians are embedded in an agrarian landscape and rely on agricultural production as their primary livelihood strategy, the ecological and landscape degradation noted above directly contributes to: Background A History of Landscape-level Land Management Efforts in Haiti 14 • Nationwide declines in agricultural productivity; The production of soil takes thousands of years but major losses of topsoil in Haiti may transpire in a single storm. As the majority of Haiti’s population is understood to be dependent on agricultural production as a primary livelihood strategy, and some 80% of rural households rely on agriculture (World Bank 2015: 27), the loss of valuable topsoil has an insidious and deleterious long-term effect on agricultural productivity. Haiti’s GDP per capita fell by an average of 0.7% between 1971 and 2013 (World Bank 2015: 1), weather-related damages and losses caused an average decline of 2% of GDP per year in Haiti from 1975 to 2012, and approximately 50% of tropical storm related losses in Haiti’s productive sector have fallen in the agricultural domain (World Bank 2015: 14). The silting of freshwater, brackish water, and reef habitats have caused a dramatic decrease in viable fish and aquatic resources, which has adversely affected the livelihoods of those rural residents dependent on fishing as their major livelihood strategy. • A resultant widespread rural out- migration; Decreased agricultural productivity in many areas of Haiti has increased rural out-migration, as few other economic opportunities currently exist in rural areas (World Bank 2015: 35). While many Haitians would rather stay in the countryside, a complex set of environmental, social, technical and economic factors incentivizes the trend to leave the countryside that started in the 1930s and accelerated in the early 1980s (ibid.). Rural Haitians have migrated to larger towns and cities, the capital city of Port-au-Prince, or have become members of the expansive Haitian diaspora abroad. Port-au-Prince, initially constructed to hold tens of thousands of people, had swelled to a greater metropolitan area of approximately three million people by the time of the 2010 earthquake. According to the US Census of the same year, approximately one million Haitians lived in the United States, and this figure represented only 43% of the entire Haitian diaspora abroad (Wah 2013: 59). • Increased exposure to environmental hazards; Regrettably, the crowded and marginalized urban spaces where many rural Haitian migrate frequently present more environmental health hazards than those found in the deteriorating ecological conditions of rural areas. The rapid spread of cholera and the chikungunya virus, and the high mortality levels of the 2010 earthquake, are testament to the fragile living conditions of many urban areas in Haiti. • Increases in hunger, malnutrition and under-nutrition. Those Haitians that abandon rural areas lose access to year-round and seasonal supplementary sources of nutrition from rivers, streams, woodlands, orchards, communal courtyards, and gardens, which may otherwise augment their diets. Instead, many urban- dwelling Haitians are forced to rely on purchased foodstuffs, frequently imported and of low nutritional value. Imports in Haiti have increased from approximately 30% of GDP in the early 1980s to some 50% of GDP in the early 2010s, with food imports representing approximately 35% of total imports over the last ten years (World Bank 2015: 17) The traditional rural breakfast of corn or sorghum porridge is increasingly replaced by imported spaghetti, topped with mayonnaise and ketchup. The traditional rice varieties of the Artibonite valley are increasingly replaced with lower-cost but less nutritious, imported white rice exported from the United States 3 . Citrus and other natural fruit juices that are available in rural areas of Haiti are increasingly replaced with high-sugar energy drinks, and rurally-produced bread is frequently replaced with high-sugar biscuits and crackers. Those Haitians that elect to remain in rural areas also suffer from increasing levels of hunger, malnutrition, and under-nutrition. The 3 Known in Haiti as ‘Miami rice,’ due to the common location of exportation. 15 deterioration of natural resources has decreased the availability of many traditional sources of dietary supplementation. One traditional strategy of mitigating crop failure—reducing the number of meals consumed a day—is fast becoming the new norm of rural Haiti, as rains come less-frequently and soil fertility declines. • Increased human vulnerabilities to tropical storms, hurricanes, and extreme climatic events. Despite large increases in rural-out migration, the majority of the Haitian population continues to reside in rural areas. Rural Haitian populations often encounter tropical storms and hurricanes that may cause landslides. Landslides also block important road systems, preventing access to hospitals, and to important transportation routes for farmers dependent on marketing their agricultural goods at regional and national markets. A World Bank report noted that Haiti has a higher number of disasters per km 2 than the average of Caribbean countries, and estimates that in 2008 alone, 15% of GDP was lost due to hurricanes and tropical storms (World Bank 2015: xi). Typically there is little governmental support for rehabilitation of damaged infrastructure, and local residents must mobilize with hand tools to clear the roads that are crucial to the economy of their internal market systems. An increased awareness and acknowledgement of these historical and contemporaneous trends in human and ecological vulnerabilities has given rise to an emphasis on landscape-level approaches to land management in Haiti, and increasingly a shift toward targeted watershed management. Outline of the Report This report reviews the available literature as an initial point-of-departure, providing a diachronic overview of sixty years of landscape management research and integrated landscape project activities in Haiti, spanning eight decades. The reporting period extends from the early efforts of the first integrated development program of the United Nations in Haiti in 1948, through the 2014 consummation of project activities associated with the USAID-funded Watershed Initiative for National Natural Environmental Resources (WINNER). Section I of this report opens with a brief environmental history of Haiti, followed by an analysis of the current state of environmental conditions, and closes with an examination of contemporary human and landscape vulnerabilities to acute and chronic environmental degradation and extreme climatic events. Section II of the report provides a brief summary of 15 regional or national landscape-level land management projects enacted in Haiti since the middle of the 20 th century. Section III concludes the report with a summary of important themes and applicable lessons that crosscut the history of landscape-level management projects in Haiti. Section IV provides a bibliography of important documents related to the issues examined in this report. This bibliography is intended to serve as a resource for future researchers, policy-makers, program administrators, project implementers, and project beneficiaries working on landscape- level land management projects in Haiti. BOX 1. Conceptual Definitions of ‘Landscape-level Land Management’ and ‘Watershed Management’ Here and throughout this report we define landscape-level land management approaches in Haiti as attempts to address the interrelated human and ecological vulnerabilities noted above, in the context of large land management projects executed at the landscape level, which focus on improving or introducing new livelihood strategies that ostensibly improve the lives of project beneficiaries. This report distinguishes watershed management as a variant of landscape-level land management approaches— increasingly embraced in the case of Haiti as the appropriate locus of project execution—and delineated by varied conceptual and operational definitions of what constitutes a watershed. A History of Landscape-level Land Management Efforts in Haiti 16 SECTION I: The Historical and Contemporary Environment of Haiti 17 The Historical Deforestation of Haiti The gradual removal of the original forests of Haiti is a complex process that spans five centuries and continues today. Popular accounts frequently rely on a gross simplification of this process, framing deforestation in Haiti as the result of an ecologically-disconnected peasantry, with little forethought to the consequences of their environmentally deleterious actions. To illustrate, deforestation in Haiti is frequently contributed to the production of charcoal, though the vast majority of Haiti’s original forests fell far before the charcoal trade roared to life in the early decades of the 20 th century. The brief history of deforestation provided here corrects some of these misconceptions, offering a pragmatically abbreviated chronology with particular focus on acute events and chronic trends that are likely to have a direct bearing on the success or failure of future landscape-level management plans for Haiti. The Colonial Period The ecology of the island of Hispaniola remained fairly unexploited for most of the 16 th century, as Spanish colonial interests were focused on the procurement of gold rather than the extraction of natural resources or agricultural production (Lindskog 1998). In the late 17 th century the environment of the island began to change rapidly with the renewed human presence of French emigrants and enslaved people from West and Central Africa. Anthropogenic influences on the environment of Haiti began to register a visibly notable change during the agricultural intensification activities promoted by 18 th century colonial powers (Moya Pons 2007). Cultivation of the initial colonial export crops—sugar, tobacco, indigo and cotton—saw the clearing of lowland plains (ibid.). The colonial plantation strategy of cultivating lowland agricultural production was pursued on several grounds: (1) lowland areas are more accessible by roads; (2) plains in Haiti are predominantly coastal and in closer proximity to maritime vessels; (3) uniform plains had higher levels of moisture or were otherwise irrigable; and (4) many of the lowlands of Haiti contain rich, alluvial soil deposits. The initial low-elevation locus of agricultural production in Haiti did spare the removal of trees and vegetative cover from surrounding mountain ranges. The booming colony required wood for a variety of purposes, but principally for construction and fuel-wood used in sugar refinement for the increasing number of plantations (Catanese 1999). Later, trees were cleared from higher elevations to cultivate coffee—a lucrative export crop (ibid.). Records from the early colony and the new Republic are replete with mention of the felling of large hardwood trees, and the encroachment of less desirable tree species (Tarter 2015b). LOOKING BACK TO MOVE FORWARD Proceeding from the logic that development initiatives seek to improve the lives of project beneficiaries, an understanding of the drivers of ecological degradation in Haiti is paramount and should be a first step for future research or landscape-level management project identification, preparation, and implementation. A History of Landscape-level Land Management Efforts in Haiti 18 In particular, the colonial period saw the commencement of a wood-extraction based economy. French ships carrying enslaved people from the African continent returned to Europe with hulls loaded with valuable timber (Diamond 2005). While the more easily accessible areas of Haiti suffered arboreal denudation during this period, many remote forest stands remained protected by their relative isolation due to the lack of accessible inroads. The Post-Independence Period Toward the end of the 13-year Haitian revolutionary period, lowland colonial plantations were burned and equipment destroyed. After the 1804 declaration of independence, the vast majority of Haitians fanned out throughout the mountainous areas of the country, establishing the traditional lakou (collective kinship habitations) of rural Haiti. Through the lakou system, Haitian farmers successfully resisted multiple attempts by early leaders of the fledgling Republic to reinstate the plantation model of agricultural production (Moral 1961). The historical dispersal of humans from lowland plains and their subsequent resettlement in mountainous areas resulted in an unusual occurrence: the reestablishment of the lowland forests of Haiti. Analyses of tree pollen from a sediment core extracted from Lake Miragoâne in southern Haiti noted a large arboreal expansion after 1804 (Brenner and Binford 1988). This regenerative arboreal expansion has direct bearing on the current state of Haiti’s ecology—a point that will be revisited subsequently, and which has policy programming implications for landscape- level land management in the current era. While the lowland areas of Haiti temporarily reestablished with trees after the destruction of the colonial plantation system and a decentralizing nationwide migration, it was only a temporary reversal of the predominant trend—an on-going removal of arboreal and vegetative covers. Deforestation during the post-colonial era was not caused by the production of charcoal, which did not commence on a large scale in Haiti until the early decades of the 20 th century. Two principal forces drove tree removal in newly- independent Haiti: (1) Haitian government payments on a war indemnity to France, financed by timber concessions to private companies; and (2) the clearing of land for agricultural production, enacted by hundreds of thousands of farm families spread throughout the rural countryside (Tarter 2015b). In 1825 Haitian president Jean-Pierre Boyer agreed that Haiti would pay France a post-war indemnity of 150  million francs—a paralyzing amount for the new Republic—as reparations to French slave and plantation owners for income lost after the world’s first successful slave revolt. The indemnity agreement, ostensibly established to ensure that France would not reinvade Haiti, is a historically curious concession, considering France’s multiple, failed attempts to retake the former colony. The Haitian government supported indemnity payments to France largely by selling forest concessions to foreign timber corporations. In 1838, the indemnity was renegotiated to nearly half of the original amount, with the requirement that the remaining amount be paid in full in 30 years (Moya Pons 2007; Bulmer-Thomas 2012). Four years later, legal mahogany exports doubled, and in 1842 exceeded 4.0 million cubic feet (Moya Pons 2007). Despite this surge in timber exportation, hardwood trees were still abundant in Haiti late into the 19th century. As one observer noted in 1878, ‘the variety of [hardwood] production seems to be almost infinite, and the supply inexhaustible’ (Stuart 1878: 267). After many of the easily accessible lowland areas of Haiti had been exploited for wood, the principal means of extracting timber shifted to dragging felled trees into the small mountain streambeds that form the many tributaries of Haiti’s multiple watersheds (Chandler 1842). With the arrival of annual torrential rainfall, these seasonally dry or low-flowing streambeds would swell, carrying the felled trees into increasingly larger rivers and eventually toward deltas where 19 they could be prepared for export by maritime vessels (ibid.). Despite this shift deeper into the interior of Haiti, the mountainous nature of the country and the difficulty of accessing remoter areas protected many remaining hardwood stands. Deforestation in the 20th Century The first three decades of the 20 th century saw the removal of most remaining hardwood forest stands in Haiti, the overexploitation of secondary arboreal growth, and the continued encroachment of former forestland by exotic tree species. The early decades of this century are marked by the US Marine occupation (1915-1934) and their establishment of a network of new roads and the improvement of earlier roads constructed by French colonialists (Leyburn 1941). The new network of improved roads in Haiti facilitated increased travel between rural and urban areas. While these new roads also encouraged urban migration, the increased urbanization during this period was not unique to Haiti—it was a phenomenon experienced globally. As Haiti increasingly urbanized—like many countries of the Caribbean—the demand for agricultural and wood products increased to meet urban needs. The changing Haitian landscape was the result of a complex interaction between immediate drivers and underlying causes of land-use change. Rapidly increasing population coupled with a decrease in agricultural productivity, spatial extensification of agricultural production and the reduction of previously fallow periods, urbanization and increasing demand for food and fuel, the expansion of road networks connecting the previously isolated hinterland to growing markets, and complementary though parallel land and tree tenure regimes all played their role in the evolution towards the highly dispersed and decentralized food and wood production system observed in Haiti today. As Haitians in areas of close proximity to urban centers intensified their agricultural production, charcoal production was shifted to the next geographically proximate location that was accessible by roads. As existing roads were limited to the national highways and lesser roads of poor condition, deforestation on a national level continued in a geospatially differential manner. Once previously forested or brush-covered areas gave away to agricultural production, charcoal production shifted again. From east of Port- au-Prince, charcoal production shifted offshore, to the island of La Gonâve; from La Gonâve, charcoal production moved to the northwest peninsula; from the northwest peninsula, charcoal production swung to the more remote southern peninsula, and to a lesser extent to the central plateau (Smucker 1981; Conway 1979; Voltaire 1979). The result of the extension of the agricultural frontier was the near-to-complete arboreal denudation of easily accessible areas, largely following in a sequential order based on geographical proximity to the ever-swelling capital city and smaller urban agglomerations. Today, the lack of access to arable land is an on- going frustration expressed by younger members of society. The latter half of the 20th century saw the establishment of hundreds of kilometers of new roads, which opened up the vast remainder of previously inaccessible areas the country. The community-based organization Harmonisation de l’Action des Communautés Haïtiennes Organisées (HACHO), operating in the northwest of Haiti, shifted to focus on road construction in 1968. By 1982 HACHO had constructed a network of over 600 kilometers of unpaved roads (Brinkerhoff et al. 1983). In the mid-1970s, USAID financed the Agricultural Feeder Roads Project, constructed over 300 kilometers of new roads by December 1982 (USAID 1983). Multiple road projects in Haiti also originated at the hands of local community councils and were spurred by the initiative of local churches or foreign missionary groups (Smucker et al. 1979). Considering just the USAID and HACHO projects, more than 900 kilometers of new roads were established in Haiti from around the 1970s through to the early 1980s. Nine hundred kilometers of new roads is a remarkable addition A History of Landscape-level Land Management Efforts in Haiti 20 to a country that does not exceed 300 kilometers at its lengthiest continuous extent (Tarter 2015b). Since new roads were constructed off of existing arterial highways, their construction opened up many of the more remote locations in Haiti. These new roads reversed the historical phenomenon of just a few accessible areas targeted for agriculture and charcoal production, and original charcoal producing areas experienced significantly less pressure than in former decades. Toward the end of the 20th century, almost all land area of Haiti had been dedicated to agricultural production (crops and livestock). Today, there is hardly any parcel of land that is not being dedicated to some kind of productive use, profoundly influencing the current state of the environment of Haiti (Tarter 2015b), with agricultural production commonly being the primary objective with sizable integration of trees and woody vegetation. 21 As early as 1830, the space created by the absence of original hardwood forests was rapidly filled by at least two exotic tree species: mesquite 4 (Prosopis juliflora), and to a lesser extent, logwood 5 (Haematoxylum campechianum) (Tarter 2015b). By 1930, the former tree species was the most common tree in the dry forests that dominate the majority of Haiti’s varied ecosystems, and the most utilized tree species for charcoal production (Gill 1931). By the 1940s, Prosopis juliflora was the most common tree in some 80 different forest-transect surveys conducted in an area north of Port-au-Prince (Curtis 1947). In 1976, the author of the first in a series of preliminary reports, which would eventually lead to the largest tree-planting project in the history of Haiti, concluded that managed Prosopis plantations could meet all of Haiti’s domestic wood needs (Earl 1976). Multiple authors of the widely-cited 1985 Haiti Environmental Profile Report noted the widespread extent of Prosopis trees and the long tradition of Haitian farmers managing the species for wood and for charcoal production (Ehrlich et al. 1985). By 1991, the World Bank reported that natural Prosopis stands 4 Bayawonn or bayahonn, in Haitian Creole. 5 Kanpèch, in Haitian Creole. dominated approximately 1/10 th of Haiti’s land surface (World Bank 1991). In a recent nation-level remote-sensing analysis of high-resolution satellite imagery from 2010 and 2011, trees were found to cover approximately 1/3 rd of the surface of Haiti, and trees and woody shrubs combined to cover 3/4ths of the surface of Haiti (Churches et al. 2014). A similar ratio of tree and shrub coverage was reported in an analogous study of the large offshore island of La Gonâve, based on high-resolution satellite photos from 1990 and 2010 (White et al. 2013). And finally, in a broader study of land changes within the Greater Antilles between 2001 and 2010, researchers discovered that 26 different Haitian municipalities underwent significant changes to woody vegetation (8 decreased and 18 increased) and 48 municipalities experienced significant changes in mixed-woody plantations (9 decreased and 39 increased) (Álvarez-Berríos et al. 2013: 88-91). In overall land percentages for the entire country, woody vegetation experienced a 1% increase while mixed-woody/plantations increased by 4%, or 368 km 2 (89). These studies combined point to an arboreal phenomenon that stands in stark contrast to popular conceptions of tree-cover in Haiti (Tarter 2016), and demands further elucidation. THE CURRENT CONDITION OF THE RURAL ENVIRONMENT OF HAITI Centuries of deforestation in Haiti have played out on an overwhelmingly mountainous terrain that receives seasonal rains and torrential downpours from frequent tropical storms and hurricanes. Exacerbated and accelerated by the exposed agricultural fields of a densely populated rural countryside, deforestation and alluvial soil erosion have produced predictable environmental consequences: the loss of topsoil, decreasing soil fertility, overall agricultural decline, and rural out-migration. A less-predictable consequence of deforestation was the widespread encroachment of exotic tree and shrub species, to collectively cover approximately 3/4ths of Haiti’s land surface by approximately 2010 (Tarter 2015b; Churches et al. 2014; White et al. 2013). A History of Landscape-level Land Management Efforts in Haiti 22 In 2012, geographers from Virginia Tech undertook a land-use/land-change analysis of Haiti’s largest offshore island of La Gonâve, using two high-resolution Landsat satellite images, twenty years apart, from 1990 and 2010 (White et al. 2013). Both images in this analysis were selected from late January—in Haiti’s dry season—to control for seasonal variability in vegetation and to diminish cloud cover. In this classification, researchers considered five different land-cover types that were mutually exclusive and exhaustive of all other land types in the area. Several complementary and higher- resolution (30 cm/pixel) satellite photographs were consulted in the accuracy assessment of the image classification, in combination with the analysts’ knowledge of the area, and a two- month field excursion to visit a stratified 6 random sample of validation points to ground- truth the classifications. The team averaged 61 points of verification for each of the five land-use categories (n=301 total), locating the random geospatial coordinates with hand-held GPS units. The overall accuracy of their 2010 classification was 87%, with a Kappa coefficient of 0.84 (White et al. 2013: 498). The results show that the percentage change in land area on La Gonâve from 1990 to 2010 for Agricultural land, Forest/DV 7 , Shrub, and Barren/Eroded land classes, were −39.73%, −22.69%, +87.37%, and −7.04%, respectively (White et al. 2013: 499). While the overall percentage of the ‘Forest/ DV’ land-cover decreased over 20 years by 22.7%, the ‘majority of 1990 agricultural lands were converted to shrub (45.01%) and forest lands (34.23%),’ and ‘56.2% of the barren/eroded land area in 1990 has been revegetated’ (White et al. 2013: 499-500). Thus, the entire land surface of La Gonâve in 2010 (excluding water and masking the < 2% cloud cover) was 40.4% covered with woody shrubs and 46% covered 6 The random sample strata were fifteen 52-meter elevation increments, to look for elevation-based influences (White et al. 2014: 498). 7 According to their classification, the latter category (Forest/DV) ‘is composed of all vegetation growth levels beyond the shrub stage’ (White et al. 2013: 496). with Forest/DV (White et al. 2013: 499). The researchers’ observations support the hypothesis of an arboreal shift in Haiti: There is strong evidence on La Gonâve indicating that previously cleared areas in the lowlands (regions of poor soil nutrients derived from limestone regolith and limited soil horizons) are beginning to revegetate initially with xerophytic species as agricultural land is abandoned after soil exhaustion. This process was discussed as a common experience in informal interviews with farmers on La Gonâve. (...) Revegetation on La Gonâve is primarily illustrated by the increase in shrub cover. Though shrub was the second largest land cover type in 2010, field observations indicate that the majority of the Shrub class cover varied along the scrub-dense vegetation gradient. Revegetated areas are composed of secondary succession forest, predominantly Acacia species [bayawonn 8 ] maturing from shrub to grove. Woody shrubs are regularly harvested for charcoal production and for export roughly at the pole stage (...) A large portion of the residents’ financial income on La Gonâve is supported by the charcoal export to the mainland. The shrub cover may be caught in a perpetual cycle of harvest and revegetation rarely reaching early succession or the pole stage in growth (White et al. 2013: 503). While forest cover on La Gonâve has decreased, it has been replaced by woody shrubs, largely in the form of Prosopis, which appear to be managed 8 White et al. (2013) now believe that the acacia species noted in this passage is in fact bayawonn, and have made this correction in a subsequent manuscript they are preparing from the same data (White et al., email communication, 3/23/2015). The authors indicate that bayawonn on La Gonâve is cut at the ground level, with some coppice limited to a few stems rather than multiple stems (ibid.). In other words, as in the colonial era, barren and abandoned agricultural lands were encroached by bayawonn from lowland areas, and these trees are now ‘harvested on a regular basis for charcoal production’ (White et al. 2013: 503). 23 for charcoal production 9 . Overall, the ‘shrub’ coverage on La Gonâve increased by 87.4% from 1990 to 2010 (White et al. 2013: 503). In a similar study, a geographer, geologist, and a natural resource management specialist collaborated to determine forest cover for the entire country of Haiti (Churches et al. 2014). The authors used five Landsat satellite images (2010-2011) from Haiti’s dry season, which provided coverage of the entire country. After a series of standard renderings and corrections, the authors reclassified their satellite images using FAO’s forest class definition, thereby creating a low (FAO) to high (authors’) range of percentage- tree-cover. These image classifications were verified through the application of a stratified 10 sample of 1,525 random reference points to higher resolution satellite imagery. Their ‘tree cover’ class had a users’ accuracy of 86% and a 0.81 Kappa coefficient, and the overall classification accuracy ranged from 78% (reference point counts) to 83% (class proportions) (Churches et al. 2014: 211). The results of the nationwide analyses showed that in 2010-2011, trees covered between 29.4% and 32.3% of Haiti’s land surface, 11 and that ‘shrub/ herbaceous’ areas covered between 45.7% and 48.6% of the land surface. Rather than a thinly dispersed arboreal covering, trees are aggregated in fragments and patches (Churches et al. 2014). It should be noted that both studies used slightly different land-use classifications: in the island of La Gonâve study, researchers combined ‘forest’ and ‘dense vegetation’ into one category, 12 and classified woody ‘shrub cover’ as separate category (White et al. 2013); in contrast, the nation-wide study aggregated ‘shrub,’ ‘herbaceous,’ and ‘agricultural’ land-uses into one 9 White, personal communication (3/23/2015). 10 Strata were based on the land use distributions from their initial classification (Churches et al. 2014: 209). 11 Includes ‘water’, ‘wetlands’, ‘bare/non-vegetated’ and ‘cloud’ categories. 12 ‘Forest/Dense Vegetation class is composed of all vegetation growth levels beyond the shrub stage. Forest lands were considered to have a tree-crown areal density of ~10% or more’ (White et al. 2013: 498). category, 13 and used a single category for ‘tree cover’ 14 (Churches et al. 2014). This aggregation is a setback for establishing nationwide percentages of managed charcoal woodlots, which would be best inferred through the woody vegetation that may have been split in both studies between tree and non-tree categories. Nevertheless, both the La Gonâve and the nationwide study use a single category that operationalizes trees in the same way, and similarly restricts smaller shrubs, and all other land uses. In the La Gonâve study, approximately 46% of the island’s land surface is covered with forest and dense vegetation beyond shrubs (White et al. 2013: 499), while the nation- wide study found approximately a third of Haiti’s land surface is tree-covered (Churches et al. 2014). In a broader study of land changes within the Greater Antilles between 2001 and 2010, 26 different Haitian municipalities (sections communales) underwent significant changes to woody vegetation 15 (8 decreased and 18 increased); 36 municipalities underwent significant changes in agriculture/herbaceous 16 (25 decreased and 11 increased); and 48 municipalities experiences significant changes in mixed-woody/plantations 17 (9 decreased and 39 increased) (Álvarez-Berríos et al. 2013: 88-91). Across municipalities of significant change, there was an 8% loss of woody vegetation, a 114% loss of agriculture/herbaceous, and a 133% increase in mixed-woody/plantation (Álvarez-Berríos et al. 13 ‘Vegetation less than 5 m in height or lacking definite structure such as stems or shoots. Woody vegetation included if crown cover is <10% and height <5 m. Includes all agriculture’ (Churches et al. 2014: 207). 14 ‘Vegetation greater than 5 m in height with a canopy cover of ≥10%. Includes mangroves. Does not include fruit-tree plantations’ (Churches et al 2014: 207). 15 “Woody vegetation was trees and shrubs with >80% cover” (Álvarez-Berríos et al. 2013: 86). 16 ‘Agriculture/herbaceous vegetation was annual crops, grasslands, and pastures with >80% cover’ (Álvarez- Berríos et al. 2013: 86). 17 “Mixed-woody/plantations was woody vegetation with a 20 to 80% cover, including agriculture/herbaceous vegetation or bare soil as background, as well as all forms of plantations and perennial agriculture” Álvarez- Berríos et al. 2013: (86). A History of Landscape-level Land Management Efforts in Haiti 24 2013: 88). Conversely, in overall land percentages for the entire country, woody vegetation increased from 1%, agriculture decreased from 4%, and mixed-woody/plantations increased from 4% (Álvarez-Berríos et al. 2013: 89). Municipalities that experienced losses or gains were widely geographically distributed. The sum of these three remote-sensing analyses suggests that the arboreal expansions that occurred in the post-independence period may be in swing again. Figures 1 and 2 represent the most recent and most empirically grounded estimates of arboreal coverage in Haiti. While most of Haiti’s original hardwood forests are gone, much of the land is now covered by a combination of trees and woody shrubs (see Figures 1 and 2). This current reality stands in stark contrast to depictions of Haiti as a lunar landscape devoid of trees. Ultimately this difference depends on how terms like ‘forest’ and ‘tree cover’ are understood. The phenomenon begs at least three important questions: (1) how did this arboreal encroachment occur; (2) how has it gone largely unnoted; and (3) what implications does this new landscape, which dominates much of contemporary rural Haiti, have on efforts to initiate landscape-level management schemes in Haiti? The arboreal landscape that currently characterizes the Haitian countryside is the result of several phenomena and their interactions in an iterative cycle (Tarter 2015b). The historical deforestation of Haiti simultaneously removed original hardwoods and depleted many of the soil profiles necessary for the regeneration of original hardwood species. Deforestation also exacerbated a drying trend on Hispaniola that began as early as the middle of the Holocene epoch (Higuera-Gundy et al. 1999). Climatic drying led to further cutting of original trees for wood and charcoal, as a necessary reaction to declining agricultural possibilities. Another permissive influence in the spread of these tree species can be indirectly traced from the Haitian land succession system, which is based on the Napoleonic Code of the colonial period. In this system, all agriculturally productive land in Haiti is inherited bilaterally and divided equally amongst all siblings, male and female. While certain unproductive familial land is considered sacred, inalienable, and is never divided (Herskovits 1937), in contrast, arable plots become smaller every successive generation at every new cycle of inheritance (Murray 1977). The inheritance of land plots that may be prohibitively small for agricultural production was initially met by an adaptive strategy on the part of Haitian farmers, which aligns with the theory of economist Ester Boserup (1965): the agricultural fallow period was shortened or eliminated (Murray 1977). While the land fallow-shortening strategy worked for a time, the associated agricultural intensification ultimately accelerated processes of declining soil fertility, climatic drying, and erosion. The exotic tree species (Prosopis juliflora and Haematoxylum campechianum) that now dominate much of the rural landscape of Haiti are highly adapted to poor soils, low moisture, and prolonged drought (Tarter 2015b). Much of the land that was eventually rendered agriculturally useless in Haiti became covered by these tree species, which are prolific seeders, propagate in a variety of novel ways, rapidly encroach into new territory, and are protected from foraging animals by thorny trunks and branches (ibid.). The predominance of these trees provides some ecosystem services, such as nitrogen fixation and the protection of the soil from chronic weathering by the elements and by acute weather from tropical storms and hurricanes. The deep taproots and lateral rooting of these species help break apart the limestone substrate that underlies much of Haiti’s topsoil, permitting a more rapid regeneration of soil in areas of extreme degradation, and a more rapid percolation of surface water into the aquifer. In some cases, fast growing-trees in Haitian farmer-managed woodlots have served as de facto nurseries, providing the necessary conditions for the regeneration of indigenous hardwoods, other trees and plants, and as habitat for animal species (Smucker and Timyan 1995). Yet the most-immediate benefit of Haiti’s current arboreal expansion is not in the form of 25 FIGURE 2. The Current Extent of Tree and Shrub Cover on the Island of La Gonâve, Haiti: 1990 and 2010. (White et al. 2013: 500) FIGURE 1. The Current Extent of Tree and Shrub Cover in Haiti in 2010. (Churches et al. 2014: 212) A History of Landscape-level Land Management Efforts in Haiti 26 ecosystem services, but as an economic input to the farmers who manage these species sustainably for continued charcoal production (Tarter 2015b). Prosopis juliflora and Haematoxylum campechianum coppice aggressively from the stump; where original hardwoods disappeared when cut, these rapidly spreading exotic plants continue to return after multiple cuts, making them ideal species for charcoal production (ibid.). Charcoal production is not easy - it is labor intensive and often pays less than agricultural food crops. But charcoal production using managed woodlots of coppiced trees is nevertheless a welcome possibility in rural Haiti in the light of declining harvests and increasingly chronic crop failures. The trees and woody shrubs that populate much of rural Haiti in the form of managed woodlots failed to spread at the national level in the first part of the 20 th century because charcoal production was centralized in a few accessible locations: not only were original forests rapidly depleted of original trees, but overexploitation of wood resources also prevented a notable establishment of these subsequent exotics. But the advent of 900 kilometers of new roads took pressure off these few original locations, decentralizing charcoal production at a national level and permitting the spread and reproduction of exotic species at a rate that eventually contributed, in part, 18 to their current extend in the rural Haitian landscape (see Figures 1 and 2). This brief history together with our understanding of current levels of tree coverage in Haiti challenge at least one misguided narrative that pertains to landscape-level development projects in Haiti: that charcoal production is responsible for the denudation of Haiti’s forests and that the key to successful reforestation is the elimination of the charcoal trade. On the contrary, 18 It is undeniable that the current arboreal expansion in Haiti may be partially contributable to other causes, such as the nationwide tree-planting efforts of the 1980s and 1990s. However, rural out-migration, the abandonment of formerly productive land, and an increase in pseudo- drought conditions all provide the necessary conditions for the spread of the species thought to dominate much of the rural countryside at this time in history. much evidence exists that the charcoal trade in Haiti occurred after much of the deforestation of original forests and that charcoal practices are currently meeting urban energy needs, decades beyond the dire warnings of analysts during the wood crises years of the 1970s and 1980s. Land Inheritance, Land Tenure and, Land Security Many common misconceptions repeatedly find their way into the development literature on Haiti, regarding land inheritance, land tenure, and land security. An illustrative example of such misconceptions is the notion that Haitian farmers do not have secure tenure of their land. However, numerous studies have taken up the issue and all have reached the same conclusion: land tenure in rural Haiti is secure and not an obstacle to landscape-level land management (Murray 1977, 1978b, 1987; White and Runge 1994, 1995; White 1992; White and Jickling 1995; Smucker et al. 2005, 2007). Several general features pertaining to land inheritance, tenure, and security, which may inform landscape-level land management processes, hold true across much of rural Haiti: Land Inheritance • All land is inherited bilaterally and equally between siblings, male and female; • While all inherited land is capable of division, and all inheritors may claim an equal share of land, not all inherited land is immediately divided; • Arable land is usually divided immediately, rendering land suitable for cropping into successively smaller and more fragmented plots each generation; • Manly arable plots of land are frequently too small to meet the needs of a single farm family; the phenomenon of fragmented arable plots is circumvented by the strategy of most farm families to acquire several such parcels through purchase; as a result, arable plots are not only 27 fragmented and small, but typically distributed in a discontinuous pattern across the landscape; most arable land plots are alienable, and the selling and purchasing of such plots is widespread and common. • In many areas of Haiti, farmers also have access to land collectively held by kin members. Such land is typically much larger in size because it has been neither informally nor formally divided, frequently owing to the land’s lower capacity for agricultural crop productivity. Such lands typically have steeper slopes, a less preferential aspect (rain and sun exposure), poorer soils, and are generally less desirable than flatter, frequently lower-elevation, arable lands. These lands are typically used for the production of charcoal, as a location to tether foraging animals, and sometimes as fruit orchards and coffee groves. These collectively owned lands also meet other needs, including non-material needs, such as a location for the veneration of the ancestors and other spirits of the Vodou (voodoo) pantheon (Tarter 2015a). The utilization of such land functions through a complex internal system that permits kin members to enter and equitably access available resources. However, due to reluctance or a lack of impetus to divide such lands, several generations of kin members may claim access to the land, complicating the management of the land’s resources, and increasing the likelihood that the land will remain undivided. Land Tenure • Rural Haiti operates on two complimentary systems of land tenure - one is informal and the other is formal (Murray 1977); • The informal tenure system has high internal security: it is widely recognized and respected within communities and often functions on agreements between neighbors or kin, in the presence of neighbors or kin, and therefore may a traceable material record of use and access; • The formal land tenure system functions and is legitimized through the legal apparatus of the Haitian Government, and has higher levels of external validity than the informal system; • Both the formal and informal tenure systems are secure in different ways, and Haitians rely on both systems for different needs; and • Rural Haitians also benefit from a land leasing system, which may be an informal agreement between the land-owner and lessee, or may be formalized by the use of a local notary. A basic understanding of land tenure is important for any effort to promote or institute landscape-level management projects in rural Haiti. The current land tenure systems are complex but generally not an impediment to project participation, investment, and ownership. The Haitian land tenure system is complex, but functional. By using the informal tenure system, rural Haitians avoid the fees and taxes associated with dividing land formally (Murray 1977, 1978b, 1987; White and Runge 1994, 1995; White 1992; White and Jickling 1995; Smucker et al. 2005, 2007). Rural Haitians utilize the formal land tenure system when land sales and purchases are undertaken outside of the kinship network or outside of the village context (ibid.). The vast majority of rural Haitian farmers are content with their current land tenure systems, which are not only functional, but are also historically adaptive (Murray 1977). Land tenure reform is rarely listed in the litany of needs expressed by Haitian peasants themselves, though access to land arable is an on-going frustration expressed by younger members of society, at least historically (Murray 1977, 1981). Land Security The situation in Haiti is different in regards to land security. Haiti has a long history of urban elites and rural gran don (a wealthy rural middle class) attempting to control the peasantry through opportunistic land purchases, ‘legal’ maneuvering, and varied strategies of displacement (Trouillot 1990; Steckley and Shamsie 2015). A subsequent case study on the Péligre Dam provides an illustrative example of this well-documented A History of Landscape-level Land Management Efforts in Haiti 28 historical trend. It is entirely possible that aspects of the land tenure system in place in rural Haiti developed in part as a strategy to mitigate land insecurity. Given that the vast majority of arable land in Haiti is represented as a patchwork of discontinuously dispersed and privately owned fragmented plots of land, the issue of land security is perhaps less than in earlier historical periods when the aggregation of arable land in the hands a fewer land-owners might engender a higher tendency to land grabs and predatory land acquisition. History suggests that while land security, land inheritance practices, and land tenure systems in rural Haiti are complex, they are fully functional and in no way impede the ‘plot-based’ approaches to landscape-level improvement projects. However, given the fragmented nature of most land plot holdings, ‘macro-level’ approaches to landscapes are unlikely to meet much success at this point in the history of rural Haiti. Soil Conditions in Haiti The loss of soil is the most immediate and observable effect from the initial deforestation of Haiti. The UNESCO Haiti Pilot Project noted degrading soil conditions as an issue of concern as early as the mid-20 th century. The decline of agricultural productivity due to soil degradation has since been systematically documented in Haiti (Zuvekas 1978; Pierre-Louis 1985; Bargout and Raizada 2013). The effects of erosion, declining soil fertility, and decreasing arable plot sizes have led to a nationwide agricultural decline. Despite this trend, the majority of the rural-dwelling Haitians continue to pursue agriculturally based livelihood strategies, supplemented heavily by remittances sent from Port-au-Prince and diaspora abroad. Soil and the Relationship to Ecological Zones of Haiti One of the most commonly employed systems for the classification of ecological zones in tropical and subtropical areas was developed by forester L.R. Holdridge in the early 1940s, but has since been adapted and applied on a global scale. Because Holdridge’s system was developed in Haiti, it continues to be the primary classificatory schemes used in much of the literature related to the ecological aspects of landscape-level land management in Haiti. The system delineates nine different zones based principally on the metrics of annual precipitation, mean annual temperature, the potential evapotranspiration ratio (PET), and considerations of elevation (Holdridge 1947, 1967). Four of the Holdridge Life Zones (HLZ) cover less than 1% of Haiti’s land, while the remaining five constitute notable percentages (Delatour et al. 1984). The largest zone, Subtropical Moist Forest, representing approximately half the entire land area of Haiti, is found at an elevation of ~800 meters. It receives 1,200-1,800 mm of annual rainfall, and supports the largest percentage of small-holding peasant farms (Delatour et al. 1984; Ehrlich 1985). The Subtropical Moist Forest zone was traditionally home to large stands of mahogany and tropical oak, now replaced principally with avocado and mango trees (ibid.). The second largest HLZ is Subtropical Dry Forest, constituting 19% of Haiti’s land, falls under 400 meters of elevation, receives an annual mean rainfall between 800-1,000 mm, experiences seasonal droughts, but has deep, irrigable soils, and is highly agriculturally productive (Delatour et al. 1984). Common trees in the Subtropical Dry Forest zone include Phyllostylon brasiliensis, Prosopis juliflora and Guaiacum officinalis. Viable crops include mangoes, limes, tobacco, cotton, plantains, sugar cane, and sisal. While these two HLZs constitute approximately 70% of Haiti’s land surface, they and the other zones are unevenly distributed across Haiti’s landscape. Thus, it remains a challenge to incorporate ‘one-size-fits-all’ program policies at the landscape level. A private consulting firm consolidated and adapted Holdridge’s ecological zones for Haiti in order to incorporate population density levels and [... middle sections omitted for long document ...] 71 Van der Plas, Robert. 2007. Haiti: Strategy to Alleviate the Pressure of Fuel Demand on National Woodfuel Resources. Energy Sector Management Assistance Program, The World Bank Group. ESMAP Technical Paper No. 112/07. Verna, Chantalle F. 2015. Haiti, the Rockefeller Foundation, and UNESCO’s Pilot Project in Fundamental Education, 1948-1953. Diplomatic History. Vol. 0, No. 0 (no pagination; advanced access). Verner, Dorte and Willy Egset
 2006 Haiti: Social Resilience and State Fragility in Haiti: A Country Social Analysis. Report to the World Bank. Report No. 36069-HT. Versluis, Anna and John Rogan. 2009. Mapping land-cover change in a Haitian watershed using a combined spectral mixture analysis and classification tree procedure. Geocarto International. Vol. 25, No. 2, pp: 85-103. Versluis, Anna J. 2008. We All Live Downstream: Disaster, Land Change and Reciprocity in A Haitian Watershed. Geography Dissertation. Clark University, Worchester, MA. Versluis, Anna. 2010. Highlands management in flood-prone watershed: Does reflexive reciprocity make a difference? Global Environmental Change. Vol. 20, pp: 333-341. Voltaire, Karl. 1979. Charcoal in Haiti. Report to USAID/Haiti. Port-au-Prince. Wha, Tatiana. 2013. Engaging the Haitian Diaspora: Emigrant Skills and Resources are needed for Serious Growth and development, not Just Charity Cairo Review, 9/2013, pp: 56-69. Watras, Joseph. 2010. UNESCO’s Programme of Fundamental Education, 1946-1959. History of Education. Vol. 39, No. 2, pp: 219-237. White, Justin, Yang Shao, Lisa M. Kennedy and James B. Campbell. 2013. Landscape Dynamics on the Island of La Gonâve, Haiti, 1990-2010. Land, 2, pp: 493-507. White, T. Anderson, and C. Ford Runge. 1994. Common Property and Collective Action: Lessons from Cooperative Watershed Management in Haiti. Economic Development and Cultural Change. Vol. 43, No. 1, pp: 1-41. White, T.A. and C.F. Runge. 1995. The emergence and evolution of collective action: Lessons from watershed management in Haiti. World Development. Vol. 23, No. 10, pp: 1683-1698. White, Thomas A. 1992. Peasant Cooperation for Watershed Management in Maissade, Haiti: Factors Associated with Participation. EPAT/MUCIA Working Paper No. 4. The Environmental and Natural Resources Policy and Training Project. Minnesota Agricutlural Experiment Station, Project Nos. MN-42-049 and MN-42-035. White, Thomas Anderson and Jon L. Jickling. 
 1995. Peasants, experts, and land use in Haiti: lessons from indigenous and project technology. Journal of Soil and Water Conservation 50, pp: 7-14. Wood, Diane Walton, and Frederick J. Conway. 1984. Assessment of Haiti’s Indigenous Environmental Non-governmental Organizations and Recommendations for a USAID Program to Support Them. Environmental Planning and Management Project. International Institute for Environment and Development. US AID. Contract No. NGO/FA/86-01. World Bank. 1982. Staff Appraisal Report: Haiti Forestry Project. Projects Department, Latin America and the Caribbean Regional Office. Report No. 3776a-HA. World Bank. 1991. Staff Appraisal Report. Haiti: Forestry and Environmental Protection Project. Country Department III: Latin America and the Caribbean Regional Office. Report No. 9307-HA. World Bank. 2015. Haiti: Towards a New Narrative. Systematic Country Diagnostic. Latin America and Caribbean Region, May 2015. Zuvekas, Clarence. 1978. Agricultural Development in Haiti: An assessment of sector problems, policies, and prospects under conditions of severe soil erosion. Report to USAID. Report Control No. PN-AAF-587. A History of Landscape-level Land Management Efforts in Haiti 72 SECTION V: Appendices 73 APPENDIX A: Challenges to this Literature Review The literature on landscape management in Haiti is voluminous, and a thorough undertaking of all literature was not possible. Development reports are frequently influenced by the mandates of the producing institution. Many reports failed to adequately describe their research design and data sources, or acknowledge potential sources of bias in their methods and conclusions. In some cases reports are clearly driven by the implicit or explicit agendas of the institution or the individual researcher, and fail to note any project shortcomings or need for policy refinement. This trend is illustrated by the authors’ review of several evaluations that praised projects widely acknowledged (in Haiti) to have failed. A final related shortcoming of evaluating literature on development projects in Haiti concerns an historical lack of data, questions of data accuracy, issues of data accessibility for particular periods of time, and uneven topical and temporal coverage. 29 The unfortunate result of this phenomenon has been that many Haiti researchers must rely on dated or outdated sources. The issue is compounded when researchers cite secondary or tertiary sources. Figures provided in a 2010 report might derive from a secondary source published in the 1980s, which relied on partial or questionable data from the 1950s. An illustrative example of this tendency can be found in the case of the 1972 OAS report that aggregated 100+ watersheds in Haiti into 30 pragmatically organized zones. This report has since been cited to the point where many researchers suggest that there are only 30 major watersheds in Haiti. Stated more succinctly, the available literature on landscape management projects in Haiti 29 While certain historical periods (eg. The period surrounding the 2010 earthquake) are replete with literature, other periods (eg. The period surrounding the Duvalier presidencies) are marked a significant decline in published research. varies not only qualitatively, but also in quality and objectivity. The solutions to these challenges are beyond the scope of this report, but point to the necessity of independent, objective project analyses, based on field-site visits and off-the- record interviews with project personnel and project beneficiaries. APPENDIX B: Landscape Level Management Schemes in Haiti (Reproduced from Smucker et al. 2007: 18-22): Landscape Engineering From the 1950s to the 1970s there were bilateral projects with the Government of Haiti using an approach commonly described as “équipement du territoire,” a landscape engineering strategy. These projects were done on both private and public lands using a top-down management approach to cover large, contiguous areas with conservation structures. Soil conservation was considered to be a strictly technical problem to be solved by engineers installing mechanical structures, mainly rock walls and contour canals, using paid (cash or food for work) labor. Neither land tenure nor the interests of peasant landowners were taken into account. In many cases this resulted in very expensive rock walls constructed on infertile, steep, and droughty garden plots. The cost of installing the structures was many times greater than the value of any added increment of crop yield due to these conservation effects. In general this model is noted for its unsustainability and its near total lack of upkeep of the structures, most of which eventually disintegrated. This model should be discarded; although, there were some limited positive results derived from teaching farmers certain techniques or introducing tree species. Civic Infrastructure and Job Creation Job creation projects that became widespread during the middle 1990s share certain characteristics with the landscape-engineering model in that public works projects were accomplished with paid labor. Some of these A History of Landscape-level Land Management Efforts in Haiti 74 projects made a serious effort to respond to the specific needs and local characteristics of rural localities, for example, the rehabilitation and expansion of irrigation works operated by grassroots water user organizations outside of Mirebalais. Others merely served as a pretext to distribute salaries for political reasons and had no lasting results. In the future, job creation projects will be required as a needed element of watershed management, especially for infrastructure repair and treatment of water courses, but such efforts must fit within an overall strategy so that resources are used efficiently. Plot-Based Models Beginning in the 1980s, donors funded agricultural extension projects through grants and cooperative agreements directly with NGOs. They specifically sought a greater understanding of peasant farming systems and tried to design field approaches accordingly. Programs emerged which engaged in tree planting and many other types of soil conservation techniques beyond the rock and canal structures favored previously. Fund management was often done by a large international NGO acting as an umbrella agency, receiving funds from the donor and in turn administering smaller agreements with local NGOs or farmer groups of various sizes and strengths. Some projects concentrated their activities in smaller regions while others spread out over larger areas. The basic unit of work within the NGO or community-based farmer organization (CBO) was the farmer work group, and project interventions were offered from a basic list of contour and gully soil conservation practices, tree planting, crop seed improvement, fruit tree grafting, marketing, and institutional strengthening activities. Extension agents were selected by the contracting groups from within their own members; these were trained by project technicians and acted as liaisons between project staff and participating farmers. These projects were not designed to cover whole watersheds. Farmers did not treat all of their plots, which varied in their slope, distance from the residence, security of access, soil fertility, contribution to household economy, and other factors. The principal reason farmers chose not to treat all plots was undoubtedly the high threshold costs for doing so. Treating degraded hillside slopes “correctly” was beyond the means of most farm families. Significant landscape-level changes did occur in some regions due to plot-based project interventions, and three such case studies are presented later in this chapter. This was an interesting period because new technical approaches and new partnerships between technicians and farmers were developed, and the implementing agencies attained greater understanding of peasant farming systems. Nevertheless, the plot-based model, although useful, was not designed to solve the fundamental problems of watershed protection. Watershed-Based Strategies The stated focus of these projects is the watershed, or more realistically, a portion of a watershed. This focus defines the geographic limits and the intended type of outcomes, which are primarily environmental and intended to benefit the broader society by reducing the damage to infrastructure. Paradoxically, design of watershed projects is often based on lessons drawn from plot-based projects where the objectives are mainly concerned with increasing agricultural income rather than producing environmental benefits for downstream users. Resolving the differences between these contrasting objectives will be crucial to designing an effective watershed project. The technical approach and outreach, funding and benefit-flow mechanisms vary across projects. Much debate has taken place in Haiti about watersheds, especially since the release of a report by Löwenstein (1984) on the deteriorating condition of the Macaya watershed; however, no satisfactory watershed-level solution has yet evolved, even though some interesting results 75 were obtained along the way. For example, the USAID-funded ASSET project…worked for three years in the Grise/Blanche watersheds and was designed to work at a landscape scale (Israel et al. 2001); however, the project did not have a major effect on flooding, sedimentation, or water supplies in the lower watershed. This was due largely to the project’s limited scale of operation in the upper watershed. In general, evaluation of watershed projects points to the importance of designing approaches that consider charcoal markets, animal markets, infrastructure, household water supply, and appropriate social structures. Most importantly, building local social capital needs to be taken to a new level when considering watershed-level interventions. It is imperative to consider off-farm externalities and build new social links that protect the interests of participating farmers whose long history of economic and social disadvantage has led to the creation of their present systems. These persistent systems will not readily be put at risk. Market-Driven Models This approach seeks to augment small farmer income by developing or improving marketing links for agricultural crops. The objective of the marketing strategy is to assist farmers to generate sufficient profits to cover their normal agricultural expenses plus the additional expenses required to invest in better germplasm and soil and water conservation. In other words, the intent of project designers is for marketing to drive improved natural resource management. A market driven strategy was introduced in the Productive Land Use Systems project (PLUS) and then substantially expanded by the Hillside Agricultural Program (HAP) since 2001. The HAP project included natural resource management components in the early years, but specific NRM interventions were eliminated in 2003 due to funding cuts. HAP marketing activities have successfully created new opportunities for tree crops (coffee, mangos, and cacao) that have benefited farmers by boosting farmgate sale prices. This may have had a positive effect on the environment, but the project has not tracked the effect of marketing on NRM. Furthermore, HAP promotes improved marketing opportunities for a given crop regardless of where it is grown, so the effects are not concentrated nor do they consider whole farming systems and their impacts on the watershed. The narrow scope of marketing projects limits their ultimate effect on natural resource management. Nevertheless, marketing is a key consideration and must be central to future watershed projects (see “Tree Products and Trade” later in this chapter). (Reproduced from Smucker et al. 2007: 18-22). APPENDIX C: Watershed Management Principals of the Ministries of Environment and Agriculture (Reproduced from Bush and Sildor 2009: 30- 32): Ministry of Environment Principle 1: The objectives set for the management of natural resources and biodiversity are society‘s choice. Principle 2: Management must be decentralized and conducted at the lowest possible level. Principle 3: The managers of an ecosystem including a watershed must consider the effects (real or potential) of their activities on adjacent ecosystems or elsewhere. Principle 4: Given the advantages of good management, it is important to understand the economic context of the ecosystem. All management programs should (a) reduce market distortions that have an negative impact on biodiversity; (b) align the incentives that favor the conservation and sustainable use of biodiversity; (c) internalize the external costs and benefits of ecosystem management. A History of Landscape-level Land Management Efforts in Haiti 76 Principle 5: The ecosystem approach should be applied only at the appropriate scale. Principle 6: Given the time scales and the different phases that characterize ecological processes, the management of ecosystems should set long term objectives. Principle 7: An ecosystem approach should determine the appropriate balance between conservation and use of biological diversity. Principle 8: An ecosystem approach should consider all the available information— including scientific data, but also local knowledge and practices. Principle 9: An ecosystem approach should include all social sectors and all the pertinent scientific disciplines (reproduced from Bush and Sildor 2009: 30-32). Ministry of Agriculture Principle 1: Intervention should be in the context of the municipality level Natural Resources Management plan established by the Conseil Municipal for setting priorities. The plan should address the technical and methodological criteria fixed by the MARNDR (before execution, all projects must go before the DDA for review, possible revisions, and a final decision made by the central office). Principle 2: Supporting structures (NGO, private sector, government agencies) should aim to help the collectivités territoriales and agricultural associations appreciate the resources in their area and to identify actions to increase their value. After an assessment of physical, land-tenure, economic, and social factors an action plan will be prepared and implemented. Principle 3: Through their representatives as members of collectivites territoriales and agricultural organizations, the local population will participate in all the steps involved: request for assistance; assessment of constraints, opportunities, and needs; seeking and participating in fund raising; deciding on priority actions (zoning, techniques, participating communities, calendar, etc.); financial management, monitoring and evaluation; making adjustments. Principle 4: All actions must be the result of negotiation at each step among the users of the resources and the supporting partners. This negotiation should eventually give rise to management agreements. Principle 5: The approach should be gradual: projects should improve the existing situation but without abrupt short term changes in social or environmental status quo. Priority should be given to supporting existing agricultural organizations. The techniques applied should be integrated technically, economically and socially into existing production systems. They should particularly take account of the characteristics of the zone, take advantage of local knowledge, be integrated into the agricultural calendar, be adapted to local materials and the local availability of labor. Principle 6: The objectives and strategies of management plans should be aligned with the priorities of the users—which is most often the maximization of revenue. Management should aim at both conservation and production: • Actions are not necessarily targeted at the most degraded areas; on the contrary, there should exist sufficient agricultural potential to allow investment costs to be recovered. • Actions should be capable of becoming economically independent. • For each agricultural product, the whole value chain should be strengthened (production, post harvest systems, processing, commercialization). • Actions must be economically viable in the short term—especially if the community situation is precarious. Principle 7: Emphasis should also be placed on non-farm income-generating activities (ecotourism, rehabilitation of access roads, stabilization of gullies, agri-processing, artisanal products) in order to reduce demographic pressures on natural resources. 77 Principle 8: Supporting measures must ensure the technical and organizational training of the communities. Principle 9: Supporting agencies should assist the communities to manage the financial resources allocated to the project in order to help the communities develop their skills and competence in financial management. However, the level of funding should be aligned with the capacity of the communities to manage these funds. Principle 10: It is a basic principle not to finance local unskilled labor for intermittent work on soil conservation (reproduced from Bush and Sildor 2009: 30-32).