Agroforestry Systems and Local Institutional Development for Preventing Deforestation in Chiapas, Mexico

The transformation of natural forest to secondary forest and pastures has been the most common process of land use change in tropical countries in recent decades (FAO, 2010). The main causes of deforestation include institutional factors, markets, public policies and global forces, which often act synergistically (Deininger and Minten, 1999; Bocco et al. 2001; Lambin et al., 2001). Mexico is a country with 64,802x103 ha of forested land, and it is one of the ten countries with the largest area of primary forest (3% of total). The annual net loss of deforestation in Mexico has been estimated to be 0.52% for the period of 1990-2010, but the net loss, on average, has decreased over the past few years (FAO, 2010). The highest deforestation rates are concentrated in the south and central regions of the country, as documented elsewhere: 8.4% in el Nevado de Toluca, state of Mexico (1972-2000) (Maass et al., 2006); 8% in Patzcuaro, Michoacan (1960-1990) (Klooster, 2000); 6.9% in some areas of Campeche (ReyesHernandez et al., 2003); 6.1% in the highlands of the state of Chiapas (Cayuela et al., 2006; Echeverria et al., 2007); and 2-6.7% in Selva Lacandona, also in Chiapas (Ortiz-Espejel & Toledo, 1998; de Jong et. al., 2000). Precisely, the states of Chiapas and Yucatan have registered the highest rate of forest conversion to grasslands and slash-and-burn cultivation over the past two decades, and Chiapas alone has contributed towards 12% of national deforestation during the period 1993-2007 (De Jong et al, 2010; Diaz-Gallegos et al., 2010). In Mexico, deforestation occurs because forests become converted to agriculture, livestock and urban areas. But also because logging activities fail to meet the requirements of forest management plans. All these processes result in the loss of forest goods and services (Lambin et al. 2003), and they contribute to ecosystem fragmentation (Ochoa-Gaona & Gonzalez Espinosa, 2000; Cayuela et al., 2006), biological invasions (Hobbs, 2000), greenhouse gas emissions (Watson et al., 2000), biodiversity loss (Lugo et al., 1993), soil degradation (Lal, 2004) and water siltation (Sweeney et al., 2004).


Introduction
The transformation of natural forest to secondary forest and pastures has been the most common process of land use change in tropical countries in recent decades (FAO, 2010). The main causes of deforestation include institutional factors, markets, public policies and global forces, which often act synergistically (Deininger and Minten, 1999;Bocco et al. 2001;Lambin et al., 2001). Mexico is a country with 64,802x10 3 ha of forested land, and it is one of the ten countries with the largest area of primary forest (3% of total). The annual net loss of deforestation in Mexico has been estimated to be 0.52% for the period of 1990-2010, but the net loss, on average, has decreased over the past few years (FAO, 2010). The highest deforestation rates are concentrated in the south and central regions of the country, as documented elsewhere: 8.4% in el Nevado de Toluca, state of Mexico (1972Mexico ( -2000 (Maass et al., 2006); 8% in Patzcuaro, Michoacan (1960-1990 (Klooster, 2000); 6.9% in some areas of Campeche (Reyes-Hernández et al., 2003); 6.1% in the highlands of the state of Chiapas (Cayuela et al., 2006;Echeverría et al., 2007); and 2-6.7% in Selva Lacandona, also in Chiapas (Ortiz-Espejel & Toledo, 1998;de Jong et. al., 2000). Precisely, the states of Chiapas and Yucatan have registered the highest rate of forest conversion to grasslands and slash-and-burn cultivation over the past two decades, and Chiapas alone has contributed towards 12% of national deforestation during the period 1993Díaz-Gallegos et al., 2010). In Mexico, deforestation occurs because forests become converted to agriculture, livestock and urban areas. But also because logging activities fail to meet the requirements of forest management plans. All these processes result in the loss of forest goods and services (Lambin et al. 2003), and they contribute to ecosystem fragmentation (Ochoa-Gaona & González Espinosa, 2000;Cayuela et al., 2006), biological invasions (Hobbs, 2000), greenhouse gas emissions (Watson et al., 2000), biodiversity loss (Lugo et al., 1993), soil degradation (Lal, 2004) and water siltation (Sweeney et al., 2004).

Agroforestry Systems and Local
Institutional Development for Preventing Deforestation in Chiapas, Mexico 335 representative area of the state's ongoing deforestation and degradation processes. The central part of the chapter highlights Scolel Te's conservation and forest management approach in our selected municipalities, and it discusses the drivers and constraints for the successful realisation of environmental, economic and social benefits in the three selected cases. The last part of the chapter discusses the overall results in the context of evolving carbon forestry markets and the through the enhancement, conservation and sustainable management of forest stocks.

The process of deforestation in Marqués de Comillas
Marqués de Comillas is one of the four municipalities of La Selva Lacandona and it comprises a total area of 203,200 ha. In the last few four decades, this municipality showed one of the highest deforestation rates in Mexico. Between 1986 and 2005 it lost approximately 81,080 ha of tropical rainforest, which represents 48% of its original forest cover ( Figure 2). This loss contributed to 1.5% of the total CO 2 emissions from land use change in Mexico (Castillo-Santiago 2009). Settlements in La Selva Lacandona were promoted during 1970-1980, and they have been recognised as the main driver of deforestation in Marqués de Comillas. Population was relocated to resolve conflicts of land scarcity and social rebellion in other Mexican states. Consequently, land in Marques de Comillas, which was completely forested at the www.intechopen.com Deforestation Around the World 336 beginning of the 1970s, was distributed in 37 ejidos 1 (Mariaca, 2002). The colonisation process was complemented by the construction of a main road, which allowed people and goods to flow between rural areas and the main cities of Tabasco and Chiapas (Harvey, 1998). Additionally, public policy favoured land use change from forest to maize agriculture, thereby financing deforestation in the 1970s. The process began with forest logging, followed by the use of fire for the cultivation of maize during three or four years. Once fertility was decreased and the land was weeded, due to intense cultivation, the natural steps were to intensify land use, fallow the land or let the grasses grow. Importantly, forested land was considered "idle land", and the process of deforestation was actually funded by government programs that encouraged cattle ranching. In Marqués de Comillas, crops, such as cardamom, cocoa, rice, vanilla and rubber, were promoted by public and private investments to diversify the agricultural system and to employ local people who mainly participated as labourers in the 1980s and 1990s. Most of these projects were unsuccessful, due to their continuous requirements of external inputs, particularly technical support and capital. Forest initiatives were discouraging as well. In 1996, the Pilot Forest Plan was launched to promote timber extraction and establish the basis for a rational use of community forests. However, only few ejidos have maintained the deforestation rate under the regional average. More recently, other public initiatives have favoured the growth of oil palm for biofuels, which are rapidly growing in this municipality, even at the expense of forests. Livestock has also gained ground. Livestock has very often represented the best economic option from the farmers' point of view because it has a relatively stable national market and a low level of investment is required. Livestock is thus still the dominant activity in the area and it occupies most of the landscape in Marques de Comillas at present ( Table 1).

Type of vegetation
Area ( Livestock in Selva Lacandona has normally been characterised by its extensiveness, and it has been devoted mainly to grow calves that are fed with naturalised grasses of low nutritional value. Moreover, it has been characterised as having a low management level, high number of animals per land unity, low capital investments, poor infrastructure, scarce technical assistance and scarce financial support (Jiménez et al, 2008;Martínez & Ruiz de Oña, 2010). Table 2 shows the main features of the livestock system in communities of Selva Lacandona. Most of the farmers (95%) have designed their systems to produce calves to sell them to intermediaries in the local and regional markets. The trees located in the pastures are mainly tropical forest remnant trees tolerated for shading cattle, including Blephariduim guatemalensis, Sabal mauritiformis, Vatairea lundellii, Guarea glabra, Albizia adinocephala, Bursera simaruba, Spondias mombin, and Swietenia macrophylla. Specific timber species, such as "popiste" (B. mexicanum) and guanacastle (Enterolobium cyclocarpum), are sometimes favoured with the purpose of being used as a lumber source in rural construction (local market or self-supply).
Recently, several studies have highlighted the importance of silvopastoral systems and other agroforestry systems for conserving biodiversity and connecting countryside landscape with reserves (Harvey et al, 2006;Rice & Greenberg, 2004). Several institutions, including the Commission for Natural Protected Areas (CONANP), the Mesoamerican Biological Corridor (CBM) and the National Commission of Knowledge and Use of Diversity (CONABIO) have launched programs in Selva Lacandona, specifically in Marques de Comillas, to improve rural production and promote conservation. One of these initiatives is the Scolel Te' project, which began in 1994 as a pilot experience in Chiapas and it is managed by the local organisation AMBIO cooperative. It originally involved a few dozens of farmers in the central highlands of the state and it has now grown to encompass more than 700 participants and their families (3500 beneficiaries) in approximately 50 communities of Chiapas and the neighbouring states of Oaxaca and Tabasco. The project is built on a participatory method that identifies rural development and forest management opportunities and constraints of each involved farmer and community. Subsequently, it helps farmers and communities to select and establish trees on individual or collective lands according to their preferences, providing technical support and paying participants for the provision of carbon offsets to national and mostly international buyers (Corbera 2010;Ruiz de Oña, 2011). Over time, the project has become a global landmark for the development of community-based payments for ecosystem services, and it has created its own design and implementation standard (www.planvivo.org). This approach has been extended to other similar projects in Uganda, Mozambique, Malawi and Cameroon

Avoiding deforestation in La Corona, Marqués de Comillas
In Ejido La Corona, a previous study estimated that 305 ha of forested land would have to be lost for agriculture purposes, as a baseline scenario during the period 2004-2009 (Quechulpa, et al., 2010). With the support of the AMBIO cooperative and financial support from Pro-Árbol (i.e., a governmental PES program developed by Mexico's National Forest Commission (CONAFOR) a participatory planning method was developed to design interventions for avoiding deforestation and increasing tree cover in deforested land. Concurrent financial resources were allocated from several programs, including Scolel Te, the Mexican Fund for Nature Conservation (FMCN) and Mesoamerican Biological Corridor (CBM). Financial resources were aimed to reduce pressure on forest land to avoid land-use change and to intensify cattle raising activities.
The following two approaches were suggested by community members and selected for project development: 1) management of secondary vegetation and forest conservation and 2) establishment of agrosilvopastoral and agroforestry systems in open and grazing areas. As a result, secondary vegetation was managed by pruning and thinning trees to eliminate competition and favour growth of the most commercially valuable species. Activities for forest conservation included the opening and maintenance of 22 km of fire protection rifts, acquiring equipment for fires, combating brigades trained for fire control, supervising and regulating agricultural burns. Communal forest conservation incorporated a wider vision of the territory and collective agreements related to resource access into the working plan, thereby establishing rules and monitoring to regulate land use change according to the plan. The establishment of 45.5 km of live fences was carried out in accordance to the work plan ( Figure 3) using "cocoite" forage trees (G. sepium) and other timber species, such as Tabebuia rosea, T. guayacan, C. odorata, S. macrophylla and Pachira aquatica. All forage trees were native species produced in communal nurseries. Other activities included the use of forage grasses such as Brachiaria, the promotion of cattle-feed supplements with multi-nutritional blocks, improvement of cattle breeds, establishment of livestock infrastructure, establishment of technical training and creation of a farmer-to-farmer exchange system (Ambio Cooperative, 2010). In parallel, other projects particularly targeted towards women, such as the substitution of traditional open stoves by fuel wood-saving stoves, improvement of house conditions and improvement of other communal infrastructure, were also launched.
In the first five years of the project, all of these activities resulted in 179 ha of prevented deforestation out of the 305 ha previously estimated to be lost, which translated in reduced CO 2 emissions from land-use change. According to inventories from permanent plots, 407 ha of passively restored secondary forests accumulated biomass with a rate of 4.4 ton ha -1 year -1 (approximately 3000 ton of CO 2 per year), and it was estimated that the establishment of live fences fixed approximately 3200 tonnes of CO 2 . Other intangible benefits included the organisation of farmers for fire prevention; the organisation and training of brigades for carrying out forest inventories; and the improvement of productive systems. Above all, the recognition that forests may contribute to environmental and socioeconomic direct benefits was one of the main gains resulting from the project.

Agroforestry in Chilón and Salto de Agua, Chiapas
Study communities of these municipalities are located in Chiapas tropical zone. Salto de Agua is located approximately 200 m a.s.l., and it has a warm and humid climate. In addition, Salto de Agua has a tropical rainforest. Meanwhile, Chilón is located in the intermediate tropical zone between 700 and 900 m a.s.l. It has a warm climate and abundant summer rains. Chilón has also a tropical rainforest, and the main soil types in this area are Regosols, Leptosols, and Cambisols (INEGI 1984;Soto-Pinto et al., 2010). Land is devoted mainly to agriculture, which is based on maize cultivation in both municipalities. While in Salto de Agua, maize cultivation is the key commercial and subsistence crop, Chilón farmers cultivate coffee as the main source of income. In both "ejidos", farmers organise around small activity groups for establishing agroforestry systems. In Chilón, farmers established maize associated to trees (Taungya rotational systems), improved fallows and shaded coffee systems; while farmers in Salto de Agua established taungya systems converted finally to silvopastoral systems. Improved fallow consists of enriching fallow lands with secondary vegetation with timber trees, so far as the latter are planted during the first five years of the fallow period. Taungya, in turn, consists of enriching maize cultivated plots with timber trees in a rotational pattern; and coffee systems were enriched with timber trees in association to other variety of previously existing native trees as shading cover. . Organic coffee cultivation translates into a higher carbon content in the upper soil layer (0-30 cm) than non-organic coffee systems, being able to store between 129.8 and 215.6 Mg C ha -1 in their components, including soil C (Soto- Pinto et al., 2010). However, the amount of aboveground biomass depends on the structure and composition of shade vegetation (Table 4). Economic analyses comparing conventional coffee with and without enrichment of timber trees and PES and organic coffee with timber trees and PES have shown positive benefit/cost ratios, as follows: 0.8, 1.2 and 1.8 for conventional management, conventional management enriched with timber trees, and organic coffee plus carbon sequestration and timber, respectively (

I m p r o v e d f a l l o w s h a v e d e m o n s t r a t e d t o b e an adequate alternative to slash-and-burn
agriculture, due to their capacity to increase biomass, productivity, economic value, complexity, carbon and diversity (Roncal 2007;Soto-Pinto et al., submitted). Tables 6 and 7 show structural and functional variables of these systems in the region. Results of these evaluations show that these systems may contribute to sedentarise the maize system, increase economic value and avoid deforestation while diversifying non-timber and timber related products.  Table 6. Structural variables of agroforestry systems in Selva Lacandona (Mexico).
In Salto de Agua and Chilón the decisions on land use are taken individually, often under the consensus of the family and the work group. The impact seems to be centered at the plot level. However the impact on the territory has not been evaluated.
In each community, regional and local technicians act as training guides. Local decisions are taken individually or by group. Regional and state-level assemblies of technicians discuss, analyse problems, and propose solutions, new projects and financial supports to resolve specific problems. Civil and academic organizations and government dependencies accompany the process offering punctual technical assistance to resolve specific questions and financial support. Academy plays an important role in knowledge management, offering training and developing human resources. All of the sectors are involved in thematic networks contributing to the building of a public policy in the thematic of forestry programs, PES and territorial developing, among other issues.

Discussion
In some places traditional communities have managed their resources sustainably for long time, even better than in many protected areas managed by the State, especially in Latin America (Bray et al., 2008;Porter-Bolland, et al., In Press). However, in the last years the effects of public policy, colonization process and urban development, among other land use change drivers led to high deforestation rates in sites which until four decades before were completely forested, this is the case of Selva Lacandona in Mexico where Marques de Comillas is a referent. Enriched shaded coffee, alternative rotational and silvopastoral systems have demonstrated benefits in the topics of food production, biodiversity and economy of livelihoods. Results demonstrate the value of agroforestry systems as a potential strategy for tree cover recovery and carbon sequestration (Haile et al., 2008;Nair et al., 2010;Soto-Pinto et al., 2010) and suggest that adequate planning, incentives and capacity-building efforts can lead to better conservation practices (Berkes, 2007). In Ejidos La Corona and communities in Chilón and Salto de Agua people have improved their local organisational activities, either in groups or collectively through preventing deforestation; intensifying the agriculture process; reforesting the deforested and open areas; controlling fire; acquiring new abilities; creating norms, sanctions, work plans, and social rearrangements; and reinforcing old capacities for developing a forest culture as a part of a new institutional development and good governance (Evans 2004;Corbera, 2005Corbera, , 2010. All this, coupled with the accompaniment of the civil society and academic institutions, and the involvement of government in a network of ecosystem services has been key in order to begin a governance development (Ruiz de Oña et al., 2011). However, the scaling up of this process is a challenge since it represents a greater organizational complexity and negotiation, a matter of governance (Swiderska et al., 2008).
In Marques de Comillas collective organization and decisions can impact broad and decisively on the territory in the short term, while in Chilón and Salto de Agua decisions taken by individuals or groups may be slower than collective decisions. Future studies on social resilience on both types of patterns should be of great importance. The lesson gained in this experience is that public programs must consider community priorities based on an integral regional vision, environmental education, strengthening of local capabilities and organisation. As a whole, public priorities may be a better appropriation and adaptation of programs in a new model of territorial development by considering institutional arrangements (Merino and Warnholtz, 2005). Farmers are the leading experts regarding their context and livelihood conditions and may or may not adopt and adapt programs that involve a territorial vision. The intense participation of communities may allow better monitoring, especially when these programs are integrated into their own communitymanagement plans (Franzel and Scherr, 2002). Territorial participatory planning of agroforestry and restoration systems, in addition to the institutional development have demonstrated their capacity to benefit environment, farmer's livelihoods and social capital, through their contribution to increase productivity, complexity, diversity, economic value, organization capacities and knowledge, which as a whole may contribute to avoid deforestation (Alburquerque, 2002;Bray, 2008;Swiderska et al., 2008). However, other products, service, process and management innovations are required to have a broader menu of options for farming systems and livelihoods to ensure the permanence of forest systems in a competitive context. Some of the needs to be resolved in order to conceive this process as a territorial development are: the best practices to achieve food self-sufficiency, development of market chains, education, training, infrastructure, financial support, policy arrangements and skills of competitiveness for production systems (Alburquerque, 2002;Ruiz de Oña et al., 2011). An integration of an agroecological matrix which enables agricultural production and natural resource conservation and a new political vision may help facilitate social and environmental synergies in rural areas (Perfecto and Vandermeer, 2010) Although the experience run by the project Scolel te' in Marqués de Comillas is not precisely REDD+, may offer a set of lessons learned about the conditions required for a good management for reducing deforestation. Some elements which need to be considered in order to contribute to the design of PES programs are the following: 1) technology adaptation, 2) ordering of land use with a territorial development vision, 3) broad participation of all stakeholders, 4) institutional development at local, regional levels and, 5) involvement in networks to launch processes of governance for the management of an adequate public policy of PES. Although these systems may contribute to avoiding deforestation, their potential to become part of REDD programs need to be better discussed since the relationship among markets and actors' rights and duties, in addition to the uncertainty regarding the factors influencing effectiveness on deforestation is unclear (Coad et al., 2008). In Chiapas, a process of institutional development around the issue of PES is emerging. In the last years groups working on ecosystem services (State Group of Ecosystem Services, GESE by its acronyms in Spanish) and deforestation and degradation (REDD) have matured in the state of Chiapas; representatives of the "Camara de Diputados y Senadores del Estado de Chiapas" (State House of Representatives and Senate) have approved the Law for Mitigation and Adaptation to Climate Change in the State; meanwhile, a group of research institutions and organizations of the civil society carry out an effort of monitoring, reporting and verification (MRV) in rural communities as a basis for a REDD planning in Chiapas. Moreover, the Government of the States of Chiapas and Campeche has recently signed an agreement with Acre Brazil and the Government of California through The Governors' Climate and Forests Task Force (GCF) for REDD + (http://tropicalforestgroup.blogspot.com/2010/11/text-of-cachiapas-acre-mou-on-redd.html) with the aim to "developing a common subnational REDD+ framework or patform for adoption and implementtion in GCF states and provinces" and "to building databases, developing options for linkage arrangements for getting financial support, technical assistance, capacity building, and advancing stakeholder involvement" (http://www.gcftaskforce.org). For its part, the Chiapas Government has allocated funds from the vehicle ownership tax payment to farmers focused on a Biodiversity Hotspot, the Selva Lacandona (Lacandon Community). This region has been widely supported by conservation projects, but from the point of view of other authors it lacks of organizational conditions for effective management of forest carbon (Castillo-Santiago et al 2009). At the national level, CONAFOR (Forestry National Comission) has convened a technical advisory board (Consejo Tecnico Consultivo CTC) aimed to "promote and deliver recommendations to the government institutions in order to influence the building of a functional mechanism for designing and implementing REDD+ in Mexico, guarantig the transference and maximization of environmental and social benefits" (http://www.reddmexico.org). However, groups of civil society organizations drew attention to the topic of indigenous peoples' rights. Hence, much effort is still needed to really get to have a shared vision. Forest alternatives will only be adopted by farmers if they respond to expressed needs of population, reduce risks, alleviate constraints, and increase productivity. Deforestation process seems to take its course if drivers persist (Cortina et al., in press). Farmers will be interested since they participate and design innovations from the beginning of the project and also if the initiatives constitute a good deal from the set of alternatives offered in the territory (Franzel & Scherr, 2002;Merino & Warnholtz, 2005;Bray, 2008). Mere academic exercises of land use ordering may be at risk of abandonment if local priorities are left out, so that the participation of farmers (men and women) in planning, designing and applying technology, norms, rules and sanctions among other local institutions must be established with broad participation of all of the actors involved.