“I came here for the land to sustain me”: land-use change and smallholder decisions in a tropical agro-forest frontier

Berget, Carolina


The Lacandona Rainforest, southern Mexico, is the largest remaining tropical rainforest in North America, a priority area for conservation and a deforestation hotspot. The region known as Marqués de Comillas located in the southern portion of the Lacandona Rainforest bordering Guatemala, was densely covered by forest and inhabited in low densities mainly by the Lacandón indigenous group. Nevertheless, colonization started in the 1970s, leading to a rapid forest conversion to mainly agriculture and cattle-ranching, giving way to the highly complex agro-forest frontier region it is today. The Marqués de Comillas region provides a living-laboratory for doing human-environment relations research given its recent history of settlement and landscape transformation.

The aim of this thesis was to assess the social-ecological factors that have shaped land-use change (historically) and smallholder land-use decisions (currently) in two neighboring villages in the Marqués de Comillas agro-forest frontier region. The studied villages, Loma and Chajul, share some similarities but also exhibit differences regarding their land-use types and composition. This thesis addressed the following general research questions: 1) what have been the historical social-ecological processes that have changed the original forested landscape in the study area, and which land-use types have emerged through time as a result?; 2) which local-household, regional-biophysical and national-institutional variables drive land-use decisions at the farm level in each village?; and 3) how do smallholder perceptions about what is most valued in their daily lives influence their land-use decisions? To address these questions this thesis was grounded on an empirical comparative case study. A mix-methods approach was used which consisted on qualitative (ethnographic work, oral histories, semi-structured interviews, farm visits, Photovoice method, among others) and quantitative (land-use change analysis using remote sensing, soils data and statistical modelling) methods.

In research Chapter 2, I reconstructed the land-use change history and assessed the resulting land-use types in the study area from 1976-2019. The results indicate that each village has its own unique land-use change and land-tenure history. As a result of its history, Chajul is almost three times larger (4,838 ha) than Loma (1,731 ha), and farm size can range between 30-150 ha/household in Chajul, while only around 20 ha/household in Loma. Loma is characterized by a dominating pasture landscape with some scattered agricultural and forest areas, while Chajul has large conserved forest areas intermixed with pastures, agriculture, oil palm and rubber plantations. It is concluded that the landscape mosaics of each village are the result of the complex historical events and intricate interactions among multiple social-ecological drivers at different spatial and temporal scales. The differential histories of each village have also had livelihood diversification implications.

In research Chapter 3, I determined the land-use composition across farms in the two villages, and evaluated the social-ecological drivers (at three scales: local, regional and national) that shape smallholder land-use decision making in the study area. The results show that at the farm level there were seven land-use types (agriculture, pasture, primary and secondary forest, rubber, oil palm and reforestation). All seven land-use types were present in Chajul, while in Loma all were present except for rubber and oil palm. The two communities significantly differ in regards to primary forest, where Chajul farms have significantly more primary forest than Loma farms. The generalized linear models indicate that primary forest was affected by drivers at all three scales analyzed, pasture and oil palm by two, agriculture and reforestation by one, and secondary forest and rubber by none. The qualitative analyses showed that some national-level subsidy programs influenced decisions on allocating the land-use rubber, oil palm and reforestation. However, clear relationships between certain land-use policies and their targeted land-uses (i.e., agriculture, pasture, primary and secondary forest) were difficult to establish. It is concluded that several social-ecological drivers at different scales shape farmers’ land-use decisions at the farm level, and that the historical context of the villages is very important to understand smallholder land-use decisions in the present.

In research Chapter 4, I explored the role of smallholder perceptions in shaping land-use decisions in the study area. The results indicate that the perceptions about what farmers value most in their everyday lives (i.e., the farm and household – including family and food) are reflected on their farm land-use decisions. Similar perceptions sometimes result in similar farm land-use compositions (agriculture and reforestation). Nevertheless, similar perceptions do not always result in similar land-use compositions (more proportion of pasture in Loma, and more forest in Chajul), because alongside perceptions other social-ecological factors, e.g., farm size, also play important roles in smallholder land-use decisions. Furthermore, positive or negative perceptions about oil palm, are highly influential on farmers’ decisions on planting or not this crop. In contrast, having a positive perception about rubber, does not always yield on decisions of planting it: some farmers prioritize land-uses with shorter-term investments and immediate cash returns (cattle-ranching), while others opt for the longer-term investment (rubber). It is concluded that perceptions influence smallholder land-use decisions, yet in some cases, other factors have to be considered also. Hence, in land-use change analysis and land-use policy formulation and design, smallholder perceptions should be taken into account as part of the set of social-ecological factors driving local land-use decisions.

The empirical results from this thesis illustrate how the landscape of the study area is a good example of a social-ecological system where multiple human-environment factors and processes at different spatial and time scales, have played a role in landscape transformation in the last 40 years. The differential historical and current human-environment interactions in the study site have shaped the original forested landscape into a highly heterogeneous and dynamic agro-forest frontier characterized by a mosaic of land-uses in each studied village. Understanding how land-use change unfolds and how smallholder land-use decisions are made has implications for the formulation and (co)design of more effective, flexible, tailored and contextual-based agricultural development and conservation policies. Such policies should not only aim to achieve specific land-use outcomes, but simultaneously aim to improve farmers’ livelihoods by addressing their history, social and ecological context, perceptions, values, motivations, needs and self-determination.