Land use systems and livelihoods in demographically heterogeneous frontier stages in the Amazon

Publicado en Environmental Development, v. 38:100587

Barbieri, A.F., Guedes, G.R., Santos, R.O.

Año de publicación 2021

Research Group on Population, Economics and the Environment (EPOPEA), Center for Regional Development and Planning (Cedeplar), Universidade Federal de Minas Gerais (UFMG), Brazil
Researcher, Research Group on Population, Economics and the Environment (EPOPEA), Universidade Federal de Minas Gerais (UFMG), Brazil



Funding for this research was provided by the follwoing sources: i) Inter-American Institute (IAI), Project CRNIII3036 &ldquoLUCIA &ndash Land Use, Climate and Infections in Western Amazonia&rdquo (CRNIII3036). IAI is an intergovernmental organization supported by 19 countries in the Americas, with the main office located in Uruguay ii) Minas Gerais Research Foundation &ndash FAPEMIG, Brazil, grant APQ-01-553-16 iii) National Research Council &ndash CNPq, Brazil, grants 447688/2014-6, 431872/2016-3, and 314392/2018-1.

Proyecto CRN3036


A unique 25-year panel dataset depicts the coevolution of demographics, livelihoods, and land use in the Amazon.
· Cohort analysis unveils the effects of household life cycles on land use choices over stages of frontier development.
· Land use systems are independent of household demographic dynamics, especially at later stages of frontier development.
· Farm livelihood diversification prevents farm households from becoming trapped in a long-term deprivation trajectory.


We investigate how Amazonian smallholders' land use systems coevolve with household-level demographic factors associated with changing livelihood strategies over the different stages of frontier development. Few micro-level studies have investigated this association, particularly due to the paucity of longitudinal data on cohorts of farm colonist households and plots. Cohort analysis is the only way to depict how the structural conditions affecting individual and household livelihood decisions differ from earlier to later stages of frontier development. Our methodological approach involves a unique dataset, based on a micro-level panel of farm households depicting 25 years of settlement in the municipality of Machadinho, in the Brazilian Amazon. We use descriptive statistics with paired t-tests, land use classification analysis, latent transition analysis, and longitudinal multinomial regressions to understand which cohorts of households thrived or failed and, most importantly, why and when. Splitting the data into panels of settlement cohorts helped us understand the effect of demographic life cycle markers on land use choices over the different stages of frontier development and the ability of farm households to adapt their livelihoods at the frontier over time. We found that, as the colonization frontier integrated into markets, the most successful original settlers were those who diversified their portfolio of capitals and livelihood strategies as a response to new local and regional market conditions. We also found a progressive change from land use systems based on subsistence agriculture to diversified land use systems that combine on- and off-farm activities. Livelihood diversification is key to preventing households from becoming trapped in a long-term deprivation trajectory, particularly when the frontier becomes more urban and market-oriented. This explains why land use has become progressively independent of household demographic dynamics in advanced stages. We contend that, as frontiers integrate into markets, diversification should not only be incentivized, but should also be used as a technical strategy to enhance access to subsidized rural credit, as it seems to increase farmers&rsquo likelihood to thrive and improve their resilience against shocks.