Decades ago, Brazil’s northeastern State of Bahia produced much of the world’s cocoa for chocolate. Most farms grew their cacao trees interspersed with other native trees, in dense agricultural forests. Children played at the forest edges.
Today those children are grown; some have inherited the farmland they once played on. Economic conditions have changed and cattle farming is now more profitable than cocoa, but the farmers’ memories of their past on the land, and their visions for the future, continue to influence how they steward their properties, finds a recent paper in Global Environmental Change. The percentage of forest that Bahia’s farmers keep, as opposed to clearing for grazing pasture, is directly associated with their memories of and aspirations for the region, according to the study.
“We want to know what drives people,” says Kira Sullivan-Wiley, a geographer who coauthored the new work while a postdoc jointly appointed through the Nature Conservancy and Brown University in Providence, RI. She and coauthor Amy Teller, a sociology PhD candidate at Brown at the time, set out to create a new, interdisciplinary approach to understand why people make the choices they do when managing their land.
Many studies of land use today don’t actually entail conversations with farmers. Satellite imagery is so high resolution that new studies are largely based on remote sensing, notes rural sociologist Thomas Rudel, now emeritus at Rutgers University in New Brunswick, NJ. While remote sensing can identify where and how agriculture is changing forests, it can’t say why people are clear-cutting some plots or allowing others to go fallow. Any work that clarifies motivations therefore “has potential to make a considerable contribution,” Rudel says.
Governments and nonprofit organizations do use simple questionnaires to gather stakeholder perspectives, before, for example, launching a new conservation campaign or policy, Sullivan-Wiley adds. These questionnaires typically pose simple socioeconomic questions to local residents, such as age, education level, and property acreage. But such queries can overlook decisions based on larger time and spatial scales than one farm in one year, she says, such as childhood memories, dreams for the future, or regional concerns.
To tease out nuance, Sullivan-Wiley and Teller traveled to Bahia, where they interviewed 49 farmers using a standardized set of questions. The interviews included the typical socioeconomic queries as well as more open-ended questions about each farmer’s perceptions of landscape change, their sense of the relationship between agriculture and environmental health, and their personal memories and hopes for the region, Teller explains. Then the researchers created a spreadsheet with each of the 49 farmers occupying one row, and a series of themes, such as optimism for the future or belief in a relationship between trees and water, for each column. They noted when an interviewee mentioned a theme.
Finally, the researchers ran a principal component analysis to identify which groups of themes were most likely to appear together in a single interview. Broadly, they found that most farmers fell into one of two groups: a cohort that was optimistic about the future and forest conservation, and a group that was pessimistic about future economic opportunities. When the researchers then ran a regression analysis to compare the groups of themes to the percentage of land that each farmer currently kept in forest versus pasture, they found that the optimistic group was significantly more likely have more forested land. The pessimistic group was more likely to have more of their land in pasture than in other uses, Teller explains. Overall, about 30% of the variability in land use was associated with narratives about the future, according to the regression analysis. Hence, belief in a given narrative had a slight-to-medium effect on how much pasture a farmer had on their land, Sullivan-Wiley explains.
Whether these methods could be useful in other contexts remains to be seen. “We put the paper out actively encouraging other people to apply it in places we don’t have access to,” Sullivan-Wiley says. Rudel, though, cautions that he would want a larger sample size, as well as interviews conducted by local people rather than outsiders, in order to trust any future results. Sullivan-Wiley is now working on a proposal to use the methodology to interview fishermen and private landholders in Massachusetts. Essentially, the goal is a more nuanced view of farmers as agents of the land’s future, she says. “They have dreams and visions,” she says, “and those need to be acknowledged and respected.”