Further Study
Despite spending the majority of a semester on this project, it seems I have more questions than answers. In particular, three ideas emerge as potential places for aditional research (some of which I include in the Policy Report): Indigenous lands, the use of remote sensing technologies, and the effects of other biomes on HDI. Additionally, this study may benefit from using a different scale.
Indigenous Lands
New research is demonstrating that forest ecosystems under the stewardship of indigenous peoples fare better. Even so, recognition of indigenous rights in Brazil is an ongoing issue (Cannon, 2021). Moreover, there is precedent that indigenous, local knowledge can be a valuable way of providing ecosystem services (Delgado-Aguilar et al., 2019; Saarikoski et al., 2017). Additionally, the growing movement for participatory GIS and participatory mapmaking has opened up more opportunities for indigenous peoples to be involved in ecological and economic policies (Sletto et al, 2020).
As applied to this research, a variable could be used to consider how much of a given state or municipality’s land is under indigenous stewardship, or at least recognized as indigenous. As a corrollary, there could also be a variable for the area desginated by the Brazilian government as the “Legal Amazon,” on the grounds that the legal status of large stretches of forest may play a role in explaining the patterns seen.
Remote Sensing
Other studies of land use and forest degredation have considered remote sensing (Delgado-Aguilar et al, 2019; Sánchez-Azofeifa et al., 2017). Not only does this allow for a more comprehensive dataset that relies on measurements from satellite images rather than calculated estimates from official sources, but the data can have a better temporal resolution. For the purposes of my study, it seems that remote sensing data would allow for a more robust study of the patterns of forest loss and urban growth beyond numbers in a spreadsheet.
Other Biomes
For the sake of scope, my study only relied on forests from the Amazon and the Mata Atlântica (Atlantic Forest) in Brazil. However, other biomes – for example, the mangrove forests along the coast and the caatinga (high mountain savannahs) are also experiencing ecological canges. These biomes could be considered as addtional variables, especially for provinces that are further from the primary forests I considered. Of note, a new report has demonstrated that seagrasses and mangroves are actually a more efficient carbon sink than forests (The Economist, 2021). As with other research here, there is a potential for policy changes based on the relationships explored.
As a corrollary, there is also a question of considering the forests I did study, but using a different variable – such as accumulated hectares forest loss or (requiring more data preparation) a percentage of forest loss (rather than forest cover). Future regressions will consdider this, at least when testing for the best possible model.
A Question of Scale
Every geographer must consider questions of scale. For this project, I used the 26 states of Brazil, plus the Federal District (analogous to Washington, DC in the US). However, each state subdivides into municipalities (analogous to counties). Especially if other ecological factors are considered, analyzing some of the data at a finer scale may demonstrate something in the data not yet visible.
Next: Policy Report