Cattle, Culture, and Feminist Ecologies in the Brazilian Amazon: Advances in Theoretical and AI-Driven Land System Science

Abstract

Land use and land use change contribute to global environmental degradation and have an impact on climate change and biodiversity loss. Although substantial progress has been made in advancing scientific theories and providing evidence-based policy recommendations, the problems of deforestation and degradation have not yet been effectively halted. This is also true for the Amazon biome, the planet’s largest tropical forest. Despite scientific consensus regarding the biome’s importance for planetary biodiversity, carbon storage, and water cycles, the Amazon continues to face numerous enduring social and environmental challenges, many of which are still not fully understood. Foremost among these understudied relationships is the role that cultural systems play in shaping land use. To address this, a systematic review approach was used, and 66 studies were examined that causally link culture and land use. Most studies focused on shared attributes such as norms, values, and practices that influence different land systems around the globe. These can lead groups to maintain a particular land use, which is coded into cultural systems, resulting in positive outcomes for the land system's resilience. Changes in economic, political or environmental factors (structural factors) or internal group events can lead to changes in norms and values that alter land use patterns or destabilize systems, leading to new system dynamics. If structural (environmental) factors change, but norms and values remain constant, this can lead to a resistance to new feedback mechanisms, which is characterized as cultural 'lock-ins'. The cattle sector in the Brazilian Amazon is described as such a ‘locked-in’ system. While there is potential for sustainable intensification, system dynamics such as deforestation, pasture degradation, and high greenhouse gas emissions are difficult to overcome. In addition to the cattle culture of extensive and low-input cattle ranching, other contributing factors include land speculation processes and low access to input credits and governmental assistance. Yet, understanding the spatial distribution of cattle remains limited. Using innovative convolutional neural network-based density estimation models on high-resolution satellite imagery, I estimate, for the first time, cattle stocking rates (heads of cattle per hectare of pasture) on 2837 properties across 44 municipalities and four states in the Brazilian Amazon. On properties where cattle were detected, I identified both recent deforestation and pasture degradation. Integrated crop and livestock systems are very rarely observed in the sample. Using a state- fixed-effects regression model, I found that stocking rates tend to be lower on land with recently deforested areas. No positive correlation between private conservation policy (zero-deforestation commitments) or between a public intensification policy (subsidized rural credits) and stocking rates is identified, but the interaction of the two policies is positively correlated with stocking rates. This suggests that a mix of private and public policies is best suited to support sustainable intensification. Cattle ranching holds cultural significance in the Brazilian Amazon; but until now, the perspectives of marginalized groups, including women and People of Color in the cattle sector, have received limited exploration. Drawing upon data gathered from 51 semi-structured interviews and a survey with 785 cattle ranchers in the state of Acre, the cultural-land use framework from Chapter 2 was extended with feminist political ecology theory and the gendered knowledge and value systems within the cattle sector was assessed. Conditions for women concerning land rights improved in the last decades. However, this analysis revealed that women and People of Color are not profiting to the same extent from cattle ranching compared more dominant groups. I found that all the groups had a similar vision for the future of the region, mainly focused on cattle farming. The interviews also revealed that invisible tasks, including subsistence farming and childcare, are linked to gender-specific pro-social and pro- environmental value systems, which could be a lever to scale up conservation and reforestation activities and break the current ‘lock-in’. This thesis undertakes a comprehensive synthesis and analysis of available evidence concerning the interplay between culture and land use, marking a pioneering effort in this area. Subsequently, it applies these insights to examine the Brazilian Amazonian cattle sector. It illustrates the cultural importance of cattle farming, also for marginalized communities. Augmenting this analysis, the thesis incorporates novel empirical data on the spatial distribution and scale of cattle production. This integrated approach aims to offer new insights into the dynamics of cattle production within this crucial ecosystem. This thesis demonstrates that to address the pressing challenges of biodiversity and carbon loss presented by land use change, agricultural and environmental policies should be both comprehensive (i.e., positively reinforcing instead of antagonistic) and gender sensitive (i.e., including groups with pro-social and pro-environmental values). Such a policy pathway has the potential to create a sustainable and equitable future for the Amazon and other regions affected by land use change.