Climate change and the sustainability of agricultural productivity in Brazil

• Ano de defesa: 2015
• Autor(a) principal: Pires, Gabrielle Ferreira
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal de Viçosa
• Palavras-chave em Português: Meteorologia agrícola; Climatologia agrícola; Desmatamento; Mudanças climáticas; Produtividade agrícola; Sustentabilidade; Agronomia
• Link de acesso: http://www.locus.ufv.br/handle/123456789/7491

There is a wide global expectation that Brazilian total agricultural output will increase like no other country in the world to meet the projected higher demand for food until 2050. While trying to meet this expectation, Brazil will face the effects of a severe climate change induced by the change in atmospheric composition. In addition, if the future increase in total production resembles the dynamics of the past and increasingly deforest natural biomes as the Amazon and the Cerrado, we run a great risk, as recent studies indicate that large-scale deforestation drives significant changes in water availability and could have implication for agricultural systems. This thesis investigates how climate change and additional deforestation may affect the productivity of the main commodities exported by the country until 2050: soybeans and cattle pasture. We used a gridded crop model to assess the effects of the climate simulated by four CMIP5 models under the IPCC AR5 RCP8.5 scenario on soybean and pasture productivity. We contrasted these results with a second group of simulations that account for the effects of more severe Amazon and Cerrado deforestation scenarios on regional climate. Soybean simulations show that, for central-northern Brazilian productive regions, the effects of climate change are dependent on the planting dates. The productivity of soybean cultivars planted in late September, sowed early by farmers who choose to adopt double-cropping systems (two crops on the same land in the same agricultural calendar) is predicted to expressively decrease. However, soybean cultivars that are planted in later dates (November- December), mainly sowed by farmers who choose to grow only one crop in the agricultural calendar, show increased productivity. The decrease in productivity for earlier dates is related to a sharper decreasing trend in precipitation during these months of the year, while the increased productivity in later dates is due to a smaller water deficit and the positive effects of an increased atmospheric CO 2 concentration. Southern Brazilian productive regions also show increased soybean productivity until the middle of the century, despite the planting date. For central-northern Brazilian productive regions, moving planting dates from September to later dates expressively increases soybean productivity, but decreases the probability of adopting double-cropping systems. In addition, increased levels of deforestation lead to increased soybean productivity loss. Pasture simulations show that, as well as in the case of soybeans, pasture productivity is predicted to decrease in central-northern Brazilian regions and slightly increase in southern regions. In addition, higher deforestation levels causes further productivity decrease, and lead to at least twice as large productivity losses. According to all simulations in this work, the regions most affected are either the major Brazilian production region (Mato Grosso) or where the exploration has begun more recently and still hold an expressive agriculture potential as MATOPIBA, indicating that government investments in these regions without the proper consideration of the climate risks are a high-risk strategy. Finally, in the face of climate change and with little evidence that deforestation in Amazonia and Cerrado is ending, Brazil needs to review its agriculture and conservation policies and immediately shift to a new standard of zero deforestation in Amazonia and Cerrado, and create mechanisms to identify and trace solutions to adapt its agriculture to climate change.