Banca de DEFESA: Bianca Ferraz Rebelatto

Uma banca de DEFESA de MESTRADO foi cadastrada pelo programa.
DISCENTE : Bianca Ferraz Rebelatto
DATA : 31/08/2023
HORA: 13:00
LOCAL: Remoto
TÍTULO:

Relationship between forests and agriculture: how does native landscape vegetation influence agricultural productivity?

PALAVRAS-CHAVES:

Soybean productivity, landscape, agricultural productivity models, ecosystem services

PÁGINAS: 39
GRANDE ÁREA: Ciências Biológicas
ÁREA: Ecologia
SUBÁREA: Ecologia Aplicada
RESUMO:

The large-scale removal of native vegetation in the tropics directly impacts various ecosystem services produced by ecosystems, such as the production of regular rainfall and climate balance. Thus, the advance of deforestation and its effects on the climate can potentially reduce the areas favorable to agricultural cultivation in Brazil and, therefore, the final productivity of crops. The main objectives of this work were: a) to create a model of soybean agricultural productivity based on climatic and reflectance variables of cultivated fields for the region that includes MATOPIBA, Mato Grosso and Goiás; b) evaluate the relationship between native vegetation in the landscape and soybean productivity for the same region. To create the agricultural productivity model, we used mixed linear models (LMM) with soybean productivity data from 63 rural properties (2009/10 and 2016/17 harvests), considering as predictors multiple climate variables and the improved vegetation index (EVI). Then, to validate the predicted productivity map, we used IBGE municipal agricultural productivity data. To evaluate the contribution of native vegetation to soybean productivity, we used the predicted soybean productivity of 100 points randomized 20 times in the region as a function of the proportion of native vegetation in the landscape in a buffer of 50 km, where we verified the point maximum proportion of forest through a segmented regression. The best soybean productivity model included as predictors the Improved Vegetation Index (EVI), the Daytime Land Surface Temperature (DLST), the Vapor Pressure Deficit (VPD) and the Number of Dry Days (NDD). The productivity model presented a coefficient of determination of 50%, with greater importance of the EVI for drier and hotter years, and with a reduction in soybean yield with the increase of DLST, VPD and NDD. The proportion of vegetation was positively correlated with soybean yield in all years of the time series, with a determination coefficient of up to 50%. Segmented regression showed a maximum breakpoint of 26% of native vegetation, indicating a strong reduction in productivity in landscapes with smaller vegetation cover. The regression slope in the second segment was positive for most of the evaluated years, indicating that the increase in the amount of native vegetation in the landscape results in greater productivity. Our results indicate that the forest contributes to the productivity of the soybean crop due to the benefits of local ecosystem services. The results of this work will allow the elaboration of regional productivity maps and the promotion of sustainable environmental practices for food production that protect the multiple ecosystem services that the forest provides.

MEMBROS DA BANCA:
Presidente - 996.005.011-49 - DIVINO VICENTE SILVÉRIO - UnB
Interno - PAJ042729 - IMMACULADA OLIVERAS MENOR - York
Externo à Instituição - ANDREA SANTOS GARCIA - USP