Correlação espacial entre renda per capita, área construída e cobertura florestal urbana em Recife - PE.

Abstract

Since cities are complex socio-ecological systems, it is highly important to study the interrelationship between socio-economic and natural indicators within the urban environment. Studies that relate urban forest cover to income have found a positive relationship, but most are concentrated in cities in developed countries. On the other hand, underdeveloped countries tend to present an irregular pattern of land occupation and social inequalities. In this way, this work aims to verify if there is a spatial correlation between per capita income, forest cover and built-up area in the city of Recife, and if there is, to analyze whether this correlation is positive or negative, in order to provide subsidies for a fairer urban environmental planning from a socio-environmental point of view. The ground cover was made by the Recife Secretariat for the Environment and Sustainability, using an orthomosaic from 2013. The neighborhoods were used as a basic spatial unit and for income data, the gross domestic product (GDP) per capita, in reais. The following analyzes were performed: Univariate Global Spatial Autocorrelation, Univariate Local Spatial Autocorrelation, Bivariate Global Spatial Correlation and Bivariate Local Spatial Correlation. The variables tested were the percentage of forest cover, percentage of constructed area and value in reais of per capita income per neighborhood. The city of Recife presented a coverage of 1% of agricultural and aquaculture activities, 4.5% of water, 2.6% without vegetation cover, 2.5% of wetland, 49.6% of constructed area, 37.6 % Forest Cover and 2.1% Herbaceous Vegetation. The Moran Index for univariate global spatial autocorrelation was 0.339, 0.476 and 0.243 for built area, forest cover and income, respectively. For the univariate local spatial autocorrelation, a significant cluster of HH Built Area in the north and central-north regions was observed, representing 31% of the neighborhoods of Recife, a cluster LL (39%) formed by neighborhoods with very low forest cover, and for Income per capita, there is a pattern of concentration in the central-north area of the city (HH cluster), surrounded by LH clusters and existence of LL clusters in the northwest, southwest and south areas (peripheries). Moran's indices for the bivariate global spatial correlation were: Forest Cover x Income (-0.119); Forest cover x Constructed Area (-0.334); Income x Built area (0.100). For the bivariate local spatial correlation, significant clusters were found for: Forest cover x Income (28% LH cluster in the central-north region of the city); Forest cover x Built area (29% LH cluster, in the central-north and north regions); Income x Built Area (22% LH cluster in the northern region). Overall, there was a negative correlation between forest and income, forest and built-up area and a positive correlation between income and built-up area, reflecting a spatial pattern that favors low-income populations in terms of proximity to forest cover, but this favoring is mainly due to the existence of remnants forests that along the historical process of urbanization of the city were being moved to the peripheral regions.