Abstract

This study uses the urban arboreal census carried out in Santiago de Cali (Colombia) between 2014 and 2015, the data of the 2005 population census and the data of the ecological structure of the municipality of 2014. The purpose is to identify if there are spatially significant patterns that provide evidence of biases in the indicators of access to environmental services of urban trees (AU) and green spaces (EV) with socioeconomic indicators, in order to spatially identify inequities in access to these types of ecosystem services.

To achieve this, spatial regression models are used to capture grouping and dispersion phenomena in spatial patterns through the inclusion of terms of spatial autocorrelation in linear regression. The spatial models are tested with two types of matrices to observe the effect of the topology in the interaction between the variables and the results of the adjustment. Statistical graphs, thematic and LISA maps are used to identify areas with poor levels of access and discuss explanation related to the causes.

The results show that there are inequities explained by status variables such as access to higher education, which are also negatively correlated with the percentage of Afro-Colombians in an urban census sector (SU). In relation to access to green spaces (EV), there is no strong evidence that the variables of ethnicity, health or status are good predictors of access. However, there was a high concentration of the EV area available in very few sectors of the city.

Keywords: canopy cover, green spaces, environmental justice, spatial regression, environmental services, socioeconomic indicators.