Predictors of the distribution of street and backyard vegetation in Montreal, Canada.

摘要

Urban vegetation is shown to be unevenly distributed across cities and there is evidence of disparities in benefits provided by vegetation and of public health problems induced by urban heat islands. In order to improve vegetation cover, it remains crucial to understand the underpinning of such unevenness. In this paper, we investigate in Montreal (Canada) how the built environment, sociodemographic factors and administrative boroughs influence tree and lawn cover in public and residential land. The analysis was conducted at the dissemination area (DA) level, a Canadian census unit containing about 400-700 people. Six vegetation indicators were used as dependent variables: the proportion of a DA covered by trees/shrubs, lawn and total vegetation; the proportion of streets covered by trees/shrubs and the proportion of residential yards covered by trees/shrubs and total vegetation. Three sets of independent variables were studied: the built environment, sociodemographics and borough names. We used spatial autoregressive models to control for dependence and the spatial autoregressive term explained a large amount of variability in vegetation cover. The built-environment variables tend to have higher effects than the socio-demographic variables when predicting the three DA vegetation indicators, backyard vegetation, and to a lesser degree, street tree/shrub cover. In particular, population density is associated negatively to all indicators but positively to street tree cover. Socio-demographics are substantial in the explanation of the distribution of street trees, especially the presence of recent immigrants (negative effect) and of university degree holders (positive effect). These findings call for appropriate greening programs adapted to the local socio-demographic profile. The significance of boroughs also suggests the need for further research on the impact of within-city administrative hierarchies on the unevenness of urban vegetation.