The urban food environment is a target for policy interventions aimed at improving diet. This study developed a workflow in the R programming language for characterising the urban food environment using standard measures of accessibility to the food environment along with methods adapted from community ecology, applying it to Perth, Western Australia.
The workflow uses the Open Route Service API to calculate 15-minute walking network buffers from the centroid of all Census Mesh Blocks in the Perth Metropolitan Region and then calculates the types and abundance of food outlets available within each network buffer. A Bray-Curtis dissimilarity matrix quantifies how similar each Mesh Block is to all other Mesh Blocks in terms of the types and abundance of food available within a 15-minute walk of the centroid, enabling clustering of the Mesh Blocks into groups according to how similar their food environment is. The workflow then calculates the total abundance, richness, and Simpson’s diversity of food outlets within each network buffer, along with standard measures of accessibility recommended in scientific literature and urban planning guidelines to produce a comprehensive description of the Perth urban food environment.
Results show that the Mesh Blocks in the Perth metropolitan region can be clustered into five groups based on the type and abundance of food outlets, with food availability and diversity decreasing from the centre of the metropolitan region towards the periphery. The workflow developed provides a more nuanced picture of the food environment compared to standard measures of accessibility alone and can quickly and easily calculate a comprehensive set of standardised indicators of the food environment in any location, thus enhancing replicability.