Selecting independent measures to quantify Melbourne's urban–rural gradient

As cities and towns increase in population and size around the world, there is a growing interest in the impact of urbanisation on humans and the environment. The use of urban–rural gradients has proven to be a useful tool for studying changes in ecological patterns and processes across urbanising landscapes. Currently, there are a wide range of measures being used to represent changes in human demographic patterns, physical structures and landscape composition and structure along urban–rural gradients. The aim of this paper was to identify a suite of measures that can be used to define an urban–rural gradient in Melbourne, Australia. Using principal components analysis, we assessed 17 commonly used measures of urbanisation that included demographic variables, physical variables and landscape metrics. Four measures captured most of the variability in the patterns of urbanisation: (1) Indexcombined; (2) the ratio of people per unit urban land cover; (3) landscape shape index; (4) dominant land-cover. We used these four measures to quantify Melbourne's urban–rural gradient and then explored their use in representing urbanisation as an environmental space rather than a geographic space. This study provides an example of how to objectively select a subset of measures to quantify urbanisation, and illustrates a novel way of combining the measures to obtain a richer understanding of ecological responses to urbanisation.