There are a variety of ways of quantifying the structural and functional connectivity of a landscape. Structural connectivity is usually quantified from aerial photographs, maps, or remote sensing data (geographic information systems (GIS) data, satellite imagery). Many metrics require that an image is rasterized by overlaying a grid of cells, with each cell having a defined size or grain. The impression of structural connectivity may vary with the grain size that is chosen, with coarser grains making it more likely that small gaps between habitat areas will be overlooked because large cells average across the gaps. Similarly, organisms may have a particular spatial scale at which they sample the environment and make decisions about movement. The grain of our sampling a landscape to measure connectivity should be sufficiently fine that it is congruent with the scale selected by the study organism. However, this is often not known prior to commencing a connectivity analysis and therefore analyses at multiple scales are of value to identify the grains that have the highest ability to describe the movement or occupancy patterns of a species across a landscape.
There is a distinction between connectivity measured between pairs or landscape elements, or other portions of a landscape, and connectivity across an entire landscape. Often movement is measured between pairs of points, or along particular paths (e.g., for radio-tracked animals), and this data is then analyzed in relation to habitat type or the occurrence of particular landscape elements (rivers, roads, etc.). A typical approach is to start with pairwise measures of connectivity between two points and then repeat such measurements across a landscape. However, there are no general guidelines available to set the spatial scale at which connectivity should be measured.
It is useful to further consider functional connectivity, which can either be based on potential or actual measurements. Potential connectivity comes from combining structural connectivity with information about the movement behavior, distance, and costs in the organism in question. Actual connectivity is that which is actually measured. Be it potential or actual, functional connectivity might be based on strict adjacency (touching) of habitat, a threshold maximum dispersal distance, a decreasing function of distance which reflects that movement frequency (or potential) is inversely related to distance, or a resistance-weighted distance function that incorporates the different costs of traversing various paths.
A useful framework for classifying connectivity metrics was presented by Justin Calabreese and Bill Fagan in 2004, and this framework is used here to organize connectivity measures. Furthermore, the description of the advantages and disadvantages of different metrics, as well as the data requirements to estimate them, draw heavily on their work. Six classes of metrics can be distinguished.
Was this article helpful?