In this class we learned about landscape metrics. Here, as with many applications of GIS, the fundamental issue is the problem of pattern and scale. Patterns and processes occur on very different scales of space, time, and ecological organization, so determining which scale to use when studying a particular phenomenon can be tricky. There is no single, natural scale that will encompass all of the different interactions and interrelations that should ideally be accounted for.
The role of geography in landscape ecology is particularly relevant because landscape ecology “is concerned with the reciprocal interactions between spatial pattern and ecological processes that occur on landscapes.” Landscape form influences processes, and processes influence form. The processes that influence spatial patterns and ecological form can range from climate, topography, natural disturbances, biotic factors, human land use, perspectives and more. For instance, topography influences forest fire behaviour, and forest fire behaviour influences the landscape through regeneration of stand cycles. Forest fire behaviour is also influenced by human land use change, forest management practices, as well as climate and biotic factors.
If the processes that influence a particular variable do not change over space (in intensity or interactions between variables, and without directional bias), they are considered stationary.
In landscape ecology, there is a fundamental assumption that things do vary across space, and that the location (and relative location) of things can have important consequences. This comes back to the concept of spatial autocorrelation; that nearer things are more similar than distant things. Apple trees are more likely to be sprouting nearer to other apple trees simply because of the way seeds are spread (and because of amenable environmental characteristics – i.e. spatial dependence).
Patterns develop when there is spatial autocorrelation, and we can study if these patterns develop in response to an environmental factor (a first-order process), or if they develop as a result of interactions between the phenomena themselves (second-order process). Landscape form and pattern varies across space; thus, they are heterogeneous and can be studied at multiple scales.
These patterns can be measured in a wide variety of ways, classified under metrics of composition, spatial configuration, fractals etc., but the best approach is to consider multiple variables, like we did in Lab 2.