ABSTRACT
Landscape connectivity has a long history, with three distinct origins. Merriam connectivity (1984) is the interaction between a species’ movement attributes and the landscape structure, which influences movement rate between habitat patches and thereby, population persistence. Noss connectivity (1987) is the extent to which patches are connected to one another by similar habitat or corridors and largely ignores species’ movement attributes. Hanski connectivity (1994) is the opposite of patch isolation: patches far from others are assumed to have a low rate of immigration and therefore, low connectivity. We review the assumptions and limitations of all three connectivity concepts. Noss connectivity is most frequently applied in landscape planning, via tools such as least cost path analysis, graph theory, and circuit theory. However, the Noss connectivity assumption, that habitat = connectivity, is not well supported by movement data. For example, many species move preferentially through the matrix (non-habitat). Where connectivity is used for land management decisions, researchers and practitioners should be aware of the assumptions and limitations of the particular concept of connectivity they are using, and they should provide evidence that the selected concept and approach are actually appropriate in their situation.
