Dear all,
Last week I presented a set of three new GRASS-AddOns for analyzing landscape connectivity based on graph-theory at the European Congress of Conservation Biology (ECCB, http://eccb2012.org/) in Glasgow . This set of tools applies methods documented (a.o.) in the following literature to the input data you might have.
Bunn, A. G., Urban, D. L. & Keitt, T. H. 2000: Landscape connectivity: A conservation application of graph theory. Journal of Environmental Management (2000) 59: 265-278. http://www.sciencedirect.com/science/article/pii/S0301479700903736
Calabrese, J. M. & Fagan, W. F. 2004: A comparison-shopper’s guide to connectivity metrics. Front Ecol Environ 2 (10): 529-536 . http://dx.doi.org/10.1890/1540-9295(2004)002[0529:ACGTCM]2.0.CO;2
Minor, E. S. & Urban, D. L. 2008: A Graph-Theory Framework for Evaluating Landscape Connectivity and Conservation Planning. Conservation Biology 22 (2): 297-307. http://www.uic.edu/labs/minor/minor&urban2008.pdf
Zetterberg, A., Mörtberg, U. M. & Balfors, B. 2010: Making graph theory operational for landscape ecological assessments, planning, and design. Landscape and Urban Planning (2010) 95: 181-191. http://www.sciencedirect.com/science/article/pii/S0169204610000204
Ranius, T. & Roberge, J.-M. 2011: Effects of intensified forestry on the landscape-scale extinction risk of dead wood dependent species. Biodiversity and Conservation 20 (13): 2867-2882. http://www.springerlink.com/content/ch9qtv2665h624q4
The input data the tools take are:
-
A polygon vector map containing habitat patches (of your species / habitat type in question) and a column for a population proxy (e.g. patch area). (required)
-
A cost raster map (for more information on that see the r.cost-manual) (optional; if no cost raster is provided Euclidean distance is used).
-
Dispersal information
o A maximum (cost) distance threshold for assumed connectivity
o A dispersal kernel (defined by three variables of a negative exponential decay function (see eg. Bunn et al. 2000)
The outputs are connectivity measures on vertex (=patch) level, edge (= the connection between two patches) level, and graph (the entire network) level. With r.connectivity.corridors you can also identify corridors for the connections (edges) you select, and you can weight their importance based on a column of your choice (e.g. connectivity measures from r.connectivity.network).
The tools (and documentation) can be found in the GRASS AddOns SVN repository (see: r.connectivity.distance, r.connectivity.network, r.connectivity.corridors in the http://grass.osgeo.org/wiki/GRASS_AddOns/). They are UNIX-shell (and one R-) scripts.
Required additional software is: Cran R with at least the packages “igraph” (version 0.6-2) and “nlme” installed. If you want to make use of parallel processing / multi-threading you need also the packages “doMC”, “foreach”, “iterators”, “codetools”, and “multicore”. But multithreading is only supported on LINUX. The tools have not been tested on Windows yet, but this will be done in near future.
Any kind of feedback would be very much appreciated!
Cheers
Stefan
P.S.: What do you think, how helpful would an example based on the spearfish dataset be? If you think so, I will add that to the documentation, as soon as the projects which triggered the development of the r.connectivity.*-tools are published.