topic Re: Habitat Suitability Models? in Data Management Questions

Habitat Suitability Models?

MarkEnglish — Thu, 04 Nov 2010 20:24:00 GMT

Hello:
I am in the preliminary stages of investigating habitat suitability models. My approach is to utilize various geospatial data such as soil type, elevation, vegetation, climate, urban areas, infrastructure, etc., and assign weighting variables associated with each targeted habitat species. Using red squirrels for example: What they forage and where they den (vegetation), their elevation range, their prefered climate, etc. This details of this information will be derived from local biologist. My first instinct is to review other projects similar to this, because it seems this work has been done. My question is, does anyone know of existing models, an appropriate place to begin, or helpful resource? Thanks so much for your time!

Re: Habitat Suitability Models?

DonovanCameron — Thu, 04 Nov 2010 22:04:06 GMT

It wouldn't be a bad idea to determine first what approach you will be using... ie, fuzzy vs static

Look at some of the documentation from Patch Analyst, it uses a fuzzy approach. Take note of the way it clusters, using hexagons, there is a specific reason for this as opposed to using squares/rectangles to define habitat patches.

http://flash.lakeheadu.ca/~rrempel/patch/
http://flash.lakeheadu.ca/~rrempel/patch/overview.htm
http://flash.lakeheadu.ca/~rrempel/ecology/papers/patchanalyst.pdf

FRAGSTATS was also used in developing Patch Analyst.

http://www.umass.edu/landeco/research/fragstats/fragstats.html

There are some new fuzzy tools with ArcGIS 10, just do a search in the help for "Fuzzy" and you should get some results for the tools that now exist.

http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#/Fuzzy_Membership/009z000000rn000000/
http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#/Fuzzy_Overlay/009z000000rp000000/

" rel="nofollow" target="_blank">http://www.youtube.com/watch?v=Hd13H0XO0LU[/video]

Re: Habitat Suitability Models?

MarkEnglish — Mon, 08 Nov 2010 21:31:57 GMT

Thanks Donavan:
I am interested in utilizing the new fuzzy tolerance tool, although will most likely need a static version as well for further statistical applications.
I believe this stage of analysis is a little far out from where I need to initially focus. I'm currently just looking for any basic habitat suitablity models first, and how to determine the weighted values through out all of the parameter datasets (climate, vegetation).
Thanks for your quick reply,
Mark

Re: Habitat Suitability Models?

DonovanCameron — Tue, 09 Nov 2010 13:56:38 GMT

When weighting those variables, it is relational to where exactly the study is taking place.

This might be something to discuss with any of the biologists you are consulting with; asking why one would have more or less weight that the other.

Once you have your rasters reclassified/coded to the appropriate values, ie, elevation ranges with codes showing level of influence in high vs low, all of these rasters can be thrown into the Weighted Overlay tool, this was used often before the new fuzzy statistics tools came in v10.

http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#//009z000000rq000000.htm

It is best to try several variations of weighting to ensure you get the results you are expecting, like making sure all areas of exclusion are actually excluded, areas of high importance are all included, etc... Try to find the best balance.

*EDIT: There is also the option of doing the weighted overlay using Map Algebra, which I prefer, its interesting to compare the results of a Map Algebra statement to the weighted overlay tool when using the same weighting preferences in each.