I've seen coots on patches of habitat smaller than 30 x 30m in China and U.K. That's all I can really offer on this, except that habitat quality is surely more important than size. One patch in particular had dense reeds over about half of the patch, open water over much of the remainder, and rich diversity of aquatic and terrestrial plants at the edges of the pond (i.e. good habitat quality). The site included a mosaic of patches, many of which were also suitable, and there were more coots in those patches. Sorry that this is just an observation that has no quantitative value. Pattern in the patch mosaic is probably important too ... but perhaps this is what he's investigating.

Thank you for your answer. That is what I figure considering what I have seen here in the United States in different parts of the country. Her model is based on a 30 x 30 grid due to some of the inputs into it which is why she cannot use a smaller size. The habitat has already been classified based on the Gap LU/LC maps. This is more of a is it reasonable quantitative number for classifying a habitat as usable at a 30 x 30m size. Her model is being used for biodiversity to determine more for presence/absence then needing to know exact population densities based on the quality of the habitat. Thank you again for your answer.

Coots are dabbling birds that prefer to lay their eggs on raised mounds in shallow water and/or reedbeds. It's no use modelling open water as breeding habitat if the bathymetry profile is steep and deep. In these kinds of model, garbage in, garbage out...

This model is not specific breeding habitat model. It is a presence and absence model so it includes areas that it may feed or use but not necessarily nest in. She is using the GAP LU/LC map which for the open water classification will include the shoreline. Is there a size of open water areas that coots would not use? She cannot include a bathymetry layer into her model. I am thinking she might want to use only open water areas as moderate habitat if the perimeter of the body water exceeds the area of it. The idea being areas like small ponds, lakes, sloughs, ect will be included as habitat for the bird but exclude larger open lakes. What do you think?

Larger open lakes will still be used, certainly by the Red-knob Coot, but only if they have some shorelines with suitable vegetation. Rather than setting a min or max size or trying to use P/A ratios, it would be better to just take out the central pixels (e.g., reclassify all LC pixels classified as water for which all 8 nearest neighbours are also water). The only challenge here would be that they would still use small vegetated islands. As suggested previously, for min size I would think they would use any body of water that shows up as water in a LCC map at 30x30m resolution. Also, you might have a look at some of the work on modelling mosquito habitat with remote sensing - it's a very similar problem - there are other tools and freely available RS data sets that you could use to get a better estimate than the GAP data are capable of.

Thanks. My coworker is going to try using the nearest neighbor. She is currently using invest examining 12 different species of birds (one of which is the American Coot) that represents variety of biodiversity and habitats through out North Dakota. She is using the GAP data because it is a good baseline for the modeling program invest. For most of the other species she has been able to track down information about minimum habitat sizes, it is only the coot that has been giving her trouble. Again thanks for your help.