The applications of RDFs above were limited in two respects. First the areal extent of some type of habitat class was considered. Second the shape used in the calculations was strictly circular. These two limitations are not always necessary and extensions for both of these aspects need to be developed and applied.
The areal extent might be replaced by various forms. Many of them are indicated in figure 7.7: perimeter (edge length), number of objects or number of different habitat types are only a few examples. In a biological context of resource allocation the biomass of food is an important factor which needs to be considered.
The circular structures used above are only a first, still relatively rough attempt to quantify the environment around a location. In the context of geographic information systems refined approaches are possible and desirable. The first and relatively easy extension is the inclusion of barriers in the calculations. Not all areas are accessible to an animal. This means that the (spatially) available biomass will produce more realistic indications for understanding the way an animal uses its habitat. Going even further with the refinements the euclidian distances used in the calculations so far can be replaced by the real travel distance or even with the energy needed to access these locations. Calculations of travel distances over surfaces including barriers and the like are techniques available in good geographic information systems. Further research could lead to RDF functions which can handle multiple aspects.
In studies where the third order habitat selection of an animal, i.e. the use of habitat within homeranges (Johnson, 1980), is being investigated, the shapes used in RDF calculations can be adjusted to the shape of the homerange.
I think it now becomes clear that RDFs are a versatile framework of how to gain closer and more detailed insights into the habitat configuration around objects, even when the objects are moving as animals do.