binary spatial data


Just working on an environmental project about vegetation mapping. I've got a regular grid of points with percentage of plants for each points. I've intepolated the different percentage values to produce distribution maps for each plant.
Anyway, now i wonder what i can do if instead of using percentage values i just use presence/absence value for each plants. I've heard about probability kriging to make some sort of probability map.
What should i hope if i use this method ?
Any other idea ?

Thanx in advance