Questions about the behavioral ecology of these savanna-woodland chimpanzees center on two (potentially related) topics: the social consequences of low population density, and the use of chimpanzee adaptation to arid, open habitats as a model for understanding early hominid adaptation to similar paleoenvironments.
Central to both is the observation that at Ugalla, chimpanzee population densities are as little as 1/50th of those seen at forested sites.
At such low densities, it would be impractical for a typical community of 20-80 individuals to patrol territorial boundaries; range sizes of more than 100 square kilometers are not defensible in the way seen at forested sites, and conspecific intruder pressure must be dramatically reduced. Variations in intruder pressure can have profound consequences for primate social structure and behavior (e.g. Moore 1999) and the UPP hopes to discover how chimpanzee party stability and (consequently) social relationships are affected by the low population density.
Above, Left: Kibwesi Forest, Kenya – “the best modern analog” for the habitat of Ardipithecus ramidus (Gibbons 2009, “Habitat for Humanity”, Science 326:40). Right: Ugalla, Tanzania.There has been controversy over using chimpanzees as “referential models” of early hominids, with some claiming that such modeling cannot be useful because modern savanna-living chimpanzees are not in fact 4 million-year-old hominids. This argument both echoes (sterile) debates in early 20th century physics and is based on a misunderstanding of what a “model” is; Moore (1996) discusses these points further. Ugalla chimpanzees are similar in body size, cranial capacity and overall biology to early (i.e., pre-Homo) hominids, living in an environment similar to that of Ardipithecus and early Australopithecus. To be useful models, insights gained from observing them must be integrated with information from the fossil record, in a manner informed by behavioral ecological theory, before drawing provisional conclusions about our ancestors. This is one goal of our research.
Referential models are a method. To claim they are inherently flawed is like claiming principal components analysis is inherently flawed.
