We examine closely the models, methods and conclusions of Doncaster and Woo
droffe (1993: Oikos 66: 88-93) who argued that den or main sett sites of cl
ans of badgers, Meles meles, are particularly important in determining terr
itory shape and size, and hence influence the size of social group. We cons
ider a realistic alternative hypothesis which allows the key assertion by D
oncaster and Woodroffe to be directly tested. We show that a Dirichlet tess
ellation model that does nor give a major role to the main setts fits data
from several studies - two of those considered by Doncaster and Woodroffe,
and a more recent and extensive one - significantly better than Doncaster a
nd Woodroffe's model. For the majority of territories, especially in the mo
st extensive data set, differences in territory shape and size under the tw
o models are substantial, suggesting that a different biological mechanism
is at work, as well as or instead of dependence on main sett locations.