Incisor size and community structure in rodents: Two tests of the role of competition

We used two different approaches to test for the effect of interspecific co mpetition on community-wide patterns in the size of the upper incisor in si x rodent communities. One tests For constancy of size ratios between adjace nt species (Barton and David's test), and the other rests for minimum mean size overlap between species pairs (randomization test). The results of the two tests were more congruent for the radius of the upper incisor than for its diameter. Although a number of tests gave significant results, they le d to a consistent rejection of the null hypothesis of a randomly generated pattern only in two communities from an African rainforest. Many factors ar e likely to disturb community structure, among which we identify recent spe cies introductions and the presence of rare species. Our results also show that Hutchinson's rule of a limiting size similarity does not hold; the mea n size ratio between adjacent species was correlated with the overall size range and the number of species in a community. (C) Elsevier, Paris.