Tests of CSR of Oklahoma dataset
The first set of analyses on the 13 study sites in the Oklahoma (OK) dataset included all living and dead redcedar, Quercus, and deciduous trees (OK all trees ) found no evidence of any site differing from CSR. When K-function tests were run on only the live redcedar, Quercus , and other deciduous trees (OK living trees ), visual checks of the envelopes indicated possible overdispersion at Sites 2, 11, and 12 between ~1-3 m but DCLF testing indicated no significant departure from CSR. Visual assessment of the K-functions indicated potential clustering of Sites 9 and 13 but DCLF testing indicated only Site 9 was found to have a significant (p = 0.048) departure from CSR between 1-3 m (not shown). However, when the OK all trees data were analyzed with K-function envelopes weighted by tree diameter, six sites showed signs of overdispersion (Figure 3). Subsequent DCLF testing adjusted for tree diameter confirmed overdispersion for sites 3, 8, 9, and 12. Visual inspection of the K-function tests weighted by tree diameter found probable overdispersion in sites 1, 2, 3, 4, 8, 9, 10, and 12 of theOK living trees dataset and DCLF testing confirmed overdispersion in site 2 (p = 0.104), site 3 (p = 0.044), site 8 (p = 0.021), site 9 (p = 0.036), and site 12 (p = 0.013).