Survival
Neighbor density. In both Q. macrocarpa and Q. ellipsoidalis , neighborhood density, both heterospecific and conspecific, was positively associated with survival likelihood. Effects were consistent across dbh classes and censuses (Figs. 2 and 3). The effect of neighbors did not diminish with distance. Effect sizes were often +0.5 up to +0.9. To reach effects > 0.8 means survival would increase from 10% per 5 years without neighbors to 90% with dense crowding.
The two species differed in detail. Indeed, both benefited most by having Q. macrocarpa neighbors, that is, the conspecific effects in Q. macrocarpa were stronger than the heterospecific, but just the opposite in Q. ellipsoidalis (Figs. 2 and 3). Note: while small sample sizes of the 2-5 cm dbh trees during the latter time intervals made it impossible to model these trees throughout the study, during the first two intervals, the survival effects of neighbors for these trees were similarly consistently positive. As in the larger trees, small Q. macrocarpa were best protected by conspecifics, while small Q. ellipsoidalis were best protected by heterospecifics.
While neighbor facilitation with respect to tree survival persisted throughout the study, and in most cases remained strong after twenty-five years, the data showed a slight but consistent downward trend in the strength of this relationship (Figs. 2 and 3).
Elevation . Survival generally was not significantly associated with elevation. Few tests were significant, and those that were significant were inconsistent in direction, so we omit the effect sizes.