A comparison of mortality rates between top height trees and average site trees

• Although comparisons between methods of selecting trees for site index estimates are well documented in the literature, little is known on mortality rates of different canopy tree cohorts used for that purpose. • This study was initiated to test the hypothesis that the mortality rates of top height trees are lower than those of codominants only or a combination of codominant and dominant trees. To test this hypothesis, we used records from a network of permanent sample plots in Quebec and studied the fate of different cohorts of site trees for five different species. • Our results did not show clear evidence of lower mortality rates for top height trees. Instead we found that depending on the species, top height trees have lower (Populus tremuloides, Pinus banksiana), higher (Picea mariana, Abies balsamea) or equal mortality rates (Betula papyrifera) than codominant trees or codominant and dominant trees combined. • These results suggest a tendency for shade intolerant species to maintain lower top height tree mortality rates over time when compared to shade tolerant species. In the latter case, it is also shown that spruce budworm epidemics (Choristoneura fumiferana) did not change the pattern of mortality rates of site trees of A. balsamea.