Site and species differences in tree mortality in southwestern mixed-conifer forests of northern Arizona

Background/Question/Methods

Increases in tree mortality have been documented in high-elevation forests in many parts of the Western U.S.   Few studies, however, have accurately quantified recent tree mortality in high-elevation forests of the southwestern U.S.   Our study established 85, 1/20^th ha permanent plots along aspect and elevational gradients located across three independent sites in southwestern mixed-conifer forests of northern Arizona; Bill Williams Mountain (BWM), San Francisco Peaks (SFP), and Sitgraeves Mountain (SIT).  Within each plot, we measured the diameter of all live and dead trees and collected topographic information, including elevation, aspect, and slope.   We quantified % mortality by basal area (m² ha^-2) for recently dead trees (decay class 1 or 2) at each plot for the most common tree species; Abies concolor, Pinus flexilis, Populus tremuloides, and Pseudotsuga menziesii.   We use this data to ask four main questions:  1) Does tree mortality differ among species?; 2) Does tree mortality differ among sites?; 3) Does tree size differ between live and recently dead trees?; and 4) What topographic and stand density factors are associated with tree mortality?

Results/Conclusions

Over all sites, average tree mortality was 7%, 28%, 28%, and, 49% for Pinus flexilis, A. concolor, P. menziesii, and Populus tremuloides, respectively.  We detected a significant site effect (F > 7.0, P < 0.005) on mortality of all species except P. menziesii.  Dead A. concolor trees were 100% larger than live trees at the BWM site (t = 3.00, P < 0.005).  Conversely, dead trees of the other three species were smaller than live trees for at least one site each (P < 0.05).  Lastly, factors related to tree mortality differed by site and species.  For example, A. concolor mortality was negatively correlated with elevation and slope at the BWM site (r² > 0.5), while P. menziesii mortality was negatively correlated to only elevation at the SIT site.  Populus tremuloides mortality was negatively correlated to live intraspecific basal area (r² = 0.61) at BWM but was weakly negatively correlated to elevation at the SFP site (r² = 0.14).  Our results highlight that tree mortality varies by site and species for southwestern mixed-conifer forests in northern Arizona, and suggest site- and species-specific causes of tree mortality.