Dynamics of beetle-killed snags following mountain pine beetle outbreaks in lodgepole pine forests

Snags (standing dead trees) are important components of forest ecosystems that, among other benefits, provide critical habitat for many species of wildlife, but also represent important safety concerns to firefighters, forest workers, and the public. We identified factors that influence the fall rates of lodgepole pines, Pinus contorta Dougl. ex Loud., killed by bark beetles during a severe regional-scale outbreak of mountain pine beetle, Dendroctonus ponderosae Hopkins, that occurred in 2004-2012. Data were obtained during annual assessments (2010-2019) of a network of 107 0.081-ha circular plots installed in P. contorta forests in Colorado, Idaho, Montana, Utah and Wyoming, U.S. A total of 3789 P. contorta snags were monitored, and in each case we recorded, among other variables, the year the snag was created and the year the snag fell to the forest floor. Among snag age classes (i.e., number of years since tree death), the highest number of snags (1046) were 12 years since death (YSD), and those 13 and 14 YSD exhibited the lowest fall rates to date (<10%) despite being the oldest in our study. Snags 13 and 14 YSD were among the largest in diameter. Snag fall occurred in every snag age class from 1 to 14 YSD, with the greatest proportion of snag fall events occurring 4-8 YSD. By 2019, 24.7% (937) of snags fell to the forest floor. We modeled snag fall using a Cox's proportional-hazards model that included six covariates of interest: elevation (m), slope aspect (categorical, cardinal direction), slope (%), canopy cover (%), snag height (m), and snag height:dbh (m, diameter at 1.37 m in height). Slope aspect had the strongest influence on fall rates. Northern aspects, increased canopy cover, and taller snag heights decreased the probability of snag fall. Conversely, southern aspects and increased height:dbh ratios (i.e., taller, skinner snags) increased the probability of snag fall. The predicted half-life (i.e., the amount of time since death required for 50% of the snag population to fall to the forest floor) was similar to 16 YSD after which the function predicted a linear, similar to 0.04/year decline in snag survival probability 15-30 YSD. The observed longevity of P. contorta snags confers important ecological benefits for some wildlife, and may offer opportunities for extended periods of salvage. However, checking (cracks) rapidly occurred, and many snags 6 YSD and older had at least one check at 1.37 m and 5.5 m in height, which can negatively impact lumber recovery. The relatively slow snag fall rates observed in our study also lengthen concerns regarding hazard trees, human safety, and protection of critical infrastructure. The implications of these and other results to management of P. contorta forests are discussed.