Predicting postfire Douglas-fir beetle attacks and tree mortality in the northern Rocky Mountains

Abstract

Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) were monitored for 4 years following three wildfires. Logistic regression analyses were used to develop models predicting the probability of attack by Douglas-fir beetle (Dendroctonus pseudotsugae Hopkins, 1905) and the probability of Douglas-fir mortality within 4 years following fire. Percent crown volume scorched (crown scorch), cambium injury, diameter at breast height (DBH), and stand density index for Douglas-fir were most important for predicting Douglas-fir beetle attacks. A nonlinear relationship between crown scorch and cambium injury was observed, suggesting that beetles did not preferentially attack trees with both maximum crown scorch and cambium injury, but rather at some intermediate level. Beetles were attracted to trees with high levels of crown scorch, but not cambium injury, 1 and 2 years following fire. Crown scorch, cambium injury, DBH, and presence/absence of beetle attack were the most important variables for predicting postfire Douglas-fir mortality. As DBH increased, the predicted probability of mortality decreased for unattacked trees but increased for attacked trees. Field sampling suggested that ocular estimates of bark char may not be a reliable predictor of cambium injury. Our results emphasize the important role of Douglas-fir beetle in tree mortality patterns following fire, and the models offer improved prediction of Douglas-fir mortality for use in areas with or without Douglas-fir beetle populations.Résumé : Des douglas de Menzies (Pseudotsuga menziesii (Mirb.) Franco) ont été suivis pendant 4 ans à la suite de trois incendies de forêt. Des analyses de régression logistique ont été utilisées pour élaborer des modèles de prédiction de la probabilité d'une attaque du dendroctone du douglas (Dendroctonus pseudotsugae Hopkins, 1905) et de la probabilité que le douglas de Menzies meure dans les quatre années suivant un feu. Le volume de cime roussi, les dommages au cambium, le diamètre à hauteur de poitrine (DHP) et l'indice de densité du peuplement de douglas de Menzies étaient les variables les plus importantes pour prédire les attaques du dendroctone. Une relation non linéaire entre le roussissement de la cime et les dommages au cambium a été observée, ce qui signifie que les dendroctones n'attaquent pas de préférence les arbres qui ont à la fois le maximum de dommages à la cime et au cambium mais plutôt un niveau intermédiaire quelconque. Les dendroctones étaient attirés par les arbres avec un degré élevé de roussissement de la cime mais sans dommages au cambium, un et 2 ans après un feu. Le roussissement de la cime, les dommages au cambium, le DHP et la présence ou l'absence d'attaques du dendroctone étaient les variables les plus importantes pour prédire la probabilité que le douglas de Menzies meure après un feu. La probabilité estimée de mortalité diminuait avec l'augmentation du DHP chez les arbres qui n'avaient pas été attaqués alors que la probabilité qu'un arbre meure augmentait avec l'augmentation du DHP chez...

Pourquoi et où les arbres adultes meurent dans une jeune forêt tempérée? Le rôle du voisinage

“…Probably, this is related to the fact that adult mortality is a rare event in long-lived species (García and Zamora, 1993). The large sample sizes and long periods of time required to obtain an appropriate number of mortality events supposes a high setback for this type of research (Hood and Bentz, 2007;Wyckoff and Clark, 2000). Not surprisingly, much of the current effort to explain causes and processes of tree mortality has been devoted to catastrophic events such as wildfires (Reggelbrugge and Conard, 1993), hurricanes (Uriarte et al, 2004;Wolf et al, 2004), root rot (Dobbertin et al, 2001) or drought (Suarez et al, 2004), where tree mortality peaks, but comparatively less effort has been dedicated to explore mortality in non-perturbed forest stands (Aakala et al, 2007;Bigler and Bugmann, 2004).…”

Proceedings of the coast redwood forests in a changing California: A symposium for scientists and managers

“…The utility of the cambium kill metric in mortality models following wildfire has been reported by others but not for coast redwood (Hood et al 2007). Measures of cambium kill significantly improved mortality predictions for trees ≤ 20.3 cm DBH with high levels of crown scorch.…”

Recognizing Women Leaders in Fire Science

“…Hood's research focuses on the mechanisms of post-fire tree mortality, the impacts of changing fire regimes on forest succession and resilience, and the impacts of fire on the susceptibility of bark beetles to attack trees. Her notable works include research to evaluate post-fire conifer mortality [90,94,268]. Her recent works have continued to build on post-fire tree mortality [269], assessing the ability of fires and thinning to help increase resistance to beetle attacks [95,96], as well as assessing the long-term impacts of fuel treatments [270].…”