Modeling post-fire mortality of Turkish pine (Pinus brutia Ten.)

Abstract

Turkish pine (Pinus brutia) is a widely distributed species in fire-prone regions of the eastern Mediterranean, where it is of economic and ecological importance. After wildfires, damaged trees are usually salvage logged out of concern that fire-injured individuals will eventually die. While fire is a major cause of Turkish pine mortality, the species has traits, such as thick bark, that should confer some resistance to low-intensity fire. The aim of this study is to develop post-fire tree mortality models for Turkish pine to support post-fire management decisions. We selected 510 Turkish pine trees from four wildfires that occurred in the Mediterranean region of Türkiye in 2018 and 2019. Only fire-damaged trees that had initially survived the fires were selected. We measured tree diameter, height, and bark thickness and fire-caused injuries (crown volume scorched, bark char height, bark char index, cambium kill rating, and ground char index) within two months after fire. Stem-related variables were measured at three different heights (0, 50, and 130 cm). We then quantified delayed mortality for two years after each fire and developed mixed effects logistic regression models to predict probability of mortality. Most Turkish pines (84 %) survived, with the majority of mortality occurring during the first year post-fire. Crown volume scorched and stem damage variables were the most important predictors of mortality. Models were selected based on their explanatory power and practical use in the field. Among these, the most explanatory model includes crown volume scorched and bark char index measured at 0 cm stem height. The suggested model for use in forestry includes crown volume scorched and diameter at breast height. Some trees were able to survive and flush new foliage during the first season after fire despite 100 % crown scorch due to low bud kill. Our results show that Turkish pine can survive fire when stem injury is low. The mortality models can assist managers in making more evidence-based decisions for post-fire management and developing more ecologically sustainable post-fire management operations.