Climate extremes during high competition contribute to mortality in unmanaged self-thinning Norway spruce stands in Bulgaria

Climatic variability often is thought to be most important for ecosystem development at ecotones, while competition is thought to be most important farther from ecotones, where neighboring plants compete for scarce resources. However, climatic variability may also modulate consequences of competition, especially under recent and future climate change. Norway spruce (Picea abies) forests are among the most important ecosystems in the mountain regions of Europe and provide various ecosystem services. Many of these forests are currently in a self-thinning, stem-exclusion phase. Understanding processes governing forest dynamics during this phase is necessary for understanding future forest structure and processes as well as effects of climatic variability on ongoing forest development. We studied growth and mortality patterns in unmanaged 100-150 years-old Norway spruce forests that originated after stand-replacing disturbances in the Parangalitsa Reserve in Bulgaria. We collected data on forest structure and tree ring samples from 648 live and dead trees (DBH \u3e 4cm) to analyze onset, pattern and duration of mortality, as well as contributing factors.We found that climate extremes acted together with competition to cause sharp growth declines lasting from a few years to several decades and, in some cases, eventually led to death. The majority of dead trees had one to several consecutive growth declines, most of which initiated in response to extreme summer droughts during periods of high within-stand competition (after trees were 40-50 years old). Our tree-ring analysis revealed that some suppressed trees that died were more drought-sensitive than living trees. Other climate extremes such as unusually cold winters or summers also contributed to sharp growth reductions in some cases. Trees that died had significantly lower initial radial growth, which suggests that in the absence of external disturbances, the outcome of mortality in the stem-exclusion stage may be pre-determined from factors that determine initial growth rates. Spatial distribution data showed that there was no significant aggregation of dead and live trees and that in almost all cases, neither live nor dead trees were clustered.Our findings contribute to understanding mortality processes in self-thinning subalpine Norway spruce forests in Europe and show that under climate change scenarios that include more frequent future droughts, even forests in which competition is thought to be the main driver of dynamics, may experience higher rates of mortality