Size-dependent mortality of Khingan fir tree in old-growth forests: a spatial point pattern analysis

Abstract

Tree mortality rates are increasing globally. Yet the finer-scale processes driving these patterns remain poorly understood. A deeper understanding of these mechanisms is critical for predicting forest responses to gradual climatic changes. Abies nephrolepis (Trautv. ex Maxim.) Maxim., a common coniferous tree species in the subalpine or boreal forests, has exhibited widespread mortality in old-growth stands. To investigate this phenomenon, we characterized the mortality spatial pattern of A. nephrolepis using data collected from two sites with varying elevations and slopes in the Mudanfeng Natural Reserve, China. Mark connection functions revealed that mortality labels (“dead” vs. “surviving”) were randomly distributed among all fir trees, consistent with a spatially random point process. Trivariate random labeling analyses indicated no influence of neighboring trees, either conspecific or heterospecific, on mortality patterns. Bivariate difference pair correlation functions provided no evidence for density-dependent mortality. However, mortality exhibited clear size dependence. We conclude that the random spatial distribution of tree sizes drove the random mortality pattern, with physiological senescence likely serving as the primary underlying mechanism.