460
ID:
Background
Spatial variation in tropical forest biomass is largely determined by differences in woody residence times and thus tree mortality. Yet, the causes and consequences of tree mortality are not well understood. Tree death involves multiple interrelated factors from drivers (e.g., drought, elevated temperature) through mechanisms (e.g., biotic outbreaks, fire, wind) that ultimately lead to the physiological processes that kill the trees (i.e., carbon starvation, hydraulic failure). Large-scale and frequent monitoring of tree mortality and associated factors will enable more accurate estimates of forest biomass and the mechanistic representation of mortality in vegetation models.
Objective
Here, we present current advances in tropical tree mortality and damage from a ForestGEO annual monitoring program. We study the relative importance of the tree-level factors involved in tree death and their implications for aboveground biomass (AGB) dynamics.
Methods
We recorded data on multiple tree-level conditions and assessed the subsequent survival of ~44,000 trees with DBH≥1 cm (>2,000 species) across seven tropical forest plots between 2016 and 2022 (29 censuses, 157,241 tree×census observations). Tree-level damage was estimated based on field-based assessments of the proportion of newly broken branches and trunk loss coupled with models describing the accumulation of woody volume with tree height. Damage estimates were then used to compare AGB losses from dead versus damaged trees.
Results
Across 19 mortality risk factors examined, we found that light limitation and crown/trunk damage were the most impactful risks (i.e., had the two highest ‘excess mortality’, relative to stand-level mortality). The mortality risks studied co-occurred strongly among trees and, as a result, limited relationships were found between the traditional modes of death (standing, broken, and uprooted) and tree-level conditions/risks. We also found that over one-third of the total AGB losses in these forests were due to damage to living trees (i.e., branch fall, trunk breakage, and/or standing wood decomposition), varying greatly across sites.
Implications
Our results show the role of tree-level damage in forest carbon dynamics. While tree-level damage is the most common condition preceding death in tropical trees, non-lethal damage can contribute to significant amounts of AGB losses that are not necessarily captured by studies focused only on tree mortality. Resolving the timing of lethal and non-lethal damage as well as their climatic drivers and physiological consequences should be a priority to better understand woody residence times and improve predictions of the fate of forests in vegetation models.
Keywords:
climate change, forest dynamics, forest carbon, mortality risk, vegetation models