Understanding the patterns and drivers of biomass allocation among organs at a broad scale is crucial for predicting the responses of plant growth and carbon sequestration to environmental change. However, the extent to which the general rules govern these patterns and the key factors affecting biomass allocation remain poorly understood. Using a global dataset of 239 tree species, we tested the two prevailing theories (i.e., the allometric partitioning theory (APT) and the optimal partitioning theory (OPT)) by investigating the scaling relationships between plant organs and how environmental factors and phylogeny shape the patterns of biomass allocation. Our results generally support APT at the global scale, with variations in biomass allocation patterns explained by OPT. As plant size increased, a significant shift in biomass allocation from leaves to roots and stems, as well as from roots to stems, was observed. Specific environmental factors (including temperature, precipitation variables, and soil properties) significantly influenced biomass allocation with distinct patterns in the angiosperms and gymnosperms, even when the allometric effects were taken into account. We conclude that tree biomass allocation among organs (i.e., the ratios of leaf to stem, leaf to root, stem to root, and aboveground to belowground) is governed by allometry but modulated by optimization at the global scale. Our findings highlight the importance of considering both the ontogenetic and environmental effects in predicting the responses of biomass sequestration to phylogenetic and environmental factors.

Tree mortality significantly influences forest structure and function, yet our understanding of its dynamic patterns among a range of tree sizes and among different plant functional types (PFTs) remains incomplete. This study analysed size-dependent tree mortality in a temperate forest, encompassing 46 tree species and 32,565 individuals across different PFTs (i.e., evergreen conifer vs. deciduous broadleaf species, shade-tolerant vs. shade-intolerant species). By employing all-subset regression procedures and logistic generalized linear mixed-effects models, we identified distinct mortality patterns influenced by biotic and abiotic factors. Our results showed a stable mortality pattern in evergreen conifer species, contrasted by a declining pattern in deciduous broadleaf and shade-tolerant, as well as shade-intolerant species, across size classes. The contribution to tree mortality of evergreen conifer species shifted from abiotic to biotic factors with increasing size, while the mortality of deciduous broadleaf species was mainly influenced by biotic factors, such as initial diameter at breast height (DBH) and conspecific negative density. For shade-tolerant species, the mortality of small individuals was mainly determined by initial DBH and conspecific negative density dependence, whereas the mortality of large individuals was subjected to the combined effect of biotic (competition from neighbours) and abiotic factors (i.e., convexity and pH). As for shade-intolerant species, competition from neighbours was found to be the main driver of tree mortality throughout their growth stages. Thus, these insights enhance our understanding of forest dynamics by revealing the size-dependent and PFT-specific tree mortality patterns, which may inform strategies for maintaining forest diversity and resilience in temperate forest ecosystems.

Seedling stage has long been recognized as the bottleneck of forest regeneration, and the biotic and abiotic processes that dominate at seedling stage largely affect the dynamics of forest. Seedlings might be particularly vulnerable to climate stress, so elucidating the role of interannual climate variation in fostering community dynamics is crucial to understanding the response of forest to climate change. Using seedling survival data of 69 woody species collected for five consecutive years from a 25-ha permanent plot in a temperate deciduous forest, we identified the effects of biotic interactions and habitat factors on seedling survival, and examined how those effects changed over time. We found that interannual climate variations, followed by biotic interactions and habitat conditions, were the most significant predictors of seedling survival. Understory light showed a positive impact on seedling mortality, and seedling survival responded differently to soil and air temperature. Effects of conspecific neighbor density were significantly strengthened with the increase of maximum air temperature and vapor pressure deficits in the growing season, but were weakened by increased maximum soil temperature and precipitation in the non-growing season. Surprisingly, seedling survival was strongly correlated with interannual climate variability at all life stages, and the strength of the correlation increased with seedling age. In addition, the importance of biotic and abiotic factors on seedling survival differed significantly among species-trait groups. Thus, the neighborhood-mediated effects on mortality might be significantly contributing or even inverting the direct effects of varying abiotic conditions on seedling survival, and density-dependent effects could not be the only important factor influencing seedling survival at an early stage.