Does life form affect tree species assembly? A demographic study across the life history of a temperate forest in Japan

In this study, the dominant mechanism that defines the species assembly in an old-growth temperate forest was investigated by analyzing the species specificity of demographic parameters in relation to the life history stage, testing the hypotheses that early life stages should be more critical for sub-canopy species than canopy species. Data for 17 major tree species in the Ogawa Forest Reserve (OFR) in Japan were analyzed, together with data from early stages (seed production, seed dispersal, and seedling emergence) to dbh-based transition probability and survival stages (5-80 cm). Both datasets were subjected to principal components analysis (PCA): one included tree of dbh < 35 cm for all species (small-sized trees of all species [SSAS]) and the other included tree species with dbh = 55 cm (large-sized species [LSS]). The most influential was the growth-survival trade-off for trees at small-to-middle-sized stages (dbh: 10-40 cm), which captured a moderate proportion of the demographic variation (24.3%-26.4%). At these stages, some Quercus species were characterized by higher growth, and some Fagus and Acer species by shade tolerance. Behavior during early-to-late life stages was the second most important dimension, for SSAS, sub-canopy tree species showed high survival at early life stages than canopy species as the second important dimension, which differed between the SSAS- and LSS-based PCAs. The relationship between functional traits and PCA axis dimensions varied with PCA type and dimension. The growth-survival trade-off in the OFR was closely associated with maximum tree height, a proxy for life form, and specific leaf area a proxy for shade tolerance. This finding implies a significant role for the growth-survival trade-off in stable species coexistence by maintaining mutual claims to vertical light resources and allowing the more efficient use of the forest canopy space. However, there was no significant association between wood density and demographic variables. Our study shows that interspecies relationships vary with the life history stage and that early life stages are more critical for sub-canopy species than canopy species in temperate forests, where the entire life history of the trees must be considered focusing on the difference in life forms.