O papel da fenologia foliar sobre a dinâmica da estrutura populacional e etária de folhas no dossel superior de uma floresta de terra firme na Amazônia Central

Leaf phenology impacts carbon, water, and nutrient cycles from local to global scales. In the Amazon forest, massive events of leaf flushing and abscission promote a seasonal shift in the age composition of upper-canopy leaves. A demographic and ontogenetic model suggests that this shift is the main trigger of the seasonality of photosynthetic capacity (PC) in Terra Firme forests in central Amazonia. However, the method (phenocams) from which data are collected and used to feed the model present limitations that enable to realize percentage of leaf turnover in evergreen tree crowns and the mean age of leaves discarded annually, thus generating uncertainties regarding the proportion of new, mature, and old leaves in the upper canopy. Leaf age dynamics emerge from trees with different habits and leaf phenological strategies. Therefore, investigating the temporal behavior of leaf dynamics in the upper-canopy tree community is necessary to understand age dynamics at the ecosystem scale. Monthly assessment (November/2015 to November/2019) of the components of leaf dynamics (leaf storage, leaf flush, and leaf fall) was conducted on branches, totaling 330, equally distributed among 33 trees at upper canopy Terra Firme forest on a plateau in central Amazonia (ATTO site). Information regarding leaf age was obtained only for those leaves for which the flushing event was observed. Similarly, leaf longevity data were extracted only for leaves for which both flushing and death events were necessarily observed throughout the study. Three phenological behaviors were identified: evergreen (29), annually brevi-deciduous (1), and biennially brevi-deciduous (3), which appear to be associated with distinct strategy of resource use and management, such as water and carbon. Evergreen canopies exhibit a complex age structure, with high variability in leaf longevity (1 to 43 months), revealing information that is hidden from phenological cameras. In contrast, leaf longevity close to and concentrated around 12 and 24 months was reported for annually and biennially brevi-deciduous trees, respectively. Leaf dynamics sampled on branches proved to be representative of the ecosystem. The number of leaves in the upper canopy does not appear to be seasonal, although annual events of massive leaf exchange by the tree community during dry months significantly alter the age composition of upper-canopy leaves. This composition is diverse and complex, shaped by individual trees with distinct leaf habits and phenological strategies. Annual canopy turnover appears to occur partially (50%), as a result of a mean leaf longevity longer than one year (≈ 20.1 months), and therefore half of the canopy population is composed of leaves that survive beyond an annual leaf phenology cycle. The empirical evidence presented here sheds new light on the understanding of leaf longevity in the upper canopy and the pattern of seasonal canopy turnover, aspects that remain poorly represented or absent in current models.