Dry‐Season Greening and Water Stress in Amazonia: The Role of Modeling Leaf Phenology

Abstract

Large uncertainties on the sensitivity of Amazon forests to drought exist. Even though water stress should suppress photosynthesis and enhance tree mortality, a green‐up has been often observed during the dry season. This interplay between climatic forcing and forest phenology is poorly understood and inadequately represented in most of existing dynamic global vegetation models calling for an improved description of the Amazon seasonal dynamics. Recent findings on tropical leaf phenology are incorporated in the state‐of‐the‐art eco‐hydrological model Thetys & Chloris. The new model accounts for a mechanistic light‐controlled leaf development, synchronized dry‐season litterfall, and an age‐dependent leaf photosynthetic capacity. Simulation results from 32 sites in the Amazon basin over a 15‐year period successfully mimic the seasonality of gross primary productivity; evapotranspiration (ET); as well as leaf area index, leaf age, and leaf productivity. Representation of tropical leaf phenology reproduces the observed dry‐season greening, reduces simulated gross primary productivity, and does not alter ET, when compared with simulations without phenology. Tolerance to dry periods, with the exception of major drought events, is simulated by the model. Deep roots rather than leaf area index regulation mechanisms control the response to short‐term droughts, but legacy effects can exacerbate multiyear water stress. Our results provide a novel mechanistic approach to model leaf phenology and flux seasonality in the tropics, reconciling the generally observed dry‐season greening, ET seasonality, and decreased carbon uptake during severe droughts.

A Mechanistic Model of Microbially Mediated Soil Biogeochemical Processes: A Reality Check

“…Numerical simulations were carried using the ecosystem model T&C (Fatichi et al, , , ; Fatichi & Pappas, ; Mastrotheodoros et al, ; Manoli et al, ; Paschalis et al, , ; Pappas et al, ) combined with new modules simulating soil biogeochemistry and plant nutrient dynamics (T&C‐BG) described in the following and extensively in the supporting information: Figure S1, Texts S1 and S2, and additional references in the supporting information (Ainsworth & Long, ; Batterman et al, ; Chapin et al, ; Curry & Schmidt, ; Daly et al, ; Daly & Porporato, ; Ebrahimi & Or, ; Friend et al, ; Farquhar et al, ; Hassink & Whitmore, ; Hanson et al, ; Jackson et al, ; Jungk, ; Kögel‐Knabner, ; Moorhead & Sinsabaugh, ; Moyano et al, ; Manzoni et al, , , ; Manzoni & Porporato, ; Manzoni, ; Poorter, ; Poorter & Villar, ; Phillips et al, ; Roumet et al, ; Sparks & Carski, ; Stewart et al, ; Smith & Read, ; Smith & Smith, ; Sinsabaugh et al, , ; Thomas & Stoddart, ; Thomas & Martin, ; Yang et al, ; Zhang et al, ). The original T&C is a mechanistic model simulating energy, water, and CO 2 exchanges at the land surface at an hourly time step.…”

“…As usual in T&C applications (Fatichi et al, ; Fatichi & Pappas, ; Mastrotheodoros et al, ), biomes were not parameterized with generic plant functional types, but for each site a parameter set able to provide satisfactory results in terms of vegetation productivity, leaf area index, soil moisture, energy and water fluxes, and local phenology was identified acting on the most sensitive parameters. The capability of the original T&C model to reproduce the observed energy and water fluxes and vegetation phenology as well as response to environmental manipulations against observations has been published before for a large number of location worldwide, and the 20 selected sites are a subset of those (e.g., Fatichi & Leuzinger, ; Fatichi & Ivanov, ; Fatichi et al, , ; Fatichi & Pappas, ; Mastrotheodoros et al, ; Manoli et al, ; Pappas et al, ). A single parameter set for the soil biogeochemistry module was selected based on literature parameters and preliminary model tests and is fully documented in the supporting information.…”

Ecohydrological dynamics in the Alps: Insights from a modelling analysis of the spatial variability

“…We used the mechanistic ecohydrological model T&C (Fatichi, Ivanov, & Caporali, ; Fatichi & Leuzinger, ; Fatichi & Pappas, ; Manoli, Ivanov, & Fatichi, ; Pappas, Fatichi, & Burlando, ), which simulates the principal processes of the hydrological cycle, such as precipitation interception, transpiration, ground evaporation, infiltration, and surface/subsurface water fluxes, including the lateral transfer of water. It further simulates plant‐related processes, such as photosynthesis, phenology, carbon allocation, and tissue turnover.…”

Cryptic phenology in plants: Case studies, implications, and recommendations

“…In these cases, time of year or “season” serves as a proxy for leaf age, which may work well for some PFTs, but not for tropical evergreen broadleaf forests where the “evergreen” canopy belies cyclic leaf turnover that the PFT rule set does not include. This case study suggests that accounting for cryptic phenology is necessary for correctly detecting, attributing, and modeling the carbon exchange dynamics of tropical forests (De Weirdt et al, ; Kim et al, ; Manoli, Ivanov, & Fatichi, ; Restrepo‐Coupe et al, ).…”