The ascent of water from the soil to the leaves of vascular plants, described by the study of plant hydraulics, regulates ecosystem responses to environmental forcing and recovery from stress periods. Several approaches to model plant hydraulics have been proposed. In this study, we introduce four different versions of plant hydraulics representations in the terrestrial biosphere model T&C to understand the significance of plant hydraulics to ecosystem functioning. We tested representations of plant hydraulics, investigating plant water capacitance, and long-term xylem damages following drought. The four models we tested were a combination of representations including or neglecting capacitance and including or neglecting xylem damage legacies. Using the models at six case studies spanning semiarid to tropical ecosystems, we quantify how plant xylem flow, plant water storage and long-term xylem damage can modulate overall water and carbon dynamics across multiple time scales. We show that as drought develops, models with plant hydraulics predict a slower onset of plant water stress, and a diurnal variability of water and carbon fluxes closer to observations. Plant water storage was found to be particularly important for the diurnal dynamics of water and carbon fluxes, with models that include plant water capacitance yielding better results. Models including permanent damage to conducting plant tissues show an additional significant drought legacy effect, limiting plant productivity during the recovery phase following major droughts. However, when considering ecosystem responses to the observed climate variability, plant hydraulic modules alone cannot significantly improve the overall model performance, even though they reproduce more realistic water and carbon dynamics. This opens new avenues for model development, explicitly linking plant hydraulics with additional ecosystem processes, such as plant phenology and improved carbon allocation algorithms.

ALBIOC (ALbedo- BIOsphere- Carbon) is an integrated terrestrial biosphere model designed as a too] to explore the effects of climate and atmospheric CO, concentration on vegetation, land-surface characteristics and carbon storage. The model is based, although designed to be simple in structure and computationally fast, on biophysical and ecophysiological principles and simulates in a fully interactive manner the potential distribution of vegetation, terrestrial carbon storage and physical land-surface properties. Testing was extensive and focused on broad spatial patterns (5 degrees resolution) of biome distribution, and variables important for the surface energy balance and hydrological cycle (seasonal snow cover, surface albedo, runoff and evaporation) and for the global carbon cycle (seasonal canopy cover, primary production and carbon storage). Because ALBIOC simulates a range of physical and biogeochemical variables in an integrated way, it was possible to test the model against a more comprehensive range of indicators than has normally been the case for terrestrial biosphere models. The simulated vegetation distribution is as accurate as more specialised biogeography models taking into account the coarse resolution of the model. ALBIOC simulates a global NPP of 57 PgC/year, which is in the range of the values found in the literature and other model estimates. Land-surface albedo. snow depth, runoff, and FPAR showed a generally good agreement with observations within the known limits of available data sets of these variables. The model's mechanistic basis would allow extension to simulate, e.g. transient response to rapid climate change (vegetation dynamics) and carbon isotopic balances. while its computational efficiency renders it suitable for inclusion in Earth system models of intermediate complexity. (C) 2001 Elsevier Science B.V. All rights reserved.