Return to search

The response of Amazonian rain forest gas exchange to reduced rainfall

We tested two interlinked hypotheses concerning forest responses to hydraulic stress using data collected at a through-fall exclusion (artificial droughting) experiment at Caxiuana in Eastern Amazonia. First, we tested the ‘isohydric’ hypothesis, that stomatal conductance, in water stressed conditions, operates to maintain leaf water potential above a certain critical threshold limit. We used the soil-plant-atmosphere (SPA) model to predict the expected ecophysiological behaviour of the trees, and tested these predictions against intensive diurnal cycle measurements of leaf water potential, stomatal conductance, sap flow and stem water potential. The data and the model predictions were largely consistent, indicating that the isohydric stomatal control may be the prevailing mechanism controlling water use of rain forest trees in drought stressed conditions. The model was parameterised using independent measurements of ecosystem properties and as such required no fitted parameters. The implication of this is that the response of rain forest ecosystems to drought may be predicted using soil, rooting and vegetation properties. The SPA model is both computationally and data requirement intensive. We assimilated the model and produced a set of empirical equations which replicate the daily modelled gas exchange predictions from daily model inputs. We extrapolated this model 100 years into the future using the latest Amazonian climate predictions and found that after -50 years, a threshold was reached when all the rainfall falling on the plot was evaporated (by interception, soil surface evaporation or evapotranspiration) and none was drained away into stream flow. The response of the forest in the first 50 years was modulated strongly by soil hydraulic properties and rooting depth, and we conclude that increased field measurements of these properties are necessary if the response of Amazonian forests to anticipated drying is to be accurately predicted.

Identiferoai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:650930
Date January 2005
CreatorsFisher, Rosie Alice
PublisherUniversity of Edinburgh
Source SetsEthos UK
Detected LanguageEnglish
TypeElectronic Thesis or Dissertation
Sourcehttp://hdl.handle.net/1842/14860

Page generated in 0.0017 seconds