CR 28:1-10 (2004)  -  doi:10.3354/cr028001

ABSTRACT: Biogeographers and ecologists often need to quantify species-environment relationships to understand the distribution of vegetation, and to assess changes in patterns resulting from environmental change. A water budget approach, which incorporates evaporative demand and moisture availability, is compared to traditional climatic variables in terms of ability to predict species growth and abundance in eastern North America. Firstly, tree growth is examined using tree-ring chronologies obtained from the International Tree-Ring Data Bank, correlated with climatic data from the nearest site in the US Historical Climatology Network. Secondly, logistic regression is used to model the range of American beech (Fagus grandifolia) using pollen records from the World Data Center for Paleoclimatology, and climatic data from NCAR’s Community Climate Model (CCM1) general circulation model (GCM) for the control, 6 ka, and 11 ka runs. Tree growth, especially for oaks and other deciduous trees, correlates more strongly with early growing-season deficit than with precipitation. Water budget variables (actual evapotranspiration and deficit) are also more effective than traditional climatic variables in modeling the range of beech. A water budget approach is attractive for modeling vegetation dynamics because it transcends geographic scale; it is able to model both local and continental-scale phenomena.

KEY WORDS: Water budget · Pollen · Tree rings · General circulation model · Scale · Beech