Tree Physiology 7,329-345 0 1990 Heron Puhlishin~Victoria, Canada Vertical variation in canopy structure and CO2 exchange of oak-maple forests: influence of ozone,nitrogen, and other factors on simulated canopy carbon gain P. B. REICH, D. S. ELLSWORTH, B. D. KLOEPPEL,’ J. H. FOWNES2 and S. T. GOWER Department of Forestry, University of Wisconsin-Madison, Madison, WI 53706, USA Summary Stand-level and physiological measurements were made for oak and maple species common in Wiscon- sin forests. Scaling relationships were identified to allow the development of a model for estimating net carbon exchange at the levels of a leaf, canopy stratum, and whole canopy. Functional relationships were determined between tissue gas exchange rates and perceived controlling variables. Vertical variation in leaf properties and in the distribution of foliage by weight, area, and species were characterized for several closed canopy forests. Forest canopies were divided into four horizontal strata to develop predictive models for canopy gas exchange. Leaf and canopy layer carbon dioxide exchange rates were predicted using leaf nitrogen concentration, leaf mass per area, ozone exposure, predawn leaf water potential, photosynthetically active radiation, and vapor pressure deficit as driving variables. Direct measurements of leaf gas exchange were used to validate the components (subroutines) of the model. Net carbon dioxide exchange was simulated for canopy layers at 5-min intervals over a diurnal time course. Simulations of canopy CO:! exchange were made for a 30-m tall, mixed oak-maple forest under hypothetical ambient and greater-than-ambient ozone pollution regimes. Daily canopy net CO2 exchange was predicted for seven forest stands and compared with estimates of aboveground net primary production, N availability, leaf area index, and canopy N. Introduction There is increasing concern about the impacts on forests of atmospheric pollution, potential climate change, and long-term changes in site productivity caused by timber harvesting. These impacts must eventually be addressed at the stand, ecosys- tem, and global scales. Unfortunately, they are difficult to evaluate solely from a physiological perspective because of the large size and complexity of the systems (Jarvis and McNaughton 1986). They may also be difficult to assess at the ecosys- tem-level, because of the difficulty of finding forests differing only in the variable of interest. Certain environmental treatments may be imposed on otherwise uniform forest stands, e.g., fumigation with atmopheric pollutants, but this is rarely feasible because of logistic problems and extremely high costs. Thus, simulation modeling may provide a valuable approach when applied in concert with controlled physiolog- ical studies (e.g., Reich 1983) and comparative field studies (e.g., Oleksyn 1983). ’ Present address: School of Forest Resources, Pennsylvania State University, University Park, PA 16802, USA. 2 Present address: Department of Agronomy and Soil Science, University of Hawaii, 1910 East-West Road, Honolulu, HI 96822, USA.