Carbon dioxide efflux rates (CER) of stems and roots of overstory and understory black cherry (<i>Prunus serotina</i> Ehrh., BC), red maple (<i>Acer rubrum</i> L., RM) and northern red oak (<i>Quercus rubra</i> L., RO) trees were monitored over two growing seasons at two contrasting sites in the Great Smoky Mountains to investigate diurnal and seasonal patterns in respiration and to develop prediction models based on environmental and plant parameters. CER of small roots (d<0-8 mm) was measured with a newly developed system which allows periodic <i>in situ</i> measurements by using permanently installed flexible cuvettes. Temperature-adjusted CER of roots showed no diel variation. The moderate long-term changes occurred simultaneously in all species and size classes, suggesting that they were driven mostly by environmental factors. Mean root CER ranged from 0.5 to 4.0 nmol g⁻¹ d.w. s⁻¹. Rates were up to six times higher for fine roots (d<2.0 mm) than for coarse roots. CER (per unit length) of boles (d>10 cm) and twigs (d<2 cm) was related to diameter by the function lnCER = a+<i>D</i>·lnd, with <i>D</i> between 1.2 and 1.8. A new, scale-invariant measure of CER, based on <i>D</i>, facilitated comparisons across diameters. Q₁₀ varied with the method of determination, and it was higher in spring (1.8-2.5) than in autumn (1.4-1.5) for all species. Daytime bole CER often fell below temperature-based predictions, likely due to transpiration. The reduction (usually <10%) was less pronounced at the drier site. Twig CER showed more substantial (often >±50%) deviations from the predictions. Deviations were higher in the canopy than in the understory. Mean bole maintenance respiration (at 20°C and d=20 cm) was 0.66, 0.43 and 0.50 μMol m⁻¹, while the volume-based growth coefficient was around 5, 6 and 8 mol cm⁻³ for BC, RM and RO, respectively. In a controlled study, BC and RM seedlings were fumigated in open-top chambers with sub-ambient, ambient and twice-ambient levels of ozone. The twice-ambient treatment reduced stem CER in BC by 50% (p=0.05) in July, but there was no treatment effect in September or in RM. Ozone reduced root/shoot ratio and diameter growth in BC, and P_max in both species. / Ph. D.