Effects of stand structure and physiology on forest gas exchange: a simulation study for Norway spruce

The process-based simulation model STAND-FLUX describes canopy water vapor and carbon dioxide exchange based on rates calculated for individual trees and as affected by local gradients in photon flux density (PFD), atmospheric humidity, atmospheric carbon dioxide concentration, and air temperature. Direct, diffuse, and reflected PFD incident on foliage elements within compartments of individual trees (defined by vertical layers and a series of concentric cylinders centered on the trunk) is calculated for a 3-dimensional matrix of points. Foliage element gas exchange rates are based on estimates of carboxylation, RuBP regeneration, and respiratory capacities as well as the correlated behavior found between stomatal conductance and assimilation rate. Because of the difficulties associated with effective sampling and description of spatial variation in structure and leaf level gas exchange parameters for trees comprising the forest canopy, the significance for canopy water and carbon dioxide exchange of varied representations of tree foliage distribution and of physiology is examined. The additional interactive effects encountered due to changes in tree density and, thus, spatial aggregation or disaggregation of foliage is also studied. The analysis is conducted within the context of observed structural and physiological variation encountered in Norway spruce (Picea abies) stands in the Fichtelgebirge region of central Germany. Potentials for simplifying the three-dimensional canopy gas exchange model without sizable influence on canopy flux rates were small. A relatively large number of sample points within the tree crowns is necessary to obtain consistent calculations of flux rates because of the nonlinear relationship between PFD and net photosynthesis. Transpiration and net photosynthesis for stands with a low leaf area index (LAI) may be obtained from single tree estimates for each tree class weighted by class frequency, while 30 or more trees per class in differing relation to neighboring trees may be necessary to calculate reliable estimates of net photosynthesis in canopies with high LAI. The complexity in structure assumed for modeled trees was important, especially when overall canopy foliage area was either high or low due to spatial heterogeneity in clumping, e.g., potential canopy overlaps or side-lighting. Effects were greater for calculated net photosynthesis than for transpiration, reflecting higher sensitivity of net photosynthesis to differences in light distribution within individual trees. Accuracy in estimates of physiological parameters is equally important, and these characteristics have profound effects on estimated canopy gas exchange rates. While one-dimensional representations of canopy structure or approximations of tree physiological characteristics from other canopies or species may often be necessary in assessing vegetation/atmosphere exchanges, especially in the study of water balance of landscapes or regions, STANDFLUX provides a tool that can aid in evaluating the limitations of these simpler approaches.