GPP in Loblolly Pine: A Monthly Comparison of Empirical and Process ModelsThis article is part of a larger document. View the larger document here.
Monthly and yearly gross primary productivity (GPP) estimates derived from an empirical and two process based models (3PG and BIOMASS) were compared. Spatial and temporal variation in foliar gas photosynthesis was examined and used to develop GPP prediction models for fertilized nine-year-old loblolly pine (Pinus taeda) stands located in the North Carolina Sandhills. Foliar gas exchange in both the upper and lower thirds of crowns was monitored monthly for a year. Based on these data, empirical models were developed for the growing and non-growing seasons and upper and lower crown levels. Common empirical models include the variables photosynthetically active radiation (PAR), Ln(PAR), and VPD. Statistical differences in model estimates for crown positions and for both the growing and non-growing seasons indicated that the use of separate empirical models was appropriate for GPP estimations, yet simulated light-response curves yield similar rates. Monthly GPP estimates derived from empirical models were compared with process model predictions. Average monthly environmental data were applied to models to estimate GPP. Both process models predicted a greater relative GPP during the growing season (80 percent) compared with the empirical model (65 percent), while the opposite trend was apparent for the non-growing season. Monthly GPP variability was greater in the 3PG and BIOMASS predictions, appearing to reflect monthly temperatures and stand growth, while the empirical analysis predicted a relatively high contribution to yearly GPP during the non-growing season. Predicted GPPs for the entire year were 192.8, 142.8, and 192.4 mol C/m2 for the empirical, BIOMASS, and 3PG models, respectively.