Phenological process models based on carbon productivity at the leaf and branch levels, and their effects on Aspen tree architecture

ECOPHYS is a complex computer model that simulates the growth of hybrid poplar and aspen trees. Its primary purposes are to identify major physiologically-based factors that influence tree growth and to predict how interactions among them and with the external environment will affect growth (Isebrands et al., 2000). ECOPHYS' capabilities and submodels of various tree ecosystem processes continue to evolve. main objective of the Aspen FACE experiment is to examine, for predictive purposes, the effects of elevated CO 2 and O 3 on ecological interactions of northern hardwood forests (Dickson et al., 2000). In this paper we describe ECOPHYS phenological process models for bud break, leaf drop, branch abscission, bud set, and bud survival and vigor. These models are of the same type and include randomness and fading memory components. The leaf-drop algorithm described in Martin et al. (2001) is a predecessor of these models, though it does not include fading memory. Each of these models is based on local photosynthate productivity within the tree at the leaf or branch level, as well as genetic factors and field data Within ECOPHYS, leaf-level photosynthate productivity is based on leaf area and photosynthetic rate. Photosynthetic rate is determined by a number of factors including levels of O 3 and CO 2 , light, temperature, and RH, as well as N and H 2 O availability and leaf age. Our phenological process models are responsive to these variables via photosynthate productivity. In developing these models we have applied phenological and growth data from the Aspen FACE site received from Jud Isebrands and Evan McDonald through personal communications in 2002. The phenological data, organized according to aspen clone and treatment, include bud break, bud set, and end-of-season leaf senescence start dates and rates for 1999 and 2000. The growth data includes stem height and stem diameter measurements for 1998 and 2000, and leaf area measurements for 2000. Stem height and stem diameter modified averages for 1999 were obtained from Isebrands et al., (2001). We have used as input Aspen FACE environmental data obtained from Warren Heilman and Jaak Sober for the years 1998, 1999, and 2000. These were obtained via personal communications and from http://www.aspenface.mtu.edu. The bud-set algorithm described below assumes that branch productivity can predict bud set adequately for tree and patch level simulation of growth and competition. It applies to all branches except for short shoots which set bud right away. At …