This paper describes a stand growth model, based on physiological processes, which incorporates a number of steps and procedures that have allowed considerable simplification relative to extant process-based models. The model, called 3-PG use of Physiological Principles in Predicting Growth., calculates total carbon fixed gross primary production; PG. from utilizable, absorbed photosynthetically active radiation fp.a.u.., obtained by correcting the photosynthetically active radiation absorbed by the forest canopy fp.a.. for the effects of soil drought, atmospheric vapour pressure deficits and stand age. PG is obtained from fp.a.u. and the canopy quantum efficiency, values of which are becoming available. The ratio of net PN. to gross primary production is emerging as relatively constant for trees. This eliminates the need to calculate respiration and is used to estimate PN—the net amount of carbon converted to biomass. 3-PG uses a simple relationship to estimate the amount of carbon allocated below ground and a procedure based on allometric ratios—widely available for many species and situations—to determine the allocation of carbon to foliage and stems and constrain tree growth patterns. The effects of nutrition are incorporated through the carbon allocation procedure; the amount of carbon allocated below ground will
increase with decreasing soil fertility. Recently acquired knowledge about the physiological factors causing decline in forest growth rates with age is used to model that decline. Changes in stem populations self-thinning. are derived from a procedure based on the y3r2 power law, combined with stem growth rates. The model requires weather data as input, works on monthly time steps and has been run for periods up to 120 years, producing realistic patterns of stem growth and stem diameter increments. The time course of leaf area index is realistic for a range of soil conditions and atmospheric constraints. 3-PG can be run from remotely-sensed estimates of leaf area index
coupled to weather data and basic, readily available information about soils and stand characteristics. It is being tested as a practical tool against forestry data from New South Wales, Tasmania, Victoria and New Zealand. Test results show excellent correspondence between stand growth measurements and simulated stem growth over 30 years. q1997 Elsevier Science B.V. Keywords: Forest model; Carbon balance; Partitioning; Physiological processes; Weather
