Application of the process-based model BIOMASS to Eucalyptus globulus ssp. globulus plantations on ex-farmland in south western Australia : II. Stemwood production and seasonal growth

Abstract

In south western Australia 31 Eucalyptus globulus plantations were surveyed. The plantations had been established on previously farmed land and were from 2 to 10 years old. They were growing on differing soils and were located on a climatic gradient with mean annual rainfall ranging from 600 to 1450 mm. Statistical analysis showed that 89% of the variation in mean DBHOB and 85% of the variation in mean tree height were accounted for by multiple regression against the age of the plantation and mean annual rainfall. Plantation age, mean annual rainfall and stocking rate together accounted for 77% of the variation in mean annual increment of stemwood production. Five of the plantations were selected for more detailed studies ranging in age from 2 to 6 years in 1990 and with mean annual rainfall ranging from 590 to 1450 mm. At these plantations mean annual stemwood production from planting to 1993 varied from 5.1 to 25.9 t ha −1. Factors affecting growth were studied by applying the BIOMASS model. Measured seasonal patterns of growth were compared with simulated patterns of growth modelled using BIOMASS. The model allows for differences in weather conditions, soil water contents, leaf area index, tree size, canopy architecture and stocking rates. The climate in south western Australia is classified as Mediterranean with cool wet winters and hot dry summers. Therefore, the capacity of the soils to store water was expected to be crucial for survival and growth over summer. At two of the plantations storage of water in the soil was limited by its depth. Measured growth rates at these plantations were slower during summer than in winter. The seasonal pattern of growth simulated using BIOMASS was growth slowed in summer due to the decrease in availability of soil water and low relative humidity (high atmospheric vapour pressure deficits). Measurements of DBHOB appear to show slow growth from October to June, whereas simulations suggest that slow growth would be expected only from January to June. However, this difference may be an artefact due to dehydration and consequent stem shrinkage in early summer (October to January). A third plantation had a slightly saline watertable at 3 to 4 m depths and the other two plantations were on soils where soil water is available at depth. For these three plantations closest agreement between the measured and simulated seasonal patterns of growth was obtained by assuming that the availability of soil water was not limiting. Net C assimilation simulated using BIOMASS was allocated to growth of aboveground tree components estimated from measurements of increases in DBHOB and tree heights. Coefficients for allocation to foliage varied among plantations from 0.10 to 0.12, allocation to stems from 0.23 to 0.60 and allocation to branches from 0.05 to 0.12. Allocation to the below-ground component was estimated by difference to range from 0.18 to 0.60. Although the variation in allocation coefficients was large, similar values have been reported in other studies. Possible reasons for the variation among plantations are discussed.