A preliminary growth and yield model for Eucalyptus globoidea Blakely plantations in New Zealand.
Background: New Zealand's plantation forest industry is dominated by the exotic species radiata pine (Pinus radiata D. Don), which comprises approximately 90% of the net stocked area. However, there is interest in introducing new species to: (a) provide wood that is naturally decay-resistant as a substitute for wood treated with preservatives; (b) match species to the wide variety of environmental conditions in New Zealand; and (c) reduce reliance on P. radiata. Some Eucalyptus species are considered as potential alternatives to P. radiata, specifically those that can survive in resource-limited conditions and produce high quality wood. While Eucalyptus species are grown in plantations in many regions of the world, limited information is available on their growth in New Zealand. Eucalyptus globoidea Blakley is of particular interest and has been planted in trials throughout New Zealand. A complete set of preliminary growth and yield models for this species will satisfy the initial information requirements for diversifying New Zealand's plantation forest industry. Methods: A set of growth and yield models was developed and validated, based on data from 29 E. globoidea permanent sample plots (PSPs) located mostly in North Island and a few in South Island of New Zealand. Trees were measured at different time intervals in these plots, with height and diameter at breast height (DBH) ranging from 0.1-39.8 m and 0.1-62.3 cm, respectively. An algebraic difference approach (ADA) was applied to model mean top height, basal area, maximum diameter, and standard deviation of DBH. Non-linear regression equations were used to project stand volume and height-diameter relationship, and Reineke's stand density index (SDI) approach was employed to model mortality. Results: Mean top height, maximum diameter, and standard deviation of DBH were best fitted by Von Bertalanffy-Richards (SE=1.1 m), Hossfeld (SE=2.4 cm), and Schumacher polymorphic (SE=1.6 cm) difference equations, respectively. Basal area data were modelled with high precision (SE=6.9 m2 ha-1) by the Schumacher anamorphic difference equation. Reineke's SDI approach was able to explain the self-thinning as a reduction in the number of stems per hectare. Stand-level volume per hectare and height-diameter relationship models were precise when including site-specific variables with standard errors of 40.5 m3 ha-1 and 3.1 m, respectively. Conclusion: This study presents a set of preliminary growth and yield models for E. globoidea to project plot-level growth attributes. The models were path invariant and satisfied basic traditional mensurational-statistical growth and yield model assumptions. These models will provide forest growers and managers with important fundamental information about the growth and yield of E. globoidea.