Hybrid growth models for Eucalyptus globoidea and E. bosistoana : explaining within and between site variability.
Degree GrantorUniversity of Canterbury
Degree NameDoctor of Philosophy
Plantation forests play a major role in satisfying many forestry needs such as demand for wood and different ecosystem services, which are projected to increase in the future. In New Zealand, the plantation forestry industry is dominated by Pinus radiata, which comprise approximately 90% of the net stocked area. Diversification of the New Zealand plantation forest estate by introducing new species is prudent, especially in arid parts of the country where Pinus radiata growth cannot achieve its full potential. Several Eucalyptus species are potential alternatives to Pinus radiata. However, there is currently very little information on their growth dynamics.
Forest growth and yield models are used to understand the growth dynamics of forest trees and are generally mensurational models for mature stands created from inventory data that span several years. Growth models of plantation trees at juvenile ages can generate information useful for plantation establishment, but such models are rarely created. Although mensurational growth and yield models project and create useful information to help management decisions, they provide little understanding of ecophysiological tree growth process. However, ecophysiological process information is important, especially in young plantations. This information can be created through process-based models, but these models are data intensive. Therefore, combining the two modelling approaches through hybridisation can give access to both mensurational and process- based modelling information, without violating basic growth and yield modelling assumptions.
Most existing growth and yield models are developed at stand level or individual tree-level, and productivity of the site is assumed to be homogenous due to silvicultural management and site preparation practices. However, in most sites growth is not homogenous throughout, especially juvenile plantation growth. Therefore, it is important to explore the factors affecting plantation growth within stands.
This doctoral thesis investigates and develops models that include within and between stand factors for juvenile Eucalyptus bosistoana and Eucalyptus globoidea by using a hybrid ecophysiological modelling approach. The study further tests and compares different hybridisation approaches. It concludes with a preliminary mature-stand mensurational growth and yield model for E. globoidea, developed from sparse available data by use of algebraic difference approach (ADA) equations.
The availability of high-resolution digital elevation models (DEMs) is inadequate for rural New Zealand, including the unproductive ex-pastoral lands where this study is sited. However, it is important to have high-resolution DEMs for hybrid ecophysiological study of growth and yield. Filed surveys conducted with global positioning system (GPS) receivers, can be an efficient, useful and simple method for creating high-resolution DEMs. This study reports on an optimisation procedure for producing DEMs by comparing three non-geostatistical interpolation procedures carried out with field collected GNSS data. Results show that the ANUDEM interpolation algorithm produced DEMs with the highest accuracy. The study also reports that data density influences final DEM resolution.
Within-stand height growth and survival proportion models indicate that topographic, wind exposure, morphometric protection, position index, and distance from ridge top significantly influenced juvenile height growth and survival proportion. These topographic indices were also found to be significant for between-site juvenile height growth and survival proportion, along with temperature. Overall, each of the final models had high precision and minimal bias, therefore they can predict juvenile tree height yield and survival proportion well.
Potentially useable light sum equations (PULSE) with augmented topographic indices were better than PULSE alone, or traditional hybridisation approaches, for explaining between-site growth. In addition to height growth and survival predictions, these hybrid models offer many other uses, including generating useful ecophysiological information, and they offer an improved understanding of tree growth processes.
Finally, the preliminary mensurational growth and yield models for E. globoidea were developed to project growth over time with high precision and minimal error. These models create useful growth dynamics information for forest managers, as well as suggesting future research avenues for growing Eucalyptus in New Zealand.