Integrating rainfall data into site-specific modelling to improve forest productivity estimation of eucalypt hybrid species in coastal Zululand, South Africa

Abstract

This study aimed to develop a dynamic dominant height model and explore the effects of rainfall on the growth dynamics of Eucalyptus grandis × Eucalyptus urophylla. This species is grown for wood production in South Africa, and accurately estimating site index from dominant height is essential for effective forest management. The study used data from E. grandis × E. urophylla plantations located in coastal Zululand, South Africa. The region is characterised by a humid subtropical climate, which is ideal for the growth of this hybrid species. Dominant height growth data was collected from 75 permanent sample plots and 7 rainfall stations within the study area, across various site conditions to capture variability, ensuring comprehensive coverage of the local climate variability. The data set comprised repeat measurements, allowing for longitudinal analysis. A growth model was developed using non-linear fixed-effects and mixed-effects methods. To account for the serial correlation in the data, a continuous autoregressive error term was included in the models. The Lundqvist–Korf function was the most reliable candidate function. The mixed-effects model outperformed the fixed-effects model, with 40% greater precision. Rainfall significantly affects the growth of this species, with a 14% increase in precision when rainfall data was incorporated into the growth model. The model developed in this study can predict stand growth and update forest inventory data considering site-specific rainfall data, thus permitting it to optimise forest management strategies in changing climates.