Silva Fennica : quarterly issues : 42(1) : sa421035.htm

Julian C. Fox, Huiquan Bi & Peter K. Ades. 2008.

Modelling spatial dependence in an irregular natural forest

Silva Fennica 42(1): 35–48

The spatial dependence present in a natural stand of Eucalyptus pilularis (Smith) dominated mixed species forest was characterised and modelled. Two wildfires imposed a significant spatial dependence on the post disturbance stand. It was hypothesised that spatial variation in the intensity of the wildfires generated the observed structures. The influence of patch formation, micro-site variability and competitive influences were also noted in the residuals of a distance-dependent individual-tree growth model. A methodology capable of modelling these complicated patterns of observed dependence was sought, and candidates included the spatial interaction, direct specification and Papadakis methods. The spatial interaction method with a moving average autoregression was identified as the most appropriate method for explicitly modelling spatial dependence. Both the direct specification and Papadakis methods failed to capture the influence of competition. This study highlights the possibility that stand disturbances such as natural and artificial fires, insect and fungal attacks, and wind and snow damage are capable of imposing powerful spatial dependencies on the post disturbance stand. These dependencies need to be considered if individual tree growth models are to provide valid predictions in disturbed stands.

Keywords
correlogram, Eucalypt, growth modelling, moving average autoregression, Moran’s I, spatial autocorrelation

Addresses
Fox and Ades, School of Forest and Ecosystem Science, University of Melbourne, Burnley Campus, 500 Yarra Blvd, Richmond, Victoria 3121 Australia; Bi, Forest Resources Research, New South Wales Department of Primary Industries, PO Box 100, Beecroft, NSW 2119 Australia. E-mail jcfox@unimelb.edu.au

Received 27 April 2007 Revised 7.11.2007 Accepted 13 November 2007
ISSN 0037-5330

Full text

Comments