Taxonomy and autecology of the grey box (Eucalyptus moluccana roxb. s. lat.) - Myrtaceae

Gillison, Andrew Napier

Taxonomy and autecology of the grey box (Eucalyptus moluccana roxb. s. lat.) - Myrtaceae

Date

1976

Authors

Gillison, Andrew Napier

Abstract

Grey box forms part of a woodland complex that is restricted to the eastern part of Australia from 170 to 35° S. Lat. and occurs within a range of about 1 x 106 km2. It is not an important timber species although it has some significant uses. It is of indirect importance as it is related to lands of high agronomic value in New South Wales and is also a significant component of woodlands in Queensland. Apart from the geographic extent of the complex it is also comprised of a number of closely related taxa that have provided considerable taxonomic problems in the past due to their clinal nature. Before any ecological investigation could be commenced it was therefore vital to make certain that the species under consideration were genecotypically distinct so that management predictions could be applied with some certainty. This required a comprehensive taxonomic revision coupled with a controlled environment study to test genotype identity. For the first part, numerical techniques were used to handle the large number of closely related morphological attributes. Pattern analysis programs available on the CSIRO CYBER 76 involving classification and ordination procedures provided a background of taxonomic pattern within grey box that was used to assist the subjectively based revision. The advantage of such a method allowed some unexpected insight into morphological relationships that were clearly related to environmental variables. 13 The revision produced three new subspecies of E. moluccana, namely subsp. pedicellata, subsp. queenslandica and subsp. crassifolia, all of which follow a geoclimatic continuum. Classifications and the accompanying ordination of growth, morphological and chemical data from seedlings of grey box grown under a range of temperature maxima (150, 21°, 270, 33° and 36°C) were compared with those of the parents. A high correlation was obtained, indicating that at an infraspecific level there is support for the hypothesis of nonspecificity. A method of comparing ordinated centroids from the classifications is put forward. The structure of the thesis is arranged so that the 14 comparison of the seedling and parent characters via ordinationwere included as part of the revision, i.e. it precedes the phytotron investigation. This facilitated the use of the revised names in the growth studies in a way that would not otherwise have been possible. Having completed the revision and established that the genecotypes were classifiable by phenotypic characters it was then possible to proceed with the establishment of a range of representative field sites. The environmental study took into account community structure at a fine level and because of the largely ubiquitous floristic components included a minimal amount of floristic data. The ensuing classification revealed a distinct tropical versus temperate division that isolated grey box communities immediately north of the Queensland border and also north of the Tropic. An analysis of soil variables between contiguous communities showed no meaningful pattern. However climatic variables were highly correlated with the distribution of seedlings of the revised taxa. The continuing use of the same broad numerical techniques facilitated a quantitative comparison between data sets in almost every case. The additional use of canonical variate analysis and canonical correlation analysis also enabled a clearer insight into group structure of the different classifications and in addition to diagnostic computer programs assisted in reducing the numbers of attributes required to recover an accepted classification. As a follow-on to the environmental study a competition experiment between E. moluccana and E. crebra carried out in a controlled environment allowed some further inferences to be made about the field situation where E. crebra occupies poorer sites to the exclusion of grey box.