Agathis is a genus of tropical conifers found in lowland and montane forest, and occasionally in scrublands, throughout much of Malesia and in the southwest Pacific islands. By virtue of its long history of taxonomic confusion, its paucity of discrete macromorphological characters, the availability of appropriate molecular markers, and its intrinsic importance as a major timber-producing tree, it is an ideal genus within which to explore a number of issues relating to the principles, methods, and utility of different types of data in species delimitation.

Despite over forty years of heated discussion within systematics over the methods employed in the reconstruction of phylogeny, and much recent interest in “species concepts”, the methods for detection and delimitation of species – the other principal task of systematics – have received remarkably little attention.  Recent work by Wiens & Servedio (2000) on a statistical approach to species delimitation and to quantifying levels of confidence in delimitation decisions has offered a start, but the underlying mathematics needs to be extensively developed in order to deal with the complexities of continuous rather than discrete data, and to understand the spatial components to taxonomically significant variation

Previous accounts of the genus Agathis (de Laubenfels 1972, 1988; Whitmore 1980) have differed widely on species boundaries and on the extent and nature of inter- and intraspecific variation, and about a thousand herbarium specimens have been examined from major US and European herbaria in an effort to understand and identify potentially useful characters.  Morphometric analyses have enabled an understanding of the nature of variation in continuously varying vegetative characters in some parts of the genus, and offered an insight into current delimitations, but an insufficiency of specimens and inadequate geographic data makes it difficult to assess variation in cone size and shape in a statistically rigorous manner.  Additionally, a number of characters that are useful in the field – such as tree architecture – are not easily understood from herbarium specimens.

Accordingly, a particular subgroup of the genus – the putatively monophyletic New Caledonian group of species – is being intensively studied to understand patterns of both molecular and morphological differentiation across landscapes and between and within populations of putatively separate species, with two months fieldwork in New Caledonia this spring.  Chloroplast microsatellite markers originally developed for forestry work by the University of Queensland, and for analysis of the genus Araucaria by the team at RBG Edinburgh, as well as a number of others developed for conifers, are being assayed for their transferability to the Agathis genome and their variability, and a suite of morphometric characters that may discriminate between species and allow the estimation of tree-architectural parameters is being developed (discussion and data online soon).

I’ve so far been successful in amplifying a few informative sequences from a range of taxa in the genus: the following pairs of primers taken from the conifer population genetics literature have been investigated accross a range of Agathis species. Those which successfully amplify and show a degree of variation between and within species are marked A in the right hand column:


Sequences of primers














Anyone interested in more details of these microsatellites should . I’m hoping to test a range of microsatellites developed for other conifer species (these have largely been developed from pine genomes) for their transferability, variability and utility in the New Caledonian species of Agathis.

Site last updated 2005-09-20