The buzzword biodiversity has gained popularity, focussing on the spectrum in species of animals and plants and their importance in stable environments. The book, The Diversity of Life, by E. O. Wilson (1992) has attracted international attention and stimulated renewed awareness of the fragmented knowledge in the identity of living organisms. Taxonomy is the 'theory and practice of classifying organisms' (Mayr 1969). Taxonomy provides a framework for classification of individuals and/or populations of organisms according to their similarities. The basis for studies of biodiversity is, therefore, sound taxonomy.

Biological control seeks to reduce by biological means the abundance of weeds, usually exotic. Since most plants are stabilised in their native range by competition and natural enemies as components of overall diversity, biological control of any weed species is most likely to succeed if programs commence by understanding the plant in its natural environment. Knowledge of the identities of the plant, its relatives and the herbivores and pathogens associated with a plant, is essential for such projects but the biological control worker is often faced with a scarcity of this information.

For new biological control projects the identity of a target weed may not be well defined, its native distribution poorly known, and potential agents not identified or properly documented. Identification of a weed is necessary at all taxonomic levels to establish its origin, distribution (indigenous and exotic), relationships with other plants (especially for specificity testing with agents) or uses. Accurate identification is also needed to enable decisions relating to searching for, or selecting, new agents for further testing. A weed at first thought to be one taxon, is sometimes found to be a complex of closely-related taxa.

At an early stage in biological control projects, taxonomic studies are also required for natural enemies associated with the target weed or related plants, and to correlate identities of the natural enemies with their biological characteristics (biosystematics). However, it is doubtful if accurate predictions can be made as to whether an agent will be effective, even when the nature of the agent - plant interaction has been studied (See Prioritising Weed Biological Control Agents). Sometimes closely-related agents have an entirely different capacity to achieve biological control of a particular weed, for example the weevils used for control of salvinia (Sands and Schotz 1985). Further, an agent may be capable of achieving control of a particular form of weed but ineffective in controlling closely-related taxa (subspecies, varieties, forms) of the same plant species, though they can develop on them. Recognition of these different plant taxa in the native range may assist in the selection of biotypes of agent species (i.e. morphologically indistinguishable populations but with different biological characteristics) better adapted to each of the forms of weeds (Sands and Harley 1981), a not uncommon situation (e.g. different agents are better adapted to different varieties of lantana). Occasionally it may be necessary to assess the host interactions of biotypes of agents as if they were different species.

If the taxonomy of a weed can be first addressed at the exploratory stage, then the more complex process of selecting effective agents will be facilitated. Identifying herbivores associated with a weed in the native and exotic range requires experts for each potential agent group (usually for an insect family). Quite often the taxa are more complex than at first thought by specialists. Moreover, agents are often located and sometimes successfully utilised before their identity has been established. The discoveries of new agents and their biology by biological control workers have frequently led to the description of new species by taxonomists.

Only those classification procedures and taxonomic complexities most commonly encountered by biological control workers on weeds are discussed on the pages listed to the left.