Future Directions in Computer-Assisted Taxonomy <B>Future Directions in Computer-Assisted Taxonomy</B><P> </B><EM><B></B></EM><B>David Chesmore</B><BR> Department of Electronic Engineering, University of Hull, Huff HU6 7RX, UK<B></B><P> Computer-assisted taxonomy (CAT) is a term describing any computer-based system for identifying taxa and can be divided into two broad categories - computer-based identification based on keys (e.g. CABIKEY) and automated systems based on techniques such as image processing or acoustics. Recent advances in computing power and image/signal processing techniques are opening up new opportunities for the development of automatic detection and identification systems for a wide range of taxa. Examples of current species or groups include frogs, birds, phytoplankton, fish, Orthoptera, Hymenoptera, Lepidoptera, leaf-miners, weeds, algae and fungal spores using techniques as diverse as molecular analysis, acoustics, flow cytometry, radar, sonar, image processing and articifial neural networks. Potential applications for such systems include automated species logging for biodiversity and ecological studies (e.g. birds, insect pests), partial or fully automated sorting of large catches, discrimination of closely similar species, training aids and "smart" traps.<BR> The purpose of this paper will be to review the current state of CAT, especially automated systems and report on a two day inaugural meeting of the Bionet International working group for computer-assisted taxonomy (BIG-CAT) at the University of Wales, Cardiff in early July 1997. The meeting's purpose was to review the current range of computer applications and focus on likely future directions. The main outcomes were:<P> SCAT is multidisciplinary and requires significant collaboration between biochemists, biologists, ecologists, computer scientists and electronic engineers;<BR> Skey systems are generally well developed and successful use depends on the development of keys for specific taxonomic groups;<BR> Sautomated systems are mainly in the research phase and tend to be application specific;<BR> Sdemonstration systems should be developed for automated CAT.<P> The paper will show that computers can be of considerably benefit in a number of areas of systematics and CAT should be included within a future ESF programme.<P>