364 Fourteenth Australian Weeds Conference Summary Prickly acacia, identified as a Weed of National Significance, now covers over seven million hectares in Australia. Biological control of this and many other weed species has not been achieved to date and there is now evidence that changes in the genetic diversity and fitness of biological control agents may be reducing the success of classical biological control efforts. This project uses a prickly acacia biological control agent as a test system to demonstrate that genetic impediments may reduce the effectiveness of current biocontrol methods. An approved host specific biological control agent, the geometrid moth Chiasmia assimilis (Warren) (Lepidoptera: Geometridae), was imported from South Africa and the genetic diversity and fecundity resulting from various rearing regimes were monitored over multiple generations. We measured changes in allele frequencies between group- reared lines of insects (the usual breeding practice) and isofemale line rearing methods. We provide evidence that there are deleterious effects from inbreeding on the biocontrol agent of prickly acacia. The genetic basis for selection, laboratory maintenance and release of insect biological control agents should receive considerable attention when developing biological control programs. Keywords Biological control, Acacia nilotica, prickly acacia, genetic impediments, isofemale lines, genetic diversity, Chiasmia assimilis. INTRODUCTION Biocontrol programs in Australia have historically had varying success with their use of insects as biocontrol agents (McFadyen 1998). As such, biocontrol scien- tists are looking to improve every stage of the process from insect selection through to breeding and release of agents. Microsatellites are a powerful technology for population genetic analysis, and will assist in the development and maintenance of genetically ‘healthy’ cultures of biological control agents. The importance of maintaining genetic diversity within the labora- tory cultures of biological control agents has been frequently emphasised (Bartlett 1985, Hopper et al. 1993, Roush and Hopper 1995, Spafford Jacob and Briese 2003). To date, little research has determined the best practices for the collection, breeding and ge- netic management of insects being reared for release and establishment as biological control agents. Chiasmia assimilis (Warren) is a geometrid moth that has been previously collected from Kenya and more recently from South Africa to control prickly acacia, Acacia nilotica subsp. indica (Benth.) Brenan (Mimosaceae). Kenyan collections of C. assimilis were reared using standard group rearing methodology and were not observed to establish in northern and cen- tral Queensland, Australia after three years of a mass release program. However South African collections have established in three coastal north Queensland sites. It is suspected that inbreeding during the rear- ing process, differences in local predacious fauna, or differences in climate may have been responsible for the failure of establishment of the Kenyan collections. The extremely host specific nature of C. assimilis and its ability to defoliate mature trees prompted us to recollect. Recent CLIMEX modelling suggested South Africa was also a suitable climate match to Australia (data not shown). Our aim was to collect C. assimilis from South Africa and maintain the maximum genetic diversity through to release using isofemale lines and hybridisa- tion between these lines immediately prior to release. At the same time we aim to replicate a standard group rearing methodology used in many biological control programs. Over multiple generations we monitored fecundity and population genetic diversity using five microsatellite markers (Wardill et al. 2004) to quantify changes in the allele frequencies between group and isofemale line rearing methodology. MATERIALS AND METHODS Field collections of larvae were made in February 2002 from four South African sites; Hoedspruit (Limpopo Province, 24.4666°S, 31.1025°E), Mkuzi (KwaZulu Natal Province, 27.6197°S, 32.1808°E), Songimvelo (Mpumalanga Province, 26.0408°S, 30.9258°E) and Ohrigstad (Northern Province 24.8133°S, 30.5536°E). Specimens were reared to moths, identified and used to start isofemale lines. At Ohrigstad, C. assimilis was not abundant. Only one isofemale line produced greater Investigating genetic diversity to improve the biological control process Trevor J. Wardill 1 , Glenn C. Graham 1 , Andrew Manners 1 , Julia Playford 2 , Myron Zalucki 1 , William A. Palmer 3 and Kirsten D. Scott 1 1 School of Life Sciences, The University of Queensland, Queensland 4072, Australia 2 Queensland Parks and Wildlife Service, PO Box 155, Brisbane Albert St., Queensland 4002, Australia 3 Queensland Department of Natural Resources, Mines and Energy, PO Box 36, Sherwood, Queensland 4075, Australia