Seasonal activity, habitat preferences and larval mortality of the leaf-mining fly Calycomyza eupatorivora (Agromyzidae), a biological control agent established on Chromolaena odorata (Asteraceae) in South Africa

Abstract

The leaf-mining fly Calycomyza eupatorivora Spencer (Diptera: Agromyzidae) has become widely established in the eastern regions of South Africa, following its release for the biological control of the invasive shrub Chromolaena odorata (L.) King and Robinson. This study was conducted to gain some insight into the impact of C. eupatorivora populations, by assessing their seasonal activity, habitat preferences, levels of leaf exploitation and extent of larval mortality in the field over a 1-year period. Leaf mining intensity was poorly synchronised with leaf availability, with leaf exploitation peaking at the end of the growing season of C. odorata. Although significantly more mines were recovered on plants growing in shaded situations, the percentage of available leaves that were exploited for mining was not significantly different between plants growing in shaded versus open situations. Overall, the levels of leaf damage were trivial with mines recovered from <5% of available leaves that were sampled during the study. Besides leaf-quality requirements, these low levels of leaf exploitation may have been influenced by high larval mortality which varied between 60 and 83%, depending on whether lower or higher estimates were used. These evaluations have verified the results of earlier laboratory studies which suggested that the impact of C. eupatorivora on mature populations of C. odorata in South Africa will be negligible.

Keywords:

• agent field impact
• larval mortality
• post-release evaluations
• triffid weed
• weed biological control

Acknowledgements

We thank the Botanical Education Trust, Tata Africa Scholarship fund and Working for Water Programme (Department of Environment Affairs, South Africa) for funding the first author, who undertook this study as part of an M.Sc. dissertation. The managers of the Sappi Cannonbrae Plantation are thanked for permitting this study on their premises. Statistical advice on the use of generalised linear modelling in SPSS Statistics 21 was provided by S.D. Johnson and H. Ramanand (both University of KwaZulu-Natal). Anonymous referees are thanked for their comments on the draft manuscript.