Impact of the biological control agent Aceria lantanae (Cook) (Acari: Trombidiforms: Eriphyidae) on the invasive weed Lantana camara L. (Verbenaceae) in South Africa.

Master of Science in Entomology.

This study was conducted to determine the establishment, dispersal, performance and impact of a recently introduced flower-galling mite, Aceria lantanae (Cook) (Acari: Trombiformes: Eriophyidae) on the inflorescence and seed production of the invasive Lantana camara L. (Verbenaceae) in Limpopo, Mpumalanga, Gauteng and KwaZulu-Natal provinces of South Africa. The climate-matching programme CLIMEX was used to predict the distribution range of the mite on the African continent. Furthermore, the influence of some climatic factors (i.e., elevation, temperature, rainfall and relative humidity) and the suitability of different L. camara varieties were also investigated. Aceria lantanae established and persisted for more than 12 months at 58.6% of the release sites in Limpopo, Mpumalanga, Gauteng and KwaZulu-Natal provinces. Continuous surveys also showed that the mite had dispersed widely throughout the geographic range of L. camara in South Africa and Swaziland, with the highest dispersal rate of 40.6 km per annum recorded between the inland area of Nkwene (Swaziland) and the coastal area of Ncotshane (KwaZulu-Natal). The performance of A. lantanae varied among sites, provinces and seasons, with the infestation levels ranging from 2.7% to 97% per site. Inflorescence and seed production declined significantly by up to 86% and 96%, respectively, on lantana stands that were infested with A. lantanae in KwaZulu-Natal compared to the control stands. The CLIMEX model predicted that the climatic conditions for A. lantanae would range from suitable to highly suitable within the distribution range of L. camara in southern Africa. Although not statistically significant, there was a slight decline in A. lantanae infestation levels, with increasing elevation and annual rainfall. Infestation levels were somewhat higher at sites receiving between 600 and 1000 mm of rainfall per year, and decreased slightly as the annual rainfall exceeds 1000 mm. This study also found that infestation levels of A. lantanae were neither related to temperature nor relative humidity. Mite infestations differed significantly amongst the 10 tested varieties of L. camara. Highly preferred varieties included 017 Orange Red, 021 White Pink and 018 Dark Pink, with infestations ranging from 50.4% to 61.2%. Those which were moderately attacked by A. lantanae included 163 Light Pink, 021 Total Pink, 165 Light Pink, 015 Yellow White, 021 Pink and 015 White Yellow varieties, with infestations ranging from 7.8% to 21.4%. Variety 010 Dark Pink was completely rejected by the mite, with no infestations recorded during the study period. Furthermore, regression analysis showed that neither plant size nor inflorescence density influenced A. lantanae infestation levels. However, there was a significant increase in A. lantanae infestation on plants already infested by other lantana biocontrol agents. This study concluded that amongst all investigated parameters, varietal resistance was the major factor that influenced the sporadic establishments and overall performance of A. lantanae throughout the distribution range of L. camara in South Africa.