Thesis (MScAgric (Conservation Ecology and Entomology))--Stellenbosch University, 2008. / ENGLISH ABSTRACT: Ant infestations comprising the Argentine ant Linepithema humile (Mayr), common pugnacious ant Anoplolepis custodiens (F. Smith) and cocktail ant Crematogaster peringueyi Emery are a widespread pest problem in South African vineyards. Integrated Pest Management (IPM) programmes aimed at suppressing the problematic honeydew excreting vine mealybug Planococcus ficus (Signoret) (Hemiptera: Pseudococcidae) on grapes must include ant control to optimize the effectiveness and efficacy of mealybug natural enemies. If ants are eliminated, natural enemies are able to contain mealybugs below the Economic Threshold Level (ETL). Current strategies for ant control are limited and generally include the application of long term residual insecticides that are detrimental to the environment, labour intensive to apply and can disrupt natural biological control if applied incorrectly. A more practical method of ant control using low toxicity baits was therefore investigated. Field bait preference and bait acceptance assessments aimed at determining bait repellency and palatability, respectively, were carried out during spring, summer and autumn in three vineyards of the Cape winelands region during 2007/08. Five toxicants comprising gourmet ant bait (0.5%), boric acid (0.5%), fipronil (0.0001%), fenoxycarb (0.5%) and spinosad (0.01%) dissolved in 25% sugar solution were tested against a 25% sucrose solution control. Gourmet ant bait was significantly more preferred and accepted by all ant species than the other baits. Laboratory bait efficacy assessments using four insecticides (gourmet, boric acid & spinosad) at concentrations of 0.25; 0.5; 1; 2 and 4 times the field dose and fipronil at 0.015625; 0.03125; 0.0625; 0.125; 0.25 times the field dose were carried out. Results revealed that boric acid (2%), gourmet ant bait (2%) and fipronil (1.0 X 10-5%) exhibited delayed toxicity for L. humile and C. peringueyi while spinosad (0.01%) showed delayed action on L. humile. Field foraging activity and food preference tests were also carried out for the three ant species during 2007/08. Foraging activity trials revealed that vineyard foraging activity of L. humile is higher relative to A. custodiens and C. peringueyi. This means fewer bait stations are required for effective L. humile control making low toxicity baits a more affordable and practical method of controlling L. humile than the other two ant species. Food preference trials showed that L. humile and C. peringueyi have a high preference for sugar while A. custodiens significantly preferred tuna over other baits. However, all ant species had a preference for wet baits (25% sugar water, 25% honey, tuna & agar) as opposed to dry ones (fish meal, sorghum grit, peanut butter & dog food). This research concludes that low toxicity baits show potential in ant pest management and can offer producers with a more practical, economical and environmentally friendly method of ant control which is compatible with vineyard IPM programmes. / AFRIKAANSE OPSOMMING: Mierbesmetting wat uit die Argentynse mier Linepithema humile (Mayr), die gewone malmier Anoplolepis custodiens (F. Smith) en die wipstertmier Crematogaster peringueyi Emery bestaan, is ’n plaagprobleem wat wydverspreid in Suid-Afrikaanse wingerde voorkom. Programme vir geïntegreerde plaagbeheer (GPB) wat daarop gemik is om die wingerdwitluis Planococcus ficus (Signoret) (Hemiptera: Pseudococcidae) – wat ’n probleem is weens die heuningdou wat dit afskei – op druiwe te beheer, moet mierbeheer insluit om sodoende die uitwerking en doeltreffendheid van die witluis se natuurlike vyande die beste te benut. As miere uitgeskakel kan word, sal dit vir die natuurlike vyande moontlik wees om die witluis sodanig te beheer dat dit onder die ekonomiese drempelvlakke (EDV) bly. Huidige strategieë om miere te beheer, is beperk en sluit gewoonlik die toediening van insekdoders in wat lank neem om in die grond af te breek, wat skadelik vir die omgewing is, waarvan die toediening arbeidsintensief is en wat die natuurlike biologiese beheer kan versteur indien dit verkeerd toegepas word. Daarom is ’n meer praktiese metode ondersoek waar miere deur die gebruik van lae toksisiteit lokase beheer word. Ondersoeke na lokaasvoorkeure en lokaasaanvaarbaarheid in die praktyk, wat daarop gemik is om te bepaal of die lokaas onderskeidelik afstootlik en smaaklik bevind word, is oor lente, somer en herfs in drie verskillende wingerde in die Kaapse wynlandstreek gedurende die 2007/08-seisoen uitgevoer. Vyf gifstowwe, bestaande uit gourmet ant bait (0.5%), boorsuur (0.5%), fiproniel (0.0001%), fenoksiekarb (0.5%) en spinosad (0.01%) wat in ’n 25%-suikeroplossing opgelos is, is getoets teenoor ’n kontrole wat uit ’n 25%-sukrose-oplossing bestaan. Al die mierspesies het gourmet ant bait bo die ander lokase verkies en aanvaar. In die laboratorium is ondersoeke gedoen om die doeltreffendheid van die lokase te bepaal deur vier insekdoders (gourmet ant bait, boorsuur en spinosad) te gebruik in konsentrasies van 0.25; 0.5; 1; 2 en 4 keer die dosis in die praktyk en fiproniel teen 0.015625; 0.03125; 0.0625; 0.125; 0.25 keer die dosis in die praktyk. Resultate het getoon dat boorsuur (2%), gourmet ant bait (2%) en fiproniel (1.0 X 10-5%) vertraagde toksisiteit getoon het vir L. humile en C. peringueyi, terwyl spinosad (0.01%) ’n vertraagde uitwerking getoon het op L. humile. Toetse om kossoekaktiwiteite in die praktyk en die voedselvoorkeure van die drie mierspesies te ondersoek, is ook gedurende die 2007/08-seisoen gedoen. Proewe oor kossoekaktiwiteite het getoon dat hierdie aktiwiteite in die wingerd by L. humile hoër is in verhouding met A. custodiens en C. peringueyi. Dit beteken dat minder lokaasstasies nodig is om L. humile doeltreffend te beheer en lei daartoe dat lae toksisteit lokaas ’n beter manier is om L. humile te beheer as die ander twee mierspesies. Proewe oor voedselvoorkeure het aangedui dat L. humile en C. peringueyi ’n groot voorkeur toon vir suiker, terwyl A. custodiens ’n duidelike voorkeur vir tuna het. Alle mierspesies het egter ’n voorkeur vir nat lokaas (25% suikerwater, 25% heuning, tuna en agar), eerder as droë lokaas (vismeel, sorghumgruis, grondboontjiebotter en hondekos) getoon. Uit hierdie navorsing word afgelei dat lae toksisteit lokaas potensiaal toon in mierbeheer en dat dit produsente ’n meer praktiese, ekonomiese en omgewingsvriendelike metode van mierbeheer kan bied wat met GPB-programme in die wingerd versoenbaar is.

This research evaluated the new red imported fire ant (RIFA) bait Advion™. Advion™ contains the active ingredient indoxacarb which, following ingestion, must be metabolized into an N-decarbomethoxyllated metabolite known as JT333 in order to become acutely toxic. Laboratory experimentation was conducted to determine the most effective chemical form and concentration of indoxacarb, as well as the most appropriate grit size, for use in Advion™. The results indicated that Advion™ containing indoxacarb was more effective than Advion™ containing JT333, that 0.10%, 0.06%, and 0.045% were the most effective concentrations of indoxacarb, and that standard sized grit (~2 mm) was more effective than small sized grit (<1 mm). Field experimentation was conducted to determine the most effective concentration and quantity of Advion™, as well as the most effective placement of the bait, for obtaining maximum control of RIFA colonies via individual mound treatments. The results indicated that 10 g (2 Tbsp) of 0.045% Advion™ placed around the mound in a circle with a radius of 0.5-3.0 m was the most effective manner in which to treat individual RIFA mounds. Field experimentation was also conducted to compare the effectiveness of label-rate broadcast treatment with Advion™ to the effectiveness of both label-rate broadcast treatment with Amdro® and pre-baiting broadcast treatment with Advion™. Both pre-baiting broadcast treatment and label-rate broadcast treatment with Advion™ ultimately resulted in 98-99% RIFA colony mortality, which was significantly greater than the 87% colony mortality resulting from broadcast treatment with Amdro®. The 6.2 d LT90 for label-rate broadcast treatment with Advion™ was nearly one half that of the LT90 for pre-baiting broadcast treatment. Laboratory experimentation was conducted to determine whether RIFA workers were capable of metabolizing indoxacarb into the toxic metabolite, and the results clearly indicated that they were. Finally, field experimentation was conducted to evaluate the effectiveness of label-rate broadcast treatment with Advion™ at controlling multiple ant species. The results indicated that label-rate broadcast treatment with Advion™ resulted in control of the RIFA and Pogonomyrmex barbatus for at least 7 wk, and Monomorium pharaonis and Dorymyrmex pyramicus for some period of time between 3 and 7 wk.