Thesis (MSc)--Stellenbosch University, 2011. / ENGLISH ABSTRACT: The Mediterranean fruit fly (medfly), Ceratitis capitata, is a highly invasive species throughout the world and considered as one of the most successful agricultural and economical pests. The increase of global trade in fruit and human travel combined with the biology of the medfly has allowed the species to spread from its proposed Afrotropical origin, to a number of locations throughout the world. In the Western Cape various control strategies have been implemented to control medfly populations, including insecticides and more environmentally-friendly techniques such as the Sterile Insect Technique (SIT). In order to be effective, however, an SIT program requires some knowledge of the population structure and the movement of individuals between pest-occupied sites. The identification of sites from which re-invasion is most likely to occur and knowledge regarding the primary routes through which pests are likely to re-establish are critical to ensure successful SIT programmes. To provide this important information to SIT and area-wide pest control programs in South Africa, sampling at two different spatial scales (regional- and fine-scale) in South Africa was undertaken. Regional scale sampling was done at 13 locations in the Western Cape and fine scale sampling was done at 13 locations within the Ceres-valley. All individuals were genotyped at 11 polymorphic microsatellite markers and selected individuals from the regional scale were sequenced for the mitochondrial gene COI. Our results show that populations at regional- and fine-scale in the Western Cape are characterized by high levels of genetic diversity (HEregional = 0.805; HEfine = 0.803). Little or weak population differentiation was detected at the regional- and fine-scales, suggesting overall high levels of gene flow among sampling locations. These findings were supported by coalescent based methods indicating sufficient levels of gene flow to prevent population differentiation between neighbouring (200m) and distant (350km) populations. However, natural dispersal in C. capitata has been shown to rarely exceed 10 km. As such, high levels of gene flow between distant populations are more likely the result of humanmediated dispersal, linked to the movement of fresh produce within South Africa. This high level of gene flow has important implications for pest management practices, as my results suggests that area-wide pest management should be undertaken at a regional scale, rather than on a farm or valley scale. My results are placed within a management framework, and I argue for more stringent control when fruit are transported within South Africa. Of particular interest for future studies is the investigation of gene flow at broader spatial scales (i.e. the whole of South Africa) and a comparison of the genetic diversity, population differentiation and gene flow patterns of C. capitata with that of Ceratitis rosa will be important to establish a successful pest management strategy in South Africa. / AFRIKAANSE OPSOMMING: Die Mediterreense vrugtevlieg (medvlieg), Ceratitis capitata, is ‘n indringerspesie wêreldwyd en word beskou as een van die mees suksesvolle ekonomiese en landbou peste. Die medvlieg het ‘n Afrotropiese oorsprong, maar die toename in wêreldwye handel en reis, gekombineer met die biologie van die medvlieg het gelei tot die verspreiding van die spesie na ‘n groot aantal bestemmings regoor die wêreld. Die Wes-Kaap provinsie van Suid-Afrika implementeer tans verskeie strategieë om medvlieg bevolkings te beheer. Hierdie strategieë sluit in die gebruik van plaagdoders sowel as meer omgewingsvriendelike tegnieke soos die Steriele Insektegniek (SIT). Om ‘n effektiewe SIT program te implementeer vereis basiese kennis ten opsigte van die genetiese struktuur van die bevolking sowel as van die beweging van individue tussen verskillende pesvoorkomsgebiede. Die identifisering van areas van waar herkolonisering mees waarskynlik kan plaasvind en kennis in verband met die primêre roetes waardeur pes spesies hervestig, is van kritiese belang om ‘n suksesvolle SIT program te verseker. Medvlieg individue is op twee verskillende ruimtelike skale (streeks- en plaaslike-skaal) versamel om die nodige inligting aan SIT en area-wye pes beheer programme in Suid-Afrika te verskaf. Streeks-skaal individue is by 13 lokaliteite regoor die Wes-Kaap versamel en plaaslike-skaal individue by 13 lokaliteite in die Ceres-vallei. Alle versamelde individue is vir 11 polimorfiese mikrosatelliet merkers gegenotipeer en DNS volgordebepaling van geselekteerde individue vanuit die streek-skaal is gedoen vir die mitochondriale geen COI. My resultate toon dat bevolkings op beide skale gekarakteriseer word deur hoë vlakke van genetiese diversiteit (HEstreeks = 0.805; HEplaaslik = 0.803) en geen of swak bevolkings differensiasie. Hierdie resultate dui daarop dat daar hoë vlakke van geenvloei tussen bevolkings is. Hierdie bevindinge word verder ondersteun deur metodes gebaseer op die statistiese eienskappe van die genealogiese verhouding tussen allele onder sekere mutasie en demografiese modelle, wat voldoende vlakke van geenvloei aandui tussen nabye (200m) sowel as verafgeleë (350km) bevolkings om bevolkings differensiasie te verhoed. Natuurlike beweging in C. capitata is egter selde meer as 10 km, sodanig kan die hoë vlakke van geenvloei toegeskryf word aan die verspreiding van individue met menslike hulp, spesifiek in die vervoer van varsprodukte in Suid-Afrika. Hierdie hoë vlak van geenvloei het verreikende implikasies vir pes beheer praktyke, omdat my resultate voorstel dat area-wye pes beheer onderneem moet word op ‘n streeks-skaal eerder as op ‘n plaas-tot-plaas of vallei wye area. Ek plaas my resultate in ‘n bestuursraamwerk, waarin ek streng beheer van vrugtevervoer in Suid-Afrika beklemtoon. Verdere navorsing moet fokus op die ondersoek van geenvloei op ‘n landswye skaal (hele Suid-Afrika) in C. capitata, sowel as die vergelyking van die genetiese diversiteit, bevolkings differensiasie en geenvloei patrone van C. capitata met die van Ceratitis rosa om ‘n suksesvolle pes beheer strategie vir Suid-Afrika te formuleer.

Várias espécies de moscas-das-frutas têm sido alvo de investigações visando sua criação massal, incluindo Ceratitis capitata (Wiedemann) (mosca-do-Mediterrâneo ou medfly) e, mais recentemente, Anastrepha fraterculus (Wiedemann) (mosca-das-frutas sul americana), sendo as espécies de maior importância econômica e quarentenária em várias áreas produtoras de frutas do Brasil, Uruguai, Argentina e Peru. Um dos fatores determinantes para que a Técnica do Inseto Estéril (TIE) seja aplicada contra uma determinada praga é a existência de métodos de criação massal da espécie. Para se alcançar a criação em larga escala dessas espécies de moscas-das-frutas, é necessário, inicialmente, o preparo de uma dieta larval adequada. Como a dieta de criação pode oferecer uma quantidade enorme de substâncias nutritivas, as larvas criadas artificialmente amadurecem mais cedo do que as selvagens, o que acelera a sucessão de gerações. Essas mudanças detectadas durante o estágio larval podem formar adultos com maiores reservas nutritivas, atingindo a maturidade sexual mais cedo. Existem várias biofábricas de mosca-das-frutas no mundo capazes de produzir machos estéreis que são utilizados para a erradicação ou supressão populacional em área-ampla. Visando-se diminuir os custos de produção dos insetos, pesquisas voltadas para a otimização e desenvolvimento de novas dietas não podem parar. Portanto, os objetivos deste trabalho foram determinar a adequação e densidade larval ideal das moscas C. capitata e A. fraterculus em novas dietas à base de farinha de milho (FM), bagaço de cana-de-açúcar (BC), pó de cenoura liofilizada (PC) e carragenina (MC), escolhendo a dieta mais eficiente e viável economicamente para produção de cada uma dessas espécies com base na quantidade e qualidade de insetos produzidos e no custo de produção. As densidades larvais ideais para a criação de C. capitata nas dietas larvais FM, BC e PC foram, respectivamente, 0,5; 2,0; e 1,0 mL de ovos/kg de dieta. A dieta larval com maior produtividade foi a dieta BC, produzindo uma média de 6.514 pupas/kg de dieta, seguida das dietas FM e PC, que produziram, respectivamente, uma média de 5.018 e 4.135 pupas por kg de dieta. A ordem decrescente de viabilidade econômica entre as três dietas larvais para C. capitata foi a mesma (BC>FM>PC), sendo necessário um investimento médio de R$ 471,29; R$ 482,26; e R$ 1.085,85, respectivamente, para a produção de 1 milhão de pupas. Os resultados com A. fraterculus mostraram que as densidades larvais ideais nas dietas FM e MC foram de 2,0/kg de dieta e 1,5 mL de ovos/L de dieta, respectivamente. A dieta larval PC foi considerada inviável devido à baixa produtividade e peso de pupas. Em termos de produtividade, a dieta larval MC (produção média de 22.451 pupas/L de dieta), foi mais eficiente do que a FM (6.688 pupas/kg de dieta). Quanto à viabilidade econômica, a dieta larval MC foi mais vantajosa, demandando um investimento médio de R$ 312,67 para a produção de 1 milhão de pupas, enquanto que a dieta FM custou R$ 1.266,36 para produzir a mesma quantidade de pupas / Several species of fruit flies have been the subject of research for their mass rearing, including Ceratitis capitata (Wiedemann) (Mediterranean fruit fly or medfly) and, more recently, Anastrepha fraterculus (Wiedemann) (South American fruit fly), being these species the ones with most economic and quarantine importance in several fruit producing areas of Brazil, Uruguay, Argentina and Peru. One of the requirements for the Sterile Insect Technique (SIT) is the existence of mass rearing methods of the target species. In order to achieve the large-scale rearing of these species of fruit flies, it is necessary initially to develop a suitable larval diet. As the rearing diet can offer a huge amount of nutrients, artificially reared larvae mature earlier than wild ones, what accelerates the succession of generations. These changes during the larval stage may form adults with higher nutritional reserves and they can reach sexual maturity earlier. There are several mass-rearing facilities for fruit flies in the world capable of producing sterile males that are used for population eradication and suppression in area-wide. In order to reduce insect production costs, the optimization or development of new diets cannot stop. Therefore, the objectives of this study were to determine the adequacy and optimal larval density of C. capitata and A. fraterculus flies in new diets based on corn flour (FM), sugarcane bagasse (BC), lyophilized carrot powder (PC) and carrageenan (MC), choosing the most efficient and less costly diet for the production of each of the species, based on the quantity and quality of insects reared and production costs. The optimal larval densities for C. capitata in larval diets FM, BC and PC were, respectively, 0.5; 2.0 and 1.0 mL of eggs/kg of diet. The larval diet that reached the highest productivity was the BC diet, giving an average of 6,514 pupae/ kg of diet, followed by FM and PC diets, which produced an average of 5,018 and 4,115 pupae per kg of diet, respectively. The decreasing order of economic viability among the three larval diets used was the same (BC>FM>PC), requiring a mean investment of R$ 471.29; R$ 482.26; and R$ 1,085.85, respectively, for the production of 1 million pupae. The results with A. fraterculus showed that the ideal larval densities in FM and MC diets were 2.0 eggs/kg of diet and 1.5 mL of eggs/L of diet, respectively. The PC larval diet was considered infeasible based on the low production and weight of pupae. In terms of productivity, the MC larval diet, which resulted in an average of 22,451 pupae produced per L of diet, was more efficient than FM, which produced a mean of 6,688 pupae per kg of diet. Considering the economic costs, the MC larval diet was also superior, requiring R$ 312,67.00 for the production of 1 million pupae, while the FM diet required R$ 1,266.36 to produce the same amount of pupae