Reproductive biology and induced sterility as determinants for genetic control of mosquitoes with the Sterile Technique

Michelle EH Helinski Promotor Prof. Dr. M. Dicke, Hoogleraar in de Entomologie, Wageningen Universiteit

Co-promotor Dr. Ir. B.G.J. Knols, Universitair Docent, Leerstoelgroep Entomologie, Wageningen Universiteit

Promotiecommissie Prof. Dr. Ir. J.L. van Leeuwen, Wageningen Universiteit Prof. Dr. M. Schilthuizen, Nationaal Museum voor Natuurlijke Historie, Naturalis, Leiden Dr. J. Hendrichs, Joint FAO/IAEA Division of Nuclear Techniques in Food and Agriculture, Vienna, Austria Dr. J.A.J. Breeuwer, Universiteit van Amsterdam

Dit onderzoek is uitgevoerd binnen de onderzoeksschool Production Ecology and Research Conservation. Michelle EH Helinski

Reproductive biology and induced sterility as determinants for genetic control of mosquitoes with the

Proefschrift

Ter verkrijging van de graad van doctor op gezag van de rector magnificus van Wageningen Universiteit, Prof. Dr. M.J. Kropff, in het openbaar te verdedigen op vrijdag 20 juni 2008 des namiddags te vier uur in de Aula Helinski, Michelle E.H. (2008)

Reproductive biology and induced sterility as determinants for genetic control of mosquitoes with the Sterile Insect Technique.

PhD-thesis Wageningen University – with references – with summary in Dutch

ISBN 978-90-8504-955-5

ABSTRACT remains an important health issue in sub-Saharan , and new methods to reduce the are needed. In this thesis, the use of the Sterile Insect Technique (SIT) against the African malaria vector arabiensis was explored. The SIT relies on the releases of large numbers of sterilised males. If the sterile males are successful in competing for mates and the females use their sperm for fertilisation, the wild population is reduced. Ultimately, this can lead to a reduction in disease incidence. Sterilisation of the sperm cells occurs by ionising radiation, resulting in the of the developing embryo after fertilisation. Somatic cells are also damaged, which can lead to a reduced mating competitiveness of the males. In this thesis I investigated 1) the relationship between dose and induced sterility for pupal or adult stage irradiation, 2) sperm quantity and sperm length and the influence of irradiation, 3) the use of stable isotopes to determine mating in mosquitoes, 4) the incidence of multiple mating in relation to irradiation, 5) the fitness of irradiated males in terms of survival and mating competitiveness, and 6) the use of a field cage for mating studies, and the small-scale feasibility of the SIT in Sudan. Results showed that even though the dose-response curves between dose and induced sterility were largely similar for pupae and adults, irradiation of adults resulted in a better competitiveness compared to pupal irradiation. A negative relationship between dose and competitiveness was observed for pupal stage irradiation. In addition, radiation during the pupal stage affected the number of spermatozoa in the testes and the distribution of sperm lengths, but no impact on the incidence of multiple mating could be observed. Stable isotopes were used successfully to determine paternity in mating. Mating competitiveness of males irradiated as pupae could be improved by a three-fold increase in their number compared to un-irradiated males, but only for the partially-sterilising dose. The small-scale irradiation and transportation of in Sudan was feasible, and the preparation of the field cage for experiments successful. It is concluded that from a biological viewpoint the irradiation of adults would be recommended; however, the feasibility of adult irradiation on a large scale is questionable. The next steps would be to scale up irradiation procedures to accommodate much larger numbers of insects, and to determine male competitiveness in the semi-field system in Sudan. However, many other factors including mass rearing, sexing, and release methodology, are of importance for an SIT programme and only when all components are in place can the true feasibility of the SIT in Sudan be determined. CONTENTS 01: General introduction 10

Part 1 Part 3 Radiation effects Mating competitiveness in genetic control studies 02: A literature review on the irradiation of mosquitoes in the context of a Sterile Insect 07: Mating competitiveness of male Anopheles Technique programme 32 arabiensis mosquitoes irradiated with a partially- or fully-sterilising dose in small and 03: Radiation-induced sterility for pupal and large laboratory cages 110 adult stages of the malaria Anoph- eles arabiensis 52 08: The influence of late-stage pupal irradiation and increased irradiated: 04: Sperm quantity and size polymorphism in un-irradiated male ratio on mating un-irradiated and irradiated males of the ma- competitiveness of the malaria mosquito laria mosquito Anopheles arabiensis Patton 66 Anopheles arabiensis Patton 124

09: A stable isotope dual-labelling approach Part 2 to detect multiple insemination in un- irradiated and irradiated Anopheles arabiensis Stable isotopes in mating mosquitoes 136 behaviour research 10: Towards a Sterile Insect Technique field 05: Stable isotope-mass spectrometric release of Anopheles arabiensis mosquitoes in determination of semen transfer in malaria Sudan: Irradiation, transportation, and field mosquitoes 82 cage experimentation 148

06: Use of the 15N stable isotope as a semen label to detect mating in the malaria mosquito Anopheles arabiensis Patton 98 11: General discussion 164

Additional material

Summary 180 Samenvatting 186 Curriculum vitae 192 Publications 194 Dankwoord/Acknowledgements 198 PE & RC 202 01 General introduction 01

General introduction

his thesis concerns the genetic control of an African malaria mosquito with the Sterile Insect Technique, and in particular the sterility and competitiveness Tof irradiated insects. Before addressing the central research question and objectives, this introduction aims to give an overview of the malaria situation in Africa, including the vector species and their ecology, and contemporary methods of malaria control. It then describes the concept of genetic control, and ends with an overview of genetic control for mosquitoes.

11 Part of this Chapter was published as: Michelle EH Helinski, Badria El-Sayed, and Bart GJ Knols The Sterile Insect Technique: Can established technology beat malaria? Entomologische Berichten 2006, 66: 13-20. Malaria The African malaria vector system Chapter 01 350-500 Million cases of clinical malaria occur Not all Anopheles species transmit malaria. Of annually, 60% of which are in sub-Saharan Af- the 422 known species, 68 are recognised as ma- rica. Moreover, 80% of all attributed to laria vectors, of which some 40 are considered malaria occur in this region. In numbers, 1 mil- important (Service 1993). Moreover, some spe- lion Africans die of the disease each year, with cies are more competent vectors than others. the vast majority of deaths occurring among In Africa highly efficient vectors are found, due children below five years of age. Pregnant wom- to their competence as vectors of the parasite, anthropophilic biting behaviour, and longevity en are another major risk group; malaria can (Collins and Besansky 1994, Miller and Green- cause low birth weight and premature delivery wood 2002), in addition to suitable climatic con- (Rogerson et al. 2007). The impact of malaria on ditions that favour parasite development in the the economic situation of endemic countries is mosquito. high (Gallup and Sachs 2001), and the correla- For malaria transmission on the African tion between poverty and malaria clearly dem- continent, the Giles species onstrated (Sachs and Malaney 2002). complex is of great importance (White 1974). Malaria is a transmit- This complex consists of seven sibling species ted by female mosquitoes of the genus Anoph- (Hunt et al. 1998) that differ in their ecology, eles. The malaria parasite is a protozoan of the host preference, and epidemiological impor- genus . There are four species of tance (Ayala and Coluzzi 2005). Two species malaria parasites that can infect under within this complex, An. gambiae sensu stricto natural conditions: , P. and An. arabiensis, are regarded as major vectors vivax, P. ovale and P. malariae. The first two spe- of malaria. Another three species, An. merus, An. cies cause the most infections worldwide and P. melas and An. bwambae, are of localised impor- falciparum Welsh is by far the most lethal para- tance, whereas the sixth and seventh species, site, and responsible for the majority of deaths An. quadriannulatus species A and B are largely that occur in sub-Saharan Africa (Breman et al. zoophilic, and hence not considered as vectors. 2007). Recent studies have shown that malaria However, a recent study in Ethiopia observed infections in humans in Malaysia attributed to anthropophilic behaviour of An. quadriannulatus P. malariae were in fact caused by P. knowlesi, Theobald species B (Pates et al. 2006). Although a malaria parasite of Old World monkeys (Cox- An. gambiae s.s. Giles and An. arabiensis Patton exist sympatrically over much of their species Singh et al. 2008), and it has been proposed to range; the latter is more drought resistant (Gray recognise P. knowlesi as the fifth malaria and Bradley 2005, Rogers et al. 2002) and thus parasite (White 2008). extends into more arid environments (Lindsay To complete its lifecycle, the malaria et al. 1998, Coetzee 2004). As a result, An. arabi- parasite requires a vector (female Anopheles ensis is present in some areas as the sole vector mosquito) and a vertebrate host (e.g. humans). of malaria. This species is believed to be a rather The insect vector (definitive host) and the verte- uniform, panmictic (i.e. freely mating) species brate host (intermediate host) are connected by as cytogenetic data from Sudan demonstrated the need of female Anopheles mosquitoes to uti- (Petrarca et al. 2000). An. gambiae s.s., on the lise a blood meal in order to produce eggs. The other hand, shows extreme genetic heterogene- 12 periodic blood-feeding behaviour of anophelines ity. It is divided in 5 chromosomal forms named results in transmission of parasites between hu- Bamako, Bissau, Forest, Mopti and Savanna mans. (Bryan et al. 1982, Touré et al. 1998, Della Torre et al. 2002) that display assortative mating in areas of sympatry (Touré et al. 1998, Wondji et 01 General introduction

Hans Smid

Fig. 1.1. Male Anopheles mosquito. al. 2005). In addition, two distinct genotypes in has been the object of a number of (cytogenet- the ribosomal DNA, designated molecular forms ic) studies and two chromosomal forms named M and S, are described (Favia et al. 1997) that Kiribina and Folonzo have been described from assort independently from the chromosomal Burkina Faso (Guelbeogo et al. 2005). Partial forms (Wondji et al. 2005). From all of the above barriers to gene flow between the two forms it is evident that substantial gene flow barriers were observed (Guelbeogo et al. 2005), and it is exist between the different chromosomal and hypothesised that incipient speciation is ongo- molecular forms, making An. gambiae s.s. a dif- ing (Michel et al. 2005a). Across the African con- ficult target species for genetic control strate- tinent, microsatellite and mitochondrial DNA gies. data divides An. funestus in three subdivisions, The other major malaria vector in Af- e.g. eastern, western and central populations rica is An. funestus. An. funestus belongs to a (Michel et al. 2005b); but at least the central complex of nine (or possibly ten) species that population appears to be largely panmictic (Co- are morphologically very similar (Coetzee and huet et al. 2005). Fontenille 2004). Of these species, An. funestus Besides the An. gambiae and An. funes- sensu stricto Giles is the only species that plays tus groups, the An. nili and An. moucheti groups a significant role in malaria transmission (Co- are considered to be major vectors in West and huet et al. 2004). An. funestus s.s. has a wide Central Africa (Fontenille and Carnevale 2006), distribution, extending through the whole of and responsible for high levels of malaria trans- sub-Saharan Africa, and is highly anthropophilic mission (Antonio-Nkondjio et al. 2002). In addi- and endophilic. Until recently, An. funestus was tion, there are a large number of secondary vec- 13 rarely studied due to colonisation problems tors in Africa that can be of localised importance and the fact that polytene stud- in malaria transmission (Antonio-Nkondjio et al. ies were more difficult to perform (Guelbeogo et 2006). al. 2005). However, in recent years An. funestus Ecology of the An. gambiae lies 1988, Takken and Lindsay 2003). However, complex Chapter 01 across the African continent, differences in feed- ing behaviour of An. arabiensis are observed. In Female anophelines lay single eggs on the wa- western Africa, An. arabiensis tends to be more ter surface or on wet mud at the edge of shallow anthropophilic and endophagic, while in the pools. Depending on the species, different water eastern part of the continent, more zoophilic bodies are preferred, and females locate suitable and exophagic behaviour is observed (Tirados et oviposition sites by semiochemical, physical and al. 2006). An. arabiensis is therefore considered visual cues (Takken and Knols 1999). In western to be an opportunistic feeder (White 1974). After Kenya, An. arabiensis and An. gambiae s.s. select- a blood meal, the female locates a resting site ed similar larval habitats, and a preference for to digest the blood meal and develop the eggs. small, temporary habitats with and little or Females of An. gambiae s.s. tend to rest indoors no aquatic vegetation was observed (Gimnig et (endophilic; Gillies 1955, Lines et al. 1986), while al. 2001). In general one or two days after ovipo- An. arabiensis females are primarily observed to sition, the eggs hatch into the first larval stage. rest outdoors (exophilic; Lines et al. 1986, Tira- Larvae are collector-filter feeders and primarily dos et al. 2006). feed on and detritus at the air/ In most anophelines, like many other water interface (Merritt et al. 1992). Larvae go Dipteran species, mating is associated with through four moults before reaching the pupal swarming (Yuval 2006). These swarms have stage. The pupae are non-feeding and after 1-2 been observed at dusk and form over cer- days the adult mosquito emerges. Depending on tain landmarks described as ‘swarm markers’ food availability, larval density and temperature, (Takken and Knols 1999). Swarms are prima- development from egg to adult takes about 7-20 rily composed of males, and can reach consid- days (Schneider et al. 2000, Service 1977). Larval erable size. In An. freeborni Aitken (Yuval et al. populations can suffer high mortalities (Service 1993) it was observed that swarming males 1977, Edillo et al. 2004), due to the drying out of were larger in size compared to non-swarming larval habitats, and other factors such as preda- males, and the largest males enjoyed a greater tion, (Lyimo 1993), cannabalism (Koen- mating success than the smaller ones. However, raadt and Takken 2003), and turbidity (i.e. rain- in An. gambiae s.s. (Charlwood et al. 2002) and fall; Paaijmans 2007). An. funestus (Charlwood et al. 2003), male size After emergence, males (Fig. 1.1) and was not related to mating success. Swarming is females replenish energy reserves by feeding costly for the males: energy reserves of swarm- on plant nectars or honeydew (Clements 1999a, ing males are depleted during swarming (Yu- Foster 1995, Gary and Foster 2001, Impoinvil et val et al. 1994), and swarms are heavily preyed al. 2004, Manda et al. 2006), and females require upon (Yuval and Bouskila 1993). The mechanism a blood meal to develop eggs. The host-seeking as to how females locate swarming males is un- behaviour of females is odour-mediated (Takken known. It has been suggested that females are and Knols 1999) and chemical compounds such actively seeking swarm sites and respond to as fatty acids and carbon dioxide, originating similar cues (e.g. light intensity, swarm mark- from the skin and breath of the host, serve as ers) as males; therefore, contact will be made cues for host-seeking females (Takken and Knols eventually (Charlwood and Jones 1980). The 1999). Host preference and feeding behaviour involvement of sex pheromones produced by of the two main vectors in the An. gambiae swarming males acting as an aggregation phe- 14 complex, An. gambiae s.s. and An. arabiensis, romone has been suggested (Clements 1999b) is somewhat different. In general, An. gambiae but has not been confirmed. A female that en- s.s. is highly anthropophilic and endophagic. An. ters the swarm is recognised by her lower wing- arabiensis, on the other hand, exhibits more zoo- beat frequency (and possibly contact pherom- philic and exophagic behaviour (White 1974, Gil- ones (Clements 1999b)) and is rapidly mated in 01 General introduction

flight. Mating pairs usually leave the swarm in preferred combination contains , de- copula (Charlwood and Jones 1980). Cuticular rived from the plant annua. Although hydrocarbons (CHCs) have been observed to act artemisinin-based combination therapies (ACTs) as contact pheromones in mate recognition in are currently the best treatments available, they Drosophila (Ferveur 2005). In anophelines, the are 10 times more expensive than the conven- role of CHCs in mate recognition is not fully un- tional monotherapies (Mutabingwa 2005). A derstood, however a laboratory study indicated number of malaria-endemic countries have now that CHCs change after mating (Polerstock et al. adopted ACTs as their first or second line drug 2002) and data from field-collected specimens treatment. However, actual implementation is suggests a possible role for CHCs in species/form still ongoing in most of these countries (UNICEF recognition (Caputo et al. 2007). Because of the and WHO/RBM 2005). highly skewed male to female ratio in a swarm, males are competing strongly for females. Male Contemporary anophelines deposit a mating plug that forms Contemporary vector control methods include a temporary barrier against further copulation the use of insecticide-treated bednets (ITNs) (Charlwood and Jones 1979). The large major- and indoor residual spraying (IRS). Other efforts ity of females mates only once and the sperm focus on larval control mainly by larvicides such lasts a life-time, however in a small proportion as Bacillus sphaericus Meyer & Neide and B. thur- of wild females polyandry is observed (i.e. 2.5% ingiensis Berliner derivatives (e.g. including the in An. gambiae s.s.; Tripet et al. 2003). Although commonly used B. thuringiensis israelensis (Bti)), swarming is the conventional observation of the which are bacterial compounds that are toxic to mating behaviour of anophelines, in the absence mosquito larvae (Fillinger et al. 2003). Coverage of swarms (i.e. as assessed by the researcher) of ITNs is increasing in Africa, although large inseminated females were observed in An. dar- differences between countries are observed. A lingi Root (Lounibos et al. 1998). Also, in the East new generation of ITNs, the long-lasting insec- African setting, swarms of male An. gambiae s.l. ticidal bednets (LLINs) are now recommended have been observed in certain areas but not in due to their much greater life span. These nets others, and it is thus likely that besides swarm- stay effective for 4-5 years, while the conven- ing, other mating strategies are employed tional nets require re-impregnation every 6-12 (Takken and Knols 1999). months. However, the cost of an ITN remains a major constraint to ownership for a large pro- portion of Africans and voucher schemes are Malaria control being introduced to improve uptake (Magesa et al. 2005). Disturbingly, in many countries re- The strategy to control malaria is to prevent sistance of mosquitoes against the insecticides death, reduce illness, and decrease social and commonly used for IRS and bednet impregna- economic loss due to the disease (WHO 1994). tion is increasing (Hargreaves et al. 2000, Etang This is achieved by two main strategies: anti- et al. 2003, Etang et al. 2004, Vulule et al. 1994). malarial drug treatment and vector control. The great advantage of larvicides is that it con- trols the mosquito before it reaches the adult Drugs stage, thus preventing disease transmission (Kil- Due to widespread resistance of the main malar- leen et al. 2002). In addition, the commonly used ia parasite P. falciparum to the affordable drugs B. thuringiensis var. strains are environmentally and sulfadoxine- safe to non-target organisms and human expo- 15 (Fansidar®), other anti-malarial therapies are sure (Fillinger et al. 2003). In spite of these ad- urgently needed. Currently, the WHO recom- vantages, the main challenge in applying larval mends combination therapies in those coun- control in an African setting is finding and treat- tries where resistance has been reported. The ing the extensive number of larval breeding sites on a regular basis (Killeen et al. 2002). This has a rural African setting was successful in reducing

Chapter 01 resulted in a low uptake of larval control for Af- mosquito populations and thus potentially ma- rican Anopheles in rural areas, however, small- laria transmission (Scholte et al. 2005, Blanford scale studies in (peri)urban settings show good et al. 2005), but this technique is not yet avail- results and are promising (Fillinger and Lindsay able for large-scale application. A malaria vac- 2006, Vanek et al. 2006). cine remains beyond reach, although significant DDT, the most controversial research progress has been made over the last on the planet, has regained ground in the fight years (Girard et al. 2006). WHO aims for an effec- against malaria. In 2006, the WHO announced tive vaccine to be ready in 2025 (WHO 2006), but its support to use DDT as one of the insecticides without any guarantees of success, the search of choice in IRS campaigns (Mandavilli 2006, Sa- for alternative control strategies should con- dasivaiah et al. 2007). DDT is cheaper (Walker tinue. Genetic control has been used during the 2000, Sadasivaiah et al. 2007), more effective and last decades against a variety of insects and cur- longer lasting than the alternative insecticides rently enjoys renewed interest to control mos- available (Mandavilli 2006) and can be effective quitoes; particularly for potential application in in areas with high levels of pyrethroid resistance. urban ‘island’ settings and vector populations However, the impact of DDT on the environ- in geographically or ecologically isolated areas ment continues to be a major cause for concern (Dame and Curtis 1996, Curtis 2002b). and opposition, even though most evidence is The use of genetic control to manage or dated from the 1950-60’s when DDT was used eradicate species was independently con- in great quantities for agricultural practices. The ceived by three inventors in the 1930-40s (Curtis impact of DDT used indoors in small quantities 2006), albeit in different environments. These for malaria control can be assessed for instance days, genetic control is a broad term that can re- in Madagascar and Kwazulu Natal, South Af- fer to a number of methodologies. The Sterile In- rica, where it has been approved and used again sect Technique (SIT) relies on ionising radiation since 1993 and 2000, respectively (Curtis 2002a). or chemosterilisation to induce sexual sterility. Use of DDT in the largely pyrethroid-resistant Other methods of inducing sterility in a popu- province Kwazulu Natal (Hargreaves et al. 2000) lation include the release of hybrids, or insects decreased the incidence of malaria with 91% with translocations or other chromosomal rear- (Maharaj et al. 2005). However, the exclusive rangements (Knipling et al. 1968). In addition use of DDT has led to the emergence of DDT to the existing methods, new approaches in ge- resistance in An. arabiensis and An. quadrimacu- netic control are under development. Germ-line latus (Hargreaves et al. 2003), urging the need transformation (transgenesis) could potentially for strategies to manage and reduce insecticide be used to replace or complement the existing resistance (Coleman et al. 2006). Moreover, the genetic control methods (Thomas et al. 2000, Al- toxicity of DDT in humans is under much debate phey 2002, Alphey et al. 2002). Gene-drive strat- (Bouwman et al. 1990, Curtis 2002a, Chen and egies based on paratransgenesis, for instance Rogan 2003, Bouwman et al. 2006), and use of bacteria that induce cytoplasmic in- DDT for malaria control requires tight regulation compatibility (Dobson et al. 2002), have recently to prevent illegal diversion to agricultural use become accessible for use in species that do not (Curtis 2002a). naturally harbour these infections (Xi et al. 2005) and could be used to genetically control insect populations. 16 Genetic control Concept of genetic control Insecticide and has led to an in- The suppression of a wild population by genetic creased interest in other methods of malaria con- control relies on the introduction of sexually trol. The use of an entomopathogenic fungus in transmitted factors that reduce the reproduc- 01 General introduction

tion rate of the population, or introduce refrac- vantageous for agricultural pests to reduce cost, toriness. These sexually transmitted factors are avoid assortative mating (i.e. mating between usually introduced by males that have been released males and females), avoid (limited) mass-produced in a factory and released into economic losses and increase the overall ef- the wild population. The released males com- ficiency of the programme (Robinson 2002a), pete with the wild males for the insemination and 8) The majority of transgenic approaches of virgin females. If a released male is successful rely on efficient mechanisms to drive in transferring his sperm, sterility or refractori- the transgenes into the wild population (James ness to disease (i.e. depending on the method of 2005, Sinkins and Gould 2006). control) is introduced in the wild population. It is Contrary to earlier interpretation important to note that the sexually-transmitted (Knipling 1955), monogamy of the females is not factors that induce sterility or some other mech- a prerequisite for genetic control (Curtis 1985). In anism of control are present in the gametes of fact the largely monogamous of anophe- the released male and thus result in reduction or lines prevents re-mating of fertile females mi- refractoriness of the next generation (Robinson grating into the release area (Curtis 1985) which 2002a). strengthens the necessity for an isolated release area or effective barriers to prevent such move- Prerequisites ment. However, this is only valid if the sperm of For genetic control to be successful, certain irradiated males is equally competitive to fertile prerequisites are needed (Vreysen 1995) that sperm of non-treated males and the chance of are identical for the majority of genetic control a female being (re-) mated by a sterile male is strategies. 1) Colonisation of the target species big. should be feasible and mass production possi- ble at a reasonable cost to provide the required The Sterile Insect Technique number of insects to be released; 2) The com- One of the established genetic control meth- petitiveness of the released males needs to be ods is the Sterile Insect Technique (SIT) (Dyck adequate and there should be no major behav- et al. 2005). Historically, the SIT relies on ionis- ioural differences between the released and wild ing irradiation to induce sterility in the released population; 3) Population density of the target male (Knipling 1955, Dyck et al. 2005). The most species needs to be low or reduced prior to re- successful and well-known SIT project is the lease to make it economically feasible to obtain eradication of the New World Screwworm Co- the desired released-to-wild male ratio; 4) De- chliomyia hominivorax Coquerel from the USA, tailed information on the target population is re- Central America (Dame 1985, Snow 1988) and quired, such as spatial and temporal dynamics, (during an outbreak in 1989) Libya (Lindquist et mating behaviour, breeding sites, flight range, al. 1992). Another example is the eradication of etc.; 5) The method needs to be applied against the Glossina austensi Newstead from the total population in the target area or part of the island of Zanzibar (Vreysen et al. 2000). On- the population that can be isolated by natural or going SIT projects are taking place against the artificial barriers to exclude immigration from Mediterranean fruit fly (medfly) Ceratitis capi- neighbouring sites; 6) The target area should tata Wiedemann in Latin and Central America, preferably contain only one pest-species or vec- , parts of southern Europe and South tor of the disease targeted (i.e. similar to the Africa (Robinson 2002b). The largest medfly one being released); 7) The release of females is production facility is in El Piňo, Guatemala and not acceptable for those species where the fe- produces around two billion sterile male flies per 17 males are vectors of disease and/or cause biting week (IAEA 2006), primarily for use in California, nuisance (Robinson and Franz 2000). Females Guatemala and Mexico. Other pests targeted therefore need to be removed from the release with SIT include the Mexican fruit flyAnastrepha population. The removal of females is also ad- ludens Loew in Southern USA and Mexico (To- Figure 1.2. Effect of releases of chemosterilised Anopheles albi-

Chapter 01 manus at Lake Apastepeque (El Salvador), 1972. Open bars illus- trate the normal seasonal rise in female density in the year prior to release (1971). The filled bars show the female density in the year sterile releases took place (1972). Sterile males were released for a period of 5 months, for more de- tails see text (Figure modified af- ter Dame et al. (1981)).

ledo et al. 2004), the Dacus cucurbitae a 5-month period, 4.3 million sterile pupae were Coquillett in Japan (Ito et al. 2003), the onion fly released around Lake Apastepeque. Results were antiqua Meigen in The Netherlands (Tichel- promising and a substantial reduction in popula- er et al. 1974, Everaarts 2006), the tion size was observed (Dame et al. 1981, Lof- Cydia pomonella Linnaeus in British Colom- gren et al. 1974). A second, more extensive trial, bia (Bloem et al. 1997), and the pink bollworm located on the Pacific coast of El Salvador, took Pectinophora gossypiella Saunders in California place between 1977-79 (Lowe et al. 1980, Dame (Lindquist et al. 1990). et al. 1981) when up to 0.5 million sterile males or 1.25 million sterile male pupae were released Genetic control against mosquitoes daily. Complete control was not achieved due Scientists have long shared an interest to ge- to the immigration of females from outside the netically control mosquitoes. Mosquitoes are in- target area, despite the introduction of a barrier teresting candidates for genetic control because zone (which consisted of a zone covered with ar- of their relative ease of culture in captivity, short ea-wide insecticide spraying; Dame et al. 1981). development time that allows for the produc- Nevertheless, re-analysis of the data (Benedict tion of large numbers, and their importance as and Robinson 2003) on An. albimanus densities disease vectors. The majority of the work on in the release and a nearby control area empha- genetic control of mosquitoes was undertaken sises how successful the sterile males were in between 1950 and 1980. The induction of domi- preventing a normal seasonal rise in vector den- nant lethality by radiation or chemicals was per- sity (Fig. 1.2; Curtis 2006). In both these trials, haps the most researched area (Dame 1985), sterilisation was induced with chemosterilants. but other forms of genetic control, e.g. translo- Other release trials with anophelines cations or other chromosomal rearrangements, were performed. For An. quadrimaculatus Say, were undertaken. Benedict and Robinson (2003) pupae irradiated with 120 Gy and released as provide a review of the release programmes per- adults were not able to induce sterility in the tar- formed. Extensive genetic control studies were get population after prolonged releases, due to carried out with Culex and Aedes mosquitoes but behavioural differences as a result of the rearing 18 these fall outside the scope of this thesis. process (Weidhaas et al. 1962). Similar results The largest SIT release programme were observed in a subsequent study that used against an Anopheles vector (An. albimanus chemosterilisation (Dame et al. 1964), although Wiedemann) was performed in El Salvador and additional reasons for the failure of the releases was initiated in 1972 (Lofgren et al. 1974). Over (e.g. immigration of females from outside the 01 General introduction

release area, short period of releases) were put Another method to sex insects is by forward. Reduced competitiveness was report- genetic sexing. Genetic sexing strains (GSSs) ed in An. culicifacies Giles (i.e. after chemosteri- have been developed for various insects includ- lisation; Reisen et al. 1981) in one experiment, ing anophelines and they rely on the linkage but sufficient competitiveness was observed of a dominant selectable marker to the male if males carrying a translocation were released determining chromosome or locus. Linkage is (Baker et al. 1980). In An. gambiae (using ster- accomplished by radiation-induced transloca- ile hybrid offspring of An. gambiae s.s. and An. tions followed by crossing and screening of the melas Theobald crosses; Davidson et al. 1970) offspring. Resistance genes, e.g. temperature dispersal of released males was acceptable but sensitive lethal genes and insecticide-resistance male competitiveness was reduced and no im- genes, have been used as selectable markers. pact on egg-batch sterility could be observed. The process of creating a GSS is very time-con- Besides field releases, a number of anopheline suming and the system must be accurate and species were researched in laboratory settings stable under mass rearing conditions. Moreover, and detailed information on the relationship be- the method is species-specific. However, once tween dose and sterility, induced by irradiation established, these strains can be very valuable. or chemicals, was obtained. In addition, longev- A successful anopheline GSS was the MACHO ity and mating competitiveness were studied. A strain of An. albimanus used in the second trial thorough review of this literature for irradiation at the Pacific coast in El Salvador (Dame et al. studies is presented in Chapter two. 1981). This strain was created by linking an in- secticide (propoxur) resistance gene to the male Sex separation chromosome, and an inversion was induced to The release of only males is a prerequisite for suppress further recombination and thus sta- any genetic control programme for mosquitoes bilise the strain. Females were removed from (Robinson and Franz 2000), thus an efficient sex the population by treatment of the eggs with a separation system is required. Male mosquitoes discriminating dose of insecticide. The effective- are generally smaller than females, resulting ness of this sexing strain was 99.9% and large in smaller pupae and a shorter development numbers of male mosquitoes (1 million per day) time (Clements 1992), both of which can be ex- could be released (Lowe et al. 1980, Dame et al. ploited. However, in anophelines, mechanical 1981). sex separation of pupae based on size will not yield satisfactory separation, because the size New approaches in genetic control distributions of both sexes are overlapping. The In recent years, germ-line transformation has effectiveness that was achieved with mechani- received interest for (Thomas cal pupal separation of An. albimanus in the et al. 2000, Alphey 2002, Alphey et al. 2002, first El Salvador trial was only 85% (Lowe et al. Scott et al. 2002, Benedict and Robinson 2003, 1980), which is too low for any operational SIT Catteruccia et al. 2005). Germ-line transforma- programme. Male and female mosquitoes have tion relies on the insertion of an alien segment a distinctly different spectrum of wing beat fre- of DNA into the (transgenesis) creating quency (Clements 1999c) and they can easily be a genetically modified organism (GMO). Genetic differentiated and separated on the basis of this. transformation can be used to engineer strains However, upscaling a system in a way that it can of mosquitoes that are refractory to Plasmodium automatically recognise and sort males and fe- parasites (Ito et al. 2002), or generate a sterilis- males with little stress remains a challenge. In ing system that is based on dominant lethal 19 such a system, females would be removed only genes expressed in females only, referred to as at the very last stage of development, reducing RIDL (Release of Insects carrying Dominant Le- the capacity of the facility and increasing the thals) (Alphey et al. 2002). In RIDL, no irradiation costs of mass rearing. is required. So far, successful germ-line trans- formation has been accomplished in a number lular Wolbachia bacteria are obligate, maternally

Chapter 01 of insects using fluorescent markers to identify inherited endosymbionts found frequently in in- transformed individuals (Robinson et al. 2004). sects and other invertebrates. However, they do However, for a sexing strain, the marker needs not occur naturally in some important mosquito to be accompanied by a gene that is condition- species (Xi et al. 2005). Wolbachia bacteria can ally lethal to females, and this has not yet been potentially act as a gene-drive system for pop- accomplished for Anopheles, though several of ulation replacement strategies as mentioned such strains have been developed in Aedes ae- previously, or can be used to spread cytoplasmic gypti L. (L. Alphey pers. comm.). Another use- incompatibility, resulting in sterility, in a wild ful application of transgenesis is the ability to population (Sinkins and O’Neill 2000, Dobson et mark insects so that they can be recognised af- al. 2002). The use of Wolbachia to induce cyto- ter release. Marking can also result in a sexing plasmic incompatibility and suppress wild popu- method. Recently, a transgenic sexing strain lations has been successfully demonstrated in was developed in An. stephensi Liston in which Culex pipiens L. in the 1960s (Laven 1967, David- male mosquitoes express a fluorescent son 1974). However, the inability to transfer the in their gonads. Females do not express the Wolbachia infection to mosquito species without protein and can be removed from the popula- a natural infection halted the use of this tech- tion by automated screening of 3rd instar larvae nique for decades. Recently, Wolbachia was suc- (Catteruccia et al. 2005). Besides the transfor- cessfully transferred to laboratory populations mation of the vector, symbionts of the vector of Ae. albopictus Skuse (Xi et al. 2006) and Ae. are also targeted for transgenesis. In paratrans- aegypti (Xi et al. 2005), using embryonic microin- genesis, genetically modified bacteria are used jection, and is currently being evaluated to con- to deliver an effector molecule that inhibits trol Ae. polynesiensis Marks in a natural setting. the disease-causing in the insect gut (Beard et al. 2002). This technology has shown promise in controlling transmission of Chagas disease (Beard et al. 2001), but has not yet been Research objectives and problem developed for mosquitoes (Riehle et al. 2003). definition Although the technologies discussed above are very promising, the release of transgenic insects The main focus of research presented in this may be problematic as no regulatory framework thesis was built around a feasibility study for the exists for the introduction of transgenic insects control of African malaria mosquitoes with the into the wild (Knols et al. 2006, Knols and Louis Sterile Insect Technique (SIT). In 2004, the Inter- 2006, Knols et al. 2007). Moreover, no suitable national Atomic Energy Agency (IAEA) initiated mechanisms are presently available that can an integrated 5-year study to develop technolo- drive the transgenes or recombinant bacteria gies for controlling malaria mosquitoes with the through the wild population. Transposable ele- SIT. The goal of the project is to develop and ments, Wolbachia, and meotic drive have been evaluate all relevant components needed for proposed among others but all face major con- such an area-wide integrated approach to vec- strictions (James 2005, Sinkins and Gould 2006) tor control (Helinski et al. 2006). Some of the thus the search for suitable systems continues. experimental work is performed at the Agency’s Recently, a novel drive system relying on a ma- laboratories in Seibersdorf, Austria, and a pilot ternal-effect selfish genetic element known as project is under development in the Northern 20 Medea has been successfully demonstrated in State of Sudan, in collaboration with the Tropi- Drosophila (Chen et al. 2007) but this technology cal Medicine Research Institute (TMRI) in Khar- remains to be transferred to disease vectors. toum. Another promising strategy for genetic The work presented in this thesis fo- control is the use of Wolbachia bacteria. Intracel- cused specifically on the effects of ionising radi- 01 General introduction

ation on male Anopheles arabiensis mosquitoes. dramatically over the last years (Hood and Knols Irradiation induces dominant lethal mutations in 2007). the germ cells of males and when sperm carry- Even though laboratory based studies ing such a mutation is used to fertilise an egg, can give valuable information needed for SIT the embryo dies (LaChance 1967, Curtis 1971). programmes, validation of these findings under Ionising radiation also affects somatic cells, field conditions is important (Knols et al. 2002). which may result in a reduced competitiveness Simulation of field conditions can be performed of the insect (Proverbs 1969). The success of an in semi-field cages, and some work was- per SIT programme largely depends on the ability formed to develop methods for the determina- of the released males to successfully inseminate tion of male fitness in these cages. In addition, wild females, and thus it is crucial to reduce the the small-scale feasibility of the SIT programme adverse effects of irradiation to a minimum (To- in the designated field site in Northern State, ledo et al. 2004, Dyck et al. 2005, Parker and Me- Sudan was determined. hta 2007). In general, damage to the germ and somatic cells increases with dose, and the opti- The specific objectives of the work presented mal radiation dose for an SIT programme should here were to: be chosen in such a way that it balances induced 1. Develop methods for the small-scale sterility with competitiveness (Parker and Me- irradiation of pupae and adults, and hta 2007). The balance between these two im- perform radiation studies in An. arabi- portant factors, sterility and fitness, has been a ensis to determine the level of sterility major theme throughout the work performed induced for a range of doses. in this thesis. Besides the effects of radiation on 2. Perform a detailed study on sperm mating competitiveness and sterility, the influ- characteristics in An. arabiensis, and ence of radiation on male biology, and in partic- determine the effects of irradiation on ular sperm quantity and size polymorphism, was these. studied. Little is known about sperm quality and 3. Develop methods to study the mating quantity as determinants of male reproductive behaviour of Anopheles using stable success in Anopheles and some of these aspects isotopes. were studied. 4. Determine the mating competitiveness Male fitness in genetic control pro- of irradiated insects in a laboratory en- grammes is usually studied in terms of mating vironment under different conditions. competitiveness, i.e. how well irradiated males 5. Test the small-scale feasibility of an are able to compete against un-irradiated males SIT programme in the designated re- for females. However, the mating behaviour of lease site and develop methods for the anophelines is difficult to study because- mat determination of male fitness in semi- ing generally takes place at dusk under low light field systems. conditions (Charlwood and Jones 1979). Further- more, Anopheles mosquitoes mate in swarms The outline of the thesis is as follows: and mating couples only last for 10-20 seconds Chapter 2 In this chapter a literature re- (Charlwood and Jones 1979, 1980). One of the ob- view on the sterilisation of mosquitoes by irra- jectives of this thesis was to develop novel tools diation is presented. to study mosquito mating behaviour. Stable iso- Chapter 3 Essential for all subsequent topes were chosen as they are naturally occur- experiments was the establishment of a dose- ring in the environment, are not radioactive and response curve of induced sterility versus dose 21 thus safe to use. Besides these attributes, stable for An. arabiensis. The two potential develop- isotopes are not species-specific, and they have mental stages for irradiation were studied (i.e. become more accessible as prices have dropped pupae or adult), and the effects of irradiation on factors such as emergence, longevity and mat- ing ability are described. References

Chapter 01 Chapter 4 All genetic control strategies rely on male sperm to inseminate the female Alphey, L. 2002. Re-engineering the sterile insect and thus introduce sterility or refractoriness. In technique. Insect Biochem. Mol. Biol. 32:1243-1247. this chapter sperm quantity and sperm length Alphey, L., C. B. Beard, P. Billingsley, M. Coetzee, polymorphism in An. arabiensis, and the influ- A. Crisanti, C. Curtis, P. Eggleston, C. Godfray, J. ence of radiation is studied. Hemingway, M. Jacobs-Lorena, A. A. James, F. C. Chapter 5 The use of stable isotopes to Kafatos, L. G. Mukwaya, M. Paton, J. R. Powell, W. study mating in Anopheles is introduced, and the Schneider, T. W. Scott, B. Sina, R. Sinden, S. Sinkins, suitability of 13C as a semen-label for An. arabien- A. Spielman, Y. Touré, and F. H. Collins 2002. Malaria sis is addressed. control with genetically manipulated insect vectors. Chapter 6 In this chapter, research on Science 298:119-121. the use of 15N as an additional semen-label is Antonio-Nkondjio, C., P. Awono-Ambene, J-C. Toto, presented. J-Y. Meunier, S. Zebaze-Kemleu, R. Nyambam, C. Chapter 7 The effects of different devel- S. Wondji, T. Tchuinkam, and D. Fontenille 2002. opmental stages (pupae or adult), doses (high or High malaria transmission intensity in a village close low), and cage types (small or large) on the mat- to Yaounde, the Capital city of Cameroon. J. Med. ing competitiveness of sterilised mosquitoes is Entom. 39:350-355. presented in this chapter. Antonio-Nkondjio C., C. H. Kerah, F. Simard, P. Chapter 8 The work presented in this Awono-Ambene, M. Chouaibou, T. Tchuinkam, chapter aimed to improve the competitiveness and D. Fontenille 2006. Complexity of the malaria of pupal-irradiated males by irradiating pupae vectorial system in Cameroon: contribution of just prior to eclosion and by using an increased secondary vectors to malaria transmission. J. Med. ratio of irradiated males. Entomol. 43:1215-1221. Chapter 9 In this chapter, the use of a Ayala, F. J., and M. Coluzzi 2005. Chromosome stable isotope dual labelling approach to deter- speciation: humans, Drosophila, and mosquitoes. mine multiple insemination events in competi- Proc. Natl. Acad. Sci. U. S. A. 102 Suppl 1:6535-6542. tion experiments in An. arabiensis is explored. Baker, R. H., W. K. Reisen, R. K. Sakai, H. R. Rathor, Chapter 10 The small-scale feasibility of K. Raana, K. Azra, and S. Niaz 1980. Anopheles the Sterile Insect Technique in a field situation in culicifacies: Mating behavior and competitiveness Sudan was assessed in some pilot experiments. in nature of males carrying a complex chromosomal These experiments as well as the preparation aberration. Ann. Entomol. Soc. Am. 73:581-588. and use of a semi-field cage for fitness experi- Beard, C. B., E. M. Dotson, P. M. Pennington, S. ments are presented. Eichler, C. Cordon-Rosales, and R. V. Durvasula Chapter 11 This chapter provides a gen- 2001. Bacterial symbiosis and paratransgenic control eral discussion. In this chapter, the results of of vector-borne Chagas disease. Int. J. Parasitol. all chapters are discussed and integrated, spe- 31:621-627. cifically in the context of the Sterile Insect Tech- Beard, C. B., C. Cordon-Rosales, and R. V. Durvasula nique feasibility study in Sudan. 2002. Bacterial symbionts of the triatominae and their potential use in control of Chagas disease Acknowledgement transmission. Annu. Rev. Entomol. 47:123-141. We thank M. Benedict for constructive com- Benedict, M. Q., and A. S. Robinson 2003. The first ments on an earlier version of this Chapter. releases of transgenic mosquitoes: an argument for 22 the sterile insect technique. Trends Parasitol. 19:349- 355. Blanford, S., B. H. Chan, N. Jenkins, D. Sim, R. J. Turner, A. F. Read, and M. B. Thomas 2005. Fungal 01 General introduction

pathogen reduces potential for malaria transmission. Mozambique. Malar. J. 2:2. Science 308:1638-1641. Chen, A., and W. J. Rogan 2003. Nonmalarial infant Bloem, S., K. A. Bloem, and L. S. Fielding 1997. deaths and DDT use for malaria control. Emerg. Mass-rearing and storing codling moth larvae in Infect. Dis. 9:960-964. diapause: a novel approach to increase production for Chen, C. H., H. Huang, C. M. Ward, J. T. Su, L. V. sterile insect release. J. Entomol. Soc. B. C. 94:75-81. Schaeffer, M. Guo, and B. A. Hay 2007. A synthetic Bouwman, H., R. M. Cooppan, A. J. Reinecke, and maternal-effect selfish genetic element drives P. J. Becker 1990. Levels of DDT and metabolites population replacement in Drosophila. Science in breast milk from Kwa-Zulu mothers after DDT 316:597-600. application for malaria control. Bull. World Health Clements, A. N. 1992. Growth and Development, Organ. 68:761-768. pp. 150-170. The biology of mosquitoes, Volume 1. Bouwman, H., B. Sereda, and H. M. Meinhardt Development, nutrition and reproduction. Chapmann 2006. Simultaneous presence of DDT and pyrethroid & Hall, London. residues in human breast milk from a malaria endemic Clements, A. N. 1999a. Feeding on plant sugars, area in South Africa. Environ. Pollut. 144:902-917. pp. 403-432. The Biology of Mosquitoes, Volume 2: Breman, J. G., M. S. Alilio, and N. J. White. 2007. Sensory reception and behaviour. CABI Publishing, Defining and defeating the intolerable burden of Wallingford. malaria III. Progress and perspectives. Am. J. Trop. Clements, A. N. 1999b. Mating, pp. 360-402. The Med. Hyg. 77(6_suppl.):vi-xi. Biology of Mosquitoes, Volume 2: Sensory reception Bryan, J. H., M. A. Di Deco, V. Petrarca, and M. and behaviour. CABI Publishing, Wallingford. Coluzzi 1982. Inversion polymorphism and incipient Clements, A. N. 1999c. Flight, pp. 287-332. The speciation in Anopheles gambiae s.str. in The Gambia, Biology of Mosquitoes, Volume 2: Sensory reception West Africa. Genetica 59:167-176. and behaviour. CABI Publishing, Wallingford. Caputo, B., F. R. Dani, G. L. Horne, S. N’Fale, A. Coetzee, M. 2004. Distribution of the African malaria Diabate, S. Turillazzi, M. Coluzzi, C. Costantini, A. vectors of the Anopheles gambiae complex. Am. J. A. Priestman, V. Petrarca, and T. A. della Torre 2007. Trop. Med. Hyg. 70:103-104. Comparative analysis of epicuticular profiles of Coetzee M., and D. Fontenille 2004. Advances in the sympatric and allopatric field populations ofAnopheles study of Anopheles funestus, a major vector of malaria gambiae s.s. molecular forms and An. arabiensis from in Africa. Insect Biochem. Mol. Biol. 34:599-605. Burkina Faso (West Africa). Insect Biochem. Mol. Biol. Cohuet, A., I. Dia, F. Simard, M. Raymond, and D. 37:389-398. Fontenille 2004. Population structure of the malaria Catteruccia, F., J. P. Benton, and A. Crisanti 2005. An vector Anopheles funestus in Senegal based on Anopheles transgenic sexing strain for vector control. microsatellite and cytogenetic data. Insect Mol. Biol. Nat. Biotechnol. 23:1414-1417. 13:251-258. Charlwood, J. D., and M. D. R. Jones 1979. Mating Cohuet, A., I. Dia, F. Simard, M. Raymond, F. behaviour in the mosquito Anopheles gambiae s.l. I. Rousset, C. Antonio-Nkondjio, P. H. Awono- Close range and contact behaviour. Physiol. Entomol. Ambene, C. S. Wondji, and D. Fontenille 2005. Gene 4:111-120. flow between chromosomal forms of the malaria Charlwood, J. D., and M. D. R. Jones 1980. Mating vector Anopheles funestus in Cameroon, Central in the mosquito, Anopheles gambiae. s.l. II. Swarming Africa, and its relevance in malaria fighting. Genetics behaviour. Physiol. Entomol. 5:315-320. 169:301-311. Charlwood, J. D., J. Pinto, C. A. Sousa, C. Ferreira, Coleman, M., B. Sharp, I. Seocharan, and J. and V. E. Do Rosario 2002. Male size does not affect Hemingway 2006. Developing an evidence-based mating success (of Anopheles gambiae in São Tomé). decision support system for rational insecticide 23 Med. Vet. Entomol. 16:109-111. choice in the control of African malaria vectors. J. Charlwood, J. D., R. Thompson, and H. Madsen Med. Entomol. 43:663-668. 2003. Observations on the swarming and mating Collins, F. H., and N. J. Besansky 1994. Vector behaviour of Anopheles funestus from southern biology and the control of malaria in Africa. Science 264:1874-1875. della Torre, A., C. Costantini, N. J. Besansky, A.

Chapter 01 Cox-Singh, J., T. M. E. Davis, K. Lee, S. S. G. Shamsul, Caccone, V. Petrarca, J. R. Powell, and M. Coluzzi A. Matusop, S. Ratnam, H. A. Rahman, D. J. Conway, 2002. Speciation within Anopheles gambiae-the glass and B. Singh 2008. Malaria is half full. Science 298:115-117. in humans is widely distributed and potentially life Dobson, S. L., C. W. Fox, and F. M. Jiggins 2002. threatening. Clin. Infec. Dis. 46:165-171. The effect of Wolbachia-induced cytoplasmic Curtis, C. F. 1971. Induced sterility in insects. Adv. incompatibility on host population size in natural and Reprod. Physiol. 5:120-165. manipulated systems. Proc. Biol. Sci. 269:437-445. Curtis, C. F. 1985. Genetic control of insect pests: Dyck, A., J. Hendrichs, and A. S. Robinson 2005 growth industry or lead balloon. Biol. J. Linn. Soc. (eds.). The Sterile Insect Technique: Principles and 26:359-374. Practice in Area-Wide Integrated Pest Management. Curtis, C. F. 2002a. Should the use of DDT be revived Springer, Dordrecht. for malaria vector control? Biomedica 22:455-461. Edillo, F. E., Y. T. Touré, G. C. Lanzaro, G. Dolo, and Curtis, C. F. 2002b. Possible ways of using transgenic C. E. Taylor 2004. Survivorship and distribution of mosquitoes for malaria and dengue control and risk immature Anopheles gambiae s.l. (Diptera: Culicidae) assessment, pp. 165-175. Beijing, Peking University. 7th in Banambani village, Mali. J Med Entomol 41:333- International Symposium on Biosafety of Genetically 339. Modified Organisms, 10-16 October. Etang, J., L. Manga, F. Chandre, P. Guillet, E. Fondjo, Curtis, C. F. 2006. Review of previous applications R. Mimpfoundi, J. C. Toto, and D. Fontenille 2003. of genetics to vector control, pp. 33-43. In C. Louis, Insecticide susceptibility status of Anopheles gambiae and B. G. J. Knols (eds.), Bridging laboratory and s.l. (Diptera: Culicidae) in the Republic of Cameroon. J. field research for genetic control of disease vectors. Med. Entomol. 40:491-497. Frontis, Springer. Etang, J., F. Chandre, P. Guillet, and L. Manga 2004. Dame, D. A., D. B. Woodard, H. R. Ford, and D. E. Reduced bio-efficacy of permethrin EC impregnated Weidhaas 1964. Field behaviour of sexually sterile bednets against an Anopheles gambiae strain with Anopheles quadrimaculatus males. Mosq. News 24:6- oxidase-based pyrethroid tolerance. Malar. J. 3:46. 14. Everaarts, T. C. 2006. De Steriele-Insecten-Techniek Dame, D. A., R. E. Lowe, and D. L. Williamson 1981. tegen de uienvlieg. Entomol. Berichten 66:21-23. Assessment of Released Sterile Anopheles albimanus Favia, G., T. A. della Torre, M. Bagayoko, A. and Glossina morsitans morsitans, pp. 231-248. In J. Lanfrancotti, N. Sagnon, Y. T. Touré, and M. B. Kitzmiller, and T. Kanda (eds.), Cytogenetics and Coluzzi 1997. Molecular identification of sympatric genetics of vectors. Elsevier Biomedical, Amsterdam. chromosomal forms of Anopheles gambiae and further Dame, D. A. 1985. Genetic control by sterilized evidence of their reproductive isolation. Insect Mol. mosquitoes, pp. 159-172. In R. Chapman, R. Barr, D. Biol. 6:377-383. E. Weidhaas, and M. Laird (eds.), Biological Control of Ferveur, J. F. 2005. Cuticular hydrocarbons: their Mosquitoes. American Mosquito Control Association, evolution and roles in Drosophila pheromonal Bull 6. communication. Behav. Genet. 35:279-295. Dame, D. A., and C. F. Curtis 1996. The potential Fillinger, U., B. G. J. Knols, and N. Becker 2003. use of the sterile insect technique and other genetic Efficacy and efficiency of new Bacillus thuringiensis control methods for control of malaria-transmitting var israelensis and Bacillus sphaericus formulations mosquitoes, pp. 1-27. IAEA, Vienna. against Afrotropical anophelines in Western Kenya. Davidson, G., J. A. Odetoyinbo, B. Colussa, and Trop. Med. Int. Health 8:37-47. J. Coz 1970. Field attempt to assess the mating Fillinger, U., and S. W. Lindsay 2006. Suppression of 24 competitiveness of sterile males produced by crossing exposure to malaria vectors by an order of magnitude 2 member species of the Anopheles gambiae complex. using microbial larvicides in rural Kenya. Trop. Med. Bull. World Health Organ. 42:55-67. Int. Health 11:1629-1642. Davidson, G. 1974. Genetic control of insect pests. Fontenille, D., and P. Carnevale 2006. Malaria and Academic, New York. dengue vector biology and control in West and Central 01 General introduction

Africa, pp. 91-99. In B. G. J. Knols and C. Louis (eds.), 2006. The Sterile Insect Technique: Can established Bridging Laboratory and Field Research for Genetic technology beat malaria? Entomol. Berichten 66: 13- Control of Disease Vectors. Frontis, Wageningen. 20. CHAPTER 1. Foster, W. A. 1995. Mosquito sugar feeding and Hood-Nowotny, R. C., and B. G. J. Knols 2007. Stable reproductive energetics. Annu. Rev. Entomol. 40:443- isotope methods in biological and ecological studies 474. of arthropods. Entomol. Exp. Appl. 124: 3-16. Gallup, J. L., and J. D. Sachs 2001. The economic Hunt, R. H., M. Coetzee, and M. Fettene 1998. The burden of malaria. Am. J. Trop. Med. Hyg. 64:85-96. Anopheles gambiae complex: a new species from Gary Jr., R. E, and W. A Foster 2001. Effects of Ethiopia. Trans. R. Soc. Trop. Med. Hyg. 92:231-235. available sugar on the reproductive fitness and IAEA 2006. International Database on Insect vectorial capacity of the malaria vector Anopheles Desinfestation and Sterilization (IDIDAS). http:// gambiae (Diptera: Culicidae). J. Med. Entomol. 38:22- www-ididas.iaea.org/IDIDAS/default.htm. 28. Impoinvil, D.E., J. O. Kongere, W. A. Foster, B. Gillies, M. T. 1955. Studies in house leaving and N. Njiru, G. F. Killeen, J. I. Githure, J. C. Beier, A. outside resting of Anopheles gambiae Giles and Hassanali, and B. G. J. Knols 2004. Feeding and Anopheles funestus Giles in East Africa. Part II. The survival of the malaria vector Anopheles gambiae on exodus from houses and the house resting population. plants growing in Kenya. Med. Vet. Entomol. 18:108- Bull. Entomol. Res. 45:375-387. 115. Gillies, M. T. 1988. Anopheline mosquitoes: vector Ito, J., A. Ghosh, L. A. Moreira, E. A. Wimmer, and behaviour and bionomics. In W. H. Wernsdorfer, and M. Jacobs-Lorena 2002. Transgenic anopheline I. Sir McGregor (eds.), Malaria. Principles and Practice mosquitoes impaired in transmission of a malaria of Malariology. Churchill Livingstone, Edinburgh. parasite. Nature 417:452-455. Gimnig, J. E., M. Ombok, L. Kamau, and W. A. Ito, Y., H. Kakinohana, M. Yamagishi, and T. Hawley 2001. Characteristics of larval anopheline Kohama 2003. Eradication of the melon fly,Bactrocera (Diptera: Culicidae) habitats in Western Kenya. J. Med. cucurbitae, from Okinawa, Japan, by means of the Entomol. 38:282-288. sterile insect technique, with special emphasis on the Girard, M. P., Z. H. Reed, M. Friede, and M. P. role of basic studies. J. Asia-Pacif. Entomol. 6:119- Kieny 2006. A review of human vaccine research and 129. development: Malaria. Vaccine 25:1567-1580. James, A. A. 2005. Gene drive systems in mosquitoes: Gray, E. M., and T. J. Bradley 2005. Physiology of rules of the road. Trends in Parasitology 21:64-67. desiccation resistance in Anopheles gambiae and Killeen, G. F., U. Fillinger, and B. G. J. Knols 2002. Anopheles arabiensis. Am. J. Trop. Med. Hyg. 73:553- Advantages of larval control for African malaria 559. vectors: Low mobility and behavioural responsiveness Guelbeogo, W. M., O. Grushko, D. Boccolini, P. A. of immature mosquito stages allow high effective Ouedraogo, N. J. Besansky, N. F. Sagnon, and C. coverage. Malar. J. 1:8. Costantini 2005. Chromosomal evidence of incipient Knipling, E. F. 1955. Possibilities of insect population speciation in the Afrotropical malaria mosquito control through the use of sexually sterile males. J. Anopheles funestus. Med. Vet. Entomol. 19:458-469. Econ. Entomol. 48:459-462. Hargreaves, K., L. L. Koekemoer, B. D. Brooke, R. H. Knipling, E. F., H. Laven, G. B. Craig, R. Pal, J. B. Hunt, J. Mthembu, and M. Coetzee 2000. Anopheles Kitzmiller, C. N. Smith, and A. W. Brown 1968. funestus resistant to pyrethroid insecticides in South Genetic control of insects of importance. Africa. Med. Vet. Entomol. 14:181-189. Bull. World Health Organ. 38:421-438. Hargreaves, K., R. H. Hunt, B. D. Brooke, J. Knols, B. G. J., B. N. Njiru, E. M. Mathenge, W. Mthembu, M. M. Weeto, T. S. Awolola, and R. Mukabana, J. C. Beier, and G. F. Killeen 2002. 25 M. Coetzee 2003. Anopheles arabiensis and An. MalariaSphere: A greenhouse-enclosed simulation quadriannulatus resistance to DDT in South Africa. of a natural Anopheles gambiae (Diptera: Culicidae) Med. Vet. Entomol. 17:417-422. ecosystem in western Kenya. Malar. J. 1:19. Helinski, M.E.H., B. El-Sayed and B.G.J. Knols Knols, B. G. J., R. Hood-Nowotny, H. Bossin, G. Franz, A. Robinson, W. R. Mukabana, and S. K. 297.

Chapter 01 Kemboi 2006. GM sterile mosquitoes—a cautionary Lounibos, L. P., D. C. Lima, and R. Lourenco-de- note. Nat. Biotechnol. 24:1067-1068. Oliveira 1998. Prompt mating of released Anopheles Knols, B. G. J., and C. Louis 2006 (eds.). Bridging darlingi in western Amazonian Brazil. J. Am. Mosq. Laboratory and Field Research for Genetic Control of Control Assoc. 14:210-213. Disease Vectors. Frontis, Wageningen. Lowe, R. E., D. L. Bailey, D. A. Dame, K. Savage, Knols, B. G. J., H. C. Bossin, W. R. Mukabana, and and P. E. Kaiser 1980b. Efficiency of techniques for A. S. Robinson 2007. Transgenic mosquitoes and the the mass release of sterile male Anopheles albimanus fight against malaria: Managing technology push in Wiedemann in El Salvador. Am. J. Trop. Med. Hyg. a turbulent GMO world. Am. J. Trop. Med. Hyg. 77(6_ 29:695-703. suppl.):232-242. Lyimo, E. O. 1993. Life history aspects of Anopheles Koenraadt, C. J. M., and W. Takken 2003. Cannibalism gambiae s.l. immature stages: breeding activity, and predation among larvae of the Anopheles gambiae density and survival, pp. 47-64. In The bionomics complex. Med. Vet. Entomol. 17:61-66. of the malaria mosquito Anopheles gambiae sensu LaChance, L. E. 1967. The induction of dominant lato in southeast Tanzania. PhD thesis, Wageningen lethal mutations in insects by ionizing radiation and University. chemicals as related to the sterile male technique of Magesa, S. M., C. Lengeler, D. De Savigny, J. insect control, pp. 617-650. In J. W. Wright, and R. Pal E. Miller, R. J. Njau, K. Kramer, A. Kitua, and A. (eds.), Genetics of Insect Vectors of Disease. Elsevier, Mwita 2005. Creating an “Enabling Environment” for Amsterdam. taking insecticide-treated nets to national scale: the Laven, H. 1967. Eradication of Culex pipiens fatigans Tanzanian experience. Malar. J. 4:34. through cytoplasmic incompatibility. Nature 216:383- Maharaj, R., D. J. Mthembu, and B. L. Sharp 2005. 384. Impact of DDT re-introduction on malaria transmission Lindquist, D. A., B. Butt, H. U. Feldmann, R. E. in KwaZulu-Natal. S. Afr. Med. J. 95:871-874. Gingrich, and A. Economopoulos 1990. Current Manda, H., L. C. Gouagna, W. A. Foster, R. R. status and future prospects for genetic methods of Jackson, J. C. Beier, J. I. Githure, and A. Hassanali insect control or eradication, pp. 69-88. In J. E. Casida 2006. Effect of discriminative plant-sugar feeding (ed.), and Alternatives. Elsevier Science on the survival and fecundity of Anopheles gambiae. Publishers B.V. Malar. J. 6:113. Lindquist, D. A., M. Abusowa, and M. J. Hall 1992. Mandavilli, A. 2006. Health agency backs use of DDT The New World screwworm fly in Libya: a review of against malaria. Nature 443:250-251. its introduction and eradication. Med. Vet. Entomol. Merritt, R. W., R. H. Dadd, and E. D. Walker 1992. 6:2-8. Feeding behavior, natural food, and nutritional Lindsay, S. W., L. Parson, and C. J. Thomas 1998. relationships of larval mosquitoes. Annu. Rev. Mapping the ranges and relative abundance of the Entomol. 37:349-376. two principal African malaria vectors, Anopheles Michel, A. P., W. M. Guelbeogo, O. Grushko, B. J. gambiae sensu stricto and An. arabiensis, using climate Schemerhorn, M. Kern, M. B. Willard, N. Sagnon, data. Proc Biol Sci 265:847-854. C. Costantini, and N. J. Besansky 2005a. Molecular Lines, J. D., E. O. Lyimo, and C. F. Curtis 1986. Mixing differentiation between chromosomally defined of indoor and outdoor-resting adults of Anopheles incipient species of Anopheles funestus. Insect Mol. gambiae and Anopheles funestus in coastal Tanzania. Biol. 14:375-387. Bull. Entomol. Res. 76:171-178. Michel, A. P., M. J. Ingrasci, B. J. Schemerhorn, M. Lofgren, C. S., D. A. Dame, S. G. Breeland, D. E. Kern, G. Le Goff, M. Coetzee, N. Elissa, D. Fontenille, 26 Weidhaas, G. Jeffery, R. Kaiser, H. R. Ford, M. J. Vulule, T. Lehmann, N. Sagnon, C. Costantini, and D. Boston, and K. F. Baldwin 1974. Release of N. J. Besansky 2005b. Rangewide population genetic chemosterilized males for the control of Anopheles structure of the African malaria vector Anopheles albimanus in El Salvador. 3. Field methods and funestus. Mol. Ecol. 14:4235-4248. population control. Am. J. Trop. Med. Hyg. 23:288- Miller, L. H., and B. Greenwood 2002. Malaria-a 01 General introduction

shadow over Africa. Science 298:121-122. Insect transgenesis and its potential role in agriculture Mutabingwa, T. K. 2005. Artemisinin-based and human health. Insect Biochem. Mol. Biol. 34:113- combination therapies (ACTs): Best hope for malaria 120. treatment but inaccessible to the needy! Acta Trop. Rogers, D. J., S. E. Randolph, R. W. Snow, and S. I. 95:305-315. Hay 2002. Satellite imagery in the study and forecast Paaijmans, K., M. O. Wandago, A. K. Githeko, and of malaria. Nature 415:710-715. W. Takken 2007. Unexpected high losses of Anopheles Rogerson, S. J., V. Mwapasa, and S. R. Meshnick gambiae larvae due to rainfall. PloS ONE 2:e1146. 2007. Malaria in : linking immunity and Parker, A., and K. Mehta 2007. Sterile insect pathogenesis to prevention. Am. J. Trop. Med. Hyg. technique: a model for dose optimization for improved 77(6_suppl.):14-22. sterile insect quality: Fla. Entomol. 90:88-95. Sachs, J., and P. Malaney 2002. The economic and Pates, H. V., W. Takken, C. F. Curtis, and H. Jamet social burden of malaria. Nature 415:680-685. 2006. Zoophilic Anopheles quadriannulatus species B Sadasivaiah, S., Y. Tozan, and J. G. Breman 2007. found in a human habitation in Ethiopia. Ann. Trop. Dichlorodiphenyltrichloroethane (DDT) for indoor Med. Parasitol. 100:177-179. residual spraying in Africa: how can it be used Petrarca, V., A. D. Nugud, M. A. Ahmed, A. M. Haridi, for malaria control? Am. J. Trop. Med. Hyg. 77(6_ M. A. Di Deco, and M. Coluzzi 2000. Cytogenetics of suppl.):249-263. the Anopheles gambiae complex in Sudan, with special Schneider, P., W. Takken, and P. J. McCall 2000. reference to An. arabiensis: relationships with East Interspecific competition between sibling species and West African populations. Med. Vet. Entomol. larvae of Anopheles arabiensis and An. gambiae. Med. 14:149-164. Vet. Entomol. 14:165-170. Polerstock, A. R., S. D. Eigenbrode, and M. Scholte, E. J., K. Ng’habi, J. Kihonda, W. Takken, J. Klowden 2002. Mating alters the cuticular K. Paaijmans, S. Abdulla, G. F. Killeen, and B. G. hydrocarbons of female Anopheles gambiae sensu J. Knols 2005. An entomopathogenic fungus for stricto and (Diptera: Culicidae). J. Med. control of adult African malaria mosquitoes. Science Entomol. 39:545-552. 308:1641-1642. Proverbs, M. D. 1969. Induced sterilization and Scott, T. W., W. Takken, B. G. J. Knols, and C. Boëte control in insects. Annu. Rev. Entomol. 14:81-102. 2002. The ecology of genetically modified mosquitoes. Reisen, W. K., R. H. Baker, R. K. Sakai, F. Mahmood, Science 298:117-119. H. R. Rathor, K. Raana, and G. Toqir 1981. Anopheles Service, M. W. 1977. Mortalities of the immature culicifacies Giles: Mating behavior and competitiveness stages of species B of the Anopheles gambiae in nature of chemosterilized males carrying a genetic complex in Kenya: comparison between rice fields sexing system. Ann. Entomol. Soc. Am. 74:395-401. and temporary pools, identification of predators, Riehle, M. A., P. Srinivasan, C. K. Moreira, and M. and effects of insecticidal spraying. J. Med. Entomol. Jacobs-Lorena 2003. Towards genetic manipulation 13:535-545. of wild mosquito populations to combat malaria: Service, M. W. 1993. Mosquitoes (Culicidae). In R. P. advances and challenges. J. Exp. Biol. 206:3809-3816. Lane, and R. W. Crosskey (eds.), Medical Insects and Robinson, A. S., and G. Franz 2000. The application Arachnids. Chapman and Hall, London. of transgenic insect technology in the sterile insect Sinkins, S. P., and F. Gould 2006. Gene drive systems technique, pp. 307-318. In A. M. Handler, and A. for insect disease vectors. Nat. Rev. Genet. 7:427-435. A. James (eds.), Insect transgenesis: Methods and Sinkins, S. P., and S. L. O’Neill 2000. Wolbachia application. CRC Press. Boca Raton, Florida, USA. as a vehicle to modify insect populations, pp. 271- Robinson, A. S. 2002a. Mutations and their use in 287. In A. M. Handler, and A. A. James (eds.), Insect insect control. Mutat. Res. 511:113-132. transgenesis: methods and applications. CRC Press, 27 Robinson, A. S. 2002b. Genetic sexing strains in Boca Raton, FL. medfly, , sterile insect technique Snow, J. W. 1988. Radiation, insects and eradication programmes. Genetica 116:5-13. in North America. An overview from screwworm to Robinson, A. S., G. Franz, and P. W. Atkinson 2004. bollworm, pp. 3-13. Modern Insect control: Nuclear techniques and Biotechnology. Proceedings of a urban Dar es Salaam, Tanzania. BMC Public Health

Chapter 01 symposium jointly organized by IAEA/FAO: November 6:154. 1987, SM-301/29. IAEA, Vienna. Vreysen, M. J. 1995. Radiation induced sterility Takken, W., and B. G. J. Knols 1999. Odor-mediated to control Tsetse flies. PhD thesis Wageningen behavior of Afrotropical malaria mosquitoes. Annu. Agricultural University. Rev. Entomol. 44:131-157. Vreysen, M. J., K. M. Saleh, M. Y. Ali, A. M. Abdulla, Takken, W., and S. W. Lindsay 2003. Factors affecting Z. R. Zhu, K. G. Juma, V. A. Dyck, A. R. Msangi, P. A. the vectorial competence of Anopheles gambiae: a Mkonyi, and H. U. Feldmann 2000. Glossina austeni question of scale, pp. 75-90. In W. Takken, and T. W. (Diptera: Glossinidae) eradicated on the island of Scott (eds.), Ecological aspects for application of Unguja, Zanzibar, using the sterile insect technique. J. genetically modified mosquitoes. Frontis, Springer, Econ. Entomol. 93:123-135. Dordrecht. Vulule, J. M., R. F. Beach, F. K. Atieli, J. M. Roberts, Thomas, D. D., C. A. Donnelly, R. J. Wood, and L. D. L. Mount, and R. W. Mwangi 1994. Reduced S. Alphey 2000a. Insect population control using a susceptibility of Anopheles gambiae to permethrin dominant, repressible, lethal genetic system. Science associated with the use of permethrin-impregnated 287:2474-2476. bednets and curtains in Kenya. Med. Vet. Entomol. Ticheler, J., M. Loosjes, J. Ph. W. Noordink, 8:71-75. J. Noorlander, and J. Theunissen 1974. Field Walker, K. 2000. Cost-comparison of DDT and experiments with the release of sterilized onion flies. alternative insecticides for malaria control. Med. Vet. Hylemya antiqua (Meig.), pp. 103-107. Proceedings Entomol. 14:345-354. of a Panel on the Practical Use of the Sterile-Male Weidhaas, D. E., C. H. Schmidt, and E. L. Seabrook Technique for Insect Control, Vienna, 13-17 November 1962. Field studies on the release of sterile males 1972. Panel proceedings series, Vienna (Austria), for the control of Anopheles quadrimaculatus. Mosq. International Atomic Energy Agency. News 22:283-291. Tirados, I., C. Constantini, G. Gibson, and S. J. White, G. B. 1974. Anopheles gambiae complex and Torr 2006. Blood-feeding behaviour of the malarial disease transmission in Africa. Trans. R. Soc. Trop. mosquito Anopheles arabiensis: implications for vector Med. Hyg. 68:278-301. control. Med. Vet. Entomol. 20:425-437. White, N. J. 2008. Plasmodium knowlesi: The Fifth Toledo, J., J. Rull, A. Oropeza, E. Hernández, and Human Malaria Parasite. Clin. Infec. Dis. 46:172-173. P. Liedo 2004. Irradiation of Anastrepha obliqua World Health Organizartion 1994. Control of (Diptera: ) revisited: Optimizing sterility Tropical Diseases: 1. Progress Report. Geneva: The induction. J. Econ. Entomol. 97:383-389. Organization, Division of Control of Tropical Diseases; Touré, Y. T., V. Petrarca, S. F. Traore, A. Coulibaly, Report No.: CTD/MIP/94.4. H. M. Maiga, O. Sankare, M. Sow, M. A. Di Deco, World Health Organization 2006. Global strategy and M. Coluzzi 1998. The distribution and inversion aims for effective by 2025. http:// polymorphism of chromosomally recognized taxa of www.who.int/mediacentre/news/notes/2006/np35/ the Anopheles gambiae complex in Mali, West Africa. en/ Parassitologia 40:477-511. Wondji, C., F. Simard, V. Petrarca, J. Etang, F. Tripet F., Y.T. Touré, G. Dolo, and G.C. Lanzaro. 2003. Santolamazza, A. della Torre, and D. Fontenille Frequency of multiple inseminations in field-collected 2005. Species and populations of the Anopheles Anopheles gambiae females revealed by DNA analysis gambiae complex in Cameroon with special emphasis of transferred sperm. Am. J. Trop. Med. Hyg. 68:1-5. on chromosomal and molecular forms of Anopheles UNICEF, and WHO/RBM 2005. World Malaria Report. gambiae s.s. J. Med. Entomol. 42:998-1005. 28 WHO/HTM/MAL/2005.1102. Xi, Z., C. C. Khoo, and S. L. Dobson 2005. Wolbachia Vanek, M. J. B. Shoo, D. Mtasiwa, M. Kiama, S. W. establishment and invasion in an Aedes aegypti Lindsay, U. Fillinger, K. Kannady, M. Tanner, and laboratory population. Science 310:326-328. G. F. Killeen 2006. Community-based surveillance Xi, Z., C. C. Khoo, and S. L. Dobson 2006. Interspecific of malaria vector larval habitats: a baseline study in transfer of Wolbachia into the mosquito disease vector 01 General introduction

Aedes albopictus. Proc. Biol. Sci. 273:1317-1322. Yuval, B., and A. Bouskila 1993. Temporal dynamics of mating and predation in mosquito swarms. Oecologia 95:65-69. Yuval, B., J. W. Wekesa, and R. K. Washino 1993. Effect of body size on swarming behaviour and mating success of male Anopheles freeborni. J. Insect Behav. 6:333-342. Yuval, B., M. L. Holliday-Hanson, and R. K. Washino 1994. Energy budget of swarming male mosquitoes. Ecol. Entomol. 19:74-78. Yuval, B. 2006. Mating systems of blood-feeding flies. Annu. Rev. Entomol. 51:413-440.

29 PART 1 Part 1

Radiation Effects PART 1 PART

31 02 Review on Anopheles irradiation 02 A literature review on the irradiation of mosquitoes in the context of a Sterile Insect Technique programme

by Michelle EH Helinski, Andrew G Parker, and Bart GJ Knols

At present, there is a renewed interest to assess the feasibility of applying the Sterile Insect Technique (SIT) against African malaria vectors in designated areas. The SIT relies on the sterilisation of males before mass release, with sterilisation in the near future being achieved through the use of ionising radiation. This paper reviews previous work on radiation sterilisation of Anopheles mosquitoes. Irradiation studies have been carried out on a number of anopheline species and in general, the 350-500pupal stage Million was cases irradiated, of clinical due malaria to ease occur of handling annually, compared 60% of which to the are adult in sub-Sa stage.- haranThe dose-response Africa. Moreover, curve 80% between of all deaths the induced attributed sterility to malaria and log occur (dose) in thiswas region.shown Into numbers, be sigmoid, 1 million and there Africans was die a marked of the disease species each difference year, with in the radiation vast majority sensitivity. of deathsCompetitiveness occurring studiesamong have children generally below been five performed years of age. under Pregnant laboratory women conditions. are an- otherThe competitiveness major risk group; of malariamales irradiated can cause at low high birth doses weight was relativelyand premature poor, butdelivery with (Rogersonincreasing et ratios al. 2007). of sterile The impact males, of hatch malaria rates on could the economic be lowered situation effectively. of endemic Males countriesirradiated isas high pupae (Gallup had anda lower Sachs competitiveness 2001), and the correlationcompared to between males irradiated poverty and as malariaadults but clearly the demonstrated use of partially-sterilising (Sachs and Malaney doses has 2002). not been extensively studied. MethodsMalaria to reduce is a somaticparasitic damage disease transmitted during the irradiation by female processmosquitoes as well of the as usege- nusof other Anopheles agents. The or malaria techniques parasite to induce is a protozoan sterility are of the discussed. genus Plasmodium We conclude. There that arethe fouroptimal species radiation of malaria dose chosenparasites for that insects can that infect are humans to be released under natural during condian SIT- tions:programme Plasmodium should falciparum ensure a ,balance P. vivax ,between P. ovale andinduced P. malariae sterility. The of firstmales two and species their fieldcause thecompetitiveness, most infections withworldwide competitiveness and P. falciparum being Welsh determined is by far the under most (semi-) 1.1 field conditions.

Chapter will be submitted in a slightly modi- fied form as: Radiation biology of mosqui- toes: an overview. Malaria Journal Background perimental phase and no regulatory framework

Chapter 02 exists for the introduction of transgenic insects into the wild (Knols and Louis 2005). The aim of he Sterile Insect Technique (SIT) for mos- this paper is to give an overview of irradiation quitoes involves the mass production, sex studies performed on anopheline mosquitoes, Tseparation, sterilisation and release of together with some information from other in- sterile males. Contemporary methods to induce sects. No attempt has been made to review all sterility in the released insects are ionising ra- the available literature on anopheline irradia- diation or chemosterilisation. Chemosterilants tion, but rather to set a baseline for future work have been used experimentally and in field trials on this subject. in the 1960-70s against mosquitoes (Dame et al. 1981, Dame 1985). The chemosterilising agents that were used were mutagenic, and thus pre- Introduction to irradiation sented a potential hazard to humans during the treatment process. Their use has been aban- When biological material is irradiated, molecu- doned after concerns were raised about the lar bonds are broken, ions created, and free effect of residues in the environment and non- radicals formed. The free radicals attack further target organisms, particularly when large num- molecular bonds, and when DNA is damaged bers of treated insects were released (Hayes this can lead to the formation of dominant le- 1968). These concerns were mainly based on thal mutations in the germ cells (LaChance the findings of one, so far unreplicated, study 1967, Curtis 1971). Damage to somatic cells also that found that spiders fed on a diet of nothing occurs, especially in cells undergoing mitosis. but chemosterilised mosquitoes subsequently In general, damage to the germ and somatic became sterile (Bracken and Dondale 1972). cells increases with dose and somatic damage Although the amount of residue released in the decreases with the progressive development of environment was very low, due to the careful the insect. As the success of an SIT campaign rinsing of pupae (LaBrecque et al. 1972), it is un- largely depends on the competitiveness of the likely that regulatory approval would be given released insects, it is crucial to reduce adverse for the release of chemosterilised insects. Ion- effects of irradiation to a minimum. Although ising radiation has therefore become the prin- it is generally believed that the released males cipal technique for sterilisation, even though it need to be fully sterile, it has been suggested has been reported to reduce competitiveness that complete sterility does not have to be in- of the males more than chemosterilisation (El- duced and that more sterility can be introduced Gazzar and Dame 1983, Dame 1985). However, into the field population using lower radiation successful SIT programmes for the eradication doses but with more competitive insects (Rob- of the New World Screwworm inson 2002). Moreover, reduced competitive- hominivorax Coquerel from the USA, Central ness of the sterile males can be partly overcome America (Snow 1988) and Libya (Lindquist et by increasing the ratio of sterile-to-wild insects al. 1992) and the tsetse fly Glossina austensi (Knipling 1955). Newstead from Zanzibar (Vreysen et al. 2000) relied on radiation-sterilised insects, as well as Radiation source and dosimetry the ongoing SIT programmes against the Medi- For the irradiation of insects, gamma rays are terranean fruit flyCeratitis capitata Wiedemann usually used due to their high energy and pen- 34 from Central and Latin America (Dyck et al. etration. The commonest sources of gamma 2005). A variety of novel sterilisation methods rays are the radioisotopes 60Co and 137Cs as both based on transgenesis are currently under de- have a long half-life and emit high-energy gam- velopment (Thomas et al. 2000, Alphey 2002). ma rays. 60Co is more easily manufactured and However, these technologies are still in the ex- is therefore more often used. In conventional 02 Review of Anopheles irradiation

1987) but the use of electron beams has not been reported. Dosimetry is used to quantify the dose received by the irradiated insects. The selec- tion of a suitable dosimetry system depends on several considerations, including dose range of interest, ease of handling, expertise available, cost, and uncertainty that is consistent with the system. For SIT programmes, a radiochro- mic film dosimetry system seems most suitable (FAO/ IAEA/USDA 2003).

Spermatogenesis To discuss the radio-sensitivity of the male germ cells, some basic knowledge on sperm development and production during different life stages is required. Sperm is produced in the testes. During the successive stages of sperma- togenesis, germ cells multiply and differenti- ate. In short, spermatogenesis includes the fol- lowing developmental stages: primordial germ cells, spermatogonia (primary and secondary), spermatocytes (primary and secondary), sper- matids and spermatozoa (mature sperm cells). Figure 2.1. Conventional self-shielded Cobalt-60 gamma Spermatogenesis is cystic and within one testis, source (Gammacell 220) used for the irradiation of insects. cysts of different stages of development can be Dimensions are 2.1 (h) x 1.0 (w) x 1.5 (l) m. found. Germ cells within a single spermatocyst are more or less at the same stage of develop- self-shielded irradiators (e.g. the Gammacell ment. As spermatocysts mature, they break 220 ®, MDS Nordion, Ottawa, Canada, Fig. 2.1) down and release spermatozoa into the sperm the sample chamber is surrounded by several reservoir, situated at the anterior section of the rods or pencils of the isotope. The dose rate testis. of the cell is determined by the activity of the Spermatogenesis occurs mainly dur- source, and the absorbed dose delivered to the ing the larval and pupal stages but mosquito insects controlled by adjusting the exposure species differ in the timing of the process (Cle- time (Bakri et al. 2005). The sample chamber ments 1992a). In An. culicifacies Giles (Mah- volume is 3.7 litres. The dose rate distribution in mood and Reisen 1994) a small number of ma- the chamber is not uniform and accordingly in- ture spermatozoa were already present in the sects receive different dose rates when placed sperm reservoir in the late pupal stage, and at different positions in the chamber although many mature spermatocysts occupied the tes- the dose rate usually changes least at the cen- tis. During the last hours of the pupal stage and tre of the chamber. Besides gamma rays, X- the first hours after emergence, spermatocyst rays and accelerated electron beams can also maturation continued and in newly emerged be used to irradiate insects. X-rays have simi- males spermatozoa made up 45% of the testis 35 lar penetration as gamma rays and they have volume in An. stephensi Liston (Mahmood and been used in a number of studies on Anopheles Reisen 1982) and 41% in An. culicifacies (Mah- irradiation (Ali and Rozenboom 1972, Curtis et mood and Reisen 1994), and spermatocysts in al. 1976, Shoukry 1980, Krafsur and Davidson various stage of development (mature or not) Table 2.1. A selection of irradiation studies performed on male anophelines. Stage of irradiation is pupa (P) or adult (A). Sterility data were obtained by mating of irradiated males (I ♂) with normal (un-irradiated) females (N ♀). Information partly gathered from the IDIDAS Database (IAEA 2008). Chapter 02 Anopheles Stage Age Dose- Induced sterility (%) Reference species (hrs) range tested (Gy) dose (I! x N") albimanus P <24 20-80 50 84.3 Ali and Rozenboom 1972 80 100 arabiensis P 23-27 120 99.6 Curtis 1976

A <16 120 99.4 A n.r. 40 75 Lines and Curtis 1985

gambiae s.s. P 0-7 120 99.5 Andreasen and Curtis 2005 24-32 99.5

A <24 120 99.5 >24 99.6 P 22-27 120 78-94 Curtis 1976

A <24 80 91 120 99.3 A 24-48 45 87.4 Krafsur and Davidson 1987 pharoensis P 24 50-80 50 95.81 Abdel-Malek et al. 1975 80 98.71 P 24 120 100 1 Abdel-Malek et al. 1967

A 72 5-50 30 76.5 Shoukry 1980 50 96.8 quadrimaculatus P 24 118 100 Davis et al. 1959

stephensi P 4-28 10-120 50 80.3* Sharma et al. 1978 80 97.2* 120 99.1* P 0-7 120 98.1 Andreasen and Curtis 2005 24-32 98.4

A <24 120 98.1 >24 98.2

A <24 120 97.5 Curtis 1976

A 60 10-80 50 87.4 Akram and Aslamkhan 1975 80 96.5 !"!#$%&'(%!)*(+,-,*.!/0)!'0-'&-0*(%!1+23!24)(+5(%!)*(+,-,*.!%0*067!8$'-(0+!,1!24)(+5(%!2+!,$%&'(%!)*(+,-,*.!/0)!%()'+,4(%6!$6+9!$2*!+('2+%(%6!

were present (Mahmood and Reisen 1982, phase, with a high maturation of sperm in the 1994, Huho et al. 2006). These spermatocysts last hours of pupal development and first hours continue to mature and release sperm into after emergence. Spermatocysts are present in the sperm reservoir during adult life, and the adult mosquitoes and maturation of new sperm 36 percentage of the testis occupied with sperm occurs continuously during adult life. reservoir increased with age (Mahmood and Reisen 1982, 1994, Huho et al. 2006). In conclu- Radiation sensitivity sion, the early stages of spermatogenesis seem Radiation-induced dominant lethal mutations to occur primarily during the larval and pupal arise as a result of chromosomal damage in the 02 Review of Anopheles irradiation

treated cells (LaChance et al. 1967). An excellent tis 1976), although much depends on the dose overview on the induction of dominant lethal administered and the age of the pupae. On the mutations by irradiation or chemosterilisation other hand, handling and irradiation of pupae is is provided by LaChance (1967). A dominant le- considered easier due to their relative robust- thal mutation occurring in a germ cell does not ness compared to the adults. affect the maturation of the cell into a gamete In Anopheles, under laboratory condi- or the participation of the gamete to form the tions, the pupal stage lasts on average between , but causes the death of the develop- 25-52 hours (Clements 1992b), depending on ing embryo (LaChance et al. 1967). The earlier species and rearing conditions. In an experi- stages of spermatogenesis (spermatocytes and ment where An. pharoensis Theobald pupae spermatogonia) are more radiosensitive than were irradiated at different ages (Abdel-Malek later stages (spermatids and spermatozoa) in et al. 1967) it was shown that emergence rate terms of irreversible damage and this can result and longevity in pupae older than 15 hours did in the death of the developing cell (Proverbs not differ from un-irradiated controls, even 1969, Anwar et al. 1971, Bakri et al. 2005). The when the dose applied was high (120 Gy). The greatest effect on induced sterility is in gen- irradiation of pupae aged 1-5 hours drastically eral achieved when the older stages of sperma- decreased emergence rate. Similar results were togenesis are irradiated, as this results in domi- obtained in An. gambiae s.s. Giles (Curtis 1976). nant lethal mutations that lead to embryonic In An. quadrimaculatus Say the irradiation of mortality after fertilisation (Sobels 1969, Anwar young pupae (1-4 hrs) with 90 Gy resulted in et al. 1971, Lecis et al. 1975). The irradiation normal emergence (Davis et al. 1959), but lon- process can also damage somatic cells, with gevity, measured as the percent survival after somatic cells in mitosis being most sensitive three days, was greatly reduced. The irradiation to radiation (Proverbs 1969). Mitotically divid- of adults can be performed soon after emer- ing cells are found in mosquitoes mostly during gence. Adults emerge in the dark, and irradia- the larval and pupal stages. Reduced longevity tion is generally performed from the next day is one of the commonest observed responses onwards (i.e. adults > 12 hrs of age). to radiation (Proverbs 1969). Other effects of irradiation can be more subtle. A study in the Handling male house flyMusca domestica L. showed that In experimental settings, pupae can be irradi- irradiation induced considerable changes in the ated in small wells or Petri dishes, lined with fine structure of the fibrillar flight muscle and wet cotton wool covered with filter paper caused damage to the flight muscle mitochon- (Abdel-Malek et al. 1966, Abdel-Malek et al. dria; the damage persisted longer in flies irra- 1967). This method allows for the irradiation of diated with higher doses (Bhakthan and Nair relatively large numbers of pupae (~ 500) at a 1972). time. Adults, however, are much more fragile than pupae and require careful handling. Prior Developmental stage to irradiation, adults can be inactivated by chill- To reduce somatic damage, insects should ing (Smittle and Patterson 1974), which allows be irradiated at, or near to, the completion of them to be confined in a small space within the their development. Mosquito stages that meet irradiator so that dose variation can be reduced these requirements are the late pupal and adult and mechanical damage to the insects mini- stage. Eggs (Tantawy et al. 1966, Abdel-Malek mised. In small-scale studies adults can usually et al. 1975) and larvae have been irradiated (Ab- tolerate the cooling and stacking for the irradi- 37 del-Malek et al. 1975), but this caused exces- ation, but in operational campaigns very large sive mortality in the treated insects. In general, numbers of insects will have to be irradiated. the negative side effects of irradiation are less For large-scale irradiation, devices need to be pronounced in adults compared to pupae (Cur- developed that can contain larger numbers. One system has been developed (Smittle and indicates a “one-hit” relationship whereas de-

Chapter 02 Patterson 1974, Curtis 1976) that allows rela- partures from linearity indicate a “multi-hit” re- tively large numbers of adults to be irradiated lationship (i.e. two or more independent events simultaneously (~ 7,000- 14,000), and for recent in the same cell interact to produce a dominant field releases of Skuse pupae lethal event; Curtis 1971). were irradiated in a device that could hold ~ 20,000 individuals (Bellini et al. 2007). Longevity As reduced longevity is often recorded as a re- Finding the optimal dose sult of somatic damage (Proverbs 1969), exper- To determine the optimal dose for released in- iments must be undertaken to investigate the sects, a wide range of doses are used to gener- effect of irradiation on insect longevity, ideally ate dose response data. Initially, it is important under conditions that induce stress to empha- to confine the batches of insects in a small vol- sise differences. Specifically, male survival dur- ume in the centre of the irradiation chamber to ing the first days of adult life is important, as ensure a dose uniformity ratio <1.1. In opera- in these days mating is expected to occur after tional programmes this precision in dose distri- release. bution cannot be obtained, as very large num- bers of insects will need to be irradiated, gener- Competitiveness ally in quite large volumes. Insects will receive The ability of irradiated males to locate, com- a range of doses, and the programme will need pete for, and successfully mate with the wild fe- to define the minimum and maximum accept- males is as important as induced sterility levels able dose. When determining the optimal dose, (Calkins and Parker 2005). Competition experi- effects on sterility, longevity, and importantly ments are performed to study the competitive- competitiveness need to be examined in detail ness of irradiated males, i.e. how well males are (Parker and Mehta 2007). able to compete against wild males for mates. Initially, these competition experiments are Radiation-induced sterility likely to take place in a laboratory setting but The level of sterility induced in irradiated field or large outdoor cage tests must be con- males is measured by mating the males with ducted as well to reveal those effects that are un-irradiated virgin females. Eggs are then col- not evident under laboratory conditions (Knols lected from individual females or en masse and et al. 2002). Ideally, irradiated males are chal- checked for hatch rate. Non-hatched eggs are lenged with wild males and compete for wild presumed to have died due to a dominant lethal females. Especially when wild insects are used mutation. Even when mated with un-irradiated these experiments need to be performed in a males, females can lay a variable proportion semi-field setting, as wild males are likely to of eggs that will not hatch and sometimes lay perform poorly under laboratory conditions. entire batches of unfertilised eggs, and the To perform competition experiments, measured hatch rate must be corrected for the un-irradiated males and females are introduced control hatch rate (Abbott 1925). Control egg into a cage in a 1:1 ratio and irradiated males hatching rate in laboratory colonies was report- are introduced at equal and higher ratios. Males ed to be 84% in An. arabiensis Patton (Krafsur will compete for the females, and hatching data and Davidson 1987), 90% in An. stephensi (An- are collected from en masse egg laying or egg dreasen and Curtis 2005) and 85% in An. gam- batches are collected from individual females 38 biae s.s. (Andreasen and Curtis 2005). that are separated after mating. When eggs are When fertility is plotted on a logarith- collected en masse, Haisch (Haisch 1970) and mic scale against dose, an insight in the number Fried (Fried 1971) developed a method for de- of dominant lethal mutations in a cell is provided termining a point estimate of competitiveness (LaChance 1967, Curtis 1971). A linear response for sterilised insects. This value, usually called 02 Review of Anopheles irradiation

the Fried index, can be determined provided eles species are shown in Table 2.1. In a majority egg-hatch data are known for control (i.e. fe- of the studies pupae were irradiated, and the males mated to normal (un-irradiated) males, age of the pupae varied between 0-32 hours. Al- Ha) and sterile matings (i.e. females mated though it is desirable to irradiate pupae as late to treated (irradiated) males, Hs). The com- as possible, most studies used pupae around petitiveness index (C) is then estimated as C= 24 hours old as this is the most convenient age ((Ha-Ee)/(Ee-Hs))* (N/S); where Ee is observed for irradiation under normal laboratory rear- hatch, N the number of normal males, and S ing conditions and light regimes. Adults were the number of treated males (Fried 1971). The irradiated from less than 24 to 96 hours old. index is thus independent of the ratio of treated The doses administered ranged between 5-120 to normal insects used. A competitiveness val- Gy, with some studies using a wide dose-range ue close to 1 indicates equal competitiveness whilst others tested fewer, depending on the of treated compared to normal males, a value goal of the study (Table 2.1). below 1 a reduced competitiveness of treated males, while values above 1 indicate a better Sterility competitiveness of treated males compared The relationship between induced sterility and to normal males. According to Fried (1971) the log (dose) in insects is sigmoid in form and fol- quantitative meaning of C is that a competitive- lows the pattern of a logistic response curve. At ness index of 0.4 indicates that the number of lower doses, an approximately linear relation- treated males should be increased 2.5 times (1/ ship between dose and induced sterility was ob- 0.4) to obtain the effect on the reduction in egg served while at higher doses the curve flattens hatch that fully competitive males would have such that increasing amounts of radiation are given. For values above 1 this does not apply required for proportionally smaller increases in and C values quickly increase. sterility (LaChance and Graham 1984, Robinson 2002). In Anopheles, doses of around 100- Experimental work 120 Gy induce more than 98% sterility (Table 2.1). In the partially-sterilising dose range, a Irradiation of Anopheles mosquitoes has been dose of 80 Gy produces more than 90% steril- performed with two aims. Firstly, to investigate ity in males of different species and at a dose the effect of different radiation doses on male of 50 Gy sterility exceeds 80%. This is confirmed sterility and competitiveness in the framework in chromosome rearrangements studies; when of an SIT programme. Secondly, to produce adult males were irradiated with 40-45 Gy, the chromosomal rearrangements and mutations level of sterility induced ranged between 75- for the development of genetic sexing systems. 87% (Lines and Curtis 1985, Krafsur and David- The latter studies are characterised by the use son 1987). of a lower dose so that progeny can be ob- A considerable amount of variation is tained for further analysis, the former studies observed when comparing sterility levels be- used higher doses with the emphasis on induc- tween species. A dose of 50 Gy applied to the ing complete sterility. In addition, a few field pupal stage induced 84% sterility in An. albi- releases of radiation-sterilised males have been manus Wiedemann, and 80% sterility in An. performed (see Benedict and Robinson (2003)). stephensi, while in An. pharoensis the level of Unfortunately, most studies do not specify sterility induced was considerably higher (96%; whether dosimetry was used to verify the ab- Table 2.1). A study that directly compared 39 sorbed dose. An. stephensi with An. gambiae s.s. found the former to be more radiation-resistant at a dose Stage and dose range of 120 Gy, even though the difference was small The levels of sterility induced in several Anoph- (Andreasen and Curtis 2005). A marked species difference was reported in a study comparing 1975) is similar to control groups; this suggests

Chapter 02 An. gambiae s.s. with An. albimanus. At a dose that sperm transfer from irradiated males to fe- of 80 Gy, 91.3% sterility was observed for An. males was normal. Insemination rates were de- gambiae s.s. while for An. albimanus this was > termined (Ali and Rozenboom 1972), and in An. 99.9% (Table 2.1). albimanus equal insemination rates for un-irra- In a few studies, both the pupal and diated females inseminated by irradiated males the adult stage were irradiated and the level of were found compared to the control. However, sterility induced in each was determined (Ta- in the irradiated males x irradiated females ble 2.1). At the high dose of 120 Gy, equal lev- group at 80 Gy, a greatly reduced insemination els of sterility were found for pupal and adult rate was observed (Ali and Rozenboom 1972). irradiation in An. gambiae s.s., although one The ability of irradiated sperm to com- experiment suggested pupae to be more radi- pete with normal sperm was tested in an exper- oresistant. In An. stephensi pupae and adults iment with An. pharoensis (Abdel-Malek et al. have been irradiated in two separate studies; 1967). Males irradiated at 120 Gy as pupae were at 50 Gy a higher level of induced sterility was allowed to mate for a number of nights with reported for the adult irradiation. At 80 Gy, this virgin females after which they were removed difference was no longer observed. and replaced with normal males. Females laid eggs of which none hatched, indicating that Longevity the irradiated sperm behaved normally, and ei- Longevity was measured in a small number of ther remating did not occur or the first mating studies and scored as daily mortality under nor- took precedence. Storage of sperm in the tes- mal rearing conditions. In most studies, differ- tis did not restore its fertility when males were ences in longevity of irradiated males compared mated at day 5 after pupal irradiation with 120 to control males were small with only trends Gy (Abdel-Malek et al. 1967), and no recovery being reported. Pupal irradiation in An. pha- of fertility was observed when males irradiated roensis resulted in a non-significant reduction as pupae or adults were remated to a second in longevity after irradiation at doses of 100-130 batch of virgin females after 8-9 days (Curtis Gy (Abdel-Malek et al. 1967); while in another 1976). In addition, storage of sperm from males study a non-significant increase in longevity af- irradiated as adults in the spermathecae did not ter irradiation with 5-70 Gy was reported (Ab- restore its fertility when females were ovipos- del-Malek et al. 1966). However, in An. stephen- ited twice (Curtis 1976). si, the longevity of males irradiated as pupae at 80 Gy was significantly reduced (Sharma et al. Competitiveness 1978). When pupae (22-27 hrs) and adults (< 24 Competition experiments in the laboratory hrs) of An. gambiae s.s. were irradiated with a have been performed in some anopheline spe- high dose of 120 Gy, an increased mortality for cies. Irradiated males were introduced in vari- the irradiated pupae 24 hrs after irradiation was ous ratios in rearing cages with un-irradiated reported compared to zero mortality in the ir- males and females present. In most cases, the radiated adults (Curtis 1976). un-irradiated males and females were laborato- ry-reared individuals, with the exception of one Mating ability study (Tantawy et al. 1967) which used wild- In general, the mating ability of irradiated caught material for one of the experiments males does not seem to be adversely affected performed. Mating was allowed for a number 40 by irradiation. The number of eggs produced by of nights and mates were introduced soon af- the females mated to males irradiated over a ter the irradiation process (Davis et al. 1959, wide dose-range as pupae (Abdel-Malek et al. Andreasen and Curtis 2005) or some days after 1967, Ali and Rozenboom 1972, Abdel-Malek emergence (Sharma et al. 1978). et al. 1975) or adults (Akram and Aslamkhan A selection of data on competitiveness 02 Review of Anopheles irradiation

Table 2.2. Overview of a selection of competitiveness studies performed on Anopheles mosquitoes under laboratory conditions. The ratio of irradiated males (I ♂) competing with normal (un-irradiated) males (N ♂) for normal (un-irradiated) females (N ♀) is given. Insects were irradiated in air. Stage of irradiation is pupa (P) or adult (A). Where eggs were collected en masse, the competitiveness value (C) is calculated according to the Fried index (Fried 1971; see text). In the lower part of the table data were collected by individual egg batches, and C= (Sterile batches/ Fertile batches) * (N/S), see text. sterile batches averaged for the three replicates. I N N Dose Anopheles Stage Age Observed Expected Competitiv Reference ! ! " (Gy) species (hrs) hatch (%) hatch (%) eness (C) 1 1 1 120 pharoensis P 20-24 39 34 0.72 Tantawy et 5 1 1 20 11 0.47 al. 1967 10 1 1 4 6 1.58 15 1 1 1 4 4.40 10 1 11 6 7 1.03 1 1 1 118 quadrimacula P 24 74 48 0.30 Davis et al.

2 1 1 tus 85 32 0.06 1959 3 1 1 42 24 0.43 4 1 1 53 19 0.20 6 1 1 42 14 0.21 10 1 1 28 9 0.24 1 1 1 80 stephensi P 4-28 54 50 0.88 Sharma et 1 1 1 120 66 50 0.51 al. 1978 Fertile Sterile Competitiv batches batches eness (C) 1 1 1 120 gambiae s.s. P 0-7 30 7 0.24 Andreasen 1 1 1 P 24-32 22 10 0.47 and Curtis 2005 2 1 1 1 A <24 24 26 1.06 1 1 1 A >24 21 23 1.10 1 1 1 120 stephensi P 0-7 32 6 0.20 1 1 1 P 24-32 29 9 0.32 1 1 1 A <24 22 22 1.07 1 1 1 A >24 19 24 1.35 1 1 1 120 arabiensis P 23-27 14 6 0.43 Curtis 1976 1 1 1 120 gambiae s.s. P 22-27 20 14 0.70 1 1 1 A <24 39 48 3.20 1 Laboratory irradiated males competed against wild males for wild females under laboratory conditions. 2 Fertile and sterile batches averaged for the three replicates.

experiments in anophelines is presented in Ta- could be lowered by increasing the ratio of irra- ble 2.2. Most studies used only one high dose diated to un-irradiated males. The competitive- to look at competitiveness, and no comparison ness of An. pharoensis irradiated with 120 Gy was made between higher and lower doses with against laboratory reared un-irradiated males one exception (Sharma et al. 1978). Irradiation was 0.47 or more when measured at lower ra- almost always had a negative impact on com- tios, and irradiated males were equally compet- 41 petitiveness of the irradiated males, resulting in itive when measured at the higher ratios. More- higher hatching rates than would be expected if over, when these males had to compete against the irradiated males were equally competitive wild-caught males for wild-caught females in with the un-irradiated males. Hatching rates the laboratory, they were equally competitive when measured at a ratio of 10:1. In An. quad- lease study in Cx. tarsalis (i.e. single release of

Chapter 02 rimaculatus, competitiveness of males irradi- 13,000 males), males irradiated with 60 Gy were ated as pupae was poor (Table 2.2). As can be effective in inducing some sterility in the target seen from Table 2.2, the competitive index of population (Asman et al. 1980). The continuous Fried (1971) does not always accurately predict releases of Ae. aegypti L. (i.e. 4.6 million in 43 the reduction in observed hatch at higher ratios weeks) radiation-sterilised as pupae did not, based on data from lower ratios, as the paper however, result in population suppression, but suggested (see above). It is therefore advised to the dose applied was high (110- 180 Gy; Morlan use the index as an indicator of competitiveness et al. 1962). only. The use of high doses has a negative effect on competitiveness. In An. stephensi it Optimising sterilisation was observed that males irradiated as pupae with 80 Gy were 1.7 times more competitive Many factors influence the competitiveness of than males irradiated as pupae with 120 Gy irradiated insects. Several strategies to reduce (Sharma et al. 1978). In Culex quinquefasciatus somatic damage during the irradiation process Say the irradiation of adults with a dose of 50 are discussed below that might be beneficial for Gy resulted in high competitiveness, but as the the irradiation of anophelines. dose increased the competitiveness decreased (El-Gazzar et al. 1983). Low oxygen environment In experiments where both the pupal An important factor during the actual radiation and the adult stage were irradiated with 120 procedure is the oxygen level. During irradia- Gy, males which were irradiated as adults were tion, oxygen molecules form free radicals that more competitive than the males irradiated as induce biological damage (Fisher 1996). Irradia- pupae (Table 2.2). In Cx. quinquefasciatus a dose tion in a low oxygen environment reduces the of 80 Gy applied to the pupal stage resulted in amount of dominant lethal mutations and so- lower competitiveness when measured at a ra- matic damage, and consequently higher doses tio of 1:1 (El-Gazzar and Dame 1983, El-Gazzar are needed to induce sterility levels comparable et al. 1983) compared to competitiveness found to those induced in air, although it is often ob- after adult irradiation with a slightly lower dose served that competitiveness and longevity are of 75 Gy (El-Gazzar et al. 1983). improved despite the higher dose required. Two Some field studies on competitiveness strategies are commonly used to reduce oxygen of radiation-sterilised mosquitoes have been levels. Irradiation under hypoxia brought about performed, although few with anophelines. In by respiration of pupae kept in sealed bags is An. quadrimaculatus, pupae irradiated with 120 routinely performed with Mediterranean fruit Gy and released as adults were not able to in- fly puparia. Nitrogen has been used experimen- duce sterility in the target population after pro- tally in tsetse (Mutika and Parker 2006) and rou- longed releases, due to behaviour differences as tinely in Western Australia for Mediterranean a result of the colonisation process (Weidhaas et fruit fly (Fisher 1997). Prior to irradiation, the al. 1962). In Cx. tarsalis Coquillett males irradi- container that holds the insects is flushed with ated with 50 Gy (95% induced sterility) as adults nitrogen for some time after which irradiation were competitive with un-irradiated males follows. Beneficial effects, i.e. long-term- sur (measured at a ratio of 1:1) from the laboratory vival, of irradiation in a nitrogen environment 42 or from field populations in, respectively, small were demonstrated in Mediterranean fruit fly cages indoors or large cages outdoors (Ainsley (Fisher 1996) and tsetse (Dean and Clements et al. 1980). Males irradiated with 70 Gy were 1969, Vreysen and Van Der Vloedt 1995, Mutika also competitive in small cages (not tested in and Parker 2006). field cages; Ainsley et al. 1980). In a small re- Irradiation of mosquitoes in nitrogen 02 Review of Anopheles irradiation

has been performed in one anopheline and two Another potential class of radioprotec- culicine species. No beneficial effects of pupae tors are anti-oxidants. Anti-oxidants neutralise radiation in An. gambiae s.s. (Curtis 1976), or free radicals and thus prevent damage. One of pupal or adult radiation in Cx. quinquefasciatus these antioxidants is nordihydroguaiaretic acid (El-Gazzar et al. 1983) in nitrogen were report- (NDGA), a reducing agent that replaces the ed. As expected, insects required much higher naturally present reducing agent glutathione, doses to achieve adequate sterility in nitrogen whose amount decreases with the age of an compared to air, but at those higher doses only organism (Richie et al. 1987). NDGA adminis- a marginal improvement in competitiveness tered to the diet of larval Ae. aegypti increased was observed in Culex adults (competitiveness the longevity of both sexes over controls by 42- was not assessed in An. gambiae s.s.). How- 64% (Richie et al. 1986). The use of antioxidants ever in Ae. aegypti, irradiation in nitrogen was has not been studied in mosquito irradiation shown to be beneficial (Hallinan and Rai 1973). but could be a potentially interesting way to re- Competitiveness of males irradiated at 35, 70 duce radiation damage caused by free-radicals. or 100 Gy in nitrogen was equal to that of un- irradiated males, while males irradiated in air at Other sterilisation methods the same doses were less competitive. Irradia- Even though it is known that irradiation causes tion in nitrogen did increase hatch rate to some a loss in competitiveness, other methods to in- extent, but at 100 Gy, in nitrogen as well as air, duce sterility are controversial or not ready for 100% sterility was achieved. implementation. Chemosterilants offer high levels of sterility with more competitive insects Radioprotectors compared to irradiation (Sharma et al. 1978, El- Somatic damage caused by the irradiation proc- Gazzar and Dame 1983), but their use requires ess could possibly be reduced with the use of ra- special safety measures which are difficult to -ac dioprotectors. Radioprotectors are substances complish under field conditions. In this regard, which when present during irradiation, dimin- self-contained gamma irradiators are very safe ish its effects. A wide range of radioprotectors and easy to use. Even though public opinion is available with various modes of action (Weiss led to the disappearance of chemosterilants and Landauer 2003). A range of protectors in- for mosquito control in the seventies, they con- cluding amino-acids, cysteamine (aminothiol), tinue to be used against other pests. In 1991, diaminoethanetetraacetic acid (EDTA), and a large field trial to eradicate introduced sea 2-aminoethyl isothiuronium bromide (AET) lampreys Petromyzon marinus L. from the Great were tested on Cx. quinquefasciatus (El-Gaz- Lakes (USA) was initiated lasting for several zar and Smittle 1984). Pupae were soaked for years. Radiation sterilisation was considered a number of hours in the compound, pre- and but yielded unsatisfactory results regarding post-irradiation. None of the tested radiopro- male survival and competitiveness, hence male tectors seemed to have a beneficial effect on lampreys were sterilised with the chemosteri- competitiveness or sterility of the irradiated lant bisazir. However, the hazards of handling males. However, these results must be consid- are acknowledged and alternative strategies ered preliminary because little absorption of have been explored to eliminate the use of mu- the radioprotector is expected in the non-in- tagenic agents (Ciereszko et al. 2004). gesting pupal stage. Another protector, dime- The use of transgenic organisms is thyl sulphoxide (DMSO), ingested in the adult promising but at an early stage of development stage before irradiation, decreased the induc- (Thomas et al. 2000, Alphey 2002), and no suit- 43 tion of dominant lethal mutations by X-rays in able drive mechanisms are available to drive the An. atroparvus van Thiel (Lecis and Orru 1974). transgenes through the wild population (James However, DSMO is toxic and even at low con- 2005, Sinkins and Gould 2006), and the search centrations a reduced life span was observed. for suitable systems continues. Although the irradiation process would be eliminated in cer- such programmes, which were estimated to be

Chapter 02 tain transgenic approaches, transgenic strains 20-40 times greater than the costs of conven- could require the release of millions of mosqui- tional methods such as insecticide spraying or toes, facing equal challenges in maintaining in- application of insecticides on cattle (Vale and sect competitiveness during mass production. Torr 2005). In addition, SIT was estimated to In addition, public opinion on the use of trans- take longer to achieve the same goals as insec- genesis is conservative, and at this moment in ticides could (Vale and Torr 2005). However, the time, the regulatory framework is lacking to eradication of the tsetse fly Glossina austensi facilitate the introduction of such strains in the Newstead from the island of Zanzibar in 1996 wild (Knols and Louis 2006, Knols et al. 2007). It with the SIT (Vreysen et al. 2000), and the main- has also been suggested that the first releases tenance of this tsetse-free status until now of transgenic organisms would require irradia- demonstrates that the method can be applied tion to sterilise the released males (Benedict successfully in areas meeting defined criteria. and Robinson 2003). Another point of criticism has been the Another potential class of sterilants aspect of elimination. While it was recognised are insect growth regulators (IGRs). IGR’s have that anything less than elimination is soon re- been used as sterilising agents against housefly garded as a failure in SIT programmes (Krafsur (Howard and Wall 1996), blowflyLucilia sericata 1998), discussions on when elimination can be Meigen (Smith and Wall 1998) (triflumuron) declared, and how to deal with re-infestation and the Mediterranean fruit fly (Navarro-Llopis after elimination were put forward as impor- et al. 2004) (lufenuron) using impregnated tar- tant constraints of the method (Krafsur 1998). gets. The compounds were effective in inducing However, it can be argued that defining elimi- some sterility, although this was dependent on nation after the use of conventional methods as the dose administered (Smith and Wall 1998). insecticide spraying would then be subjected to Males could induce sterility in unexposed fe- similar constraints. In addition, not all SIT pro- males, although in various degrees of success, grammes strive to eradicate species, with pre- and this was attributed to either the result of vention, rather than elimination as the primary direct impairment of sperm development or objective (Whitten and Mahon 2005). transfer of active ingredient during mating to The concept of behavioural resistance the females (Smith and Wall 1998). The poten- has also received considerable attention (Whit- tial use of IGR’s as a sterilant for mosquitoes re- ten and Mahon 2005). Behavioural resistance in mains unknown. SIT programmes is the development of assorta- tive mating patterns, with females actively dis- criminating against laboratory-reared males. Constraints and criticism This condition was observed in the Japanese campaign against the melon fly Bactrocera Even though the SIT has proven successful cucurbitae Coquillett, where females from an against some insect species, the concept, op- island where SIT had been applied discriminat- eration and outcome of this approach has been ed against the laboratory-reared males while criticised. A review of these limitations can be females from a neighbouring island where no found in Krafsur (1998). Some of the criticism SIT was applied did not (Whitten and Mahon has been directed at projects carried out dec- 2005). A similar phenomenon was observed in ades ago, e.g. the screwworm eradication cam- the Mediterranean fruit fly Ceratitis capitata in 44 paign in the USA and Central America (Kraf- . However, no assortative mating was sur 1998) while others are more recent, like detected in a large number of other SIT Medi- the tsetse fly campaign under way in Ethiopia terranean fruit fly programmes (Cayol 1998, (Enserink 2007). For SIT against tsetse flies an Krafsur 1998). The continuous introgression of important point was the costs associated with wild material in the release colony maintained 02 Review of Anopheles irradiation

for mass production could potentially prevent genic approaches is even more important than the development of such behaviours (Krafsur in SIT programmes, and any reduction in fitness 1998). In anopheline mating systems, it seems must be compensated by a suitable gene-drive unlikely that assortative mating of females system, however these remain to be identi- could develop, as females are not considered fied (Marrelli et al. 2006). Notwithstanding the to actively choose a copulation partner within above criticisms (as reviewed in Krafsur (1998) a swarm. It is thus vital that released males dis- and Whitten and Mahon (2005)), under specific play the appropriate mating behaviour and en- conditions SIT could be an important tool to re- sure that they are present in the same locations duce mosquito population sizes in designated as the wild males and females for swarming. areas (Curtis 2002). In addition, much criticism Finally, it is vital to SIT programmes is directed at projects initiated decades ago, that the release population is compatible with whilst these days improved technologies and the target population and that no mating bar- methods are at hand to facilitate many aspects riers exist (Whitten and Mahon 2005). For in- of SIT programmes, and lessons learned from stance in An. gambiae s.s., one of the main vec- the past can be applied to minimise future fail- tors of malaria in Africa, substantial gene flow ure (Vreysen et al. 2007). The optimal radiation barriers between the different chromosomal dose for an SIT programme should be chosen in and molecular forms exist (Favia et al. 1997, such a way that it balances induced sterility with Touré et al. 1998, della-Torre et al. 2002, Wondji competitiveness (Parker and Mehta 2007). The et al. 2005), making An. gambiae s.s. a difficult concept of inducing 100% sterility, which was species to target with SIT. Thus, before SIT can followed in the past, has led to the use of high be considered, a thorough evaluation of the irradiation doses which in some insects reduced genetic make-up of the target population is re- competitiveness to the extent that the target quired. Also, in areas where more than one spe- population was not sufficiently suppressed. cies of mosquito contribute to malaria transmis- It is now advocated that more sterility can be sion, the effects of reducing one species with induced in the target population if insects are SIT may be negated by the sustained levels of subjected to lower, partially-sterilising doses transmission by other, non-targeted species. In (Toledo et al. 2004, Dyck et al. 2005). most of sub-Saharan Africa at least three spe- The optimal developmental stage cies contribute to malaria transmission (i.e. An. for irradiation (e.g. pupa or adult) depends on gambiae s.s., An. arabiensis and An. funestus), in many factors including ease of handling on a some even up to five (incl. An. nili and An. mou- mass-production scale, competitiveness of the cheti). For SIT to be effective in such areas more insect, release methodology etc. In the studies than one species may need to be included in the presented in this paper, the irradiation of pu- campaigns, thereby significantly increasing the pae was preferred over adults because of their logistics and costs involved. robustness in handling. For the irradiation of pupae, it is recommended to delay irradiation until pupae are ~ 24 hrs old to reduce somatic Future prospects damage. The irradiation of adults can be per- formed soon after emergence. When mosqui- In the light of all the new and exciting molecu- toes are irradiated, induced sterility follows the lar techniques that are becoming available to classical pattern of a linear relationship at low create sterile insects, sterilisation by irradiation doses and saturation of the response curve at might seem a little mundane. But unlike these higher doses. Adults appeared to be slightly 45 promising techniques, irradiation has proved to more susceptible to irradiation resulting in a be a successful, safe, and accepted way to steri- higher level of induced sterility than pupal irra- lise large numbers of insects (Dyck et al. 2005). diation. There was some difference in radiation In addition, fitness of released males in trans- sensitivity between Anopheles species indicat- ing that optimal doses in SIT programmes need ods to induce sterility, e.g. chemosterilisation

Chapter 02 to be specified for each species. or transgenic approaches, are unlikely to sub- Irradiated Anopheles males have been stitute irradiation in the near future. Other po- subjected to competitiveness assays primarily tential sterilants like IGRs remain to be tested in in laboratory settings. Generally, only high dos- mosquitoes but would only be of use for an SIT es were tested, and irradiated males performed programme if they can be applied on a mass- poorly at 1:1 ratios with un-irradiated males. scale before release. Hatch rates could be lowered by increasing Gamma irradiators will remain the the ratio of irradiated males. The effect of dose most practical irradiators for some time. The on competitiveness is an inverse relationship dose delivered to a large batch of insects re- between the dose administered and the level quired for mass release is not uniform so the of competitiveness (Hooper and Katiyar 1971, minimum and maximum dose that the insects El-Gazzar et al. 1983, Parker and Mehta 2007). can receive should be determined. Quality con- Males were more competitive when irradiated trol of the system will be crucial for a successful as adults compared to pupal irradiation (Curtis outcome. Dosimetry should be made part of the 1976, Andreasen and Curtis 2005), but this was production process and doses delivered to each not studied for the partially-sterilising doses. batch need to be in the acceptable dose range. The few field studies performed with radiation- A lower dose leads to the release of insects with sterilised insects showed good competitiveness insufficient degree of sterility; a higher dose will for Cx. tarsalis (Ainsley et al. 1980, Asman et al. produce insects with insufficient competitive- 1980), and poor competitiveness in Ae. aegypti ness, which will undermine the programme’s ef- (Morlan et al. 1962); however the latter study forts and overall success of the campaign. Suc- used much higher radiation doses. The only field cessful quality control programmes have been release of a radio-sterilised anopheline mosqui- implemented in ongoing SIT campaigns (FAO/ to, An. quadrimaculatus, was not successful due IAEA/USDA 2003) and this knowledge can read- to large behavioural differences as a result of ily be transferred to other SIT programmes. the rearing process (Weidhaas et al. 1962, Dame 1985). When determining the competitiveness Conclusion of male Anopheles by conventional competi- tion assays, it has to be kept in mind that rather At present irradiation is the most obvious choice limited information on the courtship behaviour to sterilise mosquitoes in an SIT programme. of wild males is available. Critical knowledge on Substantial literature on anopheline irradiation what the important parameters are that con- is available, but should be complemented with tribute to a male’s success in the field is lacking competitiveness studies performed in a (semi-) (Ferguson et al. 2005), although experimental field setting to determine the optimal dose and work is directed increasingly to understanding developmental stage for sterilisation. The op- male mating success (Ng’habi et al. 2005, Huho timal development stage for irradiation how- et al. 2006). Converting these successful traits ever, also depends on the logistics of the irra- back to measurable parameters in the labora- diation process (e.g. the need to irradiate large tory is the subsequent challenge. numbers of insects), release methodology, and To reduce somatic damage caused costs. by the irradiation process, systems such as a hypoxic environment and radiation protecting Acknowledgement 46 agents could be useful for mosquitoes, but re- We would like to thank A. Robinson for valuable main to be tested in depth. A loss of competi- comments on earlier versions of this Chapter. tiveness can be overcome by increasing the number of released insects (Knipling 1955), but this will result in additional costs. Other meth- 02 Review of Anopheles irradiation

References A field release of irradiated male Culex tarsalis in California, pp. 64. CMVCA press, Visalia, California. Abbott, W. S. 1925. A method of computing the Bakri, A., K. Mehta, and D. R. Lance 2005. Sterilizing effectiveness of an insecticide. J. Econ. Entomol. Insects with Ionizing Radiation, pp. 233-268. In A. 18:265-267. Dyck, J. Hendrichs, and A. S. Robinson (eds.), The Abdel-Malek, A. A., A. O. Tantawy, and A. M. Sterile Insect Technique: Principles and Practice in Wakid 1966. Studies on the eradiaction of Anopheles Area-Wide Integrated Pest Management. Springer, pharoensis Theobald by the Sterile-Male technique Dordrecht. using cobalt-60. I.Biological effects of gamma Bellini, R., M. Calvitti, A. Medici, M. Carrieri, G. radiation on the different developmental stages. J. Celli, and S. Maine 2007. Use of the Sterile Insect Econ. Entomol. 59:672-678. Technique Against Aedes albopictus in Italy: First Abdel-Malek, A. A., A. O. Tantawy, and A. M. Results of a Pilot Trial. pp. 505-516. In M. J. B. Vreysen, Wakid 1967. Studies on the eradication of Anopheles A. S. Robinson, and J. Hendrichs (eds.), Area-Wide pharoensis Theobald by the Sterile-Male technique Control of Insect Pests. Springer, Dordrecht. using cobalt-60. VI. Sperm activity in males irradiated Benedict, M. Q., and A. S. Robinson 2003. The first with the sterilising dose. J. Econ. Entomol. 60:1300- releases of transgenic mosquitoes: an argument for 1302. the sterile insect technique. Trends Parasitol. 19:349- Abdel-Malek, A. A., A. M. Wakid, A. O. Tantawy, 355. and L. M. El-Gazzar 1975. Studies on factors Bhakthan, N. M. G., and K. K. Nair 1972. Fine influencing the induction of sterility in Anopheles structural damage in the somatic tissue of gamma- pharoensis Theobald by gamma radiation, pp. 161- irradiated house fly. 1. Flight muscles. Ann. Entomol. 174. In The use of isotopes and pesticides in pest Soc. Am. 65:504-508. control, Proceedings of a symposium held in Beirut- Bracken, G. K., and C. D. Dondale 1972. Fertility Lebanon, March 1974. and survival of Achaearanea tepidariorum Ainsley, R. W., S. M. Asman, and R. P. Meyer 1980. (Araneida:Theridiidae) on a diet of chemosterilized The optimal radiation dose for competitive males of mosquitoes. Can. Entomol. 104:1709-1712. Culex tarsalis (Diptera: Culicidae). J. Med. Entomol. Calkins, C. O., and A. G. Parker 2005. Sterile insect 17:122-125. quality, pp. 269-296. In V. A. Dyck, J. Hendrichs, Akram, M., and M. Aslamkhan 1975. Production of and A. S. Robinson (eds.), Sterile Insect Technique. dominant lethal mutations by gamma-irradiation in Principles and Practice in Area-Wide Integrated Pest the malaria mosquito, . Pak. J. Management. Springer, Dordrecht. Zool. 7:177-184. Cayol, J. P. 1998. World-wide sexual compatibility in Ali, S. R., and L. E. Rozenboom 1972. Observations medfly, ceratitis capitata Weid., and its implications on sterilization of Anopheles (C.) albimanus for SIT. Fifth international symposium on fruit flies Wiedemann by X-irradiation. Mosq. News 32:574- of economic importance, 1-5 June 1998, Penang, 579. Malaysia, Abstr. Alphey, L. 2002. Re-engineering the sterile insect Ciereszko, A., I. Babiak, and K. Dabrowski 2004. technique. Insect Biochem. Mol. Biol. 32:1243-1247. Efficacy of animal anti-fertility compounds against Andreasen, M. H., and C. F. Curtis 2005. Optimal sea (Petromyzon marinus) spermatozoa. life stages for radiation sterilization of Anopheles for Theriogenol. 61:1039-1050. sterile insect releases. Med. Vet. Entomol. 19:238- Clements, A. N. 1992a. Spermatogenesis and the 244. structure of spermatozoa, pp. 333-335. In The biology Anwar, M., D. L. Chambers, K. Ohinata, and R. of mosquitoes, Volume 1. Development, nutrition M. Kobayashi 1971. Radiation-sterilization of and reproduction. Chapmann & Hall, London. 47 the Mediterranean fruit fly (Diptera: Tephritidae): Clements, A. N. 1992b. Growth and development, comparison of spermatogenesis in flies treated as pp. 150-167. In The biology of mosquitoes, Volume 1. pupae or adults. Ann. Entom. Soc. Am. 64:627-633. Development, nutrition and reproduction. Chapmann Asman, S. M., F. G. Zalom, and R. P. Meyer 1980. & Hall, London. Curtis, C. F. 1971. Induced sterility in insects. Adv. El-Gazzar, L. M., and B. J. Smittle 1984. Effect of

Chapter 02 Reprod. Physiol. 5:120-165. gamma irradiation on Culex quinquefasciatus (Diptera: Curtis, C. F. 1976. Radiation sterilization. Report Culicidae) following exposure to radioprotectors. J. on mosquito research, Ross Institute of Tropical Med. Entomol. 21:91-94. Hygiene. 01.01.76-31.12.77. Enserink, M. 2007. Welcome to Ethiopia’s fly factory. Curtis, C. F., J. Akiyama, and G. Davidson 1976. A Science 317:310-113. genetic sexing system in Anopheles gambiae species FAO/IAEA/USDA 2003. Manual for Product Quality A. Mosq. News 36:492-498. Control and Shipping Procedures for Sterile Mass- Curtis, C. F. 2002. Possible ways of using transgenic Reared Tephritid Fruit Flies, Version 5.0, pp. 85 mosquitoes for malaria and dengue control and risk pp. International Atomic Energy Agency, Vienna, assessment, pp. 165-175. Beijing, Peking University. Austria. 7th International Symposium on Biosafety of Favia, G., T. A. della, M. Bagayoko, A. Lanfrancotti, Genetically Modified Organisms, 10-16 October. N. Sagnon, Y. T. Touré, and M. Coluzzi 1997. Dame, D. A., R. E. Lowe, and D. L. Williamson 1981. Molecular identification of sympatric chromosomal Assessment of Released Sterile Anopheles albimanus forms of Anopheles gambiae and further evidence of and Glossina morsitans morsitans, pp. 231-248. In their reproductive isolation. Insect Mol. Biol. 6:377- J. B. Kitzmiller, and T. Kanda (eds.), Cytogenetics 383. and genetics of vectors. Elsevier Biomedical, Ferguson, H. M., B. John, K. R. Ng’habi, and B. Amsterdam. G. J. Knols 2005. Redressing the sex imbalance Dame, D. A. 1985. Genetic control by sterilized in knowledge of vector biology. Trends Ecol. Evol. mosquitoes, pp. 159-172. In R. Chapman, R. Barr, 20:202-209. D. E. Weidhaas, and M. Laird (eds.), Biological Fisher, K. 1996. Queensland fruit fly (Bactrocera Control of Mosquitoes. American Mosquito Control tryoni): eradication from Western Australia, pp. 535- Association, Bull 6. 541. In B. A. McPheron, and G. J. Steck (eds.), Fruit Davis, A. N., J. B. Gahan, D. E. Weidhaas, and C. Fly Pests: A World Assessment of their Biology and N. Smith 1959. Exploratory studies on gamma Management. St. Lucie Press, Delray Beach, Florida. irradiation for the sterilization and control of Fisher, K. 1997. Irradiation effects in air and in Anopheles quadrimaculatus. J. Econ. Entomol. nitrogen on Mediterranean fruit fly (Diptera: 52:868-870. Tephritidae) pupae in Western Australia. J. Econ. Dean, G. J., and S. A. Clements 1969. Effect of Entomol. 90:1609-1614. gamma-radiation on Glossina pallidipes Aust. Bull. Fried, M. 1971. Determination of sterile-insect Entomol. Res. 58:755-762. competitiveness. J. Econ. Entomol. 64:869-872. della Torre, A, C. Costantini, N. J. Besansky, A. Haisch, A. 1970. Some observations on decreased Caccone, V. Petrarca, J. R. Powell, and M. Coluzzi vitality of irradiated Mediterranean fruit fly, pp. 71- 2002. Speciation within Anopheles gambiae--the 75. Sterile-Male Technique for Control of Fruit Flies. glass is half full. Science 298:115-117. IAEA, Vienna, Austria. Dyck, A., J. Hendrichs, and A. S. Robinson 2005 Hallinan, E., and K. S. Rai 1973. Radiation sterilization (eds.). The Sterile Insect Technique: Principles and of Aedes aegypti in nitrogen and implications for Practice in Area-Wide Integrated Pest Management. sterile male technique. Nature 244:368-369. Springer, Dordrecht. Hayes, W. J. 1968. Toxicological aspects of El-Gazzar, L. M., and D. A. Dame 1983. Effects of chemosterilants, pp. 315-347. In G. C. LaBrecque, combinations of irradiation and chemosterilization and C. N. Smith (eds.), Principles of insect on mating competitiveness of Culex quinquefasciatus chemosterilisation. Appleton Century Crofts, New 48 Say. J. Econ. Entomol. 76:1331-1334. York. El-Gazzar, L. M., D. A. Dame, and B. J. Smittle Hooper, G. H. S., and K. P. Katiyar 1971. 1983. Fertility and competitiveness of Culex Competitiveness of gamma-sterilized males of the quinquefasciatus males irradiated in nitrogen. J. Mediterranean fruit fly. J. Econ. Entomol. 64:1068- Econ. Entomol. 76:821-823. 1071. 02 Review of Anopheles irradiation

Howard, J., and R. Wall 1996. Autosterilization Amsterdam. of the house fly, Musca domestica, using the chitin LaChance, L. E., C. H. Schmidt, and R. C. Bushland synthesis inhibitor triflumuron on sugar-baited 1967. Radiation-induced sterilization, pp. 147-196. In targets. Med. Vet. Entomol. 10:97-100. W. W. Kilgore, and R. L. Doutt (eds.), Pest Control- Huho, B.J., K.R. Ng’habi, G.F. Killeen, G. Biological, Physical and Selected Chemical Methods. Nkwengulila, B.G.J. Knols, and H.M. Ferguson Academic Press, Inc., New York. 2006. A reliable morphological method to assess the LaChance, L. E., and C. K. Graham 1984. Insect age of male Anopheles gambiae. Malar. J 5:62. radiosensitivity: dose curves and dose-fractionation IAEA, 2008. International Database on Insect studies of dominant lethal mutations in the mature Desinfestation and Sterilization (IDIDAS). http:// sperm of 4 insect species. Mutat. Res. 127:49-59. www-ididas.iaea.org/IDIDAS/default.htm. Lecis, A. R., and G. Orru 1974. Radioprotective James, A. A. 2005. Gene drive systems in mosquitoes: effects and toxicity of DMSO in panirradiated adult rules of the road. Trends Parasitol. 21:64-67. male Anopheles maculipennis atroparvus (Diptera: Knipling, E. F. 1955. Possibilities of insect population Nematocera). Riv. Biol. 67:319-329. control through the use of sexually sterile males. J. Lecis, A. R., V. Figus, and C. Santarini 1975. Econ. Entomol. 48:459-462. Radiosensitivity curve of different stages of Knols, B. G. J., B. N. Njiru, E. M. Mathenge, W. spermatogenesis of R. Mukabana, J. C. Beier, and G. F. Killeen 2002. (Diptera:Nematocera). Parassitologia 17:145-150. MalariaSphere: A greenhouse-enclosed simulation Lindquist, D. A., M. Abusowa, and M. J. Hall 1992. of a natural Anopheles gambiae (Diptera: Culicidae) The New World screwworm fly in Libya: a review of ecosystem in western Kenya. Malar. J. 1:19. its introduction and eradication. Med. Vet. Entomol. Knols, B. G. J., and C. Louis 2006 (eds.). Bridging 6:2-8. Laboratory and Field Research for Genetic Control Lines, J. D., and C. F. Curtis 1985. Genetic sexing of Disease Vectors. Proceedings of the joint WHO/ system in Anopheles arabiensis Patton (Diptera: TDR, NIAID, IAEA and Frontis Workshop on Bridging Culicidae). J. Econ. Entomol. 78:848-851. Laboratory and Field Research for Genetic Control Mahmood, F., and W. K. Reisen 1982. Anopheles of Disease Vectors. Nairobi, Kenya 14-16 July 2004. stephensi (Diptera: Culicidae): changes in male mating Frontis, Wageningen. competence and reproductive system morphology Knols, B. G. J., H. C. Bossin, W. R. Mukabana, and associated with aging and mating. J. Med. Entomol. A. S. Robinson 2007. Transgenic mosquitoes and the 19:573-588. fight against malaria: Managing technology push in Mahmood, F., and W. K. Reisen 1994. Anopheles a turbulent GMO world. Am. J. Trop. Med. Hyg. 77(6_ culicifacies: effects of age on the male reproductive suppl.):232-242. system and mating ability of virgin adult mosquitoes. Krafsur, E. S., and G. Davidson 1987. Production Med. Vet. Entomol. 8:31-37. of semisterile mutants in the Anopheles gambiae s.l. Morlan, H. B., E. M. McCray, and J. W. Kilpatrick species complex. Iowa State J. Res. 62:85-97. 1962. Field tests with sexually sterile males for Krafsur, E. S. 1998. Sterile insect technique for control of Aedes aegypti. Mosq. News 22:295-300. suppressing and eradicating insect populations: 55 Marrelli, M. T., C. K. Moreira, D. Kelly, L. Alphey, years and counting. J. Agr. Entomol. 15:303-317. and M. Jacobs-Lorena 2006. Mosquito transgenesis: LaBrecque, G. C., W. C. Bowman, R. S. Patterson, what is the fitness cost? Trends Parasitol. 22:197- and J. A. Seawright 1972. Persistence of thiotepa 202. and tepa in pupae and adults of Culex fatigans. Bull. Mutika, G. N., and A. G. Parker 2006. Induced World Health Organ. 74:675-676. sterility of Glossina pallidipes Austen males after LaChance, L. E. 1967. The induction of dominant irradiation in a nitrogen atmosphere. Entomol. Sci. 49 lethal mutations in insects by ionizing radiation and 9:47-53. chemicals as related to the sterile male technique of Navarro-Llopis, V., J. Sanchis-Cabanes, I. Ayala, V. insect control, pp. 617-650. In J. W. Wright, and R. Pal Casana-Giner, and E. Primo-Yufera 2004. Efficacy (eds.), Genetics of Insect Vectors of Disease. Elsevier, of lufenuron as chemosterilant against Ceratitis capitata in field trials. Pest. Manag. Sci. 60:914-920. spermatozoa and late spermatids in Drosophila.

Chapter 02 Ng’habi, K. R., B. John, G. Nkwengulila, B. Mutat. Res. 8:111-125. G. J. Knols, G. F. Killeen, and H. M. Ferguson Tantawy, A. O., A. A. Abdel-Malek, and A. M. 2005. The effect of larval crowding on the mating Wakid 1966. Studies on the eradication of Anopheles competitiveness of Anopheles gambiae mosquitoes. pharoensis Theobald by the Sterile-Male technique Malar. J. 4:49. using cobalt-60. II. Induced dominant lethals in the Parker, A., and K. Mehta 2007. Sterile insect immature stage. J. Econ. Entomol. 59:1392-1394. technique: a model for dose optimization for Tantawy, A. O., A. A. Abdel-Malek, and A. M. improved sterile insect quality. Fla. Entomol. 90:88- Wakid 1967. Studies on the eradication of Anopheles 95. pharoensis Theobald by the Sterile-Male technique Proverbs, M. D. 1969. Induced sterilization and using cobalt-60. V. Mating competitiveness of control in insects. Annu. Rev. Entomol. 14:81-102. radiosterilized males. J. Econ. Entomol. 60:696-699. Richie, J. P. Jr., B. J. Mills, and C. A. Lang 1986. Thomas, D. D., C. A. Donnelly, R. J. Wood, and L. Dietary nordihydroguaiaretic acid increases the life S. Alphey 2000. Insect population control using a span of the mosquito. Proc. Soc. Exp. Biol. Med. dominant, repressible, lethal genetic system. Science 183:81-85. 287:2474-2476. Richie, J. P. Jr., B. J. Mills, and C. A. Lang 1987. Toledo, J., J. Rull, A. Oropeza, E. Hernández, and Correction of a glutathione deficiency in the aging P. Liedo 2004. Irradiation of Anastrepha obliqua mosquito increases its longevity. Proc. Soc. Exp. Biol. (Diptera: Tephritidae) revisited: Optimizing sterility Med. 184:113-117. induction. J. Econ. Entomol. 97:383-389. Robinson, A. S. 2002. Mutations and their use in Touré, Y. T., V. Petrarca, S. F. Traore, A. Coulibaly, insect control. Mutat. Res. 511:113-132. H. M. Maiga, O. Sankare, M. Sow, M. A. Di Deco, Sinkins, S. P., and F. Gould 2006. Gene drive and M. Coluzzi 1998. The distribution and inversion systems for insect disease vectors. Nat. Rev. Genet. polymorphism of chromosomally recognized taxa of 7:427-435. the Anopheles gambiae complex in Mali, West Africa. Sharma, V. P., R. K. Razdan, and M. A. Ansari 1978. Parassitologia 40:477-511. Anopheles stephensi: effect of gamma-radiation and Vale, G. A., and S. J. Torr 2005. User-friendly models chemosterilants on the fertility and fitness for sterile of the costs and efficacy of tsetse control: application male releases. J. Econ. Entomol. 71:449-452. to sterilizing and insecticidal techniques. Med. Vet. Shoukry, A. 1980. X-ray induced sterility, dominant Entomol. 19:293-305. lethality and inherited semisterility in Anopheles Vreysen, M. J., and A. M. Van Der Vloedt 1995. pharoensis Theo (Dipt., Culicidae). Z. Angew. Radiation sterilization of Glossina tachinoides Entomol. 89:498-504. Westw. pupae. I. The effect of dose fractionation and Smith, K. E., and R. Wall 1998. Effects of targets nitrogen during irradiation in the mid-pupal phase. impregnated with the chitin synthesis inhibitor Revue Elev. Méd. Vét. Pays Trop. 48:45-51. triflumuron on the blowfly Lucilia sericata. Entomol. Vreysen, M. J., K. M. Saleh, M. Y. Ali, A. M. Abdulla, Exp. Appl. 87:85-92. Z. R. Zhu, K. G. Juma, V. A. Dyck, A. R. Msangi, P. A. Smittle, B. J., and R. S. Patterson 1974. Container for Mkonyi, and H. U. Feldmann 2000. Glossina austeni irradiation and mass transport of adult mosquitoes. (Diptera: Glossinidae) eradicated on the island of Mosq. News 34:406-408. Unguja, Zanzibar, using the sterile insect technique. Snow, J. W. 1988. Radiation, insects and eradication J. Econ. Entomol. 93:123-135. in North America. An overview from screwworm to Vreysen, M. J., J. Gerardo-Abaya, and J. P. Cayol bollworm, pp. 3-13. Modern Insect control: Nuclear 2007. Lessons from area-wide integrated pest 50 techniques and Biotechnology. Proceedings of management (AW-IPM) programmes with an SIT a symposium jointly organized by IAEA/FAO: component: an FAO/ IAEA perspective, pp. 723-744. In November 1987, SM-301/29. IAEA, Vienna. M. J. Vreysen, A. S. Robinson, and J. Hendrichs (eds.), Sobels, F. H. 1969. A study of the causes underlying Area-wide control of insect pests, from research to the differences in radiosensitivity between mature field implementation. Springer, Dordrecht. 02 Review of Anopheles irradiation

Weidhaas, D. E., C. H. Schmidt, and E. L. Seabrook 1962. Field studies on the release of sterile males for the control of Anopheles quadrimaculatus. Mosq. News 22:283-291. Weiss, J. F., and M. R. Landauer 2003. Protection against ionizing radiation by antioxidant nutrients and phytochemicals. Toxicol. 189:1-20. Whitten, M., and R. Mahon 2005. Misconceptions and constraints, pp. 601-626. In V. A. Dyck, J. Hendrichs, and A. S. Robinson (eds.), Sterile Insect Technique. Principles and Practice in Area-Wide Integrated Pest Management. Springer, Dordrecht. Wondji, C., F. Simard, V. Petrarca, J. Etang, F. Santolamazza, A. della Torre, and D. Fontenille 2005. Species and populations of the Anopheles gambiae complex in Cameroon with special emphasis on chromosomal and molecular forms of Anopheles gambiae s.s. J. Med. Entomol. 42:998-1005.

51 03 Radiation-induced sterility in An. arabiensis 03 Radiation-induced sterility for pupal and adult stages of the malaria mosquito Anopheles arabiensis

by Michelle EH Helinski, Andrew G Parker, and Bart GJ Knols

In the context of the Sterile Insect Technique (SIT) radiation-induced sterility in the malaria mosquito Anopheles arabiensis Patton (Diptera: Culicidae) was studied. Male mosquitoes were exposed to gamma rays in the pupal or adult stage and dose- sterility curves were determined. Pupae were irradiated shortly before emergence (at 20-26 hrs of age) and adults < 24 hrs post emergence. Doses tested ranged between 0 and 120 Gy. The effects of irradiation on adult emergence, male survival, induced350-500 Million sterility, cases and of insemination clinical malaria capability occur annually, were evaluated. 60% of Thewhich dose-sterility are in sub- curvesSaharan for Africa. pupae Moreover, and adults 80% showed of all thedeaths classic attributed pattern to of malaria a linear occur relationship in this atregion. lower In doses numbers, and the1 million flattening Africans of die the of the curve disease at higher each year, doses. with Fertility the vast plotted onmajority a logarithmic of deaths scale occurring against among dose suggests children a below one-hit five relationship years of toage. dose; Pregnant i.e. a largewomen proportion are another of dominantmajor risk lethalgroup; mutations malaria can resulted cause lowfrom birth single weight events and in pre the- gametes.mature delivery The irradiation (Rogerson of the et pupalal. 2007). stage The had impact no effect of malariaon adult on emergence the economic for all dosessituation tested. of endemic Adult survival countries of malesis high irradiated (Gallup and as Sachs adults 2001), was similar and the compared correlation to un-irradiatedbetween poverty males. and Males malaria irradiated clearly asdemonstrated pupae had higher, (Sachs similar, and Malaney or lower 2002). survival but differencesMalaria wereis a parasiticsmall. In thedisease pupal transmitted stage a significant by female negative mosquitoes correlation of the was foundgenus betweenAnopheles insemination. The malaria and parasite dose but is the a protozoan correlation-coefficient of the genus Plasmodiumwas associated. withThere less are than four 25% species of the of total malaria variation. parasites A review that canof the infect literature humans suggests under that natural An. arabiensisconditions: is Plasmodiummore radiation-resistant falciparum, P. than vivax other, P. ovale anopheline and P. mosquitoes.malariae. The first two species cause the most infections worldwide and P. falciparum Welsh is by far the 53

Published with minimal changes in: Malaria Journal 2006, 5: 41 Introduction transgenic organisms in the wild (Benedict and

Chapter 03 Robinson 2003). Determining the optimal dose urrently there is a renewed interest to range for an SIT programme depends on the apply the Sterile Insect Technique (SIT) level of sterility induced and the competitive- Cagainst malaria vectors. The SIT has ness of the irradiated males. A low dose may successfully been applied to a number of pest result in insufficiently sterilised males whereas species (Dyck et al. 2005) and has been evalu- a high dose may undermine the insect’s ability ated against Anopheles albimanus Wiedemann to compete with wild conspecifics and may thus in the 1970s with encouraging results (Dame reduce the overall impact of the release. et al. 1981, Benedict and Robinson 2003). The In the context of SIT anopheline irradi- SIT relies on the sterilisation of insects by che- ation has been performed on a number of spe- mosterilisation (Sharma et al. 1978, Dame et cies and dose-response curves have been pub- al. 1981, Dame 1985) irradiation (Dyck et al. lished for An. albimanus (Ali and Rozenboom 2005) or modern biotechnological approaches 1972) An. pharoensis Theobald (Abdel-Malek et (Thomas et al. 2000, Alphey 2002, Alphey et al. 1975, Shoukry 1980) and An. stephensi Lis- al. 2002). Modern biotechnological approaches ton (Akram and Aslamkhan 1975, Sharma et al. based on transgenic organisms are promising 1978). An. arabiensis Patton has been studied in but at an early stage of development and no the light of genetic sexing systems (Lines and legal framework currently exists to facilitate Curtis 1985) and small-scale irradiation studies the introduction of such organisms in the wild (Curtis 1976) but no dose-response curve exists. (Scott et al. 2002, Knols and Louis 2006). Steri- Previous work has indicated that substantial lisation by irradiation or chemosterilants has inter-species variation in radiation sensitivity not been researched extensively for the last 30 is present (Curtis 1976) justifying the need for years with mosquitoes. Promising results were a dose-sterility curve for An. arabiensis. In mos- obtained with chemosterilants in terms of the quitoes both the pupal and the adult stage can level of sterility induced and competitiveness be irradiated. Pupal irradiation is easier to per- present (Dame 1985) but these have the disad- form but there is evidence of a reduced com- vantage of being mutagenic agents. They thus petitiveness when male pupae are irradiated present a potential hazard to humans during at high doses compared to adult irradiation the treatment process and non-target organ- (Andreasen and Curtis 2005). The objective of isms if residues persist in released individuals this study was to determine the dose-sterility (Dame 1985). Even though the actual amount curves for the pupal and adult stages of male of residue released in the environment was in An. arabiensis and to study the impact of irradi- fact very low due to careful rinsing of the pu- ation on a number of life-history traits including pae (LaBrecque et al. 1972) concerns about the adult emergence, longevity, and insemination possible negative effects on the environment ability. if large numbers of treated insects were to be released (Hayes 1968, Bracken and Dondale 1972) resulted in a disuse of chemosterilants for Methods mosquito control. Although it would be worth- while to identify additional compounds with Mosquitoes chemosterilant properties it remains doubtful if The mosquito strain used in the experiments the currently available ones will be acceptable is the “KGB” strain of An. arabiensis. The strain 54 for use in future genetic control programmes. originates from and has been colo- Sterilisation by irradiation remains the most nised since 1975 (Courtesy of MR4 CDC Atlanta practical way to sterilise the insects at present USA). Mosquitoes were reared at a density of and it has been argued that radiation sterilisa- ~ 750 larvae per tray (30 x 40 cm) containing ± tion should also be used to introduce the first 2 litre of deionised water (water depth 2 cm). 03 Radiation-induced sterility in An. arabiensis

Heating mats (J.B. Consulting Bratislava Slo- replicated twice and the doses between 50 and vakia) were used to maintain the water tem- 80 were replicated three times. perature at 28 ± 1°C. Larvae were fed a diet of fish-food (Aquaricare® Victor NY USA) daily (~ Parameters measured 0.3 mg/larva) that was powdered and passed Emergence through a 224 μm mesh sie Emergence from irradiated pupae was scored ve. Adults were maintained in the insectary at for both sexes as pupae were not sexed prior 28 ± 1°C and 80 ± 2% RH and all experiments to irradiation. Only mosquitoes that success- were conducted in standard 30 cm cubic cages. fully emerged were positively scored. Semi- The light regime was L10:D12 with a one hour emerged adults and dead adults were scored as simulated dusk and dawn period. Adult cages non-emerged. were continuously supplied with 10% sucrose solution [w/v]. Longevity The longevity of the irradiated mosquitoes Irradiation procedure was determined by the removal and counting Insects were exposed to gamma rays generat- of dead mosquitoes at 24 hr intervals (except ed by a cobalt-60 source (Gammacell 220 MDS during weekends when 48-72 hrs intervals were Nordion Ottawa Canada) with a dose rate of used). Cages were discarded when at least two- ca. 16 Gy/min. Insects were concentrated in the third of all the males had died. centre of the chamber to maximise dose uni- formity within the batch (Fig. 3.1). A dosimetry Sterility assessment system was used to measure the dose received Levels of sterility were observed by mating by the batch based on the Gafchromic HD-810 the irradiated males with un-irradiated virgin film (International Specialty Products NJ USA; females. To ensure virginity of females pupae FAO/IAEA/USDA 2003). Three dosimeters were were placed in individual vials prior to emer- included with each batch of insects and read af- gence. Mates were introduced into the experi- ter irradiation with a Radiachromic® reader (Far mental cages at a 1:1 ratio. For experiments West Technology Inc. California USA). with pupae mates were introduced into the cages the day after irradiation when the males Experimental setup had emerged. For the adult experiments mates Experiments were performed in series. Each were introduced after the irradiation on the series included a number of different irradiation same day. Mosquitoes were fed on the forearm doses and a control (Table 3.1). The experimen- of a volunteer for 10 min twice on consecutive tal mosquitoes in each series originated from days between day 2 and 5 after introduction of the same batch of insects and were randomly the females (with the exception of series 1 in distributed to treatment. Mosquitoes in the the adult stage when mosquitoes were fed on a control group underwent exactly the same han- membrane filled with human blood). Egg laying dling stages as the experimental mosquitoes occurred en masse in the cage. For each cage except the actual irradiation itself. Within each one egg bowl filled with water and lined with series the order of irradiation was assigned ran- wet filter paper was offered for five nights start- domly. We covered a spectrum of doses that ing two days after the first blood meal. Daily ranged from a control (no irradiation) to a dose or at 48 hr intervals eggs were removed and previously shown to yield almost complete ste- counted. For hatch rate determination the eggs rility (Abdel-Malek et al. 1967, Sharma et al. were thoroughly mixed and random samples 55 1978) and doses were 0, 25, 50, 60, 70, 80, 100, of eggs were placed in larger trays to allow for and 120 Gy for the pupal and adult experiments. hatching. Trays from control cages were filled In the pupal experiments two additional doses with ca. 200 eggs/tray and more eggs were of 35 and 45 Gy were included. Each dose was placed in one tray as the doses increased due to higher sterility levels of the eggs. Eggs were irradiated at once. Temperature during irradia-

Chapter 03 checked for hatch by counting and removing L1 tion was room temperature and between 18-21 larvae daily for seven consecutive days from the °C. After irradiation individual pupae were put trays. in small vials and left overnight for emergence. For each day of egg collection all eggs The following morning males were transferred or a sample of all eggs were used to determine to the cages according to treatment. hatch rates. Since eggs were thoroughly mixed the sample can be regarded as representative Adults of the total amount laid for that day and treat- Irradiation of adults was performed on speci- ment therefore hatch rates were weighted to mens < 24 hrs of age. Pupae were collected the total number of eggs collected for that day. from trays and allowed to emerge overnight For each treatment an average value was ob- in a standard cage. The following morning the tained per series by weighting the data for all mosquitoes were separated by sex and males egg collection days. The residual fertility was were placed in a small holding cup prior to ir- calculated as a percentage of the control fer- radiation and sugar water was offered on cot- tility of each series and subtracted from 100% ton wool. Fifty adult males were irradiated for (Abbott’s formula (Abbott 1925)) to give a value each dose. Adults were irradiated in a modified for radiation-induced sterility. version of the pupal irradiation device (Fig. 3.1). Fecundity (the average number of eggs Perma-gel® from an ice-pack (Ice-pakTM Cry- laid per female) was calculated by dividing the opak Industries Inc. Canada) surrounded a tube number of eggs laid per night by the number of in which a small vial was placed that contained females alive at the start of that night. A value the adults. The irradiation device was placed for each treatment per series was obtained by at ~ 4°C prior to irradiation to cool the gel and the sum of all egg laying nights and data were immobilise the adults during the irradiation. pooled per treatment for both stages. Adults were immobilised in an ice-box (~ 4°C) for 5 min before irradiation and transferred to Insemination the vial. The opening of the vial was closed with The proportion of females inseminated by ir- some cotton wool and placed in the irradiation radiated males was extrapolated from dissec- device. After irradiation the vial was opened in- tions of a sub-sample of females tested. After side the cage and the adults were left to recover egg laying a random selection of females was and disperse. taken from the cages and their spermathecae dissected to examine whether these had been inseminated. The presence of spermatozoa was confirmed using a compound microscope at 400x magnification.

Collection and irradiation of experimental mosquitoes Pupae Pupae were collected the day before irradiation at 3 pm from trays that had been cleared of all 56 their pupae before 9 am that day to ensure equal age of the pupae. At 11 am the next day the Figure 3.1. Insects were irradiated in a confined space pupae aged 20-26 hrs were irradiated. Pupae in the centre of the irradiation chamber to maximise were irradiated in a small plastic lid filled with dose uniformity within the batch. Devices used for the water (Fig. 3.1). For each dose 100 pupae were irradiation of pupae and adults are shown. 03 Radiation-induced sterility in An. arabiensis

Figure 3.2. Adult emergence for the different treatments of pupae irradiation. Triangles indicate individual values; means (± s.e.m.) if N> 2) are indicated by a horizontal line.

Statistical analysis Results Data were analysed using the following varia- bles: treatment (dose: 0-120 Gy) series (1-8) and Dosimetry confirmed that all doses delivered stage (pupae or adults). Unless stated otherwise lay within a 5% error range. All means given in for both stages data from the same treatments the text are ± s.e.m. between different series were pooled to get an average value per treatment. Proportional val- Emergence ues for adult emergence and insemination were Irradiation of 20-26 hrs old pupae had no effect arcsine-square-root-transformed to achieve on adult emergence. Overall emergence was normal distribution. Regression analysis was high (Fig. 3.2) on average 96 ± 0.01%. No signif- performed for adult emergence. In addition icant differences were observed in adult emer- correlation analyses (Pearson correlation-co- gence for all treatments (F1, 30= 0.11, p> 0.05). efficient) between dose and insemination and between fecundity and insemination were per- formed for both stages. Survival curves were Induced sterility analysed using Kaplan-Meier survival analyses. The total number of eggs laid in all experiments The obtained survival curves were pairwise was ~ 94,000 of which ~ 45,000 eggs were compared to the control using Mantel-Cox checked for hatching. The average number of log-rank tests. General Linear Models (GLM) eggs laid per female in all treatments was simi- were used to observe differences in fecundity lar to the control for both irradiated life stages between treatments for the two stages and (pupae: (F9, 22= 0.55, p> 0.05); adults: (F7, 3.93= to compare induced sterility levels between 1.13, p> 0.05; Table 3.2). stages. The calculated residual fertility values Control sterility (i.e. the number of were inverse-normal transformed to yield nor- eggs that naturally do not hatch) in the KGB mal equivalent deviates (N.E.D.) and irradiation colony is 23 ± 3% and was similar in both stages doses were log10 transformed to obtain a linear (t(11)= -0.37, p> 0.05). Corrected for this control relationship between dose and residual fertility. sterility the residual fertility decreased with in- A standard linear regression analysis was per- creasing radiation dose for both pupal and adult 57 formed for each stage with log10 (Dose/Gy) as stages (Fig. 3.3). A linear regression (pupae (F1, the independent and N.E.D. as the dependent 22= 159.6, p< 0.01; adults (F1, 16= 218.7, p< 0.01)) variable. All two-sided tests were performed was obtained after transformation of both axes using SPSS version 12 (SPSS Inc. Chicago IL). and the regression model can be used to predict Table 3.1. Effect of different irradiation doses on mean (± s.e.m.) survival times of An. arabiensis males irradiated in the pupal or adult stage. Log-rank tests compared treatments with the control, asterisks indicate significant differences at

Chapter 03 * p< 0.05 and ** p< 0.01. Survival time was estimated from Kaplan-Meier survival analysis (details see text). n/a: not applicable. Pupae Dose N Mean Log-rank Adult Dose N Mean Log-rank series (Gy) (± s.e.m.) series (Gy) (± s.e.m.) survival survival time time 1 0 31 11 ± 0.8 n/a 1 0 57 8 ± 0.5 n/a 25 35 7 ± 0.7 14.66** 60 55 9 ± 0.5 1.59 35 27 8 ± 0.8 6.09* 70 56 9 ± 0.5 5.10* 2 0 56 8 ± 0.5 n/a 80 52 8 ± 0.5 0.04 50 48 8 ± 0.5 0.84 2 0 54 8 ± 0.5 n/a 70 55 7 ± 0.6 0.10 50 59 8 ± 0.5 0.07 3 0 48 9 ± 0.8 n/a 60 51 8 ± 0.5 0.16 80 51 10 ± 0.9 4.93* 70 56 8 ± 0.5 0.11 100 61 10 ± 0.8 1.63 80 50 10 ± 0.8 2.60 4 0 46 9 ± 0.7 n/a 100 55 8 ± 0.5 0.00 35 54 12 ± 0.7 2.70 3 0 49 9 ± 0.7 n/a 45 58 9 ± 0.5 1.08 25 48 9 ± 0.7 0.84 50 53 11 ± 0.5 0.58 50 45 9 ± 0.7 0.05 5 0 32 9 ± 0.7 n/a 100 47 9 ± 0.7 0.44 60 45 8 ± 0.5 0.34 4 0 48 11 ± 1.0 n/a 70 35 10 ± 0.9 1.29 25 48 13 ± 1.0 1.06 80 35 10 ± 0.6 0.77 50 48 13 ± 0.9 1.79 6 0 44 7 ± 0.7 n/a 60 48 11 ± 0.9 0.29 45 42 9 ± 0.8 3.74 70 46 12 ± 1.0 0.59 60 42 7 ± 0.8 0.19 80 49 10 ± 1.0 0.00 70 41 9 ± 0.7 2.32 5 0 48 13 ± 1.0 n/a 80 41 9 ± 0.7 2.94 120 46 14 ± 0.9 0.64 100 43 8 ± 0.6 0.70 120 46 13 ± 1.0 0.72 7 0 44 7 ± 0.6 n/a 25 45 12 ± 0.9 10.63** 45 42 11 ± 0.5 11.57** 50 41 9 ± 0.5 0.07 60 45 10 ± 0.5 5.73* 50 42 11 ± 0.8 8.10** 8 0 47 15 ± 1.0 n/a 120 46 11 ± 1.0 6.51* 120 49 13 ± 0.9 2.89

induced sterility rates (= 1- residual fertility) at lationship (LaChance 1967, Curtis 1971). Graphs specific irradiation doses. for the pupal and adult stage show a predomi- For the pupal stage the equation is: nantly linear relationship (Fig. 3.3) suggesting a

N.E.D. (residual fertility) = 6.65 – 4.30 (log10(dose one-hit relationship to dose; i.e. a large propor- (Gy)) with r2= 88%; tion of dominant lethal mutations result from for the adults: single events in the gametes. As expected at

N.E.D. (residual fertility) = 6.66 – 4.42 (log10(dose higher doses the lines tend to depart from lin- (Gy)) with r2= 93%. earity suggesting that gametes carry more than 58 Plotting fertility on a logarithmic scale one dominant lethal mutation. against dose provides insight in the nature of When comparing the two stages germ dominant lethal mutations. A linear response cells of adults appear to be slightly more sus- indicates a “one-hit” relationship whereas de- ceptible to irradiation at doses between 60 and partures from linearity indicate a “multi-hit” re- 80 Gy (Fig. 3.4 and Table 3.2) compared to the

03 Radiation-induced sterility in An. arabiensis

pupal stage but differences are not significant fecundity (number of eggs/female) and insemi-

(Stage* Dose: F6, 26= 1.50, p> 0.05). nation in the pupal stage (r= 0.75, p< 0.01) and in the adult stage (r= 0.58, p< 0.01). Mosquito survival The recovery of adults after irradiation was 100%. In total survival was scored for 2585 Discussion males. Because between series survival of the controls was significantly different in the pupal Ionising radiation has over the years proven to (log-rank= 57.02, df = 7, p< 0.01) and adult stage be a safe and reliable way to induce sterility in (log-rank= 31.9, df= 4, p< 0.01) survival curves a large variety of insects (Dyck et al. 2005). The of the different treatments were analysed per potential use of the SIT against malaria vectors series against the control. For the pupal stage is currently explored and the development of the survival curve for irradiated males was not radiation sterilisation protocols is a vital part of significantly lower than the control except in such endeavour. The majority of the experimen- treatments 25 and 35 Gy in series 1 and 120 Gy tal work on Anopheles irradiation has focused in series 8 lower survival was observed (Table on the pupal stage (Abdel-Malek et al. 1967, 3.1). Higher survival was also found in certain Tantawy et al. 1967, Ali and Rozenboom 1972, treatments. In the adult stage survival of irradi- Abdel-Malek et al. 1975, Sharma et al. 1978) as ated males was similar to the control with the pupae are more robust and easier to handle. The exception of 70 Gy males in series 1 for which irradiation of pupae is preferably performed on survival was higher (Table 3.1). older (> 15 hrs) pupae since irradiation of young pupae results in a reduced emergence (Abdel- Insemination Malek et al. 1967, Curtis 1976) and shorter sur- A total number of 1149 females was dissected vival (Davis et al. 1959). In this study age of the (on average 22.1 ± 0.6 females per cage) and pupae when irradiated was between 20-26 hrs examined for insemination (except in series 1 and ~ 10 hrs before emergence. We found no for the pupal stage; no dissections were per- effect of pupal irradiation on adult emergence formed). After on average ten days of mating even when the dose applied was high and simi- control insemination was 82 ± 4% for females lar results were reported for pupal irradiation of confined with males irradiated at the pupal An. pharoensis (Abdel-Malek et al. 1967). stage and 90 ± 3% for those with males irradi- Although irradiation is intended to tar- ated in the adult stage and no significant dif- get the germ cells the process is non-specific ferences were observed between both controls and somatic cells may also be damaged. One (t(10)= -0.87, p> 0.05). For the pupae a signifi- of the commonest effects of somatic damage cant negative correlation (r= -0.49, p< 0.01) was is a reduced longevity (Proverbs 1969). In this found between dose and insemination (Fig. 3.5) study males irradiated as adults with increasing but the correlation-coefficient explained less doses had similar survival times compared to than 25% of the variation. For adults no correla- the controls. In the males irradiated as pupae a tion was observed (r= 0.05, p> 0.05). Individual somewhat more variable pattern was observed t-tests showed that insemination rates were with in general similar or higher survival and in not statistically different from the control for few cases reduced survival but differences were all treatments in both stages except at 100 Gy small. In An. pharoensis (Abdel-Malek et al. for the pupal stage (t(7)= 2.46, p< 0.05) a lower 1966) a slight increase in longevity of males ir- mean insemination was observed. When com- radiated with 5-70 Gy as pupae compared to the 59 paring the two stages at individual doses no sig- control was reported. Other studies report a re- nificant differences were observed (independ- duced longevity in An. pharoensis (Abdel-Malek ent t-tests) but overall in the adult stage higher et al. 1967) for pupae irradiated at 100-130 Gy insemination rates were found (F1, 50= 17.45, p< and in An. stephensi (Sharma et al. 1978) for 0.01). A positive correlation was found between pupae irradiated at 80 and 120 Gy. Sampling of Chapter 03

Figure 3.3. Regression line (± 95% CI for mean (thin line) and individual values (broken line)) for fertility versus irradiation dose (Gy) for male An. arabiensis pupae (A) and adults (B). Symbols indicate observed individual values. Means (± s.e.m. if N> 2) are indicated by a horizontal line.

mosquitoes occurred at 48-72 hrs intervals over similar compared to the control. Due to lower the weekends, which could have influenced con- insemination of females placed with males ir- sistency of the data to some extent. However radiated at higher doses in the pupal stage a as the controls were exposed to the same treat- positive correlation was observed between fe- ment and subsequent analyses we deem this cundity and insemination rate. Overall fecun- variation caused by different sampling intervals dity rates were variable because eggs were col- of negligible importance. The possibility that lected en masse and fecundity calculated over irradiation has a beneficial effect on longevity all females alive regardless if they had blood fed in the pupal stage cannot be excluded from our or oviposited. Differences observed in fecundity results yet longevity was not measured under are partly accounted for by the reduced insemi- stressful conditions. In the Mediterranean fruit nation rate found in certain treatments. Unlike fly Ceratitis capitata Wiedemann it is known in some other mosquito colonies un-inseminat- that under stress the possible negative effects ed females of the KGB strain used in these ex- of radiation tend to become more pronounced. periments have not been observed to oviposit Quality control testing in Mediterranean fruit resulting in fewer eggs when the insemination fly SIT programmes measures longevity after rate is low. Other inexplicable variation in egg the deprivation of food for some time (FAO/ batch sizes have been observed yet should not IAEA/USDA 2003). Although mosquitoes are influence our conclusions since these occurred 60 more sensitive to complete food deprivation throughout the experimental period and across similar tests can be devised for mosquitoes to all treatments. assess the impact of irradiation. The dose-sterility curves in An. arabien- The fecundity of the females mated sis for pupae and adults show the classic pattern with irradiated males for all treatments was found for such curves of a linear relationship at 03 Radiation-induced sterility in An. arabiensis

Figure 3.4. Comparative analysis of dose-sterility data from this study (solid line for An. arabiensis adults, dotted line for pupae) and published reports on anopheline irradiation (20-120 Gy). Induced sterility levels were calculated (Abbott’s formula (Abbott 1925)) from observed sterility levels for (Abdel-Malek et al. 1975, Sharma et al. 1978). Points for 120 Gy are spread along the x-axis for clarity purposes. Legend: +: An. albimanus pupa (Ali and Rozenboom 1972); □: An. stephensi pupa (Sharma et al. 1978, Andreasen and Curtis 2005); ■: An. stephensi adult (Akram and Aslamkhan 1975, Curtis 1976, Andreasen and Curtis 2005); : An. pharoensis pupa (Abdel-Malek et al. 1967, Abdel-Malek et al. 1975); : An. pharoensis adult (Shoukry 1980); ∆: An. gambiae pupae (Curtis 1976, Andreasen and Curtis 2005); : An. gambiae adult (Curtis 1976, Krafsur and Davidson 1987, Andreasen and Curtis 2005); : An. arabiensis pupae (Curtis 1976); – : An. arabiensis adult (Curtis 1976, Lines and Curtis 1985). lower doses and the flattening of the curve at sterility in An. arabiensis with other anophe- higher doses (LaChance and Graham 1984, Rob- lines (Fig. 3.4) An. arabiensis shows a greater inson 2002). The highest dose of 120 Gy induced radiation resistance resulting in lower induced > 99% sterility. Overall we found that fertility is sterility levels. In the pupal stage An. albimanus slightly more sensitive to exposure in the adult (Ali and Rozenboom 1972) behaved similarly stage than in the pupal stage but differences at doses > 40 Gy while An. stephensi (Sharma are small. There are only a few studies with et al. 1978) and An. pharoensis (Abdel-Malek anophelines that have irradiated both stages si- et al. 1975) overall had slightly higher induced multaneously. At 120 Gy equal levels of sterility sterility levels. In the adult stage An. pharoen- were found for pupal and adult irradiation in An. sis (Shoukry 1980) and especially An. stephensi gambiae s.s. and An. stephensi (Andreasen and (Akram and Aslamkhan 1975) showed higher Curtis 2005); while in An. gambiae s.s. pupae sterility levels. Only in An. gambiae s.s. (Curtis were more radiation-resistant than adults (Cur- 1976) at 80 Gy a lower level of induced steril- tis 1976). If we compare levels of sterility found ity was observed. In a study in which adult An. in different studies An. stephensi (Akram and arabiensis were irradiated with 40 Gy (Lines and 61 Aslamkhan 1975) and An. pharoensis (Shoukry Curtis 1985) a somewhat higher level of sterility 1980) both show higher sterility levels when ir- was reported but the causes for this remain un- radiated in the adult stage (Fig. 3.4). clear. In the majority of the studies mentioned When comparing the level of induced above no mention was made if dosimetry was

Table 3.2. Mean (± s.e.m.) induced sterility levels and fecundity of females mated with An. arabiensis males irradiated in the pupal or adult stage. A range is indicated where number of replicates is <2. N: total number of females alive at onset

Chapter 03 of egg-laying period; n/d: not determined. † fecundity of females fed on membrane not included.

Treatment Pupal irradiation Adult irradiation dose (Gy) replicates induced sterility fecundity N replicates induced sterility fecundity N

0 7 0.0 49 ± 5.9 306 4 0.0 48 ± 7.2 170

25 2 35.4 (32.3-38.5) 32 (25-38) 56 2 38.9 (38.7-39.0) 69 (62-76) 84

35 2 44.8 (35.7-53.8) 52 (49-54) 59 n/d

45 3 68.3 ± 6.9 42 ± 5.7 125 n/d

50 4 76.0 ± 4.6 34 ± 5.0 178 3 71.7 ± 5.8 45 ± 11.0 135

60 3 78.6 ± 5.7 39 ± 8.6 104 2 88.2 ± 2.2 36 (16-56)† 86

70 3 83.4 ± 2.1 41± 12.0 120 2 92.4 ± 2.5 53 (36-70)† 85

80 3 91.0 ± 3.4 42 ± 5.5 104 2 96.7 ± 0.5 44 (36-52)† 75

100 2 98.6 (98.5-98.7) 34 (26-41) 101 2 98.1 (97.2-98.9) 41 (21-60) 86

120 2 99.5 (99.3-99.7) 73 (66-79) 89 2 99.7 (99.6-99.7) 43 (18-68) 89

used to verify the absorbed dose by the batch Conclusion therefore caution in interpreting sterility data across experiments needs to be taken. In the past SIT focused on the induction of al- Although competitiveness experi- most 100% sterility and this led to the use of ments will be used to assess male fitness at dif- high radiation doses. Over the years it was ob- ferent radiation doses the level of insemination served that some insect species irradiated with in the absence of competition gives some indi- these high doses were not able to suppress the cation of the male’s ability to mate. We found natural population due to a lack of competitive- a substantial variation in insemination within ness. A revision of requiring 100% sterility was some treatments but the number of replicates needed and it is suggested that more sterility was low. The insemination of females mated can be induced in the target population if in- with males irradiated as adults was comparable sects are more competitive when subjected to to the control. For males irradiated at the pupal partially-sterilising doses (Toledo et al. 2004, stage a negative correlation was found between Dyck et al. 2005). Male competitiveness at insemination and dose but the correlation-coef- these lower doses needs to be estimated in a ficient (r2) was associated with less than < 25% semi-field setting where irradiated males com- of the total variation and only the dose of 70 pete with un-irradiated wild males for wild fe- Gy was different to the control. When the two males. On the basis of such experiments the stages were compared at individual doses no optimal dose for the released male mosquitoes differences were observed. In previous studies can be identified. On the other hand reduced equal insemination (Ali and Rozenboom 1972) competitiveness can be compensated for by or similar egg production (Abdel-Malek et al. increasing the flooding ratio of sterile males in 62 1967, Abdel-Malek et al. 1975) in females mated the field (Knipling 1955) but this would increase to males irradiated as pupae was observed. The the costs of the programme. results in this study suggest a decline in mating The choice of developmental stage for ability with increasing dose in males irradiated irradiation in an SIT programme depends on as pupae but not as adults. numerous factors including handling, survival, 03 Radiation-induced sterility in An. arabiensis

Figure 3.5. Insemination of virgin females mated with An. arabiensis males irradiated in pupal or adult stage for the different treatments. Triangles represent pupal individual values circles adult individual values. Means (± s.e.m. if N> 2) are indicated by a horizontal line. Linear regression lines (± 95% CI for mean) are given; solid line is pupae dotted line adults. For clarity purposes data points for adults have been shifted slightly to the right.

sterility, competitiveness, and release meth- that pupae irradiation reduces competitiveness odology and this study has focused on the first more so than adult irradiation (Curtis 1976, An- three factors. The mosquitoes in this study were dreasen and Curtis 2005). Competitiveness of irradiated in small numbers and both stages sur- pupae irradiated at partially-sterilising doses vived the handling and irradiation process well. has hardly been studied; but in An. stephensi Up-scaling the irradiation process for mass pro- it was observed that males irradiated as pupae duction remains a challenge. Pupal irradiation with 80 Gy were 1.7 times more competitive has a number of advantages over adults but not than males irradiated with 120 Gy (Sharma et enough resources have been directed to the de- al. 1978). Future studies will focus on compe- velopment of large-scale irradiation devices to tition experiments that will include the use of draw conclusions. fully-sterilising and partially-sterilising doses In this study fertility was slightly more for both developmental stages. sensitive to irradiation exposure in the adult stage than in the pupal stage but differences were small. The trend to reduced insemination Acknowledgements 63 rates at higher doses in the pupal stage sug- The authors would like to thank G. Germer- gests that pupae are more somatically dam- shausen for dedicated laboratory assistance. aged by the irradiation process than adults a A. Odulaja provided valuable guidance with finding supported by studies that observed the statistical analyses and we thank A. Robin- son and the two anonymous reviewers for their Sinkins, A. Spielman, Y. Touré, and F. H. Collins

Chapter 03 comments, which significantly improved the 2002. Malaria control with genetically manipulated quality of the manuscript. BGJK is supported by insect vectors. Science 298:119-121. a VIDI grant from the Netherlands Organisation Andreasen, M. H., and C. F. Curtis 2005. Optimal for Scientific Research (#864.03.004). life stages for radiation sterilization of Anopheles for sterile insect releases. Med. Vet. Entomol. 19:238- 244. Benedict, M. Q., and A. S. Robinson 2003. The first References releases of transgenic mosquitoes: an argument for the sterile insect technique. Trends Parasitol. 19:349- Abbott, W. S. 1925. A method of computing the 355. effectiveness of an insecticide. J. Econ. Entomol. Bracken, G. K., and C. D. Dondale 1972. Fertility 18:265-267. and survival of Achaearanea tepidariorum Abdel-Malek, A. A., A. O. Tantawy, and A. M. (Araneida:Theridiidae) on a diet of chemosterilized Wakid 1966. Studies on the eradiaction of Anopheles mosquitoes. Can. Entomol. 104:1709-1712. pharoensis Theobald by the Sterile-Male technique Curtis, C. F. 1971. Induced sterility in insects. Adv. using cobalt-60. I.Biological effects of gamma Reprod. Physiol. 5:120-165. radiation on the different developmental stages. J. Curtis, C. F. 1976. Radiation sterilization. Report Econ. Entomol. 59:672-678. on mosquito research, Ross Institute of Tropical Abdel-Malek, A. A., A. O. Tantawy, and A. M. Hygiene. 01.01.76-31.12.77. Wakid 1967. Studies on the eradication of Anopheles Dame, D. A., R. E. Lowe, and D. L. Williamson 1981. pharoensis Theobald by the Sterile-Male technique Assessment of Released Sterile Anopheles albimanus using cobalt-60. III. Determination of the Sterile and Glossina morsitans morsitans, pp. 231-248. In Dose and its biological effects on different characters J. B. Kitzmiller, and T. Kanda (eds.), Cytogenetics related to “Fitness” components. J. Econ. Entomol. and genetics of vectors. Elsevier Biomedical, 60:20-23. Amsterdam. Abdel-Malek, A. A., A. M. Wakid, A. O. Tantawy, Dame, D. A. 1985. Genetic control by sterilized and L. M. El-Gazzar 1975. Studies on factors mosquitoes, pp. 159-172. In R. Chapman, R. Barr, influencing the induction of sterility in Anopheles D. E. Weidhaas, and M. Laird (eds.), Biological pharoensis Theobald by gamma radiation, pp. 161- Control of Mosquitoes. American Mosquito Control 174. In The use of isotopes and pesticides in pest Association, Bull 6. control, Proceedings of a symposium held in Beirut- Davis, A. N., J. B. Gahan, D. E. Weidhaas, and C. Lebanon, March 1974. N. Smith 1959. Exploratory studies on gamma Akram, M., and M. Aslamkhan 1975. Production of irradiation for the sterilization and control of dominant lethal mutations by gamma-irradiation in Anopheles quadrimaculatus. J. Econ. Entomol. the malaria mosquito, Anopheles stephensi. Pak. J. 52:868-870. Zool. 7:177-184. Dyck, A., J. Hendrichs, and A. S. Robinson 2005 Ali, S. R., and L. E. Rozenboom 1972. Observations (eds.). The Sterile Insect Technique: Principles and on sterilization of Anopheles (C.) albimanus Practice in Area-Wide Integrated Pest Management. Wiedemann by X-irradiation. Mosq. News 32:574- Springer, Dordrecht. 579. FAO/IAEA/USDA 2003. Manual for Product Quality Alphey, L. 2002. Re-engineering the sterile insect Control and Shipping Procedures for Sterile Mass- technique. Insect Biochem. Mol. Biol. 32:1243-1247. Reared Tephritid Fruit Flies, Version 5.0, pp. 85 64 Alphey, L., C. B. Beard, P. Billingsley, M. Coetzee, pp. International Atomic Energy Agency, Vienna, A. Crisanti, C. Curtis, P. Eggleston, C. Godfray, J. Austria. Hemingway, M. Jacobs-Lorena, A. A. James, F. C. Hayes, W. J. 1968. Toxicological aspects of Kafatos, L. G. Mukwaya, M. Paton, J. R. Powell, chemosterilants, pp. 315-347. In G. C. LaBrecque, W. Schneider, T. W. Scott, B. Sina, R. Sinden, S. and C. N. Smith (eds.), Principles of insect 03 Radiation-induced sterility in An. arabiensis

chemosterilisation. Appleton Century Crofts, New Thomas, D. D., C. A. Donnelly, R. J. Wood, and L. York. S. Alphey 2000. Insect population control using a Knipling, E. F. 1955. Possibilities of insect population dominant, repressible, lethal genetic system. Science control through the use of sexually sterile males. J. 287:2474-2476. Econ. Entomol. 48:459-462. Toledo, J., J. Rull, A. Oropeza, E. Hernández, and Knols, B. G. J., and C. Louis 2006 (eds.). Bridging P. Liedo 2004. Irradiation of Anastrepha obliqua Laboratory and Field Research for Genetic Control of (Diptera: Tephritidae) revisited: Optimizing sterility Disease Vectors. Frontis, Wageningen. induction. J. Econ. Entomol. 97:383-389. Krafsur, E. S., and G. Davidson 1987. Production of semisterile mutants in the Anopheles gambiae s.l. species complex. Iowa State J. Res. 62:85-97. LaBrecque, G. C., W. C. Bowman, R. S. Patterson, and J. A. Seawright 1972. Persistence of thiotepa and tepa in pupae and adults of Culex fatigans. Bull. World Health Organ. 74:675-676. LaChance, L. E. 1967. The induction of dominant lethal mutations in insects by ionizing radiation and chemicals - as related to the sterile-male technique of insect control, pp. 617-650. In J. W. Wright, and R. Pal (eds.), Genetics of insect vectors of disease. Elsevier, Amsterdam, The Netherlands. LaChance, L. E., and C. K. Graham 1984. Insect radiosensitivity: dose curves and dose-fractionation studies of dominant lethal mutations in the mature sperm of 4 insect species. Mutat. Res. 127:49-59. Lines, J. D., and C. F. Curtis 1985. Genetic sexing system in Anopheles arabiensis Patton (Diptera: Culicidae). J. Econ. Entomol. 78:848-851. Proverbs, M. D. 1969. Induced sterilization and control in insects. Annu. Rev. Entomol. 14:81-102. Robinson, A. S. 2002. Mutations and their use in insect control. Mutat. Res. 511:113-132. Scott, T. W., W. Takken, B. G. J. Knols, and C. Boëte 2002. The ecology of genetically modified mosquitoes. Science 298:117-119. Sharma, V. P., R. K. Razdan, and M. A. Ansari 1978. Anopheles stephensi: effect of gamma-radiation and chemosterilants on the fertility and fitness for sterile male releases. J. Econ. Entomol. 71:449-452. Shoukry, A. 1980. X-ray induced sterility, dominant lethality and inherited semisterility in Anopheles pharoensis Theo (Dipt., Culicidae). Z. Angew. Entomol. 89:498-504. Tantawy, A. O., A. A. Abdel-Malek, and A. M. 65 Wakid 1967. Studies on the eradication of Anopheles pharoensis Theobald by the Sterile-Male technique using cobalt-60. V. Mating competitiveness of radiosterilized males. J. Econ. Entomol. 60:696-699. 04 Sperm quantity and polymorphism in An. arabiensis 04 Sperm quantity and size polymorphism in un-irradiated and irradiated males of the malaria mosquito Anopheles arabiensis Patton by Michelle EH Helinski and Bart GJ Knols

Anopheles mosquitoes are important candidates for genetic control strategies. However, little is known about sperm quality and quantity as determinants of male reproductive success. In this study, sperm quantity and size polymorphism was assessed in testes of normal and irradiated Anopheles arabiensis. The effects of irradiation were evaluated for males exposed in the pupal or adult stage to a full (120 Gy) or partially-sterilising dose (70 Gy). We also determined the size distribution of sperm lengths in spermathecae of inseminated females compared to that observed in350-500 male Million testes. cases Sperm of quantity clinical malaria increased occur with annually, age, and 60% 12-day-old of which are males in sub- had significantlySaharan Africa. more Moreover, sperm in 80%their oftestes all deaths(8214 ± attributed467) than tomales malaria aged occur3 days in(5022 this ± 375).region. Mosquitoes In numbers, irradiated 1 million in Africansthe pupal die stage of the had disease significantly each year, fewer with sperm the (2982 vast ±majority 125) than of deaths un-irradiated occurring males among (4950 children ± 848). below For adult five stage years irradiation of age. Pregnant similar amountswomen are of anothersperm were major observed risk group; compared malaria to can un-irradiated cause low birth males. weight Sperm and length pre- polymorphismmature delivery was (Rogerson detected et with al. 2007). sperm The lengths impact ranging of malaria between on the < 50-500 economic μm. Pupalsituation irradiation of endemic resulted countries in a significant is high (Gallup increase and inSachs sperm 2001), numbers and the in thecorrelation testes in thebetween category poverty of 100-200 and malaria μm. Sperm clearly length demonstrated distributions (Sachs in the and spermathecae Malaney 2002). were differentMalaria to those is measureda parasitic directly disease from transmitted sperm in the by testes female and mosquitoesharboured less of thecells ofgenus the smallerAnopheles (<100-200. The malaria µm), and parasite more iscells a protozoan of the larger of thecategory genus (300-400 Plasmodium µm).. TheThere findings are four that species testes of of malaria pupal irradiated parasites thatmales can produce infect fewer humans and undersmaller natural sperm inconditions: comparison Plasmodium to un-irradiated falciparum testes,, P. vivaxcombined, P. ovale with and the P. fact malariae that spermathecae. The first two selectspecies for cause larger the sperm, most are infections discussed worldwide in the context and P. of falciparum genetic control Welsh strategies, is by far the in particular the Sterile Insect Technique (SIT). 67

Submitted to Acta Tropica Introduction Our goal was to determine the range of sperm

Chapter 04 quantity in testes of a laboratory strain of An. he ability of released males to locate, arabiensis Patton over time. We hypothesised copulate with, and transfer sterile sperm that, like for Ae. aegypti, the quantity of sperm Tto wild females is of great importance in males increases with age. for the success of genetic control programmes Besides sperm quantity, variation in like the Sterile Insect Technique (SIT) (Dyck et sperm morphology (if present in a species) al. 2005). Since the mid 1990s, genetic control is another characteristic of interest from an of mosquitoes has gained renewed interest evolutionary perspective. In a wide variety of because of the rapid advances in modern bio- invertebrates, different morphological types technology (Thomas et al. 2000, Alphey 2002, of sperm have been described (Swallow and Scott et al. 2002, Benedict and Robinson 2003), Wilkinson 2002). Often, one of the sperm types and the failure of conventional control methods is incapable of fertilisation. For instance, moths (Dame and Curtis 1996, Curtis 2002, Helinski et and butterflies produce large numbers of anu- al. 2006a). However, for Anopheles mosquitoes, cleated sperm, and a number of theories on the little is known on the biology and reproduc- potential role of these sperm types exist (Swal- tive potential of males (Ferguson et al. 2005). low and Wilkinson 2002). In Drosophila species, Reproductive potential in SIT studies is usually two classes of sperm with discrete tail-lengths determined based on the competitiveness of are produced (Swallow and Wilkinson 2002). males, i.e. how well males are able to compete Both sperm morphs contain a nucleus, however against wild males for mates (Calkins and Park- studies have shown that only the larger morph is er 2005, Helinski and Knols 2008). In Anopheles, used for the fertilisation of the eggs (Snook and reproductive potential depends on age of the Karr 1998), and the role of the non-fertilizing male (Mahmood and Reisen 1982, Verhoek and short sperm has yet to be determined (Swallow Takken 1994), and his age at the time of insemi- and Wilkinson 2002). Recently, sperm polymor- nation has an impact on female reproduction phism was studied in a number of anopheline (Chambers and Klowden 2001). Larval rearing species including An. gambiae s.s. Giles (Klow- conditions also determine reproductive success den and Chambers 2004). It was observed that whereby crowding results in less competitive sperm varied in tail length, and a wide range males (Ng’habi et al. 2005). of tail lengths were observed in some species. One overlooked area of reproductive The range was greater in species belonging to success in male mosquitoes is the quality and species complex groups, implicating a possible quantity of sperm. Few studies exist on the de- role of sperm length polymorphism in specia- termination of sperm quantity in the testes of tion processes (Klowden and Chambers 2004). mosquitoes. A recent study on Aedes aegypti Sperm of all lengths in An. gambiae s.s. were L. observed an increase of the number of sper- nucleated suggesting that they were capable matozoa with age, and a positive correlation of fertilisation (Klowden and Chambers 2004). between body size and sperm quantity (Ponla- The function of the presence of sperm length wat and Harrington 2007). Direct assessment polymorphism in Anopheles mating systems is of sperm quantity in the testes of anopheline not known. The driving force behind the pres- mosquitoes has not been undertaken before. ence of sperm polymorphism in other species is A number of studies were performed on the thought to be competition among the sperm of morphology of the male reproductive system multiple males in the female (Clark 2002). How- 68 associated with aging and mating, and an in- ever, Anopheles mating systems appear to be crease in the sperm reservoir with age was ob- largely monandrous, and a low frequency (i.e. served suggesting continued spermatogenesis 2.5 %) of multiple mating was observed in field- during adult life (Mahmood and Reisen 1982, collected An. gambiae s.s. females (Tripet et Mahmood and Reisen 1994, Huho et al. 2006). al. 2003). Sperm length polymorphism has not 04 Sperm quantity and polymorphism in An. arabiensis

Figure 4.1. A) Reproductive system of male Anopheles arabiensis photographed at 100x. Testis (T), accessory gland (Ag), ejaculatory duct (Ed), seminal vesicle (Sv), and vas deferens (Vd) are indicated. B) Long and short spermatozoa of An. arabiensis. Intact sperm could be distinguished by a clear head and a tapered tail. been documented in An. arabiensis and our aim mainly during the late larval and pupal stages was to determine if it exists in this species using (Clements 1992), although the maturation of a laboratory colony of An. arabiensis and males new sperm continues during adult life (Mah- collected from the wild. We hypothesised that mood and Reisen 1994, Huho et al. 2006). In sperm length polymorphism would be present this study, the influence of irradiation on sperm in An. arabiensis because it belongs to a taxon quantity and size polymorphism in testes of ir- with seven known sibling species (i.e. the An. radiated males was determined. Males were ir- gambiae s.l. species complex, see Hunt et al. radiated in the pupal or adult stage with a low (1998)). (70 Gy), partially-sterilising or a high (120 Gy), SIT programmes for anophelines will fully-sterilising dose (Helinski et al. 2006b). initially rely on irradiation to induce sterility (Helinski et al. 2006a) and the first releases of transgenic mosquitoes have been proposed using this sterilisation method (Benedict and Material & Methods Robinson 2003). Irradiation induces dominant lethal mutations in the germ cells, and dividing Mosquitoes cells are especially prone to irradiation damage The mosquito strain used for the experiments (LaChance 1967). In general, the earlier stages was the “Dongola” strain of An. arabiensis of spermatogenesis (spermatocytes and sper- (available from MR4, CDC Atlanta, USA) colo- matogonia) are more radiosensitive than later nised in 2004 from specimens collected near stages (spermatids and spermatozoa) in terms Dongola, Northern State, Sudan. This strain has of irreversible irradiation damage resulting in been maintained in the laboratory for approxi- death of the developing cell (Proverbs 1969, mately sixty generations. Larvae were reared at Anwar et al. 1971, Bakri et al. 2005). In terms of a density of 800 L1 larvae in trays containing ± induced sterility, the older stages are more radi- 1.5 litre of deionised water. Heating mats (J.B. 69 osensitive resulting in greater embryonic mor- Consulting, Bratislava, Slovakia) were used to tality after fertilisation (Sobels 1969, Anwar et maintain the water temperature at 28 ± 1 °C. al. 1971, Lecis et al. 1975). In mosquitoes, earlier Larvae were fed a diet of fish-food (Aquari- stages of spermatogenesis are thought to occur care®, Victor, NY, USA) that was powdered and passed through a 224 μm mesh sieve, in a male were made (i.e. for a total of 11 males;

Chapter 04 aliquots of ~ 0.25 mg/larvae/day. Adults were Table 4.1) and the different counts compared. kept in standard 30 x 30 x 30 cm mosquito rear- Slides were air dried and fixed with ing cages and maintained at a temperature of 96% ethanol. After fixing, slides were stained 27 ± 1 °C and relative humidity of 80 ± 2%. The with Giemsa dye (Sigma-Aldrich, Germany) for light regime was L10:D12 with a one hour simu- 1 hr, then rinsed with distilled water, and then lated dusk and dawn period. Adult cages were allowed to air dry. Slides could be used to de- continuously supplied with 6% [w/v] sucrose termine sperm quantity and sperm length poly- solution. Wild males, used for the analysis of morphism. A wing was clipped from each male sperm polymorphism, were collected as larvae to determine wing size as a measure of body from breeding sites in Northern State, Sudan, size. A digital image of the wing was taken (CC- in 2007. Until dissection, they were maintained 12 camera, Soft Imaging System, Germany, in an insectary under similar conditions as de- mounted on a stereo microscope). Wing length scribed above. was measured between the alula notch and the wing tip, excluding scales. Measurements were Dissections performed with AnalySIS FIVE software (Soft The protocol for sperm quantification described Imaging System, Germany). by Ponlawat and Harrington (2007) for Ae. ae- Spermathecae were removed from gypti dissections was amended as follows. Male females and dissected in 2.5 µl of PBS and reproductive organs (i.e. testes, seminal vesi- covered with a coverslip. The pressure of the cles, and accessory glands; Fig. 4.1A) were dis- coverslip caused the spermathecae to burst sected using small insect pins under a dissect- and sometimes good numbers of sperm cells ing microscope in 10 µl of phosphate buffered were released. Gentle tapping on the coverslip saline (PBS) on a silane-coated (i.e. hydropho- dispersed the cells further, and sperm lengths bic) coverslip mounted on a dissection slide. were determined as described below. Testes were separated from other tissue by cut- ting the vas deferens with pins and removing Sperm quantity the other tissue (unlike Aedes, Anopheles males Mosquito sperm cells were counted under store sperm exclusively in the testes and not dark field microscopy at 100x. For each slide, in the seminal vesicle). Both testes were torn 10 spots were counted and the mean number open with the pins, and sperm separated in the of sperm cells per spot determined. The total drop of buffer by gentle stirring until no clumps amount of sperm was then extrapolated from were observed. The drop containing the sperm the mean. For every male, 2-3 slides were made was then transferred to a 1 ml Eppendorf tube, and the average count determined and used and the empty coverslip surface was rinsed 4 for data analysis. Even though dissection slides times with 10 µl of PBS to collect remaining were well rinsed, it was impossible to remove cells. Four hundred µl (2 x 200 µl) of PBS solu- all sperm cells. Therefore, dissection slides tion was added to the tube to obtain the final were stained, and an estimate of the number of stock volume of 450 µl. The sample stock solu- remaining sperm cells was made and included tion was well mixed (i.e. 15 times each) with a in the overall number of sperm cells. P200 pipette between and after adding the 2 x 200 µl solution. A sub-sample from the stock Sperm length polymorphism was then spotted on a silane-coated slide and, Sperm cells of the testes were photographed 70 on each slide, 10 separate spots of 2.5 µl were with phase-contrast illumination at 200x us- made. If multiple slides for one sample stock ing a camera (Colorview IIIu, Soft Imaging solution were spotted, the stock was mixed System, Germany) mounted on a compound again 10 times between spotting of two slides. microscope, and measurements were taken To verify the quantification method, 3 slides for with AnalySIS FIVE software (Soft Imaging Sys- 04 Sperm quantity and polymorphism in An. arabiensis tem, Germany). For each male the length of a To study the effect of irradiation on minimum of 50 sperm cells were measured. sperm quantity and size polymorphism, males To prevent any bias in selecting sperm cells for were irradiated in the pupal (i.e. 20-26 hrs since measurement, all intact sperm cells (e.g. distin- pupation) or adult stage (i.e. < 24 hrs after guished by a clear head and a tapered tail, Fig. emergence) with a partially-sterilising dose of 4.1B) in a spot were photographed before start- 70 Gy or a fully-sterilising dose of 120 Gy fol- ing the measurements for a new spot. lowing procedures as described in Helinski et al. Intact sperm cells in female spermath- (2006b). Males irradiated with the same dose ecae were photographed immediately after (i.e. in the pupal and adult stage) were from the dissection (without staining) using dark-field exact same batch of larvae, with different doses illumination. Only spermathecae where > 50 using larvae from different larval trays. As a free sperm cells were released were used; and control, males from the first replicate were used sperm from five spermathecae was analysed. (i.e. same batch of eggs and food but different larval tray). Males were dissected aged 6 days, Experimental setup and for each treatment 5 males were used. The quantity and size of sperm in the testes was followed over time. Males were sampled at 3, 6 Statistical analysis or 12 days of age (N= 5 per age group) and this Prior to analyses, data were checked for normal- was replicated thrice for different batches of ity. Sperm lengths were divided in categories of males. Slides from the first replicate were used 100 µm increments, and their proportions arc- to determine sperm length polymorphism (N= sine square root transformed to normalise the 5) over time. To determine the distribution of data for analysis (Klowden and Chambers 2004). sperm lengths in female spermathecae, 5-day- Differences between treatments were analysed old males from the second replicate were mat- with General Linear Models (GLMs), and means ed with virgin females for 1 night at a ratio of 2 were separated using Tukey’s Honestly Signifi- males versus 1 female. The following day, tes- cantly Different (HSD) and individual t-tests. tes from unmated males from the same batch Correlation tests were performed to look at (N= 5), and spermathecae from mated females sperm quantity over time and to determine the were dissected (N= 5) to determine the distribu- relationship between sperm quantity and wing tion of sperm lengths. size. All tests were two-sided and performed

Male Age Rep. 1 Rep. 2 Rep. 3 GLM Table 4.1. Verification of the proto- col to determine sperm quantity. 1 7 28.3 ± 2.5 23.8 ± 1.0 24.6 ± 2.4 F2, 23= 1.29 For each male, 3 replicates (i.e. a 2 7 25.8 ± 1.7 25.8 ± 2.3 23.4 ± 1.5 F2, 25= 0.52 slide with each 8-10 wells) were counted and mean ± s.e.m. are 3 7 48.0 ± 2.6 52.5 ± 2.2 49.6 ± 2.9 F2, 26= 0.82 given. Asterisks indicates a signifi- cant difference between replicates 4 7 36.4 ± 2.4 31.7 ± 2.0 33.9 ± 1.8 F2, 27= 1.28 at p< 0.05 (General Linear Models). 5 7 37.8 ± 3.0 34.2 ± 1.8 36.0 ± 2.1 F2, 26= 0.55 Males of equal age originated from the same batch of mosquitoes. 6 6 26.7 ± 2.8 30.8 ± 1.7 24.1 ± 1.7 F2, 26= 2.44

7 6 41.6 ± 2.7 39.7 ± 3.5 37.5 ± 1.5 F2, 27= 0.58

8 6 42.1 ± 2.8 41.0 ± 2.7 46.4 ± 3.1 F2, 27= 0.97

9 5 15.8 ± 1.0 22.2 ± 2.0 21.5 ± 1.8 F2, 27= 4.37* 71

10 5 27.5 ± 1.6 27.6 ± 1.5 24.8 ± 1.5 F2, 27= 1.07

11 5 25.2 ± 2.0 22.0 ± 1.6 20.5 ± 1.3 F2, 27= 2.03

Chapter 04

Figure 4.2. Mean ± s.e.m. number of sperm cells in testes of A) un-irradiated males dissected at age 3, 6 or 12 days (N= 15), or B) 6 day old males irradiated in the pupal or adult stage with 70 or 120 Gy, and control males (i.e. un-irradiated, N= 5). Means followed by the same letter or number are not statistically different, p< 0.05 (Tukey HSD).

using the SPSS software version 14 (SPSS Inc., number of sperm cells (r= 0.51, p< 0.01). Males Chicago, USA). aged 12 days had significantly more sperm cells than their 3-day-old counterparts and simi- lar numbers of sperm compared to six day old

males (F2, 36= 7.84, p< 0.01; Fig. 4.2A). There Results were no statistical differences in male wing

sizes across all treatments and replicates (F4,

Sperm quantity 35= 1.95, p> 0.05). A positive correlation was ob- Verification of the quantification protocol dem- served between male size and number of sperm onstrated that for the large majority of males cells when all data were combined (r= 0.37, p< the number of sperm cells observed in the three 0.05), but male size explained only 18% of the replicates were similar (Table 4.1), indicating total variation. that sperm cells were homogenously distribut- When males were irradiated in the ed in the sample. However, because this result adult stage with two different doses, the number was not always found (Table 4.1), and because of sperm cells in the testes six days after emer- of the staining, not all spots on a slide could gence were similar compared to the number

always be read, for all experimental males 2-3 observed in un-irradiated control males (F2, 12= slides per male were prepared and the average 0.38, p> 0.05; Fig. 4.2B). Males irradiated in the value used for analysis. pupal stage however, had significantly fewer The number of sperm cells present sperm cells in the testes compared to control 72 in the testes of un-irradiated males increased males for both doses (F2, 12= 4.91, p< 0.05; Fig. over time (Fig. 4.2A). There were no significant 4.2B). Because of the experimental set-up, differences between replicates and data were males irradiated with the same dose were of

combined (F2, 36= 2.40, p> 0.05). A positive cor- the exact same batch of pupae. Males irradi- relation was observed between age and the ated with 70 Gy as pupae had significantly few- 04 Sperm quantity and polymorphism in An. arabiensis

Figure 4.3. Mean ± s.e.m. distribution of sperm lengths in the testes (filled bars) and spermathecae (open bars). Asterisks indicate significant differences between bars for each category; * p< 0.05, ** p< 0.01 (independent-tests). t er sperm cells than males irradiated as adults radiation had no major effect on sperm length (t(8)= -2.66, p< 0.05); for 120 Gy, no significant polymorphism. Distributions of sperm lengths differences between pupae or adult irradiation were similar to un-irradiated control males for were observed (t(4.15)= -2.51, p> 0.05). all categories, with the exception of sperm cells with lengths of 100-200 µm, that were present Sperm length polymorphism in significantly larger numbers in males irradiat- Sperm length polymorphisms were deter- ed as pupae for both doses compared to control mined in males aged 3, 6 or 12 days and male males (F5, 24= 5.08, p< 0.01; Fig. 4.4). Wing data size based on wing length was similar for all showed that all males were equally sized (F4, 20= age groups (F2, 12= 0.69, p> 0.05). The distribu- 0.89, p> 0.05). Wild males had a similar distri- tions of sperm lengths were comparable and no bution of sperm lengths as laboratory-reared significant differences were observed between males, and no significant differences were ob- the three age groups for all length categories served compared to the distribution observed (Table 4.2). In female spermathecae, distribu- in un-irradiated control males for all categories tion of sperm lengths was significantly different (Fig. 4.4). from that in the male testes. A higher propor- tion of large sperm (300-400 µm) were found in spermathecae compared to the testes, while a greater frequency of smaller sperm (0-200 Discussion µm) were found in the testes (Fig. 4.3). The dis- 73 tribution of sperm lengths in these males was As expected, sperm size polymorphisms were significantly similar to those observed in the present in An. arabiensis and sperm varied in 6-day-old males used above (i.e. control males size between 40-495 µm. There was no variation Fig. 4.4; t-tests, p> 0.05, data not shown). Ir- in the distribution of sperm lengths for males Table 4.2. The proportion of sperm lengths in the testes of un-irradiated males determined at age 3, 6 or 12 days. N is the number of males analysed, numbers between brackets indicate the mean number of sperm cells per male for which sperm

Chapter 04 lengths were measured. For each category, means followed by the same letter are not statistically different, p< 0.05 (Tukey HSD). Treatment Proportion (%) ± s.e.m. per category sperm length (!m) N (sperm cells ± < 100 100-200 200-300 300-400 > 400 s.e.m) 3 days 4.8 ± 1.0a 31.3 ± 2.8a 25.9 ± 3.7a 38.0 ± 5.2a 5 (62 ± 0.7) 6 days 4.5 ± 1.0a 26.8 ± 4.2a 24.0 ± 0.9a 44.4 ± 4.1a 0.3 ± 0.3 5 (69 ± 4.1)

12 days 5.1 ± 2.0a 34.1 ± 2.3a 19.3 ± 1.0a 41.6 ± 3.1a 5 (64 ± 2.2)

of different ages. Female spermathecae- har sonal communication). The potential implica- boured significantly fewer sperm of the smaller tions of our findings, i.e. that males irradiated categories (i.e. 0-200 µm), and more sperm as pupae produce a larger proportion of shorter of the larger category (i.e. 300-400 µm) com- sperm, while females select for larger sperm in pared to the distribution observed in the testes. their spermathecae, for the Sterile Insect Tech- Similar results were observed in An. gambiae nique are difficult to predict based on these s.s. (Klowden and Chambers 2004), and also in results. In Drosophila, short sperm do not par- Drosophila where the long sperm was selected ticipate in fertilisation (Snook and Karr 1998), for storage (Bressac and Hauschteck-Jungen however in Anopheles the role of short sperm 1996). In An. gambiae s.s. it was shown that the has not been determined yet (Klowden and selection for larger sperm appeared to have oc- Chambers 2004), and further research is need- curred before the sperm entered the spermath- ed to determine which sperm the female uses ecae as the same distribution was observed in for fertilisation. the common oviduct immediately after insemi- The amount of sperm in the testes nation (Klowden and Chambers 2004). Male increased with age, and similar observations size did not affect the pattern of polymorphic were made in Ae. aegypti (Ponlawat and Har- sperm production in An. gambiae s.s. (Klowden rington 2007) and indirectly in An. gambiae s.s. and Chambers 2004); males in this study were because of the observed increase of the sperm reared under similar conditions and no differ- reservoir (Huho et al. 2006). The number of ences were observed in size. Irradiated males sperm present in the testes was in the same or- had a similar distribution of sperm lengths com- der of magnitude as that observed in laboratory pared to un-irradiated males with the exception strains of Ae. aegypti (Ponlawat and Harrington of sperm in the category of 100-200 µm which 2007). Males were all of similar size, and only a were significantly greater in number in males very weak positive correlation between male irradiated as pupae. Males collected from the size and sperm quantity was observed. When wild showed a similar distribution of sperm insects were reared under different crowding lengths compared to the laboratory males after regimes producing either small or large adults, sixty generations of rearing in the laboratory. this correlation was more pronounced in Ae. The range of the distribution and the propor- aegypti (Ponlawat and Harrington 2007). Irra- tions observed for each sperm length category diation had an impact on the number of sperm were very similar to what has been reported cells present in the testes; males irradiated in 74 for the closely related species An. gambiae s.s. the pupal stage had significantly fewer sperm (Klowden and Chambers 2004). Average sperm cells than control males, while this difference length for un-irradiated males in our study was was not observed for males irradiated as adults. 249 µm (N= 1327), and for An. gambiae s.s. 245 Males irradiated with the same dose in the pupal µm (N= 1762) was observed (M. Klowden, per- or adult stage came from the exact same batch 04 Sperm quantity and polymorphism in An. arabiensis

Treatment Proportion (%) ± s.e.m. per category sperm length (!m) N (sperm cells ± < 100 100-200 200-300 300-400 > 400 s.e.m) 3 days 4.8 ± 1.0a 31.3 ± 2.8a 25.9 ± 3.7a 38.0 ± 5.2a 5 (62 ± 0.7) 6 days 4.5 ± 1.0a 26.8 ± 4.2a 24.0 ± 0.9a 44.4 ± 4.1a 0.3 ± 0.3 5 (69 ± 4.1)

12 days 5.1 ± 2.0a 34.1 ± 2.3a 19.3 ± 1.0a 41.6 ± 3.1a 5 (64 ± 2.2)

Figure 4.4. Mean ± s.e.m. distribution of sperm lengths in the testes of 6-day-old males. C: control; un-irradiated laboratory- reared males, W: wild males (open bars), P: pupae irradiation, and A: adult irradiation; both with 70 (dark grey bars) or 120 Gy (light grey bars). Means followed by the same letter for each length category are not significantly different, p< 0.05 (Tukey HSD).

of pupae and for the 70 Gy dose, males irradiat- error in the protocol. Studies on sperm quanti- ed as pupae had significantly fewer sperm than fication are prone to difficulties associated with males irradiated as adults. However, for 120 Gy the clumping and binding of sperm, and the ho- the difference was no longer significant. In the mogeneously diluting of the sample for count- Mediterranean fruit flyCeratitis capitata Wiede- ing (Yuval et al. 1996, Snow and Andrade 2004), mann, a reduced number of sperm in the testes and these issues could have accounted for some was observed after pupal irradiation compared of the variation observed. Clearly there is room to adult irradiation (Anwar et al. 1971). The ob- for optimisation of sperm cell counts and size served reduction in spermatozoa is likely to be measurements in Anopheles and it is encour- the result of the increased irradiation damage aged that more studies using different tech- induced during the earlier stages of sperma- niques (e.g. sonication, fluorometry (Reichardt togenesis (Proverbs 1969, Anwar et al. 1971, and Wheeler 1995)) are explored. Bakri et al. 2005), which mainly take place dur- Ultimately, it would be interesting to ing larval and pupal development (Mahmood not just determine the quantity of sperm in the and Reisen 1982, Clements 1992, Mahmood testes, but also the amount of sperm present in and Reisen 1994). the spermathecae after mating. In Anopheles, There was quite a large variation in it has been suggested that the switch-over to sperm numbers from males in the same treat- the mated state is triggered by the presence ment. Natural variation in sperm quantity was of sperm in the spermathecae (Klowden 2006). also reported in Ae. aegypti (Ponlawat and Har- In the context of genetic control studies us- 75 rington 2007) but the magnitude of the varia- ing irradiated males, it would be important to tion was lower compared to this study. Thus, it determine if the observed reduction in sperm remains uncertain if the differences in this study quantity following pupal stage irradiation has reflected the natural variation present or some a negative impact on sperm transfer, and if the female is more likely to engage in another mat- portant implications for the effectiveness of a

Chapter 04 ing after receiving sperm from these males. In SIT programme. the Mediterranean fruit fly it was observed that sterile males transferred fewer sperm than wild males and females mated to irradiated males Acknowledgements were more likely to remate compared to fe- The authors wish to thank M. Klowden for pro- males mated to un-irradiated males (Mossin- viding training in sperm polymorphism analyses son and Yuval 2003). Protocols to determine the and for his advice and encouragement during quantity of sperm transferred to spermathecae this study, G. Chambers for training in dissec- exist for other insects (Yuval et al. 1996, Bres- tions, A. Ponlawat for discussions regarding the sac and Chevrier 1998, Mossinson and Yuval protocol, M. Dicke, M. Klowden, and L. Har- 2003, Twig and Yuval 2005), but this has not rington for constructive comments on the man- been pursued for anopheline mosquitoes due uscript, and the International Atomic Energy to difficulties associated with spermathecal dis- Agency for providing funding. BGJK is support- sections. In intact spermathecae, sperm can be ed by a VIDI grant from the Netherlands Organ- observed to rotate within the capsule after dis- isation for Scientific Research (#864.03.004). section, but upon tearing of the spermathecae, the sperm mass binds and clumps to itself and the spermathecal capsule. We found it impossi- ble to separate the sperm with any satisfactory References results, and thus the quantification of sperm in the spermathecae could not be determined, Alphey, L. 2002. Re-engineering the sterile insect nor the effects of irradiation on these. However, technique. Insect Biochem. Mol. Biol. 32:1243-1247. real time quantitative PCR techniques should Anwar, M., D. L. Chambers, K. Ohinata, and R. offer the possibility to obtain more accurate in- M. Kobayashi 1971. Radiation-sterilization of formation on the amount of sperm present now the mediterranean fruit fly (Diptera: Tephritidae): that male-specific primers have been found comparison of spermatogenesis in flies treated as (Ng’habi et al. 2007). pupae or adults. Ann. Entomol. Soc. Am. 64:627-633. Bakri, A., K. Mehta, and D. R. Lance 2005. Sterilizing Insects with Ionizing Radiation, pp. 233-268. In A. Dyck, J. Hendrichs, and A. S. Robinson (eds.), The Conclusion Sterile Insect Technique: Principles and Practice in Area-Wide Integrated Pest Management. Springer, This is the first study describing sperm quantity Dordrecht. in Anopheles and sperm length polymorphism Benedict, M. Q., and A. S. Robinson 2003. The first in An. arabiensis. As such, the work presented releases of transgenic mosquitoes: an argument for in this paper can be used to study these pa- the sterile insect technique. Trends Parasitol. 19:349- rameters in other species, and in particular for 355. those species of interest for genetic control Bressac, C., and E. Hauschteck-Jungen 1996. strategies. The observation that pupal irradia- Drosophila subobscura females preferentially select tion resulted in fewer sperm in the testes, and long sperm for storage and use. J. Insect. Physiol. significantly more sperm of the smaller cate- 42:323-328. gory (i.e. 100-200 µm) that are selected against Bressac, C. and C. Chevrier 1998. Offspring and 76 in the spermathecae, warrants the need for sex ratio are independent of sperm management in further studies. In particular it would be inter- Eupelmus orientalis females. J. Insect Physiol. 44:351- esting to see the impact of pupal irradiation on 359. sperm transfer and determine if females would Calkins, C. O. and A. G. Parker 2005. Sterile Insect be more likely to remate, which could have im- Quality. pp. 269-296. In V. A. Dyck, J. Hendricks, 04 Sperm quantity and polymorphism in An. arabiensis

A. S. Robinson (eds.), Sterile Insect Technique. Ethiopia. Trans. R. Soc. Trop. Med. Hyg. 92:231-235. Principles and Practice in Area-Wide Integrated Pest Klowden, M. J. and G. M. Chambers 2004. Management. Springer, Dordrecht. Production of polymorphic sperm by anopheline Chambers, G.M. and M. J. Klowden 2001. Age of mosquitoes and their fate within the female genital Anopheles gambiae Giles male mosquitoes at time of tract. J. Insect Physiol. 50:1163-1170. mating influences female oviposition. J. Vector Ecol. Klowden, M. J. 2006. Switchover to the mated state 26:196-201. by spermathecal activation in female Anopheles Clark, A.G. 2002. Sperm competition and the gambiae mosquitoes. J. Insect Physiol. 52:679-684. maintenance of polymorphisms. Heredity 88:148- LaChance, L. E. 1967. The induction of dominant 153. lethal mutations in insects by ionizing radiation and Clements, A.N. 1992. The biology of mosquitoes, chemicals-as related to the sterile male technique of Volume 1. Development, nutrition and reproduction. insect control, pp. 617-650. In J. W. Wright, and R. Pal Chapmann & Hall, London. (eds.), Genetics of Insect Vectors of Disease. Elsevier, Curtis, C. F. 2002. Possible ways of using transgenic Amsterdam. mosquitoes for malaria and dengue control and risk Lecis, A. R., V. Figus, and C. Santarini 1975. assessment. pp. 165-175. Beijing, Peking University. Radiosensitivity curve of different stages of 7th International Symposium on Biosafety of spermatogenesis of Anopheles atroparvus Genetically Modified Organisms, 10-16 October. (Diptera:Nematocera). Parassitologia 17:145-150. Dame, D. A. and C. F. Curtis 1996. The potential Mahmood, F. and W. K. Reisen 1982. Anopheles use of the sterile insect technique and other genetic stephensi (Diptera: Culicidae): changes in male mating control methods for control of malaria-transmitting competence and reproductive system morphology mosquitoes. pp 1-27. Vienna, IAEA. associated with aging and mating. J. Med. Entomol. Dyck, A., J. Hendrichs, and A. S. Robinson 2005 19:573-588. (eds.). The Sterile Insect Technique: Principles and Mahmood, F. and W. K. Reisen 1994. Anopheles Practice in Area-Wide Integrated Pest Management. culicifacies: effects of age on the male reproductive Springer, Dordrecht. system and mating ability of virgin adult mosquitoes. Ferguson, H. M., B. John, K. R. Ng’habi, and B. Med. Vet. Entomol. 8:31-37. G. J. Knols 2005. Redressing the sex imbalance Mossinson, S. and B. Yuval 2003. Regulation of in knowledge of vector biology. Trends Ecol. Evol. sexual receptivity of female Mediterranean fruit 20:202-209. flies: old hypotheses revisted and a new synthesis Helinski, M.E.H., B. El-Sayed and B.G.J. Knols proposed. J. Insect Physiol. 49:561-567. 2006a. The Sterile Insect Technique: Can established Ng’habi, K. R., B. John, G. Nkwengulila, B. technology beat malaria? Entomol. Berichten 66: 13- G. J. Knols, G. F. Killeen, and H. M. Ferguson 20. CHAPTER 1. 2005. The effect of larval crowding on the mating Helinski, M. E. H., A. G. Parker and B. G. J. Knols competitiveness of Anopheles gambiae mosquitoes. 2006b. Radiation-induced sterility for pupal and adult Malar. J. 4:49. stages of the malaria mosquito Anopheles arabiensis. Ng’habi, K. R., A. Horton, B. G. J. Knols, and G. Malar. J. 5:41. CHAPTER 3. C. Lanzaro 2007. A new robust polymerase chain Helinski, M. E. H. and B. G. J. Knols 2008. Mating reaction for determining the mating status of female competitiveness of male Anopheles arabiensis Anopheles gambiae mosquitoes. Am. J. Trop. Med. mosquitoes irradiated with a partially or fully- Hyg. 77:485-487. sterilising dose in small and large laboratory cages. Ponlawat, A. and L. C. Harrington 2007. Age and J. Med. Entomol. in press. CHAPTER 7. body size influence male sperm capacity of the Huho, B. J., K. R. Ng’habi, G. F. Killeen, G. dengue vector Aedes aegypti (Diptera: Culicidae). J. 77 Nkwengulila, B. G. J. Knols, and H. M. Ferguson Med. Entomol. 44:422-426. 2006. A reliable morphological method to assess the Proverbs, M. D. 1969. Induced sterilization and age of male Anopheles gambiae. Malar. J. 5:62. control in insects. Annu. Rev. Entomol. 14:81-102. Hunt, R.H., M. Coetzee, and M. Fettene 1998. The Reichardt, A. K. and D. E. Wheeler 1995. Estimation Anopheles gambiae complex: a new species from of sperm numbers in insects by fluorometry. Insectes

Chapter 04 Sociaux. 42:449-452. Scott, T. W., W. Takken, B. G. J. Knols, and C. Boëte 2002. The ecology of genetically modified mosquitoes. Science 298:117-119. Sobels, F. H. 1969. A study of the causes underlying the differences in radiosensitivity between mature spermatozoa and late spermatids in Drosophila. Mutat. Res. 8:111-125. Snook, R. R., and T. L. Karr 1998. Only long sperm are fertilization-competent in six sperm-heteromorphic Drosophila species. Current Biol. 8:291-294. Snow, L. S. E, and M. C. B. Andrade 2004. Pattern of sperm transfer in redback spiders: implications for sperm competition and male sacrifice. Behav. Ecol. 15:785–792. Swallow, J. G. and G. S. Wilkinson 2002. The long and short of sperm polymorphisms in insects. Biol. Rev. Camb. Philos. Soc. 77:153-182. Thomas, D. D., C. A. Donnelly, R. J. Wood, and L. S. Alphey 2000. Insect population control using a dominant, repressible, lethal genetic system. Science 287:2474-2476. Tripet, F., Y. T. Touré, G. Dolo, and G. C. Lanzaro 2003. Frequency of multiple inseminations in field- collected Anopheles gambiae females revealed by DNA analysis of transferred sperm. Am. J. Trop. Med. Hyg. 68:1-5. Twig, E. and B. Yuval 2005. Function of multiple sperm storage organs in female Mediterranean fruit flies (Ceratitis capitata, Diptera: Tephritidae). J. Insect Physiol. 51:67-74. Verhoek, B. A. and W. Takken 1994. Age effects of insemination rate of Anopheles gambiae s.l. in the laboratory. Entomol. Exp. Appl. 72:167-172. Yuval, B., S. Blay, and R. Kaspi 1996. Sperm transfer and storage in the Mediterranean fruit fly (Diptera: Tephritidae). Ann. Entomol. Soc. Am. 89:486-492.

78 04 Sperm quantity and polymorphism in An. arabiensis

79 PART 2 Part 2

Stable isotopes in mating behaviour research PART 2 PART

81 05 Stable isotope 13C semen label 05 Stable isotope-mass spectrometric determination of semen transfer in malaria mosquitoes

by Michelle EH Helinski, Rebecca C Hood-Nowotny, Leo Mayr and Bart GJ Knols

The potential use of stable isotopes to study mosquito mating was investigated by tracing the fate of labelled semen into spermathecae. 13C-glucose was incorporated in the diet of the malaria mosquito Anopheles arabiensis. Treatments included labelling of either the larval water or adult sugar water, or a combination of both. After mating, “spiked” spermathecae were analysed for isotope ratios using mass- 350-500 Million cases of clinical malaria occur annually, 60% of which are in sub- spectrometry. Results demonstrated that spermathecae positive for semen could Saharan Africa. Moreover, 80% of all deaths attributed to malaria occur in this successfully be distinguished from empty ones or controls (i.e. filled with unlabelled region. In numbers, 1 million Africans die of the disease each year, with the vast semen) using the raw δ13C values. Labelling during larval development and combined majority of deaths occurring among children below five years of age. Pregnant labelling of larvae and adults resulted in detectable values. The label persisted in women are another major risk group; malaria can cause low birth weight and pre- spermathecae for up to 7 days after mating, and unlabelled sugar feeding of males mature delivery (Rogerson et al. 2007). The impact of malaria on the economic labelled in the larval stage did not result in a detectable turnover of the semen situation of endemic countries is high (Gallup and Sachs 2001), and the correlation label. There were no detrimental effects of the addition of labelled glucose on larval between poverty and malaria clearly demonstrated (Sachs and Malaney 2002). development and survival, adult size, male longevity, and mating performance. We Malaria is a parasitic disease transmitted by female mosquitoes of the have proven that it is possible to label male mosquitoes and detect the semen label genus Anopheles. The malaria parasite is a protozoan of the genus Plasmodium. in females after insemination. This method offers great potential to study mating There are four species of malaria parasites that can infect humans under natural in mosquitoes and other insects and could prove useful in genetic control studies conditions: Plasmodium falciparum, P. vivax, P. ovale and P. malariae. The first two of medical or agricultural pest insects, with male mating success in the field as a species cause the most infections worldwide and P. falciparum Welsh is by far the critical verifiable indicator for a positive outcome of the intervention. 83

Published with minimal changes in Journal of Ex- perimental Biology 2007, 210: 1266-1274 05 Stable isotope 13C semen label

Introduction to radioactive solutions. This resulted in the

Chapter 05 transfer of radioactive semen during copula- tion and the successful identification thereof in ecently, stable isotopes have become spermathecae. Nowadays, the use of radioac- widely available as a labelling tool in bio- tive isotopes is rarely practiced in entomologi- Rlogical studies, largely as a result of com- cal research due to the hazards related to the paratively lower costs for both the isotopes and treatment process and environmental concerns sample analysis (Hood-Nowotny et al. 2005). In of releasing such insects, even though the half- insect studies, stable isotopes have not been life of the commonly used isotope, 32P, is short used extensively, but these can be a useful tool (i.e. 14.3 days). Other options to study mating to address issues of food preference, resource in mosquitoes include the use of mutant strains allocation, dispersal, etc. (Hood-Nowotny and (Mason 1967, Gomulski 1988, Beard et al. 1995, Knols 2007). Recent studies undertaken in our Klowden 2006), but these are not readily availa- laboratory have documented the successful ble for most stocks. A transgenic strain that can application of stable isotopes as a population be used to study mating behaviour has been marker in the context of genetic control studies developed in An. stephensi Liston (Catteruccia (e.g. Sterile Insect Technique; Hood-Nowotny et al. 2005), but this technology hinges on ethi- et al. 2006). Our current interest lies in the use cal, legal and social issues affecting the ability of stable isotopes to study mating behaviour. to release transgenic insects (Knols et al. 2006, Studies where natural abundance levels of sta- Knols and Louis 2006), and is not easily trans- ble isotopes were used in the context of mat- ferable to other species. ing have been carried out (Ponsard et al. 2004, Stable isotopes are naturally occur- Malausa et al. 2005) but to our knowledge no ring in the environment, are not radioactive work has been performed on the enrichment of and therefore do not decay. Besides these at- insect semen with stable isotopes. Semen la- tributes, stable isotopes are not species-specif- belling has recently been performed in humans; ic, which makes them attractive for use. Most 2 H2O was ingested daily and spermatogenesis elements of biological interest (including C, H, kinetics studied (Misell et al. 2006). O, N, and S) have two or more stable isotopes, In the present study we used the Af- with the lightest of these present in much great- rican malaria mosquito Anopheles arabiensis er abundance than the others. An isotope of an Patton. Of the major life history behaviours of element has the same atomic number but a dif- anopheline mosquitoes, mating remains the ferent number of neutrons and consequently least understood (Takken and Knols 1999). a different atomic mass. In general, stable iso- Studies on mosquito mating behaviour are dif- topes react chemically in a manner identical to ficult to conduct, due to its crepuscular nature, the more common isotope and thus are effec- complex constituency (i.e. swarm make-up), tive non-invasive markers in biological systems. and irregular spatial occurrence. In the context Among stable isotopes, the most useful as bio- of genetic control studies, understanding a logical tracers are the heavy isotopes of carbon male’s mating success in the field is critical for a and nitrogen. These two elements are found in positive outcome of the intervention (Ferguson soils, the atmosphere, and all living organisms. et al. 2005), and techniques that label semen For carbon, the heavy isotope 13C has a natural (i.e. spermatozoa and accessory gland fluid) abundance of 1.1% and the other isotope (l2C) would greatly facilitate the study of mating be- makes up virtually all of the remainder (carbon 84 haviour. Radioactive isotopes have been used also has a radioactive isotope, 14C). If a system in the past to study the fate of semen (Dame is enriched with the less abundant isotope, this and Schmidt 1964, Tantawy et al. 1967, Smit- element can be used as a label or tracer (Hood- tle et al. 1969, Young and Downe 1978) in which Nowotny and Knols 2007). The isotopic compo- mosquitoes were labelled by exposing larvae sition of a sample is measured by determining 05 Stable isotope 13C semen label

the ratios of the stable isotope masses. These Labelling ratios are measured on an isotope ratio mass 99 atom% 13C-glucose (U-13C6, Cambridge Iso- spectrometer, a device that separates ions of tope Laboratories Inc, Andover, MA, USA) was the element of interest on the basis of their dif- used as a label. Mosquitoes were exposed to fering mass/charge ratio (m/z) (de Groot 2004). the label either as larvae or adults. In the larval For the experiments presented in this stage, the label was added to the larval water Chapter 13C was chosen as a label. The findings on the same day as the L1 larvae were intro- of three sets of experiments that address four duced. In the adult stage, the label was added objectives are presented. The first objective to the sugar water. The level of enrichment in was to see whether it was possible to use 13C as both treatments was 20 atom% 13C (i.e. approx- a semen-labelling technique and to identify the imately 20% of all the carbon in the diet was optimal developmental stage for labelling. The 13C), and was based on findings from a previous second objective was to test the ability of males study (Hood-Nowotny et al. 2006) where lower of different ages to transfer the label. The third levels of enrichment were used (i.e. 1.11-1.46 objective was to test the persistence of the la- atom% 13C) for whole body analysis. bel in spermathecae after mating. The fourth The amount of 13C-glucose needed was objective was to study the impact of the stable based on the total amount of carbon present in isotope on the mosquito to assure that no det- the diet. Forty percent of the larval diet con- rimental effects occurred. sisted of carbon. Until pupation, 1.0 g (0.25 mg x 500 larvae x 8 days) of larval food was added to the tray, thus 0.1 g of 13C was required. As the Materials and methods percentage of carbon in glucose is 40%, 0.25 In all experiments, mosquito rearing and label- g of 99 atom% 13C glucose was added to the ling techniques were identical, as well as sam- larval trays. For adult mosquito labelling, the ple preparation and analyses. stable isotope was incorporated in the sugar solution, and similar calculations were made to Mosquitoes determine the amount of label needed. 1.00 g The Dongola strain of Anopheles arabiensis Pat- of sucrose has 0.42 g of C, thus 0.1 g of 13C was ton was used. It was collected in Northern State, required. Labelled sugar feeders received 0.25 g Sudan, in 2004 and has been reared in our labo- of 99 atom% 13C-glucose + 1.00 g sucrose in 12 ratory since then. Five hundred L1 stage larvae ml water (10.4% sugar solution); the unlabelled were counted and placed in a tray (30 x 40 cm) feeders received unlabelled glucose instead. in 1 L of deionised water, and the water level was kept constant throughout the experiment. Experimental setup Heating mats were used to maintain the water Males that emerged from the experimental temperature at 28 ± 1 °C. Larvae were fed a diet larval trays (i.e. labelled (L) and unlabelled (U)) of fish food (AquariCare Koi Floating Blend, were divided into four adult treatments: U-U, USA) daily (0.25 mg/ larva), that was ground and males unlabelled in the larval stage fed on unla- sieved through a 224 µm sieve, and mortality of belled sugar; U-L, males unlabelled in the larval the larvae was not taken into account. Adults stage fed on labelled sugar; L-U, males labelled were kept in standard 30 x 30 x 30 cm mosquito in the larval stage fed on unlabelled sugar; L-L, rearing cages and maintained at a temperature males labelled in the larval stage fed on labelled of 27 ± 1 °C and relative humidity of 80 ± 2%. sugar. When treatments were compared within The light regime was L10:D12 with a one hour an experiment, only males emerging on the 85 simulated dusk and dawn period. Adults were same day from labelled and unlabelled trays maintained on a standard 10% sucrose solution were used. Males were transferred from the lar- [w/v] unless stated otherwise. val trays to adult cages and fed on their desig- nated sugar source (i.e. labelled or unlabelled) 05 Stable isotope 13C semen label

until mating was initiated. In the experiments Experiment 2

Chapter 05 where the adult sugar water was labelled, the In experiment two, lower insemination of the unlabelled treatments received unlabelled glu- females was pursued to determine if insemi- cose; however, when adult labelling was not nated spermathecae could successfully be dis- performed, cages were maintained on the labo- tinguished from un-inseminated spermathecae ratory standard (i.e. unlabelled) sucrose solu- within the same treatment. Treatments from tion. experiment 1 (U-U, L-U, L-L) were repeated When adult labelling was performed, with the exception of adult only labelling (U-L), males were transferred to a new cage prior to and we investigated the persistence of the label mating to prevent cross-contamination of the in the females and the persistence of the label females (e.g. spills from the sugar source), and in males that mated later in life. In addition, the males in other treatments tested at the same impact of labelling on larval development and time were also transferred for comparison sake. adult longevity was studied. During and after mating, males were main- Larval trays included one tray with la- tained on standard sucrose solution. Females belled glucose, and one with unlabelled glucose. used as mates were isolated within 18 hrs after Another tray without any glucose was added to emergence to assure virginity. The age of the monitor the effect of glucose on larval survival. females in all experiments was similar to that of Adults emerging from the trays were removed males when mating was initiated. After mating, and counted daily, and trays were maintained females were either: I) dissected immediately until all larvae had pupated and emerged, or (i.e. the following day) at the end of the mat- died. ing period; or II) isolated for dissection at a later The males emerging from the labelled stage to assess the persistence of the label in glucose tray were either fed unlabelled or la- the spermathecae. belled sugar, males from the unlabelled tray were fed unlabelled sugar. After 5 days of sugar Experiment 1 feeding, 35 males were mated for 1 night with The main goal of the first experiment was to females on a 1:1 ratio. Females were dissected see if 13C could be used as a semen label, and immediately or isolated and dissected 3 days to determine the optimal treatments to deliver later (Fig. 5.2A). Mortality of the males was 13C-glucose. In addition, persistence of the label scored regularly until the majority had died. in spermathecae was studied. A second batch of males that emerged For each larval treatment (i.e. labelled from the labelled tray was used to study the glucose and unlabelled glucose) two trays were persistence of the label in spermathecae for up setup. Males emerging from the trays were to 7 days. Males were maintained on standard pooled according to treatment and divided into 10% sucrose solution and mated (N= 25) on day the four adult treatments as described above. 2 after emergence for 3 nights at a ratio of 1:2 Males were fed sugar from their designated (M:F). Females were either dissected immedi- source for four nights. On the fifth day, -mat ately, or isolated and dissected 4 or 7 days later ing was initiated. Per treatment, 62 males were (Fig. 5.2B). mated with females at a ratio of 2:1 (M:F) and A third batch of males that emerged the mating period lasted for 3 nights. After mat- from the labelled tray was used to study the ef- ing, females were immediately dissected for fect of male age on label transfer. Males were analyses, or isolated and dissected 3 days later. maintained on standard 10% sucrose solution 86 A small number of experimental males and mated (N= 25) with females at a ratio of 1:2 were removed from the cages on day 4 (L-U, (M:F) on day 4 for 3 nights, or on day 10 (N= 25) L-L treatments) and day 11 (all treatments) af- for 1 night at a ratio of 1:1. Females were dis- ter emergence and their reproductive system sected immediately or after 4-5 days of isola- dissected for isotopic determination. tion (Fig. 5.2C). 05 Stable isotope 13C semen label

Experiment 3 under a compound microscope at 100x mag- The impact of labelling on larval development nification and recorded. The spermatheca was and adult longevity, to complement initial data then transferred to a small piece of quartz fibre gathered in experiment 2, was investigated. filter paper with a fine brush and placed in a cy- In addition, the effect of 13C on the size of the lindrical tin cup of 8 x 5 mm (height x diameter). emerged adults, and the impact of the semen After each dissection, tools were cleaned with label on the hatchability of eggs was studied. ethanol to avoid contamination. The amount Larval treatments included trays with of carbon present in the spermatheca (~ 3 μg) 13C-labelled glucose, unlabelled glucose, and a was below the detection limit of the mass spec- control tray; each treatment was duplicated. trometer setup (approx. > 20 μg). Samples Adults emerging from the trays were removed were therefore “spiked” with 10 μl of a standard and counted daily, and trays were maintained sucrose solution containing ~ 23-26 μg of car- until all larvae had pupated and emerged, or bon (Dube et al. 1998). All samples were dried died. On the day the majority of pupae emerged, for ≥ 24 hrs in an oven at 50 °C before closure 50 males were collected per treatment and per of the tin cup and subsequent analyses in the replicate and placed in a standard cage to moni- mass spectrometer. Standards containing only tor survival. Males were maintained on a stand- the piece of quartz fibre filter paper with the ard 10% sucrose solution, and mortality was “spike” were included. Spermathecae from scored regularly until all males had died. virgin females were used as a control and dis- Adult body size of males and females sected similar to experimental females. emerging from the trays was determined by wing length (Lyimo and Takken 1993, Charl- Males wood et al. 2002, Lounibos et al. 1995). Day of A number of males from the first experiment emergence was noted and for each specimen a were dissected to analyse the amount of 13C in wing was clipped and mounted on a slide. A dig- their reproductive system. The testes, acces- ital image of the wing was taken (CC-12 cam- sory glands, and seminal vesicle were dissected era, Olympus Soft Imaging Solutions, Berlin, and prepared for analyses as above (i.e. includ- Germany, mounted on a stereo microscope). ing the “spike”). Wing length was measured between the alula notch and the wing tip, excluding scales; meas- Sample analysis and interpretation urements were performed with AnalySIS FIVE After drying, tin cups were sealed and contents software (Olympus Soft Imaging Solutions, analysed using a Carlo Erba (Milan) carbon ni- Germany). trogen (CN) analyser, linked to an Optima, (Mi- The effect of the 13C label on sperm vi- cromass, Manchester, UK) isotope ratio mass ability was monitored by assessing the hatch- spectrometer (IRMS), see Hood-Nowotny and ing of eggs fertilised by labelled sperm. Un- Knols (2007) for details. Briefly, organic sam- labelled virgin females (N= 50) were mated to ples are converted to gas (N2, CO2, H2, SO2) by males emerging from labelled or control trays combustion at high temperature (1800 °C) in at a 1:1 ratio and cages were maintained for 26 an appropriate preparation system linked to days. Mosquitoes were membrane blood fed on the isotope ratio mass spectrometer. Gases are multiple occasions and eggs were collected en separated on a gas chromatograph column and masse and checked for hatching. bled into the mass spectrometer. The gases are ionised on a hot filament under vacuum, accel- Sample preparation erated and separated by a magnetic field based 87 Females on their mass to charge ratio (m/z). The sepa- Females were immobilised, and their spermath- rated ions are collected in Faraday cups where ecae dissected in mosquito saline (Ephrussi and the ratios of the isotopes of interest are deter- Beadle 1936). Insemination status was checked mined. The output of the mass spectrometer is 05 Stable isotope 13C semen label

a ratio, which can be internally converted to an exceeded two, non-parametric tests were per-

Chapter 05 atom % value or a delta value (Hood-Nowotny formed. Data on spermathecae labelling were and Knols 2007). analysed using the following variables: insemi- In this thesis, most values are reported nation status (inseminated or un-inseminated as delta (δ) units parts per thousand (‰). These as determined by compound microscopy) adult delta values are calculated as the isotopic ratio labelling treatment (U-U, U-L, L-U, L-L), and of a sample standardised to the isotopic ratio of dissection history (I-II). Some outliers in δ13C a defined reference standard: values were observed in the dataset, in par-

[(RS – RR) / RR]x 1000= δ ticular in the first experiment, and these were

Where RS is the isotopic ratio of the sample and excluded to normalise the data (see results sec- 13 RR is the isotopic ratio of the reference standard tion). Differences between mean δ C values and of labelling treatments were analysed using R= [atom%13C/ atom%12C]. General Linear Models (GLMs) with planned Thus delta denotes a difference relative to a contrasts (Tukey HSD) and data on dissection standard; there are a number of internation- history were analysed with GLMs or independ- ally recognised conventional reference stand- ent t-tests. Independent t-tests were also used

ards. For carbon RR is 0.011224 (Vienna Pee to compare un-inseminated females to virgin

Dee Belemnite (VPDB)) and for nitrogen RR is control samples in all experiments, and to test 0.0036765 (atmospheric nitrogen (AIR); Hood the difference in δ13C values of inseminated and and Knols 2007). un-inseminated spermathecae for each label- All spermathecae samples were spiked ling treatment in experiment 2. A threshold val- and for data representation, the actual delta val- ue to distinguish labelled spermathecae from ue of the sample alone, i.e. without the “spike”, unlabelled spermathecae was defined as 2 or 3 was not determined because of uncertainty as- standard deviations (s.d.) above the mean δ13C sociated with calculating this value. This was (‰) value of the reference standard (Macneale due to the unknown and low amount of carbon et al. 2005), in our case virgin females. Longev- in the original un-spiked sample and proximity ity of males was analysed using Kaplan-Meier of unlabelled treatment δ13C values to the δ13C survival analyses. The obtained survival curves of the spike which made comparison of labelled were pairwise compared using Mantel-Cox and unlabelled samples difficult. Therefore, log-rank tests. Carbon data, larval survival and spiked δ13C values were used for statistical anal- wing length data were analysed with GLMs. All ysis and data representation. For the males, the two-sided tests were performed using the SPSS actual δ13C values were calculated to obtain an software version 12 (SPSS Inc., Chicago, USA). estimate of the enrichment. These values were expressed in atom%. The atom% notation is based on the ratio of the less abundant isotope of interest to the more abundant isotope Results Atom% 13C= (13C/ 12C+13C)x 100 and can be calculated from the δ definition a) Labelling 13 13 Atom% C= (δ + 1000)/[(δ + 1000 + (1000/RR)] The δ C‰ values reported are negative as For a more detailed description of calculations they are referenced to an international stand- see Hood-Nowotny and Knols (2007) and Fry ard VPDB which is more enriched in 13C than our (2006). spiked samples. 88 Statistical analysis Optimal treatment Prior to analyses, data were checked for nor- When females from experiment 1 were dis- mality. When homogeneity of variances was sected immediately after mating (I), sufficient not assumed, and the number of treatments amounts of labelled semen were transferred in 05 Stable isotope 13C semen label

labelled as larvae or as larvae and adults. The second experiment repeated the treatments from the first experiment, except for the adult only labelling, with similar results. Mean δ13C values of spermathecae inseminated by males labelled in the larval or in both stages were higher than the control males after immediate

dissection (I) (F2, 16.08= 160.53, p< 0.01), and in this experiment males labelled in both stages transferred significantly more label than larval labelled males alone (Fig. 5.2A).

Label persistence in males Unlabelled sugar feeding in the adult stage did not result in a detectable dilution of the se- men label after immediate dissection (I); males exposed to the label in the larval stage alone transferred similar amounts of label when mat-

ed at either 4 or 10 days of age (F3, 38= 1.12, p> 0.05; Fig. 5.2C).

Persistence of label in spermathecae Figure 5.1. δ13C‰ values (mean (± s.e.m.)) of inseminated After isolation of the females for 3 nights (II), spermathecae from experiment 1. The adult labelling mean δ13C values decreased in all labelled treat- treatments were: U-U: no label; U-L: adult labelling ments compared to immediate dissection (I) in only, L-U: larval labelling only, and L-L: larval and adult labelling. N, number of samples analysed. The dotted line experiment 1 (Fig. 5.1), and in the treatments indicates the threshold value of 2 s.d. above mean δ13C‰ U-L and L-U, a significant decrease was ob- of virgin females. Virgin (V) and Standard (St) samples served; U-L: (t(16)= 4.19, p< 0.01), L-U: (t(13)= are included. Values excluded from analyses: 9.97 (L-L), 2.43, p< 0.05; Fig. 5.1). When males were only -12.51 (V) and, -13.29, -18.20 (St). Dissection treatments: 13 I, females dissected immediately after mating; II, females labelled in the adult stage, δ C values of fe- isolated and dissected 3 days after mating. Values with males isolated for 3 days were still higher, but different letters are significantly different at p< 0.05 for I no longer statistically different from the control (lower case), or II (upper case), (Tukey HSD). samples (F3, 22= 27.22, p< 0.01), while in the larval labelled treatments mean δ13C values remained higher than the control (Fig. 5.1). In experiment all labelled treatments to distinguish mean δ13C 2, no decrease in δ13C values was observed after values of inseminated spermathecae from un- isolation for three days in treatments L-U and labelled samples (F3, 38= 26.88, p< 0.01; Fig. 5.1). L-L compared to immediate dissection (Fig. Labelling in the larval stage (L-U) or in both 5.2A). In treatment L-L similar values were re- stages (L-L) resulted in the highest amount of ported (t(15)= -0.12, p> 0.05), while even in the label transferred, but labelling of only the adult treatment L-U a significant increase was ob- stage (U-L) was sufficient to distinguish mean served (t(15)= -2.57, p< 0.05). δ13C values from the control (U-U). However, When females were mated to males the persistence of the label in males and fe- labelled in the larval stage and isolated for 4 89 males after adult labelling alone was not suf- or 7 days after mating, similar δ13C values were ficient (see below). Therefore, labelling in the found comparable to immediate dissection (F2, adult stage alone was not considered optimal 26= 2.23, p> 0.05; Fig. 5.2B). It was also observed and further experiments focused on the males that females inseminated by 10 day old males 05 Stable isotope 13C semen label

retained similar amounts of label after isolation

Chapter 05 (II) compared to females inseminated by 4 day old males, and δ13C values after isolation were comparable to immediate dissection (I) at both

ages (F3, 38= 1.12, p> 0.05; Fig. 5.2C).

Control samples Spermathecae from females mated with males from the unlabelled treatments in experiment 1 had similar δ13C values compared to virgin fe- males (t(36)= -1.88, p> 0.05) and the standards (t(30)= -1.06, p> 0.05). The un-inseminated fe- males from all treatments in experiment 2 had similar δ13C values as the virgin females (t(51)= -1.79, p> 0.05).

Labelled versus unlabelled samples The males from the labelled treatments in experiment 2 transferred significantly high enough amounts of label to distinguish mean δ13C values of inseminated spermathecae from un-inseminated ones (Fig. 5.2).

Threshold values To determine the accuracy of the labelling, the threshold value in each experiment was deter- mined and is represented as a dotted horizontal line in Figures 5.1 and 5.2. δ13C values of samples Figure 5.2. δ13C‰ values (mean (± s.e.m.)) of inseminated (open triangles) and un-inseminated (bold that appear above the threshold line are consid- line) spermathecae from experiment 2. For labelling ered to have been from females inseminated treatments see Fig. 1. The mating and dissection history of by labelled males, values below the line should each treatment is illustrated in the lower part of the graph. represent un-inseminated females or females Duration of mating if longer than 1 night is indicated by arrows; Female symbol with a pin indicates spermathecae mated to unlabelled males. The threshold value dissection. Subdivisions A-C refers to different batches of in experiment 1 was rather conservative due males (see text). N is the number of samples analysed for to some variation in the samples from virgin inseminated (+) and un-inseminated (-) spermathecae. females. Therefore, a small number of sper- Dotted line indicates the threshold value of 3 s.d. above mean δ13C‰ of virgin females. Virgin (V) and Standard (St) mathecae from the L-U treatment appeared samples are included. Virgins are un-inseminated females below the threshold (4/21 samples). Spermath- from treatment U-U. Value excluded from analyses: -13.02 ecae from the dual labelling group appeared all (A, L-U). Values with different letters are significantly but one above the threshold value. However, a different at p< 0.05 for I (lower case), or II (upper case) in A (Tukey HSD). Planned contrasts (p< 0.05) in B were large proportion of spermathecae inseminated made between the three values, and in C between the four by males labelled as adults only appeared be- values. Asterisks indicate significant difference between low the threshold value. Hence this treatment 90 positive and negative spermathecae at *p< 0.05, and **p< was not considered optimal. In the second ex- 0.01 (independent t-tests). periment, males labelled in the larval and adult stage transferred enough label so that the δ13C value of each inseminated spermatheca ap- peared above the 2 s.d. (standard deviation) 05 Stable isotope 13C semen label

0.62, p> 0.05) and standards (t(25)= 0.99, p> 0.05). However, the males from experiment 1 transferred more label, resulting in higher δ13C values, than the males from equal treatments in experiment 3; L-U: (t(7.50)= 2.99, p< 0.05), L-L: (t(20)= 2.50, p< 0.05). After isolation of the females for 3 days this difference was no longer observed.

Males The amount of label that was fixed in the repro- ductive system of males from experiment 1 was similar to what was found in the females after mating. Larval labelling resulted in a higher mean δ13C value than adult labelling and the labelling of both stages was superior over ei- ther singly (X2= 18.12, df= 3, p< 0.01; Fig. 5.3). In atom%, this corresponds to a mean (± s.e.m.) enrichment of 2.52 ± 0.46 atom% 13C for adult labelling (U-L), 4.64 ± 0.67 atom% 13C for larval labelling (L-U), and 7.93 ± 0.93 atom% 13C for la- belling of both stages (L-L). If males were sam- pled at an earlier interval (i.e. 4 days after emer- gence), a similar amount of label was observed Figure 5.3. δ13C‰ values (mean (± s.e.m.)) of the in treatment L-U (t(10)= 0.14, p> 0.05), and a reproductive system of males from experiment 1. higher amount was observed in treatment L-L Treatments are U-U: no label; U-L: adult labelling only, L-U: larval labelling only, and L-L: larval plus adult labelling. N (t(9)= 2.36, p< 0.05) compared to sampling on is number of males analysed. Standard samples (St) are day 11. included. Males were dissected 11 days after emergence. Values with different letters are significantly different at b) Life-history traits p< 0.05 (Tukey HSD). Values with + are males sampled at an earlier interval (i.e. 4 days after emergence). Asterisks Mating indicate a significant difference (independent t-tests) Comparable insemination rates were found for between both sample intervals for L-U and L-L at p< 0.05. labelled and control males in all experiments (Table 5.1). The highest insemination was ob- served in the first experiment when females threshold. Even if 3 s.d. was used as a threshold were introduced at the ratio of 2:1 (M:F). Ex- value, all inseminated females had higher labels periment 2 aimed for a higher proportion of and this value is indicated in Fig. 5.2. un-inseminated females, and insemination was between 57-84%, resulting in adequate num- Between experiment variation bers of un-inseminated females to validate the The mean (± s.e.m.) amount of carbon in the method. spiked samples differed significantly between the experiments (t(268)= -26.32, p< 0.01); in ex- Longevity periment 2, a higher amount of carbon was de- Labelling of males in the larval or larval and 91 tected (N= 191, M= 26.15 ± 0.07 μg) compared adult stage in experiment 2 had no detrimental to experiment 1 (N= 79, M= 22.84 ± 0.11 μg). effect on longevity. A similar (X2= 0.21, p> 0.05) Between experiments, δ13C values of control or even slightly higher (X2= 9.27, p< 0.01) sur- samples were similar for virgin females (t(16)= vival was observed compared to control males 05 Stable isotope 13C semen label

Table 5.1. Insemination of females mated with males from experiments 1 and 2. Duration is the number of nights mated (to ensure the presence of unmated females in experiment 2, females were introduced at a 1:1 ratio and mating was restricted

Chapter 05 to one night only. If mating took place over a weekend (i.e. 3 nights of mating), the number of females was doubled). Ratio: Male: Female. Treatments are: U-U: no label; U-L: adult labelling only, L-U: larval labelling only, and L-L: larval plus adult labelling. N is the number of females dissected. Experiment Duration Ratio M:F Treatment N Insemination (%) 1 3 2:1 U-U 18 100 3 2:1 U-L 18 100

3 2:1 L-U 16 94

3 2:1 L-L 18 100

2 1 1:1 U-U 28 68

1 1:1 L-U 30 60

1 1:1 L-L 30 57

3 1:2 L-U 42 69

3 1:2 L-U 32 84

1 1:1 L-U 23 65

(Table 5.2). Similar observations were made in were measured from ~ 50 individuals that experiment 3; longevity of males reared in trays emerged on the second or third day of emer- with labelled glucose was similar compared to gence (i.e. when the majority of pupae the control males (X2= 1.17, p> 0.05). Only in the emerged). No significant size differences were second replicate a lower longevity of labelled observed in wings from both days and data males compared to the control was observed were pooled. Significant differences were ob- (X2= 11.65, p< 0.01). In both replicates, males served between replicates of the same treat- from the unlabelled glucose trays had signifi- ment for males and females, therefore data was cantly higher longevity than control males (Ta- analysed per replicate. For the females, size of ble 5.2). the adults was unaffected by the label; females emerging from trays with labelled glucose or

Larval development and survival unlabelled glucose were similar (F2, 146= 1.47, p>

The rate of pupation in trays where labelled or 0.05) or larger (F2, 151= 5.09, p< 0.01) in size com- unlabelled glucose had been added was similar pared to females from the control trays (Table to those without any glucose. Pupation started 5.4). Some variation in wing length between at day 7 and continued until day 11 by which treatments and replicates was observed for the time the vast majority of L4 larvae had pupated males. The labelled trays produced the small-

(data not shown). Larval survival was not affect- est (F2, 148= 9.09, p< 0.01) and the largest (F2, 13 ed by the addition of C-labelled glucose or un- 150= 3.32, p< 0.05) males, but overall differences labelled glucose to the trays and no differences were small compared to the control males (Ta- 92 were observed between the three treatments ble 5.4).

(F2, 6= 0.07, p> 0.05; Table 5.3). Sperm viability No difference was observed in the hatch rate of Adult size the eggs from females mated to labelled males For each treatment, replicate and sex, wings compared to control males, and the number 05 Stable isotope 13C semen label

Table 5.2. Values are mean ± s.e.m. survival of males from experiments 2 and 3. N is the number of males analysed. Experiment 2: L-L, males were labelled as larvae and adult males were fed labelled sugar (i.e. for the first 5 days, see Materials and methods); L-U, as larvae males were fed labelled sugar and as adults unlabelled sugar; U-U, males unlabelled in the larval and adult stage. Experiment 3: males were reared under the three treatments shown in the larval stage, and maintained as adults on standard (unlabelled) 10% sucrose solution. For each row, values with different letters are significantly different at p< 0.01 (log-rank tests). Experiment Rep. Male mean survival in days

N L-L N L-U N U-U 2 1 34 15 ± 1.22a 32 19 ± 1.16b 33 15 ± 0.94a

N labelled N unlabelled N control glucose glucose no glucose 3 2 49 27 ± 1.31a 49 31 ± 1.26b 50 24 ± 1.45a 3 3 49 21 ± 1.56a 48 31 ± 1.65c 48 29 ± 1.21b

of eggs laid was similar (Labelled: eggs= 1786, adult labelling did label the semen but in low hatch= 0.81; Control: eggs= 2061, hatch= 0.81). amounts, therefore emphasis was put on larval and larval plus adult labelling treatments. Mass spectrometric analyses of the males showed a highly enriched reproductive Discussion system. The level of enrichment of the testis or the accessory glands separately was not de- We have shown that it is possible to use the termined. Hence, we cannot specify the rela- stable isotope 13C as a semen label. Spermath- tive contribution of the label in both fractions ecae inseminated by males labelled in the larval of the semen (i.e. spermatozoa or accessory stage have distinguishably higher δ13C values gland products). Although differential labelling than un-inseminated spermathecae and con- may be of interest at a later stage, for practi- trol samples, and the additional labelling of the cal purposes and applicability of the method adult stage resulted in even higher δ13C values developed our current methods are considered compared to larval labelling alone. Labelling adequate. The highest enrichment that was es- in the adult stage alone was not sufficient; it timated was 8.5 atom% 13C (i.e. sampled 4 days resulted in low amounts of label detected im- after emergence, L-L treatment), while the tar- mediately after mating, and the label seemed get enrichment was 20 atom% 13C. Not all 13C to diminish faster over time in spermathecae of added to the diet is recovered in the mosquito, females isolated after mating. Somewhat simi- because label is lost as a result of respiration in lar results were observed in a study with Aedes the larval trays and turnover in the insect. Respi- aegypti L., where it was found that when adult ration in the larval trays is brought about by mi- males were offered radioactive labelled honey- cro-organisms present in the water. When feed- dew no labelled semen was transferred to the ing, these micro-organisms will incorporate 13C females, even though the males themselves in their system, but due to respiration, 13C is also 13 were highly labelled (Dame and Schmidt 1964). lost from the trays as CO2. As a consequence Exposure in the larval stage did result in posi- of variations in the microbial fauna in larval en- tive semen labelling and the authors discussed vironments and the resulting levels of respira- 93 that perhaps exposure during the early stages tion, the amount of label in trays will vary, and of spermatogenesis, i.e. during the larval stage, this could account for some of the variability was necessary to incorporate the label in the observed in the amount of label transferred by semen (Dame and Schmidt 1964). In our study, males between experiments. Mosquito larvae 05 Stable isotope 13C semen label

Experiment Survival (% out of 500 L1 larvae) Table 5.3. Larval survival from experiment 2 and 3. Survival was measured by counting the labelled unlabelled control

Chapter 05 total number of emerged adults. Experiment glucose glucose no glucose 3 had two replicates per treatment. Mean 2 88 85 99 survival values are ± s.e.m.; values with dif- ferent letters are significantly different at p< 3 80 86 76 0.05 (Tukey HSD). 3 82 82 82

average 83 ± 3a 84 ± 1a 85 ± 7a

are collector-filter feeders (Clements 1992), and 99.7% confidence that all samples were classi- feed on dissolved particles in the water. The la- fied correctly. The few outliers in the dataset, bel could have been ingested directly through especially observed in experiment 1, could not the uptake of 13C glucose, or indirectly through be attributed to contamination during sample the uptake of micro-organisms that utilised the preparation. However, we cannot exclude a supplied larval diet (Merritt et al. 1992). The possible contamination by other samples dur- relative contribution of each pathway remains ing drying, or perhaps cross-contamination by speculative at this stage. the highly labelled males occurred. Even though For use of this technique in experimen- the few outliers were a cause for concern, in the tal settings, persistence of the label in the sper- subsequent experiment no such outliers were mathecae after mating is desirable, as females observed and we are confident that they do not will not always be dissected immediately after invalidate our findings. mating. It was observed that the label was de- Between experiments, the amount tectable in spermathecae up to 7 days after in- of carbon in the samples differed significantly, semination, and significantly higher δ13C values even though the “spike” solution used in all than control samples were reported. Moreover, experiments was derived from the same stock unlabelled sugar feeding of males labelled in and kept at 4 °C. The amount of carbon in the the larval stage did not result in a traceable di- experimental samples is referenced to standard lution of the semen label due to turnover of 13C samples, which are slightly different between with 12C. Males up to 10 days of age transferred experiments, causing these levels of inconsist- similar amounts of label than younger males. ency. However, as the δ13C value is a ratio and An important finding, because adult males and thus independent of the amount of carbon in females replenish energy reserves by sugar the sample, this variation has no impact on our feeding on plant nectars in nature (Foster 1995, findings. The spiking of samples resulted in a di- Clements 1999), and are maintained on sugar lution of the label and complicated calculations solutions in the laboratory. of the actual δ13C values. However, in our mass The threshold values used in the ex- spectrometry set-up this was necessary to raise periments were successful in classifying labelled the detection limit. Nonetheless, the raw δ13C spermathecae from unlabelled samples. Thresh- values could effectively be used for interpreta- old values were different between experiments tion of results and data analysis. because δ13C (‰) values of the reference stand- We have established a proof of princi- ard, in our case virgin females, varied between ple in the laboratory and shown that this tech- 94 experiments. In experiment 1, some variation nique can be used to study a variety of issues in the δ13C values of virgin females resulted in related to mating in anopheline mosquitoes a somewhat conservative threshold, but in the and other insects. Although we applied the la- subsequent experiment, the triple standard bel in the aquatic stage, in insects lacking this deviation threshold value demonstrated with stage, the label can be incorporated in the larval 05 Stable isotope 13C semen label

Sex Replicate Wing length (mm) Table 5.4. Wing lengths (mean ± s.e.m.) of males labelled unlabelled control and females from ex- glucose glucose no glucose periment 3. Each treat- ment had two replicates females 1 3.11 ± 0.01b 3.06 ± 0.01a 3.07 ± 0.01a and each value is based 2 3.04 ± 0.01a 3.07 ± 0.01a 3.06 ± 0.01a on ~ 50 individuals. For each row, values with males 1 2.88 ± 0.01b 2.86 ± 0.01a,b 2.84 ± 0.01a different letters are sig- nificantly different at p< 2 2.78 ± 0.01a 2.81 ± 0.01a 2.85 ± 0.01b 0.05 (Tukey HSD).

diet, but the optimal treatment would need to and similar results were reported when radioac- be determined (e.g. duration of labelling treat- tive isotopes were used to label semen (Young ment, formulation and amount of labelling and Downe 1978). As such, stable isotope label- diet, etc.). Besides laboratory-based studies, ling meets most criteria of an “ideal marker” for there is a great potential to use this technique insects that should be durable, non-toxic, eas- in the context of genetic control studies like ily applied, does not impact on the insects be- the Sterile Insect Technique, etc. The most im- haviour (e.g. growth, reproduction, life span), is portant parameter in these studies is the abil- clearly identifiable, and inexpensive (Hagler and ity of released males to locate and inseminate Jackson 2001). The latter could be contested in wild females, and stable isotopes can be used the case of stable isotope analysis. Even though to determine which group of males was respon- pricing of stable isotopes and sample analysis sible for the insemination. Preferably, these ex- have decreased over the last years (Hood-Now- periments take place in the field or in large field otny and Knols 2007), it is still a relative expen- cages (Knols et al. 2002) to evaluate the insects sive technique to use. One gram of 13C-glucose in their natural environment (Scott et al. 2002). 99 atom% costs 100 USD, and 0.25 g was used Because fitness of the labelled males per larval tray. Sample analyses were done in- is of high importance in these experiments, house but can be outsourced at a cost of 5 USD impact of the label on a number of life-history per sample (Hood-Nowotny and Knols 2007). traits was assessed. Exposure to the label, even Another minor drawback of mass spectrometry at the high quantities that were used, had no is that the sample analysis is destructive, leav- effect on male mating ability. Longevity of la- ing no possibility to repeatedly measure sam- belled males was similar or higher compared to ples. the control in the first two replicates; in the last replicate, a somewhat reduced longevity was observed, but in general longevity was high Conclusion and well beyond any life expectancy in a more natural situation. The labelling of mosquitoes in Larval labelling alone resulted in sufficient the larval stage by adding glucose to the trays amount of label transferred to females to distin- had no effect on larval development and a simi- guish inseminated spermathecea from control lar result was observed in a study applying the samples, and the label persisted in spermathe- same technique but with lower amounts of la- cae for at least 7 days after insemination. Males belled glucose added to the trays (Hood-Now- up to 10 days of age transfer similar amounts otny et al. 2006). Size of the females was not of label as younger males, indicating that lar- 95 affected by the label and while some variation val labelling results in a life-long signature. The in the males was observed, overall differences label had no influence on larval development were small. There is no impact of the label on and survival, longevity or mating ability and is the ability of labelled sperm to inseminate eggs therefore considered an ideal marker. The la- 05 Stable isotope 13C semen label

bel is easy to apply, the sample preparation is Dame, D. A., and C. H. Schmidt 1964. P32 labeled

Chapter 05 straightforward and cost of sample analysis is semen for mosquito mating studies. J. Econ. Ento- reasonable. Although the technology present- mol. 57:669-672. ed was tested in anopheline mosquitoes, other de Groot, P. A. 2004. Handbook of stable isotope candidate insects for genetic control studies, analytical techniques, Volume I. Elsevier Academic e.g. Aedes mosquitoes, fruit flies, tsetse flies, Press, Amsterdam. etc. are likely to benefit from the same tech- Dube, G., A. Henrion, R. Ohlendorf, and W. Rich- nology. We believe that stable isotopes offer ter 1998. Combining isotope ratio monitoring with a great potential to study mating behaviour in isotope dilution techniques for high accuracy quanti- insects. In addition, stable isotopes are envi- fication in organic chemical analysis. Rap. Commun. ronmentally safe and are thus likely to be well Mass. Spectrom. 12:28-32. accepted both within the research community Ephrussi, B., and G. W. Beadle 1936. A technique of and by the public. transplantation for Drosophila. Am. Nat. 70:225. Ferguson, H. M., B. John, K. R. Ng’habi, and B. G. J. Knols 2005. Redressing the sex imbalance Acknowledgements in knowledge of vector biology. Trends Ecol. Evol. We would like to thank S. Borovits for running 20:202-209. the isotope samples, G. Hardarson for support- Foster, W. A. 1995. Mosquito sugar feeding and re- ing the work and H. Bossin, A. Robinson, M. productive energetics. Annu. Rev. Entomol. 40:443- Dicke, and the two anonymous reviewers for 474. constructive comments during preparation of Fry, B. 2006. Stable Isotope Ecology. Springer, New the manuscript. BGJK is supported by a VIDI York. grant from the Netherlands Organisation for Gomulski, L. M. 1988. Aspects of mosquito mating Scientific Research (#864.03.004). behaviour. PhD thesis. London School of Hygiene and Tropical Medicine, University of London. Hagler, J. R., and C. G. Jackson 2001. Methods for marking insects: current techniques and future pros- References pects. Annu. Rev. Entomol. 46:511-543. Hood-Nowotny, R., R. Bol, W. Wanek, and A. Rich- Beard, C. B., M. Q. Benedict, J. P. Primus, V. Finner- ter 2005. Preface. Rap. Commun. Mass Spectrom. ty, and F. H. Collins 1995. Eye pigments in wild-type 19:1363-1364. and eye-color mutant strains of the African malaria Hood-Nowotny, R. C., L. Mayr, and B. G. J. Knols vector Anopheles gambiae. J. Hered. 86:375-380. 2006. Use of carbon-13 as a population marker for Catteruccia, F., J. P. Benton, and A. Crisanti 2005. Anopheles arabiensis in a sterile insect technique An Anopheles transgenic sexing strain for vector con- (SIT) context. Malar. J. 5:1-8. trol. Nat. Biotechnol. 23:1414-1417. Hood-Nowotny, R. C., and B. G. J. Knols 2007. Charlwood, J. D., J. Pinto, C. A. Sousa, C. Ferreira, Stable isotope methods in biological and ecological and V. E. Do Rosario 2002. Male size does not affect studies of arthropods. Entomol. Exp. Appl. 124: 3-16. mating success (of Anopheles gambiae in São Tomé). Klowden, M. J. 2006. Switchover to the mated state Med. Vet. Entomol. 16:109-111. by spermathecal activation in female Anopheles Clements, A. N. 1992. Growth and Development, gambiae mosquitoes. J. Insect Physiol. 52:679-684. pp. 150-170. The biology of mosquitoes, Volume 1. Knols, B. G. J., B. N. Njiru, E. M. Mathenge, W. R. Development, nutrition and reproduction. Chap- Mukabana, J. C. Beier, and G. F. Killeen 2002. Ma- 96 mann & Hall, London. lariaSphere: A greenhouse-enclosed simulation of a Clements, A. N. 1999. Feeding on plant sugars, pp. natural Anopheles gambiae (Diptera: Culicidae) eco- 403-432. The Biology of Mosquitoes, Volume 2: Sen- system in western Kenya. Malar. J. 1:19. sory reception and behaviour. CABI Publishing, Wall- Knols, B. G. J., R. Hood-Nowotny, H. Bossin, G. ingford. Franz, A. Robinson, W. R. Mukabana, and S. K. 05 Stable isotope 13C semen label

Kemboi 2006. GM sterile mosquitoes—a cautionary toes: Assimilation by larvae and retention and trans- note. Nat. Biotechnol. 24:1067-1068. fer during mating. J. Econ. Entomol. 62:851-853. Knols, B. G. J., and C. Louis 2006 (eds.). Bridging Takken, W., and B. G. J. Knols 1999. Odor-mediated Laboratory and Field Research for Genetic Control behavior of Afrotropical malaria mosquitoes. Annu. of Disease Vectors. Proceedings of the joint WHO/ Rev. Entomol. 44:131-157. TDR, NIAID, IAEA and Frontis Workshop on Bridging Tantawy, A. O., A. A. Abdel-Malek, and A. M. Wakid Laboratory and Field Research for Genetic Control 1967. Studies on the eradication of Anopheles pha- of Disease Vectors. Nairobi, Kenya 14-16 July 2004. roensis Theobald by the Sterile-Male technique using Frontis, Wageningen. cobalt-60. IV. Mating behaviour and its frequency in Lounibos, L. P., N. Nishimura, J. Conn, and R. the sterilized mosquitoes. J. Econ. Entomol. 60:23- Lourenco-de-Oliveira 1995. Life history correlates 26. of adult size in the malaria vector Anopheles darlingi. Young, A. D. M., and A. E. R. Downe 1978. Quanti- Mem. Inst. Oswaldo Cruz 90:774. tative assessment with radiotracers of sperm trans- Lyimo, E. O., and W. Takken 1993. Effects of adult fer by male Aedes aegypti (Diptera:Culicidae). J. Med. body size on fecundity and the pre-gravid rate of Entomol. 15:259-264. Anopheles gambiae females in Tanzania. Med. Vet. Entomol. 7:328-332. Macneale, K. H., B. L. Peckarsky, and G. E. Likens 2005. Stable isotopes identify dispersal patterns of stonefly populations living along stream corridors. Freshwater Biol. 50:1117-1130. Malausa, T., M. T. Bethenod, A. Bontemps, D. Bourguet, J. M. Cornuet, and S. Ponsard 2005. As- sortative mating in sympatric host races of the Euro- pean corn borer. Science 308:258-260. Mason, G. F. 1967. Genetic studies on mutations in species A and B of the Anopheles gambiae complex. Gen. Res. 10:205-217. Merritt, R. W., R. H. Dadd, and E. D. Walker 1992. Feeding behavior, natural food, and nutritional rela- tionships of larval mosquitoes. Annu. Rev. Entomol. 37:349-376. Misell, L. M., D. Holochwost, D. Boban, N. Santi, S. Shefi, M. K. Hellerstein, and P. J. Turek 2006. A stable isotope-mass spectrometric method for measuring human spermatogenesis kinetics in vivo. J. Urol. 175:242-246. Ponsard, S., M. T. Bethenod, A. Bontemps, L. Pélo- zuelo, M. C. Souqual, and D. Bourguet 2004. Car- bon stable isotopes: a tool for studying the mating, ovoposition, and spatial distribution of races of Euro- pean corn borer, Ostrinia nubilalis, among host plants in the field. Can. J. Zool. 82:1177-1185. Scott, T. W., W. Takken, B. G. J. Knols, and C. Boëte 97 2002. The ecology of genetically modified mosqui- toes. Science 298:117-119. Smittle, B. J., R. S. Patterson, and C. S. Lofgren 1969. P-32 Labelling of common malaria mosqui- 06 15N semen label for Anopheles mosquitoes 06 Use of the 15N stable isotope as a semen label to detect mating in the malaria mosquito Anopheles arabiensis Patton

by Michelle EH Helinski, Rebecca C Hood, Doris Gludovacz, Leo Mayr and Bart GJ Knols

In previous studies it was determined that the stable isotope 13-carbon can be used as a semen label to detect mating events. In this paper we describe the use of an additional stable isotope, 15-nitrogen (15N), for that same purpose. Both stable isotopes can be analysed simultaneously in a mass spectrometer, offering the possibility to detect both labels in one sample in order to study complex and difficult-to-detect mating events, such as multiple mating. 15N-glycine was added to350-500 larval Millionrearing cases water of and clinical the targetmalaria enrichment occur annually, was 5 60% atom% of which 15N. Males are in fromsub- theseSaharan trays Africa. were Moreover, mated with 80% unlabelled of all deaths virgin attributedfemales, and to spikedmalaria spermathecae occur in this wereregion. analysed In numbers, for isotopic 1 million composition Africans die after of matingthe disease using each mass year, spectrometry. with the vast The influencemajority of of deaths the label occurring on larval among survival children and adult below male five longevity years ofwas age. determined. Pregnant Resultswomen are showed another that major spermathecae risk group; positivemalaria can for cause semen low could birth be weight distinguished and pre- 15 frommature un-inseminated delivery (Rogerson or control et al. samples 2007). The using impact the rawof malaria δ N‰ on values. the economicThe label persistedsituation of in endemic spermathecae countries for is up high to (Gallup 5 days afterand Sachs insemination, 2001), and and the males correlation aged 10between days transferred poverty and similar malaria amounts clearly of demonstrated label as males (Sachs aged and4 days. Malaney There 2002). were no negative Malariaeffects isof a the parasitic label ondisease larval transmittedsurvival and bymale female longevity. mosquitoes Enrichment of the of teneralgenus Anopheles mosquitoes. The after malaria emergence parasite was is 4.85 a protozoan ± 0.10 atom% of the 15 N.genus A threshold Plasmodium value. Theredefined are fouras 3 species standard of malaria deviations parasites above that canthe infect mean humansof virgin under (i.e. natural un-inseminated conditions:spermathecae) Plasmodium samples wasfalciparum successful, P. vivax in classifying, P. ovale aand large P. malariaeproportion. The of firstsamples two speciescorrectly cause (i.e. the on averagemost infections 95%). We worldwide conclude and that P. falciparum15N-glycine Welsh can be is usedby far as the a semen label, and sufficient amounts of label were transferred to determine mating of labelled males. Both 15N and 13C isotopes can now be used as semen labels and 99 the pitfalls and merits of both labels are discussed.

Submitted to Journal of the American Mosquito Control Association 06 15N semen label for Anopheles mosquitoes

Introduction instance paternity in competition experiments,

Chapter 06 or more importantly, to study multiple mating events. The stable isotope composition of sam- f the major life history behaviours of ples is measured using an isotope ratio mass anopheline mosquitoes, mating re- spectrometer (de Groot 2004). Because this de- mains the least understood (Takken vice can analyse the isotopes 13C and 15N simul- O 15 and Knols 1999). The mating behaviour of taneously in one sample, the use of N for the anophelines is difficult to study with mating in labelling of semen was investigated. The per- general performed at dusk under low light con- sistence of the 15N label in males and females, ditions (Charlwood and Jones 1979). Further- and the effect of the label on larval survival and more, Anopheles mosquitoes mate in swarms adult longevity were determined. and mating couples only lasts for 15-20 seconds (Charlwood and Jones 1979, 1980). In the con- text of genetic control studies like the Sterile Materials and methods Insect Technique (SIT), male mating success in the field is of critical importance (Ferguson et Mosquitoes al. 2005), and techniques to study mating be- The Dongola strain of Anopheles arabiensis haviour are needed. Semen labelling was per- Patton was used in the experiments. Rearing formed in the past with radioactive isotopes techniques were identical to those described (Dame and Schmidt 1964, Tantawy et al. 1967, by Helinski et al. (2007); only a slightly larger Smittle et al. 1969, Young and Downe 1978) but volume of water was used in larval trays (i.e. 1.5 these are no longer used due to the hazards re- L). Briefly, 500 L1 larvae were used per tray, and lated to the treatment process. Other options fed a fixed diet of fish food (approx. 0.25 mg/ include the use of mutant strains (Mason 1967, larva/day); adult mosquitoes had continuous Gomulski 1990, Beard et al. 1995, Klowden access to a 6% sucrose solution (w/v). 2006) but these are not readily available for most mosquito species. Labelling The use of stable isotopes as markers Ninety-eight atom% 15N-glycine (NLM-202-1, for insects has become widely available in the Cambridge Isotope Laboratories Inc, Ando- last years, largely as a result of a reduction in ver, MA, USA) was used as a label. Mosquitoes the cost of labelling and sample analysis (Hood- were exposed to the label in the larval stage. Nowotny and Knols 2007). Stable isotopes are The label was added to the larval water on the naturally occurring in the environment, are not same day as the L1 larvae were introduced. The radioactive, and thus safe to use. For nitrogen, level of enrichment was 5 atom% 15N (i.e. 5% the heavy isotope 15N has a natural abundance of all the nitrogen in the diet was 15N). Control of 0.36% and the other light isotope l4N makes trays received neither labelled nor unlabelled up all of the remainder. Recent studies under- glycine. The amount of 15N-glycine added was taken in our laboratory have documented the based on the total amount of nitrogen present successful application of stable isotopes as a in the diet. Mass spectrometry confirmed that population marker in the context of genetic approximately 7.3 percent of the larval diet control studies (i.e. SIT; Hood-Nowotny et al. consisted of nitrogen. Until pupation, 1.25 g 2006), and the suitability of carbon-13 (13C) as (0.25 mg x 500 larvae x 10 days) of larval food a semen label was established (Helinski et al. was added to the tray, thus 4.63 mg of 15N was 100 2007). required. Approximately twenty percent of The aim of this study was to explore the glycine consists of nitrogen, thus 0.02348 g of use of an additional stable isotope for semen- 98 atom% 15N-glycine was added to the larval labelling besides 13C. This would allow dual-la- trays. A solution was made containing 0.1174 g belling of two groups of males to determine for 06 15N semen label for Anopheles mosquitoes

in 250 ml H2O and each tray received 50 ml. So- were maintained until all larvae had pupated lutions were kept at 4 °C. and emerged, or had died. To determine male adult longevity, Experimental setup newly emerged males of the 15N-labelled tray A total of five separate mating experiments and the control tray were placed in a stand- were performed each with different batches of ard rearing cage to monitor survival (N= 25). 15N-labelled males (Table 6.1). All males used in One replicate was performed with males from the experiments were maintained as virgins be- experiment two and a further three replicates fore mating by the removal of females from the with males from experiment six. Mortality was emergence cages within 18 hrs after eclosion. scored regularly until all males had died. For each mating experiment, males were trans- ferred to a new cage to prevent cross-contami- Sample preparation nation. All unlabelled females used as mates in Sample preparation was similar to that de- the experiment were sexed < 18 hrs after emer- scribed in Helinski et al. (2007). However for gence. After mating, females were either dis- these samples, a spike of 10 μl containing a sected immediately (i.e. the following day) at standard ammonium sulphate solution of ~ 20 the end of the mating period (I); or isolated for μg of nitrogen (Dube et al. 1998) was added. dissection at a later stage to assess the persist- Whole body analyses were performed to deter- ence of the label in the spermathecae (II). Males mine the overall enrichment of the mosquitoes; varied in age between 4 to 10 days since pupa- teneral mosquitoes (without spike) were placed tion at the start of the experiments; for an over- individually in tin cups and dried for analysis. view of ages, ratios, and mating duration see Virgin (i.e. spermathecae from virgin females) Table 6.1. If experiments were run for one night, and standard samples (i.e. tin cup containing an equal ratio of males and females were used. just the spike on quartz paper) were included. When experiments were run for more than one night, the number of females was increased Sample analysis and interpretation to generate enough un-inseminated females Sample analysis and interpretation were similar needed to validate the method. to Helinski et al. (2007), and the raw delta val- For three experiments, larval survival ues of the spiked samples were used for data was scored. For each experiment, two replicates analysis. The output from the mass spectrom- of either labelled or control (i.e. unlabelled) eter is a δ15N value, which represents the ratio trays were set-up and adults emerging from the of 15N over 14N against a reference standard. trays were removed and counted daily. Trays The δ15N‰ values reported are referenced to

Table 6.1. Overview of the five mating experiments performed. Age of the males the age at the start of the mating experiment, and the number of males (N) used is indicated. The ratio of males: females (M:F) is given, as well as the number of nights mating took place. Females were dissected the day following the last mating night (I) or isolated and dissected at a later time (II) (i.e. either 3 or 5 days after mating) with the exception of experiment 5. The percentage insemination is given, and (N) is the number of females dissected.

Exp. Age males in Ratio Nights Dissection % Insemination days (N) (M:F) mating (N) 1 7 (50) 1:1 1 I, II (3 days) 47 (34) 2 7-8 (40) 1:1 1 I, II (3 days) 54 (39) 101 3 4-6 (20) 1:2 2 I, II (5 days) 44 (39)

4 9-10 (20) 1:2 3 I, II (3 days) 67 (33)

5 4-5 (25) 1:1.6 3 I 63 (16)

06 15N semen label for Anopheles mosquitoes

the international reference standard for nitro- for completeness and statistical analysis. How-

Chapter 06 gen, i.e. atmospheric nitrogen or AIR (Chapter ever, significantly lower values were reported 5). Negative δ15N‰ values appear when the for spermathecae from virgin females from ex- standard is more enriched in 15N than the spiked periment three compared to the values in the

samples. Samples were analysed at the Seiber- other experiments (F2, 13= 6.42, p< 0.05); thus sdorf laboratories of the International Atomic the threshold value in experiment three was Energy Agency. replaced by the value from experiments 1 and 2. Means throughout the text are reported ± Statistical analysis s.e.m. Prior to analyses, data were checked for nor- mality. When data were normally distributed, Labelling differences between mean δ15N values of treat- Semen labelling ments were analysed using General Linear Labelling with 15N-glycine in the larval stage Models (GLMs) with planned contrasts (Tukey’s resulted in the detection of 15N enrichment HSD) or independent t-tests. When data were in spermathecae when labelled males were not normally distributed, Mann-Whitney U mated with unlabelled females (Fig. 6.1). When tests or Kruskal-Wallis tests were performed, females were dissected immediately after mat- and a Bonferroni correction was made when ing (I), mean δ15N‰ values of spermathecae for more than 2 groups were compared. A thresh- inseminated females were significantly higher old value to distinguish labelled spermathecae than values for un-inseminated females for all from unlabelled spermathecae was defined as 3 experiments (Mann-Whitney U or independent standard deviations (s.d.) above the mean δ15N t-tests, Fig. 6.1). When females were isolated (‰) value of the reference standard (Macneale after mating and dissected at later intervals, et al. 2005), in our case virgin females. Because δ15N‰ values of spermathecae of inseminated samples of experiments 1 and 2 were analysed females were still significantly higher compared simultaneously, their threshold value was simi- to values for spermathecae of un-inseminated lar; the same applied for experiments 4 and 5. females in all experiments ((Mann-Whitney U or Longevity of males was analysed using Kaplan- independent t-tests, Fig. 6.1), with the excep- Meier survival analyses. The obtained survival tion of experiment three (t(8)= 1.59, p> 0.05). curves were pairwise compared using Mantel- However, this was the result of one low δ15N‰ Cox log-rank tests (with Bonferroni correction). value for a spermatheca of an inseminated fe- Larval survival data were analysed with an inde- male in the dataset; without it spermathecae of pendent t-test. All tests were used two-sidedly inseminated females were significantly higher and were performed using the SPSS software compared to un-inseminated ones (t(7)= 2.87, version 12 (SPSS Inc., Chicago, USA). p< 0.05). Isolation of the females did not result in significantly lower mean δ15N‰ values com- pared to the values observed after immediate Results dissection for each experiment (Mann-Whitney U tests; data not shown), except for experiment Technical difficulties with the mass spectrom- three (t(15)= 2.54, p< 0.05; Fig. 6.1); again after eter caused variance in standard reference removal of this one low value no significant dif- samples between experimental runs. This was ferences were observed (t(7)= 2.06, p> 0.05). most pronounced for experiment three and to Males used in the experiments var- 102 a lesser extent for experiments 4 and 5. These ied in age between 4 to 10 days at the start of inconsistencies probably have resulted in a few mating and the mating period lasted between erroneous δ15N‰ values for some samples. As 1-3 nights (Table 6.1). The oldest males were we deemed it inappropriate to simply remove used in experiment four and the youngest in unwanted results, all data points were included experiment five; however, δ15N‰ values for 06 15N semen label for Anopheles mosquitoes

Figure 6.1. Mean ± s.e.m. δ15N‰ values of spermathecae of inseminated (filled symbols) and un-inseminated (open symbols) females from experiments 1-5. N is the number of spermathecae analysed for inseminated (+) and un-inseminated (-) females. The dotted line indicates the threshold value of 3 s.d. above mean δ15N‰ of virgin females. Virgin (V) and Standard (St) samples are included. Dissection treatments are: I, females dissected immediately after mating; II, females isolated and dissected 3-5 days after mating (see Table 6.1). Values with different letters are significantly different at p< 0.05 with comparisons made between dissection treatments I and II for each experiment; asterisks indicate significant difference between spermathecae of inseminated and un-inseminated females for each treatment at * p< 0.05, and ** p< 0.01 (comparisons made with independent t-tests or Mann-Whitney U tests).

inseminated spermathecae were similar (U= could be used to identify inseminated sper- 36, r= -0.38, p> 0.05). The δ15N‰ values of mathecae from un-inseminated spermathecae un-inseminated spermathecae were similar to in the large majority of cases (Table 6.2). Some samples from virgin females; and both were false positives and negatives were observed in significantly higher than the standard samples the dataset, but overall their number was low (i.e. without spermathecae; X2= 35.11, df= 2, p< (Table 6.2). In experiment three the highest 0.01). Between experiments, the δ15N‰ values proportion of wrongly classified samples was 103 of spermathecae from virgin females from ex- observed, and this is likely to be the result of periments 1-2 and 4-5 were very similar (t(11)= the difficulties experienced with the sample 0.54, P> 0.05), thus threshold values were al- analysis. most identical (Fig. 6.1). The threshold values 06 15N semen label for Anopheles mosquitoes

Exp. Inseminated Un-inseminated/ Table 6.2. The total number of samples analysed no spermatheca (N) and the number of falsely classified samples

Chapter 06 N N false N N false (N false) for inseminated and un-inseminated samples using the threshold values for all 1 16 0 14 0 experiments. Virgin (V) and standard (St) samples are included. 2 21 0 18 2

3 17 1 13 3

4 21 2 6 0

5 10 0

St 30 1

V 16 0

The proportion of females inseminat- vival; survival of larvae was high and similar to ed in the experiments is indicated in Table 6.1. survival of larvae in the control trays (t(4)= 0.32, After 1 night of mating in experiments 1 and p> 0.05; Table 6.3). The development of the lar- 2, approximately 50% of females were insemi- vae was normal and synchronous to larvae from nated. In the other experiments, mating period the control trays (data not shown). was prolonged and the number of females in- creased and insemination was between 44-67% (Table 6.1). Longevity Adult male longevity was not affected by the Whole body analysis 15N-glycine label. In experiment two, a higher Only a limited number of teneral mosquitoes longevity for 15N males was observed (X2= 6.89, (N= 4; 2 for each sex) were analysed to deter- p< 0.01). In experiment six, replicates were just mine the enrichment level, which was on aver- not statistically similar and data could not be age 4.85 ± 0.10 atom% 15N. pooled, however labelled males had a similar or higher longevity compared to control males (Table 6.4). Larval survival 15N-labelled glycine had no effect on larval sur- Discussion Table 6.3. Survival of larvae from 15N-labelled trays and control trays for three replicates. Survival was measured The data presented here confirm that 15N- by counting the total number of emerged adults. Mean glycine can be used as a semen label to detect percentage survival ± s.e.m. is given; values with different letters are significantly different at p< 0.05 (independent mating, and that sufficient amounts of label t-test). were transferred to distinguish inseminated Replicate Survival (% out of 500 L1 larvae) females from un-inseminated ones. Isolation of the females after mating did not result in a 15 N glycine control loss of label with the exception of experiment 1 73 78 three; however, this was the result of one low 104 value for a spermatheca from an inseminated 2 99 96 female skewing the dataset. The label persisted 3 81 88 in the males for up to ten days of age and simi- lar levels of label were transferred compared to average 84 ± 8a 87 ± 5a younger males. Even though some problems

06 15N semen label for Anopheles mosquitoes

Experiment Male mean survival in days Table 6.4. Mean adult survival ± s.e.m. in days of males from experiments 2 and 6 (3 N 15N-glycine N control replicates) for 15N-glycine labelled males and control (unlabelled) males. N is the 2 25 24 ± 2.14a 25 18 ± 2.33b number of males analysed. Values with different letters are significantly different 6 25 32 ± 1.74a 25 29 ± 1.15b at p< 0.01 for experiment two (log-rank tests), and p< 0.006 for experiment six 6 25 31 ± 1.48a,b 25 29 ± 1.64a,b (Bonferroni correction). 6 25 26 ± 2.09a,b 25 31 ± 1.79a,b

were observed with the sample analysis, we are with 15N-glycine was sometimes easier than 13C- confident that the data presented is a fair rep- glucose. Even though in our initial data compa- resentation since no data got excluded in the rable levels of larval survival were observed for analyses. In spite of these difficulties, 95 ± 2% larvae reared with 13C glucose, in subsequent of all samples were classified correctly. experiments a reduced growth rate and survival The addition of 15N-glycine to the lar- in 13C trays was sometimes observed compared val trays had no impact on the larvae; develop- to unlabelled or 15N-glycine trays (Helinski et al. ment and survival of labelled larvae were similar 2008). In larvae of Aedes aegypti Wiedemann to those for larvae in unlabelled trays. The label- the addition of glucose at high concentrations ling in the larval stage did not have an impact on was observed to slow development rates (Cle- adult male longevity. Mating ability of labelled ments 1992), and it is likely that such an effect males was good, and after 1 night of mating ap- occasionally took place after 13C-glucose was proximately 50% of females were inseminated. added. It should be noted that labelling with Results were comparable to previous data col- 13C-glucose is possible but adequate atten- lected for unlabelled males and males labelled tion should be given to prevent a reduction in with 13C-glucose (Helinski et al. 2007). growth. The sample analyses in the mass spec- Based on the findings in this paper trometer tends to be easier for 13C compared and those previously reported (Helinski et al. to 15N when working with the quantities used 2007), both 15N and 13C isotopes can be used as for these kind of studies (i.e. small leaks have a semen-label to detect mating. The mean dif- a much greater impact for nitrogen measure- ference in delta values between inseminated ments than carbon because of the constitution and un-inseminated samples was similar for of air (i.e. 80% N versus 0.03% C), and these is- both labels and was approximately 4 delta (‰). sues probably accounted for the difficulties ob- In terms of costs, 15N is preferable over 13C as served in the sample analyses. Even though the smaller amounts of label are required to achieve labelling with 5 atom% of 15N-glycine resulted a similar level of enrichment. For 13C, 20 atom% in a sufficiently high amount of label present for in the diet resulted in 4.64 atom% enrichment the detection of inseminated spermathecae in in the teneral mosquito, while for 15N these this study, in subsequent experiments where values were 5 atom% resulting in 4.85 atom% sample analyses were performed at a different enrichment. Carbon is lost during respiration facility, this was not always the case (Helinski et and there is some dilution of adult labelled car- al. 2008), and it is recommended that a slightly bon due to unlabelled sugar intake, however higher level (for instance double the amount) of this does not happen in the case of nitrogen. label is used in further experiments. 105 For these experiments, the labelling of one lar- val tray of 13C-glucose costs 25 USD versus 2.5 Conclusion USD for 15N-glycine. When comparing the two labels in terms of larval development, labelling The labelling of semen with 15N was successful. 06 15N semen label for Anopheles mosquitoes

The isotopes 15N and 13C can now be used in a Gomulski, L. M. 1990. Polyandry in nulliparous

Chapter 06 dual labelling system to study mating behav- Anopheles gambiae mosquitoes (Diptera: Culicidae). iour in for instance competition experiments, Bull. Entomol. Res. 80:393-396. or to determine the presence of multiple mat- Ferguson, H. M., B. John, K. R. Ng’habi, and B. ing events. G. J. Knols 2005. Redressing the sex imbalance in knowledge of vector biology. Trends Ecol. Evol. 20:202-209. Acknowledgements Helinski, M. E. H, R. C. Hood-Nowotny, L. Mayr, The authors are thankful to G. Harderson for and B. G. J. Knols 2007. Stable isotope-mass supporting to work, to M. Dicke for constructive spectrometric determination of semen transfer in comments, and to the International Atomic En- malaria mosquitoes. J. Exp. Biol. 210: 1266-1274. ergy Agency (IAEA) for funding this study. BGJK CHAPTER 5 is supported by a VIDI grant from the Nether- Helinski, M. E. H, R. C. Hood, and B. G. J. Knols lands Organisation for Scientific Research 2008. A stable isotope dual-labelling approach to (#864.03.004). detect multiple insemination in normal and irradiated Anopheles arabiensis mosquitoes. Parasites & Vectors, submitted. CHAPTER 9. Hood-Nowotny R. C., L. Mayr, and B. G. J. Knols References 2006. Use of carbon-13 as a population marker for Anopheles arabiensis in a sterile insect technique Beard, C. B., M. Q. Benedict, J. P. Primus, V. (SIT) context. Malar. J. 5:1-8. Finnerty, and F. H. Collins 1995. Eye pigments in Hood-Nowotny R. C., and B. G. J. Knols 2007. Stable wild-type and eye-color mutant strains of the African isotope methods in biological and ecological studies malaria vector Anopheles gambiae. J. Hered. 86:375- of arthropods. Entomol. Exp. Appl. 124: 3-16. 380. Klowden, M. J. 2006. Switchover to the mated state Clements, A. N. 1992. Larval Nutrition, Excretion by spermathecal activation in female Anopheles and Respiration. pp 100-118. In The biology of gambiae mosquitoes. J. Insect Physiol. 52:679-684. mosquitoes, Volume 1. Development, nutrition and Macneale K. H., B. L. Peckarsky, and G. E. Likens reproduction. Chapmann & Hall, London, United 2005. Stable isotopes identify dispersal patterns of Kingdom. stonefly populations living along stream corridors. Charlwood, J. D., and M. D. R. Jones 1979. Mating Freshw. Biol. 50: 1117-1130. behaviour in the mosquito Anopheles gambiae s.l. I. Mason, G. F. 1967. Genetic studies on mutations in Close range and contact behaviour. Physiol. Entomol. species A and B of the Anopheles gambiae complex. 4:111-120. Gen. Res. 10:205-217. Charlwood, J. D., and M. D. R. Jones 1980. Mating Smittle, B. J., R. S. Patterson, and C. S. Lofgren in the mosquito, Anopheles gambiae. s.l. II. Swarming 1969. P-32 Labelling of common malaria mosquitoes: behaviour. Physiol. Entomol. 5:315-320. Assimilation by larvae and retention and transfer Dame, D. A., and C. H. Schmid 1964. P32 labeled during mating. J. Econ. Entomol. 62:851-853. semen for mosquito mating studies. J. Econ. Takken, W, and B. G. J. Knols 1999. Odor-mediated Entomol. 57:669-672. behavior of Afrotropical malaria mosquitoes. Annu. de Groot, P.A. 2004. Handbook of stable isotope Rev. Entomol. 44:131-157. analytical techniques, Volume I. Elsevier Academic Tantawy, A. O., A. A. Abdel-Malek, and A. M. Press, Amsterdam, The Netherlands. Wakid 1967. Studies on the eradication of Anopheles 106 Dube, G., A. Henrion, R. Ohlendorf, and W. pharoensis Theobald by the Sterile-Male technique Richter 1998. Combining isotope ratio monitoring using cobalt-60. IV. Mating behaviour and its with isotope dilution techniques for high accuracy frequency in the sterilized mosquitoes. J. Econ. quantification in organic chemical analysis. Rap. Entomol. 60:23-26. Commun. Mass. Spectrom. 12: 28-32. Young, A. D. M, and A. E. R. Downe 1978. 06 15N semen label for Anopheles mosquitoes

Quantitative assessment with radiotracers of sperm transfer by male Aedes aegypti (Diptera:Culicidae). J. Med. Entomol. 15:259-264.

107 PART 3 Part 3

Mating competitive- ness in genetic control studies PART 3 PART

109 07 Competitiveness of radio-sterilised An. arabiensis 07 Mating competitiveness of male Anopheles arabiensis mosquitoes irradiated with a partially- or fully- sterilising dose in small and large laboratory cages by Michelle EH Helinski and Bart GJ Knols

Male mating competitiveness is a crucial parameter in many genetic control programmes including the Sterile Insect Technique (SIT). We evaluated competitiveness of male Anopheles arabiensis Patton as a function of three experimental variables: (i) small or large cages for mating, (ii) the effects of either a partially-sterilising (70 Gy) or fully-sterilising (120 Gy) dose, and (iii) pupal or adult irradiation. Irradiated males competed for females with an equal number of un- irradiated males. Competitiveness was determined by measuring hatch rates of individually laid egg batches. In small cages, pupal irradiation with the high dose 350-500resulted Million in the cases lowest of competitiveness clinical malaria occur whereas annually, adult 60% irradiation of which with are the in sub- low Saharandose gave Africa. the highest, Moreover, with 80%the latter of all males deaths being attributed equal in to competitiveness malaria occur into thisun- region.irradiated In males.numbers, In the 1 million large cage, Africans reduced die of competitiveness the disease each of year,males with irradiated the vast in majoritythe pupal of stage deaths was occurringmore pronounced among childrencompared below to the five small years cage. of In age. contrast, Pregnant the womenmales irradiated are another as adults major at risk both group; doses malaria performed can causesimilarly low to birth un-irradiated weight and males. pre- Unexpectedly,mature delivery males (Rogerson irradiated et al.with 2007). the high The doseimpact performed of malaria better on inthe a largeeconomic cage thansituation in a ofsmall endemic one. Acountries high proportion is high (Gallup of intermediate and Sachs hatch2001), ratesand the was correlation observed forbetween eggs collected poverty andin the malaria large cageclearly experiments demonstrated with (Sachs males andirradiated Malaney at the2002). pupal stage. It Malaria is concluded is a parasitic that irradiation disease transmitted of adult An. by arabiensis female mosquitoes with the partially- of the sterilisinggenus Anopheles dose results. The inmalaria the highest parasite competitiveness is a protozoan for of boththe genus cage designs. Plasmodium Cage. sizeThere affected are four speciescompetitiveness of malaria forparasites some that treatments. can infect humansTherefore, under competitiveness natural determinedconditions: Plasmodiumin laboratory falciparum experiments, P. must vivax be, P. confirmed ovale and P.by malariaereleases .into The simulated first two fieldspecies conditions. cause the Themost protocols infections described worldwide are andreadily P. falciparum transferable Welsh to evaluate is by far themale competitiveness for other genetic control techniques. 111

Accepted for publication in Journal of Medical En- tomology 07 Competitiveness of radio-sterilised An. arabiensis

Introduction work, mosquitoes were usually irradiated with a

Chapter 07 high (i.e. fully-sterilising) dose, even though this he use of genetic control techniques to is known to affect male competitiveness nega- suppress or replace mosquito popula- tively. In An. stephensi Liston it was observed Ttions is receiving considerable attention that males irradiated as pupae with 80 Gy were (Takken and Scott 2003, Knols and Louis 2006), almost twice as competitive as males irradiated ranging from the use of transgenic mosquitoes as pupae with 120 Gy (Sharma et al. 1978), and (Thomas et al. 2000, Alphey et al. 2002) to Wol- similar results were obtained with Culex quin- bachia-induced incompatibility (Laven 1967, quefasciatus Say (El-Gazzar et al. 1983). Steril- Dobson et al. 2002, Dobson 2003) and the more ity induced in a field population is a function of familiar methods of sterilisation by ionising ra- competitiveness of the males and their level of diation (Dyck et al. 2005). Chemosterilants have induced sterility and this relationship has to be also been used successfully (Dame et al. 1981, considered when determining the optimal irra- Dame 1985) but have fallen out of favour be- diation dose (Dame and Schmidt 1962, Parker cause of environmental and human health con- cerns (Hayes 1968, Bracken and Dondale 1972). and Mehta 2007). Therefore, in this study, the Although the above-mentioned methods rely competitiveness of males irradiated with either on different principles to create sexual sterility, a low (i.e. partially-sterilising) or high (i.e. fully- the common underlying premise is that males sterilising) dose was assessed. reared in the laboratory need to locate, com- For logistic reasons, the irradiation pete for, and successfully transfer sperm to wild of pupae is preferred in many operational SIT females for population reduction or replace- programmes (Dyck et al. 2005) because pupae ment to occur. Competitiveness of males is thus are more robust and therefore easier to handle of critical importance and assays to determine compared to adults. However, adult male mos- its level are vital to monitor the quality of the quitoes are less affected by the somatic effects insects to be released. Preferably, competition of irradiation and have a higher competitive- experiments should be performed under semi- ness after irradiation compared to males irradi- field conditions, e.g. in field cages (Knols et al. ated in the pupal stage (Curtis 1976, El-Gazzar et 2002), to assess male quality under near-natu- al. 1983, El-Gazzar and Dame 1983, Andreasen ral conditions and thus in a realistic manner. In and Curtis 2005). We therefore determined the practice, these kinds of facilities are not readily competitiveness of males irradiated with a low available, and most competition experiments or high dose both in the pupal or adult stage. are performed on a smaller scale inside the lab- We hypothesised that a low dose would result in oratory. In this study, we aimed to extend the increased competitiveness compared to a high scope of laboratory assays by including, in addi- dose and that pupal irradiation would result in tion to the routine small rearing cages normally a greater loss of competitiveness than adult ir- used for such experiments, a large cage intend- radiation. Furthermore, the influence of space ed to study males under more challenging and possibly more realistic conditions. (i.e. small versus large cages) on male competi- In the framework of an integrated ap- tiveness was studied. It was hypothesised that proach to control Anopheles arabiensis Patton, in large cages males would be challenged to a one of the three main malaria vectors in sub- greater extent compared to the small cages, Saharan Africa, irradiation studies were carried and thus evidence of differences in competi- 112 out in preparation for the future release of ster- tiveness would become more evident. ile males. Only a limited number of irradiation studies on An. arabiensis have been performed (Curtis 1976, Helinski et al. 2006) emphasising the need for a more extensive study. In previous 07 Competitiveness of radio-sterilised An. arabiensis

Material and Methods resting place for An. arabiensis during daylight hours (M. Benedict and B. Knols, unpublished Mosquitoes data). A similar tube was placed in the small The Dongola strain of Anopheles arabiensis was cage for resting. Adults were continuously sup- used and reared according to the methods of plied with 6% sucrose solution [w/v]. Helinski et al. (2006). For the small cage experi- ments, adults were kept in standard 30 x 30 x Irradiation 30 cm mosquito rearing cages maintained at a Insects were exposed to gamma rays generated temperature of 27 ± 1 °C and relative humidity by a cobalt-60 source with a dose rate of ca. 12 of 82 ± 2%. The light regime was L10:D12 (from Gy/min. Males were irradiated in the pupal or 0- 2,000 lux) with one hour computer controlled adult stage, and irradiation procedures and simulated dusk and dawn periods in between. handling methods are as described in Helinski The large cage experiments were performed et al. (2006) except that in the current experi- in a different room with similar conditions (27 ments a slightly larger container (i.e. diameter ± 1 °C and RH: 77 ± 1%), with a slightly longer 3.5 cm) was used to hold the pupae during ir- (1.5 hrs) dawn and dusk period using two lamps radiation. A dosimetry system was used to with different light intensity (i.e. 140 and 50 lux measure the dose received by the batch based measured at 1 m, respectively) set on timers. on the Gafchromic HD-810 film (International The large cages measured 1.8 (l) x 1.2 (w) x 1.0 Specialty Products, NJ, USA; IAEA 2004). Three (h) m (MoTec Pop-Up, Brettschneider, Heim- dosimeters were included with each batch of stetten, Germany; Fig. 7.1), and had a volume insects and read after irradiation with a Radia- approximately forty times greater than the chromic® reader (Far West Technology, Inc., small cages. In the large cage, a swarm marker California, USA). was added, consisting of a round piece of black Pupae were not separated by sex prior paper placed in the centre of the bottom of the to irradiation, and were irradiated 20-26 hrs af- cage, as well as resting tubes made of black con- ter pupation. After irradiation pupae emerged struction material (Fig. 7.1). We have observed overnight in a small cage, and the following that a dark resting tube is a strongly preferred morning males were removed and transferred

Figure 7.1. The large cage used in the experiments. Cages measured 1.8 (l) x 1.2 (w) x 1.0 (h) and were mounted on a wooden base with drawing pins; the edges were secured with bulldog clips. The bottom of the cage was lined with sheets of white paper. The cage was accessed through 1 large zipper at the front; 2 sugar feeders (Sf), a swarm marker (Sm) and resting tubes (Rt) were supplied. 113 07 Competitiveness of radio-sterilised An. arabiensis

to the competition cages. For adult irradia- Large cage experiments

Chapter 07 tion, males aged < 24 hrs were irradiated and In these cages 500 irradiated males, 500 un- introduced into the competition cages immedi- irradiated males and 500 un-irradiated females ately after irradiation. Based on previous work were used and mating took place for seven (Helinski et al. 2006), 70 Gy was selected as the nights. A sample of thirty females was dissect- low, partially-sterilising dose resulting in 83 ± ed prior to the first blood meal (i.e. after seven 2% and 92 ± 3% induced sterility when applied nights of mating) to determine their insemina- to the pupal or adult stage respectively. As the tion status. Cumulative mortality of males and high, fully- sterilising dose, 120 Gy was chosen females was scored for the first seven days. which resulted in > 99% sterility in both devel- To determine the mating and survival opmental stages (Helinski et al. 2006). of un-irradiated and irradiated males in the large cage in the absence of competition, a number of Experimental setup control experiments were performed. For each Irradiated males were introduced into the treatment (i.e. 70 or 120 Gy for pupal or adult competition cages together with un-irradiated stage) 500 males and 500 females were intro- males and females at the ratio of 1:1:1 (Davis et duced and mating took place for seven nights. al. 1959, Tantawy et al. 1967, Sharma et al. 1978, Cumulative mortality was scored for seven Andreasen and Curtis 2005). The un-irradiated days, and after seven nights a sample of thirty males and females in all experiments were de- females were dissected for insemination. Each rived from the same batch of insects as the ir- treatment was replicated three times. radiated males. Females used as mates were separated < 18 hrs after emergence to ensure Blood feeding and egg laying virginity and were introduced into the competi- Females were blood-fed twice on membranes tion cages on the same day as the males. Doses filled with heparinised human blood. In the were tested pair-wise per developmental stage small cage females were fed twice on consecu- (i.e. pupae or adult irradiation), and three repli- tive days between days 3-5 after the start of the cates with different batches of males (i.e. differ- experiment, and unfed females were removed ent generations) were carried out for each treat- from the cages after the last blood meal. In the ment. Small cage experiments were performed large cage, females were transferred to a small first. Subsequently the same treatments were cage after seven nights of mating and blood tested in the large cage experiments. fed twice; again unfed females were removed. Three to four days after blood feeding, approxi- Small cage experiments mately 100 females from the small cages and In these cages 125 irradiated males, 125 un- 150 females from the large cages were isolated irradiated males and 125 un-irradiated females individually for egg laying. For every egg batch, were used and mating took place for 6-7 nights the hatch rate was determined after 5-7 days. (Sharma et al. 1978). In previous experiments Across all treatments, egg batches with fewer (Helinski et al. 2006) it was demonstrated that than 20 eggs were excluded from analyses to pupal emergence and male longevity were not avoid incomplete egg laying behaviour influ- affected by irradiation and the ability of irradi- encing the statistical models applied. ated males to inseminate females was deter- The level of inherent (control) sterility mined; therefore those parameters were not in the colony, and the sterility induced by males assessed in this study. Cumulative mortality of irradiated as pupae or adults with 70 and 120 Gy 114 males (i.e. combined data for irradiated and was determined in a similar manner. For each un-irradiated males) and females was scored treatment, 2-4 replicates were set up where for the first six days. These data were compared 125-150 males were mated with virgin females with mortality data from the large cage. in a 1:1 ratio in small cages; females were blood- fed and isolated individually as above, and hatch 07 Competitiveness of radio-sterilised An. arabiensis

Figure 7.2. Distribution of hatch rates of egg batches from control matings (i.e. in the absence of competition), small cage competition, and large cage competition experiments for pupal (closed circles) and adult (grey circles) irradiation. N is the number of egg batches analysed per treatment. Dotted line for each treatment indicates the cut-off point to classify batches fathered by irradiated males (left of the line) from un-irradiated males (right of the line). This point is derived from logistic regression analysis results (see text). rates were determined. These control data were used to construct logistic regression models. treatment. If there was no heterogeneity be- tween replicates of a treatment, data were Statistical analysis pooled. Data from the control experiments were used to General Linear Models (GLMs) were develop logistic regression models (Field 2005) used to compare the proportion of un-irradi- for classifying the egg batches from competi- ated batches between treatments, and means tion experiments as either deriving from irradi- were separated using Tukey’s Honestly Signifi- ated or un-irradiated males. Competitiveness cantly Different (HSD) and individual t-tests. To was then measured by the proportion of egg facilitate comparison with similar studies in the batches derived from either irradiated or un- literature the competitiveness index C (C = # ir- irradiated males (i.e. referred to as irradiated radiated batches/ # un-irradiated batches) was or un-irradiated batches throughout the text). calculated (Haisch 1970, Fried 1971). Cumulative 115 The distribution of irradiated and un-irradiated mortality data were analysed using the GLM batches was analysed using a replicated G-test procedure. All two-sided tests were performed of goodness of fit (Sokal and Rohlf 1981), that using the SPSS software version 14 (SPSS Inc., allowed for the analysis of replicates within a Chicago, USA). 07 Competitiveness of radio-sterilised An. arabiensis

Results in the competition experiments, insemination

Chapter 07 was comparable, i.e. 75 ± 2%, and no differ- Mean values throughout the text are reported ences were observed between the treatments

± s.e.m. Figure 7.2 shows the distribution of the (F3,8= 0.22, p> 0.05; Table 7.1B). No differences egg batches and their grouping into batches were observed in cumulative male mortality af- fathered by irradiated or un-irradiated males ter seven nights for irradiated or un-irradiated using the logistic regression analysis results. males for all treatments in the absence of com-

The logistic regression analysis was necessary petition (F4,4.25= 2.96, p> 0.05), and cumulative because a substantial number of intermedi- mortality was on average 22 ± 3%. ate hatch rates were observed in the control G-test results for the number of irradi- data (Fig. 7.2). Overall, the models fitted the ated and un-irradiated batches are described in data well, as the number of experimental egg Table 7.1B. Again, replicates were statistically batches with a predicted probability between similar and data were pooled. Males irradiated 0.45- 0.55 (i.e. indicating poor fit) was < 4% in all in the pupal stage were less competitive than treatments. Dosimetry confirmed that all doses males irradiated in the adult stage regardless of

delivered were within a 5% error range. the dose received (F3,8= 39.86, p< 0.01; Fig. 7.3). Following adult irradiation with either dose, no Small cage experiments differences were observed in the proportion of G-test results for the number of irradiated and irradiated and un-irradiated batches indicating un-irradiated batches are shown in Table 7.1A. equal competitiveness (t(4)= 0.27, p> 0.05; Ta- The replicated G-test indicated that replicates ble 7.1B; Fig. 7.3). Males irradiated in the pupal were statistically similar for each treatment, and stage with either dose were significantly less data were therefore pooled (Table 7.1A). With competitive than un-irradiated males (Table the exception of the adults irradiated with the 7.1B), and the latter were responsible for 73 ± low dose, irradiated males were less competi- 2% (low dose) and 84 ± 2% (high dose) of the tive than un-irradiated males in all treatments egg batches, respectively. Males irradiated in (Table 7.1A). Especially males irradiated with the pupal stage with the low dose were more the high dose as pupae or adults were less com- competitive than males irradiated with the petitive (Table 7.1A) with 75 ± 5% and 65 ± 6% high dose (t(4)= -4.14, p= 0.01), however in the of the egg batches derived from un-irradiated GLM procedure a p-value of 0.07 was observed males, respectively (Fig. 7.3). Statistically signif- grouping both treatments in the same subset icant differences were observed in the propor- (Fig. 7.3). tion of egg batches fathered by un-irradiated The proportion of individual females that males between adults irradiated with the low oviposited in the four treatments was similar

dose and pupae irradiated with the high dose (F3,8= 0.36, p> 0.05) and was on average 59 ±

(F3,8= 5.09, p< 0.05), with the other treatments 4%. distributed between both groups (Fig. 7.3). The proportion of individual females that Comparison of large versus small cages oviposited in the four treatments averaged 69 ± Males irradiated in the pupal stage with the low

2%, and was similar between treatments (F3,8= dose were less competitive in the large cage 1.98, p> 0.05). compared to the small one (Fig. 7.3); this result was just significant (t(4)= -2.84, p= 0.047), but Large cage experiments also clearly visible from the individual G-test 116 In the absence of competition, irradiated males data (Table 7.1). When males were irradiated inseminated females at a rate comparable to the as pupae with the high dose, no differences

un-irradiated males (F4,5.19= 0.85, p> 0.05) and were observed between the performance of insemination was 74 ± 3 % after seven nights. the males in both cages (t(4)= -1.57, p> 0.05), Even though the number of males was doubled and in both cages competitiveness was low 07 Competitiveness of radio-sterilised An. arabiensis

Table 7.1. The number of batches fathered by un-irradiated or irradiated males from the small cage (A) and large cage (B) competition experiments where males irradiated as pupae or adults with 70 or 120 Gy competed with normal males for females on a 1:1:1 ratio. Individual, replicated and pooled G-tests results are given, as well as the competitiveness (C) value (mean ± s.e.m) based on the three replicates for each treatment. The mean proportion of egg batches with intermediate hatch rate (i.e. between 20-60%) is presented. Mean insemination data per treatment are given for large cage experiments (b). Asterisks indicate significant differences between batches of un-irradiated and irradiated males * p< 0.05 and ** p< 0.01. § Values followed by the same letter are not statistically different from each other, p< 0.05 (Tukey HSD). !!!!!!!!! A Stage Dose Batches G-tests C Batches with inter- rmediate hatch rate (%) un-irrad. irrad. indiv. pooled replic. pupae 70 45 26 5.15* 4.83 * 3.30 0.76 ± 10.2 ± 5 0.16 31 19 2.91

26 28 0.07

120 62 11 39.31** 55.00** 5.83 0.34 ± 5.5 ± 2 0.10 29 15 4.53*

50 17 16.98**

adult 70 30 29 0.02 0.81 0.43 0.85 ± 11.3 ± 1 0.07 27 23 0.32

23 17 0.90

120 35 23 2.50 17.79** 1.78 0.54 ± 8.1 ± 4 0.07 27 15 3.48

52 21 13.59**

B Stage Dose Batches G-tests C Batches with Inseminat § un-irrad. irrad. indiv. pooled replic. intermediate ion (%) hatch rate (%) pupae 70 70 23 24.88** 61.15** 0.48 0.37 ± 0.04 31.7 ± 3 73 ± 5a

81 37 16.81**

60 20 20.93**

120 61 9 43.33** 139.22** 0.94 0.19 ± 0.02 16.2 ± 2 72 ± 3a

92 18 54.45**

80 18 42.38**

adult 70 55 46 0.80 0.03 1.69 1.05 ± 0.11 13.2 ± 2 78 ± 6a

36 43 0.62

39 44 0.30

120 55 50 0.24 0.06 3.34 1.05 ± 0.18 7.4 ± 2 76 ± 6a 117

36 51 2.60

35 29 0.56

07 Competitiveness of radio-sterilised An. arabiensis

Figure 7.3. Proportion of egg batches (mean ± s.e.m.) fathered

Chapter 07 by un-irradiated males from small (open bars) and large (filled bars) cage competition experi- ments where males irradiated as pupae or adults with 70 or 120 Gy were competing against un- irradiated males for un-irradiated females at a ratio of 1:1:1. Bars with the same letter/number are not statistically different from each other, p< 0.05 (Tukey HSD), with comparisons made between the small cage treatments (small letters) and the large cage treat- ments (numbers). Asterisks indi- cate significant differences per treatment between small and large cages, p< 0.05 (t-tests).

with > 75% of the egg batches originating from seven days) was similar to the mortality ob- un-irradiated males (Fig. 7.3). Following irra- served in the small cages (i.e. after six days) diation with the low dose in the adult stage, no and on average 22 ± 3% and 22 ± 2% cumulative differences were observed in competitiveness mortality for males and females was observed, between small and large cages (t(4)= 1.14, p> respectively. 0.05) and equal competitiveness to the un-irra- diated males was observed. Surprisingly, a high dose given to adult males resulted in reduced Discussion competitiveness in small but not in large cages (t(4)= 3.04, p< 0.05; Fig. 7.3). It has previously been reported for a variety of The proportion of egg batches with mosquito species including anophelines that an intermediate hatch rate, defined as being radiation reduces mating competitiveness, between 20-60%, was greater in certain treat- and that irradiation of pupae causes a greater ments in the large cages than in the small cages loss than irradiation of adults (Curtis 1976, El- (Table 7.1), or the control treatments (Fig. 7.2). Gazzar et al. 1983, El-Gazzar and Dame 1983, Particularly for the pupal treatments with ei- Andreasen and Curtis 2005). Our findings sup- ther dose a significant increase in the number port these conclusions. Also the competitive- of batches with intermediate hatch rate was ness values following pupal and adult irradia- observed in the large cage compared to small tion with a high dose of 120 Gy were similar to 118 ones (70 Gy: t(4)= -3.83, p< 0.05; 120 Gy: t(4)= those reported for An. gambiae s.s. (Andreasen -3.52, p< 0.05). and Curtis 2005). The effect of a low, partially- Cumulative mortality data over the sterilising dose on both developmental stages first 6-7 nights indicated that mortality of both had not been tested in great detail, but in the males and females in the large cages (i.e. after few studies performed a reduced competi- 07 Competitiveness of radio-sterilised An. arabiensis

tiveness at higher doses has been reported periments, and in the case of the adult irradia- (Sharma et al. 1978, El-Gazzar et al. 1983). In tion with the high dose, which cage design (i.e. the present study males irradiated as pupae either the small or the large cage) was a better with the high dose had a lower competitive- indicator for field competitiveness. ness compared to males irradiated with the low Based on the hypothesis that in large dose both in small and large cages but only for cages males would be challenged to a greater the latter was the result statistically significant extent compared to the small laboratory cages, when means were compared with a t-test. For this could have resulted in an increased male the adult stage, males irradiated with the high mortality due to greater activity. However, no dose were less competitive compared to males differences were observed in cumulative- mor treated with the low dose in the small cage but tality data between the two cage types. Al- this result was not significant. In the large cage though dead insects were removed daily from both groups of males performed equally well as the cages, the number of dead insects in the the un-irradiated males. resting tubes was only scored on the 7th day The large cage had a volume approxi- in the large cages because of the disturbance mately forty times larger than the small cage. In the sampling would cause to the resting mos- the small cage, the volume per male mosquito quitoes. Thus it was necessary to compare cu- in the competition experiments was 1.08 * 10-4 mulative mortality data until the 7th day from m3 (i.e. 0.027 m3/ 250 males); thus in the large the large cages with day six of the small cages. cage a male had approximately tenfold more However, we are confident that this had little space (i.e. 1.08 * 10-3 m3/ male, calculated for impact on our findings, because daily mortality 1000 males). We hypothesised that this greater rate was low. volume would challenge the males to a larger The radiation dose used to sterilise in- extent compared to the small cage (e.g. a larger sects in an SIT program should balance induced swarming arena, increased flight activity), so sterility with competitiveness (Parker and Me- any negative effects of irradiation would be- hta 2007); with any residual fertility left in the come more pronounced. This hypothesis only male simply reducing the rate at which the pop- held true for the males irradiated as pupae: ulation is suppressed (Robinson 2002). After a their performance was significantly lower when low, partially-sterilising dose of 70 Gy, substan- tested in the larger cage after receiving a low tial residual fertility in some egg batches was dose compared to the small cage. Males irra- observed (Fig. 7.2), especially so following pupal diated with the high dose performed poorly in irradiation. To verify that these results were not both cages. In contrast, males irradiated in the due to a female being mated by a un-irradiated adult stage with the high dose performed better male before the separation of the sexes had oc- in the large cage compared to the small cage, curred, the virginity of the females was assured while males irradiated with the low dose were by the individual isolation of pupae for one of equally competitive to un-irradiated males in the replicates. However, a similar distribution both cages. It should be noted that the males in the data was observed. In other experiments used in the small and large cage studies did not with two different anopheline species, no egg originate from the same batches of insects and batches with intermediate hatch rate were re- for future studies it would be interesting to com- ported either in experimental or control treat- pare both cage types with males from the same ments (Andreasen and Curtis 2005), suggesting batch. A necessary follow-up experiment would this to be a species-specific phenomenon. be to test the males under simulated field con- Anopheles mosquitoes are largely 119 ditions (Knols et al. 2002), preferably compet- monogamous, and in the wild only in a small ing against wild males for wild females. Such proportions of females multiple mating was experiments would be able to demonstrate the observed (i.e. 1-4% in An. gambiae (Tripet et true value of laboratory competitiveness ex- al. 2001)). Intermediate hatch rates in compe- 07 Competitiveness of radio-sterilised An. arabiensis

tition experiments suggest, when observed in ter six nights (M.Helinski, unpublished data).

Chapter 07 greater numbers than can be expected based Results from a previous study also indicated on the control data, the occurrence of multiple a higher insemination rate in small cages for mating events. For the large cage experiments An. arabiensis (Verhoek and Takken 1994). The with males irradiated as pupae, an increase in slightly different lighting conditions in the room the proportion of intermediate hatch rates was holding the large cages were not responsible observed compared to the small cage results for the lower insemination observed, as similar for both doses. A low rate of multiple insemina- results were obtained when the two cage types tions was observed in competition experiments were switched between rooms for un-irradiat- with irradiated Cx. tarsalis Coquillett males ed mosquitoes (M. Osae, unpublished data & (Zalom et al. 1981), and also in Aedes aegypti M. Helinski, unpublished data). In both rooms, L. (Weidhaas and Schmidt 1963, George 1967). however, some males were observed to display These experiments, however, were performed swarming behaviour at dusk. It is likely that in- in small laboratory cages. In Anopheles mosqui- semination in large cages is lower because the toes it has been suggested that the switchover rearing procedures have selected for mating in to the mated state in the female is triggered by a small cages, and even though some swarming spermatheca filled with sperm (Klowden 2006). behaviour was observed in the large cages, this This switchover becomes more permanent, i.e. might not have been sufficient to reach high the female is less responsive to a subsequent levels of insemination. The observation that mating attempt, when > 24 hrs have passed the doubling of the number of males for com- since the previous mating (Klowden 2006). It petition experiments had no impact on insemi- has been suggested that the occurrence of mul- nation even in those treatments where males tiple mating of anophelines in the field may be were equally competitive suggests that female ascribed to incomplete transfer of sperm or ac- behaviour might also play a role. Alternatively, cessory gland fluids (Yuval and Fritz 1994). For it could indicate that males actively discard fe- Anopheles, no data are available on the amount males of a particular size, as was observed in of sperm or accessory fluids transferred in the An. gambiae s.s. with smaller females (Okanda presence or absence of irradiation, however et al. 2002). studies performed in the Mediterranean fruit fly Ceratitis capitata Wiedemann observed that irradiated males transfer fewer sperm than wild Conclusion males, and females are more likely to remate after mating with an irradiated male than af- The low, partially-sterilising dose of 70 Gy ap- ter mating with a un-irradiated male (Mossin- plied to pupae resulted in a significantly de- son and Yuval 2003). The Mediterranean fruit creased competitiveness in the large cage but fly mating system however is based on female not in the small cage. The fully-sterilising dose choice, while in Anopheles a female appears to of 120 Gy resulted in poor competitiveness after have little influence on selecting a copulation pupal irradiation in both cages. Following irra- partner upon entering in a swarm. It remains to diation in the adult stage, males irradiated with be verified whether incomplete semen transfer the higher dose of 120 Gy were significantly less has ultimately resulted in an increased number competitive than un-irradiated males in the of batches with intermediate hatch rates due to small cage, however in the large cage this dif- multiple mating events in some treatments. ference was not observed. Males irradiated as 120 Overall insemination in the large cag- adults with the low dose were equally competi- es was lower than what was observed for small tive with un-irradiated males in both cages. Al- cages. After seven nights an average of 75% of though our findings indicate that adult irradia- females were inseminated, whilst in the small tion with the low dose yielded competitive males cages > 90% insemination was observed af- in both cage designs it remains questionable 07 Competitiveness of radio-sterilised An. arabiensis

whether adult irradiation on a large scale will be References feasible. However, it should be acknowledged that little has been done to optimise existing Alphey, L., C. B. Beard, P. Billingsley, M. Coetzee, adult irradiation devices (Smittle and Patterson A. Crisanti, C. Curtis, P. Eggleston, C. Godfray, J. 1974, Curtis 1976) in recent decades, and meth- Hemingway, M. Jacobs-Lorena, A. A. James, F. C. ods to inactivate and irradiate large numbers Kafatos, L. G. Mukwaya, M. Paton, J. R. Powell, of adults should be explored. Loss of competi- W. Schneider, T. W. Scott, B. Sina, R. Sinden, S. tiveness can usually be overcome by increasing Sinkins, A. Spielman, Y. Touré, and F. H. Collins the ratio of irradiated males to un-irradiated 2002. Malaria control with genetically manipulated males and this should be pursued in subsequent insect vectors. Science 298:119-121. experiments, specifically for irradiation of the Andreasen, M. H. and C. F. Curtis 2005. Optimal pupal stage. The protocols developed here are life stages for radiation sterilization of Anopheles for readily transferable to test competitiveness sterile insect releases. Med. Vet. Entomol. 19:238- in other genetic control programmes. For in- 244. stance in transgenic insects competitiveness is Bracken, G. K. and C. D. Dondale 1972. Fertility usually measured over successive generations and survival of Achaearanea tepidariorum (Catteruccia et al. 2003, Moreira et al. 2004), (Araneida:Theridiidae) on a diet of chemosterilized and a direct assay as presented here could of- mosquitoes. Can. Entomol. 104:1709-1712. fer a valuable extension to these experiments. Catteruccia, F., H. C. J. Godfray, and A. Crisanti With the large cage design, we have attempted 2003. Impact of genetic manipulation on the to test males under more “realistic” conditions. fitness of Anopheles stephensi mosquitoes. Science Whether this has succeeded remains to be veri- 299:1225-1227. fied by carrying out these types of experiment Curtis, C. F. 1976. Radiation sterilization. Report under near-natural field conditions. It will also on mosquito research. Ross Institute of Tropical be important to assess the predictive value of Hygiene. 01.01.76-31.12.77. laboratory competition experiment especially Dame, D. A., R. E. Lowe, and D. L. Williamson 1981. for a high, fully sterilising dose given to the Assessment of Released Sterile Anopheles albimanus adult male. and Glossina morsitans morsitans, pp. 231-248. In J. B. Kitzmiller, and T. Kanda (eds.), Cytogenetics and genetics of vectors. Elsevier Biomedical, Acknowledgements Amsterdam. We would like to thank M. Benedict and A. Rob- Dame, D. A. and C. H. Schmidt 1962. The inson for discussions regarding the experimen- importance of competitiveness of radiosterilized tal protocol, A. Odulaja for guidance with statis- males in mosquito-control programmes. New Jersey tical analysis, S. Soliban for technical assistance, Mosq. Exterm. Assoc. 49:165-168. and M. Benedict, R. Hood, A. Odulaja, A. Robin- Dame, D. A. 1985. Genetic control by sterilized son, M. Dicke, and two anonymous reviewers mosquitoes, pp. 159-172. In R. Chapman, R. Barr, D. for constructive comments during preparation E. Weidhaas, and M. Laird (eds.), Biological Control of the manuscript. We are grateful to M. Osae of Mosquitoes. Am. Mosq. Cont. Assoc., Bull 6. for sharing his data. BGJK is supported by a VIDI Davis, A. N., J. B. Gahan, D. E. Weidhaas, and C. grant from the Netherlands Organisation for N. Smith 1959. Exploratory studies on gamma Scientific Research (#864.03.004). irradiation for the sterilization and control of Anopheles quadrimaculatus. J. Econ. Entomol. 52:868-870. 121 Dobson, S. L. 2003. Reversing Wolbachia-based population replacement. Trends Parasitol. 19:128- 133. Dobson, S. L., C. W. Fox, and F. M. Jiggins 2002. 07 Competitiveness of radio-sterilised An. arabiensis

The effect of Wolbachia-induced cytoplasmic of Disease Vectors. Proceedings of the joint WHO/

Chapter 07 incompatibility on host population size in natural and TDR, NIAID, IAEA and Frontis Workshop on Bridging manipulated systems. Proc. Biol. Sci. 269:437-445. Laboratory and Field Research for Genetic Control Dyck, A., J. Hendrichs, and A. S. Robinson 2005 of Disease Vectors. Nairobi, Kenya 14-16 July 2004. (eds.). The Sterile Insect Technique: Principles and Frontis, Wageningen. Practice in Area-Wide Integrated Pest Management. Laven, H. 1967. Eradication of Culex pipiens fatigans Springer, Dordrecht. through cytoplasmic incompatibility. Nature 216:383- El-Gazzar, L. M. and D. A. Dame 1983. Effects of 384. combinations of irradiation and chemosterilization Moreira, L. A., J. Wang, F. H. Collins, and M. Jacobs- on mating competitiveness of Culex quinquefasciatus Lorena 2004. Fitness of anopheline mosquitoes Say. J. Econ. Entomol. 76:1331-1334. expressing transgenes that inhibit Plasmodium El-Gazzar, L. M., D. A. Dame, and B. J. Smittle development. Genetics 166:1337-1341. 1983. Fertility and competitiveness of Culex Mossinson, S. and B. Yuval 2003. Regulation of quinquefasciatus males irradiated in nitrogen. J. sexual receptivity of female Mediterranean fruit Econ. Entomol. 76:821-823. flies: old hypotheses revisted and a new synthesis Field, A. 2005. Discovering statistics using SPSS. proposed. J. Insect Physiol. 49:561-567. Sage Publications, London. Okanda, F. M., A. Dao, B. N. Njiru, J. Arija, H. A. Fried, M. 1971. Determination of sterile-insect Akelo, Y. Touré, A. Odulaja, J. C. Beier, J. I. Githure, competitiveness. J. Econ. Entomol. 64:869-872. G. Yan, L. C. Gouagna, B. G. J. Knols, and G. F. George, J. A. 1967. Effect of mating sequence on Killeen 2002. Behavioural determinants of gene flow egg hatch from female Aedes aegypti (L.) mated with in malaria vector populations: Anopheles gambiae irradiated and normal males. Mosq. News 27:82-86. males select large females as mates. Malar. J. 1:10. Haisch, A. 1970. Some observations on decreased Parker, A. G. and K. Mehta 2007. Sterile insect vitality of irradiated Mediterranean fruit fly, pp. 71- technique: a model for dose optimization for 75. Sterile-Male Technique for Control of Fruit Flies. improved sterile insect quality. Fla. Entomol. 90:88- IAEA, Vienna, Austria. 95. Hayes, W. J. 1968. Toxicological aspects of Robinson, A. S. 2002. Mutations and their use in chemosterilants, pp. 315-347. In G. C. LaBrecque insect control. Mutat. Res. 511:113-132. and C. N. Smith (eds.), Principles of insect Sharma, V. P., R. K. Razdan, and M. A. Ansari 1978. chemosterilisation. Appleton Century Crofts, New Anopheles stephensi: effect of gamma-radiation and York. chemosterilants on the fertility and fitness for sterile Helinski, M. E. H., A. G. Parker, and B. G. J. Knols male releases. J. Econ. Entomol. 71:449-452. 2006. Radiation-induced sterility for pupal and adult Smittle, B. J. and R. S. Patterson 1974. Container for stages of the malaria mosquito Anopheles arabiensis. irradiation and mass transport of adult mosquitoes. Malar. J. 5:41. CHAPTER 3 Mosq. News 34:406-408. IAEA. 2004. Gafchromic Dosimetry system for Sokal, R. R. and F. J. Rohlf 1981. Analysis of SIT. http://www-naweb.iaea. org/nafa/ipc/public/ frequencies, pp. 691-778 Biometry. W.H. Freeman Dosimetry_SOP_v10b. pdf. and Company, New York. Klowden, M. J. 2006. Switchover to the mated state Takken, W. and T. W. Scott 2003 (eds.). Ecological by spermathecal activation in female Anopheles aspects for application of genetically modified gambiae mosquitoes. J. Insect Physiol. 52:679-684. mosquitoes. Springer, Dordrecht. Knols, B. G. J., B. N. Njiru, E. M. Mathenge, W. Tantawy, A. O., A. A. Abdel-Malek, and A. M. R. Mukabana, J. C. Beier, and G. F. Killeen 2002. Wakid 1967. Studies on the eradication of Anopheles 122 MalariaSphere: A greenhouse-enclosed simulation pharoensis Theobald by the Sterile-Male technique of a natural Anopheles gambiae (Diptera: Culicidae) using cobalt-60. V. Mating competitiveness of ecosystem in western Kenya. Malar. J. 1:19. radiosterilized males. J. Econ. Entomol. 60:696-699. Knols, B. G. J. and C. Louis 2006 (eds.). Bridging Thomas, D. D., C. A. Donnelly, R. J. Wood, and L. Laboratory and Field Research for Genetic Control S. Alphey 2000. Insect population control using a 07 Competitiveness of radio-sterilised An. arabiensis

dominant, repressible, lethal genetic system. Science 287:2474-2476. Tripet, F., Y. T. Touré, C. E. Taylor, D. E. Norris, G. Dolo, and G. C. Lanzaro 2001. DNA analysis of transferred sperm reveals significant levels of gene flow between molecular forms ofAnopheles gambiae. Mol.Ecol. 10:1725-1732. Verhoek, B. A. and W. Takken 1994. Age effects of insemination rate of Anopheles gambiae s.l. in the laboratory. Entomol. Exp. Appl. 72:167-172. Weidhaas, D. E. and C. H. Schmidt 1963. Mating ability of male mosquitoes, Aedes aegypti (L.), sterilized chemically or by gamma radiation. Mosq. News 23:32-35. Yuval, B. and G. N. Fritz 1994 Multiple mating in female mosquitoes - evidence from a field population of Anopheles freeborni (Diptera: Culicidae). Bull. Entomol. Res. 84:137-139. Zalom, F. G., S. M. Asman, and J. E. Fields 1981. Tests for multiple insemination of females involving irradiated and unirradiated male Culex tarsalis (Diptera: Culicidae). Mosq. News 41:154-156.

123 08 Competitiveness of pupal irradiated males 08 The influence of late-stage pupal irradiation and increased irradiated: un-irradiated male ratio on mating competitiveness of the malaria mosquito Anopheles arabiensis Patton by Michelle EH Helinski and Bart GJ Knols

Competitiveness of released males in genetic control programmes is of critical importance. In this paper, we explored two scenarios to compensate for the loss of competitiveness after pupal stage irradiation in the malaria mosquito Anopheles arabiensis. First, experiments with a higher ratio of irradiated versus un-irradiated insects were performed. Second, male pupae were irradiated just prior to emergence after prolonging their pupal development period. Males were irradiated in the pupal stage with a partially or fully-sterilising dose of 70 or 120 Gy, respectively. Pupae350-500 were Million irradiated cases of aged clinical 20-26 malaria hrs (young), occur annually, or the pupal 60% stageof which was are artificially in sub- prolongedSaharan Africa. by cooling Moreover, and pupae 80% of were all deathsirradiated attributed aged 42-48 to malaria hrs (old). occur Irradiated in this malesregion. competed In numbers, at a 1 ratio million of 3:1:1Africans to un-irradiated die of the disease males foreach mates year, in with a large the cagevast design.majority At of the deaths 3:1 ratio, occurring males amongirradiated children as young below pupae five were years equally of age. competitive Pregnant towomen the un-irradiated are another malesmajor forrisk most group; replicates malaria for can 70 cause Gy while low birthat 120 weight Gy males and were pre- significantlymature delivery less (Rogerson competitive et al. 2007).than un-irradiatedThe impact of malariamales. onThe the irradiation economic of older pupaesituation did of not endemic result incountries a significantly is high (Gallupimproved and competitiveness Sachs 2001), and compared the correlation to the youngerbetween pupae poverty although and malaria slightly clearly better demonstrated results were obtained. (Sachs and Our Malaney findings 2002). indicate that the lossMalaria of competitiveness is a parasitic disease after pupal transmitted irradiation by canfemale be compensated mosquitoes offor the by agenus three-fold Anopheles increase. The of malaria irradiated parasite males, is but a protozoan only for the of partially-sterilising the genus Plasmodium dose.. TheThere use are of four old pupaespecies for of irradiation malaria parasites did not resultsthat can in infecta significant humans improvement under natural in competitiveness.conditions: Plasmodium Nevertheless, falciparum cooling, P. vivax might, P. be ovale a useful and toolP. malariae to facilitate. The handlingfirst two processesspecies cause of large the mostnumbers infections of mosquitoes worldwide in genetic and P. falciparumcontrol programmes. Welsh is by far the 125

Submitted to Bulletin of Entomological Research 08 Competitiveness of pupal irradiated males

Introduction pared to the results observed at the 1:1:1 ratio.

Chapter 08 Despite the fact that adult irradiation enetic control programmes like the Ster- results in more competitive males, the irradia- ile Insect Technique (SIT) that use irradi- tion and transportation of pupae is highly pre- Gation to induce sterility often experience ferred because of their robust nature and ease a reduced competitiveness of released insects of handling. In previous studies pupae were (Dyck et al. 2005, Parker and Methta 2007). This routinely irradiated approximately 24 hrs after reduction in competitiveness results from the pupation (Davis et al. 1959, Tantawy et al. 1967, somatic damage induced during the irradiation Abdel-Malek et al. 1967, 1975, Curtis 1976), and process (LaChance 1967, Proverbs 1967), in ad- in previous work performed in our laboratory dition to the fitness costs of mass-rearing (Cayol An. arabiensis pupae were irradiated aged 20- 2000, Calkins and Parker 2005, Rull et al. 2005). 26 hrs, approximately ten hours before emer- SIT programmes can adopt different strategies gence (Helinski et al. 2006, Helinski and Knols to deal with a reduction in competitiveness. 2008). The irradiation of older pupae, i.e. closer Either the loss of competitiveness is accepted to emergence, might be beneficial in terms and the reduction compensated for by increas- of competitiveness because development of ing the ratio of irradiated males over wild ones these pupae is largely completed. In the present for releases (Dyck et al. 2005), or it is attempted study, pupal development was slowed down by to produce more competitive insects by lower- cooling and specimens were maintained in the ing the irradiation dose, or irradiating the most pupal stage for 42-48 hrs before being irradiat- competitive, normally fully mature, develop- ed. It was hypothesised that irradiation of these mental stage (Parker and Methta 2007). older pupae would improve male competitive- For mosquitoes, competitiveness of ness compared to routine pupal irradiation (i.e. males irradiated as pupae is lower compared to of pupae aged 20-26 hrs). In addition, cooling adult stage irradiation (Curtis 1976, Andreasen under hypoxia of insects prior to the irradia- and Curtis 2005), and previous results from our tion process is routinely used in Tephritid fruit laboratory confirmed these findings (Helinski fly SIT programmes (i.e. 12-20 ˚C) to slow down and Knols 2008). Males irradiated as pupae the metabolic rate and reduce somatic damage had a lower competitiveness compared to un- (FAO/IAEA/USDA 2003). Thus, cooling could irradiated males and this was more pronounced potentially contribute to a reduction of somatic for a fully-sterilising dose of 120 Gy (Helinski damage in Anopheles pupae. and Knols 2008). In contrast, adult irradiation of Anopheles arabiensis Patton with a partially (70 Gy) or fully-sterilising dose resulted in equal Material and Methods competitiveness of irradiated males compared to un-irradiated males when tested at a 1:1:1 Mosquitoes ratio in large cages (Helinski and Knols 2008). The mosquito strain used was the Dongola An increase in the ratio of irradiated males can strain of Anopheles arabiensis. For a detailed de- compensate for the loss in competitiveness and scription of rearing procedures see Helinski et this was demonstrated in An. pharoensis Theo- al. (2006). Until experiments took place, insects bald after pupal stage irradiation (Tantawy et were maintained in the insectary at 27 ± 0.5 °C al. 1967). In this study, the competitiveness and RH 82 ± 2% in normal rearing cages (30 x of males irradiated as pupae was determined 30 x 30 cm). Competition experiments were 126 when the number of irradiated males was in- performed in large pyramid shaped cages and creased three-fold (i.e. 3:1:1) compared to the measured 1.8 (l) x 1.2 (w) x 1.0 (h) m. (MoTec Pop- un-irradiated males. It was hypothesised that Up, Brettschneider, Heimstetten, Germany, see at this higher ratio the proportion of females Fig. 7.1); for a detailed description see Helinski mated to irradiated males would increase com- and Knols (2008). Large cage experiments took 08 Competitiveness of pupal irradiated males

place in the room used for larval rearing where pupae were irra- Figure 8.1. Images of a young (i.e. 20-26 hrs) and old (i.e. 42-48 conditions were slightly different (24 ± 0.5 °C diated aged 20-26 hrs) pupa used in the irradiation and RH 50 ± 3%). Previous work had shown that hrs after pupation experiments. Location of the eye survival and mating behaviour under these con- as routinely per- and wing are indicated. In the old ditions were similar to that observed in the in- formed (Helinski pupa, development was near to completion and a fully developed sectary (M. Helinski, unpubl. results). The light et al. 2006, Helin- adult (with scaled wings) was regime in both rooms was L10:D12 plus a one ski and Knols present inside the pupal skin; in hour computer-simulated dusk and dawn pe- 2008); old pupae the young pupa, development riod. All adults were continuously supplied with were irradiated was still in progress. 6% sucrose solution [w/v]. aged 42-48 hrs. Old pupae were Irradiation obtained by the artificial prolonging (i.e. - nor Insects were exposed to gamma rays gener- mally pupae emerge approximately 30-36 hrs ated by a cobalt-60 source with a dose rate of after pupation) of the pupal phase by cooling. ca. 12 Gy/min. Males were irradiated in the pu- On the morning after collection the pupae were pal stage, and irradiation procedures and han- placed at 9 am in an incubator (i.e. aged 18-24 dling methods were similar to those described hrs), and were cooled for 24 hrs until irradiation by Helinski et al. (2006) and Helinski and Knols the next day at 9 am. The light regime in the (2008). Irradiation doses were 70 Gy for a par- incubator was synchronised to light conditions tially-sterilising dose and 120 Gy for a fully- in the insectary and was L12:D12. Temperature sterilising dose. Dosimetry was used for each in the incubator was aimed at 15 °C and during irradiation to verify the doses received by the darkness, 15.7 ± 0.5 °C was observed. However, batch (Helinski et al. 2006). during the light phase temperature was 18.7 ± 127 0.5 °C as a result of the heat produced by the Collection of experimental material tube lights. After irradiation pupae emerged in Pupae were collected between 9 am and 3 pm a normal rearing cage in the insectary. Young and were not sexed prior to irradiation. Young pupae emerged in the evening following ir- 08 Competitiveness of pupal irradiated males

radiation and sexes were separated the next succession rather than simultaneously and the

Chapter 08 day. Old pupae emerged within six hours after order of experiments (i.e. dose tested first) was irradiation and sexes were separated the same alternated between replicates. The males intro- day. duced first were aged 3 days, while males intro- The effect of pupal cooling on adult duced second were aged 5 days, with the ex- emergence and longevity was tested for the fol- ception of replicate one for young pupae where lowing treatments; 1) control (i.e. no cooling), 2) all males were 2 days older. cooling (i.e. no irradiation), 3) cooling and irra- diation with 70 Gy, or 4) cooling and irradiation Blood feeding and egg laying with 120 Gy. For each treatment 100 pupae were Females were blood fed on membranes filled used and treatments were replicated thrice. All with heparinised human or defibrinated cow pupae were derived from the same batch. Adult blood. If feeding was low, the unfed females emergence was scored, and longevity of the were blood fed on the forearm of a human vol- emerged adults was determined routinely (i.e. unteer. Females were fed prior to introduction at 24-48 hr intervals) in normal rearing cages by in the cage or afterwards, depending on the the removal of dead insects until the majority time of introduction. Females introduced with of adults had died. In a separate experiment, a the first batch of males were fed afterwards, number of old pupae were fixed in 96 % etha- females introduced with the second batch be- nol and dissected to determine pupal develop- fore. Three to four days after blood feeding, fed ment. females were individually placed in tubes (Ø2.5 x 9 cm) for egg laying. For every egg batch the Experimental setup hatch rate was determined after substantial Competition experiments were performed with time was given for hatching (i.e. 5-7 days). males irradiated as young pupae. Subsequently To determine if an egg batch was fa- the same treatments were tested with males ir- thered by an irradiated or an un-irradiated radiated as old pupae. Irradiated males were in- male, the level of inherent (control) sterility in troduced in the large cage together with un-ir- the colony, and the sterility induced by males radiated males and females at the ratio of 3:1:1. irradiated as pupae with 70 and 120 Gy when 75-100 insects were used for the lower ratio, mated against un-irradiated females, were de- and mating was allowed to take place for two termined (i.e. referred to as control data). Data nights. Mortality of males and females after the for males irradiated as young pupae had already experiment was scored. Age of the males was been collected in a previous study (Helinski and 3-7 days at the start of the mating experiment. Knols 2008), and data for old pupae were col- The un-irradiated males in all experiments were lected. Males irradiated as old pupae (i.e. with derived from the same batch of pupae as the 70 or 120 Gy) were mated with virgin females irradiated males. For the experiments with old in small cages. Thereafter, females were blood- pupae, the un-irradiated males used in the com- fed and individually placed in tubes for egg-lay- petition experiments were not cooled but left ing as above, and hatch rates determined. to emerge as normal. Females used as mates were of similar age as the males and separated Statistical analysis to sex < 18 hrs after emergence to ensure vir- Prior to analyses, data were checked for nor- ginity. Females were introduced in the com- mality. Hatch data were analysed according to petition cages on the same day as the males procedures described previously (Helinski and 128 were introduced. Irradiation doses (i.e. 70 or Knols 2008). Briefly, logistic regression analy- 120 Gy) were tested pairwise, and for each pair ses were used to classify the hatch rates as ei- four replicates with different batches of males ther deriving from irradiated or un-irradiated were carried out. Because only one large cage males (i.e. using the control data). This distribu- was available, experiments were performed in tion was analysed using a replicated G-test for 08 Competitiveness of pupal irradiated males

Table 8.1. Effect of different pupal treatments (i.e. goodness of fit. If there was no heterogeneity control (no cooling) and cooling for 24 hrs with or without between replicates of a treatment, data were irradiation with 70 or 120 Gy) on mean (± s.e.m.) survival pooled. General Linear Models (GLMs) were times of An. arabiensis males. Mean survival times were estimated from Kaplan-Meier survival analysis. For used to compare the proportion of un-irradi- each replicate approximately 100 insects were used and ated batches between treatments, and means on average 54 ± 2 % were males. Replicates were not were separated using Tukey’s Honestly Signifi- statistically similar (log-rank tests, data not shown) and cantly Different (HSD) and individual t-tests. To data could not be pooled. facilitate comparison with similar studies in the Treatment Replicate Mean male survival time (days ± s.e.m.) literature the competitiveness index C (C = # ir- control radiated batches/ # un-irradiated batches) was 1 23 ± 1.2 determined (Haisch 1970, Fried 1971). Cumula- 2 18 ± 0.8 tive mortality data from the competition exper- 3 27 ± 1.1 iments were arcsine square root transformed cooling and analysed using GLM’s. Mortality data from 1 18 ± 0.6 the cooling experiment were analysed using 2 22 ± 1.0 Kaplan-Meier survival analysis with log-rank tests. All two-sided tests were performed using 3 27 ± 1.1 the SPSS software version 14 (SPSS Inc., Chi- cooling, 1 12 ± 0.8 70 Gy cago, USA). 2 24 ± 1.0

3 20 ± 1.1

Results cooling, 1 17 ± 0.9 Mean values throughout the text are reported 120 Gy ± s.e.m. Dosimetry confirmed that all doses de- 2 15 ± 0.8 livered were within a 5% error range. 3 12 ± 0.5

Cooling The older pupae (i.e. 42-48 hrs) used in the ir- radiation experiments were obtained by cool- mean survival times were similar. Survival times ing. After placing pupae for 24 hrs at the lower of males cooled as pupae and irradiated with 70 temperature, upon dissection a fully developed Gy were similar compared to un-cooled males adult was present inside the pupal skin; in con- with the exception of replicate one where sur- trast to the young pupae aged 20-26 hrs (Fig. vival was low (Table 8.1). For the cooled pupae 8.1). After irradiation of the old pupae, the large irradiated with 120 Gy, a reduction in survival majority of adults emerged within the first six was observed for all replicates compared to un- hours. The cooling of pupae for 24 hrs had no cooled males (Table 8.1). impact on pupal emergence compared to con- trol pupae without the cooling treatment, nor Sterility was there an effect on emergence of the com- The competitiveness of males irradiated as bined treatment of cooling and irradiation with pupae at 20-26 hrs with 70 Gy varied consid- either dose (F3, 8= 0.35, p> 0.05). On average, erably between replicates and data could not emergence was 96 ± 1%. The analysis of longev- be pooled (Table 8.2). Individual G-test results ity was complicated by a rather large variation showed that only in the second replicate males within treatments, and for all treatments repli- were significantly less competitive than un- 129 cates could not be grouped for statistical anal- irradiated males; in the other replicates similar ysis. However, in Table 8.1 it can be seen that competitiveness was observed. For males ir- cooling of pupae appeared to have no effect on radiated with 120 Gy, replicates were statisti- male longevity compared to un-cooled males as cally similar for each treatment, and data were 08 Competitiveness of pupal irradiated males

Table 8.2. The number of batches fathered by un-irradiated or irradiated males from the competition experiments. Males were irradiated with 70 or 120 Gy as pupae either aged 20-26 hrs, or 42-48 hrs. Irradiated males competed with

Chapter 08 un-irradiated males for females on a 3:1:1 ratio. Individual, replicated and pooled G-tests results are given, as well as the competitiveness (C) value (mean ± s.e.m) based on the four replicates for each treatment when data could be pooled (C was not corrected for the 3:1 ratio to compare with 1:1 ratio results (Chapter 7)). The mean proportions of egg batches with intermediate hatch rate (i.e. between 20 - 60%) are presented. Asterisks indicate significant differences between batches of un-irradiated and irradiated males, * p< 0.05 and ** p< 0.01. n/a: not applicable (i.e. replicates not statistically similar).

Pupal age at Dose Batches fathered by G-tests C Batches with irradiation (Gy) males intermediate un-irrad. irrad. individual pooled hatch rate (%) 20-26 hrs 70 15 9 1.52 n/a 0.60 39.4 ± 5.4

34 10 13.83** 0.29

23 22 0.02 0.96

12 22 2.99 1.83

120 27 13 5.01* 53.97** 0.26 ± 0.08 27.9 ± 3.0

17 4 8.66**

29 3 24.45**

39 8 22.27**

42-48 hrs 70 25 34 1.38 n/a 1.36 39.0 ± 10.4

12 12 0 1.00

16 27 2.85 1.69

38 20 5.68* 0.53

120 44 10 23.11** 23.41** 0.52 ± 0.10 18.0 ± 3.8

22 13 2.34

24 15 2.10

38 24 3.19

pooled. Males were significantly less competi- data obtained from young pupae (Helinski et al. tive than un-irradiated males (Table 8.2), and 2008). on average 81 ± 5% of all egg batches were fa- The distribution of egg batches fa- thered by un-irradiated males (Fig. 8.2). thered by irradiated and un-irradiated males The control data of older pupae (i.e. for the older pupae as shown in Table 8.2. For 42-48 hrs) gathered to construct regression 70 Gy, again replicates were not statistically models resulted in some interesting findings. similar and data could not be pooled. Individual Females mated to males irradiated as old pu- results showed that males irradiated with 70 pae with 120 Gy produced a small number (i.e. Gy fathered an equal number of egg batches 4%) of egg batches with a much higher hatch compared to un-irradiated males, with the ex- rate (i.e. 22-65%) than could be expected based ception of the last replicate where un-irradiated 130 on the data obtained from young pupae (Helin- males accounted for the majority of matings. ski et al. 2008). These data were removed from For 120 Gy, the individual G-test results showed the logistic regression analysis to prevent the no significant differences in competitiveness threshold from becoming too liberal. For 70 Gy, of irradiated males compared to un-irradiated the control data were similar compared to the males in all replicates but the first. However, 08 Competitiveness of pupal irradiated males

Figure 8.2. Proportion of egg batches (mean ± s.e.m.) fa- thered by un-irradiated males for young (i.e. 20-26 hrs, open bars) and old (i.e. 42-48 hrs, grey bars) pupae irradiation. Bars with the same letter are not statistically different from each other, p< 0.05 (Tukey HSD).

when data were pooled, irradiated males over- ments with 70 Gy compared to 120 Gy, but data all performed worse than un-irradiated males were just not statistically different from each

(Table 8.2), and 67 ± 5% of all egg batches were other (F3, 12= 3.58, p= 0.051). fathered by un-irradiated males (Fig. 8.2). When comparing the performance of mosquitoes that had been irradiated as young Mortality and egg laying or old pupae, irradiation of old pupae resulted Cumulative mortality of males and females af- in a greater proportion of females mated by ir- ter two nights of competition was similar for radiated males compared to young pupae for all experiments (males: (F3, 12= 1.42, p> 0.05); both doses (Fig. 8.2), and this can also be ob- females: (F3, 12= 0.48, p> 0.05)), and was 7 ± 1% served from the competitiveness indexes (i.e. and 4 ± 1% for males and females, respectively. a value of 1 indicates equal competitiveness The proportion of females laying eggs in indi- compared to un-irradiated males; Table 8.2). vidual tubes was similar for all experiments (F3,

However, when replicates were combined, the 12= 0.63, p> 0.05); and on average 55 ± 4% of the proportion of batches fathered by un-irradiated females laid eggs. males in the young compared to the old pupae experiments were not statistically significantly different for both doses (70 Gy: t(6)= 0.72, p> 0.05; 120 Gy: t(6)= 2.00, p> 0.05; Fig. 8.2). When Discussion all treatments were combined, statistically sig- nificant differences in competitiveness were At a three to one ratio, equal competitiveness of observed between young pupae irradiated with males irradiated as pupae aged 20-26 hrs with

120 Gy and old pupae irradiated with 70 Gy (F3, the partially-sterilising dose of 70 Gy was ob-

12= 4.59, p< 0.05; Fig. 8.2). served in the majority of replicates. The propor- The proportion of egg batches with an tion of females mated by un-irradiated males intermediate hatch rate, defined as being be- was on average 57 ± 9%. At the fully-sterilising 131 tween 20 – 60% (see Helinski and Knols 2008), dose of 120 Gy, competitiveness of irradiated is shown in Table 8.2 for all experiments. The males was low and 81 ± 5% of females were proportion of egg batches with intermediate mated by un-irradiated males. Previous compe- hatch rate was higher for competition experi- tition experiments on a 1:1 ratio in a large cage 08 Competitiveness of pupal irradiated males

observed 73 ± 2% and 84 ± 3% of all matings to et al. 2006). Even though the analysis of longev-

Chapter 08 be fathered by un-irradiated males when com- ity was complicated by large variation within peting with males irradiated in the pupal stage replicates, mean survival data indicated that at 70 and 120 Gy, respectively (Helinski and cooling had no effect on male longevity. Cool- Knols 2008). It thus appears that a three-fold ing in combination with irradiation resulted in increase of males irradiated with 70 Gy resulted a reduced longevity for 120 Gy, while for 70 Gy in an increase in the number of egg batches fa- this was only observed for one of the replicates. thered by irradiated males compared to the 1:1 In previous work, survival of males was similar ratio experiments. For 120 Gy, however, this dif- to the control for 70 Gy; for 120 Gy results varied ference was not observed. In males irradiated and a similar or reduced longevity compared to as pupae with 120 Gy, equal competitiveness un-irradiated males was observed (Helinski et was only observed after a tenfold increase of al. 2006). irradiated males in An. pharoensis (Tantawy et The competitiveness of mosquitoes al. 1967), while in An. quadrimaculatus Say for emerging from pupae that were irradiated with the same ratio competitiveness remained low 70 Gy at an older age (i.e. 42-48 hrs) and mated (Davis et al. 1959); in both studies competitive- at a 3:1 ratio to un-irradiated males was slightly ness was assessed in small cages. The propor- higher, but not statistically different from the tion of individual females that oviposited in competitiveness observed after the irradiation these experiments was 55 ± 4%, and compara- of younger pupae, and 49 ± 6% of the females ble to results observed for the 1:1 ratio experi- were mated by un-irradiated males. At 120 Gy, ments (i.e. 59 ± 4%; Helinski and Knols 2008). the overall competitiveness of the older pupae Some caution in interpreting results from the was lower than un-irradiated males and 67 ± different ratio experiments is used since condi- 5% of the egg batches were fathered by un- tions where not similar. In the 1:1 ratio experi- irradiated males. However, in the majority of ments, a larger number of insects (i.e. 500 per replicates competitiveness of irradiated males ratio) combined with a longer mating period of was not statistically different from un-irradi- 7 nights were used (Helinski and Knols 2008). ated males. Compared to the young pupae, For the experiments described here, conditions competitiveness was higher but results were were adjusted as the number of insects needed not statistically significantly different. In con- at the higher ratio would become too large, and trast, the irradiation of adults performed in a the mating period was reduced as for each rep- previous study observed equal competitiveness licate experiments had to be performed in suc- compared to un-irradiated males at a 1:1 ratio cession rather than simultaneously. However, it in the large cage for 70 and 120 Gy (Helinski and can be argued that a mating period of only two Knols 2008). The hypothesis that in old pupae, nights rather than seven nights is more repre- in which a fully developed adult was present sentative to the natural situation, where re- and as a result of cooling cellular activity was leased Anopheles males are likely to participate low, somatic damage would be reduced and an in few swarming events only. improved competitiveness observed, did not The cooling of pupae for 24 hrs suc- hold to the extent expected. cessfully delayed adult emergence and a fully Cooling reduces metabolic rate and developed adult was present for irradiation. this reduction is thought to result in less somatic Cooling had no impact on adult emergence, damage (Kaspi and Parrella 2003), however lit- and in combination with irradiation emergence tle evidence of the beneficial effects of cooling 132 remained high and similar to the control treat- alone on competitiveness can be found in the ment. This is in line with results from previous literature (Rananavare et al. 1991). However, work where in the absence of cooling, no im- cooling does not seem to reduce competitive- pact of irradiation on emergence was observed ness either in the tsetse fly Glossina pallidipes even when the dose applied was high (Helinski Austen (Mutika et al. 2001) or the Mediterrane- 08 Competitiveness of pupal irradiated males

an fruit fly Ceratitis capitata Wiedemann (Nes- weigh the additional handling costs. However, tel et al. 2007), and we conclude that cooling cooling can be used as a tool in operational SIT also has little effect on competitiveness of An. campaigns to facilitate pupal handling. arabiensis. The observation that cooling can be used to delay pupal emergence in An. arabiensis Acknowledgements without any impact on adult emergence or lon- The authors wish to thank S. Soliban for techni- gevity is useful for operational SIT programmes cal assistance, M. Dicke for constructive com- where pupae need to be transported (e.g. when ments on the manuscript, and the International the field site is not in the vicinity of the rearing Atomic Energy Agency for providing funding. facility) or stacked for further handling (Mutika BGJK is supported by a VIDI grant from the et al. 2001). Netherlands Organisation for Scientific Re- Surprisingly, a few egg batches with search (#864.03.004). much higher hatch rate than expected were ob- served for females mated to males irradiated as old pupae with 120 Gy. For 70 Gy, this phenom- enon was observed previously (Helinski and References Knols 2008), and again in the present study few egg batches with much higher than average Abdel-Malek, A. A., T. O. Tantawy, and A. M. hatch rate were observed. The biological rel- Wakid 1967. Studies on the eradication of Anopheles evance of these data is not clear but could indi- pharoensis Theobald by the Sterile-Male technique cate some radio-resistance in specific pupae at using cobalt-60. III. Determination of the Sterile the time of radiation. This would be interesting Dose and its biological effects on different characters to study in future experiments, but is compli- related to “Fitness” components. J. Econ. Entomol. cated by the fact that single-pair matings in An. 60:20-23. arabiensis are difficult to perform (M. Benedict, Abdel-Malek, A. A., A. M. Wakid, A. O. Tantawy, unpublished data). and L. M. El-Gazzar 1975. Studies on factors The proportion of egg batches with in- influencing the induction of sterility in Anopheles termediate hatch rates, a possible indicator for pharoensis Theobald by gamma radiation, pp. 161- multiple mating events as previously discussed 174. in M. Nehme, and A. Hassan (eds.), The use of (Helinski and Knols 2008) was higher for the isotopes and pesticides in pest control. Proceedings competition experiments with 70 Gy compared of a symposium held in Beirut-Lebanon. to 120 Gy, and in line with previous results ob- Andreasen, M. H., and C. F. Curtis 2005. Optimal served at the 1:1 ratio experiments (Helinski life stages for radiation sterilization of Anopheles for and Knols 2008). sterile insect releases. Med. Vet. Entomol. 19:238- 244. Calkins, C. O., and A. G. Parker 2005. Sterile insect Conclusion quality, pp. 269-296. In V. A. Dyck, J. Hendrichs, and A. S. Robinson (eds.), Sterile Insect Technique. The reduction in competitiveness of males irra- Principles and Practice in Area-Wide Integrated Pest diated as pupae can be compensated for by a Management. Dordrecht, Springer. three-fold increase of irradiated insects at the Cayol, J. P. 2000. Changes in sexual behavior and life partially-sterilising dose of 70 Gy, but not at the history traits of tephritid species caused by mass- fully-sterilising dose of 120 Gy. Irradiation of rearing processes, pp. 843-860. In M. Aluja, and A. L. older pupae resulted in a marginal, but non-sig- Norrbom (eds.), Fruit Flies (Tephritidae): Phylogeny 133 nificant, improvement in competitiveness com- and Evolution of Behavior. Boca Raton, CRC Press pared to the irradiation of young pupae for each LLC. dose. The beneficial effects of the irradiation of Curtis, C. F. 1976. Radiation sterilization. Report old pupae are small and probably do not out- on mosquito research. Ross Institute of Tropical 08 Competitiveness of pupal irradiated males

Hygiene. 01.01.76-31.12.77. Parker, A. G., and K. Mehta 2007. Sterile insect

Chapter 08 Davis, A. N., J. B. Gahan, D. E. Weidhaas, and C. technique: a model for dose optimization for N. Smith 1959. Exploratory studies on gamma improved sterile insect quality. Fla. Entomol 90:88- irradiation for the sterilization and control of 95. Anopheles quadrimaculatus. J. Econ. Entomol. Proverbs, M.D. 1969. Induced sterilization and 52:868-870. control in insects. Annu. Rev. Entomol. 14, 81-102. Dyck, A., J. Hendrichs, and A. S. Robinson 2005 Rananavare, H. D., M. R. Harwalkar, and G. W. (eds.). The Sterile Insect Technique: Principles and Rahalkar 1991. Influence of modifying factors on Practice in Area-Wide Integrated Pest Management. induction of sterility and mating ability of potato Springer, Dordrecht. tuberworm, Phthorimaea operculella (Zeller). J. Nucl. FAO/IAEA/USDA (2003) Manual for Product Quality Agricul. Biol. 20:199-205. Control and Shipping Procedures for Sterile Mass- Rull, J., O. Brunel, and M. E. Mendez 2005. Mass Reared Tephritid Fruit Flies Version 5.0, International rearing history negatively affects mating success Atomic Energy Agency, Vienna, Austria, 85p. of male (Diptera: Tephritidae) Fried, M. 1971. Determination of sterile-insect reared for sterile insect technique programs. J. Econ. competitiveness. J. Econ. Entomol. 64:869-872. Entomol. 98:1510-1516. Haisch, A. 1970. Some observations on decreased Tantawy, A. O., A. A. Abdel-Malek, and A. M. vitality of irradiated Mediterranean fruit fly, pp. 71- Wakid 1967. Studies on the eradication of Anopheles 75. Sterile-Male Technique for Control of Fruit Flies. pharoensis Theobald by the Sterile-Male technique IAEA, Vienna, Austria. using cobalt-60. V. Mating competitiveness of Helinski, M. E. H., A. G. Parker, and B. G. J. Knols radiosterilized males. J. Econ. Entomol. 60:696-699. 2006. Radiation-induced sterility for pupal and adult stages of the malaria mosquito Anopheles arabiensis. Malar. J. 5:41. CHAPTER 3. Helinski, M. E. H., and B. G. J. Knols 2008. Mating competitiveness of male Anopheles arabiensis mosquitoes irradiated with a partially- or fully- sterilising dose in small and large laboratory cages. J. Med. Entomol. in press. CHAPTER 7. Kaspi, R. and M. P. Parrella 2003. The feasibility of using the sterile insect technique against Liriomyza trifolii (Diptera: Agromyzidae) infesting greenhouse chrysanthemums. Ann. Appl. Biol. 143:25-34 LaChance, L. E. 1967. The induction of dominant lethal mutations in insects by ionizing radiation and chemicals-as related to the sterile male technique of insect control, pp. 617-650. In J. W. Wright, and R. Pal (eds.), Genetics of Insect Vectors of Disease. Amsterdam, Elsevier. Mutika, G. N., E. Opiyo, and A. S. Robinson 2001. Assessing mating performance of male Glossina pallidipes (Diptera: Glossinidae) using a walk-in field cage. Bull. Entomol. Res. 91:281-288. 134 Nestel, D., E. Nemny-Lavy, S. M. Islam, V. Wornoayporn, and C. Caceres 2007. Effects of Pre- irradiation Conditioning of Medfly Pupae (Diptera: Tephritidae): Hypoxia and Quality of Sterile Males. Fla. Entomol. 90:80-87. 08 Competitiveness of pupal irradiated males

135 09 Multiple mating in An. arabiensis 09 A stable isotope dual-labelling approach to detect multiple insemination in un-irradiated and irradiated Anopheles arabiensis mosquitoes by Michelle EH Helinski, Rebecca C Hood and Bart GJ Knols

In the context of a Sterile Insect Technique programme, the occurrence of multiple insemination in the malaria mosquito Anopheles arabiensis Patton was studied using a novel labelling system with the stable isotopes 15N and 13C. The incidence of multiple insemination in the absence of radiation, and when males were irradiated in the pupal stage and competed against un-irradiated males were assessed. Males used in the experiments were labelled with either 15N or 13C and the label was applied to the larval rearing water. Males with either label and virgin females were caged at a350-500 1:1:1 ratio. Million Males cases used of inclinical the irradiation malaria occur treatments annually, were 60% irradiated of which in are the in pupal sub- stageSaharan with Africa. a partially Moreover, or fully-sterilising 80% of all deaths dose ofattributed 70 or 120 to Gy, malaria respectively. occur in After this mating,region. In females numbers, were 1 million dissected Africans and inseminateddie of the disease spermathecae each year, analysed with the using vast massmajority spectrometry. of deaths occurringThe data indicateamong childrenthat about below 25% fiveof inseminated years of age. females Pregnant had beenwomen inseminated are another multiply. major risk The group; presence malaria of irradiated can cause males low in birth the experimentsweight and pre did- notmature affect delivery the incidence (Rogerson of etmultiple al. 2007). insemination. The impact In of line malaria with onprevious the economic research, irradiatedsituation of males endemic were countries generally is high less (Gallup competitive and Sachs than 2001), un-irradiated and the males.correlation The implicationsbetween poverty of these and findingsmalaria clearly for the demonstrated Sterile Insect (SachsTechnique and Malaneyare discussed, 2002). and further experimentsMalaria is a recommended.parasitic disease The transmitted dual-labelling by femalesystem mosquitoesused to determine of the paternitygenus Anopheles gave good. The results malaria for parasite13C, however, is a protozoan for 15N it is ofrecommended the genus Plasmodium to increase. Therethe amount are four of labelspecies in future of malaria studies. parasites that can infect humans under natural conditions: Plasmodium falciparum, P. vivax, P. ovale and P. malariae. The first two species cause the most infections worldwide and P. falciparum Welsh is by far the 137

Published in Parasites & Vectors 2008, 1: 9 09 Multiple mating in An. arabiensis

Introduction advantageous, as it would result in the re-mating

Chapter 09 of immigrating females from neighbouring untreated sites. In SIT programmes, irradiation olyandry, insemination of a female is routinely used to induce sterility. In previous by more than one male, is a common work it was determined that An. arabiensis Pphenomenon in many insects (Arnqvist Patton males irradiated as pupae produced and Nilsson 2000), but malaria mosquitoes fewer and smaller sperm compared to un- (Anopheles spp.) are generally believed to irradiated males (Helinski and Knols 2008a). be largely monandrous (i.e. insemination Even though it is not known if small sperm of a female by a single male) under natural contribute to fertilisation in Anopheles, the conditions (Clements 1999). However, some observation that females select for larger sperm level of polyandry (or multiple insemination) in their spermathecae (Klowden and Chambers has been observed both in laboratory and field 2004, Helinski and Knols 2008a), coupled to studies. The occurrence of multiple insemination results from studies performed in Drosophila has been assessed in field populations using where small sperm are not used for fertilisation molecular markers (Yuval and Fritz 1994, Tripet (Bressac and Haushteck-Jungen 1996), might et al. 2003). Multiple insemination occurred indicate that they are of little importance for in low frequencies (i.e. 2.5%) in Anopheles fertilisation. The implications of these findings gambiae Giles (Tripet et al. 2003) and similar on mating behaviour of anophelines is not results were observed in An. freeborni Aitken known. However, in the Mediterranean fruit fly (Yuval and Fritz 1994). In laboratory cages much Ceratitis capitata Wiedemann it was observed higher frequencies of multiple insemination can that sterile males transferred fewer sperm be observed (e.g. eye-colour mutants; Mason than wild type males and females mated to 1967, Mahmood and Reisen 1980, Gomulski irradiated males were more likely to re-mate 1988, Beard et al. 1995, Klowden 2006). In An. compared to females mated to un-irradiated gambiae s.s. Gomulski (1988) reported 12% males (Mossinson and Yuval 2003). The aim multiple insemination after a mating period of of the present study was to determine the 6 days using mosquitoes at a ratio of 1:1:2 (i.e. occurrence of multiple insemination events male: male: female), while 11-48% of multiple in a laboratory colony of An. arabiensis, and insemination was observed in another study to assess the influence of irradiation at the (i.e. depending on the timing of remating, after pupal stage with a partially or fully-sterilising or within 24 hrs of the first insemination, using dose on the occurrence of these events during a 1:1:1 ratio; Klowden 2006). In An. culicifacies competition experiments. Giles, multiple insemination was assessed for Multiple insemination events were various densities of mosquitoes (i.e. ratio of determined using a dual-labelling system of 1:1:2) and a positive correlation between insect stable isotopes. Stable isotopes are naturally density and multiple insemination was observed occurring in the environment, are not (Mahmood and Reisen 1980). radioactive, and react in general chemically Genetic control programmes including identical to the more common isotope. the Sterile Insect Technique (SIT) rely on These attributes make them effective non- the ability of released males to successfully invasive markers in biological systems (Hood- inseminate wild females and introduce their Nowotny and Knols 2007). In previous work genes into the next generation. In principal, it was established that the stable isotopes 13C 138 polyandry is not considered to be a drawback and 15N could be used individually as semen in genetic control programmes as long as the labels (Helinski et al. 2007, 2008). The labels sperm of released males is able to compete with were applied in the larval rearing water, and sperm from wild males after insemination (Curtis labelled males transferred sufficient label to 1985). Under these conditions polygamy is permit the positive identification of the label 09 Multiple mating in An. arabiensis

in spermathecae after mating (Helinski et al. ratio. For each ratio 34-40 insects were used. 2007, 2008). Because both isotopes can be The experiments were performed in standard simultaneously analysed in a sample using an rearing cages (30 x 30 x 30 cm) and the mos- elemental analyser linked to an isotope ratio quitoes were left together for 4 nights. Age of mass spectrometer, a device that separates the mosquitoes at the start of the experiment ions of the element of interest on the basis of was between 3-5 days. The two groups of males their differing mass/charge ratio (m/z) (see were labelled with either 15N or 13C. Males were Hood-Nowotny and Knols (2007) for details), in irradiated in the pupal stage following pro- this study the use of both labels to determine cedures as described in Helinski et al. (2006), multiple insemination was studied. with a partially or fully sterilising dose of 70 or 120 Gy, respectively. To determine the degree of multiple insemination in the absence of ra- Materials and Methods diation, un-irradiated males labelled with either label competed for females (i.e. control treat- Mosquitoes and labelling ment). For each dose, 2 sets of experiments The mosquito strain used in all experiments were performed and each set consisted of three was the Dongola strain of An. arabiensis Patton. treatments: Insects were reared according to the methods 1) Control: 15N un-irradiated males : 13C un-irra- described in Helinski et al (2007). Briefly, five diated males : virgin females, hundred L1 stage larvae were counted and 2) 15N irradiated males : 13C un-irradiated males placed in a tray (30 x 40 cm) in 1.5 L of deion- : virgin females, ised water, and the label was added to the lar- 3) 15N un-irradiated males : 13C irradiated males val water on the same day as the L1 larvae were : virgin females. Treatments 2 and 3 were the introduced. Larvae were fed a fixed diet of fish same except that the isotope label was alter- food (approx. 0.25 mg/larva/day, AquariCare nated between groups of males. Koi Floating Blend, USA); adult mosquitoes Males used in each set of experiments origi- were continuously supplied with a 6% sucrose nated from the same batch of eggs, and were solution [w/v]. Mosquitoes were labelled in the reared under standardised larval conditions (i.e. larval stage with 15N-glycine or 13C-glucose fol- density, food, and water depth) in their respec- lowing procedures described in Helinski et al. tive labelled trays. For each set of experiments, (2007, 2008). For both isotopes, the target en- it was attempted to use males of exactly the richment in the mosquito was 5 atom%. For 15N same age, however this was not always feasible this meant that 5 atom% of 15N was added (i.e. and in experiments three and four males differ- 5% of all the nitrogen in the diet was 15N), while ing in age by 1 day were used. for 13C 20 atom% was added due to losses re- After the mating period, females were sulting from turnover and respiration (Helinski dissected for assessment of insemination and et al. 2007, 2008). for each treatment inseminated spermathecae (N= 17-20) were prepared for sample analysis Experimental setup in the mass spectrometer (Helinski et al. 2007, Males from the labelled trays were separated 2008). For each set of experiments, spermath- from females and maintained as virgins before ecae from virgin females (N= 5) were included. mating experiments took place. All females Samples were spiked to attain sufficient N or C used as mates came from unlabelled trays and to be above the detection limit of the isotope females were sexed < 18 hrs after emergence to ratio mass spectrometer (Helinski et al. 2007, 139 assure virginity. 2008), and the spike used for these samples Competition experiments were per- consisted of 20 μl of a solution containing 20 μg formed where irradiated and un-irradiated N and 25 μg C. groups of males competed for females in a 1:1:1 09 Multiple mating in An. arabiensis

Data interpretation heterogeneity between replicates of a treat-

Chapter 09 Sample analysis and interpretation were similar ment, data were pooled. General Linear Models to Helinski et al. (2007, 2008), and the spiked (i.e. (GLMs) were used to compare the proportion of raw) delta values of the spiked samples were multiple insemination events and insemination used for data analysis. The δ values reported are rate between treatments (i.e. control and irra- referenced to the international standards for ni- diation with 70 or 120 Gy), and means were sep- trogen (i.e. AIR) and carbon (i.e. Vienna Pee Dee arated using Tukey’s Honestly Significantly Dif- Belemite). Samples were analysed at the stable ferent (HSD) or individual t-tests. All two-sided isotope facility at UC Davis (Davis, USA). To tests were performed using the SPSS software determine if a spermatheca was inseminated version 14 (SPSS Inc., Chicago, USA). by a 15N or a 13C labelled male, threshold values based on 2 (for 15N) or 3 (for 13C) standard de- viations above the mean values observed for Results virgin females were used (Macneale et al. 2005, Helinski et al. 2007). Averages were based on all After 4 nights of mating, on average 92 ± 2% virgin samples across the experiments. (± s.e.m.) of all females were inseminated and There was very little variation in mean no significant differences were observed be- δ13C‰ values for virgin samples (M= -25.86, tween treatments (i.e. treatments 1, 2 and 3;

s.d. 0.16), and the threshold value was set at 3 F2,9= 0.42, p> 0.05). Irradiation, therefore, had s.d. above the average value. The δ15N‰ values no impact on insemination in these competi- were more difficult to interpret and substantial tion experiments. variation was observed. Two δ15N‰ values of The insemination of spermathecae virgin females were removed from the dataset by either or both groups of labelled males was to normalise the data (i.e. -0.94 and -1.71). One determined based on the threshold values for high δ15N‰ was observed in a control (treat- δ15N‰ and δ13C‰. The classification of sper- ment 1) replicate (i.e. 15.88); this data point was mathecae into the different labelling groups is included as a 15N positive sample throughout the illustrated for the control, treatment 1 (Fig. 9.1). sample analysis but removed from the analysis Even though all spermathecae were inseminat- of the mean ± s.d. δ15N‰ for labelled samples. ed, in some samples the label was below the de- The standard deviation for mean virgin samples tection limit and these samples were classified remained quite large (M= -4.08, s.d. 0.68), and as not having sufficient label (i.e. indicated as a threshold value of 2 s.d. was used. none, Fig. 9.1 and Table 9.1). For all data com- bined, the mean delta values ± s.d. of samples Statistical analysis classified as 15N labelled was 0.09 ± 2.01 (N= Prior to analyses, data were checked for nor- 124), and for unlabelled samples –4.26 ± 0.98 mality. To determine competitiveness, the δ15N‰ (N= 112). For 13C, mean δ13C‰ values number of spermathecae inseminated by 15N or were –20.80 ± 2.00 (N= 129) for labelled and 13C labelled males (in the case of radiation stud- –25.86 ± 0.17 (N= 107) for unlabelled samples. ies these were categorised as irradiated versus In the absence of radiation, multiple un-irradiated; Table 9.1) were analysed using insemination was observed in three out of the a replicated G-test of goodness of fit. Even four replicates for the control treatment 1, and though in Table 9.1 the frequencies are grouped on average 14 ± 6% of inseminated spermath- separately per category (e.g. for 15N only sper- ecae were positive for both 13C and 15N (Table 140 mathecae positive for 15N are grouped), for 9.1). The competitiveness of both groups of G-test analyses spermathecae in which both males was determined, however replicates labels were present were added to obtain the differed statistically and data could not be

total number of spermathecae inseminated by grouped (G3= 8.30, p< 0.05). The individual G- either group of labelled males. If there was no tests showed that the competitiveness of 13C or 09 Multiple mating in An. arabiensis

Table 9.1. Results from the competition experiments using isotope labelled males (i.e. 15N or 13C) to determine insemination in spermathecae. Treatments were 1) 15N un-irradiated males : 13C un-irradiated males : females (i.e. control), 2) 15N irradiated males : 13C un-irradiated males : females, 3) 15N un-irradiated males : 13C irradiated males : females. Males were irradiated with 70 Gy or 120 Gy in the pupal stage. For the control treatment 1, the proportion and the number (in between brackets) of spermathecae positive for 15N or 13C, are indicated. For the radiation treatments, spermathecae inseminated by irradiated or un-irradiated males are shown. Spermathecae positive for both isotopes are classified as inseminated by ”Both”. Those with amounts of isotopes below the detection threshold are classified as ‘Neither”. N is the total number of spermathecae analysed per row. Individual and pooled (when replicates could be grouped) G-test results for each competition experiment are given (see text for details) with significant differences at * p< 0.05 and ** p< 0.01. Average values for competition experiments (insemination by irradiated or un-irradiated males only) are only given when replicates could be grouped. n/a: not applicable (replicates could not be grouped based on G-test results)

Control Exp. Treatment 1 % 15N % 13C % Both % Neither N G-test (N) (N) isotopes isotope (N) 15N vs 13C (N) 1 1 65 (13) 20 (4) 15 (3) 0 (0) 20 3.62 2 1 60 (12) 10 (2) 30 (6) 0 (0) 20 3.95* 3 1 25 (5) 60 (12) 10 (2) 5 (1) 20 2.38 4 1 45 (9) 45 (9) 0 (0) 10 (2) 20 0 average n/a n/a 14 ± 6 4 ± 2 n/a

70 Gy Exp. Treatments % Irrad. % Un- % Both % Neither N G-tests 2-3 (N) irrad. (N) isotopes isotope (N) 70 vs 0 Gy (N) 1 2 15N 35 (7) 35 (7) 30 (6) 0 (0) 20 0 1 3 13C 15 (3) 65 (13) 20 (4) 0 (0) 20 4.30* 3 2 15N 15 (3) 65 (13) 0 (0) 20 (4) 20 6.74** 3 3 13C 35 (7) 55 (11) 10 (2) 0 (0) 20 0.73 average 25 ± 6 55 ± 7 15 ± 6 5 ± 5 6.63*

120 Gy Exp. Treatments % Irrad. % Un- % Both % Neither N G-tests 2-3 (N) irrad. (N) isotopes isotope (N) 120 vs 0 Gy (N) 2 2 15N 18 (3) 65 (11) 18 (3) 0 (0) 17 3.29 2 3 13C 42 (8) 53 (10) 5 (1) 0 (0) 19 0.20 4 2 15N 0 (0) 95 (19) 5 (1) 0 (0) 20 21.07** 4 3 13C 30 (6) 65 (13) 5 (1) 0 (0) 20 2.38 average n/a n/a 8 ± 3 0 n/a

15N labelled males was similar with the excep- the spermathecae were positive for both labels tion of replicate two where 15N labelled males (Table 9.1). inseminated significantly more females than13 C For 120 Gy, the distributions of fe- labelled males. males inseminated by irradiated or un-irradiat- When males were irradiated with 70 ed males were not similar and replicates could

Gy and competed against un-irradiated males, not be pooled (G3= 10.59, p< 0.01). In three out replicates were statistically similar and data of four replicates males irradiated with 120 Gy were grouped (G3= 5.14, p> 0.05). Un-irradiated inseminated fewer females than un-irradiated males had a higher competitiveness than irradi- males, however, only in the third replicate a ated males (Table 9.1), and on average 55 ± 7% significant result was observed (Table 9.1). In of females were inseminated by un-irradiated replicate 2 irradiated males inseminated similar males only compared to 25 ± 6% by irradiated numbers of females compared to un-irradiated 141 males only. Cases of multiple mating were ob- males. Multiple insemination occurred in all served in three out of four replicates; levels of replicates and on average 8 ± 3% of the sper- multiple insemination observed were compara- mathecae were positive for 15N and 13C. ble to treatment 1 and on average 15 ± 6% of When data were combined, there were 09 Multiple mating in An. arabiensis Chapter 09

Figure 9.1. δ13C‰ and δ15N‰ values of spermathecae from treatment 1 (the four symbols indicate samples from the 4 replicates). Threshold values were defined as 2-3 standard deviations above the mean value of virgin spermathecae samples, and are indicated for δ15N‰ (solid line) and δ13C‰ (broken line). Experimental samples could thus be defined as 13C labelled (top left quadrant), 15N labelled (bottom right quadrant), labelled with both isotopes (Both, top right quadrant) or no label was detected (Neither, bottom left quadrant).

no significant differences in the proportion multiple insemination occurs in a laboratory of multiple insemination events for all treat- colony of An. arabiensis. On average 12 ± 3% 15 ments (F2, 9= 0.42, p> 0.05), thus the irradiation of spermathecae were labelled with both N of one group of males with either dose had no and 13C. It is assumed that undetected double impact on the proportion of multiple insemina- mating events (e.g. 13C followed by a second tion compared to the control treatment. The 13C insemination) occurred to the same extent alteration of the label for the irradiated males (Gomulski 1990, Klowden 2006), thus the ac- group did not have an impact on the proportion tual number of multiple insemination is about of multiple insemination events, and similar re- two-fold higher, and approximately 25% of all sults were observed for treatments 2 and 3 for females were inseminated more than once. both doses (t(6)= 0.43, p> 0.05). A higher frequency of multiple insemination was observed compared to results for An. cu- licifacies ((i.e. 12-14%; Mahmood and Reisen 142 1980), and An. gambiae s.s. (i.e. 12%; Gomulski Discussion 1990), however in those studies the proportion of females used was two-fold higher compared Using an entirely new method of stable isotope to this study. At a similar ratio of insects com- labelling, the findings of this study show that pared to this study (i.e. 1:1:1), a higher frequen- 09 Multiple mating in An. arabiensis

cy of multiple insemination was observed in An. males was lower compared to un-irradiated gambiae s.s. after 1 night of mating (i.e. 48%; males in the majority of replicates for 70 and Klowden 2006). 120 Gy. For 70 Gy, replicates could be pooled Irradiation appeared to have no im- and 55 ± 7% of the females were inseminated pact on the frequency of multiple insemination by un-irradiated males only. In previous studies, in this experimental design; for 70 Gy similar hatch data of individual egg batches were used frequencies were reported compared to control to determine paternity, and un-irradiated males treatments, while for 120 Gy lower frequencies accounted for 58 ± 5% and 75 ± 5% of the egg were found but differences were not signifi- batches for 70 and 120 Gy experiments, respec- cant. Even though pupal irradiated males have tively (Helinski and Knols 2008b), thus results fewer and smaller sperm in their testes (Helin- were comparable for 70 Gy. Hatch data of egg ski and Knols 2008a), in treatments where these batches permits the identification of batches males competed against un-irradiated males with intermediate hatch, which is assumed to for females, no increase in multiple insemina- be the result of a multiple insemination event tion was observed. Females inseminated by (Weidhaas and Schmidt 1963, George 1967, males irradiated in the pupal stage were thus Zalom et al. 1981, Helinski and Knols 2008b). not more likely to engage in another mating Data from small cage competition experiments event with an un-irradiated male. This is in con- on a 1:1 ratio observed 10% and 5% of the egg trast to results for the Mediterranean fruit fly, batches with intermediate hatch rate (i.e. be- where females inseminated by irradiated males tween 20-60%) for 70 and 120 Gy, respectively were more likely to re-mate (Mossinson and (Helinski and Knols 2008b), which is similar to Yuval 2003). However, males were introduced the values observed in this study, (i.e. 12.5% sequentially (i.e. between 1-3 days after first and 8% for 70 and 120 Gy, respectively). mating) while in the present study males were A review of the literature shows that introduced simultaneously. In addition, mating different results on multiple insemination are behaviour of both species is completely differ- observed between field and laboratory -stud ent (i.e. the Mediterranean fruit fly mating sys- ies. In the former, multiple insemination only tem is based on female-choice, while in Anoph- occurred in very low frequencies as assessed eles a female appears to have little influence using molecular markers (Yuval and Fritz 1994, on the selection of a copulation partner upon Tripet et al 2003), while in the latter a high pro- entering a swarm). However, for both species portion of multiple insemination can be found a signal from the sperm is thought to regulate using mutant strains (Mahmood and Reisen female mating behaviour (Miyatake et al. 1999, 1980, Gomulski 1990, Klowden 2006). In the Klowden 2006), and it thus seems that pupal ir- field, mating couples usually leave the swarm radiated An. arabiensis males transfer enough in copula (Charlwood and Jones 1980), and af- sperm to trigger this behaviour. As discussed, ter a successful mating females probably leave polyandry in anophelines would be advanta- the swarm site. In laboratory cages, females are geous for Sterile Insect Technique programmes confined with males throughout the swarm- on the premise that sperm of released males is ing period, usually in small cages (i.e. 30 cm equally competitive to wild males (Curtis 1985). square cages are common) at high densities. The competitiveness of irradiated sperm com- It is therefore likely that females are forced to pared to un-irradiated sperm for fertilisation of fly into the swarm on several occasions due to the eggs was not determined in this study, and disturbances in the cage, which facilitates re- only the incidence of multiple insemination was mating. Alternatively, it has been suggested 143 scored. In future studies it would be necessary that in small cages mating is interrupted (i.e. to determine which sperm (i.e irradiated or un- due to space restriction when flying in copula, irradiated) the female uses to fertilise her eggs. or other males disturbing the couple) resulting The competitiveness of irradiated in the incomplete transfer of semen making fe- 09 Multiple mating in An. arabiensis

males more responsive to a subsequent mating Culex pipiens L. glycine was essential to achieve

Chapter 09 (Mahmood and Reisen 1980). This suggestion normal growth rates (Clements 1992). No such was based on results from a study in An. culici- data for anophelines are available, however the facies, which found that when cage size was in- standard diet provided to the larvae in our in- creased and insect density decreased, multiple sectary always resulted in normal growth and insemination decreased (Mahmood and Reisen development; and it thus seems unlikely that 1980). From the above it is evident that studies 13C-glucose labelled males were deprived of gly- should be repeated in much larger cages (e.g. cine. In future experiments it might neverthe- semi-field systems; Knols et al. 2002) to exclude less be worthwhile to add unlabelled glycine to disturbances and space issues as the possible the 13C-glucose labelled trays and vice versa. explanations for the observed results. The threshold values used in this study The dual-labelling approach with two classified some spermathecae as unlabelled different isotopes assumed equal mating abil- even though they were visually inseminated; ity of both groups of labelled males. In previous however, only about 3% of the spermathecae studies, the mating ability of 13C-glucose and were in this category. A careful look at the data 15N-glycine labelled males were similar to un- indicated that the majority of the difficulties in labelled males (Helinski et al. 2007, 2008), and the data analysis were caused by the δ15N‰ for 13C it was determined that the use of 13C-la- values, and a greater variation in unlabelled belled glucose or unlabelled glucose had no im- samples was observed compared to the δ13C‰ pact on larval development and male longevity values. Even though in previous work an en- compared to a control with no glucose (Helin- richment of 5 atom% 15N was sufficient to dis- ski et al. 2007). In three out of four replicates in tinguish inseminated spermathecae from un- the control treatment, 15N or 13C labelled males inseminated ones (Helinski et al. 2008), in the were equally competitive, while in the other present study we conclude that a higher enrich- replicate 15N males were just significantly more ment would have resulted in a better spread of competitive (p= 0.047). In the irradiation experi- the data for labelled and unlabelled samples. ments no specific pattern was observed to indi- Samples in this study were analysed at a differ- cate that males of a particular label performed ent facility compared to previous work (Helinski better compared to the other group (Table 9.1). et al. 2008), and this could have accounted for For the different labelling treatments, different the differences observed. It is recommended in carrier compounds for the isotopes were used, future experiments to increase the enrichment i.e. glycine and glucose. Even though in our of 15N labelled males, for instance double the initial data comparable levels of larval survival amount of label can be used. This would only were observed for larvae reared with 13C-glu- result in a marginal increase in costs (i.e. 5 USD cose (Helinski et al. 2007), in these experiments versus 2.5 USD per larval tray) compared to 13C a reduced growth rate and survival in 13C trays labelling (i.e. 25 USD per larval tray). The im- was sometimes observed compared to unla- pact of an increase of 15N-glycine on larval de- belled or 15N-glycine trays. In larvae of Aedes velopment however, needs to be tested, as high aegypti Wiedemann the addition of glucose at amino-acid concentrations could be deleterious high concentrations was observed to slow de- (Clements 1992). velopment rates (Clements 1992), and it is likely that such an effect occasionally took place after 13C-glucose was added. It should be noted that Conclusion 144 labelling with 13C-glucose is possible but ade- quate attention should be given to prevent a re- In conclusion, the first time use of a dual-label- duction in growth. Glycine did not belong to the ling system of stable isotopes in mating studies class of essential amino-acids needed for devel- gave evidence for multiple insemination events opment in Aedes aegypti L. larvae; however, in in a laboratory strain of An. arabiensis. Further 09 Multiple mating in An. arabiensis

optimisation of the method is possible, and mosquitoes, Volume 1. Development, nutrition and the use of a higher enrichment of 15N-labelled reproduction: London, Chapmann & Hall. males is recommended for future experiments. Clements, A. N. 1999. Mating, pp. 360-402. In The Discrepancies between field and laboratory Biology of Mosquitoes, Volume 2: Sensory reception collected data indicate that these experiments and behaviour. CABI Publishing, Wallingford. should be repeated in larger (semi-field) cages Curtis, C. F. 1985. Genetic control of insect pests: to avoid artificially forcing females into repeat- growth industry or lead balloon. Biol. J. Linn. Soc. ed contact with males. In the context of genetic 26:359-374. control studies with the Sterile Insect Technique, George, J. A. 1967. Effect of mating sequence on experiments were performed to determine the egg hatch from female Aedes aegypti (L.) mated with effects of irradiation on the frequency of multi- irradiated and normal males. Mosq. News 27:82-86. ple insemination. In small cages, the irradiation Gomulski, L. M. 1990. Polyandry in nulliparous of males had no impact on the proportion of Anopheles gambiae mosquitoes (Diptera: Culicidae). multiple insemination events when these males Bull. Entomol. Res. 80:393-396. competed with un-irradiated males for mates. Helinski, M. E. H., A. G. Parker, and B. G. J. Knols Further studies on the sequential introduction 2006. Radiation-induced sterility for pupal and adult of mates and the determination of sperm use stages of the malaria mosquito Anopheles arabiensis. for fertilisation are recommended. Malar. J. 5: 41. CHAPTER 3. Helinski, M. E. H., R. C. Hood-Nowotny, L. Mayr, and B. G. J. Knols 2007. Stable isotope-mass Acknowledgements spectrometric determination of semen transfer in We would like to thank D. Harris for the sample malaria mosquitoes. J. Exp. Biol. 210:1266-1274. analysis, and M. Dicke and two anonymous re- CHAPTER 5. viewers for constructive comments during prep- Helinski, M. E. H., and B. G. J. Knols 2008a. Sperm aration of the manuscript. BGJK is supported by quantity and size polymorphism in normal and a VIDI grant from the Netherlands Organisation irradiated males of the malaria mosquito Anopheles for Scientific Research (#864.03.004). arabiensis Patton. Acta Trop. submitted. CHAPTER 4. Helinski, M. E. H., and B. G. J. Knols 2008b. Mating References competitiveness of male Anopheles arabiensis mosquitoes irradiated with a partially- or fully- Arnqvist, G., and T. Nilsson 2000. The evolution sterilising dose in small and large laboratory cages. of polyandry: multiple mating and female fitness in J. Med. Entomol. in press. CHAPTER 7 insects. Animal Behav. 60:145–164. Helinski, M. E. H., R. C. Hood, and B. G. J. Knols Beard, C. B., M. Q. Benedict, J. P. Primus, V. 2008. Use of the 15N stable isotope as a semen label Finnerty, and F. H. Collins 1995. Eye pigments in to detect mating in the malaria mosquito Anopheles wild-type and eye-color mutant strains of the African arabiensis Patton. Bull. Entomol. Res. submitted. malaria vector Anopheles gambiae. J. Heridity 86:375- CHAPTER 6. 380. Hood-Nowotny, R. C., and B. G. J. Knols 2007. Bressac, C., and E. Hauschteck-Jungen 1996. Stable isotope methods in biological and ecological Drosophila subobscura females preferentially select studies of arthropods. Entomol. Exp. Appl. 124:3-16. long sperm for storage and use. J. Insect Physiol. Klowden, M. J., and G. M. Chambers 2004. 42:323-328. Production of polymorphic sperm by anopheline Charlwood, J. D., and M. D. R. Jones 1980. Mating mosquitoes and their fate within the female genital 145 in the mosquito, Anopheles gambiae s.l. II. Swarming tract. J. Insect Physiol. 50:1163-1170. behaviour. Physiol. Entomol. 5:315-320. Klowden, M. J. 2006. Switchover to the mated state Clements, A. N. 1992. Larval Nutrition, excretion by spermathecal activation in female Anopheles and respiration, pp. 109-111. In The biology of gambiae mosquitoes. J. Insect Physiol. 52:679-684. 09 Multiple mating in An. arabiensis

Knols, B. G. J., B. N. Njiru, E. M. Mathenge, W.

Chapter 09 R. Mukabana, J. C. Beier, and G. F. Killeen 2002. MalariaSphere: A greenhouse-enclosed simulation of a natural Anopheles gambiae (Diptera: Culicidae) ecosystem in western Kenya. Malar. J. 1:19. Macneale, K. H., B. L. Peckarsky, and G. E. Likens 2005. Stable isotopes identify dispersal patterns of stonefly populations living along stream corridors. Freshw. Biol. 50:1117-1130. Mahmood, F., and W. K. Reisen 1980. Anopheles culicifacies: the occurrence of multiple insemination under laboratory conditions. Entomol. Exp. Appl. 27:69–76. Mason, G. F. 1967. Genetic studies on mutations in species A and B of the Anopheles gambiae complex. Gen. Res.10:205-217. Miyatake, T., T. Chapman, and L. Partridge 1999. Mating-induced inhibition of remating in female Mediterranean fruit flies Ceratitis capitata. J. Insect Physiol. 45:1021-1028. Mossinson, S., and B. Yuval 2003. Regulation of sexual receptivity of female Mediterranean fruit flies: old hypotheses revisted and a new synthesis proposed. J. Insect Physiol. 49:561-567. Tripet, F., Y. T. Touré, G. Dolo, and G. C. Lanzaro 2003. Frequency of multiple inseminations in field- collected Anopheles gambiae females revealed by DNA analysis of transferred sperm. Am. J. Trop. Med. Hyg. 68:1-5. Weidhaas, D. E. and C. H. Schmidt 1963. Mating ability of male mosquitoes, Aedes aegypti (L.), sterilized chemically or by gamma radiation. Mosq. News 23:32-35. Yuval, B., and G. N. Fritz 1994. Multiple mating in female mosquitoes - evidence from a field population of Anopheles freeborni (Diptera: Culicidae). Bull. Entomol. Res. 84:137-139. Zalom, F. G., S. M. Asman, and J. E. Fields 1981. Tests for multiple insemination of females involving irradiated and unirradiated male Culex tarsalis (Diptera: Culicidae). Mosq. News 41:154-156.

146 09 Multiple mating in An. arabiensis

147 10 Towards a SIT field release in Sudan 10 Towards a Sterile Insect Technique field release of Anopheles arabiensis mosquitoes in Sudan: Irradiation, transportation, and field cage experimentation by Michelle EH Helinski, Mo’awia M Hassan, Waleed M El-Motasim, Colin A Malcolm, Bart GJ Knols and Badria El-Sayed

The work described in this chapter forms part of a feasibility study to suppress a population of the malaria vector Anopheles arabiensis in Northern State, Sudan, with the Sterile Insect Technique. Mosquitoes were irradiated in Khartoum and transported as adults by air to the field site earmarked for future releases (400 km from the laboratory). Experiments under near-natural conditions were performed in a large field cage, and the preparation of this cage for experiments, and some pilot experiments on mating and survival are described. Minor problems were experienced350-500 Million with cases the of irradiation clinical malaria of insects, occur mostly annually, associated 60% of withwhich the are absence in sub- ofSaharan a rearing Africa. facility Moreover, in close 80% proximity of all deaths to the attributed irradiation to source. malaria The occur small-scale in this transportationregion. In numbers, of adult 1 million mosquitoes Africans to the die release of the sitedisease resulted each in year, minimal with mortality the vast (

Published in Malaria Journal 2008, 7: 65 10 Towards a SIT field release in Sudan

Introduction view of the release programmes performed.

Chapter 10 The largest SIT release programme against an he Sterile Insect Technique (SIT) entails Anopheles vector (An. albimanus Wiedemann) the mass production, sex separation, was executed in El Salvador in the 1970s (Lof- Tsterilisation, and subsequent release of gren et al. 1974). Over a 5-month period, 4.3 sterile male insects into a target population million mosquito pupae were mass-produced, in an area-wide, and usually integrated, pest sterilised, and released around Lake Apaste- management strategy. The released males in- peque. Analysis of An. albimanus population seminate wild females with sterile sperm. The data (Benedict and Robinson 2003) from the females subsequently fail to produce viable release and a nearby control area demon- offspring leading to an overall size reduction of strates how successful the sterile males were the target population. Over the years, SIT has in preventing a normal seasonal rise in vector proven to be a safe, effective and environmen- density (Curtis 2006). A subsequent, more ex- tally sound method to suppress, eliminate or tensive trial, located on the Pacific coast of El contain particular insect pest populations (Dyck Salvador, took place from 1977-79 (Lowe et al. et al. 2005). The International Atomic Energy 1980, Dame et al. 1981) when up to 0.5 million Agency (IAEA) has a long history of supporting sterile males or 1.25 million sterile male pupae SIT programmes against major insect pests, in- were released daily. Complete control was not cluding fruit flies, tsetse flies and codling moths. achieved due to the immigration of females In 2004, the IAEA initiated a five-year feasibility from surrounding areas, despite the introduc- study to develop technologies for controlling tion of a barrier zone (Dame et al. 1981). malaria mosquitoes with the SIT (Helinski et al. The Tropical Medicine Research Insti- 2006a, IAEA 2008). tute (TMRI) in Khartoum is leading a Republic The use of the SIT as a genetic control of Sudan project to explore the possibility of strategy for mosquitoes is not new. The major- using SIT as part of an integrated area-wide ap- ity of studies on genetic control of mosquitoes proach to control Anopheles arabiensis Patton were conducted from 1950-1980. The induction in the north west of the country. The IAEA in of dominant lethality by radiation or sterilising collaboration with many partners is involved in chemicals was perhaps the most researched the development and evaluation of the neces- area (Dame 1985), but other forms of genetic sary components needed for an area-wide in- control, e.g. translocations or other chromo- tegrated approach to vector control of African somal rearrangements, were also undertaken. malaria vectors using the SIT. Some of the ex- Benedict and Robinson (2003) provide a re- perimental work is performed at the Agency’s

Figure 10.1. Satellite image of the project area along the Nile, situated between Dongola and Merowe, in Northern State, Sudan. The position of the Merowe dam and the reservoir lake are shown.

150 10 Towards a SIT field release in Sudan

laboratories in Seibersdorf, Austria, and a pilot Nile. Temperatures in the area fluctuate greatly project area is under development in Sudan. The and are between 7-47 ˚C (Malcolm et al. 2007), field site of the pilot project is situated in North- and relative humidity ranges from five to 70%. ern State, where pockets of breeding sites of the An. arabiensis is the only malaria vector present malaria vector An. arabiensis occur on the banks (Malcolm et al. 2007). of the Nile in an area otherwise surrounded by The rearing and sexual sterilisation irrigated land and desert. The area stretches by gamma radiation occurs in Khartoum, situ- from Dongola in the north to Merowe in the ated approximately 400 km south of Merowe. south and is about 350 km long following the The field site and Khartoum are connected by Nile (Figure 10.1). Upriver from Merowe, a dam air with two to three commercial 1 hr flights a is near to completion that will create a reservoir week (i.e. to Dongola), or 6 hours by road. Ir- lake of approximately 200 km in length. An. ara- radiation studies had been undertaken at the biensis is very rare along this stretch of the Nile IAEA (Helinski et al. 2006b) but to date, no ir- and the lake will eliminate it altogether. The radiation of anophelines had been performed in lake will be surrounded by rocky desert unsuit- Sudan, and no irradiated males had been trans- able for human habitation and so it should form ported from Khartoum to the field site. Here we a barrier to prevent migration of mosquitoes report our initial experiences on the irradiation downriver. To the north of Dongola, conditions and subsequent transportation of adult mos- become far less favourable for An. arabiensis, quitoes to the field site by air. due to the rocky terrain, sparse human popula- A field cage was constructed in Don- tion and colder climate. In addition, the Gam- gola to perform a variety of experiments under biae Control Project jointly run by the Egyptian near-natural conditions (Knols et al. 2002), with and Sudanese Ministries of Health has been op- emphasis on the survival and mating competi- erating since 1970 along the Nile from Aswan in tiveness of irradiated males against wild mos- Egypt to Abu Fatma in Sudan. This programme quitoes collected from the surrounding area. has maintained and extended an An. arabiensis Only limited data on survival had previously free zone that now reaches Abri (Malcolm et been collected of insects placed in small rear- al. 2007). All mosquito breeding sites between ing cages in the field cage (M. Hassan, personal Merowe and Dongola are found within 5 km of communication). These indicated that survival the river banks. Breeding sites close to the river was poor, and conditions in the field cage had are primarily natural breeding sites associated to be improved to create more favourable con- with the seasonal flooding of the river, while the ditions for mosquitoes. We here describe the majority of breeding sites inland is man-made preparation of the field cage for experiments and associated with agricultural practices. The and present data on mating and survival of wild human population in Northern State is around and laboratory-reared mosquitoes. Of key im- 600,000 and heavily concentrated along the portance was to assess if the laboratory-adapt- ed males would withstand transportation to Table 10.1. Mortality of adult males during transport the field site and survive and mate under field from Khartoum to Dongola by air. In experiment 1 and 2 irradiated males were used, in experiment 3 un-irradiated conditions. In addition, two small-scale compe- males. tition experiments were performed to investi- gate the competitiveness of irradiated males. Experiment # Adult males Mortality transported (%) 1 259 0 Methods 151 2 300 4 Mosquitoes 3 600 6 The laboratory strain used for all experiments is the Dongola strain of Anopheles arabiensis. The

10 Towards a SIT field release in Sudan

Figure 10.2. Schematics of the field cage in Dongola. The

Chapter 10 cage was divided in three equal sections (a-c); only for section a the various items are displayed but other sections were identical; Pictures 1-4 indicate the resting sites: 1) kuseba, 2) corner site, 3) tree trunk, 4) zir; position in field cage indicated by numbers. Picture 5 shows plant beds seeded with local crops and plants rooted in soil-containing sacks.

strain was collected from breeding sites close was situated in Soba, approximately 45 min by to the town of Dongola and taken into culture car from the rearing facility in downtown Khar- in 2004. It was maintained both in the insectary toum; thus insects had to be transported to and in Seibersdorf (T: 27 ± 1 °C, RH: 82 ± 2%) and from the source. Insects were irradiated with in Khartoum for sixty generations (T: 26 ± 2 °C, a partially-sterilising dose of 70 Gy (Helinski et RH: 60 ± 10%) for use in the irradiation, trans- al. 2006b, Helinski and Knols 2008a, 2008b), portation and field cage experiments. All wild and two separate irradiation sessions with dif- mosquitoes used in the field cage experiments ferent batches of insects were performed. Ap- were collected from breeding sites in the vicin- proximately 500 males were irradiated in each ity of the field cage, either as larvae or pupae. session. A dosimetry system was used to verify They emerged in the Dongola insectary (i.e. the dose received by the batch (Helinski et al. situated in a building next to the field cage) and 2006b, Helinski and Knols 2008a). Males were were held in standard rearing cages (30 x 30 x irradiated in the pupal or adult stage, and pu- 30 cm) until release into the field cage. All mos- pae were sexed before irradiation. Pupal sexing quitoes were sexed < 18 hrs after emergence to was done manually by looking at the terminalia ensure virginity. Adult mosquitoes were main- under a stereoscope; adult sexing was also done tained in the insectary on sugar water (i.e. 10% manually by visual determination. Pupae were sucrose solution). transported to the source in a small holding 152 container with a screw top to prevent spilling. Irradiation and transport source Adults were transported in a standard rearing Insects were irradiated in a Cobalt-60 Gam- cage, placed in a stryrofoam box and covered macell (Nordion 220) following procedures with moist towels to avoid overheating of the described in Helinski et al. (2006b). The source mosquitoes. 10 Towards a SIT field release in Sudan

Transport Khartoum-field site covered with clay, with a small opening where Un-irradiated males and two batches of irradi- mosquitoes could enter), corner sites (i.e. brick/ ated males were taken by air as adults in sepa- clay structures in the corners of the section), a rate trips to the field station in Dongola (Table tree trunk, and a zir (i.e. local clay pot closed 10.1). For transportation, adults were placed partly with bricks; Fig. 10.2). The kuseba and the fifty at a time in standard paper drinking cups zir were filled with a shallow amount of water to covered with cotton mesh. Sugar solution was increase humidity. To further increase humidity provided in cotton wool secured on top of the and shelter, various types of vegetation were cup. The cups were then fixed in place in a styro- used. These consisted of two large plant beds foam box and covered with moist towels. Upon in the middle of each section seeded with local arrival in Dongola the boxes with mosquitoes crops such as bean and maize (Fig. 10.2), and were transported to the local insectary by car. some smaller plant beds on the east side. Ad- ditionally, plants rooted in soil-containing sacks Field cage experiments were distributed across the area (Fig. 10.2). Preparation of field cage The field cage (18 x 8 x 2.75 m) used in the ex- Experimental setup periments consisted of a metal structure fixed Before introduction of mosquitoes, plant beds with thick green shade netting that allowed were flooded, and this was done every morn- for air and light exchange to simulate ambient ing when experiments were run for more than conditions. The field cage was portioned into one night. Two sugar feeders with a sucrose three identical sections (6 x 8 m); this allowed solution and a drop of honey were placed in for the simultaneous performance of three ex- each section; one in the kuseba and one close periments (Fig. 10.2). Sections were made by to a corner site, and refreshed every other day. gluing netting material to the metal structure, Mating experiments lasted for 1 or 2 nights. In and clay and bricks were used to fix the netting the latter case, mosquitoes were introduced in at the bottom. Each section was accessible by the early evening, in the former in the early af- an outside door. A datalogger, recording am- ternoon. After mating, males and females were bient temperature and relative humidity, was recaptured from the field cage by searching the situated in each section at 1 m height, and re- resting sites and the whole section during the cordings were made for two days (Table 10.2). day. Mostly, mosquitoes were concentrated in Each section was equipped with resting sites a few places and these were checked carefully consisting of a kuseba (i.e. a brick structure for approximately 2 hours. For the competition experiments, after the day collections, human Table 10.2. Climatic conditions measured for two days in landing catches were performed in the early two resting sites in Section B, i.e. the kuseba and one of evening (i.e. from 8-10 pm). the corner sites in the left side of the field cage; ambient Recaptured males were counted and a conditions are logged by the permanent logger in Section wing was clipped for size determination (Lyimo B at 1 m height. Mean (± s.d.), maximum, and minimum values are given. *logger got too wet to deliver precise and Takken 1993, Charlwood et al. 2002, Louni- readings. bos et al. 1995). Recaptured females were ei- 4-5th April 2007 ther dissected to determine insemination sta- Site kuseba corner ambient tus by the examination of the spermathecae for sperm, and a wing was collected, or kept for mean T (˚C) 21.9 ± 3.1 24.8 ± 4.7 29.7 ± 6.8 blood feeding and egg laying (i.e. in the com- Tmax 27.5 34.4 41.1 petition experiments). With the exception of 153 Tmin 17.5 19.4 19.0 the first competition experiment wings from mean RH (%) >95* 62.8 ± 10.7 16.3 ± 7.3 females were classified as coming from insemi- RHmax 100* 89 37.6 nated or un-inseminated females to determine RHmin 57.6 35.9 6.5 a possible size preference of males. A digital

10 Towards a SIT field release in Sudan

Table 10.3. Data from control experiments where wild or laboratory reared males were introduced with wild females. The number and age (i.e. between brackets) of the mosquitoes introduced is given, as well as the ratio of males and females

Chapter 10 used. Recapture rates of males and females from the field cage after mating, as well as insemination rate and number of females dissected (N) are presented. In experiment 1 the mating period was two nights, in experiments 2 and 3 mating lasted for one night. Experiment # ! lab # ! wild # " wild Ratio Recapture (%) Insemination (age in d) (age in d) (age in d) !: " ! " N %

1 wild males - 275 (2-3) 254 (2-3) 1:1 51 28 55 96 section A

2 wild males - 60 (3) 100 (3) 1: 1.66 25 55 42 81 section A

2 lab males 60 (5) - 100 (3) 1: 1.66 46 55 39 72 section B

3 lab males 60 (4) - 100 (3-4) 1: 1.66 67 60 56 46 section A

3 wild males - 60 (4) 100 (3-4) 1: 1.66 72 50 40 43 section B

3 lab males 60 (4) - 100 (3-4) 1: 1.66 97 57 44 18 section C

image of the wing was taken (CC-12 camera, in the experiments was between 2-5 days, and Soft Imaging System, Germany, mounted on a mosquitoes were introduced at either a 1:1 ra- stereo microscope). Wing length was measured tio, or a higher ratio of females was used (Ta- between the alula notch and the wing tip, ex- ble 10.3). In the first experiment mating was cluding scales; measurements were performed allowed for two nights; in the second and third with AnalySIS FIVE software (Soft Imaging Sys- experiment mating lasted for one night. tem, Germany). In some experiments the males or females were dusted with fluorescent pow- Competition experiments der prior to release. Mosquitoes were placed Mosquito irradiation for these experiments was in a plastic cup that was sealed at the top, and performed in Khartoum and adults were trans- powder was applied with a syringe. Dusting was ported by air (Table 10.1). Because the number done to distinguish groups of mosquitoes after of irradiated mosquitoes was low, only two recapture and to identify potential survivors experiments with different batches of insects from the previous experiment (i.e. when experi- were performed. Age of the mosquitoes at the ments were performed in close succession). start of the experiment was between 2-4 days (Table 10.4), and mating lasted for 2 nights. Ir- Control experiments radiated males competed against wild males Control experiments were performed to test for wild virgin females. The mating period was the mating performance and survival of wild 2 nights. For the first experiment, males irradi- or laboratory reared males when confined with ated with 70 Gy as pupae or adults were com- wild virgin females. For each group of males, bined, and introduced at a ratio of ~ 2 irradiated 154 three replicates were performed (Table 10.3) males versus 1 wild male. In the second experi- with different batches of mosquitoes (with ment, all males were irradiated as pupae with the exception of experiment 3 where the same 60 Gy, and introduced at a 1:1 ratio with wild batch of laboratory males was used for the two males (Table 10.4). In experiment two, wild and replicates). Age of the males and females used irradiated males were dusted with different col- 10 Towards a SIT field release in Sudan

ours to distinguish them after recapture. The source, and a mortality of approximately 50% recovered females were blood fed on a human upon arrival and after cooling and irradiation arm, and females were placed in individual vials was observed. For the second session, older for egg laying (experiment 1) or egg laying oc- adults were irradiated and transportation and curred en masse (experiment 2). cooling went well, however spills of the sugar source on the return journey caused substantial Statistical analysis mortality. Dosimetry confirmed that the dose General Linear Models (GLMs) or individual received during the first irradiation session was t-tests were used to compare the results from close to 70 Gy, however in the 2nd session the different treatments, and means were sepa- dose received by the batch was slightly lower rated using Tukey’s Honestly Significantly Dif- (i.e. 60 Gy). ferent (HSD) tests. All tests were two-sided and Transportation of adults to the field performed using the SPSS software version 14 site by air was successful. For the first batch (SPSS Inc., Chicago, USA). of males no mortality was observed upon ar- rival and insects spent approximately 5 hrs in the cups (Table 10.1). In the second and third Results batch, mortality was slightly higher because males spent more time in the cups (i.e. 7 hrs), Irradiation and transportation and in the last replicate a cardboard box was The first pupal irradiation session resulted in used instead of stryrofoam, but overall mortal- some pupal mortality after irradiation and this ity remained below 6% (Table 10.1). was attributed to problems associated with the container used (i.e. some spilling had oc- Field cage experiments curred during transport). In the second session Data from the three loggers indicated that the the container used for holding was properly three sections were identical in ambient con- sealed and pupal mortality was low (i.e. < 5%). ditions (data not shown). The fluctuation of The adults used in the first irradiation session temperature and relative humidity over 24 hrs had barely emerged upon transportation to the is shown in Fig. 10.3. Overall, humidity was low

Figure 10.3. Typical tempera- ture (bold line) and relative humidity patterns in the field cage in April logged at 1 m height in section C over 24 hrs. The increase in humidity in the morning is associated with the flooding of the plant beds.

155 10 Towards a SIT field release in Sudan

Table 10.4. Competition experiments in the field cage. The number and age (i.e. in days between brackets) of the mosquitoes introduced is given, as well as the ratio of irradiated males (! *)! versus wild males versus wild females (for the 2nd batch nd

Chapter 10 of females in exp. 1 this ratio was not known). Mating period lasted for 2 nights (with the exception of the 2 batch of females in experiment 1 where mating lasted for 1 night). In experiment 1, first a batch of wild females was introduced, and secondly laboratory females. For females the number recaptured by daytime checking (am) and human landing catches in the evening (pm) is given. The proportion of recaptured males and females after mating, as well as insemination rate (%) and number of females dissected (N) after recapture are given. n/d: not determined; n/n: not known

Experiment 1 # "* # " wild # ! wild # ! lab Ratio # ! Inseminated irradiated (age in d) (age in d) (age in d) "*:":! recaptured (N) (age in d) am pm section A 256 (2) 110 (2-3) 170 (2-3) ~ 2:1:0.6 19 2 first batch ! recapture (%) n/d 12 100 (3)

section A 175 n/n 43 3 second batch !

recapture (%) 5 26 63 (30)

experiment 2 223 (4) 223 (2-4) 223 (2-4) 1:1:1 51 2 section B recapture (%) 34 38 24 72 (18)

and dropped during the day when temperatures seminated (10. 3). In the third experiment, peaked, and temperatures fluctuated by 20 ˚C conditions were similar as in experiment two; over 24 hrs. The east side of the field cage was however insemination rate was lower. Wild and cooler and the majority of resting mosquitoes laboratory males in sections A and B of the field sampled during the day were observed in this cage inseminated 43% and 46% of the females, part for all sections. Preferred resting sites were respectively, and insemination of females mat- the kuseba, the corner sites, and the lower parts ed to laboratory males in section C was only of the vegetation close to these sites. Data log- 18%. The reasons for this lower and variable gers placed in these sites showed that tempera- insemination are not clear. No significant differ- ture was lower, and humidity higher compared ences were observed in mean insemination for to ambient conditions (Table 10.2). All mosqui- laboratory and wild males when combining the toes were recaptured close to the ground, i.e. three replicates (t(4)= 1.26, p> 0.05). at less than 0.5 m height. The human landing After two nights of mating in experi- catches performed in the competition experi- ment one, 28% of the females and 51% of the ments yielded less than 5 females (Table 10.4). wild males were recaptured (Table 10.3). More mosquitoes were recaptured the following Control experiments day (i.e. around 20), thus some mosquitoes When wild males and females were introduced escaped collection, and this was observed in at a 1:1 ratio, insemination of the females was other experiments as well. In experiments two 96% after 48 hrs (Table 10.3). Even though the and three, mating was restricted to one night 156 ratio of females was increased and mating re- and the recapture rate of females was between duced to one night only in the second experi- 50-60%. Recapture of wild males was highly ment, wild males inseminated 81% of the fe- variable and ranged from 25-72%, and the same males, and the laboratory males performed applied to laboratory males with rates ranging only slightly less with 72% of the females in- from 46-97% (Table 10.3). 10 Towards a SIT field release in Sudan

Competition experiments 18), however, only 102 eggs were laid with a After two nights of mating in experiment one, hatch rate of 11%. The unfed females (N= 18) only a small fraction of the wild females intro- were dissected for insemination and 72% of duced was recaptured (i.e. 12%; Table 10.4). these were inseminated (Table 10.4). Because a substantial number of males were observed in the cage a second batch of females Wing length data (i.e. dusted laboratory) were introduced (Table For all experiments (i.e. control and competi- 10.4). The following day around 26% of dusted tion) combined, the wild males were smaller females were recaptured. Recapture of the than laboratory reared males (t(220)= -6.62, p< males after 84 hrs was only 5%. Blood feeding of 0.01; Table 10.5). However, no size-differences the recaptured females was difficult; in the first were observed between wild and laboratory batch, none of the wild females fed blood after males in control experiment two (t(6.35)= 0.24, several opportunities to do so. Few females p> 0.05) or three (t(62)= 0.37, p> 0.05; Table from the second batch fed (N= 10), and only 10.5). Although no samples were available of ir- three egg batches were obtained. Hatch data radiated males in competition experiment one, showed that 1 batch had a hatch rate of 90% the males recaptured after mating probably and was thus fathered by a wild male, while the were irradiated laboratory males as their wing other 2 batches had 0% hatch and were likely size was significantly larger (i.e. 2.70 ± 0.05 to be the result of a mating with an irradiated mm, N= 16; t(21)= -2.60, p< 0.05) compared to male. Unfed alive females were dissected to the wild males sampled before introduction. determine insemination status. From the first The wild females from competition batch of wild females, three were dissected for experiment one (i.e. 2.62 ± 0.03 mm, N= 17) insemination and all were inseminated; in the were strikingly smaller than the second batch second batch 63% of females (N= 30) were in- of laboratory females introduced (i.e. 2.95 ± seminated after 1 night of mating (Table 10.4). 0.05 mm, N= 21; t(36)= -5.46, p< 0.01). The In the second experiment, an equal wild females from all other experiments were proportion of irradiated and wild males was larger and similar in size compared to the labo- recaptured after mating. Of the recaptured ratory females from competition experiment females 56% fed (N= 30). Three days after the one (t(204)= -1.41, p> 0.05). No size differences blood meal, the alive fed females were flown were observed between inseminated or un-in- back to Khartoum for en masse egg laying (N= seminated females when data were combined

Table 10.5. Mean (± s.e.m.) wing Experiment Wild Laboratory lengths (mm) of males from com- irradiated†/ un-irradiated petition and control experiments. N N size N size is the number of wings measured. Mean values for wild and laborato- competition ry (i.e. irradiated and un-irradiated) are given. Means without letters in 1 8 2.47 ± 0.07 n/d common are significantly different 2 47 2.67 ± 0.02b 39 2.88 ± 0.02a† at p< 0.05 for each row (independ- ent t-tests or Tukey HSD). n/d: not control determined; n/a: not applicable 1 41 2.60 ± 0.02 n/a

2 7 2.75 ± 0.09a 16 2.73 ± 0.02a 157

3 25 2.74 ± 0.03a 14 2.77 ± 0.04a section A 25 2.70 ± 0.04a section C overall 128 2.65 ± 0.01a 94 2.78 ± 0.02b

10 Towards a SIT field release in Sudan

Experiment Section Females mean wing length (mm) ± s.e.m. Table 10.6. Mean (± s.e.m.) wing lengths (mm) of wild

Chapter 10 females recaptured from inseminated un-inseminated competition and control ex- N size N size periments. N is the number of wings measured. Wings competition were grouped according to insemination status. Means 2 11 2.86 ± 0.05a 6 2.83 ± 0.08a without letters in common are significantly different at control p< 0.05 for each row (inde- § pendent t-tests); except for 1 40 2.88 ± 0.02 2 2.91 ± 0.01 §, too few un-inseminated females to allow for statis- 2 A 25 2.94 ± 0.02a 6 2.90 ± 0.07a tical comparison. B 22 2.98 ± 0.03a 9 2.95 ± 0.06a

3 A 13 2.81 ± 0.05a 12 3.03 ± 0.03b

B 5 2.83 ± 0.10a 11 2.81 ± 0.08a

C 6 2.84 ± 0.13a 17 2.73 ± 0.06a

overall 122 2.90 ± 0.01a 63 2.86 ± 0.03a

(t(183)= 1.13, p> 0.05; Table 10.6). When data cluded as the likely cause of problems as simi- were analysed per experiment similar results lar experiments performed in a well-controlled were observed (Table 10.6), except for con- laboratory environment showed no impact of trol experiment three section A where the un- pupal irradiation on emergence or of adult cool- inseminated females were significantly larger ing on recovery (Helinski et al. 2006b). than the inseminated females (t(23)= -3.83, p< The transportation of adults to the 0.01; Table 10.6). field site by air in relatively small numbers was successful. For future release programmes, much larger numbers of insects will have to Discussion be transported, and the next step would be to scale up transportation for these kinds of We report here on the first series of attempts to numbers. The distance of the field site from irradiate and transport anopheline mosquitoes the irradiation source is likely to result in adult from the laboratory to a remote field site in the transportation even if pupal irradiation is per- context of a SIT study in Sudan. Logistically, the formed, and devices that can transport adults project in Sudan is a challenging one with time or allow for adult emergence during transport consuming travel between the rearing facility should be explored. The transportation and and the irradiation source and considerable dis- ground release of adults was performed in the tances to the field site. The irradiation of insects El Salvador release trials in the 1970s. A special performed in this chapter was the first irradia- “flat cage” was developed that could hold up to tion of Anopheles mosquitoes in Sudan or in Af- 2,000 adults and cages could easily be stacked rica overall. Some difficulties were observed for transport (Bailey et al. 1979). Mortality was 158 for pupal and adult stage irradiation, and these acceptable; however, the handling was inten- were associated with transportation and cool- sive and caused considerable stress to the mos- ing of the insects, and a general lack of expe- quitoes. Releases were difficult and, due to the rience with performing the experiments under weather conditions, adults had to be released local conditions. The irradiation process was ex- after sunset (Bailey et al. 1979). It is known that 10 Towards a SIT field release in Sudan

cooling can be used to slow down pupal devel- ing sites during the day. Climate data showed opment (Helinski and Knols 2008b), however, that even when ambient conditions were harsh, it remains to be tested if after irradiation the with high temperatures and low humidity, mi- development of the pupae can be slowed down cro-climates could be created where conditions successfully to allow for pupal releases. The re- were favourable for mosquitoes. Preferred rest- lease of pupae was also performed in El Salva- ing sites were the brick/clay structures and the dor. Pupae were released in cups or pans and left vegetation in the corners, and the kuseba, both to emerge in the field (Dame et al. 1974, Bailey in the east side of the field cage. Human land- et al. 1979, Lowe et al. 1980), and a cup could ing catches only caught few females when per- hold around 1,500 pupae. Cups were either put formed during the early evening (i.e. 8-10 pm). in floating containers that were released on wa- It remains uncertain whether this was due to ter surfaces of breeding sites or on land when the fact that all females had been recaptured placed in release shelters (Lowe et al. 1980). during daytime or that the period for collection Field cage experiments were success- was not suitable. In a study done in Ethiopia, the ful in demonstrating mating and survival of re- majority of An. arabiensis collected with human leased mosquitoes in the field cage. Data clearly landing catch were caught after 10 pm (Taye et indicated that mating occurred in the field cage, al. 2006), and similar observations were made in and a large proportion of females (i.e. on aver- Kenya (Githeko et al. 1996). However in another age 60%) was inseminated after only one or two study in Eritrea peak activity of An. arabiensis nights of mating. Wild males appeared to per- was observed between 8-10 pm and 1-3 am (Shi- form slightly better than laboratory males but lilu et al. 2004). The proportion of mosquitoes no significant differences were observed due recaptured after one night of mating was larger to the variation observed between replicates. than after two, with recapture rates for the fe- It is recommended that additional experiments males between 50-60%, and males between 25- are performed to understand the source of vari- 97%, although the latter value could have been ation. The observation that laboratory males overestimated by some males still present from were capable of inseminating wild females sug- a previous experiment. Recapture rates of labo- gests that no major behavioural differences due ratory males were similar to those of wild males. to the rearing process existed that impacted on A low recapture rate was observed in the first mating and this is of great importance for an competition experiment, and wild males and SIT project. To maintain this behaviour under females were strikingly smaller in size, suggest- mass-rearing conditions, introgression of wild ing that small insects suffered greater mortality material in the rearing colony, a strategy rou- in the field cage. Data on male survival in the tinely performed in other mass-reared insects field are scarce, but data from field-collected (Fisher and Caceres 2000), could be adopted. An. gambiae s.l. females suggested that larger There was no preference of males to mate fe- females had a higher probability of survival males of a particular size, and wing lengths of compared to smaller ones (Lyimo and Takken inseminated and un-inseminated females were 1993, Ameneshewa and Service 1996). The similar. This is in contrast to a laboratory study smaller size of the wild females in competition performed with An gambiae s.s. Giles, where experiment one was attributed to the fact that males were observed to select larger females as these females had been collected from a differ- mates (Okanda et al. 2002), but in agreement ent breeding site compared to the rest of the with a study performed in the wild where the females. Only few dead mosquitoes were found size distribution of mated An. funestus Giles fe- in the field cage, notably in the breeding sites, 159 males was similar to the distribution observed thus it remained uncertain if recapture rates re- after emergence (Charlwood et al. 2003). flected true survival. In some experiments, the Recapture of mosquitoes from the field day following recapture more mosquitoes were cage was virtually all done by sampling the rest- found, however, their numbers were low and 10 Towards a SIT field release in Sudan

we are reasonably certain that a large propor- strains (GSSs) have been developed for various

Chapter 10 tion of mosquitoes alive in the sections were insects including anophelines and they rely on recaptured. Experiments were performed in the linkage of a dominant selectable marker to close succession due to time limitations, but it the male determining chromosome or locus. is advisable in future experiments to allow more Linkage is accomplished by radiation-induced time between experiments. translocations followed by crossing and screen- Unfortunately, the number of eggs ob- ing of the offspring. Resistance genes, e.g. tem- tained from the two competition experiments perature sensitive lethal genes and insecticide- performed was too low to draw any meaningful resistance genes, have been used as selectable conclusions on the competitiveness of irradi- markers. A successful anopheline GSS was the ated males. This was due to the low number of MACHO strain of An. albimanus used in the females introduced and recaptured and prob- second trial at the Pacific coast in El Salvador lems associated with the feeding and egg lay- (Dame et al. 1981). Several GSSs for An. arabi- ing of wild females in the laboratory. In the first ensis are under development at the IAEA and experiment, two out of the three egg batches promising results were observed with some. obtained were classified as resulting from an ir- These strains will be evaluated for their use in radiated male, and the ratio of irradiated males SIT programmes in the near future. in this experiment was twofold compared to wild males. In the second experiment, larger numbers of fed females were obtained but egg Conclusion laying was low. Hatch rate of the eggs was only 11% suggesting that the 1-2 females that laid The irradiation and transportation of insects to eggs were inseminated by an irradiated male. the field site as performed in this study is fea- In future experiments, it is recommended that sible. Some difficulties were experienced with only one night of mating is used to increase the irradiation process (i.e. transportation and recapture rates. Preferably, a strain is used for cooling) but these will be greatly reduced in the competition experiments in which mating can near future when the new rearing facility is es- be detected by PCR or other molecular meth- tablished in close proximity to the irradiation ods (e.g. for instance the use of a (transgenic) source. In the meantime, lessons learned from genetic sexing strain (Catteruccia et al. 2005)), adult transportation to the field site can be ap- or alternatively stable isotopes can be used to plied to minimise mortality. Ultimately, the irra- label the semen (Helinski et al. 2007); thus ex- diation and transportation of much larger num- cluding the need for egg batches to determine bers of insects is needed for a release, and focus sterility. The survival of irradiated males was should lay on the development of such tools. comparable to wild males in experiment two, The field cage experiments demonstrated that and also in experiment one the wing length data mating occurred in high numbers and recovery suggested that most recovered males belonged of released insects was satisfactory. Irradiated to the irradiated males group. Thus it appeared males survived and mated in the field cage and that irradiation does not impact on survival in further experiments on the competitiveness of the first nights and this is important for future these insects should be pursued. studies on competitiveness. Insects in this study were separated to Acknowledgements sex by manual determination of either pupae or The authors are grateful for the assistance of R. 160 adults. For a large-scale SIT programme, this Taq-Elsir and D. Aldirdeery for the preparation method is obviously not feasible and automated of mosquitoes for irradiation and transport, to methods of sex-separation are required. In the S. Soliban for shipments of larvae and eggs, near future sex-separation will be performed to A. Hameed Nugud for his help during the using a genetic sexing strain. Genetic sexing experiments, to the staff of Sudan Atomic En- 10 Towards a SIT field release in Sudan

ergy Commission for operation of the gamma Release of chemosterilized males for the control of source, and to T. Ageep, M. El-Dirdeiry, T. Ad- Anopheles albimanus in El Salvador. II. Methods of ams, A. Hassan and F.i Mansour for their help rearing, sterilization, and distribution. Am. J. Trop. with the collections of wild mosquitoes, and Med. Hyg. 23:282-287. to M. Dicke, and two anonymous reviewers Dame, D. A., R. E. Lowe, and D. L. Williamson 1981. for constructive comments during preparation Assessment of Released Sterile Anopheles albimanus of the manuscript. Funding for this study was and Glossina morsitans morsitans, pp. 231-248. In provided by the Sudanese National Centre for J. B. Kitzmiller, and T. Kanda (eds.), Cytogenetics Research as part of the activities of the IAEA and genetics of vectors. Elsevier Biomedical, Regional Project (RAF 5/052), and through Amsterdam. the Technical Cooperation Department of the Dame, D. A. 1985. Genetic control by sterilized IAEA. BGJK is supported by a VIDI grant from mosquitoes, pp. 159-172. In R. Chapman, R. Barr, D. the Netherlands Organisation for Scientific Re- E. Weidhaas, and M. Laird (eds.), Biological Control search (#864.03.004). of Mosquitoes. Am. Mosq. Cont. Assoc., Bull 6. Dyck, A., J. Hendrichs, and A. S. Robinson 2005 (eds.). The Sterile Insect Technique: Principles and References Practice in Area-Wide Integrated Pest Management. Springer, Dordrecht. Ameneshewa, B., and M. W. Service 1996. The Fisher, K., and C. Caceres. 2000. A filter rearing relationship between female body size and survival system for mass reared genetic sexing strains of rate of the malaria vector Anopheles arabiensis in Mediterranean fruit fly (Diptera: Tephritidae), pp. Ethiopia. Med. Vet. Entomol. 10:170-172. 543-550. In K. H. Tan (ed.), Area-wide Control of Bailey, D. L., R. E. Lowe, J. E. F. Fowler, and D. Fruit Flies and Other Insect Pests, Joint Proceedings A. Dame 1979. Sterilizing and packaging males of of the International Conference on Area-wide Anopheles albimanus Wiedemann for field release. Control of Insect Pests and of the Fifth International Am. J. Trop. Med. Hyg. 28:902-908. Symposium on Fruit Flies of Economic Importance, Benedict, M. Q., and A. S. Robinson 2003. The first Penang, Malaysia, 1-5 June 1998. Penerbit Universiti releases of transgenic mosquitoes: an argument for Sains Malaysia, Penang. the sterile insect technique. Trends Parasitol. 19:349- Githeko, A. K., N. I. Adungo, D. M. Karanja, W. A. 355. Hawley, J. M. Vulule, I. K. Seroney, A. V. Ofulla, F. Catteruccia, F., J. P. Benton, and A. Crisanti 2005. K. Atieli, S. O. Ondijo, I. O. Genga, P. K. Odada, P. An Anopheles transgenic sexing strain for vector A. Situbi, and J. A. Oloo 1996. Some observations control. Nat. Biotechnol. 23:1414-1417. on the biting behavior of Anopheles gambiae s.s., Charlwood, J. D., J. Pinto, C. A. Sousa, C. Ferreira, Anopheles arabiensis, and Anopheles funestus and and V. E. Do Rosario 2002. Male size does not affect their implications for malaria control. Exp. Parasitol. mating success (of Anopheles gambiae in Sao Tome). 82:306-315. Med. Vet. Entomol. 16:109-111. Helinski, M. E. H., B. El-Sayed, and B. G. J. Knols Charlwood, J. D., R. Thompson, and H. Madsen 2006a. The Sterile Insect Technique: can established 2003. Observations on the swarming and mating technology beat malaria? Entomol. Berichten 66:13- behaviour of Anopheles funestus from southern 20. CHAPTER 1. Mozambique. Malar. J. 2:2. Helinski, M. E. H., A. G. Parker, and B. G. J. Knols Curtis, C. F. 2006. Review of previous applications 2006b. Radiation-induced sterility for pupal and adult of genetics to vector control, pp. 33-43. In C. Louis stages of the malaria mosquito Anopheles arabiensis. and B. G. J. Knols (eds.), Bridging laboratory and Malar. J. 5:41. CHAPTER 3. 161 field research for genetic control of disease vectors. Helinski, M. E., R. Hood-Nowotny, L. Mayr, and B. Frontis, Springer. G. J. Knols 2007. Stable isotope-mass spectrometric Dame, D. A., C. S. Lofgren, H. R. Ford, M. D. determination of semen transfer in malaria Boston, K. F. Baldwin, and G. M. Jeffery 1974. mosquitoes. J. Exp. Biol. 210:1266-1274. CHAPTER 10 Towards a SIT field release in Sudan

5. Akelo, Y. Touré, A. Odulaja, J. C. Beier, J. I. Githure,

Chapter 10 Helinski, M. E. H., and B. G. J. Knols 2008a. Mating G. Yan, L. C. Gouagna, B. G. J. Knols, and G. F. competitiveness of male Anopheles arabiensis Killeen 2002. Behavioural determinants of gene flow mosquitoes irradiated with a partially- or fully- in malaria vector populations: Anopheles gambiae sterilizing dose in small and large laboratory cages. males select large females as mates. Malar. J. 1:10. J. Med. Entomol. in press. CHAPTER 7. Shililu, J., T. Ghebremeskel, F. Seulu, S. Mengistu, Helinski, M. E. H., and B. G. J. Knols 2008b. H. Fekadu, M. Zerom, G. E. Asmelash, D. Sintasath, The influence of late-stage pupal irradiation and C. Mbogo, J. Githure, E. Brantly, J. C. Beier, and R. increased irradiated: un-irradiated male ratio on J. Novak 2004. Seasonal abundance, vector behavior, mating competitiveness of the malaria mosquito and malaria parasite transmission in Eritrea. J. Am. Anopheles arabiensis Patton. Bull. Entomol. Res. Mosq. Control. Assoc. 20:155-164. submitted. CHAPTER 8. Taye, A., M. Hadis, N. Adugna, D. Tilahun, and IAEA 2008. Development of the Sterile Insect R. A. Wirtz 2006. Biting behavior and Plasmodium Technique for mosquitoes. http://www. infection rates of Anopheles arabiensis from Sille, iaea.org/OurWork/ST/NA/NAAL/agri/ent/ Ethiopia. Acta Trop. 97:50-54. entMOSQUITOmain.php. Knols, B. G. J., B. N. Njiru, E. M. Mathenge, W. R. Mukabana, J. C. Beier, and G. F. Killeen 2002. MalariaSphere: A greenhouse-enclosed simulation of a natural Anopheles gambiae (Diptera: Culicidae) ecosystem in western Kenya. Malar. J. 1:19. Lofgren, C. S., D. A. Dame, S. G. Breeland, D. E. Weidhaas, G. Jeffery, R. Kaiser, H. R. Ford, M. D. Boston, and K. F. Baldwin 1974. Release of chemosterilized males for the control of Anopheles albimanus in El Salvador. 3. Field methods and population control. Am. J. Trop. Med. Hyg. 23:288- 297. Lounibos, L. P., N. Nishimura, J. Conn, and R. Lourenco-de-Oliveira 1995. Life history correlates of adult size in the malaria vector Anopheles darlingi. Mem. Inst. Oswaldo Cruz 90:769-774. Lowe, R. E., D. L. Bailey, D. A. Dame, K. Savage, and P. E. Kaiser 1980. Efficiency of techniques for the mass release of sterile male Anopheles albimanus Wiedemann in El Salvador. Am. J. Trop. Med. Hyg. 29:695-703. Lyimo, E. O., and W. Takken 1993. Effects of adult body size on fecundity and the pre-gravid rate of Anopheles gambiae females in Tanzania. Med. Vet. Entomol. 7:328-332. Malcolm, C. A., D. A. Welsby, and B. B. El Sayed 2007. SIT for the malaria vector Anopheles arabiensis 162 in Northern State, Sudan: an historical review of the field site, pp. 361-372. In M. J. B. Vreysen, A. S. Robinson, and J. Hendrichs (eds.), Area-Wide Control of Insect Pests. Springer, Dordrecht. Okanda, F. M., A. Dao, B. N. Njiru, J. Arija, H. A. 10 Towards a SIT field release in Sudan

163 11 General discussion 11

General discussion

he main findings of the research described in this thesis have been discussed at length at the end of each chapter. A brief overview of Tthe main research findings and conclusions can be found in Table 11.1. The work in this thesis centred around the research question of what the optimal stage and dose is for irradiation of Anopheles arabiensis males that will balance induced sterility with male fitness. From this overarching aim a number of objectives were formulated, and in this section the different findings are discussed and synthesised in more detail and specifically in the context of the Sterile Insect Technique (SIT) feasibility study in Sudan. 350-500 Million cases of clinical malaria occur annually, 60% of which are in sub- Saharan Africa. Moreover, 80% of all deaths attributed to malaria occur in this region. In numbers, 1 million Africans die of the disease each year, with the vast majority of deaths occurring among children below five years of age. Pregnant women are another major risk group; malaria can cause low birth weight and pre- mature delivery (Rogerson et al. 2007). The impact of malaria on the economic situation of endemic countries is high (Gallup and Sachs 2001), and the correlation between poverty and malaria clearly demonstrated (Sachs and Malaney 2002).

165 11 General discussion

Mosquito irradiation pae were collected within a time frame of 6 hrs

Chapter 11 (i.e. between 9 am and 3 pm). The handling of In the reported studies, insects were irradiated pupae for irradiation was easy, and no effects in a Cobalt-60 gamma source irradiator on a on adult emergence after the handling and irra- relatively small scale, and both for pupae and diation were observed (Fig. 3.2). The cooling of adults a maximum of approximately one thou- pupae for 24 hrs at 15-18 ºC had no influence on sand insects were irradiated at once. For these adult emergence either (Chapter 8). In contrast, relatively small numbers, it was attempted to adults needed to be inactivated during the han- maximise dose-uniformity in the irradiation dling and irradiation process to avoid excessive chamber by means of concentrating the insects damage. Adults were inactivated by cold treat- in a small volume in the centre of the chamber. ment, since the use of gasses like nitrogen and In future Sterile Insect Technique (SIT) pro- carbon dioxide have a major influence on the grammes like the feasibility study in Sudan, outcome of the irradiation process (see Chap- much larger numbers of insects need to be ir- ter 2). The cooling of adults for 4-5 min on ice radiated to achieve sufficient throughput. The did not result in increased mortality (Chapters 3 concentration of insects in the centre of the and 7). However, the transfer of adults to cool- chamber cannot be maintained, and accord- ing containers was time-consuming when larg- ingly insects will receive variable doses of irra- er numbers of adults were irradiated. If adult ir- diation. The optimal dose in an SIT programme radiation would become the method of choice, is thus not a dose but rather a dose range. The ideally a container should be developed that is largest number of mosquitoes that can be irra- suitable for irradiation as well as subsequent diated simultaneously with existing devices is transport to the release site. approximately twenty thousand pupae (Bellini et al. 2007) and ~ 7,000- 14,000 adults (Smittle and Patterson, 1974, Curtis 1976). For compari- Dose-sterility son, in Mediterranean fruit fly Ceratitis capitata Wiedemann SIT programmes, 1.8 litres of pu- The dose-sterility curves for male Anopheles pae can be irradiated simultaneously, which arabiensis Patton irradiated in the pupal or corresponds to ~100,000 flies (i.e. with an adult stage were largely similar (Fig. 3.4), al- emergence of 85%; C. Caceras, personal com- though it appeared that pupae were slightly munication). Therefore, new devices that can more radio-resistant at certain doses (i.e. be- accommodate larger numbers of mosquitoes tween 60-80 Gy; Table 3.2), and this became should be developed. New approaches in insect apparent when females were placed in individ- irradiation include the use of a new generation ual vials for egg laying (Fig. 7.2). Compared to of X-ray machines that can generate a much other anophelines, An. arabiensis appeared to higher dose rate than conventional machines be slightly more radio-resistant (Fig. 3.4). Dose- (i.e. 45 Gy/min; Radsource, Alpharetta, GA, sterility curves of male An. arabiensis followed USA). One of these machines is currently being the classical pattern for irradiation damage in evaluated by the International Atomic Energy germ cells of a linear relationship at lower doses Agency (IAEA). These devices are likely to re- and the flattening of the curve at higher doses place gamma source irradiators in the future, as (LaChance and Graham 1984, Robinson 2002). these are no longer being manufactured, and Fertility was plotted on a logarithmic scale problems associated with transportation and against irradiation dose in Chapter 3 to provide 166 disposal of radioactive material are becoming insight in the nature of the dominant lethal increasingly difficult. mutations (Fig. 3.3). A predominantly linear re- For the irradiation of pupae, synchro- sponse was observed for both stages at lower nised development is required to obtain pupae doses suggesting a “one-hit” relationship, i.e. of a similar age, and in the reported studies pu- one mutation in the cell (LaChance 1967, Cur- 11 General discussion

tis 1971). At the higher doses the lines departed stage with 70 Gy, this phenomenon was less from linearity suggesting that gametes carried clearly observed (Fig. 7.2). At 120 Gy, no egg more than one dominant lethal mutation, as is batches with higher than average hatch were routinely observed in other insects (LaChance observed for pupal or adult radiation (Chapter 1967, LaChance et al. 1967, Curtis 1971). The 7; Fig. 7.2). However, after the irradiation of old- sterility in Chapter 3 was presented as induced er cooled pupae with 120 Gy in Chapter 8, again sterility, i.e. corrected for the control sterility some egg batches with higher hatch were ob- naturally occurring in the population using Ab- served. The biological relevance of these data bott’s formula (Abbott 1925). The level of ob- is not clear. As suggested in Chapter 9, they served sterility was thus slightly higher; e.g. for could potentially indicate some radio-resist- 70 Gy 2-5% higher for adult or pupal stage irra- ance in certain pupae at the time of radiation. diation, respectively. In Chapters 7 and 8, steril- This would be interesting to study in future ex- ity in individual egg batches was not corrected periments, but it is complicated by the fact that for control sterility as the regression model was single-pair matings in An. arabiensis are difficult built from the observed sterility levels for egg to perform (M. Benedict, unpublished data). To batches laid by females inseminated by un-irra- determine if pupal age at the time of irradiation diated or irradiated males. is a cause for these findings, pupae could be ir- Lethality after a dominant lethal mu- radiated in age groups collected over a much tation often occurs during early development shorter period of time (e.g. collection of pupae of the embryo (LaChance 1967). In Chapter 3, every 30 min), followed by determination of lethality was scored by counting L1 stage lar- sterility levels in these males. vae developing from a defined number of eggs. However, for the adult stage irradiation experi- ments eggs where also examined microscopi- Sperm, and the influence of cally for hatching and no differences were ob- irradiation served in hatch rate when data were compared between both methods (data not shown). Be- The Sterile Insect Technique relies on the trans- cause the microscopic examination of the eggs fer of sterile gametes to the females and the was less time-consuming, in subsequent Chap- subsequent use of these gametes for fertilisa- ters this was the method of choice. In this thesis tion of eggs. The quality and quantity of the work, no attempt was made to determine the semen transferred by irradiated males is impor- precise moment of lethality in the developing tant, especially so in anophelines, where it was embryo. hypothesised that a spermatheca containing Sterility data were obtained by en semen is the trigger for females to change their masse egg laying in Chapter 3. Even though behaviour from mating to blood feeding and this method accurately reflects the effects at egg laying (Klowden 2006). the cage population level, individual differ- The irradiation of pupae resulted in ences cannot be determined and therefore in the production of fewer spermatozoa com- further competition experiments, egg laying pared to un-irradiated pupae whilst males irra- of individual females was the method of choice diated as adults had similar amounts of sperm (Chapters 7 and 8). The individual egg batches as un-irradiated counterparts (Fig. 4.1b). In the of females mated to irradiated males showed pupal stage, earlier stages of spermatogenesis some interesting results. While on average the are present compared to the adult stage, and observed sterility in females mated to males ir- the irradiation of these stages can result in cell 167 radiated with 70 Gy in the pupal stage was 86 ± death. This phenomenon most likely accounted 1%, a substantial number of egg batches were for the reduction in sperm cells after pupal ir- observed with a higher than average hatch rate radiation and similar results were observed in (Fig. 7.2). For irradiation of males in the adult the Mediterranean fruit fly (Anwar et al. 1971). 11 General discussion

Table 11.1. Overview of the main research findings and conclusions. Research question Ch Findings Conclusions What are the effects of 3 • Dose-sterility curves for pupae and adults show a linear • Pupal and adult stage irradiation on irradiation on sterility, relationship at lower doses and the flattening of the curve at a small-scale is feasible. Chapter 11 emergence and longevity? higher doses. • Mating ability of males irradiated as • Adult emergence was not affected by irradiation of the pupal pupae with high doses was slightly stage. compromised. • Survival of irradiated males was largely similar to un-irradiated • An. arabiensis was quite radio- males. resistant compared to other • Reduced insemination after pupal stage irradiation at higher anophelines. doses was observed.

Does sperm length 4 • The quantity of sperm in the testes increased with age; 6-day- • Sperm quantification in Anopheles morphology exist in An. old males had on average 4950 ± 848 spermatozoa in their is feasible, but methods can be arabiensis and what is the testes. optimised. quantity of sperm in the • Sperm length polymorphism was present (< 50-500 µm), and • Sperm length polymorphism is testes? distributions over length classes were similar to wild males. present, and the role of • A significantly greater proportion of longer sperm were present polymorphisms in the anopheline in spermathecae compared to the testes. mating system is discussed.

What are the effects of 4 • Pupal irradiation resulted in significantly less sperm, and • Irradiation of the pupal stage irradiation on sperm length proportionally more sperm of the smaller category compared resulted in fewer spermatozoa due morphology and quantity? to un-irradiated males. to the irradiation of the early stages • For adult irradiated males no difference in the distribution of of spermatogenesis. sperm polymorphisms and number of sperm compared to un- • Implications of these findings for the irradiated males was observed. SIT are discussed.

Can stable isotopes be used 5, 6 • Chapter 5: 13C-glucose can be used to label semen; a • Both labels can be used to study to label semen of An. substantial loss of label during fixation was observed. mating behaviour. arabiensis males? • Chapter 6: 15N-glycine can be used to label semen; no loss of • The use of 13C is preferred for label was observed but sample analysis was more sample analysis; however, 15N is complicated. preferred for larval development. • No impact of either label on life-history traits as larval and • The use of 15N was approx. 5 times adult survival, and mating ability were observed. cheaper than the use of 13C.

What is the competitiveness 7 • Adult irradiation resulted in better competitiveness than pupal • Adult irradiation gave the best of pupal or adult stage stage irradiation. results in terms of mating irradiated males when • Males irradiated as pupae with the partially-sterilising dose (70 competitiveness. competing against un- Gy) had a better competitiveness than males irradiated with • Lower doses resulted in better irradiated males at a 1:1 ratio the fully-sterilising dose (120 Gy). competitiveness for pupal stage in small and large cages? • Mixed results were obtained for large or small cage irradiation. experiments. Males irradiated as pupae with 70 Gy performed • Cage size affected competitiveness better in small cages compared to larger cages, while males for some treatments. irradiated as adults with 120 Gy had a better competitiveness • Experiments should be repeated in large cages compared to small cages. under semi-field conditions.

Can the reduction in pupal 8 • A three-fold increase in the ratio of irradiated males resulted in • The loss in competitiveness after competitiveness be a better competitiveness compared to the 1:1 ratio pupal stage irradiation can be overcome by a three-fold experiments for the partially-sterilising dose but not for the overcome by a threefold increase in increase in the ratio of fully-sterilising dose. the ratio of irradiated males for the irradiated males, or by • Irradiation of older pupae did not result in a significant low (70 Gy) but not for high (120 irradiating older pupae? improved competitiveness compared to young pupae. Gy) dose. • The cooling of pupae for 24 hrs had no effect on adult • Cooling might be a useful tool in emergence or male longevity. operational SIT programmes to facilitate handling processes.

Does multiple insemination 9 • Stable isotopes were successful in detecting multiple • Dual-labelling approach of stable occur using a stable isotope insemination events. isotopes can be used to study dual-labelling approach, and • After 4 nights of competition, 25% of the females were multiple insemination. what is the influence of inseminated multiply. • Studies should be repeated in semi- radiation on the incidence of • Irradiation had no effect on the incidence of multiple field cages to exclude space issues multiple insemination? insemination. as the possible cause of polyandry. • Competitiveness of pupal irradiated males lower than un- • The level of 15N enrichment should irradiated males on a 1:1 ratio in small cages. be increased in future studies to obtain a better spread of the data.

Is the small-scale study of 10 • Minor problems, associated with a lack of experience, were • Small-scale irradiation and SIT in Sudan feasible? observed with the irradiation process. transportation of insects in Sudan is • The transportation of small numbers of adult mosquitoes by feasible. air to the field site resulted in minimal mortality (i.e. < 6 %). • Scaling up of procedures to accommodate larger numbers of insects is the next challenge.

Can the semi-field system in 10 • Mating occurred in the field cage and laboratory reared males • Mating occurred in the field cage Sudan be used for fitness were able to inseminate wild females at rates comparable to and survival was good, indicating studies? wild males. that the cage is suitable for • Survival of wild and laboratory males and females in the field experiments. 168 cage was good. • Studies to test the competitiveness • No data on competitiveness was obtained due to the low of irradiated males can be pursued. number of eggs collected.

11 General discussion

Another parameter of sperm was studied, i.e. males for fertilisation (Curtis 1985). In future ex- sperm length. Sperm length polymorphism was periments, sperm use by multiply inseminated present in An. arabiensis and the sperm length females would be an important parameter to distributions observed in laboratory males were determine. similar to those observed in wild males (Fig. Although not examined in great de- 4.2.). Irradiation of males in the pupal stage had tail, the sperm from irradiated testes appeared an effect on the distribution of sperm lengths, similar in motility compared to the sperm from and significantly more smaller sperm in the cat- un-irradiated males. Data from other insects egory of 100- 200 μm were observed in these show that inactivation of sperm does not occur males compared to un-irradiated males. For until complete dominant lethality is attained adult irradiated males this difference was not (LaChance et al. 1967). Studies performed in observed (Fig. 4.2.). The implications of these Aedes aegypti L. did not observe a loss of sperm findings for the SIT are not fully understood. It motility even when the dose applied was far remains to be determined whether irradiated higher than needed for sterilisation (i.e. 300 males transferred fewer sperm than un-irradi- Gy; Terzian and Stahler 1958). Sperm inactiva- ated males. However, at present, this analysis tion by irradiation might also result in a loss of is hampered by the difficulty to quantify sperm fertilising capacity (LaChance et al. 1967). In after transfer to the spermatheca (Chapter 4). Chapter 4 it was observed that irradiated males In addition, the distribution of sperm lengths in produced more sperm of the smaller category females mated to irradiated males was not de- (100-200 µm) while females harboured signifi- termined, and it remains unknown if the greater cantly more sperm cells of the larger categories proportion of small sperm in the testis of pupal in the spermathecae (300-400 µm). The role of irradiated males can also be observed in sper- short sperm in fertilisation is not understood in mathecae after mating. anophelines but in Drosophila short sperm was To determine if a female was more not used for fertilisation of the eggs (Bressac likely to engage in another mating after receiv- and Haushteck-Jungen 1996). In contrast, in ing irradiated sperm, the incidence of multiple other genera fertilisation success was biased mating was determined (Chapter 9). Multiple towards males with relatively short sperm (e.g. mating occurred frequently (i.e. 25% of all fe- in the dung beetle Onthophagus taurus Schreb.; males were mated multiply) in standard rearing Garcia-Gonzalez and Simmons 2007). Also in cages under laboratory conditions. Irradiation experiments with the cricket Gryllus bimacula- did not appear to impact on the frequency of tus de Geer in double-male insemination exper- multiple insemination in small laboratory cag- iments males producing numerous and small es, and in experiments where irradiated males sperm won the competition for fertilisation competed with un-irradiated males for mates (Gage and Morrow 2003). Thus experiments to the frequency of females inseminated by two understand the role of short sperm in anophe- males was similar compared to a control treat- line fertilisation are urgently needed. ment without irradiation (Table 9.1). However, The driving force behind the pres- a review of the literature indicated that these ence of sperm polymorphism in other species is results might be an overestimation of the true thought to be competition among the sperm of proportion of multiple insemination in the wild, multiple males in the female (Clark 2002). How- most likely due to space restrictions in labora- ever, in Anopheles, females appear to be largely tory cages (Chapter 9), and it was suggested to monandrous in the wild (Tripet et al. 2003, Yu- repeat these experiments in semi-field systems val and Fritz 1994) and little competition in the 169 to exclude the influence of space. For SIT pro- spermathecae is therefore expected. A possible grammes, multiple insemination is not consid- role of sperm length polymorphism in specia- ered to be a drawback as long as the sperm of tion processes in anophelines was implicated released males is equally competitive to wild after it was observed that species belonging to 11 General discussion

species complex groups had a greater range of 6). After semen transfer the label could be iden-

Chapter 11 sperm lengths compared to the ones that be- tified in the spermathecae of the females using long to firmly demarcated taxa (e.g. An. free- mass spectrometry. The merits and pitfalls of borni Aitken; Klowden and Chambers 2004). The both labels were discussed in Chapters 5 and sperm length distributions in An. arabiensis was 6. In the experiments reported in Chapter 9, very similar to the distribution observed in the 13C and 15N were used simultaneously to study closely related sibling species An. gambiae s.s. multiple mating events. This approach was suc- Giles (Chapter 4; Klowden and Chambers 2004). cessful in identifying spermathecae with both This appears to contradict with the above hy- labels. However, for 15N substantial variation in pothesis of speciation, as both species belong unlabelled spermathecae was observed, and it to the same species complex (i.e. An. gambiae was recommended to increase the amount of s.l. species complex, see Hunt et al. (1998)) and 15N label in future studies. This conclusion could a different distribution might have been- ex only be drawn after all experiments had been pected. However, it is not unlikely to assume completed (i.e. due to delays in sample analysis that optimal sperm length polymorphisms at the mass spectrometry facilities) thus no cor- might have evolved similarly for both species rections could be made at the time the experi- (M. Klowden, pers. communication). The study ments were conducted. of Klowden and Chambers (2004) also observed In this thesis, two different approaches that females of species that had polymorphic to analyse paternity, for instance in competi- sperm lengths tended to have larger spermath- tion experiments, were used. Individual egg ecae, which could give females the opportunity batches were collected in Chapters 7 and 8 and for female cryptic choice (i.e. the ability of fe- hatch rates were determined, while in the ex- males to manipulate the outcome of fertilisa- periments in Chapters 5, 6 and 9 stable isotopes tion by selectively choosing which sperm is were used. Both approaches have certain ben- used). The occurrence of female cryptic choice efits and limitations. For determining insemina- and sperm polymorphism has been observed tion, the accuracy of stable isotopes was high in a number of species and is usually based on and paternity could be assigned with 99.7% polyandrous mating systems (Snook 2005). confidence (Chapters 5 and 6). Determination Clearly, more studies are needed to determine of paternity of individual egg batches was as- the role of sperm polymorphism in the largely signed based on hatch rates using logistic re- monandrous Anopheles species. Only then can gression analyses. This analysis was complicat- the possible implications for a SIT programme ed by the overlapping distributions of the con- be determined. trol data (i.e. in the absence of competition, Fig. 7.2). The collection of egg batches and determi- nation of hatch rate is straightforward but time Stable isotopes in mating consuming but no additional costs are made. behaviour research Stable isotopes on the other hand are relatively expensive, but their use is less time consuming. One of the objectives of this thesis was to de- Multiple mating can be studied directly by us- velop novel tools to study mating behaviour in ing isotopes, however the indirect measure, i.e. anophelines. In Chapters 5 and 6 the use of the the proportion of egg batches with intermedi- stable isotopes 13C and 15N as semen (i.e. sperm ate hatch, appeared to be quite a good indica- and accessory gland fluids) labels was reported. tor of multiple insemination in small laboratory 170 Stable isotopes were chosen because they are cages (Chapters 7 and 9). An additional advan- non-invasive, safe for use, relatively cheap, and tage of the stable isotope semen labels is that can be easily applied. Semen could effectively the males themselves are highly labelled, thus be labelled in the larval stage by applying the creating opportunities to study, besides mat- label to the larval rearing water (Chapters 5 and ing behaviour, a number of parameters that are 11 General discussion

important for control programmes, such as lon- irradiated mosquitoes. Under more challenging gevity, ratios (i.e. released versus wild males) conditions, i.e. in the field cage in Sudan (see and dispersal, if these males would be released below), laboratory males had a similar survival in the field. The quantities of the labels used will compared to wild males (Table 10.3) when re- result in a life-long isotopic signature, as in pre- covered after one night of mating. The survival vious studies much lower quantities of 13C could of males irradiated as pupae with approximate- be detected up to 21 days after emergence ly 60 Gy was similar compared to the survival (Hood et al. 2006). of males from the wild after two nights of mat- ing (Table 10.4). However, this observation was based on only one replicate, and more data on Male mating competitiveness survival in relation to irradiation doses in the field cage are needed. The fitness of released males is of critical im- Male mating competitiveness experi- portance for SIT programmes. It should be not- ments were performed where irradiated males ed that fitness of released males in the context competed with un-irradiated males for females of a SIT study is not the same as the fitness of (Chapters 7, 8 and 9). A number of parameters organisms from a biological evolutionary point that could influence competitiveness were of view (i.e. a measure of the relative contri- studied. These included developmental stage bution of an individual to the gene pool of the during irradiation (pupae or adult), radiation next generation), as in that context the fitness dose (i.e. partially (70 Gy) or fully-sterilising of fully-sterilised males would be zero. Thus, fit- (120 Gy)), cage type (small or large cages), ness in SIT studies is defined as the ability of the and the ratio of irradiated versus un-irradiated released males to survive, find mates, compete males used (1:1 or 3:1). When insects were in- with wild males and successfully inseminate troduced on a 1:1:1 ratio (Chapter 7), males ir- the females. Parameters important for the fit- radiated as adults had a better competitiveness ness of wild males need to be tested in labora- than males irradiated as pupae (Table 7.1), and tory males to determine how “fit” they are in were equally competitive compared to un-irra- comparison to the wild males. However, male diated males when tested in large cages (Table Anopheles biology, especially in terms of mat- 7.1, Fig. 7.3). In small cages, males irradiated as ing, is essentially a black box (Ferguson et al. adults with 120 Gy had a lower competitiveness 2005). It is known that within a swarm, the male compared to un-irradiated males (Table 7.1) to female ratio is highly biased towards males while for 70 Gy this was not observed. Males ir- and thus fierce competition for females occurs radiated as pupae had a lower competitiveness (Charlwood and Jones 1980). However, the pa- compared to un-irradiated males (Table 7.1) rameters that make a male successful in obtain- and this was more pronounced for males irradi- ing a mate remain largely unknown. Recently, a ated with the high dose. The hypothesis that in number of studies on male biology and behav- the large cages males would be challenged to a iour (Huho et al. 2006, 2007, Ng’habi et al 2006, larger extent compared to the small cage (e.g. 2008) have been performed that will ultimately a larger swarming arena that necessitated in- lead to a better understanding of male fitness. creased flight activity to perform matings) that The parameters used in this thesis to determine would result in a reduced competitiveness was male fitness were survival and mating competi- only observed for pupal irradiation with 70 Gy tiveness. (Fig. 7.3). Because cage size affected competi- Survival of irradiated males in labora- tiveness for some treatments, competitiveness 171 tory cages was largely similar to that of un-irra- determined in laboratory experiments must diated males in Chapter 3 (Table 3.1). However, be confirmed by releases into semi-field con- in Chapter 8, the irradiation of pupae with 120 ditions where irradiated males are challenged Gy resulted in a lower survival compared to un- 11 General discussion

with wild males and compete for wild females 2005). Acclimatisation of insects to laboratory

Chapter 11 (Knols et al. 2002). conditions may select for certain traits that are The reduction in competitiveness after beneficial to the artificial laboratory setting, pupal stage irradiation could be compensated but most likely compromise fitness when labo- for by a three-fold increase in the ratio of irradi- ratory males are released back into the field, ated males for the lower dose of 70 Gy (Table as was shown for the Mediterranean fruit fly in 8.2). However, for the higher dose competitive- competitiveness assays when comparing mass- ness remained low (Table 8.2). The irradiation reared strains to wild-type strains (Cayol 2000). of older pupae as performed in Chapter 8 did The genetic sexing strain (GSS) used to remove not result in a dramatic improvement in com- females from the release population could also petitiveness, and no significant results were have an impact on fitness. However, in mos- observed compared to the irradiation of young quitoes, competitiveness of genetically altered pupae (Fig. 8.2). It can be concluded that the male mosquitoes was high in Culex pipiens L. competitiveness of males irradiated as pupae, under natural conditions (Grover et al. 1976a), even when performed minutes away from and in Ae. aegypti under both natural (Grover et adult eclosion, resulted in a lower competitive- al. 1976b) or field cage conditions (Grover et al. ness compared to adult irradiation as a larger 1979). Nevertheless, males from the MACHO number of irradiated males were needed to strain of An. albimanus Wiedemann released in achieve a similar level of competitiveness (Ta- El Salvador proved competitive under field con- ble 7.1 and 8.2). ditions (Kaiser et al. 1979). The mating competitiveness of stable isotope labelled irradiated males in Chapter 9 was in line with previous results (Chapter 7), and A SIT feasibility study in Sudan irradiated males were less competitive than un- irradiated males for the majority of replicates The work performed in this thesis was part of (Table 9.1). For 70 Gy, the proportion of females a larger study to determine the feasibility of inseminated by un-irradiated males was similar the SIT against Anopheles arabiensis in North- between both studies, i.e. 58 ± 5% and 55 ± 7% ern State, Sudan (Chapter 10). Logistically, for Chapter 7 and Chapter 9, respecively. For 120 the project in Sudan is challenging with large Gy in Chapter 9 replicates were not statistically distances between the rearing and sterilisa- significant and could therefore not be grouped. tion facilities in Khartoum and the release site However, the proportion of females inseminat- in Northern State. However, on a small-scale, ed by un-irradiated males appeared only slight- insects could be irradiated and transported to ly lower compared to the 75 ± 5% observed in the release site without much difficulty (Chap- Chapter 7. Two small-scale competition experi- ter 10). The scaling up of insect numbers cou- ments with partially-sterilised males were per- pled to a reliable and consistent production line formed in the field cage in Sudan. Regretfully, is the next challenge. This will be facilitated by very limited data were collected due to the low the use of a new rearing facility in close proxim- number of eggs obtained (Chapter 10), and no ity to the irradiation source. Commercial flights predictions on male competitiveness under were used to transport insects to the field site semi-field conditions could be made. but these flights are only available 2-3 times per Loss of fitness (i.e. survival and com- week, which poses important restrictions on petitiveness) of irradiated males in an SIT con- the flexibility of a release programme. 172 text is not merely due to the irradiation proc- The field cage in Dongola was pre- ess. The entire production process including pared for experiments by creating resting sites colonisation, mass production, sexing, irradia- with suitable climatic conditions where mos- tion and handling, has an influence on the qual- quitoes could survive (Fig. 10.2). The most im- ity of the males produced (Calkins and Parker portant observation from the field cage experi- 11 General discussion

Figure 11.1. Overview of the different components of a Sterile Insect Technique (SIT) programme. Filled boxes represent the essential steps needed for the release and monitoring of sterile males. On their right is the contribution of the work performed in this thesis indicated. The open boxes represent in random order the additional components of importance to a SIT programme. ment was that mosquitoes were observed to discussed in Chapter 10). It was therefore sug- mate. Wild mosquitoes mated in high numbers gested that other methods, for instance stable in the cage (Table 10.3), indicating that field- isotopes (Chapters 5 and 6) should be used that cage conditions simulated natural conditions will facilitate data collections in these kind of and the cage was suitable for competitiveness experiments. studies. Laboratory-reared males were able to Mosquitoes in the field cage were re- inseminate wild females at comparable rates to leased from holding cups. In a future release wild males (Table 10.3), which indicated that no programme in Sudan, strategies need to be major behavioural differences due to the colo- developed that can release large numbers of in- nisation process had occurred. Survival of the sects in designated places. The ground releases mosquitoes in the field cage was good (Tables performed for pupae and adult An. albimanus 173 10.3 and 10.4) but the number of wild females mosquitoes in the El Salvador release trials collected from competition experiments was were reviewed in Chapter 10. Briefly, ground re- not sufficient to obtain adequate numbers of leases required good access to the release sites, egg batches to allow statistical analyses (as and a large number of personnel to perform the 11 General discussion

releases (Dame et al. 1974, Bailey et al. 1979, reviewed by Silver (2008)). Observed dispersal

Chapter 11 Lowe et al. 1980). Pupal releases were pre- distances by females differ between studies. ferred over the release of adults, as adults had In Tanzania the large majority of An. gambiae to be released after sunset since release during s.l. females (laboratory reared specimens) re- daytime temperatures had a negative effect on leased were recaptured within 3.2 km of the survival (Bailey et al. 1979). In the field site in release site (Gillies 1961), while the majority of Sudan, mosquito breeding sites are limited to recaptures in another study along the Kenyan an area of approximately 5 km on either side coast with An. gambiae and An. funestus Giles of the Nile (Chapter 10). It has been suggested (collected as larvae from natural breeding sites) to release mosquitoes directly from the Nile to occurred within 400 m of the release site (Mide- cover the populations closest to the riverbanks ga et al. 2007), while mosquitoes dispersed a (C. Malcolm, personal communication). Besides bit further in a study with An. gambiae s.l. in ground releases, insects in SIT programmes can Tanzania (i.e. 400-1000 m; Takken et al. 1998). also be released by air. Aerial releases, although However, in both studies mosquitoes were only never tried with mosquitoes, have a number of recovered close to the release sites (i.e. with- potential benefits over ground releases (Dow- in approx. 1 km), and ad-hoc collections in a ell et al. 2005). The release sites can be further neighbouring village (3 km) recaptured another away from the facilities, extending the geo- two females (Takken et al. 1998). A study that graphical scope of the operation greatly. The investigated male dispersal observed a maxi- need for good ground access to the field sites mum flying distance of 3.2 km in An. gambiae is no longer valid for daily releases, although s.l. in Tanzania (Gillies 1961). In a study with An. for monitoring purposes it would still be desir- gambiae s.l. mosquitoes in a savannah area in able. In addition, the number of staff required Burkina Faso only four males were recaptured for aerial releases is lower, and aerial releases of which one An. arabiensis male had dispersed can benefit from existing on-board navigation the furthest (i.e.1 km; Constantini et al. 1996). equipment to accurately release the mosqui- To determine male and female dispersal, adult toes in designated areas. However, costs asso- population structure, and adult population size ciated with aerial releases are high, and landing in the release site in Northern State, mark-re- strips/platforms etc. need to be in place (Dow- lease-recapture studies should be performed. ell et al. 2005). Aerial releases are performed in A study on the dry-season biology of An. ara- the large Mediterranean fruit fly programmes bienis populations in two areas in Sudan indi- where flies are kept immobile during- pack cated that in the arid area mosquitoes survived ing and transport by chilling, and are released as nulliparous females at very low population through the bottom of the aircraft (Dowell et densities, while in the valley of the White Nile al. 2005). However, unlike the robust Mediter- year-round breeding at low-level was observed ranean fruit flies, mosquitoes are rather fragile (Omar and Cloudsley-Thompson 1968, 1970). A creatures. Handling, packing and release meth- study performed in the Sahelian region of Sen- ods for mosquitoes need to be developed and egal concluded that An. arabiensis populations tested to assess the impact of aerial release on maintained themselves throughout the dry male behaviour and longevity (Dame and Curtis season in a scenario described as diffuse-deme; 1996). i.e. permanent populations that pass the dry For SIT release programmes, the dis- season in low numbers in each dwelling (below persal rate of the released males, and of the the detection limit of sampling methods), but 174 wild population is of critical importance, in ad- across their distribution area overall total num- dition to detailed knowledge on the adult popu- bers are high (Simard et al. 2000). lation size and structure. The large majority of Males in the field cage could be recov- dispersal studies in African anophelines have ered by extensive searches of the cage. How- evidently been undertaken with females (as ever, the collection of males from the field is 11 General discussion

complicated as no specific tools are available optimal irradiation dose for An. arabiensis, it is to capture male anophelines in the field. Nei- imperative that the optimal dose should bal- ther do we have information on the habitat ance sterility with fitness. Results presented use of males that would facilitate developing here show that the fitness of males irradiated male trapping methods (Ferguson et al. 2005). as adults is superior over pupal stage irradia- However, some preliminary data on mating be- tion under laboratory conditions. Specifically, haviour of An. arabiensis in Northern State in- for pupal stage irradiation it was observed that dicated that swarms can be regularly observed at a higher dose longevity and competitiveness (O. Seidahmed, personal communication) and appeared to be compromised more than at a these sites can be used to sample males. For- lower radiation dose. Thus, for pupal irradiation tunately, the most important parameter in any it is advised that radiation doses are carefully genetic control programme (i.e. the ability of selected to avoid the use of higher than neces- the released males to inseminate wild females), sary doses. However, adult irradiation would be can be assessed by the trapping of wild females preferred in terms of competitiveness and de- (Vreysen 2005). Generally, females are collect- vices that can be used to irradiate and transport ed in the field, transported to the laboratory adults to release sites should be developed. for egg laying, and egg hatch (or another out- However, the same question posed put depending on the mechanism of control) from the perspective an SIT programme for is determined. Alternatively, oviposition sites Anopheles is far more complicated. The opti- in the wild, either existing or man-made (i.e. mal development stage for irradiation within ovitraps; Reiter et al. 1991), can be monitored, an SIT programme also depends on the feasi- but this method has not been developed yet for bility of large-scale handling methods, which anophelines. in itself depends on the characteristics of the In this thesis only one aspect of an release site (i.e. distance, release methodol- SIT programme, the irradiation of insects on a ogy), and the numbers of insects to be released small-scale, and the competitiveness of these on a weekly or even daily basis. The number of insects under primarily laboratory studies, were insects that need to be released depends on the investigated. Large-scale irradiation devices competitiveness of the males, the natural pop- still need to be developed, and competitive- ulation density and the desired release ratio. ness of irradiated males under (semi-) field The better the competitiveness of the males conditions should be thoroughly tested. How- under field conditions, the fewer males are re- ever, only when all other components of an quired. Based on our findings, males irradiated SIT programme, i.e. mass production, sexing, as pupae had a lower competitiveness than release methodology, and characterisation of males irradiated as adults and larger numbers the release site among others (Fig. 11.1), are of males would need to be released if pupal ir- in place can the true feasibility of the tech- radiation would be preferred over adult irradia- nique be established. Good progress has been tion. The release ratio of released versus wild made over the last years in all of these areas, males needed for anopheline SIT programmes and the implementation of these techniques to is not known at present. However, it is desirable operational levels is the next challenge before to initiate releases when the wild population the feasibility of SIT in Sudan and other African density is either naturally low or artificially re- countries can be established. duced through an additional intervention (Man- gan 2005). Natural mosquito populations often follow a seasonal pattern, and releases can be 175 The optimal stage and dose? initiated when the adult population is at its low- est, a technique practiced in the El Salvador To return to the central research question, i.e. release trials (Dame et al. 1981). Alternatively, determination of the most suitable stage and the active reduction of the wild population 11 General discussion

prior to release is performed, and this is usually In the case of Anopheles, eradication

Chapter 11 achieved by insecticides (Mangan 2005). The of all vectors of malaria by means of the SIT optimal dose, or more accurately, the optimal alone seems highly unlikely to achieve on the dose-range, depends on the level of competi- African continent due to its species-specifity tiveness observed, coupled to a threshold of and the vast areas that are difficult to separate observed residual fertility that is acceptable in a from neighbouring untreated sites. In most of given programme. If insects are not fully sterile, sub-Saharan Africa at least three species con- then this simply reduces the rate at which the tribute to malaria transmission (i.e. An. gambi- population is suppressed (Robinson 2002). ae s.s., An. arabiensis, and An. funestus), and in Currently, any discussions on costs are some areas even up to five (incl. the An. nili and highly speculative as no data exist on the costs An. moucheti groups). In addition, the existing of mass-production of anophelines, the costs gene-flow barriers within some species, most of the actual implementation of an SIT pro- importantly within An. gambiae s.s. (della Torre gramme (including releases and surveillance), 2002) poses important restrictions of the use of and the fixed costs associated with running an SIT (Chapters 1 and 2). Only in certain (urban) operational programme. Cost-benefit analyses island settings or localities on the edges of the of existing SIT programmes were performed for distribution of the species where a single vec- the Mediterranean fruit fly, tsetse flies, screw- tor can be found (e.g. Northern State, Sudan, worm flies and codling moth (Mumford 2005). the island of La Reunion) can SIT be applied as Benefits of SIT programmes vary with the spe- a stand-alone technique. In other parts of Af- cies concerned, the scale of the project and the rica, SIT could become part of an integrated ap- objectives of the programme (Mumford 2005). proach towards the elimination of the disease. However, for the majority of species benefits After the failed eradication campaign in the can be clearly expressed in terms of economic 1960s, only recently the eradication of malaria gains (e.g. livestock productivity, and trade in is back on the agenda of the World Health Or- agricultural produce). For SIT against malaria ganisation (EMRO/WHO 2007). Some say the vectors, benefits can ultimately only be- ex tools that are available today, i.e. the artemisi- pressed in terms of human health gains. How- nin based drugs, insecticidal long-lasting bed- ever, even though the correlation between the nets, and indoor residual spraying can achieve economic situation of endemic regions and ma- this goal, where most others believe they won’t laria has been demonstrated (Gallup and Sachs suffice for a global eradication and other strat- 2001, Sachs and Malaney 2002), these benefits egies are needed (Roberts and Enserink 2007). are much less easy to express in hard numbers. Indoor residual spraying specifically targets The cost/ benefit analysis of SIT programmes mosquitoes that exhibit endophilic behaviour should be compared with costs and benefits as- (i.e. indoor resting), while exophilic species (i.e. sociated with other methods of malaria control outdoor resting) are much less affected (Pates including insecticide spraying, bednets, drug and Curtis 2005). Insecticide-treated bednets administration etc. It should be kept in mind (ITNs) rely on the nocturnal biting behaviour of that one of the great benefits of SIT is that it is anophelines. In an area with prolonged ITN use, an environmentally safe method to use, and re- a behavioural adaptation towards earlier feed- sistance (i.e. in terms of behavioural resistance ing was reported (Braimah et al. 2005) which of females towards released males (Itô and could seriously affect the effectiveness of ITNs. Yamamura 2005)) is much less likely to develop The SIT could be used in combination with ex- 176 compared to conventional methods. However, isting methods to specifically target those spe- SIT programmes are a long-term commitment cies that are difficult to control by conventional and even after a vector has been successfully methods, or is used against populations that eliminated from a region, monitoring and fol- have developed resistance. Thus, the Sterile In- low-up is needed to maintain this status. sect Technique and other genetic control strate- 11 General discussion

gies could become a key component in an inte- and Evolution of Behavior. CRC Press LLC, Boca grated vector management (IVM) programme, Raton. with the ultimate goal to eradicate malaria. Charlwood, J. D., and M. D. R. Jones 1980. Mating in the mosquito, Anopheles gambiae s.l. II. Swarming Acknowledgements behaviour. Physiol. Entomol. 5:315-320. I thank B. Knols and M. Dicke for constructive Clark, A.G. 2002. Sperm competition and the comments on earlier versions of this Chapter. maintenance of polymorphisms. Heredity 88:148- 153. Costantini, C., S. Li, A. della Torre, N’F. Sagnon, M. References Coluzzi, and C. E. Taylor 1996. Density, survival and dispersal of Anopheles gambiae complex mosquitoes Abbott, W. S. 1925. A method of computing the in a West African Sudan savanna village. Med. Vet. effectiveness of an insecticide. J. Econ. Entomol. Entomol. 10:203–219. 18:265-267. Curtis, C. F. 1971. Induced sterility in insects. Adv. Anwar, M., D. L. Chambers, K. Ohinata, and R. Reprod. Physiol. 5:120-165. M. Kobayashi 1971. Radiation-sterilization of Curtis, C. F. 1976. Radiation sterilization. Report the mediterranean fruit fly (Diptera: Tephritidae): on mosquito research, Ross Institute of Tropical Comparison of spermatogenesis in flies treated as Hygiene. 01.01.76-31.12.77. pupae or adults. Ann. Entomol. Soc. Am. 64:627-633. Curtis, C. F. 1985. Genetic control of insect pests: Bailey, D. L., R. E. Lowe, J. E. F. Fowler, and D. growth industry or lead balloon. Biol. J. Linn. Soc. A. Dame 1979. Sterilizing and packaging males of 26:359-374. Anopheles albimanus Wiedemann for field release. Dame, D. A., C. S. Lofgren, H. R. Ford, M. D. Am. J. Trop. Med. Hyg. 28:902-908. Boston, K. F. Baldwin, and G. M. Jeffery 1974. Bellini, R., M. Calvitti, A. Medici, M. Carrieri, G. Release of chemosterilized males for the control of Celli, and S. Maine 2007. Use of the Sterile Insect Anopheles albimanus in El Salvador. II. Methods of Technique Against Aedes albopictus in Italy: First rearing, sterilization, and distribution. Am. J. Trop. Results of a Pilot Trial, pp. 505-516. In M. J. B. Vreysen, Med. Hyg. 23:282-287. A. S. Robinson, and J. Hendrichs (eds.), Area-Wide Dame, D. A., R. E. Lowe, and D. L. Williamson 1981. Control of Insect Pests. Springer, Dordrecht. Assessment of Released Sterile Anopheles albimanus Braimah, N., C. Drakeley, F. Mosha, M. Helinski, and Glossina morsitans morsitans, pp 231-248. In J. H. Pates, C. Maxwell, T. Massawe, C. Curtis 2005. B. Kitzmiller, and T. Kanda (eds.), Cytogenetics and Tests of bednet traps (Mbita traps) for mosquito genetics of vectors. Elsevier Biomedical, Amsterdam, monitoring in Tanzania. J. Trop. Insect Sci. 25:208- Netherlands. 213. Dame, D. A., and C. F. Curtis 1996. The potential Bressac, C., and E. Hauschteck-Jungen 1996. use of the sterile insect technique and other genetic Drosophila subobscura females preferentially select control methods for control of malaria-transmitting long sperm for storage and use. J. Insect. Physiol. mosquitoes, pp. 1-27. IAEA, Vienna. 42:323-328. della Torre, A, C. Costantini, N. J. Besansky, A. Calkins, C. O., and A. G. Parker 2005. Sterile insect Caccone, V. Petrarca, J. R. Powell, and M. Coluzzi quality, pp. 269-296. In V. A. Dyck, J. Hendrichs, 2002. Speciation within Anopheles gambiae--the and A. S. Robinson (eds.), Sterile Insect Technique. glass is half full. Science 298:115-117. Principles and Practice in Area-Wide Integrated Dowell, R. V., J. Worley, and P. J. Gomes 2005. Pest Management. Springer, Dordrecht, The Sterile insect supply, emergence, and release, pp. Netherlands. 297-324. In A. Dyck, J. Hendrichs, and A. S. Robinson 177 Cayol, J. P. 2000. Changes in sexual behavior and life (eds.), Sterile Insect Technique: Principles and history traits of tephritid species caused by mass- Practice in Area-Wide Integrated Pest Management. rearing processes, pp. 843-860. In M. Aluja and A. L. Springer, Dordrecht. Norrbom (eds.), Fruit Flies (Tephritidae): Phylogeny EMRO/WHO 2007. Guidelines on prevention of 11 General discussion

the reintroduction of malaria. EMRO Technical physiological fitness of free-living and laboratory-

Chapter 11 Publications Series 34. http://www.emro.who.int/ reared male Anopheles gambiae s.l. J. Exp. Biol. dsaf/dsa743.pdf. 210:2939-2947. Ferguson, H. M., B. John, K. R. Ng’habi, and B. Hunt, R.H., M. Coetzee, and M. Fettene 1998. The G. J. Knols 2005. Redressing the sex imbalance Anopheles gambiae complex: a new species from in knowledge of vector biology. Trends Ecol. Evol. Ethiopia. Trans. R. Soc. Trop. Med. Hyg. 92:231-235. 20:202-209. Itô, Y., and K. Yamamura 2005. Population and Gage, M. J. G. and E. H. Morrow 2003. Experimental behavioural ecology, pp. 177-208. In V. A. Dyck, J. evidence for the evolution of numerous, tiny sperm Hendrichs, and A. S. Robinson (eds.), Sterile Insect via sperm competition. Curr. Biol. 13:754-757. Technique. Principles and Practice in Area-Wide Gallup, J. L., and J. D. Sachs 2001. The economic Integrated Pest Management. Springer, Dordrecht, burden of malaria. Am. J. Trop. Med. Hyg. 64:85-96. The Netherlands. García-González, F., and L. W. Simmons, L.W. Kaiser, P. E., D. L. Bailey, R. E. Lowe, J. A. Seawright, 2007. Shorter sperm confer higher competitive and D. A. Dame 1979. Mating competitiveness of fertilization success. Evol. 61:816-824. chemosterilized males of a genetic sexing strain of Gillies, M. T. 1961. Studies on the dispersion and Anopheles albimanus in laboratory and field tests. survival of Anopheles gambiae in East Africa, by Mosq. News 39:768-775. means of marking and release experiments. Bull. Klowden, M. J. and G. M. Chambers 2004. Entomol. Res. 52:99–127. Production of polymorphic sperm by anopheline Grover, K. K., C. F. Curtis, V. P. Sharma, K. R. P. mosquitoes and their fate within the female genital Singh, K. Dietz, H. V. Agarwal, R. K. Razdan, tract. J. Insect Physiol. 50:1163-1170. and V. Vaidyanathan 1976a. Competitiveness Klowden, M. J. 2006. Switchover to the mated state of chemosterilised males and cytoplasmically by spermathecal activation in female Anopheles incompatible translocated males of Culex pipiens gambiae mosquitoes. J. Insect Physiol. 52:679-684. fatigans Wiedemann (Diptera, Culicidae) in the field. Knols, B. G. J., B. N. Njiru, E. M. Mathenge, W. Bull. Entomol. Res. 66:469-480. R. Mukabana, J. C. Beier, and G. F. Killeen 2002. Grover, K. K., S. G. Suguna, D. K. Uppal, K. R. MalariaSphere: A greenhouse-enclosed simulation P. Singh, M. A. Ansari, C. F. Curtis, D. Singh, of a natural Anopheles gambiae (Diptera: Culicidae) V. P. Sharma, and K. N. Panicker 1976b. Field ecosystem in western Kenya. Malar. J. 1:19. experiments on the competitiveness of males LaChance, L. E. 1967. The induction of dominant carrying genetic control systems for Aedes aegypti lethal mutations in insects by ionizing radiation and (L.). Entomol. Exp. Appl. 20:8-18. chemicals-as related to the sterile male technique of Grover, K. K., H. V. Agarwal, S. G. Suguna, R. insect control, pp. 617-650. In J. W. Wright, and R. Pal S. Patterson, and V. P. Sharma 1979. Studies on (eds.), Genetics of Insect Vectors of Disease. Elsevier, chemosterilization of Aedes aegypti (L.): I. Evaluation Amsterdam. of thiotepa as a sterilant in laboratory and field cages. LaChance, L. E., C. H. Schmidt, and R. C. Bushland Mosq. News 39:490-500. 1967. Radiation-induced sterilization, pp. 147-196. In Hood-Nowotny, R. C., L. Mayr, and B. G. J. Knols W. W. Kilgore, and R. L. Doutt (eds.), Pest Control- 2006. Use of carbon-13 as a population marker for Biological, Physical and Selected Chemical Methods. Anopheles arabiensis in a sterile insect technique Academic Press, Inc., New York. (SIT) context. Malar. J. 5:1-8. LaChance, L. E., and C. K. Graham 1984. Insect Huho, B. J., K. R. Ng’habi, G. F. Killeen, G. radiosensitivity: dose curves and dose-fractionation Nkwengulila, B. G. J. Knols, and H. M. Ferguson studies of dominant lethal mutations in the mature 178 2006. A reliable morphological method to assess the sperm of 4 insect species. Mutat. Res. 127:49-59. age of male Anopheles gambiae. Malar. J. 5:62. Lowe, R. E., D. L. Bailey, D. A. Dame, K. Savage, Huho, B. J., K. R. Ng’habi, G. F. Killeen, G. and P. E. Kaiser 1980. Efficiency of techniques for Nkwengulila, B. G. J. Knols, and H. M. Ferguson the mass release of sterile male Anopheles albimanus 2007. Nature beats nurture: a case study of the Wiedemann in El Salvador. Am. J. Trop. Med. Hyg. 11 General discussion

29:695-703. Sachs, J., and P. Malaney 2002. The economic and Mangan R. L. 2005. Population suppression in social burden of malaria. Nature 415:680-685. support of the Sterile Insect Technique, pp. 407-425. Silver, J. B. 2008. Measuring adult dispersal, pp. 1377- In V. A. Dyck, J. Hendrichs, and A. S. Robinson (eds.), 1424. In Mosquito ecology- field sampling methods. Sterile Insect Technique. Principles and Practice in Springer, Dordrecht, The Netherlands. Area-Wide Integrated Pest Management. Springer, Simard, F. T. Lehmann T, J. J. Lemasson, M. Diatta, Dordrecht, The Netherlands. Am. J. Trop. Med. Hyg. and D. Fontenille 2000. Persistence of Anopheles 29:695-703. arabiensis during the severe dry season conditions Midega, J. T., C. M. Mbogo, H. Mwambi, W. D. in Senegal: an indirect approach using microsatellite Wilson, G. Ojwang G, J. M. Mwangangi, J. G. Nzovu, loci. Insect Mol. Biol. 9:467-479. J. I. Githure, G. Yan, and J. C. Beier 2007. Estimating Smittle, B. J., and R. S. Patterson 1974. Container for dispersal and survival of Anopheles gambiae and irradiation and mass transport of adult mosquitoes. Anopheles funestus along the the Kenyan coast by Mosq. News 34:406-408. using mark–release–recapture methods. J. Med. Snook, R. R. 2005. Sperm in competition: not playing Entomol. 44:923-929. by the numbers. Trends Ecol. Evol. 20:46-53. Mumford, J. D. 2005. Application of benefit/cost Takken, W., J. D. Charlwood, P. F. Billingsley, and analysis to insect pest control using the sterile insect G. G. Gort 1998. Dispersal and survival of Anopheles technique, pp. 481-498. In V. A. Dyck, J. Hendrichs, funestus and An. gambiae s.l. (Diptera: Culicidae) and A. S. Robinson (eds.), Sterile Insect Technique. during the rainy season in Southeast Tanzania. Bull. Principles and Practice in Area-Wide Integrated Entomol. Res. 88:561–566. Pest Management. Springer, Dordrecht, The Terzian, L. A., and N. Stahler 1958. A study of some Netherlands. effects of gamma irradiation on the adults and eggs Ng’habi, K. R., B. John, G. Nkwengulila, B. of Aedes aegypti. Biol. Bull. 115: 536-550. G. J. Knols, G. F. Killeen, and H. M. Ferguson Tripet, F., Y. T. Touré, G. Dolo, and G. C. Lanzaro 2005. The effect of larval crowding on the mating 2003. Frequency of multiple inseminations in field- competitiveness of Anopheles gambiae mosquitoes. collected Anopheles gambiae females revealed by Malar. J. 4:49. DNA analysis of transferred sperm. Am. J. Trop. Med. Ng’habi, K.R., B. J. Huho, G. Nkwengulila, G. F. Hyg. 68:1-5. Killeen, B. G. J. Knols, and H. M. Ferguson 2008. Vreysen, M. J. 2005. Monitoring sterile and wild Sexual selection in mosquito swarms: May the best insects in area-wide intergrated pest management man lose? Anim. Behav. in press. programmes, pp. 325-361. In A. Dyck, J. Hendrichs, Omar, S. M., and J. L. Cloudsley-Thompson 1968. and A. S. Robinson (eds.), Sterile Insect Technique: Dry Season Biology of Anopheles gambiae Giles in the Principles and Practice in Area-Wide Integrated Pest Sudan. Nature 217:879-880. Management. Springer, Dordrecht. Omar, S. M., and J. L. Cloudsley-Thompson 1970. Yuval, B. and G. N. Fritz 1994. Multiple mating in Survival of female Anopheles gambiae Giles through female mosquitoes - evidence from a field population a 9-month dry season Sudan. Bull. World Health of Anopheles freeborni (Diptera: Culicidae). Bull. Organ. 42:319-330. Entomol. Res. 84:137-139. Pates, H., and C. F. Curtis 2005. Mosquito behavior and vector control. Annu. Rev. Entomol. 50:53-70. Reiter, P., M. A. Amador, and C. Nelson 1991. Enhancement of the CDC ovitrap with hay infusions for daily monitoring of Aedes aegypti populations. J. Am. Mosq. Control Assoc. 7:52-55. 179 Roberts, L, and M. Enserink 2007. Did they really say...eradication? Science 318:1544-1545. Robinson, A. S. 2002. Mutations and their use in insect control. Mutat. Res. 511:113-132. SU MM ARY

180 Introduction (Chapter 1)

Malaria is a parasitic disease caused by protozoans of the genus Plasmodium that are transmitted to humans by mosquitoes of the genus Anopheles. Malaria remains one of the main health problems in sub-Saharan Africa. 350-500 Million cases occur annually, with 1 million Africans, mostly children under five, dying of the disease each year. The economic impact of the disease in affected countries is high, and the relation between poverty and malaria has been demonstrated. Insecticide and drug resistance in recent years has led to an increased interest in other methods of malaria control. An established genetic control method is the Sterile Insect Technique (SIT). This technique has been successfully applied against a number of pest insects during the last decades. It relies on the sterilisation of factory mass-produced males, where sterilisation is induced by radiation. These males are released in the wild, and compete with the wild males for the insemination of wild females. If a female uses sterile sperm for fertilisation no progeny is obtained. If this happens for sufficient numbers of females the population size is reduced. This technique has been applied on a small-scale against an Anopheles mosquito in El Salvador in the 1970s, with encouraging results. The main focus of the research presented in this thesis was built around a feasibility study for the control of African malaria mosquitoes with the SIT. The work focused specifically on the effects of ionising radiation on male Anopheles arabiensis Patton mosquitoes. This species is one of the main vectors of malaria in Africa. Radiation induces dominant lethal mutations in the germ cells leading to sterility. Somatic cells, however, are damaged as well and this can result in a decreased fitness. Because the success of an SIT programme is largely dependent on the ability of released males to inseminate females, the balance between sterility and fitness has been the main focus of study throughout the work performed for thesis.

Irradiation effects (Chapters 2-4)

A review of the literature in Chapter 2 indicated that irradiation studies have been carried out on a number of anopheline species. For An. arabiensis, however, only limited data were available. In general, irradiation of the pupal stage was preferred over adult stage irradiation, due to ease of handling of the pupal stage. There was a marked species difference in radiation sensitivity between the anopheline species, indicating that dose-sterility curves should be determined for each species separately. Dose-response curves for An. arabiensis mosquitoes were determined for the pupal and adult stage in Chapter 3, and the influence of a range of doses on male-induced sterility was assessed. Doses ranged from a control (no irradiation, 0 Gy) to a dose yielding almost complete sterility (120 Gy). The dose response curves between the induced sterility and log (dose) were similar for pupae and adults and were sigmoid. At lower doses, an approximately linear relationship between dose and induced sterility was observed while at higher doses the curve flattened such that increasing amounts of radiation were required for proportionally smaller increases in sterility. Pupae were slightly more radiation-resistant than adults. Irradiation of pupae had no effect on eclosion, and longevity of irradiated males was overall similar to un-irradiated males. For males irradiated in the 181 pupal stage, a significant negative correlation was found between insemination and dose but the correlation coefficient was small (i.e. .25). In Chapter 4, the reproductive potential of male An. arabiensis was determined using sperm quantity and sperm length polymorphisms in the testes as parameters. Sperm quantity increased with age, and 12-day-old males had significantly more sperm in their testes (8214 ± 467) than males Summary Summary

aged 3 days (5022 ± 375). Males irradiated in the pupal stage with 70 or 120 Gy had significantly dance than the others. If a system is enriched fewer sperm (2982 ± 125) than un-irradiated with the heavier and less abundant isotope, males (4950 ± 848). For adult stage irradiation this can be used as a label. The stable isotope similar amounts of sperm were observed composition of samples is measured using an compared to un-irradiated males. Sperm length isotope ratio mass spectrometer. 13 polymorphism was present in An. arabiensis, In Chapter 5, C-glucose was incorpo- with sperm tails ranging in length between < rated in the diet of An. arabiensis. Mosquitoes 50-500 μm. In males collected from the wild, were labelled in the larval or adult stage, or a similar length distributions were observed. combination of both. Results demonstrated Sperm length distributions in the spermathecae that spermathecae positive for labelled semen of females were different to those measured could successfully be distinguished from empty directly from sperm in the testes and harboured ones or controls (i.e. filled with unlabelled se- 13 13 less cells of the smaller (<100-200 µm), and men) using the raw d C values. Not all C added more cells of the larger category (300-400 to the diet was recovered in the mosquito, and µm). Adult irradiation had no impact on the label was lost as a result of respiration in the lar- sperm length distribution observed. However, val trays and turnover in the insect. Labelling of pupal irradiation resulted in a significant mosquitoes in the larval stage was considered increase in small sperm (i.e. in the category to be the optimal treatment. The label persisted of 100-200 μm) compared to un-irradiated in spermathecae for up to 7 days after mating, males. The implications of these findings for and unlabelled sugar feeding of males labelled SIT programmes cannot be predicted yet and in the larval stage did not result in a detectable experiments to determine the role of short turnover of the semen label. There were no det- sperm in fertilisation, and the amount of sperm rimental effects of the addition of labelled glu- transferred by irradiated males, should be cose on larval development and survival, adult performed. size, male longevity, and mating performance. The use of a threshold value of three standard deviations above the mean value for virgin (i.e. unlabelled) females was used to demonstrate Stable isotopes in mosquito mat- with 99.7% confidence that samples could be ing research (Chapters 5, 6, and 9) classified correctly. Apart from the application of a semen-label in laboratory-based studies, In the context of genetic control studies male’s the potential of the technique to study mating mating success in the field is critical for a posi- behaviour of insects in the field was discussed. tive outcome of the intervention, and studies In Chapter 6 similar experiments were that can lead to a better understanding should performed with 15-nitrogen to determine be pursued. However, studies on the mating be- its suitability as a semen-label. 15N-glycine haviour of Anopheles mosquitoes are difficult to was added to larval rearing water and results conduct, with mating taking place at dusk under showed that spermathecae positive for semen low light conditions. Furthermore, Anopheles could be distinguished from un-inseminated or mosquitoes mate in swarms and mating cou- control samples using the raw δ15N‰ values. ples only last for 15-20 seconds. Stable isotopes The label persisted in spermathecae for up to were chosen as novel tools to study mating in 5 days after insemination, and males aged 10 An. arabiensis. Stable isotopes are naturally oc- days transferred similar amounts of label as 182 curring in the environment, are not radioactive males aged 4 days. There were no negative ef- and therefore do not decay. Most elements of fects of the label on larval survival and male lon- biological interest (including carbon and nitro- gevity. Enrichment of teneral mosquitoes after gen) have two or more stable isotopes, with the emergence was 4.85 ± 0.10 atom% 15N, and no lightest of these present in much greater abun- loss of label due to respiration or turnover was Summary Summary

observed. The threshold value was successful in pensate for the loss of competitiveness after classifying a large proportion of samples cor- pupal stage irradiation were explored. First, rectly (i.e. on average 95%), and the possibilities experiments with a higher ratio of irradiated and constraints of both labels were discussed. versus un-irradiated insects were performed. In Chapter 9, both 15N-glycine and 13C- Second, male pupae were irradiated just prior glucose were used as semen labels to study to emergence after prolonging their pupal de- multiple insemination events. The approach velopment period by cooling. The outcome of was successful in identifying females insemi- competition experiments could be improved nated by both groups of males and 25% of all by a three-fold increase in the ratio of pupal ir- females were inseminated at least twice under radiated males versus un-irradiated males for laboratory conditions. For 15N some variation in the lower dose of 70 Gy compared to competi- unlabelled spermathecae was observed, and it tiveness observed at an equal ratio (Chapter 7). was recommended to increase the amount of However, for the higher dose of 120 Gy compet- 15N label in future studies. itiveness remained low. The irradiation of older pupae did not result in a dramatic improvement in competitiveness. Nevertheless, cooling could Male mating competitiveness be a useful tool to facilitate handling processes (Chapters 7, 8, and 9) of large numbers of mosquitoes in genetic con- trol programmes. Male mating competitiveness is of critical In Chapter 9, irradiation did not appear importance in SIT programmes, and mating to impact on the frequency of multiple insemi- competitiveness experiments were performed nation. In experiments where males irradiated where irradiated An. arabiensis males competed as pupae competed with un-irradiated males with un-irradiated males for females. In Chap- for mates the frequency of females inseminat- ter 7, a number of parameters were studied, in- ed by two males was similar compared to a con- cluding developmental stage (pupae or adult), trol treatment without irradiation. However, dose (partially (70Gy) or fully-sterilising (120 a review of the literature indicated that these Gy)), and cage type (small or large). Overall, experiments should be repeated in semi-field adult irradiation resulted in better competitive- cages, as space restrictions in laboratory cages ness compared to pupal irradiation, and males could have accounted for an overestimation of were equally competitive against un-irradiated the true proportion of multiple insemination in males at a 1:1 ratio for both doses in large cages. the wild. In the majority of replicates un-irradi- In smaller cages, males irradiated as adults with ated males had a better competitiveness than 120 Gy had a lower competitiveness compared irradiated males, and for 70 Gy the proportion to un-irradiated males, while for 70 Gy this was of females inseminated was similar to previous not observed. Males irradiated as pupae had a results (Chapter 7). lower competitiveness compared to un-irradi- ated males, and this was more pronounced for males irradiated with the high dose. Males irra- Small-scale feasibility study of the diated as pupae with 70 Gy had a significantly SIT in Sudan (Chapter 10) lower competitiveness when tested in large cages compared to results observed for the The work described in Chapter 10 forms part of small cage experiments. The fact that cage size a feasibility study to suppress a population of affected competitiveness for some treatments the malaria vector An. arabiensis in Northern 183 indicated that competitiveness determined in State, Sudan, with the SIT. The small-scale fea- laboratory experiments must be confirmed un- sibility of the irradiation and transportation of der semi-field conditions. mosquitoes was assessed. An. arabiensis mos- In Chapter 8, two scenarios to com- quitoes were irradiated in Khartoum, and this Summary Summary

was the first irradiation of Anopheles mosqui- sibility of the SIT in Sudan be determined. The toes in Africa. Minor problems were associated future of the SIT against African malaria vectors with the irradiation process, largely as a result is largely seen as a component of an integrated of inexperience of performing the experiments vector management approach. under local conditions. After the irradiation, in- sects were transported as adults by air to the field site earmarked for future releases (400 km from the laboratory), and this resulted in mini- mal mortality (< 6%). Experiments under near-natural condi- tions were performed in a large field cage situ- ated in the release site. The cage was prepared for experiments, and micro-climates were cre- ated where conditions were favourable for mos- quitoes. A number of experiments on survival and mating were performed, using males from the laboratory in Khartoum (i.e. transported by air) and wild males collected as larvae or pupae from breeding sites. Mosquitoes were recovered from the field cage by the sampling of resting sites during the day. Survival of mosquitoes in the field cage was high, and approximately 60% of the insects were recaptured after one night. Survival of laboratory reared males was similar to wild males, and the same applied for irradi- ated males; however this was only replicated once. Mating occurred in high frequencies (i.e. an average of 60% insemination of females af- ter 1-2 nights of mating). Importantly, labora- tory reared males (i.e. sixty generations) were able to inseminate wild females at rates com- parable to wild males; indicating that no major behavioural difference due to the rearing proc- ess had occurred. Two small-scale competition experiments were performed; however, due to problems observed with the feeding and egg laying of low numbers of wild females, no con- clusions on competitiveness could be drawn. It was concluded that mosquitoes survived and mated in the field cage and further experiments on the competitiveness of these insects should be pursued. From a purely biological viewpoint the 184 irradiation of adults would be recommended in SIT programmes, however logistically this is challenging. The work in this thesis only com- prised a small area of the SIT and only when all other components are in place can the true fea- Summary Summary

185 SA M EN - I NG VATT

186 Introductie (Hoofdstuk 1)

Malaria is een parasitaire ziekte die wordt veroorzaakt door protozoën van het genus Plasmodium. De parasieten worden overgebracht op mensen door muggen van het genus Anopheles. Malaria blijft één van de grootste gezondheidsproblemen in Afrika ten zuiden van de Sahara. Daar komen jaarlijks 350-500 miljoen gevallen van de ziekte voor, waaraan 1 miljoen Afrikanen, meestal kinderen onder de vijf jaar, sterven. De ziekte drukt een stempel op de economieën van getroffen landen, en de relatie tussen armoede en malaria is duidelijk aanwezig. De resistentie van de Afrikaanse malariamug en de parasiet tegen insecticiden en medicijnen in afgelopen jaren heeft geleid tot een toegenomen interesse in andere manieren van malariabestrijding. De Steriele Insecten Techniek (SIT) is een gevestigde genetische bestrijdingsmethode en is succesvol toegepast tegen een aantal plaaginsecten gedurende de afgelopen decennia. Het berust op de sterilisatie van grote aantallen mannetjes die in een fabriek gekweekt worden. Steriliteit wordt teweeggebracht door middel van ioniserende straling. Deze mannetjes worden daarna losgelaten in de vrije natuur, en concurreren met wilde mannetjes om de wilde vrouwtjes te insemineren. Als een vrouwtje vervolgens het steriele sperma gebruikt om haar eieren te bevruchten produceert het vrouwtje geen nakomelingen. Als dit bij voldoende vrouwtjes gebeurt, dan zal de populatiegrootte afnemen. Deze techniek is in de zeventiger jaren van de vorige eeuw op kleine schaal succesvol toegepast tegen malariamuggen in El Salvador. Het onderzoek in dit proefschrift vormde een onderdeel van een haalbaarheidsstudie naar de mogelijkheid van het gebruik van de SIT voor het bestrijden van Afrikaanse malariamuggen. Het werk richtte zich specifiek op de effecten van ioniserende straling op mannetjesmuggen van de soort Anopheles arabiensis Patton. Deze soort is één van de belangrijkste vectoren (overdragers) van malaria in Afrika. Straling resulteert in dominante letale mutaties in de kiemcellen die kunnen leiden tot steriliteit. Somatische lichaamscellen worden echter ook bestraald en dit kan leiden tot een verminderde fitness van de mannetjes. Het succes van een SIT programma hangt voornamelijk af van het vermogen van de gesteriliseerde mannetjes om vrouwtjes te bevruchten. Daarom is de balans tussen steriliteit en fitness het belangrijkste thema geweest in het voor dit proefschrift verrichte werk.

Effecten van bestraling (Hoofdstuk 2-4)

Een terugblik op de literatuur in Hoofdstuk 2 leert dat bestralingsstudies zijn uitgevoerd voor een aantal anopheline soorten. Echter, voor An. arabiensis was slechts een beperkte hoeveelheid informatie beschikbaar. Over het algemeen heeft de bestraling van muggen in het popstadium de voorkeur boven bestraling in het volwassen stadium, omdat poppen robuuster zijn en dus makkelijker in gebruik. Tussen verschillende anopheline soorten werd een duidelijk verschil in bestralingsgevoeligheid gevonden, en voor iedere soort zal de relatie tussen dosis en steriliteit apart moeten worden vastgesteld. De invloed van ioniserende straling op de mate van geïnduceerde steriliteit wordt beschre- ven in Hoofdstuk 3. An. arabiensis mannetjes werden bestraald in het pop of volwassen stadium en de invloed van een serie doses werd onderzocht. Deze varieerden van een controle (geen be- straling) tot een dosis die vrijwel volledige steriliteit opleverde (120 Gy). De relatie tussen dosis en 187 geïnduceerde steriliteit was hetzelfde voor het pop en volwassen stadium en was sigmoide in vorm. Bij lagere doses werd een bijna lineaire verhouding tussen dosis en geïnduceerde steriliteit gevon- den terwijl voor de hogere doses de curve vlakker werd zodat grotere hoeveelheden straling no- dig waren voor proportioneel kleinere toenames in steriliteit. Het popstadium had een iets hogere Samenvatting Samenvatting

resistentie dan het volwassen stadium. Het Stabiele isotopen in paringsonder- bestralen van poppen had geen invloed op het zoek bij malariamuggen (Hoofd- uitkomen van de adulten, en de overleving van stukken 5, 6, and 9) bestraalde mannetjes was over het algemeen gelijk aan die van onbestraalde mannetjes. Een Het paringssucces van mannetjes is essentieel significante negatieve correlatie tussen insemi- voor het slagen van een genetische bestrij- natie en dosis werd gevonden voor mannetjes dingsprogramma en studies die dit succes kun- die in het popstadium bestraald waren, maar de nen doorgronden moeten worden uitgevoerd. correlatiecoëfficiënt was zwak (nl. .25). Echter, het paringsgedrag van Anopheles mug- Hoofdstuk 4 beschrijft onderzoek naar gen is moeilijk te bestuderen omdat de paring de reproductieve capaciteit van An. arabiensis plaatsvindt in de schemering en dus bij weinig mannetjes, daarbij gebruikmakend van de pa- licht. Daarbij komt dat Anopheles muggen pa- rameters spermakwantiteit en spermalengte- ren in zwermen en een paring maar 15 tot 20 variatie in de testes. De hoeveelheid sperma seconden duurt. Stabiele isotopen werden ge- nam toe met leeftijd, en 12 dagen oude man- kozen als een nieuwe methode om paring in An. netjes hadden significant meer sperma in hun arabiensis te bestuderen. Stabiele isotopen ko- testes (8214 ± 467) dan mannetjes die 3 dagen men van nature voor in de omgeving, zijn niet oud waren (5022 ± 375). Mannetjes bestraald in radioactief en vervallen daarom niet. De mees- het popstadium met 70 of 120 Gy hadden sig- te chemische elementen die biologische rele- nificant minder spermacellen (2982 ± 125) dan vantie hebben (inclusief koolstof en stikstof) onbestraalde mannetjes (4950 ± 848). Manne- hebben twee of meer stabiele isotopen waar- tjes bestraald in het volwassen stadium hadden van de lichtste isotoop aanwezig is in veel gro- evenveel sperma als onbestraalde mannetjes. tere hoeveelheden dan de andere isotopen. Als Het sperma van An. arabiensis varieerde in een systeem verrijkt wordt met de zwaardere lengte, met staartlengtes tussen de < 50 en 500 en minder vaak voorkomende isotoop, dan kan μm. Mannetjes die in het wild verzameld waren deze gebruikt worden als label. De samenstel- hadden een gelijke verdeling in spermalengtes ling van stabiele isotopen in een monster wordt in vergelijking met mannetjes van de laborato- gemeten met behulp van een isotoop-ratio- riumkweek. Er was een verschil in de verdeling massaspectrometer. van spermalengtes in spermathecae (zaadop- In Hoofdstuk 5 wordt de toevoeging slagorgaan in vrouwtjes) ten opzichte van de van een 13C-glucose label aan het dieet van An. testes: minder cellen van de kleinere lengtes arabiensis beschreven. Muggen werden gela- (<100-200 µm) en meer cellen van de grote beld in het larvale of het volwassen stadium, lengtes (300-400 µm) waren aanwezig. De be- of in beide. De resultaten lieten zien dat sper- straling van het volwassen stadium had geen mathecae met gelabeld zaad succesvol onder- invloed op spermalengte, echter de bestraling scheiden konden worden van spermathecae van het popstadium resulteerde in een signifi- zonder zaad of met ongelabeld zaad, daarbij cante toename van het aantal kleinere sperma- gebruikmakend van de ruwe d13C waardes. Niet cellen (namelijk in de lengte van 100-200 μm) alle 13C die werd toegevoegd aan het dieet kon in vergelijking tot onbestraalde mannetjes. De worden teruggevonden in de mug, en verlies implicaties van deze resultaten voor een SIT van het label door respiratie en omzetting in programma kunnen op basis van deze experi- het insect vond plaats. De toevoeging van het menten nog niet voorspeld worden. Experimen- label in het larvale stadium bleek de optimale 188 ten die de rol van kort sperma in bevruchting en behandelingsmethode te zijn. Het label kon de hoeveelheid sperma dat overgebracht wordt worden teruggevonden in spermathecae tot door bestraalde mannetjes kunnen vaststellen minstens zeven dagen na de paring. De opna- moeten daarvoor worden uitgevoerd. me van ongelabeld suikerwater door mannetjes gelabeld in het larvale stadium resulteerde niet Samenvatting Samenvatting

in een verlies van het zaadlabel door omzetting Echter, in deze studie werd voor de 15N waarden in het insect. Er werden geen negatieve effec- van ongelabelde spermathecae wat variatie ge- ten gevonden van de toevoeging van de gela- vonden en het is aanbevolen om in vervolgstu- belde glucose op de ontwikkeling en overleving dies de hoeveelheid 15N label te vergroten. van de larven, de grootte van de adulten, of de paringsprestaties van de mannetjes. Een drem- pelwaarde van drie standaarddeviaties boven Concurrentiekracht van mannetjes de gemiddelde waarde voor spermathecae van maagdelijke (ongelabelde) vrouwtjes kon wor- met betrekking tot paring (Hoofd- den gebruikt om monsters met 99.7% betrouw- stukken 7, 8, and 9) baarheid correct te classificeren. Naast het ge- bruik van het zaadlabel in laboratoriumstudies Een bepaling van de competitiekracht van werd ook de potentie van de methode voor de mannetjes met betrekking tot paring is van es- bestudering van paringsgedrag van insecten in sentieel belang voor een SIT programma, en het veld besproken. experimenten waar bestraalde mannetjes in In Hoofdstuk 6 worden vergelijkbare competitie waren met onbestraalde mannetjes experimenten als in Hoofdstuk 5 beschreven voor vrouwtjes werden uitgevoerd. In Hoofd- maar nu met 15-stikstof als zaadlabel. 15N-gly- stuk 7 werden een aantal parameters bestu- cine werd toegevoegd aan het water waarin de deerd waaronder ontwikkelingsstadium (pop of larven gekweekt werden en resultaten geven adult), dosis (gedeeltelijk (70 Gy) of volledig ste- aan dat spermathecae met gelabeld zaad on- riliserend (120 Gy)), en kooitype (klein of groot). derscheiden konden worden van spermathecae Over het algemeen genomen resulteerde de zonder zaad of met ongelabeld zaad gebruik- bestraling van het volwassen stadium in een be- makend van de ruwe δ15N‰ waardes. Het label tere competitiekracht van mannetjes in vergelij- was aanwezig in spermathecae tot minstens king met bestraling in het popstadium. Boven- vijf dagen na de inseminatie, en tien dagen dien hadden mannetjes een concurrentiekracht oude mannetjes droegen evenveel label over die vergelijkbaar was met die van onbestraalde aan vrouwtjes als vier dagen oude mannetjes. mannetjes in een 1:1 ratio voor beide doses in De toevoeging van het label in het larvale sta- grote kooien. In de kleinere kooien hadden man- dium had geen negatief effect op de overleving netjes, die in het volwassen stadium waren be- van larven of van de volwassen mannetjes. Vol- straald met 120 Gy een lagere competitiekracht wassen muggen die net verpopt zijn waren voor ten opzichte van onbestraalde mannetjes ter- 4.85 ± 0.10 atoom% verrijkt met 15N. Er trad wijl dit voor 70 Gy niet het geval was. Mannetjes geen verlies op van het 15N label door respiratie die in het popstadium bestraald waren hadden of omzetting. De drempelwaarde kon succes- een lagere competitiekracht in vergelijking met vol gebruikt worden voor het correct classifice- onbestraalde mannetjes en dit was duidelijker ren van een groot aantal monsters (gemiddeld voor de mannetjes bestraald met de hogere do- 95%), en de mogelijkheden en beperkingen van sis. Mannetjes bestraald als pop met 70 Gy had- het gebruik van de 15N en 13C labels werden be- den een significant lagere competitiekracht als sproken. ze getest werden in grote kooien maar niet in In Hoofdstuk 9 werden 15N-glycine and kleine kooien. Kooigrootte dus had een invloed 13C-glucose samen gebruikt als zaadlabel om de op competitiekracht in sommige experimenten aanwezigheid van meervoudige inseminaties en dit geeft aan dat resultaten behaald in labo- te bestuderen. Beide labels konden succesvol ratoriumexperimenten bevestigd moeten wor- 189 worden toegepast om spermathecae met zaad den onder (semi-)veld condities. van beide labelgroepen te identificeren. In de In Hoofdstuk 8 werden twee scenario’s gebruikte laboratoriumopzet was 25% van de ter compensatie van een verlies in competitie- vrouwtjes minstens tweemaal geïnsemineerd. kracht na de bestraling van het popstadium Samenvatting Samenvatting

onderzocht. Ten eerste werden experimenten deel uit van een haalbaarheidsstudie om een uitgevoerd waarin het aantal bestraalde man- populatie van An. arabiensis muggen in het netjes ten opzichte van het aantal onbestraalde noordelijk deel van Soedan te onderdrukken mannetjes werd verdrievoudigd. Ten tweede met de SIT. De haalbaarheid van het bestra- werd de ontwikkeling van het popstadium ver- lingsproces en transport van muggen werd op traagd door de poppen te koelen en werden kleine schaal onderzocht. An. arabiensis mug- deze oudere poppen bestraald vlak voor het gen werden bestraald in Khartoem, en dit was uitkomen. Een verdrievoudiging van het aantal de eerste keer dat Anopheles muggen werden bestraalde mannetjes resulteerde in meer geïn- bestraald in Afrika. Enige problemen werden semineerde vrouwtjes door bestraalde manne- ondervonden met het bestralingsproces, maar tjes in vergelijking tot de 1:1 ratio experimenten deze waren grotendeels terug te voeren op een uit Hoofdstuk 7. Echter, dit effect was alleen gebrek aan ervaring met het verrichten van zichtbaar als mannetjes bestraald werden met deze experimenten onder lokale condities. De 70 Gy; voor 120 Gy bleef de competitiekracht insecten werden na de bestraling in het volwas- laag. De bestraling van de oudere poppen re- sen stadium met het vliegtuig vervoerd naar het sulteerde niet in een dramatische verbetering gebied dat was geselecteerd voor de toekom- van de competitiekracht van de mannetjes. stige loslatingen (400 km van het laboratorium Daarentegen kan koeling wel een nuttig mid- verwijderd). Dit resulteerde in slechts geringe del zijn om grote aantallen poppen in een ge- mortaliteit (< 6%). netisch bestrijdingsprogramma gemakkelijker Experimenten onder bijna-natuurlijke te hanteren. omstandigheden werden uitgevoerd in een In Hoofdstuk 9 bleek bestraling geen grote veldkooi aanwezig in het loslaatgebied. invloed te hebben op het voorkomen van meer- De kooi werd ingericht voor experimenten door voudige inseminaties. In experimenten waar middel van het creëren van geschikte microkli- mannetjes, bestraald in het popstadium, con- maten voor muggen. Een aantal experimenten curreerden met onbestraalde mannetjes om werden uitgevoerd om de overleving en paring vrouwtjes was het aantal vrouwtjes dat geïn- te bepalen van in Khartoem gekweekte man- semineerd was door twee mannetjes gelijk aan netjes en van wilde mannetjes die als larve of het aantal in de controle-experimenten zonder pop verzameld waren op de plek van de experi- bestraling. Echter, uit een literatuuroverzicht menten. Muggen werden teruggevangen uit de bleek dat deze experimenten opnieuw moeten kooi door het bemonsteren van rustplekken ge- worden uitgevoerd in semi-veld kooien, waar durende de dag. De overleving van de muggen ruimtebeperking minder een rol speelt, om een in de kooi was aanzienlijk, en ongeveer 60% van overschatting van de echte hoeveelheid meer- de insecten werd teruggevangen na één nacht. voudige inseminaties in het wild te voorkomen. Overleving van mannetjes die aan laboratori- In de meeste herhalingen hadden de onbe- umomstandigheden zijn aangepast was gelijk straalde mannetjes een betere competitie- aan die van wilde mannetjes. Ook bestraalde kracht ten opzichte van de mannetjes bestraald mannetjes hadden een vergelijkbare overle- in het popstadium, en was de proportie geïnse- ving, al berust dit maar op één herhaling. Mug- mineerde vrouwtjes voor de 70 Gy experimen- gen paarden in de kooi in grote aantallen (ge- ten gelijk aan de resultaten in Hoofdstuk 7. middeld 60% van de vrouwtjes was geïnsemi- neerd na 1-2 nachten). Erg belangrijk voor een toekomstig SIT programma was de observatie 190 De toepassing van de SIT in Soe- dat mannetjes die gedurende zestig generaties dan: een haalbaarheidsstudie op in het laboratorium gekweekt waren vrouwtjes kleine schaal (Hoofdstuk 10) in gelijke aantallen insemineerden als wilde mannetjes. Hieruit kan worden geconcludeerd Het werk beschreven in Hoofdstuk 10 maakt dat geen grote gedragsveranderingen als ge- Samenvatting Samenvatting

volg van langdurig kweken onder laboratorium- omstandigheden waren geïntroduceerd. Twee competitie-experimenten werden uitgevoerd op kleine schaal. Er kon echter geen conclusie met betrekking tot de competitiekracht van de bestraalde mannetjes worden getrokken omdat het aantal wilde vrouwtjes dat zich voedde en eieren legde te laag was. De conclusie uit deze experimenten was dat muggen overleefden en paarden in de veldkooi en dat experimenten om de competitiekracht van bestraalde mannetjes te bepalen onder deze omstandigheden kun- nen worden uitgevoerd. Vanuit een puur biologisch standpunt wordt de bestraling van het volwassen stadium voor een SIT programma aangeraden maar dit kan logistieke problemen opleveren. Het werk beschreven in dit proefschrift omvat alleen een klein gedeelte van een volledig SIT pro- gramma en alleen wanneer alle componenten gereed zijn kan de werkelijke haalbaarheid van de methode in Soedan worden aangetoond. De toekomst van een SIT aanpak tegen Afrikaanse malariavectoren wordt grotendeels gezien als een component van een geïntegreerde aanpak van vectorbestrijding.

191 M CULU I V I TAE

192 CURR On the 26th of May 1980 was I, Michelle Helinski, born in Heerlen, and I grew up in the town of Nuth. After completing secondary school in Sittard I started my biology studies in 1998 at Wageningen University. During the course of these studies, I became interested in medical entomology and started my first MSc project with Dr. Willem Takken and others on the phenology of Ixodus ricinus ticks, vectors of Lyme disease, in the Netherlands. After this project I spent 6 months in Muheza, Tanzania, to work together with Caroline Maxwell from the London School of Medicine and Tropical Hygiene on a project studying the use of different collection methods for malaria mosquitoes. After returning to Wageningen my next short project was done at the Genetics department where, under the supervision of Dr. Fons Debets, I worked on the characterisation of the het-s gene of the fungus Podospora anserina. My final project was performed at the University of Edinburgh in Scotland where I spent five months studying the dynamics and evolutionary aspects of virulence of malaria parasites in a rodent malaria model, under the supervision of Dr. Jaap de Roode and Prof. Andew Read. In March 2004 I graduated from my biology studies. Around that time I came in touch with Dr. Bart Knols who was looking for new people to join his group at the International Atomic Energy Agency (IAEA) in Vienna, Austria. Only a few weeks later, in June 2004, I started a 7-month contract at the Agency’s laboratories in Seibersdorf to study the effects of radiation on the biology of males of the malaria mosquito Anopheles arabiensis. During these first months the idea was born to continue beyond the duration of the initial contract and submit the work for a PhD degree at Wageningen University. Four contract extensions and 3.5 years later I finished my job at the IAEA in November 2007. The first months of 2008 I spent at the Entomology Department in Wageningen to complete the writing of this thesis, the result of which you now hold in your hands. My next job has, when this thesis is printed, already started at Cornell University in Ithaca, NY State, USA to work in the group of Dr. Laura Harrington on fitness aspects of genetically modified mosquito vectors of dengue.

193 194 I ONS PUBL I CAT The chapters (or parts thereof) presented in this thesis have been or will be published as:

Helinski, M. E. H., A. G. Parker, and B. G. J. Knols. The Sterile Insect Technique for African Anopheles. V. Radiation biology of mosquitoes, a review. Malar. J. in prep.

Helinski, M. E. H., and B. G. J. Knols. Sperm quantity and size polymorphism in un-irradiated and irradiated males of the malaria mosquito Anopheles arabiensis Patton. Acta Trop. submitted.

Helinski, M. E. H., and B. G. J. Knols. The influence of late-stage pupal irradiation and increased irradiated: un-irradiated male ratio on mating competitiveness of the malaria mosquito Anopheles arabiensis Patton. Bull. Entomol. Res. submitted.

Helinski, M. E. H., R. Hood, D. Gludovacz, L. Mayr, and B. G. J. Knols. Use of the 15N stable isotope as a semen label to detect mating in the malaria mosquito Anopheles arabiensis Patton. J. Am. Mosq. Con. Assoc. submitted.

Helinski, M. E. H., M. M. Hassan, W. M. El-Motasim, C. A. Malcolm, B. G. J. Knols, and B. El- Sayed. 2008. Towards a Sterile Insect Technique field release of Anopheles arabiensis mosquitoes in Sudan: irradiation, transportation, and field cage experimentation. Malar. J., 7:65.

Helinski, M. E. H., and B. G. J. Knols. 2008. Mating competitiveness of Anopheles arabiensis males irradiated with a partially-sterilizing and a fully sterilizing dose in small and large laboratory cages. J. Med. Entomol. in press.

Helinski, M. E. H., R. Hood, and B. G. J. Knols. 2008. A stable isotope dual-labelling approach to detect multiple insemination in un-irradiated and irradiated Anopheles arabiensis mosquitoes. Parasit. Vectors 1:9. doi: 10.1186/1756-3305-1-9.

Helinski, M. E. H., R. Hood, L. Mayr, and B. G. J. Knols. 2007. Stable isotope-mass spectrometric determination of semen transfer in malaria mosquitoes. J. Exp. Biol. 210:1266-1274. doi: 10.1242/ jeb.002642.

Helinski, M. E. H., A. G. Parker, and B. G. J. Knols. 2006. Radiation-induced sterility for pupal and adult stages of the malaria mosquito Anopheles arabiensis. Malar. J. 5:41. doi: 10.1186/1475- 2875-5-41.

Helinski, M. E. H., B. El-Sayed, and B. G. J. Knols. 2006. The Sterile Insect Technique: Can established technology beat malaria? Entomol. Ber. 66:13-21. Other publications: de Roode, J. C., M. E. H. Helinski, M. Ali Anwar, and A. F. Read. 2005. Dynamics of Multiple Infection and Within-Host Competition in Genetically Diverse Malaria Infections. Am. Natural. 166:531-542. doi: 10.1086/491659 195 de Roode, J. C., R. Pansini, S. J. Cheesman, M. E. H. Helinski, A. S. Bell, B. H. K. Chan, D. Walliker, and A. F. Read. 2005. Virulence and competitive ability in genetically diverse malaria infections. PNAS 102:7624-7628. doi: 10.1073/pnas.0500078102. Publications Publications

Braimah, N., C. Drakeley, F. Mosha, M. Helinski, H. Pates, C. Maxwell, T. Massawe, and C. Curtis. 2005. Tests of Bednet Traps (Mbita traps) for Mosquito monitoring in Tanzania. J. Trop. Insect Sci. 25:208-213. doi: 10.1079/IJT200576.

196 Publications Publications

197 198 D ANKWOOR A cknowle d ge m ents When I started working for the International Atomic Energy Agency (IAEA) in June 2004 on a 7-month contract I could have never imagined that 4 years later I would hold this thesis in my hands. The truth is that I thought it would be nice to do something else for a while before I would start thinking about a PhD. But 7 months became 3,5 years and when I finished my last contract with the IAEA in November 2007 I was very grateful this PhD degree was within reach. However, it did not materialise out of thin air, and a lot of hard work was put into it. However, even with all the hard work in the world it would have not been possible without the support of many people, whom I’d like to thank here.

Foremost, I would like to thank Bart. For if it wasn’t for you I would have never even thought about doing a PhD within the IAEA. I think it is truly exceptional to find people that are not just good in what they do, but also in the way they do it. With or without the MBA, you have the ability to create a very open and positive atmosphere that I have always greatly appreciated. So many thanks for all the time you have invested into this thesis, and thanks to Ingeborg and yourself for making me feel (probably too) welcome in your home. I would also like to express my gratitude to Marcel for giving me the opportunity to submit this thesis in Wageningen and for the time you invested in giving valuable feedback on my work.

The majority of the work in this thesis was done within the Mosquito group of the Entomology Department in the Seibersdorf laboratories. I have many good memories of the time I spent there, not in the least because of the nice people I ended up working with. Genevieve, I think us two brought the average age of the Agency down with a few years, thanks for all the fun in the lab, in Vienna and at the ball. I am looking forward to work together with you in the future. Hervé, you must be one of the friendliest French people around;), thanks a lot for all. Gerti, auch sehr vielen Dank für die schöne Zeit. Becky, well, what can I say, you and I shared more than just the crappy car, you are a great friend! Many thanks as well for introducing me to the wonderful world of stable isotopes. Janis, thanks for all the spass we had in the lab, I have great memories from Venice and our Friday afternoon sessions. Sharon, our mosquito mama, you taught me what clean really is! But more important you were always there for me to help out with experiments and listen to my rambling, for which I cannot thank you enough, you are a wonderful friend. A special thanks to Mark as well, you’ve been a great help throughout the work performed for this thesis; thanks for your continuous support and all your invested time. Thanks as well to Colin for organising the field work in Sudan which resulted in a very nice chapter. A number of students joined our lab over the years and thanks to Safia, Osama, Hazar, Samual, Mike, Mr. Ali, Ms. Nurhayati, Taif and Wolfgang for the hard work and nice times. Mike, I truly hope I get to visit Ghana one day! Safia, Hazar and Osama, many thanks for making me feel very welcome in Sudan.

In the Entomology Department, I am greatly indebted to Alan, Gerald, Carlos and Andrew for all your support, for looking at drafts of manuscripts, helping me find references, thinking about experimental protocols and tolerating my jumping around. Andrew, thanks for making me familiar with the gamma source and its dosimetry system. Also many thanks to all Entomology staff for creating such a friendly working environment, vielen Dank dafür! Thanks as well to the Medfly boys Neil and Diego for the after-work beers and pool, to the workshop people and Miklos for their kind cooperation with all that we needed built, and to Anne and Anna for the administrative duties. 199 Adedapo, your help in statistical matters is much appreciated. On this note I would also like to thank Ms. Voigt for supporting my work during my time in the Agency. Also a very special thanks to Badria, Mo’awia, Tellal, Waleed, Rania, Dalia, and all other staff at TMRI for making me feel very welcome in Sudan and for all the hard work you put into the project.

Dankword/ Acknowledgements

Then there was the Monday evening dinner club, although the people kept on changing, Joanna, Damian, Marcin, to name a few, it was always great fun and a nice break from work, even though we always ended up talking about it;). Much of the work stress was vented off by kicking against a ball, so many thanks to the girls from the football team and our wonderful coach for preventing me from sleeping in on Saturday mornings and for keeping me fit.

Living in Vienna was a truly great experience and I am glad I could share this with all those people that came to visit me, on average one visit every two months, over the years. Thanks to all my friends at home as well for keeping in touch, I greatly appreciate our friendship and I am confident we can keep it up even with an ocean in between! In Vienna I would like to thank Maarten; bedankt voor alle gezellige avondjes, voetbal kijken en hachee eten, and Friderike: I have great memories of all our movie and cocktail evenings, thanks a lot for all!

My final months in Wageningen in 2008 at the Department of Entomology have been very pleasant. Unfortunately 4 months were not really enough to get to know you all well, but I have greatly appreciated the friendly working environment. Many thanks to Fedor for tolerating my laptop voyager and for all those print jobs on your PC. Kim, many thanks for the great cover illustration. Last but not least I am greatly indebted to Nina for making this thesis look as beautiful as it is, and for taking care of printing matters while I was away, I cannot thank you enough!

Tenslotte wil ik mijn familie bedanken voor alles wat jullie voor mij gedaan hebben en doen, een leukere familie kan ik mij niet wensen! Roman, nog bedankt voor je hulp met het nakijken van de samenvatting.

Alright, this chapter is almost closed and a new one has begun, but I won’t say goodbye but rather a see you around, either at home, in Ithaca, or anyplace else. Take care!

200 Dankword/ Acknowledgements

201 PE&RC PhD Education Certificate

With the educational activities listed below the PhD candidate has complied with the educational requirements set by the C.T. de Wit Graduate School for Production Ecology and Resource Conservation (PE&RC) which comprises of a minimum total of 32 ECTS (= 22 weeks of activities)

Review of Literature (5.6 ECTS) - The sterile insect technique: can established technology beat malaria? (2006) - Sterile insect technique for African anopheles V. Radiation of mosquitoes (2007)

Writing of Project Proposal (7 ECTS) - Reproductive biology and induced sterility as determinants for genetic control of mosquitoes with the sterile insect technique (2005)

Laboratory Training and Working Visits (5.3 ECTS) - Sperm morphology; University of Idaho, Department of Entomology, USA (2005) - Competitiveness of irradiated males in a semi-field setting; Tropical Medical Research Institute, Khartoum, Sudan (2007)

Post-Graduate Courses (3 ECTS) - 18th Biology of disease vectors course; TDR/Fiocruz, Manaus, Brazil (2007)

Competence Strengthening / Skills Courses (1 ECTS) - Basic and advanced security in the field; IAEA (2005 and 2007)

Discussion Groups / Local Seminars and Other Meetings (6.4 ECTS) - Entomology Unit, local seminars & lunch meetings; Entomology Department (2004-2007) - Mosquito group and steering committee meeting (2004-2007) - Scientific meetings including coordinated research projects: Aedes meeting, Sudan planning meeting and SIT planning meeting (2004-2007)

PE&RC Annual Meetings, Seminars and the PE&RC Weekend (0.6 ECTS) - NERN days (2008)

International Symposia, Workshops and Conferences (5.8 ECTS) - International conference on area-wide control of insect pests: integrating the sterile insect and related nuclear and other techniques; Vienna, Austria (2005) - 15th European society of vector ecology (SOVE) meeting; Serres, Greece; poster presentation (2006) - 5th International conference on applications of stable isotope techniques to ecological studies; Belfast, Northern Ireland; oral presentation (2006)

202 The research presented in this thesis was funded by the International Atomic Energy Agency (IAEA) in Vienna, Austria.

Cover design and thesis layout: Nina Fatouros, Bugsinthepicture.com Visual Studios, The Netherlands

203 Cover illustration: Kim Vermeer

Cover pictures: Michelle Helinski

Printed by Ponsen en Looijen BV, The Netherlands