Malahlapanga – Exploiting Nature’s Bounty for Malaria Control

David N. Durrheim (MBChB, DTM&H, DCH, MPH&TM, FACTM)1, John M. Govere (PhD)1, L.E.O. Braack (PhD)1, Anton Gericke (MSc)1, Rick Speare (BVSc, MBBS, PhD, MACVS, FACTM, FAFPHM)2

1 Communicable Disease Control, Mpumalanga Department of Health, South Africa. orion@mweb.co.za.

2 Tropical Infectious and Parasitic Diseases Unit, School of Public Health and Tropical Medicine, James Cook University, Townsville Qld 4811, Australia. Richard.Speare@jcu.edu.au.

Malahlapanga, a remote spring in the Kruger National Park, South Africa, has proven a valuable site for seminal research on the behavior of Anopheles arabiensis, the most important malaria vector in southern Africa. Findings have had direct application for improved malaria prevention and control. This unique oasis supports a large perennial population of An. arabiensis mosquitoes.  Although adult An. arabiensis are readily collected while biting humans, they are usually obliged to feed on wild mammals that are abundant in the area, as the nearest human habitation is more than 9 km distant and the area is inaccessible to tourists.  No other members of the An. gambiae complex are present and An. arabiensis mosquitoes are free of known human pathogens, including Plasmodium spp. Studies conducted at Malahlapanga have found that the peak biting activity of An. arabiensis occurs during the predawn period and that 81% of bites on humans occur on the ankles or feet. Wearing closed shoes or application of small doses of DEET containing insect repellent to the feet and ankles dramatically reduces vector contact. Elucidation of endophilic vector resting behavior and the value of repellents derived from local plants for reducing bites are currently being assessed. Operational research findings from Malahlapanga have influenced national guidelines, directed larviciding around residential camps in nature reserves, been employed for malaria outbreak response and offer great potential for cost-effective personal protection against An. arabiensis in low incidence malaria areas.

Malaria and Africa

The African Continent bears the brunt of the malaria scourge, malaria being responsible for 9% of the continent’s human disease burden and over one million deaths each year (WHO, 1996). The past decade has seen a global increase in malaria prevalence and malaria-specific mortality that has particularly affected Africa (Nchinda, 1998). This imminent disaster has prompted demands for novel approaches and more effective implementation of proven strategies (Marsh, 1998).

Recent malaria trends in South Africa, where Plasmodium falciparum infection accounts for more than 90% of all malaria cases, are particularly disconcerting. In South Africa, the approach of combining vector control utilizing intradomicillary spraying with a residual insecticide, and prompt diagnosis and therapy of malaria disease through an extensive clinic system, saw a massive eighty percent reduction in the malaria risk area (Gear et al., 1981).   This was sustained since the early 1950s with malaria occurrence restricted to seasonal epidemics in the extreme northeastern portion of the country bordering Swaziland, Mozambique and Zimbabwe, and large-scale agricultural and tourism development became possible in areas previously ravaged by endemic malaria (Sharp and le Sueur, 1996). 

However in the past two decades, the incidence of malaria has escalated from 2,343 malaria cases and 9 deaths notified to the National Health Department in 1981, to 51,535 cases and 402 deaths in 1999 (Durrheim et al., 1999b).

Various contributory factors have been identified. Of particular note is the evolution of resistance in Plasmodium falciparum parasites to chloroquine in the three malaria-affected provinces, KwaZulu-Natal, Mpumalanga and Northern Province, and to sulfadoxine-pyrimethamine in KwaZulu-Natal (Freese et al., 1988, 1991; Wernsdofer, 1991; Kruger et al., 1996; Govere et al., 1999; Bredenkamp et al., in press; Durrheim et al., in press b).

Burgeoning travel by non-immune South Africans into malaria areas and an increased volume of cross-border migration by humans carrying Plasmodium parasites and of infected vectors from malaria endemic neighboring countries has fueled transmission (Durrheim, 1995; Wilson, 1997). Additional complicating factors described by White (1974), Sharp et al. (1984), Sharp and le Sueur (1991), Bouma et al. (1994), Colwell et al. (1998), Mnzava et al. (1998), Govere et al. (2000a,b) and Hargreaves et al. (2000) include:

This altered malaria epidemiology has had a marked impact on health service utilization and perception of increased malaria risk appears to have detrimentally affected the tourism industry (Durrheim et al., 1999b; Durrheim et al., in press a). Malaria prevention in tourists is hampered by the absence of a completely effective chemoprophylactic agent and troubling adverse events profile of available chemoprophylaxis. For example, a postal survey of Kruger National Park visitors found that 23.8% of travelers using chemoprophylaxis experienced adverse events, which they attributed to the medication, a frequency similar to that experienced by European visitors to tropical Africa (Durrheim et al., 1999a; Steffen et al., 1990). Adverse events were cited as a major reason for poor compliance with prescribed medication.

These evolving challenges to malaria prevention and management highlight the importance of a responsive and integrated approach to control (Durrheim and Whittaker, 1996). The need for a comprehensive operational research agenda directed at all vulnerable links in the chain of transmission cannot be overemphasized.

Entomological epidemiology

The simple mathematical models constructed to capture the transmission dynamics of malaria offer an understanding of the importance of effectively targeting vectors for optimal malaria control (Anderson and May, 1992). The basic reproductive ratio (R0), the number of new cases of malaria generated by one case introduced into a population of fully susceptible hosts during the duration of the case, may be quantified by multiplying the transmission rate factor from vector to human during the life-span of the vector (TH) with the transmission rate factor from human to vector during the duration of infection in the human (TV).

In this model,    TH = V/H a bH LV  

while                TV = a bV DH

where:

Important assumptions underpinning this model include a lack of immunity in humans, which is a relatively robust assumption in the South African setting where immunity is not thought to exist due to the seasonal nature of malaria, and an absence of a parasite-induced effect on vector survival or behavior (Dye, 1986).  Thus R0 = TH * TV = V/H a2 bH bV DH LV (Dye, 1994). It can be seen from this equation that the biting rate (a) and life expectancy (LV) of the vector are major determinants of the final value of R0 (Garret-Jones and Shidrawi, 1964). Spraying programs reduce anopheline life expectancy, but the potential value of effective personal protection in reducing the mosquito biting-rate is evident.  

Personal protection and Anopheles arabiensis

Personal protection against mosquito bites falls within the ambit of community involvement as envisaged by the Alma Ata declaration and has proven an effective first line of defense against malaria in tourists (Primary Health Care, 1978; Rifkin, 1981; Schoepke et al., 1998; WHO, 2000). However this strategy has seldom been utilized for malaria control in malaria-endemic areas, except where bed-nets are common, and is underutilized amongst travelers. A recent study of community knowledge and perceptions about Malaleveva (literally the sleep-and-shake illness) in Tonga, one of the highest-risk malaria districts in South Africa, found that although 92% of respondents indicated that mosquito bites were the cause of malaria, only 51% reported ever using personal prevention measures against malaria (Govere et al., 2000a). Only 28.4% of study participants had burnt mosquito coils, while 16.7% had used commercial repellents. Missed opportunities for personal protection commonly exist among tourists to malaria areas.  A postal survey of visitors to the Kruger National Park during the seasonal high-risk period found that 13% of visitors used no personal protection measures and only 17.1% used four or more measures (Durrheim et al., 1998).

Indoor house spraying with residual insecticides is optimally effective against anthropophilic indoor feeding and resting malaria vectors, like Anopheles gambiae s.s. and Anopheles funestus s.s., as proven by their elimination through the house-spraying program from South Africa (Gillies and Meillon, 1968). However Anopheles arabiensis, which is the main malaria vector in South Africa and much of southern Africa, is catholic in its feeding and resting behavior, feeding on both humans and animals and resting both indoors and outdoors (White, 1974; Sharp and le Sueur, 1991). This has limited the effectiveness of indoor house spraying and supplementary effective methods for controlling this vector are urgently required.

Current knowledge and understanding of An. arabiensis behavior in southern Africa is limited. Vector behavioral research is complicated by the relative scarcity of the vector in malaria affected areas, despite locally occurring malaria cases. The small numbers of An. arabiensis collected during routine vector surveillance in Mpumalanga and neighboring Provinces provide study samples of feral mosquitoes that are too small to allow meaningful field studies on mosquito behavior (Govere et al., 2000b; la Grange and Coetzee, 1997).

The greater relative abundance of other members of the An. gambiae complex, including An. merus and An. quadriannulatus, which occur sympatrically with An. arabiensis in South Africa, confounds field studies because sophisticated laboratory expertise is required to reliably differentiate catches.

Malahlapanga (“the place of the long black dagger”)

Malahlapanga (22053’S 31002’E) is a fresh water geothermal spring in a pristine wilderness area in the remote north of the Kruger National Park, South Africa (Figure 1). 

The Kruger National Park at over 1,949 million hectares is the largest wildlife haven in a single African country. It is situated in the northeastern corner of South Africa and falls within the seasonal malaria-risk area (Durrheim et al., 1998). Extensive mosquito collections made in the Kruger National Park during the early 1990s led to the discovery of this remote warm water spring with an unique breeding colony of Anopheles arabiensis Patton mosquitoes (Coetzee et al., 1993). The spring is located in mixed Colophospermum mopane (Kirk) and Acacia nigrescens (Oliver) savanna. It has a high mineral content, although not saline, and ranges in temperature between 370C and 380C where it emerges from the ground (Coetzee and Braack, 1991). The spring is located in a clearing of approximately 200 by 600 meters.  This oasis supports a large perennial population of An. arabiensis mosquitoes (Braack et al., 1994). Although adult An. arabiensis are readily collected biting humans, they are usually obliged to feed on wild mammals that are plentiful in the area, as the nearest human habitation is more than 9 km distant. Mammal species include buffalo, elephant, white rhinoceros, lion, leopard, hyena, eland, kudu, impala and zebra. The area is inaccessible except by four-wheel drive vehicles and entry to this wilderness zone is prohibited to tourists.

Polymerase chain reaction investigation has confirmed the unsurpassed research potential of this perennial and abundant mosquito colony.  Not only is it a pure colony restricted to An. arabiensis mosquitoes with no other members of the An. gambiae complex present to complicate identification but it is also free of known human pathogens, including Plasmodium spp. (Braack et al., 1994; Govere, 2000). This has allowed a number of important studies that have meaningfully enhanced the role of personal protection and malaria control in South Africa. It is envisaged that the unfolding research agenda that builds on insights already gained will further improve personal protection and control measures directed against the malaria vector An. arabiensis in South Africa.

Preferred feeding period

Although there is a marked difference in human biting catch rates between the wet and dry seasons, with wet season catches being approximately double in size, the proportional hourly nocturnal biting pattern is similar throughout the year (Braack et al., 1994). An. arabiensis invariably commences biting after sunset but well before last light. Biting activity is relatively low in the early evening, persists throughout the night but peaks during the predawn period.

This finding has important implications for advice provided to travelers. An evaluation conducted in Mpumalanga found that N,N-diethyl-m-toluamide (DEET) provides complete protection for five hours after application against An. arabiensis but that this effect then fades (Govere et al, 2000c). Results from a field evaluation of DEET in Kenya confirm these findings and demonstrate that An. arabiensis is sensitive to DEET, which provided greater than 80% protection for three hours but only 60% protection after nine hours (Walker et al., 1996). Where application of repellents to exposed skin is performed at dusk, as generally advocated, the repellent effect of a single application will have greatly diminished by late evening and, in particular, the pre-dawn maximum biting frequency period. Advice provided to tourists by the only yellow fever vaccination and travel medicine center in Mpumalanga has been tailored to include a second application prior to retiring to sleep at night with reapplication on venturing out of a screened environment prior to sunrise. This repeated application is feasible due to the excellent safety profile of DEET during 40 years of extensive use (Fradin, 1998).

Biting pattern and distance from breeding site

Capture stations were placed at 30 m intervals up to 90 m, 60 m intervals up to 180 m and 200 m intervals up to 600 m from the spring into the prevailing wind direction.  Mosquito numbers were found to be relatively high in the immediate vicinity of the spring, low in the surrounding trampled area, at a maximum approximately 60 m from the spring, and then rapidly declined to 600 m (Braack et al., 1994).

This finding was used to plan control interventions in the Kruger National Park where human dwellings are localized within tourist camps.  Larviciding with Bacillus thuringiensis of anopheline breeding sites is performed within a diameter of 300 m from the outer fences of residential camps and extended to a diameter of 800 m during high incidence years, an intervention that has been directly guided by Malahlapanga operational research findings (Durrheim et al., in press a).

Preferred anatomical feeding sites

The preferential biting patterns of malaria vectors have largely been neglected when designing personal protection measures against mosquito vectors, despite evidence that selection of biting sites on humans by mosquitoes is non-random.  Anopheles gambiae s.s., for example prefers to bite the feet of seated humans  and An. arabiensis mosquitoes have demonstrated a preference for biting the ankles and feet of motionless humans (Curtis et al., 1987; De Jong and Knols, 1995). This was postulated to be the result of odor or chemical attraction of mosquitoes, but Malahlapanga research suggests otherwise for local An. Arabiensis (Knols et al., 1996).

Repeated human-biting catches made with collectors seated on platforms so that their bare feet were either at 0, 0.72-0.85, and 1.44-1.78 m above the ground, found that biting activity was significantly greater at ground level and decreased dramatically with increasing altitude (Braack et al., 1994). Random biting on the human body when lying on the ground supports the finding that proximity to the ground is an important determinant of An. arabiensis biting frequency.

Prevention strategies targeted to ankles and feet

The finding of preferential feeding on humans close to the ground was the foundation of research into the role of mechanical barriers to biting. Research conducted at Malahlapanga demonstrated that the number of An. arabiensis biting motionless humans decreased dramatically when their feet and ankles were covered with plastic bags to 2 cm above the ankle without proportional shift to other parts of the legs or body (Braack, pers. comm.). A total of 1,118 bites were recorded on 10 volunteers during alternating 45 min collection sessions with feet covered or bare over a 3-night period. Volunteers with feet and ankles covered received 23.8% (266) of all bites while bare-footed volunteers received 76.2% (852) bites.

As a consequence, particular emphasis was placed on the use of shoes and socks as a personal measure for protection against anopheline bites in the South African Department of Health malaria prevention guidelines (Department of Health, 1996). The recommendation of donning shoes and socks to prevent mosquito bites when venturing outside at night appears to have influenced practice.  This measure was the second most commonly reported, after skin application of insect repellent, by a large cohort of Kruger National Park visitors responding to a postal questionnaire (Durrheim and Leggat, 1999).

A seminal study was conducted at Malahlapanga to compare the biting behavior of feral An. arabiensis mosquitoes on eight adult human volunteers whose ankles and feet had either been treated or not treated with DEET. The cream formulation of DEET (195 mg/ml AI Tabardä Shell South Africa, Pty Ltd) commercially available in South Africa was used during human night-biting catches. The method, described in detail elsewhere, consisted of subjects dressed only in short pants, being randomly paired into four teams, two teams randomly selected for routine application of DEET on their ankles and feet while the remaining two teams served as controls (Govere, 2000). Catches on four consecutive evenings, with each team being treated on two evenings and hourly rotation of teams through four positions, 10 m apart, parallel to and about 30 m from the main breeding mosquito area at the spring, compensated for positional differences in mosquito density, personal differences in catching ability and attractiveness to mosquitoes, and the hour of night. Total catches for each night were summed and percentage protection, defined as the number of bites received by individuals treated with DEET relative to those experienced by controls, was 69.2% (Mehr et al., 1985). Most bites (76.4% [519/679]) were on untreated subjects (c2=43.978; P<0.001) (Table 1) (Govere et al., in press a).  Eighty-one percent (421/519) of mosquitoes attempting to bite untreated subjects did so on the feet and ankles, and 97.5% of all catches on untreated subjects were made on the area below the knee. No mosquitoes were caught from treated feet and ankles.

Table 1.  Distribution of total mosquito bites on human volunteers with and without skin application of DEET to ankles and feet

Body part

With DEET application [N (%)]

Without DEET application [N (%)]

Ankle/feet

Legs

Arms

Body

Total

0 [0]

141 [88.1]

4 [2.5]

15 [9.4]

160 [100]

421 [81.1]

85 [16.4]

7 [1.3]

6 [1.2]

519 [100]

These findings were used to plan a successful public health intervention.  Following the severe flooding in the Mpumalanga Lowveld in early February 2000, there was an outbreak of malaria in the village Albertsnek Village.  In comparison to the average weekly incidence of 18 malaria cases in the village during the high-risk month, January to May, and the average of 12 malaria cases per week during January and early February 2000, from mid-February to mid-March an average of 33 cases per week were diagnosed.  Many of the homes that had been sprayed with a residual synthetic pyrethroid insecticide prior to the wet season were flooded and although floodwater had begun subsiding, walls were not suitable for reapplication of insecticide.  A donation of DEET cream to the Mpumalanga Department of Health provided a novel opportunity for control.  Rough calculations indicated that if all non-infant members of the village applied DEET twice per evening to their ankles and feet only, the volume of DEET available would be sufficient for residents’ use for the remainder of the malaria season.  After agreement was reached with local leaders and provincial authorities, DEET cream was distributed with concomitant careful demonstration of the application technique at public meetings in the village during the week 13-17th March 2000. Malaria case numbers decreased rapidly following this intervention, being 14 and 11 cases for the subsequent two weeks and with an average of 16 cases per week for the remainder of the high-risk period. 

Evaluation of vector control using contact bioassays

Where intradomicillary wall spraying is practiced for malaria vector control, contact bioassays of insecticide sprayed walls using a sensitive laboratory mosquito colony are advocated by the World Health Organization (WHO, 1975). An insecticide-susceptible mosquito colony is maintained in the Driekoppies Insectary, Mpumalanga, at 240C temperature and 80% relative humidity for this purpose. This colony was founded from the Malahlapanga breeding-colony, as this area of the Kruger National Park is not exposed to insecticides. Standardized monthly assays in Mpumalanga Province have allowed assessment of the residual efficacy of the insecticide used against vector mosquitoes, quality of insecticide application on wall surfaces and persistence of insecticide on these surfaces, that has proven advantageous for program planning and quality assurance (Govere, 2000).

Current research agenda

Present studies of An. arabiensis behavior at Malahlapanga, include investigations into the impact of varying the height of the sleeping surface, modifying bed supports, tucking in bed linen and altering bed placement relative to wall and other solid surfaces, on the biting behavior of An. arabiensis.

In addition, a detailed study is in progress that is exploring the resting habits of endophilic mosquitoes within two experimental huts erected at Malahlapanga. The relative importance of eaves and windows as mosquito entrance routes into the huts is also being studied.

In recent years, three plants, fever tea (Lippia javanica), rose geranium (Pelargonium reniforme) and lemon grass (Cymbopogon excavatus) believed by local Mpumalanga inhabitants to be effective repellents against mosquitoes, were subjected to controlled laboratory experiments (Govere, 2000). Alcohol plant extracts of the three plants provided varying protection against An. arabiensis mosquito bites. L. javanica provided 77% protection after a four-hour period compared to 59% protection by P. reniforme and C. excavatus (Govere et al., in press b). L. javanica will now be evaluated using standardized methodology under field conditions against feral mosquitoes at Malahlapanga (Curtis et al., 1987).

Mosquitoes captured at Malahlapanga continue to be used to replenish the laboratory colony of An. arabiensis maintained for susceptibility and bioassay testing as part of the evaluation of the malaria control program in Mpumalanga Province.

Conclusions

The origins of epidemiology applied to communicable disease control stem from a natural experiment dating back to the 1850s (Snow, 1965). In Malahlapanga, nature has again provided a unique epidemiological oasis for conducting operational research into mosquito vector behavior that may be applied to malaria prevention and control.

The research agenda in Malahlapanga has been driven by the progressive failure and escalating cost of traditional chemical malaria prevention and control strategies, and a recognition that innovative approaches, guided by operational research, are needed into all facets of malaria control (Schapira, 1989; Speare et al., 1999b).

Although malaria vector characteristics, including place of feeding (exophagic or endophagic), resting behavior (exophilic or endophilic), and host preference (opportunistic or zoophilic), have been successfully manipulated for malaria vector control, almost no research of this nature has previously been conducted on southern African Anopheles arabiensis, the major and highly efficient malaria vector in this region.  Malahlapanga’s unique location, unusual single disease-free species and large vector population have provided an unrivaled opportunity to address this deficiency.  

Current progress is heartening with research findings already finding application in South Africa. Findings on flight and biting behavior have resulted in refined barrier spraying strategies for vector control, particularly in nature reserves and cost-effective repellent application for personal protection and outbreak control in Mpumalanga Province. 

A reality of field research sites is their vulnerability to nature’s vagaries and Malahlapanga is no exception. Devastating flooding in north-eastern South Africa and Mozambique during the first quarter of 2000 destroyed all land access routes to Malahlapanga and effectively retarded efforts to further evaluate supplementary cost-effective malaria control measures by 12 months.

The malaria research agenda in Africa should be appraised by its ability to contribute to cost-effective prevention of infection or management of disease at community level (Speare et al., 1999a). The challenge is one of developing appropriate, cost-effective, simple personal protection and population control measures against one of humankind’s greatest foes.  Components of the solution may well be found at a remote spring in the African wilderness.

References

Anderson RM, May RM. (1992) Infectious Diseases of Humans: Dynamics and control. Oxford University Press, Oxford.

Bouma MJ, Sondorp HE, Van der Kaay HJ. (1994) Climate change and periodic epidemic malaria. Lancet 343: 1440.

Braack LEO, Coetzee M, Hunt RH, Biggs H, Cornel A, Gericke A. (1994) Biting pattern and host-seeking behavior of Anopheles arabiensis (Diptera: Culicidae) in northeastern South Africa.  Journal of Medical Entomology 31: 333-339.

Bredenkamp BLF, Sharp BL, Mthembu D, Durrheim DN, Barnes K. (in press) Failure of Sulphadoxine–pyrimethamine in treating Plasmodium falciparum malaria in KwaZulu-Natal Province.  South African Medical Journal.

Coetzee M, Braack LEO. (1991) Anopheles arabiensis in the Kruger National Park, pp. 19. In: Proceedings of the 8th Entomological Congress of the Entomological Society of Southern Africa, Bloemfontein.

Coetzee M, Hunt RH, Braack LEO, Davidson D. (1993) Distribution of mosquitoes belonging to the Anopheles gambiae complex, including malaria vectors, south of the latitude 15°S. South African Journal of Science 89: 227-231.

Colwell R, Epstein P, Gubler D, Hall M, Reiter P, Shukla J, Sprigg W, Takafuji E, Trtanj J. (1998) Global climate change and infectious diseases. Emerging Infectious Disease 4: 451-452.

Curtis CF, Lines JD, Ijumba J, Callaghan A, Hill N, Karimzad MA. (1987) The relative efficacy of repellents against mosquito vectors of disease.  Medical and Veterinary Entomology 1: 109-119.

De Jong R, Knols BGJ. (1995) Selection of biting sites on man by two malaria mosquito species. Experimentia 51:80-84.

Department of Health. (1996) Guidelines for the prophylaxis of malaria. South African Government, Pretoria.

Durrheim DN. (1995) Taxi rank malaria.  British Medical Journal 311: 1507.

Durrheim DN, Whittaker RK. (1996) Malaria control--the changing paradigm.  South African Medical Journal 86:978-979.

Durrheim DN, Braack LE, Waner S, Gammon S.  (1998) Risk of malaria in visitors to the Kruger National Park, South Africa.  Journal of Travel Medicine 5:173-177.

Durrheim DN, Leggat PA. (1999) Prophylaxis against malaria. Preventing mosquito bites is also effective. British Medical Journal 318:1139.

Durrheim DN, Gammon S, Waner S, Braack LE. (1999a) Antimalarial prophylaxis use and adverse events in visitors to the Kruger National Park. South African Medical Journal 89:170-175.

Durrheim DN, Ogunbanjo G, Blumberg L. (1999b) Managing re-emergent malaria in South Africa. South African Family Practice 21:19-24.

Durrheim DN, Braack LEO, Grobler D, Bryden H, Speare R, Leggat PA.  (in press a) Safety of Travel in South Africa: The Kruger National Park.  Journal of Travel Medicine.

Durrheim DN, Sharp BL, Barnes K. (in press b) Sentinel malaria surveillance – merely research or indispensable public health tool? South African Medical Journal.

Dye C. (1986) Vectorial capacity: must we measure all of the components?  Parasitology Today 2:203-209.

Dye C. (1994) The epidemiological context of vector control.  Transactions of the Royal Society of Tropical Medicine and Hygiene 88:147-149

Fradin MS. (1998) Mosquitoes and mosquito repellents: a clinicians guide.  Annals of Internal Medicine 128:931-940.

Freese JA, Sharp BL, Ngxongo SM, Markus MB. (1988) In vitro confirmation of chloroquine-resistant Plasmodium falciparum malaria in KwaZulu. South African Medical Journal 74: 576-578.

Freese JA, Markus MB, Golenser J. (1991) In vitro sensitivity of southern African reference isolates of Plasmodium falciparum to chloroquine and pyrimethamine. Bulletin of the World Health Organization 69:707-712.

Garret-Jones C, Shidrawi G. (1964) Malaria vectorial capacity of a population of Anopheles gambiae. An exercise in epidemiological entomology. Bulletin of the World Health Organization 40:531-545.

Gear JHS, Hansford CF, Pitchford RJ. (1981) Malaria in South Africa: Department of Health, Welfare and Pensions, Pretoria.

Gillies MT, de Meillon B. (1968) The Anophelinae of Africa south of the Sahara. Publication of South African Institute for Medical Research No. 54.

Govere JM. (2000) Anopheline mosquitoes, malaria transmission and control in Mpumalanga Province, South Africa. PhD thesis. University of Witwatersrand.

Govere JM, la Grange JJP, Durrheim DN, Freese JA, Sharp BL, Mabuza A, Mngomezulu N, Bredenkamp BLF. (1999) Sulfadoxine-pyrimethamine effectiveness against Plasmodium falciparum malaria in Mpumalanga Province, South Africa. Transactions of the Royal Society of Tropical Medicine and Hygiene 93:644.

Govere JM, Durrheim DN, la Grange K, Mabuza A, Booman M. (2000a) Community knowledge and perceptions about malaria, and practices influencing malaria control in Mpumalanga Province, South Africa. South African Medical Journal 90:611-616.

Govere J, Durrheim DN, Coetzee M, Hunt RH, la Grange JJP. (2000b) Captures of mosquitoes of the Anopheles gambiae complex (Diptera: Culicidae) in the Lowveld Region of Mpumalanga Province, South Africa. African Entomology 8:91-99.

Govere JM, Durrheim DN, Baker L, Hunt RH, Coetzee M. (2000c) Efficacy of three insect repellents against the malaria vector Anopheles arabiensis.  Medical and Veterinary Entomology 14:369-375.

Govere JM, Braack LEO, Durrheim DN, Coetzee M, Hunt RH. (in press a) The effect of treating human ankles and feet with mosquito repellent on the biting pattern of Anopheles arabiensis (Diptera: Culicidae) mosquitoes in Kruger National Park, South Africa. Medical and Veterinary Entomology.

Govere JM, Durrheim DN, Du Toit N, Hunt R, Coetzee M. (in press b) Local plants as repellents against Anopheles arabiensis in Mpumalanga Province, South Africa.  Central African Medical Journal.

Hargreaves K, Koekemoer LL, Brooke BD, Hunt RH, Mtembu J, Coetzee M. (2000) Anopheles funestus is resistant to pyrethroid insecticides in South Africa. Medical and Veterinary Entomology 14:190-194.

Knols BGJ, van Loon JJA, Cork A, Robinson RD, Adam W, Meijerink J, de Jong R, Takken W. (1996) Behavioural and electrophysiological responses of the female malaria mosquito Anopheles gambiae (Diptera: Culicidae) to limburger cheese volatiles.  Bulletin of Entomological Research 87:151-159.

Kruger P, Durrheim DN, Hansford CF. (1996) Increasing chloroquine resistance-the Mpumalanga Lowveld story, 1990-1995. South African Medical Journal 86:280-281.

la Grange JJP, Coetzee M. (1997) A mosquito survey of Thomo village, Northern Province, South Africa, with special reference to the bionomics of exophilic members of the Anopheles funestus group (Diptera: Culicidae). African Entomology 5:295-299.

Marsh K. (1998) Malaria disaster in Africa.  Lancet 352:924.

Mehr ZA, Rutledge LC, Morales FL, Meixsell VE, Korte DW. (1985) Laboratory evaluation of controlled-release insect formulations. Journal of the American Mosquito Control Association 1:143-147.

Mnzava AEP, Ntuli MV, Sharp BL, Mthembu JD, Ngxongo S, Le Sueur D. (1998) House replastering as a reason to shift from DDT spraying to synthetic pyrethroids. South African Medical Journal 88:1024-1028.

Nchinda TC. (1998) Malaria: A reemerging disease in Africa. Emerging Infectious Diseases 4:398-403.

Primary Health Care. (1978) Report of the International Conference on Primary Health Care, Alma-Ata, USSR, 6-12 September. WHO, Geneva.

Rifkin SB. (1981) The role of the public in the planning, management and evaluation of health activities and programmes, including self help. Social Science and Medicine 15:377-386.

Schapira A. (1989) Chloroquine resistant malaria in Africa: the challenge to the health services. Health Policy and Planning 4:17-28.

Schoepke A, Steffen R, Gratz N.  (1998) Effectiveness of personal protection measures against mosquito bites for malaria prophylaxis in travelers.  Journal of Travel Medicine 5:188-192.

Sharp BL, le Sueur D. (1991) Behavioural variation of Anopheles arabiensis (Diptera: Culicidae) populations in Natal, South Africa. Bulletin of Entomological Research 81:107-110.

Sharp BL, le Sueur D. (1996) Malaria in South Africa – the past, the present and selected implications for the future.  South African Medical Journal 86:83-89.

Sharp BL, Quicke FC, Jansen EJ. (1984) Aspects of the behaviour of five anopheline species in the endemic malaria area of Natal. Journal of the Entomological Society of Southern Africa 47:251-258.

Snow J. (1965) On the mode of communication of cholera. 2nd edition.  London: Churchill, 1885.  Reproduced in Snow on cholera. Hafner, New York.

Speare R, Durrheim DN, Govere J. (1999a) Roll back malaria: the African perspective.  South African Journal of Epidemiology and Infection 14:65-66.

Speare R, Govere J, Durrheim DN, Mngomezulu N. (1999b) Malaria Control in South Africa: Symposium in the Wilderness.  Journal of Travel Medicine 6:149-151.

Steffen R, Heusser R, Machler R, Bruppacher R, Naef U, Chen D, Hofmann AM, Somaini B. (1990) Malaria chemoprophylaxis among European tourists in tropical Africa: use, adverse reactions and efficacy.  Bulletin of the World Health Organization 68:313-322.

Walker TW, Robert LL, Copeland RA, Githeko AI, Wirtz RA, Githure JK, Klein TA. (1996) Field evaluation of arthropod repellents, DEET and piperidine compund, A13-37220, against Anopheles funestus and Anopheles arabiensis in Western Kenya.  Journal of the American Mosquito Association 12:172-176.

Wernsdorfer WH. (1991) The development and spread of drug resistant malaria. Parasitology Today 7:297-303.

White GB. (1974) Anopheles gambiae complex and disease transmission in Africa. Transactions of the Royal Society of Tropical Medicine and Hygiene 68:278-301.

Wilson ME. (1997) Population movements and emerging diseases. Journal of Travel Medicine 4:183-186.

WHO. (1975) Manual on Practical Entomology in Malaria.  World Health Organization, Geneva.

WHO. (1996) Investing in health research for development. Report of the Ad Hoc Committee on Health Research Relating to Future Intervention Options. Report No: TDR/Gen/96.1. World Health Organization, Geneva.

WHO. (2000) International travel and health: vaccination requirements and health advice. World Health Organization, Geneva