SEARCH

SEARCH BY CITATION

Keywords:

  • Anopheles;
  • ecology;
  • mosquitoes;
  • males;
  • population dynamics;
  • survival;
  • dispersal

ABSTRACT:

  1. Top of page
  2. ABSTRACT:
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgments
  8. REFERENCES CITED

Little is known about the fitness of wild male mosquitoes, the females of which are vectors of malaria. The problem of studying male biology has been exacerbated by difficulties associated with catching them. In southern Mozambique, however, almost the entire adult population of An. funestus and An. gambiae s.l. rest inside houses. They leave in a dusk exodus, which makes them easy to collect. In 8,348 exit collections from a village from 2003 to 2009, 567,195 male An. funestus and 34,591 male An. gambiae s.l. were collected. During the study, numbers of An. funestus increased but numbers of An. gambiae s.l. declined to the point of extinction. Overall numbers of An. gambiae s.l. were positively correlated with temperature, whilst the relationship between temperature and numbers of An. funestus changed from an initially positive one in the first three years of the study to a negative one in the last three years. Marked males were recaptured up to 300 m from the release site, with most recaptures occurring within 150 m. Estimated mean daily survival of male An. funestus was 0.86 (95% C.I. 0.869–0.850). For the years 2003–2007, estimated mean daily survival of male An gambiae s.l. was 0.660 (95% C.I. 0.682–0.638). For either species, there was no relationship between mean weekly temperature and estimated daily survival. These results imply that males of An. funestus live as long as females but have a relatively short flight range. They are discussed in the light of possible release strategies of sterile or genetically modified mosquitoes.


INTRODUCTION

  1. Top of page
  2. ABSTRACT:
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgments
  8. REFERENCES CITED

Little is known of the biology of male anopheline mosquitoes in the field. The lack of data has, in part, been due to the difficulty of obtaining a suitable sample and because the male's role in disease transmission and vector control has previously been unimportant. The development of possible control techniques using genetically modified male mosquitoes to introduce genes into the population has highlighted our ignorance on this sex (Ferguson et al. 2005). For such techniques to succeed, released males need to be as fit as wild males (Alphey et al. 2002). Whilst fitness is best measured by assessing lifetime mating success (Charlwood 2003), other, secondary, measures may provide some indication of targets that need to be reached by released mosquitoes for their successful implementation. These include dispersal, especially for techniques that involve the release of sterile males, and survival. Although often considered to be more fragile than females, in mark-release studies with Anopheles gambiae males were recaptured for as long as females and, despite a generally lower dispersal, the mosquito with the greatest flight range was a male (Gillies 1961). In a much smaller study Charlwood et al. (1999) also found that recapture rates among male An. funestus Giles were similar to females.

Malaria vectors in Africa have evolved with humans and are exceptionally synanthropic. Anopheles funestus in particular is extremely endophilic (Gillies and DeMeillon 1968). Males, newly emerged virgin females, and older females at all stages of gonotrophic development rest inside houses. The males, young females, and gravid females leave at dusk via the lightest part of the house: the males and young females to swarm (and mate), the gravid females to oviposit. The insects are positively phototropic at this time, which makes them easy to catch. Both An. funestus and An. gambiae s.l. occur in southern Mozambique. A series of collections from 2003–2009 were therefore undertaken from a village where they are common to examine the population dynamics of the males and to determine environmental factors that might affect them. A novel method for estimating survival in such insects is presented, as are estimates of dispersal from capture-recapture experiments. Results imply that males live as long as females but have a relatively short flight range. The results are discussed in the light of possible release strategies of sterile or genetically modified mosquitoes.

MATERIALS AND METHODS

  1. Top of page
  2. ABSTRACT:
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgments
  8. REFERENCES CITED

The village of Furvela (24º 43'S and 35º 18'E), 475 km north of the capital Maputo, straddles EN1, the road that goes from the south to the north of the country. The approximately 5×4 km village has been described by Charlwood et al. (2003a) and Kampango et al. (2010). It is bordered on two sides by the alluvial plain of two river systems. Both provide ample sites for the larvae of Anopheles sp. The Furvela River valley to the north of the village in particular has a considerable amount of local irrigation that provides a large and relatively stable number of small canals that are typical An. funestus breeding sites. The Inhanombe River valley to the east of the village consists largely of beds of the reed used for house construction and sugar cane, used in the production of local alcohol. The river does not flow as fast as the Furvela and is not used for washing or bathing.

Houses in the village are generally built with reed walls and palm thatch roofs. Although most houses have no windows, the majority have a (circa 15 cm) gap between the roof and walls at either end of the house. This provides illumination and ventilation. Doors and doorframes are also generally badly fitting; hence mosquitoes can easily gain access to the inside of the house. Other styles of house include those with zinc sheets for the roof and those made of concrete blocks (which do have windows). Houses are built either in family compounds of three to six houses or as relatively evenly spaced individual homes. Houses, which are often just a single room, have separate kitchens. The life span of an individual house, from just a couple of years to more than ten, depends very much on the ability and care of the constructor, and to the relative effects of termites and bad weather. In 2003, cyclone Japhert passed through Furvela blowing many houses down.

Anopheles funestus is common and perennial. It is an important vector of malaria, and despite control attempts malaria remains a serious health problem. Three members of the An. gambiae complex also occur in the village, but at lower densities. All species are endophilic with males and young females as well as post-prandial females resting inside houses.

Prior to the study no formal interventions against the vectors had been undertaken. Project staff sold a small number of mosquito nets in the village from the end of 2004 and in 2007 a cordon sanitaire, of permanently impregnated nets inside all bedrooms up to 300m inland, was established on both sides of the Furvela River valley.

Mosquito collections

Resting collections. In order to determine the proportion of the population resting inside houses, both indoor and outdoor resting collections were performed between 07:30 and 10:00. Twenty-min indoor collections were made, using a flashlight and aspirator (Charlwood et al. 1995a) in a variety of man-made shelters including houses, kitchens, and latrines. Outdoor collections, at the same time of day, included the examination of tree roots and shaded areas among vegetation. In 2004, pit-shelters for the collection of outdoor resting mosquitoes (WHO 1975) were built at three different locations (two in the shade one in the open) and during their operation searched on an almost daily basis. Collected female mosquitoes were classified as being unfed, part-fed, engorged, semi-gravid, or gravid according to the external appearance of their abdomen.

Exit collections. Collections of mosquitoes as they attempted to leave houses at dusk were undertaken at the open doorway of the house. The opening was covered with a white mosquito-netting curtain and mosquitoes manually aspirated as they landed on it. In a small number of collections, nets were simultaneously placed over the gables and the open doorway to determine the proportion leaving from each available exit. In addition to sets of sentinel houses, in which more than 100 collections were made, different sets of houses (depending on the number of collectors available) were sampled for about four days every week. Although new houses were used in most years, their construction, including such things as the size of the opening between roof and walls and number of inhabitants, was similar to the house they replaced. For most of the collections, mosquitoes were identified, sexed, and separated on the morning after collection, by which time most were dead. In these samples, females were separated into unfed and gravid categories. Occasionally, samples were identified on the night following collection. These insects were separated into unfed, part-fed, blood-fed, semi-gravid, and gravid categories.

On 63 occasions, those leaving from the doorway of a sentinel house were separated into three-min intervals. Intervals were initiated when the first insect was collected and continued until no insects were seen for a period of at least three min. For these collections a torch was used to confirm the absence of mosquitoes arriving once darkness had fallen.

Capture-recapture

Mosquitoes exiting a house situated in the middle of a set of other houses were marked with fluorescent powder and released in a series of experiments undertaken in November and December, 2004. The protocol adopted was similar to that described by Rawlings et al. (1981) with female An. culicifacies. Since the aim was primarily to determine dispersal, mosquitoes were marked with the same color on successive days whilst, on one day, a pulse of a second color was used. The decline in numbers marked with this color provided an indication of survival. Recaptures were undertaken in available houses around the release site. Their distance and direction from the release site was estimated using measurements made with a GPS handheld receiver (Garmin eTrex). On the occasions when the owner was absent from a selected house, a nearby alternative was used. The numbers recaptured per house per day of collection by distance from the release site, rather than absolute numbers recaptured, were used in calculations of dispersal distance. Collections from a small number of swarms were also made.

Collected anophelines were separated into species or species group according to the keys of Gillies and De Meillon (1968) and Gillies and Coetzee (1987). Culicines were identified using the keys of Edwards (1941) or Service (1990). A small sample of anophelines was identified to species by PCR.

Estimates of survival

The daily survival of male mosquitoes was determined assuming the following three conditions:

There is an equal sex ratio.

Males (of all ages) and young females have a similar tendency to rest inside houses and leave at dusk.

The sampling efficiency for the two sexes is the same.

Young females comprise the unfed portion of the female population leaving at dusk (Charlwood et al. 2003b). Despite a possible day or two of difference in their age (since males tend to emerge as adults before females), these are therefore equivalent to the number of newly emerged males (assumption 1). The excess males in the collection are therefore survivors from previous days (equivalent to the parous female population in a biting or light-trap catch). Unlike the females, the males and newly emerged females are available for sampling every day. Hence,

Daily survival = (total number of males-number of newly emerged females)/Total number of males.

Survival was calculated on a weekly basis for both An. funestus and, when they occurred in sufficient number, An. gambiae s.l.

Environmental monitoring

A digital weather station (Delta-T Devices, Cambridge. U.K.) monitored soil and air temperature, rainfall and humidity, wind speed and direction, and solar insolation every hour for some, but not all, of the study. For periods when the weather station was not operational, mean daily air temperatures were obtained from the meteorological stations of Vilanculos (200 km to the north of Furvela) and Inhambane 30 km to the south (http://www.tutiempo.net). Rainfall data was available from the town of Maxixe, 20 km to the south of Furvela. Since the distribution of rainfall is important (20 mm falling on seven consecutive days in a week is likely to have a different effect than 140 mm falling on a single day), in addition to examining the effects of rain and rain days, a modified measure of rainfall was used to estimate effects:

Modified weekly rain = Rain (mm) ×# rain days/7

Data from Vilanculos were used for many of the longitudinal comparisons between mosquito densities and temperature, whilst data from Maxixe was used for the comparison of rainfall and mosquito numbers, since these were the most complete data sets available.

Data were entered into Excel spreadsheets. Different time scales were used for different analyses. Weekly means were used for much of the analysis. Data were also amalgamated into monthly (calendar month) or trimester (13-week) intervals. Lagged cross-correlations between environmental factors and mosquito numbers were undertaken in Minitab14. In order that the program can perform such cross-correlations, a dataset without missing periods is required. For the ten weeks with missing data, the mean of the previous and subsequent weeks were used for the interpolation. On the three occasions when the interval of missing data was two or three weeks, the mean of the two previous and subsequent weeks was used.

RESULTS

  1. Top of page
  2. ABSTRACT:
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgments
  8. REFERENCES CITED

Species identification

Of the 30 males and 407 females from an unselected sample of the An. gambiae complex identified by PCR between 2002 and 2005, 86%, and 83%, respectively, comprised Anopheles gambiae, the other species being An. arabiensis (13%) and An. merus (1%) (J. Pinto, personal communication). Thus, even though the species may have different patterns of behavior, the predominant one observed in the present study would always have been that of An. gambiae.

All of An. funestus examined morphologically had a single pale spot on the upper branch of the 5th vein and did not have a pale spot at the tip of the 6th vein and corresponded to An. funestus, and even though only 71 females of the An. funestus group were identified by PCR, all were An. funestus (A.L Szalanski and J. Austin, unpublished data). Given that this is the endophilic member of the species group, and that it was endophilic behavior that we studied, we also assume that this was the only member of the species group present in our collections.

Resting

Both An. funestus and An. gambiae were readily collected inside houses (Table 1) and to a lesser extent in disused latrines and other buildings. In sites close to the valley edge, approximately 20% of males had partially rotated terminalia. The only outdoor sites in which mosquitoes were regularly collected were clay pots that had some water in them. These were always adjacent to, or close by, houses and the behavior cannot be considered to be different to indoor resting. After 60 days of mosquito collection from the pit-shelters, fewer than 20 were collected.

Table 1.  Mean number (± 95% confidence intervals) of Anopheles funestus and An. gambiae s.l by trimester collected per 20 min resting collection inside houses from Furvela, southern Mozambique.
  Anopheles funestusAnopheles gambiae s.l.
PeriodCountUnfedPart-fedEngorgedSemi-gravidGravidMaleUnfedPart-fedEngorgedSemi-gravidGravidMale
Jan-Mar2482.39 (0.65)0.60 (0.16)4.90 (0.90)0.23 (0.10)3.83 (0.67)6.82 (1.03)2.44 (0.68)0.53 (0.17)3.98 (0.74)0.22 (0.15)1.79 (0.37)2.58 (0.52)
Apr-Jun4332.42 (0.30)0.93 (0.19)7.03 (0.91)0.96 (0.25)4.59 (0.59)7.97 (0.81)0.96 (0.24)0.62 (0.17)2.32 (0.47)0.35 (0.13)0.96 (0.25)2.30 (0.42)
Jul-Sep2993.31 (0.47)0.60 (0.26)2.07 (0.67)0.61 (0.27)3.30 (0.59)13.70 (1.98)0.14 (0.05)0.06 (0.04)0.20 (0.07)0.02 (0.02)0.11 (0.06)0.13 (0.06)
Oct-Dec4121.96 (0.25)0.84 (0.46)3.11 (0.67)0.27 (0.13)3.39 (0.38)8.27 (0.90)0.59 (0.17)0.24 (0.07)1.57 (0.28)0.06 (0.04)0.48 (0.12)1.49 (0.28)

There was a deficit of male An. gambiae in the indoor collections in relation to unfed females compared to the number of male and unfed female An. funestus (Chi sq, 324, p >0.001). This may reflect a greater tendency to exophiliy in at least one of the An. gambiae species present or a difference in catchability between the sexes. (Males of An. gambiae may be more easily disturbed than females). The small numbers of An. gambiae s.l. collected and the great variety of possible sites outside may have been responsible for their absence in outdoor collections.

Exit collections

In 8,348 exit collections, 567,195 male An. funestus and 34,591 male An. gambiae s.l. were collected. Also collected were 76,980 unfed and 95,347 gravid female An. funestus and 13,109 unfed and 19,047 gravid female An. gambiae s.l. Totals of 8,696 female mosquitoes (86% being Culex quinquefasciatus) and 6,625 males (96% being Cx. quinquefasciatus) comprising eight other species were also collected. The great majority of collections (8,031) were performed on the Furvela river valley side of the village and it is these that will be discussed further. The number of males collected from the Furvela valley side in each year of the study, mean temperatures, and estimated number of rain days at Maxixe are given in Table 2.

Table 2.  Mean annual temperatures (measured at Vilanculos), number of rain days, and rainfall (recored at Maxixe), correlation co-efficient between temperature and mean number of male mosquitoes collected per week; mean number per house per day, and estimated survival of male Anopheles funestus and An. gambiae s.l. from Furvela, southern Mozambique.
    An.funestusAn. gambiae
YearTemp (°C)Rain daysRain (mm)r2Number collectedsurvivalr2Number collectedsurvival
200324.69831322.70.56144.320.800.3594.940.69
200425.9473973.20.48851.280.830.6008.210.58
200526.01111652.50.26839.170.880.68112.580.67
200625.68651036.30.29974.390.840.5442.290.77
200725.13861232.2–0.15294.200.850.6010.50 
200825.1783914–0.61683.950.860.4670.06 
200925.5995924.2–0.25857.370.890.5100.12 

Time of exit

Males of both An. funestus and An. gambiae s.l started leaving approximately one min before sunset on clear nights and up to six min before sunset on cloudy nights. Although it was not possible to precisely collect all of the insects arriving at the net during any one three-min period during peak exit times, exit patterns clearly differed between sexes but not between species. Figure 1 shows exit patterns of male, unfed, and gravid An. funestus and An. gambiae s.l., relative to the time of the first mosquito collected, from the doorway. The number of males collected reached a maximum in the first six min of collection coincident with the first swarms of either species being seen outside. Both gravid females and unfed females had similar exit patterns with a peak between six to ten min after the males. The number of insects collected by the time it was dark for almost all nights of collection was zero. Even at high densities, none of the males appeared to attempt to mate with females arriving at the net. An average of 4.6 (95% C.I. 6.2–3.1) males were collected at the side compared to 45.4 (95% C.I. 50.7–40.1) from the door in 107 paired collections from houses with a gable gap in excess of 15 cm. Thus, the proportion of mosquitoes leaving via the gable ends was less than 11% of those at the door when nets were placed over both the open doorway and the gable ends of houses.

image

Figure 1. Time of collection of male, unfed, and gravid mosquitoes, relative to the time of the first mosquito collected, attempting to leave the open door of reed houses from the village of Furvela, Mozambique. A) Anopheles funestus B) Anopheles gambiae s.l.

Download figure to PowerPoint

Population dynamics

The mean number of males and unfed females exiting per house per week from March, 2003 to December, 2009 and weekly mean temperatures recorded at Vilanculos are shown in Figure 2a for An. funestus and Figure 2b for An. gambiae. The An. gambiae population declined to extinction during the study.

image

Figure 2 A). Mean number of males and unfed female An. funestus and B) male An. gambiae s.l. exiting per house per week from March, 2003 to December, 2009 and weekly mean air temperatures recorded at Vilanculos.

Download figure to PowerPoint

In all years the number of An. gambiae collected were positively correlated with temperature (Table 2). The pattern observed in the An. funestus was more complex. Density was positively correlated with temperature in the first years of the study, but from 2006 onwards numbers were greatest in the cooler periods of the year, leading to a negative correlation between temperature and mosquito numbers (Table 2).

Overall, densities of An. funestus showed little seasonality, whilst densities of An. gambiae were lower in the cooler months of the year (Figure 3). Lagged correlation coefficients between An. gambiae and temperature (measured at Vilanculos) were positive and peaked with a three-week delay, whilst overall they were negative for An. funestus, reaching a minimum at 14 weeks (Figure 4).

image

Figure 3. Weekly mean densities of male Anopheles funestus and An. gambiae attempting to leave the open door of reed houses from the village of Furvela, Mozambique, and weekly mean air temperatures recorded at Vilanculos. (dotted lines are 95% confidence intervals).

Download figure to PowerPoint

image

Figure 4. Pearson correlation co-efficient between weekly mean numbers of male mosquitoes collected and weekly mean air temperatures recorded at Vilanculos by weekly lag.

Download figure to PowerPoint

The relationship between rainfall, rain days per week, or the adjusted measure of rain measured at Maxixe and mosquito numbers in Furvela followed a lower, more irregular but similar pattern to temperature. Numbers of An. gambiae were positively correlated and An. funestus negatively correlated with rainfall for lags up to 14 weeks but these relationships were weaker and less regular than those of temperature (Figure 5).

image

Figure 5. Pearson correlation co-efficient between weekly mean number of male mosquitoes collected attempting to leave the open door of houses from the village of Furvela, Mozambique, and adjusted rainfall measured at Maxixe, 21 km to the south, by weekly lag.

Download figure to PowerPoint

In the early years of the study, during periods when other environmental conditions were stable, both An. funestus and An. gambiae showed similar trends as temperatures increased. Numbers of male Anopheles funestus and An. gambiae s.l. collected attempting to leave a sentinel house in the village of Furvela, Mozambique, from the 1st of July to the 22nd of October, 2005 and mean daily soil temperatures measured 150 m distant are shown in Figure 6. In this period, mean daily temperatures rose from 18° to 31° C, numbers of An. funestus collected rose ten-fold (from 20 to 200 per collection), whilst numbers of An. gambiae twenty-fold (from one to 20 per collection). The regression equation for the expected number of An. funestus collected was (13.184 × Temperature) – 274.08 (r2= 0.599), whilst numbers of An. gambiae were (15.173 × Temperature) – 295.28 (r2= 0.614). The relationship between log number collected and temperature for this period is shown in Figure 7.

image

Figure 6. Numbers of male Anopheles funestus and An. gambiae s.l. collected attempting to leave a sentinel house in the village of Furvela, Mozambique, from 1st July to 22nd October 2005 and mean daily soil temperatures measured 150 m distant.

Download figure to PowerPoint

image

Figure 7. Number of male An. funestus (left) and An. gambiae s.l. (right) collected attempting to leave a sentinel house in the village of Furvela, Mozambique, according to mean daily soil temperature.

Download figure to PowerPoint

Survival

For both species, or species group, there was no relationship between temperature and estimated survival, which was always higher in the An. funestus than the An. gambiae s.l. and similar to estimates of female survival from a number of other studies (Gillies and Wilkes 1965, Charlwood et al. 1995b, Mutero and Birley 1989) (Figures 8a and 8b). Overall estimated survival of male An. funestus from exit collections was 0.849 (95% C.I. 0.858–0.839) per day. Survival could only be calculated among male An. gambiae for the first three years of the study. From 126 weeks of observation, daily survival was estimated to be 0.660 (95% C.I. 0.682–0.638). The annual mean estimates of daily survival among the An. funestus and An. gambiae are given in Table 2.

image

Figure 8. A) Estimated daily survival rate of male Anopheles funestus and B) Anopheles gambiae s.l. collected attempting to leave houses in the village of Furvela, Mozambique, according to mean daily air temperatures recorded at Vilanculos.

Download figure to PowerPoint

Capture- recapture

The number of mosquitoes recaptured relative to the estimated number released was small. Nevertheless, a total of 85 marked mosquitoes was recaptured over a period of 25 days. Recapture rates declined rapidly with distance from the release site (Figure 9), most insects being caught in houses within 100 m of the release point. Two marked mosquitoes were also collected in a sample of 33 mosquitoes collected from the swarm close to the release site. No marked mosquitoes were collected from samples made from two houses 500 and 950 m distant from the release site.

image

Figure 9. Mean number of marked male Anopheles funestus per collection by distance from the release site, Furvela, Mozambique.

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. ABSTRACT:
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgments
  8. REFERENCES CITED

Here, for the first time, we present longitudinal data on the male population of malaria vectors in southern Africa. The data indicate that males of An. funestus survive as long as females but have a limited dispersal and that males of An. gambiae survive less than females. The dearth of mosquitoes from the pit-shelters and the frequency with which they were collected in houses indicates that human-made structures are the resting site for the great majority of An. funestus in Furvela. As pointed out by Gillies (1954), the search for outdoor resting mosquitoes is an arduous one that can provide little reward and the location of artificial resting shelters will affect their catch, those that collect mosquitoes often acting as surrogate houses. Thus, in the present studies, the mosquitoes collected from clay pots in the vicinity of houses would otherwise have been collected indoors. Alternative outdoor sites are rare in Furvela especially close to houses, which are often clear of all vegetation other than the occasional coconut palm or mango tree. The An. gambiae population consisted of at least two species, each of which may have different tendencies to rest indoors. This may influence survival rate estimates, which were consistently lower in An. gambiae s.l. than An. funestus. Both species (or species group) leave houses at dusk. Exit collections are an easy way of sampling these insects. The technique is simple, efficient, risk free, does not require an external energy source, is easy to supervise, and cheap, the main cost being the small remuneration given to the collectors. The caveat is that if one is late, then the sample will be lost. Although mosquitoes will leave through the opening between the gable ends and the roof of a house when the door is closed, when it is open the numbers leaving by that route are small compared to the numbers leaving through the doorway.

Males of both An. funestus and the An. gambiae complex showed similar exit behaviors. Males of both species (or species group) started leaving and peaked before the females. In addition to being endophilic, An. funestus is eurygamic, mating in open areas (Charlwood et al. 2005). The absence of any attempts to mate, and the dearth of recently mated females in light-traps following an evening of heavy rain (Charlwood, unpublished data) implies that little, if any, mating takes place inside houses in An. funestus as it might do in An. gambiae (Dao et al. 2010). Exit patterns at dusk of females were independent of abdominal condition or mated status. This is a further indication that mated status does not affect activity in anophelines (Klowden 2001, Charlwood et al. 2003b). Close to almost all the houses in Furvela, males form swarms at dusk at the same time that gravid females fly out to oviposit. Swarms of males may distract potential predators, such as dragonflies that were seen to prey on swarming mosquitoes, from these females.

In the laboratory, males of the different members of the An. gambiae complex show slightly different activity patterns following an artificial dusk, with An. gambiae being active three or four min earlier than An. arabiensis. Although it would be interesting to know if differences also occur in exit times between the species, unfortunately the An. gambiae from these collections were not identified beyond species group.

The dynamics of An. funestus and An. gambiae differed. The change in fortunes of the species is all the more puzzling since in other measured behaviors, they were very similar. The less common An. gambiae were consistently positively associated with temperature measured 200 km to the north and, to a lesser extent, rainfall 20 km to the south, even during the latter part of the study when numbers collected were small. The decline in numbers of An. gambiae may have been due to there being an increasing number of bednets in the village during the study (culminating in a cordon sanitaire around the valley in 2007). This does not, however, explain the changes observed in the An. funestus. The population structure of the An. funestus that were identified by PCR in 2002 and 2008 were similar (A.L. Szalanski and J. Austin, personal communication), indicating that no profound population change had occurred as a result of the introduction of the nets. Nevertheless, the An. funestus that were sampled in 2009 were resistant to the insecticide used on the nets (Charlwood et al. unpublished observations). The transition between the positive correlation between the number of male An. funestus and temperature (and the estimated number of rain days per month) and the rise in numbers collected is therefore difficult to explain. Whether this was due to the increasing numbers of mosquito bednets in use is unknown. Nets should not affect males and young females since these insects only go into houses to rest on the walls not the net with its relatively non-volatile insecticide. The results therefore indicate a considerable degree of plasticity in the ecology of An. funestus. The population at any moment may depend on small differences in initial conditions (i.e., be chaotic). In other, earlier, studies of female An. funestus, a variety of seasonal patterns have been described. In general, there is one seasonal peak that occurs late in the wet season (Gillies and DeMeillon 1968). In some areas, however, there is little or no seasonality in numbers (Lamborn 1925, Garnham 1929, quoted in Gillies and DeMeillon 1968, Charlwood et al. 1997). Availability of suitable water for the larvae is the predominant factor in their dynamics (Gillies and DeMeillon 1968) and presumably the changes observed (in both gambiae and funestus) in the present study must have something to do with changing habitat availability. As long as sufficient water is available, a reduction in rainfall might reduce the flushing and drowning effects that rain can have on mosquito larvae whilst reducing the availability of habitat used by An. gambiae. In Furvela, local irrigation provided such conditions. No obvious changes in agriculture were, however, recorded in the Furvela valley until 2008 when upstream irrigation was initiated, reducing the level of water in the canals close to the study area.

The greater rate of increase in the An. gambiae compared to the An. funestus recorded in 2005 probably reflects the faster development time of the larvae of the former species compared to the latter. All members of the Anopheles gambiae complex are known for their rapid larval development, whilst An. funestus is known to have a long larval development time (Gillies and DeMeillon 1968). Even though they may share the same water, Anopheles gambiae larvae may seek out shallow sunlit surfaces and An. funestus shaded deeper ones, effectively exposing them to two different temperature regimes. In the present study, temperature measured 200 km distant was a better predictor of mosquito numbers than rainfall measured 20 km distant. How far the effective predictive radius of temperature might be has yet to be determined. Given the changes observed in the relationship between An. funestus and temperature during the study, it appears that the information cannot be used to predict the dynamics of An. funestus nor can it be used to estimate actual densities of An. gambiae since mean annual temperatures were similar throughout the study period yet the An. gambiae steadily declined.

Mating may have a cost in male mosquitoes (Dao et al. 2010), especially in those males that contribute mating plugs as part of the process, as do An. gambiae and An. funestus. Nevertheless, although fragile and dying easily when collected, male mosquitoes have an easier life than females, so apart from the problems associated with mating, there should be no reason for them not to survive as well, and estimated daily survival rates of male An. funestus at all times and temperatures were similar to those of females from many other studies (Gillies and Wilkes 1965, Charlwood et al. 1985a, 1995b, 1997, Magbity and Lines, 2002, Lines et al. 1991).

Given the greater fragility of males during collection, marking and releasing may have had an impact on both dispersal and survival. The powders used in the present experiments were those used in the experiments of Birley and Charlwood (1989). They point out that the quality and type of dust used in marking studies can affect adherence to the mosquito and thus affect the results. The data from the release-recapture experiments need, therefore, to be viewed with caution. Dusting itself may enhance dispersal (Mallett et al. 1987). Measured dispersal in males was almost certainly influenced by the density as well as the distance (but not apparently the direction) of the houses surrounding the release site and may differ in other areas. A leptokurtic distribution (with most individuals staying close to the release site but a few making relatively long flights) is a relatively common one for many animal populations (Endler 1977). Dispersal in An. funestus females is thought to be relatively limited (DeMeillon 1934, Charlwood et al. 1998). Unlike females, who must go there to lay their eggs, there is no apparent reason for males to return to the river valley. Hence a greater amount of random drift in their dispersal compared to females might be expected. It makes sense, however, for the males to behave as virgin females do and it behooves males to remain in the vicinity of virgin females if their chances of mating are to be enhanced. Although it is not surprising that dispersal measured by capture-recapture was limited, in other routine collections males in considerable numbers were caught from houses more than 800 m from the valley edge, the putative site of emergence.

We do not know what the males feed on. Plant sugar is an important food source for females and is mandatory for males (Gu et al. 2011). If virgin females take a sugar meal and have a preference for a particular plant that is limited in its distribution, then males that have the same preference may be at a selective advantage of mating.

Although at present the main bottleneck in genetically modified mosquito technology is the lack of adequate gene drive mechanisms, our results indicate that for release strategies to be successful in areas like Furvela, the number of locations and frequency of releases will need to be considerable for successful implementation. If adults are released, then for area wide control individual releases will need to be made at intervals of 150 m, and if the modified insects have a reduced fitness compared to wild ones, releases will also need to be frequent and take place for a protracted period. Thus, even when a suitably modified mosquito is available in the laboratory, their effectiveness in the field remains in question.

The dearth of An. funestus from outdoor resting collections implies that outdoor resting is not a usual alternative for this mosquito in Furvela (as it is for An. gambiae in São Tomé (Souza et al. 2001) or An. farauti from Papua New Guinea (Charlwood et al. 1985b). Since the local An. funestus is resistant to the insecticide used on nets (Charlwood et al. unpublished data) spraying the inside walls of houses with a residual insecticide, perhaps as a cordon sanitaire (Charlwood et al. 1998), remains the most logical method for its control.

Acknowledgments

  1. Top of page
  2. ABSTRACT:
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgments
  8. REFERENCES CITED

I thank the staff of the MOZDAN project for their help in collecting and identifying the mosquitoes and the collectors and villagers who cheerfully allowed us to catch mosquitoes from their houses. Thanks to Joao Pinto of CMDT, Portugal, for identifying the members of the An. gambiae complex and Alan Szalanski and James Austin for identifying the An. funestus. I would like to thank the ARI-Sul for supplying information on the rainfall from Maxixe. The study was financed by DBL-Centre for Health Research & Development, Denmark.

REFERENCES CITED

  1. Top of page
  2. ABSTRACT:
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgments
  8. REFERENCES CITED
  • Alphey, L., C.B. Beard, P.F. Billingsley, M. Coetzee, A. Cristiani, C.F. Curtis, P. Eggleston, C. Godfray, J. Hemingway, M. Jacobs-Lorena, A.A. James, F.C. Katatos, L.G. Mukwaya, M. Paton, J.R. Powell, W. Schneider, T.W. Scott, B. Sina, R. Sinden, S. Sinkins, A. Spielman, Y. Touré, and F.H. Collins. 2002. Malaria control with genetically manipulated insect vectors. Science 298: 119121.
  • Birley, M.H. and J.D. Charlwood. 1989. The effect of moonlight and other factors on the ovipositon cycle of malaria vectors in Madang, Papua New Guinea. Ann. Trop. Med. Parasitol. 8: 415422.
  • Charlwood, J.D. 2003. May the force be with you: measuring mosquito fitness in the field. In: T. Scott and W. Takken (eds.) Proceedings of a Workshop on Transgenic Mosquitoes. Wageningen University Press, Vol. 2. 450 pp.
  • Charlwood J.D., J. Alcântara, J. Pinto, C. Sousa, H. Rompão, V. Gil, and V. de Rosário. 2005. Do bednets reduce malaria transmission by exophagic mosquitoes? Trans. R. Soc. Trop. Med. Hyg. 99: 901904.
  • Charlwood, J.D., P.F. Billingsley, W. Takken, E.O.K. Lyimo, T. Smith, and J.M.E. Th. Meuwissen. 1997. Survival and infection probabilities of anthropophagic Anophelines from an area of high prevalence of Plasmodium falciparum in humans. Bull. Entomol. Res. 87: 445453.
  • Charlwood, J.D., M.H. Birley, H. Dagaro, R. Paru, and P.R. Holmes. 1985a. Assessing survival rates of Anopheles farauti (Diptera, Culicidae) from Papua New Guinea. J. Anim. Ecol. 54: 10031016.
  • Charlwood, J.D., H. Dagaro, and R. Paru. 1985b. Blood-feeding and resting color in the Anopheles punctulatus Donitz complex (Diptera, Culicidae) from coastal Papua New Guinea. Bull. Entomol. Res. 75: 463475.
  • Charlwood, J.D., J. Kihonda, S. Sama, P.F. Billingsley, H. Hadji, J.P. Verhave, E.O. Lyimo, P.C. Luttikhuizen, and T. Smith. 1995b. The rise and fall of Anopheles arabiensis (Diptera, Culicidae) in a Tanzanian village. Bull. Entomol. Res. 85: 3744.
  • Charlwood, J.D., C. Mendis, R. Thompson, K. Begtrup, N. Cuamba, M. Dgedge, A. Gamage-Mendis, R.H. Hunt, R.E. Sinden, and B. Høgh. 1998. Cordon-sanitaire or laissez-faire, differential dispersal of young and old females of the malaria vector Anopheles funestus in southern Mozambique. African Entomol. 6: 16.
  • Charlwood, J.D., J. Pinto, C.A. Sousa, C. Ferreira, V. Gil, and V. de Rosario. 2003b. Mating does not affect the biting behaviour of Anopheles gambiae from the islands of São Tomé and Príncipe, West Africa. Ann. Trop. Med. Parasitol. 97: 751756.
  • Charlwood, J.D., T. Smith, J. Kihonda, P.F. Billingsley, and W. Takken. 1995a. Density independent feeding success of malaria vectors in Tanzania. Bull. Entomol. Res. 85: 2935.
  • Charlwood, J.D., R. Thompson, and H. Madsen. 2003a. Swarming and mating in Anopheles funestus from southern Mozambique. Malar. J. 2: 3.
  • Charlwood, J.D., R. Vij, and P.F.B. Billingsley. 1999. Dry season refugia of malaria-transmitting mosquitoes in a dry savannah zone of East Africa. Am. J. Trop. Med. Hyg. 62: 726732.
  • Dao, A., Y. Kassogue, A. Adamou, M. Diallo, A.S. Yaro, S.F. Traore, and T. Lehmann. 2010. Reproduction-longevity trade-off in Anopheles gambiae (Diptera: Culicidae). J. Med. Entomol. 47: 769777.
  • DeMeillon, B. 1934. Entomological studies – Observations on Anopheles funestus and Anopheles gambiae in the Transvaal. Pub. S. African Inst. Med. Res. Johannesburg, South Africa 32: 195248.
  • Edwards, F.W. 1941. Mosquitoes of the Ethiopian Region. III Culicine Adults and Pupae. British Museum (Natural History), London . 499 pp.
  • Endler, J.A. 1977. Geographic Variation, Speciation and Clines. Monograph in Population Biology, No. 10, 295 pp.
  • Ferguson, H.M., B. John, K. Ng'habi, and B.G.J. Knols. 2005. Redressing the sex imbalance in knowledge of vector biology. Trend. Ecol. Evol. 20: 202209.
  • Gillies, M.T. 1954. Studies of the house-leaving and outside resting of Anopheles gambiae Giles and Anopheles funestus Giles in East Africa. II – The exodus from the houses and the house resting population. Bull. Entomol. Res. 45: 375387.
  • Gillies, M.T. 1961. Studies on the dispersion and survival of Anopheles gambiae Giles in East Africa, by means of marking and release experiments. Bull. Entomol. Res. 52: 99127.
  • Gillies, M.T. and M. Coetzee. 1987. A supplement to the Anophelinae of Africa south of the Sahara (Ethiopian zoogeographical region). Pub. S. African Ins. Med. Res. No. 55. Johannesburg, South Africa.
  • Gillies, M.T. and B. DeMellion. 1968. The Anophelinae of Africa South of the Sahara (Ethiopian Zoogeographical Region). 2nd Ed. Publications of the South African Institute for Medical Research No. 54. Johannesburg, South Africa .
  • Gillies, M.T. and T.J. Wilkes. 1965. A study of age composition of population of Anopheles gambiae Giles and Anopheles funestus Giles in North-eastern Tanzania. Bull. Entomol. Res. 56: 129135.
  • Gu, W., G. Müller, Y. Schlein, R.J. Novak, and J.C. Beier. 2011. Natural plant sugar sources of Anopheles mosquitoes strongly impact malaria transmission potential. PLoS ONE 6: e15996. doi: 10.1371journal.pone.0015996.
  • Kampango, A., N. Cuamba, and J.D. Charlwood. 2010. Does moonlight influence the biting behaviour of Anopheles funestus (Diptera: Culicidae)? Med. Vet. Entomol. doi: 10.1111j.1365-2915.2010.00917.x.
  • Klowden, M.J. 2001. Sexual receptivity in Anopheles gambiae mosquitoes: absence of control by male accessory gland substances. J. Insect Physiol. 47: 661666.
  • Lines J.D., T.J. Wilkes, and E.O. Lyimo. 1991. Human malaria infectiousness measured by age-specific sporozoite rates in Anopheles gambiae in Tanzania. Parasitology 102:167177.
  • Magbity, E.B. and J. Lines. 2002. Spatial and temporal distribution of Anopheles gambiae s.l. (Diptera: Culicidae) in two Tanzanian villages: implication for designing mosquito sampling routines. Bull. Entomol. Res. 92: 483488.
  • Mallett, J., J.T. Longino, D. Murawski, A. Murawski, and A. Simpson de Gamboa. 1987. Handling effects in Heliconius: where do all the butterflies go? J. Anim. Ecol. 56: 377386.
  • Mutero, C.M. and M.H. Birley. 1989. The effect of pre-gravid development on the estimation of mosquito survival rates. J. Appl. Ecol. 107: 96101.
  • Rawlings, P., C.F. Curtis, M.B. Wickramasinghe, and J. Lines. 1981. The influence of age and season on dispersal and recapture of Anopheles culicifaces in Sri Lanka. Ecol. Entomol. 6: 307319.
  • Service, M.W. 1990. Handbook to the Afrotropical Toxorhynchitine and Culicine Mosquitoes, Excepting Aedes and Culex. British Museum (Natural History), London , 207 pp.
  • Sousa, C.A., J. Pinto, P.G. Almeida, C. Ferreira, V.E. do Rosario, and J.D. Charlwood. 2001. Dogs as a favored host choice of Anopheles gambiae sensu stricto (Diptera, Culicidae) of São Tomé, West Africa. J. Med. Entomol. 38: 122125.
  • W.H.O. 1975. Manual on Practical Entomology in Malaria. Part I. Division of Malaria and Other Parasitic Diseases, World Health Organization, Geneva , Switzerland , 196 pp.