Anopheles gambiae mates in flight. Males gather at stationary places at sunset and compete for incoming females. Factors that account for male mating success are not known but are critical for the future of any genetic control strategy. The current study explored variations in nutritional reserves (sugars, glycogen, lipids, and proteins) in wild-caught swarming and resting males and evaluated the effect of body size and wing symmetry on male mating success. Our results showed that glycogen and sugar reserves are mobilized for flight. Males consume proportionally 5.9-fold as much energy derived from sugars in swarming activities than when they are at rest. Mated males were on average bigger than unmated ones (P<0.0001). A strong correlation between the left and right wings in both mated and unmated males was found and additional analysis on fluctuating asymmetry did not show any indication of mated males being more symmetrical than unmated ones. The distribution of wing size of mated males was focused around a central value, suggesting that intermediate size of males is advantageous in the An. gambiae mating system. The results are discussed in the context of sexual selection.
Malaria vector control programs rely heavily on the use of insecticide-treated nets (ITNs) and indoor residual spraying (IRS) with insecticide. Unfortunately, resistance to almost all classes of insecticide has emerged in anopheline mosquitoes and its rapid spread suggests that it will become a major hindrance to malaria control. Alternative measures are urgently needed. New control approaches envision rendering wild vector populations less susceptible to infection by releasing mosquitoes that are genetically modified (GM) to resist infection (Ferguson et al. 2005). Another approach that may be effective is based on sterile male release. However, availability of these tools does not necessarily guarantee success. One concern is the presence of reproductive barriers that limit the spread of the refractoriness genes between subpopulations. Another concern is the possibility that laboratory-adapted mosquitoes will not be able to compete for mates in the wild and therefore the genes of interest will not be integrated into the natural population gene pool. Success will depend on understanding mosquito reproduction systems, a facet of mosquito biology that is poorly understood. Failures in releasing sterile males in the past (Benedict and Robinson. 2003 and references therein) and a growing awareness of the importance of understanding male reproductive success have led to repeated calls for addressing this neglected area in mosquito biology (Ferguson et al. 2005).
In mosquitoes, nutritional reserves accumulated during larval development and from sugar-feeding as adults are critical determinants of adult survival and mating success (Van Handel 1988, Briegel 1990, Ng'habi et al. 2005). For instance, lipids are required for long-term maintenance and are primarily acquired from feeding during larval development and from sugar feeding as adults (Briegel et al. 2001, Ziegler and Ibrahim. 2001, Pascini et al. 2011). Flight is a requirement for mosquito mating and this activity is fuelled by sugars and glycogen derived from sucrose or its components fructose and glucose, obtained from plants (Foster 1995). Hence, a comprehensive study on energy variation levels of mosquitoes taken at different physiological states and time of day would estimate energy allocated to essential activities such as swarming and mating.
In this study, we measured variations in energetic reserves (proteins, sugars, glycogen, and lipids) in mosquitoes from the field (swarming and resting males). Further, we looked at the correlation between male body size, wing symmetry, and male mating success. Finally, the correlation between males and females caught in copula was also assessed to find out if size-assortative mating operates in An. gambiae.
MATERIALS AND METHODS
The survey was conducted in Vallée du Kou 7 (4°25′ W, 11°24′ N), one of the seven discrete districts of the rice growing area of Bama, northwest of Burkina Faso. The irrigation system had been installed in the 1970s and covers 1,200 ha with a total of 4,470 inhabitants. The rainy season extends from June to October and the dry season from November to May. The Kou River is a permanent source of irrigation water and there are two rice crops per year from July to November and from January to May. The VK7 district is located at the periphery of the rice fields and has about 600 inhabitants, mainly farmers. Livestock is mainly composed of sheep, goats, pigs, and cows. Cotton and maize fields surround the village. The irrigation system and rice fields provide year-round mosquito larval sites. Additional larval sites are created by rains in the depressions and ponds (Baldet et al. 2003, Diabaté et al. 2005). The two molecular forms of An. gambiae occur in sympatry, however the M form is predominant (Diabaté et al. 2003).
Male mosquitoes were collected from swarms (July to August, 2010) using a sweeping net. Early morning and late afternoon collections were also done in resting places indoor using a mouth aspirator. All samples were anesthetized and transported in a cold box. They were frozen at –80° C 12 h after they were collected. Newly emerged males were collected from breeding sites using emergence traps. Total sugars, glycogen, lipids, and protein contents were assessed in individual specimens. Each individual male was briefly crushed in methanol and separated into two fractions: one for protein analysis and the other for lipids, sugar, and glycogen. Chloroform–methanol (1:2) was added in the second fraction then centrifuged and the supernatant was separated into two fractions: one for lipid analysis and the other one for sugar analysis. The precipitate was kept for glycogen analysis. A colorimetric test was used to quantify the different metabolites following a modified spectrophotometric protocol (Van Handel, 1985a,b) by Rivero and Ferguson (2003).
Standard curves for converting absorbency readings into quantities of total sugars, glycogen, lipids, and proteins were obtained from the absorbency of known concentrations. Protein, lipid, sugar, and glycogen concentrations were respectively obtained from a standard curve based on bovine serum albumin (Sigma-Aldrich), vegetal oil, and glucose (Sigma-Aldrich).
A total of 162 free flying males and 80 males in copula was collected in six swarms over nine days for wing symmetry measurements. Of the free flying males, 82 males' energetic reserves were assessed and compared to 30 morning resting and 30 evening resting males. Fifty-three newly emerged males from eleven breeding sites were analyzed. All the samples were identified as An. gambiae M molecular form.
Body size and symmetry
Wing lengths (left and right) were measured as described previously (Huestis et al. 2011, Diabaté et al. 2011). Both wings were dissected, mounted dried on microscope slides, and photographed with Leica EZ4 D (Leica Microsystems, Suisse). Their size was then measured using the software Image J1.41.0 (Wayne Rasband, National Institutes of Health, U.S.A.) from the annular notch to the end of the radius vein (excluding fringe scales). This length raised to the cube (WL3) was considered an index of mosquito size (Briegel 1990, Fernandes and Briegel 2005).
Data were analyzed using R 2.12 and GraphPad Prism 5.0 software. The energy values in Joules (J) of metabolites were obtained by multiplying the corresponding metabolite amount obtained from calibration curves by 16.74 J per mg for sugars, glycogen, and protein and 37.65 J per mg for lipids (Clements 1992). These energy values were then divided by the cube of the size of the wing (WL3 in mm3) to take into account the size of the specimen and so standardize the measurements. Analysis of the different metabolites indicated that they did not follow a normal distribution, hence the non-parametric Mann-Whitney test was used to assess differences in energetic reserves between newly emerged mosquitoes and other adult mosquitoes collected indoor and in swarms. Comparison of the different metabolites was also done between the morning (07:00) and the evening (17:00) indoor male and swarm collections (18:30–19:00) using the Kruskall Wallis test. Differences in wing size as a proxy of body size between males caught in copula and free-flying males in swarms were tested by the Student t test. The coefficients of variation (CV) of the size distribution for males caught in copula and the free-flying males in swarms were compared to test for the magnitude of size variation in the two groups using the Fligner-Killeen test in PAST version 2.10 software. The Pearson's correlation coefficient was used to test for wing symmetry in males caught in copula as well as those collected free-flying in swarms. In addition, we used least-square regression analysis to look at the relationship between fluctuating asymmetry and wing length. Fluctuating asymmetry was calculated as the absolute difference between right and left wing lengths. Slopes of the two regression lines were then compared on the assumption that they were equal. A comparison of the mean differences of wing size (left and right) between mated males and unmated ones in fluctuating asymmetry was carried out with the Student t test. Each wing was measured three times to separate measurement error from asymmetry. The Pearson correlation analysis was also used to test for association between size of females and males in copula and the magnitude of size variation between males and females caught in copula was tested by comparing the coefficient of variation (CV) of their body size.
Variation in nutritional reserves with respect to time and activity
The reserves of total sugars, glycogen, lipids, and proteins converted in Joule were determined in different groups of mosquitoes. Total sugars were significantly higher in indoor males than in newly emerged males (Figure 1a, Mann-Whitney U=218, P<0.0001) as well as males from swarms (U= 1204, P<0.0001). No significant difference was found between newly emerged males and the groups of males from swarms (U=716, P= 0.38). However, sugar content was significantly higher in swarms when newly emerged males were compared to males collected in swarms only five min after the swarms had started (U= 110, P= 0.0028 data not shown). A significant difference in glycogen content was observed only between indoor males and males from swarms (Figure 1b, U=1879, P= 0.0146), but not between newly emerged males and indoor males (U= 521, P= 0.37), nor between newly emerged males and males from swarms (U=706, P= 0.32). The lipid and protein contents did not significantly vary between groups (Figures 1c and 1d, P>0.05).
Energy reserves of the above metabolites were then compared among indoor morning males, indoor evening males, and males collected in swarms (Table 1). Significant differences in total sugar amounts (P<0.0001, Kruskal Wallis test) and glycogen (P=0.039 Kruskal Wallis test) were observed among the three groups. The morning indoor males exhibited the highest reserves content, suggesting that 59% of the total sugars are burned at resting sites before males swarm at sunset. Similarly, glycogen content is higher in the morning males and decreased in the afternoon by 32% at swarming time (Table 1). No differences in lipid and protein contents were observed among the different groups (P>0.05, Table 1). Energy reserves could not be assessed in males caught in copula due to poor quality conservation of these mosquitoes.
Table 1. Mean variation of energy reserves between groups of male mosquitoes collected at different places and times (± standard error).
Indoor morning males
Indoor evening males
P (H ;df)
H and df refer to H-value and the relative degree of freedom, respectively.
J denotes energy unit (Joule); * indicates significance.
Total sugar (J)
4.4 × 10–3±0.0005
1.8 × 10–3±0.0002
<0.0001* (40.47; 2)
1.9 × 10–3±0.0005
1.3 × 10–3±0.0003
0.9 × 10–3±0.0002
0.039* (6.48; 2)
39 × 10–3±0.003
34 × 10–3±0.003
34 × 10–3±0.002
0.121 (4.22; 2)
2.9 × 10–3±0.0004
2.5 × 10–3±0.0003
2.6 × 10–3±0.0002
0.876 (0.26; 2)
Wing length as a proxy of male body size was measured and tested for normality. Figure 2a indicated that the distribution of wing length follows a normal distribution (P>0.05, Kolmogorov-Smirnov test). Mean wing size significantly differed between males caught in copula (2.88±0.016mm) and free-flying males in swarms (2.78±0.011mm), indicating that overall, mated males were bigger than unmated ones (Figure 2b, t=4.84, P<0.0001). A comparison of the coefficients of variation of wing size showed that the magnitude of the variation in wing size was similar between the two groups (T=73.8, P=0.71 Fligner Killeen test). The distribution of female wing length followed a normal distribution (P>0.10, Kolmogorov-Smirnov test) and their mean wing size was significantly greater than that of males (t=6.61, P= 0.0001, data not shown).
Wing symmetry and size assortative mating
A Pearson correlation test was used to test for wing symmetry in mated and unmated males (Figure 3). Wings in both mated and unmated males were strongly correlated (r=0.99, P<0.0001, mated males and r=0.98, P<0.0001, unmated males) indicating that mated males were not more symmetrical than unmated ones (t=0.24, df= 240, P=0.80). Furthermore, a least square regression analysis to look at fluctuating asymmetry between the two groups indicated that the slopes of the two regression lines were not only not different from zero (F=0.048, P=0.826 mated males and F=0.197, P=0.657) but were equal (Figure 4, F=0.193, P=0.66). To test for associations among sizes of females and males in copula, mated female wing size was plotted against the wing size of males caught in copula (Figure 5). A significant but weak correlation (r=0.39, P=0.0003) among wing sizes of the two groups was observed suggesting that size assortative mating may occur to some extent in An. gambiae. However, comparison of the coefficient of variation of wing length attested that the magnitude of size variation was significantly greater in females (T=94.78, P=0.028, Fligner Killeen test). The distribution of wing size of mated males is mostly focused around a central value (Figures 2a, 5), indicative of an optimal intermediate size of males that are chosen by females of any size.
The role of sugar meals in shaping the reproductive outcome of mosquitoes is well known in females but has been overlooked in males. Part of the disinterest is due to males not being involved in disease transmission. However, at the onset of the release of genetically engineered and sterile male mosquitoes, it becomes critical to explore male biology and identify factors that account for their mating success. Our main goals in the present study were to describe the variation in energy reserves in male mosquitoes and determine the role of body size and wing symmetry in male mating success. We found significant differences in total sugar content between swarming and indoor resting males vs newly emerged males. Similarly, significant differences in total sugar and glycogen content were observed between males in the early morning and males from swarms at sunset. No variation in lipid and protein content was observed between the different groups. Our results suggest that total sugars and glycogen are the main energy reserves that fuel male reproductive activities.
The role of sugars in the biology of males has been investigated in a few studies (Müller and Schlein 2006, Gouagna et al. 2010). It is the only food for male adults and for females of some species (Foster 1995, Gu et al. 2011). In addition, sugar provides both sexes with a ready source of flight energy (Nayar and Van Handel 1971) and can, in some cases, improve fecundity (Foster 1995). Males are especially vulnerable to starvation if they are unable to feed on sugar soon after emergence (Foster 1995). They need a source of sugar for survival and reproductive guarantee commitment. Our results showed that both total sugar and glycogen contents were higher in morning indoor males than in evening and males from swarms and reflects the foraging behavior of An. gambiae. Dao et al. (2008), using entry and exit traps to assess indoor mating, showed that both males and females of An. gambiae are active at sunset, leaving their refuges to look for mates, blood, and sugar meals. They mostly return to their refuges in the early morning after replenishing their energy reserves. Similarly Yuval et al. (1994) showed that sugar-feeding in An. freeborni takes place after swarming activities and before males enter resting sites. Accordingly they found a significant variation of sugars and glycogen contents of males An. freeborni collected in the early morning and those collected at sunset in swarms (Yuval et al. 1994). The carbohydrates thus acquired serve to sustain the daily activities and are mostly invested in swarming (Yuval 1992, Yuval et al. 1994). The authors estimated that 50% of the total energy reserves available to An. freeborni males was used to fuel swarming activities. The current study suggests that 59% of the total sugar and 32% of glycogen are burned when An. gambiae are at rest and 15.9% of the total sugars energy is allocated to swarming activities. Considering that swarming activities last for ∼30 min and that males spend ∼660 min in resting sites (from 06:00, time of early morning collection, to 17:00. time of evening collection) that means that males consume proportionally 5.9-fold as much as energy derived from sugars in swarming activities than when they are at rest. Consequently, sugars are extremely important in male biology and it has been shown that males deprived of sugar meals were unable to inseminate females in a recent experimental study leading to population extinction (Stone et al. 2009, Gary et al. 2009). Huho et al. (2007) found that sugar and glycogen contents of field-collected An. gambiae s.l were nil in contrast to laboratory-reared males, probably because field-collected males are in constant demand of these metabolites and may have burned them at the time of collection. The difference in cost of swarming discussed by Yuval et al. (1994) and also here probably reflects differences in the biology of An. freeborni and An. gambiae as well as differences in the collection places, since the energy requirements may vary with respect to ecology and season. That both sugar and glycogen are required to fuel flight is indicative of a regulatory mechanism to mobilize reserves from both sources. However, our results showed that sugar reserves are more necessary to fuel flight activities than is glycogen. This is possibly due to the fact that sugars accumulates in high concentration in the hemolymph and can diffuse rapidly through tissues (Clements 1999), while only changes in the concentrations of hemolymph sugars of insects trigger the activity of glycogen phosphorylase (Clements 1999 and references therein).
The lack of significant changes in protein content among the different collection groups indicates that proteins play a more structural role. Proteins are mainly obtained during the larval stages in breeding sites (Foster 1995). Larvae feed on particulate organic matter that enter the aquatic habitat and are degraded by fungi and bacteria. The nutritive reserves accumulated from this food is essentially made of proteins, lipids, and vitamins and studies on the nutritional requirements of mosquito larvae have shown that larvae of Aedes aegypti failed to develop when the food was lacking proteins. The necessary proteins needed for their development could be made of ten essential amino acids. Similarly, lipid contents did not significantly vary from one group to another. Nayar and Van Handel (1971) established that carbohydrates are used by both male and female Aedes sollicitans to sustain flight and that lipids are not used. However, in exceptional high stress conditions, when males persistently exhibit insufficient levels of sugars, lipids can be transformed into glucose and glycogen to compensate for the lack of sugars (Yuval et al. 1994).
That differences in the sugar and glycogen content, but not lipid and proteins, are found between newly emerged males and the other collections is not surprising because adult mosquitoes emerge with low energy reserves and build it up by sugar-feeding (Yuval et al. 1994). Fluctuating asymmetry is known to be associated with variation in male mating success. Males with perfect wing symmetry are selected by mate-seeking females because this trait might be perceived as an estimate of genetic health. Asymmetrical individuals are less fit and exhibit high mortality, low fecundity, and slow growth rate (McLachlan and Cant 1995 and references therein). Our results showed a very strong correlation between the left and right wings on both mated and unmated males, and additional analysis on fluctuating asymmetry did not find any difference between the two groups of mosquitoes. Our results suggest that both groups are symmetrical, though more samples and more powerful analytical methods are needed to confirm this observation.
Body size is another trait that is associated with male mating success. Our results showed that mated males were significantly bigger than non-mated ones. Similarly, Yuval et al. (1998) found that An. freeborni males that attend swarms were bigger than the non-swarming ones. In contrast, Charlwood et al. (2002) could not find any difference in body size between mated and non-mated An. gambiae, while Ng'habi et al. (2008) showed that intermediate-sized males of the same species were mated more often than the others. Pecharsky et al. (2002) showed that intermediate male size of the mayfly Baetis bicaudatus may be optimal during mating because of their flight agility. Mating systems characterized by male territoriality, lead to males fighting over territory to have the control of incoming females (Diabaté et al. 2011 and references therein). Large body size males are advantageous in such a mating system. Whether or not male mosquitoes fight over territory is not known and that deserves to be investigated. While these discrepancies between the different studies are due to variation in mosquito species mating characteristics and/or to ecological differences, they stress the need to further investigate the role of body size in An. gambiae male mating success (Diabaté et al. 2011).
A weak but significant relationship between male and female body size caught in copula was observed. However, the magnitude of size variation in mated males was smaller than that of females. The size of mated males was focused around a central value suggesting an advantage of intermediate size in this mating system. That intermediate size males mate more often than others can be explained by the fact that they are more agile in flight and can make contact with females more quickly (Crompton et al. 2003, Ng'habi et al. 2008).
A caution should be made with regards to our method of couples collection. An. gambiae mates on wing. Once the pairing couple is formed, one individual flies while its partner remains still. Mate-carrying impairs flight ability. Some couples fly up to the sky when leaving the swarms, and others just drop to the floor straight from the swarms. Whether or not these events are determined by the flight capacity of the individual dragging its partner is not known. Our sampling scheme targeted mostly couples flying up to the sky, hence an exhaustive sampling scheme is needed.
Many lek mating systems are characterized by a high skew in male mating success. The factors that account for female preferences to mate with these few males are not known in most cases. It is not known whether the mating system of An. gambiae also exhibits a high skew in male mating success, however, it has been shown that only 7% of the male population mate in any single day, suggesting that the competition for mating in swarms is very high (Diabaté et al. 2011). However, it should be noted that the same study showed that though mating increased with swarm size, the mating prospect of an individual male was the same irrespective of swarm size. This finding does not necessary imply that all males are equal (Yuval et al. 1998, Ng'habi et al. 2008).
The current study showed that sugar and glycogen are the main energy sources that fuel male mating activities. Mated males were bigger than unmated ones, but whether or not body size accounts for male mating success remains to be tested. Similarly, an optimal intermediate size advantage seems to operate within An. gambiae and deserves to be further investigated. A better knowledge of these different parameters will help understand factors that account for male mating success. Further studies of this kind are needed as they are critical to the success of genetic control strategies.
We are grateful to Seni Ilboudo and Hyacinthe Guel for technical assistance and Dr. Frédéric Simard, Dr. Karine Mouline, Dr. Louis Clement Gouagna, and Cécile Brengues for guidance on energy reserve analysis. Many thanks to the mosquito collectors and the villagers who allowed us to collect mosquitoes in their houses. This investigation was funded by the Multilateral Initiative on Malaria (Grant ID: tdr-mim a80690) and the MRC/DfID African Research Leadership Award (Grant ID 97014).