Interspecific mating bias may drive Aedes albopictus displacement of Aedes aegypti during its range expansion

Abstract Aedes albopictus is the most invasive mosquito in the world and often displaces Ae. aegypti in regions where their populations overlap. Interspecific mating has been proposed as a possible cause for this displacement, but whether this applies across the range of their sympatry remains unclear. Aedes albopictus and Ae. aegypti collected from allopatric and sympatric areas in China were allowed to interact in cage experiments with different crosses and sex-choices. The results confirm that asymmetric interspecific mating occurs in these populations with matings between allopatric Ae. albopictus males and Ae. aegypti females being significantly higher (55.2%) than those between Ae. aegypti males and Ae. albopictus females (27.0%), and sympatric mosquitoes showed a similar but lower frequency bias, 25.7% versus 6.2%, respectively. The cross-mated females can mate second time (remate) with the respective conspecific males and the 66.7% remating success of female Ae. albopictus was significantly higher than the 9.3% of Ae. aegypti females. Furthermore, 17.8% of the matings of Ae. albopictus males exposed to mixed pools of Ae. albopictus and Ae. aegypti females and 9.3% of the matings of Ae. aegypti males with mixed Ae. aegypti and Ae. albopictus females were interspecific. The difference in the length of clasper between male Ae. albopictus (0.524 mm) and Ae. aegypti (0.409 mm) may be correlated with corresponding mates. We conclude that stronger Ae. albopictus male interspecific mating and more avid female intraspecific remating result in a satyr effect and contribute to competitive displacement of Ae. aegypti as allopatric Ae. albopictus invade during range expansion.


Introduction
Aedes aegypti is the primary vector of dengue, Zika, and yellow fever viruses and is distributed mainly in the global tropical regions (1)(2)(3). Aedes albopictus is the most important vector of chikungunya virus and also can transmit dengue viruses. Furthermore, it is the most invasive mosquito species worldwide and is distributed widely in tropical, subtropical, and temperate regions. It is now found in all continents except Antarctica (4)(5)(6). Aedes mosquitoes and their transmitted viral pathogens represent major threats to public health (7).
Aedes aegypti and Ae. albopictus share similar life cycles and ecological niches and often are found distributed sympatrically (8,9). Aedes aegypti was found in the state of Florida in the United States, since 1880 (10). When Ae. albopictus invaded the state in the 1980s (11,12), it gradually replaced Ae. aegypti as the local dominant species (13,14), even causing total displacement in some regions (15)(16)(17). Similarly, Ae. aegypti used to be the predominant species during the 1980s in Hainan, China (18) and caused several dengue fever pandemics on the island and in the neighboring Leizhou peninsula, a joint area among Guangdong, Hainan, and Guangxi Provinces (19). At the time, Ae. albopictus was found rarely on the island. Several recent studies report that Ae. aegypti is now rare in the region and that Ae. albopictus is widespread (20). Both Ae. aegypti and Ae. albopictus have been on the Leizhou peninsula since the 1980s (21). Since then, the population and density of Ae. aegypti has been decreasing in this area and now only small numbers are found in the limited regions of Wushi town and Qishui town of Zhanjiang city, and there is a trend of Ae. albopictus completely replacing Ae. aegypti in the near future (22). Similar trends of species replacement also have been reported in the southeastern United States (23,24) and Bermuda (25), where Ae. albopictus displaced Ae. aegypti with comparable rapidity. It is unclear how Ae. albopictus replaces Ae. aegypti although competition for breeding sites, and asymmetries in interspecific mating and the effects of male accessory gland proteins have been proposed (24,(26)(27)(28).
Satyrization is a form of mating interference in which males of one species mate with females of another species, produce no hybrid progeny, and significantly decrease the reproductive fitness of the species from which the females originate (27,29,30). Satyrization has been proposed as the probable cause of competitive displacements of resident mosquitoes by invasive species, especially of Ae. aegypti by Ae. albopictus (24). Cage experiments and field observations indicated that Ae. albopictus males are capable of satyrizing females of other species of the Stegomyia subgenus, potentially leading to competitive displacements, and possible extinctions, especially of endemic species on islands (27). However, the general dynamics of interspecific matings between Ae. albopictus and Ae. aegypti are not known.
Collections of Aedes albopictus and Ae. aegypti from allopatric and sympatric regions in China were used to investigate the dynamics of satyrization ( Figure S1 and Table S1, Supplementary Material). Interspecific and intraspecific matings and rematings between Ae. albopictus and Ae. aegypti show a strong bias for favoring Ae. albopictus in regions where they might overlap with Ae. aegypti. These data confirm and extend what was seen with populations in Florida, USA (24,31).

Asymmetric interspecific matings occur at different frequencies in laboratory-and field-derived Ae. albopictus and Ae. aegypti in China
A total of two interspecific mating groups for both laboratory-and field-collected mosquitoes were examined. Intermating I consists of male Ae. albopictus crossed with female Ae. aegypti and Intermating II has male Ae. aegypti crossed with female Ae. albopictus. In total, two control mating groups also were done: Control I had male Ae. albopictus crossed with female Ae. albopictus, and Control II had male Ae. aegypti crossed with female Ae. aegypti. Interspecific mating was observed in both groups ( Fig. 1; Table  S2, Supplementary Material). The interspecific mating rate for the laboratory mosquitoes was significantly higher in group I (55.2 ± 2.2%) compared to group II (27.0 ± 3.3%; t = 7.15, d.f. = 4, and P = 0.0020). This asymmetry also was observed and significant in the sympatric field-collected mosquitoes, 25.7 ± 1.0% and 6.2 ± 1.3% for Intermatings I and II, respectively, (t = 12.07, d.f. = 4, and P = 0.0003) but was less frequent. These data support the conclusion that interspecific matings between Ae. albopictus and Ae. aegypti are asymmetric, with Ae. albopictus males having a much higher rate of successful intermating than Ae. aegypti males. Furthermore, the higher interspecific mating rates between allopatric laboratory populations and sympatric Guangdong field-derived populations (Student t tests, all P < 0.01) support the hypotheses that prolonged sympatry may lead to selection for more robust premating barriers between the two species, especially for Ae. aegypti (32).

Interspecific-mated female mosquitoes do not produce viable offspring
The overall oviposition rates were high for females in all mating groups (χ 2 = 6.51, d.f. = 5, and P = 0.0852), regardless of being unmated or mated interspecifically or intraspecifically (Table S3, Supplementary Material). However, the number of eggs per female varied among mating groups, and the egg-hatching rates among all cross-species-mated were zero, and consequently, no progeny were produced. The control intraspecific egg-hatching rates were 74.1% to 79.4% and these combined data support and confirm previous reports that interspecies matings in either direction do not result in viable progeny ( Fig. 1) (24).

Remating between interspecific mated females and conspecific males
Remating experiments between previously interspecific mated females and conspecific males were carried out only for the laboratory-reared mosquito colonies. As expected, remating rates were low, 9.3 ± 1.0%, for Ae. aegypti for which it is known that the first mating inhibits subsequent mating (33, 34) ( Fig. 1; Table S4, Supplementary Material). Aedes albopictus rematings were significantly higher,66.7 ± 4.2%, (t = 11.24, d.f. = 4, and P = 0.0004) and the odds ratio of remating was 25.6 (95% CI: 7.3 to 90.1) for Ae. albopictus over Ae. aegypti. Similarly, the hatching rate of eggs produced by those previously cross-species mated females was significantly highly for Ae. albopictus (35.8 ± 4.2%) compared to Ae. aegypti (4.9 ± 1.8%; t = 6.78, d.f. = 4, and P = 0.0025). These data support the conclusion that cross-mated female Ae. albopictus and Ae. aegypti can remate with their conspecific males with the former being more likely to contribute to the subsequent generation. However, a basal difference in remating rates between the species could explain these observed differences (35).

Male Ae. albopictus and Ae. aegypti mate differentially with conspecific and interspecific females in a competition assay
The results of a series of two mating experiments ("male-choice"), Male-alb, comprising Ae. albopictus males crossed with mixed Ae. albopictus and Ae. aegypti females, and Male-aeg, Ae. aegypti males crossed with Ae. aegypti and Ae. albopictus females, revealed interspecific mating rates of 17.8 ± 0.1% and 9.3 ± 1.3% (t = 5.25, d.f. = 3, and P = 0.0135), respectively, for laboratory populations and 11.8 ± 0.6% and 2.0 ± 0.4% (t = 13.77, d.f. = 4, and P = 0.0002), respectively, for Guangdong field-derived populations ( Fig. 2; Table  S5, Supplementary Material). Control intraspecific mating rates were 100%. Male Ae. albopictus had 2-to 5-fold higher proportion of interspecific matings compared to male Ae. aegypti, and the allopatric populations (laboratory colonies) were higher than the sympatric populations (Guangdong field-derived). These findings support the conclusion that male mosquito choice is a significant factor in interspecific matings and is consistent with prolonged sympatry selecting for more robust premating barriers (31, 32). Interspecific mating and hatching rates between Ae. albopictus and Ae. aegypti. Lab: laboratory strain; Fgd: field Guangdong strain. (E) Eggs per female mosquito were detected for male specific Nix gene by PCR to confirm whether the mosquito had been interspecific mated or not. (F) Remate I: intermated female Ae. aegypti × male Ae. aegypti, Remate II: intermated female Ae. albopictus × male Ae. albopictus. " " represents females that had previously mated interspecifically. In total, one interspecific mated female was exposed to one conspecific male in a cup for 5 days, then laid eggs and hatching situation from every female mosquito were recorded. (G) Remating and hatching rate between intermated females and conspecific males. The black mosquito icons represent Ae. albopictus and brown mosquito icon represents Ae. aegypti. Gray columns represent intermating or remating rate; black columns represent hatching rate. Bars represent the standard error of mean. Statistics were performed using Student t test. * * * P < 0.001 and * * P < 0.01. These assays were repeated three times.

Ae. aegypti
Similar experiments with females ("female choice") Female-alb, in which Ae. albopictus females were crossed with Ae. albopictus and Ae. aegypti males, and Female-aeg, crosses of Ae. aegypti females with Ae. aegypti and Ae. albopictus males, showed similar intraspecific mating rates, ranging from 83.2 ± 5.2% to 93.9 ± 3.9% ( Fig. 2; Table S6, Supplementary Material). Furthermore, the interspecific mating rates also were similar, ranging from 3.5 ± 2.5% to 8.5 ± 4.3% (Student t tests, all P > 0.05). These results support the conclusion that female choice of either Ae. albopictus or Ae. aegypti does not play a major role as a premating barrier for subsequent interspecific mating. Those males and females were exposed in a cage for 7 days, then female spermathecae were dissected to determine their mating status. (B) Female-choice I: 100 female Ae. albopictus × (100 male Ae. albopictus + 100 male Ae. aegypti), Female-choice II: 100 female Ae. aegypti × (100 male Ae. aegypti + 100 male Ae. albopictus). Those females and males were exposed in a cage for 7 days, then female mosquitoes were used to detect male specific Nix gene by Nest PCR to identify whether the female had been intraspecific or interspecific mated. (C) Choice mating in field mosquitoes. Both Ae. albopictus and Ae. aegypti mosquitoes collected in the field were used to detect male specific Nix gene by Nest PCR to identify whether the female had been intraspecific or interspecific mated. (D) Male-choice mating rates between Ae. albopictus and Ae. aegypti. (E) Female-choice mating rates between Ae. albopictus and Ae. aegypti. (F) Intraspecific and interspecific mating rates of Ae. albopictus and Ae. aegypti collected from the fields. The black mosquito icons represent Ae. albopictus and brown mosquito icon represents Ae. aegypti. Lab: laboratory stain (green columns). Fgd: field Guangdong strain (Zhanjiang; red columns). Fyn: field Yunnan strain (Jinghong; blue columns). (D)-(E) Bars represent standard error of mean, and statistics were performed using Student t test. (F) Bars represent 95% CI, and statistics were performed using χ 2 -test or Fisher exact test. * P < 0.05, * * P < 0.01, * * * P < 0.001, and ns = not significant. Male-and female-choice mating assays were repeated three times. The field mosquitoes were collected five times in Zhanjiang, Guangdong and Jinghong, Yunnan, respectively.

Evidence of interspecific mating between Ae. albopictus and Ae. aegypti in field-derived mosquitoes
Examination of field-collected adult females showed intraspecific mating rates ranging from 46.9 ± 3.6% to 57.5 ± 7.2% for Ae. albopictus and Ae. aegypti, respectively (χ 2 tests, all P > 0.05; Fig. 2; Table S7, Supplementary Material). In contrast, the interspecific mating rates between male Ae. albopictus and female Ae. aegypti or male Ae. aegypti and female Ae. albopictus were 4.3 ± 2.9% and 1.6 ± 0.9%, respectively, in Zhanjiang samples, and 9.1 ± 3.1% and 5.5 ± 1.6%, respectively, in Jinghong samples. Only the matings biases of Ae. aegypti males with Ae. albopictus females between Zhanjiang and Jinghong were significant (P < 0.05). Interestingly, Ae. albopictus and Ae. aegypti have been sympatric in Zhanjiang for more than 40 years (21) but less than 10 years in Jinghong (36), and these data are consistent with the hypothesis that prolonged sympatry selects for premating barriers between these two species (32,37).

Aedes albopictus males have longer claspers than Ae. aegypti males
Morphological characteristics may be among the factors contributing to the observed differentials in interspecific matings between Ae. albopictus and Ae. aegypti. While female mosquitoes are generally larger than their conspecific males, the sizes and coloration of both male and female Ae. albopictus and Ae. aegypti are similar ( Fig. 3; Table S8, Supplementary Material). The average weight, body length, wing length, and leg length in samples of 50 each male and female mosquitoes also are similar. In contrast, the average length of the clasper in male Ae. albopictus is 0.524 ± 0.003 mm and 0.518 ± 0.003 mm in laboratory and Zhanjiang field-derived mosquitoes, respectively, and these are significantly . Lab: laboratory strain. Fgd: field Guangdong strain. alb♂:Ae. albopictus male (light gray); aeg♂: Ae. aegypti male (light brown); alb♀:Ae. albopictus female (dark gray); aeg♀: Ae. aegypti female (dark brown). Bars represent standard error of mean. Statistics were performed using Student t test. * P < 0.05, * * P < 0.01, * * * * P < 0.0001, and ns = not significant. A total of 10 mosquitoes were assigned to each group. The measurements were repeated five times.

Aedes albopictus copulation bouts are of longer duration than those of Ae. aegypti
Video imaging of interspecific mating bouts of Intermate I: Ae. albopictus males with Ae. aegypti females and Intermate II: Ae. aegypti males with Ae. albopictus female revealed that the grasping episodes (n = 261; males attempting to engage females) of Ae. albopictus males with Ae. aegypti females were significantly lower than Ae. aegypti males against Ae. albopictus females (n = 2650), indicating that Ae. aegypti males were more active in pursuit of a heterospecific females than male Ae. albopictus ( Fig. 4; Table  S9, Supplementary Material). The copulating attempts, successful copula, and copulate-fail rates were 101, 49, and 51.49%, respectively, in Intermate I, and were 1,365, 687, and 49.67%, respectively, in Intermate II. These data support the conclusion that the acceptances of both female mosquito species to the interspecific males are similar and the females do not play the dominant role in mate choice. Interestingly, the copulating duration and insemination rate in Intermate I was 14.78 ± 2.12 s and 17.54 ± 0.47%, which are significantly longer and higher than 6.54 ± 0.19 s and 7.66 ± 0.34% in Intermate II (Student t tests, all P < 0.01). The copulating duration in control I (Ae. albopictus males with Ae. albopictus females) was 20.60 ± 0.95 s, which also significant longer than 7.55 ± 0.32 s (Student t tests, P < 0.0001) in control II (Ae. aegypti males with Ae. aegypti females; Figure S3 and Table S9, Supplementary Material). These results are consistent with interspecific mating rates and coincident with male clasper lengths between Ae. albopictus and Ae. aegypti.

Discussion
The results of these experiments demonstrate that asymmetric interspecific matings can occur between Ae. albopictus and Ae. aegypti mosquito samples collected in China, and these findings are consistent with studies in southeastern United States (24,38,39) and Bermuda (25). The interspecific matings between Ae. albopictus males and Ae. aegypti females were significantly more frequent than the reciprocal matings of Ae. aegypti males with Ae. albopictus females. Interspecific matings also were more frequent in allopatric strains than in sympatric strains. Furthermore, females that had mated to interspecific males produced no progeny. This combination of results provide a basis for the observed gradual displacement of Ae. aegypti by Ae. albopictus in Zhanjiang. From these results, we conclude that Ae. albopictus suppresses the reproduction of Ae. aegypti by interspecific mating and this is an example of satyrization in species competition. These observations extend the conclusions of the previous findings to include largescale arenas of invasion and displacement in China.
The insemination rate is typically used to assess male performance in mosquitoes (40)(41)(42), but the role of female choice and the degree to which it influences mating outcomes is not known (43,44). The interspecific mating frequencies in male choice groups observed in this study were significantly different and Ae. albopictus males mated more frequently with Ae. aegypti females, even in the presence of conspecific females, than Ae. aegypti males did with Ae. albopictus females under similar choice conditions. In contrast, no significant differences were observed in the female choice experiments. These findings support the conclusion that interspecific matings between Ae. albopictus and Ae. aegypti is dominated by male behavior, and male Ae. albopictus engage in interspecific mating more frequently than male Ae. aegypti. While the evidence for female choice in mating bias is not evident here, it may arise as a premating barrier as the species experience longer periods of sympatry (32,37). This asymmetric mating coupled with the infertility of interspecific-mated females would have a significant impact on species displacement.
Male mating behavior and morphology may provide a partial explanation for the asymmetrical interspecific mating bias. One correlated factor is that the average length of the clasper in Ae. albopictus males is significantly longer than that of Ae. aegypti males. The claspers are used to hold the female tightly during copulation and prevent easy disengagement during sperm transfer, and therefore, result in a more successful mating. Longer copulation times and increased insemination rates of Ae. albopictus male compared with Ae. aegypti were observed and detected in video experiments. We do not propose that these morphological differences alone account for the male-driven asymmetry in mating success, but they may have an additive effect in combination with other factors including sex pheromones (45) and wing-beat frequency (46,47). Further studies are needed to clarify these and other potential mechanism for the asymmetric interspecific matings.
Female mosquitoes usually mate only once in their reproductive lifetime (48)(49)(50) because the sperm from the first mating can be stored in spermatheca and used throughout subsequent gonotropic cycles (51). In addition, substances secreted by the accessory glands and passed to the female along with the sperm alter female mating behavior to prevent remating (33,52). However, we showed here that Ae. albopictus or Ae. aegypti females previously experiencing an interspecific mating could remate with a conspecific male and produce viable offspring. It is note-worthy that the remating and hatching rates of Ae. albopictus females were significantly higher than those of female Ae. aegypti. This is likely due to the differential effectiveness of male accessory gland protein suppression of remating in the interspecific crosses (24). The accessory protein HP-1 from the semen of Ae. albopictus could impose enforced monogamous paternity on Ae. aegypti females and inhibit secondary matings, but in contrast, secretions of Ae. aegypti males did not have the effect on female Ae. albopictus (53). Other factors correlated with the female activity in remating possibly exist in the semen of the male (33,54) and need to be clarified with more study.
Aedes albopictus and Ae. aegypti share similar life cycle characteristics and ecological habits (8,55,56), as well as the similar size, body weight, and wing length. As the result, both mosquitoes also may have similar swarm behavior, wing beat frequencies, and other premating factors that make possible occasional interspecific mating. Because the reproductive cost of interspecific mating is high, no viable progeny, resistance to interspecific mating would be expected to be selected against during divergence of the ancestral mosquitoes that give raise to these two species (32,37). Consistent with this expectation, interspecific matings among field-derived Zhanjiang samples where Ae. albopictus and Ae. aegypti have been sympatric for at least 40 years were significantly lower than those samples where both mosquitoes have only been sympatric for less than 10 years. Similar observations have been reported for these two species in sympatric locales in Florida, USA (32,37). It is possible that more prolonged close contact of both mosquitoes leads to stronger premating barriers, so that the errant interspecific matings could be avoided and decreased. This observed asymmetry also may account for the circumstances where Ae. albopictus as a species invading regions where Ae. aegypti is already extant leads to the suppression and elimination of the latter.

Conclusions
We can conclude from these studies that the asymmetric interspecific matings between Ae. albopictus and Ae. aegypti characterized by Ae. albopictus males exhibiting more frequent interspecific mating than Ae. aegypti males, and female Ae. albopictus having a greater frequency of conspecific remating is a general phenomenon in areas where these mosquitoes are sympatric. This results in a species competition known as satyrization in which Ae. albopictus can displace Ae. aegypti. This asymmetry is correlated with the length of clasper and female monogamy. These findings highlight some of the potential factors and mechanism of interspecific mating and species competition between Ae. albopictus and Ae. aegypti, two important vector and invasive species. Furthermore, the observed biases may complement and enhance the efficacy of sterile insect technologies (SIT) for impacting pathogen transmission dynamics in regions where the two species are sympatric and able to transmit the same pathogens.

Mosquito strains and rearing
Laboratory Ae. albopictus and Ae. aegypti strains have been colonized in our laboratory for many years. Field-derived colonies of Zhanjiang Ae. albopictus and Ae. aegypti were collected from artificial containers (discarded tires and buckets) in several places in Wushi Town, Zhanjiang City, Guangdong Province in 2019 to 2021, and these colonies used in the mating experiments were second to fourth generation (F 2 to F 4 ). All mosquitoes were reared at 27 ± 1 • C, 70 ± 10% humidity and under 14 h light/10 h dark cycles. To obtain experimental mosquitoes, the larvae (200 larvae/l water) were reared in stainless steel trays containing dechlorinated water and were provided daily with yeast and turtle food. Pupae of both species were collected individually and secured in 2 ml Eppendorf tubes with 1 ml water. When adults emerged, the species and sex of adults were determined by examination of the scutum and antennae, respectively. Adults of each sex and species were placed separately in paper bowls (9.5 × 6.7 × 6.2 cm 3 ) with a mesh cover and offered a 10% sucrose solution on a cotton wick.

Noncompetitive (no-choice) interspecific mating experiments
Experiments were conducted in microcosms (20 × 20 × 30 cm 3 ) using virgin males and females of each species, originating from the laboratory colonies and the field-derived colonies of Guangdong Province. All mosquitoes were between 2-and 3-d-old when used for mating experiments. A total of 120 males of one species was crossed with 120 unmated females of the other species in each enclosure (microcosm) using laboratory (lab) or field-derived mosquitoes. Mosquitoes were left to cohabit for 7 days. Conspecific microcosms containing males and females of either Ae. albopictus and Ae. aegypti were set up as normal mating controls. In addition, unmated females of each species were held alone in microcosms as egg-laying controls. All females were anesthetized with CO 2 and spermathecae dissected to determine insemination status. The presence of sperm in spermathecae was recorded as an insemination event ( Figure S2, Supplementary Material). Triplicate replicates were carried out for every cross combination.

Gene amplification (PCR) to detect the male-specific gene, Nix, in mated females
Nix is a male-specific gene present in both Ae. albopictus (57) and Ae. aegypti (58) and can be used to detect mated females. If female mosquitoes mated, the sperm of male mosquitoes would be transferred to the spermathecae of female mosquitoes. PCR assays were used to detect Nix in eggs to determine whether females had mated. Moreover, we established a sensitive and specific nested PCR assay to detect Nix in female mosquitoes to identify whether the female had been intraspecifically or interspecifically mated. The sequences of primers and program used in the study are given in Table S10 (Supplementary Material). AlbNix and AegNix were used to detect Nix in eggs as well as the first fragment in female mosquitoes, and Nest-AlbNix and Nest-AegNix were used to amplify the second fragment which is inside of the first fragment. For the first run, female genomic DNA was used for template. For the second run, the first run of PCR product was diluted 100 times as a template. Positive and negative controls were included in each first run experiment. Amplification products were observed under UV light after electrophoresis in 1% agarose gel containing and confirmed by DNA sequencing.

Remating experiment
The remating assays recapitulated the noncompetitive (nochoice) experiment described above. Females from the intercross experiments were offered a blood meal after a week of expose to the interspecific males. Restrained mice were put in the cages, in compliance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the NIH. Individual engorged females were transferred to a 250-ml paper cup with filter paper. Females were given 3 days to lay eggs. The pools of eggs were counted and the DNA extracted used as a template for gene amplification to detect the presence of Nix gene. Only mosquitoes positive for Nix-eggs were used for remating to ensure female had indeed mated with an interspecific male. These females then were transferred to a paper cup and mated for 5 days with conspecific males at a ratio of 1:1. Blood fed again. Every blood-fed female was transferred to a new paper cup with filter paper. Eggs collected from each female were counted and hatched to determine if a female deposited viable or nonviable eggs. Production of viable eggs was used as a proxy for successful intraspecific insemination. In the controls, 15-d-old virgin females were exposed to 3-d-old conspecific males for 5 days and subsequently allowed to blood feed. Each blood-fed female was allowed to oviposit. There were 10 to 20 females per replicate and three repetitions were performed.

Male-choice and female-choice mating experiment
A total of 100 males were aspirated into microcosms (20 × 20 × 30 cm 3 ) containing 100 conspecific and 100 heterospecific females in the male-choice experiment. They were left to cohabit for 7 days. After that, the females were removed and morphologically identified as Ae. albopictus and Ae. aegypti. The spermatheca of female mosquitoes then were dissected and examined the sperm microscopically. A total of 100 females were aspirated into microcosms containing 100 conspecific and heterospecific males for the female-choice experiments. After 7 days, females were removed. Each female mosquito was transferred into a 1.5-ml Eppendorf tube containing 50 μl of lysis buffer from MiniBEST Universal Genomic DNA Extraction Kit (Takara-Bio, Shiga, Japan). The mosquito was digested by proteinase K and RNase A overnight at 56 • C according to the manufacturer's instructions. Following DNA extraction, samples were stored at −20 • C until PCR analysis.

Detection of cross-mating in natural populations
Adult mosquitoes were collected from randomly selected locales in two regions in Southern and Southwestern China, where Ae. albopictus and Ae. aegypti coexist to determine potential interspecific matings in natural populations. Human landing catches were carried out using power aspirators at two different cities in China. Wushi town (109 • 86"E, 20 • 56" N), located in the southwest of Leizhou Peninsula, is one of the main fishing ports in Guangdong Province. The other city, Jinghong (101 • 31"E,22 • 36"N), is located in the southern part of Yunnan Province and adjacent to Myanmar. Both cities have larval habitats for Ae. albopictus and Ae. aegypti. Wild-caught females were identified morphologically as Ae. albopictus and Ae. aegypti and were stored in ethanol. They were transferred to a plastic plate and washed three times in deionized water to remove ethanol. In order to further determine which male mosquitoes the females mated with, Nix (Ae. albopictus, AlbNix and Ae. aegypti, and AegNix) diagnostic fragments were amplified from their genomic DNA using the previously described nested PCR system.

Measurement of adult mosquito
The weight, body length, wing length, and leg length of adult male and female Ae. albopictus and Ae. aegypti (laboratory strains), as well as the length of the male mosquito clasper (laboratory and field-derived Guangdong strains) were measured. Mosquitoes were placed in an oven to dry for 1 h, and their weight was measured in groups of 10 mosquitoes (repeated five times). Forceps and dissecting scissors were used to separate the legs, wings, and body of mosquitoes and the male mosquito claspers. These were examined microscopically and images recorded using a computer and camera. Image-pro Plus software was used to measure the length of the organs. Measurements of 50 mosquitoes of each species and sex were taken.

Video observation of interspecific mating interactions
Newly emerged 2-to 3-d-old mosquitoes of laboratory strains were used for mating behavioral video recording. A total of 50 males and 50 females were transferred into a custom video cage (7 × 11 × 12 cm 3 ) and a digital camera (Logitech Capture, 1080P, 60fps) was used to record the mating process between Ae. albopictus and Ae. aegypti. The videos were recorded at intervals from ZT0-3 (3 h after light on) and ZT11-14 (3 h before light off) for five consecutive days. The number of grasp (male grasps female), copulation attempt (male rolls its abdomen to try copulate), copulation (male copulates with female successfully), copulation fail rate, and mean of copulate duration and insemination rate were observed and counted.

Statistical analysis
All statistical analyses were performed using SPSS version 20.0 (IBM, Chicago, IL). Mating rates, remating rates, hatching rates, mosquito eggs, body size, and male clasper length were compared using the Student t test (significant level of α = 0.05). χ 2test or Fisher exact test (if any n < 5) (significant level of α = 0.05) was used to compare the mating rate in wild-caught female mosquitoes. The odds ratio of remating success (after interspecific mating) was calculated for Ae. albopictus against Ae. aegypti.
Odds ratio : Percentage deviation is calculated as Observed − expercted expected × 100.