Temporal genetic variation and dispersal patterns of Aedes albopictus (Diptera: Culicidae) among three different temperature regions of China

Background Aedes albopictus is an indigenous and primary vector for Dengue and Zika viruses in China. Compare with its insecticide resistance, biology, and vector competence; little was known about its genetic variation, corresponding to environmental variations. Thus, the present study aims to discuss how Ae. albopictus population varies among different temperatures regions of China and decipher its potential dispersal patterns. Methods The genetic variation and population structure of all 17 Ae. albopictus populations, collected from three temperature regions of China, were investigated with 11 microsatellite loci and mitochondrial COI gene. Results 11 pairs out of 44 isolated microsatellite markers were chosen for genotyping analysis with the average PIC value of 0.713, which was high polymorphism. The number of alleles was high for each population, with the n e value increased from the Temperate region (3.876) to the Tropical region (4.144). 25 COI Haplotypes were detected, and the highest diversity was observed among the Tropical region. The mean Ho value (ca. 0.557) of all temperature regions, was signicantly lower than the mean He values (ca. 0.684), with nearly all populations signicantly departed from the HWE test and displayed signicant population expansion ( p-value < 0.05).Two genetically isolated groups and three Haplotype clades were evaluated via STRUCTURE and Haplotype phylogenetic analyses, with Tropical populations isolated from other regions, signicantly. Meanwhile, the majority genetic variation of Ae. albopictus populations were detected within populations and individuals at 31.40% and 63.04%, respectively, via AMOVA test, and a relatively signicant positive correlation was merely observed among populations from the temperate region via Isolation by distance (IBD) analysis (R 2 = 0.6614, p = 0.048). Recent dispersions were observed among different Ae. albopictus populations and a total of four major migration trends were rebuilt between the Tropical and the other two regions with the high genetic ows (Nm>0.5). Environmental factors, especially temperature and rainfall, may be the leading cause of genetic diversity differences of different temperature regions.

Ae. aegpti. Meanwhile, its strong ecological plasticity and global trades such as the movement of used tires and "lucky bamboo" also accelerate the global spread of this species from the original South-eastern Asian to nearly every continent except Antarctica [6]. Now, it is among the list of the top 100 worst invasive species, threatening the health of people all around the world [7].
With the development of urbanization in China, the natural environment was altered tremendously, more people have moved from rural areas to the cities, and trades increased between cities and areas, all these processes create suitable habitats for Ae. albopictus and facilitate its breeding and diffusion across the whole country [8,9,10,11]. Meanwhile, the thriving international traveling and trades, providing the essential routines for the continual importing of arboviruses from other countries, especially from Southeast Asian, where Ae. albopictus originated [12], putting all people at risk. During the past ten years, at least 22 ZIKV imported cased were con rmed by the China CDC, and the persistent emergence of Dengue in Southern China, especially in Guangdong and Yunnan province, is positively related to the widespread of Aedes mosquitoes among these regions [13]. In contrast to Ae. aegypti (an invasion mosquito, only distributed in the Southern area of China), Ae. albopictus is an indigenous mosquito of China, ranging from Dalian in the north to Hainan in the south [14,15]. Moreover, it also has been considered the sole voter responsible for numerous recent dengue fever outbreaks in China. Its wider distribution and high competence for numerous arboviruses and nematode parasites emphasize the need to study more extensively the biology, distribution, and dispersion patterns of this species.
Natural environmental variation is responsible for the uctuation of insects' population dynamics, distribution, and biology, including the abundance of population, intensity, and feeding behavior [16]. As a powerful and rapid adaptive insect with a high fecundity rate and short life cycle, the population dynamics and vector competence of Ae. albopictus are also greatly in uenced by environmental conditions [17,18]. Numerous previous studies discussed the in uence of environmental variations on the Ae. albopictus population, mainly based on their in uence on mosquito's abundance, survival, sizefecundity, and competence for certain arbovirus [19,20,21,22,23]. However, less focused on monitoring the changes in genetic diversity and population structure of the Aedes albopictus population, corresponding to the various environmental conditions, which are essential for rebuilding the dispersion patterns of Ae. albopictus among certain regions and providing necessary information for the subsequent mosquito control.
Genetic variation is ubiquitous in the vector population, especially for invasive vectors, such as Ae. aegypti and Ae. albopictus, in helping them to occupy diverse niches and respond quickly to evolutionary challenges [24]. Microsatellites are the preferred markers in studying the genetic variation of vectors for its co-dominant, highly informative, and vast abundance throughout vector genomics [25]. Up to the present, multiplies microsatellite loci associated with Ae. albopictus were successfully isolated and employed in Ae. albopictus population study on a global scale and continue to be a popular choice of genetic marker [22,26,27,28]. Simultaneously, independently from the Ae. albopictus genome, mitochondrial COI gene was also used for the mosquito barcoding and monitoring invasion species, frequently, for its conservation and accuracy of distinguishing sequence variation more su ciently at the inter-species [29,30,31,32].
To evaluate how Ae. albopictus populations vary with the environmental variations among different temperatures regions of China genetically and decipher the potential dispersal patterns of Ae.albopictus among these regions, It wasexamined 17 Ae. albopictus population based on 11 microsatellite loci and COI gene. The results can provide some basic guidelines for the future vector control of Ae. albopictus in China.

Mosquitoes sampling and DNA isolation
In the present study, ca. 600 Ae. albopictus larvae were sampled from seventeen geographically sites across three environmentally distinct regions of China from June to August 2018, and all the sampling sites information as described in Fig. 1 and Additional le 1: Table S1. For each of the samples, larvae were reared and emerged, independently, at 25℃ ± 1℃ at 75% ± 5% relative humidity (RH) under a 14 hlight/10 h-dark (LD) photoperiod and all the Female adult mosquitoes were identi ed under the microscope before DNA isolation. In order to avoid the inbreeding interference, each pooled female mosquito for a given locality was picked up from at least ve wild breeding places within 500 meters. The following DNA isolation work was conducted with Qiagen DNA isolate Kit (No. 69504) under the manufactory's protocol, and all the DNA samples were stored under − 80℃.
Microsatellite isolation, processing, and COI gene ampli cation Referring to CHAMBERS E.W. et al. [33], microsatellite markers were isolated from the whole genome of Ae. albopictus by magnetic-bead enrichment and PCR screening method, and all markers were tested as high polymorphism via Denatured Polyacrylamide Gel Electrophoresis (D-PAGE). A set of 11 microsatellite loci were employed for genotyping 17 Ae. albopictus populations. Detailed microsatellite primers information was list in Table 1. All PCR reactions were performed on a T100 Thermal Cycler (Bio-Rad) under a 50 µl reaction system containing ten ng of DNA, 0.25 U of PrimeStar HS DNA Polymerase (TaKaRa), six µM of dNTPs and ca.5 µM of each primer, with the program set as 35 cycles of 95 °C for 30 sec, 57 °C for 30 sec and 72 °C for 1 min and nal elongation at 72 °C for 10 min. All products were then checked with 2% agarose gel electrophoresis under UV light and run on a 3730XL DNA Genetic Analyzer (Applied Biosystems, California, USA). As Kamgang et al. described [29], COI sequence polymorphism of each locality was investigated among at least 20 individuals. Brie y, DNA ampli cation of a 550-bp fragment of COI was performed on a T100 Thermal Cycler (Bio-Rad) with the following two sets of primers: 5'-GGAGGATTTGG-AAATTGATTAGTTC-3' (F-COI) and 5'-CCCGGTAAAATTAAAATATAAA-CTTC-3'(R-COI) in a 50 ul reaction mix containing 10 ul PCR reaction Buffer (TaKaRa), 4 ul of dNTPs (TaKaRa), 1 ul of Primers (10 pmol/ul), 0.5 ul PrimeStar HS DNA Polymerase (TaKaRa), respectively. The PCR ampli cation program was set as Pre-denaturation at 94℃ for 3 min, followed by 35 cycles of denaturation at 94℃ for 30 sec, annealing at 54℃ for 45 sec, and elongation at 72℃ for 1 min, with the nal elongation at 72℃ for 10 min. All the PCR products were detected and separated by 2% agarose gel electrophoresis. The target fragments for COI were then cut from the gel under the UV light, and puri ed with GenElute™ PCR Clean-Up Kit (NA1020, Sigma-Aldrich).
Each puri ed PCR product was then cloned to pCR™2.

Microsatellite maker isolation and assessment
In the present study, a total of 44 pairs of microsatellite markers were isolated from the whole genome of Ae. albopictus, 11 pairs of which were tested as high polymorphism and chosen for the microsatellite genotyping analysis ( Table 1) Table S5). Linkage disequilibrium (LD) test showed that a total of 302 pairs of loci out of 1870 (16.15%) across all locations were tested signi cantly after Bonferroni correction, while no consistency was found among them (Additional File 5: Figure S1).

Genetic diversity and variation
The observed number of alleles (n a ) of each Ae. albopictus population was very high, and the Mean n a value of each temperature region ranged from 6.909 to 8.091 without signi cant difference. In contrast, the effective number of alleles (n e ) ranged from 3.501 to 4.525, and the n e value increased from the

Population structure and differentiation
In the present study, all Ae. albopictus populations were adequately allocated to two groups with signi cant genetic differences, and the best K value, assessed via Evanno et al.'s ΔK methods, was also equal to two (Fig. 2a). Combined with STRUCTURE bar plots analysis, the Bayesian clustering analysis displayed that Ae. albopictus populations from the Tropical region were genetically isolated with the subtropical and temperature regions (Fig. 2b&c). Moreover, a total of 86.4% of variation was explained by 50 PCs in the DAPC analysis and the results revealed that two genetically isolated groups were obtained, and there was no clear relationship between Ae. albopictus population structure and their distribution temperature areas (Fig. 2d).
AMOVA test revealed that the majority genetic variation of Ae. albopictus populations were detected within populations and individuals at 31.40% and 63.04%, respectively, with the signi cant Fixation indices (F IS =0.33253, F IT =0.36962, and all p = 0.0000; Table 3 Table S4). In contrast to the individual differentiation, a relatively signi cant positive correlation was merely observed among populations from the temperate region via Isolation by distance (IBD) analysis (R 2 = 0.6614, p = 0.048), while no such evidence was observed among the other two regions (Additional le 6: Figure S2).  (Fig. 3b).
A phylogenetic tree, combined with heatmap analysis, was well-established with sequences of all 25 Haplotypes, and it also demonstrated that all 25 haplotypes were divided into three major well-supported clades. As expected, Clade was separated from Clade and Clade with a 100% bootstrap support rate, including Haplotype H9, H10, H11, H12, H19, and H20, which was only observed at the tropical region. In comparison, Clade was isolated from Clade with a relatively lower bootstrap support rate (70.74%), including ve Haplotypes (H2-H3, and H5-H7, respectively) that distributed merely at South subtropical region. Nevertheless, Clade is an admixture group containing all the rest 14 Haplotypes, whereas only two of them were observed in the tropical region and four at the temperature region (Fig. 3a).

Migration and the correlation analyses between genetic indices and environmental factors
Migration patterns were rebuilt using divMigrate networks among all 17 Ae. albopictus populations. As expected, Ae. albopictus was observed migrating back and forth between the Tropical and Temperate areas frequently (Fig. 1). A total of four major migration trends were observed among different temperature regions with the high genetic ows (Nm > 0.5). Two routines were observed originated from the Tropical and subtropical areas and destined to Hebei and Beijing of the Temperate area, while another two destined to Guangdong and Guangxi of the Southern subtropical area. Compared with the south to north routines, the latter migration routines were substantially higher. Meanwhile, nearly all Tajima's D   and Fu's Fs values were tested negatively with no statistically signi cant p values except population  JKCH (Additional le 3: Table S3). Based on the COI sequences, mismatch analysis results showed that Harpending's Raggedenes indexes for all three Haplotype clades were relatively low (ranged from 0.1054 to 0.1812, p > 0.05) and unimodal mismatch distributions were observed among different Ae. albopictus populations (Additional le 7: Figure S3).
As illustrated in Fig. 4, all ve genetic indices and two environmental factors (i.e., Temperature and Rainfall) contributed equivalently to the rst axis of the PCA up to 95.2% except the environmental factor Latitude, which contributes more to the second axis with a proportion of 17.5%. Combined with Hierarchical clustering, performed via multiple factor analysis (MFA), all 17 Ae. albopictus populations were then clustered into three groups, with cluster II and cluster III closely related to each other. When environmental factors were regarded as the major in uencing factors, molecular diversity indices of the Tropical populations were signi cantly higher than that of other regions. It also demonstrated that environmental factors, especially temperature and rainfall, were the leading cause of genetic diversity differences of different temperature regions.

Discussion
For their high mutation rate, co-dominant expression, and universal distribution throughout the eukaryotic genome, microsatellite loci have been widely used to evaluate the genetic variation and population structure of vectors, especially for those without fully annotated genome [26,33]. Some studies mentioned that the vast existence of the null allele might have effects on classical estimates of population differentiation, especially for the assessment of heterozygosity de cient [34,35,36]. In the present study, even though nearly all loci were tested high informative (PIC > 0.5), the null allele was still observed at all loci ranged from 0.02 to 0.158, with an average of 0.078. In order to improve the accuracy of the ndings, the mitochondrial COI gene was also employed to investigate the genetic variation of all individuals with microsatellite loci together.
Consistent with Zhong et al. [37], Ae. albopictus population of the Tropical region showed higher diversity than the other two regions. Continuous dispersion and a better survival environment may be the best explanation for this phenomenon. Geographically, Southern china directly borders with many countries of Southeast Asian where Ae. albopictus originates, including Laos, Philippines, and Myanmar et al. [12].Frequent border trades and personnel exchanges among these areas result in the continuous dispersion of Ae. albopictus, in turn, enrich the diversity of the local population, which is also con rmed by the results of bottleneck analysis that nearly all populations from Temperate and Subtropical regions experienced bottleneck effect except the populations of the Tropical region. Moreover, the hot and humid climate of Southern China is more suitable for the breeding and development of Ae. albopictus [38,39,40]. Stimulously, all the mean Ho values of the Temperate and Subtropical regions were observed signi cantly lower than the mean He values, and nearly all populations signi cantly depart from the HWE test. It may be closely related to the application of mosquito control programs in these areas, and the continuous dispersion of Ae. albopictus from Southeast Asian countries to the Tropical region helps the recovery of the local mosquito population.
The population differentiation analyses showed that nearly all pairwise F ST among populations were signi cant but not high, which indicates the potential communications between populations across different temperature regions, especially among populations ZGND, NJTH, NJDX, and QDDX. Correspondingly, two genetically separated groups were observed among these populations, and molecular variations within populations and individuals contribute to the differentiation between populations. As a global invasion species, Ae. albopictus has developed several strategies to cope with a broader range of temperatures and adapt to local thermal conditions [17,40], which helps this mosquito disposal and colonize at different locations successfully. Migration analysis results displayed four major migration trends that were observed among different temperature regions with the high genetic ows (Nm > 0.5). All these results revealed that human-aided dispersion might be the main reason for the similarity of populations among different regions. This hypothesis is also veri ed via IBD analysis that long-distance is not signi cantly associated with the genetic variation of individuals.
Among all the 25 haplotypes detected in the present study, nearly 44% (11 out of 25) of these haplotypes were only observed at the Tropical and Subtropical regions, the p-value of the mantle test for the Tropical region was signi cant. It means continuous dispersal from the neighborhood Southeast Asia countries maybe the reason for the higher diversity for the Ae. albopictus populations of the Tropical and Subtropical regions and the origination for the Ae. albopictus populations of these two regions may be different. Furthermore, the mismatch analysis results also revealed that recent population dispersal among different temperature regions, which is also the main reason for the universal distribution of the rest haplotypes. As a powerful and rapid adaptive insect, the biology of Ae. albopictus is greatly in uenced by environmental conditions [17,18]. In the present study, The PCA analysis among genetic indices and environmental factors revealed that temperature and rainfall were the leading cause of genetic diversity differences of Ae. albopictus populations among different temperature regions.
Overall, energetic dispersal patterns are not only observed only in the southern part of China, but frequent diffusions also exist from the Tropical and Subtropical regions to the Temperate region. Since Dengue fever continues to break out in southern China every year, the potential diffusion routines between the south and north regions may put the rest areas of China in danger.

Conclusion
The present study systematically evaluated the genetic variation, population structure, and haplotype phylogenetic relationships of Ae. albopictus populations across different temperature regions of China. In accordance with the locality or region, all 17 Ae. albopictus populations were genetically structured, and three major Haplotype clusters were observed via COI phylogenetic analysis, which suggests different evolutionary histories under changing environments, especially temperature and rainfall. Meanwhile, four major migration trends were observed among different temperature regions with high genetic ows, which contributes to the similarity of all Ae. albopictus populations. Overall, the results of the present study suggest that more effective control strategies should be employed to prevent the multiple dispersion patterns of Ae. albopictus in China.  Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.