Population genetic structure of the Asian bush mosquito, Aedes japonicus (Diptera, Culicidae), in Belgium suggests multiple introductions

Abstract

Background: Aedes japonicus japonicus has expanded beyond its native range and has established in multiple European countries, including Belgium. In addition to the population located at Natoye, Belgium, locally established since 2002, specimens were recently collected along the Belgian border. The first objective of this study was therefore to investigate the origin of these new introductions, which were assumed to be related to the expansion of the nearby population in western Germany. Also, an intensive elimination campaign was undertaken at Natoye between 2012 and 2015, after which the species was declared to be eradicated. This species was re-detected in 2017, and thus the second objective was to investigate if these specimens resulted from a new introduction event and/or from a few undetected specimens that escaped the elimination campaign.

Methods: Population genetic variation at nad4 and seven microsatellite loci was surveyed in 224 and 68 specimens collected in Belgium and Germany, respectively. German samples were included as reference to investigate putative introduction source(s). At Natoye, 52 and 135 specimens were collected before and after the elimination campaign, respectively, to investigate temporal changes in the genetic composition and diversity.

Results: At Natoye, the genotypic microsatellite make-up showed a clear difference before and after the elimination campaign. Also, the population after 2017 displayed an increased allelic richness and number of private alleles, indicative of new introduction(s). However, the Natoye population present before the elimination programme is believed to have survived at low density. At the Belgian border, clustering results suggest a relation with the western German population. Whether the introduction(s) occur via passive human-mediated ground transport or, alternatively, by natural spread cannot be determined yet from the dataset.

Conclusion: Further introductions within Belgium are expected to occur in the near future, especially along the eastern Belgian border, which is at the front of the invasion of Ae. japonicus towards the west. Our results also point to the complexity of controlling invasive species, since 4 years of intense control measures were found to be not completely successful at eliminating this exotic at Natoye.

Keywords: Aedes japonicus japonicus; Introduction; Invasive mosquito; Microsatellites; Nad4 haplotypes; Population genetics; Temporal changes.

Fig. 1

Clusters for both K = 2 and K = 6 (K = number of genotypic clusters), inferred with Structure v2.3.4 software, after Evanno et al. [60] correction. The cluster membership of each individual is shown by the colour composition of the vertical lines, with the length of each coloured part of the line being proportional to the estimated membership coefficient. WG, western Germany; SWG, southwestern Germany; NG, northern Germany; B, Belgium. Colours of the pie chart represent the mean assignment probabilities for all individuals collected at Natoye to each of the clusters per collection year

Fig. 2

Map of Belgium and Germany showing the Bayesian cluster analysis results for K = 6 per sampling locality, based on seven microsatellite loci. Each pie chart represents one sampling location. Colours of the pie chart represent the mean assignment probabilities for all individuals collected at that location to each of the clusters, following the same colour code as in Fig. 1

Fig. 3

Principal coordinates analysis (PCoA) performed with GenAlEx v6.51b2 (first two axes explaining 56.18% of the genetic variability). Red indicates Belgian sampling localities; blue, violet and green indicates northern Germany, southwestern Germany and western Germany sampling localities, respectively

Fig. 4

Clusters for K = 2 at Natoye (Belgium), inferred with Structure v2.3.4 software, after Evanno et al. [60] correction. The cluster membership of each individual is shown by the colour composition of the vertical lines, with the length of each coloured part of the line being proportional to the estimated membership coefficient