Since 2000, Ae. albopictus has been detected at ten PoE’s spread throughout the Belgian territory. Based on the observations made during the period 2007 – 2020, the species was not yet able to establish given that no evidence for overwintering was found. Yet, indoor and outdoor summer reproduction of Ae. albopictus, as indicated by detection of immature life stages, occurred at and around some PoE’s where control measures were not implemented, or only implemented two to three months after the first detection of Ae. albopictus (Additional file 1: Table S3). As reproduction of Ae. albopictus in Belgium is possible, the time gap between the detection and control should be as small as possible to lower the risk of possible spread and establishment.
The earliest detection of Ae. albopictus during the mosquito season (April-November) in a non-sheltered environment was in May 2018 at the used tyre import company AB in Kallo. This finding precedes the earliest detection of the species in a non-sheltered environment in the Netherlands, applying similar methods and covering the same sampling period . This early observation might be due to an early introduction combined with suitable climatic conditions in April – May 2018, which were exceptionally warm  and might have favoured the development and survival of the introduced individuals.
The last specimens of Ae. albopictus were predominantly collected in September up to the end of October, comparable to the situation in the Netherlands . Indoors, introduced Ae. albopictus were collected once as larva in February 2016, which was probably linked to the increased import of lucky bamboo from Guangdong (southern China) for the Chinese New Year. Further, twice Ae. albopictus was collected in November, again linked to the import of lucky bamboo. These findings point to the fact that the species can be introduced through the lucky bamboo trade at any time of the year .
The companies trading used tyres and lucky bamboo plants are well-known introduction routes [53, 56, 57]. Since the first detection of Ae. albopictus in Belgium in 2000 , 12 Ae. albopictus introductions through used tyre and lucky bamboo trade were recorded. When preventive larviciding was implemented at these well-known and high risk PoE’s, no new detections were made anymore. Other preventive measures such as obligate covering of used tyres and obligate treatment of lucky bamboo water should be taken to regulate the used tyre and lucky bamboo import in Belgium. In the Netherlands such a regulation for prevention of Ae. albopictus is already in place in the form of covenants and a legislation for import of these commodities (used tyres and lucky bamboo) . But even with these regulations in place, active surveillance remains necessary to early detect introduction of IMS and evaluate control measures .
The detection of Ae. albopictus at two Belgian parking lots in 2018 points towards a short-distant introduction pathway, already observed in other European countries [56, 58, 59, 60, 61, 62] but new for Belgium. For example, ongoing introductions of Ae. albopictus in southern England occurred via ground vehicular traffic through train and ferry from nearby established populations in France . The expanding population of Ae. albopictus in northern France (departments of Val-de-Marne, Seine-Saint-Denis, Hauts-de-Seine, Seine-et-Marne, Essonne, Aisne Bas-Rhin) or in Germany (Baden-Württemberg and Hesse) [64, 65, 66] most likely are the source of the introduced Ae. albopictus found at the parking lots in Belgium, as these are located within a 2 – 2.5 hours’ drive from these populations . For vehicles coming from colonised areas in northern France and Germany, the monitored parking lots were often the first stop after crossing the border. Also the recent detections at parking lots in the Netherlands at the northern Belgian border  suggest that Ae. albopictus can be introduced anywhere in Belgium. Certainly, Ae. albopictus populations are progressively establishing closer to Belgium and more frequent short-distant introductions will happen in the coming years.
Since 2002, Ae. japonicus has been locally established at a used tyre import company at Natoye . At that time tyres were imported from its home range in Japan and from its invasive range in the USA, where the species is established, making the international used tyre trade the probable introduction pathway . During more than ten years, Ae. japonicus was reported as locally established species in a 2 km perimeter around the tyre import company, without further spread. This was in strong contrast to the fast spread of the species observed in other European countries . The abundance and range of the species decreased during the first Belgian IMS elimination campaign undertaken between 2012 and 2015 [24, 50, 51], but a new southwest spread was recorded between 2017 and 2019, associated with a progressive increase in abundance. It seems that a small forest located southwest of the tyre company forms a good ‘shrub-corridor’ for the spread of the mosquito [38, 69]. At the northeast side, the company is surrounded by open meadows, which possibly hamper the spread of the population in that direction.
An investigation of the genetic variation at seven microsatellite loci indicated that remnants of the original population were still present, although no detections were done in 2015 and 2016 during the control campaign . Natural larval habitats (tree holes) in the small forest probably played a role as ‘refuge’ [21, 38, 70]. The current admixed population displayed a higher allelic richness than the original population, which points towards at least one re-introduction of Ae. japonicus from an external source population after the elimination campaign . Tyres are regularly imported from an Ae. japonicus colonised area in Germany , which might be the possible source of this new introduction. Currently, the tyre company only imports tyres from Europe (mainly Luxembourg, France, Italy, Spain, Germany and the Netherlands) (HA, personal communication). German Ae. japonicus populations display high invasiveness potential as indicated by their fast spread after establishment all over the country . High genetic diversity might be a component of their invasiveness success [72, 73]. A new introduction possibly increased the genetic diversity of the Belgian population, which could explain the apparently faster spread seen in 2019 compared to that before 2012. Controlling Ae. japonicus at HA was proven to be challenging and given the continuous tyre trade with colonised countries like Germany, new introductions are likely to occur, with further genetic admixture.
In the Netherlands, Ae. japonicus was associated with allotment gardens [74, 75]. Likewise, this species was found at an allotment garden along the country border with Germany. The monitoring results, supported by a genetic investigation , point to the phenomenon of multiple introductions in Belgium from the close-by population in West-Germany [64, 68, 76]. In fact, the late detection in June 2018 and the limited number of specimens collected, together with the absence of adults, nor of egg laying females (i.e., no eggs were collected with OT) in 2019, suggest the absence of an established population in the allotment garden. Yet, the detections of Ae. japonicus around the allotment garden later in the year confirmed summer reproduction of the species. The range expansion of Ae. japonicus in Austria, Italy and Switzerland, seems to be mainly driven by active dispersion, next to passive ground transport [62, 77, 78, 79]. Whether the introductions in Belgium occurred via one or the other pathway could not be determined from the genetic dataset . Yet, the users of the allotment garden sometimes cross the Belgian-German border to buy plants for their garden. Hence, introduction through hitchhiking mosquitoes seems to be plausible.
At the industrial area in Maasmechelen (MM) close to the German border, adults of Ae. japonicus were found only between mid-June and mid-July 2018, and thus a single introduction is hypothesised, which is supported by the genetic results . Interestingly, it was the first time that Ae. japonicus and Ae. koreicus co-occurred at a PoE in Belgium. Co-occurrences of Ae. japonicus and Ae. koreicus have been previously reported from Germany, Switzerland, Slovenia and Italy [35, 79].
Since 2008, Ae. koreicus has established at the industrial area 'Op de Berg' (MM) . Although the import route of this IMS into Belgium remains unknown, international trade was speculated because of the large industrial zone surrounding the area . Aedes koreicus uses a variety of man-made breeding sites [35, 80]. In Belgium these included metal containers (such as old construction equipment (mainly excavator heads)), small and large tyres, plastic containers (such as buckets, trays) and plastic sheets . Aedes koreicus was also found once in temporary muddy road tracks (2008) . The frequent observations of this species at more than 2.8 km from the point of first detection indicates a widespread Ae. koreicus population. However, most specimens can be found at the industrial area itself which remains the hot spot for this species. The detection of some larvae and adults at the used tyre import company in Dilsen-Stokkem (MB) at 5.4 km from the industrial area MM was the furthest detection of this species. Additionally, the fact that mainly adults were collected at MB, often late in the reproduction season (September – October), suggests seasonal spread from the established population at MM. Although introduction by tyre trade at MB cannot be ruled out since the company imports tyres from a wide number of countries (MB, personal communication). It remains unclear why the species did not disperse faster, in contrast to other populations in Europe where it also spreads passively by road transport [35, 62, 80]. The spread found in the northeast of the industrial area might be explained by the presence of the forest ‘Mechelse Bos’, which forms a good ‘shrub-corridor’ for the mosquito to spread compared with the open terrain of the sand quarry and the heath at the other sides of the industrial area. In general, mosquitoes prefer to fly through bushes and shrubs and avoid dry and open terrain . Yet, low availability of PLH, other than tree holes, in the forest ‘Mechelse Bos’ next to the industrial area MM might have slowed down the active spread. We recommend to further investigate the host preferences of the Belgian Ae. koreicus population, as there have never been biting complaints from the people working and living at and around the industrial area . Hypothetically, the population could feed mainly on non-human mammals, in contrast to other studies , which might explain the apparent lack of passive spread through ground traffic.
Implications for the future
Based on the current spread of Ae. albopictus, particularly in France and Germany, on the increasing number of interceptions of this species in Belgium and on the suitability models developed in Europe [10, 82, 83], establishment of this species in Belgium is to be expected. Not only Ae. albopictus, but also Ae. japonicus is gaining more and more territory in neighbouring countries: Germany [58, 61, 64, 84], the Netherlands , Luxembourg , France [60, 66], United Kingdom (UK) [63, 86]. Anno 2020, both species have established populations at less than 200 km from Belgium, and Ae. japonicus is even reaching the Belgian border in the east. We observed a clear increasing trend of the percentage of positive PoE’s for Ae. albopictus in Belgium over the last 13 years. On top of the still important long-distance IMS import with well-defined PoE’s (especially the used tyre and lucky bamboo import companies), now also short-distance import with less-defined PoE’s occurs. For Belgium, the passive ground transport is a new and important introduction pathway for especially Ae. albopictus and probably also for Ae. japonicus. Natural dispersal for the latter species into Belgium is expected as well. Belgium literally is at the front of their invasion range and highlights the need for an integrated management programme in contrast to the current short-term project-based monitoring. Such a programme should include not only active and passive surveillance for the detection of IMS at both well-known and unknown PoE’s, but also a clear and complete control management plan at national and regional levels setting out clear criteria for action, control methods and strategies with appropriate implementation, supervision and evaluation. The decision and implementation of control of IMS species in Belgium differs between regions and is not always based on epidemiological and entomological risk scenarios such as described by Roiz et al. . The IMS control in Belgium is often too ad hoc as it is done reactively or proactively depending on the available biocides, budget and sometimes political priorities, and it also depends on what the policy is in neighbouring countries, e.g., the lack of Ae. japonicus management in Germany. Finally, involvement of local authorities (municipalities, provinces and local Public Health units), social mobilisation, cross-border collaboration and a link with arbovirus surveillance should be added to the surveillance and control management plan.