Geographic distribution and future expansion of Aedes albopictus in the Democratic Republic of the Congo

Aedes albopictus with an Asian origin has been reported from central African countries. The establishment of this mosquito species poses a serious threat as the vector of various infectious diseases. Since information about Ae. albopictus in Democratic Republic of the Congo (DRC) is scarce, we investigated the current distribution of this mosquito species. Based on the factors affecting the distribution, we predicted future distribution. We conduced entomological surveys in Kinshasa and three neighboring cities from May 2017 to September 2019. The survey was extended to seven inland cities. A total of 19 environmental variables were examined using the maximum entropy method to identify areas suitable for Ae. albopictus to establish a population. We found Ae. albopictus at 21 of 23 sites in Kinshasa and three neighboring cities. For the first time Ae. albopictus was also found from three of seven inland cities, while it was not found in four cities located in the eastern and southeastern parts of DRC. A maximum entropy model revealed that the occurrence of Ae. albopictus was positively associated with maximum temperature of the warmest month, and negatively associated with wider mean diurnal temperature range and enhanced vegetation index. The model predicted that most parts of DRC are suitable for the establishment of the mosquito. The unsuitable areas were the eastern and southeastern highlands, which have low temperatures and long dry seasons. We confirmed that Ae. albopictus is well established in Kinshasa and its neighboring cities. The expansion of Ae. albopictus to the inland is ongoing, and in the future the mosquito may establish in most parts of DRC.

Curative treatments and vaccines are not available for dengue and chikungunya [31,32], and 67 thus vector control is a valuable available tool for reducing infections [ 33]. As such, 68 understanding the current distribution of Ae. albopictus in DRC is an essential step for the 69 control. Global level distribution models based on environmental variables indicate that 70 almost the entire area of DRC is suitable for A. albopictus establishment [3, 34, 35]. These 71 models were constructed without entomological data from DRC, and thus the provided 72 information was too coarse to apply to local vector control. In the present study, we described 73 the current distribution of Ae. albopictus in DRC based on locally available data. In particular, 74 we provided detailed information for Kinshasa and the neighboring areas where chikungunya 75 outbreaks recently occurred. We also revealed important environmental variables related to 76 the distribution, and attempted to determine if the present distribution is static.    (Table 1). First, we focused on the western region in which Ae. albopictus has been 106 recorded [13,25]. The survey in the western region included 14 sites within Kinshasa and 107 nine sites in the three cities, Kasangulu, Kisantu, and Matadi, in Kongo Central Province.

108
Since human-mediated dispersal of Ae. albopictus was an immediate concern, the survey also 109 included nine sites along the major transportation routes (Congo River and national roads) in 110 the other three regions (Fig 1). These sites were three sites within Mbandaka in the western  Inc, Regensburg, Germany) from 3:00 pm to 6:00 pm for three to seven consecutive days at 122 each site. Sampled mosquitoes were identified morphologically to species according to 123 Huang's identification keys [36]. When at least one Ae. albopictus was collected, the site was  Lubumbashi, in the southeastern region (Table 1).

181
A total of 19 environmental variables were selected based on a literature review (Table 2). 12 217 An asterisk indicates that that the difference was statistically significant (p < 0.05) with 218 Wilcoxon-Mann-Whitney tests. Of 12 selected variables, excluding dry season length, five pairs were highly correlated 233 among eight variables (S1 File). We chose annual mean temperature, mean diurnal 234 temperature range and the EVI over the others because the past studies showed that they were 235 more important. As a result, seven variables were included in the full MaxEnt analysis (Table   236 2). After the model selection, the optimal model contained three variables, maximum 237 temperature of the warmest month, mean diurnal temperature range, and EVI. Mean diurnal 238 temperature range was the most important variable, followed by maximum temperature of 239 warmest month, and EVI ( Table 2). The AUC of the optimal model was 0.975. The response 240 curves revealed that the highest suitable area was predicted with EVI below -0.017, 13 241 maximum temperature of the warmest month above 34.3 o C, and mean diurnal temperature 242 below 6.5°C (Fig 4). The model predicted that most of DRC is suitable for Ae. albopictus establishment (Fig 5).

250
The suitability was high in the most parts of the western region; however, it varied between 0 251 to 75% in the southern area of the region. The suitability was also high in the central region  We also confirmed that Ae. albopictus has extended its distribution to the inland cities. This   The environmental variables used in the present study were selected based on studies 382 conducted mostly in temperate areas, because few studies were conducted in Africa.

383
Appropriate variables for the African situation might be different.