The Crimean-Congo haemorrhagic fever tick vector Hyalomma marginatum in the south of France: Modelling its distribution and determination of factors influencing its establishment in a newly invaded area

Abstract

We developed a correlative model at high resolution for predicting the distribution of one of the main vectors of Crimean-Congo haemorrhagic fever virus (CCHFV), Hyalomma marginatum, in a recently colonised area, namely southern France. About 931 H. marginatum adult ticks were sampled on horses from 2016 to 2019 and 2021 in 14 southern French departments, which resulted in the first H. marginatum detection map on a large portion of the national territory. Such updated presence/absence data, as well as the mean number of H. marginatum per examined animal (mean parasitic load) as a proxy of tick abundance, were correlated to multiple parameters describing the climate and habitats characterising each collection site, as well as movements of horses as possible factors influencing tick exposure. In southern France, H. marginatum was likely detected in areas characterised by year-long warm temperatures and low precipitation, especially in summer and mostly concentrated in autumn, as well as moderate annual humidity, compared to other sampled areas. It confirms that even in newly invaded areas this tick remains exclusively Mediterranean and cannot expand outside this climatic range. Regarding the environment, a predominance of open natural habitats, such as sclerophyllous vegetated and sparsely vegetated areas, were also identified as a favourable factor, in opposition to urban or peri-urban and humid habitats, such as continuous urban areas and inland marshes, respectively, which were revealed to be unsuitable. Based on this model, we predicted the areas currently suitable for the establishment of the tick H. marginatum in the South of France, with relatively good accuracy using internal (AUC = 0.66) and external validation methods (AUC = 0.76 and 0.83). Concerning tick abundance, some correlative relationships were similar to the occurrence model, as well as the type of horse movements being highlighted as an important factor explaining mean parasitic load. However, the limitations of estimating and modelling H. marginatum abundance in a correlative model are discussed.

Keywords: France; Hyalomma marginatum; Tick; correlative distribution modelling.

FIGURE 1

Localisation of the sampling sites of Hyalomma marginatum across 14 French departments. Coloured circles represent the sites where H. marginatum is present and circle size corresponds to the average number of H. marginatum per horse in each equestrian structure. Red circles correspond to the surveys conducted from 2016 to 2019, while blue circles correspond to the 2021 survey. White circles correspond to sites where H. marginatum was considered absent during the 2016–2019 surveys and white triangles for the 2021 survey. Black crosses represent the sites where only one specimen of H. marginatum was found during the 2016–2019 surveys (not included in the modelling)

FIGURE 2

Principal component analysis of the climatic data, under the principal components 1 (PC1) and 2 (PC2) (2A) and the principal components 2 (PC2) and 3 (PC3) (2B). Colours represent the squared sum of the coordinates of the climatic variables on the principal components. Red corresponds to high correlation whereas blue means low correlation. The horizontal axis corresponds to PC1 in Figure 2a and PC2 in Figure 2b, and the vertical axis to PC2 in Figure 2a and PC3 in Figure 2b. Tmax: maximum temperature, Tmin: minimum temperature, ETP: potential evapotranspiration, Prec: precipitation, Humr: relative humidity. The numerical suffix corresponds to the month (1 for January to 12 for December)

FIGURE 3

Conditional effects of climate synthetic variables PC1, PC2 and PC3 on the conditional tick abundance (3A, 3C, 3E) and the probability of tick presence (3B, 3D, 3F) using the most parsimonious model, when all other parameters are fixed to their mean. The X‐axis on conditional tick abundance curves (left side) represents the range of PCs values on observed tick presence

FIGURE 4

Predicted probability of presence for H. marginatum in the south of France, using the most parsimonious model (dark blue: null probability to bright yellow: highest probability). Red circles represent the observed presence and white circles observed absences from testimonies and surveys