Describing fine spatiotemporal dynamics of rat fleas in an insular ecosystem enlightens abiotic drivers of murine typhus incidence in humans

Abstract

Murine typhus is a flea-borne zoonotic disease that has been recently reported on Reunion Island, an oceanic volcanic island located in the Indian Ocean. Five years of survey implemented by the regional public health services have highlighted a strong temporal and spatial structure of the disease in humans, with cases mainly reported during the humid season and restricted to the dry southern and western portions of the island. We explored the environmental component of this zoonosis in an attempt to decipher the drivers of disease transmission. To do so, we used data from a previously published study (599 small mammals and 175 Xenopsylla fleas from 29 sampling sites) in order to model the spatial distribution of rat fleas throughout the island. In addition, we carried out a longitudinal sampling of rats and their ectoparasites over a 12 months period in six study sites (564 rats and 496 Xenopsylla fleas) in order to model the temporal dynamics of flea infestation of rats. Generalized Linear Models and Support Vector Machine classifiers were developed to model the Xenopsylla Genus Flea Index (GFI) from climatic and environmental variables. Results showed that the spatial distribution and the temporal dynamics of fleas, estimated through the GFI variations, are both strongly controlled by abiotic factors: rainfall, temperature and land cover. The models allowed linking flea abundance trends with murine typhus incidence rates. Flea infestation in rats peaked at the end of the dry season, corresponding to hot and dry conditions, before dropping sharply. This peak of maximal flea abundance preceded the annual peak of human murine typhus cases by a few weeks. Altogether, presented data raise novel questions regarding the ecology of rat fleas while developed models contribute to the design of control measures adapted to each micro region of the island with the aim of lowering the incidence of flea-borne diseases.

Télédétection et modélisation spatiale

“…Exploration du comportement du modèle À l'échelle d'un territoire insulaire tropical soumis à différents risques vectoriels comme l'île de La Réunion, des réussites récentes illustrent tout le potentiel de ces approches de modélisation appliquées comme une boîte à outils. La caractérisation spatiale et saisonnière du risque lié au typhus murin ou aux « bavites » 1 permet de mettre en place des campagnes de sensibilisation ciblées et adaptées auprès des professionnels de santé ou des éleveurs (Grimaud et al, 2021(Grimaud et al, , 2019Tran et al, 2021). Des outils cartographiques de prédiction des densités vectorielles et du risque de transmission de la dengue sont appropriés et utilisés depuis plusieurs années par l'opérateur de lutte antivectorielle (ARS La Réunion) pour optimiser la surveillance entomologique et les actions de prévention et de contrôle .…”

Modelling the Potential Human Exposure to Japanese Encephalitis Virus (JEV) in Case of Introduction into Reunion Island

“…Since 2018, dengue seems to have become endemic in the island with (i) an uninterrupted circulation of the virus including during the austral winter, (ii) a cocirculation of diferent serotypes, and (iii) an increase in the number of cases with severe forms and deaths [18]. Vector-borne zoonotic diseases have recently emerged in Reunion Island, such as fea-borne murine typhus [19]. Two vector-borne zoonotic viruses are under special surveillance because they circulate in the region: Rift Valley fever virus (circulating in Madagascar [20] and in Mayotte [21]) and West Nile fever virus, circulating in Madagascar and for which antibodies have been detected in horses in Reunion Island, with no observed indigenous human cases [22,23].…”