Environmental factors associated with the distribution of Loa loa vectors Chrysops spp. in Central and West Africa: seeing the forest for the trees

Abstract

Background: Loiasis is caused by the filarial parasite Loa loa, which is widespread through Central and West Africa and largely confined the tropical equatorial rainforests. The tabanid flies Chrysops silacea and Chrysops dimidiata are the main vectors driving transmission. This study aimed to better define the spatial distribution and ecological niche of the two vectors to help define spatial-temporal risk and target appropriate, timely intervention strategies for filariasis control and elimination programmes.

Methods: Chrysops spp. distributions were determined by collating information from the published literature into a database, detailing the year, country, locality, latitude/longitude and species collected. Environmental factors including climate, elevation and tree canopy characteristics were summarised for each vector from data obtained from satellite modelled data or imagery, which were also used to identify areas with overt landcover changes. The presence of each Chrysops vector was predicted using a maximum entropy species distribution modelling (MaxEnt) method.

Results: A total of 313 location-specific data points from 59 published articles were identified across seven loiasis endemic countries. Of these, 186 sites were included in the climate and elevation analysis, and due to overt landcover changes, 83 sites included in tree canopy analysis and MaxEnt model. Overall, C. silacea and C. dimidiata were found to have similar ranges; annual mean temperature (24.6 °C and 24.1 °C, respectively), annual precipitation (1848.6 mm and 1868.8 mm), elevation (368.8 m and 400.6 m), tree canopy cover (61.4% and 66.9%) and tree canopy height (22.4 m and 25.1 m). MaxEnt models found tree canopy coverage was a significant environmental variable for both vectors.

Conclusions: The Chrysops spp. database and large-scale environmental analysis provides insights into the spatial and ecological parameters of the L. loa vectors driving transmission. These may be used to further delineate loiasis risk, which will be important for implementing filariasis control and elimination programmes in the equatorial rainforest region of Central and West Africa.

Keywords: Africa; Chrysops dimidiata; Chrysops silacea; Climate; Ecology; Environment; Loa loa; Loiasis; MaxEnt; Rainforest.

Fig. 1

Diagram of methodology used in Chrysops data points

Fig. 2

Map of Chrysops locations in the Central African countries

Fig. 3

MaxEnt model results plots. a, b Area under the curve (AUC) plots of both species models. Red line shows the mean of the 30 replicate MaxEnt runs and blue area the mean ± one standard deviation. c, d Jackknife test of regularized training gain. Dark blue columns show how would be the model gain using each variable in isolation. Light blue columns show how would change the model gain if the variable was excluded. The longest dark blue column turns to be the variable to have the most useful information by itself. The shortest light blue column appears to be the variable which has the most information that is not present in other variables

Fig. 4

Response curves of environmental variables in the two MaxEnt models for C. silacea (a, c, e, g) and C. dimidiata (b, d, f, h). The plots represent a MaxEnt model created using only the corresponding variable. The curves show the mean response of the 30 replicate MaxEnt runs (red line) and the mean ± one standard deviation (blue area)

Fig. 5

Predicted distribution maps for C. silacea (a) and C. dimidiata (b) vectors obtained from MaxEnt model data. Probability of occurrence is depicted in the form of percentages