Search

The Ebola virus is the causing agent of Ebola virus disease (EVD), an emerging infectious disease that can reach a 90% mortality rate in infected populations. The largest epidemic took place in West Africa in 2014-2016, whereas previously outbreaks had been confined to Central Africa. Multiple ecological and social factors may interact to cause an Ebola virus outbreak. Increased human – wildlife contacts in degraded forests or modified landscapes may lead to a higher risk of outbreaks. In this study we combine ecological knowledge and Earth observation to model Ebola virus transmission in a bat species (Hypsignathus monstrosus) suspected to be a reservoir for the virus. Land use maps, derived from satellite images, and information on habitat use by bats were integrated to a spatially explicit model to study how land use changes may influence virus transmission in this species. We used land use maps from 2005 and 2015 of an area of approximately 190km2 in Guinea, where bat-human interactions are common, as a starting point for our simulations. During this period in the study area, forest area decreased while the surface of forest-agricultural mosaics and savannahs increased, reflecting a common landscape transformation in the region. Habitat use by bats was modeled based on previous and on on-going ecological studies on the species. Virus transmission among bats is more likely when groups of individuals come together, such as during roosting and foraging. To integrate this information into our model, we assigned indexes of habitat suitability to each land use for these two behaviors. Results of our simulations show how land use changes affect animal behavior of an Ebola virus host, and the impact this virus can have on transmission within host populations. Deforestation is considered an important risk factor in the emergence of some infectious diseases, notably Ebola. Western Africa has undergone important landscape and human population changes in the last few decade

The unexpected Ebola virus outbreak in West Africa in 2014 involving the Zaire ebolavirus made clear that other regions outside Central Africa, its previously documented niche, were at risk of future epidemics. The complex transmission cycle and a lack of epidemiological data make mapping areas at risk of the disease challenging. We used a Geographic Information System-based multicriteria evaluation (GIS-MCE), a knowledge-based approach, to identify areas suitable for Ebola virus spillover to humans in regions of Guinea, Congo and Gabon where Ebola viruses already emerged. We identified environmental, climatic and anthropogenic risk factors and potential hosts from a literature review. Geographical data layers, representing risk factors, were combined to produce suitability maps of Ebola virus spillover at the landscape scale. Our maps show high spatial and temporal variability in the suitability for Ebola virus spillover at a fine regional scale. Reported spillover events fell in areas of intermediate to high suitability in our maps, and a sensitivity analysis showed that the maps produced were robust. There are still important gaps in our knowledge about what factors are associated with the risk of Ebola virus spillover. As more information becomes available, maps produced using the GIS-MCE approach can be easily updated to improve surveillance and the prevention of future outbreaks.

Spatial scaling to some extent determines biodiversity patterns in larger organisms, but its role in microbial diversity patterns is much less understood. Some studies have shown that bacterial community similarity decreases with distance, whereas others do not support this. Here, we studied soil bacterial communities of tropical rainforest in Malaysia at two spatial scales: a local scale with samples spaced every 5 m over a 150-m transect, and a regional scale with samples 1 to 1,800 km apart. PCR-amplified soil DNA for the bacterial 16S rRNA gene targeting the V1–V3 region was pyrosequenced using Roche/454 GS FLX Titanium platform. A ranked partial Mantel test showed a weak correlation between spatial distance and whole bacterial community dissimilarity, but only at the local scale. In contrast, environmental distance was highly correlated with community dissimilarity at both spatial scales, stressing the greater role of environmental variables rather than spatial distance in determining bacterial community variation at different spatial scales. Soil pH was the only environmental parameter that significantly explained the variance in bacterial community at the local scale, whereas total nitrogen and elevation were additional important factors at the regional scale. We obtained similar results at both scales when only the most abundant OTUs were analyzed. A variance partitioning analysis showed that environmental variables contributed more to bacterial community variation than spatial distance at both scales. In total, our results support a strong influence of the environment in determining bacterial community composition in the rainforests of Malaysia. However, it is possible that the remaining spatial distance effect is due to some of the myriad of other environmental factors which were not considered here, rather than dispersal limitation.