Refine
Year of publication
- 2019 (2) (remove)
Document Type
- Article (2) (remove)
Language
- English (2)
Has Fulltext
- yes (2)
Is part of the Bibliography
- no (2) (remove)
Keywords
- Aedes aegypti (1)
- Aedes albopictus (1)
- Biogeography (1)
- Biogeography of disease (1)
- Climatic habitat suitability (1)
- Ecological modelling (1)
- Maxent (1)
- Species distribution modelling (1)
- Tiger mosquito (1)
- Vector borne diseases (1)
Background: Zika is of great medical relevance due to its rapid geographical spread in 2015 and 2016 in South America and its serious implications, for example, certain birth defects. Recent epidemics urgently require a better understanding of geographic patterns of the Zika virus transmission risk. This study aims to map the Zika virus transmission risk in South and Central America. We applied the maximum entropy approach, which is common for species distribution modelling, but is now also widely in use for estimating the geographical distribution of infectious diseases.
Methods: As predictor variables we used a set of variables considered to be potential drivers of both direct and indirect effects on the emergence of Zika. Specifically, we considered (a) the modelled habitat suitability for the two main vector species Aedes aegypti and Ae. albopictus as a proxy of vector species distributions; (b) temperature, as it has a great influence on virus transmission; (c) commonly called evidence consensus maps (ECM) of human Zika virus infections on a regional scale as a proxy for virus distribution; (d) ECM of human dengue virus infections and, (e) as possibly relevant socio-economic factors, population density and the gross domestic product.
Results: The highest values for the Zika transmission risk were modelled for the eastern coast of Brazil as well as in Central America, moderate values for the Amazon basin and low values for southern parts of South America. The following countries were modelled to be particularly affected: Brazil, Colombia, Cuba, Dominican Republic, El Salvador, Guatemala, Haiti, Honduras, Jamaica, Mexico, Puerto Rico and Venezuela. While modelled vector habitat suitability as predictor variable showed the highest contribution to the transmission risk model, temperature of the warmest quarter contributed only comparatively little. Areas with optimal temperature conditions for virus transmission overlapped only little with areas of suitable habitat conditions for the two main vector species. Instead, areas with the highest transmission risk were characterised as areas with temperatures below the optimum of the virus, but high habitat suitability modelled for the two main vector species.
Conclusion: Modelling approaches can help estimating the spatial and temporal dynamics of a disease. We focused on the key drivers relevant in the Zika transmission cycle (vector, pathogen, and hosts) and integrated each single component into the model. Despite the uncertainties generally associated with modelling, the approach applied in this study can be used as a tool and assist decision making and managing the spread of Zika.
Environmental niche modelling is an acclaimed method for estimating species’ present or future distributions. However, in marine environments the assembly of representative data from reliable and unbiased occurrences is challenging. Here, we aimed to model the environmental niche and distribution of marine, parasitic nematodes from the Pseudoterranova decipiens complex using the software Maxent. The distribution of these potentially zoonotic species is of interest, because they infect the muscle tissue of host species targeted by fisheries. To achieve the best possible model, we used two different approaches. The land distance (LD) model was based on abiotic data, whereas the definitive host distance (DHD) model included species-specific biotic data. To assess whether DHD is a suitable descriptor for Pseudoterranova spp., the niches of the parasites and their respective definitive hosts were analysed using ecospat. The performance of LD and DHD was compared based on the variables’ contribution to the model. The DHD-model clearly outperformed the LD-model. While the LD-model gave an estimate of the parasites’ niches, it only showed the potential distribution. The DHD-model produced an estimate of the species’ realised distribution and indicated that biotic variables can help to improve the modelling of data-poor, marine species.