Wetlands, wild Bovidae species richness and sheep density delineate risk of Rift Valley fever outbreaks in the African continent and Arabian Peninsula

Michael G. Walsh; Allard Willem de Smalen; Siobhan M. Mor

doi:10.1371/journal.pntd.0005756

Abstract

Rift Valley fever (RVF) is an emerging, vector-borne viral zoonosis that has significantly impacted public health, livestock health and production, and food security over the last three decades across large regions of the African continent and the Arabian Peninsula. The potential for expansion of RVF outbreaks within and beyond the range of previous occurrence is unknown. Despite many large national and international epidemics, the landscape epidemiology of RVF remains obscure, particularly with respect to the ecological roles of wildlife reservoirs and surface water features. The current investigation modeled RVF risk throughout Africa and the Arabian Peninsula as a function of a suite of biotic and abiotic landscape features using machine learning methods. Intermittent wetland, wild Bovidae species richness and sheep density were associated with increased landscape suitability to RVF outbreaks. These results suggest the role of wildlife hosts and distinct hydrogeographic landscapes in RVF virus circulation and subsequent outbreaks may be underestimated. These results await validation by studies employing a deeper, field-based interrogation of potential wildlife hosts within high risk taxa.

Author summary

Rift Valley fever (RVF) is a vector-borne zoonotic disease that imparts a substantial burden to the economy and public health of pastoralist communities across the African continent and Arabian Peninsula. Furthermore, RVF is also an emerging pathogen of growing global concern. Knowledge of the epidemiological and ecological factors that influence the geographic distribution of RVF outbreaks and determine risk for humans and animals is incomplete. The current study examined the distribution of RVF outbreaks from 1998 to 2016 and modeled their occurrence as a function of climate, surface water, land cover, livestock density, wild mammalian species richness, and human migration. The results indicate that wetlands, Bovidae species richness, and sheep density were associated with increased risk of RVF outbreaks. Our findings contribute to improved understanding of the spatial and ecological dynamics of RVF risk with a particular emphasis on the distribution of wetlands and potential wildlife reservoirs in designing RVF surveillance programs.

Citation: Walsh MG, Willem de Smalen A, Mor SM (2017) Wetlands, wild Bovidae species richness and sheep density delineate risk of Rift Valley fever outbreaks in the African continent and Arabian Peninsula. PLoS Negl Trop Dis 11(7): e0005756. https://doi.org/10.1371/journal.pntd.0005756

Editor: Benjamin Althouse, Institute for Disease Modeling, UNITED STATES

Received: December 7, 2016; Accepted: June 29, 2017; Published: July 25, 2017

Copyright: © 2017 Walsh et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: RVF occurrence files are available from the Figshare database (https://doi.org/10.6084/m9.figshare.4272620.v1).

Funding: The authors received no specific funding for this work.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Rift Valley fever (RVF) is an emerging, vector-borne viral zoonosis that causes significant morbidity in humans and their livestock. The etiologic agent, Rift Valley fever virus (RVFV), is a Phlebovirus in the Bunyaviridae family and is transmitted by several mosquito species that facilitate viral maintenance (Aedes spp.) or amplification (Culex spp.) [1,2]. Human infections are invariably asymptomatic or mild in early stages, however, severe cases can manifest as hemorrhagic fever or encephalitis [3,4]. Sheep, goats, and cattle experience fetal abortions as a result of RVFV infection, and the disease contributes to substantive economic losses to pastoralist communities during outbreaks [5,6].

Historically, most outbreaks in humans and domestic animals have occurred in the African continent, and in eastern Africa these typically follow periods of excessive rain in poorly draining arid or semi-arid landscapes [7]. The resultant flooding is conducive to the breeding and hatching of infected mosquitoes which transmit the virus to ruminant hosts followed by eventual secondary transmission to other ruminants and humans [5]. In more recent years, RVF has progressively expanded east into the Arabian Peninsula, with outbreaks in Saudi Arabia and Yemen [8].

The epidemiology and infection ecology of RVFV is complex and our knowledge of these incomplete. Several species in two distinct mosquito genera transmit infection. As primary vectors, Aedes mosquitoes maintain RVFV transovarially during dry periods; during these times there are little to no reported human or livestock infections. Aedes mosquito population explosion following wet periods leads to localized transmission to mammalian hosts [9,10]. Following this, Culex mosquitoes can expand (amplify) transmission to more dispersed livestock and human populations distant from the areas of local Aedes transmission [2,11,12]. Once RVFV becomes amplified in livestock, ongoing human infection occurs primarily through zoonotic transmission as a result of direct or indirect contact with animal tissues and body fluids, such as occurs during slaughtering or through performing obstetrical procedures on infected animals. Transmission from mosquitoes that feed on infected animals is also a viable though less important source of human infection [1,13].

While the role of vectors in RVFV infection ecology is well-established, the extent to which wildlife contributes to transmission as possible maintenance or amplification hosts is not well understood. Field investigations suggest that wild ruminants and rodents are the most likely RVFV reservoirs [14]. Nevertheless, data from these field surveys are limited, so definitive mammalian natural reservoirs for RVFV are not described [11,14].

The landscape epidemiology of RVFV is also incomplete with respect to abiotic systems of influence. For example, periods of excessive rain are strongly associated with RVF outbreaks in East Africa [7,15–18], however very little is known regarding the interaction between climate and terrestrial or hydrogeographic profiles in mediating RVF outbreaks [19]. In addition, there has been a lack of attention to land cover characteristics, which have the potential to influence mosquito habitat, sylvan reservoir habitat, and the movement of domestic livestock through the landscape. Finally, anthropogenic influence, such as human migration, may introduce novel, or increase existing, exposures among pastoralist and/or other rural and peri-urban communities [20].

The study sought to expand our current understanding of RVF epidemiology and infection ecology by investigating the role of diverse hydrogeographic features and wild Bovidae and Muridae species richness in delineating the landscape suitability of future outbreaks across the African continent and Arabian Peninsula.

Methods

Data sources

Occurrence data for RVF outbreaks in humans and livestock animals were obtained from the ProMED-mail electronic surveillance system. This surveillance system is maintained by the International Society of Infectious Diseases and provides near real time and archival documentation of formal and informal reports of infectious diseases [21]. The database was searched using the keywords “rift valley fever”, “rift valley fever virus”, “rvf”, and “rvfv”. Only those reports documenting RVF outbreaks in humans or livestock in unique locations were included (i.e. duplicate outbreaks were not included). One hundred and three reports of laboratory confirmed, geolocated outbreaks of RVF in humans and livestock were documented by the ProMED system between January 1, 1998 and August 31, 2016. Google Maps was used to capture the geographic coordinates for each outbreak and cross-checked against Open Street Map. Centroids of the reported outbreak locations were recorded to a spatial resolution of 4km².

To test our landscape suitability model (see Statistical Analysis section), a second source of RVF outbreak data were obtained from the World Organization for Animal Health (OIE). OIE maintains an official biosurveillance mechanism for RVF in livestock. These data have been archived since 2004 and can be accessed via the World Animal Health Information System (WAHIS) web portal [22]. Reports included the location of each event by place name, the date, type of livestock affected, and the number of infected animals identified. Between January, 2005 and August, 2016 a total of 50 "immediate notification" and subsequent “follow-up” RVFV outbreak reports were submitted to OIE. The geographic coordinates for these events were obtained with Google Maps as above. Outbreaks from ten of these reports could not be located within this coordinate reference system. This left a total 40 OIE reports with 102 unique outbreak occurrences. Twenty-three of the OIE documented outbreaks were also recorded in the ProMED surveillance and therefore were not included in this testing dataset to prevent inflation of model performance. Thus, the final OIE sample of 79 was used for model testing.

Altitude and four Bioclim climate rasters were obtained from the WorldClim Global Climate database and used as climate indicators for this investigation [23]. Aggregate spatio-temporal weather station data between 1950 and 2000 were used to calculate the mean temperature during the hottest and coldest quarters, and the mean precipitation during the wettest and driest quarters, and extracted as 30 arc second (approximately 1 km²) resolution rasters [24].

Vegetation cover was assessed using the MODIS-based Maximum Green Vegetation Fraction (MGVF), which is a data product from the United States Geologic Survey's Land Cover Institute [25]. The MGVF records the percentage of green vegetation cover per pixel as a function of the normalized difference vegetation index at a resolution of 1 km²[26]. Rasters were obtained at two time points, years 2001 and 2010, and the difference between them calculated to determine vegetation loss over this 10 year period. Change in MGVF over this time period was considered a more robust representation of vegetation cover than mean MGVF, and therefore more appropriate in assessing its influence on RVF landscape suitability.

The Global Lakes and Wetlands Database [27] was used to define surface water. This raster was derived from three discrete components. The first two comprised vector data of polygons. Component 1 represented lakes with area ≥ 50 km² and controlled water reservoirs with volume ≥ 0.5 km³, while component 2 represented all surface water with area ≥ 0.1 km². The third component combined and rasterized the polygon data from the first two components, while supplementing the wetland data. The final 1 km² raster based on component three was used here. The surface water categories were: lake, controlled water reservoir, river, freshwater marsh, swamp, coastal wetland, brackish, bog, or intermittent wetland [28]. The surface water types were extracted and new distance rasters created. Distance was calculated in the QGIS geographic information system using the proximity function to produce separate 1 km² resolution rasters for each water category[29]. Pixel values in these rasters convey the distance in kilometers between a given pixel and the nearest pixel occupied by each unique category of surface water. In this way the models can incorporate a spectrum of proximity to diverse hydrogeography across the metacontinent (see Statistical Analysis section).

Net human population migration was obtained as a 30 arc-second raster from the Socioeconomic Data and Applications Center (SEDAC), which is part of the National Aeronautics and Space Agency's Earth Observing System Data and Information System [30]. This raster describes the net change (increase vs. decrease) in persons per km² from the period 1990 to 2000 [31].

The global densities of cattle, sheep, and goats were represented as 1 km² resolution rasters from the Gridded Livestock of the World (GLW) [32]. The GLW also classified ruminant livestock production systems by system (livestock-only, mixed rain fed, and mixed irrigated) and climate regime (Hyper-arid, Arid, Humid, and Temperate/Tropical Highlands) comprising 12 production system categories plus one additional category classified as Urban [33].

Rasters of Bovidae and Muridae species richness at 1 km² resolution were acquired from the International Union for Conservation of Nature (IUCN) and Center for International Earth Science Information Network (CIESIN)[34].

Finally, all species of Aedes mosquitoes observed across the geographic range of RVF outbreaks were extracted from the Global Biodiversity Information Facility (GBIF)[35]. There were 215 field observations of Aedes mosquitoes geolocated within the African continent and Arabian Peninsula. However, of these 215 mosquito observations, 151 observations recorded the genus only without species designation, while 57 were Ae. africanus, and 7 were Ae. albopictus. As such, there was not sufficient species representation in the GBIF to produce valid models of the ecological niche of Aedes vectors. One generic ecological niche model of Aedes mosquitoes was included in an exploratory analysis, but this contributed very little to the loss function when modeling RVF landscape suitability (see Statistical Analysis section), further suggesting that the mosquito data were insufficient to include in the current investigation. Similarly, this analysis did not include potential Culex amplification vectors as there were too few GBIF specimens across the region and those that were present were of too diverse an ecological and behavioral spectrum to be pooled for analysis.

Statistical analysis

This study used maximum entropy (Maxent) machine learning to model the landscape suitability of RVF outbreaks in human and livestock hosts across Africa and the Arabian Peninsula at a resolution of 4 km². In the current study, risk is defined explicitly as the probability of landscape suitability to RVF outbreaks. Machine learning in general, and Maxent in particular, is analytically appealing because a specific model form is not assumed. Instead algorithms create rule-based data partitions that optimize homogeneity between predictors and outcomes [36]. Further, the Maxent machine learning algorithm does not require the locations of RVF outbreak absences which are effectively unknowable [37,38].

The full Maxent model (based on ProMED data) comprised the following landscape features: mean dry quarter precipitation; mean warm and cold quarter temperature; change in vegetation cover; proximity to the surface water features; wild Bovidae and Muridae species richness; cattle, sheep, and goat densities; and net human migration between 1990 and 2000. Correlation between most of the landscape factors acquired for this study was low. However, there were a few exceptions (wet quarter precipitation, ruminant production systems, swamp, and altitude), all of which were correlated with several other landscape factors and provided generally redundant information. Therefore, these factors were dropped from the original 22 predictors acquired from our data sources described above. Ten thousand background points were sampled, weighted according to human population density to adjust for any potential sampling bias in RVF occurrences derived from ProMED. A value of 1.0 was selected for the regularization parameter, to correct for overfitting of the model predictions. The Maxent models were trained using five-fold cross-validation. This approach divides the training set into k = 5 subsets, iteratively fits the model to 4-subset combinations, and then tests against the 5^th. Each of the five subsets included approximately 20 RVF outbreaks selected randomly from the total number of available observations in the training dataset (ProMed; n = 103 outbreaks).

Landscape features used in the full Maxent model were ranked according to their permutation importance, which randomly permutes the values of the landscape factors between background and presence points in the training dataset. This is preferred over the direct percent contribution to the loss function because it is non-heuristic and more robust to any residual correlation in assessing the influence of individual features on RVF landscape suitability [37,39].

Finally, as a robust evaluation of prediction error, the trained models were tested against the data obtained from OIE. The difference in model predictions based on the training and testing data was used to assess the model prediction error, which was reported as the area under the curve (AUC). The models were fit using the maxent function (dismo package; v. 0.9–3) setting the distribution to Bernoulli [38,40,41]. All analyses were performed using R statistical software version 3.1.3 [42].

Results

The distributions of RVF outbreaks captured by ProMED and OIE are presented in Fig 1. The clustering of these outbreaks in the Sahel and in eastern and southern African is demonstrated, as is the more recent emergence of RVFV in the Arabian Peninsula.

Download:

Fig 1. The distribution of Rift Valley fever outbreaks across the African continent and the Arabian Peninsula between 1998 and 2016 as reported by ProMED-mail (blue) and between 2005 and 2016 as reported by the World Organization for Animal Health (red).

Countries affected are highlighted in salmon and include: Botswana, Burundi, Egypt, Kenya, Madagascar, Mali, Mauritania, Mayotte (France), Namibia, Niger, Saudi Arabia, Senegal, Somalia, South Africa, Sudan, Tanzania, Uganda, Yemen.

https://doi.org/10.1371/journal.pntd.0005756.g001

All landscape features used in the ecological niche modeling are presented separately for the abiotic (climate, vegetation change, and surface water) and biotic (livestock densities, Bovidae and Muridae species richness, and human population migration) features in Figs 2, 3 and 4.

Download:

Fig 2. Distributions of driest and wettest quarter precipitation (upper panels) and coldest and warmest quarter temperature (lower panels) across Africa and the Arabian Peninsula.

https://doi.org/10.1371/journal.pntd.0005756.g002

Download:

Fig 3. Distributions of surface water (left panel) and vegetation loss (right panel) across Africa and the Arabian Peninsula.

https://doi.org/10.1371/journal.pntd.0005756.g003

Download:

Fig 4. Distributions of livestock (animals per km²), wild Bovidae and Muridae species richness (number of species per km²), and net human migration (number of people into [red] or out of [blue] each km²).

https://doi.org/10.1371/journal.pntd.0005756.g004

The predicted landscape suitability of the African continent and Arabian Peninsula to RVF outbreaks is presented in Fig 5. High risk landscapes were identified in Mauritania extending eastward into the Sahel, as well as in large portions of Sudan, Kenya, Tanzania, South Africa, Madagascar, and a corridor adjacent to the Red Sea in Saudi Arabia and Yemen. Moderate landscape suitability was predicted for northern parts of the Maghreb, the Horn of Africa and the broader Arabian Peninsula.

Download:

Fig 5. Landscape suitability (%) to Rift Valley fever outbreaks.

The risk surface is derived from the ecological niche of RVF outbreaks using the Maxent model.

https://doi.org/10.1371/journal.pntd.0005756.g005

The Maxent model identified proximity to intermittent wetlands (permutation importance = 18%), Bovidae species richness (11.7%), sheep density (11%), dry quarter precipitation (10.2%), and proximity to freshwater marsh (9.1%) as the most influential features to RVF landscape suitability (Fig 6). Muridae species richness was not as influential to suitability as Bovidae richness, but was impactful in the model with 8.6% permutation importance.

Download:

Fig 6. Permutation importance describing the relative importance of each feature to RVF landscape suitability as derived from the Maxent model.

https://doi.org/10.1371/journal.pntd.0005756.g006

Response curves for these features and RVF outbreak risk are presented in S1 Fig. Increasing wild Bovidae species richness was associated with an increase in landscape suitability, as was sheep density up to an approximate average of 100 animals per km², after which it decreased and remained constant at 250 animals per km². Muridae richness demonstrated a V-shaped relationship with high landscape suitability in areas of low and high species richness. Close proximity to intermittent wetlands and freshwater marsh was associated with greater suitability to RVF outbreaks, as was low precipitation during the driest quarter. The model performed well when tested against the OIE data with the AUC equal to 83%.

Discussion

This is the first study to explore and identify the influence of diverse surface water types on RVF outbreaks at a continental scale. The distribution of hydrogeographic features, particularly intermittent wetlands, contributed to suitable landscape conditions for RVF outbreaks. Our model also indicated that Bovidae species richness, sheep density and (to a lesser extent) Muridae species richness were predictive of RVF landscape suitability. This supports field observations that suggest that wild ruminants and rodents are the most likely wild reservoirs for RVFV [14]. These two important landscape features, i.e. wetland hydrogeography and Bovidae species richness, are novel contributions to our understanding of the epidemiology and infection ecology of RVF outbreaks in the African continent and the Arabian Peninsula.

Proximity to intermittent wetlands was particularly important to RVF landscape suitability, as identified by its permutation importance (18%). Hydrogeography has not previously been investigated across the diverse spectrum of surface water types in association with RVF occurrence in continent or country-wide analyses. One comprehensive study did find that proximity to rivers was an important determinant of landscape suitability across the African continent [43]. Our model, which is based on a larger number of epidemic occurrences, identified other wetland features–namely intermittent wetlands and freshwater marsh–as more important contributors to RVF risk. Nonetheless, the two continent-wide studies are in general agreement with respect to high risk landscapes in the Sahel, and eastern and southern Africa.

A regional study in the Ferlo area of Senegal examined proximity to transient ponds in hyperlocal settings of mixed vegetation cover and found that the juxtaposition of these small bodies of water with dense vegetation was strongly associated with positive serology in sheep and goats [44]. Moreover, this same group demonstrated that these landscape features corresponded to fluctuations in Aedes and Culex species during RVF epidemic years [45]. These kinds of transient surface water features were the most influential features to RVF landscape suitability in our model as well. Another study in the Mbeya region of Tanzania, found that proximity to Lake Malawi was very strongly associated with antibody evidence of past infection in humans [46,47], which is consistent with our finding of the importance of fixed freshwater sources such as freshwater marshes.

Our understanding of the landscape epidemiology of RVFV is, perhaps, most deficient with respect to reservoir hosts. Several potential mammalian RVFV hosts have been studied in both field and experimental settings to identify natural reservoirs in RVFV infection ecology [14]. While several species across multiple orders may be implicated as reservoirs, one review identified ruminants and rodents as likely natural reservoir(s) [14]. Evidence from the same study suggests that ruminants are more effective in virus amplification rather than maintenance and this may pose the greatest risk to humans in proximity. Consistent with this, our model demonstrated that wild Bovidae species richness was important in delineating the landscape suitability of RVF (permutation importance 11.7%). Interestingly, this landscape feature was as important as sheep density and more important than density of goats and cattle (see below). Muridae species richness was not as influential as Bovidae species richness, but the former did contribute to RVF landscape suitability (permutation importance = 8.6%). The V-shaped response curve also suggested high risk landscapes associated with low and high Muridae species richness and lower risk landscapes across intermediate species richness. While some previous work has identified the possibility of individual Muridae species as possible maintenance [14] or amplification hosts [48–54], the findings from the current study cannot attribute either role to the Muridae.

The density of sheep was strongly associated with landscape suitability to RVF outbreaks (permutation importance = 11%). This is not surprising given the high susceptibility of sheep [5] and their focus within most of the RVF outbreaks that occurred between 1998 and 2016 [55,56]. Moreover, it is through contact with animals and animal products that most human RVF occurs [13,57,58]. Goat and cattle density did not contribute substantively to RVF landscape suitability(permutation importance < 3 for both). While goats certainly have been affected in many RVF outbreaks, their lesser susceptibility relative to both sheep and cattle has been previously demonstrated [56].

Dry quarter precipitation was a moderately strong contributor to landscape suitability. Low levels of precipitation were associated with increased landscape suitability during the driest period of the year. However, as precipitation increased during this time, RVF suitability sharply decreased (S1 Fig). This may suggest that climatically drier areas are more susceptible when periodically inundated with sporadic episodes of rain [56]. Indeed, barren, arid landscapes punctuated with temporary ponds were associated with the greatest proliferation of both maintenance and amplification vector mosquitoes in West Africa [55,59]. Furthermore, regions with sensitive hydrogeographic dependency demonstrate high RVF susceptibility to precipitation variability attributable to El Nino southern oscillations [18,60]. Indeed, the latter, more recent, study identified a clear sensitivity to both seasonal climatic variation and El Nino oscillation across the whole of the African continent [60]. While the current findings may suggest climatic variations in arid to semi-arid conditions, or extremes of precipitation between dry and wet seasons, we emphasize that our study did not assess the relationship between RVF outbreaks and specific weather events.

This investigation identified landscapes suitable to RVFV beyond the historical extent of transmission. This suggests that conditions may be favorable for transmission should the virus be introduced into these naïve geographic spaces. In essence this allows us to conceive of how a realized niche in one space may be related to a potential niche in another, and what structures may be necessary or useful to prevent the realization of the niche in the novel space. For example, the introduction of RVFV from infected cattle herds to non-infected susceptible herds has been well documented in both local [61] and global [62,63] scales. More specifically, network structure and resulting dissemination of herd animals across villages has been shown to drive the introduction or re-introduction of RVFV to susceptible livestock herds [61]. Translocations of RVFV at more global scales can follow international trade in livestock [62], as well national-level trade in livestock [63]. Surveillance mechanisms should target transboundary livestock movement and mosquito control along livestock migratory corridors for effective prevention of RVF transmission. Indeed, some success has been documented with respect to cross-disciplinary, systems-based approaches to RVFV surveillance in East Africa as organized by the Armed Forces Health Surveillance Center, Division of Global Emerging Infections Surveillance and Response System Operations (AFHSC-GEIS)[64].

A perennial problem in ecological niche modeling, in general, and in emerging disease risk mapping, in particular, is the lack of independent sample data for testing trained models [65]. Typically only one sample is available, which is partitioned into training and testing datasets. A strength of the current study is that we utilized two distinct sources of RVF outbreak data; one (ProMED) to train the landscape suitability model and the other (OIE) to test the prediction error of the model. Nonetheless, the findings from the current study must be interpreted with caution due to its limitations. First, temperature and precipitation were measured as single composites over the period 1950 to 2000. Therefore, while the spatial resolution of these measurements was high (~1 km² rasters), the analysis was simultaneously constrained by coarse temporal resolution, given that the climate rasters represent 50 year averages. Second, the number of documented RVF occurrences is small. Therefore, with only 103 ProMED-reported occurrences and 78 unique OIE-reported occurrences, this collection of RVF outbreaks may not represent the true incidence across the continent of Africa and the Arabian Peninsula. Furthermore, the model may produce an underfit representation of the ecological niche of RVF. Nevertheless, the current analysis attempted a robust validation of models trained on limited data by using a second vetted source of RVF occurrence data from OIE for testing. Third, due to a lack of adequate presence data, the current investigation was unable to evaluate the influence of Aedes and Culex mosquitoes. Because these vectors are part of the RVFV infection cycle, the lack of Aedes and Culex mosquitoes in our model necessarily demarcates a less refined picture of RVF landscape suitability. Nevertheless, the hydrogeographic features identified by the model as influential to RVF suitability correspond to surface water features documented to be most favorable to the Aedes mosquitoes [44,45,66] and Culex mosquitoes [2,59,67] and so the model still appears to identify landscapes appropriate to the relevant vector ecology.

The findings of this study will help to improve our understanding of the landscape epidemiology of RVF outbreaks. The model of RVF landscape suitability will require further development as more data become available to validate the results. Moreover, because this study is observational rather than experimental, we cannot assign causation to the associations between RVF outbreaks and landscape factors, including surface water features and mammalian hosts. More definitive causal inference demands better assessment of species in the landscapes where outbreaks are occurring, and crucially at high spatial resolution and in real time. This will require thorough field studies that combine animal serology and viral RNA detection, more detailed habitat description, observation of wild Bovidae movement patterns, wildlife-domestic animal interaction, and maintenance and amplification vectors, as well as cultural and economic drivers of the livestock industries.

Conclusion

This study found that proximity to wetlands in landscapes that are rich in wild Bovidae species and high in sheep density, delineate the most suitable landscapes for RVF outbreaks. Moreover, this study found that the RVF risk surface may extend to regions beyond the historical range of past zoonotic experience should the virus be introduced to these regions via livestock transport or local invasion by infected mosquitoes.

Supporting information

S1 Fig. Variable response curves for the most influential landscape factors in identifying the landscape suitability of Rift Valley fever outbreaks based on the Maxent model.

https://doi.org/10.1371/journal.pntd.0005756.s001

(PNG)

References

1. Wilson ML. Rift Valley fever virus ecology and the epidemiology of disease emergence. Ann N Y Acad Sci. 1994;740: 169–80. Available: http://www.ncbi.nlm.nih.gov/pubmed/7840448 pmid:7840448
- View Article
- PubMed/NCBI
- Google Scholar
2. Sang R, Kioko E, Lutomiah J, Warigia M, Ochieng C, O’Guinn M, et al. Rift Valley fever virus epidemic in Kenya, 2006/2007: The entomologic investigations. Am J Trop Med Hyg. 2010;83: 28–37.
- View Article
- Google Scholar
3. Madani TA, Al-Mazrou YY, Al-Jeffri MH, Mishkhas AA, Al-Rabeah AM, Turkistani AM, et al. Rift Valley Fever Epidemic in Saudi Arabia: Epidemiological, Clinical, and Laboratory Characteristics. Clin Infect Dis. 2003;37: 1084–1092. pmid:14523773
- View Article
- PubMed/NCBI
- Google Scholar
4. Adam AA, Karsany MS, Adam I. Manifestations of severe Rift Valley fever in Sudan. Int J Infect Dis. 2010;14: 179–180. pmid:19570703
- View Article
- PubMed/NCBI
- Google Scholar
5. Nanyingi MO, Munyua P, Kiama SG, Muchemi GM, Thumbi SM, Bitek AO, et al. A systematic review of Rift Valley Fever epidemiology 1931–2014. Infect Ecol Epidemiol. 2015;5: 1–12. Available: http://www.ncbi.nlm.nih.gov/pubmed/26234531
- View Article
- Google Scholar
6. Sindato C, Karimuribo ED, Mboera LE. The epidemiology and socio-economic impact of rift valley fever epidemics in Tanzania: a review. Tanzania J Heal Res. 2011;13: 305–18.
- View Article
- Google Scholar
7. Linthicum KJ, Anyamba A, Tucker CJ, Kelley P, Myers M, Peters CJ. Climate and Satellite Indicators to Forecast Rift Valley Fever Epidemics in Kenya. Science (80-). 1999;285: 397–400. pmid:10411500
- View Article
- PubMed/NCBI
- Google Scholar
8. Centers for Disease Control and Prevention. Outbreak of Rift Valley fever—Saudi Arabia, August—October, 2000. Morb Mortal Wkly Rep. 2000;49: 905–908.
- View Article
- Google Scholar
9. Linthicum KJ, Davies FG, Kairo a, Bailey CL. Rift Valley fever virus (family Bunyaviridae, genus Phlebovirus). Isolations from Diptera collected during an inter-epizootic period in Kenya. J Hyg (Lond). 1985;95: 197–209.
- View Article
- Google Scholar
10. Linthicum K, Bailey C, Davies F, Kairo A, Logan T. The horizontal distribution of Aedes pupae and their subsequent adults within a flooded dambo in Kenya: implications for Rift Valley fever virus control. J Am Mosq Control Assoc. 1988;4: 551–554. pmid:3225577
- View Article
- PubMed/NCBI
- Google Scholar
11. Bird BH, McElroy AK. Rift Valley fever virus: Unanswered questions. Antiviral Res. Elsevier B.V; 2016;132: 274–280. pmid:27400990
- View Article
- PubMed/NCBI
- Google Scholar
12. Turell MJ, Linthicum KJ, Patrican L a, Davies FG, Kairo A, Bailey CL. Vector competence of selected African mosquito (Diptera: Culicidae) species for Rift Valley fever virus. J Med Entomol. 2008;45: 102–108. pmid:18283949
- View Article
- PubMed/NCBI
- Google Scholar
13. Sumaye RD, Abatih EN, Thiry E, Amuri M, Berkvens D, Geubbels E. Inter-epidemic acquisition of Rift Valley fever virus in humans in Tanzania. PLoS Negl Trop Dis. 2015;9: e0003536. pmid:25723502
- View Article
- PubMed/NCBI
- Google Scholar
14. Olive M-M, Goodman SM, Reynes J-M. The role of wild mammals in the maintenance of Rift Valley fever virus. J Wildl Dis. 2012;48: 241–266. pmid:22493102
- View Article
- PubMed/NCBI
- Google Scholar
15. Anyamba A, Linthicum KJ, Small J, Britch SC, Pak E, De La Rocque S, et al. Prediction, assessment of the Rift Valley fever activity in east and southern Africa 2006–2008 and possible vector control strategies. Am J Trop Med Hyg. 2010;83: 43–51. pmid:20682905
- View Article
- PubMed/NCBI
- Google Scholar
16. Anyamba A, Linthicum KJ, Mahoney R, Tucker CJ, Kelley P. Mapping Potential Risk of Rift Valley Fever Outbreaks in African Savannas Using Vegetation Index Time Series Data. Photogramm Eng Remote Sensing. 2002;68: 137–145.
- View Article
- Google Scholar
17. Anyamba A, Linthicum KJ, Tucker CJ. Climate-disease connections: Rift Valley fever in Kenya. Cad Saude Publica. 2001;17: S133–S140.
- View Article
- Google Scholar
18. Anyamba A, Linthicum KJ, Small JL, Collins KM, Tucker CJ, Pak EW, et al. Climate teleconnections and recent patterns of human and animal disease outbreaks. PLoS Negl Trop Dis. 2012;6: e1465. pmid:22292093
- View Article
- PubMed/NCBI
- Google Scholar
19. Anyamba A, Small JL, Britch SC, Tucker CJ, Pak EW, Reynolds CA, et al. Recent weather extremes and impacts on agricultural production and vector-borne disease outbreak patterns. PLoS One. 2014;9: e92538. pmid:24658301
- View Article
- PubMed/NCBI
- Google Scholar
20. Lambin EF, Tran A, Vanwambeke SO, Linard C, Soti V. Pathogenic landscapes: interactions between land, people, disease vectors, and their animal hosts. Int J Health Geogr. BioMed Central Ltd; 2010;9: 54. pmid:20979609
- View Article
- PubMed/NCBI
- Google Scholar
21. International Society for Infectious Diseases. ProMED-mail. Available: http://www.promedmail.org/
22. Health WO for A. World Animal Health Information System [Internet]. Available: http://www.oie.int/wahis_2/public/wahid.php/Diseaseinformation/Immsummary
23. WorldClim—Global Climate. Data for current conditions (~1950–2000) | WorldClim—Global Climate Data. In: WorldClim—Global Climate Data [Internet]. [cited 23 Oct 2014]. Available: http://www.worldclim.org/current
24. Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A. Very high resolution interpolated climate surfaces for global land areas. Int J Climatol. 2005;25: 1965–1978.
- View Article
- Google Scholar
25. The USGS Land Cover Institute. MODIS-based Maximum Green Vegetation Fraction [Internet]. [cited 23 Oct 2014]. Available: http://landcover.usgs.gov/green_veg.php
26. Broxton PD, Zeng X, Scheftic W, Troch PA. A MODIS-Based Global 1-km Maximum Green Vegetation Fraction Dataset. 2014; Available: http://journals.ametsoc.org/doi/abs/10.1175/JAMC-D-13-0356.1
27. Fund WW. Global Lakes and Wetlands Database [Internet]. Available: http://www.worldwildlife.org/pages/global-lakes-and-wetlands-database
28. Lehner B, Döll P. Development and validation of a global database of lakes, reservoirs and wetlands. J Hydrol. 2004;296: 1–22.
- View Article
- Google Scholar
29. QGIS Development Team. QGIS Geographic Information System [Internet]. Open Source Geospatial Foundation; 2009. Available: http://www.qgis.org/
30. Socioeconomic Data and Applications Center | SEDAC. Data Sets | SEDAC [Internet]. [cited 23 Dec 2014]. Available: http://sedac.ciesin.columbia.edu/data/sets/browse
31. De Sherbinin A, Levy M, Adamo S, MacManus K, Yetman G, Mara V, et al. Migration and risk: net migration in marginal ecosystems and hazardous areas. Environ Res Lett. 2012;7: 045602.
- View Article
- Google Scholar
32. Robinson TP, Wint GRW, Conchedda G, Van Boeckel TP, Ercoli V, Palamara E, et al. Mapping the global distribution of livestock. PLoS One. 2014;9: e96084. pmid:24875496
- View Article
- PubMed/NCBI
- Google Scholar
33. Robinson T, Thornton P, Franceschini G, Kruska R, Chiozza F, Notenbaert A, et al. Global livestock production systems. Rome—Food and Agriculture Organization of the United Nations (FAO) and International Livestock Research Institute (ILRI).pdf. 2011.
34. Socioeconomic Data and Applications Center | SEDAC. Global Mammal Richness Grids [Internet]. Available: http://sedac.ciesin.columbia.edu/data/set/species-global-mammal-richness-2015
35. Facility GBI. Global Biodiversity Information Facility [Internet]. Available: http://www.gbif.org/
36. Elith J, Leathwick JR, Hastie T. A working guide to boosted regression trees. J Anim Ecol. 2008;77: 802–13. pmid:18397250
- View Article
- PubMed/NCBI
- Google Scholar
37. Phillips SJ, Anderson RP, Schapire RE. Maximum entropy modeling of species geographic distributions. Ecol Modell. 2006;190: 231–259.
- View Article
- Google Scholar
38. Franklin J. Mapping Species Distributions: Spatial Inference and Prediction [Internet]. First. Cambridge: Cambridge University Press; 2010. Available: https://books.google.com/books?id=CkshAwAAQBAJ&pgis=1
39. Phillips SJ, Dudík M. Modeling of species distribution with Maxent: new extensions and a comprehensive evalutation. Ecograpy. 2008;31: 161–175.
- View Article
- Google Scholar
40. Hijmans RJ, Phillips S, Leathwick JR, Elith J. Package “dismo.” In: The Comprehensive R Archive Network [Internet]. 2014 [cited 15 Jan 2015] pp. 1–65. Available: http://cran.r-project.org/web/packages/dismo/dismo.pdf
41. Fielding AH, Bell JF. A review of methods for the assessment of prediction errors in conservation presence/absence models. Environ Conserv. 1997;24: 38–49. Available: http://climate.calcommons.org/bib/review-methods-assessment-prediction-errors-conservation-presenceabsence-models
- View Article
- Google Scholar
42. Team RC. R: A language and environment for statistical computing [Internet]. Vienna: R Foundation for Statistical Computing; 2016. Available: https://www.r-project.org/
43. Clements ACA, Pfeiffer DU, Martin V. Application of knowledge-driven spatial modelling approaches and uncertainty management to a study of Rift Valley fever in Africa. Int J Health Geogr. 2006;5: 57. pmid:17156467
- View Article
- PubMed/NCBI
- Google Scholar
44. Soti V, Chevalier V, Maura J, Bégué A, Lelong C, Lancelot R, et al. Identifying landscape features associated with Rift Valley fever virus transmission, Ferlo region, Senegal, using very high spatial resolution satellite imagery. Int J Health Geogr. 2013;12: 10. pmid:23452759
- View Article
- PubMed/NCBI
- Google Scholar
45. Soti V, Tran A, Degenne P, Chevalier V, Lo Seen D, Thiongane Y, et al. Combining Hydrology and Mosquito Population Models to Identify the Drivers of Rift Valley Fever Emergence in Semi-Arid Regions of West Africa. PLoS Negl Trop Dis. 2012;6. pmid:22928055
- View Article
- PubMed/NCBI
- Google Scholar
46. Heinrich N, Saathoff E, Weller N, Clowes P, Kroidl I, Ntinginya E, et al. High seroprevalence of Rift Valley fever and evidence for endemic circulation in Mbeya region, Tanzania, in a cross-sectional study. PLoS Negl Trop Dis. 2012;6. pmid:22479657
- View Article
- PubMed/NCBI
- Google Scholar
47. Clements AC a, Pfeiffer DU, Martin V, Pittliglio C, Best N, Thiongane Y. Spatial risk assessment of Rift Valley fever in Senegal. Vector Borne Zoonotic Dis. 2007;7: 203–16. pmid:17627440
- View Article
- PubMed/NCBI
- Google Scholar
48. Hoogstraal H, Meegan JM, Khalil GM, Adham FK. The rift valley fever epizootic in Egypt 1977–78. 2. Ecological and entomological studies. Trans R Soc Trop Med Hyg. 1979;73: 624–629. pmid:44038
- View Article
- PubMed/NCBI
- Google Scholar
49. Diop G, Thiongane Y, Thonnon J, Fontenille D, Diallo M, Sall S, et al. The potential role of rodents in the enzootic cycle of Rift Valley fever virus in Senegal. Microbes Infect. 2000;2: 343–346. pmid:10817634
- View Article
- PubMed/NCBI
- Google Scholar
50. Zeller HG, Fontenille D, Traore-lamizana M, Thiongane Y, Digoutte J. Enzootic activity of Rift Valley fever virus in Senegal. Am J Trop Med Hyg. 1997;56: 265–272. pmid:9129528
- View Article
- PubMed/NCBI
- Google Scholar
51. Kark JD, Aynor Y, Mordechai Y Ben, Ron-Kuper N, Peleg B a., Peters CJ. A serological survey of Rift Valley fever antibodies in the northern Sinai. Trans R Soc Trop Med Hyg. 1982;76: 427–430. pmid:6926758
- View Article
- PubMed/NCBI
- Google Scholar
52. Saluzzo J, Chartier C, Bada R, Martinez D, Digoutte J. Rift Valley fever in Western Africa. Rev Elev Med Vet Pays Trop. 1987;40: 215–223. pmid:3453507
- View Article
- PubMed/NCBI
- Google Scholar
53. Pretorius a, Oelofsen MJ, Smith MS, van der Ryst E. Rift Valley fever virus: a seroepidemiologic study of small terrestrial vertebrates in South Africa. Am J Trop Med Hyg. 1997;57: 693–698. Available: http://mutex.gmu.edu/login?url=http://search.ebscohost.com/login.aspx?direct=true&db=mnh&AN=9430529&site=ehost-live&scope=site pmid:9430529
- View Article
- PubMed/NCBI
- Google Scholar
54. Swanepoel R, Coetzer J. Rift Valley fever. Infectious diseases of livestock. 2004. pp. 1037–1070.
55. Glancey MM, Anyamba A, Linthicum KJ. Epidemiologic and Environmental Risk Factors of Rift Valley Fever in Southern Africa from 2008 to 2011. Vector Borne Zoonotic Dis. 2015;15: 502–511. pmid:26273812
- View Article
- PubMed/NCBI
- Google Scholar
56. Sindato C, Pfeiffer DU, Karimuribo ED, Mboera LEG, Rweyemamu MM, Paweska JT. A Spatial Analysis of Rift Valley Fever Virus Seropositivity in Domestic Ruminants in Tanzania. PLoS One. 2015;10: e0131873. pmid:26162089
- View Article
- PubMed/NCBI
- Google Scholar
57. LaBeaud a. D, Muchiri EM, Ndzovu M, Mwanje MT, Muiruri S, Peters CJ, et al. Interepidemic Rift Valley fever virus seropositivity, northeastern Kenya. Emerg Infect Dis. 2008;14: 1240–1246. pmid:18680647
- View Article
- PubMed/NCBI
- Google Scholar
58. Anyangu AS, Gould LH, Sharif SK, Nguku PM, Omolo JO, Mutonga D, et al. Risk factors for severe rift valley fever infection in Kenya, 2007. Am J Trop Med Hyg. 2010;83: 14–21. pmid:20682901
- View Article
- PubMed/NCBI
- Google Scholar
59. Diallo D, Talla C, Ba Y, Dia I, Sall AA, Diallo M. Temporal Distribution and Spatial Pattern of Abundance of the Rift Valley Fever and West Nile Fever Vectors in Barkedji, Senegal Temporal distribution and spatial pattern of abundance of the Rift Valley fever and West Nile fever vectors in Barkedji, Sen. J Vector Ecol. 2011;36: 426–436. pmid:22129415
- View Article
- PubMed/NCBI
- Google Scholar
60. Redding DW, Lo Iacono G, Bett B, Tiedt S, Jones K.E. Spatial, seasonal and climatic predictive models of Rift Valley Fever disease across Africa. Philos Trans R Soc B Biol Sci. 2017;Accepted. pmid:28584173
- View Article
- PubMed/NCBI
- Google Scholar
61. Nicolas G, Durand B, Duboz R, Rakotondravao R, Chevalier V. Description and analysis of the cattle trade network in the Madagascar highlands: Potential role in the diffusion of Rift Valley fever virus. Acta Trop. Elsevier B.V.; 2013;126: 19–27. pmid:23295911
- View Article
- PubMed/NCBI
- Google Scholar
62. Abdo-Salem S, Waret-Szkuta A, Roger F, Olive MM, Saeed K, Chevalier V. Risk assessment of the introduction of Rift Valley fever from the Horn of Africa to Yemen via legal trade of small ruminants. Trop Anim Health Prod. 2011;43: 471–480. pmid:20967567
- View Article
- PubMed/NCBI
- Google Scholar
63. Xiao Y, Beier JC, Cantrell RS, Cosner C, DeAngelis DL, Ruan S. Modelling the effects of seasonality and socioeconomic impact on the transmission of rift valley Fever virus. PLoS Negl Trop Dis. 2015;9: e3388. pmid:25569474
- View Article
- PubMed/NCBI
- Google Scholar
64. Witt CJ, Richards AL, Masuoka PM, Foley DH, Buczak AL, Musila LA, et al. The AFHSC-Division of GEIS Operations Predictive Surveillance Program: a multidisciplinary approach for the early detection and response to disease outbreaks. Spec Issue Dep Def Glob Emerg Infect Surveill Response Syst an Updat 2009. 2011;11: S10. Available: http://www.biomedcentral.com/1471-2458/11/S2/S10
- View Article
- Google Scholar
65. Peterson AT. Mapping Risk of Nipah Virus Transmission Across Asia and Across Bangladesh. Asia Pac J Public Health. 2013; pmid:23343646
- View Article
- PubMed/NCBI
- Google Scholar
66. Himeidan YE, Kweka EJ, Mahgoub MM, El Rayah EA, Ouma JO. Recent Outbreaks of Rift Valley Fever in East Africa and the Middle East. Front Public Heal. Frontiers; 2014;2: 169. pmid:25340047
- View Article
- PubMed/NCBI
- Google Scholar
67. Talla C, Diallo D, Dia I, Ba Y, Ndione JA, Sall AA, et al. Statistical modeling of the abundance of vectors of West African Rift Valley Fever in Bark??dji, Senegal. PLoS One. 2014;9: 1–23. pmid:25437856
- View Article
- PubMed/NCBI
- Google Scholar

[ref1] 1. Wilson ML. Rift Valley fever virus ecology and the epidemiology of disease emergence. Ann N Y Acad Sci. 1994;740: 169–80. Available: http://www.ncbi.nlm.nih.gov/pubmed/7840448 pmid:7840448
View Article
PubMed/NCBI
Google Scholar

[2] View Article

[3] PubMed/NCBI

[4] Google Scholar

[ref2] 2. Sang R, Kioko E, Lutomiah J, Warigia M, Ochieng C, O’Guinn M, et al. Rift Valley fever virus epidemic in Kenya, 2006/2007: The entomologic investigations. Am J Trop Med Hyg. 2010;83: 28–37.
View Article
Google Scholar

[6] View Article

[7] Google Scholar

[ref3] 3. Madani TA, Al-Mazrou YY, Al-Jeffri MH, Mishkhas AA, Al-Rabeah AM, Turkistani AM, et al. Rift Valley Fever Epidemic in Saudi Arabia: Epidemiological, Clinical, and Laboratory Characteristics. Clin Infect Dis. 2003;37: 1084–1092. pmid:14523773
View Article
PubMed/NCBI
Google Scholar

[9] View Article

[10] PubMed/NCBI

[11] Google Scholar

[ref4] 4. Adam AA, Karsany MS, Adam I. Manifestations of severe Rift Valley fever in Sudan. Int J Infect Dis. 2010;14: 179–180. pmid:19570703
View Article
PubMed/NCBI
Google Scholar

[13] View Article

[14] PubMed/NCBI

[15] Google Scholar

[ref5] 5. Nanyingi MO, Munyua P, Kiama SG, Muchemi GM, Thumbi SM, Bitek AO, et al. A systematic review of Rift Valley Fever epidemiology 1931–2014. Infect Ecol Epidemiol. 2015;5: 1–12. Available: http://www.ncbi.nlm.nih.gov/pubmed/26234531
View Article
Google Scholar

[17] View Article

[18] Google Scholar

[ref6] 6. Sindato C, Karimuribo ED, Mboera LE. The epidemiology and socio-economic impact of rift valley fever epidemics in Tanzania: a review. Tanzania J Heal Res. 2011;13: 305–18.
View Article
Google Scholar

[20] View Article

[21] Google Scholar

[ref7] 7. Linthicum KJ, Anyamba A, Tucker CJ, Kelley P, Myers M, Peters CJ. Climate and Satellite Indicators to Forecast Rift Valley Fever Epidemics in Kenya. Science (80-). 1999;285: 397–400. pmid:10411500
View Article
PubMed/NCBI
Google Scholar

[23] View Article

[24] PubMed/NCBI

[25] Google Scholar

[ref8] 8. Centers for Disease Control and Prevention. Outbreak of Rift Valley fever—Saudi Arabia, August—October, 2000. Morb Mortal Wkly Rep. 2000;49: 905–908.
View Article
Google Scholar

[27] View Article

[28] Google Scholar

[ref9] 9. Linthicum KJ, Davies FG, Kairo a, Bailey CL. Rift Valley fever virus (family Bunyaviridae, genus Phlebovirus). Isolations from Diptera collected during an inter-epizootic period in Kenya. J Hyg (Lond). 1985;95: 197–209.
View Article
Google Scholar

[30] View Article

[31] Google Scholar

[ref10] 10. Linthicum K, Bailey C, Davies F, Kairo A, Logan T. The horizontal distribution of Aedes pupae and their subsequent adults within a flooded dambo in Kenya: implications for Rift Valley fever virus control. J Am Mosq Control Assoc. 1988;4: 551–554. pmid:3225577
View Article
PubMed/NCBI
Google Scholar

[33] View Article

[34] PubMed/NCBI

[35] Google Scholar

[ref11] 11. Bird BH, McElroy AK. Rift Valley fever virus: Unanswered questions. Antiviral Res. Elsevier B.V; 2016;132: 274–280. pmid:27400990
View Article
PubMed/NCBI
Google Scholar

[37] View Article

[38] PubMed/NCBI

[39] Google Scholar

[ref12] 12. Turell MJ, Linthicum KJ, Patrican L a, Davies FG, Kairo A, Bailey CL. Vector competence of selected African mosquito (Diptera: Culicidae) species for Rift Valley fever virus. J Med Entomol. 2008;45: 102–108. pmid:18283949
View Article
PubMed/NCBI
Google Scholar

[41] View Article

[42] PubMed/NCBI

[43] Google Scholar

[ref13] 13. Sumaye RD, Abatih EN, Thiry E, Amuri M, Berkvens D, Geubbels E. Inter-epidemic acquisition of Rift Valley fever virus in humans in Tanzania. PLoS Negl Trop Dis. 2015;9: e0003536. pmid:25723502
View Article
PubMed/NCBI
Google Scholar

[45] View Article

[46] PubMed/NCBI

[47] Google Scholar

[ref14] 14. Olive M-M, Goodman SM, Reynes J-M. The role of wild mammals in the maintenance of Rift Valley fever virus. J Wildl Dis. 2012;48: 241–266. pmid:22493102
View Article
PubMed/NCBI
Google Scholar

[49] View Article

[50] PubMed/NCBI

[51] Google Scholar

[ref15] 15. Anyamba A, Linthicum KJ, Small J, Britch SC, Pak E, De La Rocque S, et al. Prediction, assessment of the Rift Valley fever activity in east and southern Africa 2006–2008 and possible vector control strategies. Am J Trop Med Hyg. 2010;83: 43–51. pmid:20682905
View Article
PubMed/NCBI
Google Scholar

[53] View Article

[54] PubMed/NCBI

[55] Google Scholar

[ref16] 16. Anyamba A, Linthicum KJ, Mahoney R, Tucker CJ, Kelley P. Mapping Potential Risk of Rift Valley Fever Outbreaks in African Savannas Using Vegetation Index Time Series Data. Photogramm Eng Remote Sensing. 2002;68: 137–145.
View Article
Google Scholar

[57] View Article

[58] Google Scholar

[ref17] 17. Anyamba A, Linthicum KJ, Tucker CJ. Climate-disease connections: Rift Valley fever in Kenya. Cad Saude Publica. 2001;17: S133–S140.
View Article
Google Scholar

[60] View Article

[61] Google Scholar

[ref18] 18. Anyamba A, Linthicum KJ, Small JL, Collins KM, Tucker CJ, Pak EW, et al. Climate teleconnections and recent patterns of human and animal disease outbreaks. PLoS Negl Trop Dis. 2012;6: e1465. pmid:22292093
View Article
PubMed/NCBI
Google Scholar

[63] View Article

[64] PubMed/NCBI

[65] Google Scholar

[ref19] 19. Anyamba A, Small JL, Britch SC, Tucker CJ, Pak EW, Reynolds CA, et al. Recent weather extremes and impacts on agricultural production and vector-borne disease outbreak patterns. PLoS One. 2014;9: e92538. pmid:24658301
View Article
PubMed/NCBI
Google Scholar

[67] View Article

[68] PubMed/NCBI

[69] Google Scholar

[ref20] 20. Lambin EF, Tran A, Vanwambeke SO, Linard C, Soti V. Pathogenic landscapes: interactions between land, people, disease vectors, and their animal hosts. Int J Health Geogr. BioMed Central Ltd; 2010;9: 54. pmid:20979609
View Article
PubMed/NCBI
Google Scholar

[71] View Article

[72] PubMed/NCBI

[73] Google Scholar

[ref21] 21. International Society for Infectious Diseases. ProMED-mail. Available: http://www.promedmail.org/

[ref22] 22. Health WO for A. World Animal Health Information System [Internet]. Available: http://www.oie.int/wahis_2/public/wahid.php/Diseaseinformation/Immsummary

[ref23] 23. WorldClim—Global Climate. Data for current conditions (~1950–2000) | WorldClim—Global Climate Data. In: WorldClim—Global Climate Data [Internet]. [cited 23 Oct 2014]. Available: http://www.worldclim.org/current

[ref24] 24. Hijmans RJ, Cameron SE, Parra JL, Jones PG, Jarvis A. Very high resolution interpolated climate surfaces for global land areas. Int J Climatol. 2005;25: 1965–1978.
View Article
Google Scholar

[78] View Article

[79] Google Scholar

[ref25] 25. The USGS Land Cover Institute. MODIS-based Maximum Green Vegetation Fraction [Internet]. [cited 23 Oct 2014]. Available: http://landcover.usgs.gov/green_veg.php

[ref26] 26. Broxton PD, Zeng X, Scheftic W, Troch PA. A MODIS-Based Global 1-km Maximum Green Vegetation Fraction Dataset. 2014; Available: http://journals.ametsoc.org/doi/abs/10.1175/JAMC-D-13-0356.1

[ref27] 27. Fund WW. Global Lakes and Wetlands Database [Internet]. Available: http://www.worldwildlife.org/pages/global-lakes-and-wetlands-database

[ref28] 28. Lehner B, Döll P. Development and validation of a global database of lakes, reservoirs and wetlands. J Hydrol. 2004;296: 1–22.
View Article
Google Scholar

[84] View Article

[85] Google Scholar

[ref29] 29. QGIS Development Team. QGIS Geographic Information System [Internet]. Open Source Geospatial Foundation; 2009. Available: http://www.qgis.org/

[ref30] 30. Socioeconomic Data and Applications Center | SEDAC. Data Sets | SEDAC [Internet]. [cited 23 Dec 2014]. Available: http://sedac.ciesin.columbia.edu/data/sets/browse

[ref31] 31. De Sherbinin A, Levy M, Adamo S, MacManus K, Yetman G, Mara V, et al. Migration and risk: net migration in marginal ecosystems and hazardous areas. Environ Res Lett. 2012;7: 045602.
View Article
Google Scholar

[89] View Article

[90] Google Scholar

[ref32] 32. Robinson TP, Wint GRW, Conchedda G, Van Boeckel TP, Ercoli V, Palamara E, et al. Mapping the global distribution of livestock. PLoS One. 2014;9: e96084. pmid:24875496
View Article
PubMed/NCBI
Google Scholar

[92] View Article

[93] PubMed/NCBI

[94] Google Scholar

[ref33] 33. Robinson T, Thornton P, Franceschini G, Kruska R, Chiozza F, Notenbaert A, et al. Global livestock production systems. Rome—Food and Agriculture Organization of the United Nations (FAO) and International Livestock Research Institute (ILRI).pdf. 2011.

[ref34] 34. Socioeconomic Data and Applications Center | SEDAC. Global Mammal Richness Grids [Internet]. Available: http://sedac.ciesin.columbia.edu/data/set/species-global-mammal-richness-2015

[ref35] 35. Facility GBI. Global Biodiversity Information Facility [Internet]. Available: http://www.gbif.org/

[ref36] 36. Elith J, Leathwick JR, Hastie T. A working guide to boosted regression trees. J Anim Ecol. 2008;77: 802–13. pmid:18397250
View Article
PubMed/NCBI
Google Scholar

[99] View Article

[100] PubMed/NCBI

[101] Google Scholar

[ref37] 37. Phillips SJ, Anderson RP, Schapire RE. Maximum entropy modeling of species geographic distributions. Ecol Modell. 2006;190: 231–259.
View Article
Google Scholar

[103] View Article

[104] Google Scholar

[ref38] 38. Franklin J. Mapping Species Distributions: Spatial Inference and Prediction [Internet]. First. Cambridge: Cambridge University Press; 2010. Available: https://books.google.com/books?id=CkshAwAAQBAJ&pgis=1

[ref39] 39. Phillips SJ, Dudík M. Modeling of species distribution with Maxent: new extensions and a comprehensive evalutation. Ecograpy. 2008;31: 161–175.
View Article
Google Scholar

[107] View Article

[108] Google Scholar

[ref40] 40. Hijmans RJ, Phillips S, Leathwick JR, Elith J. Package “dismo.” In: The Comprehensive R Archive Network [Internet]. 2014 [cited 15 Jan 2015] pp. 1–65. Available: http://cran.r-project.org/web/packages/dismo/dismo.pdf

[ref41] 41. Fielding AH, Bell JF. A review of methods for the assessment of prediction errors in conservation presence/absence models. Environ Conserv. 1997;24: 38–49. Available: http://climate.calcommons.org/bib/review-methods-assessment-prediction-errors-conservation-presenceabsence-models
View Article
Google Scholar

[111] View Article

[112] Google Scholar

[ref42] 42. Team RC. R: A language and environment for statistical computing [Internet]. Vienna: R Foundation for Statistical Computing; 2016. Available: https://www.r-project.org/

[ref43] 43. Clements ACA, Pfeiffer DU, Martin V. Application of knowledge-driven spatial modelling approaches and uncertainty management to a study of Rift Valley fever in Africa. Int J Health Geogr. 2006;5: 57. pmid:17156467
View Article
PubMed/NCBI
Google Scholar

[115] View Article

[116] PubMed/NCBI

[117] Google Scholar

[ref44] 44. Soti V, Chevalier V, Maura J, Bégué A, Lelong C, Lancelot R, et al. Identifying landscape features associated with Rift Valley fever virus transmission, Ferlo region, Senegal, using very high spatial resolution satellite imagery. Int J Health Geogr. 2013;12: 10. pmid:23452759
View Article
PubMed/NCBI
Google Scholar

[119] View Article

[120] PubMed/NCBI

[121] Google Scholar

[ref45] 45. Soti V, Tran A, Degenne P, Chevalier V, Lo Seen D, Thiongane Y, et al. Combining Hydrology and Mosquito Population Models to Identify the Drivers of Rift Valley Fever Emergence in Semi-Arid Regions of West Africa. PLoS Negl Trop Dis. 2012;6. pmid:22928055
View Article
PubMed/NCBI
Google Scholar

[123] View Article

[124] PubMed/NCBI

[125] Google Scholar

[ref46] 46. Heinrich N, Saathoff E, Weller N, Clowes P, Kroidl I, Ntinginya E, et al. High seroprevalence of Rift Valley fever and evidence for endemic circulation in Mbeya region, Tanzania, in a cross-sectional study. PLoS Negl Trop Dis. 2012;6. pmid:22479657
View Article
PubMed/NCBI
Google Scholar

[127] View Article

[128] PubMed/NCBI

[129] Google Scholar

[ref47] 47. Clements AC a, Pfeiffer DU, Martin V, Pittliglio C, Best N, Thiongane Y. Spatial risk assessment of Rift Valley fever in Senegal. Vector Borne Zoonotic Dis. 2007;7: 203–16. pmid:17627440
View Article
PubMed/NCBI
Google Scholar

[131] View Article

[132] PubMed/NCBI

[133] Google Scholar

[ref48] 48. Hoogstraal H, Meegan JM, Khalil GM, Adham FK. The rift valley fever epizootic in Egypt 1977–78. 2. Ecological and entomological studies. Trans R Soc Trop Med Hyg. 1979;73: 624–629. pmid:44038
View Article
PubMed/NCBI
Google Scholar

[135] View Article

[136] PubMed/NCBI

[137] Google Scholar

[ref49] 49. Diop G, Thiongane Y, Thonnon J, Fontenille D, Diallo M, Sall S, et al. The potential role of rodents in the enzootic cycle of Rift Valley fever virus in Senegal. Microbes Infect. 2000;2: 343–346. pmid:10817634
View Article
PubMed/NCBI
Google Scholar

[139] View Article

[140] PubMed/NCBI

[141] Google Scholar

[ref50] 50. Zeller HG, Fontenille D, Traore-lamizana M, Thiongane Y, Digoutte J. Enzootic activity of Rift Valley fever virus in Senegal. Am J Trop Med Hyg. 1997;56: 265–272. pmid:9129528
View Article
PubMed/NCBI
Google Scholar

[143] View Article

[144] PubMed/NCBI

[145] Google Scholar

[ref51] 51. Kark JD, Aynor Y, Mordechai Y Ben, Ron-Kuper N, Peleg B a., Peters CJ. A serological survey of Rift Valley fever antibodies in the northern Sinai. Trans R Soc Trop Med Hyg. 1982;76: 427–430. pmid:6926758
View Article
PubMed/NCBI
Google Scholar

[147] View Article

[148] PubMed/NCBI

[149] Google Scholar

[ref52] 52. Saluzzo J, Chartier C, Bada R, Martinez D, Digoutte J. Rift Valley fever in Western Africa. Rev Elev Med Vet Pays Trop. 1987;40: 215–223. pmid:3453507
View Article
PubMed/NCBI
Google Scholar

[151] View Article

[152] PubMed/NCBI

[153] Google Scholar

[ref53] 53. Pretorius a, Oelofsen MJ, Smith MS, van der Ryst E. Rift Valley fever virus: a seroepidemiologic study of small terrestrial vertebrates in South Africa. Am J Trop Med Hyg. 1997;57: 693–698. Available: http://mutex.gmu.edu/login?url=http://search.ebscohost.com/login.aspx?direct=true&db=mnh&AN=9430529&site=ehost-live&scope=site pmid:9430529
View Article
PubMed/NCBI
Google Scholar

[155] View Article

[156] PubMed/NCBI

[157] Google Scholar

[ref54] 54. Swanepoel R, Coetzer J. Rift Valley fever. Infectious diseases of livestock. 2004. pp. 1037–1070.

[ref55] 55. Glancey MM, Anyamba A, Linthicum KJ. Epidemiologic and Environmental Risk Factors of Rift Valley Fever in Southern Africa from 2008 to 2011. Vector Borne Zoonotic Dis. 2015;15: 502–511. pmid:26273812
View Article
PubMed/NCBI
Google Scholar

[160] View Article

[161] PubMed/NCBI

[162] Google Scholar

[ref56] 56. Sindato C, Pfeiffer DU, Karimuribo ED, Mboera LEG, Rweyemamu MM, Paweska JT. A Spatial Analysis of Rift Valley Fever Virus Seropositivity in Domestic Ruminants in Tanzania. PLoS One. 2015;10: e0131873. pmid:26162089
View Article
PubMed/NCBI
Google Scholar

[164] View Article

[165] PubMed/NCBI

[166] Google Scholar

[ref57] 57. LaBeaud a. D, Muchiri EM, Ndzovu M, Mwanje MT, Muiruri S, Peters CJ, et al. Interepidemic Rift Valley fever virus seropositivity, northeastern Kenya. Emerg Infect Dis. 2008;14: 1240–1246. pmid:18680647
View Article
PubMed/NCBI
Google Scholar

[168] View Article

[169] PubMed/NCBI

[170] Google Scholar

[ref58] 58. Anyangu AS, Gould LH, Sharif SK, Nguku PM, Omolo JO, Mutonga D, et al. Risk factors for severe rift valley fever infection in Kenya, 2007. Am J Trop Med Hyg. 2010;83: 14–21. pmid:20682901
View Article
PubMed/NCBI
Google Scholar

[172] View Article

[173] PubMed/NCBI

[174] Google Scholar

[ref59] 59. Diallo D, Talla C, Ba Y, Dia I, Sall AA, Diallo M. Temporal Distribution and Spatial Pattern of Abundance of the Rift Valley Fever and West Nile Fever Vectors in Barkedji, Senegal Temporal distribution and spatial pattern of abundance of the Rift Valley fever and West Nile fever vectors in Barkedji, Sen. J Vector Ecol. 2011;36: 426–436. pmid:22129415
View Article
PubMed/NCBI
Google Scholar

[176] View Article

[177] PubMed/NCBI

[178] Google Scholar

[ref60] 60. Redding DW, Lo Iacono G, Bett B, Tiedt S, Jones K.E. Spatial, seasonal and climatic predictive models of Rift Valley Fever disease across Africa. Philos Trans R Soc B Biol Sci. 2017;Accepted. pmid:28584173
View Article
PubMed/NCBI
Google Scholar

[180] View Article

[181] PubMed/NCBI

[182] Google Scholar

[ref61] 61. Nicolas G, Durand B, Duboz R, Rakotondravao R, Chevalier V. Description and analysis of the cattle trade network in the Madagascar highlands: Potential role in the diffusion of Rift Valley fever virus. Acta Trop. Elsevier B.V.; 2013;126: 19–27. pmid:23295911
View Article
PubMed/NCBI
Google Scholar

[184] View Article

[185] PubMed/NCBI

[186] Google Scholar

[ref62] 62. Abdo-Salem S, Waret-Szkuta A, Roger F, Olive MM, Saeed K, Chevalier V. Risk assessment of the introduction of Rift Valley fever from the Horn of Africa to Yemen via legal trade of small ruminants. Trop Anim Health Prod. 2011;43: 471–480. pmid:20967567
View Article
PubMed/NCBI
Google Scholar

[188] View Article

[189] PubMed/NCBI

[190] Google Scholar

[ref63] 63. Xiao Y, Beier JC, Cantrell RS, Cosner C, DeAngelis DL, Ruan S. Modelling the effects of seasonality and socioeconomic impact on the transmission of rift valley Fever virus. PLoS Negl Trop Dis. 2015;9: e3388. pmid:25569474
View Article
PubMed/NCBI
Google Scholar

[192] View Article

[193] PubMed/NCBI

[194] Google Scholar

[ref64] 64. Witt CJ, Richards AL, Masuoka PM, Foley DH, Buczak AL, Musila LA, et al. The AFHSC-Division of GEIS Operations Predictive Surveillance Program: a multidisciplinary approach for the early detection and response to disease outbreaks. Spec Issue Dep Def Glob Emerg Infect Surveill Response Syst an Updat 2009. 2011;11: S10. Available: http://www.biomedcentral.com/1471-2458/11/S2/S10
View Article
Google Scholar

[196] View Article

[197] Google Scholar

[ref65] 65. Peterson AT. Mapping Risk of Nipah Virus Transmission Across Asia and Across Bangladesh. Asia Pac J Public Health. 2013; pmid:23343646
View Article
PubMed/NCBI
Google Scholar

[199] View Article

[200] PubMed/NCBI

[201] Google Scholar

[ref66] 66. Himeidan YE, Kweka EJ, Mahgoub MM, El Rayah EA, Ouma JO. Recent Outbreaks of Rift Valley Fever in East Africa and the Middle East. Front Public Heal. Frontiers; 2014;2: 169. pmid:25340047
View Article
PubMed/NCBI
Google Scholar

[203] View Article

[204] PubMed/NCBI

[205] Google Scholar

[ref67] 67. Talla C, Diallo D, Dia I, Ba Y, Ndione JA, Sall AA, et al. Statistical modeling of the abundance of vectors of West African Rift Valley Fever in Bark??dji, Senegal. PLoS One. 2014;9: 1–23. pmid:25437856
View Article
PubMed/NCBI
Google Scholar

[207] View Article

[208] PubMed/NCBI

[209] Google Scholar

Figures

Abstract

Author summary

Introduction

Methods

Data sources

Statistical analysis

Results

Discussion

Conclusion

Supporting information

S1 Fig. Variable response curves for the most influential landscape factors in identifying the landscape suitability of Rift Valley fever outbreaks based on the Maxent model.

References