Following a request from the European Commission, the Panel on Animal Health and Welfare (AHAW) was asked to deliver a scientific opinion on the risk of Rift Valley fever (RVF). The first term of reference (ToR) requested an update on the global occurrence of Rift Valley fever and possible changes in the distribution during the last 10 years. Although the cyclic occurrence of RVF in endemic areas (once every 5–15 years) makes it hard to observe possible changes in the spatial distribution of RVF over a 10-year period, comprehensive literature review and screening of OIE (World Organisation for Animal Health) outbreak reports indicates that RVF moved north within Mauritania, in a desert area, but no new countries have become infected during the past 10 years. Further, a strong increase in reported outbreaks in South Africa was observed in the last 10 years, which may be partly due to better reporting and registration of RVF outbreaks and does not necessarily indicate an increased risk.
The second ToR requested maps of the Mediterranean Basin displaying the geographical distribution of potential invertebrate hosts, taking into account their vector competence and seasonal variation in abundance. First, nine potentially competent RVFV vectors were identified: (i) Aedes vexans, (ii) Ochlerotatus caspius, (iii) Ochlerotatus detritus, (iv) Culex pipiens, (v) Culex theileri, (vi) Culex perexiguus, (vii) Culex antennatus, (viii) Culex tritaeniorhynchus and (ix) Aedes albopictus. A systematic literature review was then conducted to compile presence/absence data and relevant environmental and eco-climatic data for these nine species. All information extracted from the literature was used to generate reported and predicted presence maps. The predicted presence maps show the probability of occurrence of the vectors, which is mainly determined by the amplitude of the daytime and night-time land surface temperature, the set of ecological determinants derived from the systematic literature review, the reported presence and absence data for the vectors as well as two vegetation indices.
The predicted presence maps show that the probability of the presence of Aedes vexans and Aedes albopictus appears to be medium across large areas of the countries around the Mediterranean Basin. The probability of the presence of Ochlerotatus caspius, Ochlerotatus detritus, Culex pipiens and Culex theileri appears to be medium to high in the coastal areas and deltas of the countries around the Mediterranean Basin. The probability of the presence of the Culex perexiguus and Culex antennatus appears to be high around the Nile Delta.
In the Mediterranean Basin, the largest number of mosquito species and the highest population density are found during summer and autumn (from the beginning of June to the end of September). During winter (from November to March), there is reduced mosquito activity. Geo-referenced data on the abundance of vectors in the southern Mediterranean area are scarce. Only for Cx. pipiens were sufficient abundance data found in the literature to generate predicted abundance maps. This species is estimated to be abundant in the coastal areas of the region of concern (RC). To provide improved maps of the seasonal variation in vector abundance in the countries around the Mediterranean Basin, vector collection programmes in these areas would need to be initiated while existing ones should be intensified. Further, detailed laboratory investigations to determine the vector competence of each of the potential vector candidates are needed and field studies are needed to gain further insight on their vector capacity.
The third ToR asked AHAW to assess the risk of introduction of Rift Valley fever virus (RVFV) into the RC, especially through the movements of live animals and vectors. The RC, as defined in the mandate, comprises Mauritania, Morocco, Algeria, Tunisia, Libya, Egypt, Jordan, Israel, the Palestinian Territories, Lebanon and Syria. Since RVFV was introduced and is probably still present in Egypt and Mauritania, these two countries were excluded from this assessment. The Veterinary Services of the RC reported that currently (2012–2013) no official import of live animals from RVF-infected countries into the RC is allowed. Consequently, this assessment concerned only undocumented movements of RVFV-infected animals, using a quantitative model, parameterised by expert knowledge elicitation (EKE). The EKE model indicates that some hundreds of RVFV-infected animals may be introduced without documentation into RC when an epidemic in the source areas occurs. The number of infected animals moving into the RC from the east source (the Arabian Peninsula and East Africa) would be higher than the number of infected animals moving into the RC from the west source (Central and West Africa). This is mainly due to the higher number of expected movements into the RC, and the shorter duration of the journey from the east source than from the west source. This results in a higher probability of infected animals remaining infectious when entering the RC. Additional to the risk of introduction of RVFV through undocumented movements of infected animals, the possibility that RVFV can be introduced into the RC by vectors moved by wind was assessed. It was concluded that, although introduction of RVFV into the RC by vectors through wind cannot be quantified, it is expected to be small in comparison with the risk of introduction by infected animals.
The fourth ToR requested that the risk of RVF becoming endemic in the RC be assessed. The transmission model developed by Fischer et al. (2012) was used to determine the initial epidemic growth rate of RVFV infections, which is an indicator of the potential occurrence of RVFV spread to following virus introduction. The model was assessed using parameters for host, vector and pathogen derived from literature and using the predicted relative abundances of Cx. pipiens obtained for ToR 2. Estimates of host densities, which are also needed to assess the risk, were obtained from the FAO (2007) livestock grid. Two scenarios for two different host preferences were applied in the model, and, consequently, two risk maps were generated. The first scenario concerned vectors biting livestock only and the second scenario concerned vectors biting both livestock and other, refractory, hosts. Both scenarios showed that there is a potential for the occurrence of RVFV spread in the coastal areas of the RC as well as on the banks of the River Nile. Because of lack of quantitative information on the seasonality of vector abundance and on the vertical transmission of RVFV within local vector species, survival of the virus during the period with limited vector activity could not be assessed, although this is necessary to assess the risk of endemicity. When these seasonal abundance data and data for other species than Cx. pipiens become available in the future, they will need to be included in the model, and then the risk of endemicity can be modelled. Furthermore, when more detailed spatial information is available, the picture may also change, and more detailed risk zones may be distinguished.