During October–December 2010, surveillance and investigation teams collected 80 clinical samples from 36 humans with suspected RVF and 44 of their contacts in Atar, Aoujeft, Akjoujt, Chingetti, Kobeni, and Tintane Moughatas (Figure 1). A human with a suspected case of RVF was defined as any patient with fever associated with or without hemorrhage, jaundice, or neurologic symptoms during October–December. A human contact was defined as a healthy person living within the immediate environment (i.e., same household or neighborhood) of a human or animal with a confirmed case of RVF. Serum samples from suspected case-patients were tested by using real time reverse transcription PCR, ELISA (for detection of IgM), and virus isolation.

Samples were obtained from animals living near suspected or confirmed case-patients, and a questionnaire about the species, sex, age and reproductive status of female animals was completed by their owners. Sampling of mosquitoes was conducted in December 2010 by using animal-baited traps placed in and around houses in which suspected and/or confirmed case-patients were identified. RVFV isolation and identification was performed on mosquitoes and human serum samples as described (9). Coding regions for RVFV nonstructural protein and glycoprotein 2 were amplified by using reverse transcription PCR and described sets of primers (10,11) and then purified after electrophoresis on agarose gels, and sequenced for phylogenetic analyses.

IgM against RVFV or viral RNA genome was detected in 30 (37.5%) of 80 patients, and 6 RVFV isolates were recovered (Table 1). The dates of symptom onset for confirmed cases of RVF in humans indicated that the outbreak probably started at the end of October, peaked in November, and decreased the second week of December. Among 30 patients (8 contacts and 22 suspected case-patients) with acute RVFV infections, the male: female ratio was 5:1 (25:5), and the median age was 35 years (range 17–70 years); 3 deaths were reported. Except for 1 man (a teacher), all infected men were herdsmen and all infected women were housewives.

For the animal investigation, serum samples were obtained and processed from 83 small ruminants (70 goats and 13 sheep) with a median age 4 years (range 1–10 years) from all localities tested (Table 2, Appendix). IgM against RVFV was detected in 23 (27.7%) of 83 animals. In addition, samples from 5 sick camels were collected in neighborhoods of persons with suspected and confirmed cases of RVF, and RVFV RNA was detected in 3 of these samples. Information for populations living in Adrar indicated that the first probable case of the disease among animals was in a sick camel in Aoujeft during the last week of October (8).

A total of 2,741 mosquitoes belonging to 5 genera and 11 species were collected for entomologic investigation. Culex antennatus was the most abundant mosquito species (83%), followed by Anopheles gambiae (9%) and An. rufipes (2%) mosquitoes (Table 3). Three RVFV strains were isolated from Cx. antennatus mosquitoes collected in Safia (Adrar Province).

Figure 2

Figure 2. . Maximum-likelihood trees for the RNA A) small and B) medium segments of Rift Valley fever virus. Tees show relationships among strains isolated from different localities and countries. Samples from Mauritania...

Phylogenetic analysis based on RVFV partial small and medium RNA sequences showed a low level of variation between isolates from humans, animals, and mosquitoes. Isolates belonged to the West Africa lineage and were closely related to strains from Diawara (1998) and Hodh El Garbi (1998) in Mauritania and from Kenya (1965), Zimbabwe (1974), and Zambia (1975) (Figure 2).

After a 4-fold increase in rainfall in Mauritania from 2009 through 2010 (59 mm vs. 196 mm; (Aoujeft Weather Station). RVF clinical cases were confirmed in Adrar and Inchiri Provinces. Although RVF was reported in some provinces of Mauritania in 1987, 1998, and 2003 (5–7), cases in humans from Adrar and Inchiri Provinces have not been previously reported. However, in 2010 a total of 30 laboratory-confirmed human case-patients were recorded in these 2 provinces; 3 of the case-patients died. Both areas had a drought in recent years or their average rainfall was not >60 mm before 2010. The study showed animal abortion rates of 26.13% (23/88) and 30.68% (27/88) in 2 groups of small ruminants living near patients with recent RVFV infections. Three camels were infected with RVFV, and the cases were probably linked to human infections. The first human death from RVF was reported on November 11 in the village of Tawaz, in which a high mortality rates were observed in camel populations (11).

Serologic evidence of RVF in camels from northern Sudan and southern Egypt in 1977 (12) led to camels being suspected as key contributors in RVFV transmission (13). The involvement of camels in human infections could explain the high male:female ratio (25:5) for infections in humans because men are more involved in animal husbandry. In 2006, a similar sex difference for infection was observed in Kenya, where men had a prevalence rate of IgG against RVFV that was 3 times higher than that in women (14).

Entomologic investigations in Mauritania showed that only Cx. antennatus mosquitoes were infected with RVFV. Similar finding were reported for previous RVF outbreaks in Mauritania. Furthermore, Cx. antennatus mosquitoes were found to be infected with RVFV in Nigeria in 1967 and in Kenya during 1981–1984, and vector competence was demonstrated in laboratory studies (15). Phylogenetic analysis suggests reemergence of the local RVFV focus and emphasizes the need to strengthen RVF surveillance and other control measure in Mauritania after heavy rainfall.