Population dynamics, host biting networks and dispersal distances of potential vectors of Rift Valley fever virus in north-eastern KwaZulu-Natal and western Free State provinces, South Africa

Abstract

Rift Valley fever virus (RVFV) is a single stranded RNA zoonotic mosquito-borne virus that causes periodic outbreaks with significant animal health, public health and economic consequences. The virus is endemic in the sub-Saharan Africa but has also spread to the Arabian Peninsula and other African regions. Three major Rift Valley fever (RVF) outbreaks have been reported in South Africa, from 1950-1951, 1973-1976 and 2010-2011. Recent circulation of RVFV has been reported in north-eastern KwaZulu-Natal (neKZN), South Africa, where the virus is suspected to be endemic, although outbreaks have not been reported in the area. The vectors of RVFV in this area are poorly understood, although several mosquito species such as Aedes (Aedimorphus) durbanensis, Aedes (Neomelaniconion) circumluteolus, Culex (Culex) tritaeniorhynchus and Culex (Lasioconops) poicilipes are the suspected vectors since they have previously been associated with transmission of RVFV. The first aim of the study was to describe the species diversity, abundance and seasonal population dynamics of potential RVFV vectors in neKZN, to identify potential RVFV vectors. A total of 141,594 mosquitoes consisting of 46 species from 8 genera including Aedes (n = 2,610), Culex (89,143), Mansonia (44,029), Anopheles (4,709) and others (n = 1,103) were sampled over a three-year period from pans located in three sites (Namaneni, Mpala and Bumbe). Culex tritaeniorhynchus (n = 55,491, 39.19%), Mansonia (Mansonioides) uniformis (n = 35,453, 25.04%) and Cx. poicilipes (n = 14,192, 10.02%) were the most frequently collected species. The most common Aedes species were Ae. durbanensis (n = 1,473), Ae. mcintoshi (n = 526) and Ae. circumluteolus (n = 265). The mosquito abundance changed significantly over the study period (p < 0.001). Overall mosquito abundance peaked significantly between January and April, while a substantial increase in Aedes spp. was observed in May. Namaneni (n = 107,204) had the highest species abundance and diversity (H = 1.79 and 1-D = 0.77) while Mpala (n = 32,213) had the highest species richness (S = 39). The second aim was to determine the vertebrate blood meal sources of potential RVFV mosquito vectors in neKZN and to characterize the host-biting network. A total of 561 blood-fed mosquitoes were collected from rural areas (n = 445) and Ndumo Game Reserve (n = 116). Nyala (Tragelaphus angasii) was the most frequent host from the reserve (n = 61) while cattle (Bos taurus) was the most common vertebrate blood meal sources from the rural area (n = 195). The rural network comprised 42 species (24 mosquitoes and 18 hosts) with 314 interactions and the reserve network consisted of 13 species (4 mosquitoes and 9 hosts) with 95 interactions. Most of the blood meals for floodwater Ae. circumluteolus and Ae. mcintoshi in the rural areas originated from cattle while nyala was the most common vertebrate blood meal source in the reserve and a source of blood for majority of Ae. circumluteolus meals. The third aim was to detect RVFV in mosquitoes using real time reverse transcriptase polymerase reaction (RT-PCR). A total of 55,113 mosquitoes (Aedes = 2,606, Culex = 36,624, Mansonia = 15 149 and others = 734) from neKZN, including blood-fed mosquitoes, were screened for RVFV and all tested negative. Low RVFV detection rates in mosquitoes have also been reported in other studies. The fourth aim was to estimate the dispersal distances of various potential RVFV mosquito vector species from a seasonally flooded wetland in the western Free State, where large RVF outbreaks occur. A capture mark release recapture (CMRR) study was conducted in a seasonal pan after a period of heavy rainfall. A total of 5,622 mosquitoes were marked with fluorescent dye and released. A total of 281,254 mosquitoes, of which 163 (0.06%) were marked, were collected using 30 CDC traps at distances between 250 and 1000 m over the following 6 days. Neomelaniconion spp. accounted for 19% of the recaptures, while Cx. theileri (43%) and other Culex mosquitoes (32%) were the most frequently recaptured mosquitoes. The mean distance travelled by Neomelaniconion spp. and Ae. (Ochlerotatus) caballus/juppi was 250 m while for Cx. theileri it was 489 m. This study has demonstrated that several mosquitoes such as Cx. tritaeniorhynchus, Cx. poicilipes, Cx. antennatus and Mansonia spp., that have been implicated in RVFV transmission are abundant in neKZN and could play an important role in transmission. Other RVFV suspected vectors including Ae. circumluteolus, Ae. mcintoshi and Ae. durbanensis were also collected. Vertebrate host species such as cattle, goats and nyala likely also play a central role in RVFV transmission and maintenance. Results from the dispersal study contribute to the knowledge of dispersal of potential RVFV mosquito vectors, which is important for predicting the local spread of mosquito-borne diseases.