Evidence for the presence of African swine fever virus in an endemic region of Western Kenya in the absence of any reported outbreak

Abstract

Background: African swine fever (ASF), caused by African swine fever virus (ASFV), is a severe haemorrhagic disease of pigs, outbreaks of which can have a devastating impact upon commercial and small-holder pig production. Pig production in western Kenya is characterised by low-input, free-range systems practised by poor farmers keeping between two and ten pigs. These farmers are particularly vulnerable to the catastrophic loss of livestock assets experienced in an ASF outbreak. This study wished to expand our understanding of ASFV epidemiology during a period when no outbreaks were reported.

Results: Two hundred and seventy six whole blood samples were analysed using two independent conventional and real time PCR assays to detect ASFV. Despite no recorded outbreak of clinical ASF during this time, virus was detected in 90/277 samples analysed by conventional PCR and 142/209 samples analysed by qPCR. Genotyping of a sub-set of these samples indicated that the viruses associated with the positive samples were classified within genotype IX and that these strains were therefore genetically similar to the virus associated with the 2006/2007 ASF outbreaks in Kenya.

Conclusion: The detection of ASFV viral DNA in a relatively high number of pigs delivered for slaughter during a period with no reported outbreaks provides support for two hypotheses, which are not mutually exclusive: (1) that virus prevalence may be over-estimated by slaughter-slab sampling, relative to that prevailing in the wider pig population; (2) that sub-clinical, chronically infected or recovered pigs may be responsible for persistence of the virus in endemic areas.

Keywords: ASFV real time PCR; African swine fever virus; Epidemiology; Genotype IX; Kenya; Slaughter house; p72 PCR.

Fig. 1

Map depicting study site showing divisional pig population density and location of registered porcine slaughter facilities at time of sampling. This map was produced using ArcMap^TM version 9.1 with geographical data provided by ILRI GIS unit http://www.ilri.org/gis and pig population data provided by the District Livestock and Production Office 2009 figures and overlaid with the location of slaughter facilities collected in the field using a hand held Garmin® eTrex GPS unit

Fig. 2

Phylogenetic tree based on the 3‘-variable end of the B646L gene. Indicates the 20 nucleotide sequences analyzed in this study in comparison to 35 reference sequences obtained from Genbank. The 20 sequences clustered within ASFV genotype IX. The evolutionary history was inferred using the Minimum Evolution (ME) method after initial application of the Neighbor-joining algorithm. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) are shown next to the branches. The tree was drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The p-distance method was used to compute evolutionary distances and the Close-Neighbor-Interchange (CNI) algorithm at a search level of 1 was used to determine the strength of the ME tree

Fig. 3

Phylogenetic tree based on the full length E183L gene. Indicates the 20 sequences analyzed in this study that cluster within genotype IX in comparison to 16 reference sequences obtained from Genbank. The evolutionary history was inferred using the Minimum Evolution method after initial utilization of the Neighbor-joining algorithm. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1000 replicates) are shown next to the branches. The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. The evolutionary distances were computed using the p-distance method and are in the units of the number of base differences per site. The ME tree was searched using the Close-Neighbour-Interchange (CNI) algorithm at a search level of 1