2.1 Study area

Nigeria is located in West Africa, bordered by the Benin Republic to the west, the Niger Republic to the north, Cameroun to the east and the Atlantic Ocean to the south. The country is divided into 36 states and the federal capital territory (FCT) Abuja. The 36 states are further divided into local government areas (LGA) with a total of 774 LGAs. Pigs are reared in 30 states of the country. Nigeria has two seasonal periods, the rainy season which is from April to October, and the dry season from November to March. The common breeds of pigs in Nigeria are the Iberian type Nigerian Indigenous Pig (NIP) which is usually black or pied; others are exotic breeds including Large White, Landrace, Duroc, Hampshire and their crossbreeds (Fadiga et al., 2013).

2.2 Selection and description of live pig markets

Organized LPMs are common in northern parts of Nigeria, while in the southern part of Nigeria, pigs are sold live at farm gates or as processed pork at stores. Due to limited information on the LPM operation in Nigeria, selection of LPM sites were done purposively based on information from pig farmers, easy access, literature search, logistics, visit to LPM, the network of distribution, security challenges and the number of live pigs traded at markets visited. Hence, four live pig markets were purposefully selected, all located in Northern Nigeria, namely, Katsit pig market (also called Kafanchan pig market) in Zagon Kataf LGA of Kaduna State, Pandam pig market in Qua Pan LGA and Dawaki pig market in Kanke LGA both located in Plateau State and Numan pig market in Numan LGA of Adamawa State (Figure 1). Katsit pig market is located on latitude N 9°34255 and longitude E 8°164936 at the outskirts of Kafanchan, Kaduna State. It is the largest LPM in Nigeria (Table 1); it is once a week market with trading activities taking place on Thursdays only. Located within the premises of the Katsit pig market is a slaughter slab where pigs bought within the market are slaughtered and sold to traders and the surrounding communities. The second market is Pandam pig market located on latitude N9°.52519 and longitude É8°.50510, it is also a once in week market with sales of pigs taking place on Mondays (Table 1). There is a slaughter point 500 m to the market. The third market is Dawaki pig market, which is located on latitude N 9.429581° and longitude E 9.571361°. Apart from pigs, other animals sold at the Dawaki pig market including sheep, goats, chickens and dogs. The Dawaki pig market holds weekly on Thursdays. The slaughter point at Dawaki pig market is located directly behind the perimeter fence of the market with an access point for free movement of butchers and traders. The fourth market is the Numan Pig market located in Numan LGA of Adamawa State, the market holds on Tuesdays with GPS coordinates N 09°.27440′ E12.002043′

2.3 Data collection

Data were collected based on key informant interviews of government and pig traders' union officials at the market sites. Data collected were average weekly population of pigs in the markets. Towns and villages from where pigs are brought into the markets and locations where the pig purchased are transported to. Government regulations in the markets and attached slaughter slabs and the fate of pigs unsold on at the market. Data on the average number of pigs slaughtered at the slabs were also collected.

2.4 Field observation

Due to poor records at the selected LPMs, observations of activities of the value chain actors (traders, farmers, transporters) and slaughter slab operators were carried out using a checklist. These include pre-entry inspections of pigs going in and out of the market site, mode of transportation of the pigs, holding facilities at the markets, for example, pens, activities of the slaughter slab operators and access of people and other animals to the market sites. Others include common breeds of pigs at the market, estimated age of pigs brought to the market, sanitation and hygiene at the market sites. To carry out this aspect of the study, arrival at market sites was from 7 am to 8 am depending on location to maximize observation of activities, sample collections, and travel time back to the laboratory. At least two cycles of visits were planned per month for data and sample collection.

2.5 Sample size determination and collections

Samples were collected between April 2019 and February 2020, and October–November 2020 at the Dawaki, Katsit, Numan and Pandam pig markets in Nigeria. With a population of 7 million pigs, the minimum recommended sample size for such a population using a confidence level of 95% and a margin of error of 5% gives a calculated sample size of 384 (Dohoo et al., 2003). The samples were collected using a convenient sampling method, based on the cooperation of the pig traders and slaughter slab operators. The samples collected were whole blood, tissue and sera samples. Five millilitres of blood in heparinized and plain sample bottles were collected from pigs at LPMs and their slaughter slabs. Tissue samples (liver, spleen and mesenteric nodes) were collected in labelled sample bottles from all pigs slaughtered at the slaughter slabs within the premises of the LPMs. Criteria for sample collection at the slaughter slab were as follows: pigs bought at LPMs and pigs from external sources were excluded. Both apparently healthy and sick pigs were sampled. Sample collection was disrupted between March 2020 and September 2020 due to COVID-19 lockdown restrictions in Nigeria.

2.6 Detection of African swine fever antibodies

Sera samples collected were analysed using commercially available ID Screen ASF Indirect kit (IDVet, Grabels, France) specific for the presence of antibodies against ASFV p32, p62 and p72 proteins according to the manufacturer´s instructions.

2.7 Detection of African swine fever virus

Total DNA was extracted from whole blood and homogenized tissue samples using the Qiagen Mini kit (Qiagen, Hilden Germany) following the manufacturer's protocol. A 257 bp region corresponding to the C-terminal region of the B646 gene encoding the p72 major capsid protein was amplified using the diagnostic primers PPA1 and PPA2 as previously reported to confirm the presence of the ASFV genome (Agüero et al., 2003).

2.8 ASFV genotyping, serogrouping and CVR sequence analysis

Representative ASFV positive samples from the four selected LPMs were selected for molecular characterization targeting four regions of the virus. The regions were the C-terminal end of the B646L gene encoding the p72 protein, the full length of E183L gene encoding p54 protein, central variable region (CVR) within the B602L and CD2v protein encoding the EP402R gene (Bastos et al., 2003; Gallardo et al., 2009; Malogolovkin et al., 2015). ASFV sequences were generated commercially by Macrogen Inc. (Netherlands, Europe) and LGC Genomic (Germany). The raw sequence data were assembled using Staden Package (http://staden.sourceforge.net/) and multiple alignments were performed using the CLUSTALX algorithm implemented in MEGA X (Kumar et al., 2018). Sequences for four gene segments of ASFV were compared to available sequences in the GenBank using the Basic Local Alignment Search Tool (https://blast.ncbi.nlm.nih.gov/Blast) with default settings. MEGA X was used for phylogenic analysis to a neighbour-joining tree inferred using the Maximum Composite Likelihood method for p72 protein with data resampled 1000 times in the bootstrap method. For the p54 protein encoded in E183L gene, a Minimum Evolution method tree was inferred using the Maximum Composite Likelihood method and data resampled 1000 times in the bootstrap method (Kumar et al., 2018). The CD2v protein sequences were inferred using neighbour-joining method and data resampled 1000 times in the bootstrap method. Sequences of ASFV genotypes and serogroups were retrieved from the Genbank for the construction of the phylogenetic trees for both p72, p54 and CD2v protein genes. The tandem repeat sequences (TRS) of CVR (B602L gene) from this study were translated to protein sequences using ExPASy (https://web.expas y.org/translate/) with default software settings. The translated amino acid tetramers were matched with previously reported codes (Achenbach et al., 2017; Nix et al., 2006). ASFV TRS from this study were compared with 21 TRS of CVR previously reported in Nigeria and other West African countries. Sequences were deposited into GenBank with the following accession numbers: OL692379-OL692391 (p72 gene), OL692397-OL692406 (p54), OL692410-OL692419 (B602L gene) and OL780773- OL780778 (CD2v gene).

2.9 Data analysis and mapping of pig movement

Results of data and analysed samples collected were tabulated and maps of pig movements were generated to show the inflows and outflows of pigs at the four selected LPMs in Nigeria.

2.10 Ethics approval

This study was approved by the National Veterinary Research Institute Animal Ethics Committee Vom, Nigeria (AEC/03/26/16).

3.1 Data on weekly population and destinations of pigs at the four selected live pig markets

A total of 34 visits were carried out to the four selected LPMs for data and sample collection within 13-month period. The common breeds at the four selected live pig markets visited were the crossbreed of Nigeria Indigenous Pig (NIP), Large White, Landrace and Duroc with an age range from 6 months to 5 years. Weekly population of pigs at the selected LPMs was as follows: Katsit (3000–6000 pigs) (Figure 1a, Table 2), Pandam (1500–3000 pigs) (Table 2, Figure 3b), Dawaki (400–500 pigs) and Numan pig markets (1000–1500). The pig population on a market day depends on the season of the year; there are more pigs during the dry season compared to the rainy season. Overall, it is estimated that 500,000−700,000 pigs transit through these markets yearly. Data collected based on the interview of the pig market officials show that the pigs are also brought to the market and transported to several locations (Figures 1 and 2). Pigs bought from Katsit pig market are supplied to 18 towns or cities in 13 States of Nigeria (Table 2, Figure 2). Pigs are brought into Katsit pig market from several towns or villages in Kaduna (14), Niger (4), Nasarawa (3) and Plateau (2) States (Table 2, Figure 1). While pigs at the Pandam market are sourced from villages or towns in Plateau State (20) and Nasarawa State (4). Interestingly, pigs from Pandam pig market are also transported to the Katsit pig market. Furthermore, pigs sold at Dawaki pig market are mostly sourced from villages and towns in Bauchi (2) and Plateau (8) States, and transported to 5 States in Nigeria (Table 2). Live pigs traded at the Numan pig market are sourced from towns and villages in Adamawa (9), Gombe (7), and Taraba (2) States as well as the Republic of Cameroon. From the Numan pig market, the pigs are transported to five states of Nigeria (Table 2, Figure 2). Based on the data collected at the slaughter slabs within or closed to the LPMs, averagely, 15–40 pigs are slaughtered at Katsit, while 4–10 pigs are slaughtered at Dawaki, Numan and Pandam markets. The number of pigs slaughtered is dependent on the season of the year and disease outbreak. Postmortem meat inspection is usually carried out by a veterinary assistant with emphasis on zoonotic diseases like tuberculosis and porcine cysticercosis which are partially or wholly condemned for destruction.

3.1.1 Patten of trade at live pig markets

Apart from Numan and Katsit pig markets, there are no holding pens in all the three other LPMs. Pigs are kept under trees on the bare ground which is not cemented and they tethered to pegs very close to each other (Figure 3a and b). Although at Katsit LPM there is a private withhold facility, within the market, there are no pens. The traders transport their pigs in trucks, vans and motorcycles to the LPMs. The pigs are collected from different communities or villages close to the markets. Pigs are bought based on physical appearance and size without weighting. There were no pre-entry antemortem inspections before pigs were allowed into the market's sites. Although, inside the LPMs antemortem inspection is usually carried by traders to rule out porcine cysticercosis by examination of the tongue of pigs before any transaction. All the four pig markets in this study were fenced, but the perimeter fencing of Dawaki and Katsit was partial, and free-roaming pigs or dogs were observed at the two markets. Another observation is the washing of transport vehicles within the premises of the markets. Further observation is the slaughtering of moribund pigs in the premises of the market sites (Figure 3d). Other observations include free movement of slaughter slab operators between the slaughter slabs and the market area where live pigs are kept. Pigs are slaughtered on every market day and sales of pork were observed at all live pig market sites. Pigs showing clinical signs suggestive of ASF were observed at three LPMs (Dawaki, Katsit, Pandam) cyanotic skin and ASFV genome was detected by PCR in samples collected from some of these pigs (Figure 3c).

3.2 Laboratory results

3.3 Phylogenetic analysis of B646L, E183L and EP402R genes of African swine fever virus detected and sequence in samples collected

The targeted fragments of the ASFV samples were amplified and sequenced successfully as follows: B646L (13), E183L (10) and B602L (10) genes. While only six ASFV samples were successfully amplified and sequenced for the EP402R. The samples were selected from the LPMs as follows: Katsit (8), Dawaki (3), Pandam (1) and Numan (1).

The phylogeny of B646L gene revealed ASFV genotype I was circulating in all the selected LPMs within the study period (Figure 2). Phylogenetic analysis of the E183L (p54) gene of ASFV revealed that live pig market ASFV sequences clustered with genotype 1a sequences of Nigeria and West African countries (Figure 3). Furthermore, the phylogeny of the EP402R (CD2v) gene shows all the ASFV sequences from Nigeria clustered within serogroup 4 alongside isolates from Mali (Figure 5)

3.4 Sequence analysis of tandem repeat sequences of CVR (B602L)

The analysis of TRS of CVR (Table 4) revealed eight TRS variants, namely, Tet-10, Tet-13, Tet-18, Tet-20a, Tet-22, Tet-26, Tet-33 and Tet-41 from samples collected between April 2019 and November 2020. Six of the TRS variants were new and different from previously reported variants reported in Nigeria. The TRS variants were recovered as follows: Dawaki (1), Katsit (6) Numan (1) and Pandam (1). Katsit and Pandam share similar TRS variants (Tet-13) (Table 4); similarly Tet-13 was recovered from Katsit and Pandam LPMs.