Tel.+49 30 838 62550 Fax.+49 30 838 46029
This study was conducted from December 2004 – May 2005 in Chiang Mai and Lamphun provinces in Northern Thailand. In a larger framework, 193 live pigs (from 22 cohorts) were sampled individually, followed up into an abattoir and further investigated on cutting and R for Salmonella enterica. The dynamic of agent transfer in a pork chain including pork products in Northern Thailand was studied. For that, 1,000 isolates of Salmonella enterica were available. Overall prevalence of Salmonella in samples from pigs and associated environments of the pork chain was 48.9 % (971/1982 samples). Drinking and cleaning water from farms had a similar prevalence (13.6 %). Wastewater was Salmonella positive almost every time and overshoe samples indicated comparable high positive results (95.5 % and 94.8 %, respectively). At slaughterhouse level, the highest percentage was found during cutting procedures, 23% of samples were positive, At individual pig level, the lowest prevalence was obtained from carcasses after washing (12.9 %). Between farm faeces and mesenteric lymph nodes samples, no considerable difference was noticed, 61.4 % and 63.9 %, respectively. The prevalence after splitting was about 2.5 times higher than that after washing. Caecal content yielded the highest percentage of positive samples (83.1 %). The number of Salmonella positive results was different depending on sites and cohorts of investigation. The highest positive correlation coefficient was found between carcasses after washing and FP (rs = 0.66; P = 0.0014), indicating that the carcass quality after splitting related to FP quality. Here, a relative risk (1.64; 95%CI: 1.294 - 2.089) was observed with statistical significance. Detecting of Salmonella on CW increased the odds (OR = 3.9; P-value = 0.039) of FP. Salmonella from cutting boards increased odds of TP (OR = 3.9; P = 0.042) significantly. Additionally, Salmonella on overshoes at fattening farms increased odds of Salmonella being positive in TP (OR = 5.5; P = 0.015). The detection of Salmonella in mesenteric lymph nodes increased the odds of Salmonella findings in CC (OR = 2.3; P = 0.045) and of contaminated FP (OR = 2.0, P = 0.030). Overall, 26 serovars were identified. Salmonella Rissen was the predominant serovar (45.9%). 7 serovars (S. Anatum, Krefeld, Panama, Rissen, Stanley, Typhimurium and Weltevreden) were identified throughout the complete chain (farms, slaughterhouses and R samples). 11 serovars (Salmonella Afula, Agona, Alfort, Bovismorbificans, Chittagong, Corvallis, Derby, Hato, Israel, Langensalza, Regent and Rideau) were detected only in farm samples including faeces from animals entering slaughterhouse, whereas 3 serovars (Salmonella Eppendorf, Livingstone and Tsevie) were detected only from samples from the slaughterhouses. One serovar (S. Enteritidis) was isolated only from 1 R product sample and could not be detected from somewhere else. Dendrograms of PFGE patterns (pulsotypes) of S. Krefeld, S. Panama and S. Bovismorbificans were highly similar, while those of S. Stanley, S. Typhimurium, S. Rissen and S. Corvallis were highly diverse. The highest number of isolates was obtained from overshoes. The ratio of 3.1 isolates per pulsotype indicated the smallest variability, which was observed in isolates from FP. Transfer/ trace back to the farm and evidence for contamination during slaughter could be observed. The pulsotype SRX02 could be traced back to the farm of origin; it was obtained from wastewater, overshoes, F, CC, ML, CS, CW, FP, and R samples. Retail: frequently, Salmonella isolates from R pork were more closely related to mesenteric lymph nodes and/or samples from environment during cutting and/or FP and/or TP.