A Bayesian inference approach to quantify average pathogen loads in farmyard manure and slurry using open-source Irish datasets

Abstract

Farm-to-fork quantitative microbial risk assessments (QMRA) typically start with a preliminary estimate of initial concentration (C_initial) of microorganism loading at farm level, consisting of an initial estimate of prevalence (P) and the resulting pathogen levels in animal faeces. An average estimation of the initial concentration of pathogens can be achieved by combining P estimates in animal populations and the levels of pathogens in colonised animals' faeces and resulting cumulative levels in herd farmyard manure and slurry (FYM&S). In the present study, 14 years of data were collated and assessed using a Bayesian inference loop to assess the likely P of pathogens. In this regard, historical and current survey data exists on P estimates for a number of pathogens, including Cryptosporidium parvum, Mycobacterium avium subspecies paratuberculosis (MAP), Salmonella spp., Clostridium spp., Campylobacter spp., pathogenic E. coli, and Listeria monocytogenes in several species (cattle, pigs, and sheep) in Ireland. The results revealed that Cryptosporidium spp. has potentially the highest mean P (P_mean) (25.93%), followed by MAP (15.68%) and Campylobacter spp. (8.80%) for cattle. The P_mean of E. coli is highest (7.42%) in pigs, while the P_mean of Clostridium spp. in sheep was estimated to be 7.94%. C_initial for Cryptosporidium spp., MAP., Salmonella spp., Clostridium spp., and Campylobacter spp. in cattle faeces were derived with an average of 2.69, 4.38, 4.24, 3.46, and 3.84 log₁₀ MPN g ^-1, respectively. Average C_initial of Cryptosporidium spp., Salmonella spp., Clostridium spp., and E. coli in pig slurry was estimated as 1.27, 3.12, 3.02, and 4.48 log₁₀ MPN g ^-1, respectively. It was only possible to calculate the average C_initial of Listeria monocytogenes in sheep manure as 1.86 log₁₀ MPN g ^-1. This study creates a basis for future farm-to-fork risk assessment models to base initial pathogen loading values for animal faeces and enhance risk assessment efforts.

Keywords: Animal manure; Bayesian inference; Exposure assessment; Pathogen load; Risk Assessment Approach; Slurry.