Farm-to-fork changes in poultry microbiomes and resistomes in Maputo City, Mozambique

Abstract

Increasing demand for poultry has spurred poultry production in low- and middle-income countries like Mozambique. Poultry may be an important source of foodborne, antimicrobial-resistant bacteria to consumers in settings with limited water, sanitation, and hygiene infrastructure. The Chicken Exposures and Enteric Pathogens in Children Exposed through Environmental Pathways (ChEEP ChEEP) study was conducted in Maputo City, Mozambique from 2019 to 2021 to quantify enteric pathogen exposures along the supply chain for commercial and local (i.e., scavenger) chicken breeds. Here, we performed metagenomic sequencing of total DNA from banked ChEEP ChEEP samples to characterize fecal and carcass microbiomes and resistome diversity between chicken breeds and along the supply chain. Fecal samples (n = 26) were collected from commercial and local chickens at production sites and markets and carcass (n = 49) and rinse bucket samples (n = 26) from markets. We conducted taxonomic profiling and identified antimicrobial resistance genes (ARGs) from metagenomic sequence data, focusing especially on potential human pathogens and "high-risk" ARGs. We estimated alpha diversity for each sample and compared by site and breed. We estimated Bray-Curtis dissimilarity between samples and examined clustering. We found that commercial and local chickens harbored distinct fecal potential pathogens and resistomes at production and market sites. Many potentially pathogenic bacteria and ARGs present in chicken fecal samples are also present on carcasses sold to consumers. Finally, commercial chicken carcasses contain high-risk ARGs that are not necessarily introduced from chicken feces. These results indicate markets are an important site of exposure to potentially pathogenic bacteria and high-risk ARGs.

Importance: While chicken eggs and meat are a critical protein source in low-income settings, antibiotics are routinely fed to chickens with consequences for selection of antimicrobial resistance. Evaluating how poultry gut bacterial communities, including potential human pathogens and high-risk antimicrobial resistance genes, differ from farm to market could help identify where to target interventions to minimize transmission risks to human populations. In this study in Maputo City, Mozambique, we found compositional differences between commercial and local chicken breeds at production and market sites. We also found that while all potentially pathogenic bacteria and many high-risk antimicrobial resistance genes persisted from production and market through processing, some resistance genes were detected on carcass samples only after processing, suggesting human or environmental contamination is occurring within markets. Overall, our findings indicate that open-air markets may represent a critical juncture for human exposures to pathogens and antimicrobial resistance genes from poultry and poultry products.

Keywords: East Africa; antibiotic resistance; markets; metagenomics; poultry.

Fig 1

Diagram of samples collected along the poultry value chain. Samples were collected from commercial (top) and local (bottom) chickens. Comparisons were made between breeds for fecal samples at production and market sites and carcass processing samples. Samples within each breed were compared between farm and market and market and processing.

Fig 2

Commercial and local chicken fecal microbiomes and resistomes are compositionally distinct at production sites. (A) NMDS plot of Bray–Curtis distances for pathogen composition. (B) Differences in relative abundance of pathogens among local versus commercial (reference) chickens, modeled using beta-binomial regression. (C) Heat map of HR-ARG abundances among commercial and local chickens.

Fig 3

Commercial and local chicken fecal microbiomes and resistomes are compositionally distinct at market sites. (A) NMDS plot of Bray–Curtis distances for pathogen composition. (B) Differences in relative abundance of pathogens among local versus commercial (reference) chickens, modeled using beta-binomial regression. (C) Heat map of high-risk HR-ARG abundances among commercial and local chickens.

Fig 4

Consumers are exposed to clinically important HR-ARGs via chicken carcasses that are not necessarily introduced from chicken feces. (A) Violin plot depicting HR-ARG abundance measured in FPGE of local and commercial chicken samples at production, market, and processing. HR-ARGs are abundant at multiple points in the supply chain, especially among commercial chickens. Note that local carcass samples are missing from this plot as no HR-ARGs were detected in these samples. (B) Heat map of HR-ARG abundances among commercial fecal and carcass samples. HR-ARGs ant(2)-Ia, aph(3)-IIIa, and mphE were identified in commercial carcass but not market fecal samples.