Global spread of *Salmonella Enteritidis* via centralized sourcing and international trade of poultry breeding stocks

Poultry products are a major source of protein globally. Industrialized poultry production utilizes a breeding pyramid structure, where a small number of breeding stocks supply the entire industry. This centralized system, coupled with advancements in breeding selection and technology since the 1940s, has led to significant consolidation in the poultry industry. This consolidation, characterized by merging and acquisition of breeds and brands, resulted in most commercial poultry worldwide originating from a limited number of nucleus breeding flocks. The success of industrialized poultry farming has unfortunately been accompanied by the widespread dissemination of *Salmonella enterica*, particularly *Salmonella Enteritidis*. A pandemic of *Salmonella Enteritidis* infections linked to poultry simultaneously emerged in North America, South America, and Europe in the 1980s, later spreading to Asia and Africa. Despite its significant impact on public health, the reasons behind the rapid and widespread dissemination of this pathogen remained largely a mystery. The study hypothesizes that the synchronized and expansive spread of *Salmonella Enteritidis* can be explained by the centralized sourcing and international trade of infected breeding stocks, a hypothesis supported by circumstantial evidence from various parts of the world.

Numerous reports and speculations linked the global spread of *Salmonella Enteritidis* to international trade of breeding stocks. Detections of *Salmonella Enteritidis* in primary breeding flocks in the UK and US, major exporters of breeding stocks, were documented. Studies suggested that effective *Salmonella* control in imported grandparent stocks protected Sweden from the pandemic and helped eliminate the pathogen from the Netherlands' national flock. Evidence from Brazil and Egypt also indicated *Salmonella Enteritidis* infections linked to imported day-old chicks. However, despite these reports, a comprehensive chain of evidence directly linking the global spread of the pathogen to the international trade of breeding stocks remained lacking. Two significant gaps existed: (1) lack of evidence demonstrating the ability of a foreign strain of *Salmonella Enteritidis* to penetrate and sustain itself within the domestic poultry production hierarchy of an importer, and (2) absence of a spatiotemporal evaluation to determine the sufficiency of this transmission mode to trigger a pandemic.

This study integrated global genomic surveillance data of *Salmonella Enteritidis* with international live poultry trade data to test its hypothesis. The researchers analyzed whole-genome sequencing (WGS) data of 30,015 *Salmonella Enteritidis* strains from 98 countries (1949–2020) obtained from EnteroBase. They identified genetically related isolates from different continents, using core genome multi-locus sequence typing (cgMLST) to assess genetic similarity. Specifically, they identified nearly identical isolates from the US and Suriname, suggesting a common source of contamination linked to breeding stock imports. Further analysis of population structure using cgMLST revealed that poultry isolates displayed tighter clustering and lower genomic diversity compared to isolates from other sources. Phylodynamic analysis using a representative subset of 914 genomes from 46 countries (1954–2020) was conducted to investigate the spatiotemporal spread and evolutionary origins of circulating *Salmonella Enteritidis* lineages. This involved constructing a maximum likelihood phylogeny, identifying major lineages, inferring MRCA ages, and estimating effective population sizes using Bayesian methods. International trade data on live poultry from the FAO, OEC, and FAS were used to assess the concordance between trade patterns and the geographic distribution of *Salmonella Enteritidis*. Finally, a generalized linear model (GLM) extension of Bayesian phylogeographic inference was employed to quantitatively assess the relative contributions of different trade variables to the spread of the Global and Atlantic lineages.

The study discovered several key findings supporting the hypothesis that international trade of breeding stocks contributed significantly to the global spread of *Salmonella Enteritidis*. First, the identification of genetically near-identical *Salmonella Enteritidis* isolates from poultry in Suriname and the US strongly suggests intercontinental transmission via breeding stocks. The Suriname isolate was from a domestically raised bird, eliminating the possibility of contamination from imported processed products. The US is a major supplier of breeding stocks to Suriname, indicating a likely link between trade and the pathogen's introduction. The discovery of closely related isolates of other *Salmonella* serotypes in the Caribbean and the US further reinforces the hypothesis. Second, analysis of population structure revealed that poultry *Salmonella Enteritidis* isolates exhibited lower genomic diversity than isolates from other sources, suggesting a limited number of common origins. Third, phylodynamic analysis identified three major lineages of poultry *Salmonella Enteritidis*: a global lineage, an Atlantic lineage (predominantly US and European), and a US lineage. The estimated emergence times of these lineages aligned with the timeline of the *Salmonella Enteritidis* pandemic and the intensification of global poultry breeding stock trade. Fourth, the study demonstrated a strong concordance between international trade of live poultry (especially breeding stocks and hatching eggs) and the geographic distribution of *Salmonella Enteritidis*. Fifth, the GLM analysis revealed a strong contribution of egg trade (specifically hatching eggs) to the international dispersal of the Global lineage. Importantly, this finding points to a crucial link between hatching eggs used as breeding stocks and the pandemic's spread.

The study's findings provide strong evidence supporting the hypothesis that centralized sourcing and international trade of infected breeding stocks played a significant role in the global spread of the *Salmonella Enteritidis* pandemic. The discovery of near-identical strains in geographically distant locations, coupled with the temporal concordance between the pandemic and the growth of the global breeding stock trade, provides compelling support. The quantitative assessment using GLM further emphasizes the importance of hatching eggs as a driver of the pandemic's spread. While the lack of direct isolates from breeding stocks limits definitive proof, the indirect evidence presented, including the Suriname case study, is highly suggestive. The results highlight the need for enhanced surveillance and improved biosecurity measures within the global poultry breeding industry to mitigate the risk of future outbreaks.

This study provides strong evidence that the global *Salmonella Enteritidis* pandemic was significantly driven by the international trade of poultry breeding stocks. The findings highlight the vulnerability of the centralized poultry breeding system and underscore the importance of strengthening biosecurity measures within this sector. Future research should focus on enhancing surveillance of *Salmonella* in breeding stocks, investigating the effectiveness of different control strategies, and developing improved risk assessment models to predict and prevent future outbreaks. The study’s novel integration of genomic and trade data offers a valuable framework for investigating the global spread of other foodborne pathogens.

The study’s conclusions rely heavily on indirect evidence, as direct sampling of *Salmonella Enteritidis* from breeding stocks is limited due to biosecurity concerns and trade secrecy. While the study uses a substantial dataset of genomic and trade data, the available information might not fully represent all relevant factors influencing *Salmonella Enteritidis* spread. Data biases, such as uneven geographical sampling of *Salmonella Enteritidis* isolates or potential inaccuracies in trade data, could influence the results. The GLM analysis also had a limited sample size for the Atlantic lineage, potentially affecting its statistical power.