Bacterial genomic variation plays a crucial role in adaptation and transmission across different environments. Understanding this variation is particularly vital for foodborne pathogens like *Listeria*, which can inhabit both natural and food-processing environments. Previous research has been limited by a lack of large-scale genomic studies encompassing diverse ecological contexts. This study aimed to address this gap by analyzing a comprehensive genomic dataset of *Listeria* isolates from soil, agricultural water, and produce processing facilities across the United States. The research question focuses on identifying the extent of genomic variation within *Listeria* species associated with these different environments, and determining the factors driving this variation. The study's purpose is to improve our understanding of *Listeria*'s adaptation and transmission dynamics between natural and food-associated environments. This knowledge is important for enhancing public health surveillance and food safety strategies, particularly for fresh produce where pathogen inactivation is often minimal. The importance of the study stems from the potential to identify key genomic markers and environmental factors that can be used to predict the risk of *Listeria* contamination in food production environments. Understanding the genomic basis for environmental adaptation in *Listeria* will inform the development of more effective control strategies to reduce the occurrence of foodborne illnesses.

The introduction cites several studies emphasizing the importance of genomic versatility in bacterial adaptation to various ecological niches, citing examples of gene gain, loss, and homologous recombination driven by environmental selection and dispersal. It highlights the limited understanding of genomic variation in human pathogens across different environments, emphasizing the need for investigations beyond human-associated settings. The literature review implicitly references prior studies on *Listeria* distribution in natural and agricultural settings and on its stress response mechanisms, including its σ-dependent general stress response and stress survival islets (SSI-1 and SSI-2). This context establishes *Listeria* as a suitable model for examining genomic variation linked to environmental adaptation and its role in pathogen transmission dynamics.

This study utilized a large collection of 839 *Listeria* isolates, encompassing 449 soil isolates, 115 isolates from agricultural water, and 275 isolates from produce processing facilities. The isolates represented four species: *L. monocytogenes*, *L. seeligeri*, *L. innocua*, and *L. welshimeri*. Environmental data (latitude, longitude, soil properties, climate, land use) were collected for the soil isolates. 16S rRNA gene amplicon sequencing was performed on soil samples to analyze bacterial community composition. Whole-genome sequencing was conducted for all isolates. Core single nucleotide polymorphisms (SNPs) were identified using kSNP3, and phylogenetic trees were constructed using RAxML. Core genome multilocus sequence typing (cgMLST) was performed for *L. monocytogenes* isolates. Genome annotation and identification of accessory genes were performed using established bioinformatics tools. Plasmids, stress survival islets (SSI-1 and SSI-2), and virulence genes were detected using PlasmidFinder and BLASTN. Statistical analyses, including Fisher's exact tests and Phi correlation, were used to identify environment-associated genes and assess correlations between genes and environmental variables. Machine learning (LightGBM) was employed to predict isolation sources based on cgMLST profiles. A partial Mantel test was used to assess correlations between environmental variables and average nucleotide identity (ANI). Finally, co-occurrence networks were constructed to investigate associations between *Listeria* species and other bacterial species in the soil.

The core genomes of *Listeria* species showed a strong association with isolation sources. *L. monocytogenes* lineages I, II, and III exhibited environment-associated subclades. Machine learning models accurately predicted isolation sources for lineages II and III based on cgMLST profiles. Accessory genomes also differed significantly between environments, particularly genes involved in cell envelope biogenesis and carbohydrate metabolism. Virulence genes, plasmids, and stress survival islets (SSI-1 and SSI-2) displayed significant associations with isolation sources. For example, plasmids were over-represented in soil isolates of *L. monocytogenes* but over-represented in food-associated isolates of *L. innocua*. Genomic similarity of soil isolates correlated with soil properties, climatic factors, and land use. Positive and negative correlations were observed between *Listeria* species and other bacterial species in the soil, particularly Proteobacteria and Actinobacteria.

The findings demonstrate a strong link between *Listeria* genomic diversity and environmental context, highlighting the significant adaptation of *Listeria* populations to different niches. The environment-associated genes identified, particularly those involved in cell envelope biogenesis and carbohydrate metabolism, underscore the importance of these pathways in mediating interactions with the environment. The differing prevalence of plasmids and stress survival islets in different species and environments suggests their crucial roles in adaptation to specific environmental challenges. The association between genomic diversity and various abiotic (soil properties, climate, land use) and biotic factors (co-occurring bacterial species) highlights the interplay between intrinsic (genetic) and extrinsic factors in shaping *Listeria* evolution. The limited number of closely related isolates found in different environments suggests barriers to transmission, even when geographic proximity is considered. However, the exceptions observed indicate that transmission can occur, albeit infrequently. These observations inform risk assessment and source-tracking strategies for foodborne outbreaks caused by *Listeria*.

This study provides comprehensive evidence of differential adaptation in *Listeria* species between natural and food-associated environments, shaped by both intrinsic genetic factors and extrinsic environmental pressures. The ability to accurately predict isolation sources using machine learning based on core genome data holds significant promise for source tracking and risk assessment. The identification of specific abiotic and biotic factors influencing *Listeria* genomic diversity suggests novel strategies for mitigating the risk of contamination in food production. While transmission between environments is possible, the data suggest that it is infrequent, emphasizing the significance of environmental barriers and genetic adaptations in shaping *Listeria* ecology and evolution.

The study primarily focused on isolates from specific regions of the United States, limiting the generalizability of the findings to other geographic locations. The temporal aspect of isolate collection could be further investigated to determine the timeframe of environmental adaptation. The study mostly relied on correlations, and further investigation is needed to fully understand the causal relationships between *Listeria* genomics, environmental factors, and interactions with other bacterial species. Finally, not all isolates in the food-associated group are from the same time, which could lead to some bias.