- The feedlot and pen environments are not homogenous, and each environmental location has its own unique microbiome. The “micro-environments” within each location (i.e., surface layer, interface layer and soil layer) have a large influence on the microbial population that persists and thrives in that given environment. Thus, any scientific interpretations or risk factor analyses of environmental Salmonella persistence need to consider the potential confounding influence of sample type/location.
- It is likely that environmental interventions for Salmonella will need to be tailored, much like feedlot management is often tailored to the specific environmental conditions and populations of cattle within the feedlot.
- The presence of Salmonella within a given sample type is dictated by both macro-level factors (season and feedlot) and micro-level factors (interactions with specific microbial genera), interactions between which likely influence the ability of Salmonella to persist within a given environment. However, because these interactions are complex and dynamic (generating a lot of statistical “noise”), larger studies are needed to uncover the mechanistic underpinnings of Salmonella persistence given certain environmental and microbiome conditions.
control within the beef supply chain is a complex issue that the beef industry tackles through numerous strategies, including before and after harvest. One area that is gaining interest as a potential interventional target is the feedlot environment. Specifically, it is unknown how Salmonella
might persist within the feedlot environment, and whether this could act as a persistent source of Salmonella
(re)infection in live cattle destined for harvest. Environmental persistence of Salmonella
is a complex phenomenon undergirded by environment-microbe and microbe-microbe interactions. However, little is known about the beef feedlot environment and whether specific environmental and/or microbial dynamics support Salmonella
persistence within, for example, the pen floor. Identifying microbiome-environment risk factors for Salmonella
within the pen ecosystem could provide insight into potential feedlot-based interventions for Salmonella
control across the beef production chain.
Four feedlot operations were selected for the study. At each feedlot, two pens were randomly selected for inclusion at each of 4 sampling time points, corresponding to each season over a single year. At each sampling time point, soil layers in each pen were sampled as follows: boot socks (BS) were used to sample the most superficial layer; then, a large hole was dug and material was carefully collected from the surface layer (SL), interface layer (IL), and soil layer (SLL). In addition, at each sampling time point, boot sock samples of the shipping pen (SP) were collected, along with water samples from the lagoon and soil samples from the perimeter of each feedlot. A total of 447 samples were subjected to total DNA extraction and 16S rRNA sequencing, as well as culture for Salmonella
serotyping using CRISPR-SeroSeq was performed on Salmonella
-positive samples. The DADA2 pipeline was used to process raw sequence data and to generate amplicon sequence variants (ASVs) and taxonomic assignments. The effect of season, feedlot, sample type, and Salmonella
status on overall soil microbial community was assessed using permutational multivariate analysis of variance (PERMANOVA) testing. In addition, Dirichlet multinomial mixture modeling (DMMs) was performed to determine microbial community types (i.e., clusters) and to assess potential association with Salmonella
status. Finally, Bayesian network (BN) analysis was used to evaluate associations between Salmonella
status (including specific serotypes) and the following risk factors: specific microbial taxa, sample type, season, feedlot ID and pen ID.
The largest difference in overall microbial community composition was driven by sample type, with lagoon and perimeter samples clustering separately from BS, SL, IL and SLL samples. The large influence of sample type was also reflected in DMM analysis, which showed segregation of DMM cluster membership by sample type. Overall, sample type explained 32% of microbiome variation, compared to <4% for season, feedlot ID, and pen ID. Salmonella
prevalence differed between feedlots and sample types, and the presence of Salmonella
was significantly associated with differences in microbiome composition, although this effect was small (PERMANOVA R2<4%, P<0.01). These results suggest that the microbiome of each soil layer and each feedlot is unique, which could differentially impact Salmonella
dynamics. Bayesian network analysis suggests that the presence of Salmonella
within a sample is associated with numerous macro- and micro-level factors including specific microbes, soil characteristics, feedlot and season. Importantly, none of the microbiome data contained sequences that could be assigned to the Salmonella
genus, suggesting that this workflow is not sufficient to detect Salmonella
within these types of samples. Furthermore, Enterobacteriaceae sequences were detected, however these detections were not correlated with culture-based Salmonella
status, again suggesting that these sequences are not a sufficient proxy for Salmonella
status of a sample.
Differential microbiome dynamics may explain some of the variability in Salmonella
prevalence across feedlots and across different layers of the pen floor soil/manure. However, additional research is necessary to better understand these dynamics and the mechanism by which they may support Salmonella
persistence within the pen environment – especially given the large confounding effects of sample type (i.e., soil layer), season and feedlot. Given the complexity of the soil microbiome and the fact that it is significantly different between feedlots and soil layers, Salmonella
control efforts focused on pen soil microbiome manipulation or remediation may need to be tailored to both the feedlot and the soil layer being targeted.