Project Summary

Longitudinal Evaluation of Salmonella in Environmental Components and Peripheral Lymph Nodes of Fed Cattle from Weaning to Finish in Three Distinct Feeding Locations

Principle Investigator(s):
T. Matthew Taylor, Kerri B. Gehring, J. E. Sawyer, Ashley Arnold
Department of Animal Science, Texas A&M University
Completion Date:
July 2021



  • Rates of Salmonella-positive environmental samples differed by location of selected commercial feedyards that historically produced cattle with “low” and “high” levels of Salmonella prevalence in lymph nodes (Locations A and B, respectively). 
  • Serovars varied by location and sample type with minimal overlap between lymph nodes (LN) and environmental samples.
  • No Salmonella-positive LNs were recovered from cattle managed at Location A. 30% of lymph node samples from Location B returned a Salmonella-positive result. Salmonella-positive results did not differ across feeding stages for Location B. 


Numerous studies have shown that harborage of Salmonella in peripheral lymph nodes (LNs) is a pre-harvest phenomenon. Peripheral LNs are commonly associated with the fat that accompanies lean trimmings destined for ground beef products. After several years without a salmonellosis outbreak attributed to beef products, two major outbreaks occurred, highlighting the need for an increased understanding of the relationship between Salmonella and beef. Preliminary data collected by our research team have shown management practices vary among feed yards, and that salmonellae prevalence can vary by environmental source. Furthermore, clear differences between feeding locations regarding Salmonella prevalence in the LNs of cattle post-harvest have been identified, as well as an increase in Salmonella prevalence coinciding with duration of exposure to a given production environment. While each of these studies have added valuable data to the increasing body of knowledge in this area, until the present study, longitudinal sampling of environmental components (water, pen soil, individual feed ingredients, prepared rations, and fecal drop samples) and LNs from specific groups of cattle over time had yet to be conducted.

Calves (n = 120) were raised at the Texas A&M University McGregor Research Center; in lieu of beginning the backgrounding/stocker phase, thirty weanling calves were transported to Rosenthal Meat Science and Technology Center (RMSTC) for harvest. Of the remaining ninety calves, sixty were transported to one of two commercial feedyards (thirty calves each) in Texas to complete their feeding program, and thirty were retained at the McGregor Research Center. The selected commercial feedyards have historically produced cattle with greatly differing levels regarding Salmonella prevalence in lymph nodes (one “high” and one “low”). In two previous studies, weanling calves from McGregor returned Salmonella-negative LN samples, determining our inclusion of McGregor Research Center as a third feeding location to allow investigation of a second “low-prevalence” feeding environment.   

Ten calves from each feeding location were later transported to RMSTC for harvest: (1) at the conclusion of the backgrounding/stocker phase, (2) 60d into their feeding period, and (3) upon reaching market weight. On each harvest day, left and right subiliac and superficial cervical lymph nodes were collected from each carcass, pooled by type within animal (n = 238 total samples), and subjected to an initial screening for Salmonella prevalence using a Hygiena BAX Q7 unit. Samples returning a positive BAX result were subjected to cultural analyses and confirmation per the USDA-FSIS Microbiology Laboratory Guidebook (MLG). Three isolates from confirmed positive samples were submitted to the USDA-APHIS National Veterinary Service Laboratory for serotyping. Sampling of environmental components (water, pen soil, individual feed ingredients, prepared rations, and fecal drop samples) were conducted at each of the three feeding locations. Samples of each component were collected in triplicate prior to initial placement of calves in preconditioning pens at the start of the backgrounding/stocker feeding stage and at the conclusion of their time in the preconditioning (approximately 45 d). After preconditioning, calves at the commercial feeding locations spent 120 to 130 d on pasture before being placed in feedlot pens; feed samples were obtained on d0 and upon conclusion of the stocker phase. Collection of all environmental components then were repeated prior to initial placement in feedlot pens, and again every 30d until market weight is reached. Environmental samples were subjected to Salmonella procedures as previously described. Additionally, pen soil samples were submitted to the Texas A&M AgriLife Extension Soil, Water and Forage Testing Laboratory for routine analysis including micronutrients (pH, NO3-N, Conductivity and Mehlich III by ICP P, K, Ca, Mg, Na, S, DTPA ZN, FE, CU, and Mn).

Although 5.0% of samples from weanling heifer carcasses returned Salmonella-positive results, all other LN samples analyzed from cattle managed at McGregor were found to be Salmonella-negative. As expected, LNs samples obtained from carcasses of cattle managed at Location A returned low proportions (0/20 LN samples per stage) of Salmonella-positive LN samples across all feeding stages. At Location B, 35.0 (7/20), 20.0 (4/20), and 35.0 (7/20)% Salmonella-positive samples were confirmed from stages 2, 3, and 4, respectively. Overall, 30% of LN samples from Location B returned a Salmonella-positive result. Salmonella-positive results did not differ (P > 0.05) across feeding stages for Location B. However, higher proportions of Salmonella-positive LN samples were seen for stages 2 and 4 when compared to other feeding locations.   

Overall, rate of Salmonella-positive environmental samples differed by location (P < 0.017). McGregor returned the lowest rate (11.5%), followed by Location A (23.6%), and Location B at 41.0%. At McGregor, Salmonella-positive samples were most often collected from pen soil (33.3%) and trough water (42.4%), while Salmonella was rarely (3.7 and 0.8% for feces and individual feed ingredients, respectively) or never (0%; prepared ration from the bunk) recovered from samples in the other environmental component categories. Salmonella-positive samples were recovered for all sample types at both Locations A and B. The largest (P < 0.017) quantity of Salmonella-positive pen soil and freshly voided fecal samples were recovered from Location B. Salmonella presence in individual feed components collected from each location prior to ration mixing varied (P < 0.017) based on location. Salmonella-positive individual feed ingredients were least common at McGregor (0.8%), occasional at Location B (12.1%), and recovered most frequently from Location A (28.2%). This trend does not translate directly to prepared rations as sampled from the feed bunk. The highest percentage discrepancy between these two categories is seen for Location B, as prepared ration samples returned Salmonella-positive results 46.2% of the time. This differs (P < 0.017) from the 0% Salmonella-positive sample recovery documented for McGregor, but not (P > 0.05) from Location A (24.2%).   

In total, 36 unique serovars were identified from environmental component samples. Six serovars comprised 61.83% (128/207) of total serovars reported, these were: Anatum (18.36 %), 6,7:g,m,s:e,n,z15 (11.59%), Montevideo (11.11%), Muenchen (8.21%), Mbandaka (7.25%) and Cerro (5.31%). Notably however, is the proportion of these data that are represented by serovars isolated from environmental component samples at Location B. The same six serovars comprise nearly 81% (97/120) of serovars reported for environmental samples from this Location B. This equates to over 75% (97/128) of the total for these six serovars isolated from environmental samples across all locations. From McGregor, Meleagridis is the only serovar recovered from both a LN and environmental components (1 LN; 1 feces; 7 pen soil). For Location B, three serovars were isolated from both LNs and environmental samples: Anatum (11 LN; 26 environmental), Muenchen (1 LN; 16 environmental), and Cerro (1 LN; 11 environmental). Overall, serovars varied by location and sample type with minimal overlap between LN and environmental samples.

Previous studies have shown differing levels of Salmonella prevalence in bovine lymph nodes due to seasonality, geographic location, and feedyard environment. By overlaying Salmonella prevalence in bovine lymph nodes with prevalence in environmental components associated with general beef production, future research and intervention development and implementation can be better planned for. Data from this study provide insight into Salmonella prevalence differences among previously evaluated feeding locations and the possible influence of environmental and/or management practices at each. While the goal for both researchers and the industry would be prevention or mitigation of salmonellae uptake by LNs of cattle in commercial feeding environments, a more practical outcome related to this work would be the development of best practices to reduce Salmonella prevalence in cattle feeding operations. This would result in decreased prevalence of Salmonella in lymph nodes, and thereby, decreased risks to human health.