Longitudinal assessment of Salmonella enterica prevalence, incidence, and resistance patterns in relation to feedlot cattle management, health, and food safety

Experiment 1: The objective was to investigate the influence of cattle origin and region of finishing on the prevalence of Salmonella, Escherichia coli O157:H7, and select antimicrobial resistance in E. coli populations. Yearling heifers (n = 190) were used in a 2 × 2 factorial arrangement. After determining fecal Salmonella prevalence, heifers were sorted into 1 of 4 treatments: heifers originating from SD and finished in SD (SD-SD); heifers originating from SD and finished in TX (SD-TX); heifers originating from TX and finished in SD (TX-SD); heifers originating from TX and finished in TX (TX-TX). Fecal, pen, and water scum line (WSL) samples were collected longitudinally throughout the study; hide swab and subiliac lymph node (SLN) samples were collected at study end. A treatment × time interaction was observed (P ≤ 0.01) for fecal Salmonella prevalence, with prevalence being greatest for TX-TX and TX-SD heifers before transport. From d 14 through study end, prevalence was greatest for TX-TX and SD-TX heifers compared with SD-SD and TX-SD heifers. Salmonella prevalence on hides was greater (P ≤ 0.01) for heifers finished in TX compared with SD. Salmonella prevalence in SLN tended (P = 0.06) to be greater in TX-TX and SD-TX heifers compared with TX-SD and SD-SD. Fecal E. coli O157:H7 prevalence showed a treatment × time interaction (P = 0.04), with SD-TX prevalence being greater than TX-SD on d 56 and SD-SD and TX-TX being intermediate. A treatment × time interaction also was observed for fecal trimethoprim-sulfamethoxazole-resistant and cefotaxime-resistant E. coli O157:H7 prevalence (P ≤ 0.01). Overall, these data suggest region of finishing influences pathogenic bacterial shedding patterns, with the initial 14 d after feedlot arrival being critical for pathogen carriage. Experiment 2: The objective was to longitudinally assess the prevalence of F. necrophorum subsp. necrophorum and subsp. funduliforme, F. varium, and Salmonella enterica in the nasal cavity, ruminal fluid, and feces of finishing beef steers with and without liver abscesses (LA). Crossbred steers (n = 227; 353 ± 39.6 kg) were transported to a feedlot and fed a high-concentrate diet. Nasal, ruminal fluid, and fecal samples were collected following feedlot arrival (d 5), 1 wk after adaptation to a finishing diet (d 35), and the day before harvest (study end). Livers were collected at harvest and examined for LA, and cattle were subsequently assigned into either control or liver abscess groups. Overall LA prevalence was 18.7%. The concentration and prevalence of Salmonella decreased in ruminal fluid and increased in feces with days on feed (P < 0.01). Conversely, ruminal fluid prevalence of F. necrophorum subsp. necrophorum and F. varium increased with days on feed (P < 0.01). Fusobacterium abundance in ruminal fluid and feces was not indicative of LA development except for F. varium being more abundant in the ruminal fluid of steers with LA (P < 0.01). Abundance of F. necrophorum subsp. necrophorum was greater in abscessed liver tissue than healthy tissue (P = 0.03), although no other differences in bacterial abundance or prevalence were observed in livers. Fusobacterium and Salmonella prevalence in the nasal cavity, ruminal fluid, and feces were affected by days on feed, but their prevalence and abundance were not indicative of LA occurrence. Experiment 3: The objective was to investigate the effects of random metaphylaxis administered with tulathromycin at feedlot arrival to 0, 33, 66, and 100% of high-risk beef steers on health outcomes, growth performance, complete blood cell count, Salmonella prevalence, and antimicrobial use during a 35-d receiving period. Crossbred beef steers (n = 232; arrival BW = 213 kg ± 6.3 kg) were used in a generalized complete block design consisting of 2 source blocks, each with 4 BW blocks and 4 treatments. Steers were randomly assigned to pen within source block, and experimental treatments were: 1) negative control; s.c. injection with sterile saline (0M); 2) tulathromycin administered at random to 33% of steers within a pen (33M); 3) tulathromycin administered at random to 66% of steers within a pen (66M); or 4) conventional metaphylaxis with tulathromycin administered to 100% of steers within a pen (100M). Body weight, whole blood, and fecal samples were collected on d 0, 14, and 35. Metaphylactic treatment was included in the model as a fixed effect, and BW block within source block was included as a random effect. The percentage of steers treated for bovine respiratory disease once was 108% and 62.2% greater on average for 0M and 33M, respectively, than for 66M and 100M steers (P = 0.01). Neither BW on d 35 nor ADG, DMI, or G:F from d 0 to 35 differed among treatments (P ≥ 0.63). Only reticulocyte count and percentage of eosinophils differed among treatments (P ≤ 0.03), whereas most hematology variables were affected by day (P < 0.01). Fecal Salmonella concentration and prevalence (P ≥ 0.40) did not differ among treatments but increased from d 0 to 14 before equalizing through d 35 (P < 0.01). These data suggest that metaphylaxis can be administered to 66% of cattle without increasing bovine respiratory disease morbidity or decreasing growth performance. Moreover, random metaphylaxis to 66% of steers at arrival decreased the amount of active drug compound and the number of antimicrobial injections administered in each pen without increasing pathogen shedding compared with administering conventional metaphylaxis to 100% of the pen in high-risk beef steers.