Search
Author
Title
Vol.
Issue
Year
1st Page

Journal of Animal Science - Article

 

 

This article in JAS

  1. Vol. 90 No. 6, p. 1929-1939
     
    Received: Aug 15, 2011
    Published: January 20, 2015


    2 Corresponding author(s): bwhite@vet.ksu.edu
 View
 Download

doi:10.2527/jas.2011-4599

Associations between the distance traveled from sale barns to commercial feedlots in the United States and overall performance, risk of respiratory disease, and cumulative mortality in feeder cattle during 1997 to 20091

  1. N. Cernicchiaro*,
  2. B. J. White 2,
  3. D. G. Renter*,
  4. A. H. Babcock*,
  5. L. Kelly and
  6. R. Slattery§
  1. *Department of Diagnostic Medicine/Pathobiology, Kansas State University, Manhattan 66506
    †Department of Clinical Sciences, Kansas State University, Manhattan 66506
    ‡Work completed while at Kansas State University, Manhattan 66506
    §Work completed while at Kansas State University, Manhattan 66506

Abstract

Most beef cattle are transported at least once during their lives, and this potentially stressful practice may affect subsequent health and performance. Limited research is available quantifying the effects of transport on feedlot performance and health, and particularly the risk of bovine respiratory disease complex (BRD), which is the most common disease of weaned calves after arrival to the feedlot. The objective of this retrospective study was to determine potential associations between distance traveled (DTV) during transportation with health (cumulative BRD morbidity and mortality of all causes) and performance (ADG and HCW) parameters in cattle cohorts (n = 14,601) that arrived to 21 U.S. commercial feedlots from 1997 to 2009. Multivariable mixed-effects negative binomial and linear regression models were employed to determine associations between health and performance outcomes with DTV and other cohort-level demographic variables. Cattle were transported a median of 552 km from origin to feedlot with a mean (± SEM) of 698 ± 4.4 km. The mean (±SEM) cumulative BRD morbidity was 4.9% ± 0.01% (median = 1.1%; range: 0 to 100%) whereas the mean (±SEM) cumulative mortality due to all causes was 1.3% ± 0.01% (median = 0.8%; range: 0 to 28.7%). Distance traveled was significantly associated (P < 0.05) with BRD morbidity, overall mortality, HCW and ADG, and its effects were modified by demographic characteristics (i.e., cohort region of origin, mean arrival BW, gender, and the season of the year) of the cohort. Knowledge of the distance traveled during transportation could allow a more precise prediction of cattle feedlot health and performance.



INTRODUCTION

Transportation of cattle between locations by truck is a routine management practice in the U.S. beef cattle industry. Cattle handling and commingling during transportation, in addition to pre-transit production activities surrounding feeder cattle (e.g., weaning, handling, re-grouping), have been associated with increased risk of morbidity and mortality in calves, shrink BW loss, decreases in carcass quality, and increased economic costs (Ribble et al., 1995b; Grandin, 1997; Swanson and Morrow-Tesch, 2001; Fike and Spire, 2006; White et al., 2009).

Accurately predicting the expected bovine respiratory disease complex (BRD) incidence in a cohort is challenging, and the expected disease risk influences the implementation of interventions such as metaphylaxis utilization at arrival (Nickell and White, 2010). Previous research has illustrated that the distance cattle were shipped was positively associated with increased risk for BRD morbidity (Sanderson et al., 2008). Approximately 39% of feedlots with more than 8,000 cattle modify their initial processing procedures on the basis of the distance cattle traveled (USDA, 2000); however, there is little quantitative information to determine how data on characteristics of the journey can be used to predict health risks and expected performance. The objective of this retrospective observational study was to determine potential associations between distance traveled during transportation with cattle health (BRD morbidity and overall mortality risks) and performance (HCW and ADG) after feedlot arrival and to evaluate if the effects of distance traveled (DTV) on the health and performance outcomes varied by cohort demographics. Quantifying effects of the distance cattle traveled before arrival may be useful to augment current systems for classification of cattle as high or low risk for respiratory disease and may allow more accurate predictions of cattle performance.


MATERIALS AND METHODS

Our study population was a convenience sample from 21 feedlots with available operational data on cohort-level demographic, health, and performance variables. A cohort was considered as a group of cattle that were purchased and managed similarly but not necessarily allocated together Study inclusion criteria included: cohorts classified as either male or female (not mixed) with an arrival mean BW greater than 227 kg that contained more than 20 cattle and had available information on cattle origin. The primary transportation factor of interest was the DTV during the journey to the feedlot. Cohorts that did not have a listed origin were removed, and for all others, the distance from the origin to the feedlot was calculated by using the origin zip code and the known location of the feedlot using a commercial program (Microsoft MapPoint, Microsoft Corp., Redmond, WA).

Data available on each cohort at arrival included mean arrival BW, gender, arrival date, and the specific feedlot to which cattle arrived. The arrival date was used to create an arrival quarter variable (or season) for cattle that arrived in January through March (winter), April through June (spring), July through September (summer), and October through December (fall). The region of the country that cattle originated from was categorized using the geographic regions described in the most recent United States Department of Agriculture National Animal Health Monitoring System (USDA NAHMS) cow-calf survey (USDA, 2008) which divides the country in 4 regions on the basis of the beef population (Central, South-central, East, and West). A fifth region was created by including cohorts originated from the remaining U.S. states and Canada.

Health outcomes of interest included cumulative (over the entire feeding period) BRD morbidity and overall mortality risks. A BRD case was defined as an animal identified for the first time by feedlot personnel with clinical signs of BRD and subsequently treated with an antimicrobial. Overall mortality considered cattle death related to any cause between the time of arrival and final cohort closeout. Performance outcomes of interest included cohort-level ADG, calculated as total BW gain·number of shipped cattle−1·days on feed−1 over the entire feeding period and mean HCW (unchilled carcass weight taken shortly after slaughter).

Statistical Analyses

Associations between DTV and other cohort-level predictors with health outcomes (cumulative BRD morbidity and overall mortality risks) were evaluated using mixed-effects negative binomial regression models (xtnbreg, STATA 10, StataCorp LP, College Station, TX) with a random intercept for feedlot to account for the hierarchical structure of the data (cohorts within feedlots). Morbidity attributed to BRD was modeled as the count of initial respiratory disease cases in each cohort with the initial number of cattle per cohort at arrival modeled as the exposure variable. Overall mortality was modeled as the counts of deaths of any cause with the initial number of cattle per cohort as the exposure variable of the model. Mixed-effects linear regression models (xtmixed, STATA 10) using maximum likelihood estimation and a random intercept for feedlot were employed to evaluate performance outcomes (HCW, ADG).

During model building, initial main effects models were built including all independent variables. Manual backward elimination was then conducted until only statistically significant (Wald χ2 test, P < 0.05) main effects remained in the model. A correlation analysis was performed using the Spearman's rank correlation statistic to identify variables that may be collinear in the models. If the value of the correlation statistic between 2 independent variables was |0.8| or greater at a 5% significance level (P < 0.05), only 1 of the variables was selected for inclusion in the multivariable model based on biological plausibility or the completeness and quality of the data (Dohoo et al., 2009). The linearity assumption between the log odds of the predicted rate of the outcome (for negative binomial models) or the non-transformed outcome (for linear models) and continuous predictors was assessed using graphical methods (i.e., lowess smoothing of the outcome on the continuous predictor). If the assumption was not met, depending on the shape of these relationships, the predictor variable was categorized unless it was more appropriately transformed (Dohoo et al., 2009).

All possible 2-way interaction terms between the predictor of primary interest (DTV) and variables that might interact because of their potential confounding effect (i.e., mean arrival BW, gender, season, and region) were tested for statistical significance at the 5% level using manual forward selection. Variables that were part of significant interaction terms (P < 0.05) were retained in the model regardless of their individual levels of statistical significance.

Independent variables evaluated included distance in kilometers (categorized in the following categories: 0 to 250, 251 to 500, 501 to 750, 751 to 1000, and >1000 km), region [coded as: Central, South-central, East, West (USDA, 2008), and other States and Canada], mean arrival weight in kilograms (categorized as: 227 to 271, 272 to 317, 318 to 362, and > 362 kg), gender (male and female), season (or arrival quarter: winter = January to March, spring = April to June, summer = July to September, and fall = October to December), arrival year (1997 to 2009, with observations from 1997 to 2001 collapsed in 1 category due to the small number of cohorts in some of those years), and cohort size (or total number of cattle per cohort: 20 to 100, 101 to 150, 151 to 200, and > 200 cattle).

Distance traveled, which was recorded in a continuous scale, was categorized given it did not meet the linearity assumption in the BRD morbidity model. Categorization based on different cut-point values was attempted for this and other cohort-level variables, and the selection of appropriate categories was performed by comparing models including these variables as single covariates using Akaike's and Bayesian information criteria (AIC and BIC, respectively).

Diagnostics of residuals from the final multivariable models included the evaluation of the predicted values of the random effects (i.e., feedlot-level residuals) in the model or BLUP and Pearson and Deviance residuals for observations at the lowest level (i.e., cohort). Normal quantile plots of BLUP and residual plots of cohort-level residuals were visually examined to assess general model fit and to identify potential outliers and influential observations. Finally, probabilities and incidence rate ratios (IRR) and their respective 95% confidence intervals were estimated for predictors included in the final multivariable negative binomial models for cumulative BRD morbidity and overall mortality risks. Similarly, coefficients and 95% confidence intervals were computed for the mixed-effects linear regression models constructed for the ADG and HCW outcomes.


RESULTS

The final DTV dataset consisted of 14,601 cohorts from 21 feedlots located in the central and south plains of U.S. Descriptive statistics of continuous predictors and outcomes analyzed in this study are presented in Table 1.


View Full Table | Close Full ViewTable 1.

Descriptive statistics for continuous independent variables and outcomes observed for the study population1

 
Variable Mean SEM Median IQR2 Range
Distance traveled, km 697.9 4.44 552.0 338.0 to 951.1 0 to 3086.7
Mean arrival weight, kg 332.8 0.34 336.1 304.8 to 364.7 226.8 to 408.2
Days on feed, days 159.4 0.27 156.0 137.0 to 177.0 2.0 to 330.0
Cohort size at arrival, cattle 171.2 0.77 153.0 95.0 to 239.0 20.0 to 1439.0
Number of BRD3 first cases 8.4 0.13 2.0 0 to 10.0 0 to 185.0
BRD morbidity (number of first cases/cohort size at arrival), % 4.9 0.01 1.1 0 to 5.9 0 to 100
Number of deaths from any cause 1.8 0.02 1.0 0 to 3.0 0 to 64.0
Mortality of any cause (number of deaths/ cohort size), % 1.3 0.01 0.8 0 to 1.6 0 to 28.7
Mean HCW, kg 367.8 0.24 371.1 347.4 to 391.0 156.3 to 454.7
ADG, kg 1.4 0.002 1.4 1.3 to 1.5 0.2 to 2.4
1Data from 14,601 cohorts from 21 feedlots
2IQR (Interquartile range: Quartile 75 to Quartile 25)
3Bovine respiratory disease complex

Cattle were transported a median of 552 km from origin to feedlot with a mean (±SEM) of 698 ± 4.4 km. Forty-five percent (n = 6,635) of cohorts originated from the South-central region, followed by 23.6% (n = 3,442), 18.0% (n = 2,628), and 9.8% (n = 1,432) from the Central, East, and West regions, respectively. The remaining 3.2% (n = 462) of cohorts came from the other States not included in the regions comprised in the USDA classification of U.S. beef regions (USDA, 2008) and Canada. Cohorts were 71.7% (n = 10,472) and 28.3% (n = 4,129) male and female cohorts, respectively. The mean (±SEM) BRD morbidity risk was 4.9 ± 0.01% (median = 1.1%), with a range of 0 to 100% at the cohort level.

The main effects of DTV, region of origin, mean arrival BW, gender, season, cohort size, and arrival year were significantly associated (P < 0.05) with BRD morbidity risk; in addition, there were significant interactions between DTV and region, DTV and mean arrival BW, DTV and gender, and DTV with season (Table 2). Mortality risk due to all causes ranged from 0 to 28.7% among the DTV cohorts with a mean (±SEM) of 1.3 ± 0.01% and a median of 0.8%. The risk of overall mortality was significantly (P < 0.05) associated with DTV, region, mean arrival BW, gender, season, cohort size, and arrival year. In addition, 2-way interactions between DTV and region, DTV and mean BW at arrival, DTV and gender, and DTV and season were significantly associated (P < 0.05) with overall mortality risk (Table 2).


View Full Table | Close Full ViewTable 2.

Results of the final multivariable regression models of associations between distance traveled (DTV) and several demographic factors with bovine respiratory disease complex (BRD) morbidity, overall mortality, HCW, and ADG

 
BRD morbidity Mortality HCW, kg ADG, kg
Variable IRR 1 95% CI 2 IRR 95% CI Coef 3 95% CI Coef 95% CI
Distance traveled, km
    ≤250 ref4 ref ref ref ref ref ref ref
    251 to 500 1.70 1.25 to 2.31 1.43 1.10 to 1.86 6.14 1.85, 10.43 −0.01 −0.06, 0.03
    501 to 750 1.73 1.28 to 2.33 1.53 1.17 to 1.99 0.69 −3.51, 4.89 −0.05 −0.10, −0.01
    751 to 1000 1.91 1.41 to 2.58 2.11 1.62 to 2.74 6.64 2.18, 11.10 −0.04 −0.08, 0.005
    >1000 1.59 1.19 to 2.12 1.84 1.45 to 2.34 6.50 2.49, 10.51 −0.01 −0.05, 0.03
Region
    Central ref ref ref ref ref ref ref ref
    South-central 0.80 0.70 to 0.90 1.03 0.94 to 1.14 −0.13 −1.58, 1.32 0.008 −0.006, 0.02
    East 0.69 0.22 to 2.24 0.83 0.65 to 1.07 −5.08 −7.75, −2.41 0.08 0.05, 0.11
    West 1.24 1.09 to 1.42 0.99 0.88 to 1.11 7.90 6.17, 9.62 0.06 0.05, 0.08
    Other states/CAN 1.38 0.69 to 2.76 0.54 0.19 to 1.51 −3.40 −14.56, 7.75 0.02 −0.09, 0.13
Gender
    Female ref ref ref ref ref ref ref ref
    Male 1.13 1.01 to 1.27 1.16 1.06 to 1.28 35.40 34.02, 36.79 0.15 0.13, 0.16
Mean arrival BW, kg
    227 to 271 ref ref ref ref ref ref ref ref
    272 to 317 1.08 0.86 to 1.35 0.99 0.82 to 1.22 9.52 6.41, 12.63 0.006 −0.03, 0.04
    318 to 362 0.69 0.56 to 0.86 0.71 0.58 to 0.86 19.46 16.45, 22.48 0.05 0.02, 0.08
    >362 0.55 0.44 to 0.69 0.52 0.43 to 0.64 26.01 22.93, 29.10 0.09 0.06, 0.12
Cohort size, cattle
    20 to 100 ref ref ref ref ref ref ref ref
    101 to 150 0.80 0.75 to 0.85 0.94 0.89 to 0.99 −1.36 −2.07, −0.65 −0.01 −0.02, −0.007
    151 to 200 0.63 0.59 to 0.67 0.88 0.83 to 0.94 −2.77 −3.60, −1.95 −0.03 −0.04, −0.02
    >200 0.51 0.48 to 0.54 0.96 0.82 to 0.91 −3.01 −3.72, −2.31 −0.04 −0.04, −0.03
Season, months
    Winter (Jan to March) ref ref ref ref ref ref ref ref
    Spring (Apr to Jun) 0.83 0.73 to 0.93 0.86 0.78 to 0.96 0.80 −0.77, 2.36 0.03 0.01, 0.04
    Summer (Jul to Sep) 1.04 0.91 to 1.17 0.84 0.75 to 0.93 −2.38 −3.97, −0.79 −0.01 −0.03, 0.003
    Fall (Oct to Dec) 1.07 0.95 to 1.21 1.04 0.94 to 1.16 −7.95 −9.52, −6.39 −0.11 −0.12, −0.09
Year of arrival
    2001 ref ref ref ref ref ref ref ref
    2002 2.49 1.85 to 3.35 1.12 1.01 to 1.24 −0.73 −2.23, 0.76 0.01 −4 × 105, 0.03
    2003 8.04 6.08 to 10.62 1.20 1.08 to 1.32 −1.99 −3.49, −0.50 0.02 0.007, 0.04
    2004 7.24 5.46 to 9.61 1.25 1.13 to 1.38 3.12 1.58, 4.66 0.003 −0.01, 0.02
    2005 2.93 2.17 to 3.96 1.05 0.94 to 1.16 9.38 7.82, 10.94 0.02 0.007, 0.04
    2006 6.92 5.12 to 9.36 1.21 1.08 to 1.35 8.94 7.23, 10.64 −0.04 −0.06, −0.03
    2007 26.44 20.28 to 34.47 1.30 1.17 to 1.44 11.92 10.33, 13.52 −0.02 −0.03, −0.0003
    2008 26.96 20.65 to 35.20 1.14 1.02 to 1.27 19.80 18.05, 21.54 0.05 0.04, 0.07
    2009 34.21 26.05 to 44.92 1.43 1.26 to 1.61 27.13 25.14, 29.12 0.07 0.05, 0.09
DTV*Region
    ≤250*Central ref ref ref Ref ref ref ref ref
    251 to 500*Central 1.70 1.25 to 2.31 1.43 1.10 to 1.86 6.14 1.85, 10.43 −0.01 −0.06, 0.03
    501 to 750*Central 1.73 1.28 to 2.33 1.53 1.17 to 1.99 0.69 −3.51, 4.89 −0.05 −0.10, −0.01
    751 to 1000*Central 1.91 1.41 to 2.58 2.11 1.62 to 2.74 6.64 2.18, 11.10 −0.04 −0.08, 0.005
    >1000*Central 2.59 1.19 to 3.12 1.84 1.45 to 2.34 6.50 2.48, 10.51 −0.01 −0.05, 0.03
    ≤250*S-central ref ref ref ref ref ref ref ref
    251 to 500* S-central 1.62 1.21 to 2.17 1.51 1.17 to 1.95 1.42 −2.80, 5.63 −0.07 −0.11, −0.03
    501 to 750*S-central 1.96 1.47 to 2.62 1.37 1.05 to 1.78 −1.03 −5.24, 3.17 −0.09 −0.13, −0.05
    751 to 1000*S-central 1.50 1.13 to 2.00 1.57 1.21 to 2.03 3.63 −0.68, 7.94 −0.07 −0.11, −0.03
    >1000*S-central 1.01 0.77 to 1.43 1.23 0.96 to 1.59 −3.73 −7.79, 0.34 −0.15 −0.19, −0.10
    ≤250*East ref ref ref ref ref ref ref ref
    251 to 500*East na5 na na na na na na na
    501 to 750*East na na na na 15.49 9.94, 21.03 na na
    751 to 1000*East 2.23 0.89 to 6.73 1.93 1.05 to 2.65 na na na na
    >1000*East 1.69 0.51 to 5.61 1.70 1.21 to 2.40 −0.03 −4.55, 4.49 −0.16 −0.21, −0.11
    ≤250*West ref ref ref ref ref ref ref ref
    251 to 500*West 0.96 0.70 to 1.32 1.26 0.96 to 1.67 −5.55 −10.02, −1.09 −0.10 −0.15, −0.06
    501 to 750*West 0.50 0.36 to 0.70 1.06 0.79 to 1.41 −7.80 −12.54, −3.45 −0.14 −0.18, −0.09
    751 to 1000*West 0.98 0.65 to 1.27 2.04 1.53 to 2.71 −1.74 −6.57, 3.09 −0.11 −0.16, −0.06
    >1000*West 1.77 0.55 to 2.28 1.42 1.08 to 1.88 −8.75 −13.11, −4.38 −0.08 −0.12, −0.03
    ≤250*Other /CAN ref ref ref ref ref ref ref ref
    251 to 500*Other /CAN 0.82 0.39 to 1.73 1.92 0.52 to 7.09 17.90 2.86, 32.9 0.16 0.009, 0.31
    501 to 750*Other /CAN 0.40 0.16 to 5.11 3.84 1.12 to 13.14 25.89 11.25, 40.54 −0.09 −0.24, 0.05
    751 to 1000*Other /CAN 0.41 0.19 to 0.90 2.24 0.76 to 6.60 1.08 −10.99, 13.16 −0.14 −0.27, −0.02
    >1000*Other /CAN 3.00 0.99 to 9.02 2.96 1.02 to 8.59 8.45 −3.37, 20.27 −0.04 −0.16, 0.08
DTV*Gender
    ≤250*Male vs. ≤250*Female 1.13 1.01 to 1.27 1.16 1.06 to 1.28 35.40 34.02, 36.79 0.15 0.13, 0.16
    251 to 500*Male vs. 251 to 500*Female 1.12 1.02 to 1.23 1.04 0.96 to 1.13 35.98 34.76, 37.21 0.13 0.12, 0.14
    501 to 750*Male vs. 501 to 750*Female 1.30 1.17 to 1.44 1.13 1.03 to 1.23 33.57 32.23, 34.91 0.13 0.11, 0.14
    751 to 1000*Male vs. 751 to 1000*Female 1.27 1.13 to 1.42 1.04 0.94 to 1.15 33.89 32.26, 35.51 0.14 0.13, 0.16
    >1000*Male vs. >1000*Female 1.44 1.32 to 1.57 1.19 1.11 to 1.26 30.86 29.75, 31.97 0.10 0.09, 0.11
DTV*Arrival BW
    ≤250*227 to 271 ref ref ref ref ref ref ref ref
    ≤250*272 to 317 1.08 0.86 to 1.35 0.99 0.82 to 1.22 9.52 6.41, 12.63 0.006 −0.03, 0.04
    ≤250*318 to 362 0.69 0.56 to 0.86 0.71 0.58 to 0.86 19.46 16.44, 22.48 0.05 0.02, 0.08
    ≤250*>362 0.55 0.43 to 0.69 0.52 0.43 to 0.64 26.01 22.93, 29.10 0.09 −0.12, −0.06
    251 to 500*227 to 271 ref ref ref ref ref ref ref ref
    251 to 500*272 to 317 0.83 0.70 to 0.90 0.84 0.73 to 0.98 10.77 8.07, 13.48 0.09 0.06, 0.12
    251 to 500*318 to 362 0.53 0.45 to 0.63 0.55 0.47 to 1.58 18.77 16.13, 21.42 0.15 0.12, 0.18
    251 to 500*>362 0.36 0.31 to 0.43 0.38 0.33-0.45 26.32 23.61, 29.03 0.21 0.18, 0.23
    501 to 750*227 to 271 ref ref ref ref ref ref ref ref
    501 to 750*272 to 317 0.73 0.63 to 0.86 0.72 0.62 to 0.84 12.11 9.61, 14.62 0.10 0.07, 0.12
    501 to 750*318 to 362 0.47 0.40 to 0.54 0.54 0.47 to 0.63 22.03 19.58, 24.48 0.16 0.13, 0.18
    501 to 750*>362 0.34 0.28 to 0.40 0.36 0.31 to 0.43 29.10 26.51, 31.69 0.21 0.18, 0.23
    751 to 1000*227 to 271 ref ref ref ref ref ref ref ref
    751 to 1000*272 to 317 0.60 0.52 to 0.69 0.57 0.50 to 0.65 8.39 6.05, 10.73 0.09 0.07, 0.11
    751 to 1000*318 to 362 0.40 0.35 to 0.47 0.42 0.36 to 0.48 18.57 16.21, 20.93 0.15 0.12, 0.17
    751 to 1000*>362 0.32 0.27 to 0.38 0.32 0.27 to 0.37 24.15 21.51, 26.80 0.17 0.14, 0.19
    >1000*227 to 271 ref ref ref ref ref ref ref ref
    >1000*272 to 317 0.78 0.70 to 0.87 0.74 0.39 to 0.81 11.24 9.74, 12.73 0.08 0.07, 0.10
    >1000*318 to 362 0.49 0.44 to 0.56 0.50 0.46 to 0.55 21.05 19.49, 22.61 0.15 0.13, 0.16
    >1000*>362 0.35 0.30 to 0.41 0.36 0.32 to 0.41 30.51 28.59, 32.44 0.22 0.20, 0.24
DTV*Season
    ≤250*Winter ref ref ref ref ref ref ref ref
    ≤250*Spring 0.83 0.73 to 0.93 0.86 0.78 to 0.96 0.79 −0.76, 2.36 0.03 0.01, 0.04
    ≤250*Summer 1.04 0.91 to 1.17 0.84 0.75 to 0.93 −2.38 −3.97, −0.79 −0.01 −0.03, 0.003
    ≤250*Fall 1.07 0.95 to 1.21 1.04 0.94 to 1.16 −7.95 −9.52, −6.39 −0.11 −0.12, −0.09
    251 to 500*Winter ref ref ref ref ref ref ref ref
    251 to 500*Spring 0.83 0.75 to 0.92 0.85 0.78 to 0.92 −0.57 −1.90, 0.75 0.02 0.007, 0.03
    251 to 500*Summer 1.02 0.93 to 1.12 0.90 0.83 to 0.98 −3.38 −9.38, −6.39 −0.02 −0.04, −0.01
    251 to 500*Fall 0.97 0.87 to 1.10 1.09 0.99 to 1.20 −7.88 −1.05, 2.63 −0.10 −0.12, −0.09
    501 to 750*Winter ref ref ref ref ref ref ref ref
    501 to 750*Spring 0.74 0.64 to 0.84 0.82 0.73 to 0.93 0.79 −1.05, 2.63 0.03 0.01, 0.05
    501 to 750*Summer 0.90 0.80 to 1.01 0.85 0.76 to 0.94 −2.01 −3.66, −0.36 −0.01 −0.03, 0.007
    501 to 750*Fall 1.11 0.98 to 1.25 1.10 0.98 to 1.23 −5.43 −7.18, −3.67 −0.09 −0.11, −0.07
    751 to 1000*Winter ref ref ref ref ref ref ref ref
    751 to 1000*Spring 1.11 0.95 to 1.30 0.92 0.80 to 1.06 −0.53 −2.76, 1.70 0.01 −0.01, 0.03
    751 to 1000*Summer 1.35 1.17 to 1.55 1.11 0.98 to 1.26 −4.50 −6.62, −2.38 −0.06 −0.08, −0.04
    751 to 1000*Fall 1.16 1.00 to 1.34 1.11 0.98 to 1.26 −5.58 −7.72, −3.44 −0.01 −0.12, −0.07
    >1000*Winter ref ref ref ref ref ref ref ref
    >1000*Spring 0.96 0.83 to 1.10 0.83 0.75 to 0.93 3.49 1.75, 5.24 0.04 0.02, 0.05
    >1000*Summer 1.14 1.02 to 1.28 1.14 1.04 to 1.25 −1.65 −3.20, −0.09 −0.04 −0.06, −0.03
    >1000*Fall 0.92 0.82 to 1.04 1.06 0.97 to 1.16 −5.89 −7.42, −4.35 −0.09 −0.10, −0.07
1IRR = Incidence Rate Ratio
2CI = Confidence Interval
3Coef = Coefficient
4ref: referent category
5na: estimates not available

For cattle originating from the Central region, DTV was associated with a significantly (P < 0.05) greater incidence risk of BRD morbidity and overall mortality for DTV longer than 250 km, compared with cattle originated from the Central region traveling shorter distances (<250 km). For cattle originated from the South-central region, the incidence risk of BRD morbidity was greater for DTV categories of 250 to 750 km and less for DTV of >750 km compared with cattle from the same region traveling <250 km. However, the risk of overall mortality was greater for DTV longer than 250 km compared with cattle from the South-central region traveling distances <250 km. For cattle originated from the West region, conversely, the risk of BRD morbidity was decreased for DTV of 251 to 1000 km compared with cattle from the same region traveling <250 km. However, for cattle originated from the West region the risk of overall mortality was significantly (P < 0.05) greater after traveling longer distances (>750 km) compared with cattle originating from the same region with DTV less than 250 km. The BRD morbidity and overall mortality risks for cattle originated from other States and Canada varied as the DTV increased, with both risks being greater for distances longer than 1000 km compared with cattle originated from the same region with DTV shorter than 250 km (Table 2, Figures 1 and 2).

Figure 1.

Model predicted mean bovine respiratory disease complex (BRD) morbidity risk by distance traveled from origin to feedlot (DTV) and region of origin from a mixed-effects negative binomial regression model. Region: Central: Iowa, Kansas, Missouri, Nebraska, North Dakota, South Dakota; South-central: Oklahoma, Texas; East: Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi; West: California, Colorado, Idaho, Montana, New Mexico, Oregon, Wyoming, Tennessee, Virginia. Source: USDA 2008, Beef 2007-08, Part I: Reference of beef cow-calf management practices in the United States, 2007-08 USDA-APHIS-VS, CEAH, Fort Collins, CO #N512-1008. Other States and Canada: Arizona, Illinois, Indiana, North Carolina, Nevada, Ohio, Pennsylvania, South Carolina, Utah, Washington, Wisconsin, West Virginia, Canada (Alberta, Ontario and Saskatchewan). Negative binomial regression model included significant main effects of DTV, region, arrival weight class, gender, season, cohort size, arrival year, and interactions between region and DTV, gender and DTV, arrival weight and DTV, and season with DTV. The model also included a random effect (intercept) to account for lack of independence among cohorts (n = 14,601) within feedlots (n = 21) and an exposure variable corresponding to the initial number of cattle per cohort.

 
Figure 2.

Model predicted mean overall mortality risk by distance traveled from origin to feedlot (DTV) and region of origin from a mixed-effects negative binomial regression model. Region: Central: Iowa, Kansas, Missouri, Nebraska, North Dakota, South Dakota; South-central: Oklahoma, Texas; East: Alabama, Arkansas, Florida, Georgia, Kentucky, Louisiana, Mississippi; West: California, Colorado, Idaho, Montana, New Mexico, Oregon, Wyoming, Tennessee, Virginia. Source: USDA 2008, Beef 2007-08, Part I: Reference of beef cow-calf management practices in the United States, 2007-08 USDA-APHIS-VS, CEAH, Fort Collins, CO #N512-1008. Other States and Canada: Arizona, Illinois, Indiana, North Carolina, Nevada, Ohio, Pennsylvania, South Carolina, Utah, Washington, Wisconsin, West Virginia, Canada (Alberta, Ontario, and Saskatchewan). Negative binomial regression model included significant main effects of DTV, region, arrival weight class, gender, season, cohort size, arrival year, and interactions between region and DTV, gender and DTV, arrival weight and DTV, and season with DTV. The model also included a random effect (intercept) to account for lack of independence among cohorts (n = 14,601) within feedlots (n = 21) and an exposure variable corresponding to the initial number of cattle per cohort.

 

Heavier BW cattle (>272 kg) showed significantly (P < 0.05) reduced incidence risks of BRD morbidity and overall mortality across the different DTV categories compared with their lighter counterparts (227 to 271 kg) traveling the same distances (Table 2). Although an interaction was identified between gender and DTV, male cohorts showed significantly (P < 0.05) greater risks of BRD morbidity and overall mortality across the different categories of DTV compared with female cohorts (Table 2). Increasing DTV affects BRD morbidity similarly for cattle arriving to the feedlot during winter, summer, and fall months (Figure 3). A significantly (P < 0.05) greater risk of BRD morbidity in cattle arriving during summer (July through September) months was identified after traveling distances longer than 750 km compared with cattle arriving during winter months traveling the same distances (Table 2). The data indicated that increasing DTV affects the overall mortality risk similarly for 3 of the seasons (winter, fall, and spring), but for the summer season, there appears to be a much more dramatic increase once DTV is above a threshold (between 500 and 750 km; Figure 4).

Figure 3.

Model predicted mean bovine respiratory disease complex (BRD) morbidity risk by distance traveled from origin to feedlot (DTV) and season of arrival from a mixed-effects negative binomial regression model. Seasons were based on months as follows: winter = January through March; spring = April through June; summer = July through September; fall = October through December. Negative binomial regression model included significant main effects of DTV, region, arrival weight class, gender, season, cohort size, arrival year, and interactions between region and DTV, gender and DTV, arrival weight and DTV, and season with DTV. The model also included a random effect (intercept) to account for lack of independence among cohorts (n = 14,601) within feedlots (n = 21) and an exposure variable corresponding to the initial number of cattle per cohort.

 
Figure 4.

Model predicted mean overall mortality risk by distance traveled from origin to feedlot (DTV) and season of arrival from a mixed-effects negative binomial regression model. Seasons were based on months as follows: winter = January through March; spring = April through June; summer = July through September; fall = October through December. Negative binomial regression model included significant main effects of DTV, region, arrival weight class, gender, season, cohort size, arrival year, and interactions between region and DTV, gender and DTV, arrival weight and DTV, and season with DTV. The model also included a random effect (intercept) to account for lack of independence among cohorts (n = 14,601) within feedlots (n = 21) and an exposure variable corresponding to the initial number of cattle per cohort.

 

The final multivariable models for HCW and ADG included the following significant (P < 0.05) variables: DTV, region, mean arrival BW, gender, season, cohort size, year of arrival and 2-way interactions between DTV and region, DTV and mean arrival BW, DTV and gender, and DTV with season (Table 2).

Cattle that originated from the West region showed significantly (P < 0.05) lighter HCW with greater categories of DTV compared with those hauled less than 250 km. However, the HCW observed in cattle originating from the Central and East regions and from other States and Canada were significantly (P < 0.05) greater as DTV increased compared with cattle originating from the same regions traveling shorter distances (<250 km) (Table 2). Contrarily, ADG in cattle from all regions was significantly (P < 0.05) less with greater DTV compared with cattle from the same regions traveling distances shorter than 250 km (Table 2).

Despite the presence of a significant (P < 0.05) interaction term between DTV and mean arrival BW, heavier weight cattle at arrival had greater HCW and ADG values than their lighter counterparts (227 to 271 kg) across all the different categories of DTV (Table 2). Similarly, male cohorts showed significantly (P < 0.05) greater HCW and ADG across all DTV categories compared with female cohorts traveling the same distances (Table 2).

Cattle arriving at the feedlots in fall and summer displayed significantly (P < 0.05) lighter HCW and ADG values across all categories of distance traveled compared with winter cattle that were hauled the same distances. In contrast, cattle arriving in the spring showed significantly (P < 0.05) greater HCW for DTV longer than 1000 km and greater ADG values across all categories of distance, compared with cattle arriving in winter months that traveled less than 250 km (Table 2).


DISCUSSION

Results from this study illustrate that knowledge of DTV during transport may be useful for more accurately describing the expected health and performance of specific cohorts of feeder cattle. Furthermore, these data indicated that BRD morbidity, mortality due to all causes, HCW, and ADG were associated with DTV, but the effects were modified by the region the cattle originated from, cohort gender, cohort mean arrival BW, and the season cattle arrived at the feedlot.

The increase in stress after transportation has been documented through work showing a modification of the acute phase protein response (Arthington et al., 2003) and an increase in cortisol release (Buckham Sporer et al., 2008). These changes in physiologic function also influence the ability of the calf to respond to disease challenge as immune function may be impaired after transportation because of decreased leukocyte numbers (Stanger et al., 2005) and an increased neutrophil:leukocyte ratio (Kent and Ewbank, 1986; Murata, 1989). Moreover, highly stressed, weaned, and transported beef cattle are at greater risk of experiencing increased morbidity and mortality due to respiratory disease (Ribble et al., 1995b). Implementation of specific health interventions such as metaphylaxis is often based on the predicted disease risk of a cohort (Nickell and White, 2010); therefore, more accurate health risk predictions at feedlot arrival are important for the feeding industry to reduce economic losses associated to respiratory disease and to improve animal welfare.

Transportation stress may cause transient changes in physiologic indices (Stanger et al., 2005; Gupta et al., 2007) or BRD risk (White et al., 2009); yet our current work indicates that the effect of DTV also can be associated with cumulative BRD morbidity risk. Previous work illustrates a cumulative effect of DTV on health (Mormede et al., 1982; Pinchak et al., 2004), and a study of commercial U.S. feedlots found that the risk of BRD morbidity increased by 10% for each additional 100 miles traveled to the feedlot (Sanderson et al., 2008). Although our study also found positive associations between DTV and BRD morbidity, we categorized DTV and found the effect was neither linear nor consistent across all cohort types and that there seemed to be thresholds of biological importance in DTV when evaluating health and performance variables. The indication that DTV may not have a linear effect on animal health is supported by previous research where measures of transport stress plateaued during both simulated (Sartorelli et al., 1992) and actual transportation (Warriss et al., 1995). In our study, the DTV variable was categorized on the basis of approximately equal ranges of distance with similar distribution of cohorts among those categories. This is one of the first evaluations where the distance variable was categorized; however, the selected DTV categories were somewhat arbitrary as there are no actual biological cut-points to define this variable. In one report, no differences in risk of fatal pneumonia were identified on the basis of distance traveled; however, the data used in this research are greater than 20 yr old and production systems and transport effects were likely different (Ribble et al., 1995b). The findings of our study indicated that distance traveled was a significant predictor of BRD morbidity and overall mortality risks; however, distance may be a proxy of other predictors such as placement on the truck (White et al., 2009), handling of cattle during loading and unloading, ventilation (and fume levels), stocking density, food and water deprivation, commingling, sanitation of the vehicles, climatic conditions (e.g., temperature and humidity), or driving conditions, which may be the actual stressors (Eicher, 2001; Swanson and Morrow-Tesch, 2001).

Distance traveled was significantly associated with all health and performance outcomes, but the effect of DTV varied by specific cohort characteristics as indicated by the presence of several significant interaction effects between DTV and cohort-level demographic variables. A common perception exists that cattle health may be influenced not only by DTV but also by the region of origin of the cattle, and this study is the first report quantifying how region of origin modifies the effect of DTV on BRD morbidity and overall mortality risks. Differences in health and performance variables in cattle originating from varied regions may be related to environmental (e.g., weather conditions) or management (e.g., conditioning practices) factors that are specific to each region. Cattle originated from the Central and South-central regions, regions that are in close proximity to the feeding industry in the study area, tended to travel, on average, shorter distances (mean distance ± SEM = 460.8 ± 5.6 km, 503.7 ± 3.4 km, and 515.9 ± 11.8 km for cattle originated from the Central, South-central, and West regions, respectively) than cattle originated from the East and other States and Canada (mean distance ± SEM = 1,510.1 ± 9.4 km and 1,198.3 ± 18.3 km, respectively); however, the risks of BRD morbidity and overall mortality differed by the distance traveled across all regions. Thus, the stress of transportation may affect health outcomes irrespective of the region of origin of the cattle after traveling long distances. Nonetheless, the values of HCW observed in cattle that originated from certain regions (e.g., East and other States and Canada) was significantly greater with longer traveling distances (<250 km). Although this seems counterintuitive, animals may be able to adjust their BW gain after traveling long distances as previous evidence indicates the major physiological changes and losses in BW occur during the early stages of transit (Sartorelli et al., 1992; Barnes et al., 2004). Knowledge of cattle movement, area of origin, and distance hauled may be important indicators of the risk profile of cattle and might help preventing disease and performance losses by adjusting feedlot protocols at the arrival of the cohort.

Male cohorts showed significantly greater risks of BRD morbidity and overall mortality and greater HCW and ADG values across all DTV categories compared with female cohorts traveling the same distances. Similarly, previous studies have shown that steers had increased respiratory morbidity (Alexander et al., 1989; Snowder et al., 2006) and mortality (Cusack et al., 2007) risks, compared with heifers. In our dataset, male cattle were not classified as steers or bulls; therefore, we were unable to account for potential health risks associated with castration. The stress associated with castration has been indicated as a risk factor for increased morbidity (Pinchak et al., 2004; Snowder et al., 2006), and the increase in morbidity and mortality risks in our study for male cohorts across the DTV categories may be the result of an additive effect of stresses from transport and castration.

Mean cohort arrival BW is considered an important factor for predicting subsequent health risks (Lechtenberg et al., 1998; Babcock et al., 2010). In our study, the effect of mean arrival BW was significantly associated with BRD morbidity and overall mortality, but effects depended on the DTV; heavier BW cattle showed significantly decreased risks of BRD morbidity and overall mortality across the different DTV compared with their lighter counterparts. These findings indicate that the stresses of transport may be more pronounced in lighter cattle. Therefore, changes in disease risk may occur with shorter DTV, or there may be underlying problems with the lighter cattle before shipping (e.g., being younger or having a poor performance) that are exacerbated by transport stress. Arrival BW is often considered a rough proxy for animal age, and previous authors have suggested that young cattle may be more susceptible to the stress of transportation likely because of an incomplete development of the hypothalamo-pituitary-adrenal axis (Eicher, 2001; Swanson and Morrow-Tesch, 2001). In addition, after short-term road transport, young cattle are more susceptible to respiratory diseases because of an alteration in the number and function of immune cells from the bronchoalveolar fluid (Ishizaki et al., 2005).The effect of DTV on health and performance outcomes also depended on the season that cattle arrived at the feedlot. Consistent with previous reports (Ribble et al., 1995a,b), we identified a significantly (P < 0.05) greater risk of BRD morbidity in cattle arriving during summer months after traveling long distances (>750 km), compared with cattle arriving in winter; however, the overall trend was that as DTV increased, the risk of BRD morbidity also increased. Similarly, increasing DTV affected the overall mortality risk in all seasons, with cattle arriving during summer months experiencing a dramatic increase with DTV greater than 500 km. These changes may be related to the fact that associations have been observed between daily temperature fluctuations and the occurrence of BRD (Cusack et al., 2007). Season could be a proxy for many things including weather conditions such as temperature, precipitation, wind, and other weather-related events; thus, perhaps it is not surprising that fall arrivals displayed significantly decreased HCW and ADG. The summer arrivals traveling long distances also may have been subject to adverse heat events or prolonged dehydration, both of which might contribute to greater morbidity and mortality and decreased HCW and ADG.

Some limitations of this study are related to the retrospective, observational nature of the study design; thus, when interpreting the results, it is important to recognize that direct causal inferences cannot be drawn between the characteristics of the journey (DTV) and subsequent outcomes. However, this study does offer the advantage of a good representation of cohorts from different regions of the country as well as the inclusion of cohorts of different size. The data were limited to available feedlot operational information, and the stated cattle origin may not have been the farm of origin. From these data, it is not possible to estimate the total DTV of some of the cohorts; however, our findings are directly applicable to the current level of data available to feedlot managers. Some of the cohorts in the study showed a cumulative BRD morbidity risk of 100%. On the basis of certain cohort characteristics at arrival, feedlot personnel may opt for treating all animals in the cohort (i.e., mass medication or metaphylaxis), even though not all animals are morbid. Since diagnosis of BRD is subjective and may not be consistent among feedlots, misclassification of the study outcome can occur. The existence of different case definitions related to management practices can affect disease detection. If knowledge of the DTV during transit affects the distribution of manpower or the allocation of cattle and that is associated with identification of BRD cases, it might affect the accuracy of diagnosing daily BRD incidence and bias perceived BRD morbidity. In addition, if cattle that were transported long distances were perceived as at increased risk by feedlot personnel, data on BRD morbidity may be biased as knowledge of distance traveled may lead to an increase in rates of BRD treatment. Mortality data are likely less biased than morbidity data; however, our data comprised information on overall mortality and not on BRD-specific mortality, as the majority of the study feedlots did not record information on the cause of death.

Our data demonstrated that the characteristics of the journey cannot be assessed independently in their association with health and performance variables as their effects depend on specific characteristics of the cohort (region the cattle originated from, cohort gender, cohort mean arrival weight, and the season cattle arrived at the feedlot). Practitioners and feedlot producers should be aware that multiple pre- and post-feedlot arrival effects should be taken into consideration when predicting health and performance in beef cattle early in the feeding period.

Knowledge of how DTV is associated with health and performance could allow more precise classification of feeder cattle cohorts. Thus, it may be possible to achieve more knowledge-based economic and management decisions, including timely application of disease intervention measures such as metaphylaxis, more informed purchasing and marketing decisions, or implementation of differential purchase price based on risk of respiratory disease and mortality or projected performance of cattle that were transported short distances. Further research on transportation characteristics would be useful for determining potential causal associations between transit and detrimental health and performance variables as well as for identifying interventions to mitigate transportation stress and its consequences.

 

References

Footnotes