Associations between immune competence phenotype and feedlot health and productivity in Angus cattle

Abstract

Genetic strategies aimed at improving general immune competence (IC) have the potential to reduce the incidence and severity of disease in beef production systems, with resulting benefits of improved animal health and welfare and reduced reliance on antibiotics to prevent and treat disease. Implementation of such strategies first requires that methodologies be developed to phenotype animals for IC and demonstration that these phenotypes are associated with health outcomes. We have developed a methodology to identify IC phenotypes in beef steers during the yard weaning period, which is both practical to apply on-farm and does not restrict the future sale of tested animals. In the current study, a total of 838 Angus steers, previously IC phenotyped at weaning, were categorized as low (n = 98), average (n = 653), or high (n = 88) for the IC phenotype. Detailed health and productivity data were collected on all steers during feedlot finishing, and associations between IC phenotype, health outcomes, and productivity were investigated. A favorable association between IC phenotype and number of mortalities during feedlot finishing was observed with higher mortalities recorded in low IC steers (6.1%) as compared with average (1.2%, P < 0.001) or high (0%, P = 0.018) IC steers. Disease incidence was numerically highest in low IC steers (15.3 cases/100 animals) and similar in average IC steers (10.1 cases/100 animals) and high IC steers (10.2 cases/100 animals); however, differences between groups were not significant. No significant influence of IC phenotype on average daily gain was observed, suggesting that selection for improved IC is unlikely to incur a significant penalty to production. The potential economic benefits of selecting for IC in the feedlot production environment were calculated. Health-associated costs were calculated as the sum of lost production costs, lost capital investment costs, and disease treatment costs. Based on these calculations, health-associated costs were estimated at AUS$103/head in low IC steers, AUS$25/head in average IC steers, and AUS$4/head in high IC steers, respectively. These findings suggest that selection for IC has the potential to reduce mortalities during feedlot finishing and, as a consequence, improve the health and welfare of cattle in the feedlot production environment and reduce health-associated costs incurred by feedlot operators.

Introduction

Maintaining consumer confidence is key to the sustainability of the beef industry. Consumers are increasingly concerned with the use of antibiotics to treat disease in food-producing animals and are demanding the highest standards of health and welfare for animals. The Australian feedlot industry is actively investing in research aimed at developing strategies to enhance animal welfare by reducing the incidence of disease in feedlot cattle and, consequently, reduce the use of antibiotics. To achieve this aim, a multifaceted approach is required which combines strategies to 1) reduce the pathogen load that animals are exposed to in their production environment through targeted environmental management, 2) minimize detrimental effects on immune system function imposed by stressful management procedures, 3) protect animals against disease using effective vaccination protocols and immune system stimulants, and 4) employ genetic strategies to improve the ability of animals to both resist disease and minimize the impact of disease should they become infected.

Genetic strategies aimed at improving general disease resistance have the potential to significantly reduce disease incidence and severity (Mallard et al., 2015). However, for such strategies to be implemented, phenotyping tools to allow targeted selection must first be developed. The majority of Angus seedstock beef cattle producers in Australia market their animals with accompanying estimated breeding values (EBVs) for economically important traits (https://www.angusaustralia.com.au). This allows potential purchasers to compare the genetic merit of individual animals both within and across herds. Currently, EBVs are available for multiple traits in several breeds including those related to growth, reproduction, carcass, temperament, and feed efficiency. Although EBVs related to resistance to specific disease are now available for some beef cattle breeds (such as high altitude pulmonary hypertension in American Angus cattle), EBVs for general health and fitness traits have not been available, thereby limiting the ability of beef producers to improve these traits through selection.

Selection for resistance to specific diseases has been successfully demonstrated in beef cattle, including selection for resistance to brucellosis (Adams and Templeton, 1993) and to cattle ticks (Frisch and O′Neill, 1998). However, selection for resistance to a specific disease has the potential to inadvertently increase the susceptibility to other diseases. This is because different arms of the adaptive immune system are more effective at eliminating different types of pathogens. For example, enhanced macrophage function, important for the control of intracellular pathogens, is associated with resistance to Brucella abortus, whereas high antibody production, important for the control of extracellular pathogens, is associated with susceptibility to the B. abortus (Price et al., 1990). Although we acknowledge that selection for resistance to specific diseases of economic importance will remain important, genetic strategies aimed at improving general disease resistance in beef cattle are urgently required to moderate any inadvertent selection pressure that unbalances immune system function. Such approaches are particularly appealing for managing multifactorial diseases, such as bovine respiratory disease (BRD), which are caused by variable combinations of infectious agents and are, therefore, difficult to control through vaccination alone. Further, the ever-increasing climatic variability and the associated threat of exposure to novel diseases in the production environment are best addressed by enhancing general disease resistance.

The concept of breeding for general disease resistance, by selecting farm animals with an enhanced ability to mount both antibody (Ab-IR) and cell-mediated immune responses (Cell-IR), was first proposed by Wilkie and Mallard (1999) and has been used successfully as a selection tool to reduce the incidence of disease in both pigs and dairy cattle (Mallard et al., 2015). We have developed a modified methodology, based on that described by Wilkie and Mallard (1999), to assess the immune competence (IC) or ability to mount both Ab-IR and Cell-IR of beef cattle, which is practical to apply on-farm without restricting the future sale of tested animals (Hine et al., 2019). Key modifications to the methodology of Wilkie and Mallard (1999) include the use of a commercial vaccine to induce measurable immune responses during testing and the overlaying of testing procedures on the yard weaning period to identify animals with an enhanced ability to mount an immune response while exposed to the stressors associated with yard weaning (Hine et al., 2019). Here, we describe associations between IC phenotype and feedlot performance, in terms of health and growth. We hypothesized that high IC steers would have reduced health-related issues and incur reduced health-related costs during feedlot finishing, when compared with their low IC counterparts, and that these health benefits would have a positive impact on growth performance in the feedlot.

Materials and Methods

Animals

Experimental procedures were approved by the CSIRO, Animal Ethics Committee Armidale, NSW, Australia (application number 12/31). A total of 1,149 calves were enrolled in the original study, comprising 1,062 Angus steers and 87 Angus heifers (as described previously, Hine et al., 2019). Of the steers, 978 were progeny of the Australian Angus Sire Benchmarking Program (ASBP), representing years 2 and 3 of the program (ASBP calves, birth year cohorts 1 & 2). The ASBP is a major initiative of Angus Australia with support from Meat & Livestock Australia (MLA) and industry partners that aims to generate progeny test data on contemporary Angus bulls, particularly for hard-to-measure traits, such as feed efficiency, carcass measurements, meat quality attributes, and female reproduction (https://www.angusaustralia.com.au/sire-benchmarking/about/general-information/). The ASBP steers were IC tested on-farm in cooperator herds during the weaning period at various locations across NSW, Australia. A total of 839 of these ASBP steers that were phenotyped for IC at weaning, and subsequently finished at commercial feedlots situated in Northern NSW, Australia, were used in this study.

Experimental design

The assessment of IC was undertaken on-farm during the weaning period when calves were 5 to 9 mo old as described previously (Hine et al., 2019). Briefly, calves received a 7in1 clostridial and leptospira multivalent vaccine (Ultravac 7in1, Zoetis) on day 0 of testing, coinciding with the start of yard weaning, to induce measurable immune responses. Serum immunoglobulin G1 antibody responses to the tetanus toxoid component of the Ultravac vaccine were measured between 8 and 21 d post-vaccination (dependant on prior vaccination history and aligned with the expected peak circulating antibody levels) and used to assess the ability of calves to mount Ab-IR. The magnitude of delayed-type hypersensitivity (DTH) reactions to vaccine antigens (Ultravac 7in1) injected intradermally into the skin was used to assess the ability of calves to mount Cell-IR. The magnitude of DTH responses was calculated by comparing the change in double skin-fold thickness (DSFT) observed at test reaction sites (injected with Ultravac 7in1) vs. control reaction sites (injected with saline) at 48 h post-injection as described previously (Hine et al., 2019).

Defining IC phenotypes

The IC phenotype assigned to each steer was based on their combined rankings for Ab-IR and Cell-IR, generated using model residual (observed minus predicted) values for antibody and DTH responses, respectively (Hine et al., 2011). Residuals for ranking on Ab-IR and Cell-IR were generated independently using the models described in the Statistical Analysis section. Relevant fixed effects (as described in Hine et al., 2019) were fitted to statistical models when generating residual values. Residual values for each respective trait (Ab-IR and Cell-IR) were then standardized, by dividing each residual value by the standard deviation of all residual values for that trait. Standardized residual values for Ab-IR and Cell-IR were then used to determine the IC phenotype for each individual steer. For categorical data analyses, steers with a standardized residual value, which was >0.5 for both Ab-IR and Cell-IR, were classified as high IC phenotype steers; steers with a standardized residual value < −0.5 for both Ab-IR and Cell-IR were classified as low IC phenotype steers; and all other steers were classified as average IC phenotype steers. The threshold value of 0.5 was chosen as it resulted in approximately 10% of steers being categorized as low and high IC phenotype animals, respectively, and would allow meaningful analysis to investigate associations between IC and feedlot performance to be undertaken. Further studies will be required to determine optimal threshold values to target when selecting for IC. For continuous data analyses, the standardized residual values for Ab-IR and Cell-IR were averaged to generate a single IC index value named ZMEAN.

Feedlot finishing

The effects of IC phenotype on the health and productivity of the 838 ASBP steers were assessed during the finishing period. Following IC testing at weaning, calves remained on their property of origin and were grown out on pasture until purchased by the feedlot. A proportion of the steers from each birth year cohort (n = 108 in birth year cohort 1 and n = 190 in birth year cohort 2) were purchased by the commercial feedlot operator prior to reaching feedlot entry weights and left their property of origin to be grown out on pasture by the feedlot operator (a practice commonly known as “backgrounding”) for a period of 280 and 132 d, respectively, prior to entering the feedlot. As backgrounding cattle on pasture prior to feedlot entry is common practice for Australian feedlot operators, and detailed health and productivity data were recorded on steers during both the backgrounding and feedlotting periods by the feedlot operator, data collected during the backgrounding phase were included in the results presented here. Once steers had reached feedlot entry weight (approximately 400 kg body weight), they were transported from their property of origin, or the backgrounding facility, to a research feedlot (1,000 head capacity ran under commercial conditions) where they remained on feed for approximately 100 d so that individual feed intakes could be measured for the estimation of sire EBV for feed efficiency (not related to the current study). At the conclusion of this initial 100-d feeding period, steers were transported to a large-scale commercial feedlot (>30,000 head capacity) where they remained on feed for approximately 180 d (minimum 171 d) until slaughtered. Both feedlots were situated in Northern NSW, Australia.

All steers were pre-vaccinated against BRD prior to entering the research feedlot. Steers received a primary vaccination with Bovilis MH+IBR (Coopers Animal Health) 4 to 6 wk prior to entering the feedlot followed by a boost vaccination at feedlot induction. Upon arrival at the feedlot, all steers were inducted using a standard induction protocol which, in addition to boost vaccination with Bovilis MH-IBR, involved tagging, weighing, chemical treatment for internal and external parasites, and vaccination against the clostridial disease. Steers remained in their herd cohort throughout the feedlot finishing at both the research and commercial feedlot and were only mixed with unfamiliar cattle at the commercial feedlot (after a minimum of 100 d in the research feedlot) where cohort size was smaller than pen capacity.

Feedlot entry and exit dates, which were the same for a given herd cohort, were recorded for all steers along with feedlot entry and exit weights to allow the computation of days on feed (DOF) and average daily gain (ADG). Details of the initial purchase price paid by the feedlot operator for individual steers entering backgrounding or the feedlot were also obtained. All health-related issues observed in the steers during backgrounding and feedlot finishing at both the research and commercial feedlot were recorded. Data collected included details of the date when any animal was removed from its paddock/feedlot pen and placed in a hospital pen for veterinary treatment, the diagnosed disease/illness, therapeutics administered, including dose rate, frequency of administration, and cost of therapeutics (estimated as cost price + 70% retail markup), the date when animals either succumbed to disease/illness or returned to the paddock/feedlot pen, and any mortalities including date and cause of death including necroscopy findings where necroscopies were undertaken. Diagnosis of disease/illness and cause of death were determined by highly trained feedlot staff in consultation with the relevant feedlot veterinarian. When the cause of death was not obvious, necroscopies were undertaken by highly trained feedlot staff under the guidance of the feedlot veterinarian.

Statistical Analysis

IC phenotype

Univariate animal models were tested in ASReml (Gilmour et al., 2009) to estimate variance components for IC traits. Traits were tested for normality using a Shapiro–Wilk test in R (R Core Team, 2013) and data transformed were required to approximate normality (as described in Hine et al., 2019). Contemporary group (CG, with 28 levels) was defined as the combination of property of origin, birth year cohort, management group (birth to weaning), and herd IC testing cohort. For logistical reasons, steers in large management groups were randomly split and tested across consecutive days. CG and dam age were tested as fixed effects in models. Covariates fitted and assessed in models included age at measurement and DSFT at test site/DSFT at control site at T0 (for Cell-IR). Details of fixed effects and covariates assessed and retained in models when analyzing each IC trait are detailed in Hine et al. (2019). The main effect of CG, along with relevant covariates, was retained in models regardless of their significance. However, models were reduced by removing other fixed effects, which were not significant (P > 0.05). Residual values from models for the IC traits, Ab-IR and Cell-IR, were then used to assign individual steers to IC phenotype categories low, average, and high for categorical data analyses or used to calculate an IC index value, ZMEAN, for continuous data analyses (as described above).

Categorical data analyses

Average daily gain

Least squares means (LSM) for ADG within IC phenotype groups (low, average, and high) were generated from a linear model, fitting CG and age of dam as a fixed effects and both entry weight and DOF as covariates. Separate analyses were undertaken to estimate LSM for ADG using either data collected from the start of backgrounding (where the feedlot had purchased steers prior to entering the feedlot) through to exit from the feedlot or solely during the feedlotting period. The decision to analyze the data separately was based on the knowledge that it is common practice for some feedlot operators in Australia to background steers (purchase animals prior to reaching feedlot entry weight and background them on pasture prior to entering the feedlot) and not others and was aimed at presenting data of relevance to the broader industry. When analyzing the combined data from the backgrounding and feedlotting periods, weights collected at the commencement of backgrounding were used as entry weight and the sum of days spent at the backgrounding facility and the feedlot was used as DOF for steers, which were backgrounded on pasture before entering the feedlot (n = 298), whereas, weights collected at induction at the research feedlot were used as entry weight and the sum of days spent at the feedlot was used as DOF for steers not backgrounded on pasture by the feedlot (n = 541). When analyzing data solely from the feedlotting period, weights collected at induction at the research feedlot were used as entry weight and the sum of days spent at the feedlot was used as DOF for all steers (n = 839). The ADG of individual steers was calculated as ((exit weight – entry weight) / DOF).

To investigate the influence of mortalities on ADG, a “Deads-in” vs. “Deads-out” analysis was undertaken as described by Gaylean and Elam (2009). For the analysis of ADG using the “Deads-in” approach, an ADG of zero was assigned to any animal that died during the finishing period. Feedlot entry weight and DOF (at time of death) were included as covariates for these steers. Alternatively, for the analysis of ADG using the “Deads-out” approach, all data from steers that died during the finishing period were excluded from the analysis. When interpreting the results from “Deads-in” vs. “Deads-out” analysis methods, findings from the “Deads-in” analysis are expected to better reflect the actual productivity output of a feedlot enterprise.

Disease/illness incidence and mortalities

The Pearson’s chi-square test (Excel) was used to assess the effects of IC phenotype on the incidence of disease/illness (cases per 100 animals) and the percentage of mortalities observed during the finishing period. For disease/illness incidence data, where a steer was diagnosed with the same disease/illness on more than one occasion, only cases where the steer was reported to have recovered from the disease/illness before being diagnosed a second time were included in the data. Further, a maximum of two cases per steer per disease/illness were included in the data to minimize the potential influence that individual steers could have on results.

Continuous data analyses

Average daily gain and days on feed

To complement the categorical data analyses based on low, average, and high IC phenotype classes, we undertook a series of continuous data analyses aimed at characterizing the strength of the significance of the fitted effects and covariates, including ZMEAN, on the response variables of ADG and DOF. Feedlot entry and exit dates, and therefore DOF, were the same for steers in a given herd cohort, with the exception of when a small number of steers were sold early as they were not performing or died during backgrounding or feedlot finishing. Therefore, DOF were analyzed as an indirect measure of underperformance/mortality during finishing. Similar to analyses of categorical data described above, separate analyses were undertaken using data collected from the start of backgrounding through to exit from the feedlot or solely during the feedlotting period. A “Deads-in” vs. “Deads-out” analysis was also undertaken to investigate the influence of mortalities on ADG and DOF. The Procedures CORR and GLM of SAS (version 9.4; SAS Institute, Cary, NC) were employed to undertake these analyses. The linear model for the analysis of ADG contained the fixed effect of CG and the covariates of age of dam, entry weight, DOF, and ZMEAN. Similarly, the linear model for the analysis of DOF contained CG and the linear regression covariates of age of dam, entry weight, and ZMEAN.

Results

Of the total number of steers (n = 839) entering the feedlot, 98 (12%) were classified as low, 653 (78%) as average, and 88 (10%) as high for IC phenotype. The effects of the IC phenotype group on mortality rates during the finishing are presented in Table 1. Although the overall mortality rate was low (1.67%), the rate of mortalities was higher in low IC steers (6.1%) as compared with average (1.2%, P < 0.001) or high (0%, P = 0.018) IC steers. Associations between the IC phenotype index, ZMEAN, and ADG and DOF of steers during finishing are presented in Table 6. A significant positive association between ZMEAN and DOF (r = 2.061 +/− 0.711, P = 0.004) was observed during feedlot finishing when data from animals that died at the feedlot were included in the analysis (“Deads-in” analysis), suggesting that steers with a lower IC index score are more likely to exit the feedlot earlier due to them not performing or dying at the feedlot.

Table 1.

The effects of IC phenotype on the incidence of mortalities observed in steers (n = 839) during finishing¹

	IC phenotype
	Low	Average	High	All
Total mortalities	6	8	0	14
Total animals	98	653	88	839
% Mortalities	6.12% a	1.23% b	0% b	1.67%

¹Data are presented as the number of mortalities observed in steers from each of the IC phenotype groups, low (n = 98), average (n = 653), and high (n = 88).

^a,bValues that differ significantly are depicted using different superscript letters. Low vs. average IC steers (P < 0.001) and low vs. high IC steers (P = 0.018).

Table 6.

Analysis of variance (ANOVA) table with P-values for the association of fitted effects and covariates on dependent variables and goodness of fit (R²) for each respective model

		Fitted effect/covariate
Dependent variable	Analysis	CG	Age of dam	Entry weight	DOF	ZMEAN	R 2, %
Backgrounding and feedlot finishing periods combined
ADG	Deads-in	<0.001	0.088	0.023	<0.001	0.613	48.0
	Deads-out	<0.001	0.735	<0.001	0.079	0.427	57.3
DOF	Deads-in	<0.001	0.052	0.468	NA	0.150	89.8
	Deads-out	<0.001	0.810	0.917	NA	0.881	97.9
Feedlot finishing period only
ADG	Deads-in	<0.001	0.148	0.024	<0.001	0.837	39.0
	Deads-out	<0.001	0.563	0.001	<0.001	0.882	39.6
DOF	Deads-in	<0.001	0.638	0.393	NA	0.004	23.0
	Deads-out	<0.001	0.474	0.282	NA	0.495	54.2

Details of disease/illness incidence cases observed in steers during finishing are summarized in Table 2. Only a single case of disease incidence was recorded in those steers that were backgrounded on pasture, being a foot abscess in a steer classified as average for IC phenotype. Although the incidence of disease/illness was not significantly affected by IC phenotype, the low IC phenotype steers had numerically more cases (15.3) of disease/illness per 100 animals compared with average (10.1) and high (10.2) IC phenotype steers. In contrast to cases of disease/illness, the numbers of sick days were lowest for low IC phenotype animals (80 d/100 animals) and highest for high IC phenotype steers (115 d/100 animals). When interpreting the “sick days” data, it is important to consider that “sick days” included both the number of days between date when an animal removed from paddock or feedlot pen and recovery (return to paddock or feedlot pen) or death where an animal succumbed to disease/illness. Also, the increased mortality rate observed in the low IC phenotype steers is expected to have contributed to the reduced numbers of “sick days” observed for this group as average “sick days” for steers that died was generally less than average “sick days” for animals that recovered.

Table 2.

Summary of disease/illness incidence observed in steers (n = 839) during backgrounding and feedlot finishing¹

	Low		Average		High		Total
Disease/illness	n	/100 head	N	/100 head	n	/100 head	n	/100 head
Foot abscess/lame	1	1.02	11	1.69	2	2.28	14	1.67
Bloat	6	6.12	30	4.59	2	2.27	39	4.65
Cast2	1 (1)	1.02	3	0.46	0	0	4 (1)	0.48
Respiratory/pneumonia	1	1.02	7 (1)	1.07	3	3.41	11 (1)	1.31
Digestive	1	1.02	0	0	1	1.14	2	0.24
Cellulitis2	0	0	1	0.15	0	0	1	0.12
Ascites2	1 [1]	1.02	0	0	0	0	1 [1]	0.12
Heart failure	0	0	1(1)	0.15	0	0	1 (1)	0.12
Unknown (disease/illness could not be determined)	4 {1}(3)	4.08	13 {1}(5)	1.99	1	1.14	18 {2}(8)	2.15
TOTAL	15    {1}[1](4)	15.31 a	66    {1}(7)	10.11 a	9	10.23 a	91    {2}[1](11)	10.85
Sick days3	78	80	602	92	101	115	781	93

¹For each disease/illness, data are presented as the number of observed cases in steers from each of the IC phenotype groups, low (n = 98), average (n = 653), and high (n = 88). The numbers in curly, square, and curved brackets indicate the number of animals that died from that disease/illness during backgrounding on pasture, at the research feedlot, and the commercial feedlot, respectively.

²Cast—animal in lateral or dorsal recumbency unable to recover ventral recumbency, Cellulitis—soft tissue swelling usually caused by a bacterial infection, and Ascites—accumulation of fluid in the peritoneal cavity.

³Sum of days between date removed from paddock/feedlot pen and recovery (return to paddock/feedlot pen) or death for all sick animals within the group.

^aValues that differ significantly are depicted using different superscripts letter. Low vs. average IC steers (P < 0.122) and low vs. high IC steers (P = 0.303).

The effects of the IC phenotype group on productivity are presented separately for the combined backgrounding and finishing periods (Table 4) and only the finishing period (Table 5). In both tables, the productivity data are presented separately based on the inclusion (“Deads-in” analysis) and exclusion (“Deads-out” analysis) of steers that died during the finishing period. A “Deads-in” vs. “Deads-out” analysis of feedlot exit weight data was undertaken to investigate the influence of IC phenotype on feedlot exit weight when the impacts of mortalities were captured (“Deads-in” analysis) or ignored in the analysis (“Deads-out” analysis). No significant differences in ADG were observed between IC phenotype groups using either the “Deads-in”, (P = 0.832 for productivity data from backgrounding and feedlot finishing combined and P = 0.822 for productivity data from feedlot finishing only) or “Deads-out” (P = 0.434 for productivity data from backgrounding and feedlot finishing combined and P = 0.439 for productivity data from feedlot finishing only) analysis method (Table 3). In support of these findings, no associations between ZMEAN and ADG were observed during the combined backgrounding and feedlot finishing periods of feedlot finishing period only using either the “Deads-in” or “Deads-out” analysis method (Table 6).

Table 4.

Summary of feedlot performance for steers (n = 839) during backgrounding and feedlot finishing¹

IC phenotype	n 2	Mean entry weight, kg	Mean feedlot exit weight, kg	Mean weight gain2, kg	Mean DOF, d	Mean feedlot exit weight, kg	ADG2,3, kg/d
Deads-in4
Low	98	406	782	376	337	808	1.16
Average	653	409	812	402	347	816	1.16
High	88	389	822	433	372	822	1.18
Deads-out4
Low	92	404	833	429	350	833	1.23
Average	645	409	822	413	349	821	1.20
High	88	389	822	433	372	822	1.21

¹Data are presented for each of the IC phenotype groups low, average, and high when data were analyzed using both the Deads-in vs. Deads-out approaches. Entry weights were recorded at either commencement of backgrounding on pasture (n = 298) or induction into the research feedlot (n = 541). Feedlot exit weights were recorded at the commercial feedlot after a minimum of 271 DOF.

² n = number of animals; weight gain = feedlot exit weight – entry weight; ADG = weight gain/DOF.

³LSM for ADG adjusted for CG, entry weight, and DOF.

⁴For Deads-in calculations, feedlot entry weight and DOF data for animals that died were included in calculations (see Gaylean and Elam, 2009); for Deads-out calculations, feedlot entry weight and DOF data for animals that died were not included in calculations (see Gaylean and Elam, 2009).

Table 5.

Summary of feedlot performance for steers (n = 839) during feedlot finishing¹

IC phenotype	n 2	Mean feedlot entry weight, kg	Mean feedlot exit weight, kg	Mean feedlot weight gain2, kg	Mean DOF, d	Mean feedlot exit weight, kg	ADG2,3, kg/d
Deads-in 4
Low	97	446	808	367	276	808	1.27
Average	652	449	816	368	282	816	1.26
High	88	436	822	385	283	822	1.28
Deads-out4
Low	92	446	833	387	282	833	1.31
Average	645	449	821	372	282	821	1.28
High	88	436	822	385	283	822	1.29

¹Data are presented for each of the IC phenotype groups, low, average, and high, when data were analyzed using both the Deads-in vs. Deads-out approaches. Entry weights were recorded at induction into the research feedlot. Feedlot exit weights were recorded at the commercial feedlot after a minimum of 271 DOF.

² n = number of animals; feedlot weight gain = feedlot exit weight – feedlot entry weight; ADG = total feedlot weight gain/total DOF.

³LSM for ADG adjusted for CG, feedlot entry weight, and DOF.

⁴For Deads-in calculations, feedlot entry weight and DOF data for animals that died were included in calculations (see Gaylean and Elam, 2009); for Deads-out calculations, feedlot entry weight and DOF data for animals that died were not included in calculations (see Gaylean and Elam, 2009).

Table 3.

Analysis of variance (ANOVA) table with P-values describing the influence of fitted effects and covariates on ADG assessed during the backgrounding and feedlot finishing periods combined or feedlot finishing period only

		Fitted effect/covariate
Variable	Analysis	CG	Age of dam	Entry weight	DOF	IC phenotype
Backgrounding and feedlot finishing periods combined
ADG	Deads-in	<0.001	0.339	0.008	<0.001	0.832
	Deads-out	<0.001	0.530	<0.001	0.012	0.434
Feedlot finishing period only
ADG	Deads-in	<0.001	0.334	0.024	<0.001	0.822
	Deads-out	<0.001	0.583	<0.001	<0.001	0.439

As a consequence of the increased mortalities observed in the low IC steers, when productivity data from backgrounding and feedlot finishing were combined, low vs. high IC steers had an ADG of 1.23 and 1.21 kg/d, when data from steers that died were excluded from the analysis, and 1.16 kg/d compared with 1.18 kg/d when steers that died were included in analysis. When performance during feedlot finishing only was considered, low vs. high IC steers had an ADG of 1.31 kg/d compared with 1.29 kg/d when data from steers that died were excluded from the analysis and 1.27 kg/d compared with 1.28 kg/d when data from steers that died were included in the analysis.

The effects of IC phenotype on estimated health costs of steers during the finishing period are presented in Table 7. The total costs were estimated based on the value of the feeder steer on feedlot entry and the costs of feed and the therapeutic agents used to treat disease. Lost production costs were calculated by summing the total costs to feed and manage steers that died. Lost capital costs were calculated by summing the purchase cost of steers that died. Disease treatment costs were calculated by summing the total cost of therapeutic agents used to treat disease. The results from this analysis estimated the costs incurred by feedlot operators due to disease/illness and mortalities during finishing to be AUS$103/head for low, AUS$25/head for average, and AUS$4/head for high IC phenotype steers (Table 7). It is important to consider when interpreting these results that the direct labor costs associated with administering treatments and monitoring of animals along with lost productivity as a result of animals being sick could not be calculated and, therefore, were not factored into estimates. At the same time, the opportunity cost associated with having a sick animal or an animal that dies taking up pen space that could otherwise have been occupied by a healthy animal that would have been gaining weight and generating income for the feedlot was also not factored into estimates.

Table 7.

Estimated disease and mortality costs incurred for steers (n = 839) classified as low (n = 98), average (n = 653), or high (n = 88) IC phenotype animals during finishing

	Estimated health-associated costs
	Lost production costs (/100 head)		Lost capital investment (/100 head)		Disease treatment costs (/100 head)	Total cost (/100 head)
IC phenotype	Days fed at the time of death (A)1	Cost (B)1	Deaths (C)1	Cost (D)1	Cost (E)1	Total cost (F)1
Low	[776] (14)	AUS$3,801	6.12	AUS$6,336 (AUS$2.37)	AUS$145	AUS$10,282
Average	[173] (82)	AUS$926	1.23	AUS$1,434 (AUS$2.56)	AUS$146	AUS$2,506
High	0	AUS$0	0	AUS$0	AUS$353	AUS$353

¹A, number in square brackets indicates the sum of days fed in the feedlot at time of death for all animals within the group which died (includes days in hospital pen). The number in curved brackets indicates the sum of days fed pasture only at the time of death for all animals within the group which died; B, total lost production costs for all animals within the group. Costs to feed on pasture were estimated at a flat rate of AUS$1/head/d. Costs to feed at feedlot were estimated by the feedlot operator and are based on an average dry matter intake of 13.5 kg/head/d at a cost of AUS$308/tonne (AUS$4.16) + daily direct costs per head of AUS$0.72 to give total cost of AUS$4.88/head/d. Therefore, B = (number of days fed in the feedlot [see column A] * AUS$4.88) + (number of days fed pasture only [see column A] * AUS$1.00); C, deaths observed per 100 animals within each group; D, total cost associated with lost capital investment for all animals within the group. Cost represents the sum of purchase costs obtained from the feedlot for individual animals that died within each group. Figure in brackets represents average cents/kilogram liveweight paid at purchase; E, total cost associated with the purchase of therapeutic agents used to treat disease. Figures represent estimated retail cost of therapeutic agents calculated at cost price + 70%; F, total cost associated with lost production days, lost capital investment, and disease treatment costs for all animals within the group (B + D + E).

Discussion

The critical first step in the development of genetic strategies aimed at improving general disease resistance is to establish methodologies to phenotype animals that are predictive of health outcomes in their commercial production environments. We have developed a testing procedure to assess the IC phenotype of beef cattle, which is both practical to apply on-farm and does not restrict the future sale of tested animals for human consumption. The aim of the current study was to investigate associations between IC and feedlot performance, in terms of both health and growth, in an industry-relevant commercial setting.

Steers were categorized as low, average, or high for IC phenotype, and differences in disease incidence and mortalities observed between groups during finishing were investigated. A favorable association between IC phenotype group and number of mortalities was observed in the current study with 6.1% mortalities recorded in low IC phenotype steers, 1.2% mortalities in average IC phenotype steers, and no mortalities observed in high IC phenotype steers. Mortalities in low IC phenotype steers were significantly higher than in their average (P = <0.001) and high (P = 0.018) IC phenotype counterparts. A significant positive association between ZMEAN and DOF (P = 0.004) was also observed during feedlot finishing, when data from animals that died at the feedlot were included in the analysis, suggesting that steers with a low ZMEAN score are more likely to exit the feedlot earlier due to them not performing or dying at the feedlot.

Combined these results suggest that there may be significant benefits for the Australian feedlot sector, in terms of improved animal health and welfare, to be gained through genetic improvements in IC. Although these results are encouraging, it is acknowledged that the overall incidence of disease (10.85 cases/100 animals) and the number of mortalities (1.67%) observed in the current study were low and, further, that it was not possible to determine the cause of death in a high proportion of the mortalities observed during feedlot finishing (10/14). These limitations highlight the need for further large-scale studies to validate these findings and confirm the potential benefits for the feedlot sector in identifying animals with improved IC.

Disease incidence was numerically higher in low IC phenotype steers (15.3 cases/100 animals) and similar in average IC phenotype steers (10.1 cases/100 animals) and high IC phenotype steers (10.2 cases/100 animals); however, differences between groups were not significant. In contrast, associations between general immune responsiveness and disease incidence in large commercial dairy herds have been reported previously (Thompson-Crispi et al., 2012). In that study, the incidence of mastitis was higher in average, as compared with high, immune responder cows. Mastitis incidence also tended to be higher in low, as compared with high, immune responder cows but the observed difference was not statistically significant. The incidence of displaced abomasums and retained fetal membranes was also higher in low immune responder cows; however, the incidence of metritis and ketosis was not influenced by immune responder phenotype. The low incidence of disease observed in the current study may have contributed to the differing associations between IC phenotype and disease incidence observed between these studies.

In contrast to the trend observed for mortality rates, which showed that fewer health mortalities were recorded in high as compared with low IC phenotype steers, the numbers of reported sick days were higher in high (115 d/100 head) as compared with average (92 d/100 head) and low (80 d/100 head) IC phenotype steers. Sick days were calculated as the sum of days between date removed from paddock/feedlot pen to enter hospital pen and recovery (return to paddock/feedlot pen) or death. Therefore, steers that succumbed to disease in the hospital pen and did not return to the paddock/feedlot pen, or died in the paddock/feedlot pen itself, generally contributed fewer sick days than those steers that recovered and returned to their paddock/feedlot pen. On this basis, the higher number of sick days recorded for high IC steers could have been influenced by the fact that no mortalities were observed in this group and all animals that were removed from paddock/feedlot pens for treatment recovered and returned to their paddock/feedlot pen. The ability of an animal to cope with challenges faced in its production environment and return to being productive can be defined as resilience (Colditz and Hine, 2016). Livestock respond to challenges from infectious agents and other environmental stressors through immunological, physiological, and behavioral defense reactions and these modalities of host defense are highly integrated (Hine et al., 2014). The results from the current study suggest that high IC phenotype animals have an improved ability to recover from disease challenges and, therefore, are expected to be more resilient. Research over a number of years has highlighted that the level of activity of the immune system is associated with an animal’s ability to thrive in the face of environmental stressors and can be an indicator of future health and performance (Schmid-Hempel, 2003), highlighting the important contribution of IC to an animal’s resilience (Dantzer et al., 2018).

The health-associated costs incurred by steers categorized as low, average, or high for IC phenotype during finishing were calculated to quantify the potential economic benefits for feedlot operators when selecting for IC. Health-associated costs were calculated as the sum of lost production costs, lost capital investment costs, and disease treatment costs. Labor costs associated with treating sick animals and opportunity costs associated with having a profitable animal in the feedlot in the place of an animal that died during feedlot finishing were not considered in the calculation. The health-associated costs were estimated at AUS$4, AUS$25, and AUS$103, (per head) for high, average, and low IC phenotype animals, respectively. Furthermore, the results showed that low IC phenotype animals, which represented only 11.7% of steers entering the feedlot, were responsible for 35% of the total estimated health-associated costs incurred by the entire cohort of steers during finishing. These results suggest that significant economic benefits could be gained through strategies aimed at improving the general IC of cattle destined for feedlot finishing.

Similar to the disease incidence and mortality data, estimates of the potential economic benefits of selecting for improved IC described here will require further validation in future large-scale commercial trials. However, it is important to consider that the current study was conducted in what could be considered a “low disease risk feedlot scenario” and, therefore, the potential economic benefits of selecting for IC estimated here may be conservative. BRD is the largest health issue facing the Australian feedlot industry, being responsible for the majority of veterinary treatments (up to 82%) and health-related mortalities (~80%) in Australian feedlots (Gonzalez et al., 2018). Known risk factors for BRD include animals not being pre-vaccinated against BRD prior to feedlot entry and stressors experienced during the feedlot induction period, including human handling, transportation, commingling with unfamiliar animals, and acute metabolic changes (Taylor et al., 2010; Hay et al., 2016). All steers enrolled in the current study were pre-vaccinated against BRD prior to entering the feedlot, yard weaned, and regularly handled during the backgrounding phase and were not mixed with unfamiliar animals during the first 100 d of feedlot finishing. Together, these factors are expected to have significantly reduced the risk factors for disease.

Due to the perceived contribution of agriculture to the global issue of antimicrobial resistance, there is mounting global pressure from the supply chain and consumers to reduce the use of antibiotics to prevent and/or treat disease in food-producing animals. As the use of antibiotics comes under increasing scrutiny, our ability to treat diseased animals will be restricted. This is expected to have significant implications for animal health and welfare, highlighting a critical need for alternative strategies to reduce disease incidence in livestock. When estimating the potential economic benefits of such strategies, the indirect benefits realized through the ability to maintain consumer confidence in products of the livestock industries, as a result of improved animal health and welfare and reduced reliance on the use of antibiotics, should not be underestimated.

It could be speculated that the ability to mount an enhanced immune response to a disease challenge may incur a production cost as a consequence of nutrients being redirected from growth, for instance, to support immune function. However, counterbalancing any cost of enhanced immune function is the metabolic cost of disease (reviewed by Colditz 2002, 2008). Along with clinical disease and mortalities, chronic activation of immune defense pathways during subclinical infection is expected to lead to reduced efficiency of production. To investigate associations between IC phenotype and productivity, an ADG was calculated for all steers and a “Deads-in” vs. “Deads-out” analysis of the data, considering the IC phenotype as a categorical or continuous variable, was undertaken. The “Deads-in” vs. “Deads-out” analysis used in this study allowed the influence of IC phenotype on productivity to be estimated when animals that died were either included (“Deads-in”) or excluded (“Deads-out”) from the analysis. When productivity data from backgrounding and feedlot finishing were combined and steers that died were excluded, ADG LSM for steers categorized as having a low, average, and high IC phenotype were 1.23, 1.20, and 1.21 kg/d, respectively. In comparison, when steers that died were included, ADG LSM for low, average, and high IC phenotype steers were 1.16, 1.16, and 1.18 kg/d, respectively. When productivity data from feedlot finishing only were considered and steers that died were excluded, ADG LSM for steers categorized as having a low, average, and high IC phenotype were 1.31, 1.28, and 1.29 kg/d, respectively. In comparison, when steers that died were included, ADG LSM for low, average, and high IC phenotype steers were 1.27, 1.26, and 1,28 kg/d, respectively.

No significant differences in ADG were observed between IC phenotype groups when IC phenotype was analyzed as a categorical variable using either the “Deads-in” or “Deads-out” analysis method. Supporting these findings, no associations between ZMEAN and ADG were observed using either the “Deads-in” or “Deads-out” analysis methods. Together, these results suggest that selection for improved IC does not incur a significant penalty to production. Associations between immune responsiveness and productivity have been investigated previously in other species. In pigs, high immune responder animals were found to have higher growth rates than their average and low immune responder counterparts (Mallard et al., 1998). While in dairy cattle, the relationship between antibody-mediated immune responsiveness and milk production was reported as being favorable in multiparous cows but unfavorable in first-parity cows (Wagter et al., 2003).

There is mounting evidence to suggest that selection for productivity, with little or no emphasis on health and fitness, has led to an increase in susceptibility to disease in many production animal species (Rauw et al., 1998). Although we acknowledge the need for beef producers to continue to target genetic gains in economically important production traits, it is critical that they start to place some selection emphasis on health and fitness traits if they wish to breed highly productive animals that can cope with biotic and abiotic challenges within their production environment. The results from the current study suggest that selection for IC has the potential to reduce mortalities in feedlot cattle while not negatively impacting the productivity. Such health benefits are expected to reduce health-related costs for feedlot operators and improve the health and welfare of animals in the feedlot production system.

Future studies will be aimed at 1) exploring alternative methods to combine both IC metrics into a single index, which can be used to develop genomically enhanced EBVs for IC that can inform breeding decisions in industry and 2) validating the favorable associations between IC and feedlot health reported here. We also see the potential for genomic tools to be developed, based on data for health, growth, and carcass traits, which can predict the genetic potential of commercial animals to perform in the feedlot environment and inform management decisions such as identifying the most suitable finishing path for individual animals entering the feedlot.

Glossary

Abbreviations

Ab-IR

antibody-mediated immune response

ADG

average daily gain

BRD

bovine respiratory disease

Cell-IR

cell-mediated immune response

DOF

days on feed

DTH

delayed-type hypersensitivity

EBV

estimated breeding value

IC

immune competence

Conflict of interest statement

The authors declare no real or perceived conflicts of interest.