Introduction

Foot and leg disorders can negatively influence longevity and therefore cause economic loss in both beef (McGuire et al., 2019) and dairy cattle (Enting et al., 1997; Boettcher et al., 1998). The type-trait evaluation has been long established in dairy industry. Even though the importance of phenotypic foot soundness traits is well established among beef cattle producers, the genetic evaluation for such traits was only introduced recently and improvements are ongoing (Jeyaruban et al., 2012; Australian Angus Association, 2016,, 2017; Vargas et al., 2017; Wang et al., 2017; Giess et al., 2018; BIF Guidelines Wiki, 2019). Foot angle (FA) and claw set (CS) are important foot soundness structural traits that can be collected on beef cattle. Phenotype for both traits is a subjective score from 1 to 9 with a score of 5 being an intermediate optimum (American Angus Association, 2017). Foot structural soundness scores are indicator traits for longevity and should not be considered as economically relevant traits (BIF Guidelines Wiki, 2019). Moderate heritability estimates for FA and CS indicate that genetic improvement of foot score can be achieved (Wang et al., 2017). Our objective was to estimate genetic parameters and predict expected progeny differences (EPD) for FA and CS scores using the data from Leachman Cattle of Colorado (LCoC).

MATERIALS AND METHODS

Animal Care and Use Committee approval was not required for this study as the data were obtained from LCoC database.

Animals and Data

Conforming to the American Angus Association (AAA) guidelines (American Angus Association, 2017), phenotypes of FA and CS were recorded using a nine-point scoring system. For FA, a score of one indicates extremely straight pasterns and short toes; a score of nine indicates extremely shallow heels, weak pasterns, and long toes; and a score of five is the intermediate optimum. For CS, a score of one corresponds to weak, open, divergent CS; a score of nine corresponds to extreme scissor claw; and a score of five is ideal.

Raw data obtained from LCoC (n = 6,090 records) included animals with repeated measures of FA and CS. Most animals (n = 4,333) had one observation reported for each trait. There were 865 animals that had multiple records with most having two records per trait per animal. Only data on yearlings and 2-yr-old animals (i.e., measurement age ranged from 320 to 810 d) were kept. The Australian Angus Association collected structural soundness scores before hoof trimming to avoid mis-scoring animals. In the genetic evaluation of structural soundness scores in Australian Angus, only animals that are younger than 750 d (i.e., 25 mo) were included (Australian Angus Association, 2016). In a preliminary assessment of LCoC data, we decided to include data from animals younger than 810 d (i.e., 27 mo) for two reasons: 1) heritability using data from yearlings and 2-yr-old animals was higher than heritability with no age restriction based on preliminary analyses and 2) the majority of data were collected on yearlings and 2-yr-old animals. Following the AAA and Beef Improvement Federation (BIF) guidelines for foot score evaluation (American Angus Association, 2017; BIF Guidelines Wiki, 2019) for animals with multiple observations, only the most recent recorded observation (i.e., as a 2 yr old if available) was kept for further analysis. There were 285 and 99 animals with scores < 5 for FA and CS, respectively. Such observations were removed from data as recommended by BIF Guidelines Wiki (2019). Wang et al. (2017), in their study on American Angus, reported that analyzing FA and CS using scores ≥ 5 gave higher heritability estimates compared with using scores ≤ 5. Removing low scores also helps producers to easily understand and interpret EPD (i.e., lower EPD are favorable).

Leachman Cattle of Colorado is a multi-breed breeding program. Hence, the retained hybrid vigor was calculated using the following breed percentage designations: 1) Angus (Black and Red Angus), 2) South Devon, 3) British Other (British breed percentages that were not in 1 and 2 designations), 4) Simmental, 5) Charolais, 6) Gelbvieh, 7) Continental Other (Continental breed percentages that were not in 5, 6, and 7 designations), and 8) Other (breed percentages that were not in 1 to 7 breed designations).

Fixed effects included measurement age, herd–birthyear–season of birth, retained hybrid vigor, and contemporary group, which were created by combining sex, birth, weaning, and yearling group. Contemporary groups with zero variance were removed from data. Resulting data included 4,800 animals with at least one usable record. Beginning with those animals, a three-generation pedigree was constructed (15,677 animals, 2,204 unique sires, and 9,060 unique dams).

Estimation of Covariances and EPD Prediction

A bivariate, linear, animal model was used to estimate direct genetic and residual (co)variance parameters for FA and CS. The analysis was performed using ASReml 3.0 software (Gilmour et al., 2009). Fixed effects included measurement age, herd–birthyear–season of birth, retained hybrid vigor, and contemporary group. Fixed effects were tested for significance using the Wald F test. Subsequently, the retained hybrid vigor was found to be not significant (P > 0.05) for both traits, so it was removed from the model. A direct genetic effect was the sole random effect included in this model. The two-trait animal model used can be described as:

[ y 1 y 2 ] =  ​​ ​​ [ X 1  ​​ β ​​ 1 X 2  ​​ β ​​ 2 ] +  ​​ ​​ [ Z 1 u 1 Z 2 u 2 ] +  ​​ ​​ [ e 1 e 2 ] ,

In the above equation, subscripts 1 and 2 denote FA and CS, respectively; X i and Z i are known incidence matrices related to unknown fixed (βi) and direct genetic (ui) effects, respectively, to observations in yi; and ei is a random residual term. The first and second moments of the model were assumed to be:

E [ y ] =  X β ,  E [ u ] =  E [ e ] = 0

var [ u e ] =  ​​ ​​ [ G  ​​ ​​ ⊗  ​​ ​​ A 0 0 R  ​​ ​​ ⊗  ​​ ​​ I ]

where G = a 2 × 2 matrix of direct genetic variances and covariances, A = numerator relationship matrix, R = a 2 × 2 residual covariance matrix, I = identity matrix of order appropriate to the numbers of observations, and ⨂ = Kronecker product.

Estimated direct genetic and residual variances and covariances were used to predict EPD for FA and CS. Prediction of EPD was carried out using the same data, pedigree, and model that were used to estimate genetic parameters. The genetic evaluation of both traits was performed using the Animal Breeder's Tool Kit (Golden et al., 1992). Predicted EPD for FA and CS were averaged by birth year and then plotted against birth year to produce the genetic trends graph. Further, for each trait, estimated EPD were also regressed on birth year to estimate the intercept and slope (i.e., the annual genetic change) of the regression line, which were tested for difference from zero (R Core Team, 2019).

RESULTS AND DISCUSSION

Summary statistics for the raw data are presented in Table 1. Data included a total of 12,179 observations, on FA and CS combined, collected on 5,197 animals. Age of measurement ranged from 253 to 6,059 d with an average of 1,020 d and SD of 882.53 d. Distributions of FA and CS raw scores showed similar means and SD (Table 1). Average scores for FA and CS were 5.94 and 6.06, respectively.

Table 1.

Raw data (multiple observations per animal) summary statistics

Trait Animals Observation per animal Total†
1 2 3 4 Mean SD Minimum Maximum
FA 5,197 4,332 838 26 1 6,090 5.94 0.96 1 9
CS 5,197 4,333 837 26 1 6,089 6.06 0.94 3 9
Measurement age, d 5,197 6,090 1,020 882.53 253 6,059
Trait Animals Observation per animal Total†
1 2 3 4 Mean SD Minimum Maximum
FA 5,197 4,332 838 26 1 6,090 5.94 0.96 1 9
CS 5,197 4,333 837 26 1 6,089 6.06 0.94 3 9
Measurement age, d 5,197 6,090 1,020 882.53 253 6,059

†Total = total count of observations (allowing multiple observations per animal).

Table 1.

Raw data (multiple observations per animal) summary statistics

Trait Animals Observation per animal Total†
1 2 3 4 Mean SD Minimum Maximum
FA 5,197 4,332 838 26 1 6,090 5.94 0.96 1 9
CS 5,197 4,333 837 26 1 6,089 6.06 0.94 3 9
Measurement age, d 5,197 6,090 1,020 882.53 253 6,059
Trait Animals Observation per animal Total†
1 2 3 4 Mean SD Minimum Maximum
FA 5,197 4,332 838 26 1 6,090 5.94 0.96 1 9
CS 5,197 4,333 837 26 1 6,089 6.06 0.94 3 9
Measurement age, d 5,197 6,090 1,020 882.53 253 6,059

†Total = total count of observations (allowing multiple observations per animal).

Final data used to estimate genetic and residual parameters and predict EPD for FA and CS are summarized in Table 2. Average CS and FA scores were 6 and 5.93, respectively. Because of using only data on animals at age of 27 mo or younger, the distribution of measurement age (SD = 115 d; Table 2) was expectedly narrower than that of raw data (SD = 882.53 d Table 1).

Table 2.

Summary statistics for foot scores data

Trait n Mean SD Minimum Maximum
FA 4,133 5.93 0.77 5 9
CS 4,410 6.00 0.84 5 9
Measurement age, d 4,520 437.6 115 320 810
Trait n Mean SD Minimum Maximum
FA 4,133 5.93 0.77 5 9
CS 4,410 6.00 0.84 5 9
Measurement age, d 4,520 437.6 115 320 810

Table 2.

Summary statistics for foot scores data

Trait n Mean SD Minimum Maximum
FA 4,133 5.93 0.77 5 9
CS 4,410 6.00 0.84 5 9
Measurement age, d 4,520 437.6 115 320 810
Trait n Mean SD Minimum Maximum
FA 4,133 5.93 0.77 5 9
CS 4,410 6.00 0.84 5 9
Measurement age, d 4,520 437.6 115 320 810

The multi-breed nature of the LCoC data required incorporating retained hybrid vigor as another fixed effect beside contemporary groups, herd–birthyear–season of birth, and measurement age. Breeds and biological types used to calculate retained hybrid vigor were described earlier. All fixed effects, used to estimate genetic parameters and EPD for FA and CS, were tested for significance using the Wald F test available in ASReml 3 of Gilmour et al. (2009). Measurement age, herd–year–season, and contemporary groups showed a significant effect (P < 0.05) on both FA and CS. However, the retained hybrid vigor was not significant (P > 0.05) and, therefore, dropped from the model.

There were few articles in literature that published genetic parameters for foot structure in beef cattle (Jeyaruban et al., 2012; Wang et al., 2017; Giess et al., 2018), and therefore this study adds to the base of knowledge. In the United States, the AAA was among the first to estimate genetic parameters (Wang et al., 2017) and publish EPD for FA and CS scores. Giess et al. (2018), in a study on Red Angus cattle, estimated genetic parameters for foot and leg traits and concluded that they were lowly to moderately heritable. Genetic and residual parameters for FA and CS are presented in Table 3 with the heritability estimate for FA moderate (0.26) and within the range (0.17 to 0.34) of previously reported estimates (Jeyaruban et al., 2012; Wang et al., 2017; Giess et al., 2018). In a study on yearling American Angus (320 to 460 d), Wang et al. (2017) reported a heritability estimate of 0.34 for FA, which was higher than our estimate using both yearling and 2-yr-old animals (320 to 810 d). However, Jeyaruban et al. (2012) in their study on Australian Angus with measurement age limits (320 to 750 d) close to those of the present study reported heritability estimates of 0.32 and 0.29 for front and rear feet angle, respectively. However, if an animal has two observations (i.e., as a yearling and a 2 yr old), Jeyaruban et al. (2012) used the first observation instead of using the most recent one. Giess et al. (2018) evaluated foot and leg traits in Red Angus and did not put any restrictions on age of females. Authors estimated heritabilities for front and rear hoof angle to be 0.18 and 0.17, respectively, where the age at measure ranged from 1 to 2 yr for bulls and from 1 to 18 yr for cows. Heritability for CS was estimated to be 0.29 (Table 3), which was similar to heritabilities (0.33 and 0.29 for front feet and rear feet angle, respectively) estimated in Australian Angus (Jeyaruban et al., 2012). Wang et al. (2017) estimated a lower heritability of 0.21 for yearling American Angus. Red Angus front and rear claw shape heritability estimates were 0.08 and 0.15, respectively, (Giess et al., 2018). The residual correlation between FA and CS was 0.16 (Table 3). To the best of our knowledge, there were no published residual correlations between FA and CS in beef cattle. The genetic correlation of 0.32 between FA and CS was higher than that (0.22) reported by Wang et al. (2017) for yearling American Angus. For Australian Angus, Jeyaruban et al. (2012) reported a strong genetic correlation between front feet angle and front CS (0.79) and between rear feet angle and rear CS (0.62). Due to the low genetic correlation (0.22) between FA and CS, the AAA currently evaluates the two traits separately using single-trait models. The higher genetic correlation from the current study makes evaluating the two traits jointly an attractive choice.

Table 3.

Genetic parameters† for FA and CS scores

FA CS
FA 0.26 ± 0.06 0.16 ± 0.05
CS 0.32 ± 0.16 0.29 ± 0.06
FA CS
FA 0.26 ± 0.06 0.16 ± 0.05
CS 0.32 ± 0.16 0.29 ± 0.06

†Heritability on diagonal (boldfaced), genetic correlation below diagonal, and residual correlation above diagonal.

Table 3.

Genetic parameters† for FA and CS scores

FA CS
FA 0.26 ± 0.06 0.16 ± 0.05
CS 0.32 ± 0.16 0.29 ± 0.06
FA CS
FA 0.26 ± 0.06 0.16 ± 0.05
CS 0.32 ± 0.16 0.29 ± 0.06

†Heritability on diagonal (boldfaced), genetic correlation below diagonal, and residual correlation above diagonal.

Estimated EPD for FA and CS using a bivariate linear animal model are summarized in Table 4. The FA EPD averaged −0.026 (Table 4), and the annual genetic change was favorable (−0.0028; Table 5 and Figure 1), which corresponds to 0.88% of the genetic standard deviation. The average EPD for CS showed an increasing unfavorable rate of genetic change (0.0022; Table 5 and Figure 1). This rate of change corresponds to 0.58% of the genetic standard deviation for CS. Both show genetic change without the availability of EPD, as EPD were not previously available in this breeding program. Such change suggests breeders might be using phenotypic selection when evaluating potential breeding animals. Even though a positive moderate genetic correlation of 0.32 was estimated between the two traits, their genetic trends seem to slightly follow opposite directions. This might be a consequence of the genetic correlation being far from unity.

Table 4.

Summary statistics for FA and CS EPD

Trait n Mean SD Minimum Maximum
FA 15,677 −0.026 0.073 −0.33 0.39
CS 15,677 0.028 0.077 −0.34 0.56
Trait n Mean SD Minimum Maximum
FA 15,677 −0.026 0.073 −0.33 0.39
CS 15,677 0.028 0.077 −0.34 0.56

Table 4.

Summary statistics for FA and CS EPD

Trait n Mean SD Minimum Maximum
FA 15,677 −0.026 0.073 −0.33 0.39
CS 15,677 0.028 0.077 −0.34 0.56
Trait n Mean SD Minimum Maximum
FA 15,677 −0.026 0.073 −0.33 0.39
CS 15,677 0.028 0.077 −0.34 0.56

Table 5.

Regression coefficients† (±SE) for FA and CS EPD on birth year

Trait n Intercept Slope
FA 15,171 0.028 ± 0.001 −0.0028 ± 0.001
CS 15,171 −0.014 ± 0.001 0.0022 ± 0.001
Trait n Intercept Slope
FA 15,171 0.028 ± 0.001 −0.0028 ± 0.001
CS 15,171 −0.014 ± 0.001 0.0022 ± 0.001

†All regression coefficients were different from zero (P < 0.001).

Table 5.

Regression coefficients† (±SE) for FA and CS EPD on birth year

Trait n Intercept Slope
FA 15,171 0.028 ± 0.001 −0.0028 ± 0.001
CS 15,171 −0.014 ± 0.001 0.0022 ± 0.001
Trait n Intercept Slope
FA 15,171 0.028 ± 0.001 −0.0028 ± 0.001
CS 15,171 −0.014 ± 0.001 0.0022 ± 0.001

†All regression coefficients were different from zero (P < 0.001).

Figure 1.

Genetic trends for FA and CS scores in LCoC.

Genetic trends for FA and CS scores in LCoC.

Figure 1.

Genetic trends for FA and CS scores in LCoC.

Genetic trends for FA and CS scores in LCoC.

Conclusions

The moderate heritability estimates for FA and CS suggest that there is sufficient genetic variability that can be exploited to select for foot soundness and, therefore, improve longevity by reducing involuntary culling in beef cattle. Estimated genetic correlation between FA and CS showed a moderate genetic relationship between traits. Therefore, a two-trait genetic evaluation could be a viable option to jointly evaluate FA and CS scores.

Acknowledgments

We gratefully acknowledge Leachman Cattle of Colorado for providing the data.

Conflict of interest statement. The authors declare no conflict of interest related to this work.

LITERATURE CITED

Boettcher

,

P. J.

,

J. C.

Dekkers

,

L. D.

Warnick

, and

S. J.

Wells

.

1998

.

Genetic analysis of clinical lameness in dairy cattle

.

J. Dairy Sci

.

81

:

1148

–

1156

. doi:10.3168/jds.S0022-0302(98)75677-2

Enting

,

H.

,

D.

Kooij

,

A. A.

Dijkhuizen

,

R. B. M.

Huirne

, and

E. N.

Noordhuizen-Stassen

.

1997

.

Economic losses due to clinical lameness in dairy cattle

.

Livest. Prod. Sci

.

49

:

259

–

267

. doi:10.1016/S0301-6226(97) 00051-1

Giess

,

L. K.

,

B. R.

Jensen

,

R. L.

Weaber

,

J. M.

Bormann

, and

W. A.

Fiske

.

2018

.

Feet and leg traits are moderately to lowly heritable in Red Angus cattle

.

Kans. Agric. Exp. Stn. Res. Rep

.

4

. doi:10.4148/2378-5977. 7533

Gilmour

,

A. R.

,

B. J.

Gogel

,

B. R.

Cullis

, and

R.

Thompson

.

2009

.

ASReml user guide release 3.0

.

VSN International Ltd,

Hemel Hempstead, UK

.

Golden

,

B. L.

,

W. M.

Snelling

, and

C. H.

Mallinckrodt

.

1992

.

Animal Breeder's ToolKit user's guide and reference manual

.

Colorado State Univ. Agric. Exp. Stn. Tech. Bull., LTB92-2,

Fort Collins

,

CO

.

Jeyaruban

,

G.

,

B.

Tier

,

D.

Johnston

, and

H.

Graser

.

2012

.

Genetic analysis of feet and leg traits of Australian Angus cattle using linear and threshold models

.

Anim. Prod. Sci

.

52

:

1

–

10

. doi:10.1071/AN11153

McGuire

,

C. L.

,

J. G.

Powell

,

R.

Cauble

,

B.

Kutz

,

K.

Anschutz

,

T. D.

Lester

, and

W.

Gragg

.

2019

.

Foot scoring effects on cow and calf performance at breeding and weaning

.

J. Anim. Sci

.

97

(

Suppl. 1

):

50

. (Abstr.). doi:10.1093/jas/skz053.112

R Core Team.

2019

.

R: A Language and environment for statistical computing

.

R Foundation for Statistical Computing,

Vienna (Austria)

.

Vargas

,

G.

,

H. H. R.

Neves

,

V.

Cardoso

,

D. P.

Munari

, and

R.

Carvalheiro

.

2017

.

Genetic analysis of feet and leg conformation traits in Nelore cattle

.

J. Anim. Sci

.

95

:

2379

–

2384

. doi:10.2527/jas.2016.1327

Wang

,

L.

,

S. P.

Miller

,

K. J.

Retallick

, and

D. W.

Moser

.

2017

.

Genetic parameter estimation for foot structure in American Angus cattle

.

J. Anim. Sci

.

95

(

Suppl. 4

):

100

. (Abstr.). doi:10.2527/asasann.2017.203

© The Author(s) 2021. Published by Oxford University Press on behalf of the American Society of Animal Science.

This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial License (https://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com

© The Author(s) 2021. Published by Oxford University Press on behalf of the American Society of Animal Science.