Genetic parameters for mouth size and their influence on growth traits in pasture-raised Nelore cattle

Abstract

Evaluating traits that allow breeders to increase production efficiency in beef production systems is important. The mouth size (MS) score is a trait easily measured and implemented by breeders. Bite size is related to MS in beef cattle and is a determinant of daily feed intake of pasture-raised animals, influencing their growth. The aim of this study was to estimate genetic parameters for MS, weaning weight (WW) and postweaning weight gain (PWG) of Nelore cattle and to evaluate the influence of the interaction between MS and WW on PWG. Phenotypic records of 134,282 Nelore animals born between 1995 and 2019 were used. Variance components were estimated using multitrait animal model with the Bayesian method. The model included the contemporary group as fixed effect, age at measurement of the trait as linear covariate, and direct additive genetic and residual effects as random effects. For WW, random maternal and maternal permanent environmental effects were added to the model. A Bayesian approach was used to analyze the interaction between WW clusters and MS. The heritability estimates were 0.24, 0.15, and 0.23 for MS, WW, and PWG, respectively. The genetic correlation between variables studied ranged from 0.24 to 0.46. The results suggest that animals with a larger mouth tend to have greater PWG, demonstrating the positive influence of MS score on the postweaning performance of cattle. The direct heritability estimates confirm the possibility of selecting animals for the traits studied.

In this study, we estimated genetic parameters for mouth size (MS) and the growth traits of weaning weight (WW) and postweaning weight gain (PWG) of pasture-raised Nelore cattle. We evaluated the influence of the interaction between MS and WW on PWG using the Bayesian method. This approach is relevant because it helps breeders to identify and select more efficient animals. Our results showed that to accelerate animal development, mount size can be used as a selection criterion together with growth traits in breeding programs of beef cattle.

Introduction

The growth of an animal is a complex biological phenomenon that involves genetic and environmental factors and that breeders can evaluate by monitoring body weight gain. Growth traits such as weaning weight (WW) and postweaning weight gain (PWG) are widely used as selection criteria in genetic improvement programs of beef cattle because they are easily obtained and are considered important for evaluating herd efficiency (Kamei et al., 2017).

In an attempt to promote greater individual weight gains and to increase herd efficiency under pasture conditions, it is important to evaluate mouth size (MS) since animals with a larger mouth tend to have a larger bite size, reducing grazing time and increasing rumination time (Erlinger et al., 1990). The use of visual scores is one approach to evaluate this trait. These scores are easy to measure during handling but are also suitable for camera measurement, which has gained space in precision agriculture.

Bite size or intake per bite is the main determinant of daily feed intake in pasture-raised cattle (Dougherty et al., 1990; Brâncio et al., 2003) and is affected by forage height and leaf density (Brâncio et al., 2003; Teixeira et al., 2010). On quality pasture with adequate height and high leaf density, animals take bites of sufficient size to ensure dry matter intake, resulting in greater weight gains per animal (Brâncio et al., 2003). It would therefore be interesting to evaluate animals with different MSs reared under the same pasture conditions in order to determine whether or not MS influences the productive performance of cattle.

Estimating genetic parameters, such as the heritability of MS and its correlation with other traits already used as selection criteria, would permit strategies to be developed for the selection of animals in breeding programs in pasture-raised cattle. This would assist breeders identify more productive animals for a given farming system. Within this context, knowledge about the genetic parameters for MS, WW, and PWG would be important to decide on the use of MS in breeding programs of pasture-raised beef cattle.

The aim of the present study was divided into two steps: 1) to estimate genetic parameters for MS, WW, and PWG of pasture-raised Nelore cattle, and 2) to evaluate the influence of the interaction between MS and WW on PWG.

Materials and Methods

Ethics statement

Approval of the Animal Care and Use Committee was not required for this study because the data were acquired from an existing beef industry database.

Data description

Phenotypic records of 134,282 Nelore animals born between 1995 and 2019 were used. The animals belong to farms that participate in the Qualitas Nelore Breeding Program, located in eight different Brazilian states. The pedigree file contained information from six generations, totaling 200,958 individuals born to 1,537 sires and 69,541 dams.

The following traits were analyzed: WW, measured close to 210 d of age (minimum of 160 and maximum of 300 d); PWG, calculated as the difference between weight adjusted to 450 d of age and weight adjusted to 210 d of age, and MS, evaluated by visual scoring close to 450 d of age (minimum of 390 and maximum of 510 d) in which a score of 1 to 5 was attributed, where 1 corresponds to a small mouth and 5 to a large mouth (Figure 1). This trait was evaluated by two expert technicians with previous training. The distribution of the phenotypes of MS is presented in the Supplemental Material (Figure S1).

Figure 1.

Mouth size illustration. MS1, mouth size score 1; MS2, mouth size score 2; MS3, mouth size score 3; MS4, mouth size score 4; MS5, mouth size score5.

The contemporary group (CG) for WW was defined by the effects of farm, birth year (animals born from March of year ‘x’ to February of year ‘x+1’, where x ranged from 1995 to 2019), date of measurement, sex, and management group at weaning. The same CG was used for PWG and MS but date of measurement and management group postweaning were amended.

The effects included in the CGs had a significance of less than 0.01 in previous analyses. For all traits, CGs with fewer than three animals and phenotypes with 3.5 standard deviations above or below the mean of the CG were excluded from the analyses. Table 1 shows the structure of the database evaluated.

Table 1.

Description of the database structure of mouth size and growth traits of Nelore cattle and of the weaning weight clusters formed

	Unit	N	Mean (SD)	Min	Max	CV%	N_CG
Trait; abbreviation
WW; weaning weight	kg	134,282	194.8 (32.9)	74	388	16.9	3,532
PWG; postweaning weight gain	kg	81,316	96.3 (40.1)	1	399	41.6	5,532
MS; mouth size	score	76,178	3.0*	1	5	–	5,532
Cluster_WW; abbreviation
C_LWW; cluster of low weaning weight	kg	24,236	169.4 (14.5)	80.0	187	8.6	–
C_MWW; cluster of medium weaning weight	kg	34,774	205.0 (10.3)	187.3	224	5.0	–
C_HWW; cluster of high weaning weight	kg	16,713	243.6 (16.7)	224.5	364	6.9	–

N, = number of observations; SD, standard deviation; CV%, coefficient of variation in percent; N_CG = number of contemporary groups. *Mode.

To determine the influence of the interaction between MS and WW on PWG, the animals with phenotypes for PWG were divided into three clusters based on WW (Table 1) using nonhierarchical cluster analysis performed with the kmeans procedure implemented in the R software (R Core Team, 2020). The three WW clusters were formed to ­evaluate the interaction between WW clusters and MS scores. The hypothesis was that animals of C_LWW (cluster of low WW) with a larger mouth (score 5) would have greater PWG than animals of the same WW cluster but with a smaller mouth (score 1).

Estimation of genetic parameters

The (co)variance components were estimated by the Bayesian approach under a multitrait animal model using the GIBBS2F90 program (Misztal et al., 2002). The mixed model adopted can be written in matrix form as:

$${\displaystyle \begin{array}{l}{\displaystyle y=X\beta +Z\alpha +Mm+Wc+e}\end{array}}$$

where y is the vector of dependent variables (WW, PWG, and MS); $\beta$ is the vector of systematic effects (CG and age at measurement as linear covariate) for each trait; X is the incidence matrix associating $\beta$ with y; $\alpha$ is the vector of random direct additive genetic effects; Z is the incidence matrix associating $\alpha$ with y; m is the vector of random maternal additive genetic effects; M is the incidence matrix associating m with y; c is the vector of random maternal permanent environmental effects; W is the incidence matrix associating c with y (the last two vectors and incidence matrices were used only for WW); and e is the vector of random residual effects.

Maternal additive and permanent environmental effects were not considered for PWG and MS because these phenotypes were evaluated after weaning; it was therefore assumed that the maternal effect was diluted during postweaning growth of the animal (Albuquerque and Meyer, 2001). The data structure of commercial herds is inadequate to estimate the covariance of direct and maternal genetic effects (Texeira and Albuquere, 2005); thus, the covariances with the maternal genetic effect of WW were considered to be zero for the traits included in the model.

A uniform a priori distribution was defined for the systematic effects (β). Gaussian and inverted Wishart distributions were adopted as a priori distributions for random effects and (co)variance components, respectively (Van Tassel and Van Vleck, 1996). The marginal distributions are represented as follows:

β α constant;

$${\displaystyle \begin{array}{l}{\displaystyle \left(\begin{array}{l}a\\ m\end{array}\right)|G_a,G_m,G_{am}\sim MVN(\left(\begin{array}{l}0\\ 0\end{array}\right),\left(\begin{array}{l}A\otimes G_{a}A\otimes G_{am}\\ A\otimes {G^{\prime }}_{am}A\otimes G_m\end{array}\right))}\end{array}}$$

$${\displaystyle \begin{array}{l}{\displaystyle p|\mathrm{P}\sim MVN[0,\left({\mathrm{I}}_{\mathrm{n}}\mathrm{P}\right)];}\end{array}}$$

$${\displaystyle \begin{array}{l}{\displaystyle {\mathrm{G}}_{\mathrm{a}}|S_a,v_a\sim IW[S_a{v}_a,v_a];}\end{array}}$$

$${\displaystyle \begin{array}{l}{\displaystyle {\mathrm{G}}_{\mathrm{m}}|S_m,v_m\sim IW[S_m{v}_m,v_m];}\end{array}}$$

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{P}|S_p,v_p\sim IW[S_p{v}_p,v_p];}\end{array}}$$

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{R}|S_r,v_r\sim IW[S_r{v}_r,v_r],}\end{array}}$$

where A is the relationship matrix; G, P and R are the (co)variances matrices of direct and maternal genetics effects, permanent environmental effects and residual effects, respectively; $I_n$ is the identity matrix; $\otimes$ is the Kronecker product; $S_a$ and $v_a$, $S_v$ and $v_m$, $S_p$ and $v_p$, and $S_r$ and $v_r$ are the a priori values and degrees of freedom for direct additive genetic, maternal additive genetic, permanent environmental and residual (co)variances, respectively (Van Tassel and Van Vleck, 1996). Genetic (co)variance estimates reported in the literature were used as a priori values in the analyses.

A single chain with a length of 100,000 samples was defined for Gibbs sampling, discarding the first 10,000 samples and collecting samples at every 50 cycles, resulting in 1,800 samples for inference.

The convergence was verified assessing the effective chain size and the test proposed by Geweke (1991). The convergence diagnosis by the effective chain size is based on means and variances, and its efficiency is confirmed if the evaluated components present posterior Gaussian distribution (Gelman et al., 2013).

The posterior (co)variances were used to estimate measures of central tendency of direct and maternal additive heritabilities (h²a and h²m, respectively), as well as the posterior standard deviation and higher density intervals to 95%.

The following equation was applied to determine whether indirect selection would be more efficient than direct selection for the traits studied (adapted from Van Vleck et al., 1987):

Response efficiency (%):

$${\displaystyle \begin{array}{l}{\displaystyle \frac{\mathrm{C}{\mathrm{R}}_{1(2)}}{\mathrm{D}{\mathrm{R}}_1}=\left[{\mathrm{r}}_{\mathrm{A}1.\mathrm{A}2}.\frac{\surd {\mathrm{h}}_2^2}{\surd {\mathrm{h}}_1^2}\text{-}1\right]\mathrm{x}100}\end{array}}$$

where CR₁₍₂₎ is the correlated response of trait 1 when selected for trait 2; DR₁ is the direct response of trait 1; r_A1.A2 is the genetic correlation between the two traits; $h_1^2\mathrm{a}\mathrm{n}\mathrm{d}{h}_2^2$ are the estimated heritabilities for traits 1 and 2, respectively.

A new analysis was used to evaluate the interaction between WW clusters and MS using the GIBBS2F90 software ­(Misztal et al., 2002) and applying a single-trait animal model for PWG. The model included the systematic effects of CG and age of animal as linear covariate, as well as the effect of the WW cluster-MS interaction. The variances estimated in the previous analysis were defined as known values and added to the “OPTION fixed_var all” command line in the parameter file of the GIBBS2F90 program.

A sample chain was requested to obtain the solutions of the effects included in the model. For each level of systematic effects, 9,000 solutions were obtained in a Markov chain of 100,000 cycles, with a burn-in period and thinning interval of 10,000 and 10 cycles, respectively. Comparisons of the different levels of the effect of WW cluster-MS interaction according to PWG were made using the t-test, considering values of less than 5% to be significant.

Results

Genetic parameter estimates

According to the samples of the variance posterior densities, the central tendency measures such as mean, median, and mode of the later marginal densities of heritability estimates showed similarity, demonstrating symmetry and convergence (Table 2). The heritability estimates (h²) obtained for the traits were of low to moderate magnitude. The h² coefficient for MS was 0.24, suggesting a moderate response to direct selection. The h² obtained for the maternal component of WW was 0.12 ± 0.01 and the maternal permanent environmental variance effect for WW was 111.03 ± 3.31. The genetic correlations between traits ranged from 0.24 to 0.46. The genetic correlation between MS and WW was moderate (0.46), indicating pleiotropy among these traits. The standard errors for all estimates were less than 0.05.

Table 2.

Descriptive statistics of the posterior estimates of genetics parameters for growth traits and mouth size in Nelore cattle

Trait	Parameter	Mean ± SE	Min	Max	Median	Mode	HPD 95%
WW	h2	0.15 ± 0.01	0.12	0.18	0.15	0.15	0.15	to	0.15
WW	h2m	0.12 ± 0.01	0.09	0.14	0.12	0.12	0.12	to	0.12
PWG	h2	0.23 ± 0.01	0.21	0.27	0.23	0.23	0.23	to	0.23
MS	h2	0.24 ± 0.01	0.20	0.27	0.24	0.24	0.24	to	0.24
WW - PWG	RG	0.33 ± 0.01	0.19	0.47	0.33	0.33	0.33	to	0.34
WW - MS	RG	0.46 ± 0.01	0.33	0.58	0.47	0.45	0.46	to	0.47
PWG - MS	RG	0.23 ± 0.01	0.13	0.36	0.23	0.24	0.23	to	0.24
WW - PWG	RE	−0.14 ± 0.01	−0.18	−0.11	−0.14	−0.14	−0.14	to	−0.14
WW - MS	RE	0.11 ± 0.01	0.08	0.14	0.11	0.11	0.11	to	0.11
PWG - MS	RE	0.10 ± 0.01	0.07	0.13	0.10	0.10	0.10	to	0.10
WW	PV	513.28 ± 0.06	504.50	523.77	513.35	512.13	513.16	to	513.40
PWG	PV	331.88 ± 0.05	325.99	337.89	331.88	331.98	331.79	to	331.97
MS	PV	0.33 ± 0.01	0.32	0.34	0.33	0.33	0.33	to	0.33

h², direct additive heritability; h²m, maternal additive heritability; RG, genetic correlation; RE, residual correlation; PV, phenotypic variance; HPD 95%, 95% highest density intervals.

Response efficiency

The relative response efficiency of indirect selection for MS using WW and PWG indicated that direct selection will always be more efficient than indirect selection (Table 3).

Table 3.

Efficiency of correlated response of growth traits and mouth size in Nelore cattle

Selected trait1	Correlated response in:	Efficiency (%)
MS	WW	−41.8
MS	PWG	−75.5
WW	MS	−63.6
PWG	MS	−76.5

¹See Table 1 for abbreviations of trait names.

Influence of the interaction between mouth size and weaning weight on postweaning weight gain

Analysis of the influence of different MS scores within the same WW cluster on PWG showed significant differences (P < 0.01), in which animals with higher scores tended to obtain greater PWG (Tables S1 and 4).

The results shown in Table 4 indicate that regardless of whether they had low, medium, or high WW, a larger mouth resulted in greater gains than a small mouth, indicating that this trait is a decisive factor for PWG. This gain related to a large as compared to a small mouth is greater in animals that are heavier at weaning, i.e., the best set is to select heavier animals at weaning with a high MS score.

Table 4.

Differences between adjusted means of postweaning weight gain (PWG) according to weaning weight cluster (Cluster_WW) and mouth size, with 95% prediction interval (95% PI)

Type of contrast	Cluster_WW * MS contrasts	Adjusted means of differences for PWG	Differences with 95% PI for PWG (kg)
Within-cluster contrasts	C_LWW * MS5 vs. C_LWW * MS1	12.30	12.24 to 12.35
	C_MWW * MS5 vs. C_MWW * MS1	13.38	13.34 to 13.42
	C_HWW * MS5 vs. C_HWW * MS1	15.21	15.14 to 15.28
Between-cluster contrasts	C_LWW * MS1 vs. C_HWW * MS1	5.72	5.65 to 5.79
	C_LWW * MS5 vs. C_HWW * MS1	18.02	17.94 to 18.10
	C_HWW * MS5 vs. C_LWW * MS1	9.49	9.44 to 9.54
	C_LWW * MS5 vs. C_HWW * MS5	2.81	2.75 to 2.87

¹See Table 1 for abbreviations of trait names. All comparisons were statistically significant at the 1% significance level (t-test, P < 0.01).

Discussion

The MS score is a trait that is easily measured and implemented without costs for breeders, because it is not necessary to invest in specific equipment and it can be done while handling the cattle for weighing. The heritabilities and genetic correlations demonstrate that MS is favorably associated with growth traits and can be used as a selection criterion, causing no losses in WW or PWG. The genetic correlations suggest that animals with a larger mouth are heavier at weaning and gain more weight after weaning on pasture.

A bite is the act of grasping forage with the teeth and bite size is related to MS (Erlinger et al., 1990). Feed intake in grazing cattle is the sum of forage consumed per bite; thus, the maximization of each bite during grazing allows an animal to maximize feed intake and consequently animal production (Teixeira et al., 2010). Variations in bite size may occur due to the size of the animals (Erlinger et al., 1990) and the height and leaf density of the forage species (Dougherty et al. 1992). According to Chacon et al. (1978), animals compensate for variations in bite size by changes in grazing time (smaller bite size, longer grazing time). However, animals may not be able to compensate for reductions in bite size with increased grazing time and bite rate (Dougherty et al., 1990). Consequently, bite size or intake per bite is a determinant of daily feed intake in grazing animals. The genetic correlation and cluster analysis results of our study showed that animals with a higher MS score tend to gain more weight during the postweaning period, suggesting an influence of MS on animal growth. This superior PWG of animals with a larger mouth suggests that a greater MS allows to maximize forage intake, reducing grazing time (Chacon et al., 1978) and consequently increasing the PWG of animals. Thus, MS score could be used in defining production efficiency in cattle, as an indicator of forage intake varying in mature size. However, further studies including the mentioned parameters are needed to elucidate what actually happens. It is important to highlight that selection for growth traits can generate an increase in mature size in cattle that may not necessarily be advantageous (Arango and Van Vleck, 2002). The positive genetic correlation between birthweight and mature weight meant that the high mature size tended to be heavier weights at birth, increasing the incidence of dystocia (Hickson et al., 2006). Furthermore, changes in the mature size may consequently affect nutritional requirements (Arango and Van Vleck, 2002). Therefore, caution should be taken when selecting traits related to growth.

Comparison of MS between the different WW clusters showed that having heavier calves is not enough to obtain greater PWG since animals of the high WW cluster with a small mouth (C_HWW*MS1) gained 18.02 kg less weight than animals with a low WW and a large mouth (C_LWW*MS5). This finding indicates a possible compensatory gain of light weaned animals, probably because these animals have less access to milk due to reduced maternal ability when compared to heavy animals. Our results showed higher mean estimates to maternal additive of cows that weaned calves with high WW (4.23 kg) compared to cows weaned calves with low WW (−1.26 kg), indicating the influence of the cow in the calves WW. Thus, after a period of nutritional restriction, animals undergo compensatory growth if they have access to a diet whose quantity and quality are adequate (Silva et al., 2022).

A difference of 5.72 kg was observed for PWG among animals from the C_LWW and C_HWW clusters with the same mouth score (MS1), in which C_LWW had the higher value. The difference can be related to compensatory gain. Comparing the previous clusters, but with the C_LWW with MS5 score (C_LWW*MS5 vs. C_HWW*MS1), the C_LWW cluster presented higher PWG (18.02 kg). If the value attributed to compensatory gain (5.72 kg) is removed from the mean, the resulting value is 12.3 kg for PWG for animals in the C_LWW cluster with MS5 score, which highlights the importance of larger MS for better postweaning growth.

Future studies should be performed to assess variations in MS over time. We had only one MS measurement for our research, so it was not possible to estimate its repeatability. However, due to the different growth rates of body tissues, e.g., muscle tissue grows faster than bone tissue (Lawrence et al., 2012), not to be expected large variance in short periods for MS. In addition, as animals reach maturity, head weight tends to be lower in proportion to body weight (Lawrence et al., 2012), indicating that weight gain occurs mainly due to increased muscle mass, so it would not be expected significant variations of MS during the evaluation time of the animals in this study.

The adoption of traits related to the growth and development of individuals, such as body weight and weight gain, as selection criteria is justified by the fact that they are routine traits that can be easily obtained, are directly linked to meat production, and show an expressive response to selection, resulting in high genetic progress (Souza et al., 2018). WW is used to evaluate the growth potential of calves and the maternal ability of cows by determining the effect of the animal’s own genes on growth (direct effect) and the effect of maternal genes on offspring development (maternal effect). The direct (0.15) and maternal (0.12) heritability estimates for WW obtained here indicate a substantial additive genetic effect (27%) influencing the phenotypic variance of WW. Still, nonadditive and environmental effects have a greater influence on the expression of this trait in the population. Similar results have been reported by Silva Neto et al. (2020) and Oliveira et al. (2021) for Nelore cattle, with estimates of 0.14 and 0.13 for direct effects and of 0.11 and 0.09 for maternal effects, respectively. Conversely, Schmidt et al. (2019) and Silveira et al. (2021) reported higher heritability (0.23 and 0.24, respectively) for direct effect of WW in the same breed. The differences in the magnitude of heritability may be due to the sizes of the herds analyzed and data structure, because these factors can lead to changes in frequencies and, consequently, in the genetic variances of the population (Falconer and Mackay, 1996).

The heritability estimates observed for PWG and MS indicate that a part of the phenotypic variation in these traits is influenced by direct additive genetic effects and can therefore be transmitted to subsequent generations. Similar to the estimate found in the present study for PWG, Santana et al. (2015) and Prestes et al. (2019), using different models, reported heritability estimates for PWG ranging from 0.15 ± 0.01 to 0.29 ± 0.02 (in a multibreed population) and from 0.27 ± 0.01 to 0.30 ± 0.06 (in Nelore cattle), respectively. In Nelore cattle, Araujo Neto et al. (2018) obtained higher estimates (0.26 to 0.52). According to the authors, the differences between their results and those reported in the literature are probably due to differences in the models applied and to the use of different herds.

The genetic correlations between MS and the growth traits (WW and PWG) were favorable and of low to moderate magnitude (0.24 and 0.46, respectively). Thus, selection for greater MS should benefit animal development, increasing WW and PWG. Although the traits are genetically correlated, the relative efficiency of indirect selection for the different traits indicates that direct selection is more efficient than indirect selection. These results suggest that, to accelerate animal development, MS can be used as a selection criterion together with growth traits in breeding programs of beef cattle. Using PWG as the goal trait and constructing indexes that either included or excluded MS as a selection criterion alongside WW and PWG, the extent of that potential advantage could be tested.

Since MS is evaluated based on visual scores, its measurement using more accurate methods may also improve the heritability estimates for the trait. Given the advances in precision agriculture and in the large-scale acquisition of phenotypic data using cameras, MS is suitable for this purpose and further in-depth studies focusing on this trait should be conducted.

Conclusion

WW and MS score influenced PWG. The results showed that a larger mouth is a decisive factor to obtain greater PWG in the beef cattle population studied. The direct heritability estimates confirmed the possibility of selecting animals for the traits studied, especially MS which is economically important considering its favorable genetic correlation with WW and PWG, the main traits used in selection indices for pasture-raised cattle.

Abbreviations

C_HWW

cluster of high weaning weight

C_LWW

cluster of low weaning weight

C_MWW

cluster of medium weaning weight

MS

mouth size

MS1

mouth size score 1

MS5

mouth size score 5

PWG

postweaning weight gain

WW

weaning weight

Conflict of Interest Statement

The authors declare no real or perceived conflicts of interest.