Archivos Latinoamericanos de Producción Animal. 2020. 28 (1­2)  
Additive and non­additive effects for mature weight in beef cattle  
Ana Claudia Guillenea1  
Mario Lema2  
Diego Gimeno3  
Olga Ravagnolo2  
Ana Carolina Espasandín3  
Facultad de Agronomía, Universidad de la República, Avenida Garzón 780, CP 12900.  
Abstract: Selection has emphasized animal growth, leading to an increase in their mature size affecting in some  
cases the pregnancy of the cows and the efficiency of the systems. Usually, crossbreeding improve productivity  
because of the genetic effects that the cows exploit, but the impact on mature weight (MW) has not been studied. The  
present study aimed at estimating MW and genetic parameters associated with the MW in crosses between two  
British breeds: Hereford (H/H) and Angus (A/A), a Continental: Salers (S/S), and a Zebu: Nelore (N/N). MW was  
analyzed at 4; 4.5; 5; 5.5 and 6 years of age using a repeated­measure sire model. For parameters estimation, an  
additive – dominant model was used including the fixed effects of breed group, contemporary group, and age as  
covariate linear and quadratic, with the linear regression fitted by breed group. Permanent environmental and sire  
were included as random effects. According to the results, it is expected to observe heterosis between H/H and N/N,  
however, the structure of the data may not be enough for estimate accurately the genetic parameters in this trait. The  
A/H, N/H, S/H, S/SH and H/NH cows were heavier than the H/H cows. All the breed groups continue gaining  
weight until six years of age. The results revealed that British crossbred animals are heavier than H/H at the first  
crossing but not in the following. Crossbred cows with proportions of 0.5 and greater for the Continental breed are  
heavier than H/H cows. Crosses between British and Zebu cows have higher mature weight than H/H at the first  
crossing and in backcrosses toward the British in all ages.  
Key words: Crossbreeding, Heterosis, Cows, Maturity  
Efectos aditivos y no aditivos para el peso adulto en ganado de carne  
Resumen. La selección ha enfatizado el crecimiento de los animales llevando a un aumento en el tamaño adulto,  
que afecta en ocasiones la preñez de las hembras y la eficiencia del sistema. Los cruzamientos generalmente  
mejoran la productividad debido a los efectos genéticos que explotan las hembras, pero su impacto en el peso adulto  
PA) no se ha estudiado. El objetivo del presente estudio fue estimar los parámetros genéticos asociados al PA y el  
PA en cruzamientos entre razas británicas: Hereford (H/H) and Angus (A/A), una continental: Salers (S/S) y una  
cebú: Nelore (N/N). El PA fue analizado a los 4; 4.5; 5; 5.5 y 6 años de edad utilizando un modelo padre de medidas  
repetidas. Para la estimación de los parámetros, se utilizó un modelo aditivo dominante que incluyó los efectos fijos  
del grupo racial, grupo contemporáneo y edad como covariable lineal y cuadrática, con la regresión lineal ajustada  
por grupo racial. El ambiente permanente y el padre fueron incluidos como efectos aleatorios. Según los resultados,  
se espera observar heterosis entre H/H y N/N, sin embargo, la estructura de los datos puede no haber sido  
suficiente para estimar con precisión los parámetros genéticos en este rasgo. Las vacas A/H, N/H, S/H, S/SH y  
H/NH superaron a las H/H. Todos los grupos raciales continuaron aumentando de peso hasta los seis años de edad.  
Los resultados revelaron que los animales cruzas británicos son más pesados que H/H en el primer cruce (F1), pero  
no en los siguientes. Las vacas cruzas con proporciones de 0.5 y mayores para la raza Continental son más pesadas  
que las vacas H/H. Los cruzamientos entre vacas británicas y cebú tienen un mayor peso adulto que las H/H en el  
primer cruzamiento y en retrocruzas hacia la británica en todas las edades.  
Palabras clave: Cruzamientos, Heterosis, Vacas, Madurez  
Recibido: 2020­07­05. Aceptado: 2020­09­22  
Autor para la correspondencia:  
Instituto Nacional de Investigación Agropecuaria, INIA Treinta y Tres, Ruta 8 Km 282, CP 33000. Treinta y Tres, Uruguay.  
Secretariado Uruguayo de la Lana, Camino Servando Gómez 2408, CP 11200. Montevideo, Uruguay  
Instituto Nacional de Investigación Agropecuaria, INIA Las Brujas, Ruta 48 Km 10, CP 90200. Canelones, Uruguay.  
Guillenea et al.  
The mature cow weight (MW) is a crucial trait to  
selection and complimentary breed matings to choose  
the most appropriate genetic resource to obtain the  
highest economic profit. Crossbreeding parameters  
are estimated using genetic models to establish the  
difference in genetic merit of breeds and to select the  
most suitable breed group (Gregory and Cundiff,  
determine the profitability and sustainability in the  
production system (Costa et al., 2011). About 70 % of  
the energy requirements are needed for maintenance  
in the beef breed cycle (Ferrell and Jenkins, 1985). A  
bioeconomic model in Uruguay determined that  
increasing maternal weight at weaning has a negative  
economic impact on the beef cattle system since it led  
to higher feeding requirements of cows without  
increasing incomes (Pravia et al., 2014). Nevertheless,  
the authors found that carcass weight was the second  
trait in economic importance because of its  
contribution to income. MW have medium to high  
heritability and a positive genetic correlation with  
growth traits (Junior et al., 2019). This could lead to  
an increase in MW when selecting animals for higher  
growth rates. However, these animals are more  
efficient only when food is not restricted, but under  
limited resources, as in grazing systems, they are the  
least productive (Jenkins, 2009). An important point  
to manage the resources in a calf­producer enterprise  
is to understand in depth the effects involved in MW  
of different breed groups.  
The most common models used in large beef cattle  
populations are the additive – dominant model  
(Gardner and Eberhart, 1966) and the Dickerson  
model (Dickerson, 1969, 1973). The additive –  
dominant model assumes that heterosis occurs for  
dominance effects. Riley and Crockett (2006) suggest  
that the expression of heterosis is determined by  
dominance effects at many genes, so that the heterosis  
is proportional to heterozygosity. The first cross of two  
animals belonging to two pure breeds results in a  
heterozygous animal at all its loci. Consequently,  
according to the additive – dominant model, F1  
animals would be expected to express maximum  
heterosis, while F2 would be expected to retain half of  
the heterosis observed in the F1 generation. Additive  
and dominance effects are relatively simple to model;  
conversely, epistasis effects are more complex. The  
estimability of these parameters depends on complex  
designs that include many animals and breed groups  
(Kinghorn and Vercoe, 1989). Despite its importance,  
research studies to estimate crossbreeding parameters  
in beef cattle are scarce, particularly in mature traits.  
The objectives of this study were: (1) to estimate MW  
for Hereford and different crossbred cows, and (2) to  
estimate crossbreeding parameters in terms of  
additive and heterotic effects for MW of Hereford,  
Angus, Salers, and Nellore breed in temperate climate  
under grazing conditions.  
The main tools to improve animal production  
through genetic are selection within and among breeds  
and crossbreeding. Crossbreeding are widely used for  
their ability to take advantage of complimentary for  
many traits and to exploit additive and non­additive  
genetic differences among breeds. Biological variation  
has been reported to exist among cattle breeds for  
birth weight, meat quality, growth, and final weight  
Kuehn and Thallman, 2016). The breed origins of  
beef cattle are widely bunched in Bos indicus and Bos  
taurus, and the last ones, in turn, in British and  
Continental breeds. Different breed origins allow  
Materials and Methods  
Management of the experiment  
Experimental design  
The data used in this study were records of cow  
weight from beef herds of several experiments carried  
out in a commercial farm called “Capilla Vieja”, located  
in the department of Paysandú, Uruguay that was part  
of a research project between Caja de Notarial de  
Seguridad Social and Facultad de Agronomía,  
Universidad de la República. The experiments were  
developed to estimate crossbreeding genetic  
parameters to evaluate the importance of different  
gene action (additive and non­additive) affecting  
economically important traits, as well as to assess the  
relative production of crossbreds with a different breed  
composition (Gimeno et al., 1995).  
Four breeds of beef cattle were used: Hereford (H/H)  
and Aberdeen Angus (A/A) as representatives of  
British breeds, Salers (S/S) as Continental and Nellore  
(N/N) as Zebu breed.  
The initial herd consisted of only H/H females and  
S/S and N/N females were not available. Therefore,  
the crossbreeding scheme was optimized in order to  
design the best possible experiment given the  
restriction of having only one initial maternal breed to  
use and of not being able to have all the breed groups  
at the same time (Sölkner and James, 1990a). Gimeno  
et al. (1995) describe the strategy used in the design of  
the experiments. The optimization was carried out  
ISSN 1022­1301. Archivos Latinoamericanos de Producción Animal. 2020. 28 (1­2): 20­28  
Crossbreeding parameters for beef cow  
using the D­Optimally criterion of the Optimum  
Design Crossbreed Experiment program (Sölkner and  
Fucks, 1994), considering the Dickerson full genetic  
model with maternal effects.  
pure females generated and analyzed in this study  
correspond only to H/H cows.  
The mating season stretched over twelve weeks  
usually from 1 December to 20 February). The 698  
The experiment started in 1993 with a herd of 500  
adult H/H cows. Two years later pure A/A cows were  
also included in the experiment. Few weights of pure  
A/A cows were recorded, therefore there were not used  
in the analyses. Given the lack of maternal S/S and  
cows analyzed in this study are progeny of 71 bulls: 24  
A/A, 20 H/H, 14 N/N, and 13 S/S. The sires were  
selected prioritizing minimizing relationship, and they  
were used with different maternal breeds to obtain  
different breed groups of half­sibs from the same sire  
as recommend Sölkner and James (1990b). At least  
one bull was repeated per breed per year to connect  
the experiments. The herd remained under grazing  
conditions with 5 500 kg dry matter/ha/year of  
production approximately. No supplementary feed was  
provided during the experimental period. See Lema et  
al. (2011) for a detailed description of general  
management on the herd.  
The H/H cows were annually mated with A/A, H/H,  
N/N, and S/S. In 1995, the first F1 and pure H/H cows  
bred in the experiment were inseminated with the four  
purebreds and F1 bulls (A/H, N/H and S/H) born in  
the experiment. Consequently, F1 and backcrosses  
were produced during eight years (Table 1). Because of  
the lack of pure S/S and N/N females, the records of  
Table 1. Number of cows born between 1993 and 2004 by maternal and paternal breed group.a  
Sire breed group  
Dam breed group  
aSire and dam breed group at the left and right of the slash, respectively. Breed group: Angus (A/A), Hereford (H/H), Nellore (N/N), Salers  
S/S), Angus/Hereford (A/H), Nellore/Hereford (N/H) and Salers/Hereford (S/H)  
locked up approximately 12 hours. In grazing  
conditions with medium to a high quality of forage,  
these fasting hours represent a reduction of between 5  
and 7 % of the total live weight (Di Marco, 2006). The  
records of MW per breed group per age at  
measurement are presented in Table 2. The differences  
in age by breed group are common in the designs of  
crossbreeding experiments, where certain breeds are  
initially available, and they are required to generate  
subsequent generations. The H/H and F1 cows were  
born in 1993, and they were weighted during the whole  
experiment. The first backcrosses daughters of F1 dams  
were born in 1996, consequently, they do not reach the  
The mature weight was analyzed after 4 yr of age.  
After editing, data consisted of a set of 7651 weight  
measures of 698 cows of 10 different breed groups:  
H/H, A/H, N/H, S/H, H/AH, H/NH, H/SH, A/AH,  
N/NH, and S/SH. As H/H was the base­maternal  
breed, 93.5 % of the cows had H/H dam. The average  
number of records per cow was 7 (ranging from 2 for  
N/NH to 12 for S/H). Records of cows without date of  
birth or weighing, and unidentified parents or breed  
group were excluded from the analyses. Additionally,  
records outside the range ± 3 studentized residual (in  
absolute value) were deleted. Records of cows of less  
than 4 yr of age and older than 8.5 yr of age were also  
eliminated, as well as breed groups without  
observations from 6 yr of age (except N/NH because it  
was needed to estimate the crossbreeding parameters).  
Thus, the range of ages used for the analyzes was from  
yr of age in the trials. The experiments ended after  
weaning in autumn 2002, therefore the most advanced  
crossbred groups did not reach five years old. Thus  
recombination losses were not possible to estimate  
since the needed breed groups (Dickerson, 1969) did  
not reach necessary age for the expression of this trait.  
to 8.5 yr of age (5.4 yr on average), and the least­  
square means and crossbreeding parameters by breed  
group were estimated at 4, 4.5, 5, 5.5 and 6 yr of age.  
The first edition was conducted using R Software (R  
Development Core Team, 2016), and analyses were  
carried out using SAS (SAS Institute, 2014).  
The cows were weighed roughly every 45 days between  
997 and 2002. Before each measurement, they were  
ISSN 1022­1301. Archivos Latinoamericanos de Producción Animal. 2020. 28 (1­2): 21­28  
Guillenea et al.  
Table 2. Number of weight records according to age of the cow per breed group.a  
Age of cows (yr) at measurement  
Breed group  
7651 1775  
2617 2009  
Sire and dam breed group at the left and right of the slash, respectively. Breed group: Hereford (H/H), Nellore (N/N), Salers (S/S),  
Angus/Hereford (A/H), Nellore/Hereforf (N/H), Salers/Hereford (S/H), Hereford/Angus­Hereford (H/AH), Hereford/Nellore­Hereford  
H/NH), Hereford/Salers­Hereford (H/SH), Angus/Angus­Hereford (A/AH), Nellore/Nellore­Hereford (N/NH) and Salers/Salers­Hereford  
Two codes were defined to assign to each record the  
with less than five observations were excluded from the  
analysis. Age as covariate linear and quadratic was also  
included in the analyses, with the linear regression  
fitted by breed group.  
physiological and lactation status affecting MW. A  
pregnancy code took a value of 1 when the record was  
in the range of 300 days before calving (pregnant) and  
for the other records (empty). This range was taken  
as a reference since it was reported for Continental  
Model of analysis of mature weight  
breeds (Sobek et al., 2015) and Bos indicus breeds  
Chenoweth, 1994). The lactation status was assigned  
A repeated­measure sire model (Model 1) was used  
for the analyses of MW:’  
with another code that took a value of 1 when the  
record was between the calving date and one day after  
weaning (cows suckled), the records outside it took a  
value of 0 (dry). The records were assigned to their  
respective seasons, which were defined as summer  
from December to February, autumn from March to  
May, winter from June to August, and spring from  
September to November.  
yjklmn = μ + BG + CG + c + s + β x+ βx + e  
where yjklmn were the observations of MW of the l  
cow, of the j breed group, in the k contemporary  
group, progeny of the m sire, µ was the general mean,  
BG was the effect of the j breed group (j = 1, …, 10),  
CG was the effect of the k contemporary group (k = 1,  
Models of analysis of means and estimation of  
crossbreeding genetic parameters  
…, 55), c was the permanent environmental effect of  
the l cow ~ (0, σ2c) (l = 1, …, 698), s was the random  
effect of m sire ~ (0, σ ) (m = 1, …, 71), x was the age  
Preliminary analyses of variance were carried out to  
determine the relevance of fixed effects in the models  
of the lth cow in the n moment, β was the regression  
coefficient of weight on age x (linear) of the jth breed  
not shown). Significant effects were subsequently  
group, β was the regression coefficient of weight on age  
included in the analyses. As the mating was seasonal,  
calves were born in spring and early summer  
x (quadratic), e  
the jklmn observation ~ (0, σ ).  
was the random residual effect of  
September to December, mainly). They were allowed  
suckling and grazing with the cows, and weaned from  
their dams at about 6 months of age, with one weaning  
date for the whole herd. Due to this management,  
environmental effects and cow weights might be  
confounded. For this reason, contemporary groups  
The relationships among sires were not considered in  
these analyses. Statistical analyses were conducted  
using the PROC MIXED of SAS (SAS Institute, 2014),  
with the ESTIMATE statement to create the contrast  
and the LSMEANS statement to predict the means by  
breed group and age.  
CG) were defined by a combination of lactation and  
pregnancy status, year, and season of measure. CGs  
ISSN 1022­1301. Archivos Latinoamericanos de Producción Animal. 2020. 28 (1­2): 22­28  
Crossbreeding parameters for beef cow  
Model of analysis of crossbreeding effects  
the first model (Model 1), µ was an intercept, g was the  
additive effect of the breed i , h was the heterosis  
The estimation of crossbreeding parameters was  
carried out applying the additive – dominant model.  
Maternal effects were not included in these statistical  
analyses due to they being considered important only  
until weaning (Koch, 1972; Sölkner and James, 1990a).  
effect because of crossing ith with H/H, e was the  
random residual effect (var(e ) = σ V)). V was the  
error variance­covariance matrix of BGM , α was the  
proportion of breed i contribution to the cow, which  
was calculated as α = ½ (α + α ), and was expressed  
as difference from H/H: α =α ­αH/H, δiH denotes the  
The least­square means and variance­covariance  
matrix per breed per age obtained from Model 1 were  
subsequently used to estimate the crossbreeding  
parameters by generalized least­squared analysis  
probability that at a randomly chosen locus of an  
individual, one allele come from the ith breed and the  
S D  
other from the H/H. It was derived as δ = α α  
αH/HSα , where αiS and αiD were the breed  
GLS). To avoid linear dependencies between breed  
contribution of the sire and dam of the individual, and  
αH/HS and αH/HD were the H/H contribution of the sire  
and dam of the individual, respectively (Wolf et al.,  
proportions, the additive effect coefficients were  
calculated as deviation from H/H.  
The regression model used for the estimation of  
crossbreeding parameters (Model 2) was the following:  
These analyses were performed using the MIXED  
procedure of SAS (SAS Institute, 2014) with the  
ESTIMATE statement.  
BGM = μ + α* g + δ h + e  
iH iH  
where BGM was the vector of least­square means for  
MW for each breed group at each age estimated with  
Results and Discussion  
Average daily gain of different breed groups  
0.21 kg/d (H/NH). The growth rates tended to be more  
similar according to the H/H percentage than  
according to the breed origin. Only two breed groups  
had a significantly­different growth rate from H/H  
between 4 and 8.5 yr of age: S/H (0.16 kg/d, P < 0.001)  
and H/NH (0.21 kg/d, P = 0.05) (Figure 1).  
The linear regression coefficient of weight on age x (β)  
represents the average daily gain of the cows of each  
breed group. For the breed groups analyzed from 4 yr  
of age to 8.5 yr of age, these coefficients β are in the  
range from 0.15 kg/d (N/NH, N/H, H/H, and A/H) to  
Sire and dam breed group at the left and right of the slash, respectively. Breed group: Hereford (H/H), Angus/Hereford (A/H), Nellore/Hereford  
N/H), Salers/Hereford (S/H), Hereford/Angus­Hereford (H/AH), Hereford/Nellore­Hereford (H/NH), Hereford/Salers­Hereford (H/SH),  
Angus/Angus­Hereford (A/AH), Nellore/Nellore­Hereford (N/NH) and Salers/Salers­Hereford (S/SH)  
Figure 1. Average daily gain (β) from 4 to 8.5 yr of age per breed group. Error bars represent the s.e. of the means, while different  
letters indicate significant differences among the breed groups.  
ISSN 1022­1301. Archivos Latinoamericanos de Producción Animal. 2020. 28 (1­2): 23­28  
Guillenea et al.  
Previous reports have shown that Bos indicus × Bos  
breed (Wiltbank et al., 1969). The level of nutrition in  
this study is likely more similar to the low plane given  
grazing condition of our herd, so our result differs  
from this study since the average daily gain of  
crossbred cows involving A/A, and H/H did not differ  
significantly from H/H.  
taurus crossbred steers aged about 10 mo present  
higher average daily gain than Bos taurus × Bos  
taurus crossbred animals (0.18 and 0.07 kg/d,  
respectively) feeding ad libitum pasture hay diet  
Frisch and Vercoe, 1977). Other study compared  
average daily gain of H/H, A/A and their crossbred  
heifers on two levels of feed (high and low plane of  
nutrition) and they report gains of 0.82 and 0.73 kg/d  
for crossbred and purebred heifers, respectively, on  
the high plane, and 0.30 and 0.36 kg/d on the low,  
thus with an interaction between level of nutrition and  
Estimated mature weight of different breed  
The results of the least­square means by breed group  
from 4 to 6 yr of age will be presented in Table 3.  
Table 3. Mature cow weight (kg) for breed group by age.  
Age of cows (yr)  
Breed group  
357 ± 3.4  
368 ± 3.7  
376 ± 3.5  
383 ± 3.7  
387 ± 4.0  
367 ± 3.4*  
378 ± 3.3*  
386 ± 3.5*  
393 ± 3.7*  
397 ± 4.0*  
398 ± 3.9***  
377 ± 3.5***  
409 ± 3.8*** 417 ± 3.9***  
390 ± 3.5*** 400 ± 3.6***  
423 ± 4.0***  
409 ± 3.8***  
428 ± 4.3***  
415 ± 4.1***  
335 ± 13.8  
417 ± 13.2***  
360 ± 12.5  
353 ± 15.2  
375 ± 25.2  
395 ± 15.5*  
348 ± 13.0  
358 ± 13.4  
367 ± 14.9  
374 ± 17.3  
437 ± 12.2*** 456 ± 13.2*** 472 ± 16.0*** 487 ± 19.7***  
378 ± 11.4  
368 ± 14.5  
385 ± 16.8  
406 ± 14.7*  
394 ± 13.2  
380 ± 14.7  
393 ± 29.0  
415 ± 15.1*  
408 ± 16.9  
391 ± 15.8  
399 ± 48.1  
422 ± 16.5*  
420 ± 21.6  
399 ± 17.6  
404 ± 68.5  
427 ± 18.8*  
Breeds: A/A= Angus, H/H=Hereford, N/N=Nellore, S/S=Salers. Sire and dam breed group at the left and right of the slash,  
Asterisks indicate significant differences between breed group and Hereford: ***, P < 0.001 or * P < 0.05.  
All breed groups increased in their weight from 4 to 6  
yr of age. The weight gain between 4 and 6 yr of age of  
H/H and F1 was about 8 %. A similar pattern was  
observed in the backcrosses to N/N and S/S. On the  
other hand, the increase in that period in H/H  
backcrosses was between 12 and 17 %, which indicates  
that they reach the MW at ages more advanced. These  
patterns did not necessarily imply differences in MW.  
were also consistently heavier (P < 0.001) than H/H at  
every age by differences that remained at 41 kg (11 %).  
A similar situation was reported by Arango et al.  
(2004). They reported N/H to be heavier than H/H  
from 45 kg at 2 yr of age to 35 kg at 6 yr of age. The  
H/NH backcrosses showed the highest superiority  
from H/H ranging from 60 at 4 yr of age to 100 kg at 6  
yr of age (P < 0.001). The S/H cows were significantly  
heavier than H/H, with differences that tended to  
increase with age ranging from 20 (6 %) to 28 kg (7 %)  
(P < 0.001). The S/SH cows also showed a superiority  
from H/H of roughly 40 kg (P < 0.05). Arango et al.  
(2004) also founded that S/H outweighed H/H by  
approximately 60 kg from 2 to 6 yr of age.  
Estimates of breed­group means for MW were  
slightly lower than other estimates for the same breeds  
at similar ages for other authors (Arango et al., 2004).  
However, unlike the present study, they used  
improved pastures and supplemental silage or hay  
during the experimental period.  
The number of records used for estimating means of  
H/H and F1 was higher than for backcrosses, this is  
reflected in the standard error of the estimates. While  
for the F1 groups the standard errors are from 3.3 kg  
(A/H at 4.5 yr of age) to 4.0 kg (N/H at 6 yr of age), for  
backcrosses they are from 11.4 kg (H/SH at 4.5 yr of  
age) to 68.5 kg (N/NH at 6 yr of age).  
Weights of A/H were significantly heavier than H/H  
P < 0.05) by a difference of about 10 kg at every age,  
which is roughly 3 % of MW1. Arango et al. (2002)  
reported similar results, in their study reciprocal H/A  
cows were heavier than both A/A and H/H (P < 0.01)  
by differences that tended to increase with age from 18  
kg (2 yr of age) to 29 kg (7 yr of age). The N/H cows  
ISSN 1022­1301. Archivos Latinoamericanos de Producción Animal. 2020. 28 (1­2): 24­28  
Crossbreeding parameters for beef cow  
Table 4. Estimates of crossbreeding parameters for mature weight (kg) by age.  
Age of cow (yr)  
361.0 ± 3.3  
368.9 ± 3.3  
377.9 ± 3.4  
384.4 ± 3.6  
389.0 ± 4.0  
Additive effects  
­20.1 ± 30.8  
­10.6 ± 50.7  
47.7 ± 31.1  
­9.6 ± 29.4  
­9.3 ± 34.0  
53.0 ± 29.8  
­2.4 ± 29.8  
­10.0 ± 58.2  
51.5 ± 30.6  
5.1 ± 32.0  
­10.6 ± 96.3  
50.6 ± 33.5  
12.7 ± 35.6  
­10.8 ± 137.1  
49.6 ± 38.0  
14.9 ± 16.1  
44.5 ± 25.9  
­4.0 ± 16.4  
12.5 ± 15.4  
47.0 ± 17.8†  
­5.6 ± 15.7  
8.8 ± 15.7  
46.7 ± 29.6  
­3.1 ± 16.1  
5.5 ± 16.8  
46.8 ± 48.5  
­0.5 ± 17.6  
1.9 ± 18.7  
46.3 ± 68.8  
1.8 ± 19.9  
Hereford least­square means for mature weight ± standard error of the additive – dominant model. g , g , and gI are the individual additive  
effects of Angus, Nellore, and Salers as deviation from Hereford. hI , hI , and hI are the individual heterosis between Angus – Hereford,  
Nellore – Hereford and Salers – Hereford, respectively. The mean followed by † significantly differs from Hereford (H/H) with P < 0.1.  
Additive and non­additive genetic effects  
The h between N/N and H/H had a significant effect  
at 4.5 yr of age being of 47.0 kg (P < 0.1). Other  
Crossbreeding effects from 4 to 6 yr of age are  
crossbreeding experiments carried out at Texas  
presented in Table 4.  
University reported similar estimates for Brahman –  
H/H crosses 47.9 kg for F1 and 34.9 kg for F2 (Boenig,  
The g of A/A did not have significant differences to  
011) and 32.7 kg for F1 and 42.1 kg for F2 (Key,  
H/H for MW at any age. Morris et al. (1987)  
comparing these breeds reported from 6 to 14 kg of  
superiority from H/H. Melucci et al. (2006) obtained  
The estimate of h between S/S and H/H was non­  
gI of ­43.5 kg when comparing to H/H. The g of N/N  
significant. Theunissen et al. (2013) indicate 39.5 kg of  
heterosis for cow weight at calving for H/H –  
Simmental crosses and, ­27.1 kg for H/H – Charoláis  
crosses. These results differ from the obtained in our  
study, but Continental breeds have a wide range of  
weights, and the authors did not find similar studies  
involving Salers breed.  
had a non­significant effect for MW at any age, which  
is in agreement with the result of Boenig (2011) for  
crosses between H/H and Brahman. Non­significant  
estimates for g of S/S were found for MW in the range  
of ages studied. Theunissen et al. (2013) estimated  
additive effects of cow weight at calving for Afrikander  
and cows of different breed origins. The direct effects  
were 62.6 kg for Brahman (Zebu breed), 10.2 kg and  
The effects of heterosis are known to be higher for  
traits with low heritabilities, such as reproduction  
traits (Arthur et al., 1999). Nevertheless, MW has been  
reported within the groups of traits with medium­high  
heritabilities, ranged from 0.28 in Hereford using  
repeated measure mean at 4 yr of age (Meyer, 1995) to  
80.1 kg for Simmental and Charoláis (Continental  
breeds), and 48.8 for H/H as deviation from  
Afrikander. These results show a significant difference  
in size among Continental breeds.  
The h between A/A and H/H had a non­significant  
.56 for Angus at the same age (Costa et al., 2011).  
effect even though the absolute values were as high as  
those reported by other authors. Cundiff (1970)  
reported 12.5 kg of hI  
Despite the reported heritability values, heterosis  
levels have been important both in this study for N/H  
at 4.5 yr of age and in those found in the literature.  
for adjusted weight by the  
condition score of cows from 6 to 9 yr of age (P <  
.05). Melucci et al. (2006) reported the highest value  
27.6 kg; P < 0.05) for these breeds. Similar estimates  
were found by Morris et al. (1987), reporting between  
3 kg and 26 kg. Arango et al. (2004) indicated that  
the levels of heterosis between A/A and H/H tended to  
decrease with age. At 2 yr of age they found 19 kg of  
heterosis (P < 0.01), at 3 and 4 yr of age the value was  
Regarding the evolution of heterosis with age, we did  
not find differences. Other studies emphasize that the  
heterosis effects on size­related traits are largely  
exhibited by the age of 1 yr and are maintained to  
maturity (Gregory et al., 1992) or that the levels  
decrease with age (Smith et al., 1976; Arango et al.,  
7 kg (P < 0.01) and at 6 yr of age the estimate was 5 kg  
P < 0.05).  
ISSN 1022­1301. Archivos Latinoamericanos de Producción Animal. 2020. 28 (1­2): 25­28  
Guillenea et al.  
The use of an animal model, which considers all  
Regarding the accuracy of the estimates, the additive  
genetic effect presented high standard errors, roughly  
twice the standard error of heterosis estimates. It has  
been reported that the accuracy of the estimate of  
heterosis is always higher than the accuracy of the  
additive genetic effects (approximately 20 % fewer  
animals are required to obtain the same accuracy)  
(Cunningham and Connolly, 1989). The lack of MW  
records of purebred cows and reciprocal crosses may  
have led to high estimates of standard errors. The  
structure of the data may not be enough for the  
accurate estimation of parameters in this  
characteristic. More animals and breed groups would  
likely have obtained greater precision in the estimates,  
and more parameters would have had a significant  
genetic relationships, is a more complete method of  
analysis for crossbreeding data. The relationship  
matrix was not available for the data set used in this  
study. However, since sires were balanced across the  
cow­breed group and at least one was repeated per  
breed per year, in addition to the fact that the cows of  
the general herd (without known genealogy) were  
rotating every year, the use of an animal model is  
expected to provide little additional information.  
The crossbreeding parameters in this study were  
estimated for different ages since the time when the  
cows reach the MW is not clear, and different growth  
patterns have been reported for different breeds. Some  
authors have reported that the cows would reach their  
mature weight at 5 yr of age (Kaps et al., 1999;  
Goldberg and Ravagnolo, 2015), 6 yr of age (Boenig,  
Further analyzes that allow reducing the standard  
error or quantifying the probability of differences in  
the mature weight of the cows at the ages analyzed can  
provide information to consider in the design of  
commercial crossbreeding.  
011), or at more advanced ages: 7 yr of age (Gregory  
et al., 1992; Choy et al., 2002). Contrary, Arango et al.  
2004) analyzing the weight of British, Continental,  
and Zebu cows found that by 4 yr of age, the cows had  
accumulated most of their final weight (98.6 %). Other  
authors have reported that Nellore reaches maturity at  
about 3 to 5 yr of age (Mercadante, 2001; Rosa et al.,  
In this study, we aimed to estimate MW and  
our results, for herds in which crossbred animals are  
kept to maturity either as dams or for marketing,  
improved environmental management may be needed  
to allow the crossbred animals to express its full  
potential. Further investigation of the causes  
underlying the differences between breed groups and  
the application of alternative statistical methods are  
warranted for a more comprehensive assessment of  
crossbreeding parameters for the same trait in  
crossbred animals of breeds from divergent origins.  
The use of crosses increased the MW, especially for the  
F1 animals, with a pattern that increased when the  
racial groups were genetically more diverse. The  
crossbred animals with H/H sire and F1 H/N dam  
showed the highest MW, which supports the previous  
and current evidence that more heterosis is generated  
in the Bos taurus × Bos indicus crosses. According to  
We gratefully acknowledge to Caja Notarial de  
Seguridad Social for the financial support and Facultad  
de Agronomía (Uruguay) for providing the data for  
this study. We greatly appreciate the editor and  
reviewers for the careful work and valuable comments,  
which helps for improving the manuscript. We also  
thank Gabriel Ciappesoni and Jorge Urioste for the  
nice suggestions and discussion during the data  
analyses. The research that generated the results  
presented in this publication received financial  
support from the National Agency of Research and  
Innovation of Uruguay (POS_NAC_2016_1_129902).  
Literature Cited  
Arango, J. A., L. V. Cundiff, and L. D. Van Vleck. 2002.  
Breed comparisons of Angus, Brahman, Hereford,  
Pinzgauer, Sahiwal, and Tarentaise for weight, weight  
adjusted for condition score, height, and body  
condition score. J. Anim. Sci. 80: 3142­3149.  
and Shorthorn breeds for weight, weight adjusted for  
condition score, height, and condition score of cows.  
J. Anim. Sci. 82: 74­84.  
Arthur, P. F., H. Hearnshaw, and P. D. Stephenson.  
999. Direct and maternal additive and heterosis  
Arango, J. A., L. V. Cundiff, and L. D. Van Vleck. 2004.  
Comparisons of Angus, Charolais, Galloway,  
Hereford, Longhorn, Nellore, Piedmontese, Salers,  
effects from crossing Bos indicus and Bos taurus  
cattle: Cow and calf performance in two  
environments. Livest Prod Sci 57: 231­241.  
ISSN 1022­1301. Archivos Latinoamericanos de Producción Animal. 2020. 28 (1­2): 26­28  
Crossbreeding parameters for beef cow  
Boenig, L. 2011. Heterosis and heterosis retention for  
reproductive and maternal traits in Brahman x  
Hereford crossbred cows. Doctoral thesis, Texas A&M  
University, 79 pp.  
Jenkins, T. G. 2009. Interbreed evaluation of beef cattle  
productivity under low and moderate dry matter  
availables. In: Procedings of the 18th Association for  
the Advancement of Animal Breeding and Genetics. p  
Chenoweth, P. J. 1994. Aspects of reproduction in  
female Bos indicus cattle: a review. Aust Vet J 71:  
Junior, C. K., Roso, V. M., Fávaro, P. D., Pereira, G. R.,  
Borges, M. H., Junior, F. A., Barcellos, J. O., and  
Ribeiro, E. L. . 2019. Heritability estimation and  
genetic correlations for mature weight, visual scores,  
and growth traits in Nellore cattle. R. Bras. Zootec.  
Choy, Y., J. Brinks, and R. M. Bourdon. 2002.  
Repeated­measure animal models to estimate genetic  
components of mature weight, hip height, and body  
condition score. J. Anim. Sci. 80: 2071­2077.  
Costa, R., Misztal, I., Elzo, M. A., Bertrand, J. K., Silva,  
L. O. C., and Łukaszewicz, M.. 2011. Estimation of  
genetic parameters for mature weight in Angus cattle.  
J. Anim. Sci. 89: 2680­2686.  
Kaps, M., W. O. Herring, and W. R. Lamberson. 1999.  
Genetic and environmental parameters for mature  
weight in Angus cattle. J. Anim. Sci. 77: 569­574.  
Cundiff, L. V. 1970. Experimental Results on  
Crossbreeding Cattle for Beef Production. J. Anim.  
Sci. 30: 694­705.  
Key, K. L. 2005. Heterosis and heterosis retention for  
reproductive and maternal traits in Brahman ­  
British crossbred cows. Doctoral thesis, Texas A&M  
University, 96 pp.  
Cunningham, E. P., and J. Connolly. 1989. Efficient  
design of crossbreeding experiments. Theor. Appl.  
Genet. 78: 381­386.  
Kinghorn, B. P., and P. E. Vercoe. 1989. The effects of  
using the wrong genetic model to predict the merit  
of crossbred genotypes. Anim. Prod 49: 209­216.  
Di Marco, O. 2006. Rendimiento de res. La industria  
cárnica latinoamericana 26.  
Koch, R. M. 1972. The Role of Maternal Effects in  
Animal Breeding: VI. Maternal Effects in Beef  
Dickerson, G. E. 1969. Experimental approaches to  
utilizing breed resources. In: Animal Breeding  
Abstract. p 191­202.  
Cattle. J. Anim. Sci. 35: 1316­1323.  
Kuehn, L., and R. Thallman. 2016. Across­breed EPD  
tables for the year 2016 adjusted to breed  
differences for birth year of 2014. Procedings of the  
Beef Improvement Federation: 127­154.  
Dickerson, G. E. 1973. Inbreeding and heterosis in  
animals. Proceedings of the Animal Breeding and  
Genetics Symposium in Honour of J.L. Lush.: 54­77.  
Lema, O. M., D. Gimeno, N. J. Dionello, and E. A.  
Navajas. 2011. Pre­weaning performance of  
Hereford, Angus, Salers and Nellore crossbred  
calves: Individual and maternal additive and non­  
additive effects. Livest Prod Sci 142: 288­297.  
Frisch, J. E., and J. E. Vercoe. 1977. Food intake, eating  
rate, weight gains, metabolic rate and efficiency of  
feed utilization in Bos taurus and Bos indicus  
crossbred cattle. Anim. Prod 25: 343­358.  
Gardner, C. O., and S. A. Eberhart. 1966. Analysis and  
Interpretation of the Variety Cross Diallel and  
Related Populations. Biometrics 22: 439­439.  
Melucci, L. M., C. A. Mezzadra, and E. L. Villarreal.  
006. Genetic components for breeding system  
traits in Angus­Hereford crossing. Proceedings of  
the 8th World Congress on Genetics Applied to  
Livestock Production, Belo Horizonte, Minas Gerais,  
Brazil, 13­18 August, 2006: 03­51.  
Gimeno, D., S. Avendaño, and R. Severino. 1995.  
Elección de un diseño óptimo de cruzamientos en un  
experimento con cuatro razas bovinas. In: Memorias  
de la XIV Reunión Latinoamericana de Producción  
Animal. p 914–918.  
Mercadante, M. E. 2001. Análise de um experimento de  
seleção para crescimento em bovinos Nelore:  
Goldberg, V., and O. Ravagnolo. 2015. Description of  
the growth curve for Angus pasture­fed cows under  
extensive systems1. J. Anim. Sci. 93: 4285­4290.  
respostas direta no peso ao sobreano  
correlacionadas no tamanho e reprodução das  
matrizes. Tesis Doctoral, Universidade de São Paulo.  
Gregory, K. E., and L. V. Cundiff. 1980. Crossbreeding  
in Beef Cattle: Evaluation of Systems1. J. Anim. Sci.  
Meyer, K. 1995. Estimates of genetic parameters for  
mature weight of Australian beef cows and its  
relationship to early growth and skeletal measures.  
Livest Prod Sci 44: 125­137.  
51: 1224­1242.  
Gregory, K. E., L. V. Cundiff, and R. M. Koch. 1992.  
Breed effects and heterosis in advanced generations  
of composite populations on actual weight, adjusted  
weight, hip height, and condition score of beef cows.  
J. Anim. Sci. 70: 1742­1754.  
Morris, C. A., R. L. Baker, A. H. Carter, and J. C.  
Hunter. 1987. Reciprocal crossbreeding of angus  
and hereford cattle 3. cow weight, reproduction,  
ISSN 1022­1301. Archivos Latinoamericanos de Producción Animal. 2020. 28 (1­2): 27­28  
Guillenea et al.  
maternal performance, and lifetime production.  
New Zealand Journal of Agricultural Research 30:  
Sobek, Z., A. Nienartowicz­Zdrojewska, J. Różańska­  
Zawieja, and I. Siatkowski. 2015. The evaluation of  
gestation length range for different breeds of Polish  
dairy cattle. Biometrical Letters 52: 37­45.  
Pravia, M. I., O. Ravagnolo, J. I. Urioste, and D. J.  
Garrick. 2014. Identification of breeding objectives  
using a bioeconomic model for a beef cattle  
production system in Uruguay. Livest Prod Sci 160:  
R Development Core Team. 2016. A language and  
environment for statistical computing.  
Sölkner, J., and W. Fucks. 1994. ODCE Optimun Design  
Design of Crossbreeding Experiments. Version 1.O.  
Sölkner, J., and J. W. James. 1990a. Optimum design of  
crossbreeding experiments: I. A basic sequential  
procedure. J. Anim. Breed. Genet 107: 61­67.  
Sölkner, J., and J. W. James. 1990b. Optimum design of  
crossbreeding experiments: II. Optimum relationship  
structures of animals within and between genetic  
groups. J. Anim. Breed. Genet 107: 411­420.  
Foundation for Statistical Computing, Vienna,  
Austria. ISBN 3­900051­07­0.  
Riley, D., and J. Crockett. 2006. Heterosis retention  
and the dominance model in Florida beef research.  
Florida Cattleman 70: 42­46.  
Theunissen, A., M. M. Scholtz, and F. W. C. Neser.  
013. Crossbreeding to increase beef production:  
additive and non­additive effects on weight traits. S.  
Afr. J. Anim. Sci 43: 143­152.  
Rosa, A. D. N., R. B. Lôbo, H. N. D. Oliveira, L. A. F.  
Bezerra, and A. D. L. Reyes Borjas. 2001. Peso adulto  
de matrizes em rebanhos de seleção da raça Nelore no  
Brasil. Revista Brasileira de Zootecnia 30: 1027­1036.  
Wiltbank, J. N., C. W. Kasson, and J. E. Ingalls. 1969.  
Puberty in crossbred and straightbred beef heifers on  
two levels of feed. J. Anim. Sci. 29: 602­605.  
SAS Institute. 2014. 13. 2 User´s Guide: High­  
Performance Procedures, Cary, NC.  
Wolf, J., O. Distl, J. Hyánek, T. Grosshans, and G.  
Seeland. 1995. Crossbreeding in farm animals. V.  
Analysis of crossbreeding plans with secondary  
crossbred generations. J. Anim. Breed. Genet 112:  
Smith, G. M., H. A. Fitzhugh, L. V. Cundiff, T. C.  
Cartwright, and K. E. Gregory. 1976. Heterosis for  
maturing patterns in Hereford, Angus and Shorthorn  
cattle. J. Anim. Sci. 43: 380­388.  
ISSN 1022­1301. Archivos Latinoamericanos de Producción Animal. 2020. 28 (1­2): 28­28