para$el$agro$

ASSOCIATION BETWEEN REPRODUCTIVE TRAITS AND MILK

YIELD, IN CROSSBRED HEIFERS

ASOCIACIÓN ENTRE CARACTERÍSTICAS REPRODUCTIVAS Y

PRODUCCIÓN DE LECHE, EN NOVILLAS MESTIZAS

Rafael Maria Román Bravo

, José Atilio Aranguren Méndez

, Rogelio Garcidueñas Piña

, Benjamín Gómez Ramos

Alfredo Nicanor García Gavidia

, Edelmira Criseida Márquez Carrera, Soto Belloso, Eleazar

Facultad de Ciencias Veterinarias, Universidad del Zulia, Maracaibo, Venezuela

Tecnologico Nacional de México/IT del Valle de Morelia, Morelia – Michoacán, México

Email: rafael.roman@fcv.luz.edu.ve

Información

del artículo

Tipo de artículo:

Artículo original

Recibido:

12/03/2023

Aceptado:

13/07/2023

Licencia:

CC BY-NC-SA 4.0

Revista

ESPAMCIENCIA

14(2):63-70

DOI:

https://doi.org/10.512

60/revista_espamcien

cia.v14i2.358

Abstract

This research was carried out in heifers from a crisscrossing program between the red Holstein and Brahman

breeds, in Venezuela. (Co)variance components were estimated and the relationship between milk yield (P305),

age (AGE1) and fertility at first service (FERT1) in virgin heifers was investigated. The analysis methodology

was restricted maximum likelihood (REML), under a multivariate animal model. The mixed model included:

the random effect of the animal; and the fixed effects of racial group, year-seasons of the event occurrence; for

P305, the duration of lactation was also included as a covariate in quadratic form. Heritabilities were: 0.498 ±

0.022; 0.436 ± 0.020 and 0.024 ± 0.008, for P305, AGE1 and FERT1 respectively. Genetic correlations

between: P305-AGE1; P305-FERT1 and AGE1-FERT1, were: -0.281±0.040; 0.238 ± 0.117 and -0.076 ±

0.123, these results suggest a favorable genetic association between P305-AGE1, as well as, between P305-

FERT1. Weighted regression analysis of the averages breeding values on years, showed an improvement in

P305 of 10.85 kg/year and a reduction of -0.1150 month/year in AGE1 representing 0.60 and 0.41% of the

population means for those traits (p < 0.01); the annual genetic change for FERT1 was not significant (p >

0.05). The phenotypic changes in P305 and AGE1 were not significant (p > 0.05). It is concluded that there is

no genetic antagonism between P305 and FERT1 and that it is possible to reduce AGE1 for decreasing the

costs for generating replacement heifers, improving the rate of genetic change in P305.

Keywords: Milk yield, fertility, age first service, crossbred cattle.

Resumen

Esta investigación se realizó con novillas de un programa de cruzamiento entre las razas Holstein y Brahman

roja, en Venezuela. Se estimaron los componentes de (co)varianza y se investigó la relación entre producción

de leche (P305), edad (AGE1) y fertilidad al primer servicio (FERT1) en novillas vírgenes. La metodología de

análisis fue bajo máxima verosimilitud restringida (REML), bajo un modelo animal multivariado. El modelo

lineal incluía: el efecto aleatorio del animal; y los efectos fijos de grupo racial, años-temporadas de ocurrencia

del evento; para P305, la duración de la lactancia también se incluyó como covariable en forma cuadrática. Las

heredabilidades fueron: 0,498 ± 0,022; 0,436 ± 0,020 y 0,024 ± 0,008, para P305, AGE1 y FERT1

respectivamente. Las correlaciones genéticas entre: P305-AGE1; P305-FERT1 y AGE1-FERT1, fueron: -

0,281±0,040; 0,238 ± 0,117 y -0,076 ± 0,123, estos resultados sugieren una asociación genética favorable entre

P305-AGE1, así como entre P305-FERT1. El análisis de regresión ponderada de los valores genéticos

promedios en años, mostró una mejora en P305 de 10.85 kg/año y una reducción de -0.1150 mes/año en AGE1

lo cual representa 0.60 y 0.41% de las medias poblacionales para esos rasgos (p < 0.05), el cambio genético

anual para FERT1 no fue significativo (p > 0,05). Los cambios fenotípicos en P305 y AGE1 no fueron

significativos (p > 0,01). Se concluye que no existe antagonismo genético entre P305 y FERT1 y que es posible

reducir AGE1 para disminuir los costos de generación de vaquillas, mejorando la tasa de cambio genético en

P305.

Palabras clave: Producción de leche, fertilidad, edad primer servicio, ganado mestizo.

Román et al. (2023) Vol. 14 Nº 2. pp: 63-70. ISSN:1390-8103! ! !

INTRODUCTION

Reproduction is a complex multifactorial process, with

high phenotypic variance, very little genetic variation,

hence is very sensitive to environmental effects. The

parturition of the cow triggers lactation, which is the

fundamental source of income in dairy farms, therefore,

breeder's objective must be focused on having efficient

reproduction in the herd, this includes an early pregnancy

with the least number of services. The cost of raising

replacement heifers in farms are between 15 to 20% of the

total costs in dairy farms (Fodor et al., 2020). In an

experiment with dairy tropical creole heifers managing the

feeding system has proved to improve age and weight at

puberty and at the same time follicular dynamics

(Severino et al., 2017). An age at first service of 17.7

months with a range between 14.5 and 23.8 months was

reported for heifers (Muller et al., 2017). In Israel it has

been suggested that the optimal age at first calving in

Holstein is between 22 and 24 months (Weller et al.,

2022a).

Heritability for milk yield at 305 days of lactation have

been reported in the range of 0.15 to 0.50 in different

populations of dairy cattle (Shalaby, 2005; Montaldo et

al., 2010; Ayalew et al., 2017; Weller et al., 2022b). For

reproductive traits, on the contrary, the additive variance

is very low: a heritability of 0.012 has been reported for

the 56-day non-return rate (Sun et al., 2009); for age at

first service, values close to 0.06 have been published

(Ayalew et al., 2017; Weller et al., 2022b), although a

value of 0.128 was previously published (Abe et al.,

2009). Heritability for first service fertility for this

crossbred cattle and age group was estimated to be 0.03 by

using a Gaussian model and 0.07 with a threshold model

(Román et al., 2010). Despite having low heritability,

reproductive and health traits are being incorporated into

selection programs in some countries, it is interesting to

see the evolution of selection indices in dairy cattle in

Israel from 1985 to the present (Weller et al., 2022a;

2022b).

In dual purpose cattle, evidence of deterioration

reproductive behavior was found when classifying the

cows, according to the level of milk yield and recording

first service fertility, a significant reduction was observed,

with the increase of production level (Soto & Perea, 2014).

Previously, studying a cause-effect relationship between

production level and fertility, evaluating the effects of

milk yield at 120 days in milk, on the number of services

per conception, it was found that more services were

required in Holstein, Jersey and Guernsey cows per each

100 kg of extra milk (Olds et al., 1979).

Negative estimates of the genetic correlation between age

at first artificial insemination and milk yield, and therefore

favorable, were reported previously (Abe et al., 2009;

Weller et al., 2022b). However, some studies suggest a

genetic antagonism between milk yield and reproductive

performance (Shalaby, 2005; Windig et al., 2006; Getahun

and Beneberu, 2023). Reports of the genetic correlations

between age at first service and age at first calving are high

and positive (Abe et al., 2009; Brzáková et al., 2019;

Weller et al., 2022b; Getahun and Beneberu, 2023).

At the experimental level, working with the Jersey breed

a genetic change of 39.90 Kg/year was reported by Román

et al. (1999), representing 1.48% of the population

average, which is very close to the maximum possible for

this trait as it was suggested previously (Rendel and

Robertson, 1950). Despite the change achieved on milk

yield, correlated favorable changes, for age at first heat

and duration of the period from parturition to first service

were observed, only the change in the number of services

per conception showed a positive slope, but very close to

zero (Román et al., 1999).

The foregoing has motivated this research with the aims

of estimating the genetic, environmental and phenotypic

correlations between: milk production at 305 days; age

and fertility at first service, as well as, the changes with

time, in virgin heifers from a crisscrossing program in a

tropical environment.

MATERIAL AND METHODS

This research was carried out using the databases of the

“Mompox” farm which belongs to the Santa Ana farming,

during the period 1989-2016. The records are kept

electronically with the GanSoft© Software. The

production unit is located in a sub-humid zone in the Lake

Maracaibo basin, at 14 m.a.s.l, at 9º 30' 3'' north latitude

and 72º 20' 39'' west longitude, further details may be

found on previous work (Soto and Perea, 2014).

The animals belong to a population generated by a

crisscrossing program between the Holstein and the red

Brahman breeds mainly, as a strategy to combine the

production levels of Bos taurus, with the adaptation ability

of Bos indicus to hard weather conditions, hence, in this

work, the records from the two resulting subpopulations at

the time of the stabilization of the crossbreeding program

were used.

For this research, only records belonging to first calving

animals were used, corresponding to the variables: Milk

yield at 305 days (P305), which was estimated by monthly

supervision, but truncated to a maximum duration of 305

days; Age at first service (AGE1) by artificial

insemination (AI), which was estimated as the difference

between the date of the first recorded service and the date

of birth expressed in months; finally, the fertility at first

service (FERT1) was evaluated as a binomial variable

coded as 0, for empty and 1 for pregnant, these last two

variables were taken when the heifers were virgin. In

Román et al. (2023) Vol. 14 Nº 2. pp: 63-70. ISSN:1390-8103! ! !

addition, lactation length (LL) and age at parturition

(AGE1P) were reordered.

The model included the fixed effects of: predominant

breed group in the animals with two levels, 2/3 Holstein

1/3 Brahman and 2/3 Brahman 1/3 Holstein; The months

of the year were grouped into four seasons, based on the

rainfall records of the farm, namely: season 1, months of

December and January; season 2, months of February

March; season 3, months of April, May and June; season

4, months of July, August, September and October.

Contemporary groups were structured concatenating years

and seasons, groups with less than four observations in

each contemporary group were excluded.

The multivariate model can be written as:

!"#$%&'%(

If we partition the data vector as

)

+++!

)

++!

)

; and

conveniently the incidence matrices of the fixed effects as:

)

+++#

)

+++#

)

; and the random effects matrix as:

)

+++&

)

+++&

)

and if 1, 2 and 3, represent the

variables: P305; AGE1 and FERT1, the statistical model

can be expressed as the following linear combination

(Searle, 1971; Mrode 2005):

3 3

3 #

3 3 #

3 3

3 &

3 3 &

(

The sub-matrices for the allocation of the fixed effects for

P305 had the effect of racial group with two levels as

above, the effect of year-calving season with 93 levels, and

the LL as a quadratic covariate; the sub-matrices for the

allocation of the fixed effects for AGE1 and FERT1, in

addition to racial group, included 111 and 105 levels of

year-seasons of the event occurrence, respectively.

For the estimation of the (co)variance components, the

Wombat software was used (Meyer, 2022), selecting in the

estimation strategy the option - -pxai, a hybrid algorithm

consisting of a few initial expectation-maximization

iterations followed by iterations with the REML average

information algorithm. First, univariate analyzes were

performed, estimating the additive variance of each trait

from

, where

is the phenotypic variance of

the random variable, estimated from the data of table 1 for

each trait; the environmental component was estimated by

difference. Subsequently, the univariate estimators were

used as initial values and the genetic covariance

components were estimated by using the

equation:

70:+

=>;0:?

7>;0:?

, analogously, the initial

values for the environmental covariance components were

estimated. The convergence criterion established a priori

was 1x10

-6

For the description of the dispersion of the random effects

of the model, it was considered that

is the additive

genetic variance covariance matrix among the three traits

submitted to the multivariate analysis, and likewise,

the variance covariance matrix for the residuals; A is the

matrix of the additive genetic relationships between the

animals, I is an identity matrix and C represents the

Kronecker product, the variances of the random effects are

defined by the expression:

DE<

(

CJ K

K B

The required data files were built with the Statistical

Analysis System (SAS, 2016), including one with the

information of the three traits for the multivariate analysis.

RESULTS AND DISCUSION

The Pedigree file contained the identification of 17,544

animals with their ancestors, of the total number of

animals 2024 belonged to the base population. There were

680 descendants of related animals, with an average

inbreeding coefficient of 0.2976%, the average inbreeding

among inbreed animals was 7.68%.

Table 1 shows the averages and standard deviations for:

P305, FERT1 and AGE1, used as dependent variables in

these analyses, in addition, LL and AGE1P are included.

The P305 of these animals is below the previously

published value of 3504.02 for Holstein cattle in Ethiopia

(Ayalew et al., 2017) and very inferior to publications for

the same breed in Israel, Egypt and Japan (Shalaby, 2005;

Abe et al., 2009; Weller et al., 2022a) and in the Mexican

republic with an average of 11499 kg milk equivalent adult

(Montaldo et al., 2010). In tropical conditions it is difficult

to reach these figures because the diet is mainly based on

forage, with supplementation only at the milking time

according to the level of production.

The age at which these heifers begin their reproductive life

is considerably high, which is affecting the costs of raising

replacements, compromising genetic progress and

contributing to shortening the length of the lifespan of the

animals. The AGE1 average is much higher than those

published for Holsteins (Muller et al., 2017; Brzáková et

al., 2019; Weller et al., 2022a; Abe et al., 2009); This

variable responds satisfactorily to feeding programs as has

been published in creole dairy cattle in Mexico (Severino

et al., 2917). Despite the delay in reaching the weight

required to be included in the reproductive program, the

FERT1 in these heifers of 62% is higher than that

published in Israel for Holsteins and slightly lower than

the 69% reported for Holstein heifers in Japan (Abe et al.,

2009).

Román et al. (2023) Vol. 14 Nº 2. pp: 63-70. ISSN:1390-8103! ! !

The dairy aptitude of this crossbreed cattle is reflected in

LL with an average of 295.09 ± 23.54 days, which

indicates that the dairy temperament of the Bos taurus

breed has been inherited in this population, despite the

introduction of Brahman genes, which is a breed that tends

to have shorter lactations.

Table 1. Means ± standard deviations for productive and

reproductive traits of heifers from a crisscrossing

program between the Holstein and Red Brahman

breeds

Heritability estimates

The additive and environmental genetic (co)variance

components are shown together with their standard errors,

for each trait on tables 2 and 3. The corresponding

proportions and their standard errors are in tables 4 and 5.

In the case of P305, the ratio of the additive variance to the

total was 50%. Our estimator is identical to the reported

for Holstein heifers in Israel (Weller et al., 2022b), higher

than the reported for the same type of animals in Egypt,

Mexico and Ethiopia (Shalaby, 2005; Montaldo et al.,

2010; Ayalew et al., 2017).

Table 2. Estimates of the additive (co)variance

components ± standard errors, for a population

resulting from a crisscrossing program between

the Holstein and Red Brahman breeds!

Component

Genetic

Estimates

S.E

#$%&'()

81614.9000

4716.6900

%&'(,#-./

)

00000

-168.8640

24.8790

#$%&'(,1.23/)

5.0695

2.4707

#$#-./)

4.4157

0.2485

#$#-./,1.23/)

0.0119

0.0191

#$1.23/)

0.0056

0.0018

#$4)

=!additive!genetic!variance!for!the!i

!trait,!

4,5

)

=!additive!

genetic!covariance!between!the!i

!and!j

!traits.!

Table 3. Estimates of the environmental (co)variance

components ± standard errors for a population

resulting from a crisscrossing program between the

Holstein and Red Brahman breeds

Component

Environmetal

Estimates

S.E

.$%&'()

82332.4000

3006.5200

%&'(,#-./

)

00000

-3.3747

16.8075

.$%&'(,1.23/)

-7.0711

2.2601

.$#-./)

5.7043

0.1640

.$#-./,1.23/)

0.0618

0.0164

.$1.23/)

0.2222

0.0029

.$4)

=!environmental!variance!for!the!i

!trait,!

4,5

)

environmental!covariance!between!the!i

!and!j

!traits.!

Table 4. Estimates of the heritabilities and genetic

correlations ± standard errors, for a population

resulting from a crisscrossing program between

the Holstein and Red Brahman breeds

Estimates

Genetic

Proportion

S.E

0$%&'()

0.498

0.022

#$%&'(,#-./)

-0.281

0.040

#$%&'(,,1.23/)

0.238

0.117

$#-./)

0.436

0.020

#$#-./,1.23/)

-0.076

0.123

$1.23/)

0.024

0.008

0$4)

=!heritability!for!the!i

!trait,!

4,5

)

=!additive!genetic!covariance!

between!the!i

!and!j

!traits.!

Table 5. Estimates of the proportion of the environmental

(co)variance components ± standard errors and

environmental correlation for a population

resulting from a crisscrossing program between the

Holstein and Red Brahman breeds&

Estimates

Environmental

Proportion

S.E

0$%&'()

0.502

0.022

;$%&'(,#-./)

-0.005

0.025

;$%&'(,,1.23/)

-0.052

0.017

$#-./)

0.564

0.020

;$#-./,1.23/)

0.055

0.015

$1.23/)

0.976

0.008

0$4)

= proportion of the environmental variance from the phenotypic

variance for the ith trait,

;$4,5)

environmental correlation between

between the i

and j

traits.

In the particular case of this population, having a high

genetic variability was to be expected, if we consider that

these animals came from crossing individuals of very

extreme genotypes: firstly, the best red Holstein bulls

available on the market of semen, with positive and high

values in their expected progeny difference for milk yield;

secondly, bulls of the Brahman breed, paying attention

mainly to have high weaning weights, which are raised in

the Brahman breeding center of the “Mompox” farm.

The phenotypic values, in their simplest form, are due to

the action of the genotype and the environment (Falconer

and MacKay, 2001), we must say that the management and

feeding conditions, as well as, medical-veterinary

assistance and control of production records in this herd,

is well above average for the herds in this sub-tropical

zone; observation which is made, because it could be

thought that, since we are dealing within a region in a

tropical weather usually with poor management practices,

the tendency would be for the heritabilities to be low.

These results are encouraging and hence: livestock

associations, development corporations and entities that

promote livestock development in the country, should see

this population as the strategy to follow in national

programs, to consolidate the livestock that sustains more

than 75% of the national supply of milk and around 45%

of that corresponding to meat. From here the initiatives for

the formation of local tropical dairy breeds could come

Traits

Statistics

Mean

S.D

P305

10599

1798.65

534.90

AGE1

14300

27.14

3.67

FERT1

15427

0.62

0.49

10599

295.09

23.54

AGE1P

10599

37.30

4.44

Román et al. (2023) Vol. 14 Nº 2. pp: 63-70. ISSN:1390-8103! ! !

out; an issue that has not been achieved because simple

decisions have not been taken yet, such as reducing the

number of breeds to the minimum required.

For AGE1 the heritability estimator indicates that 44% of

the variability in this trait has an additive genetic

component and therefore a substantial amount of response

is to be expected if this trait is included in a selection index

to obtain replacements for the farm. Our estimator is much

higher than those previously reported (Brzáková et al.,

2019; Weller et al., 2022b). This must be looked at

carefully, if the aim is to improve the productivity of this

livestock, since heifers are beginning their reproductive

life, when they should already be lactating. This is

compromising: productivity of the farm, duration of the

productive life of cows and genetic progress. This trait is

a developmental characteristic and is simply a function of

growth rate, combined with the onset of puberty, an event

that must have occurred well in advance. It can be inferred

that with rational feeding plans AGE1 can be reduced and

weight increased at puberty (Severino et al., 2017). We

could expect that if we manage to make heifer’s grow, as

close as, possible to the "optimum", they will have an

efficiently developed hypothalamus-pituitary-gonad-

uterus axis so that the feedback mechanisms that

determine cyclicity and pregnancy occur.

The heritability for FERT1 was 0.024; ratifying the almost

null genetic variability in this trait; only slightly lower than

the previously reported estimated for this population

(Román et al., 2010), however, in this case, the estimate is

more precise due to the increased number of observations.

Previously comparing linear and threshold models a very

low heritability for fertility was found, observing slightly

higher estimates with the latter model; however, the

correlations between the predictions by both

methodologies were greater than 0.99 (Weller & Ron,

1982).

There are sufficient reasons to suppose that the additive

variance for reproductive traits to be low, since these

characters have been subject to natural selection for

hundreds of years, because they are determinants in the

survival of the species, and as a result, any gene that affect

the proportional contribution of offspring of a particular

genotype to the next generation is dropped from the

population (Falconer & MacKay, 2001). It is striking that

in some countries reproductive and other health

characteristics have been added into the selection plans;

knowing that the magnitude of the response to selection

must be negligible given the high homozygosity. In the

case of tropical herds, like this one, the crisscrossing

program has as objectives: to give stability to the

production system generating within it the replacement

females; to maintain maximum expression of heterosis

and to combine the desirable characteristics of two

populations. Genetic improvement must be supported both

in the crossing program and in the selection plans. Over

the years, efforts have been made in both directions in this

population.

The fact that some animals stand out from their

contemporaries, achieving acceptable levels of production

in P305, and that they also get pregnant at the right time,

with fewer services, makes us suppose that they have a

genetic system in harmony with the environment in which

that unfold; therefore, they should be left for replacement.

This justifies the inclusion of reproductive and health

characters as have done in Israel (Weller et al., 2022b).

Another way of interpreting this is the search for an

indirect response to selection for "adaptation" to the

environment and its inclusion in the improvement plans

would be justified; Otherwise, it would be worthless to

include them, even in an index, because it compromises

the annual progress in the traits of interest (Rendel &

Robertson, 1950), mainly affecting the selection

differential for the economic traits.

Genetic, environmental and phenotypic correlations

The genetic correlation between P305 and AGE1 was -

0.28, rather low but favorable and much higher than that

found for Holstein cattle in Israel and Japan respectively

(Abe et al., 2009; Weller et al., 2022b). It has been

reported that younger heifers at the time of pregnancy had

a higher probability of calving, produced more milk, and

less likely to be culled from the herd (Fodor et al., 2010).

On the other hand, a moderate antagonism of 0.29 has

been reported in crossbred cattle in Ethiopia (Getahun &

Beneberu, 2023) and a very extreme one for Holstein

cattle in Egypt of 0.93 (Shalaby, 2005).

The genetic correlation between P305 and FERT1 was

0.24 with a high standard error. A negative estimate of

0.34 was found for Holstein heifers in Israel (Weller et al.,

2022b); a positive but small value was found for heifers of

the same breed in Japan (Abe et al., 2009). For practical

purposes our positive estimate is also favorable, since it

suggests an increase in fertility at first service with the

increase in milk yield.

In a review on the causes of poor fertility in high

producing dairy cows, authors make a clear distinction

between virgin heifers with a high non-return rate

compared to adult cows, and they also point out the

importance of non-genetic factors as responsible for the

possible antagonism between reproduction and milk

production, exposed by some authors (Walsh et al., 2011).

On the other hand, it has been pointed out that many of the

studies in this field are observational and consequently

there is the possibility of entanglement from the statistical

point of view among the multiple factors affecting the

complex reproductive process and lactation (Bello et al.,

2012). The genetic correlation between AGE1 and FERT1

was -0.08 and due to the magnitude of the standard error,

it is essentially zero, therefore, it indicates that there are

Román et al. (2023) Vol. 14 Nº 2. pp: 63-70. ISSN:1390-8103! ! !

very few genes that act on both physiological processes at

the same time.

The environmental correlations were close to zero in

agreement with previous work (Abe et al., 2009; Weller et

al., 2022b). The phenotypic correlations were very low,

the highest being -0,13 between P305 and AGE1, the

others were almost zero.

AGE1 is strongly correlated with AGE1P phenotypically

and genetically, the latter being close to 0.99 (Abe et al.,

2009; Brzáková, et al., 2019; Weller et al., 2022b;

Getahun and Beneberu, 2023). This implies that they are

the same trait genetically. That was the reason for

incorporating AGE1 in the multivariate analysis instead of

AGE1P, if we intend to reduce the costs of generating

replacements on the farm and to shorten the interval

between generations in order to accelerate annual genetic

progress. With these data, in a bivariate analysis including

AGE1 and AGE1P, the phenotypic correlation was 0.876

± 0.002 and the genetic one 0.991 ± 0.003, confirming the

above rationale, futhermore, AGE1 is less subject to bias

(Weller et al., 2022b).

Genetic and phenotypic changes

Table 6 shows the estimates of the weighted regression

coefficients of the average genetic values for each trait on

years, along with their standard errors. The weighting

factor was the reciprocal of the standard errors of each

year. With all the selective effort put into milk yield, the

regression slope for P305 may represent the direct genetic

response for P305, with a significant increase of 10.85

kg/year (p<0.01), which represents 0.60% of the

population average, being therefore much lower than that

achieved in Jersey (Román et al., 1999) or the maximum

possible previously reported (Rendel and Robertson,

1950). This implies that the selection program in this

population must be reviewed, especially at the level of

females and in Brahman bulls, seeking in this case to

identify within this breed the best genotypes with dairy

aptitude, since at the Holstein breed level they are using

the best bulls available. In table 7 are the phenotypic

changes for milk P305 was only 5.89 kg/year, but not

significantly different from zero (p > 0.05). This implies,

the possibility of improving management and feeding

conditions with the purpose that animals adequately

express their genotypic value and hence minimizing

erroneously selecting the wrong females.

Table 6. Annual genetic change for P305 and correlated

response for AGE1 and FERT1, in a population resulting

from a crisscrossing program between the Holstein and

Red Brahman breeds

Genetic trends

Trait

#>?;@A

S.E

“t”

P305

10.8475

1.4189

7.41

AGE1

-0.1150

0.0097

-11.84

FERT1

0.0008

0.0004

2.02

#>?;@A

= Annual genetic change, S.E= standard error, t=computed value

for “t”

Table 7. Annual phenotypic changes P305, AGE1 and

FERT1, in a population resulting from a

crisscrossing program between the Holstein and

Red Brahman breeds

Phenotypic trends

Trait

%>?;@A

“t”

P305

5.8882

6.3657

!!0.88

EDAD1

0.0014

0.0016

!!0.83

FERT1

-0.0104

0.0022

-4.71

#>?;@A

= Annual phenotypic change, S.E= standard error, t=computed

value for “t”

The regression slopes for AGE1 and FERT1 in table 6

would represent correlated genetic changes; In the first

case, the regression coefficient is negative, therefore

favorable, suggesting a downward trend in the age at

which heifers reach puberty and are ready to enter to the

reproductive program, the percentage magnitude of the

change is, however, very low, representing only 0.41% of

the population average (p < 0.001). The regression

coefficient for the genetic change in FERT1 was not

significant; this seems to be associated with the low

additive variability estimated in this population for this

trait (p > 0.05), however, the sign of the regression

coefficient is favorable.

The phenotypic changes for P305 and AGE1 in table 7

were not significant (p > 0.05), the change for FERT1

represents 1.69% of the population mean (p < 0.01), this

change is unfavorable, however, since it contributes to the

reduction of the probability of pregnancy of the heifers,

this implies reviewing the management factors, since

favorable changes can be canceled by changes in

management.

From the review by (Walsh et al., 2011), we conclude that

in countries like the USA and the Netherlands, the genetic

improvement of dairy cattle has been very accelerated;

consequently, it is relatively easy to obtain replacement

heifers of high or similar genotypic value and therefore it

is justified to include in the selection programs health and

reproductive traits. Furthermore, according to these

authors, the trend should be that by the year 2050 it will

be possible to maintain the same supply in milk

production, reducing the number of cows. It is to be

expected, therefore, an increase in the metabolic

expenditure of these animals, which must be compensated

through management and at the same time improve

reproductive programs in order to avoid reproductive

failures.

In the Venezuelan case, in recent years, a series of events

have discouraged the dual-purpose cattle production

sector, with a notable reduction in the operations of the

dairy industry, this is reflected by the drop in the supply of

fluid or powdered milk on the shelves in the market. Many

dual-purpose livestock producers have dedicated their

activity to cheese production, some of them have shifted

their activities to buffalo farming.

Román et al. (2023) Vol. 14 Nº 2. pp: 63-70. ISSN:1390-8103! ! !

CONCLUSIONS

Milk yield in this population has a high heritability, with

possibilities of improvement through the selection

procedures in males and females.

The heritability for age at first service suggests enough

variability to reduce the average age at first breeding

through a selection process. This should improve the

development of females in the growth phase prior to

weaning and around puberty, in order to reduce the age for

starting reproduction.

The age at first service could be considered in the

elaboration of selection indices for the incorporation of

replacement heifers in this population, due to a possible

favorable correlated response due to adaptation to the

tropical environment.

The heritability for fertility at first service is very close to

zero and consequently little genetic change is to be

expected if this trait is included in selection plans.

There is no evidence of genetic antagonism between milk

yield and reproductive behavior measured by the

association of this variable with age and fertility at first

service.

The crisscrossing program, like the one developed in this

population should be taken as a model for the

consolidation of dual-purpose livestock, a fundamental

axis for supplying the national demand and the country’s

milk and meat.

ACKNOWLEDGEMENTS

The authors wish to express their gratitude to the owners

of the Santa Ana livestock farming, for allowing us to use

the databases of the ”Mompox” farm.

CONFLICTS OF INTEREST

None of the authors of this paper had a personal or

financial interest whit organizations that could

inappropriately influence or bias the results of this

research.

LITERATURE CITED

Abe, H., Masuda, Y., and Suzuki, M. 2009. Relationships

between reproductive traits of heifers and cows and

yield traits for Holsteins in Japan. J. Dairy Sci.

92:4055-4062.

Ayalew,W., Aliy, M, and Negussie, E. 2017. Estimation

of genetic parameters of the productive and

reproductive traits in Ethiopian Holstein using multi-

trait models. Asian-Australas. J. Anim. Sci. 30:1550-

1556.

Bello, N. M., Stevenson, J. S., and Tempelman R. J. 2012.

Invited review: Milk production and reproductive

performance: Modern interdisciplinary insights into an

eduring axiom. J. Dairy Sci. 95: 5461-5475.

Brzáková, M., Zavadilová, L., Přibyl, J., Pešek, P., Kašná,

E, Kranjčevičová, A. 2019. Estimation of genetic

parameters for female fertility traits in the Czech

Holstein population. Czech Journal of Animal

Science.64:199–206.

https://doi.org/10.17221/51/2018-CJAS

Falconer, D. S. y MacKay T.F. 2001. Introduccion a la

Genetica Cuantitativa. Editorial Acribia, S. A.

Zaragoza, España. 469 p.

Fodor I, Lang Z, and Ózsvári, L. 2020. Relación de la

reproducción de vaquillas lecheras con supervivencia

al primer parto, producción de leche y riesgo de

sacrificio en la primera lactancia. Asia-Australas. J.

Anim. ciencia 33:1360–1368.

Getahun K and Beneberu N. 2023. Genetic and phenotypic

relationship between fertility and lactation traits in

crossbred dairy cows in Ethiopia. Livestock Research

for Rural Development. Volume 35, Article #13.

Retrieved February 26, 2023, from

http://www.lrrd.org/lrrd35/2/3513bnib.html

Meyer, K. 2022. WOMBAT: A programs for mixed

Model Analyses by Restricted Maximum Likelihood.

USER NOTES. Animal Genetics and Breeding Unit,

University of New England Armidale, NSW 2351,

Australia. 136 p.

Montaldo, H. H., Castillo-Juarez, H., Valencia-Posadas,

M., Cienfuegos- Rivas, E.G., and Ruiz-Lopez, F.G.

2010. Genetic and environmental parameters for milk

production, udder health, and fertility traits in Mexican

Holstein cows. J. Dairy Sci. 93:2168-2175.

Mrode, R. A. 2005. Linear Models for the Prediction of

Animal Breeding Values. Second Edition. CABI

Publishing. 344 p.

Muller, C. J., Cloete, S.W., Botha, J.A. 2017. Fertility in

dairy cows and ways to improve it. South African

Journal of Animal Science. 48:858-868.

Olds, D., Cooper, T., and Thrift, F.A. 1979. Relationships

between Milk Yield and Fertility in Dairy Cattle. J.

Dairy Sci. 62:1140-1144.

Rendel, J. M., and A. Robertson. 1950. Estimation of

genetic gain in milk yield by selection in a closed herd

of dairy cattle. J. Genet. 50:1–8.

Román et al. (2023) Vol. 14 Nº 2. pp: 63-70. ISSN:1390-8103! ! !

Román, R. M., Wilcox, C. J., and Littell, R. C. 1999.

Genetic Trends for Milk Yield of Jerseys and

Correlated Changes in Productive and Reproductive

Performance. J. Dairy Sci. 82:196-204.

Román, R. M., Aranguren, J. A., Villasmil, Y., Yánez, L.

F, y Soto Belloso, E. 2010. Análisis de la fertilidad al

primer servicio en novillas de doble propósito bajo un

modelo Animal. Revista Científica, FCV-LUZ.

XX:383 – 389.

SAS Institute Inc. 2016. Base SAS® 9.4 Procedures

Guide: Statistical Procedures, Sixth Edition. Cary, NC:

SAS Institute Inc. 288-549.

Searle, S. R. 1971. Linear Models. John Wiley & Sons,

Inc., Printed in USA. 532 p.

Severino Lendechy, V. H., Montiel-Palacios F, Pérez

Hernández P. 2017. Age and weight at puberty in

tropical dairy criollo heifers with and without

supplementary feeding. Ecosist. Recur. Agropec.

4:555-561.

Shalaby, N.A. 2005. Genetic Evaluation for Milk

Production, Reproduction Traits and Persistency of

Lactation Using Single and Two-trait Animal Model

Analyses for Friesian Cows in Commercial Herds in

Egypt. J. Agric. Sci. 30:3637-3653.

Soto-Belloso, E. y Perea – Ganchou, F. 2014.

Comportamiento de los cruces entre razas Bos – tauris

y Bos indicus en una finca de Doble propósito. En:

Logros & Desafios de la Ganaderia de Doble

Proposito. C. Gonzalez-Stagnaro, N Madrid-Bury, E

Soto-Belloso (eds). Fundacion GIRARZ. Ediciones

Astro Data S.A. Maracaibo, Venezuela, Cap. LXIII:

594-606.

Sun, C. Madsen, P., Nielsen, U. S., Zhang, Y., Lund, M.

S. and Su, G. 2009. Comparison between a sire model

and an animal model for genetic evaluation of fertility

traits in Danish Holstein population. J. Dairy Sci.

92:4063–4071.

Walsh, S.W., Williams, E. J., and Evans, A, C. 2011. A

review of the causes of poor fertility in high milk

producing dairy cows. Animal Reproduction Science.

123:127-138.

Weller, J. I. and Ron, M. 1982. Genetic analysis of fertility

traits in Israeli Holstein by linear and threshold

models. J. Dairy Sci. 75:2541-2548.

Weller, J.I., Ezra, E., Gershoni, M. 2022a. Genetic and

genomic analysis of age at first insemination in Israeli

dairy cattle. J. Dairy Sci. 105:5192-5205.

Weller, J. I. Gershoni, M., Ezra, E. 2022b. Breeding Dairy

Cattle for Female Fertility and Production in the Age

of Genomics. Vet. Sci. 2022,9,434

https://doi.org/10.3390/vetsci9080434

Windig, J.J.; Calus, M.P. Beerda, B.; and Veerkamp, R.

F. 2006. Genetic Correlations Between Milk

Production and Health and Fertility Depending on

Herd Environment. J. Dairy Sci. 89:1765–1775.