Genetic evaluation of multibreed populations of milking bovines

Evaluación genética de poblaciones multirraciales de bovinos lecheros

Arelis Hernández,^I Raquel Ponce de León ,^I Sonia Maria García,^II

^IInstituto de Ciencia Animal, Apartado Postal 24, San José de Las Lajas, Mayabeque, Cuba.
^IIEmpresa Pecuaria Genética de Matanzas. Finca San Andrés. Limonar. Matanzas, Cuba.

ABSTRACT

In order to fulfill this study, a total of 130,111 records of accumulated milk production, up to 305 days (L305), from three Cuban genetic enterprises and between 1984 and 2013, were used. Out of a total of 61,471 cows, 12,496 belonged to Mambí de Cuba breed (3/4 Holstein ¼ Zebu), 17,178 Siboney de Cuba breed (5/8 Holstein 3/8 Zebu) and 31,797 Holstein. The purpose of the study was to compare the multibreed genetic evaluation for milk production with genetic evaluations of each breed. To estimate the genetic indicators and breeding values of L305, a univariate animal model was used, which included, as fixed effects, the combination herd-year-four-month period of parturition. The age at parturition was used as linear and square covariable, random effects of the animal, the effect of the environment and residues. Breed was also included as fixed effect in the multibreeding genetic evaluation. Heritabilities and repetitions were  0.31 ± 0.01 y 0.40 ± 0.01 for the multibreeding genetic evaluation, and 0.18 ± 0.05 and  0.41±0.02, 0.14± 0.05 and 0.41 ± 0.02, 0.12 ± 0.01 and 0.25 ± 0.01 for the independent racial evaluations of  Mambí de Cuba, Siboney de Cuba and Holstein, respectively. In the multibreeding genetic evaluation, the estimation of heritability and the accuracy of breeding values for milk production of the three studied breeds increased regarding the evaluations of each individual breed, mainly because of a higher amount of information and higher connectivity among breeds, which contribute to decrease the environmental variance. It can be concluded that the inclusion of multibreeding genetic evaluations on the program of genetic improvement of Cuban milking bovines allows to obtain a better accuracy of breeding values, which will also contribute to the increase of the genetic progress for milk production.

Key words: genetic evaluation, multibreed, crossing.

RESUMEN

Se utilizaron  130 111 registros de producción de leche acumulada hasta los 305 días (L305), procedentes de tres empresas genéticas cubanas, durante los años 1984 al 2013. De un total de 61 471 vacas, 12 496 fueron de la raza Mambí de Cuba (3/4 Holstein ¼ Cebú), 17 178 Siboney de Cuba (5/8 Holstein 3/8 Cebú) y 31 797 Holstein. El propósito del estudio fue comparar la evaluación genética multirracial para la producción de leche con las evaluaciones genéticas para cada raza de manera independiente. Para estimar los parámetros genéticos y valores genéticos (VG) de L305 se empleó un modelo animal univariado que incluyó como efectos fijos la combinación rebaño-año-cuatrimestre de parto, la edad  al parto como covariable lineal y cuadrática, los efectos aleatorios del animal, el efecto del ambiente permanente y los residuales. En la evaluación genética multirracial se incluyó además como efecto fijo la raza. Las heredabilidades y  repetibilidades fueron: 0.31 ± 0.01 y 0.40 ± 0.01 para la evaluación genética multirracial, y de 0.18 ± 0.05 y  0.41 ±  0.02, 0.14 ± 0.05 y 0.41 ± 0.02, 0.12 ± 0.01 y 0.25 ± 0.01 para las evaluaciones raciales independientes del Mambí de Cuba, Siboney de Cuba y Holstein, respectivamente. En la evaluación genética multirracial se incrementó el estimado de heredabilidad y la exactitud de los VG para la producción de leche de las tres razas estudiadas con respecto a las evaluaciones de cada raza independiente, debido fundamentalmente a la mayor cantidad de información y mayor conectividad entre las razas, que contribuye a disminuir la varianza ambiental. Se concluye que la incorporación de las evaluaciones genéticas multirraciales, en el programa de mejoramiento genético de los bovinos lecheros cubanos, permite obtener mayor exactitud de los VG, lo cual contribuirá al incremento del progreso genético para la producción de leche.

Palabras clave: Evaluación genética, multirrazas, cruzamiento.

The beginning of genetic improvement in Cuban cattle rearing dates back to the 60´s, after being proof, through the use of semen, different breeds specialized in milk production (Holstein, Jersey, and Brown Swiss), which were mated with Zebu cows in order to study the performance of crossed F1 females, under the same conditions. These researches originated the Siboney de Cuba (5/8 Holstein 3/8 Zebu) and Mambí de Cuba (3/4 Holstein ¼ Zebu) breeds. Between 1981 and 1991, animal husbandry, which was mainly developed for meat animals (88%), changed because 84 % of the animals started to be used for milk production and 16 % of them for meat. From 1992 to 2002, the insemination started with more rudimentary genotypes like Criollo, Milking Zebu (3/4 Zebu 1/4 Holstein) and Zebu, until reaching the current genotypes, a herd under insemination, with the best participation of Siboney de Cuba. An amount of 56.4 % of all the females from milking breeds included on the genetic herds are Siboney de Cuba and Siboney crossbred, 15.5% are Mambí de Cuba and Mambí crossbred and 6.5 % are Holstein (DG 2012).

Procedures for multibreed genetic evaluations, according to Elzo and de los Reyes (2004), allow to perform more precise additive genetic predictions, a direct comparison of animals from different breed composition, as well as the additive and non-additive genetic improve. It also allows to consider the heterogeneity of variances and covariances among breed groups (pure and crossed)

Nowadays, in Cuban milking breeds (Mambí de Cuba, Siboney de Cuba and Holstein), the accumulated milk production at 305 days of lactation are used as selection criteria, and a uni-trait model is used for the genetic evaluation, through the Animal Model BLUP methodology. This model performs an independent evaluation of each breed, disregarding the fact that populations are multibreed because there is more than one breed in the same herd. On the other hand, breeds obtained from crossing have common ancestors with Holstein breed, so the animals are related in the pedigree. In addition, studies carried out by Acosta et al. (2013) reported that Mambí de Cuba and Siboney de Cuba breeds are genetically related. Due to all the previous reasons, the objective of this study was to compare the results of genetic evaluations in milking bovines, using each breed, to those obtained from multibreed evaluations.

MATERIALS AND METHODS

Records of accumulated milk production, up to 305 days, from Mambí de Cuba (¾ Holstein ¼ Zebu), Siboney de Cuba (5/8 Holstein 3/8 Zebu) and Holstein, were used. This data was provided by three genetic enterprises (Empresa Pecuaria Genética de Matanzas, Empresa Los Naranjos and Empresa Camilo Cienfuegos) and it includes the period between 1984 and 2013. All these enterprises are located in the western region of the Republic of Cuba, at the Gulf of Mexico, between 20 and 23° N and 74 and 85° W.

In order to guarantee a better precision for estimating parameters, some data was removed including those lactations with less than 100 days (4.57 % of data), ages at parturition inferior to 24 months (0.31 % of data), milk productions with less than 300 kg (3.98 % of data) and lactations superior to the tenth lactation (0.16% of data). The groups of contemporaries, formed by less than 3 animals, were also removed. The combination of herd-year-four-month period of parturition was considered as group of contemporaries.

The final sample included a total of 130,111 lactations (52,511 Holstein, 35,725 Mambí de Cuba and 41,875 Siboney de Cuba) from 61,471 cows (31,797 Holstein, 12,496 Mambí de Cuba and 17, 178 Siboney de Cuba). Males were represented by a total of 1,300 bulls (686 Holstein, 245 Mambí de Cuba and 339 Siboney de Cuba).

In a first analysis, a group of data was used with the information of the three breeds together. Components of variance, heritabilities (h²), repetitions (r) and breeding values were estimated through ASREML program (Gilmour et al. 2003), using the following univariate animal model:

Where, y: vector of accumulated milk production up to 305 days, b: vector of fixed effects that contain breed effect, groups of contemporaries and age at parturition as linear and square covariable, a: vector of random effects of the animal, p: vector of the effect of permanent environment of cow,  X, Z and W: matrixes of design or of incidence that relate fixed, random and permanent environment effects with data, respectively, e: vector of random residual effects.

Later, data were divided into three samples, where each database contained the information of each individual breed. Components of variance, heritabilities (h²), repetitions (r) and breeding values were estimated per each breed using a similar model. However, the fixed effect of breed was not considered.

The pedigree file was formed by 153,963 individuals. The pedigree information included the three breeds together and up to the grandparents through maternal and paternal line.

RESULTS AND DISCUSSION

Table 1 shows the statistics (mean, standard deviation and coefficient of variation) estimated from variances, heritabilities and repetitions of milk production up to 305 days, in the analysis of three breeds together (multibreed) and each of them individually. The estimates of additive genetic variances, heritabilities and repetitions were superior in the multibreed analysis. These results come from the use of a sample composed by a higher number of animals related, which included the information of female offspring of Holstein bulls, belonging to Mambí de Cuba and Siboney de Cuba breeds. Therefore, there is an increase of the number of female offspring per bull, contributing to the genetic evaluation of each bull.

Table 2 shows breeding values for milk production up to 305 days and their accuracy in evaluated sires through multibreed analysis and in each breed individually. It demonstrates that using multibreed analysis, there is a better accuracy of the genetic evaluation, mainly because of the inclusion of records of crossed breeds (Mambí de Cuba and Siboney de Cuba) and those of the pure breed (Holstein). Therefore, there is database that contains a larger amount of information and better connectivity among animals that contributes to decrease environmental variance.

Elzo and de los Reyes (2004) stated that the methodologies for the genetic evaluation of only one breed assume that all animals in the population belong to a single breed. However, most livestock populations are produced by crossing, either by backcrossing one of the parental breeds, or composed by fractions of several breeds. Crossbred animals function as a connection to pure parental populations, generating a multibreed population consisting of pure and crossbred animals.

 In the studied multibreed population, breeds obtained from the crossing (Mambí de Cuba and Siboney de Cuba) have common ancestors of Holstein breed, because 16 Holstein sire are parents of Mambí de Cuba and Siboney de Cuba cows, so the animals of these three breeds are related in the pedigree. Therefore, by forming the pedigree file with the three breeds, there is a higher connectivity among the animals is obtained.

In the multibreed genetic evaluation, there was an increase of the estimations of heritability, repetition and accuracy of breeding values for milk production, regarding the evaluations of each individual breed, mainly because a higher amount of information and better connectivity among breeds, which contributes to decrease the environmental variance. It can be concluded that the inclusion of multibreed genetic evaluations on the genetic improvement program of Cuban milking bovines allows to obtain a better accuracy of breeding values, which will contribute to the increase of genetic progress for milk production.

REFERENCES

Acosta, A. C., Uffo, O., Sanz, A., Ronda, R., Osta, R., Rodellar, C., Martin-Burriel, I. & Zaragoza, P. 2013. “Genetic diversity and differentiation of five Cuban cattle breeds using 30 microsatellite loci”. Journal of Animal Breeding and Genetics, 130 (1): 79–86.

DG 2012. Sistema integral para la política genética animal. (ser. Technical Report), Ministerio de la agricultura,  Dirección de genética animal.

Elzo, M. A. & de los Reyes, A. 2004. “Perspectives for multibreed genetic evaluation of cattle in Brazil”. Ciencia Animal Brazileira, 5: 171–185.

Gilmour, A. R., Gogel, B. D., Cullis, B. R., Welham, S. J. & Thompson, R. 2003. ASREML. User Guide Release. 1.10 ed.

Received: November 24, 2015
Accepted: January 1, 2016

Arelis Hernández, Instituto de Ciencia Animal, Apartado Postal 24, San José de Las Lajas, Mayabeque, Cuba. Email: arelishdez@ica.co.cu