Genetic and phenotypic relationship between fertility and lactation traits in crossbred dairy cows in Ethiopia

Kefale Getahun and Nibo Beneberu

Ethiopian Institute of Agricultural Research, Holetta Research Center. P O Box 2003 Addis Ababa or 31 Holetta, Ethiopia
bnibo1984@gmail.com

Abstract

This study aimed to estimate genetic and phenotypic relationships or correlations for fertility and lactation traits of crossbred dairy cows at Holetta research center. Recorded data on 11331 calving and 8123 lactations were collected during the last 43 years (1974-2017). The genetic and phenotypic correlation values for fertility and lactation traits were estimated using WOMBAT software through multivariate analysis. The genetic correlations between fertility traits varied from -0.31±0.08 to 0.98±0.03 while the values of phenotypic correlations were ranged from -0.03±0.001 to 0.84±0.04. For lactation traits, the genetic correlations varied from 0.11±0.03 to 0.77±0.12 while the values of phenotypic correlation were ranged from 0.07±0.02 to 0.36±0.07. The direct genetic correlations estimate for age at first service-age at first calving, calving interval-lactation length, lactation milk yield-lactation length and days open-lactation length were 0.98±0.03, 0.79±0.07, 0.77±0.12 and 0.68±0.2, respectively.. The positive direct genetic correlations among traits in the present study indicate that the selection of one trait might improve the other trait. Knowing the genetic and phenotypic relationship between fertility and milk yield traits are significant for animal performance improvement and selection should be done by showing traits high favorable correlations.

Keywords: crossbred, Ethiopia, genetics, lactation, multivariate analysis

Introduction

The interaction among two or multiple genes in two or more traits produces genetic relationship while multiple genes interact with multiple environmental variables produces phenotypic relationship. Therefore, genetic and phenotypic relationships or correlations between in two or more traits are raised from these biological facts. Correlations are measures of the strength of the relationship between two variables or traits. A high genetic or phenotypic correlation values between two or more traits implies a strong relationship between them and vice versa (Bourdon 2000). Correlations are important as an aid in the prediction of response to selection in one trait due to selection in another and are partitioned into phenotypic and genotypic correlations. The genetic correlation expresses the extent to which two characters are influenced by the same genes or by genes located in the same chromosome and it is important when selecting for net merit involving several traits. Estimates of genetic correlation between any pair of traits suggest that selection for one trait can lead to an indirect genetic response in the other trait (Edward et al 2013, Gebeyehu et al 2014).

The potential for genetic improvement of a trait largely depends upon genetic variation existing in the population (Zeleke 2019). The most common genetic parameters are heritability, repeatability, genetic and phenotypic correlation (Yibrah 2008).

Furthermore, the development of effective genetic improvement programs require advanced knowledge of the genetic variation of economically important reproductive and production traits and accurate estimates of genetic and phenotypic correlations of economically important traits (Solomon et al 2002, Juma and Alkass 2006). However, there is limited information on genetic and phenotypic correlations among reproductive and milk production traits of crossbred dairy cows under different dairy management systems in Ethiopia. Therefore, the objective of this study was to estimate the genetic and phenotypic correlation between fertility and milk yield traits for crossbred dairy cows at Holetta research center.

Materials and methods

Description of the study area

This research was conducted at Holetta Agricultural Research Center (HARC). Holetta is located in the central highland of Ethiopia at 35 km west of Addis Ababa (3°24´N to 14°53´N latitude and 33°00´E to 48°00´E longitude) with an altitude of 2400 meter above sea level. The average annual rainfall is 1100 mm and average annual temperature is 15 ⁰C with minimum 6 ⁰C and maximum 24 0C, respectively (Yohannes et al 2016). The average monthly relative humidity is 60% (Gebregziabher et al 2013).

Animal management

The cattle were herded based on breed, pregnancy, lactation stage, sex and age. Uniform feeding and management practices were adopted for all animals within each category. Natural grazing, hay and concentrate supplement constituted the major feed supply. During the daytime animalswere allowed to graze on pastureland from early morning, 8.00 AM to 4.00 PM. Natural pasture hay was provided as additional feed during the evening. Concentrate mixture composed of wheat middling (32%), wheat bran (32%), noug (Guizocia abyssinica) cake (34%) and salt (2%) was supplemented based on their body weight, productivity, and physiological categories. Milking cows, heifers and calves were supplemented with concentrate mixture at a rate of 4, 1-1.5 and 0.25-1 kg per day, respectively, depending up on availability of supplemental feed. The cows had free access to clean tap water all the time. Calves were allowed to suckle their dam immediately after birth for about four days to receive colostrum. Weighting and ear tagging were also engaged within 24 hours after birth. After 4 days, calves were taken into calf rearing pen and continued to feed recommended amount of whole milk for 98 days through artificial rearing system (bucket feeding) except the F1 calves, which have been suckling their dams until winning since 2002. Weaned calves were transferred to other pen and kept indoor until 6 months. The animal management was also supported with vaccination against major diseases and treatment to control any incidence of diseases. Feeding, disease and other management practices were limited by the government budget. Therefore, there are no constant feeding and management practices implemented across seasons in the year or across in the whole years.

Data source and data collection

The data for this study was obtained from long-term records of crossbred dairy cows that has been kept for crossbreeding experiment at Holeta dairy research farm. The recorded data for the last 43 years (1974-2017) to estimate genetic parameters (genetic and phenotypic correlations) for fertility and lactation data were used for this study. Identity number of the animal was sequenced by pedigree viewer software package (version 6.5) for arranging animals ID and pedigree identity in chronological orders and to clear any mistake in identity number to obtain the relationship matrix. The pedigree data includes animal ID, dam and sire of a cow. Numbers of records on pedigree characteristics for estimation of genetic and phenotypic correlations for fertility and lactation traits were included in table 1.

Traits studied

The traits included in this study were categorized into fertility and lactation traits. The fertility traits include age at first service (AFS) months, age at first calving (AFC) months, calving interval (CI) days, days open (DO) days and number of service per conception (NSPC) repetitions. The lactation traits include lactation milk yield (LMY) liters, daily milk yield (DMY) liters and lactation length (LL) days.

Statistical analysis

The genetic and phenotypic correlations were estimated by using WOMBAT software (Meyer 2012) fitted an animal model using multivariate analysis. Fixed factors (Genetic group, year, season and parity) that have a significant effect were included in the model for estimation of genetic and phenotypic correlations. Additive (animal) and permanent environment for repeated records were fitted as random effects in the model. For CI, DO, NSPC, LMY, DMY and LL a repeatability animal model was fitted, where direct additive effects and permanent environmental effects were fitted as random effects due to repeated records per cow. For the other traits (AFS and AFC), however, direct additive genetic effect was the only random effect fitted. The representation of multi trait animal model used to estimate genetic and phenotypic co (variances) for fertility (AFS, AFC, CI, DO and NSPC) and lactation (LMY, DMY and LL) traits are as follow:

Y = Xb + Za + Wd + e where;

Y is a vector of records/ observations for the traits of interest (AFS, AFC, CI, DO, NSPC, LMY, DMY and LL),

b is a vector of fixed effects which had a significant effect (breed/Genotype, year, season and parity),

a is a vector of random individual direct additive genetic effects (animal),

X is a matrix relating records to fixed effects,

Z is an incidence matrix for direct additive genetic effect,

W matrices of permanent environmental effects,

e is a vector of random residual effect (error term).

This model assumed as expected value of Y to be Xb. The vector random individual additive effects, permanent environmental effects and residual effects are assumed to be uncorrelated and have expected mean of zero and variances σa², σc² and σe², respectively. From these expectations, WOMBAT is estimated the direct (co) variance, genetic and phenotypic correlations for each trait from multigenerational pedigree data.

Therefore, the genetic and phenotypic correlations are calculated by using the following formulas:

Where;

Result and discussion

Genetic and phenotypic relationships

The estimates of direct genetic and phenotypic correlations between five fertility traits (AFS, AFC, CI, DO and NSPC) and three lactation traits (LMY, DMY and LL) are shown in table 2. The differences observed among the estimated genetic and phenotypic correlations with that of others reported could be attributed to differences in sample size, breeds (genetic makeup) used, models and procedures employed to estimate parameters and environmental conditions such as feeding, climate and management practices. Genetic correlations between the traits in the present study were higher than the corresponding phenotypic correlations among most traits except for some fertility traits that have observed negative values in both genetic and phenotypic correlations. However, in this study, traits which have shown a positive genetic correlation had shown a negative phenotypic correlation or vice versa. As shown below the table, antagonistic genetic and phenotypic correlation were observed between AFS and AFC with CI, AFC and CI with NSPC, AFC and DO with LL and AFC with DMY.

Genetic relationships

The present study shows negative, weak positive and strongly positive genetic relationships among fertility traits. A negative relationship between AFS and DO (-0.001±0.003), AFC and DO (-0.05±0.01), AFS and NSPC (-0.02±0.01), AFC and NSPC (-0.29±0.06) and CI and NSPC (-0.31±0.08) were observed. The negative genetic relationship or correlation of these traits indicates that genes that influenced one trait in the favorable direction which influenced other traits in the unfavorable ways. Strong and positive genetic relationships (0.98±0.03) were appeared between AFS and AFC traits. The positive genetic relationships might be arising due to the pleiotropic effect of gene and some linkage among genes. However, a moderate (0.36±0.04) genetic relationship was observed between CI and DO. The relationships between AFS and AFC with DO and NSPC in this study disagree with those found by Haile et al (2009b) who reported 0.51 for AFS and DO, 0.19 for AFC and DO, 0.38 for AFS and NSPC and 0.65 for AFC and NSPC, respectively. This difference might be due to variation in size of the data set, estimation procedure and analysis type. Other study carried out by Gutterez et al (2002) found comparatively higher relationship or correlation (0.23) between AFC and CI for beef crossbred. Strong positive relationship or correlation (0.99±0.00) was reported by Tadesse (2014) between CI and DO. However, a perfect positive genetic correlation (1) between AFS and AFC and CI and DO for Fogera crosses were reported by Belay et al (2016). These differences might be due to variation in size of the data set, breed, estimation procedure and analysis type.

The genetic relationship or correlations among lactation traits in the present study were positive and ranges from low (0.11±0.03) to slightly high (0.77±0.12). High genetic correlation was observed between LMY and LL (0.77±0.12). This value was closely similar with the study by Gebregziabhere et al (2013) who reported 0.73 correlations between LMY and LL. Furthermore, Ashutosh et al (2013) reported an estimated correlation 0.31 for LMY and LL and 0.30 for LMY and DMY. Haile et al (2009a) reported genetic correlation between LMY and LL (0.55±0.12) and between DMY and LL (0.78±0.12). Tadesse (2014) also reported moderate to very strong genetic correlation (0.59, 0.96 and 0.99) between DMY and LL, LMY and DMY and LMY and LL, respectively.

The genetic relationship or correlations between fertility and lactation traits were also examined in this study and there were closely associated with each other in some traits. Accordingly, slightly higher genetic correlations were appeared between CI and LL (0.79±0.07) and DO and LL (0.68±0.2). AFS with DMY and AFS with LL traits have shown negative genetic correlation (-0.55±0.04 and -0.11±0.09) while no genetic relationship or correlation was observed between AFS and DMY traits. The relationship between CI with LL and LMY estimated for this study was different from the report of Tadesse (2014) who found a correlation of 0.81±0.06 and 0.59±0.06, respectively but a negative genetic correlation between CI and LL was reported by (Gutterez et al 2002). This study was also deviated from Ashutosh et al (2013) who reported a correlation of -0.14, 0.12, 0 and 0.03 between AFC and LMY, CI and LMY, AFC and LL and CI and LL, respectively for Bangladesh HF crosses with Sahiwal and local bred. The positive genetic correlation among traits (LMY with DMY, LL with DMY, LMY with CI, and AFS with AFC and DMY with DO) in the present study indicated that selection of one trait might be important for the improvement programme of other traits. However, traits which have shown negative or zero values (AFS with DO, AFC with NSPC, CI with NSPC and AFS with DMY) in the present study is considered as uncorrelated of each other and the result of independent gene action and/or selection to improve single trait might disfavour other traits.

Generally, the positive direct genetic correlations among traits in the present study indicated that the selection of one trait might be important for the improvement of other traits. Besides, these high genetic correlation values are due to the phenomenon of a single gene affecting more than one trait and due to the occurrence of two or more loci that affect the same trait on the same chromosome (Bourdon 2014). However, traits that have shown negative direct genetic correlations in the present study indicate that as one trait increases, the other trait tends to decrease, which might be favorable or unfavorable depending on the combination of traits considered.

Phenotypic relationships

The phenotypic relationships among fertility traits were indicated in table 2. Strong phenotypic relationship or correlation was observed between AFS and AFC whereas inversely correlated traits (AFS with CI, DO and NSPC and AFC with CI and DO) were existed. Like the genetic relationship or correlation of fertility traits, the positive and negative phenotypic relationships or correlations of these traits are strongly disagreed with the finding of Belay et al (2016) who found very strong phenotypic correlation between AFS and AFC (0.86) and CI and DO (0.99) and also Beneberu et al (2021) report strong phenotypic relationship between AFS and AFC (0.98±0.00) and CI and DO (0.89±0.00) for pure Jersey cows. Tadesse (2014) also reported very strong phenotypic correlation between CI and DO (0.99) for crossbred cattle. The variation of the present study from others literature might be due to breed, number of observations studied and software procedure used for analysis.

Phenotypic relationships or correlations were also estimated among lactation traits were very low (0.07±0.02 between DMY and LMY), low (0.18±0.05 between DMY and LL) and moderately low (0.36±0.07 between LMY and LL). The phenotypic relationship between LMY and LL agreed with the report of (Ashutosh et al 2013). Relative to the present study, Tadesse (2014) found higher phenotypic correlations between DMY and LL (0.39), DMY and LMY (0.86) and LMY and LL (0.76) for crossbred cattle.

The phenotypic relationships or correlations between fertility and lactation traits were weak positive and negative (uncorrelated between them). Lactation length was a negative phenotypic correlation with AFS, AFC and DO. However, the correlation between AFC with DMY and AFS with LMY were good compared to other traits. The present study suggested that both random environmental and genetic (additive and non- additive) effects could influence phenotypic correlation of the studied traits. The negative phenotypic correlation between AFC and LL in the present study was agreed with the finding of (Ashutosh et al 2013). Phenotypic correlations between fertility and lactation traits in the present study were lower than direct genetic correlation estimates. Both environmental and genetic effects could influence phenotypic correlations of the fertility and lactation traits.

Conclusion

Genetic interaction between traits in the present study was high and important which means, majority of traits were positive relationships among them. For phenotypic relationships, traits showed smaller values and an indicator of management attentions. Milk yield traits are positive genetic and phenotypic relationships. From fertility traits, ages at first service-age at first calving were found to be strong positive genetic and phenotypic correlations. The genetic and phenotypic relationship between fertility and lactation traits were negative in mostly which would be humper the acceleration of genetic progress through selection of total traits. The only positive genetic correlations between all the fertility and lactation traits in this study implies that they all are being controlled by similar gene and indicating that selection for one trait will improve other correlated traits in a desired direction, helping the breeding process as a whole by improvement in all the traits correlated with each other. In most of our selection and breeding program, fertility traits are not included in the selection processes. This showed great economic losses in the dairy industry as many research reports showed. Thus, the relationships between fertility and lactation traits in the present study are an important outset for consideration of selection and designing in dairy cattle genetic improvement program.

Acknowledgment

This study was done by using long-term crossbred data which is available at Holetta dairy research farm. The authors would like to thank the research farm.

References

Ashutosh Das, Gous Miah, Mukta Das Gupta and Kabiru Islam Khan 2013 Genetic parameters of Holstein crossbred on commercial dairy farms in Chittagong, Bangladesh. Indian Journal of Animal Research. 47 (4): 327-330. https://arccjournals.com/journal/indian-journal-of-animal-research/ARCC149

Belay Zeleke, Kefelegn Kebede and Ajay Kumar Banerjee 2016 Estimation of genetic parameters for reproductive traits of Fogera and Holstein Friesian crossbred cattle at Metekel ranch, Amhara region, Ethiopia. Online Journal of Animal and Feed Research. 6: 90-95. https:// www.ojafr.ir/main/attachments/article/122/Online%20J.%20Anim.%20Feed%20R es.,%206(4)%2090-95,%202016.pdf

Beneberu N, Alemayehu K, Mebratie W, Getahun K, Wodajo F and Tesema Z 2021 Genetic and phenotypic correlations for reproductive and milk production traits of pure Jersey dairy cows at Adea-Berga, central highland of Ethiopia. Livestock Research for Rural Development. Volume 33, Article #46. Retrieved February 3, 2022, from http://www.lrrd.org/lrrd33/3/bnibo3346.html.

Bourdon R M 2000 Understanding animal breeding. 2nd ed. Upper Saddle River, New Jersey 07458, USA.

Bourdon R M 2014 Understanding animal breeding. 2nd ed. Harlow: Pearson education Limited.165, 279 & 281 pp.

Edward Missanjo, Venancio Imbayarwo-Chikosi and Tinyiko Halimani 2013 Estimation of genetic and phenotypic parameters for production traits and somatic cell count for jersey dairy cattle in zimbabwe. ISRN Vet Sci. 2013. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3728504/

Gebeyehu Goshu, Harpal Singh, Karl-Johan Petersson and Nils Lundeheim 2014 Heritability and correlation among first lactation traits in Holstein Friesian cows at Holeta Bull Dam Station, Ethiopia. International Journal of Livestock Production, 5(3), 47-53. https://academicjournals.org/journal/IJLP/article-abstract/A700B5F44832

Gebregziabher Gebreyohannes, Skorn Koonawootrittriron, Mauricio A, Elzo and Thanathip Suwanasopee 2013 Variance components and genetic parameters for milk production and lactation pattern in an Ethiopian multi-bred dairy cattle population. Asian Australasian Journal of Animal Science. 26 (9): 1237-1246. https:// www.ncbi.nlm.nih.gov/pmc/articles/PMC4093399/

J P Gutierrez, I Alvarez, I Fernandez, L J Royo, J Diez and F Goyache 2002 Genetic relationships between calving date, calving interval, age at first calving and type traits in beef cattle. Livestock Production Science. 78: 215-222. https://ria.asturias.es/RIA/bitstream/123456789/727/1/Archivo.pdf

Haile Aynalem, B K Joshi, Workneh Ayalew, Azage Tegegne and A Singh 2009a Genetic evaluation of Ethiopian Borena cattle and their crosses with Holstein Friesian in central Ethiopia: milk production traits. Animal, 3(4): 486-493. https:// www.researchgate.net/publication/221973164.

Haile Aynalem, B K, Joshi, Workneh Ayalew, Azage Tegegne and A Singh 2009b Genetic evaluation of Ethiopian Borena cattle and their crosses with Holstein Friesian in central Ethiopia: Reproductive traits. Journal of Agricultural Science, 147: 81-89. https://www.researchgate.net/publication/248620160.

Juma K H and Alkass J E 2006 Genetic and Phenotypic Parameters of Some Economic Characteristics in Awassi Sheep of Iraq: A Review. Egyptian Journal of Sheep, Goat and Desert Animals Sciences. 1: 15-29.

Meyer K 2012 WOMBAT, a program for mixed model analyses by restricted maximum likelihood. User notes. Animal Genetics and Breeding Unit, University of New England Armidale, Australia.

Solomon Abegaz, Negussie E, Duguma G and Rege J E O 2002 Genetic parameter estimates for growth traits in Horro sheep. Journal of Animal Breeding and Genetics. 119: 35-45. https://onlinelibrary.wiley.com/doi/abs/10.1046/j.1439-0388.2002.00309.x

Tadesse Birhanu 2014 Estimation of crossbreeding parameters in Holstein Friesian and Ethiopian Boran-crosses for milk production and reproduction traits at Holeta agricultural research center, Ethiopia. MSc. Thesis, Haramaya University, Ethiopia, 83 pp.

Yibrah Yacob 2008 Environmental and genetic parameters of growth, reproductive and survival performance of Afar and Blackhead Somali sheep at Werer Agricultural Research Center. Fellowship report submitted to International Livestock Research Institute (ILRI) and Ethiopian Institute of Agricultural Research (EIAR). Ethiopia. 70Pp.

Yohannes Gojam, Million Tadesse, Kefena Effa and Direba Hunde 2016 Performance of crossbred dairy cows suitable for smallholder production systems at Holetta Agricultural Research Centre. Ethiopian Journal of Agricultural Science 27(1): 121-131. https:// www.ajol.info/index.php/ejas/article/view/150378

Zeleke Tesema 2019 On-station and on-farm performance evaluation and genetic parameters estimation of Boer x Central Highland Crossbred Goat in North Wollo Zone, Ethiopia. M.Sc. Thesis, Bahir Dar University, Ethiopia, 150 pp.