- Email: [email protected]
Direct and maternal effects for body measurements at birth and weaning in Muzaffarnagari sheep of India
Available online at www.sciencedirect.com
Small Ruminant Research 75 (2008) 123–127
Direct and maternal effects for body measurements at birth and weaning in Muzaffarnagari sheep of India Ajoy Mandal ∗ , R. Roy, P.K. Rout Genetics and Breeding Division, Central Institute for Research on Goats, Makhdoom, Mathura 281122, Uttar Pradesh, India Received 13 July 2007; received in revised form 17 August 2007; accepted 27 August 2007 Available online 24 October 2007
Abstract Estimates of co(variance) components were obtained for body measurements at birth and weaning in Muzaffarnagari sheep maintained at the Central Institute for Research on Goats, Makhdoom, Mathura, India, over a period of 27 years (1978 through 2004). Records of 5821 lambs descended from 176 rams and 1698 ewes were used in the study. Analyses were carried out by REML fitting an animal model and ignoring or including maternal genetic or permanent environmental effects. Six different animal models were fitted for all traits. The best model was chosen after testing the improvement of the log-likelihood values. Direct heritability estimates were inflated substantially for all traits when maternal effects were ignored. The direct heritability estimates for body length, height at withers and heart girth of lambs at birth were 0.14, 0.14 and 0.07, respectively. The corresponding maternal heritability estimates for these traits at birth were 0.13, 0.15 and 0.13, respectively. Moderate estimates of the direct heritability (h2 ) and the fraction of variance due to maternal permanent environmental effects (c2 ) for body length (h2 = 0.12, c2 = 0.08), height at withers (h2 = 0.16, c2 = 0.08) and heart girth (h2 = 0.15, c2 = 0.09) was observed at weaning. Results suggest that maternal additive effects were only important at birth for these traits but permanent environmental maternal effect had some influence on body measurements at weaning. These results indicate that modest rates of genetic progress appear possible for body measurements at birth and weaning. © 2007 Elsevier B.V. All rights reserved. Keywords: Genetic parameters; Maternal effects; Body measurements; Muzaffarnagari sheep
1. Introduction Body conformation and growth rate of animals are important criteria for selection of breeding animals in meat producing species. In many livestock species, conformation traits have been included in the genetic evaluation procedures and selection programmes can incorporate estimated breeding values for these traits. Body dimensions or linear measurements have been used
∗
Corresponding author. Tel.: +91 565 2763280x555; fax: +91 565 2763246. E-mail addresses: [email protected], [email protected] (A. Mandal). 0921-4488/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.smallrumres.2007.08.004
as indicators of body size and weight, and live body weight can be predicted from body measurements (Aziz and Sharaby, 1993; Valdez et al., 1997; Varade and Ali, 1999; Atta and El khidir, 2004). The phenotype of an animal for size and conformation is the result of the genetic potential and the influence of environment as well as maternal effects. Measurements of body dimensions of animals may be taken at a relatively early age; therefore the influence of maternal effects on these traits needs to be quantified in order to formulate optimal breeding programmes. Willham (1980) postulated that knowledge of the magnitude of the (co)variance components for traits of economic importance is critical for the genetic evaluation of animals and the development of sound breeding programs. Studies on traits measured
124
A. Mandal et al. / Small Ruminant Research 75 (2008) 123–127
at an early age in farm animals have shown that both direct and maternal influences are important for animal growth (Mar´ıa et al., 1993; Snyman et al., 1995; N¨asholm and Danell, 1996; Matika et al., 2003; Mandal et al., 2006a,b). Further, it was observed that exclusion of genetic or environmental factors may lead to overestimation of the (co)variance components fitted in the model. Thus, N¨asholm and Danell (1994) reported that when maternal genetic effects are of importance, but not accounted for, heritability estimates are biased upward and the realized efficiency of selection is reduced. Many of the published heritability estimates for body measurements of Indian sheep breeds were derived from sire models and based on small sets of data (Bhadula et al., 1980) that did not take account of additive maternal effects. However, no attempt has yet been made to apply REML procedures for estimation of genetic parameters for body dimensions of sheep. Therefore, the present study was conducted to estimate variance and covariance components due to direct genetic effects, maternal genetic effects and maternal permanent environmental effects for body measurements in Muzaffarnagari sheep.
seasons in May–June and October–November, with lambing in October–November and March–April. At birth each lamb was identified and date of birth, sex, type of birth and weight were recorded. Lambs were normally weaned at 3 months of age. A more detailed description of the breed and of its natural habitat, husbandry practices and production performances were provided by Mandal et al. (2000, 2003). 2.2. Data Data on birth (5821) and weaning (5060) records of lambs produced by 176 sires and 1698 and 1589 dams, respectively, and spread over a period of 27 years (1978–2004) were analyzed. The traits studied were body length (BL), height at withers (HW) and heart girths (HG) of animals, measured at birth and at weaning. BL was measured by taking the average of left and right side measurements (using a tape) of the distance between the point of the shoulder (lateral tuberosity of the humerus) and the pinbone (tuber ishii). HW was measured on the dorsal midline at the highest point on the withers by using a tape. HG was measured with tape by taking body circumference immediately posterior to the front leg. All body dimensions were taken when the animal was standing naturally with head raised and body weight on all four feet. 2.3. Statistical analyses
2. Materials and methods 2.1. Study area and flock management This study was carried out at the Central Institute for Research on Goats, Makhdoom, Mathura, located in India, 169 m above mean sea level and 10◦ N, and 78◦ 02 E with a semi-arid climate. The annual rainfall in the area averages 750 mm, while the average daily temperature varies from 6 ◦ C (winter) to 44.5 ◦ C (summer). The Muzaffarnagari sheep, an important mutton breed of India, is known for its relatively rapid growth rate (Singh, 1995), high feed conversion efficiency (Mandal et al., 2000) and very good adaptability (Mandal et al., 2003) in the semi-arid region of the country. The institute currently maintains a flock of 250 breeding ewes. The flock was generally reared under semi-intensive feeding. All animals grazed during the day (6–7 h) on natural pasture with supplementation depending upon the status and age category of the animals and were penned at night. The mean body weights of animals were 3.49 ± 0.04, 15.00 ± 0.20, 26.50 ± 3.56, 30.4 ± 0.40 and 33.4 ± 0.46 kg at birth, 3, 6, 9 and 12 months of age, respectively (Mandal et al., 2000). The ADG pre-weaning (0–3 months) was 155 ± 2.2 g and postweaning (3–6 months) 101.0 ± 2.9 g per day. The average feed conversion efficiency of this breed is reported to vary from 10% to 16%. Generally, controlled mating was practiced. Ewes in heat were mated with pre-assigned sires in the morning. Sheep were bred twice at each oestrus. Ewes were first exposed to rams at about 12 month of age. One breeding ram was allowed to mate with 20–25 ewes. There were two breeding
(Co)variance components were estimated by restricted maximum likelihood (REML) using a derivative-free algorithm fitting an animal model. Data were first analyzed by least-squares analysis of variance (Harvey, 1990) to identify the fixed effects to be included in the model. The statistical model included the fixed effects of birth year, season of birth, parity of dam, sex and birth status (single vs. twin) of lambs. All these effects were significant (P < 0.05) for all traits. The models used to estimate genetic parameters included random effects and all fixed effects that were found significant in least-squares analysis. Convergence of the REML solutions was considered reached when the variance of function values (−2 log L) in the Simplex was less than 10−8 . To ensure that a global maximum was reached, analyses were restarted for several other rounds of iterations using results from the previous round as starting values. When estimates did not change, convergence was confirmed. Standard errors were calculated for the estimated parameters as a part of the DFREML programme (Meyer, 2000). Univariate animal models were fitted to estimate the genetic parameters for each trait. By ignoring or including various combinations of maternal genetic and permanent environmental effects, the following six different models were used: y = Xb + Za a + e
(1)
y = Xb + Za a + Zc c + e
(2)
y = Xb + Za a + Zm m + e
with Cov(a, m) = 0
(3)
y = Xb + Za a + Zm m + e
with Cov(a, m) = Aσam
(4)
A. Mandal et al. / Small Ruminant Research 75 (2008) 123–127
y = Xb + Za a + Zm m + Zc c + e
with Cov(a, m) = 0
y = Xb + Za a + Zm m + Zc c + e
with Cov(a, m) = Aσam
(5)
(6) where y is the vector of observations for each trait; b, a, m, c and e are vectors of fixed effects (birth year, season of birth, parity of dam, sex and birth status of lamb), direct additive genetic effects (animal), maternal additive genetic effects, permanent environmental effects of dam and the residual effects, respectively; X, Za , Zm , Zc are the corresponding incidence matrices relating these effects to y; A is the numerator relationship matrix between animals; and σ am is the covariance between additive direct and maternal genetic effects. The analysis used the standard assumptions and definitions, i.e., V (a) = Aσa2 , V (e) = Iσe2 ,
V (m) = Aσm2 ,
V (c) = Iσc2 ,
Cov (a, m) = Aσam
where I is an identity matrix and σa2 , σm2 , σc2 and σe2 are direct additive genetic variance, maternal additive genetic variance, maternal permanent environmental variance and environmental variance, respectively. Estimates of heritability (h2 ), maternal heritability (m2 ) and permanent maternal environmental (c2 ) effects were calculated as ratios of estimates of σa2 , σm2 and σc2 , respectively, to the phenotypic variance (σp2 ). The direct-maternal correlation (ram ) was computed as the ratio of the estimates of direct-maternal covariance (σ am ) to the product of the square roots of estimates of σa2 and σm2 . The total maternal effect, tm = (1/4)h2 + m2 + c2 + mram h was calculated to estimate repeatability of ewe performance. The heritability estimates for the total genetic component was calculated (Willham, 1980) as h2t = h2 + 0.5m2 + 1.5mram h, to estimate the expected response to phenotypic selection. The most appropriate model for each trait was selected based on the likelihood ratio tests (Meyer, 2000). An effect was considered to have a significant influence when its inclusion caused a significant increase in log-likelihood, com-
125
pared to a model in which it was ignored. Significance was tested at P < 0.05 by comparing differences in log-likelihoods (−2 log L) to values for a chi-square distribution with degrees of freedom equal to the difference in the number of (co)variance components fitted for the two models. If −2 log L values were not significantly different (P > 0.05), the model with fewest random terms was chosen.
3. Results and discussion 3.1. Characteristics of data The least-squares means along with the standard errors for BL, HW and HG of animals were 33.65 ± 0.08, 37.54 ± 0.07 and 34.61 ± 0.07 cm, respectively at birth and the corresponding figures were 56.93 ± 0.14, 58.02 ± 0.13 and 57.40 ± 0.14 cm, respectively, at weaning. In these data, 51.0% of the lambs were males and 49.0% were females. Single and twin-born lambs represented 91.0% and 9% of the data, respectively. Coefficients of variation for body measurements ranged from 7.84% at birth to 9.48% at weaning. 3.2. Body measurements Estimates of (co)variance components and genetic parameters for BL, HW and HG of lambs for the most appropriate model at birth and weaning are summarized in Table 1. 3.2.1. At birth Estimates of direct additive heritabilities for body measurements at birth depended on the model used. Ignoring maternal effects (Model 1) produced higher estimates of σa2 and h2 than other models for all traits. Fit-
Table 1 Estimated parameters and their standard errors for body measurements from the best model for each traita Traits
Model
h2
m2
c2
σa2
2 σm
σc2
At birth BL HW HG
3 3 3
0.14 (0.04)b 0.14 (0.04) 0.07 (0.01)
0.13 (0.02) 0.15 (0.02) 0.13 (0.02)
– – –
0.967 0.975 0.458
0.942 1.029 0.908
– – –
At weaning BL HW HG
2 2 2
0.12 (0.03) 0.16 (0.03) 0.15 (0.03)
– – –
0.08 (0.02) 0.08 (0.02) 0.09 (0.02)
2.427 3.081 3.385
– – –
1.731 1.527 1.951
σe2
σp2
5.090 5.009 5.427
6.999 7.013 6.793
16.717 14.385 17.381
20.876 18.992 22.717
a BL is the body length, HW is the height at withers, HG is the heart girth, σ 2 is the direct additive genetic variance; σ 2 is the maternal additive a m genetic variance; σc2 is the maternal permanent environmental variance; σe2 is the environmental variance; σp2 is the phenotypic variance; h2 is the heritability; m2 is the maternal heritability; c2 = σc2 /σp2 . b Figures in parentheses are standard errors of the estimate.
126
A. Mandal et al. / Small Ruminant Research 75 (2008) 123–127
ting a permanent environmental maternal effect (Model 2) explained 9%, 10% and 10% of total phenotypic variance for BL, HW and HG of lambs, respectively, and the corresponding reduction in direct heritability was 33%, 30% and 44% in comparison with Model 1. Model 3, which included only direct and maternal additive effects, yielded an maternal effects contributing 13%, 15% and 13% for BL, HW and HG, respectively and also resulted a substantial reduction in the estimates of h2 relative to Model 2. Fitting a non-zero covariance (σ am ) along with a maternal genetic effect (Model 4) resulted in similar estimates of h2 and m2 as observed in Model 3 with the corresponding estimates of ram of 0.24, −0.02 and 0.21 for BL, HW and HG, respectively. In Model 5, which attempted to disentangle genetic and environmental components of the dam effects, the estimates of m2 were smaller than those of Model 3 for all traits, and there was no significant improvement in likelihood compared to Model 3. Allowing a direct-maternal genetic covariance (σ am ) in Model 6 yielded estimates of ram ranging from −0.03 to 0.24 for these traits, and the resulting likelihood gave no significant improvement compared to Model 5. The model with maternal genetic effects (Model 3) was the preferred model for BL, HW and HG of lambs at birth (Table 1). Our estimates for direct heritability for BL (0.14), HW (0.14) and HG (0.07) at birth were generally low to medium in magnitude. The present estimates for body length for the direct genetic effect was similar to the estimates obtained for body length in cattle (Van MarleK¨oster et al., 2000; Maiwashe et al., 2002). The low to moderate estimates for body measurements at birth can be explained by generally poor nutritional level of ewes creating a large environmental variation. The maternal heritability estimates for BL (0.13), HW (0.15) and HG (0.13) at birth suggesting that maternal genetic effects are important for these traits. The relatively low productivity and poor nutritional environment, resulting in lambs not expressing their genetic potential, could partially explain moderate maternal effects in Muzaffarnagari sheep. At younger ages, insufficient milk production by the ewes may also give rise to variation in environmental effects and result in low heritabilities. The medium estimates of total heritability (h2t ) and maternal repeatability (tm ) for body length (h2t = 0.14, tm = 0.17), height at withers (h2t = 0.21, tm = 0.19) and heart girth (h2t = 0.13, tm = 0.15) indicates that a response to phenotypic selection can be expected for these traits. 3.2.2. At weaning Estimates of heritability was biased upward, when maternal effects were ignored (Model 1), while inclusion
of permanent environmental maternal effects (Model 2) led to a reduction in additive heritability of 27–32% for body measurements at weaning compared to Model 1; this effect accounted for 8–9% of total phenotypic variance for these traits. Incorporating maternal additive effects in the model (Model 3) produced an estimate of m2 that explained 6–8% of phenotypic variance for all body measurements. Model 5 attempted independent estimation of estimation of m2 and c2 , but the estimate of m2 converged to 0.00, 0.02 and 0.04 for BL, HW and HG indicating very low additive maternal variance for these traits at weaning in these data. Adding σ am in Model 6, the estimate of the direct heritability increased for all measurements and yielded direct heritability estimates ranging from 0.15 to 0.18 along with the negative estimates of ram , which ranged from −0.21 to −0.59 but there was no significant improvement in likelihood compared to Model 5. Therefore, the most appropriate model (Table 1) for BL, HW and HG of lambs at weaning included only permanent environmental effects due to dam (Model 2). The direct heritability estimates for BL (0.12), HW (0.16) and HG (0.15) of lambs at weaning was much lower than the estimates of Janssens and Vandepitte (2004) who reported the direct heritability for BL, HW and HG of adult sheep in Blue du Maine, Suffolk and Texel sheep breed. Horstick (2001) found heritability estimates for BL (0.72), HW (0.70) and HG (0.56) in East Friesian and Black-Brown milksheep, which were also higher than our estimates. The estimates of permanent environmental effect due to dam for BL, HW and HG ranged from 8% to 9% of the total phenotypic variance indicating the importance of maternal effects for these traits from birth to weaning. The medium estimates of total heritability (h2t ) and maternal repeatability (tm ) for BL (h2t = 0.12, tm = 0.11), HW (h2t = 0.16, tm = 0.12) and HG (h2t = 0.15, tm = 0.13) of lambs at weaning suggest the existence of scope of genetic improvement for these traits. 4. Conclusion The modest estimate of heritability for body measurements at birth and weaning in this study indicates that the genetic progress for these traits is possible by selection. The maternal genetic effect was only important for body measurements at birth but permanent environmental maternal effect had some influence on body measurements at weaning. Therefore, both direct additive effects and maternal effects need to be considered for improving these traits by selection.
A. Mandal et al. / Small Ruminant Research 75 (2008) 123–127
Acknowledgements The authors thank Dr. Karin Meyer, Principal Biometrician, University of New England, Armidale, New South Wales, Australia, for granting permission to use the DFREML programme and Prof. D.R. Notter, Department of Animal and Poultry Sciences, Virginia Polytechnic Institute and State University, Blacksburg, USA, for his critical comments and suggestions for preparing this manuscript. We wish to acknowledge the contribution of the all former incharges, associated with the Muzaffanagari sheep project, for management and recording of data. The support extended by Director, CIRG, in providing facilities to carry out this study is also gratefully acknowledged. References Atta, M., El khidir, O.A., 2004. Use of heart girth, wither height and scapuloischial length for prediction of liveweight of Nilotic sheep. Small Rumin. Res. 55, 233–237. Aziz, M.A., Sharaby, M.A., 1993. Collinearity as a problem in predicting body weight from body dimensions of Najidi sheep in Saudi Arabia. Small Rumin. Res. 12, 117–124. Bhadula, S.K., Bhat, P.N., Garg, R.C., 1980. Note on the estimates of heritability and phenotypic correlations of weight and linear body measurements in Muzaffarnagari sheep. Ind. J. Anim. Sci. 50, 573–575. Harvey, W.R., 1990. User’s Guide for LSMLMW MIXMDL, PC-2 Version. Columbus, OH, USA. Horstick, A., 2001. Populationagenetische Untersuchung von Milchleistungs und Exterieurmerkmalen beim ostFriesischen und swarzbraunen Milchschaf. Ph.D. Thesis. Nannover, p. 254. Janssens, S., Vandepitte, W., 2004. Genetic parameters for body measurements and linear type traits in Belgian Blue du Maine, Suffolk and Texel sheep. Small Rumin. Res. 54, 13–24. Maiwashe, A.N., Bradfield, M.J., Theron, H.E., van Wyk, J.B., 2002. Genetic parameter estimates for body measurements and growth traits in South African Bonsmara cattle. Small Rumin. Res. 75, 293–300.
127
Mandal, A., Singh, L.B., Rout, P.K., 2000. The Muzaffarnagari Sheep, a Mutton Breed in India. Anim. Genet. Resour. Inf., vol. 28. Food and Agricultural Organization, Rome, pp. 19–25. Mandal, A., Pant, K.P., Nandy, D.K., Rout, P.K., Roy, R., 2003. Genetic analysis of growth traits in Muzaffanagari sheep. Trop. Anim. Health Prod. 35, 271–284. Mandal, A., Neser, F.W.C., Rout, P.K., Roy, R., Notter, D.R., 2006a. Estimation of direct and maternal co(variance) components for preweaning growth traits in Muzaffarnagari sheep. Livest. Sci. 99, 79–89. Mandal, A., Neser, F.W.C., Rout, P.K., Roy, R., Notter, D.R., 2006b. Genetic parameters for direct and maternal effects on body weights of Muzaffarnagari sheep. Anim. Sci. 82, 133–140. Mar´ıa, G.A., Boldman, K.G., Van Vleck, L.D., 1993. Estimates of variances due to direct and maternal effects for growth traits of Romanov sheep. J. Anim. Sci. 71, 845–849. Matika, O., van Wyk, J.B., Erasmus, G.J., Baker, R.L., 2003. Genetic parameter estimates in Sabi sheep. Livest. Prod. Sci. 79, 17–28. Meyer, K., 2000. DFREML programs to estimate variance components by restricted maximum likelihood using a derivative-free algorithm—user notes. ¨ 1994. Maternal genetic effects on lamb N¨asholm, A., Danell, O., weights. In: Proc. 5th World Congress Genet. Appl. Livest. Prod., Guelph, Canada, 18, pp. 163–166. ¨ 1996. Genetic relationships of lamb weight, N¨asholm, A., Danell, O., maternal ability, and mature ewe weight in Swedish finewool sheep. J. Anim. Sci. 74, 329–339. Singh, R.N., 1995. Status Paper, Network Project on Sheep Improvement. Central Sheep and Wool Research Institute, Avikanagar (Rajasthan), India, p. 85. Snyman, M.A., Erasmus, G.J., van Wyk, J.B., Olivier, J.J., 1995. Direct and maternal (co)variance components and heritability estimates for body weight at different ages and fleece traits in Afrino sheep. Livest. Prod. Sci. 44, 229–235. Valdez, C.A., Tupas, D.G.A., Matias, J.B., 1997. Determination of body weight in sheep using external body measurements. Philipp. J. Vet. Med. 34, 24–31. Van Marle-K¨oster, E., Mostert, B.E., Van der Westhuizen, J., 2000. Body measurements as selection criteria for growth in South African Hereford cattle. Arch. Tierzuecht. 43, 5. Varade, P.K., Ali, S.Z., 1999. Body measurements of sheep in field conditions. Ind. J. Anim. Sci. 5, 59–61. Willham, R.L., 1980. Problems in estimating maternal effects. Livest. Prod. Sci. 7, 405–418.
Small Ruminant Research 75 (2008) 123–127
Direct and maternal effects for body measurements at birth and weaning in Muzaffarnagari sheep of India Ajoy Mandal ∗ , R. Roy, P.K. Rout Genetics and Breeding Division, Central Institute for Research on Goats, Makhdoom, Mathura 281122, Uttar Pradesh, India Received 13 July 2007; received in revised form 17 August 2007; accepted 27 August 2007 Available online 24 October 2007
Abstract Estimates of co(variance) components were obtained for body measurements at birth and weaning in Muzaffarnagari sheep maintained at the Central Institute for Research on Goats, Makhdoom, Mathura, India, over a period of 27 years (1978 through 2004). Records of 5821 lambs descended from 176 rams and 1698 ewes were used in the study. Analyses were carried out by REML fitting an animal model and ignoring or including maternal genetic or permanent environmental effects. Six different animal models were fitted for all traits. The best model was chosen after testing the improvement of the log-likelihood values. Direct heritability estimates were inflated substantially for all traits when maternal effects were ignored. The direct heritability estimates for body length, height at withers and heart girth of lambs at birth were 0.14, 0.14 and 0.07, respectively. The corresponding maternal heritability estimates for these traits at birth were 0.13, 0.15 and 0.13, respectively. Moderate estimates of the direct heritability (h2 ) and the fraction of variance due to maternal permanent environmental effects (c2 ) for body length (h2 = 0.12, c2 = 0.08), height at withers (h2 = 0.16, c2 = 0.08) and heart girth (h2 = 0.15, c2 = 0.09) was observed at weaning. Results suggest that maternal additive effects were only important at birth for these traits but permanent environmental maternal effect had some influence on body measurements at weaning. These results indicate that modest rates of genetic progress appear possible for body measurements at birth and weaning. © 2007 Elsevier B.V. All rights reserved. Keywords: Genetic parameters; Maternal effects; Body measurements; Muzaffarnagari sheep
1. Introduction Body conformation and growth rate of animals are important criteria for selection of breeding animals in meat producing species. In many livestock species, conformation traits have been included in the genetic evaluation procedures and selection programmes can incorporate estimated breeding values for these traits. Body dimensions or linear measurements have been used
∗
Corresponding author. Tel.: +91 565 2763280x555; fax: +91 565 2763246. E-mail addresses: [email protected], [email protected] (A. Mandal). 0921-4488/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.smallrumres.2007.08.004
as indicators of body size and weight, and live body weight can be predicted from body measurements (Aziz and Sharaby, 1993; Valdez et al., 1997; Varade and Ali, 1999; Atta and El khidir, 2004). The phenotype of an animal for size and conformation is the result of the genetic potential and the influence of environment as well as maternal effects. Measurements of body dimensions of animals may be taken at a relatively early age; therefore the influence of maternal effects on these traits needs to be quantified in order to formulate optimal breeding programmes. Willham (1980) postulated that knowledge of the magnitude of the (co)variance components for traits of economic importance is critical for the genetic evaluation of animals and the development of sound breeding programs. Studies on traits measured
124
A. Mandal et al. / Small Ruminant Research 75 (2008) 123–127
at an early age in farm animals have shown that both direct and maternal influences are important for animal growth (Mar´ıa et al., 1993; Snyman et al., 1995; N¨asholm and Danell, 1996; Matika et al., 2003; Mandal et al., 2006a,b). Further, it was observed that exclusion of genetic or environmental factors may lead to overestimation of the (co)variance components fitted in the model. Thus, N¨asholm and Danell (1994) reported that when maternal genetic effects are of importance, but not accounted for, heritability estimates are biased upward and the realized efficiency of selection is reduced. Many of the published heritability estimates for body measurements of Indian sheep breeds were derived from sire models and based on small sets of data (Bhadula et al., 1980) that did not take account of additive maternal effects. However, no attempt has yet been made to apply REML procedures for estimation of genetic parameters for body dimensions of sheep. Therefore, the present study was conducted to estimate variance and covariance components due to direct genetic effects, maternal genetic effects and maternal permanent environmental effects for body measurements in Muzaffarnagari sheep.
seasons in May–June and October–November, with lambing in October–November and March–April. At birth each lamb was identified and date of birth, sex, type of birth and weight were recorded. Lambs were normally weaned at 3 months of age. A more detailed description of the breed and of its natural habitat, husbandry practices and production performances were provided by Mandal et al. (2000, 2003). 2.2. Data Data on birth (5821) and weaning (5060) records of lambs produced by 176 sires and 1698 and 1589 dams, respectively, and spread over a period of 27 years (1978–2004) were analyzed. The traits studied were body length (BL), height at withers (HW) and heart girths (HG) of animals, measured at birth and at weaning. BL was measured by taking the average of left and right side measurements (using a tape) of the distance between the point of the shoulder (lateral tuberosity of the humerus) and the pinbone (tuber ishii). HW was measured on the dorsal midline at the highest point on the withers by using a tape. HG was measured with tape by taking body circumference immediately posterior to the front leg. All body dimensions were taken when the animal was standing naturally with head raised and body weight on all four feet. 2.3. Statistical analyses
2. Materials and methods 2.1. Study area and flock management This study was carried out at the Central Institute for Research on Goats, Makhdoom, Mathura, located in India, 169 m above mean sea level and 10◦ N, and 78◦ 02 E with a semi-arid climate. The annual rainfall in the area averages 750 mm, while the average daily temperature varies from 6 ◦ C (winter) to 44.5 ◦ C (summer). The Muzaffarnagari sheep, an important mutton breed of India, is known for its relatively rapid growth rate (Singh, 1995), high feed conversion efficiency (Mandal et al., 2000) and very good adaptability (Mandal et al., 2003) in the semi-arid region of the country. The institute currently maintains a flock of 250 breeding ewes. The flock was generally reared under semi-intensive feeding. All animals grazed during the day (6–7 h) on natural pasture with supplementation depending upon the status and age category of the animals and were penned at night. The mean body weights of animals were 3.49 ± 0.04, 15.00 ± 0.20, 26.50 ± 3.56, 30.4 ± 0.40 and 33.4 ± 0.46 kg at birth, 3, 6, 9 and 12 months of age, respectively (Mandal et al., 2000). The ADG pre-weaning (0–3 months) was 155 ± 2.2 g and postweaning (3–6 months) 101.0 ± 2.9 g per day. The average feed conversion efficiency of this breed is reported to vary from 10% to 16%. Generally, controlled mating was practiced. Ewes in heat were mated with pre-assigned sires in the morning. Sheep were bred twice at each oestrus. Ewes were first exposed to rams at about 12 month of age. One breeding ram was allowed to mate with 20–25 ewes. There were two breeding
(Co)variance components were estimated by restricted maximum likelihood (REML) using a derivative-free algorithm fitting an animal model. Data were first analyzed by least-squares analysis of variance (Harvey, 1990) to identify the fixed effects to be included in the model. The statistical model included the fixed effects of birth year, season of birth, parity of dam, sex and birth status (single vs. twin) of lambs. All these effects were significant (P < 0.05) for all traits. The models used to estimate genetic parameters included random effects and all fixed effects that were found significant in least-squares analysis. Convergence of the REML solutions was considered reached when the variance of function values (−2 log L) in the Simplex was less than 10−8 . To ensure that a global maximum was reached, analyses were restarted for several other rounds of iterations using results from the previous round as starting values. When estimates did not change, convergence was confirmed. Standard errors were calculated for the estimated parameters as a part of the DFREML programme (Meyer, 2000). Univariate animal models were fitted to estimate the genetic parameters for each trait. By ignoring or including various combinations of maternal genetic and permanent environmental effects, the following six different models were used: y = Xb + Za a + e
(1)
y = Xb + Za a + Zc c + e
(2)
y = Xb + Za a + Zm m + e
with Cov(a, m) = 0
(3)
y = Xb + Za a + Zm m + e
with Cov(a, m) = Aσam
(4)
A. Mandal et al. / Small Ruminant Research 75 (2008) 123–127
y = Xb + Za a + Zm m + Zc c + e
with Cov(a, m) = 0
y = Xb + Za a + Zm m + Zc c + e
with Cov(a, m) = Aσam
(5)
(6) where y is the vector of observations for each trait; b, a, m, c and e are vectors of fixed effects (birth year, season of birth, parity of dam, sex and birth status of lamb), direct additive genetic effects (animal), maternal additive genetic effects, permanent environmental effects of dam and the residual effects, respectively; X, Za , Zm , Zc are the corresponding incidence matrices relating these effects to y; A is the numerator relationship matrix between animals; and σ am is the covariance between additive direct and maternal genetic effects. The analysis used the standard assumptions and definitions, i.e., V (a) = Aσa2 , V (e) = Iσe2 ,
V (m) = Aσm2 ,
V (c) = Iσc2 ,
Cov (a, m) = Aσam
where I is an identity matrix and σa2 , σm2 , σc2 and σe2 are direct additive genetic variance, maternal additive genetic variance, maternal permanent environmental variance and environmental variance, respectively. Estimates of heritability (h2 ), maternal heritability (m2 ) and permanent maternal environmental (c2 ) effects were calculated as ratios of estimates of σa2 , σm2 and σc2 , respectively, to the phenotypic variance (σp2 ). The direct-maternal correlation (ram ) was computed as the ratio of the estimates of direct-maternal covariance (σ am ) to the product of the square roots of estimates of σa2 and σm2 . The total maternal effect, tm = (1/4)h2 + m2 + c2 + mram h was calculated to estimate repeatability of ewe performance. The heritability estimates for the total genetic component was calculated (Willham, 1980) as h2t = h2 + 0.5m2 + 1.5mram h, to estimate the expected response to phenotypic selection. The most appropriate model for each trait was selected based on the likelihood ratio tests (Meyer, 2000). An effect was considered to have a significant influence when its inclusion caused a significant increase in log-likelihood, com-
125
pared to a model in which it was ignored. Significance was tested at P < 0.05 by comparing differences in log-likelihoods (−2 log L) to values for a chi-square distribution with degrees of freedom equal to the difference in the number of (co)variance components fitted for the two models. If −2 log L values were not significantly different (P > 0.05), the model with fewest random terms was chosen.
3. Results and discussion 3.1. Characteristics of data The least-squares means along with the standard errors for BL, HW and HG of animals were 33.65 ± 0.08, 37.54 ± 0.07 and 34.61 ± 0.07 cm, respectively at birth and the corresponding figures were 56.93 ± 0.14, 58.02 ± 0.13 and 57.40 ± 0.14 cm, respectively, at weaning. In these data, 51.0% of the lambs were males and 49.0% were females. Single and twin-born lambs represented 91.0% and 9% of the data, respectively. Coefficients of variation for body measurements ranged from 7.84% at birth to 9.48% at weaning. 3.2. Body measurements Estimates of (co)variance components and genetic parameters for BL, HW and HG of lambs for the most appropriate model at birth and weaning are summarized in Table 1. 3.2.1. At birth Estimates of direct additive heritabilities for body measurements at birth depended on the model used. Ignoring maternal effects (Model 1) produced higher estimates of σa2 and h2 than other models for all traits. Fit-
Table 1 Estimated parameters and their standard errors for body measurements from the best model for each traita Traits
Model
h2
m2
c2
σa2
2 σm
σc2
At birth BL HW HG
3 3 3
0.14 (0.04)b 0.14 (0.04) 0.07 (0.01)
0.13 (0.02) 0.15 (0.02) 0.13 (0.02)
– – –
0.967 0.975 0.458
0.942 1.029 0.908
– – –
At weaning BL HW HG
2 2 2
0.12 (0.03) 0.16 (0.03) 0.15 (0.03)
– – –
0.08 (0.02) 0.08 (0.02) 0.09 (0.02)
2.427 3.081 3.385
– – –
1.731 1.527 1.951
σe2
σp2
5.090 5.009 5.427
6.999 7.013 6.793
16.717 14.385 17.381
20.876 18.992 22.717
a BL is the body length, HW is the height at withers, HG is the heart girth, σ 2 is the direct additive genetic variance; σ 2 is the maternal additive a m genetic variance; σc2 is the maternal permanent environmental variance; σe2 is the environmental variance; σp2 is the phenotypic variance; h2 is the heritability; m2 is the maternal heritability; c2 = σc2 /σp2 . b Figures in parentheses are standard errors of the estimate.
126
A. Mandal et al. / Small Ruminant Research 75 (2008) 123–127
ting a permanent environmental maternal effect (Model 2) explained 9%, 10% and 10% of total phenotypic variance for BL, HW and HG of lambs, respectively, and the corresponding reduction in direct heritability was 33%, 30% and 44% in comparison with Model 1. Model 3, which included only direct and maternal additive effects, yielded an maternal effects contributing 13%, 15% and 13% for BL, HW and HG, respectively and also resulted a substantial reduction in the estimates of h2 relative to Model 2. Fitting a non-zero covariance (σ am ) along with a maternal genetic effect (Model 4) resulted in similar estimates of h2 and m2 as observed in Model 3 with the corresponding estimates of ram of 0.24, −0.02 and 0.21 for BL, HW and HG, respectively. In Model 5, which attempted to disentangle genetic and environmental components of the dam effects, the estimates of m2 were smaller than those of Model 3 for all traits, and there was no significant improvement in likelihood compared to Model 3. Allowing a direct-maternal genetic covariance (σ am ) in Model 6 yielded estimates of ram ranging from −0.03 to 0.24 for these traits, and the resulting likelihood gave no significant improvement compared to Model 5. The model with maternal genetic effects (Model 3) was the preferred model for BL, HW and HG of lambs at birth (Table 1). Our estimates for direct heritability for BL (0.14), HW (0.14) and HG (0.07) at birth were generally low to medium in magnitude. The present estimates for body length for the direct genetic effect was similar to the estimates obtained for body length in cattle (Van MarleK¨oster et al., 2000; Maiwashe et al., 2002). The low to moderate estimates for body measurements at birth can be explained by generally poor nutritional level of ewes creating a large environmental variation. The maternal heritability estimates for BL (0.13), HW (0.15) and HG (0.13) at birth suggesting that maternal genetic effects are important for these traits. The relatively low productivity and poor nutritional environment, resulting in lambs not expressing their genetic potential, could partially explain moderate maternal effects in Muzaffarnagari sheep. At younger ages, insufficient milk production by the ewes may also give rise to variation in environmental effects and result in low heritabilities. The medium estimates of total heritability (h2t ) and maternal repeatability (tm ) for body length (h2t = 0.14, tm = 0.17), height at withers (h2t = 0.21, tm = 0.19) and heart girth (h2t = 0.13, tm = 0.15) indicates that a response to phenotypic selection can be expected for these traits. 3.2.2. At weaning Estimates of heritability was biased upward, when maternal effects were ignored (Model 1), while inclusion
of permanent environmental maternal effects (Model 2) led to a reduction in additive heritability of 27–32% for body measurements at weaning compared to Model 1; this effect accounted for 8–9% of total phenotypic variance for these traits. Incorporating maternal additive effects in the model (Model 3) produced an estimate of m2 that explained 6–8% of phenotypic variance for all body measurements. Model 5 attempted independent estimation of estimation of m2 and c2 , but the estimate of m2 converged to 0.00, 0.02 and 0.04 for BL, HW and HG indicating very low additive maternal variance for these traits at weaning in these data. Adding σ am in Model 6, the estimate of the direct heritability increased for all measurements and yielded direct heritability estimates ranging from 0.15 to 0.18 along with the negative estimates of ram , which ranged from −0.21 to −0.59 but there was no significant improvement in likelihood compared to Model 5. Therefore, the most appropriate model (Table 1) for BL, HW and HG of lambs at weaning included only permanent environmental effects due to dam (Model 2). The direct heritability estimates for BL (0.12), HW (0.16) and HG (0.15) of lambs at weaning was much lower than the estimates of Janssens and Vandepitte (2004) who reported the direct heritability for BL, HW and HG of adult sheep in Blue du Maine, Suffolk and Texel sheep breed. Horstick (2001) found heritability estimates for BL (0.72), HW (0.70) and HG (0.56) in East Friesian and Black-Brown milksheep, which were also higher than our estimates. The estimates of permanent environmental effect due to dam for BL, HW and HG ranged from 8% to 9% of the total phenotypic variance indicating the importance of maternal effects for these traits from birth to weaning. The medium estimates of total heritability (h2t ) and maternal repeatability (tm ) for BL (h2t = 0.12, tm = 0.11), HW (h2t = 0.16, tm = 0.12) and HG (h2t = 0.15, tm = 0.13) of lambs at weaning suggest the existence of scope of genetic improvement for these traits. 4. Conclusion The modest estimate of heritability for body measurements at birth and weaning in this study indicates that the genetic progress for these traits is possible by selection. The maternal genetic effect was only important for body measurements at birth but permanent environmental maternal effect had some influence on body measurements at weaning. Therefore, both direct additive effects and maternal effects need to be considered for improving these traits by selection.
A. Mandal et al. / Small Ruminant Research 75 (2008) 123–127
Acknowledgements The authors thank Dr. Karin Meyer, Principal Biometrician, University of New England, Armidale, New South Wales, Australia, for granting permission to use the DFREML programme and Prof. D.R. Notter, Department of Animal and Poultry Sciences, Virginia Polytechnic Institute and State University, Blacksburg, USA, for his critical comments and suggestions for preparing this manuscript. We wish to acknowledge the contribution of the all former incharges, associated with the Muzaffanagari sheep project, for management and recording of data. The support extended by Director, CIRG, in providing facilities to carry out this study is also gratefully acknowledged. References Atta, M., El khidir, O.A., 2004. Use of heart girth, wither height and scapuloischial length for prediction of liveweight of Nilotic sheep. Small Rumin. Res. 55, 233–237. Aziz, M.A., Sharaby, M.A., 1993. Collinearity as a problem in predicting body weight from body dimensions of Najidi sheep in Saudi Arabia. Small Rumin. Res. 12, 117–124. Bhadula, S.K., Bhat, P.N., Garg, R.C., 1980. Note on the estimates of heritability and phenotypic correlations of weight and linear body measurements in Muzaffarnagari sheep. Ind. J. Anim. Sci. 50, 573–575. Harvey, W.R., 1990. User’s Guide for LSMLMW MIXMDL, PC-2 Version. Columbus, OH, USA. Horstick, A., 2001. Populationagenetische Untersuchung von Milchleistungs und Exterieurmerkmalen beim ostFriesischen und swarzbraunen Milchschaf. Ph.D. Thesis. Nannover, p. 254. Janssens, S., Vandepitte, W., 2004. Genetic parameters for body measurements and linear type traits in Belgian Blue du Maine, Suffolk and Texel sheep. Small Rumin. Res. 54, 13–24. Maiwashe, A.N., Bradfield, M.J., Theron, H.E., van Wyk, J.B., 2002. Genetic parameter estimates for body measurements and growth traits in South African Bonsmara cattle. Small Rumin. Res. 75, 293–300.
127
Mandal, A., Singh, L.B., Rout, P.K., 2000. The Muzaffarnagari Sheep, a Mutton Breed in India. Anim. Genet. Resour. Inf., vol. 28. Food and Agricultural Organization, Rome, pp. 19–25. Mandal, A., Pant, K.P., Nandy, D.K., Rout, P.K., Roy, R., 2003. Genetic analysis of growth traits in Muzaffanagari sheep. Trop. Anim. Health Prod. 35, 271–284. Mandal, A., Neser, F.W.C., Rout, P.K., Roy, R., Notter, D.R., 2006a. Estimation of direct and maternal co(variance) components for preweaning growth traits in Muzaffarnagari sheep. Livest. Sci. 99, 79–89. Mandal, A., Neser, F.W.C., Rout, P.K., Roy, R., Notter, D.R., 2006b. Genetic parameters for direct and maternal effects on body weights of Muzaffarnagari sheep. Anim. Sci. 82, 133–140. Mar´ıa, G.A., Boldman, K.G., Van Vleck, L.D., 1993. Estimates of variances due to direct and maternal effects for growth traits of Romanov sheep. J. Anim. Sci. 71, 845–849. Matika, O., van Wyk, J.B., Erasmus, G.J., Baker, R.L., 2003. Genetic parameter estimates in Sabi sheep. Livest. Prod. Sci. 79, 17–28. Meyer, K., 2000. DFREML programs to estimate variance components by restricted maximum likelihood using a derivative-free algorithm—user notes. ¨ 1994. Maternal genetic effects on lamb N¨asholm, A., Danell, O., weights. In: Proc. 5th World Congress Genet. Appl. Livest. Prod., Guelph, Canada, 18, pp. 163–166. ¨ 1996. Genetic relationships of lamb weight, N¨asholm, A., Danell, O., maternal ability, and mature ewe weight in Swedish finewool sheep. J. Anim. Sci. 74, 329–339. Singh, R.N., 1995. Status Paper, Network Project on Sheep Improvement. Central Sheep and Wool Research Institute, Avikanagar (Rajasthan), India, p. 85. Snyman, M.A., Erasmus, G.J., van Wyk, J.B., Olivier, J.J., 1995. Direct and maternal (co)variance components and heritability estimates for body weight at different ages and fleece traits in Afrino sheep. Livest. Prod. Sci. 44, 229–235. Valdez, C.A., Tupas, D.G.A., Matias, J.B., 1997. Determination of body weight in sheep using external body measurements. Philipp. J. Vet. Med. 34, 24–31. Van Marle-K¨oster, E., Mostert, B.E., Van der Westhuizen, J., 2000. Body measurements as selection criteria for growth in South African Hereford cattle. Arch. Tierzuecht. 43, 5. Varade, P.K., Ali, S.Z., 1999. Body measurements of sheep in field conditions. Ind. J. Anim. Sci. 5, 59–61. Willham, R.L., 1980. Problems in estimating maternal effects. Livest. Prod. Sci. 7, 405–418.