- Email: [email protected]
Heritability of semen traits in German Warmblood stallions
G Model
ARTICLE IN PRESS
ANIREP-5393; No. of Pages 5
Animal Reproduction Science xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Animal Reproduction Science journal homepage: www.elsevier.com/locate/anireprosci
Heritability of semen traits in German Warmblood stallions M. Gottschalk a , H. Sieme b , G. Martinsson c , O. Distl a,∗ a b c
Institute for Animal Breeding and Genetics, University of Veterinary Medicine Hannover (Foundation), Hannover, Germany Unit of Reproductive Medicine—Clinic for Horses, University of Veterinary Medicine Hannover (Foundation), Hannover, Germany Lower Saxon National Stud Celle, Celle, Germany
a r t i c l e
i n f o
Article history: Received 8 December 2015 Received in revised form 7 March 2016 Accepted 13 March 2016 Available online xxx Keywords: Horse Semen quality Genetic parameters Estimated breeding value
a b s t r a c t The objectives of the present study were to evaluate genetic parameters for semen quality traits of 241 fertile German Warmblood stallions regularly employed in artificial insemination (AI). Stallions were owned by the National Studs Celle and Warendorf in Germany. Semen traits analyzed were gel-free volume, sperm concentration, total number of sperm, progressive motility and total number of progressively motile sperm. Semen protocols from a total of 63,972 ejaculates were collected between the years 2001 and 2014 for the present analysis. A multivariate linear animal model was employed for estimation of additive genetic and permanent environmental variances among stallions and breeding values (EBVs) for semen traits. Heritabilities estimated for all German Warmblood stallions were highest for gel-free volume (h2 = 0.28) and lowest for total number of progressively motile sperm (h2 = 0.13). The additive genetic correlation among gel-free volume and sperm concentration was highly negative (rg = −0.76). Average reliabilities of EBVs were at 0.37–0.68 for the 241 stallions with own records. The inter-stallion variance explained between 33 and 61% of the trait variance, underlining the major impact of the individual stallion on semen quality traits analyzed here. Recording of semen traits from stallions employed in AI may be recommended because EBVs achieve sufficient accuracies to improve semen quality in future generations. Due to favorable genetic correlations, sperm concentration, total number of sperm and total number of progressively motile sperm may be increased simultaneously. © 2016 Elsevier B.V. All rights reserved.
1. Introduction There is considerable variation in semen quality among stallions (Rousset et al., 1986; Parlevliet et al., 1994; Van Eldik et al., 2006; Labitzke et al., 2013, 2014). Probable reasons for this are consistent stallion specific and genetic effects (Pickett et al., 1976; Pattie and Dowsett, 1982; Rousset et al., 1986; Parlevliet et al., 1994; Van Eldik et al., 2006; Ducro et al., 2011; Labitzke et al., 2014). The inter-individual variation seems to be much larger
than the variation among breeds (Brito, 2007). Several studies confirmed the stallion accounting for a significant proportion of the variance in semen quality. Evaluating data of 30Hanoverian stallions, Labitzke et al. (2013) demonstrated the large degree of inter-stallion variability accounting for 49–85% of the variance in semen quality traits. In a French study including 80 stallions of the breeds Thoroughbred, French Trotter, Selle Francaise and Breton draught horse, Rousset et al. (1986) found that 37–69% of the variance was due to stallions. Pattie and Dowsett (1982) estimated repeatabilities at 26–48% for semen quality traits of 168 stallions. Gottschalk et al. (2015) showed that stal-
∗ Corresponding author at: Bünteweg 17p, 30559 Hannover, Germany. E-mail address: [email protected] (O. Distl). http://dx.doi.org/10.1016/j.anireprosci.2016.03.004 0378-4320/© 2016 Elsevier B.V. All rights reserved.
Please cite this article in press as: Gottschalk, M., et al., Heritability of semen traits in German Warmblood stallions. Anim. Reprod. Sci. (2016), http://dx.doi.org/10.1016/j.anireprosci.2016.03.004
G Model ANIREP-5393; No. of Pages 5
ARTICLE IN PRESS M. Gottschalk et al. / Animal Reproduction Science xxx (2016) xxx–xxx
2
Table 1 ¯ standard deviation (SD), minimum (Min) and maximum (Max) for raw data of semen traits from 241 German Warmblood stallions. Means (x), Trait
x¯
SD
Min
Max
Gel-free volume (ml) Sperm concentration (×106 /ml) Progressive motility (%) Total number of sperm (×109 ) Total number of progressively motile sperm (×109 )
37.9 213.1 60.8 7.3 4.5
18.1 88.9 9.8 3.0 2.0
2 1 1 0.6 0.3
290 695 95 36.0 21.9
lions accounted for 55–65% of the total variance in semen traits of 106Hanoverian Warmblood stallions. For an improvement of semen quality through selective breeding it is important to know whether the trait is heritable. Several studies suggested that an improvement in semen quality might be achieved by means of targeted selection. Genetic analyses of semen traits were performed for Dutch Warmblood horses (Parlevliet et al., 1994), Shetland Ponies (Van Eldik et al., 2006) and Friesian horses (Ducro et al., 2011) based on data from breeding soundness examinations prior to studbook registration. In contrast to these Dutch studies, semen data for genetic analyses in Hanoverian Warmblood horses were collected from stallions regularly employed in artificial insemination (AI) (Labitzke et al., 2014). Comparing semen quality among paternal half-sibs, Parlevliet et al. (1994) determined a significant sire effect for gel-free volume, progressive motility and sperm concentration in 66 maiden Dutch Warmblood stallions. A study on 285 Shetland Pony stallions revealed high heritabilities for ejaculate volume (h2 = 0.57) and progressive motility (h2 = 0.46) with the remaining semen traits showing low to moderate estimates (Van Eldik et al., 2006). Ducro et al. (2011) investigated data from 1146 Friesian stallions. Heritability estimates were moderate for gel-free volume (h2 = 0.16) and progressive motility (h2 = 0.27), while higher heritabilities were found for percentages of normal sperm cells (h2 = 0.52) and abnormal acrosomes (h2 = 0.60). In a study on 30Hanoverian stallions, Labitzke et al. (2014) identified high heritabilities for gel-free volume (h2 = 0.43) and moderate estimates for total number of sperm (h2 = 0.29) and progressive motility (h2 = 0.20). In Friesian stallions, sperm motility showed a positive genetic correlation to gel-free volume, sperm concentration and sperm morphology, indicating that selection on progressive motility may improve these parameters (Ducro et al., 2011). A highly positive genetic correlation was found among gel-free volume and total number of sperm in Hanoverian stallions, whereas gel-free volume and sperm concentration showed a highly negative genetic correlation (Labitzke et al., 2014). The objective of the current study was to estimate permanent environmental and genetic variances for semen quality traits of 241 fertile German Warmblood stallions routinely employed in AI. Permanent environmental and genetic variation could be distinguished as recording extended over 14 consecutive years and included 63,972 fresh semen reports. Estimated breeding values (EBVs) and their reliabilities should show whether breeders may have the option to improve semen quality in stallions through breeding measures.
2. Materials and methods 2.1. Stallions Semen traits were recorded between 2001 and 2014 for 241 stallions routinely used in AI on the Lower Saxon National Stud Celle and the North Rhine-Westphalian National Stud Warendorf. The present study only included stallions approved for AI and with a conception rate of at least 70% in >9 mares bred from a stallion (Klug, 2002). Therefore, all stallions under study were considered as fertile. Stallions were housed in boxes on straw without contact to mares and fed hay and oats three times daily at both national studs. Water was freely available. Stallions were 3–30 years old and were registered by five different German Warmblood horse breeding associations including Hanoverian, Holsteiner, Oldenburger, Rhinelander and Westphalian. Stallions from the different horse breeding organizations were connected through their pedigrees with common ancestors and exchange of stallions. 2.2. Evaluation of semen traits Semen samples were collected once daily on six consecutive days every week in the months February to August using a phantom and an artificial vagina (Hanover model). Prior to evaluation of semen traits, semen was passed through a sterile filter to remove the gel portion of the ejaculate. Afterwards, gel-free volume, sperm concentration, total number of sperm (TNS), progressive motility and total number of progressively motile sperm (TNM) were evaluated. The sperm concentration, presented as millions per ml (106 /ml), was assessed by means of photometry using a SpermaCue photometer (Minitube, Tiefenbach, Germany). For calculation of the TNS (x109 ), gel-free volume and sperm concentration were multiplied. After dilution to a concentration of 25 × 106/ml, progressive motility was determined by experienced observers through subjective visual estimation using a phase-contrast microscope with a stage heater at 200–400X magnification (Olympus CH-II, Olympus Optical, Hamburg, Germany). All semen parameters were documented immediately after fresh semen examination at the respective national stud. A total of 63,972 fresh semen reports were used for analyses. Means, standard deviations, minima and maxima of semen traits from 241 fertile German Warmblood stallions are presented in Table 1. 2.3. Statistical analyses Mixed linear models were employed to analyze the fixed effects of year, month and stud as well as the ran-
Please cite this article in press as: Gottschalk, M., et al., Heritability of semen traits in German Warmblood stallions. Anim. Reprod. Sci. (2016), http://dx.doi.org/10.1016/j.anireprosci.2016.03.004
G Model
ARTICLE IN PRESS
ANIREP-5393; No. of Pages 5
M. Gottschalk et al. / Animal Reproduction Science xxx (2016) xxx–xxx
3
Table 2 Heritabilities (on the diagonal in bold), additive genetic (above the diagonal) and residual correlations (below the diagonal) with their standard errors for semen traits in 241 fertile German Warmblood stallions. Trait
Vol
Vol Conc Mot TNS TNM
0.28 −0.37 −0.09 0.55 0.47
Conc ± ± ± ± ±
0.04 0.00 0.00 0.00 0.00
−0.76 0.21 0.02 0.40 0.37
Mot ± ± ± ± ±
0.07 0.03 0.00 0.00 0.00
−0.10 0.02 0.14 −0.06 0.34
TNS ± ± ± ± ±
0.06 0.08 0.02 0.00 0.00
0.43 0.25 −0.18 0.14 0.89
TNM ± ± ± ± ±
0.12 0.15 0.09 0.01 0.00
0.36 0.28 0.20 0.93 0.13
± ± ± ± ±
0.14 0.15 0.11 0.01 0.01
Vol: gel-free volume, Conc: sperm concentration (×106 /ml), Mot: progressive motility (%), TNS: total number of sperm (×109 ), TNM: total number of progressively motile sperm (×109 ).
dom effects of the stallion on semen traits. These analyses were performed using the MIXED procedure of SAS, version 9.4 (Statistical Analysis System, SAS Institute, Cary, NC). The age of the stallion at the time of semen collection was considered as covariate. The effects of age, year and month when semen collection took place were significant for all semen traits. The random permanent environmental effect among stallions was included to account for repeated records for each stallion. For estimation of genetic parameters and EBVs for German Warmblood stallions, an additive genetic effect of the animal was added to the model. Estimation was done with a multivariate linear animal model and restricted maximum likelihood (REML). At least 10 generations of known ancestors were included in the pedigree files. The multivariate linear animal model employed was as follows: Y ijklmnop = + Yeari + Monthj + b1 (Age)k + b2 (Age)2 k + b3 log(Age)k + Studl + Regm + stallionn + animalo + eijklmnop
with Yijklmnop = semen trait of the ijklmno-th ejaculate including gel-free volume, sperm concentration, progressive motility, total number of sperm and total number of progressively motile sperm; = model constant; Yeari = fixed effect of the year (i = 1–14); Monthj = fixed effect of the month (j = 1–5; 1 = January–March, 2 = April, 3 = May, 4 = June, 5 = July–August); Age = age of the stallion in months at semen collection; b1 , b2 , b3 = linear, quadratic and logarithmic regression coefficients; Studl = fixed effect of the national stud (l = 1–2; 1 = Celle, 2 = Warendorf); Regm = fixed effect of the horse breeding register (l = 1–5; 1 = Hanoverian, 2 = Westphalian, 3 = Holsteiner, 4 = Oldenburger, 5 = Rhinelander); stallionn = random permanent environmental effect of the stallion (n = 1–241); animalo = random additive genetic effect of the animal (o = 11,673) and eijklmnop = random residual effect. Variance and covariance parameters as well as their standard errors were estimated using the Variance Components Estimation (VCE 6) package, version 6.0.2 (Groeneveld et al., 2010). The variance of semen traits among stallions can be divided into a permanent environmental and an additive genetic proportion. The total variance among stallions corresponds to the repeatability of the phenotypic values. The permanent environmental variance among stallions shows the proportion of variance which is caused by environmental factors with consistent and long lasting effects on stallions. These effects may include management, housing and nutrition. The vari-
ance within stallions corresponds to the residual variance. All randomly acting factors increase the variance within stallions. The additive genetic variance can be estimated via groups of related stallions in the data set. The larger differences among groups of closely related stallions can be retrieved the higher the additive genetic variance will be estimated. Genetic correlations reflect the correlations among groups of related stallions. The genetic correlations have practical impact for breeding decisions. Positive genetic correlations among traits inform the breeder that these traits can be simultaneously improved as expected changes flow in the same direction among the stallions to be intended for breeding. However, negative genetic correlations indicate that these traits can be hardly improved at the same time as EBVs of most stallions show high values for one trait and low values for the EBVs of the other trait. EBVs were estimated with multivariate models using PEST (Prediction and Estimation), version 4.2 (Groeneveld et al., 1990) and the (co-)variance component estimates of the REML analysis. EBVs were standardized onto a mean of 100 and a standard deviation of 20 using the 241 fertile stallions with semen records as reference. EBVs >100 for semen traits indicate stallions transmitting higher abilities for the respective trait than the average of the 241 stallions. Reliabilities (r2 ) of EBVs were calculated as follows: r2 = 1-(PEV/2 a ) with PEV = predicted error variance and 2 a = additive genetic variance.
3. Results Heritabilities estimated for German Warmblood stallions were moderate with the highest estimates for gel-free volume (h2 = 0.28) and the lowest estimates (h2 = 0.13) for TNM (Table 2). Additive genetic correlations were highly negative among gel-free volume and sperm concentration (rg = −0.76) indicating strongly antagonistic trait values, particularly EBVs, for almost all stallions. Moderately negative genetic correlations among progressive motility and TNS (rg = −0.18) allow to find some stallions with high EBVs for progressive motility and TNS but on average stallions with higher EBVs for progressive motility show lower EBVs for TNS. Additive genetic correlations among gel-free volume and TNS as well as TNM were moderately positive. Among sperm concentration and TNS as well as TNM a moderately positive genetic correlation was obvious. These moderate positive genetic correlations will make it possible to find stallions with high positive EBVs for these trait combinations.
Please cite this article in press as: Gottschalk, M., et al., Heritability of semen traits in German Warmblood stallions. Anim. Reprod. Sci. (2016), http://dx.doi.org/10.1016/j.anireprosci.2016.03.004
G Model
ARTICLE IN PRESS
ANIREP-5393; No. of Pages 5
M. Gottschalk et al. / Animal Reproduction Science xxx (2016) xxx–xxx
4
Table 3 Proportion of the permanent environmental (PERM), additive genetic (ADD) and total variance among stallions (STALLION) for the phenotypic variance of semen traits (2 ) estimated in a multivariate animal model for 241 German Warmblood stallions. Trait1
PERM
ADD
STALLION
Trait variance (2 )
Vol Conc Mot TNS TNM
0.10 0.38 0.19 0.18 0.19
0.28 0.21 0.14 0.14 0.13
0.38 0.59 0.33 0.32 0.32
260.15 8697.00 92.36 7.91 3.38
1
See Table 2 for abbreviations.
Table 4 Minima (Min), maxima (Max) and range of estimated breeding values (EBVs) and their reliabilities (r2 ) for semen traits of 241 stallions with own records. Trait1
Min
Max
Range
r2 mean
r2 Min
r2 Max
Vol Conc Mot TNS TNM
56 51 26 43 35
189 173 156 177 165
133 122 130 134 130
0.68 0.37 0.41 0.43 0.41
0.54 0.29 0.29 0.31 0.30
0.76 0.53 0.57 0.58 0.56
1
See Table 2 for abbreviations.
The inter-stallion variance accounted for 32–59% of the total variance in semen traits (Table 3). Estimates for the additive genetic variance showed the highest values for gelfree volume (28%) and the lowest values for TNM (13%). The permanent environmental variance among Warmblood stallions was highest for sperm concentration (38%) and lowest for gel-free volume (10%). EBVs ranged from 26 to 189 on a scale of 100 ± 20 for the 241 stallions with own records (Table 4). Average reliabilities of EBVs were at 0.37–0.68 for stallions with own records with the highest values for gel-free volume and the lowest values for sperm concentration. 4. Discussion The objective of the current study was to estimate genetic parameters for semen quality in AI stallions of German Warmblood breeds. The present study allowed differentiation among permanent environmental and additive genetic stallion effects through collection of all fresh semen reports during the breeding season. All stallions considered in this study were fertile and their semen did not show abnormal morphology. Heritability estimates for semen quality traits of the investigated stallions were moderate and in the range reported by previous studies. The size of the total interstallion variance in the present study underlined the large impact of the stallion on semen quality traits in Warmblood horses which is in agreement with previous studies (Labitzke et al., 2014; Gottschalk et al., 2015). Heritability estimates for gel-free volume seem to be higher in Warmblood stallions compared to Friesian stallions (Ducro et al., 2011). In Shetland Pony stallions (Van Eldik et al., 2006) estimates were at h2 = 0.57 but with large standard errors. Sperm concentration in the present study showed rather consistent heritability estimates in comparison with studies in Friesian Horses and Shetland
Ponies. Across all studies, heritability estimates were in a range from 0.14 (Hanoverian, Labitzke et al., 2014) to 0.21 (present study and Friesian stallions, Ducro et al., 2011) and 0.24 in Shetland Pony stallions (Van Eldik et al., 2006), respectively. Heritability estimates for progressive motility were at h2 = 0.14 and thus, were smaller than estimates in Friesian (Ducro et al., 2011) and Shetland Pony (Van Eldik et al., 2006) stallions. Taking standard errors from both latter studies into account, the estimates of our study are still in the same range. For TNM, heritabilities in the present study were slightly smaller compared to previous reports in Friesian (Ducro et al., 2011) and Shetland Pony (Van Eldik et al., 2006) stallions. Computer assisted sperm analysis (CASA) provides a more objective analysis of sperm motility and therefore shows more repeatable estimates of sperm motility than subjective estimations (Ball et al., 2003; Love, 2011). A limitation of the present study may be the use of subjective evaluations of progressive sperm motility. Possibly, heritabilities and repeatabilities might have been increased for progressive motility and TNM when CASA data would have been available. A general difference in the studies on Dutch Warmblood horses (Parlevliet et al., 1994), Shetland Pony (Van Eldik et al., 2006) and Friesian (Ducro et al., 2011) stallions in the Netherlands stems from the procedure how data were collected. In The Netherlands, young stallions have to pass a breeding soundness examination prior to studbook registration including evaluation of semen traits. As this data from unselected stallions was employed in the studies by Parlevliet et al. (1994), Van Eldik et al. (2006) and Ducro et al. (2011), additive genetic variation may be larger than in stallions, like in the present study, selected for fulfilling minimum requirements of semen quality. Another issue may explain further differences among the Dutch studies and the present analysis. Semen quality of the first two ejaculates from stallions not yet regularly used in AI may show larger random variation as semen parameters tend to normalize only after daily semen collection for at least 5–7 days (Sullivan and Pickett, 1975; Thompson et al., 2004). This effect may influence variation among stallions and could have increased residual variation but decreased additive genetic variation in these previous studies. Larger datasets and multivariate animal models give estimates with lower standard errors and should be more representative for the actual population under analysis. Standard errors for heritability estimates of the present study were lower than in the previous studies on Shetland Ponies (Van Eldik et al., 2006), Friesian (Ducro et al., 2011) and Hanoverians (Labitzke et al., 2014). The smaller standard errors in our study are due to the large number of stallions with repeated records of semen traits. All heritability estimates for semen traits were three times larger than their corresponding standard errors and thus, can be considered as significant. Therefore, heritability estimates of the present study indicate that there is a large additive genetic variation in semen traits of stallions regularly used in AI. Comparisons with a previous study on 30Hanoverian stallions (Labitzke et al., 2014) makes visible that heritability estimates depend on the number of stallions included
Please cite this article in press as: Gottschalk, M., et al., Heritability of semen traits in German Warmblood stallions. Anim. Reprod. Sci. (2016), http://dx.doi.org/10.1016/j.anireprosci.2016.03.004
G Model ANIREP-5393; No. of Pages 5
ARTICLE IN PRESS M. Gottschalk et al. / Animal Reproduction Science xxx (2016) xxx–xxx
in the analysis. Therefore, population specific estimates should be based on all available stallions used in AI. In accordance with previous studies of Parlevliet et al. (1994), Van Eldik et al. (2006), Ducro et al. (2011) and Labitzke et al. (2014), the results of our study support the assumption that semen quality in Warmblood stallions can be improved through breeding measures. Ducro et al. (2011) suggested selection for TNM because this trait was highly genetically correlated with sperm concentration, progressive motility, normal sperm morphology and frequency of abnormal acrosomes. However, we could not confirm the high positive genetic correlation among TNM and progressive motility. We propose gel-free volume, sperm concentration, progressive motility and TNS as traits to be included in a selection index. Our study confirmed that selection on phenotypic records might be less effective than the use of EBVs (Ducro et al., 2011; Labitzke et al., 2014). In the present study the large dataset with repeated records of stallions markedly increased reliabilities of EBVs when comparing heritabilities and reliabilities of EBVs. Nevertheless, stallions are primarily selected on performance traits and minimum requirements for semen quality standards (Colenbrander et al., 2003). Introducing a systematic breeding program based on EBVs for semen quality should allow the option for selective breeding of sons in the next generation for fertility without compromising performance levels. 5. Conclusion Summarizing our results, we can conclude that semen quality traits show moderate heritabilities and publication of EBVs may inform breeders on semen quality of stallions. Along with EBVs, the composite estimate of the permanent environmental and additive genetic effect of the stallion may be of considerable importance for the management of stallions and their semen production capability for AI. Conflict of interests The authors declare that they have no competing interests.
5
References Ball, B.A., Gravance, C.G., Wessel, M.T., Sabeur, K., 2003. Activity of angiotensin converting enzyme (ACE) in reproductive tissues of the stallion and effects of angiotensin II on sperm motility. Theriogenology 59, 901–914. Brito, L.F., 2007. Evaluation of stallion sperm morphology. Clin. Tech. Equine Pract. 6, 249–264. Colenbrander, B., Gadella, B., Stout, T., 2003. The predictive value of semen analysis in the evaluation of stallion fertility. Reprod. Domest. Anim. 38, 305–311. Ducro, B., Bovenhuis, H., Stout, T., van Arendonk, J., 2011. Genetic parameters and founder inbreeding depression on semen quality in Friesian horses. In: Thesis. University of Wageningen, The Netherlands, pp. 97–112. Gottschalk, M., Sieme, H., Martinsson, G., Distl, O., 2015. Spermamerkmale bei Hannoveraner Deckhengsten. Pferdeheilkunde 31, 357–362. Groeneveld, E., Kovac, M., Wang, T., 1990. PEST, a general purpose BLUP package for multivariate prediction and estimation. Proceedings of the 4th World Congress on Genetics Applied to Livestock Production (XIII. Plenary lectures, molecular genetics and mapping, selection, prediction and estimation., pp. 488–491). Groeneveld, E., Kovaˇc, M., Mielenz, N., 2010. VCE 6. Users guide and reference manual, Version 6. Klug, E., 2002. Untersuchung von Stute und Hengst auf Zuchttauglichkeit. Pferdeheilkunde 18, 225–229. Labitzke, D., Sieme, H., Martinsson, G., Distl, O., 2013. Analyse fruchtbarkeitsassoziierter Spermaqualitätsparameter bei Hannoveraner Warmbluthengsten. Züchtungskunde 85, 354–366. Labitzke, D., Sieme, H., Martinsson, G., Distl, O., 2014. Genetic parameters and breeding values for semen characteristics in Hanoverian stallions. Reprod. Domest. Anim. 49, 584–587. Love, C., 2011. Relationship between sperm motility, morphology and the fertility of stallions. Theriogenology 76, 547–557. Parlevliet, J., Kemp, B., Colenbrander, B., 1994. Reproductive characteristics and semen quality in maiden Dutch Warmblood stallions. J. Reprod. Fertil., 183–187. Pattie, W., Dowsett, K., 1982. The repeatability of seminal characteristics of stallions. J. Reprod. Fertil., 9–13. Pickett, B., Faulkner, L., Seidel, G., Berndtson, W., Voss, J., 1976. Reproductive physiology of the stallion. IV. Seminal and behavioral characteristics. J. Anim. Sci. 43, 617–625. Rousset, H., Chanteloube, P., Magistrini, M., Palmer, E., 1986. Assessment of fertility and semen evaluations of stallions. J. Reprod. Fertil., 25–31. Sullivan, J., Pickett, B., 1975. Influence of ejaculation frequency of stallions on characteristics of semen and output of spermatozoa. J. Reprod. Fertil., 29–34. Thompson, J.A., Love, C.C., Stich, K.L., Brinsko, S.P., Blanchard, T.L., Varner, D.D., 2004. A Bayesian approach to prediction of stallion daily sperm output. Theriogenology 62, 1607–1617. Van Eldik, P., Van Der Waaij, E., Ducro, B., Kooper, A., Stout, T., Colenbrander, B., 2006. Possible negative effects of inbreeding on semen quality in Shetland pony stallions. Theriogenology 65, 1159–1170.
Acknowledgements The authors would like to thank the Lower Saxon National Stud Celle and the North Rhine-Westphalian National Stud Warendorf for supporting this study.
Please cite this article in press as: Gottschalk, M., et al., Heritability of semen traits in German Warmblood stallions. Anim. Reprod. Sci. (2016), http://dx.doi.org/10.1016/j.anireprosci.2016.03.004
ARTICLE IN PRESS
ANIREP-5393; No. of Pages 5
Animal Reproduction Science xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Animal Reproduction Science journal homepage: www.elsevier.com/locate/anireprosci
Heritability of semen traits in German Warmblood stallions M. Gottschalk a , H. Sieme b , G. Martinsson c , O. Distl a,∗ a b c
Institute for Animal Breeding and Genetics, University of Veterinary Medicine Hannover (Foundation), Hannover, Germany Unit of Reproductive Medicine—Clinic for Horses, University of Veterinary Medicine Hannover (Foundation), Hannover, Germany Lower Saxon National Stud Celle, Celle, Germany
a r t i c l e
i n f o
Article history: Received 8 December 2015 Received in revised form 7 March 2016 Accepted 13 March 2016 Available online xxx Keywords: Horse Semen quality Genetic parameters Estimated breeding value
a b s t r a c t The objectives of the present study were to evaluate genetic parameters for semen quality traits of 241 fertile German Warmblood stallions regularly employed in artificial insemination (AI). Stallions were owned by the National Studs Celle and Warendorf in Germany. Semen traits analyzed were gel-free volume, sperm concentration, total number of sperm, progressive motility and total number of progressively motile sperm. Semen protocols from a total of 63,972 ejaculates were collected between the years 2001 and 2014 for the present analysis. A multivariate linear animal model was employed for estimation of additive genetic and permanent environmental variances among stallions and breeding values (EBVs) for semen traits. Heritabilities estimated for all German Warmblood stallions were highest for gel-free volume (h2 = 0.28) and lowest for total number of progressively motile sperm (h2 = 0.13). The additive genetic correlation among gel-free volume and sperm concentration was highly negative (rg = −0.76). Average reliabilities of EBVs were at 0.37–0.68 for the 241 stallions with own records. The inter-stallion variance explained between 33 and 61% of the trait variance, underlining the major impact of the individual stallion on semen quality traits analyzed here. Recording of semen traits from stallions employed in AI may be recommended because EBVs achieve sufficient accuracies to improve semen quality in future generations. Due to favorable genetic correlations, sperm concentration, total number of sperm and total number of progressively motile sperm may be increased simultaneously. © 2016 Elsevier B.V. All rights reserved.
1. Introduction There is considerable variation in semen quality among stallions (Rousset et al., 1986; Parlevliet et al., 1994; Van Eldik et al., 2006; Labitzke et al., 2013, 2014). Probable reasons for this are consistent stallion specific and genetic effects (Pickett et al., 1976; Pattie and Dowsett, 1982; Rousset et al., 1986; Parlevliet et al., 1994; Van Eldik et al., 2006; Ducro et al., 2011; Labitzke et al., 2014). The inter-individual variation seems to be much larger
than the variation among breeds (Brito, 2007). Several studies confirmed the stallion accounting for a significant proportion of the variance in semen quality. Evaluating data of 30Hanoverian stallions, Labitzke et al. (2013) demonstrated the large degree of inter-stallion variability accounting for 49–85% of the variance in semen quality traits. In a French study including 80 stallions of the breeds Thoroughbred, French Trotter, Selle Francaise and Breton draught horse, Rousset et al. (1986) found that 37–69% of the variance was due to stallions. Pattie and Dowsett (1982) estimated repeatabilities at 26–48% for semen quality traits of 168 stallions. Gottschalk et al. (2015) showed that stal-
∗ Corresponding author at: Bünteweg 17p, 30559 Hannover, Germany. E-mail address: [email protected] (O. Distl). http://dx.doi.org/10.1016/j.anireprosci.2016.03.004 0378-4320/© 2016 Elsevier B.V. All rights reserved.
Please cite this article in press as: Gottschalk, M., et al., Heritability of semen traits in German Warmblood stallions. Anim. Reprod. Sci. (2016), http://dx.doi.org/10.1016/j.anireprosci.2016.03.004
G Model ANIREP-5393; No. of Pages 5
ARTICLE IN PRESS M. Gottschalk et al. / Animal Reproduction Science xxx (2016) xxx–xxx
2
Table 1 ¯ standard deviation (SD), minimum (Min) and maximum (Max) for raw data of semen traits from 241 German Warmblood stallions. Means (x), Trait
x¯
SD
Min
Max
Gel-free volume (ml) Sperm concentration (×106 /ml) Progressive motility (%) Total number of sperm (×109 ) Total number of progressively motile sperm (×109 )
37.9 213.1 60.8 7.3 4.5
18.1 88.9 9.8 3.0 2.0
2 1 1 0.6 0.3
290 695 95 36.0 21.9
lions accounted for 55–65% of the total variance in semen traits of 106Hanoverian Warmblood stallions. For an improvement of semen quality through selective breeding it is important to know whether the trait is heritable. Several studies suggested that an improvement in semen quality might be achieved by means of targeted selection. Genetic analyses of semen traits were performed for Dutch Warmblood horses (Parlevliet et al., 1994), Shetland Ponies (Van Eldik et al., 2006) and Friesian horses (Ducro et al., 2011) based on data from breeding soundness examinations prior to studbook registration. In contrast to these Dutch studies, semen data for genetic analyses in Hanoverian Warmblood horses were collected from stallions regularly employed in artificial insemination (AI) (Labitzke et al., 2014). Comparing semen quality among paternal half-sibs, Parlevliet et al. (1994) determined a significant sire effect for gel-free volume, progressive motility and sperm concentration in 66 maiden Dutch Warmblood stallions. A study on 285 Shetland Pony stallions revealed high heritabilities for ejaculate volume (h2 = 0.57) and progressive motility (h2 = 0.46) with the remaining semen traits showing low to moderate estimates (Van Eldik et al., 2006). Ducro et al. (2011) investigated data from 1146 Friesian stallions. Heritability estimates were moderate for gel-free volume (h2 = 0.16) and progressive motility (h2 = 0.27), while higher heritabilities were found for percentages of normal sperm cells (h2 = 0.52) and abnormal acrosomes (h2 = 0.60). In a study on 30Hanoverian stallions, Labitzke et al. (2014) identified high heritabilities for gel-free volume (h2 = 0.43) and moderate estimates for total number of sperm (h2 = 0.29) and progressive motility (h2 = 0.20). In Friesian stallions, sperm motility showed a positive genetic correlation to gel-free volume, sperm concentration and sperm morphology, indicating that selection on progressive motility may improve these parameters (Ducro et al., 2011). A highly positive genetic correlation was found among gel-free volume and total number of sperm in Hanoverian stallions, whereas gel-free volume and sperm concentration showed a highly negative genetic correlation (Labitzke et al., 2014). The objective of the current study was to estimate permanent environmental and genetic variances for semen quality traits of 241 fertile German Warmblood stallions routinely employed in AI. Permanent environmental and genetic variation could be distinguished as recording extended over 14 consecutive years and included 63,972 fresh semen reports. Estimated breeding values (EBVs) and their reliabilities should show whether breeders may have the option to improve semen quality in stallions through breeding measures.
2. Materials and methods 2.1. Stallions Semen traits were recorded between 2001 and 2014 for 241 stallions routinely used in AI on the Lower Saxon National Stud Celle and the North Rhine-Westphalian National Stud Warendorf. The present study only included stallions approved for AI and with a conception rate of at least 70% in >9 mares bred from a stallion (Klug, 2002). Therefore, all stallions under study were considered as fertile. Stallions were housed in boxes on straw without contact to mares and fed hay and oats three times daily at both national studs. Water was freely available. Stallions were 3–30 years old and were registered by five different German Warmblood horse breeding associations including Hanoverian, Holsteiner, Oldenburger, Rhinelander and Westphalian. Stallions from the different horse breeding organizations were connected through their pedigrees with common ancestors and exchange of stallions. 2.2. Evaluation of semen traits Semen samples were collected once daily on six consecutive days every week in the months February to August using a phantom and an artificial vagina (Hanover model). Prior to evaluation of semen traits, semen was passed through a sterile filter to remove the gel portion of the ejaculate. Afterwards, gel-free volume, sperm concentration, total number of sperm (TNS), progressive motility and total number of progressively motile sperm (TNM) were evaluated. The sperm concentration, presented as millions per ml (106 /ml), was assessed by means of photometry using a SpermaCue photometer (Minitube, Tiefenbach, Germany). For calculation of the TNS (x109 ), gel-free volume and sperm concentration were multiplied. After dilution to a concentration of 25 × 106/ml, progressive motility was determined by experienced observers through subjective visual estimation using a phase-contrast microscope with a stage heater at 200–400X magnification (Olympus CH-II, Olympus Optical, Hamburg, Germany). All semen parameters were documented immediately after fresh semen examination at the respective national stud. A total of 63,972 fresh semen reports were used for analyses. Means, standard deviations, minima and maxima of semen traits from 241 fertile German Warmblood stallions are presented in Table 1. 2.3. Statistical analyses Mixed linear models were employed to analyze the fixed effects of year, month and stud as well as the ran-
Please cite this article in press as: Gottschalk, M., et al., Heritability of semen traits in German Warmblood stallions. Anim. Reprod. Sci. (2016), http://dx.doi.org/10.1016/j.anireprosci.2016.03.004
G Model
ARTICLE IN PRESS
ANIREP-5393; No. of Pages 5
M. Gottschalk et al. / Animal Reproduction Science xxx (2016) xxx–xxx
3
Table 2 Heritabilities (on the diagonal in bold), additive genetic (above the diagonal) and residual correlations (below the diagonal) with their standard errors for semen traits in 241 fertile German Warmblood stallions. Trait
Vol
Vol Conc Mot TNS TNM
0.28 −0.37 −0.09 0.55 0.47
Conc ± ± ± ± ±
0.04 0.00 0.00 0.00 0.00
−0.76 0.21 0.02 0.40 0.37
Mot ± ± ± ± ±
0.07 0.03 0.00 0.00 0.00
−0.10 0.02 0.14 −0.06 0.34
TNS ± ± ± ± ±
0.06 0.08 0.02 0.00 0.00
0.43 0.25 −0.18 0.14 0.89
TNM ± ± ± ± ±
0.12 0.15 0.09 0.01 0.00
0.36 0.28 0.20 0.93 0.13
± ± ± ± ±
0.14 0.15 0.11 0.01 0.01
Vol: gel-free volume, Conc: sperm concentration (×106 /ml), Mot: progressive motility (%), TNS: total number of sperm (×109 ), TNM: total number of progressively motile sperm (×109 ).
dom effects of the stallion on semen traits. These analyses were performed using the MIXED procedure of SAS, version 9.4 (Statistical Analysis System, SAS Institute, Cary, NC). The age of the stallion at the time of semen collection was considered as covariate. The effects of age, year and month when semen collection took place were significant for all semen traits. The random permanent environmental effect among stallions was included to account for repeated records for each stallion. For estimation of genetic parameters and EBVs for German Warmblood stallions, an additive genetic effect of the animal was added to the model. Estimation was done with a multivariate linear animal model and restricted maximum likelihood (REML). At least 10 generations of known ancestors were included in the pedigree files. The multivariate linear animal model employed was as follows: Y ijklmnop = + Yeari + Monthj + b1 (Age)k + b2 (Age)2 k + b3 log(Age)k + Studl + Regm + stallionn + animalo + eijklmnop
with Yijklmnop = semen trait of the ijklmno-th ejaculate including gel-free volume, sperm concentration, progressive motility, total number of sperm and total number of progressively motile sperm; = model constant; Yeari = fixed effect of the year (i = 1–14); Monthj = fixed effect of the month (j = 1–5; 1 = January–March, 2 = April, 3 = May, 4 = June, 5 = July–August); Age = age of the stallion in months at semen collection; b1 , b2 , b3 = linear, quadratic and logarithmic regression coefficients; Studl = fixed effect of the national stud (l = 1–2; 1 = Celle, 2 = Warendorf); Regm = fixed effect of the horse breeding register (l = 1–5; 1 = Hanoverian, 2 = Westphalian, 3 = Holsteiner, 4 = Oldenburger, 5 = Rhinelander); stallionn = random permanent environmental effect of the stallion (n = 1–241); animalo = random additive genetic effect of the animal (o = 11,673) and eijklmnop = random residual effect. Variance and covariance parameters as well as their standard errors were estimated using the Variance Components Estimation (VCE 6) package, version 6.0.2 (Groeneveld et al., 2010). The variance of semen traits among stallions can be divided into a permanent environmental and an additive genetic proportion. The total variance among stallions corresponds to the repeatability of the phenotypic values. The permanent environmental variance among stallions shows the proportion of variance which is caused by environmental factors with consistent and long lasting effects on stallions. These effects may include management, housing and nutrition. The vari-
ance within stallions corresponds to the residual variance. All randomly acting factors increase the variance within stallions. The additive genetic variance can be estimated via groups of related stallions in the data set. The larger differences among groups of closely related stallions can be retrieved the higher the additive genetic variance will be estimated. Genetic correlations reflect the correlations among groups of related stallions. The genetic correlations have practical impact for breeding decisions. Positive genetic correlations among traits inform the breeder that these traits can be simultaneously improved as expected changes flow in the same direction among the stallions to be intended for breeding. However, negative genetic correlations indicate that these traits can be hardly improved at the same time as EBVs of most stallions show high values for one trait and low values for the EBVs of the other trait. EBVs were estimated with multivariate models using PEST (Prediction and Estimation), version 4.2 (Groeneveld et al., 1990) and the (co-)variance component estimates of the REML analysis. EBVs were standardized onto a mean of 100 and a standard deviation of 20 using the 241 fertile stallions with semen records as reference. EBVs >100 for semen traits indicate stallions transmitting higher abilities for the respective trait than the average of the 241 stallions. Reliabilities (r2 ) of EBVs were calculated as follows: r2 = 1-(PEV/2 a ) with PEV = predicted error variance and 2 a = additive genetic variance.
3. Results Heritabilities estimated for German Warmblood stallions were moderate with the highest estimates for gel-free volume (h2 = 0.28) and the lowest estimates (h2 = 0.13) for TNM (Table 2). Additive genetic correlations were highly negative among gel-free volume and sperm concentration (rg = −0.76) indicating strongly antagonistic trait values, particularly EBVs, for almost all stallions. Moderately negative genetic correlations among progressive motility and TNS (rg = −0.18) allow to find some stallions with high EBVs for progressive motility and TNS but on average stallions with higher EBVs for progressive motility show lower EBVs for TNS. Additive genetic correlations among gel-free volume and TNS as well as TNM were moderately positive. Among sperm concentration and TNS as well as TNM a moderately positive genetic correlation was obvious. These moderate positive genetic correlations will make it possible to find stallions with high positive EBVs for these trait combinations.
Please cite this article in press as: Gottschalk, M., et al., Heritability of semen traits in German Warmblood stallions. Anim. Reprod. Sci. (2016), http://dx.doi.org/10.1016/j.anireprosci.2016.03.004
G Model
ARTICLE IN PRESS
ANIREP-5393; No. of Pages 5
M. Gottschalk et al. / Animal Reproduction Science xxx (2016) xxx–xxx
4
Table 3 Proportion of the permanent environmental (PERM), additive genetic (ADD) and total variance among stallions (STALLION) for the phenotypic variance of semen traits (2 ) estimated in a multivariate animal model for 241 German Warmblood stallions. Trait1
PERM
ADD
STALLION
Trait variance (2 )
Vol Conc Mot TNS TNM
0.10 0.38 0.19 0.18 0.19
0.28 0.21 0.14 0.14 0.13
0.38 0.59 0.33 0.32 0.32
260.15 8697.00 92.36 7.91 3.38
1
See Table 2 for abbreviations.
Table 4 Minima (Min), maxima (Max) and range of estimated breeding values (EBVs) and their reliabilities (r2 ) for semen traits of 241 stallions with own records. Trait1
Min
Max
Range
r2 mean
r2 Min
r2 Max
Vol Conc Mot TNS TNM
56 51 26 43 35
189 173 156 177 165
133 122 130 134 130
0.68 0.37 0.41 0.43 0.41
0.54 0.29 0.29 0.31 0.30
0.76 0.53 0.57 0.58 0.56
1
See Table 2 for abbreviations.
The inter-stallion variance accounted for 32–59% of the total variance in semen traits (Table 3). Estimates for the additive genetic variance showed the highest values for gelfree volume (28%) and the lowest values for TNM (13%). The permanent environmental variance among Warmblood stallions was highest for sperm concentration (38%) and lowest for gel-free volume (10%). EBVs ranged from 26 to 189 on a scale of 100 ± 20 for the 241 stallions with own records (Table 4). Average reliabilities of EBVs were at 0.37–0.68 for stallions with own records with the highest values for gel-free volume and the lowest values for sperm concentration. 4. Discussion The objective of the current study was to estimate genetic parameters for semen quality in AI stallions of German Warmblood breeds. The present study allowed differentiation among permanent environmental and additive genetic stallion effects through collection of all fresh semen reports during the breeding season. All stallions considered in this study were fertile and their semen did not show abnormal morphology. Heritability estimates for semen quality traits of the investigated stallions were moderate and in the range reported by previous studies. The size of the total interstallion variance in the present study underlined the large impact of the stallion on semen quality traits in Warmblood horses which is in agreement with previous studies (Labitzke et al., 2014; Gottschalk et al., 2015). Heritability estimates for gel-free volume seem to be higher in Warmblood stallions compared to Friesian stallions (Ducro et al., 2011). In Shetland Pony stallions (Van Eldik et al., 2006) estimates were at h2 = 0.57 but with large standard errors. Sperm concentration in the present study showed rather consistent heritability estimates in comparison with studies in Friesian Horses and Shetland
Ponies. Across all studies, heritability estimates were in a range from 0.14 (Hanoverian, Labitzke et al., 2014) to 0.21 (present study and Friesian stallions, Ducro et al., 2011) and 0.24 in Shetland Pony stallions (Van Eldik et al., 2006), respectively. Heritability estimates for progressive motility were at h2 = 0.14 and thus, were smaller than estimates in Friesian (Ducro et al., 2011) and Shetland Pony (Van Eldik et al., 2006) stallions. Taking standard errors from both latter studies into account, the estimates of our study are still in the same range. For TNM, heritabilities in the present study were slightly smaller compared to previous reports in Friesian (Ducro et al., 2011) and Shetland Pony (Van Eldik et al., 2006) stallions. Computer assisted sperm analysis (CASA) provides a more objective analysis of sperm motility and therefore shows more repeatable estimates of sperm motility than subjective estimations (Ball et al., 2003; Love, 2011). A limitation of the present study may be the use of subjective evaluations of progressive sperm motility. Possibly, heritabilities and repeatabilities might have been increased for progressive motility and TNM when CASA data would have been available. A general difference in the studies on Dutch Warmblood horses (Parlevliet et al., 1994), Shetland Pony (Van Eldik et al., 2006) and Friesian (Ducro et al., 2011) stallions in the Netherlands stems from the procedure how data were collected. In The Netherlands, young stallions have to pass a breeding soundness examination prior to studbook registration including evaluation of semen traits. As this data from unselected stallions was employed in the studies by Parlevliet et al. (1994), Van Eldik et al. (2006) and Ducro et al. (2011), additive genetic variation may be larger than in stallions, like in the present study, selected for fulfilling minimum requirements of semen quality. Another issue may explain further differences among the Dutch studies and the present analysis. Semen quality of the first two ejaculates from stallions not yet regularly used in AI may show larger random variation as semen parameters tend to normalize only after daily semen collection for at least 5–7 days (Sullivan and Pickett, 1975; Thompson et al., 2004). This effect may influence variation among stallions and could have increased residual variation but decreased additive genetic variation in these previous studies. Larger datasets and multivariate animal models give estimates with lower standard errors and should be more representative for the actual population under analysis. Standard errors for heritability estimates of the present study were lower than in the previous studies on Shetland Ponies (Van Eldik et al., 2006), Friesian (Ducro et al., 2011) and Hanoverians (Labitzke et al., 2014). The smaller standard errors in our study are due to the large number of stallions with repeated records of semen traits. All heritability estimates for semen traits were three times larger than their corresponding standard errors and thus, can be considered as significant. Therefore, heritability estimates of the present study indicate that there is a large additive genetic variation in semen traits of stallions regularly used in AI. Comparisons with a previous study on 30Hanoverian stallions (Labitzke et al., 2014) makes visible that heritability estimates depend on the number of stallions included
Please cite this article in press as: Gottschalk, M., et al., Heritability of semen traits in German Warmblood stallions. Anim. Reprod. Sci. (2016), http://dx.doi.org/10.1016/j.anireprosci.2016.03.004
G Model ANIREP-5393; No. of Pages 5
ARTICLE IN PRESS M. Gottschalk et al. / Animal Reproduction Science xxx (2016) xxx–xxx
in the analysis. Therefore, population specific estimates should be based on all available stallions used in AI. In accordance with previous studies of Parlevliet et al. (1994), Van Eldik et al. (2006), Ducro et al. (2011) and Labitzke et al. (2014), the results of our study support the assumption that semen quality in Warmblood stallions can be improved through breeding measures. Ducro et al. (2011) suggested selection for TNM because this trait was highly genetically correlated with sperm concentration, progressive motility, normal sperm morphology and frequency of abnormal acrosomes. However, we could not confirm the high positive genetic correlation among TNM and progressive motility. We propose gel-free volume, sperm concentration, progressive motility and TNS as traits to be included in a selection index. Our study confirmed that selection on phenotypic records might be less effective than the use of EBVs (Ducro et al., 2011; Labitzke et al., 2014). In the present study the large dataset with repeated records of stallions markedly increased reliabilities of EBVs when comparing heritabilities and reliabilities of EBVs. Nevertheless, stallions are primarily selected on performance traits and minimum requirements for semen quality standards (Colenbrander et al., 2003). Introducing a systematic breeding program based on EBVs for semen quality should allow the option for selective breeding of sons in the next generation for fertility without compromising performance levels. 5. Conclusion Summarizing our results, we can conclude that semen quality traits show moderate heritabilities and publication of EBVs may inform breeders on semen quality of stallions. Along with EBVs, the composite estimate of the permanent environmental and additive genetic effect of the stallion may be of considerable importance for the management of stallions and their semen production capability for AI. Conflict of interests The authors declare that they have no competing interests.
5
References Ball, B.A., Gravance, C.G., Wessel, M.T., Sabeur, K., 2003. Activity of angiotensin converting enzyme (ACE) in reproductive tissues of the stallion and effects of angiotensin II on sperm motility. Theriogenology 59, 901–914. Brito, L.F., 2007. Evaluation of stallion sperm morphology. Clin. Tech. Equine Pract. 6, 249–264. Colenbrander, B., Gadella, B., Stout, T., 2003. The predictive value of semen analysis in the evaluation of stallion fertility. Reprod. Domest. Anim. 38, 305–311. Ducro, B., Bovenhuis, H., Stout, T., van Arendonk, J., 2011. Genetic parameters and founder inbreeding depression on semen quality in Friesian horses. In: Thesis. University of Wageningen, The Netherlands, pp. 97–112. Gottschalk, M., Sieme, H., Martinsson, G., Distl, O., 2015. Spermamerkmale bei Hannoveraner Deckhengsten. Pferdeheilkunde 31, 357–362. Groeneveld, E., Kovac, M., Wang, T., 1990. PEST, a general purpose BLUP package for multivariate prediction and estimation. Proceedings of the 4th World Congress on Genetics Applied to Livestock Production (XIII. Plenary lectures, molecular genetics and mapping, selection, prediction and estimation., pp. 488–491). Groeneveld, E., Kovaˇc, M., Mielenz, N., 2010. VCE 6. Users guide and reference manual, Version 6. Klug, E., 2002. Untersuchung von Stute und Hengst auf Zuchttauglichkeit. Pferdeheilkunde 18, 225–229. Labitzke, D., Sieme, H., Martinsson, G., Distl, O., 2013. Analyse fruchtbarkeitsassoziierter Spermaqualitätsparameter bei Hannoveraner Warmbluthengsten. Züchtungskunde 85, 354–366. Labitzke, D., Sieme, H., Martinsson, G., Distl, O., 2014. Genetic parameters and breeding values for semen characteristics in Hanoverian stallions. Reprod. Domest. Anim. 49, 584–587. Love, C., 2011. Relationship between sperm motility, morphology and the fertility of stallions. Theriogenology 76, 547–557. Parlevliet, J., Kemp, B., Colenbrander, B., 1994. Reproductive characteristics and semen quality in maiden Dutch Warmblood stallions. J. Reprod. Fertil., 183–187. Pattie, W., Dowsett, K., 1982. The repeatability of seminal characteristics of stallions. J. Reprod. Fertil., 9–13. Pickett, B., Faulkner, L., Seidel, G., Berndtson, W., Voss, J., 1976. Reproductive physiology of the stallion. IV. Seminal and behavioral characteristics. J. Anim. Sci. 43, 617–625. Rousset, H., Chanteloube, P., Magistrini, M., Palmer, E., 1986. Assessment of fertility and semen evaluations of stallions. J. Reprod. Fertil., 25–31. Sullivan, J., Pickett, B., 1975. Influence of ejaculation frequency of stallions on characteristics of semen and output of spermatozoa. J. Reprod. Fertil., 29–34. Thompson, J.A., Love, C.C., Stich, K.L., Brinsko, S.P., Blanchard, T.L., Varner, D.D., 2004. A Bayesian approach to prediction of stallion daily sperm output. Theriogenology 62, 1607–1617. Van Eldik, P., Van Der Waaij, E., Ducro, B., Kooper, A., Stout, T., Colenbrander, B., 2006. Possible negative effects of inbreeding on semen quality in Shetland pony stallions. Theriogenology 65, 1159–1170.
Acknowledgements The authors would like to thank the Lower Saxon National Stud Celle and the North Rhine-Westphalian National Stud Warendorf for supporting this study.
Please cite this article in press as: Gottschalk, M., et al., Heritability of semen traits in German Warmblood stallions. Anim. Reprod. Sci. (2016), http://dx.doi.org/10.1016/j.anireprosci.2016.03.004