The relationship between rabbit semen characteristics and reproductive performance after artificial insemination

Transcription

1 Animal Reproduction Science 70 (2002) The relationship between rabbit semen characteristics and reproductive performance after artificial insemination Jean-Michel Brun, Michéle Theau-Clément, Gérard Bolet INRA, Station d Amelioration,Genetique des Animaux, B.P. 27, Castanet-Tolosan Cédex, France Received 26 April 2001; received in revised form 12 November 2001; accepted 13 December 2001 Abstract The relationships between several rabbit buck semen traits, concerning either the ejaculate or the dose inseminated (volume, mass motility, ph, percentage of motile sperm (PMS), concentration, number of total and motile sperm per ejaculate and per insemination dose) and the reproductive performance of does was investigated in 839 inseminations involving 54 bucks and 111 does. Four genetic types were involved: INRA1601 strain (A), INRA2066 strain (B) and the two reciprocal crossbreds (AB and BA). The mating design was A A, B B, (AB or BA) (AB or BA). Semen was diluted (1:9) and a constant volume of 0.5 ml was inseminated 2 4 h after collection. Therefore, the total number of spermatozoa per dose was variable and proportional to the initial concentration. Mass motility significantly influenced the kindling rate. Taken separately, volume, PMS and concentration did not influence the kindling rate but their product, the number of motile sperm per ejaculate, did. Litter size (total born) was significantly influenced by concentration and all variables depending on it, particularly the number of total and motile sperm per dose. However, reproductive performances were predominantly influenced by the physiological status of the does at insemination (lactation stage and receptivity) Elsevier Science B.V. All rights reserved. Keywords: Rabbit; Sperm quality; Fertility; Litter size; Artificial insemination 1. Introduction The relationships between semen characteristics and fertility have been investigated in many domestic species, with the general objective of predicting fertility from semen traits (Cherms, 1968, in turkeys; Hirao, 1975, in cattle; Wilson et al., 1979, in broilers; Colas, 1984, in sheep; Woelders, 1991, in pig; Bencheikh, 1993, in rabbit). Although some authors Corresponding author. Tel.: ; fax: address: brun@germinal.toulouse.inra.fr (J.-M. Brun) /02/$ see front matter 2002 Elsevier Science B.V. All rights reserved. PII: S (01)00197-X

2 140 J.-M. Brun et al. / Animal Reproduction Science 70 (2002) used in vitro fertilization, Amann (1989) stressed the importance of the in vivo fertility test through artificial insemination (AI). The semen traits investigated covered a wide range of traits including e.g. biochemical constituents (enzymes, fructose, etc.), qualitative traits of the spermatozoa such as acrosome quality or their mobility and characteristics of the insemination dose. In rabbits, the use of AI along with the cycle production system (Facchin et al., 1988) is a common practice in meat production. Insemination is performed either with fresh semen collected and conditioned in the farm or with cooled semen provided by a semen production farm. In general, semen from different males is pooled before insemination. Is the fertilizing ability dependent on the qualitative and quantitative traits of the semen? In rabbit, few authors (More O Ferral and Meacham, 1968; Foote et al., 1991; Bencheikh, 1993, 1995) have investigated the relationships between qualitative and quantitative traits of semen and reproductive performances after AI. Any information about such a relationship will help in selecting semen in order to optimize insemination doses. The present study, using monospermic insemination and constant dilution, aims at contributing to that knowledge. 2. Materials and methods 2.1. Breeding system The experiment was performed between August 1996 and February 1997 at the experimental rabbitry at the INRA Centre of Toulouse. The data analysed involved 54 bucks and 111 does of four genetic types. They came from the monitoring of the formation of a synthetic strain (Brun et al., 1998, 1999), when the semen traits and reproductive performances of the two reciprocal F1 crossbred breeders were compared to those of the parental strains: the INRA1601 strain, descended from the Verde strain of the University of Valencia (Spain) and the INRA2066 strain. Both strains have been selected for litter size (Estany et al., 1989; Rochambeau, 1998). Here onwards, strains INRA1601 and INRA2066 will be called A and B, respectively. Offspring from the cross m A f B will be called AB, and those from the reciprocal cross BA. Purebred males only mated with females of their own type while crossbred males could mate with both types of crossbred dams. Reproduction was performed by AI with fresh semen. On the day of insemination, buck semen was collected twice, at a 15 min interval. Semen with urine, or of a volume below 0.3 ml, or with a mass motility value below 6 (Table 1) was discarded. The better ejaculate in terms of mass motility was used for insemination. Immediately after collection, the semen was diluted (1:9) in a commercial extender (Galap, IMV, France) at 32 C, packed in French rabbit straws of 0.5 ml (IMV, France) and stored at C. Prior to insemination, the sexual receptivity of the does was tested. A doe was considered receptive if she adopted a lordosis position in the presence of a vasectomized buck. The does were inseminated immediately after the test, with a constant volume of 0.5 ml of diluted semen, using a bent AI gun (IMV, France). At each insemination, females were allocated to males so as to balance the physiological status of females (lactating or not, receptive or not) between males. Neither hormone nor biostimulation were used to induce rabbit doe receptivity. Immediately after insemination, the does were given an intra muscular injection of 0.2 ml Receptal (Hoechst Roussel Vet,

3 J.-M. Brun et al. / Animal Reproduction Science 70 (2002) Table 1 Notation scale used to measure mass motility (according to Petitjean, 1965) Note Description 0 No spermatozoa or motionless spermatozoa (spz) 1 Few stirring spz without notable movement 2 Important number of stirring spz without any movement 3 Few spz moving slowly 4 Few motionless spz, few stirring spz without any movement, few mobile spz 5 Same as 4, but higher proportion of mobile spz. Rather high, but not homogeneous motility in the observed field 6 Nearly all spz moving. High and homogeneous motility 7 Same as 6, wavelike movements begin 8 Same as 7, with slow wave movements 9 Strong waves; whirlwind appearance Romainville, France) to induce ovulation. Two batches of does were inseminated every 21 days. Non-pregnant does (palpation on day 12 of pregnancy) were inseminated 21 days later Traits recorded Immediately after semen collection, volume, ph, mass motility and percentage of motile sperm (PMS) were estimated according to the methods described by Boussit (1989). Concentration was estimated using a Thomas Zeiss cell counter (final dilution 1:200). All the measurements were made by the same expert technician during the whole experiment. In the sample of semen used for AI, mass motility, ranging from 6 to 9, was coded into an all-or-none trait with two values, 0 and 1, corresponding respectively to the absence (initial codes 6 and 7) and the presence (initial codes 8 and 9) of marked waves of spermatozoa. The total number of spermatozoa per ejaculate (TSE) was calculated as the product of the volume and the concentration. The number of motile sperm per ejaculate (MSE) was calculated as the product of TSE and PMS. The total number of spermatozoa per insemination dose (TSD) is proportional to the concentration of the ejaculate (TSD = 0.05 concentration). Therefore, the analysis of one variable (say concentration ) can be used for both. On the other hand, the number of motile spermatozoa per dose (MSD = TSD PMS) is an additional variable. Fertility of insemination was defined at kindling as a Bernoulli variable (1 if kindling followed insemination, 0 if no kindling). The corresponding variable is the kindling rate (KR). Litter size (LS) was recorded at birth (total born) Statistical analyses The characteristics of the inseminated semen were analyzed for all inseminations by analysis of variance with the fixed effects of the genetic type of the male (four levels), the insemination batch (grouped into three batches: 1 4, 5 7, 8 11) and their interaction. The effect of the insemination batch was a composite effect including age of mates, season and any specific effect of a single batch. Phenotypic correlations between sperm traits were

4 142 J.-M. Brun et al. / Animal Reproduction Science 70 (2002) Table 2 Definition of discrete variables (class 1 and 2) from the initial sperm variables Variable Threshold Means of class 1 Means of class 2 (<threshold) (>threshold) ph Volume (ml) Motile sperm (%) Concentration ( 10 6 spz/ml) TSE ( 10 6 spz) MSE ( 10 6 spz) TSD ( 10 6 spz) MSD ( 10 6 spz) estimated as the residual correlation from an analysis of variance of semen traits in the sample of ejaculates with the same fixed effects as previously. The relationships between semen traits and the discrete variable kindling rate were analyzed using a generalized linear model adapted to discrete data. A logistic model was used with a logit link function using the Genmod procedure of SAS (1993). The model includes one sperm trait (transformed into a discrete variable, subdividing the initial continuous variable into two equal-sized classes separated by the median of the trait distribution, Table 2), the genetic type of mates (pooling purebred types on the one hand and crossbred types on the other hand) and the physiological status of the does at insemination (four levels, crossing receptivity status receptive or not and lactation status lactating or not). Genetic types were pooled to fulfil the condition that elementary class numbers are greater than 5 and after having checked the similarity in reproductive traits between types A and B on the one hand and between types AB and BA on the other hand. The relationships between semen traits and litter size were studied using two kinds of analysis of variance models, both using the GLM procedure of SAS (1990) and considering the sperm variables one by one as independent variables. For the first kind of model, all continuous sperm variables were entered as covariates into a linear regression. In the second one, they were transformed into binary variables as previously described in the analysis of KR and entered as fixed effects. The general model was: LS = type of mating (four levels: A A, B B, AB (AB or BA) and BA (AB or BA)) physiological status of the rabbit does + insemination batch (three levels) + the semen trait considered. Interaction between semen trait and genetic type, tested as non-significant, was not kept in the model. The linearity of the relationship had been previously checked. 3. Results 3.1. General patterns of variation in semen traits and fecundity traits The semen used for the 11 insemination batches came from 54 bucks and 216 ejaculates and were used in 839 inseminations. All semen traits varied according to the genetic type of the bucks (Table 3). The BA males displayed the highest percentage of ejaculates with high

5 J.-M. Brun et al. / Animal Reproduction Science 70 (2002)

6 144 J.-M. Brun et al. / Animal Reproduction Science 70 (2002) Table 4 Phenotypic correlations between sperm characteristics ph Volume Motile sperm (%) Concentration TSE MSE MSD and TSD Motility ph Volume Motile sperm (%) Concentration TSE MSE 0.86 P<0.05. mass motility (51.5% versus 27.4% on average for the other three types). Concentration of crossbred male semen was on average higher than that of purebreds. The MSE and MSD were on average higher in crossbred males than in purebred ones, with marked superiority of the BA type. This type cumulated higher volumes, concentrations and percentages of motile spermatozoa, although not always significantly. Conversely, type B males performed the worst. This type was also characterized by the highest semen ph. As shown in Table 4, correlations between elementary sperm traits were generally low, except that between motility and concentration (0.43) and that between concentration and ph ( 0.30). The synthetic criteria TSE, MSE and MSD were highly correlated with concentration, moderately correlated with volume and motility and negatively (around 0.30) correlated with ph. Apart from concentration and ph, the elementary traits were not influenced by insemination batch but the synthetic criteria, TSF, MSE and MSD, were higher in later insemination batches, due to increasing concentration with batch number. Kindling rate and litter size were higher in mating types involving F1 crossbred males and females (Table 5), thus revealing heterosis for both traits. Kindling rate varied with insemination batch (lowest value for the second batch) and litter size was highest in the third insemination batch. However, the physiological status of the does at the time of insemination played a major role in determining both variables Relationships between semen traits and the two components of fecundity Kindling rate Of all the semen traits, mass motility had the most influence on kindling rate (Table 6). The KR means were 52.4±2% for low mass motility and 66.7±3% for high mass motility. There tended to be a ph effect (P = 0.09), with the lower ph class corresponding to the higher kindling rate. Taken separately, volume, PMS and concentration did not significantly influence kindling rate. However, when combined into synthetic criteria such as TSE or MSE, they had a significant effect on KR (P = 0.07 and 0.03, respectively). As regards MSE, the mean KR corresponding to the low level of MSE was 54.5 and 60.4% for that corresponding to the high level. The number of total or motile sperm in the dose did not influence KR significantly.

7 J.-M. Brun et al. / Animal Reproduction Science 70 (2002) Table 5 Characteristics of kindling rate and litter size according to the genetic type of bucks and does and the insemination batch Kindling rate Litter size Mean (S.D.) 57.4 (45.3) 10.5 (3.5) Statistical significance (χ 2 ) (F-test) Genetic type of bucks + Insemination batch Physiological status of doesa Genetic types c AB bucks (AB + BA does) b 59.9 (3.3) 10.9 a (0.4) BA bucks (AB + BA does) b 63.6 (3.0) 10.6 a (0.3) A bucks (A does) 54.9 (3.2) 8.9 b (0.4) B bucks (B does) 49.4(3.5) 9.5 b (0.4) Insemination batch (2.6) 9.5 b (0.3) (3.1) 9.6 c (0.4) (3.0) 10.8 a (0.3) Physiological status of the does a NR-NL 42.8 (3.8) 9.9 b (0.4) NR-L 22.2(3.8) 8.4 c (0.6) R-NL 74.8 (2.8) 10.3 b (0.3) R-L 65.8 (3.0) 1l.3 a (0.3) a R/NR, receptive/non-receptive to mating; L/NL, lactating/non-lactating. b AB and BA does not differ significantly. c The number of inseminations was 217, 203, 244 and 175 and the number of kindlings was 118, 118, 150 and 86 for bucks A, AB, BA and B, respectively. + P<0.10. P<0.05. P< Litter size The R 2 of the analysis of variance of litter size before including any sperm trait was around 11%. Including a sperm trait improved the R 2 only up to 12.1%, which means that sperm traits accounted for a small amount of variation in litter size. It must be stressed first that, when using sperm traits as covariates in regression models of litter size, the heterogeneity of slopes between genetic types of bucks was not significant. Therefore, the relationships between any sperm trait and litter size were independent of the genetic type of the bucks. Analyses of variance and regression were consistent in indicating a significant effect of semen concentration on litter size (Table 6). The two classes of concentration corresponded to a difference of 0.7 rabbits born per litter. In terms of regression, an increase in concentration by one standard deviation induced an increase in litter size of 0.40 ± 0.15 pups. Concentration dependent sperm variables, like TSE, MSE, TSD and MSD, also influenced litter size at birth. Thus, an increase in these variables by one standard deviation will

8 146 J.-M. Brun et al. / Animal Reproduction Science 70 (2002) Table 6 Influence of semen characteristics on kindling rate and litter size Kindling rate Litter size (total born) Regression Analysis of variance P a Class Estimated P a Delta b P a Class Estimated Mass motility c ns d ns d ph ns d ns d Volume ns d ns d ns d Motile sperm (%) ns d ns d ± ± 0.3 Concentration ns d ± ± ± 0.3 TSE ± MSE ± ± ± 0.3 TSD ns d ± ± ± 0.3 MSD ns d ± ± ± 0.3 a P, probability of type 1 error, given only if less than 015. b Delta, increase in litter size per one standard deviation increase in the semen trait. c Percentage of ejaculates with high motility. d Not significant. generate a increase in litter size. Analysis of variance gave the same trends although the significance levels were different. The percentage of motile spermatozoa tended to influence litter size (P = 0.09): the two classes of this variable corresponded to a difference in litter size of 0.6 rabbits born. Mass motility, exerting the main influence on kindling rate, did not influence litter size significantly Synthesis On the whole, relationships between sperm traits and either kindling rate or litter size were evidenced. However, kindling rate seems to depend more on sperm quality (such as sperm motility) plus the cumulative effects of some other traits, while litter size seems to be more dependent on quantitative aspects such as number of spermatozoa in the dose, via concentration. 4. Discussion Genetic type influenced all sperm characteristics (mass motility, ph, PMS, volume and concentration). Semen from the BA bucks performed best in terms of mass motility, PMS

9 J.-M. Brun et al. / Animal Reproduction Science 70 (2002) and concentration (not different from BA), but had the lowest ph (not significantly). Conversely, semen from B bucks exhibited the lowest values (not always different from A) for concentration, motility and PMS associated with the highest ph. This opposition between genetic types is consistent with the phenotypic correlations between sperm traits and particularly the negative correlation between ph and concentration and also with MSD, the product of concentration and PMS. An opposition between ph and other sperm traits, particularly concentration and mass motility, had been found previously by More O Ferral and Meacham (1968) and Bencheikh (1993). It can be interpreted in terms of the metabolic activity of the spermatozoa, utilizing fructose as the major source of energy (Klein and Mayer, 1963) and releasing lactic acid which, according to Coffey (1988), would decrease the ph. In our experiment, however, the time lag between semen collection and ph determination was low and significant generation of lactic acid may be questionable. Differences in ph could also be accounted for by the contribution of secretions from accessory glands fluids. Our results, where fertility varied from 22 to 75% and litter size from 8.4 to 11.3 rabbits according to the physiological stage of the does at the time of insemination, confirm previous results from Theau-Clément and Roustan, 1992 and Castellini and Lattaioli (1999). As regards the relationships between semen traits and reproductive performance, our results are in general agreement with other data for rabbits, although with particular emphasis on the influence of mass motility on fertility. Bencheikh (1993) found relationships between number of spermatozoa, motility, ph and percentage of motile spermatozoa on the one hand and fertility and litter size on the other hand. More O Ferral and Meacham (1968) evidenced a relationship between ph and fertility (r = 0.51). Vrillon et al. (1979) reported a link between pregnancy rate and mass motility, evaluated using the same method as ours. In the latter study, the pregnancy rate dropped when motility went below 6 (scale of Petitjean, 1965), which justifies the discarding of semen with a mass motility value below 6. Several studies in rabbit have focused on the question of the optimum number of spermatozoa to inseminate (Wales et al., 1965; Foote et al., 1991; Pizzi et al., 1996; Alvarino et al., 1996; Viudes-de-Castro and Vicente, 1997; Castellini and Lattaioli, 1999). The conclusions varied between authors due to differences in the physiological stages (reproduction rhythm, receptivity induced or not) and also in the conditions of sperm storage and treatment. However, they all reported a relationship between the number of spermatozoa inseminated and the subsequent rate of fertility and prolificacy with a threshold effect and a curvilinear relationship. In the present study, a curvilinear regression did not fit litter size data better than a linear regression. According to Nadir et al. (1993), the higher the dose, the greater the competition between spermatozoa at the site of fertilization, increasing the probability of an ovum being fertilized by normal sperm, ensuring both a good fertilization rate and normal subsequent embryonic development. This hypothesis could explain why the number of spermatozoa acts both upon fertilization rate and litter size. As regards the influence of the number of motile spermatozoa on fertility, it is surprising that it was significant when related to the ejaculate (MSE) but not when related to the dose (MSD). As MSD is proportional to concentration PMS while MSE is proportional to concentration PMS volume, this discrepancy can be explained by the cumulative significance of non-significant elementary traits.

10 148 J.-M. Brun et al. / Animal Reproduction Science 70 (2002) Compared to the physiological stage of does when inseminated, the quality of semen is less determining for the outcome of AI. However, any contribution to the optimization of reproductive performance ought to be utilized. 5. Conclusion Relationships between sperm traits and reproductive performance afier insemination have been evidenced. Kindling rate seemed to depend rather on sperm quality, such as sperm motility plus the cumulative effects of some other traits (concentration, PMS, volume) while litter size seemed to be more dependent on quantitative features such as the number of spermatozoa in the dose, via concentration. Our results stress the importance of selecting the ejaculates according to mass motility prior to insemination, with regard to fertility. Attention must also be paid to the characteristics of the insemination dose, and particularly the number of spermatozoa per dose which influences reproductive performance. Acknowledgements The authors are grateful to all the technicians of the experimental rabbitry of the Station d Amélioration Génétique des Animaux, namely Chabal G., Girotto H. and Esparbié J., and to the laboratory technician Jacky Falières. References Alvarino, J.M.R., Lopez, F.J., Del Arco, J.A., Bueno, A., Torres, R., Effect of semen concentration on rabbit artificial insemination with fresh or 24 h stored semen. In: Lebas, F. (Ed.), INRA. Proceedings of the sixth World Rabbit Congress, Vol. 2, pp Amann, R.P., Can the fertility potential of a seminal sample be predicted accurately? J. Androl. 10, Bencheikh, N., Production de sperme et fertilité du lapin mâle Oryctolagus cuniculus. Effet de la fréquence de collecte et du type génétique. Thése d Etat. Ecole Nationale Agronomique de Toulouse. Bencheikh, N., Effet de la fréquence de collecte de la semence sur les caractéristiques du sperme et des spermatozoïdes récoltés chez le lapin. Ann. Zoot. 44, Boussit, D., Reproduction et insémination artificielle en cuniculture. Assoc. Fr. Cuniculture, Lempdes, France, p Brun, J.M., Bolet, G., Baselga, M., Esparbié, J., Falières, J Comparaison de deux souches de lapins sélectionnées sur la taille de portée intérêt de leur croisement, 7èmes Journ. Rech. Cunicole Fr., Lyon, Brun, J.M., Bolet, G., Theau-Clément, M., Esparbié, J., Falières, J Constitution d une souche synthétique à l INRA. 1 Evolution des caractères de reproduction et du poids des lapines dans les premières générations. 8èmes Journ. Rech. Cunicole Fr., Paris. Castellini, C., Lattaioli, P., Effect of number of motile sperms inseminated on reproductive performance of rabbit does. Anim. Reprod. Sci. 57, Cherms, F.L., Variations in semen quality and relationships of semen quality to fertility in turkeys. Poultry Sci. 47, Coffey, D.S., Androgen action and the sex accessory tissues. In: Physiology of reproduction, Knobil, E., Neill, J.D. (Eds.), Raven Press, NewYork, pp Colas, G., Menissier, F., Courot, M., Paquignon, M., Technologie de l I.A. et fertilité du mâle. 1.A. et amélioration génétique: Bilan et perspectives critiques. Les colloques de l INRA no. 29. Toulouse, Auzeville, France, November 1983.

11 J.-M. Brun et al. / Animal Reproduction Science 70 (2002) Estany, J., Baselga, M., Blasco, A., Camacho, J., Mixed model methodology for the estimation of genetic response to selection in litter size of rabbits. Livestock Prod. Sci. 21, Facchin, E., Zanirato, M.G., Gualterio, L., Valentini, A., AI in rabbit breeding. Note 1: AI service for meat rabbit breeding. In: Proceedings of the fourth World Rabbit Congress, Budapest, Vol. 2, 1988, pp Foote, R.H., Farrell, P.B., Simkin, ME., Rabbit semen quality, number of sperm inseminated and fertility. J. Androl. 49, (Suppl. 1), Abstract 100. Hirao, K., A multiple regression analysis on six measurements of bovine semen characteristics for fertility. Int. J. Fertil., pp Klein, M., Mayer, G., Fertilité male. In Physiologie. III. Les grandes fonctions. Kayser, C. (Ed.), Flammarion, Paris, pp More O Ferral, G.J. Meacham, T.N., Relationships between ph, other semen traits and fertility in rabbits. In: Proceedings of the 6ème Congrès International de Reproduction Animale et Insémination artificielle. INRA, Vol. 2, Paris, July 1968, pp Nadir, S., Saacke, R.G., Bame, J., Mullins, J., Degelos, S., Effect of freezing semen and dosage of sperm on number of accessory sperm, fertility and embryo quality in artificially inseminated cattle. J. Anim. Sci. 71, Petitjean, M., Recherches sur l estimation du pouvoir fécondant des coqs. Mémoire de fin d é tudes d ingénieur, CNAM, Paris, p. 85. Pizzi, F., Guaita, N., Luzi, F., Biffi, B., Brivio, R., Crimella, C., Effect of the number of spermatozoa and spermatozoal quality on the fertility in rabbits. In: Lebas, F. (Ed.), INRA. Proceedings of the sixth World Rabbit Congress, Vol. 2, pp Rochambeau, H. de, La femelle parentale issue des souches expérimentales de l INRA: évolutions génétiques et perspectives, 7èmes Journ. Rech. Cunicole Fr., Lyon, pp SAS, 1990, User s Guide. Version 6. SAS/STAT Software. Cary, NC, USA. SAS, 1993, Release 6.09 SAS Technical Report P-243. SAS/STAT Software. Cary, NC, 14 USA, p. 88. Theau-Clément, M., Roustan, A., A study on relationships between receptivity and lactation in the doe and their influence on reproductive performances. J. Appl. Rabbit Res. 15, Viudes-de-Castro, M.P., Vicente, J.S., Effect of sperm count on the fertility and prolificity rates of meat rabbits. Anim. Reprod. Sci. 46, Vrillon, J.L., Donal, R., Poujardieu, B., Rouvier, R., Sélection et testage des lapins mâles de croisement terminal In: Bulletin Technique du Département de Génétique Animale. INRA. Vol. 28, p Wales, R.G., Martin, L., O Shea, T., Effect of dilution rate and the number of spermatozoa inseminated on the fertility of rabbits ovulated with chorionic gonadotropin. J. Reprod. Fertil. 10, Wilson, H.R., Piesco, N.P., Miller, E.R., Nesbeth, W.G., Prediction of the fertility of potential of broiler breeder males. World Poultry Sci. J. 35, Woelders, H., Reproduction in Domestic Animals. (Suppl. 1) pp