Data on various semen parameters (semen volume and initial motility), reproduction parameters and the causes of disposal of HF-crossbred cattle and Murrah buffalo bulls, during the period of 1999 to 2010, were compiled from the available calving records, daily semen collection register, auction lists and the performance record-registers (GADVASU Dairy Farm). Data on male reproductive traits were available for a total of 129 bulls. Animals which are under experimental trials and those with incomplete records were not included in the study as it adds to the effect of different genetic and non-genetic factors on the fertility parameters. The semen production parameters, generated from the primary data, includes: age at first semen collection (AFSC), age at first semen freezing (AFSF), semen production period (SPP), frozen semen production period (FSPP), age at last semen collection (ALSC), age at last semen freezing (ALSF) and age at disposal (AD). The subfertility traits included poor libido, poor semen quality (initial motility <45%) and poor semen freezability (post thaw motility <35%) which lead to disposal of bulls.

To study the effect of species (cattle and buffalo), the period vis-à-vis the season of birth the data generated on male specific breeding soundness traits were subjected to least squares analysis. Seasons of birth were further divided into four categories: summer (May to July); rainy (August to October); winter (November to February) and spring (March to April). The periods of semen donation were taken as six years interval, 1985 to 1990; 1991 to 1996; 1997 to 2002 and last eight years i.e. from 2003 to 2010, corresponding to the subsequent changes in the managemental practices. The least squares model that incorporated the fixed effects (of species, birth period and birth season) is : Y_ijkl = μ+Sp_i+S_j+P_k+I_Sp.P+e_ijkl; where, Y_ijkl = character variable of l^th bull under i^th species, j^th season of birth and k^th period of birth; μ = population mean; Sp_i = fixed effect of i^th species (i = 1, 2); S_j = fixed effect of season of birth (j = 1, 2...4); P_k = fixed effect of k^th period of birth (k = 1, 2...4); I_Sp.P. = effect of interaction between species and period; e_ijkl = random error for l^th observation under i^th species, j^th season and k^th period of birth.

Chi-square test was done for the test of heterogeneity (Snedecor and Cochran, 1994) between the species for different subfertility traits. The statistical software used for data analysis was SPSS 11.

Linear regression equation has been constructed to predict the semen donation parameters (AFSF, AD, SPP, FSPP) using the age at first semen collection as predictor variable. The regression equation constructed separately for each of the species, for prediction of the andrological parameters is: Y_i = b₀+b₁*X_i+e_i ; where, Y_i = dependent variable of i^th observation; b_o = intercept constant; b₁ = regression coefficient; e_i = random error for i^th observation and X_i = explanatory variable.

The total number of cattle and buffalo bulls reserved for breeding and number of bulls produced freezable semen, during the period of 1999 to 2010 has been represented in Table 1. Percent production of freezable semen out of the total number of semen-donating bulls was higher in buffalo (30.4%) than in cattle (25.641%). The number of bulls producing freezable and non-freezable semen differed significantly (p<0.01) between the two species (Table 2).

The subfertility traits leading to the disposal of dairy bulls maintained at the dairy farm, GADVASU, has been tabulated in Table 3. Lack of libido, poor semen quality and freezability were the important reasons for the bull wastage both in cattle (26.70%) and buffalo (25.84%). One cryptorchid (FC 1367) and two bulls with penile deviation were encountered in cattle during the tenure. Disposal pattern of the bulls showed that poor growth rate was specific to cattle. Chi-square test revealed that the subfertility traits were significantly important reasons (p<0.01) leading to disposal of bulls (Table 4).

Least squares analyses of the andrological parameters revealed that species (Sp) and periods of birth (P) had significant effect on most of the traits studied (Table 5). The effect of season is both direct and indirect. It affects the animal directly through macro and micro climatic factors, which are the temperature, humidity, rainfall and photo-period. Indirectly it acts by affecting the vegetation, forage quality and soil-plant-animal interaction. The least squares mean along with standard error of the species/breeds for different variables have been given in Table 6. In cattle, semen volume (S.Vol) was higher than buffalo. While age at first semen collection (AFSC), age at first semen freezing (AFSF), age at disposal (AD), age at last semen collection (ALSC), freezable semen production period (FSPP) and semen production period (SPP), initial motility (IM) and age at last semen freezing (ALSF) was higher in buffalo. The least squares means of all the reproductive parameters except for initial motility were showing a decreasing yet drifting trend over the periods (Table 7).

The prediction equation for the parameters, namely, SPP, AFSF, AD and FSPP using AFSC as the predictor variable, were significant (p<0.01) in cattle (Table 8), however, in buffalo only the AFSF was found to be significant (p<0.01) as the dependent variable (Table 9). The coefficient of determination (R²) value was 50% or more for prediction of AFSF (64.5%) and AD (57.9%) in cattle, vis-à-vis AFSF (49.9%) in buffalo, using AFSC as the predictor variable.

Selection of male calves based on expected predicted difference (EPD) has led to higher culling rate in cattle followed by buffalo due to differences in respective herd-size in GADVASU dairy farm. The increasing use of EPD information by beef cattle producers has placed greater value on early puberty, larger testis size, and increased sperm production by individual bulls (Casas et al., 2007). The initial selection of the male calves has no impact on the reproductive potential of the future bulls. One of the remarkable observations in this regard is that higher frequency of subfertility in the cattle among the dairy breeding bulls. Concurrently, the freezable semen production capacity was also less in crossbred cattle. Earlier reports also show that the seminal parameters are poor in crossbred bulls than the indigenous breeds (Sethi et al., 1989; Rao et al., 1995; Khate, 2005; Mukhopadhyay et al., 2011). In a study on 173 Frieswal bulls at PDC, Meerut, India, Tyagi et al. (2000) found that 54.91% and 45.09% of the crossbred bulls produced semen of poor quality and poor freezability, respectively. This account was much higher than our observation at GADVASU herd. Available reports show that about 50% of the crossbred bulls produce semen which is not cryopreservable (Sethi et al., 1989; Bhavsar, 1993; Sahni and Mohan, 1998). Variations in spermatozoal resistance to hypertonic solutions may affect the differential cryopreservability of different bulls (Maule, 1962; Tyagi et al., 2000). Increase in exotic inheritance may improve the freezability of crossbred semen (Methew et al., 1982).

The present investigation shows that the significantly important cause of disposal of breeding bulls is subfertility (p<0.05). Due to sub-optimal reproductive efficiency almost half to one-third of the reserved stock has been disposed. A number of reports related to the vital records on semen production parameters and bull wastage (Aehnelt et al., 1963; Walter and Weiser, 1966; Djimde and Weniger, 1984) specify that infertility and subfertility are the major reasons for disposal of breeding bulls of dairy animals in the tropical and temperate countries (Mukhopadhyay et al., 2011). The major reasons for disposal of breeding bulls are the poor semen quality in crossbred and Murrah buffaloes (Suryaprakasam and Rao, 1993) and the poor semen freezability in case of exotic and crossbred bulls (Mathew et al., 1982; Suryaprakasam and Rao, 1993). Roman et al. (1968) indicated that reproductive incompetence and failure to serve accounted for 47% and 16%, respectively, of all reasons of disposal of beef sires in US and Canada during 1949 to 1964. While in India, the most prevalent reasons given for culling of 469 bulls (at three AI centers in Kerala, India) were poor semen quality (45.42%), poor libido (36.46%), attainment of production targets (14.07%) and poor semen freezability (10.87%) (Sudheer and Xavier, 2000). This becomes a cause for huge wastage and consequently results in huge economic losses, as these bulls are discarded at around 2 to 3 years of age. Apart from this, there is also a reduction in the genetic gain due to disposal of bulls selected for breeding purpose. Often, even the bulls with good libido and producing freezable semen, they tend to produce low grade ejaculates under stressful conditions (Bhakat et al., 2009). Kodagali et al. (1980) reported that very high proportion (i.e. 23.81% and 38.09%) of Surti buffalo bulls (at Amul AI centers, India) were culled due to poor semen quality and overall sexual derangements, respectively, while the occurrence of poor libido was only 2.38%, during the four years tenure at Anand, Gujarat, India. A total of 1 case of cryptorchidism was recorded in cattle bull in the present study, whereas no such cases were found in buffalo bull. Collection of semen with artificial vagina is very challenging when the bulls suffer from penile deviation; in our case two cattle bulls with penile deviation but acceptable seminal profile were thus culled.

Season of birth bears its effect up to few months of age of the animal and gradually fades away. In the present study, the seasons of birth didn’t differ significantly for any of the seminal parameters, however, in a study on buffalo, the male calves born during spring season were found to have higher culling rate (Kodagali et al., 1980). Effects of period of birth and species×period interaction are entirely herd specific. In the present context improvement over the period has been noted for almost all the parameters. Most reproductive traits are heavily influenced by differences in herd management practices and other environmental factors. Better managemental practices and regular monitoring of reproductive health have contributed to improvement of the reproductive parameters. Wilcox et al. (1966) found that the periods were affecting the semen production parameters significantly, while analyzing the data on 3,774 males of US and Canada. Sethi et al. (1989) concluded that bulls donating larger volume of neat semen with higher mass activity are supposed to produce freezable semen. The mean age of first semen collection (AFSC) and freezing (AFSF) were lower in cattle since these are crossbred, hence attain maturity earlier than buffalo bulls. The maximum herd life of the buffalo bulls, as evident from the mean Age at disposal (AD), is explained by the delayed initiation of semen donation and freezable semen production. Our findings on the mean age at first semen collection in buffalo are lesser than what was reported for Murrah bulls by Suryaprakasham and Rao (1993) but comparable to Sahiwal and crossbred bulls as reported by the above authors and Rao (1995) on Jersey×Ongole crossbred and Ongole bulls. Average age of first semen donation could be reduced by subjecting the bulls on training at an early age, without affecting the semen quality and freezability (Sethi et al., 1989). The age of starting semen donation is lower in crossbred than indigenous humped breed of cattle (Mukhopadhyay et al., 2010). The results in our study have revealed a lower breeding period of the buffalo bulls than the findings of Rao et al. (1995) and Sudheer and Xavier (2000) in various indigenous and crossbred bulls in other herds. Kodagali et al. (1980), while studying the culling percentage of Surti buffalo bulls, reported the mean age at disposal (AD) and breeding tenure as 2,330 and 480 d, respectively. Mathur et al. (2002) observed that there was no significant seasonal difference in mass motility, progressive motility and sperm concentration except for semen volume, which was the lowest during winter (3.84 vs 4.17, 4.18 ml), as compared to summer and rainy seasons in Frieswal bulls.

The coefficient of determination (R²) of the regression equation is equal to 50% in case of buffalo (for AFSF) and more than 50% in case of cattle (for AFSF and AD). This indicates that the AFSF and AD can be predicted by AFSC. To the best of our knowledge, no such prediction equations have been constructed to foresee the possible herd-life (i.e. age at disposal) of the cattle bulls. This warrants further investigation using more number of animals in order to validate the prediction equation.