Application of liquid semen technology improves conception rate of sex-sorted semen in lactating dairy cows

Transcription

1 J. Dairy Sci. 97 : /jds American Dairy Science Association, Application of liquid semen technology improves conception rate of sex-sorted semen in lactating dairy cows Z. Z. Xu 1 LIC, Private Bag 3016, Newstead, Hamilton 3240, New Zealand ABSTRACT The objective was to compare reproductive performance of liquid sex-sorted (SS) semen with that of conventional (CON) semen in lactating dairy cows. Between 2011 and 2013, commercial dairy herds (n = 101, 203, and 253 for 2011, 2012, and 2013, respectively) with predominantly Holstein-Friesian cows were enrolled in a contract mating program to produce surplus heifers for export using liquid SS semen. During the spring mating period, each herd was allocated with liquid SS semen at 50% of its daily requirement and the remaining daily requirement was allocated with CON liquid semen. Sperm for producing SS semen was sorted by Sexing Technologies NZ Ltd. (Hamilton, New Zealand) and then packaged using the liquid semen technology of LIC (Hamilton, New Zealand) at a dose of sperm. Artificial insemination (AI) with liquid SS semen was carried out between 43 and 46 h after collection. Conventional semen straws contained , , or sperm for semen to be used on d 1, 2, or 3 after collection, respectively. Only CON inseminations on the same days as when SS semen was used were included in the comparison. Herd managers biased usage of SS semen toward cows with a longer postpartum interval before the mating start date (64.0 vs d), cows of higher genetic merit (NZ$107.0 vs. NZ$98.4), younger cows (5.1 vs. 5.2 yr), and cows in which they had more confidence of being genuinely in estrus as measured by a lower percentage of short returns between 1 and 17 d (5.3 vs. 7.5%). After adjusting for these factors, the estimated difference in nonreturn rate between AI with SS and CON semen over the 3 seasons was 3.8 percentage points (SS = 70.2% vs. CON = 74.0%; SS/CON = 94.9%). The estimated maximum difference in calving rate per AI between SS and CON semen was 3.1 percentage points for 2011 (SS = 51.2% vs. CON = 54.3%; SS/CON = 94.3%) and 3.0 percentage points for 2012 (SS = 49.7% vs. CON = 52.6%; SS/CON = 94.5%). Calving data for Received June 19, Accepted August 10, Corresponding author: zxu@lic.co.nz were not yet available. The percentage of heifer calves born to AI with SS semen was 87.0% for 2011 and 85.8% for 2012, both of which were lower than the expectation of 90% mainly due to misidentification of calf dams in seasonal dairy herds calving on pasture. In summary, results in this report showed that liquid SS semen only required half the dose rate of frozen SS semen to achieve a reproductive performance of over 94% of CON semen in lactating dairy cows. Careful planning and a robust distribution network are required to avoid semen wastage and to maximize the benefit of liquid SS semen. Key words: liquid sex-sorted semen, reproductive performance, lactating dairy cow INTRODUCTION The ability to preselect the sex of the offspring at the time of breeding offers economic benefits to the dairy industry where commercial dairy farmers are primarily focused on breeding heifer replacements. Currently, the only commercially viable method for producing sexsorted (SS) mammalian sperm is by flow cytometry to exploit the small difference in DNA content between X and Y chromosomes (Garner and Seidel, 2008; Seidel, 2012). Although this technology has undergone rapid development in the past decade, the current method of sperm sorting has a relatively slow throughput and, therefore, a suboptimal number of about SS sperm per dose is commonly used as a compromise between cost and acceptable semen conception rate (Seidel, 2007). In addition, the sorting process causes damage to sperm (Seidel, 2007; Garner and Seidel, 2008). As a consequence of both a suboptimal dose rate and damage to sperm during the sorting process, conception rate from inseminations with frozen SS semen is typically between 70 and 80% of that from conventional (CON) semen at the normal dose rate (DeJarnette et al., 2009, 2011; Frijters et al., 2009). However, the reduction in conception rate with frozen SS semen cannot be fully overcome through increasing the sperm dose rate (Seidel and Schenk, 2008; DeJarnette et al., 2010, 2011; Lucena et al., 2014). The reduction in conception rate with frozen SS semen would negate some of the po-

2 OUR INDUSTRY TODAY 7299 tential benefits from using frozen SS semen in seasonal dairy production systems where cows need to conceive within a short period of time during spring (Verkerk, 2003; Butler et al., 2014). In New Zealand, most cows are bred with liquid semen and the typical dose rate for liquid semen is between and sperm compared with the typical dose rate of sperm for frozen semen (Vishwanath, 2003). It is known that the freezing and thawing processes damage sperm and this effect may be greater for SS sperm because SS sperm have already been compromised during the sorting process (DeJarnette et al., 2011; Gosálvez et al., 2011a,b). Therefore, we hypothesized that the use of liquid SS semen to avoid damage during the freezing and thawing processes would overcome much of the decrease in conception rate with frozen SS semen. Following a successful pilot trial in 2010, a large number of inseminations with liquid SS semen was carried out between 2011 and Here, the results and experience are presented from using liquid SS semen at sperm per dose in lactating dairy cows on commercial dairy farms. MATERIALS AND METHODS Cows and Trial Procedures Data were collected over a 3-yr period between 2011 and 2013, involving lactating cows in 101, 203, and 253 dairy herds, respectively. These commercial dairy herds with predominantly Holstein-Friesian cows were enrolled in a contract mating program to produce surplus heifers for the export market using liquid SS semen. During the spring breeding period, each enrolled herd was allocated liquid SS semen at 50% of the estimated daily requirement. The remaining semen requirement was allocated with CON liquid semen. Artificial insemination was carried out by experienced technicians employed by LIC (Hamilton, New Zealand). Both the herd managers and AI technicians were aware which semen straws contained SS semen. The decision on which cows were to be bred with SS or CON semen was made by the herd managers. Bulls and Semen Processing Procedures Bulls (n = 10, 11, and 12 for 2011, 2012, and 2013, respectively) that were for producing SS semen were genomically selected young (2 or 3 yr) Holstein-Friesian bulls and these were different from the Holstein-Friesian bulls (n = 39, 44, and 34 for 2011, 2012, and 2013, respectively) used for producing CON semen, which were a mixture of mature bulls that had graduated from progeny testing and genomically selected young bulls (1 to 3 yr). Ejaculates for producing SS semen were collected using an artificial vagina per standard operation procedures between 1300 and 1400 h. Ejaculates that met the required quality standards were delivered to Sexing Technologies NZ Ltd. (Hamilton, New Zealand) to be sorted according to their standard operating procedures. The targeted purity for the X-sperm fraction was 90%. After centrifugation (undisclosed conditions) of the sorted sperm, the sperm pellet was rediluted using a Caprogen liquid semen extender (LIC) to sperm/ml (Shannon, 1965). The diluted semen was delivered to the LIC semen production laboratory the following morning (about 0600 h). The SS semen was filled and sealed in liquid semen straws (0.25 ml) and dispatched to AI technicians for insemination in the following morning. The interval between semen collection and insemination was mostly between 43 and 46 h, depending on when insemination took place after the morning milking. Ejaculates for producing CON semen were collected during early morning and diluted using Caprogen liquid semen extender to , , or sperm/ml for semen to be used on d 1, 2, or 3 after collection, respectively. The CON semen was filled and sealed in liquid semen straws (0.25 ml) and dispatched to technicians every 3 d and used according to their target usage days. Data and Statistical Analyses Data on cows, AI, calving, and culling were extracted from the National Dairy Database maintained by LIC. Because reliable pregnancy diagnosis data were not available for most herds, reproductive performance outcome was measured using 24-d nonreturn rate (NRR), which was defined as the percentage of AI that were not followed by another AI between 18 and 24 d after the previous AI. An AI was counted as not returned if the following AI was >24 d after the previous AI. Artificial inseminations that were followed by another AI between 1 and 17 d were excluded from the NRR calculation because these short returns were either due to cows having short estrous cycles or due to estrus detection errors, but were not causally related to the performance of the first AI. Artificial inseminations within 24 d of the end of the AI breeding period were also excluded from NRR calculation because of the poor accuracy of mating records during the bull breeding period. The average AI breeding period for herds in the trial was 42 d (between 27 and 88 d). The short AI breeding period precluded the use of an interval longer than 24

3 7300 XU d for NRR calculation. Although use of a 24-d interval had the potential to inflate NRR result compared with a longer interval, its effect was minimized by the high estrus detection rate (>90%) in dairy herds in New Zealand (Xu et al., 1998; Xu and Burton, 2000) To verify if the NRR results reflected actual conception results, calving dates in the following season were used to estimate the conception date for cows that had not been culled before calving. An AI was considered to have produced a pregnancy if the interval between the date of AI and date of calving was between 270 and 292 d (mean ± 2 SD for gestation length: 281 ± 11 d; Grosshans et al., 1997). Although a small proportion of AI would be assigned an incorrect pregnancy status based on the above rule, it was expected that the false positives and false negatives would cancel each other due to the normal distribution for gestation length. Differences in postpartum interval, genetic merit and cow age between cows bred with SS and CON semen were analyzed using PROC GLM of SAS (SAS Institute, 2010). The NRR data were analyzed using PROC GLIMMIX of SAS, which fitted generalized linear mixed models to binomial data with the default logit link function. For the NRR analyses, season, herd, and cow age (2 to 7 and 8 yr) were included in the model as fixed effects and postpartum interval and genetic merit were included as covariates. The interactions between season and treatment were also included to test if the treatment effect was consistent across seasons. Initially, herd was included in the model as a random factor, but the model failed to converge. Because only inseminations in herds that used both CON and SS semen on the same days were included in the analyses and a large number of herds were involved, the inclusion of herd as a fixed effect in the model would not invalidate the comparison between treatments. Comparisons of frequency results for calving and sex ratio between treatment groups were analyzed within season using the Pearson chi-squared test. RESULTS AND DISCUSSION SS Semen Usage on Farms Because of the high cost and limited shelf life of liquid SS semen, a major challenge to implementing liquid SS semen on a large scale is to minimize semen wastage after dispatch from the semen production laboratory. By allocating SS semen at 50% of the expected daily demand, a utilization rate of over 96% was achieved over the 3 seasons (Table 1). Inability to accurately predict daily requirement at a herd level was one cause for semen wastage. Daily semen requirement was estimated using historical average proportion of cows Table 1. Statistics on the number of dispatched straws, inseminations, and utilization rate of liquid sex-sorted semen in 2011, 2012, and 2013 Dispatch, no. Inseminations, no. Utilization, % ,880 16, ,362 34, ,741 47, Total 101,983 98, bred on different days of the breeding period. Variation among herds and among days within a herd, especially when the herd size was small, resulted in the number of dispatched SS semen straws being greater than the number of cows to be bred on a small number of occasions (<5%). Other main causes for semen wastage included sporadic operational errors, such as delivery of semen to the incorrect location. When SS semen was allocated at 50% of the expected daily requirement, herd managers had the opportunity to target SS semen usage on cows that would give them the most benefits. Table 2 shows the differences between cows bred with SS and CON semen in the postpartum interval to each herd s breeding start date, breeding worth index (the genetic merit index used in New Zealand), and cow age. Except for cow age in 2013, all differences between CON and SS were statistically significant (P < ) due to the large number of observations involved. Herd managers biased SS semen usage toward cows that calved earlier (by 1.2 d), cows of higher genetic merit (by NZ$8.6), and younger cows (by 0.1 yr). Among cow age groups, SS semen was used on fewer cows 8 yr and over (14.0 vs. 17.5%; P < ), more cows between 4 and 6 yr (44.2 vs. 38.8%; P < ), fewer 2-yr-old cows (17.0 vs. 19.4%; P < 0.001), and similar percentages of cows aged 3 (16.0%) and 7 yr (8.6%). However, these biases were small in absolute values relative to the standard deviations for these parameters. The greatest bias was for cow genetic merit at 0.17 standard deviation, compared with the possible maximum bias of 0.8 standard deviation for when 50% of the top cows were selected for breeding with SS semen. The smaller-than-expected bias in cow selection based on genetic merit could be due to the difference in genetic merit of bulls used for producing SS (second-tier genomically selected bulls) and CON semen (progeny-tested and first-tier genomically selected bulls). In addition, herds in this study were contracted to produce surplus heifer calves for the export market, which might have reduced the drive to maximize the genetic merit of offspring through using SS semen. The percentage of return inseminations between 1 and 17 d (short returns) was higher (P < 0.001) after AI with CON semen than after AI with SS semen

4 OUR INDUSTRY TODAY 7301 Table 2. Differences (means ± SD) between cows bred with sex-sorted (SS) and conventional (CON) semen in the postpartum interval (PPI) to each herd s breeding start date (d), breeding worth index (BWI; NZ$), cow age (yr), and percentage of AI followed by another AI between 1 and 17 d of the previous AI (short return; %) 1 Semen type n PPI, d BWI, NZ$ Age, yr Short return, % 2011 CON 18, ± ± ± SS 15, ± ± ± CON 33, ± ± ± SS 33, ± ± ± CON 45, ± ± ± SS 46, ± ± ± Total CON 98, ± ± ± SS 95, ± ± ± All differences between CON and SS are statistically significant (P < ), except for cow age in 2013, which is not significant. (Table 2). Although some short returns could be due to cows with genuine short estrous cycles, the majority of short returns were probably caused by estrus detection errors. The lower percentage of short returns after AI with SS semen compared with CON semen suggests that herd managers biased SS semen usage toward cows they had more confidence of being genuinely in estrus. The potential effect of this bias on NRR comparison was minimized by excluding AI with short returns from NRR calculations. NRR The NRR for SS semen (69.1%) over the 3 seasons was significantly lower (P < ) than that for CON semen (73.1%; Table 3)., herd, cow age, and calving dates all had significant (P < ) effects on NRR, but no significant effects of cow genetic merit or season by treatment interactions were detected. The least squares means estimates for NRR were 70.2% for SS semen compared with 74.0% for CON semen, a difference of 3.8 percentage points, or 5.1%. Results in this report from a large number of inseminations over 3 yr show that liquid SS semen at sperm per dose was able to consistently achieve a NRR that was over 94% of CON semen. The performance of liquid SS semen relative to CON semen was higher than the 70 to 80% reported in other studies for frozen SS semen at sperm per dose (Seidel and Schenk, 2008; DeJarnette et al., 2009, 2011; Frijters et al., 2009). In New Zealand, not many inseminations with frozen SS semen have been conducted. For comparison, the NRR for 2,172 inseminations with frozen SS semen in 2013 was 53.0%, which was 20 percentage points lower than the NRR of CON semen in 2013 (Z. Z. Xu, unpublished data). On a percentile basis, the NRR of frozen SS semen in New Zealand was 73% of CON semen, consistent with the relative performance of 70 to 80% reported in other studies (Seidel and Schenk, 2008; DeJarnette et al., 2009, 2011). The performance of liquid SS semen in the current report is higher than the performance of liquid SS semen reported for another study in Ireland, which showed that the conception rates for liquid SS semen at or sperm per dose were 75 and 87% of CON semen, respectively (Butler et al., 2014). A difference in the liquid semen extenders used in New Zealand and Ireland could be one reason for the difference in performance of liquid SS semen. Calving Performance To verify if the difference in NRR observed above is a valid indication of the performance of liquid SS semen in terms of conception and calving, we followed the fate of the trial cows in 2011 and 2012 until they either Table 3. Nonreturn rate (NRR) for AI with liquid sex-sorted (SS) or conventional (CON) semen SS CON SS CON SS/CON n NRR, % n NRR, % NRR, % % , , , , , , Total 51, ,

5 7302 XU were culled or calved in the following season (Table 4). The calving rate per AI was similar between SS and CON semen in both seasons. However, differences existed in culling rate between cows inseminated with SS and CON semen, which would invalidate the comparison in calving rate per AI. The percentage of AI with CON semen in culled cows was higher (P < 0.001) than that with SS semen, suggesting that farmers were less likely to cull cows that had conceived to inseminations with SS semen. The percentage of AI with SS semen in nonpregnant cows was lower (P < 0.001) than that of AI with CON semen in both seasons. Although the magnitudes of these differences were small (1% or less), they indicated that the use of SS semen did not increase the nonpregnancy rate at the end of the breeding season. Therefore, the observed difference in culling rate between cows after breeding with SS and CON semen was mainly due to differences in voluntary culling for reasons not related to reproductive performance. This difference in culling practice was expected because herd owners wanted to maximize the number of surplus heifer calves for the export market. By assuming that all the extra AI with CON semen in cows that had been culled for reasons other than nonpregnancy had actually produced a conception, we can estimate the maximum difference in calving rate between SS and CON groups, these being 3.1 percentage points for 2011 and 3.0 percentage points for 2012 (adjusted calving/ai in Table 4). These differences in calving rates are slightly lower than the difference of 4.2 percentage points for NRR in these same 2 seasons (Table 3). However, on a percentile basis, the adjusted calving rate per AI with SS semen was 94.5% of that with CON semen, very similar to the 94.6% for NRR (Table 3). Therefore, results for both NRR and calving rate suggest that the liquid SS semen technology implemented by LIC can achieve a reproductive performance that is over 94% of CON semen. Calf Sex Results for the sex of calves born to AI with SS and CON semen are shown in Table 5. For both seasons, the percentages of heifer calves born to AI with SS semen were significantly (P < 0.001) less than the 90% target set for the sorting stage, whereas the percentages of heifer calves born to AI with CON semen were significantly (P < 0.001) higher than the 48.6% reported previously for CON semen under New Zealand conditions (Xu et al., 2000). The main reason for these differences was due to misidentification of calf dams under the seasonal production system in New Zealand, where many cows could calve together on the same paddock (Verkerk, 2003). An unpublished study in 2011 using DNA parentage testing on more than 20,000 cows in 97 herds showed that 23% of the animals had incorrect sire information. For the present data set, an error rate of about 20% in dam identification would be sufficient to explain the differences between the observed heifer percentages and their expected values. General Discussion Results in this report demonstrate the effectiveness of using liquid semen technology to improve the performance of SS semen. At 50% of the dose rate of frozen SS semen, liquid SS semen was able to achieve a performance of over 94% of CON semen, which was between 15 and 25% better than the 70 to 80% achieved by frozen SS semen relative to CON semen. Therefore, the efficiency gain from liquid over frozen SS semen is 2-fold: twice as many straws per unit time of sorting and 15 to 25% increase in conception rate. This improved conception rate is required to make SS semen economical for the pasture-based, seasonal dairy production system in New Zealand (McMillan and Newman, 2011). The mechanism by which liquid SS semen improves concep- Table 4. Calving and culling statistics for AI with sexed-sorted (SS) or conventional (CON) semen in 2011 and SS CON SS CON SS CON SS CON No. of AI 14,239 17,372 31,051 31,294 Calving/AI, % % of AI in all culled cows a a % of AI in nonpregnant cows a a Adjusted calving/ai, 3 % a a a Differences between SS and CON differ significantly (P < 0.001; chi-squared test). 1 Number of AI in culled cows/total number of AI Number of AI in nonpregnant cows/total number of AI Adjusted calving/ai for the CON group was calculated by assuming that the extra AI with CON semen in cows that had been culled for reasons other than nonpregnancy had actually produced a conception. For example, adjusted calving/ai for CON in 2011 = [( ) ( )].

6 OUR INDUSTRY TODAY 7303 Table 5. Percentage of heifer calves born to AI with sex-sorted (SS) or conventional (CON) semen 1 SS tion rate over frozen SS semen is probably by avoiding the damage to sperm during the freezing and thawing processes. This is similar to the situation with CON semen where liquid semen at one-tenth of the dose rate of frozen semen is able to achieve equivalent conception rate (Shannon and Vishwanath, 1995; Vishwanath and Shannon, 2000). It is possible that the sorting process might have made SS sperm more sensitive to the damage caused by freezing and thawing. Reduction in the dose rate from to sperm caused an 8.3% decrease in conception rate in frozen CON semen, but a 13.6% decrease for frozen SS semen (DeJarnette et al., 2011). In addition, it has been shown that the increases over time in sperm DNA fragmentation in frozen-thawed semen samples was faster for SS semen than CON semen (Gosálvez et al., 2011b). Implementation of liquid SS semen is not without its own challenges. Due to its short lifespan, implementation of liquid SS semen on a large scale requires careful planning and an efficient distribution network to avoid wastage of SS semen. During the 3 seasons, a utilization rate over 96% was consistently achieved. In this regard, the seasonal production system in New Zealand with a short, concentrated breeding period in spring enables the benefits offered by liquid SS semen to be efficiently exploited. In New Zealand, insemination is usually carried out once per day after the morning milking, regardless of when cows are first detected in estrus. In situations where insemination follows the a.m.-p.m. rule, the number of cows to be bred is known 12 h ahead of time. With modern communication tools and an efficient distribution network, it may be possible to dispatch the exact number of liquid SS semen straws over a 12-h period to farmers, thus solving the semen wastage problem. ACKNOWLEDGMENTS CON n % n % , , , , Total 23, , A gestation length of between 270 and 292 d was used to determine if an AI had resulted in the recorded calving. The author thanks Marcelo Stiaque of Sexing Technologies NZ Ltd. (Hamilton, New Zealand) for his cooperation to adapt the sorting operation to the requirements for liquid SS semen, staff at LIC (Hamilton, New Zealand) who contributed to the successful implementation of the liquid SS semen in the field, and the herd owners. REFERENCES Butler, S. T., I. A. Hutchinson, A. R. Cromie, and L. Shalloo Applications and cost benefits of sexed semen in pasture-based dairy production systems. Animal 8(Suppl. 1): DeJarnette, J. M., M. A. Leach, R. L. Nebel, C. E. Marshall, C. R. McCleary, and J. F. Moreno Effects of sex-sorting and sperm dosage on conception rates of Holstein heifers: Is comparable fertility of sex-sorted and conventional semen plausible? J. Dairy Sci. 94: DeJarnette, J. M., C. R. McCleary, M. A. Leach, J. F. Moreno, R. L. Nebel, and C. E. Marshall Effects of 2.1 and sex-sorted sperm dosages on conception rates of Holstein cows and heifers. J. Dairy Sci. 93: DeJarnette, J. M., R. L. Nebel, and C. E. Marshall Evaluating the success of sex-sorted semen in US dairy herds from on farm records. Theriogenology 71: Frijters, A. C. J., E. Mullaart, R. M. G. Roelofs, R. P. van Hoorne, J. F. Moreno, O. Moreno, and J. S. Merton What affects fertility of sexed bull semen more, low sperm dosage or the sorting process? Theriogenology 71: Garner, D. L., and G. E. Seidel Jr History of commercializing sexed semen for cattle. Theriogenology 69: Gosálvez, J., M. A. Ramirez, C. López-Fernández, F. Crespo, K. M. Evans, M. E. Kjelland, and J. F. Moreno. 2011a. Sex-sorted bovine spermatozoa and DNA damage: I. Static features. Theriogenology 75: Gosálvez, J., M. A. Ramirez, C. López-Fernández, F. Crespo, K. M. Evans, M. E. Kjelland, and J. F. Moreno. 2011b. Sex-sorted bovine spermatozoa and DNA damage: II. Dynamic features. Theriogenology 75: Grosshans, T., Z. Z. Xu, L. J. Burton, D. L. Johnson, and K. L. Macmillan Performance and genetic parameters for fertility of seasonal dairy cows in New Zealand. Livest. Prod. Sci. 51: Lucena, J. A., A. G. Kenyon, J. P. Reynolds, J. F. Moreno, R. W. Lenz, D. Carroll, T. W. Lehenbauer, J. D. Champagne, and S. S. Aly Comparison between low-dose, high-sort and high-dose, low-sort semen on conception and calf sex ratio in Jersey heifers and cows. J. Dairy Sci. 97: McMillan, W. H., and M. G. Newman An economic appraisal of sex-sorted semen in New Zealand dairy herds. Proc. N.Z. Soc. Anim. Prod. 71: SAS Institute The SAS System for Windows, Release 9.3. Version 13. SAS Institute Inc., Cary, NC. Seidel, G. E., Jr Overview of sexing sperm. Theriogenology 68: Seidel, G. E., Jr Sexing mammalian sperm Where do we go from here? J. Reprod. Dev. 58: Seidel, G. E., Jr., and J. L. Schenk Pregnancy rates in cattle with cryopreserved sexed sperm: Effects of sperm numbers per inseminate and site of sperm deposition. Anim. Reprod. Sci. 105: Shannon, P Contribution of seminal plasma, sperm numbers, and gas phase to dilution effects of bovine spermatozoa. J. Dairy Sci. 48: Shannon, P., and R. Vishwanath The effect of optimal and suboptimal concentrations of sperm on the fertility of fresh and frozen bovine semen and a theoretical model to explain the fertility difference. Anim. Reprod. Sci. 39:1 10. Verkerk, G Pasture-based dairying: Challenges and rewards for New Zealand producers. Theriogenology 59: Vishwanath, R Artificial insemination: The state of the art. Theriogenology 59:

7 7304 XU Vishwanath, R., and P. Shannon Storage of bovine semen in liquid and frozen state. Anim. Reprod. Sci. 62: Xu, Z. Z., and L. J. Burton Estrus synchronization of lactating dairy cows with GnRH, progesterone, and prostaglandin F 2α. J. Dairy Sci. 83: Xu, Z. Z., D. L. Johnson, and L. J. Burton Factors affecting the sex ratio in dairy cattle in New Zealand. Proc. N.Z. Soc. Anim. Prod. 60: Xu, Z. Z., D. J. McKnight, R. Vishwanath, C. J. Pitt, and L. J. Burton Estrus detection using radiotelemetry or visual observation and tail painting in dairy cows on pasture. J. Dairy Sci. 81: