Authors
Gary J. Nie, DVM, MS, PhD, Diplomate ACT, Diplomate ABVP; Barton W. Rohrbach, VMD, MPH, Diplomate ACVPM; James G.W. Wenzel, DVM, MS, PhD, Diplomate ACT, Diplomate ACVPM; John B. Bennett, DVM; J. Eric Martin, DVM; David F. Davenport, DVM, MS (Ag Econ), MS (Nutr), CNS.

Materials and Methods
This study was conducted using Tennessee Walking Horses at a commercial breeding farm in middle Tennessee. Stallions (n=12) standing at stud and the mares booked to them and managed on the farm during the 2000 (n=338) and 2001 (n=255) seasons were included in the study. Stallions were grouped by age and then randomized to a treatment or control group for the 2001 season. Farm records were reviewed to provide fertility data from the 2000 breeding season and data was used to compare with reproductive performance in the 2001 season.

All mares were bred by artificial insemination, therefore effects on conception rates were not due to changes in total sperm per ejaculate. Only mares bred on farm using fresh semen and whose 16 day pregnancy status confirmation was preformed on farm were used. These selection criteria resulted in a higher proportion of “problem mares” which explains the below normal conception rates for both the control and treatment groups.

Fertility data was analyzed using a mixed model ANOVA procedure (PROC Mixed) with repeated measures. The model consisted of a dependent variable percent pregnant and independent class variables treatment, year and time period (1-5). Stallion was included in the model as a random factor. Semen quality data was analyzed using PROC GLM with repeated measures.

Results
Stallion ages ranged from 5 to 26 years. Control stallions were bred to 127 and 100 mares and the treatment stallions to 211 and 155 mares in the 2000 and 2001 seasons, respectively. The control and treatment stallions settled mares in 65/215 (30%) and 83/351 (24%) cycles in 2000 and in 53/179 (30%) and 86/223 (39%) cycles in 2001 during the T2-T5 periods, respectively. The proportion of pregnancies was not different between the 2000 and 2001 season in the T2-T4 periods for the control stallions (P >0.05). Treatment stallions achieved a significantly higher proportion of pregnancies during the same time periods T2-T4 in 2001 when compared with the same period in 2000 when they were not receiving the supplement (P <0.05). The increase in proportion of pregnancies was consistent across all the treatment stallions (n=6) while all control stallions (for which 2000 fertility data was available, n=3) experienced a slight decline in fertility during 2001. The proportion of pregnancies did not differ statistically between treatment and control stallions during T2-T4 in 2001 (P >0.05). Semen was available from all treatment (n=6) and control (n=4) stallions for WK1-15 of the evaluation period in 2001. However, between WK16 and WK20 motility, concentration and total number of sperm data were unavailable in some weeks for some of the stallions. Therefore the results reported for WK16-20 are based on fewer stallions in each group. Mean motility, concentration and total number of sperm were not significantly different (P = 0.888, 0.958, and 0.567, respectively) between the treatment and control stallions during the 20 week evaluation period in the 2001 season. Morphology and viability data was available for all stallions during the entire 20 week semen evaluation period. Mean normal morphology and sperm viability varied considerably during the evaluation period, however neither were different (P = 0.929 and 0.774, respectively) between the treatment and control stallions.

2001 season. Arrow indicates time that treatment and placebo were started. WK16-20 data points represent fewer than all stallions.

Discussion
Many nutrients appear to play a role in the development and function of sperm (1-9). Detailed knowledge of the role played by various nutrients or the level required for optimal development and function of sperm is not well understood. It is logical to believe that correcting a key nutrient deficiency through supplementation would improve the process. This study evaluated the effect of a nutritional supplement designed to provide a combination of ingredients to benefit the reproductively active stallion. Fertility was improved among stallions receiving the nutritional supplement over that observed during the previous breeding season. Loss of 2 control group stallions may have contributed to an inability to demonstrate a significant difference in proportion of pregnancies among treatment and control groups for the 2001 season. The fact remains that the proportion of pregnancies increased significantly during the treatment period, T2-T4 in 2001 compared with the same time period in 2000 while fertility among control stallions remained unchanged.

Conflicting anecdotal information painted a confused picture of the effect some ingredients may have on semen quality and prompted this investigation. Benefit derived from the supplement was not reflected in routine semen analyses as quality did not differ during a 20 week evaluation period between treated and control stallions. Most importantly there was no detectable decrease in quality observed.

We acknowledge the limitations of this type of study. The limited number of stallions in the control group decreases our ability to distinguish pregnancy rates among treated and control groups. Time period T5 was eliminated from some of the analyses due to the fact that it was the end of the breeding season and several stallions had few or no observations during this period. Although no specific procedures were implemented to assure the mares bred to each of the treatment groups were similar, randomization of stallions is assumed to satisfy this assumption. It was not possible within the scope of this study to define a mechanism for the improved fertility observed. Neither was identifying which ingredient or combination of ingredients affected the change. However it was a starting point. Logically it would seem there was a positive impact of the supplement on sperm development and/or function that was not detected by routine methods of semen analysis. Pursuant to this need a follow up study is underway to further define the effect of the supplement on fertility and semen quality.

The study cited above was designed by and the sample and data analysis done by a board certified theriogenologist and equine practitioner at a Veterinary Teaching Hospital. The data is being collected at a commercial breeding facility using 11 of their stallions and is supervised by their resident veterinarian. Only cycles bred on farm with corresponding on farm pregnancy examinations were used for the per cycle conception rate data [1,437 cycles over years one and two]. This selection criteria yielded a mare population that contained more problem breeders than the general population serviced by these stallions. No changes were made to the way any mares were handled at the facility. All mares were bred AI with a dose of semen that was controlled for the amount of sperm delivered. Therefore, the observed increase in total sperm and the increase in conception rate are independent effects.