Assessment of fertility and semen evaluations of stallions.

(1) Various estimations of motility (subjective appreciation, count on video record, Doppler laser apparatus and optic-microcomputer analyser) or percentage of live spermatozoa (eosin-nigrosin staining) showed a higher repeatability for measurements of a same sample than for straws of a same ejaculate: the values were high (respectively greater than 0.78 and greater than 0.69) except for the optical analyser and staining. (2) Semen samples were collected from 80 stallions 5 times at 24-h intervals. The repeatability varied from 0.37 to 0.69 for gel-free volume, concentration, total sperm number, % of motile spermatozoa in raw and extended semen at collection or after storage. Therefore, to assess a stallion, 2-7 ejaculates, according to the characteristic being measured, are needed to get a repeatability of 0.80 for the mean value. (3) A study on 79 stallions serving mares over 4-152 oestrous cycles per year showed that 17 seasons are necessary to measure fertility based on the pregnancy rate per oestrous cycle with the same accuracy. If they served more than 25 cycles every year, 6 seasons only were required. In both cases, it was less than for % of live foals per season.