Influence of random factors and two different staining procedures on computer-assisted sperm head morphometry in bulls

Computer-assisted sperm analysis has the potential to improve reproducibili ty and objectivity in the assessment of sperm morphology. The aim of this s tudy was to evaluate the use of a computer-assisted sperm morphometry analy sis system for the determination of sperm head dimensions in bulls. Two exp eriments were performed to determine the variability caused by random facto rs and the influence of two different staining procedures. In the first exp eriment, three ejaculates were collected from each of five clinically healt hy bulls. Air-dried semen smears were stained using a modified Farelly stai ning. The slides were observed via bright field microscopy with green filte r using a 100 x oil immersion objective. A video camera attached to the mic roscope transmitted images to a personal computer. Each sperm head was iden tified and analysed by the computer software (Morphology Analyzer V. 1.5; M ika Medical GmbH, Ismaning, Germany). Area, length and width of each sperm head were calculated and stored in a database for further statistical analy sis. A minimum of 100 sperm heads were evaluated per slide. In experiment 2 , the influence of two different staining procedures (Farelly and Papanicol aou) on sperm head dimensions was determined. The mean spermatozoal head measurements across all slides for area, length and width were 40.49 mu m(2), 9.70 mu m and 5.30 mu m, respectively. On the basis of these results, the variability between slides, ejaculates and bul ls using variance component estimation was calculated. All random factors ( bull, ejaculate and slide) had a significant effect (p < 0.001) on sperm he ad dimensions. However, the variability attributable to bull (18.89-51.72%) was considerably higher compared with that of slide and ejaculate (0.17-5. 27%). Additionally, differences existed between bulls concerning the shape and normality of histograms of their sperm head dimensions. The minimum num ber of spermatozoa required for analysis of sperm head dimensions was found to be about 60 spermatozoa per sample. The use of Papanicolaou stain resul ted in significantly smaller sperm head dimensions, e.g, sperm head area 31 .48 versus 38.35 mu m(2) (p < 0.001). In conclusion, computer-assisted sper m head morphometry provides an objective, precise and reproducible tool. Co mparisons of results from different studies should consider the influence o f random and experimental factors to avoid misinterpretation.