STATISTICAL-METHODS TO PRODUCE GOOD BOVINE FROZEN-SEMEN

A ''good'' bovine frozen semen must have: high fertility, low costs, g ood genetic values. The genetic value is associated to the sire and th e Artificial Insemination (AI) centre can only optimize the management of the progeny test program. The high fertility and the low costs are strictly linked to the qualitative level of the technical management of the AI centre. The optimization of the management of an AI centre c an be done by the choice of the best systems of semen handling and ana lysis, using the methods of the Statistical Quality Control (SQC). To support the production in an AI centre, it becomes important to set up some statistical instruments aiming to performing a Statistical Quali ty Control. These instruments (such as Control Charts) should allow th e monitoring of the production because the various lots, produced by t he same centre, should have a basically constant number of motile sper matozoa after thawing (the quality parameter with the better correlati on with the fertility). Moreover, this number should range within the tolerance limits established by the centre. Only in this way the produ ction activity of an AI centre may become of industrial type. The syst em that aims to maximize the production level, in an AI centre, as a f unction of the number of motile spermatozoa after thawing, consists of predicting forward motility and concentration, after thawing, of a ba tch of semen and using these values to compute the correct number of s traws to produce for the batch (for each bull). The ARIMA models can b e used to forecast the percentage of motile sperms after thawing for b ulls with continue production, while to forecast the concentration it is necessary to verify the loss of spermatozoa during the semen handli ng.