Pregnancy testing of cows by determining progesterone levels in milk

Abstract

In six herds, each comprising 400 to 600 cows, the progesterone test (PT) was compared at different numbers of milk samplings from one cow and at a reduced number of days of milk sampling per week (from seven to two per week). There were three variants. In four herds the foremilk was sampled on the 23rd day after insemination in the period from October 1978 to September 1979 (variant A). In the remaining two herds the samples were taken from the whole amount of milk from single milking in May to July 1979 (variant B) and in August to November 1979 (variant C). In variant B the samples were taken on the day of insemination or the first day after insemination, and on the 19th and 23rd days after insemination. In variant C also three samples were taken after each insemination, but only on two days in the week, Monday and Thursday (1st sampling 19–22, 2nd sampling 22–25 and 3rd sampling 29–32 days after insemination). A radioimmunological method was used to determine progesterone (tritium-labelled progesterone). Pregnancy was diagnosed by rectal palpation between the 60th and 90th days after insemination. In comparison with a single sampling, the use of two or three samples after insemination in variants B and C did not improve the proportion of correct results in the group of non-pregnant cows (progesterone level ⩽ 2.4 ng ml −1) nor in the overall results. Of the clinically identified non-pregnant cows, single milk sampling correctly determined 42.7 and 43.0%, the combination of two samplings 58.9 to 62.3%, and the combination of three samplings 74.7% of the cows. The differences were significant ( P < 0.01). The reduction of sampling days from seven to two in a week did not influence adversely the proportion of the results which were in agreement in the PT and clinical examination.