Br. vet.
3.
(1990). 146, 504
A ROUTINE 20-22 DAYS POSTSERVICE MILK PROGESTERONE MONITORING IN DAIRY COWS . ECONOMIC EVALUATION
O . MARKUSFELD*, H . ADLERt, N. NAHARIt and D . KASTNERt * `Hachaklait' Mutual Society for Cattle Insurance and Veterinary Services in Israel Ltd, 57 Balfour Street, ,Nahariya 22426, Israel ; tIMMB Milk Laboratory, Acco
SUMMARY Three hundred and sixty inseminated Holstein cows were monitored in a controlled trial for low milk progesterone concentration 21-23 days postbreeding . Information about low progesterone cows was given to herdsmen for the experimental group only . Of the cows in the experimental group with milk progesterone of < 4 .5 ng/ml, 56 .0% were inseminated within 2 days of monitoring compared to 28 .8% of the control animals . Of the cows reaching the 42day rectal pregnancy diagnosis, 13 .7% and 43 .6% were negative for the experimental and the control groups respectively . Inter-service intervals were 34 .5 ± 14 .5 days and 44 .3 ± 17 .7 days for the experimental and control groups respectively . Pregnancy rates were the same for 264 cows served on routine progesterone monitoring and for 318 cows bred on behavioural oestrus when matched to farm and day of breeding . A gain of US $32 .5 (an equivalent of 85 1 milk) per lactation for a cow in the monitored herd was established using the data in a model of a 250-cow herd with a total pregnancy goal of 82 .4% .
INTRODUCTION
The development of the enzyme linked immunoassay (ELISA) system for the rapid evaluation of progesterone concentration in whole milk made possible the massive field use of progesterone assay in milk, as first suggested by Laing & Heap (1971) . Practical applications suggested for the system in commercial dairy herds include the early detection of pregnancy, confirmation of oestrus, identifying problem cows before service, servicing cows without oestrus detection, confirmation of the presence of corpus luteum before injecting prostaglandins, identification of all non-cycling cows, and the prediction of the onset of behavioural oestrus from the decreased concentrations of milk progesterone (Munro et al., 1982 ; Drew, 1986 ; Stanley et al., 1986) . In a previous study by the authors (Markusfeld et al., 1988) the accuracy of pregnancy and non-pregnancy diagnosis was evaluated in all non-returning cows 21-23 days postservice . The purpose of the present study is to evaluate the economic advantages of a routine monitoring of all non-returning cows 20-22 days post-service .
PROGESTERONE MONITORING : ECONOMIC EVALUATION
505
MATERIALS AND METHODS The study was carried out in two stages . In the first part, a conducted prospective trial was carried out on seven Israeli-Holstein herds during the summer of 1986 . All cows which were served 21-23/4 days before and did not return to oestrus were sampled . Whole milk (20 ml) was taken from each cow with the first stripping removed . All samples were preserved with potassium bichromate and examined in the laboratory the same day . Progesterone monitoring was carried out using a Compur M 2000 S/SC instrument (Compur Electronic) using an 'Ovucheck' TM (Cambridge Life Science PLC, England) set . Cows were allocated before sampling into two groups according to the time interval from the last service . A list of cows with progesterone concentration of 4 .5 ng/ml from the experimental group only was given to the herdsmen, the cows presented the next day to the artificial inseminator without discriminating remarks . No information regarding the control group was given . Cows judged by the inseminator to be in oestrus were inseminated, while the others were presented for a rectal examination on the next routine weekly veterinary visit . Cows of both groups were served if returned to oestrus during the trial . A rectal pregnancy examination was carried out 40-47 days from service on all non-returning cows . Cows in the experimental group with progesterone concentration of < 4 .5 ng/rnl which were not served were consecutively examined at weekly intervals and treated for anoestrus when indicated . The second part of the study was carried out in the years 1988 and 1989 on three farms . The farms were characterized by a wide non-return to pregnancy rates ratio (> 3 : 1) . Milk was collected by herdsmen from all non-returning cows 20-22 days post-breeding and examined in the laboratory . A list of all cows having a progesterone concentration of 5 1 .0 ng was given to the herdsmen ; cows were then treated in the manner described above . Monitoring was carried out three times weekly . A model was computed from the data gathered in the two studies, so that the economical evaluation of a routine monitoring could be made . `Status waste days' (SWD), the sum of days of first service to conception intervals, was calculated for a 250(ow size herd with a total pregnancy goal of 82 .4/,, . SWI) were compared for monitored versus non-monitored herds according to the following :
J SWD=AC X SI /N-nj i where : SWD=status waste days, sum of days from insemination to conception for all conceiving cows ; AC=inter-insemination interval (days) ; .N==number of cows in the herd ; j=number of iterations (cycles) (0, 1, 2 . . .) ; n=number of pregnant cows ; i=insemination number (1st, and > Ist) ; and :
nj =nj - 1 -±-JN-nj-1JXPRi where : PR=pregnancy rate . The total number of progesterone assays was computed as follows :
la =ti X NR
506
BRITISH VETERINARY JOURNAL, 146, 6
where : to=number of progesterone assays ; ti=total cows inseminated ; NR=non-return rate in 21 days post-breeding .
Scope and statistical analysis Two hundred and sixty-two non-returning out of 360 inseminated cows were monitored in the first trial and were used for the evaluation of the effect of monitoring on the inter-service interval . Data were analysed by routine `u ' and `t ' tests for the significance of differences between proportions and between means respectively (Arkin & Colton, 1950) . Two hundred and sixty-four out of 1154 monitored cows with low milk progesterone concentrations and 318 cows bred at behavioural oestrus were used to evaluate the respective overall pregnancy rates of the two groups . Data were analysed by the method described by Mantel & Haenszel (1959) for the evaluation of odds ratio, cows matched to farm and day of breeding .
RESULTS Of 1564 cows from the farms studied which reached the 42-day rectal pregnancy diagnosis in the years 1988 and 1989, 25 .3% were negative . Pregnancy rate to first service in those years was 36 .0% for 967 cows inseminated . Of 360 cows in the first trial, 72 .8% did not return to oestrus by 21 days post-service . Of 62 cows in the experimental group with progesterone level of S 4 .5 ng/ml, 56 .5% were served within 2 days from monitoring compared to 28 .8% of 66 cows of the control group (P< 0.0001) . The mean inter-service intervals were 34 .5 ± 14.5 and 44 .3 ± 17 .7 days for the experimental and control groups respectively, the difference in days (9 .8) being highly significant (P
3.
(1990). 146, 504
A ROUTINE 20-22 DAYS POSTSERVICE MILK PROGESTERONE MONITORING IN DAIRY COWS . ECONOMIC EVALUATION
O . MARKUSFELD*, H . ADLERt, N. NAHARIt and D . KASTNERt * `Hachaklait' Mutual Society for Cattle Insurance and Veterinary Services in Israel Ltd, 57 Balfour Street, ,Nahariya 22426, Israel ; tIMMB Milk Laboratory, Acco
SUMMARY Three hundred and sixty inseminated Holstein cows were monitored in a controlled trial for low milk progesterone concentration 21-23 days postbreeding . Information about low progesterone cows was given to herdsmen for the experimental group only . Of the cows in the experimental group with milk progesterone of < 4 .5 ng/ml, 56 .0% were inseminated within 2 days of monitoring compared to 28 .8% of the control animals . Of the cows reaching the 42day rectal pregnancy diagnosis, 13 .7% and 43 .6% were negative for the experimental and the control groups respectively . Inter-service intervals were 34 .5 ± 14 .5 days and 44 .3 ± 17 .7 days for the experimental and control groups respectively . Pregnancy rates were the same for 264 cows served on routine progesterone monitoring and for 318 cows bred on behavioural oestrus when matched to farm and day of breeding . A gain of US $32 .5 (an equivalent of 85 1 milk) per lactation for a cow in the monitored herd was established using the data in a model of a 250-cow herd with a total pregnancy goal of 82 .4% .
INTRODUCTION
The development of the enzyme linked immunoassay (ELISA) system for the rapid evaluation of progesterone concentration in whole milk made possible the massive field use of progesterone assay in milk, as first suggested by Laing & Heap (1971) . Practical applications suggested for the system in commercial dairy herds include the early detection of pregnancy, confirmation of oestrus, identifying problem cows before service, servicing cows without oestrus detection, confirmation of the presence of corpus luteum before injecting prostaglandins, identification of all non-cycling cows, and the prediction of the onset of behavioural oestrus from the decreased concentrations of milk progesterone (Munro et al., 1982 ; Drew, 1986 ; Stanley et al., 1986) . In a previous study by the authors (Markusfeld et al., 1988) the accuracy of pregnancy and non-pregnancy diagnosis was evaluated in all non-returning cows 21-23 days postservice . The purpose of the present study is to evaluate the economic advantages of a routine monitoring of all non-returning cows 20-22 days post-service .
PROGESTERONE MONITORING : ECONOMIC EVALUATION
505
MATERIALS AND METHODS The study was carried out in two stages . In the first part, a conducted prospective trial was carried out on seven Israeli-Holstein herds during the summer of 1986 . All cows which were served 21-23/4 days before and did not return to oestrus were sampled . Whole milk (20 ml) was taken from each cow with the first stripping removed . All samples were preserved with potassium bichromate and examined in the laboratory the same day . Progesterone monitoring was carried out using a Compur M 2000 S/SC instrument (Compur Electronic) using an 'Ovucheck' TM (Cambridge Life Science PLC, England) set . Cows were allocated before sampling into two groups according to the time interval from the last service . A list of cows with progesterone concentration of 4 .5 ng/ml from the experimental group only was given to the herdsmen, the cows presented the next day to the artificial inseminator without discriminating remarks . No information regarding the control group was given . Cows judged by the inseminator to be in oestrus were inseminated, while the others were presented for a rectal examination on the next routine weekly veterinary visit . Cows of both groups were served if returned to oestrus during the trial . A rectal pregnancy examination was carried out 40-47 days from service on all non-returning cows . Cows in the experimental group with progesterone concentration of < 4 .5 ng/rnl which were not served were consecutively examined at weekly intervals and treated for anoestrus when indicated . The second part of the study was carried out in the years 1988 and 1989 on three farms . The farms were characterized by a wide non-return to pregnancy rates ratio (> 3 : 1) . Milk was collected by herdsmen from all non-returning cows 20-22 days post-breeding and examined in the laboratory . A list of all cows having a progesterone concentration of 5 1 .0 ng was given to the herdsmen ; cows were then treated in the manner described above . Monitoring was carried out three times weekly . A model was computed from the data gathered in the two studies, so that the economical evaluation of a routine monitoring could be made . `Status waste days' (SWD), the sum of days of first service to conception intervals, was calculated for a 250(ow size herd with a total pregnancy goal of 82 .4/,, . SWI) were compared for monitored versus non-monitored herds according to the following :
J SWD=AC X SI /N-nj i where : SWD=status waste days, sum of days from insemination to conception for all conceiving cows ; AC=inter-insemination interval (days) ; .N==number of cows in the herd ; j=number of iterations (cycles) (0, 1, 2 . . .) ; n=number of pregnant cows ; i=insemination number (1st, and > Ist) ; and :
nj =nj - 1 -±-JN-nj-1JXPRi where : PR=pregnancy rate . The total number of progesterone assays was computed as follows :
la =ti X NR
506
BRITISH VETERINARY JOURNAL, 146, 6
where : to=number of progesterone assays ; ti=total cows inseminated ; NR=non-return rate in 21 days post-breeding .
Scope and statistical analysis Two hundred and sixty-two non-returning out of 360 inseminated cows were monitored in the first trial and were used for the evaluation of the effect of monitoring on the inter-service interval . Data were analysed by routine `u ' and `t ' tests for the significance of differences between proportions and between means respectively (Arkin & Colton, 1950) . Two hundred and sixty-four out of 1154 monitored cows with low milk progesterone concentrations and 318 cows bred at behavioural oestrus were used to evaluate the respective overall pregnancy rates of the two groups . Data were analysed by the method described by Mantel & Haenszel (1959) for the evaluation of odds ratio, cows matched to farm and day of breeding .
RESULTS Of 1564 cows from the farms studied which reached the 42-day rectal pregnancy diagnosis in the years 1988 and 1989, 25 .3% were negative . Pregnancy rate to first service in those years was 36 .0% for 967 cows inseminated . Of 360 cows in the first trial, 72 .8% did not return to oestrus by 21 days post-service . Of 62 cows in the experimental group with progesterone level of S 4 .5 ng/ml, 56 .5% were served within 2 days from monitoring compared to 28 .8% of 66 cows of the control group (P< 0.0001) . The mean inter-service intervals were 34 .5 ± 14.5 and 44 .3 ± 17 .7 days for the experimental and control groups respectively, the difference in days (9 .8) being highly significant (P
