Studies on the replacement policies in dairy cattle

English. <p/>In The Netherlands dairy farmers replace on average 25-30% of their cows each year. The decision to replace instead of to keep a cow is based mainly on economic considerations rather than because a cow is no longer able to produce.<p/>The investigations described in this thesis were directed towards the economic optimization of the policy for replacement and insemination of dairy cows. The following three items were treated:<br/>1. The evaluation of techniques to determine the optimum policy of inseminating and replacing dairy cows.<br/>2. The consequences for the optimum policy of variation between cows in milk production, time of conception and seasoig of calving.<br/>3. The development of a management guide for individual cows to support the farmer in making decisions concerning replacement.<p/>In the first chapter the optimization techniques which have been applied to dairy cattle replacement were described and evaluated. The optimum decision in all methods was derived essentially from the comparison of the expected revenues from the cow presently in the herd with those from her replacement at different times in the future. The latter revenues can be interpreted as the opportunity costs of keeping a cow. In the marginal net revenues approach it has to be assumed that subsequent heifers are identical. This assumption makes it impossible to account for seasonal variation in traits or prices. This limitation does not hold for the more general dynamic programming technique. Further, the technique allows the variation in, and the repeatability of, the performance of present and replacement cows to be accounted for. Most techniques offer the opportunity to rank cows according to their future profitability. Future profitability was defined as the profit expected by keeping the cow until the optimum time for replacement instead of replacing it immediately.<p/>A model was constructed for estimating the performance, revenues and costs of individual cows on a monthly basis (chapter 2). The revenues came from milk, which depended on the fat and protein content, and the sale of calves; while costs originated from feeding and sundries. In addition, the course of the carcass value and the probability of, and the financial loss associated with, involuntary disposal were considered. The parameters and prices in the model were chosen as representing the Black and White cows in The Netherlands at the normalized price level of 1981-1982. The influence on the model's results of variation in age, stage of lactation, milk production, number of days open and season of calving between cows was validated.<p/>The optimum policy for replacement of cows, taking into account the variation in milk production, was determined by dynamic programming (chapter 3). The objective in determining the optimum decision was the maximization of the expected present value of the cash flow from the present and subsequent replacement cows. In the optimization, cows were characterized by their lactation number, stage of lactation and the level of milk production during the previous and present lactations. Given the probabilities of involuntary replacement, the optimum replacement decisions corresponded with an optimum average herd life of 43 months. The optimum average herd life was greatly affected by the difference between the replacement heifer price and the carcass value of culled cows. A reduction of this difference resulted in a higher rate of voluntary replacement. The optimum replacement policy was not very dependent on the price of milk, calves or feed, the production level of the herd or the rate of genetic improvement. The financial advantage of a proportional decrease in the involuntary, replacement rate of 20% increased from 32 Dfl. to 48 Dfl. per cow per year if the variation in milk production was taken into account. This was due to an increase in the voluntary replacement rate of cows which, however, reduced the lengthening of the optimum average herd life. When no involuntary replacement occurred, the optimum average herd life of cows was 5.6 years and the average annual income per cow was 251 Dfl. higher compared to the basic situation. The increase in profit from voluntary replacement accounted for 133 Dfl. of this difference. It was concluded that variation in production should be included for a correct evaluation of the consequences of changes in the production or price situation for the optimum replacement policy.<p/>In chapter 4 the optimization was extended by taking into account also variation in the length of the calving interval. Between 2 and 7 months in lactation it was determined whether a cow should be inseminated, with a calculated probability of conception, or kept open and replaced at the optimum time during the lactation. Insemination was optimum as long as the expected loss from a prolonged calving interval was less than the loss from replacing the cow.<p/>The minimum production level for insemination to be the optimum choice increased with the time in lactation and, to a lesser extent, the age of the cow. The optimum policy resulted in an average calving interval of 371 days, while 13% of the cows had an interval of 14 months or longer. Changes in the persistence of milk production and the level of the conception rates of all cows had a significant influence on the optimum policy. Insemination of cows according to the optimum policy gave a correlation of 0.09 between the production of the cow, corrected for the influence of gestation, and the length of the calving interval. The correlation increased to 0.35 when the total 305-day production was used. The correlations agreed closely with estimates from data analysis reported in the literature. Therefore the farmer's policy to inseminate cows could explain to a great extent the reported phenotypic correlations between production and calving interval.<p/>Seasonal variation in performance and prices was found to have a great impact on the expected income and the optimum policy for inseminating and replacing cows (chapter 5). Heifers calving in October were expected to produce a 277 Dfl. higher income than those calving in June. The minimum productions for insemination to be optimum were the lowest when insemination took place from February to April. After conception the cows were expected to calve during the relatively profitable autumn months. The optimum policy resulted in an average 44 day longer calving interval for August calving heifers compared with heifers calving in March. Further it corresponded with an optimum average herd life of 37 and 46 months for heifers calving in April and October, respectively.<p/>The seasonal differences in production, feed costs and calf prices were the main sources of the differences in the expected income from heifers. The optimum policy for replacing and inseminating cows was greatly affected by the seasonal variation in production.<p/>Three alternatives to simplify the calculation of the expected cash flows, by excluding the variation in traits, were studied (chapter 6). A simulation model was constructed to determine the average performance of cows corresponding to the optimum policy resulting from each of the alternatives. The average annual income per cow was reduced by 19 Dfl. and 11 Dfl. per cow when a fixed carcass value or fixed feed costs, respectively, was used for all cows. When insemination and replacement decisions were based on the expected milk revenues, instead of on the expected income, the annual income per cow was reduced by an average of 13 Dfl. The alternatives did not affect the amount of information which has to be registered for each cow to arrive at the optimum decisions. It was therefore concluded that it is not justified to make any of the simplifications.<p/>Two management guides should be given to the farmer to help him in making decisions concerning the replacement of individual cows (chapter 7). In the first place, the future profitability should be calculated to support the decisions on immediate replacement and veterinary treatment of cows. Secondly, the expected profit in case conception of a cow takes place should be given to support the farmer in deciding whether or not to inseminate a cow. These management guides indicate the loss anticipated when taking a non-optimum decision. This expectation can be used by the farmer to weigh additional information about the cow or the herd.<p/>Two methods to calculate the management guides for individual cows were described. For the calculations information is needed on the production and reproduction status of the cows and on the expected herd situation. It was concluded that the future profitability and the expected profit in case of conception may lead to an inprovement of the farmer's income.

Dutch; Flemish. In Nederland wordt jaarlijks 25 tot 30% van de melkkoeien vervangen, waarbij het in het merendeel van de gevallen om een economische beslissing gaat, van de vervangende koe wordt meer verwacht. Het hier beschreven onderzoek richt zich op de economische optimalisatie van het beleid ten aanzien van inseminatie en vervanging van melkkoeien. Hiervoor zijn ten eerste de verschillende methoden geevalueerd voor het vaststellen van het optimale beleid voor inseminatie en vervanging van melkkoeien. Ten tweede is de invloed van verschillen in melkproduktie, tussenkalftijd en seizoen van afkalven op het optimale beleid nagegaan. Tenslotte zijn de mogelijkheden nagegaan voor de berekening van een kengetal voor elke koe, dat dient ter ondersteuning bij het nemen van de beslissingen met betrekking tot inseminatie en vervanging door de veehouder