Studies on fertility in dairy cattle based on analysis of AI data

英语. Fertility is one of the non-yield traits which is of great economic importance in dairy herds. Reduced fertility results in prolonged calving intervals and an increased culling rate. Both are undesirable. Fertility problems account for about 30% of all disposals.In this thesis it was researched whether the use of routinely collected and processed AI data could be improved both for breeding and management purposes. Much research has been done on breeding aspects of fertility. The main items were summarised in the literature review in chapter one. Three aspects were considered for further research.Firstly the correction of insemination results of technicians and Friesian bulls in AI was studied. Data were available on 87112 first inseminations by 283 bulls and 37 technicians. The traits studied were 28 and 56 days non-return rates. In the analyses a distinction was made among a monthly and a yearly presentation of results. Differences in age at first service in heifers and calving to first service interval in cows did not affect the ranking of and differences between sires and technicians. The routine monthly evaluation of non-return results of sires required a correction for parity of inseminated animals and month of insemination. Also the percentage HosItein was of interest. Deviations from the stud average had to be regressed to account for sampling. In obtaining a reliability of 50% for sires at least 300 first inseminations were required. The monthly evaluation of technicians could be based on deviations within studs.The annual avaluation of both technicians and sires could be further improved if the herd effect was also accounted for. The correlation between solutions corrected for and not corrected for the herd effect was 0.94. After correction for environmental effects the range in solutions for 56 days non-return rate of sires and technicians was about 9%.Secondly the relationship between insemination results of sires and of their daughters was researched. The success of an insemination is influenced by the quality of the sperm the sire produces (direct effect) as well as the environment in the uterus and the quality of the egg. The latter are affected by the sire indirectly via genes transmitted to his daughters. The traits considered were 56 days nonreturn rate, conception rate, days open in cows and age at conception in heifers. Traits were analysed by parity. The number of records was 6497, 3704, 6883 and 4711 animals in heifers and cows in parity 1, 2 and 3, respectively.The relationship between the direct and the indirect effect of the sire changed with the age of the cow: from -0.29 in heifers to 0.59 in parity 3 for 56 days non-return rate and from -0.74 to 0.48 for conception rate. Changes in days open were smaller and not consistent. It was concluded that use of sires with non-return rates above average would be favourable for the insemination results of older cows, but somewhat adverse on results in heifers.The direct and indirect effect can be further distinguished among a direct genetic and maternal genetic effect. The first concerns the genetic make up of the embryo, the second the genetic value of the environment in which conception takes place. The observed interaction for the direct and indirect effect with parity number, was also found for the direct genetic and maternal genetic effect. The negative relationships in heifers were even stronger. The cause for this negative relation is not known. The maternal and direct genetic component were about equal for 56 days non-return and conception rate. Days open showed a larger maternal component.Chapter 4 deals with the relationships between fertility traits in sire progeny groups at different ages. Sire evaluation and selection for fertility based on records of heifers and first parity cows is only effective if the relation with fertility at older ages is high. Insemination results decrease at older ages and culling due to fertility problems increases. The number of records available to study these questions varied between 2307 and 12145 records per parity.The genetic correlation between age at first insemination in heifers and interval to first service in first parity cows was 0.67, that for age at conception and days open 0.66. The genetic relations between interval to first service in parity 1, 2 and 3 respectively varied from 0.78 to 0.89, that for days open from 0.68 to 1.06. The genetic correlations between 56 days non-return in different parities were above unity.The results of chapter 3 and 4 seemed to be conflicting. In chapter 3 the relationship between 56 days non-return of the sire and that of his progeny changed with age, whereas in chapter 4 it was found that the genetic relationship between different parities was one. Further analyses as discussed in the general discussion chapter showed that when the genetic correlation between parities was about 0.75, the realised change in the genetic relation between the direct and indirect effect was possible. If the relationship between parities was higher the possible range should be smaller.In the discussion chapter it is also discussed to what extent selection for daughter fertility would be possible. The trait 56 days non-return rate would be the best to select for, because this is less biased by decisions of dairy farmers. The traits days open and number of inseminations are also affected by the production capacity of the cow. However, current size of progeny groups in first lactation is about 75 to 100 daughters, which is too low to detect sufficiently large differences between daughter groups for non-return rate.Finally it was studied in which way AI data could be used to monitor herd fertility. Data of 65 herds collected over a period of 3 years, were used. The following herd averages were calculated to characterise herd fertility: interval calving to first service, 56 days non-return rate after first service, percentage reinseminations done in the period 18-24 days after a previous service, fertility status, oestrus index (a measure for the quality of heat detection) and number of inseminations per average cow present in the herd.These figures were related to loss from prolonged calving interval and loss from culling due to fertility problems. The oestrus index and the fertility status showed the highest correlation with loss from prolonged calving intervals: 0.79 and -0.67, respectively. Interval to first insemination was next with 0.52. Relationships with loss from culling were never significant. The oestrus index, interval calving to first service and 56 days non-return rate were also rather repeatable: 0.63, 0.58 and 0.52, respectively. The repeatability of loss from prolonged calving intervals could be characterised well with three figures. The oestrus index was the most important one, explaining 63% of the differences, the interval to first service explained a further 10% and 56 days non-return rate a further 6%. Loss from culling could not be characterised by figures calculated from AI data. It was concluded that the oestrus index, the interval calving to first service and 56 days non-return rate were to be presented to the farmer as management aids.

荷兰语； 佛兰德语. Nagegaan is op welke wijze KI-gegevens, die routinematig verzameld en bewerkt worden, beter aangewend kunnen worden voor zowel foktechnische als management-doeleinden. De auteur concludeert o.a. dat bij de presentatie van de resultaten van KI-stieren rekening gehouden zou moeten worden met effect van seizoen, bedrijf en leeftijd van de geinsemineerde koe. Het gebruik van goed bevruchtende stieren blijkt ietwat nadelig te kunnen uitwerken op de bevruchtingsresultaten van de nakomelingen als pink, maar juist positief op de resultaten bij oudere koeien. Uit het onderzoek op welke wijze de KI-gegevens beter benut zouden kunnen worden voor de karakterisering van de vruchtbaarheidssituatie op de melkveebedrijven blijkt dat door koppeling van deze gegevens aan de melkcontrole-gegevens een aantal kengetallen routine-matig ten behoeve van het bedrijfsbeheer berekend kunnen worden