Ardilasunu Wicaksono

To sustain optimal production levels in dairy cows, reproductive performance is an important factor. Nevertheless, reproductive diseases commonly occur on dairy farms, posing an impact to reproductive performance (e.g., a reduced pregnancy rate and a longer calving interval), cow productivity (e.g., a reduced milk production and an increased risk of culling), and farm profitability (economic losses associated with reduced cow productivity and treatments). Several reproductive diseases can occur in dairy cows, predominantly occurring from calving until the end of the post-partum period, when cows are re-inseminated. For instance, dairy farmers generally initiate breeding after the voluntary waiting period ends, but post-partum ovarian disorders may occur during this critical phase of the lactation. To mitigate those reproductive diseases, hormonal treatments can be used. In addition to their therapeutic applications, hormones can also be administered to induce oestrus in cows. However, the use of hormones in dairy farming does generate societal concerns, particularly regarding the demand to reduce medicine use and to produce naturally. These societal developments necessitate increased transparency regarding the application of hormonal treatments for fertility in dairy farms. Furthermore, the routine reliance on hormones may obscure underlying reproductive management issues, potentially overlooking the fundamental causes that need to be addressed.

Reproductive hormones are used in dairy farms to improve cow reproductive performance, but the evidence on their effectiveness remains inconclusive. The effectiveness of reproductive hormones is commonly observed at the cow level in randomized clinical trials, while their impact at the herd level under field conditions has not been investigated. The overall herd level effect of hormone use on reproductive performance may differ from the cow level, because the treatment effect of problematic cows might be alleviated by the reproductive performance of healthy cows. In addition, to investigate the use of hormones in dairy farms, it is crucial to explore farmers' mindsets and behaviour related to hormone use. This will help to understand the extent of their knowledge about the hormones farmers use, their attitude for using them, and whether their mindsets influence their behaviour regarding hormone use. Additionally, the occurrence of reproductive disorders implies the needs dairy farmers for preventive measures. The estimation of the economic impact of interrelated reproductive disorders can provide useful insights for dairy farmers regarding the prevention of reproductive disorders. Besides the evaluation of the economic impact of reproductive disorders, it is essential to assess economic impact of the use of hormones. Evaluating this will provide insights for dairy farmers to decide which reproductive management strategy provides them the highest economic benefit in relation to their hormone use intensity.

Hence, the aim of this thesis was to support decision-making on the use of reproductive hormones to treat reproductive diseases in Dutch dairy farms, through an integrated epidemiological and socio-economic analysis. Therefore, four sub-objectives were specified to reach the aim of this thesis:

1. To investigate the herd level association of reproductive hormone use with reproductive performance in Dutch dairy herds.
2. To determine and associate dairy farmers’ knowledge and attitude regarding the use of reproductive hormones with their hormone use behaviour.
3. To estimate the costs of reproductive disorders, including dystocia, post-partum puerperal disorders and post-partum ovarian disorders, in dairy farms.
4. To compare the economic impact of reproductive management programs applying systematic hormonal treatments to individual cows with a reproductive management program applying cow-specific hormonal treatments based on a veterinary diagnosis.

In Chapter 2, an ecological study was conducted to associate hormone use for reproductive diseases and oestrus induction with reproductive performance at herd level. Hormone sales data from 754 representative Dutch dairy farms belonging to 5 large veterinary practices from 2017 to 2019 (1,679 observations in total) were utilized. Hormone use was classified into prostaglandin, gonadotropin-releasing hormone, and progesterone, and was expressed at herd level as the annual number of hormone doses per 100 adult dairy cows. Hormone use was categorized into 4 levels (no usage, low, medium, and high use), following the 33 and 66 percentiles of herds that applied them. Three herd-level reproductive performance indicators (calving interval, calving to first insemination interval, number of inseminations per cow) were analysed using multivariable General Estimating Equations models. The median annual total hormone use was 36.1 doses per 100 adult dairy cows in all herds while the median was 39.2 doses per 100 adult dairy cows among the herds that applied hormones. The final statistical models identified that herds with a high hormone use had a calving interval and a calving to first insemination interval that was 9.3±2.6 and 16.4±2.1 days shorter than that of non-user herds (424.0±2.7 and 114.0±2.1 days), respectively. Furthermore, high-user herds needed on average 0.3±0.04 inseminations more to get their cows pregnant compared to non-user herds (1.83±0.04 number of inseminations per cow). In conclusion, using a large representative herd-level dataset, hormone use was associated with a better reproductive performance in terms of calving interval and calving to first insemination interval but gave extra average number of inseminations per cow.

In Chapter 3, an online questionnaire among Dutch dairy farmers was undertaken to determine and associate dairy farmers’ knowledge and attitude with their reproductive hormone use behaviour. Data on herd characteristics, knowledge, attitude and hormone use behaviour were collected. Hormone use behaviour data were determined by self-reported based data on the questionnaire and actual hormone sales data. A knowledge score (0-10) was determined based on 10 objective knowledge questions. An explanatory factor analysis was conducted to identify latent structures among 19 attitude variables, resulting in 4 factor variables. Those attitude factors were defined as the attitude towards the benefits of using hormones, the attitude towards following a treatment protocol, the attitude towards non-hormonal fertility management, and the attitude towards barriers of using hormones. Several regression analyses were conducted to associate knowledge with attitude (multivariate multiple regression analysis), and knowledge and attitude with self-reported behaviour (logistic regression analysis) and with actual hormone use (negative binomial regression analysis). All models were corrected for 8 herd characteristic variables. The mean knowledge score was 5.9 ± 1.8 and 83% of the farmers indicated that they used hormones if their cows were not in oestrus after exceeding some expected days in milk. No significant association was found between knowledge and the 4 attitude factors. However, a higher knowledge score was positively associated with hormone use behaviour (self-reported and actual hormone use). The same direction was shown for a stronger attitude towards the benefits of using hormone with both higher hormone use behaviours. A stronger attitude towards following the treatment protocol was positively associated with a higher actual hormone use while a stronger attitude towards non-hormonal fertility management was associated with a lower actual hormone use.

In Chapter 4, a stochastic bio-economic simulation model of a 200-cow herd with daily time steps was developed to estimate the costs of dairy cow reproductive disorders. The simulation model included dystocia, retained placenta, acute metritis, chronic endometritis, anovulation, cystic ovarian disease and sub-oestrus, thereby taking into account their complex interrelations. The parameterization of the probabilities on developing reproductive disorders were based on scientific literature and expert opinion. Nine scenarios were simulated and included a 1) default scenario where all reproductive disorders were included in the simulation model. The second to eighth scenarios were simulated with zero (0) probability of each specific disorder, including 2) dystocia, 3) retained placenta, 4) acute metritis, 5) chronic endometritis, 6) anovulation, 7) cystic ovarian disease, and 8) sub-oestrus. In the ninth scenario, all disorders were absent. The annual net economic return (NER) of the herd was calculated for all the scenarios. Subsequently, the NER of the scenarios with zero probability of one of the disorders (scenarios 2-9) were compared to the NER of the default scenario and its difference was considered the cost of each specific reproductive disorder (or of all 7 reproductive disorders combined). This study showed that taking into account all disorders resulted in the mean annual cost of €100 per average cow in the herd per year. At herd level, the highest mean annual cost was observed for acute metritis (€30/cow/year) while the lowest mean annual cost was observed for dystocia (€4/cow/year). The highest cost per case was observed for acute metritis (€257/case) while the lowest cost per case was observed for cystic ovarian disease (€58/case). Given the interrelationships among reproductive disorders, preventing one disorder could reduce the occurrence of others, thereby lowering their overall economic impact.

In Chapter 5, using an adapted version of the simulation model of Chapter 4, the economic impact of several hormone-based reproductive management programs was evaluated. The updated simulation model included systematic hormonal treatments to individual cows with a specific day in milk. It also included a reproductive management program using cow-specific hormonal treatment based on a veterinary diagnosis during a fertility check. Four hormone-based reproductive management programs were modelled: 1) default reproductive management program reflects the current reproductive management of Dutch dairy herds, 2) Fixed Time Artificial Insemination (FTAI) reproductive management program, 3) FTAI with detection of oestrus reproductive management program (FTAIHED), and 4) detection of oestrus followed by timed AI (EDHTAI) reproductive management program. The NER was calculated for each of the 4 reproductive management programs and compared to a default reproductive management program. The highest NER was observed for the FTAIHED reproductive management program with €23,764 higher net revenues, followed by the FTAI and the EDHTAI reproductive management programs with €19,550 and €14,314 higher net revenues, respectively. Overall, systematic hormone-based reproductive management programs gave higher costs due to the higher use of hormones administered as well as higher calving and feed costs per year due to more pregnant cows. Nevertheless, the additional revenues of milk and calves in the systematic hormone-based reproductive management programs outweighed the cost of intervention. Reproductive management programs where hormones were systematically used gave economic advantages over the current default reproductive management program in which hormones are administrated to individual cows based on a veterinary diagnosis of ovarian disorders during a fertility check.

In the general discussion (Chapter 6), the results of the previous chapters are synthesized, and the used data and methods are discussed. Future directions to use reproductive hormones in Dutch dairy farms are described in this chapter, suggesting that dairy farmers continue implementing reproductive management programs that reasonably apply reproductive hormones, based on veterinary advice and in alignment with the fertility guideline's recommendations. Potential future research in the Netherlands regarding the use of reproductive hormones and reproductive disorders are described. Additionally, the future insights of this topic for smallholder dairy farmers are discussed.

Based on all the findings, the main conclusions of this thesis are:

1. At herd level, an increased hormone use was associated with a better reproductive performance in terms of shorter calving interval and calving to first insemination interval but with an extra number of inseminations per cow (Chapter 2).
2. Farmers’ knowledge was not associated with their attitude towards hormone use. Meanwhile, farmers’ knowledge and a stronger farmers’ attitude on benefits of using hormones and on following the hormonal treatment protocol were positively associated with a higher hormone use (Chapter 3).
3. A stronger farmers’ attitude towards non-hormonal fertility management was associated with a lower hormone use (Chapter 3).
4. By taking into account their complex interrelations, the mean annual cost of dairy cow reproductive diseases was €100/cow/year. At herd level, the highest mean annual cost was observed for acute metritis at €30/cow/year, while the lowest mean annual cost was observed for dystocia at €4/cow/year. In terms of cost per case, the highest cost was observed for acute metritis (€257/case) while the lowest cost was observed for cystic ovarian disease (€58/case; Chapter 4).
5. Reproductive management programs where hormones were systematically used gave economic advantages over the reproductive management program in which hormones are administered to individual cows based on a veterinary diagnosis during a fertility check (Chapter 5).

Overall, in comparison with the current use of hormones in the Netherlands, an increased use of reproductive hormones improves cow reproductive performance, increases milk production, and reduces culling rates, thereby offering economic benefits to dairy farmers. However, a farmer’s decision to use more hormones must be balanced with their preference to implement non-hormonal fertility management. Moreover, these decisions should also meet societal concerns regarding the use of hormones and their preference for natural dairy production practices. Therefore, in order to balance economic and ethical considerations regarding reproductive management in dairy herds, fertility guidelines should align the use of hormones in reproductive programs with societal demands.