Achmad Fadillah

studies with bio-economic simulation modelling, it addresses critical knowledge gaps and provides practical solutions for smallholder dairy systems. Each chapter contributes to a better understanding of the challenges and solutions related to milk quality and mastitis in smallholder dairy farms, moving from a better awareness of the actual on-farm milk quality to the implementation of cost-effective strategies for improving udder health management. The following is a summary of the key findings from Chapters 2-6 of this thesis.

Chapter 2 highlights the crucial role dairy cooperatives play in enhancing Indonesian smallholder dairy farmers' awareness of key milk quality parameters. The study found that factors such as the milk cooperative, distance to other farmers, technology adoption, buyer priority on total plate count (TPC), milk production information, and cow health information were associated with farmers' awareness of TPC. Additionally, milk cooperatives, dairy business experience, and milk quality test adoption were associated with farmers' awareness of somatic cell count (SCC). Chapter 3 identified several factors associated with SCC, including manure removal frequency, mastitis treatment training, washing udders with soap, the number of workers on the farm, and pasture area ownership. Furthermore, manure removal frequency and the contribution of dairy farming to household income were associated with TPC. The economic analyses in Chapter 4 reveal that improved milk quality, better udder health, a higher ratio of adult cows per livestock unit, and more frequent mastitis treatment trainings were associated with increased gross margins. Moreover, improved udder health and more frequent mastitis treatment trainings were also linked to increased technical efficiency. Bio-economic modelling in Chapter 5 underscores the significant economic burden of mastitis on smallholder farms, emphasizing the need for effective control strategies to enhance profitability. Among the evaluated strategies, post-milking teat disinfection emerged as the most cost-efficient intervention for reducing mastitis incidence and associated costs. Finally, Chapter 6 showed that monitoring SCC at the milk collection point is a feasible option to provide dairy farmers with information on their udder health. Although the diagnostic performance (in terms of sensitivity and specificity) of a theoretical future diagnostic equipment was lower than other measurement equipment, their relatively low costs make them an interesting option for SCC monitoring for smallholder dairy farms. Based on these key findings of Chapters 2-6, the next section synthesizes the results presented in this thesis.

Chapter 2 examined milk quality awareness and its associated risk factors among Indonesian smallholder dairy farmers, revealing that their awareness of milk quality parameters was generally low. Only 44% of these smallholder farmers were aware of TPC and 4% of SCC, with limited knowledge of the actual values. This finding underscores the need for cooperative-led programs focused on enhancing farmers' awareness of milk quality to make subsequent milk quality improvement efforts (Chapter 3) more impactful. To enhance awareness, cooperatives should prioritise extension programs that educate farmers about the relationship between milk production, milk quality, and animal health (Groot and van ’t Hooft, 2016). Furthermore, effective communication strategies, particularly peer communication, can positively influence farmers’ mindsets and behaviours regarding animal health management (Lam et al., 2011, 2017; Jansen and Lam, 2012; Tschopp et al., 2015).

A thorough understanding of the risk factors associated with milk quality and mastitis is essential for designing targeted interventions. For example, research by McDougall et al., (2009) emphasizes that identifying specific risk factors, such as farm hygiene and animal health management, enables more tailored control measures. Knowledge of these factors aids in developing farm-specific interventions that align with the unique challenges faced by individual farms, as risk factors can vary widely depending on environmental conditions and management practices (Green et al., 2007). Chapter 3 explores the relationship between risk factors and milk quality parameters, specifically SCC and TPC, in smallholder dairy farms in Indonesia. By identifying baseline risk factors, we were able to provide insights that form a foundation for improving milk quality and udder health in these farms. Proper hygiene practices are crucial for improving udder health, reducing the risk of mastitis, and minimising bacterial contamination in milk (Ingawa et al., 1992; Gleeson et al., 2009). By implementing these practices, it is possible to improve both milk quality and udder health in smallholder dairy farms. However, because of the very different production circumstances of smallholder dairy farms in tropical circumstances, we should be careful to extrapolate the findings of the majority of risk factor studies, carried out in high-income countries towards low and middle-income countries. Our epidemiological analysis of risk factors for Indonesian dairy farmers provided a foundation for the economic impact analysis conducted in Chapter 4 by offering a clear understanding of the underlying causes of milk quality and udder health issues. The results also provide a data-driven basis for targeting interventions, which formed the basis for designing udder health improvement strategies used in the modelling study in Chapter 5, aimed at reducing udder health issues and enhancing milk quality.

The economic impact of mastitis and its control is another important area of research. For instance, Gill et al., (1990) and Richardet et al., (2023) underscore the potential for mastitis control practices to enhance milk yield, improve milk quality, and generate economic benefits. The decision-making process regarding mastitis control requires an understanding of both mastitis impact and its associated control costs (Halasa et al., 2007; Yalcin and Stott, 2000). Building on the epidemiological study in Chapter 3, Chapter 4 highlights that better farm management may lead to improved udder health and as a result enhanced economic performance. This aligns with previous studies (Skevas and Cabrera, 2020; Huijps et al., 2010), which found an association between udder health management and both technical and cost efficiency. These economic findings complement the epidemiological findings from Chapter 3, providing insights to smallholder dairy farmers and emphasizing the importance of milk quality and udder health. This integration of epidemiological and economic analysis reflects the thesis’s interdisciplinary approach.

Prevention and treatment programs taking into account the local circumstances are crucial for reducing the incidence of mastitis and its associated impacts (Ruegg, 2017; Green et al., 2007; Dingwell et al., 2003). Various empirical and simulation studies have explored mastitis interventions, including hygiene practices, treatment methods, and dry cow management (Stevens et al., 2016; Beekhuis-Gibbon et al., 2011; Allore and Erb, 1998). Omore et al., (1999) highlight the economic benefits of adopting these interventions in smallholder dairy farms with limited resources, based on their empirical study. However, there is a knowledge gap regarding the cost-efficiency of these interventions through simulation studies, specifically in the tropical circumstances of Asian countries, where smallholder farmers and cooperatives play a major role in dairy production and supply chains. The findings of the empirical studies (Chapters 3 and 4) culminate in Chapter 5 further underscoring the significant economic burden of mastitis on Indonesian smallholder dairy farms. The total costs of mastitis were 14% (€175 per farm per year) of the annual gross margin (€1,271 per farm per year; Chapter 4). This may rise to 36% (€461 per farm per year) when considering the 95th percentile of mastitis costs. These findings highlight the critical need for efficient control strategies in smallholder dairy farms. Among the evaluated strategies, post-milking teat disinfection emerged as the only cost-efficient strategy, reducing the incidence of clinical cases by 39%, subclinical mastitis cases by 38%, milk production losses by 38%, SCC at the farm level by 21%, and associated costs of mastitis by 15% compared to the default control strategy. These percentage reductions represent the epidemiological and economic benefits of implementing post-milking teat disinfection. Post-milking teat disinfection has been shown to be an effective practice for improving udder health, milk quality, and the economic benefits of dairy farms (Martins et al., 2017; Huijps et al., 2010). However, this practice is not widely adopted on smallholder farms in Indonesia, highlighting the potential for improvement and increased farm profitability through its implementation.

Implementing udder health monitoring programs allows farmers to detect and address mastitis promptly (Zigo et al., 2021). Hillerton et al. (1995) and Britten (2012) illustrate how regular monitoring and diagnostic tools can enhance control efforts, leading to better milk quality and udder health. Similarly, Nielsen and Emanuelson (2013) and Down et al. (2016) discuss the role of performance indicators in guiding effective intervention decisions. SCC is widely recognized as a reliable indicator of milk quality and udder health (Ruegg and Pantoja, 2013; Schukken et al., 2003; van Schaik et al., 2002). Regular SCC monitoring provides critical insights for the early detection of subclinical infections, enabling timely interventions to prevent the deterioration of udder health (Neculai-Valeanu and Ariton, 2022; Jayarao et al., 2004). However, almost all studies and implementations of SCC monitoring systems have been conducted in high-income countries. Chapter 6 of this thesis addresses this gap by evaluating the cost-effectiveness of SCC monitoring systems for smallholder dairy farmers in a lower-middle-income country, Indonesia, where affordability and accessibility of these systems are key considerations. Using the farm-level bio-economic modelling approach of Chapter 5, a novel stochastic simulation model was developed in Chapter 6. This simulation model was used to evaluate the cost-effectiveness of SCC monitoring and the role of dairy cooperatives in collectively managing these systems. Because of a lack of suitable performance data for SCC monitoring equipment, the performance of two types of diagnostic equipment, with differing accuracy and precision as well as differing costs, were estimated as good as possible. The study shows that SCC monitoring using a theoretical future diagnostic equipment should be possible and affordable for smallholder dairy farms. The SCC monitoring system using a theoretical future diagnostic equipment with weekly measurements and applying the first algorithm was the most cost-effective among the monitoring systems, with a cost of approximately €10 per farm per year. This represents only 6% of the total costs due to mastitis (€175 per farm per year; Chapter 4). Despite this, SCC monitoring still represents an additional cost for smallholder farmers. This strategy should therefore be complemented by hygiene practices, for instance like those identified in Chapter 3 and other control strategies presented in Chapter 5, to generate greater benefits for smallholder dairy farmers and cooperatives. Furthermore, implementing SCC monitoring (Chapter 6) can help smallholder farmers improve their awareness of milk quality and mastitis (Chapter 2) and lead to better farm management (Chapter 3). These findings also provide valuable insights for dairy cooperatives on the potential implementation of SCC monitoring for smallholder farms.

7.3 Data and Methodological Approaches
In this section, we reflect on the data and methodologies used in this thesis. This section also discusses the constraints related to data collection, comprehensive data availability, and study design limitations, which reflect the challenges of conducting research on smallholder farms. Several data collection approaches were utilised throughout. First, in collaboration with the University of Adelaide, Australia, and IPB University, Indonesia we utilised data from the IndoDairy Smallholder Households Survey data (Umberger et al., 2020), covering 600 farms in West Java province (Chapter 2). Secondly, we conducted fieldwork to collect data from 119 dairy farms in Cianjur, West Java. This fieldwork gathered data on farmers’ socio-demographics, farm characteristics, and management practices for empirical studies, with additional bio-sampling and laboratory tests conducted to obtain milk quality and milk composition data. This Cianjur dataset was employed in Chapter 3. Additionally, economic data collected during the fieldwork, combined with cooperative recording data, was analysed in Chapter 4.

During the empirical studies in Chapters 2-4, we encountered challenges in data collection due to the limited availability of comprehensive, year-round data on milk quality, individual cow data, socio-demographics, farm characteristics, and farm economics in Indonesian smallholder dairy farms. The limited amount of routinely available data may also hinder decision-making and limit farmers’ recognition of potential issues with milk quality and their economic impacts. This constraint further hinders their ability to make informed decisions on interventions to address these challenges. These challenges were further compounded by the foot-and-mouth disease outbreak in Indonesia in 2022, which occurred during the data collection phase. The outbreak of this highly contagious disease forced us to cancel the 1-year longitudinal data collection that was originally scheduled. Based on the data constraints encountered during the empirical studies, we propose more regular collection, recording, and analysis of the recorded data, ideally coordinated by dairy cooperatives. Regular data collection on milk quality, individual cow data, socio-demographics, farm characteristics, and farm economics is crucial, as it enables a better understanding of the current status, trends, variations, and relationships within the data. Consistent and reliable records would support both future dairy research and informed decision-making by smallholder dairy farmers and cooperatives.

The analytical methodologies in these studies included multinomial and Firth-type logistic regression models (Heinze and Schemper, 2002; Frankena and Graat, 2001) for Chapter 2, Generalized Estimating Equations (GEE) for Chapters 3 and 4, and Data Envelopment Analysis (DEA; Banker et al., 1984) in combination and truncated bootstrap regression modelling (Simar and Wilson, 2007) for Chapter 4. The studies employed a cross-sectional design, which limited causal inference. Despite these limitations, these empirical studies provide valuable insights into the current challenges and potential solutions for smallholder farmers. We employed the GEE population-averaged model to generalise findings beyond the investigated dairy cooperative and enable reliable inferences about the associations between explanatory and outcome variables at the population level (Burton et al., 1998; Hu et al., 1998; Neuhaus et al., 1991). While random effects models were used in Chapter 3 to determine the intraclass correlation for the farmers' group, these models result in subject-specific estimates, which are not easily generalisable beyond the investigated farmers' groups whereas GEE models do.

Empirical data, information from literature, and expert opinions were used as input parameters to develop a bio-economic simulation model for estimating the mastitis impacts and its control strategies in Chapter 5. This model was adapted from a previously developed stochastic dynamic simulation model that assessed the impact of mastitis in Ethiopian smallholder dairy farms (Getaneh et al., 2017). Some of the mastitis control strategies identified in Chapter 3 and also available in the literature (Khasanah and Widianingrum, 2021; Ashraf and Imran, 2018; Ferronatto et al., 2018; National Mastitis Council, 2016) were utilised in this chapter to evaluate their cost-effectiveness in Indonesian settings. The study reported in Chapter 5 used simulation partly parameterized by the findings of empirical research. This way, my thesis integrates empirical (Chapter 3) and modelling (Chapter 5) studies to generate results that are more relevant and better reflect the real conditions of smallholder dairy farms in Indonesia. However, the developed simulation model was relatively straightforward and did not include pathogen transmission dynamics due to data availability constraints, which may result in an underestimation of the impact of control strategies. Gussmann et al. (2019) and van den Borne et al. (2010), for instance, have shown that the level of transmission within the herd affects the economic impact of control strategies. Including the dynamics of transmission in the study would provide a more accurate estimation of the benefits of the control strategies. Despite this limitation, the study in Chapter 5 provides valuable insights into cost-efficient mastitis control strategies that can be particularly adapted to smallholder farms.

Simulated data from Chapter 5 in combination with data about several diagnostic tools (Hanuš et al., 2011) then utilised in Chapter 6 to evaluate the cost-effectiveness of SCC monitoring systems in a dairy cooperative. This second simulation model complements the empirical studies by providing a predictive framework for assessing the costs of SCC monitoring systems. For Chapter 6 we wanted to simulate the theoretical SCC monitoring systems consisted of a measuring device, either the DeLaval cell counter (DCC) or a theoretical future diagnostic equipment (FDE), combined with three alert algorithms and measuring frequencies. However, it seemed impossible to find the correct input for such a study, consisting of the accuracy and precision of SCC measuring equipment over a range of real SCC levels. Limited field data and studies on mastitis impact and SCC monitoring in Indonesian smallholder farms posed challenges in obtaining accurate input parameters to develop models that reflect actual conditions. To address this, we incorporated insights from local cooperative staff, veterinarians, farmers, university experts, industry experts, and authors' expertise. These simulation studies offer practical insights for smallholder farms and cooperatives, enabling them to evaluate the costs and benefits of implementing these strategies.

7.4 Policy and Business Implications
Mastitis control requires a multifaceted approach that addresses national programs, farmers’ mindset and behaviour, risk factors, economic impacts, prevention and treatment strategies, and diagnostic and monitoring systems. National mastitis control programs play an important role in promoting mastitis control across dairy farms and provide a structured framework that encourages standardised practices and resources at a broad level. Hence, they establish a foundation for coordinated mastitis reduction efforts across regions. Studies from the Netherlands and Norway highlight the effectiveness of these programs in reducing mastitis-related costs and enhancing farm profitability (Lam et al., 2013; Østerås and Sølverød, 2009). By providing structured guidelines and resources, these programs not only address financial aspects but also improve farmers' mindset regarding mastitis management (Jansen et al., 2010).

Farmers' mindset is a critical component of decision-making in mastitis management (Valeeva et al., 2007). Mindset change encompasses openness to acquiring new knowledge, willingness to adopt new practices, and motivation to modify existing habits (Sok et al., 2021; Lam et al., 2017; Jansen and Lam, 2012b). Effective training and communication have been shown to improve farmers’ mindset and therefore decisions related to mastitis prevention and management (Erskine et al., 2015). Enhancements in farmers' mindset and understanding of udder health and its management have subsequently been associated with a reduction in mastitis incidence and bulk milk SCC during the implementation of the Dutch mastitis control program (van den Borne et al., 2014; Jansen et al., 2010). Brightling et al., (2009) and Tschopp et al. (2015) highlight the importance of enhancing the capacity of local extension officers and facilitating knowledge exchange within peer groups of farmers to support mastitis management. According to the theory of planned behaviour, mindset and behavioural change are driven by intention and motivation, both internal and external, shaped by attitudes, subjective norms, and perceived behavioural control (Ajzen, 1991). Building on this framework, Lam et al., (2017) developed the RESET framework, which identifies key cues for mastitis control programs: clear rules and regulatory systems, effective education and training, social pressure through peer and veterinary influence, economic incentives and consequences, and accessible tools and facilities.

Mastitis control strategies on individual farms may vary depending on the specific farm conditions and causative pathogens being present (Down et al., 2016; Britten, 2012). Farm-specific interventions are thus needed to maximise the effectiveness of control programs (Green et al., 2007). Within this framework, enhancing farmers' mindset enables them to adopt recommended practices and make informed decisions tailored to their farms' specific needs (McDougall et al., 2009).

Economic impact assessments reinforce the value of mastitis control by demonstrating its effects on milk yield, cost reduction, farm profitability, and long-term sustainability, motivating farmers and policymakers to invest in these strategies (Lam et al., 2013; Halasa et al., 2007; Gill et al., 1990). Integrating prevention and treatment strategies reduces the incidence of mastitis, while diagnostic and monitoring systems provide early warnings that support timely management decisions and interventions (Ruegg, 2017; Beekhuis-Gibbon et al., 2011; Hillerton et al., 1995). These interconnected elements create a comprehensive, data-informed approach to mastitis control that enhances both animal health and economic outcomes in the dairy sector (Zigo et al., 2021; Nielsen and Emanuelson, 2013).

Programs that focus on understanding and aligning with the intentions of dairy farmers are essential for designing interventions that are both practical and feasible for mastitis control (Mekonnen et al., 2017). This is particularly important in smallholder contexts, where resources are limited (Omore et al., 1999). These interventions would not only enhance smallholder farmers’ knowledge and awareness but also strengthen their motivation and perceived ability to adopt and sustain improved milk quality and mastitis management practices.

Indonesian Mastitis Control Program – A Proposal
In the Indonesian context, the absence of a national mastitis control program not only leaves milk quality and udder health challenges unaddressed but also risks production losses, farm inefficiencies, and economic consequences for smallholder dairy farmers (Hetherington et al., 2023a; Umberger et al., 2020; Chapter 4; Chapter 5). This gap underscores the need for a mastitis control program and presents an opportunity to develop such a program. Systematic insight in milk quality and the udder health situation (Chapter 2 and Chapter 6) should form the basis of such a program. Furthermore, the existing infrastructure of para-veterinarians, employed by the dairy cooperatives can provide follow-up to farmers with udder health problems on their farm. Such an approach can ensure sustainable improvements in milk quality, udder health, and farm profitability (Groot and van’t Hooft, 2016). This thesis (Chapters 2–6) can serve as a basis and a reference for developing a national mastitis control program, as it provides critical insights and practical guidance for adopting an integrated, interdisciplinary approach to addressing the complex challenges of milk quality and mastitis in smallholder dairy farms. The following are specific contributions from this thesis that can be utilised in shaping such a program in Indonesia.

One of the key components and foundations of the program is fostering farmer awareness and mindset (Valeeva et al., 2007). Chapter 2 of this thesis highlights the low level of awareness among smallholder farmers regarding essential milk quality parameters, such as SCC and TPC. Cooperatives play a crucial role in knowledge dissemination, and the program should build on this by organizing regular training sessions for smallholder farmers (Chapter 3) focused on improving milk quality, udder hygiene, mastitis prevention, and treatment. Farmer-focused training materials should be developed to simplify complex concepts, including milk quality, mastitis, and their economic implications (Chapters 4 and 5). Furthermore, a widespread adoption of milk quality testing and monitoring (Chapter 6) should be promoted and implemented in dairy cooperatives, enhancing smallholder farmers’ awareness and mindset and enabling them to address milk quality and mastitis issues proactively and timely.

The program should also prioritise risk-based and farm-specific interventions informed by the epidemiological findings described in Chapter 3. Critical risk factors such as manure removal frequency, mastitis treatment training, and udder hygiene practices should guide the development of tailored recommendations for smallholder farmers. By promoting improved manure management (including dry cleaning), routine hygiene practices, and easy-to-understand mastitis prevention and treatment protocols, the program can systematically address key milk quality and mastitis challenges in smallholder farms (Mekonnen et al., 2017; Omore et al., 1999). Cooperative staff (veterinarians and paramedics) should receive comprehensive training in mastitis prevention and treatment to enhance knowledge dissemination within farmer peer groups and to support smallholder farmers in effectively implementing these practices. To ensure the implementation of the training programs, a small coordinating team with expertise in mastitis control should be established. This approach would facilitate the improvement of milk quality and the reduction of mastitis prevalence (Erskine et al., 2015). This will likely increase both gross margins and the technical efficiency of the farmers involved (Chapter 4).

To ensure affordability and feasibility for smallholder farmers, cost-sharing mechanisms should be integrated into the program. Chapter 4 highlights the economic benefits of improved milk quality and udder health for farm efficiency and gross margins, indicating a strong incentive for farmers to adopt better practices, even though SCC has not yet been used as the basis for milk quality-based payments in cooperatives. Incentives based on SCC as a basis for quality-based milk payment systems, along with improved hygiene practices, can motivate farmer participation and ensure compliance (Botaro et al., 2013a; Nightingale et al., 2008). Furthermore, interventions such as post-milking teat disinfection have been shown to be cost-effective (Chapter 5). The program should establish cooperative-led cost-sharing schemes. To implement these schemes, the dairy cooperative and processors can subsidise essential intervention equipment costs and cover training costs, while farmers may contribute to costs associated with labour and consumables. Additionally, partnerships with government and dairy companies should be pursued to develop financial support mechanisms tailored to the resource constraints of smallholder farms.

Another essential component of implementing a mastitis control program is regular monitoring and data collection of milk quality and udder health status (Zigo et al., 2021; Down et al., 2016; Hillerton et al., 1995). Chapter 6 emphasises the practicality and cost-effectiveness of using a theoretical future diagnostic equipment for bulk tank milk SCC monitoring, which aligns with the affordability and accessibility considerations for smallholder dairy farmers. The program should focus on implementing such an SCC monitoring system, led by cooperatives, to guide control strategies (Chapter 5). It will provide early warning signals to increase smallholder farmers’ awareness of milk quality and mastitis (Chapter 2), allowing them to intervene in a timely manner. Training cooperative staff to adopt and manage these monitoring systems and interpret results is critical for ensuring accurate and comprehensive data and timely interventions. Furthermore, establishing a centralised database at the national or regional level will facilitate the assessment of farm- and MCP-level milk quality trends and measure the effectiveness of the mastitis control program over time.

By synthesising and implementing the findings of this thesis (Chapters 2–6) into a structured mastitis control program, Indonesia can effectively address milk quality and mastitis challenges faced by smallholder dairy farms. The proposed udder health program utilises cooperative leadership, targeted interventions, and support from dairy companies to improve milk quality, udder health, and farm profitability. This approach not only benefits farmers but also strengthens the foundation for the sustainable development of the dairy sector in Indonesia. It would ultimately imply a reduced dependency on imported milk and enhance overall milk quality.

7.5 Research outlook
This thesis provides insight into the epidemiology and economics of milk quality and mastitis control strategies in Indonesian smallholder dairy farms. Therefore, future research should analyse the benefits of combining SCC monitoring systems with interventions (i.e., surveillance programs) on smallholder dairy farms. Additionally, randomized field trial study interventions, preferably with a control group to correctly evaluate the effect, such as implementing hygiene practices identified in this thesis and literature (e.g., post-milking teat dipping), should assess their impacts in daily practice. Linked to continued data collection and studies to evaluate the effectivity of interventions, the developed simulation models (Chapter 5 and Chapter 6) can be updated and expanded, for instance, to include pathogen transmission dynamics (Gussmann et al., 2019b; van den Borne et al., 2010). Since mastitis results in multiple burdens (economic, environmental, public health, and animal welfare burdens) (Steeneveld et al., 2024), research on epidemiology and economic analysis should be expanded to include other sustainability aspects, such as animal welfare, public health concerns related to antimicrobial use, and the environmental impact of implementing control strategies on smallholder farms. The development of affordable and accurate SCC monitoring systems that can be applied in MCPs has proven to be valuable (Chapter 6). Therefore, research investments towards the development of such systems are warranted.

All of these future research directions should be supported by regular data monitoring, recording, and analysis at the cow, herd, and cooperative levels to ensure informed decision-making within smallholder farms and cooperatives, ultimately supporting the development of smallholder dairy systems.

To ensure that the proposed mastitis control program becomes a reality and remains sustainable, further research should also focus on developing cost-effectiveness strategies tailored to the unique constraints of smallholder farms in Indonesia, including limited financial resources, smaller herd sizes, low awareness of milk quality and mastitis, and reliance on cooperative infrastructure. Key areas for further research focusing on the implementation of the national control program include designing practical mastitis training modules for farmers and cooperative staff, assessing cost-sharing mechanisms to improve access to affordable hygiene tools and interventions, piloting cooperative-led SCC monitoring systems to assess their feasibility, and evaluating the impact of implementing mastitis control programs. Strong collaboration among researchers, governments, farmers, cooperatives, and industry stakeholders is crucial to addressing the multifaceted challenges of milk quality and mastitis management, ensuring the program's viability and fostering sustainable improvements in Indonesia's smallholder dairy sector. This thesis will lay a foundation for such an improvement.

7.6 Conclusions
The most important conclusions of this thesis are the following:
Combining empirical studies (Chapters 2-4) and simulation studies (Chapters 5-6) provides a comprehensive understanding of milk quality and mastitis control in smallholder dairy farms, highlighting the value of an interdisciplinary approach that integrates epidemiological and economic analyses to address complex challenges and offers practical solutions.
Findings from Chapter 2 highlight the need to enhance farmers’ awareness of milk quality (SCC and TPC) as a precursor to successful milk quality and udder health improvement efforts in smallholder dairy farms.
Chapter 3 identifies five risk factors—manure removal frequency, receiving mastitis treatment training, washing the udder with soap, number of labour, and ownership of the pasture area—that are associated with SCC. Additionally, manure removal frequency and dairy income contribution were found to be associated with TPC. These epidemiological findings lay the foundation for economic performance analysis (Chapter 4) and ensure a clear understanding of the risk factors associated with milk quality and udder health in Indonesian smallholder dairy farms.
Chapter 4 reveals that improving milk quality and udder health can increase technical efficiency and the gross margin of smallholder farms. These findings complement the epidemiological findings from Chapter 3, offering more valuable insights for mastitis control in smallholder dairy farms.
Chapter 5 quantifies the significant economic burden of mastitis on smallholder dairy farms, showing that the total cost of mastitis can represent up to 36% of the annual gross margin of farms (Chapter 4). This underscores the need for efficient control strategies in smallholder dairy farms. Among the evaluated strategies, post-milking teat disinfection was identified as the only cost-efficient strategy for reducing mastitis incidence, milk production losses, SCC, and mastitis-related costs.
Investment in the development of an affordable yet accurate diagnostic monitoring system is advisable (Chapter 6). This SCC monitoring system should be combined with findings from empirical and modelling studies (Chapters 2-5) to generate greater benefits for smallholder dairy farmers and cooperatives.
The findings from Chapters 2-6 serve as a reference and guidance for developing and implementing national mastitis control programs in Indonesia and other tropical regions where smallholder farms play a significant role in dairy production. The proposed program addresses the constraints faced by smallholder farmers by prioritising cost-effectiveness, cooperative-led implementation of initiatives, and collective approaches to milk quality and mastitis monitoring and interventions.

These integrated findings provide practical recommendations for national mastitis control program implementation, aiming to improve mastitis control and milk quality in smallholder dairy systems, with broader implications for regional, national, and global level strategies.

References
Aal, A. El, 2014. Correlation between some direct and indirect tests for screen detection of subclinical mastitis. International Food Research Journal 21, 1249–1254.
Adkins, P.R.F., Middleton, J.R., 2018. Methods for Diagnosing Mastitis. Veterinary Clinics of North America - Food Animal Practice. doi:10.1016/j.cvfa.2018.07.003
Ahmedsham, M., Amza, N., Tamiru, M., 2018. Review on milk and milk product safety, quality assurance and control. International Journal of Livestock Production 9, 67–78. doi:10.5897/ijlp2017.0403
Ajzen, I., 1991. The Theory of Planned Behavior. Organizational Behavior and Human Decision Processes 50, 179–211.
Akzar, R., Umberger, W., Peralta, A., 2023. Understanding heterogeneity in technology adoption among Indonesian smallholder dairy farmers. Agribusiness 39, 347–370. doi:10.1002/agr.21782
Allendorf, J.J., Wettemann, P.J.C., 2015. Does animal welfare influence dairy farm efficiency? A two-stage approach. Journal of Dairy Science 98, 7730–7740. doi:10.3168/jds.2015-9390
Allore, H.G., Erb, H.N., 1998. Partial Budget of the Discounted Annual Benefit of Mastitis Control Strategies. Journal of Dairy Science 81, 2280–2292. doi:10.3168/jds.S0022-0302(98)75808-4
Anas, M. Al, Hasanah, H., Agus, A., 2021. The potency of traditional market vegetable waste as ruminant feed in the special region of Yogyakarta. Advances in Animal and Veterinary Sciences 9, 1416–1423. doi:10.17582/journal.aavs/2021/9.9.1416.1423
Anugrah, I.S., Purwantini, T.B., Erwidodo, N., 2021. Milk Collection Points: Innovation and Partnership of the Dairy Cattle Farming in Pangalengan-South Bandung. Analisis Kebijakan Pertanian 19, 1. doi:10.21082/akp.v19n1.2021.1-18
Ashraf, A., Imran, M., 2018. Diagnosis of bovine mastitis: from laboratory to farm. Tropical Animal Health and Production. doi:10.1007/s11250-018-1629-0
Atkinson, A.A., Balakrishnan, R., Booth, P., Cote, J.M., Groot, T., Malmi, T., Roberts, H., Uliana, E., Wu, A., 1997. New Direction in Management Accounting Research. Journal of Management Accounting Research 9, 79–108.
Badunenko, O., Tauchmann, H., 2019. Simar and Wilson two-stage efficiency analysis for Stata. Stata Journal 19, 950–988. doi:10.1177/1536867X19893640
Banker, R.D., Charnes, A., Cooper, W.W., 1984. Some Models for Estimating Technical and Scale Inefficiencies in Data Envelopment Analysis. Management Science 30, 1078–1092.
Bao, K.L.N., Sandjaja, S., Poh, B.K., Rojroongwasinkul, N., Huu, C.N., Sumedi, E., Aini, J.N., Senaprom, S., Deurenberg, P., Bragt, M., Khouw, I., Khouw, I., 2018. The consumption of dairy and its association with nutritional status in the South East Asian nutrition surveys (SEANUTS). Nutrients 10. doi:10.3390/nu10060759
Barkema, H.W., Schukken, Y.H., Lam, T.J.G.M., 1998. Management practices associated with low, medium, and high somatic cell counts in bulk milk. Journal of Dairy Science 81, 1917–1927. doi:10.3168/jds.S0022-0302(98)75764-9
Bartlett, P.C., Miller, G.Y., Lance, S.E., Heider, L.E., 1992. Environmental and managerial determinants of somatic cell counts and clinical mastitis incidence in Ohio dairy herds. Preventive Veterinary Medicine 14, 195–207. doi:https://doi.org/10.1016/0167-5877(92)90016-
Bassotto, L.C., Lima, A.L.R., Carvalho, F. de M., Lopes, M.A., Nascimento, E.F.R., Netto, E.P.L., 2023. Characteristics of dairy farms with different levels of technical efficiency. Ciencia e Agrotecnologia 47. doi:10.1590/1413-7054202347019122
Beekhuis-Gibbon, L., Devitt, C., Whyte, P., O’Grady, L., More, S.J., Redmond, B., Quin, S., Doherty, M.L., 2011. A HACCP-based approach to mastitis control in dairy herds. Part 2: Implementation and evaluation. Irish Veterinary Journal 64. doi:10.1186/2046-0481-64-7
Beuchat, L.R., Copeland, F., Curiale, M.S., Danisavich, T., Gangar, V., King, B.W., Lawlis, T.L., Likin, R.O., Okwusoa, J., Smith, C.F., Townsend, D.E., 1998. Comparison of the simplate(TM) total plate count method with petrifilm(TM), redigel(TM), and conventional pour-plate methods for enumerating aerobic microorganisms in foods. Journal of Food Protection 61, 14–18. doi:10.4315/0362-028X-61.1.14
Bicudo, J.R., Goyal, S.M., 2003. Pathogens and manure management systems : A review. Environmental Technology 24, 115–130. doi:10.1080/09593330309385542
Biffa, D., Debela, E., Beyene, F., 2005. Prevalence and Risk Factors of Mastitis in Lactating Dairy Cows in Southern Ethiopia. Journal of Applied Research in Veterinary Medicine 3, 189–198.
Bobbo, T., Cipolat-Gotet, C., Bittante, G., Cecchinato, A., 2016. The nonlinear effect of somatic cell count on milk composition, coagulation properties, curd firmness modeling, cheese yield, and curd nutrient recovery. Journal of Dairy Science 99, 5104–5119. doi:10.3168/jds.2015-10512
Böker, A.R., Bartel, A., Do Duc, P., Hentzsch, A., Reichmann, F., Merle, R., Arndt, H., Dachrodt, L., Woudstra, S., Hoedemaker, M., 2023. Status of udder health performance indicators and implementation on farm monitoring on German dairy cow farms: results from a large scale cross-sectional study. Frontiers in Veterinary Science 10. doi:10.3389/fvets.2023.1193301
Bonestroo, J., Fall, N., Hogeveen, H., Emanuelson, U., Klaas, I.C., van der Voort, M., 2023. The costs of chronic mastitis: A simulation study of an automatic milking system farm. Preventive Veterinary Medicine 210. doi:10.1016/j.prevetmed.2022.105799
Bortolami, A., Fiore, E., Gianesella, M., Corro, M., Catania, S., Morgante, M., 2015. Evaluation of the udder health status in subclinical mastitis affected dairy cows through bacteriological culture , somatic cell count and thermographic imaging 18, 799–805. doi:10.1515/pjvs-2015-0104
Botaro, B.G., Gameiro, A.H., dos Santos, M.V., 2013. Quality based payment program and milk quality in dairy cooperatives of Southern Brazil: an econometric analysis. Scientia Agricola 70, 21–26. doi:10.1590/S0103-90162013000100004
Bradley, A.J., Leach, K.A., Green, M.J., Gibbons, J., Ohnstad, I.C., Black, D.H., Payne, B., Prout, V.E., Breen, J.E., 2018. The impact of dairy cows’ bedding material and its microbial content on the quality and safety of milk – A cross sectional study of UK farms. International Journal of Food Microbiology 269, 36–45. doi:10.1016/j.ijfoodmicro.2017.12.022
Bravo-Ureta, B.E., Solís, D., Moreira López, V.H., Maripani, J.F., Thiam, A., Rivas, T., 2007. Technical efficiency in farming: A meta-regression analysis. Journal of Productivity Analysis 27, 57–72. doi:10.1007/s11123-006-0025-
Brightling, P.B., Dyson, R.D., Hope, A.F., Penry, J., 2009. A national programme for mastitis control in Australia : Countdown Downunder. Irish Veterinary Journal 62. doi:10.1186/2046-0481-62-S4-S52
Britten, A.M., 2012. The role of diagnostic microbiology in mastitis control programs. Veterinary Clinics of North America - Food Animal Practice 28, 187–202. doi:10.1016/j.cvfa.2012.03.006
Burton, P., Gurrin, L., Sly, P., 1998. Extending the simple linear regression model to account for correlated responses : An introduction to generalized estimating equations and multi-level mixed modelling. Statistics in Medicine 17, 1261–1291. doi:https://doi.org/10.1002/(SICI)1097-0258
Busanello, M., Rossi, R.S., Cassoli, L.D., Pantoja, J.C.F., Machado, P.F., 2017. Estimation of prevalence and incidence of subclinical mastitis in a large population of Brazilian dairy herds. Journal of Dairy Science 100, 6545–6553. doi:10.3168/jds.2016-12042
Chye, F.Y., Abdullah, A., Ayob, M.K., 2004. Bacteriological quality and safety of raw milk in Malaysia. Food Microbiology 21, 535–541. doi:10.1016/j.fm.2003.11.007
Cooper, W.W., Seiford, L.M., Tone, K., 2002. Data Envelopment Analysis: A Comprehensive Text with Models, Applications, References and DEA-Solver Software, 1st ed. Springer, New York. doi:https://doi.org/10.1007/b109347
Cui, J., 2007. QIC program and model selection in GEE analyses. The Stata Journal 7, 209–220.
Daraio, C., Simar, L., 2007. Advanced Robust and Nonparametric Methods in Efficiency Analysis: Methodology and Applications, Springer. ed. New York. doi:10.1007/978-0-387-35231-2
De Andrade Ferrazza, R., Lopes, M.A., de Oliveira Prado, D.G., de Lima, R.R., Bruhn, F.R.P., 2020. Association between technical and economic performance indexes and dairy farm profitability. Revista Brasileira de Zootecnia 49. doi:10.37496/RBZ4920180116
De Vries, A., Kaylegian, K.E., Dahl, G.E., 2020. MILK Symposium review: Improving the productivity, quality, and safety of milk in Rwanda and Nepal. Journal of Dairy Science 103, 9758–9773. doi:10.3168/jds.2020-18304
De Vries, M., Wouters, B., 2017. Characteristics of small-scale dairy farms in Lembang, West-Java 36.
De Vries, M., Wouters, B., Vellinga, T., 2017. Environmental impacts of dairy farming in Lembang, West Java 37.
Dekkers, J.C.M., Erp, T. Van, Schukken, Y.H., 1996. Economic Benefits of Reducing Somatic Cell Count under the Milk Quality Program of Ontario. Journal of Dairy Science 79, 396–401. doi:10.3168/jds.S0022-0302(96)76378-6
Delgado, C.L., 2003. Rising consumption of meat and milk in developing countries has created a new food revolution. Journal of Nutrition 133, 3907–3910. doi:10.1093/jn/133.11.3907s
DeLong, K.L., Lambert, D.M., Schexnayder, S., Krawczel, P., Fly, M., Garkovich, L., Oliver, S., 2017. Farm business and operator variables associated with bulk tank somatic cell count from dairy herds in the southeastern United States. Journal of Dairy Science 100, 9298–9310. doi:10.3168/jds.2017-12767
Deng, Z., Hogeveen, H., Lam, T.J.G.M., van der Tol, R., Koop, G., 2020. Performance of Online Somatic Cell Count Estimation in Automatic Milking Systems. Frontiers in Veterinary Science 7. doi:10.3389/fvets.2020.
Devendra, C., 2001. Smallholder dairy production systems in developing countries: characteristics, potential and opportunities for improvement - review. Asian-Australasian Journal of Animal Sciences 14, 104–113. doi:https://doi.org/10.5713/ajas.2001.104
Didanna, H.L., Mengistu, A., Kuma, T., Kuma, B., 2018. Improving milk safety at farm-level in an intensive dairy production system: Relevance to smallholder dairy producers. Food Quality and Safety 2, 135–143. doi:10.1093/fqsafe/fyy009
Dingwell, R.T., Kelton, D.F., Leslie, K.E., 2003. Management of the dry cow in control of peripartum disease and mastitis. Veterinary Clinics of North America: Food Animal Practice 19, 235–265. doi:10.1016/S0749-0720(02)00072-5
Dohmen, W., Neijenhuis, F., Hogeveen, H., 2010. Relationship between udder health and hygiene on farms with an automatic milking system. Journal of Dairy Science 93, 4019–4033. doi:10.3168/jds.2009-3028
Dong, F., Hennessy, D.A., Jensen, H.H., 2012. Factors determining milk quality and implications for production structure under somatic cell count standard modification. Journal of Dairy Science 95, 6421–6435. doi:10.3168/jds.2012-5522
Douphrate, D.I., Hagevoort, G.R., Nonnenmann, M.W., Lunner Kolstrup, C., Reynolds, S.J., Jakob, M., Kinsel, M., 2013. The Dairy Industry: A Brief Description of Production Practices, Trends, and Farm Characteristics Around the World. Journal of Agromedicine 18, 187–197. doi:10.1080/1059924X.2013.796901
Down, P.M., Bradley, A.J., Breen, J.E., Hudson, C.D., Green, M.J., 2016. Current management practices and interventions prioritised as part of a nationwide mastitis control plan. Veterinary Record 178, 449. doi:10.1136/vr.103203
Erskine, R.J., Martinez, R.O., Contreras, G.A., 2015. Cultural lag: A new challenge for mastitis control on dairy farms in the United States. Journal of Dairy Science 98, 8240–8244. doi:10.3168/jds.2015-9386