Farmers rely on their daily experience and practical knowledge of how their livestock will optimally produce, which is something they positively relate to AHW (Balzani and Hanlon, 2020). Productivity and AHW status of farm animals are seen as a means to ensure economic viability of the farm system. Thereby, farmers are confronted with a high degree of complexity, which emerges from the interactions between manifold factors on different scales within the farm system and from interactions with the external world. In particular, livestock farmers have to deal with volatile market prices for raw goods and production tools such as feed, which they can hardly influence but have to adapt to. Sales prices are the result of a price competition on a global level, forcing livestock farmers to produce raw goods to the lowest possible costs. Although being concerned about the economic viability of the farm, livestock farmers are not necessarily guided by rational thinking and evidence. Especially negative side effects of the production processes such as production diseases and the economic consequences are often disregarded (Hogeveen et al., 2019).

The whole farm is the reference system when it comes to decisions about the allocation of resources. From the perspective of farmers, investments in the AHW issue are only justified when a return of investment can be expected to directly pay off, whereas it cannot be expected from them to operate against their own interests. Correspondingly, limitations in the availability of relevant resources (e.g. high-quality feed, labour capacity, investments, know how) are causing profound trade-offs when decisions have to be taken regarding their allocation between partial goals. On the one hand, there is a large extend of overlap between the goal of productivity and the success in protecting farm animals against severe AHW problems. On the other hand, an excessive and unilateral pursuit of productivity can be counteracted by negative side effects and failure costs, particularly due to AHW problems (Farm Animal Welfare Committee, 2011). It is also possible that measures to improve AHW are so costly that the overall productivity diminished. While the state of physical and mental health is an end in itself for the individual animal, it is only a means to an end for the farmer. He/she is not well recommended to strive for the highest possible level of AHW but for the farm specific optimum that balances the failure costs due to AHW problems with the costs to prevent losses by appropriate countermeasures (Hogeveen et al., 2019). The same applies for strategies to increase the productivity of farm animals, e.g. by enhancing the performance level. When expenses exceed the optimum level of the cost-to-benefit ratio for the individual animal, farmers yield a negative marginal profit.

To improve the level of AHW, appearing dysfunctions, disturbances, disorders or diseases have to be treated and modifications in the allocation of resources have to be made. The expertise to meet the comprehensive challenges is not aggregated in one but divided up over different disciplines. Concerning the treatment of diseased animals, veterinarian surgeons are the trained specialists from which is expected that they are able to contribute to substantial improvements. However, to do so, some preconditions must be fulfilled. Inter alia, they have to be consulted at the onset of diseases, charged to identify the causes of diseases, enabled to exploit the possible treatment options and to assess treatment success, and last but not least enabled to eliminate the farm specific weak spots and main causes of recurrent diseases. Not only the consultation but particularly their recommendations are a relevant cost factor. To improve treatment success, veterinarian surgeons have to be charged, paid, and finally, farmers have to implement the recommendations. According to a study, aimed to qualitatively and quantitatively describe farmers’ reasons for adherence and nonadherence with veterinary recommendations, trust, feasibility, and priorities were identified as the main themes (Svensson et al., 2019). However, there is a lot more at stake than the adherence and nonadherence with the recommendations of experts.

A systemic perception of the need for alterations requires a farm management that focuses on shaping the environment (Wells and McLean, 2013). From the recognition of animals failing to cope follows the operational requirement to design an environment that is better adapted to the needs of the animals. This can only be achieved effectively if the causes of dysfunctions are identified and repaired. However, it is also necessary to consider the individuality and self-referentiality of farm animals and their context-variant interactions with the living conditions. This is illustrated in further detail by different examples of AHW problems in dairy farming, but applies in a modified form also to other species in livestock production.

Mastitis is a main production disease of dairy cows. Its aetiology is complex and multifactorial as are the options to treat or prevent mastitis effectively and cost-efficiently (Vliegher et al., 2018). Factors with explanatory power are inter alia the time interval between infection and treatment, knowledge about the pathogenicity of the pathogen germs, the degree of parenchyma alterations, the frequency of previous treatments, and the energy balance with respect to the nutrient supply. Success to reduce mastitis is above all depending on the implementation of a surveillance programme and an appropriate treatment strategy. An on-farm survey revealed that therapeutic success ranged between 18% to 59% while the total costs of mastitis per cow and year varied between € 158 to € 483 (Doehring and Sundrum, 2019). The large variation between and within farms indicates that average figures risk giving misleading information about effective and cost-efficient measures. Due to a lack of profound data, failure costs of mastitis and investments to prevent them are seldom estimated appropriately (van Soest et al., 2019).

Another example is given by metabolic disorders. They are a key problem in the transition period of dairy cows and often appear as onset of further health problems (Sundrum, 2015; Webster, 2021). They mainly derive from varying gaps between nutrient supply and demand and cause variations in activity and feeding behaviour patterns (Jawor et al., 2012; Itle et al., 2015), collectively called “sickness behaviour” (Harden et al., 2015). Metabolic disorders indicate an overstressed ability to balance input, partitioning and output variables and to adapt to imbalances in the supply with nutrients. The earlier endangered and diseased animals are identified and treated accordingly, the lower are the impairments and the higher the probability that countermeasures succeed. A low prevalence of metabolic disorders can most likely be achieved when the farm management gets an overview of the nutritional status of the single animals, e.g. by making use of balance sheets, for example with respect to energy balance (Thorup et al., 2018). By this means, endangered animals can be identified and modifications of the individual supply or demand side engineered. Such an approach suggests a more effective and cost-efficient procedure than strategies that primarily rely on the adjustment of a diet which is fed to all animals of a feeding group while disregarding the intra- and interindividual variation of nutrient requirements (Sundrum, 2020b).

Culling of farm animals indicates the ultimate failure of adaptation. This applies for both, the farm animals and the farmers who were not capable to adjust the living conditions to meet the individual requirements of the farm animals. Culling of dairy cows is primarily due to infertility or production diseases (Compton et al., 2017; Rilanto et al., 2020). Often, more or less extended phases of pain, suffering and damage preceded the way towards the final end. Time and circumstances of culling are not only associated with severe AHW problems, but are also main drivers of economic results (Dechow and Goodling, 2008). Farmers often tend to underestimate costs of disease, culling and replacement (Jones et al., 2016). Individual cows can only contribute to farmers income, if their individual revenues from milk and slaughter override their individual costs for rearing, feeding and their share of the fix costs of the farm. Thus, culling is not economically desirable before cows reach their productive zenith. The cow-specific calculation of individual income over service life cost (ⁱIOLC) regards the individual cow as both the basic biological and the basic economic unit of a dairy farm, which allows for allocation of all costs and revenues of the dairy branch to single cow’s service lives (Habel et al., 2021). In a comprehensive study on 32 dairy farms, lifetime profitability of individual cows varied considerably between farms as well as between the culled cows on the same farm. Median ⁱIOLC was negative for 59% of the investigated farms and varied largely between farms. The majority of the investigated farms failed to take appropriate decisions to better sustain the lives of their animals and to generate sufficient profit from the investments.

There is no doubt that it costs efforts and money to solve AHW problems. Although there is growing literature on farmer decision making and ‘barriers to uptake’ with regard to the different production sectors (Palczynski et al., 2016; Svensson et al., 2019; van Klompenburg and Kassahun, 2022), it often remains unclear for the farmer where investments and efforts could prove essential to survival, where they might be effective and worthwhile, and where they seem counterproductive. Correspondingly, farmers may be prevented from acting because of the complexity at hand. Only few costs are obvious: e.g., the reconfiguration of pens, and administration of improved feeding, hygienic and treatments strategies. Some interventions may require ongoing additional personnel costs or additional veterinary expertise. Not less problematic is the fact, that the projected benefits are less certain than the costs, and are more difficult to assess in financial terms (Fernandes et al., 2021). In a review, various financial benefits of good AHW through, for example, reduced mortality, improved health, improved resistance to disease and reduced medication, and lower risk of zoonoses and animal-borne infections have been outlined (Dawkins, 2017). Although these parameters might directly affect the profitability of the livestock farm businesses, they are neither expressed in economic terms nor in terms of cost-to-benefit ratios. While it is widely recognized that animals with poor welfare are unlikely to produce at optimal levels, the cost-to-benefit ratio of investments are seldom known, because they are highly context-variant. What makes it even more complex and confusing for the farmer is the fact that investments are underlying the general law of diminishing marginal utility.

Due to the heterogeneity of farm conditions, to reduce AHW problems, farmers cannot rely on general recommendations. They have to find a farm centric approach to organise and regulate the processes within the farm system. In their role, farmers are both observer and part of the farm system, primarily guided by a self-referential perspective. To strive for a balance between the goals of a high productive efficiency and a low rate of dysfunctions, e.g. in terms of AHW problems, they could learn from perceiving their farm as an agroecosystem. Concerning the characteristics of self-organizing autopoietic systems (Razeto-Barry, 2012), the structure of agroecosystems can be described as a hierarchy of scales as illustrated in Figure 1 .

Cells as autopoietic systems of 1^st order are interacting with available material and elements in their environment in a self-referential way with respect to what they need to sustain and function when organised with other cells in tissues and organs. Furthermore, they fulfil demands of the individual organism as the autopoietic system of 2^nd order. Individual animals are part of a herd and subsystems of the agroecosystem. In this way, each superordinate system acts as the environment for the subsystem, which cannot exist or be sustained without the resources and the protection provided by the superordinate system. Autopoietic systems continuously produce the components, which constitute it by itself, while these components steadily sustain and regenerate the superordinate system (Varela, 1997). Individual animals are an integral and functional system and represent an entity with a functional integrity. In striving for maintenance, the uptake of external material is regulated following own purposes when integrating and transforming the material within the system. They select components from the environment and give them a meaning of incentives, objects of perception, and interests.

Because the requirements and needs of farm animals and their ability to cope vary considerably even in largely homogenized groups, it is of high importance to consider the individuality of farm animals, interacting within a superordinate system based on the functionality and integrity of its own subsystems. The more supply and allocation of nutrient resources and protection against biotic and abiotic stressors are adapted to the individual requirements, the higher the chance that the individual is able to cope. Thus, when striving for a reduction of AHW problems, farmers are well recommended to recognise the farm and the subsystems as autopoietic, being self-referential in what they need to cope, in order to be able to support the functionality the superior system in which they are embedded.

The enormous growth in scientific activities and knowledge has inevitably led to a fragmentation of scientific knowledge. Differentiation of disciplines and specialisation of research areas are still considered as a recipe for success. By answering key research questions, new questions arise and generate a self-sustaining momentum, not unlike the path dependency in farm business. The focus on disciplinary research programs entail that large parts of the real world complexity are disregarded, associated with a vague expectation that these are covered by other disciplines (Krohn et al., 2017). Each discipline follows its own goals and problem framing, encompassing the definition of system boundaries, the role of actors and the selection of design tasks. They refer to different reference systems such as different groups of animals, housing systems or production methods. Differentiation and specialisation of scientific approaches have resulted in what Alrøe and Noe (2011) called perspectival knowledge asymmetries. Scientists have their own discipline-born perspective and see complex matters differently. Experts perceive knowledge gaps from within their professional paradigms and are widely incapable of forming judgements beyond their specific expertise (Millgram, 2015).

According to Fraser (2008), “different scientists adopted the different value-based views of animal welfare - basic health and functioning, natural living, and affective states – as the rationale for different scientific approaches to assessing and improving animal welfare”. The different approaches are associated with “a dilemma that threatens to throw animal welfare science into disarray”. This is particularly the case, when different scientific approaches arrive at opposite conclusions as exemplified in the case of housing sows in gestation stalls. Fraser (2008) identified the reason for contradictions in the fact that tools of science are used within a framework of values and that the understanding of animal welfare and its scientific assessment is both values-based and science-based. Thus, scientists are self-referential when choosing a specific perspective on the AHW issue or apply specific methodological tools they have taken ownership of in their specialist discipline. While the disciplinary focus through the inherent methodological lenses provides a deeper insight in the interconnectedness of processes within sub-areas, the focus becomes even more narrowed in search for further gaps beyond the knowledge that is already available. The drivers of specialisation in animal science like in other scientific areas have inevitably led to a reductionist approach while the whole picture is becoming more and more lost out of sight, leading to a decline in the skills to solve complex problems (Deutsche Forschungsgemeinschaft, 2005). This applies above all for problems that develop beyond the boundaries of disciplinary specialities, which is particularly the case for most of AHW problems. Such can only be solved when the underlying causes are understood not only from the ankle of single disciplines and from a focus on single factors. To grasp the outcomes of complex interactions of various factors involved on different scales, these have to be examined from different perspective. Understanding the multifactorial genesis of AHW problems and contributing to their mitigation requires an integrative and thus inter- and transdisciplinary solution approach. This is most likely to be achieved when funding bodies specify schemes to encourage inter- and transdisciplinary thinking. However, this alone is not sufficient. As scientific findings always have some generality and some specificity, it is of high importance for the chance that solution proposals are implemented in farm practice to know more about the ambiguity and context dependence of methods and measures. Consequently, there is an urgent need to extend tests on the external validity of measures and results obtained under specific conditions.

Scientists are not only self-referential concerning the research subjects but have to consider also their own benefits when conducting research work. When striving for a scientific career they will not be able to avoid relying on publications in scientific journals. According to the guide for authors of many scientific journals, studies may be considered for publication only if the results are likely to be of international interest, i.e. it must be possible to generalise the findings using scientifically based approaches (Elsevier publisher, 2023). Of course, this applies not for all scientific journals. However, the Impact Factor of the journals that make the such demands tends to be higher, and thus are of higher attractiveness for scientists. As scientists compete for research funding and for comparatively few permanent positions, promising scientists only have a chance to pursue a career when they are careful to demonstrate their publishing skills. The general framework of research work affects the areas where scientists are looking for gaps of knowledge suited to generate generalisable and therewith publishable results. In contrast, scientists are not often rewarded for translating their results to be useful to end-users.

Publications must satisfy the major scientific criteria of validity and reliability. Validity describes how well an instrument does what it is supposed to do. Reliability describes the consistency with which results are obtained (Andrade, 2018). Concerning validity, it can be distinguished between internal, external and ecological validity. Internal validity examines whether the study design, conduct, and analysis answer the research questions without bias, and allows trustworthy answers to the research questions in the study. For example, improper randomization, inadvertent unblinding, or missing data can all undermine the fidelity of the results and conclusions of a randomized controlled trial (RCT). External validity examines whether the study findings can be generalized to other contexts, for example other patients. Ecological validity refers to the degree to which research findings reflect real-life related functions and examines how closely an experiment aligns with real-world phenomena (Fahmie et al., 2023). Ecological validity is evaluated within a given study, external validity of single-case research is evaluated across a collection of studies.

While standardisation of experimental conditions is of high importance when striving for a high internal validity and for generating publishable results, study results can become increasingly determined by the specific experimental conditions in which they are conducted. Thus, environmental standardisation is a cause of, rather than a cure for, poor reproducibility of experimental outcomes (Richter et al., 2009). To reduce within-experiment variation, experimenters are usually advised to standardize the conditions of their experiments as rigorously as possible. Standardising the experimental conditions is believed to render the animals’ responses to experimental treatments more homogeneous, thereby reducing within-experiment variation and increasing test sensitivity. As a consequence, study results might appear as generalisable and context-invariant from a disciplinary perspective, whereas they can appear as highly context-specific when being examined with regard to external validity. Not less important is the fact, that highly standardised conditions can be very different from what is found on commercial farms. The degree by which farm animals pay for the pursuit to increase productivity with limitations in the availability of resources and in the protection against a large variety of biotic and abiotic stressors vary considerably. The variety of disruptive factors includes high stocking density, provoking conflicts with other animals, high concentrations of microbiological pathogens, parasites or toxins, and harmful gases. In addition, the variety of the interactions between these factors are subject of dynamic changes, whereas the combination of the respective factors affects the animals not separately but jointly.

In contrast, it is inherent for a disciplinary perspective to focus on selected effects, necessarily implying that other aspects are disregarded. The focus on single aspects or the selection of specific criteria touches upon a core problem of biology (Cassirer, 1974): the ‘part-whole-problem’. Within an organism, the different parts and subsystems are not separate from each other but are self- and mutually reinforcing and work together towards the same goal: to keep the organism alive. There is no part that does not need the support and cooperation of nearly all other parts. A living system cannot be fully understood without considering the purpose behind the biological processes. This applies particularly for adaptation as a functional and target-oriented process. Whether the whole organism responds adequately to challenges cannot be deduced from single parts or by indicators. Whether adaptive measures and processes are beneficial or not depends primarily on the context in which they take place. They do not follow a one-size-fits-all approach as is for instance conducted in randomized controlled trials, but require an external validation with regard to the effectiveness in farm practice (Krauss, 2018).

Different scientific disciplines prioritise different indicators to assess the state of physical and mental health of farm animals. While partial views reveal only partial insights, assessments based on partial results and on an inductive approach are at risk for misjudgements and inductive fallacies. The piecemeal approach of animal science blinds actors to the possible larger conflict constellation for AHW problems. Whereas in experimental studies scientist are striving for the greatest possible narrowing of the variation, variation between animals can be quite extended in farm practice. What is statistically valid for a group of farm animals under highly standardised conditions does not necessarily work for individual animals, deviating from what might be supportive when assessed in terms of means values. This applies for instance when providing feeding rations formulated to meet the average nutrient requirements of a group of farm animals, although animals show a high degree of variation in their requirements (Rumphorst et al., 2021). Nevertheless, effects of nutrient supplies are often validated against the mean effects revealed in relation to a control group, but not against the state of physical health of individuals, being the valid reference system in the AHW context. The state of physical and mental health refers to the individual animals at any point in their lifetime. Thus, the state cannot be averaged.

Possible counter-examples are epidemiological studies which are analysing on-farm risks for AHW problems (e.g. lameness, mastitis, pecking, fractures, abnormal behaviour) and explore real-world variation, although they seldom consider individual animals as the reference unit. Benefits and limitations of epidemiological studies are exemplified with a study investigating risk factors for lameness in dairy cows (Dippel et al., 2009). The results revealed that lying comfort and nutrition were identified as key risk areas for lameness across regions, breeds, and farming systems, therewith providing information about the priority of factors that have to be considered. However, the farm-level factors in this study, as is the case in many other studies (Rittweg et al., 2023), were restricted to few easily detectable indicators, whereas many of those factors that are also relevant for the development of lameness, such as the quality of the floor, hygienic conditions and preventive hoof trimming (Sogstad et al., 2005), comorbidities (Daros et al., 2020) or the energy balance of the individual animal (Collard et al., 1999) have been disregarded. In addition, a priority of risk factors that might be valid for the majority of dairy farms is not necessarily true for a specific farm situation. The multifactorial genesis of lameness varies not only between farms but first of all between the individual animals. While epidemiological results offer some orientation about the most probable risk factors that should be considered with regard to possible preventive measures, they seldom lead to concrete options for action. These would require a sufficient persuasive power towards the farmers to convince them that the recommended measures can be expected to be effective and cost-efficient in the farm-specific context. Above all, epidemiological results do not save the efforts of a thorough diagnostic procedure to identify the main causes in the farm and animal specific situation when it comes to the foremost task to treat the suffering animals in due time.

While many scientific studies, technologies and practices claim to result in positive AHW (Bertenshaw et al., 2008; Cronin et al., 2014; Westerath et al., 2014; Rentsch et al., 2023), without physiological and behavioural evidence from the individual animal, we have no way of knowing whether these claims are valid (Fernandes et al., 2021). Thus, what is target-oriented when striving for a scientific career is not necessarily suited for solving AHW problems that emerge from highly context-dependent dysfunctions. Striving for generalisable and publishable results widely disregards the inherent goals of organisms and the inner logic of farm systems. Thus, a fundamental trade-off is seen in the discrepancy between what animal scientists are striving for and what is required to grasp the complexity of processes within the farm systems. The outlined considerations explain why the predominant approaches in animal science primarily deliver piecework. Not unlike puzzle pieces, new findings generated from a disciplinary focus under specific experimental conditions are to the disposal of all who have access to scientific journals and are in search of information that might be of interest for them. This knowledge about tools, causes and effects which has emerged under specific condition can be classified as disposal knowledge (Mittelstraß, 2001). Disposal knowledge seems like a missing puzzle piece in search of a gap within a large mosaic picture where it might fit. Here again, a trade-off becomes obvious. To be generalisable, the shape of the disposal knowledge should be rather unspecific, whereas envisaged effects can be expected in particular, when the impetus fits to a high degree to the needs in a specific dysfunctional situation.

In the past, the overarching goal of livestock production has been to increase productive efficiency. Often this goal is equated with a continuous increase in the performance level and with benefits created by economies of scale. Disposal knowledge deemed suited to contribute to this goal was perceived with a high priority, explaining inter alia the high attractiveness of breeding strategies (Baumgard et al., 2017). In general, breeding strategies follow an inductive approach, moving from specific phenotypic traits to broader generalisations, and being widely context-invariant, i.e., a progress in relevant traits is expected to occur independent from the farm specific conditions. Some counter-examples of shifts in the approach, e.g. considering social factors (Ellen et al., 2014), try to ensure that selection takes place in farm environments and that the antagonisms between productivity and AHW is more considered than has been the case in the past (Brito et al., 2021).

The goal to reduce AHW problems is quite different from the predominant breeding approach. It requires a deductive and iterative approach, respectively. In addition, it is not an end in itself but a means to the overall goal of the sustainability of the farm system, setting a farm specific framework for all options of action. The reference systems are the individual animals and the respective farm system. Being a subordinated goal, the envisaged level of AHW is farm specific. Knowledge to transform a farm into a healthy agroecosystem must be designed to realise a win-win-win situation which conflates the interests of the farmer with those of the farm animals and public goods. However, before the relevance of reducing AHW problems for the economic viability of the farm can be estimated, various questions arise for the farmer, e.g.: Where are we now and where do we want to stand in the future? What has to be changed? How do we proceed and how can we assure that the implemented measures are effective and cost-efficient? And last but not least, how can improvements be measured?

Orientation concerns what is needed for an appropriate functionality of the subsystems, and what is provided by the superordinate system in terms of resources and what is demanded from the subsystem to contribute to the functionality of the superior system (see Figure 1 ). Correspondingly, both an external and internal benchmark system has to be considered. Being a relative parameter ranging from a low to a high grade, the AHW level can be benchmarked against the AHW levels of other farms, therewith classifying the farm’s own rank position in comparison to other farms. An example for an external benchmark is provided by a monitoring system for the assessment of cattle health in Dutch dairy herds based on routinely available data (Brouwer et al., 2015). This can be extended and combined with other dates, e.g. pathological meat inspections findings for a more comprehensive cattle health surveillance (Veldhuis et al., 2021). It is beyond the scope of this article to refer to the assets and drawbacks of surveillance systems and the use of iceberg variables (Farm Animal Welfare Council, 2009) for the assessment of the AHW level in further detail.

In the current context, it is emphasized that it is generally possible and already implemented in several countries to benchmark farm systems according to their degree of fulfilling public interests in relation to the AHW issue. In this respect, animal science has considerably contributed to the development of benchmarks (Mullan et al., 2016; Sandøe et al., 2020; De Luca et al., 2021). Benchmarking provides orientation. This applies in the first place for the farmers, who get to know where their farm system is classified concerning the AHW issue. Furthermore, this knowledge is essential in guiding farmers to define future target values with respect to the AHW level. Farm specific target values are the lighthouses for a reorientation and realignment of the farm practice and also relevant for designing transformation knowledge. Due to the general law of diminished marginal utility, the current rank position as well as the gap between the current and the envisaged rank position have a considerable influence on the cost-to-benefit-ratio of measures. Furthermore, benchmarking provides orientation for superordinate institutions with regard to the degree of variation between farms and the identification of those farms that are particularly in need of a transformation process.

The internal benchmark refers to the position of the AHW issue within the farm system and in relation to the overall goal of economic viability. For the farmers it should be of high importance to know, how far the current economic losses through AHW problems undermine the economic viability of the farm system. This also includes knowledge about the degree of AHW problems, the additional expenditures needed to solve AHW problems and to create a win-win situation that corresponds with both, the own economic as well as the public interests. While it certainly goes beyond the scope of the capacities of the farm management to deal with all farm animals in a comprehensive individual manner, the starting point when trying to reduce AHW problems should not rely on average figures but should focus on those animals, which are affected with the most severe AHW problems. In a similar way, preventive measures should first be addressed to those animals being exposed to the greatest risks in failing to cope due to the largest gap between what is required and what is provided by the living conditions. Thus, it is the variation and the animals at both ends of the distributional width that count both in relation to AHW problems and cost-to-benefit ratio.

These hints may suffice to understand that solving AHW problems on commercial livestock farms is above all an economic question. However, generalizations of economic effects may be fallacious for the individual farm (Jarvis and Valdes-Donoso, 2018). The economic impacts associated with differences in productivity often remain hidden under average figures. The various types of data from a farm, e.g. economic and biologic data, commonly lack connectivity and compatibility at animal level (Habel et al., 2021). Thus, neither agronomists nor other specialists, let alone the farmers themselves generally have adequate knowledge to their disposal to take appropriate decisions to strive for an optimal cost-to-benefit ratio with respect to the AHW level. What makes the situation even more complicated is that being aware of the core AHW problems on the farm and having defined new AHW target values does not necessarily mean the reasons and causes behind the problems are known, let alone the most effective and cost-efficient measures to reduce them.

Farmers will only agree in the implementation of measures to solve AHW problems, when they see a chance to benefit from such investments (Jansen et al., 2010). Measures can only be expected to be effective when they consider the individual gap between what is required and what is offered by the farm management and what is needed to support the individual animal to cope. Thus, the first step to improve the AHW level from a systemic perspective and by making use of deductive knowledge (from general to particular) is to get an overview of the possible shortcomings and AHW problems in the herd and to identify the most threatened animals. In the past, there have been tremendous efforts to develop technical options, e.g. by the use of ‘smart’ sensors and the principles and technology of process engineering, to identify possible deviations of numerous indicators from reference values (Berckmans, 2022). However, identifying deviations is not equivalent with a profound interpretation about their meaning. Interpretation is, among other things, aggravated due to a more or less extended overlapping area between false positive and false negative diagnostic results. Not all animals without obvious deviations are without AHW problems and not all deviations indicate an AHW problem. Above all, the information deriving from such techniques seldom generates knowledge about the causes behind the deviations.

Abnormal behaviour, symptoms of a clinical disease or deviations from physiological reference values indicate the need for further diagnostic procedures to assess the possible causes behind the disturbances and the degree of severity and potential threat for the recovery or survival of the animals. Diagnosis in individuals is deductive and goes from general signs to specific aetiologies (Slenning, 2001). When faced with a sick individual, the practitioner begins by making a list of “normal” comparisons. Each “abnormality” is a clue to the cause of the illness. Thus, a deductive process can be depicted as a series of questions. The answers to each question take the examiner down an increasingly branching decision process. Finally, a process of pattern recognition leads to a clear diagnosis when the disease pattern matches with official definition of the respective disease.

Dealing with all possible dysfunctions and disturbances that might affect farm animals in an appropriate diagnostic and treatment procedure up to a state of restitutio ad integrum is very time-consuming and costly and thus beyond what the farmers think is affordable. Instead, when striving for effort and cost saving strategies, a reductionist approach has been created. This tries to narrow AHW problems to main factors in relation to the affected animals within a group of animals while widely disregarding the individuality of AHW problems. Thus, diagnosis in herds or groups is in general inductive and goes from the specifics in a few individuals to generalities about the group (Slenning, 2001). Herd-based testing is done by subsampling a number of cows, representative of the animals at risk and then evaluating the proportion of cows above a certain cut-point. However, the signs are rarely precise. All the possible aspects which might have caused the current prevalence rate of AHW problems on the herd level will probably lead to a long list of factors. However, the list of factors does not provide certainties or probabilities with respect to the ranking of these factors in relation to the occurrence of AHW problems in the farm specific situation. To reduce AHW problems, disposal knowledge has to be contextualized. However, this does not work by the predominant inductive approach but requires an iterative approach that alternates between deductive and inductive conclusions, thereby strengthening synergies and resolving trade-offs.

Reducing efforts as part of a reductionist approach is not promising when simultaneously considerable negative side effects are disregarded, and particularly when these expand to a degree that undermines the economic viability of the farm system. Often, these become only assessable when biological and economic data are connected with the individual animal as reference system. Many farmers save themselves the effort of a time-consuming data acquisition and rely only on average figures. A study about the economic losses in connection with mastitis in dairy cows showed that it may be worthwhile to invest in a comprehensive documentation and analysis of data, so that it is clear where action is required and which investments can be expected to be financially feasible (Doehring and Sundrum, 2019). In contrast, average or incomplete figures risk giving the wrong impression and leading to false conclusions. As there is no general pathway to reduce AHW problems, also the detail depth of data acquisition underlies a context-specific cost-to-benefit ratio. The farm-specific degree and nature of AHW problems determine the efforts needed to overcome a negative marginal utility. The same applies for the disciplinary expertise needed to identify the main causes of AHW problems and to support individual farm animals in their ability to cope.

The trend in farm practise following the general idea of economies of scale when striving for the overarching goal of productive efficiency coincides with the reductionist approach and trend towards specialisation in animal science. Having widely lost sight of the whole farm system, animal science in its current alignment is neither able to provide orientation nor appropriate action-guiding knowledge. When focusing on selected sub-aspects from a mono-disciplinary perspective, the context in which AHW problems emerge as dysfunctional outcomes of the interactions between the individual animals and their living conditions is largely disregarded. Based on a reductionist approach, large quantities of detailed information have been accumulated with little attempt to incorporate it into a broad view of the whole subject and integrate it into a coherent understanding of complex systems. Thus, one of the biggest challenges for animal scientists is to interpret findings in a meaningful fashion in relation to the overall functionality of an agroecosystem. Otherwise, it is difficult to provide estimations about the effectiveness of measures in reducing AHW problems. On the other hand, these estimations are the starting point in finding the optimal cost-to-benefit ratio of measures. The options to improve this ratio is a main financial source and at the same time a main driver that might encouraged farmers to engage in efforts to reduce AHW problems.

It is a truism that conditions that require improvements need to be assessed, measured and the success of implemented measures whenever possible validated. Otherwise, it will not be possible to act against the barriers of established ways of thinking, of production processes or clashing interests within or outside the farm system. As AHW is a very complex issue, where substantial progress has not yet been adopted in many areas, possible improvements are associated with fundamental challenges, requiring systemic changes as well as an orientation towards external yardsticks. In light of the necessity to follow a context-variant approach when trying to solve AHW problems, it could be argued that the transformation of livestock systems is not a task for animal science but for advisory services in the first place. These have the most nearly access to the conditions on the respective farms and might be able to link what is available from animal science as disposal knowledge with the specific living conditions. The fact is, however, that the transformation of livestock systems to healthy agroecosystems cannot succeed when it is not evidence based, being a core area of science.

Concerning the validation of progress in reducing AHW problems, benchmarking does not only provide orientation but functions as a yardstick, by which a possible progress can be assessed. Any changes in the rank position serve as feedback for the farmers whether they are on the right track or have to realign their efforts. Thus, the previous suggestions in the literature on how to benchmark farm systems in relation to the AHW level (Brouwer et al., 2015; Nienhaus et al., 2020) need to be developed further and to be transferred to all farm animal species. Concerning the manifold options and strategies that might be considered when trying to reduce AHW problems, it should not be left to the primary producers and the advisory services to find solutions in a trial and error approach. This could be very expensive as well as inefficient and might prevent many farmers to walk the path towards improvements. With respect to the effectiveness of measures to reduce AHW problems, it is not the internal validity within highly standardised studies that counts in the first place but the external and ecological validity of scientific results.

External validity refers to the robustness of a causal relation outside the narrow circumstances in which it was established and thus defines the extent to which a scientific result can be generalised (Richter et al., 2016). A low reproducibility of results calls for research into practicable and effective ways of systematic environmental heterogenization. Concerning AHW problems, it should not be the primary research question whether various parameters show a significant difference in mean values when compared in a ceteris-paribus approach, but if the strategies are successful in reducing the portion of animals with AHW problems or in increasing the portion of animals that are able to cope, respectively. Future research should aim to collect longitudinal data on single animals and single farms concerning a wide range of AHW problems to be able to follow individual animals and farms and to identify possible patterns over time.

Ecological validity is aligned to real-world phenomena. In light of the complexity of AHW problems within farm systems, it is the whole system that functions as a reference system. Instead of large-number trials, focussing on single interrelationships in cross-sectional studies (Rittweg et al., 2023), single case studies are suggested to advanced understanding of systemic interactions (Teixeira de Melo et al., 2020). This approach does not only fit to the complexity of farm systems and the development of multifactorial AHW problems but also to the fact that many dysfunctions and problems by which farmers are challenged have a number of equivalent solutions (Sundrum et al., 2016). In order to capture and reduce the complexity of AHW related factors on farm level, a farm centric and equifinal approach is suggested to reduce AHW problems by making use of a participatory approach (Krieger et al., 2017). Such a concept was developed to enable farmers to identify effective measures and to invest resources more efficiently. Case studies can be the basic to subdivide systems and subsystems according their starting and their boundary conditions and according to their needs and regarding the effectiveness of measures to improve AHW. In a further step, the external validity of single-case research should be evaluated across a collection of comparable studies in a meta-analysis to examine whether study findings can be generalized to other contexts. Outcomes from a meta-analysis may include a more precise estimate of the effect of treatment or risk factor for disease, or other outcomes, than any individual study contributing to the pooled analysis (Lean et al., 2009).

While farmers are striving for the most cost-efficient measures when considering to reduce AHW problems, there is little knowledge to be gained from traditional animal science. Those studies that linked animal health measures to farm level economic outcomes reveal a heterogeneous body of evidence in terms of both methods, animal health measures and economic outcome measures used. In a recent review, none of the included studies made explicit causal claims between AHMs and economic outcomes (Sandeberg et al., 2023). From an economic point of view, farmers can select between different options to deal with AHW problems. E.g., they can treat animals according to their individual AHW problems and support them in their ability to cope. It is also possible, to abandon them to their fate, or to slaughter them ahead of schedule. Furthermore, they can take the occurrence of specific AHW problems as a reason to rearrange the living conditions and management strategies and to change the allocation of available resources in order to prevent or reduce the re-emergence of AHW problems. In this context, regulatory authorities as well as food processors and retailers play an important role to provide feedback and make clear that decisions about the farm specific degree of AHW problems are not completely left to the self-referential estimations of the farmers.

How such decisions are made is seldom a careful balancing between the different options and their respective cost-to-benefit ratio. To do so, farmers would need data, which they in general don’t have and a perspective that differs considerably from a strategy that seeks for productive efficiency primarily by measures going along with reduced efforts. Those enclose, among other things, a limited willingness to invest in comprehensive data collection and processing as well as in diagnostic procedures on the animal and farm level to gain knowledge and insight into the animal and farm specific background of AHW problems. Instead, it is about considering the alternatives. It would request a change in the mindset to focus on the disadvantages, farmers will face if they don’t implement what would be appropriate in the case at hand, as opposed to the disadvantages they will face if they do. Instead, farmers often apply rules of thumb, which they seldom change over the years, for instance when considering a culling strategy regarding the most common reasons for culling (Kulkarni et al., 2023).

On the other hand, regular veterinary visits might be seen as an obvious strategy to provide opportunities for constructive conversations between veterinarians and farmers and for shifting management from a reactionary approach to proactively optimizing health and welfare. However, a recent research study in dairy farming revealed that veterinary visits intended to improve herd health and production management are not necessarily effective in questioning predominant management strategies (Ritter et al., 2021). In a study, a mean of 51% of the visit duration was dedicated to transrectal pregnancy and fertility diagnostics, and a considerable amount of time (30%) was spent on visit preparation, transitions between tasks, and leaving. Mean length of discussions was 2 minutes, and only 17% of the visit duration was spent discussing dairy-specific topics. Such insights in the relationship between farmers and advisory services suggest that it cannot be left to the practice alone to deal with AHW problems, when they are at the same time considered being of public interests.

Instead, a retrospective approach seems a promising option to understand the background behind the processes. For instance, raising the proportion of ‘profit animals’ seems a suitable strategic goal to provide orientation for the farm management and validation for implemented measures that consider heterogeneous farming conditions (Habel et al., 2021). While retrospective analyses exclude the possibility to react in real-time for individual cows, the monetary valuation associated with service life characteristics can help to prioritize when choosing between different options of investments. E.g., a retrospective examination might conclude that the living conditions need to be shaped according to the individual requirements of the animals to allow for a desired outcome, rather than to strive for measures that generalize animal and farm performance. Evaluation of the overall lifetime profitability along with identification of economically poor farm-specific herd characteristics can help to pinpoint problems, optimize herd management, and prioritize investment necessities. Joining economic results and AHW of individual animals is a way to change the perception of AHW problems from collateral damage to a cause for severe economic losses. The corresponding knowledge might foster farm individual, iterative change processes, aiming for less AHW problems and for a higher cost-to-benefit ratio to the benefit of all stakeholders involved (Hoischen-Taubner et al., 2021).