Perspective Chapter: Exploring Multifaceted Approaches to Enhance Dairy Cow Welfare

Abstract

This chapter covers welfare issues observed worldwide in dairy cattle farms and outlines as its main objective, welfare indicators, including animal as well as resource-based concerns. Appropriate microclimate conditions and some useful indicators that can be numerically evaluated are highlighted e.g., body hygiene score, gait score, and dairy cow pain scale. As in other species, the ultimate goal of welfare auditing in cows is to use integrative systems or protocols, including those most relevant to the aforementioned indicators, and in particular numerical integrative ones, allowing even for the comparison of different farm conditions or husbandry systems in terms of overall herd welfare. Future challenges for improving cows’ welfare are the expansion of publicly available databases resulting in different in-field assessments worldwide, the widespread use of environmental enrichment and smart farming technologies such as automated monitoring systems and virtual fencing, as well as refining the assessment protocols and using dedicated apps to make implementation faster.

1. Introduction

According to the current largely accepted definition, animal welfare is the crossing zone of three areas of influence: animal physiology, animal behavior and the environment [1]. Only by approaching this concept in such a manner can one objectively evaluate it at the individual level (the key elements being the intensity of poor welfare and the time the animal experiences such negative situations) and at group level (the additional element being the number of affected animals), as well as develop different assessment systems allowing for the comparison of different groups of animals or different farms in terms of their welfare status.

While the notion of welfare refers to a specific, distinctive state of animals, the notion of animal protection includes all the measures taken to ensure and preserve the optimal welfare at both individual and herd level. This begins with having a legislative framework in place, and is managed with high standards and practices. This includes good farming, hygiene, biosecurity and transportation practices and ends with the manner in which they are implemented by all involved parties.

Animal welfare is a matter of major concern and debates in modern society and has been addressed by governmental as well as non-governmental organizations in the field. In this regard, the Council of Europe issued many conventions for animal protection e.g., for farming purposes, during transportation and slaughtering, for animals used for experiments and scientific purposes. The World Organization for Animal Health (WOAH) issues and revises on a regular basis, the welfare standards in the seventh section of its Terrestrial Animal Health Code [2]. World Animal Protection (the former World Society for the Protection of Animals) created a synthetic tool for evaluating the animal protection level in different countries and regions worldwide: Animal Protection Index [3]. Some other organizations pertinent to this matter include: International Society for Animal Hygiene, United Nations, Food and Agriculture Organization, World Health Organization, and European Food Safety Authority (EFSA). The Directorate-General for Environment of the European Commission collects the welfare-related data on animals used in scientific research activities, draws systematic reports and manages online Alures database [4], while International Air Transport Association delineates and constantly updates its own Live Animals Regulations.

One Health has already become a well-known concept, yet it continues to be refined and further developed. As a pinnacle concept, derived or included in the holistic, integrated approach of One Health, the One Welfare concept transfers the problem of animal welfare to a much wider ideological field. The concept of One Welfare intertwines the welfare of animals, the physical and mental welfare of human society, as well as the protection and preservation of the environment [5, 6].

Similar to other species, in order to optimize the workflow in cattle, farmers have had to handle significant management difficulties that arose from intensive dairy farming and technology. Nevertheless, the ultimate goal in designing and approving new animal housing systems, technological or handling systems, and in issuing new welfare standards, is to avoid neglecting the main character: the dairy cow.

Worldwide, working protocols have been established to guarantee that all the parties engaged in dairy farming adhere to at least the minimum welfare standards. Some of these protocols have been extended to other related cattle species, including exotic and wild animals.

2. Welfare issues in dairy cows

Even though the life expectancy of a dairy cow under optimal welfare conditions can easily exceed 20 years, in some farms, only 25% of this average is achieved, representing the productive lifespan. This is due to the intensive measures applied to increase milk production, which results in cows being culled and slaughtered after they complete their third lactation (2.5–4 years of age) [7]. Culling decisions are frequently influenced by the availability of genetically superior replacement heifers and by declining welfare standards prior to culling. Although management aspects suggest an annual culling rate per dairy farm of over 20%, farmers are frequently forced to cull females due to issues related to subfertility (17.7%), locomotion (14.3%) and udder health (12.1%) [8]. Technologies that enhance culling decisions will contribute to maximizing each cow’s economically ideal productive lifespan.

According to the Terrestrial Animal Health Code of the WOAH [2], dairy cows are kept in confinement (free stall, tie stall), semi-confinement, pasture-intensive or extensive housing conditions. In line with the scientific opinion developed by the EFSA, there are several ways to house cattle, including tie stalls, cubicles (free stalls), open-bedded housing systems (with the litter consisting of straw, compost, or dry manure), and systems with animal outdoor access (to pasture or outdoor resting areas) [9].

Generalizing from any farm to the whole husbandry system is dangerous since systems frequently succeed or fail due to the finer points, such as the methods in which they are set and operated. Numerous factors, such as the degree of social interaction the cows experience, whether they live indoors or outdoors, and the amount of space allotted, are highly variable throughout housing systems [10]. For instance, the cows kept in tie stalls, also known as stanchions, are confined to an area with an elevated floor for laying down and standing, allowing also individual access to water and food. Typically, the stall is the location in which the cows are also milked. However, this housing type restricts the animal’s freedom [11]. Cows may remain tied for months if they are milked in the stall. Moreover, animals have limited opportunities for intimate physical contact with their conspecifics. The animals may not be able to fully groom every area of their body when restrained in the stall, as they can not turn around. Animals in free stalls, also known as cubicles, are allowed unrestricted access to a laying bedded area, but also to other distinct areas for feed and water. The cows are often milked in an automated milking system or a central milking parlor. Such systems permit animal interactions and the opportunity to avoid dominant, aggressive conspecifics more than in the case of open-bedded systems. For open-bedded systems, the so-called straw yards, the two biggest risks in reducing cow welfare are inadequate space allocation per cow, generating an increased rate of antagonistic behavior and unhygienic resting places [12, 13].

The primary risks to animal welfare associated with managing cows on pasture are: inadequacies in protection from harsh weather, water availability, nutrient supply, and control of parasites, as well as unkempt walking trails or roads, and cattle charging each other [9]. Rarely is individual feeding feasible; therefore, animals may have to fight for few feed resources and must consume a diet designed for the typical cow rather than for the individual.

In the free stalls system, the females are usually cleaner and cleaning is made easier by the installation of individual stalls. The animals can also retreat from hostility in the stalls, but there might not be enough cubicles for each to lie down, further adding to the competition for resources [14, 15].

A recent study described that the risk of developing various mammary gland infections was decreased in females reared in the tie stall system compared to those reared in the free stall system. Moreover, in the case of the latter, somatic cell count, the most frequently used indicator of subclinical mastitis, shows higher values [16].

One of the most significant tasks a farmer completes each day is milking [17]. The cows must participate in this maneuver daily; therefore, stressors, such as long-term inflammation or improperly set milking equipment, must be minimized. Special attention should be paid to teat lesions and clinical mastitis, especially when cows reach peak lactation. It is commonly recognized that the milker’s stimulation of the mammary gland causes the posterior pituitary gland to produce more oxytocin, which in turn triggers the process of milk ejection [18]. Stressful stimuli including discomfort, an unfavorable setting (new females or heifers being milked for the first time), fear of milkers, and social isolation, prevent cows from releasing oxytocin, which results in forced, insufficient, and decreased milk production. Females will react to these stressors by vocalizing, urinating, defecating, becoming agitated and worried, and even trying to strike when they are in pain. Additionally, milkers will need to take extra precautions to prevent injury to themselves [19, 20, 21].

It is advised to reduce harsh handling and crowding of animals in the milking parlors and lanes, increase contact with the animals, and decrease anxiety by approaching them calmly and utilizing positive verbal interactions. Such behaviors help to avoid such strikes and, more importantly, to prevent compromising their welfare. By choosing robot milking, stress related to personnel or overcrowding during milking can be eliminated completely.

Besides mastitis, the most common mammary gland pathology in this species, other conditions such as lameness and nutritional metabolic disorders e.g., hypocalcaemia, ketosis, and displacement or torsion of the abomasum, are also encountered on dairy farms [22, 23, 24]. In addition to these diseases, reproductive disorders are the main cause of animal culling. It is well known that clinical mastitis is a condition that induces discomfort, especially those caused by contagious pathogens [24, 25]. However, treatment and dry cow management have succeeded in reducing such stressful stimuli. Lameness is most often caused by localized lesions in the soles or walls of the hooves. Injuries often occur due to unsuitable floor surfaces, poor maintenance and potentially contagious infections.

Other welfare concerns that garner a great deal of attention include undernutrition and heat stress. Negative energy balance (NEB) is of particular concern. This frequently arises peri-parturition, when feed intake is decreased and milk production increases [26, 27, 28]. Heat stress is becoming more significant as a result of global warming.

Regardless of which of these disorders is under discussion, their severity and duration must be taken into account. The more pronounced they are, the more the welfare is altered.

Dehorning and disbudding are leading causes of stress and pain in calves and adults due to restraint and painful stimulation that accompany these maneuvers [17]. Nevertheless, these procedures benefit long-term welfare and prevent pain-causing punctures and trauma between the congeners or towards the workers. However, they must be conducted in such a manner that pain and distress are minimized. Whether we discuss the application of this method to adult or juvenile animals and regardless of the technique used (guillotine or Gigli saw vs. cautery with hot iron or paste), pain management must be taken into account in order to limit discomfort (local anesthesia, blockage of horn nerves, sedatives and non-steroidal anti-inflammatory drugs) [12, 29, 30].

Dehorning is a recommended procedure in dairy cows; however, tail docking is not justified unless the aim is to limit the fouling of the udder with feces. Alternatively, this can be achieved by simply trimming the tail hair. Tail docking can cause pain similar to that experienced by people with phantom limb post-amputation pain. In addition, this surgical procedure will induce a long-term restlessness in animals and limit their desire to interact positively with humans [30].

3. Objective welfare indicators used in dairy cows

Auditing the welfare level on farms involves tracking and rating various objective welfare indicators (OWI). These indicators can be classified in different terms: depending on the duration in either long- or short-term indicators, depending on their effect on animal welfare in either poor or good welfare indicators, or depending on their nature in animal-based, resource-based and management-based indicators. The class of the animal-based OWI is represented by those that can be directly measured in animals e.g., body condition, physiological indicators, the percentages of different disorders, and the prevalence of non-adaptative behavior patterns. The engineering-based indicators (resource- and management-based) are the variables that are measured indirectly, not in animals, but in their environment. This includes the size, design and technical conditions of their enclosures, the state and efficiency of equipment in the barn, microclimate quality, and the provision of fodder and water. Also included are farm management protocols such as: environmental enrichment, biosecurity, disease control, animal handling, dehorning, breeding, and record keeping [2, 14, 31, 32].

In the following sections we delineate the major categories of indicators for assessing animal welfare in the dairy cow industry, beginning with resource and management-based indicators then discussing animal-based indicators.

3.1 Resource and management-based indicators

3.1.1 Space allowance and resting area design elements

The amount of usable space is the most important criteria for ensuring optimal welfare in animals. Providing adequate space helps to maintain the normal social structure within the herd and helps to mitigate aggressive behavior patterns commonly seen in overcrowded conditions. Similar to humans, animals have personal and social space. Using the concept of distinctive pressure zone, flight zone and fight zone should be the key element in issuing procedures and practices for approaching animals and handling them properly, while minimizing distress. The three mentioned zones are concentric areas, around the animal body:

• The pressure zone is the widest zone in which another animal, a person or an object passing the borders is perceived or observed.

• The flight zone is narrower. Here the approaching of other animals, persons or unfamiliar objects will trigger the avoidance, or flight response.

• The fight zone (the paradoxical attack zone) is the area closest to the animal’s body. The fight response is caused by an intense level of fear generated when an unfamiliar or sudden movement of an object or person is unexpectedly perceived inside this zone.

The three areas are not circular but rather elliptical in shape. The borders of these areas are further in the front of the animal (the section with the highest visual perception) and closer on the sides. The basic approach for appropriate handling and leading the animal, is that the stockperson or the handler moves in a triangle pattern, entering the flight zone from the direction of the animal’s head, then passing by the balance point to the hind quarters, then leaving the pressure zone and returning to the starting point. Such a movement pattern (Figure 1) will effectively use the pressure zone to naturally direct the animal to move in the opposite direction [33].

Figure 1.

Using the flight zone for driving the animals – after T. Grandin concept, modified.

The appropriate use of the flight zone concept significantly reduces the negative interactions between the stockperson and the cow and will prevent wilful acts of abuse when driving the animals (such as slamming the gates over the animals, hitting or beating them, using electric prods on sensitive parts of the animals – eyes, ears, muzzle, belly, anus or vulva – for forcing them to move). Such wilful acts of abuse are included among the criteria of Hazard Analysis and Critical Control Points (HACCP) – based welfare auditing protocols in cattle during transportation or before slaughtering, and they are considered major incompliances, with a zero-tolerance policy [34].

Increasing the radius of the flight distance in dairy cows is a reliable indicator of poor welfare, revealing an increased level of fear of humans and unsuitable stockmanship conditions. According to Temple Grandin [33], the common flight zone boundaries in cows are different when the person approaching is on foot or on horseback. A person on horseback can approach as close as 2 meters before the cow will attempt to flee, while for a person on foot the distance becomes even longer at 4–6 meters. Some other sources [35] used this specific indicator to evaluate good human-animal relationships, based on the distance and the percentage of animals in the herd with withdrawal reactions. The operating procedure described in the integrative Welfare Quality assessment protocol for cattle, namely for lactating and dry (pregnant) cows, consists of approaching an animal using the manger by an assessor, with the movement starting point in front of the animal, at a distance of 250 cm. The assessor will advance in an angle of 45° to the animal’s longitudinal body axis, at a speed of one step per second, with palm up, trying to touch with the posterior hand the cow’s muzzle. The animal avoidance (withdrawal) distance is defined as the space between the assessor and the cow, in the moment of observing that the animal moves back, turns its head to the side, shakes it, or pulls it back trying to move away from the manger. Based on the withdrawal distances recorded for all the animals in the herd, various relevant calculations have been made. These include the percentages of animals that can be touched, of animals that can be approached closer than 50 cm but not touched, of animals that can be approached without avoidance at a distance ranging 50 to 100 cm, and of animals that cannot be approached as closely as 100 cm.

Regarding the dairy cows’ space requirements in barns and the resting area design elements, recommendations from the scientific European and North American literature [9, 14, 36, 37, 38, 39, 40, 41] are synthesized in Table 1, depending on the applied solution for inner space division in the barn:

• straw yard system (or the so-called open bedded conventional system, or free walking/deep bedding system). This is the most unrestrictive, welfare-friendly system, giving the animal complete freedom to perform lying down/rising behaviors and allowing them expression of a larger spectrum of social and exploratory behaviors. The inner space of the barns with straw yards/collective pens can be organized in different manners: two area straw yard system (a feeding area completely separated by the straw-bedded area) and multiple area straw yard system (three distinctive areas in the dairy barns: straw yard, exercise or resting area, and feeding area);

• free stalls (cubicles). Such a solution for the barn inner space division implies the use of these designated areas only for resting, no feeding or watering equipment is provided in the stalls, so the cows need to leave the stalls and walk to other specific barn areas for feeding or drinking. Each cow must be provided at least one cubicle. The recommendation is to provide a number of extra cubicles equal to 5% of the number of dairy cows that will be accommodated in the barn.

Straw yard system	Free stalls (cubicles)	Tie stalls (stanchions)
Minimum 6 m^2 per cow, recommended 8–10 or even more for horned cows – to be possible to avoid the conspecifics [9] At least 5 m^2 per animal for dehorned dairy cows, at least 6 m^2 per animal for horned dairy cows (optimal more than 8 m^2 per animal regardless horned or dehorned) [14] At least 5 m^2 in cattle with the body weight of 350–500 kg, for heavier cattle 1 m^2 per 100 kg of live body weight [36] Minimum lying areas: for two area straw yard system the surface per cow is calculated by applying the eq. 1.65 x CH x [0.85 (DL + CH)] and for multiple area straw yard system 1.38 x CH x [0.85 (DL + CH)] [41]	Cubicle width: 0.83 x CH [37] Cubicle length: for the space sharing ones (e.g. open front, head-to-head) 1.8 x CH or 1.06 x cow DL + 0.48 x CH; for the closed ones (with the front against a wall) 2 x CH or 1.06 x cow DL + 0.65 x CH [9, 41] Cubicle resting length (from the rear curb/kerb to the brisket board = the area covered by bedding or rubber mats): 1.1 x DL Head and lunging space (available free space in front of the brisket locator for head movement in lying cows and for forward lunging when cows are rising): 0.65 x CH Neck rail distance (distance between the top of the neck rail or neck strap and the kerb): 1.2 x CH Neck rail height: 0.8–0.9 x cow DL or cubicle resting length + 0.1 m or 0.83 x RH [9, 37, 39, 41] Brisket board height: less than 10 cm above the stall surface, to prevent diagonal lying and lunging The distance from the top of the brisket locator to the loop (the bottom pipe of the stall partition/divider): more than 12–13 cm [38, 40]	Stall length = 0.92 x DL + 0.3 m Stall width = 0.86 x CH for cows with a 550 kg body weight: stalls of 1.79 m length and 1.15 m width at minimum, for cows weighting 750 kg stalls of 1.9 m length and 1.25 m width [9, 41]; Tie rail height above the cows’ feet (tie rail or head rail is the pipe for attaching the tying chain) = 0.80 x RH Tie rail mounted at 1.12–1.22 m higher than the bedding and 20–30 cm forward from centre of the manger curb, over the manger [38].
Advantages: more freedom to express lying down and resting behaviors, improved health of the limbs and hooves, increased animal locomotion possibility; Drawbacks: reduced animal hygiene state (lower hygiene scores due to the dirtier bed in the lying area), increased risk of udder issues [9]	Advantages: better hygiene state, cleaner animals; improved udder health in some cases; Drawbacks: issues in expressing resting and lying down/rising up behaviors even generating hooves injuries and lameness, increased risk of skin alterations [9]	Advantages: reduced antagonistic interactions between the herd conspecifics, reduced prevalence of some hoof disorders as digital dermatitis and white-line disease; Drawbacks: very restrictive for animal locomotion possibility, inhibiting social behavior, lying behavior and oestrus behavior, increased risks of skin lesions, hooves’ issues and lameness

Table 1.

European and north American recommendations for dairy cows resting area size.

CH = cow height at withers, DL – denoted by L in some standards = diagonal length (between the shoulder and the tail head), W = cow width between shoulders, RH = rump height.

Depending on the design, the cubicles can be classified as closed-front (against the barn’s wall), open front, deep bedded, mattress bed or head-to-head (opened) free stalls [42]. The design of the cubicles is based on optimization between two opposite elements: cow comfort and stall hygiene and maintenance. Two specific elements in the stall (Figure 2), namely the brisket locator (band/tube) – placed on the floor of the stall, and the neck rail or band – placed on top of the stall divider (loop), will minimize the soiling in the rear part of the stall by forcing the animals to land feces in the dunging alleys while they are lying, and when they are standing, respectively. Brisket locator position should also allow the cows to extend their front limbs when they rest and to step forward when they are lunging to rise [39]. An unsuitable position of the brisket board, closer to the rear curb/kerb (the rear border of the stall) can generate diagonal lunging or side lying;

• tie stalls (stanchions). The use of this system is often criticized due to close confinement conditions, and its use is declining more and more in favor of the loose-housing systems. The animals are tethered in the barn, year-round or only during cold periods, and are provided different degrees of access to pastures for grazing in the summer period.

Figure 2.

Dairy cows’ housing solutions: Straw yard system (left) and free stalls (right, with division elements highlighted). 1 = brisket locator, 2 = neck rail, 3 = partition (loop/divider), 4 = rear curb (kerb).

In accordance with the provisions of the EU Regulation 2018/848, in organic dairy farms, tethering is accepted only for farms with less than 50 animals (excluding the young stock) and if the animals are untethered at least twice weekly to access outdoor resting areas when grazing is not possible during the summer period.

The actual perspective in drawing welfare standards addressing usable space and divisions design and dimensions for dairy cows is to express the divisions’ space requirements based on different animal measurements. This is a reasonable perspective, given the premise that each division must ensure good conditions for the animal, and must fit its size. In some cases, using only numerical absolute values for surfaces or distances can often generate technical difficulties or design errors and poor welfare. The current literature in the field recommends to use equations based on ratios of body dimensions.

3.1.2 The comfort around the resting area

The importance of comfort in the lying area is undeniable, considering that dairy cows spend 10–14 hours a day resting. Aside from meeting the space and design detail requirements previously presented, the comfort of the resting area is highly influenced by its degree of thermal insulation, softness, dryness, cleanliness and slipperiness.

The use of bedding in the resting areas is essential, significantly increasing the air effective temperatures felt by the cows, and reducing the stress on limbs and hooves via the shock-absorbing capacity. In compliance with European standards [9], the following materials and technical solutions may be used for the floor in the resting areas:

• a straw bedding at least 6 cm thick when compacted, completely refreshed at least once daily at 2.5 kg per animal. Soiled and wet material is removed twice daily (for the deep bedding system, a straw bed of 30–40 cm thickness, completely removed and replaced with fresh material every 6 months and stirred twice a day for aeration);

• an inorganic bedding consisting of sand at least 15 cm thick [14, 39, 40, 43], refreshed once a day. This bedding has the advantage of reducing bacterial overgrowth;

• compost or manure bedding for the new compost-bedded-pack systems (CBP) ranging from 20 to 100 cm thick (recommended 50 cm, to create enough heat and maintain the fermenting process to effectively minimize the bacterial load);

• rubber mats (either soft foam rubber or plastic mats 3 cm thick is preferred to conventional hard rubber mats, or cow mattresses – two-layer type or shredded rubber filled type). Mats should be covered with chopped straw at 0.4–0.8 kg/cow in order to keep them dry;

• concrete, metal and plastic grids or concrete slats (slat width at least 8 cm and gap width no more than 3.5 cm [14, 31] – compared to solid floors, slatted floors generate stress on limbs and may lead to locomotion problems).

The cleanliness in the resting area is reflected in the cows’ body hygiene state. Inadequate dryness of the area reduces thermal insulation of the bedding, maximizes poor hygiene of the divisions, and promotes bacterial overgrowth with an increased production of barn gases due to fermentation processes. Resting area reduced grip may lead to altered rising/lying down movements and incidents such as cows slipping, falling or stepping on teats, which can create major problems in such large animals. To assess the quality of the flooring, some quick subjective tests can be performed: grip test, drop knee test and wet knee test. To test the grip, the assessor abruptly turns when pressing the floor with the heel of their rubber boot, using their body weight. Drop knee test consists of either crouching or dropping to the knees, any pain felt denotes an uncomfortable lying area. Wet knee test is conducted by kneeling in the stall for 10–30 seconds while wearing clean overalls or kneeling in the stall on a paper towel: if after this period of time the knees or the paper is wet, the dryness of the bedding is unsatisfactory [14, 43].

In the case of dairy cows’ tether housing system, the ease of movement in the stalls is also influenced by the length of the tie-chain. The tethered animal must have enough space to move side by side a distance of 0.4 m and back and forth a distance of 0.6 m; such motions’ magnitude indicates a minimum length of the tie-chain of 1 m. Research has shown that a tie-chain length of 1.4 m has beneficial effects on cows [14, 44].

The significant lack of comfort in the stalls’ resting area can be revealed by some animal-related indicators: abnormal postures (sitting dog postures, side or diagonal lunging); increasing of the time taken for a standing animal to lie down to more than 6.30 seconds; more than 30% of animals hitting the divisions’ components and equipment in the process of lying down; more than 10% of cows standing in the lying area; more than 5% of cows lying down outside the resting area; more than 50% of animals with dirty lower legs; and more than 19% of animals with soiled udder/flank and upper leg regions [35, 45].

3.1.3 The quality of activity-, feeding- and outdoor-areas

In compliance with the European requirements, to increase grip in heavy traffic areas (dunging alleys, passage and feeding lanes, alleys to the outdoor resting areas or to the dairy hall milking parlors), the concrete floor can be designed with draining grooves of 8–10 mm width and 2–3 mm depth, at distances of 10–12 mm. To allow enough ease of movement for the cows, the width of one-way passage areas must be more than 2 m, as cows require 0.9 m width to walk forward unimpeded, 1.3 m for turning 90 degrees and 1.7 m for a U-turn. For multiple activities’ areas (walking areas with watering or feeding facilities), the width should be more than 3.5 m to provide enough space for the cows to walk without interfering with other cows using the feeders or waterers. In the areas where the feed is delivered to the troughs by using a mixer wagon, the width of the tractor passage needs to be at least 4 m if the fodder is released on one side and 5 m when it is released on both sides (in front of two feeding barriers). Floors with poor design and maintenance generate locomotion problems and injuries, e.g. floors with high curbs, uneven or abrasive surface, or holes or sharp edges [14, 41, 46]. The use of slats, meeting the technical conditions described at 3.1.2., for the floor of this area ensures the reduction of slipperiness and more efficient drainage of feces.

According to the US American Humane Farm Program standards for dairy cows, a critical control point included in the welfare auditing system is slips and falls. It stipulates that the incidence of animals slipping is not to exceed 3% and incidence of animals falling is not to exceed 1% [47].

Technical conditions and the quality of areas for cow feeding (feeding barriers, racks or cribs, mangers, throughs) are also subject to requirements for optimal welfare level, as cows can spend 5 to 9 hours daily in this part of the barn. The feeding facilities should allow the animals to easily reach the fodder. This helps to mitigate antagonistic behaviors, thereby minimizing injuries and has the added advantage of preventing fodder wastage.

Ideally, each animal must have its own feeding location (post), however, in the case of ad libitum feeding systems, it is acceptable that up to 2.5 animals may use the same feeding place. According to European scientific literature [9, 41], the minimum length of an animal’s feeding post is estimated by multiplying the chest width of the animal by 1.3, the resulting values range from 65 cm (in cows of 550 kg body weight) to 83 cm (in cows weighting 850 kg). The diet composition and fodder hygiene and quality are the basic elements determining nutritional state of the cows. In addition to roughage, concentrates are given to cows in parlors during milking, or in outdoor parlor feeders/feeding stations throughout the day (each station providing concentrate for 25–30 cows, releasing the meals based on a computer-controlled system in relation to the milk production level, the animals wear transponders and are identified as they enter the stations), or added to the recipes of the total mixed rations rather than the use of feeding stations [41]. Adequate animal nutrition and feeding results in too many inputs to assess in a timely manner. This wide range of inputs involves the design, condition, and hygienic state of the feeding equipment and the nutritional and microbiological features of the fodder. Thus, the quality of animal nutrition is usually assessed based on a more practical approach, namely on body condition scoring of the animals in the herd.

The drinking area must provide permanent access to potable water fulfilling all quality provisions in the legislation: in EU member states the provisions of the 2020/2184 Directive. The recommended drinking water temperature range for dairy cows is 10–20°C, water temperatures lower than 10°C reduce milk production by 0.8 kg per day. Each animal should have access to at least two watering points, and be provided with clean, functional, and well-operating watering facilities. Regarding available space at the watering throughs, at least 6 cm must be allotted per cow for facilities with a water flow rate of more than 20 L/min and double the space for equipment with a flow rate less than 20 L/min. When water bowls are used, one bowl with a water flow rate of 10 L/min must be provided for 7–10 cows, and for 3–5 cows when the facilities have a lower flow rate [41, 45].

Because of the overwhelming beneficial effect of outdoor exercise on dairy cows, the quality of grazing areas and of barns’ outdoor resting areas contribute immensely to the optimal welfare of cows.

To reduce the welfare risks associated with extreme weather conditions for cows on pastures, the areas must be provided with windbreaks, shade and shelters. The shaded area, protecting the animals from the negative effects of direct sunlight, may be arranged in various technical options (permanent or portable shade structures, natural shade created by trees, or even solar panel shades charging cow feeders for sustainability) and ensures 1.8–2.5 m² per cow (3.7 m² in some standards) [47, 48, 49].

In order to promote animal locomotion outside the barns, the paddocks (outdoor resting areas) must allocate at least 3 m² per animal (for barns using a loose system) [14], and must be available at all times. In addition, the outdoor areas must have provisions for sun and wind protection as well as access to feeding and drinking equipment, and they must be clean, paved, and with a good surface grip. Research has shown that cows spent more time during the day in the paddocks when the outdoor space allowance was increased from 4 to 16 m² per animal. Moreover, it was highlighted the need to provide an outdoor bedded pack, especially for the housing system with no access to pastures [50].

3.1.4 The barn microclimate conditions

Even if cattle tolerate air temperatures as low as −20°C without major effects on health, the thermal neutral zone (TNZ) in dairy cows – the range of air temperatures for which only minor energy compensations are necessary for regulating core body temperature, and also the air temperature range compatible with an optimal health state and production level – is between +5 and + 20°C (at maximum 25°C). Calves require a higher air temperature. The upper critical temperature is 25°C, with the maximum permissible temperature at 28°C. For the air relative humidity, the recommended values are 40–80% in dairy cows, and 50–70% in calves, respectively [38, 41, 46, 51]. Because the effect of these two physical environmental factors in animals is highly interconnected, in order to interpret their overall effect, the synthetic Temperature Humidity Index (THI or heat index) is often used. In Europe, it is considered that THI values for the optimal welfare of dairy cows must be lower than 72 (69–75). Values between 72 and 79 reveal mild heat stress, moderate heat stress is seen between 80 and 89, and values higher than 80 result in severe stress [41, 52]. To estimate the heat or cold stress risk, some researchers promote the use of even more complex integrative indices, including the air temperature, relative humidity, and also wind/air draught speed, and solar radiation – Comprehensive Climate Index CCI [53].

To keep the apparent temperature (the effective temperature felt by animals) within acceptable ranges and to mitigate respiratory disorders and other adverse effects, the air draught speed should not exceed 0.2–0.5 m/s, the most dangerous location for recording air movements with high velocity is the cows’ resting area, where they potentiate the highest degree of body heat loss. However, during the summer, due to the positive effect of air draughts’ velocity on the apparent temperature in dairy cows, up to 1 m/s is an acceptable value [46]. Such speed values contribute to a reduction of the apparent temperature compared to the actual air temperature of up to 2°C in animals with long hair coats and up to 4°C in those with short hair coats.

The light quality is both a basic element in conditioning animal health and normal body development and a farm technology parameter used to control the level of animal production. The lighting systems in animal housing must take into account the peculiarities of vision in different farm species. Unlike humans (which are trichromatic), birds are tetrachromatic (with retinal cone cells sensitive to blue, green, and red portions of the visible spectrum, and some cone cells sensitive to UV type A radiation), and most farm mammal species are dichromatic (cone cells are only sensitive to blue and green colors in the visible spectrum, making them insensitive to red light). This is the scientific basis of using red light (red light bulbs of 7.5 W mounted 6–9 m apart and 3 m above the floor) for cows at night. This minimizes disturbances to the animals’ physical and behavioral biorhythms while allowing the stockpersons and veterinarians to perform simple activities and inspections in the barns. Sudden changes in light intensity should be avoided at all cost. Because cows take four to five times longer to adapt to changes in lighting than humans, a dimmable lighting system would have a positive influence [54, 55, 56].

Regarding the lighting period, for dry (pregnant) cows, short day photoperiods (SDPP) of no more than 12 hours of light per day are used. For lactating cows, long day photoperiods (LDPP) of 16–18 hours of light (produced by prolonging the daylight with a period of artificial lighting) and 6–8 hours of uninterrupted darkness are provided each day. The increasing of the photoperiod is done gradually over 2–4 weeks. Applying SDPP in non-lactating cows helps to “reset” the cows’ ability to increase milk production in their next lactation [56].

Regarding daytime illuminance, the European minimum value for dairy barns (resting areas, feeding areas, walking passages) is 100 lux, with a recommended range of 150–200 lux. During the night time hours, illumination of 5 lux should be provided so that the animals can recognize objects and eat. In milking parlors and dairy halls, light of 200–300 lux should be provided, and in areas where procedures requiring high visual precision are conducted (i.e. medical treatments), 500 lux is recommended. For natural lighting systems, the transparent openings’ area relative to floor area (windows to floor area ratio WFR) ranges from 8 to 15% (1/7 to 1/13). In the case of wider barns, a third of this area is represented by the surface of the transparent roof openings: skylights, ridge openings, or ridge lights [14, 41, 46].

For the noise level, the limits set for humans (workplace exposure limits WEL) seems to be valid in cattle too. According to European requirements, the noise intensity values for dairy barns and dairy halls must not exceed 65 decibels for a continuous exposure and 80 decibels for short-time exposure. In barns, and especially in the milking parlors, in order to create normal responses, the noise level must be kept as low as possible. Unexpected noises generated by equipment, machinery, doors, gates, feeding barriers, stockpersons and visitors shouting or speaking loudly, should be avoided [41].

For the gases in the barn air, the maximum permissible exposure limits issued by European recommendations are 3000 ppm for carbon dioxide (recommended up to 1000 ppm for very well ventilated or open barns), 20 ppm for ammonia and 0.5 ppm for hydrogen sulphide – concentrations up to 5 ppm of hydrogen sulphide in the air are acceptable while handling manure and during the slurries’ drainage and removal [46].

The biological air factors permissible exposure limits must also meet the cows’ welfare standards and the legal provisions concerning the occupational hazards for the personnel working in dairy farms. The statutory maximum permissible exposure values for stockpersons’ workplace safety vary depending on countries and are usually more restrictive than the values suggested for good hygiene and animal welfare.

Due to the very large variations of values for barn bioaerosols and airborne bacteria in relation to the husbandry technology and the quality of the housing conditions, there are undefined yet generally valid international standards. Some Romanian recommendations state the maximum estimative values in barn air, regardless of farm species, of 15 mg/m³ for total suspended particulates (TSP); of 17 g/m²/30 days for settling dust (SD); 250,000 colony forming units (CFU)/m³ for mesophilic aerobic bacteria, 125,000 CFU/m³ for staphylococci; 62,500 CFU/m³ for streptococci and Gram-negative bacteria, and 12,500/m³ for yeasts and molds respectively [57]. The European threshold limits set specifically for workplace exposure in the cattle husbandry sector, are the following: 3 mg/m³ at maximum, recommended for the total suspended particulates, among the TSP no more than 2.4 mg/m³ for inhalable particles and 0.23 mg/m³ for the respirable particles, respirable endotoxin no more than 50 endotoxin units EU/m³, total airborne bacteria no more than 100,000 CFU/m³ of air [41, 46]. Set limits in the US include a Time-Weighted Average Exposure Value (TWAEV) of no more than 10 mg/m³ for total dust and 4 mg/m³ for grain dust: oat, wheat, or barley [58].

3.1.5 Other farm flow management indicators

Good herd health entails defining and efficiently implementing clear management protocols for procedures such as: supplying good quality water and fodder to the herd from safe and trusted sources; suitable management of manure and slurries with a reduced risk for contamination and emissions into the environment; appropriate environmental enrichment protocols for maximizing stimulation and minimizing the incidence of non-adaptative or stereotypic behavior. It is equally important to develop and enforce protocols to ensure reliable farm biosecurity and biocontainment, diseases’ control, animal restraint, and pain management, for all animals on the premises.

For increasing production performance, a common practice in large conventional and organic farms, is to segregate the animals into groups based on age or production stage/purpose. Calves are separated from their dams after the colostrum period (24 hours) and fed with milk replacers (in conventional farms) or with whole milk, preferably from their own mothers (in organic farms) over a period or 3–5 months (early weaning) and 6–9 months (natural weaning). In addition to addressing the balance between calves’ welfare and financial concerns around the time of weaning, the cow-calf contact period is another major topic of debate. In some small-scale family farms (backyard cow farms) and some organic farm systems, social interactions are improved by keeping the natural cattle group structure (i.e. family herds) together. The dairy cows are not separated from their offspring and the bulls. Allowing calves to suckle on their dams or on foster cows (for economic reason the cows can be milked in addition to suckling) prolongs the calf-cow contact period beyond the colostrum period, with undeniable benefits on the welfare of both calves and cows [14, 59, 60].

A principle in reducing antagonistic cows’ behavior and competition for resources, is to minimize as much as possible, any change in the group structure (by adding new-comer cows or cows returned to the farm). In addition, some findings suggest that the incidence of antagonistic patterns (head-butting, head-to-head fights, displacement of other cows from the feeding areas) significantly decreases when they are housed during the post-partum phase, in pens of smaller groups (6 animals compared to 24) [61].

The presence and the accuracy of farm records are directly proportional to the quality of farm management. Farm registrars and electronic databases may be used to track and rate data regarding the numbers and description of all types of conditions and disorders. For instance, numbers of lameness treatments (they are not counted twice if they address the same limb of the cow within 14 days), lameness scoring (gate scoring), nutrition related health concerns (bloats and metabolic disorders), reproduction disorders (assisted calving, infertility cases or conception rates), and mastitis cases may all be tracked. Such records may also delineate body condition scoring, results of any fodder exam (e.g., residues, minerals) and metabolic profiles.

In farms, written programs for the prevention and control/treatment of specific conditions must exist: for lameness and limb disorders, for metabolic disorders, reproductive ones, mastitis, for contagious/parasitic diseases. Also, the farms must have good implemented general health plans, created with professional guidance and suitable to the particular features of the farm. The protocols should be reviewed and updated at least annually, and they should include information on diagnosis, treatments and the routes of administration, as well as the target or intervention levels. For organic farms, such general health plans include the strategies and evidence for decisions to reduce the use of routine medications (e.g., closed herds, rotation of land). The mechanism by which recognized problems on the farm are solved is important. This necessitates the creation of protocols defining the problems’ management, contingency plans for crisis situations and diseases outbreaks, and describing actions to be taken and their results [62].

Farm biocontainment and biosecurity programs are key elements in improving the welfare level, largely by preventing and reducing the incidence of infectious diseases associated with pain and suffering. Some sources use a metaphor useful in understanding and setting up biosecurity programs: all the farm premises are “castles” (e.g., the barns for housing the animals, the buildings for feedstuff or other sensitive supplies and material storage), surrounded by “moats” (uninterrupted line of separation LOS between the dirty, off-farm zone and the clean, on-farm zone), and the only access is permitted by “drawbridges” (LOS entry and exit points) [63].

The biosecurity programs rely on the most effective application of space/time segregation (separation). This separation applies not only to animals, fodder and water supplies, and milk delivering routes, but also secondary sources of contamination and fomites, such as stockpersons and farm visitors/suppliers/transporters, vehicles, clothing, or tools and equipment shared with other farms or between different animals or barns.

The provisions regarding the minimum safety distance between the cow farms and residential areas or other protected areas are very different depending on country and the number of animals on the farm. This can range from 50 m for small farms to 500 m or even 2 km for large commercial farms with more than 500 animals.

Premises with the higher risk of spreading pathogens (quarantine areas, manure and deceased animals’ storage) are always located at the farm boundaries, as far as possible from the dairy barns, and downwind. The farm layout must ensure a clear distinction, with no crossing between the farm paths serving the clean and dirty zones. Items such as farm tools and instruments and personal protective equipment (PPE) are also designated for either the clean or dirty zone, visibly marked (different colors) and not shared between the zones. If possible, the staff is distinct for each production group of animals or for each group requiring another level of biosecurity [64]. Animal inspection, handling or treatments is carried out in a specific sequence of most at-risk animals to least at-risk animals. Begin with young, then old, then healthy adults, then quarantined and sick animals. New animals added to the herds will undergo a quarantine period of at least 30 days according to EU 2020/692 Commission delegated Regulation. To reduce the risk of spreading pathogens between different production series after general prophylactic disinfection, a period of farrowing (a downtime) will be ensured, to break the pathogens’ cycles and to increase the contact time of disinfectants. Farm traffic is strictly monitored for vehicles and visitors, avoiding direct contact with the animals or feedstuff areas (by displaying warning signs or by taping) and directing them primarily to administrative or parking areas. The premises are protected from wild animals, other species of domestic animals, and from pests by installing bait stations and fences around buildings and using electric fly zappers, and sticky straps inside buildings [64].

For vehicle biosecurity at each farm, a wheel disinfection bath or disinfection gate is in place, and at each barn or animal area’s entry (LOS access point) other safety precautions are provided. This includes footwear disinfection baths, handwashing stations and hand sanitizers, and specific farm clothing or disposable overalls and farm boots/boot covers. Before accessing the calves’ rearing area, some authors recommend the use of a two-zone Danish entry (sanitary lock). This includes a physical barrier (a bench) set up in the middle of a separate space and delineating a clean and a dirty area. Individuals coming from outside the farm doff their street coveralls, sanitize their hands, remove their street footwear while sitting on the bench in the dirty zone, then swing their feet over the bench to the clean zone (without touching the dirty zone with their socks), where they put on farm footwear then dress in farm overalls or suits, and finally wash their hands before entering the animal area [65].

The efficiency of rodent and insect control, and disinfection procedures are vital elements included in the biosecurity programs. Additionally, vaccination and parasite control protocols, and good hygiene and safety of feedstuffs are equally important. It is also imperative that water is provided to dairy cows only from sources with known status.

3.2 Animal-based indicators

Numerous minimally invasive/no contact animal-based indicators can be used to assess the welfare level in individuals, then by summarizing the findings, at herd level. Table 2 shows the animal-based indicators regularly used to determine welfare level in dairy cows. Computer and mobile apps have recently been developed to shorten time required for observations, interpretation and generation of results during the in-field assessments of cows’ herd welfare level. These software (e.g., Body Condition Scorer, BCS Tracker, Loco Score, Healthy Hoof app, Step-Up, Locomotion Scorer, Cow notes) are able to collect and process the results of the evaluation of one or more parameters presented in Table 2.

Body condition score	Body condition scoring (BCS) is an animal-based parameter largely used to establish the nutritional state and energy balance (prerequisites for good welfare) in various species, including wild ruminants in natural habitats or in zoological parks e.g., buffalo, bison, deer, giraffes, elk [66]. There are some variants of scales applicable in domestic cattle: a 5-point scale (with increments of 0.25 points), an 8-point scale, a 9-point scale or a 10-point scale. The 5-point scale is the most commonly used [66, 67, 68]. It consists of evaluation of fat deposition in different anatomical regions: short ribs (transverse processes of the spine), backbone, hook and pin bones, thurl (the angle of the area delineated by the pin and hook bones of the hip), tail head and the visibility of sacral and tailhead ligaments. The recommended condition is “fit and not fat”, evaluations are carried out in the herd for all the productive stages: for cows at calving (target BCS of 2.5–3), at 60 days after calving (target score of 2–2.5), at 100 days before drying off and at drying off (target scores of 2.5–3). In emaciated, thin cows the bones are sharply defined, with no discernible fat pad, visible short ribs giving the loin a sawtooth, shelf-like appearance, the backbone vertebrae in the chine, loin and rump regions are easily seen as individual bones, the region above the pelvis – between the hook and pin bones, forms a deep V-shaped concavity (for cows having scores more than 3 thurl region is U-shaped), in the tailhead region are hollows on the sides, with obvious skin folds and prominent sacral and tailhead ligaments. Animals scoring 4 or 5, have obvious fat deposits, with a flat thurl area, the bone rays are rounded, being felt only with firm pressure. At herd level, thin cows (cows with score less than 2) and fat cows (with scores of 4 and 5), are in unsuitable states and indicate poor welfare [62].; A related indicator is represented by the rumen fill score (RFS) on a scale 1 to 5 points, which shows the cow feed intake in the last 2–6 hours. The system implies only observation, but can be completed by touching the left flank area, to feel the firmness of the rumen. The cows need to be evaluated daily. For the pregnant cows, the target score is at least 4 points – taking into account the large volume of the uterus, and for lactating cows, at least 3. Lower scores indicate that animals have not eaten, or are probably sick or injured. If more than 15% of the herd does not reach these targets, intervention is required. Animals with empty rumen show an obvious hollow in the left flank, after the last rib, a concavity deeper than one hand-width (score 1) or with the depth of one hand-width (score 2), having skin folds under the lumbar vertebrae [69].
Hygiene scoring	Body hygiene scoring (HS) reveals inappropriate management of manure collection and disposal or some technical deficiencies concerning barn layout and divisions. There are variants for determining the overall cows’ body hygiene or udder and teats hygiene. The first system consists of observing the extent of the soiling, manure or bedding material covering three body regions: lower rear leg – above the coronary band; udder; upper leg and flank. Each region can be rated with 1 point (no soiled areas), with 2 points (minor splashes), 3 points (distinct plaques of soiling), or with 4 points (confluent plaques, covering the hair). At herd level, for each of the three regions, the percentage of animals with scores more than 3 points is determined. The suggested benchmarks (best hygiene target levels) for cows in free stalls housing system is 24% on lower legs, 5% on udder and 6% on upper legs and flanks [70].; For establishing the udder and teats’ hygiene only, a dedicated scoring chart can be used. Scores of 1 point are given when the area is free of soiling, 2 points when it is slightly soiled on up to 10% of the surface, 3 points when it is covered with soiling on 10–30% of the surface and 4 points when more than 30% is soiled. Teat cleanliness scorecards have also been developed (clean swabs rubbed around the cow teat ends then estimating the amount of soiling in these swabs by using a 4-point scale – prior to attaching the milking units) [71].
Gait scoring (locomotion scoring)	The gait scoring or lameness/locomotion scoring (GS or LS) is a useful, reliable tool in herds for controlling hoof and limb condition, generated by unsuitable environmental conditions and to quickly formulate improvements. Performed at least once at 2 months intervals (recommended monthly frequency), with different variants (3-point scales, 4-point scales, and 5-point scales), the gait scoring protocol tracks different signs of pain, making the animal reluctant to walk or stand: arched back, head bobbing, shortened or uneven steps, limping, avoidance to bear weight on some limbs. Lower scores are given to animals walking and standing normally, whilst the highest scores designate seriously affected animals, such as those refusing to walk, with pronounced arched back and a complete weight transfer from the affected limbs [72, 73].; The Welfare Quality assessment protocol in dairy cows has established standards for the percentage of animals in the herd affected by one or several mild (not severe) alterations, and the percentage of animals affected by one or more severe alterations. It is considered that the severe gait alterations have five times worse effect on animal welfare (in the calculated weighted sum, the coefficient used is 1 for the percentage of cows with mild alteration, and 5 for the ones with severe alterations, respectively) [35].
Grimace score (cow pain scale)	The grimace score is based on tracking 7 criteria, each scoring either 0 (normal cow, no pain) or 1 point/2 points (animals with signs of pain). The first criterion is represented by the cow’s attention to the surroundings – the animals being classified as focused, attentive or absent. The second criterion is the position of the head compared to a horizontal line crossing on top of the withers – a lower position indicating pain. For the next criterion, the position of the ears – it is taken into account that cows in pain keep their ears backwards or even with the pinna facing back and downward, like in lambs. Regarding the fourth criterion, the facial expression – the specific pain face is revealed by tension of muscles above the eyes – seen as “furrow lines” and on the cheeks, tight lips, dilated nostrils, with lines above the muzzle. For the criterion response to approach – it is considered that animals in pain are trying to avoid interaction. Concerning the criterion the back position – it is considered that an arched back indicates pain. The last criterion is represented by the presence of lameness – indicated by careful walking, with nonrhythmic or shorter strides, with non-weightbearing due to limb pain. The overall score is reached by summing up the scores for all these criteria; values >5 points indicate animals in pain, requiring further medical examination [74].
Incidence of injuries, symptoms of diseases or other types of disorders, mortality	The percentages of animals in the herd showing some symptoms of diseases and lesions can be compared to warning thresholds (values beyond which decisions are made addressing the health plan at the farm level) and alarm thresholds (unacceptable values).; Different animal welfare assessment systems [35, 62] approach the following percentages in relation to the presence of injuries and of various diseases in the herd: animals with mild and severe integument and adjacent tissue alterations (lesions, swellings, or erythema and alopecia on a surface with the diameter greater than 2 cm) on any body region except on hocks. For rating an alteration as severe, at least one lesion or swelling must be identified; percentage of animals with swollen hocks (hygromas); animals with overgrown hooves (the well-trimmed hoof has a front wall length of about a palm width: 7.5 cm, angled at 45 degrees); percentage of the overall number of coughs recorded within 15 min from the total number of animals in the herd (warning value 3%, alarm value 6%); cows with nasal discharge (warning value 5%, alarm value 10%); percentage of cows with ocular discharge (warning value 3%, alarm value 6%); cows with abnormal respiration – deep, labored or difficult breathing – and those with diarrhea (warning values 3.25%, alarm values 6.5%); animals with bloated rumen; percentage of cows with mastitis – milk somatic cell counts of 400,000/ml or above within 3 months (warning values 8.75%, alarm values 17.5%); animals with vulvar discharge – purulent effluent or pus plaques (warning values 2.25%, alarm values 4.5%); percentage of animals with dystocia (warning values 2.75%, alarm values 5.5%); percentage of downer cows – non ambulatory cows during the last 12 months (warning values 2.75%, alarm values 5.5%); mortality on the farm during the last year (warning values 2.25%, alarm values 4.5%).
Behavior problems	Classified as breeding related, nursing related or miscellaneous behavior problems, some are indicators of poor welfare, either non-adaptative or stereotypical behavior (stable sequence with no useful result or function), or vices: inter sucking or self-sucking in adult cows (in some reports with an incidence as high as 25.2% in primiparous cows); bar biting (reported in some sources with frequencies up to 10% in bulls and even higher in early-weaned, individually-housed calves without access to resting area, and no bedding material to allow for oral activities); pica (licking or eating foreign objects such as the wall and the floor surface of the pen or manger, wood, rags, rubber); tongue playing or tongue rolling (animal thrusting their tongue out of the mouth and repeatedly moving, curling/uncurling outside and inside the mouth because they cannot completely satisfy their instinct of prehension, especially if they are not allowed to graze); feeding and watering related vices (feed tossing and feed dropping out of the manger, water lapping – cows licking water instead of drinking it), caused by the same exercise deprivation and lack of stimulation, making it impossible to express natural exploratory behavior on pastures; kicking and pawing, appeared because of inappropriate treatment and handling, or due to painful lesions on the udder/teats generating fear of milker manipulation; eye rolling (especially in cows without enough stimulation: animals are motionless, with repeated rolling of the eyes in the orbits not related to actual visual stimuli); head shaking or nodding (mostly in adult cows in close confinement); rubbing (sometimes generating local skin irritations); neonatal calves’ rejection (in first calvers with dystocia, in cows kept during the gestation in large groups, facing great competition for food, water and social ranking) [75]. In addition, a high rate of antagonistic behavior (chasing, pushing, head butting, hitting, thrusting, displacement of other animals in feeding/drinking areas) and of the number of animals suffering it’s effect is considered a useful poor welfare indicator, and is tracked in some welfare assessment protocols. Even the general behavioral states or expressions revealing animals’ positive or negative feelings (sociability, pleasantness, calmness, playfulness, agitation, boredom, fear, frustration), the manner they interact with the herd or with the environment (their body language) can be monitored by using a visual analog scale to conduct a qualitative behavior assessment [35].
Animal vocalizations	Vocalization rate, number of animals vocalizing or vocalization score (VS) is a formally used critical control point included in the US cattle welfare auditing system in slaughterhouses, revealing animal distress, agitation, inappropriate handling, restraining or stunning, using electric prods on sensitive body regions. The maximum accepted percentage of cattle vocalizing is 3% from at least 100 animals in large plants and 50 in smaller ones. Values of 10% indicate serious problems [34].; Lower-frequency vocalizations (closed-mouth vocalizations, murmuring calls, during resting and ruminating periods) is considered a good welfare indicator, as they indicate positive emotions [76]. They may be recorded by with microphones in the barn or collar-mounted acoustic recorders allowing individual assessment.

Table 2.

Examples of frequently used animal-based objective indicators in dairy cows.

4. Dairy cow welfare assessment

There are different systems and protocols used to rate the welfare of cows at the group level. Some assessment systems integrate only data collected on farms, by monitoring behavior indicators, auditing farm management systems and practices, or approaching a wider range of indicators from a holistic perspective: HACCP-based systems and integrative systems. Other systems involve processing of farm-collected data, refined by performing laboratory exams. Eventually, there are welfare assessment systems addressing only laboratory analysis (based on the interpretation of blood hematology and biochemistry panel, measuring manure cortisol concentration, endogenous opioid concentration, and animal immune status exams) [77, 78, 79, 80]. HACCP-based and integrative systems are most often used, as they have the major advantages of being reliable and cost-effective.

HACCP-based systems have greatly evolved, especially in North America. Temple Grandin made a great contribution to their introduction and development [34, 81]. At the beginning, they were used only for assessing welfare level of different species of meat-producing animals during slaughter and transportation. Over time, the use of these protocols extended to on-farm welfare assessment, or at least some of their critical control points (CCPs) became indicators included in the larger numeric systems dedicated to dairy cow on-farm welfare level assessment [47]. The CCPs suggested for their repeatability, specificity, and utility in cattle include: percentage of lame cows and of cows with leg and hock lesions and swellings, percentage of cows too thin (BCS less than 2), the percentage of calves that do not receive colostrum, percentage of cows and calves with poor body hygiene, percentage of downer cows. Other valid CCPs in cattle are: compliance with welfare guidelines when conducting surgical procedures such as dehorning, disbudding or castration, ammonia concentration in the barn air, complying with usable space requirements and housing technical specifications. Furthermore, CCPs in cattle are considered: percentage of cows with neglected disorders and health issues such as bad eyes, bald spots caused by external parasites, severe injuries, emergency milking protocols or systems in case of power failures, the concern about the welfare of young calves and heifers leaving the farm [81].

Integrative welfare assessment systems, especially numeric, are used in many species (farm-producing or wild animals, even dolphins: the C-Well assessment system [82]). Thus the welfare level is obtained in a numeric form as an overall, global result, so that comparing animals’ welfare levels between different groups, farms, or husbandry conditions becomes possible [83]. Different integrative numeric systems and protocols were developed during the last decades for establishing the dairy cow herds’ welfare level, including the Bristol Welfare Assurance Program protocol, the Animal Needs Index system, the Welfare Quality System, and the American Humane Certified Welfare Standards Audit Tool. The major challenges in developing such systems, with an indisputable scientific foundation, are to establish the weight, the influence of each included criterion (of each welfare indicator) for the unique output – which is the welfare of the entire group, and the potential ways the indicators may compensate for and interfere with each other in a joint result. Therefore, when designing such a system, multiple teams of experts and various professionals are involved, to address all three components of animal welfare: animal physical state, farm and outdoor environment quality, and eventually mental status (animal behavior). A good starting point for issuing welfare assessment protocols is to follow the model introduced in human psychology in 1954 by Abraham Maslow and to create hierarchies of needs based on all the scientific data available. This includes results of behavior tests for deciding the intensity of motivation, the preference and aversion for different resources and stimuli, and interpreting ethograms in the species of interest [80].

In this regard, the ANI 35 L system was introduced by Helmut Bartussek in 1990, initially for organic cattle, poultry, and pig farms. The abbreviation stands for Animal Needs Index, 35 is the maximum possible overall score in the initial version, and L denotes the “Long”, detailed version. However, in the subsequent revisions, the final score changed. In the final version of the protocol used for cattle welfare assessment (the revision from 2000), it ranges from −9 points to +46 points. The ANI 35 L cattle assessment procedure addresses a total of 25 welfare indicators or factors (28 factors for cattle with access to pastures), grouped within five areas of influence: locomotion, social interactions, flooring, light and air, and stockmanship. The system stresses the influence of housing conditions on animals (it is centered on management-based and resource-based parameters). Only seven factors are related to animals, namely the herd structure, calf and younglings’ management, condition of integument, cleanliness of the animals, condition of hooves, health state at group level, and the incidence of diseases generated by faulty farm technologies (injuries or damage resulting from any part of the barn equipment). Within the ANI 35 assessment sheets, the weight of the cattle access to outdoor resting areas and pastures represents about 35.7% (15 points) of the overall score, demonstrating the great importance of daily exercise and environment exploration for the welfare of the cows. Used in many field assessments [78, 83, 84, 85, 86, 87, 88], with a simple manner of interpretation, the protocol can be improved by including a hygiene score and a gait score for increased accuracy in the evaluation of some factors. To exclude the possibility of masking major issues by compensation of factors recording very low scores compared to highly rated scores, a check for compliance with the minimum welfare standards according to national legislation is also performed. Additionally, in cases where legal requirements are not met, the final score is considered provisional until the farm management solves the major deficiencies [14].

The Welfare Quality (WQ) system protocol used for assessing cattle welfare requires a more complex level of processing the data collected on farms. This includes an elaborate manner of interpretation, including the computation of weighted sums and means, elaborate equations and indexes (e.g., for Qualitative Behavior Assessment, a weighted index resulting from 20 descriptors), and the interpretation of l-spline functions. These procedural stages are usually performed with the help of template spreadsheets, calculators, or by using online databases [89]. The protocol tracks 31 measurements (welfare indicators), the majority of which (24 measurements) belong to the group of direct, animal-related factors, the WQ system is opposite to the case described for the ANI 35 system. The measurements are gathered into 12 welfare criteria (12 groups of measurements) made up of four principles: good feeding and watering, good housing, good health, and appropriate cattle behavior [35, 90, 91]. Used for many in field assessments, WQ protocols were developed within the European Welfare Quality Project for establishing the welfare level in pigs, poultry, and cattle farms and were adapted by some researchers for use in wild animals in nature reserves and zoos, e.g., adaptation of the cattle WQ assessment protocol in establishing the welfare level of Dorcas gazelles [92].

The cattle welfare auditing protocols for the American Humane Certified Standards summarize methodologies for welfare assessment in different locations: on farm, during transportation, or at slaughter. The auditing tool for farm level assessment (E-section: Environment/On Site) is a numeric integrative system [47], comprising of 48 welfare indicators (E1-E48), mostly regarding farm premises and equipment conditions and technical details, or farm management. Only a few factors are animal-related: udder condition score, incidence of slips and falls, locomotion score, hygiene score, leg condition score (hair loss patches, swelling), coat condition (bald spots in the cows’ coat), and tail condition (undamaged, unbroken tail).

Similar to other species, new trends in dairy cows’ welfare assessment at the farm level are generated by applying a fully automated, computer-assisted methodology (via dedicated computer software and barn sensors, fully exploiting the benefits of livestock precision farming technologies) or by the use of portable, mobile phone applications. [93, 94]

5. Dairy cow livestock precision farming

Labelling animal products with the level of welfare the animals experienced, has recently gained momentum in European markets [95]. However, existing animal welfare assessment protocols have certain limitations, such as: (1) they are mainly applied at the group level; (2) they are not able to permanently monitor animal welfare levels and highlight any potential changes over time; and (3) they are performed by human assessors, implying some level of subjectivity. Furthermore, these protocols are not reliable for detecting early signs of disease, which can inhibit the implementation of preventive measures in a timely manner. These limitations can be partially solved by the application of precision livestock farming (PLF) or smart farming systems and technologies. Precision Livestock Farming describes the combined use of sensor technology, and related algorithms, interfaces, and applications in animal husbandry [96]. Different PLF systems, some of which are available on the market, have been developed for monitoring either outdoor and barn microclimate data [93, 94, 97] or, by using wearable devices and sensors attached to the animals, for monitoring some animal-related parameters. Other useful data such as GPS location of animals on pastures for virtual fencing technology [98] or their position indoors, reproductive or milk production-related parameters, smart feeding, rumination, and nutrition parameters as well as health status parameters [94, 99, 100] may also be monitored. Abrupt changes in daily activity, feeding and watering behaviors, body condition, and health can be detected by different sensors (e.g., radio-frequency identification, accelerometers, transducers, and video footage). Device-recorded changes in behaviors or physical status may indicate management problems or could be caused by different disease states, or could indicate the cow has entered estrus.

Cows exhibit cyclical periods of sexual receptivity (estrus), during which behavioral changes that are particular to the hormonal profile are followed by ovulation, approximately 24 hours after the end of estrus. The use of sensors (particularly accelerometers) to identify sexual receptivity and later predict the time of ovulation is recommended, considering the widespread use of artificial insemination in dairy cows. The most convenient method of detecting estrus in these animals is through the use of pedometers, which measure the number of steps each animal takes and allow, by comparison, the overall assessment of locomotor activity [101, 102]. The use of pedometers to identify estrus represents an improvement in comparison to remote visual observations, with a detection success rate of over 90%, compared to 60–70% in the conventional observation system [103]. Other commercially available devices for estrus detection are pressure-sensitive detectors, gobblers that count inertia, thermometers that measure body or milk temperature [102], thermographs that highlight skin temperature [104], and devices that measure progesterone levels in milk (which are reduced before estrus) [102].

Animal behavior detection plays an important role in intelligent herd management. Some studies on this topic have focused on the classification of automatic, accurate, and multiple behavioral patterns in cattle. Monitoring herd behavior based on AI-assisted video recording and sensors attached to the animal’s body are two of the important approaches that have emerged with the development of modern technology.

The measurement of body temperature is a common method used to monitor the health status of animals. Veterinarians and farmers determine rectal temperature in animals showing obvious symptoms of disease or as part of the screening protocols. However, the method is time-consuming and potentially disruptive to animals. Therefore, other technologies have been developed for continuous recording of body temperature, e.g., rumen thermography boluses [105] and subcutaneously implanted transmitters [106]. These methods appeared to be promising, but they are all invasive and more suitable for research use. Infrared thermography cameras can be used to detect body temperature in specific body regions, offering the advantage of being fast, non-invasive, and avoids contact with the animal. They can be installed in the barn or at the milking parlors/maternity buildings to measure the body temperature of each animal. Additionally, recent studies have shown that infrared thermography can be a useful method for analyzing animal stress and welfare [107] and is promising for detecting mastitis and testicular disorders [108]. Investigations in calves indicated that infrared thermography can provide early detection of infection, and ocular temperature was most accurate [109].

PLF technologies can add value to the on-farm management process by improving data processing, decision-making, and the day-to-day implementation of management decisions in herds [110]. In addition, PLF technologies are applied in high-tech farms to carry out automatic animal welfare assessments [111] or by transposing visionary projects in a flawless manner to solve modern social crises (e.g., the dairy cow floating farm in the Netherlands [112]). On the other hand, as demonstrated in large-scale studies [113], investments in sensor systems may not necessarily lead to direct economic gains for farmers.

6. Environmental enrichment in dairy farming

Even though the term “environmental enrichment” (EE) lacks a clear definition, it can be defined as modifications that aim to raise the degree of social and physical stimulation offered by confinement settings [114]. Regarding dairy cows, the worldwide trend is to be reared in a zero-grazing system because this can lead to improved welfare by protecting them from natural predators, reducing exposure to extreme climatic conditions and ingestion of toxic plants, and limiting internal and external parasites [114]. In addition, this husbandry system is beneficial in terms of ensuring a balanced feed supply throughout the year. However, keeping females in barns can expose them to stressors such as environmental noise (banging of metal fences, milking parlor, slamming doors) or long periods of boredom (e.g., feeding time can be halved compared to cows kept on pasture) [115]. The multitude of these stressors can lead to frustration to the animal.

While environmental enrichment is crucial for preserving the health of lab animals, zoo animals, and some livestock, its use on cattle farms is relatively limited and has not kept up with the gradual transition to year-round indoor housing and the difficulties it raises for cows. Each of the five areas of environmental enrichment – social, occupational, physical, sensory, and nutritional – could provide a unique contribution to the animal’s welfare [114]. For instance, social enrichment can satisfy calves’ need for company and improve their ability to handle environmental stressors, while physical enrichment, such as giving a cow a private space in a maternity pen, can satisfy a cow’s need to hide during labor or during illness. In addition, sensory enrichment can be achieved by synchronizing milking with music or ambient noises, and nutritional stereotypies can be limited through feed management [115].

7. Conclusion

To summarize, the topic of dairy cow welfare is very complex and continues to evolve, concerning both the organizations in the field and the general public. The welfare assessment in herds implies the tracking of numerous animal-based, resource-based, and management-based indicators, requiring high degrees of both technical engineering skills and medical expertise.

The path to the future improvement of dairy cow welfare level is achieved by blending training and developing stockpersons’ abilities and awareness to face everyday farm welfare issues with the use of new solutions like electronic and information technologies that are reliable at the farm level.

Acknowledgments

We are grateful to our colleagues and students from the Faculty of Veterinary Medicine within the University of Agronomical Sciences and Veterinary Medicine of Bucharest. We owe them who we are now, as professionals.

Conflict of interest

The authors declare no conflict of interest.

Thanks

Deepest thanks to our mentors, families and faithful friends.