The management of feed and water is important, with adequate accessibility and quality of these resources being fundamental to animal welfare (Appleby and Waran, 1997). Our Pen-side protocol contained animal output measures that were considered appropriate for assessment from outside of the pen (body condition score (BCS), feeding behaviour score (FBS), drinking behaviour and faecal pat consistency) in addition to several resource- and management-measures, including feed contamination, feed bunk length, water trough number, water trough access, water trough contamination, and water trough fill.

Feedlot rations are developed in conjunction with nutritionists to meet requirements and reduce metabolic issues (Salvin et al., 2020), and there is a clear management focus on the delivery and monitoring of feed at a pen level in feedlots. Furthermore, considerable literature focuses on the nutritional management of cattle within the feedlot setting (e.g., Duff and Galyean, 2007; Nagaraja and Lechtenberg, 2007). However, fulfilment of the basic needs of an animal, in this case food, does not necessarily guarantee good welfare (Webster, 2005b). To address this, inclusion of animal output measures such as FBS that captures short-term feeding and BCS that captures medium- to long-term feeding were proposed in our protocol. FBS is a novel animal-based measure to inform on hunger status, social competition for food and response to thermal stress, and was first proposed for use in the live export industry (Dunston-Clarke et al., 2020). While feedlot pens are designed to give most if not all cattle access to feed simultaneously, the assessment of both resource-based and animal output measures of Good Feeding is important and may be useful in the identification of possible issues. For example, crowding or competition at feed troughs may indicate stocking density is too high, or that the feed bunks were void of feed for too long and action may be needed to manage welfare outcomes.

BCS is a commonly used measure of feeding that is easy to collect under field conditions and was included in four of the five reviewed beef cattle protocols. BCS offers a subjective assessment of body reserves (subcutaneous fat and muscular reserves) independent of frame-size, thus providing valuable information regarding nutrient intake relative to the animals’ requirements (overall nutrition status) (Roche et al., 2004; Kenyon et al., 2014). The monitoring of BCS at feedlot would be useful to identify cattle in poor condition with low body reserves (emaciated) or ‘poor doers’ (captured under ill-thrifty animals in Good Health), the implication of which is inadequate nutrition informing on feed and animal management. Equally, the monitoring of heavy (fat or obese) cattle may also be useful as these cattle not seen in other points within the beef cattle supply chain (e.g., calf-cow operations, live exports), are present at feedlots. Weight or excess fat cover in cattle is a risk factor for heat stress (Brown-Brandl et al., 2006a; Dikmen et al., 2012), hoof disease (Schöpke et al., 2013), slips and falls during unloading (Gregory, 2008), and lameness (Wells et al., 1993). Heavier cattle may also be harder to move within facilities or require additional care when handling to manage risks of injury to staff and animal, which have both welfare and profitability impacts. From a management perspective, these risks may be harder to mitigate, and as such the pushing of cattle to extreme weight may not be ideal. Recording BCS may inform on other measures that identify animals in a negative welfare state (e.g., animals unfit to load by industry standards), aiding animal management efforts (e.g., feed management).

The accessibility and quality of water, alongside the appropriate management of water during climatic extremes, is an important welfare consideration in feedlots (Animal Health Australia, 2016; Salvin et al., 2020). From a welfare and legal perspective, it is vital to ensure that all cattle have access to water (Animal Health Australia, 2016; Grandin, 2016), particularly during heat stress conditions, when cattle have higher water requirements (Arias and Mader, 2011). Thus, measures to indicate good water accessibility were included in our protocol (e.g., water trough number, water trough length, water trough fill, and time off water for transported cattle). The provision of clean water is essential for good welfare (Grandin, 2016), and is reported to influence water intake in feedlot cattle (Sparke et al., 2001; Schütz et al., 2019), thus water contamination and temperature are important to assess.

Animal-based measures for water availability and utilisation were not present in any of the welfare protocols reviewed. Previous protocols focused on input (resource- and management-based) measures to assess Good Feeding, reflecting the ease of collection under field conditions and reliability of such measures (Main et al., 2001; Rushen et al., 2011). However, animal output based measures are needed to provide a direct assessment of animal welfare (Main et al., 2001; Webster et al., 2004). Therefore, a drinking behaviour measure was created to capture pen-level crowding around water trough(s). During periods of heat stress, cattle alter their behaviour, increasing water intake (Arias and Mader, 2011), standing over water troughs (Sparke et al., 2001), and seeking shade created by troughs (Mitlöhner et al., 2001; Castaneda et al., 2004; Lees et al., 2020). As only a small number of cattle are able to access water trough(s) at any one time, monitoring cattle drinking behaviour, particularly when heat stress conditions are occurring or expected is paramount. Crowding may be an indicator of poor welfare as it indicates that there are cattle that are thirsty but are unable to access water. Such observations can inform management decisions, ensuring appropriate and prompt action is taken, such as providing additional water points. The monitoring of water accessibility and quality measures is important from both a welfare and management perspective.

Good Housing in our Pen-side, Yard, and Transport protocols includes the assessment of pen condition and environment, with measures selected to inform both cattle comfort and thermal challenge. Animal output measures of coat cleanliness, panting score, shivering, huddling, grouped, and agitation associated with flies were included, with input measures necessary to inform these output measures also captured (e.g., head in pen, stocking and loading density, surface moisture, mud depth, access to shade, WBGT, breed, class, and coat colour).

Thermal challenge, particularly heat stress, is considered a major welfare issue in feedlot cattle (Salvin et al., 2020). The thermal environment, including temperature, humidity, solar radiation, air flow, and precipitation all directly impact cattle comfort and their ability to effectively thermoregulate (Mader et al., 2006; Gaughan et al., 2008; Grandin, 2016). Heat stress can lead to decreased feed intake and efficiency, cattle discomfort, and death in extreme cases (Brown-Brandl et al., 2006b; Mader and Griffin, 2015; Tucker et al., 2015; Grandin, 2016). Panting score is an animal-based measure of heat stress widely used in cattle and is routinely assessed under feedlot conditions (Mader and Griffin, 2015; Lees et al., 2020). The assessment of cattle demeanour (proposed under Appropriate Behaviour) offers further direct assessment of the impact of thermal challenge on cattle and is both simple and time efficient. For example, observations of demeanour together with panting scores are reported to capture animal responses to heat under live export conditions (Willis et al., 2021b). Stockperson stock appraisal skills may benefit from routine assessment of animal demeanour, improving awareness and sensitivity to how animals are coping with their environment and the identification of issues such as thermal stress. With the incidence of heat stress in feedlot cattle influenced by several factors, including climate, water accessibility, shade, breed, coat colour, weight, diet, and current health status (Brown-Brandl et al., 2006a; Tucker et al., 2015; Grandin, 2016; Salvin et al., 2020), the collection of this input data is vital to inform output measures.

The reviewed international cattle welfare protocols currently do not include animal-based measures to quantify cold stress in cattle, therefore, the assessment of shivering and huddling at a pen level was developed and proposed in the Pen-side, Yard, and Transport protocols. This was deemed important as some feedlots are located in southern Australia and can experience cold winter conditions. There is a heavy focus by industry on heat stress; however, the impact of cold stress on cattle thermal comfort and welfare needs to be considered. Evidence from US feedlots shows that cold stress impacts production and behaviour (Gonyou et al., 1979), and if not appropriately managed can result in mortalities (Belasco et al., 2015). Cattle are reported to huddle under heat stress conditions (Gaughan et al., 2010; Lees et al., 2020), and both huddle and/or shiver under cold stress conditions (Gonyou et al., 1979; Graunke et al., 2011). Recording these behaviours is considered simple and would better inform stakeholders on the impact of cold stress on cattle comfort and welfare, and provide insight into intervention points and the effectiveness of management strategies which could include use of bedding, temporary wind breaks, and/or feed management (Mader, 2003; Mader and Griffin, 2015; Tucker et al., 2015).

Mud has been identified as one of the top three welfare problems in cattle feedlots, particularly in areas of high rainfall (Grandin, 2016). High mud levels have been reported to impact resting behaviour and cattle comfort (Muller et al., 1996; Fisher et al., 2003; Tucker et al., 2015; Chen et al., 2017), and is associated with increased risk of lameness (Stokka et al., 2001; Marti et al., 2016) and heat stress through contributions to humidity in hot conditions (Petrov, 2007; Tucker et al., 2015; Salvin et al., 2020). Lying behaviour and the number of steps taken have been shown to be reduced in beef cattle confined to a surface with a high mud depth within a feedlot (Dickson et al., 2022). Measures of coat cleanliness are routinely employed as an indirect evaluation of environment (sanitation) and comfort in cattle and was included in all but one of the protocols reviewed. Overall, coat cleanliness can inform on pen surface, indicating muddy conditions or whether a dry area for lying is available (Hauge et al., 2012; Grandin, 2016; Chen et al., 2017). The maintenance of facilities and targeted pen surface management are vital to control mud and cattle cleanliness (Mader, 2003; Mader and Griffin, 2015; Grandin, 2016). Therefore, measures which assess manure pad integrity (e.g., surface moisture, mud depth or animal mud depth measures) should be recorded, and relevant management practices (e.g., stocking density) and environmental measures (e.g., precipitation) should be monitored. Importantly, even when good pen surface management practices are employed, pens can still get muddy (Grandin, 2016). Hence, monitoring cattle lying behaviour (captured as posture in Appropriate Behaviour) and cleanliness in the Pen-side and Yard assessments would be informative.

Animal health is central to animal welfare and forms a vital component of the proposed protocol. Routine daily pen walks/rides by feedlot staff in Australian feedlots ensure that cattle health is monitored, and appropriate treatment action is taken where required. Here, important health issues relevant to the feedlot context are prioritised for further standardised formal recording at a pen level in the Pen-side protocol (lameness, non-ambulatory animals, nasal discharge, coughing, ocular discharge, ill-thrift), and within the Yard (lameness) and Transport protocols (animals unfit for transport, animals unfit for transport on arrival, animals dead on arrival, tender-footed animals). Additional measures of Good Health available for assessment at a feedlot level from readily accessible records were also considered (routine husbandry practices, births/abortions, treatment pulls, case fatality rate, mortality/euthanasia, abattoir report data).

Mortality represents the ultimate endpoint of compromised welfare and potentially poor welfare management (Colditz et al., 2014), and is an informative indicator of welfare under commercial conditions. Four of the protocols evaluated included a measure for cattle mortality, however, they considered only mortality at its endpoint, reporting the percentage of animals that died and/or were euthanised, regardless of cause (Welfare Quality®, 2009). Importantly, identifying the cause of death is considered a vital element of a successful welfare assessment protocol (Waiblinger et al., 2001). This enables action to safeguard welfare rather than simply reporting the status quo. Hence, the collection of more detailed mortality measures, including cause and case fatality rate following treatment, were proposed to be captured at a feedlot level. In addition to mortality, other measures of importance to be collected at a feedlot level included routine husbandry practices, births and abortions, treatment pulls, and abattoir report data.

Lameness, injury, and respiratory illnesses (e.g., Bovine Respiratory Disease (BRD)) are common and important issues in feedlot cattle (Salvin et al., 2020), and could reflect the suitability of the pen environment, feed and animal management, and stockpersonship. For example, pen condition and pen surface are considered as the two most common factors contributing to lameness caused by infection in feedlot cattle, followed by weather patterns and handling (Terrell et al., 2013). BRD is a major cause of morbidity and mortality in feedlot cattle (Perkins, 2013; Vogel et al., 2015), with age, stress at entry to feedlot environment (e.g., weaning, transport), immune status, nutritional status, climate and management (e.g., stocking density) important risk factors in feedlot cattle (Snowder et al., 2006; Duff and Galyean, 2007). Cattle requiring treatment for illness or injury likely experience negative welfare and to capture this, signs of these common health issues in cattle are included in the proposed protocol (lameness, nasal discharge, coughing). The collection and reporting of such information in a more targeted, routine, and standardised manner will inform on Good Health and would be beneficial. For example, the routine recording of nasal discharge and coughing at a pen level may have the added benefit of improving the accuracy of BRD diagnosis by pen walkers/riders, which is considered poor (~ 60%) (White and Renter, 2009). Overall, the Pen-side protocol captures several major health concerns applicable to feedlots, providing information valuable from both welfare assessment and management standpoints.

Cattle unfit to load at exit, unfit for transport at arrival, and dead-on-arrival in the Transport protocol, and lameness in the Yard protocol, represent additional measures of Good Health proposed here. The presence of animals that are unfit to load represents an assessment of moderate to severe health issues at the point of exit from feedlot, with the welfare of cattle unfit to load likely to be compromised (e.g., non-weight bearing lame, severe injury, or severe distress) (Animal Health Australia, 2012). The MLA ‘fit to load guide’ (MLA, 2019) is available to producers to ensure the relevant legislative standards (Animal Health Australia, 2012) are met and best practice animal welfare is achieved. Cattle are routinely passed through the yards for drafting into lots for transport within weeks to days of exit from the feedlot prior to loading on trucks. This represents a practical time to collect valuable information captured in the Yard and Transport protocols proposed, including the formal recording of unfit, tender-footed, and lame animals.

It has been suggested that many on-farm welfare issues (e.g., cleanliness, ease of handling) can be assessed at the abattoir (Grandin, 2017; Knock and Carroll, 2019). While it is generally considered that ‘lead’ indicators, those that facilitate corrective or preventative actions to be taken, are most informative from a welfare perspective (Barnett and Hemsworth, 2009), the benefit of ‘lag’ indicators such as those that could be assessed at the abattoir should not be overlooked. The collection of this data could capture issues that have been missed including injuries and bruises (Grandin, 2017; Knock and Carroll, 2019) or issues that may not be apparent externally such as liver abscesses (Galyean and Rivera, 2003) or pneumonic (lung) lesions (Fernández et al., 2020). This information could then be used to inform targeted action at the feedlot. For example, consistent reports of liver abscesses could indicate ruminal acidosis, suggesting that current feed management needs to be reviewed (Nagaraja and Chengappa, 1998; Galyean and Rivera, 2003). In this way, data from received in abattoir feedback can be applied to refine management, thus advancing animal welfare at a feedlot level.

Behaviour has an important role in the diagnosis and early detection of health issues at feedlots (e.g., diagnosis of BRD (White and Renter, 2009)), and behavioural measures are recognised as meaningful indicators of welfare across numerous production industries and systems (Mench and Mason, 1997; Mench, 1998; Webster, 2005a). The Pen-side protocol prioritises the behavioural observation of cattle in their home pen environment, capturing both positive and negative behaviours (through posture and activity), to provide a direct assessment of welfare outcomes. The assessment of cattle demeanour in the Pen-side protocol also offers a valuable assessment of both positive and negative welfare and mental state within a single measure. The assessment of animal handling and Human-animal relationship (HAR) in all three protocols is considered an advantage of this protocol, with the collection of several output (e.g., Pen-side; approach test, reactivity index, Yard and Transport; cattle slapped/hit or tails twisted, handling aid/electric prodder use, mis-caught, slips, falls, choking, animal flow through facility and to/from home pen; see Table 3 and Table 4 for details), and relevant input measures (e.g., staff generated noise, facility generated noise and/or use of dogs).

Both antagonistic social behaviours (e.g., displacement, aggression) and abnormal behaviours (e.g., buller syndrome, tongue rolling) can be observed in feedlot cattle (Blackshaw et al., 1997; Mitlohner et al., 2002; Gonzalez et al., 2008; Val-Laillet et al., 2009). While it is important to recognise that many factors influence behaviour (e.g., DOF, health status, climate, stocking density), the observation of negative behaviours can relate to competition and social stress, indicating that the social environment may be unstable (e.g., inadequate resource availability (Gonzalez et al., 2008; Val-Laillet et al., 2009)), or that animals lack adequate stimulation (Park et al., 2019a). Conversely, the absence or low incidence of these behaviours could be considered as an indication of neutral or positive welfare states. For these reasons, measures to capture negative social and abnormal behaviour are presented in the Pen-side protocol (see Table 2 for details).

The identification of measures of positive welfare is considered extremely valuable in on-farm monitoring protocols (Farm Animal Welfare Council, 2009; Webster, 2011; Edgar et al., 2013). The collection of information indicating that animals are experiencing a positive welfare state is paramount to the assessment of quality of life (see Mattiello et al., 2019). The assessment of positive welfare here involved measuring maintenance behaviours (e.g., self-grooming, resting, lying, ruminating), positive social behaviours (e.g., allogrooming, social play), and positive interactions with the environment (e.g., engaged; object play, locomotor play, exploration) listed in the ethogram and posture measures. Allogrooming, play (social and non-social), self-grooming and behavioural synchrony are indicators of positive welfare in cattle (Napolitano et al., 2009). For example, cattle naturally display synchrony in their feeding, resting/lying and ruminating behaviours (Rook and Huckle, 1995; Stoye et al., 2012; Tuomisto et al., 2019). High levels of synchrony in lying and ruminating behaviours may indicate comfort, social stability and/or that space allowance and stocking density are appropriate to avoid competition (see Napolitano et al., 2009; Asher and Collins, 2012). However, as indicators of positive welfare, some of these behaviours have limitations. For example, cattle may also demonstrate grooming when parasitised (ectoparasites including flies, ticks, lice) or when dirty (Napolitano et al., 2009), therefore high frequencies of these behaviours and consideration of these outputs with other measures may be required. Further, the use of pen infrastructure (e.g., fence posts, water troughs and/or enrichment) for grooming may damage the pen which could lead to injury or escape, a concern for welfare and management. These behaviours may also occur infrequently (e.g., allogrooming; (Napolitano et al., 2009) so careful consideration surrounding the most appropriate time for recording is required. For example, there is some evidence that allogrooming in cattle occurs in higher frequency during periods of feeding and overnight (Val-Laillet et al., 2009). For these reasons, measures to capture positive behaviour and mental state would be beneficial at feedlots; however, care should be given when interpreting these behaviours as indicators of positive welfare.

In addition to assessing quantitative measures of positive behaviour, the assessment of cattle demeanour is considered a particular advantage in the Pen-side protocol. Assessments of animal demeanour or body language offer a qualitative, whole animal assessment that is useful for the interpretation of welfare measures (Wemelsfelder et al., 2000; Wemelsfelder and Lawrence, 2001; Wemelsfelder, 2007), and are quick to capture compared to other behavioural measures (Knierim and Winckler, 2009). This is important under commercial conditions, with collection burden a major limitation to the use of behavioural measures (Barrell, 2019). The assessment of demeanour using Qualitative Behavioural Assessment (QBA) principals is considered valuable under commercial conditions, particularly to assess positive state (Boissy et al., 2007; Rutherford et al., 2012; Murphy et al., 2014; Fleming et al., 2016). Two of the reviewed protocols incorporate QBA which involves a panel of observers scoring the behavioural expression of cattle using a predetermined list of descriptive terms ( Table 7 ). A modified-QBA, or ‘demeanour’ was proposed in the Pen-side protocol which includes one stockperson observing cattle real-time scoring four positive (curious, content, lively, settled), four negative (agitated, dull, nervous, uncomfortable) and two neutral terms (active, alert) ( Table 7 ). Importantly, producers in the Australian red meat industry believe that assessing animal welfare is an innate ability they possess (Buddle et al., 2021). Scoring animal demeanour could be considered fundamentally similar to this, capturing what a good stockperson does when they survey their livestock in a formal, numerical manner (Fleming et al., 2016). The application of including demeanour to feedlots is novel; however, the full QBA approach has been validated in cattle under numerous contexts (see; Rousing and Wemelsfelder, 2006; Brscic et al., 2010; Stockman et al., 2013; de Boyer des Roches et al., 2018; Vindevoghel et al., 2019; Rizzuto et al., 2020), with the modified-QBA (demeanour) approach recently reported useful and valid for welfare assessment within the Australian live export industry (Willis et al., 2021a; Willis et al., 2021b). At feedlots, demeanour may provide evidence of positive affective state, and an early indicator of issues, such as thermal stress, ill-health, and unstable social housing, allowing earlier mitigation than otherwise would be possible. It would also aid in the interpretation of other measures, informing on how cattle are coping under certain conditions, allowing users to more easily interpret animal welfare states. In these ways, the assessment of cattle demeanour pen-side would be valuable; however, piloting under feedlot conditions is required.

Our protocols contain many measures of animal handling and the human-animal relationship (HAR). The quality of the HAR is widely considered an important factor that impacts animal welfare (Hemsworth, 2003; Waiblinger et al., 2006; Hemsworth and Coleman, 2011). In feedlots, good stockpersonship is an essential component of good cattle welfare (Grandin, 2016), and handling can influence cattle’s fear of humans (Breuer et al., 2003; Petherick et al., 2009a; Petherick et al., 2009b). Recording cattle responses to handling (e.g., slips, falls, crush exit) and human resource use (e.g., use of electric prodders and handling aids, noise, mis-caught animals) informs on stockpersonship, animal stress (Grandin, 1993; Fell et al., 1999; King et al., 2006; Anderson and Miller, 2019), the design of the yards, and can be used to demonstrate improvements in handling practices over time (Grandin, 2016; Grandin, 2018). Feedlot cattle are routinely drafted through the yards and crush for treatments, performance weighing and/or immediately prior to transport, which represent important opportunities to easily and unobtrusively assess HAR. For example, poor flow through yards results when animals refuse to move forward or attempt to back up or turn around; thus, an assessment of animal flow can indicate that there are problems with facilities and/or handling (Grandin, 2018).

An assessment of HAR is also included in the Pen-side protocol. Observations concerning the reactions of cattle to an approaching human are commonly used to assess HAR and inform on how animals perceive humans, and whether this changes over time (Waiblinger et al., 2006). The incorporation of the ‘Reactivity Index’ measure used in the Live Export Protocol (Dunston-Clarke et al., 2020) was considered to be most suitable to pen-side collection at feedlots. The capture of this information pen-side ensured that HAR is considered at all applicable points within the feedlot system and the lack of this could be considered a limitation in previous protocols. Importantly, cattle responses to humans are influenced by social context, the surrounding environment, the novelty of human exposure and type (Grignard et al., 2000) and loud machine and human noise (Weeks, 2008). Thus, these measures should be considered when observations of cattle are made Pen-side and also during handling (e.g., Yard and Transport protocols). Any measure based on the observation of cattle both pen-side and in the yards must be feasible to capture in a fast-paced environment and integrate within the feedlot system efficiently. The collection of those measures proposed is considered to achieve this; however, piloting will ultimately determine the practicality of these measures.

The proposed protocol includes 41 additional measures identified as suitable for inclusion ( Table 6 ), with 28 of these incorporated at the request of stakeholders and during manuscript review for a more comprehensive assessment of transport (loading, unloading) and animal handling than initially proposed and the remaining 13 were added to capture issues relevant to feedlots. For example, nuisance flies and their impact on cattle was identified as a specific seasonal problem due to the abundance of food and ideal breeding conditions (Urech et al., 2004). Flies can disrupt behaviour and reduce animal welfare (Machtinger et al., 2021), thus the measure was incorporated in the Pen-side protocol. Likewise, it was deemed necessary to incorporate a measure of ‘ill-thrifty’ animals to capture ‘poor doers’ that may have a poor demeanour (e.g., dull), have hollow sides or suboptimal body condition and/or poor coat condition. This addition is not anticipated to have a marked impact on assessment burden as it can be captured during pen walks/rides. Other measures were included as alternative methods to capture relevant information, either offering collection of information in a simpler manner than initially proposed (e.g., faecal pat consistency vs. diarrhoea), or one that was more familiar to feedlot staff, thus considered more easily understood and adoptable (e.g., THI vs. HLI). Regarding measures of coat cleanliness and pen manure pad integrity, multiple methods of measurement were determined necessary to identify the most appropriate for use in a feedlot context. Pilot testing will determine the most appropriate measures to capture this information.

A total of 51 measures were excluded during the advisory process. Engagement with industry stakeholders and animal welfare scientists ensured that issues relevant to all stakeholders were addressed, ensuring measure feasibility and practicality were considered (Sørensen and Fraser, 2010). Some tools to assess animal welfare developed in a research capacity may be too complex for practical use (Grandin, 2018), thus this step is considered vital in developing a successful protocol. Overall, the main reasons for exclusion were related to feasibility and practicality. For example, of the extensive list of potential health issues proposed under Good Health (n = 27; Table 1 ), 22 (81.4%) were excluded from observations in the Pen-side protocol on the basis of stakeholder feedback but retained at a feedlot level collected from feedlot records (see Table 5 ). In light of the difficulty in identifying many of these health issues accurately when pen-side (e.g., acidosis) and the time to collect data, it was considered more appropriate to prioritise the formal collection of indicators of BRD (e.g., nasal discharge, coughing) and injury (e.g., lameness, non-ambulatory). This approach was deemed more comprehensive than collecting disease incidence data from only a sample of pens, with the added benefit of minimising assessment burden during pen-side assessments. Likewise, although measures of water temperature are included in four of the reviewed protocols, they were excluded here due to stakeholder concern that the time needed to manually assess them was prohibitive. Perhaps with advances in automated water monitoring systems in the future these measures may be feasible. Deliberation by the advisory board also identified measures not applicable to the feedlot environment (e.g., ammonia), or as is the case for those measures under Good Feeding relating to rations (ration number, ration type, ration MMEF), not considered pertinent to include in a feedlot welfare assessment. The reason behind these exclusions was the fact that feedlots are considered to inherently address nutritional requirements at an appropriate level because diets are specially formulated with nutritionists and carefully managed. In addition, several measures were considered by stakeholders to capture duplicative information and discussion determined which of these was most appropriate for inclusion. For example, for the Yard protocol, crush agitation and crush exit speed measures were argued to provide the same evidence to inform HAR as the crush exit measure, the latter of which was identified as the easiest to capture by feedlots, thus retained under Appropriate Behaviour. Likewise, the collection of the water trough number, water trough length and water trough fill measures were considered sufficient to inform water accessibility resulting in the exclusion of water trough circumference, position of water trough/s and water functionality measures from Good Feeding.

Consultation with industry stakeholders indicated that 46 (46.5%) of the revised list of proposed measures are presently collected at feedlots to some extent. This means that the data is either readily accessible in the proposed format (e.g., days on feed, breed, mortalities, climatic data) or the data is presently collected in a different form (e.g., panting scores, slick bunks, water trough fill, health data). For those in the latter category, careful consideration during the advisory process was given as to how measures could be obtained without feedlot staff being required to collect duplicative information which would strain resources or be viewed as intrusive. Pilot testing will determine whether those measures presently collected can be easily transferred to the collection method proposed. Aligning data collection methods with those already collected at feedlot means that the protocol is considerate of resources (staff time and labour), and it is likely that the monitoring will not impede on normal staff responsibilities.

The proposed protocol appears practical and feasible within the feedlot context. The next step is to pilot test the protocol on representative feedlot premises across Australia to further refine the measures and their collection methods. The final selection of measures to be included will ultimately be a trade-off between comprehensiveness and validity, and practicality under commercial feedlot conditions. The piloting, which is currently underway, will further develop the protocol, answering many questions vital to ultimate adoption and success, including: