The effect of pasture-based dam-rearing on attention bias after disbudding in dairy calves

Introduction

Under typical Australian commercial conditions, dairy calves are removed from their dam within 24 h of birth and reared artificially in groups of similarly aged animals. The practice of separating dairy calves from their dams soon after birth aims to reduce cow–calf disease transmission, ensure that calves consume enough high-quality colostrum, gain control of calf health and avoid cow–calf bonding that leads to a future stressful separation (Flower and Weary 2003; Godden 2008; Beaver et al. 2019; Meagher et al. 2019). Keeping cows and calves together allows for more natural maternal behaviours, which can enhance calf growth, improve social development, and reduce stress in both cow and calf, ultimately contributing to better welfare outcomes and potentially improved long-term productivity (Meagher et al. 2019). Beyond the welfare benefits already highlighted in literature, increasing public interest in humane animal practices has driven interest in cow–calf contact systems, which may support not only calf welfare but also reinforce the maternal bond, benefiting both the cow and calf. However, few studies have explored how cow–calf contact (CCC) affects calf responses to routine procedures such as disbudding, leaving an important gap in understanding the potential role of CCC in reducing stress during such events.

Disbudding is painful as indicated through measures of behaviour (Faulkner and Weary 2000), physiology (Stafford and Mellor 2011) and affective states, with induction of negative emotional states and pessimism, even in sedated and locally anaesthetised calves (Neave et al. 2013; Daros et al. 2014). The administration of a local anaesthetic before the procedure delays the rise in cortisol following disbudding by approximately 2 h, but the stress hormone increases rapidly once the anaesthetic wears off (McMeekan et al. 1999). Behavioural evidence suggests calf aversion to hot-iron disbudding by using a place-conditioning paradigm, even with the use of a local anaesthetic (Ede et al. 2019), and wound sensitivity for up to 75 h following hot-iron cautery disbudding (Mintline et al. 2013). Painful dehorning procedures can also lead to negative emotional responses such as reduced play occurring for up to 44 h after the event (Huber et al. 2013; Mintline et al. 2013; Costa et al. 2019).

Social context can alter stress and pain perception in animals via social buffering. Social buffering refers to the reduction of an animal’s behavioural and physiological stress responses owing to the presence of affiliative social partners, such as a peer or dam (Kiyokawa and Hennessy 2018). This effect has been observed in several species, including rodents, canines, fish, primates and livestock (Hennessy et al. 2009). For example, pair-housed dairy calves resume eating more quickly after disbudding than do individually housed calves, suggesting faster recovery (Bučková et al. 2022). Similarly, mother-reared monkeys have shown increased social contact, reduced stereotypic behaviour and lower cortisol responses to a novel environment when a social companion is available, compared with nursery-reared monkeys (Winslow et al. 2003). The mother’s presence is particularly effective at reducing stress-induced cortisol concentrations in offspring. This has been reported in juvenile humans, monkeys, guinea pigs, rats, sheep, dogs and zebra finch (reviewed by Hennessy et al. (2009)). Cow–calf dairy systems may reduce the emotional impact of stressful or painful husbandry procedures such as disbudding on the calf, but this has not been studied.

Affective states, such as anxiety, can be assessed through an attention bias test (ABT). The ABT assesses the allocation of attention towards a potential threat (i.e. dog) and has been validated to assess anxious states in cattle (Lee et al. 2018). This provides insights into the emotional state of the subjects by measuring behaviours such as vigilance, the frequency and duration of gaze towards a stimulus (i.e. attention), and latency to eat. Increased vigilance and direction of attention towards a threat indicate heightened anxiety, whereas quicker resumption of eating and less attention towards a threat indicate lower anxiety levels (Lee et al. 2016). Animals experiencing threatening stimulus (i.e. predator threat) exhibit increased vigilance and altered attention patterns, validating the sensitivity of ABTs to detect these states (Monk et al. 2023).

The aim of this study was to explore the effect of pasture-based dam-rearing on the behavioural responses of dairy calves in an ABT after undergoing disbudding. It was hypothesised that dam rearing would reduce calves’ negative emotional response to disbudding during an ABT.

Materials and methods

The attention bias test

Six hours post-disbudding (a time point when the effects of the drugs would likely have worn off (Neave et al. 2013), meaning that the calves would be experiencing pain during the ABT), calves dehorned on that day were separated from their dam or group peers and transported in a mesh-walled box trailer (1.5 m in width × 3 m in length × 1.7 m in height) in groups of five to six calves at a time exclusively with others from their experimental group, to an indoor arena. The arena was closer to the commercial calf pen, but they were transported a similar distance as the dam-reared calves. Calves remained with their treatment peers in the arenas pre-test holding pen for 30 min to allow calves to settle. The methodology of the ABT for calves was adapted from the procedures described by Lee et al. (2018) and Monk et al. (2018), and is described below.

The test arena

The ABT was conducted in an indoor arena purposely constructed for this project and contained within an enclosed shed. The test arena had a 4 × 5 m area with solid 1.5 m high wall to visually and physically isolate the calf, albeit enabling audio contact with the remaining calves in the group who were housed nearby. Grain concentrate (~2 kg), with which calves were already familiar because of its availability in the dam-reared and commercially reared calf pens from the beginning of the rearing period, was placed in a trough in the middle of the arena. The left side of the arena had a mobile window (80 × 60 cm) that opened to show a sitting dog. Three cameras (DJI Osmo Action) around the arena recorded the calves’ behavioural responses, and a 1.25 × 1.33 m grid was painted on the ground of the arena to divide the arena area into zones (see Fig. 1).

Fig. 1.

Schematic diagram of the attention bias test (ABT) arena adapted from Monk et al. (2018). A 4× 5 m arena, visually isolated from the environment by a 1.3 m wall, entrance/exit gate, a mobile window to provide dog exposition and three action cameras distributed around the arena. White lines represent a 1.25 × 1.33 m grid painted on the ground (not to scale).

ABT procedures

Individual calves entered the arena and, once the arena door was closed, they were exposed to a quietly seated live dog (Border Collie) through the window for 10 s. The 10 s commenced once the calf had first sighted the dog. The window was then closed to visually remove the dog. The calf remained within the arena for a further 3 min, after which it was removed from the arena and returned to the holding pen with its peers. The test was terminated for any calf that showed high-severity fear behaviours (i.e. calf makes ≥4 attempts to escape the test arena or repeated and constant high-pitched calls). One calf from the dam-reared group attempted to escape the test arena more than four times, so was excluded from the analysis. On testing completion, the calves were collectively transported back to their treatment housing.

Behavioural data collection

Behavioural response data for the calves was collected by a single trained observer using Behavioural Observation Research Interactive Software (BORIS, ver. 7.12.2, see https://github.com/olivierfriard/BORIS/releases/download/v7.12.2/boris-7.12.2-win64-setup.exe; Friard and Gamba 2016). An ethogram describing the behaviours observed from video recordings is presented in Table 1. Specific calf behaviours related to the dog (e.g. attention directed towards the dog, and approaches or withdrawals) that measured calf fear-related behavioural responses to the threat, were recorded during the 10 s when the dog was visible. The 3-min period when the dog was absent was intended to measure anxiety, reflecting the calf’s perception of the threat once it was no longer visible. However, the duration of time the calf was visually focused on the closed window was recorded for the first minute following the removal of the unfamiliar dog, following procedures previously used on Angus steers (Lee et al. 2018). All other behaviours were collected during the 3-min period immediately after the dog window was closed. Intra-observer reliability was assessed using Cohen’s kappa (κ), resulting in a coefficient of 0.90.

Table 1.Ethogram of the behaviours observed in the attention bias test (ABT) performed on dairy calves.

Behaviour	Data recorded	Description
Behaviours for the 10 s that dog is visible (fear-related behaviours)
Attention towards dog	Duration	The time the calf stands visually oriented towards the dog during the 10 s of exposure to the moment the calf moves her head in a different direction
Calf approaches the dog	Frequency	The number of times that the calf takes one or more steps towards the dog while visually oriented to the dog
Withdrawal from the dog	Frequency	The number of steps the calf takes away while visually oriented to the dog
Behaviours for the 1 min following the window closing (anxiety-related behaviours)
Focus on dog window	Duration	The time the calf spends visually oriented towards the dog window once the window is closed
Behaviours for the total 3 min following the window closing (anxiety-related behaviours)
Vigilance	Duration	The time the calf stands with her head positioned above the withers and ears erected, visually oriented to random places within the arena
Locomotion	Frequency	The number of zones (grid painted) crossed by all four feet
Exploration	Frequency	The number of times the calf is sniffing or touching the arena ground, walls or trough
Latency to walk	Duration	The time for the calf to move at least a foot and place it a minimum of 20 cm away from the initial place
Latency to eat	Duration	The time to the first collection or chewing of the food from the trough
Standing	Duration	The time in which the calf stands with all four feet on the ground, with no movement reaction. The calf stand immobile, with no head movement
Escape attempt	Frequency	The number of times the calf makes attempts to escape the test arena (i.e. jumping walls)
Vocalisations	Frequency	The number of times the calf produces calls
Elimination	Frequency	The number of times the calf urinates or defecates during the test
Eating frequency	Frequency	The number of times the calf actively eats grain

Duration refers to time spent performing the behaviour and frequency to a count of the number of observed behavioural events. Adapted from Lee et al. (2018) and Monk et al. (2018).

Results

Dam-reared calves ate grain less frequently than did commercial calves in the 3 min after the window exposing the dog was closed (F_1,19 = 8.10, P = 0.01). There were no significant differences for the attention-related behaviours vigilance (F_1,19 = 0.101, P = 0.75) and attention towards dog (F_1,2 = 0.002, P = 0.97). No other significant differences were found between the groups (see Table 2 and Supplementary material Table S1).

Overall, 76.2% (n = 16) of the calves did not eat grain, whereas 23.8% (n = 5) of the calves did and four of these were commercial calves. The mean latency to eat for the commercial calves that ate grain was 77.3 ± 58.05 s and for the dam-reared calf was 76.3 s. According to the Cox proportional hazard model, there were no significant differences in the likelihood of eating the grain during ABT between the groups (hazard ratio: 0.205 [95% CI: 0.023–1.836], P = 0.17).

Discussion

The purpose of this study was to explore the effect of pasture-based dam-rearing on the behavioural responses of dairy calves in an ABT after undergoing disbudding. The key behavioural indicators associated with negative affect in an ABT are increased attention directed towards an area where a threat had been, increased vigilance and increased latency to consume feed after the threat is removed. We did not observe an effect of dam versus commercial rearing treatment on any of these measures in the calves. There was a rearing effect on the frequency of eating, with dam-reared calves eating less frequently. However, given the overall low proportion of calves eating and evidence that free-suckling calves consume almost no concentrate before weaning, even when compared with calves on high milk allowance (Fröberg et al. 2011; Ospina Rios et al. 2023), these results should be interpreted with caution. These results may suggest potential differences in the calves’ behavioural responses to ABT owing to their rearing treatment and painful disbudding, but this requires further investigation.

Except for eating grain, dam rearing did not affect the overall behavioural response of disbudded calves in the ABT, as assessed using the behaviours listed in the ethogram. This contrasts with findings in other species. For example, Winslow et al. (2003) found that 36-month-old mother-reared rhesus macaques were able to use a companion for social buffering in a novel environment test, but nursery-reared macaques were not. Similar results have been also reported in humans (Hostinar et al. 2015). It is possible that benefits of dam-rearing do not manifest in changes in calf behaviour during an ABT after disbudding, or the ABT may not have been sensitive enough to capture subtle differences in anxiety or pain perception between groups. However, there are differences in how the present research was designed, compared with previous studies, which warrant discussion. First, in the studies of Winslow et al. (2003) and Hostinar et al. (2015), testing occurred when subjects were out of the juvenile phase of the development. By contrast, the calves in this study were young at time of the testing. Conducting the ABT after a longer period of dam-rearing may yield different results. Second, previous studies demonstrated social buffering in animals that were in the presence of a companion, but the calves in this study were isolated during the ABT. Research on domestic hens (Edgar et al. 2015) and fish (Culbert et al. 2019) found evidence of social buffering when animals were faced with a stressor in the presence of their mother or social grouping, compared with in isolation. In guinea pigs, the presence of the dam during a novel environment test had a greater effect on reducing cortisol concentrations in the offspring than did the presence of an unfamiliar adult female (Graves and Hennessy 2000). We suggested that future studies observe calves in the presence of their dam or peers to fully elucidate the effect of social buffering on pain perception post-disbudding.

It is also possible that the design and conditions of the test arena used in this study affected calf behaviour. The behavioural responses of calves in the ABT presumably reflected their emotional states at the time of testing (6 h post-disbudding), when the calves were experiencing pain. The effects of sedation (Stafford et al. 2003), anaesthesia (Stafford and Mellor 2005) and pain relief (Heinrich et al. 2010) during disbudding are reported to gradually wear off over the next 6 h post-procedure. All calves were transported for the first time prior to the test, and this is known to be stressful. Indeed, cortisol concentrations peak within 30–60 min after loading calves for transport and decline over the next 2–6 h, and the highest cortisol response often occurs during shorter transport durations (reviewed by Eicher (2001)). In addition to pain and transportation, the dam-reared calves were experiencing separation from the dam and a fully enclosed indoor environment for the first time, complete with cement flooring. Thus, the results of this study may be reflecting the emotional state of calves to the additive effects of disbudding, separation from the dam, transportation, and a novel environment, rather than the effect of disbudding per se.

Willingness to eat is a validated indicator of anxiety in the ABT, but its validity requires all calves to consistently consume the food before testing. The trade-off between anxiety and willingness to feed has been seen in sheep (Lee et al. 2016), starlings (Brilot and Bateson 2012) and beef cattle (Lee et al. 2018). Whereas dam-reared calves were not consuming grain at 6 weeks, commercial calves consumed an average of 40 g/day, although 5 of 10 were not eating grain at all (Ospina Rios et al. 2023), and none had prior experience eating from a bucket. Notably, no differences were found in vigilance or attention towards the threat, key indicators in attentional-orienting assessments (Bethell et al. 2012; Brilot and Bateson 2012; Lee et al. 2016). Given that only a few animals ate the grain at all (4 of 10 commercial calves, 1 of 11 dam-reared calves), and the absence of vigilance differences, we recommend caution in interpreting grain consumption as an anxiety indicator in this study. To improve validity, the best approach would be to ensure that all animals have equal experience with the selected food before the ABT. One potential design could involve a ‘food familiarity phase’ where both groups are introduced to a highly palatable food item (e.g. hay, apples, carrots) outside their regular diet. This approach would ensure consistent exposure across groups and confirm that all calves are comfortable consuming the food before testing. A pilot study with non-experimental calves could further confirm intake levels under novel conditions. By prioritising equal experience with the food, these steps would enhance the reliability of feeding behaviour as an indicator of emotional state in future ABTs.

To the best of our knowledge, this is the first study to assess the behaviour of young calves in an ABT. Lee et al. (2018) found that a dog visible for only 10 s was threatening to cattle. Animals treated with an anxiogenic drug showed increased attention towards where the dog had been, increased vigilance and had a greater core temperature than did cattle treated with an anxiolytic or saline. Future studies incorporating physiological measures, such as stress-induced hyperthermia, could further validate the distinction between neophobia and threat perception in this context. The cattle in Lee et al. (2018) also spent 6.7 of 10 s focused on the dog while visible, which is similar to the 7.2 s the calves from both treatments spent in this study. Thus, our results suggest that it is likely that calves in this study perceived the dog as a potential threat, as evidenced by their reduced exploration, increased vigilance, and alignment with previous findings on fear responses.

Behavioural variability in pasture-based systems may be underestimated when using power analyses based on indoor data. Variability in observed behaviours highlights the influence of individual factors, such as temperament and prior experiences in determining behavioural responses to challenging situations (Bushby et al. 2018). This variability could suggests that power analyses based on indoor systems may fail to account for the broader behavioural diversity influenced by the external environment in pasture-based systems. Research from indoor dairy systems reports positive effects of cow–calf rearing systems on calf growth, play behaviour, social behaviour and stress resilience when using 10–14 calves per treatment (Roth et al. 2009; Chen et al. 2015; Wagner et al. 2015; Santo et al. 2020; Waiblinger et al. 2020). Chen et al. (2015) found significant differences in serum cortisol concentrations and behaviour during a novel object behaviour test between mother-reared and artificially reared calves (n = 5 per treatment). The higher (P < 0.05) duration of staying near to a novel object (s) on mother-reared calves when than on artificially reared calves (682.5 ± 258.5 and 192.8 ± 112.9 respectively) could be detected with four animals in each group at 95% power. Therefore, we proceeded with a sample size of 10 and 11 calves per treatment in this analysis of calf behaviour in an ABT. Recognising the increased variability in pasture-based environments is essential for refining future sample-size calculations and ensuring robust behavioural assessments in outdoor systems.

Study limitations

This is the first study to assess emotional states in pasture-based dairy systems. Several factors are likely to have influenced the testing outcomes and should be considered when interpreting the results. We aimed to assess the ABT response of disbudded calved reared either with the dam or conventionally. Latency of calves to eat in the ABT may have been affected by treatment differences in familiarity with the grain, as commercial calves consumed more grain pre-test than did dam-reared calves. Additionally, conventionally reared calves were also more accustomed to cement flooring and indoor housing than were dam-reared calves. These limiting factors are directly associated with the rearing treatments imposed (dam-reared outdoors versus commercially reared indoors). The study sample size was determined using a power analysis and data derived from other species and indoor-housing systems; however, our results suggest increased variability potentially owing to pastoral rearing systems. Larger sample sizes are needed to determine the true effects of dam-rearing on calf affect post-disbudding specifically, and potentially for pasture-based behaviour studies generally.

Conclusions

Under the conditions of this study, dam-rearing did not alter behavioural responses indicative of anxiety in an ABT. The effect of the dam-rearing on pain perception may have been confounded by separation from the dam, a novel environment and transportation stress. This study has highlighted complexities in measuring affective states of dam-reared calves in an ABT. Development of a method to test calf affective states without removing them from their rearing environment will help elucidate the emotional experiences of calves in a pasture-based dam-rearing system.

Supplementary material

Supplementary material is available online.