Benefits of tactile stimulation and environmental enrichment for the welfare of crossbred dairy calves

ABSTRACT The aim of this study was to evaluate behavioural and physiological variables of 18 Gir x Holstein calves housed in groups submitted to two treatments: presence (EE) or absence (NE) of environmental enrichment. Behavioural variables (walking, feeding, leisure and play behaviour) were analyzed continuously in the first and second hours after brushing using generalized mixed models. Physiological parameters (cortisol, lactate and oxytocin) measured before and after brushing were analyzed using linear models with repeated measures over time. Models included treatment and day fixed effects and treatment x day interaction. The brushing duration was shorter (P = 0.01) for the EE treatment (1.93 ± 0.08 min) compared to the NE treatment (2.47 ± 0.08 min). The animals in the NE treatment allowed longer tactile stimulation time than the EE animals throughout the management days. There was no treatment effect on serum lactate levels before toothbrushing (P > 0.05). After brushing, lactate concentration was lower for the EE treatment (P < 0.01). There was no treatment effect on cortisol before or after brushing. Mean oxytocin levels were higher for EE treatment before and after toothbrushing (P < 0.01). Brushing positively influences the well-being of group-housed calves raised in an enriched environment. Environmental enrichment is a low-cost tool with important positive effects.


Introduction
Heifer rearing is one of the most important stages of the milk production system since heifers will replace old and unproductive animals which are culled.However, this stage represents one of the greatest bottlenecks in production since mortality is high during this phase and the production costs until first calving impact the economic viability of milk production systems (Le Le Cozler et al. 2008).Machado Neto et al. (2004) reported a mortality rate of dairy calves of around 14.9% in the state of Minas Gerais, mainly due to failures in nutritional and health management.
The welfare of dairy calves is sometimes neglected by breeders due to effort to decrease the costs of calf rearing (Vasseur et al. 2010).During the rearing period, the adoption of positive handling practices (e.g.gentle contact, brushing and stroking, human presence during feeding) contribute to reduce animal's reactivity to handling, which in turn results in improved welfare and productivity (Lürzel et al. 2015;Hemsworth et al. 2000;Ujita et al. 2020a).Calf rearing is particularly important for tropical milk production systems that use more temperamental cattle such as the zebu (Bos indicus) and crossbreed F1 Gyr x Holstein, named Girolando breed (Silva et al. 2017).
According to Beaver et al. (2018), the separation of calves from their mothers immediately after birth is stressful and the environment where the animals are raised is important for their welfare, especially in the case of feedlot animals.Tactil stimulation is an approach to improve or reduce the negative effects of confinement (Schmied et al. 2008a;Silva-Antunes and Costa 2021) as it triggers the oxytocin release, the hormone related to the formation of social bonds, and reduces the stress suffered by the animal (Carter 1998;Ferguson et al. 2001;Bates et al. 2004).There is a relationship of oxytocin with positive social behaviour, since the higher the oxytocin concentration (modulated by general management practices, such as natural suckling, maternal care, and grazing; enrichment, such as manual brushing; and intranasal administration) the greater the degree of induction of positive social behaviour, which in turn increases oxytocin concentration (Chen and Sato 2017).
Environmental enrichment is a strategy used to reduce the stress caused by confinement, reduce frustration, and enable animals to perform natural behaviours (Mandel et al. 2016).Rearing environments in restricted, confined spaces and still without stimuli cause boredom and restrict animals from satisfying their behavioural needs, generating frustration that, in the long term, results in the appearance of abnormal behaviours, degrades the physical and mental health of the animal, reflecting in economic losses (Mason and Burn 2011;Ninomiya 2014).
The present study tested a rearing system for dairy heifer calves consisting of the inclusion of environmental enrichment in the facilities and the use of tactile stimulation to improve the welfare of animals as well as their living condition.For this purpose, the behavioural changes of the animals were evaluated based on the hypothesis that both environmental enrichment and tactile stimulation provide benefits for heifer welfare and/or development.
The aim of this study was to assess behavioural and physiological variables related to stress and welfare of calves housed in an environment enriched or not with toys.

Material and methods
The Animal Use Ethics Committee of the Instituto de Zootecnia approved the experiment under protocol CEUA/IZ 266/18 and it was carried out at the Dairy Cattle Research Center, Instituto de Zootecnia, Ribeirão Preto, São Paulo state, Brazil (21o42' S, 47o24' W, altitude of 535 m), between November 2018 to January 2019, during summer season.According to the Köppen classification, the climate of the region is tropical, characterized by warm and wet summers and dry winters, the average of temperature in the experimental period was 23.7°C.
The 18 F1 Dairy Gyr x Holstein heifer calves were allocated to two treatments: presence (EE) and absence (NE) of environmental enrichment.The heifers were obtained by embryo transfer using Nellore beef breed females as recipients.The animals remained on pasture with their mothers from birth to the day preceding transport to the farm of the experiment, when they were on average 30 days old.In this age the mean body weight of the calves was 51.94 ± 1.15 kg.Upon arrival, the animals were allocated to two barns (presence (EE) or absence (NE) of environmental enrichment, randomly assigned) and were allowed to adapt to the environment and to the automatic feeder for 7 days.The experimental period comprised 31 days.The barns were separated and isolated so that the animals of each treatment could not see the others.The total area of the barns contained 700 m 2 (Figure 1).
The heifers received daily 6 litres of commercial milk replacer composed of 20% crude protein, 44% lactose and 16% fat, supplied in an automatic feeder (DeLaval CF1000S) and diluted to contain 14% of dry matter.The computer program of the feeder provided daily individual data of suckling speed and number of visits of the calves when there was food available and when no food was available, as well as the total number of visits.Roughage (hay), concentrate (pelleted ration) in the trough, and water were available ad libitum for the two groups throughout the experimental period.The composition of hay and concentrate were respectively: crude protein 12 and 24%; neutral detergent fibre 75 and 28% and total digestible nutrients 59 and 75%.
All calves were submitted to tactile stimulation, and for the tactile stimulation two handlers, both female, brown hair, same height and using overall brown, applied the brushing once a day, alternating between the two treatments (EE and NE).On the first two day, the animals were restrained outside the feeder and brushed for 4 min.From the third day, brushing was performed for the maximum of 4 min once the animal had entered for suckling.If the animal refused the treatment, brushing was stopped and the time the animal allowed to be brushed was recorded.
Days with adverse situations were excluded, (i.e.days of training the calves in the automatic feeder and or a few days with equipment damage, or lack of electricity in the sheds) and that could influence the results.For the brushing, the handler approached the animal and started at the shoulder blade, loin, and hindquarter using a medium hard brush (slow brushing movements from top to bottom and from front to back in the region of the neck and dorsum).
Environmental enrichment (treatment EE) was provided using playful objects (ball, tyre, hanging chain, and scratcher) that were made available continuously to all animals of the respective treatment.The same playful objects were fixed and used throughout the experiment, except for the balls that remained loose in the environment.Two rubber balls measuring 65 cm in diameter left loose in the environment were used, as well as two tyres with a rope and two hanging chains swinging at 60 cm from the ground.Two scratchers made on the farm were mounted inside the pens.One scratcher was made of a eucalyptus post (56 cm in diameter and height of 1 m) wrapped in a thick sisal rope.The other scratcher was made of a T-shaped fence post (56 cm in diameter and height of 1 m) wrapped with brooms measuring 45 × 10 cm with polyester bristles of medium hardness (Figure 1(A)).In the control treatment (NE), the heifer calves were also group housed similar to the EE treatment but without playful objects (Figure 1(B)).
Behaviour at first and second hour after tactile stimulation, blood cortisol, lactate and oxytocin levels before and after tactile stimulation, at the beginning of the experiment (day 1) and the end (day 31) were recorded.For behavioural assessment (first and second hours after tactile stimulation), video cameras were installed in all areas to record the activities.To facilitate identification of the calves during filming, white cotton shirts with numbers painted in black were fabricated.A 3-cm thick elastic band was attached to the hem and the middle of the shirt to better adapt it to the animal's anatomy and to prevent the shirt from interfering with the animals' movements.The shirts were always placed on the animals the day before filming.The activities recorded, for each animal individually, were: lying down, walking, eating, idling, cross-sucking, and play behaviour (composed of exploration of the environment, inter-relationship between animals, and playing using toys).The films were stored and subsequently analyzed.An imaging laboratory with four computers was set up with the aim of simultaneously individualizing the different environments, with each calf being read individually by two previously trained people.For statistical analysis, the dataset comprised 24 days of tactile stimulation data from 18 heifers divided into two treatments (EE and NE) considering the time spent continuously (in minutes) in each activity.
Blood samples were collected from the jugular vein into heparinized vacuum Vacutainer tubes of Vacuette® and were kept refrigerated until the end of the procedure.In the laboratory, the tubes were centrifuged at 1000 ×g for 5 min for serum separation.The supernatant (serum) was transferred to a 1.5-mL Eppendorf tube and stored at −20°C for subsequent analysis.
Plasma cortisol and oxytocin levels were measured by enzyme immunoassays (EIA) using commercial kits Monobid and Mybiosource, respectively, from Diagnostic Systems Laboratories (Enzo, USA).Lactate was measured by an enzymatic method using Laborlab kits.The assays were read in an ELISA microplate reader at the Laboratory of Animal Physiology, FZEA/USP.

Statistical analysis
The variables were analyzed using the SAS MIXED or GLIMMIX procedure (SAS Institute, Inc., Cary, NC, USA).When the response variables were not normally distributed, such as behaviour traits, generalized linear mixed models were employed in analysis.The effects of treatment (EE or NE) and day of observation were included as a fixed effect, and data were evaluated as repeated measures over time (represented by day of observation).The interaction treatment × day was also included in the models.The covariance matrices were tested for each trait and the structure showing the best fit was used (Crowder and Hand 1990).Each animal was considered an experimental unit.When significant differences were found (P ≤ 0.05), estimated means were compared using the Tukey test at a level of significance of 5%.Table 1 describes in detail the procedures, models and covariance structures used in the analysis.
For serum cortisol, oxytocin and lactate levels on days 1 and 31, which were measured before and after tactile stimulation, a paired t-test was also applied using the TTEST procedure (SAS, Institute, Inc., Cary, NC, USA) to identify differences in the means of these parameters within the same treatment (EE or NE).The estimated means were compared by the Tukey test, with P < 0.05 being considered significant and P < 0.10 indicating a tendency.

Results
In general, the heifer calves did not voluntarily allow to be brushed for 4 min (Figure 2); thus, except for the first 2 days, brushing was performed only for as long as they allowed it.The duration of tactile stimulation varied over the experimental period (Figure 2(A)).For the EE treatment, only 16.7% of the observations lasted 3-4 min; this percentage was 34.7% for the NE treatment.The duration of tactile stimulation ranged from 1 to 3 min in more than 50% of the observations.Heifer calves receiving EE allowed a shorter time of brushing (1.93 ± 0.08 min) than those kept under the NE conditions (2.47 ± 0.08 min) (p ≤ 0.01).There was a significant effect of day (p ≤ 0.05) and of the interaction treatment x day (p ≤ 0.001).The time allowed for tactile stimulation increased over the days of handling for NE animals (p ≤ 0.05) compared to animals of the EE treatment (Figure 2(B)).
With respect to the behaviours of heifer calves, the activities were more uniformly distributed in the first hour after tactile stimulation, although the animals spent most of the time lying down.In the second hour, the animals also spent more time lying down compared to the other activities (Table 2).The estimated means of the behavioural characteristics in the first hour after tactile stimulation indicate that EE animals spent more time lying down and playing (P < 0.0001) than NE animals (Table 2).Heifer calves of the NE treatment spent more time idling (P < 0.0001), walking (P < 0.0001), and eating (P = 0.0108).The time spent cross-sucking in the first hour after tactile stimulation was 3.97 ± 0.38 and 4.10 ± 0.41 min for the EE and NE treatments, respectively, with no differences between means (P = 0.845).For all activities, there was a significant effect of observation day (P < 0.05) but not of the interaction treatment × day.
In the second hour after tactile stimulation, NE animals spent more time idling (P = 0.0107), as well as lying down (P < 0.0001); thus, there was an inversion of the latter behaviour from the first to the second hour after tactile stimulation (Table 2).Treatment exerted no effect on the time spent walking, eating or playing (P > 0.05).Cross-sucking was not observed in the second hour after tactile stimulation.
There was no difference in mean suckling speed (Figure 3  (A)) between treatments (P = 0.387), but observation day (P = 0.004) and the interaction day × treatment (P < 0.0001) exerted significant effects.During the final third of the experiment, animals of the EE treatment tended to feed faster (1,033.43± 24.98 ml/min) than NE animals (1,000.97± 26.50 ml/min).This result suggests that, as animals in the EE treatment interacted more, they used the feeder only to satisfy their hunger and not as an attractive.
There was no effect of treatment (P > 0.05) on the number of visits to the feeder when food was available (Figure 3(B)), indicating that animals of both treatments equally visited the feeder when food was available.The mean number of visits/ day was 6.48 ± 0.59 for treatment EE and 7.59 ± 0.63 for treatment NE.However, there was an effect of observation day (P < 0.001) and of the interaction day × treatment (P = 0.002).
When no food was available, animals of the NE treatment visited the feeder more often (5.83 ± 0.80 visits/day) than EE animals (4.51 ± 0.76 visits/day), with significant differences on some days of the experiment (Figure 3(B)).No difference was observed in the total number of visits to the feeder between treatments (P = 0.20), but there was an effect of observation day and of the interaction day x treatment (P < 0.0001).Animals of the NE treatment visited the automatic feeder more often in the first and last third of the experiment (Figure 3(C)).
There was no effect of treatment on serum cortisol levels, regardless of time, before or after (P > 0.05).Unlike cortisol, mean serum oxytocin levels differed between treatments before and after brushing (P < 0.01), with means always being higher for the EE treatment (Table 3).
Paired analysis within treatment comparing the time points before and after brushing (Table 4) showed a decrease in serum lactate levels after brushing only for the EE treatment (P < 0.05), an increase in cortisol levels for both treatments (P < 0.01), and an increase in oxytocin levels only for the EE treatment (P < 0.01).

Discussion
Over the 31 days of tactile stimulation, heifer calves of the EE treatment allowed less time to be brushed than those maintained in the NE environment.This result does not indicate the rejection of brushing nor a negative effect of environmental enrichment in EE heifers since the animals of this treatment exhibited a decrease in lactate and a lower increase in cortisol level compared to the NE group.A possible explanation would be that EE heifers received greater stimulation from the environment to play and spent their time using toys, followed by resting.On the other hand, heifer calves of the NE treatment were more bored and allowed longer brushing.The cortisol levels were low in both treatments, indicating a low level of stress as animals subjected to gentle handling exhibit cortisol levels between 11 and 22 ng/mL (Grandin and Deesing 2008).
Afterwards, they spent more time standing still or eating, activities related to idling.Park et al. ( 2020) also found beneficial effects of environmental enrichment using brushes in feedlot steers.According to the authors, enrichment did not influence lying, eating or drinking but reduced aggressive and stereotypic behaviours, suggesting that a brush provided mental and physical stimulation and consequent welfare to the animals.
The duration of tactile stimulation used in studies on cattle is approximately 5 min (Schmied et al. 2008b;Bertenshaw et al. 2008;Silva et al. 2017;Shahin 2018;Ujita et al. 2020b).In this experiment, the pre-established brushing duration of 4 min was found to be longer than that allowed voluntarily by the   heifers.Finally, the duration was established by the animal itself since the animals were not restrained for tactile stimulation.
Considering very young animals, this duration can be reduced to 2-2.5 min, which was the preference of heifers.Tactile stimulation through brushing had positive effects in the present study, in agreement with studies showing a reduction of stress and improvement of negative behavioural reactions caused by aversive management and veterinary procedures (Waiblinger et al. 2004).These authors used tactile stimulation to reduce the stress of Brown-Swiss and Simmental cows during veterinary procedures and observed a significant difference in heart rate (P < 0.05) between cows handled before and during artificial insemination (75.3 ± 1.7 bpm) and control cows (82.1 ± 2.7 bpm).
Tactile stimulation does not only benefit young animals as observed in the present study but can also modify social behaviour.Bertenshaw et al. (2008), applying tactile stimuli for 5 min to the neck, shoulder, and head of unrestrained Holstein heifers, reported that the animals were less reactive during milking after calving.Similarly, Ujita et al. (2020b) observed improvement in the behaviour of Dairy Gyr cows in response to brushing before and after calving, with positive changes in the animal's behaviour in the milking parlour and improvement of milk production.
Environmental enrichment is a strategy used to reduce the stress caused by confinement, to decrease frustration, and to trigger natural behaviours of the animals (Mandel et al. 2016).Environments that stimulate play behaviours in animals provide better welfare conditions (Held and Špinka 2011).The behaviour of the Girolando heifer calves was influenced by environmental enrichment, which significantly improved animal welfare as demonstrated by the longer time the animals were engaged in play behaviour after brushing (tactile stimulation) when submitted to an enriched environment.In a study providing physical enrichment items to calves reared in individual pens, Pempek et al. (2017) observed positive this behavioural pattern effects of EE on growth, behaviour and response to novelty after weaning.Calves receiving environmental enrichment spent more time (P = 0.03) engaged in locomotor play (galloping, jumping, hopping, kicking, or turning) and played more frequently (P = 0.04) than calves of the control group.
The environmental enrichment provided in the present experiment stimulated the animals to spend more time engaged in play behaviour and/or lying down, while NE animals spent more time in activities such as walking and idling.Zobel et al. (2017), who applied environmental enrichment (brush and hanging rope) to pair-housed calves, observed a similar pattern of interaction with the enrichment items.The authors suggested that the use of brush by young calves could be related to the lack of maternal contact and that the rotating brush may mimic maternal grooming.Ude et al. (2011) evaluated the effect of environmental enrichment on the time spent in the feeding area and crosssucking behaviour of group-housed calves.The animals were submitted to a control treatment and to an environmental   enrichment treatment.Both groups had access to automatic feeders.The time the animals spent in the feeding area after sucking increased with age (P < 0.05) in the control group, while this duration decreased from 26.0 to 8.0 s in the group submitted to environmental enrichment.These results agree with those of the present study in which NE heifer calves increased the number of visits to the automatic feeder when no food was available.On the other hand, the results reported by Ude et al. (2011) regarding cross-sucking differed from our findings.The authors observed a lower percentage of crosssucking animals for the environmental enrichment treatment, while there were no differences between treatments in the present study (Table 2).Using a drum can containing hay as environmental enrichment for beef cattle, Ishiwata et al. (2006) evaluated the behaviour of these animals at intervals of 10 min over a period of 2 h after morning and evening feeding for 3 consecutive days per month, totalling a period of 5 months.The use of the drum can increased (P < 0.01) active behaviours such as eating, drinking, investigating and moving, but there were no significant differences between treatments (P > 0.05) for cortisol, glucose or plasma adrenaline levels.The results regarding the active behaviours of the animals are consistent with those of the present study, as were the cortisol results which exhibited the same magnitude in the two treatments.On the other hand, cortisol levels could be considered of low magnitude.
The present study clearly demonstrated that brushing calmed the animals, leaving them more relaxed when the environment was enriched, as indicated by the lower lactate and higher oxytocin concentrations for the EE treatment after brushing.The level of oxytocin was higher for the EE treatment, compared to the NE, even before the brushing, which indicates that the EE was very important to increase the sensation of pleasure for calves.
Comparing the difference before and after tactile stimulation within each treatment, EE animals exhibited a lower lactate concentration and a higher oxytocin concentration after handling.Lactate is produced immediately after a sudden fright or tension, while cortisol is related to chronic stress or suffering (Smith 2000;Stockham and Scott 2002).In the present study, a lower lactate concentration was observed in heifers receiving EE after brushing, but not for heifer calves of the NE treatment.As the serum levels of lactate were higher before and after tactile stimulation for NE treatment, a possible explanation would be that these animals were more stressed while handling the calves for the collection of blood samples, when compared to the animals in the EE treatment.
According to Chen et al. (2015), brushing may exert similar effects as maternal contact on serum oxytocin concentration in calves, as does the act of licking another animal, a behaviour resembling tactile stimulation of the animals.This behaviour characterized by social licking is a reliable indicator of friendship among animals since it is a care behaviour (Val-Laillet et al. 2009;Wasilewski 2003;Boissy et al. 2007).Tactile stimulation exerted a positive effect since it reduced the stress of calves, demonstrated by the increased release of oxytocin (a hormone related to the formation of social bonds).Similar results have been reported in several studies (Carter 1998;Ferguson et al. 2001;Bates et al. 2004).
The release of oxytocin is related to natural behaviours of animals and brushing increases the concentration of this hormone through a positive feedback mechanism, triggering social behaviours and positive emotions (Chen and Sato 2017;Lee et al. 2009).This was observed in our study in which paired analysis showed a significant increase in the concentration of this hormone after brushing.Yayou et al. (2014) evaluated the behaviour of dairy calves in open-field tests and in response to a novel object.The authors suggested that oxytocin is positively associated with positive behavioural response to novel environmental stimuli, such as greater investigative and exploratory behaviour.The longer time spent playing in the environmentally enriched area agrees with the findings of Yayou et al. (2014) and Coulon et al. (2013) who associated a more active behaviour of the animals (in terms of time) with higher oxytocin and lower cortisol concentrations.These parameters were also evaluated in our study on heifer calves and the results showed that the treatment combining tactile stimulation and environmental enrichment exerted a significant and positive effect on stress parameters, suggesting improvement of animal welfare during the period when tactile stimulation was applied.

Conclusion
Brushing positively influences the welfare of group-housed heifer calves reared in an enriched environment and the duration of brushing heifer calves aged 30 days can be 2-3 min per day.Environmental enrichment is a low-cost tool with important positive effects on heifer rearing.

Disclosure statement
No potential conflict of interest was reported by the authors.

Figure 1 .
Figure 1.Layout of the area where the experiment was carried out.(A) Area with environmental enrichment; (B) non-enriched area; (C) feeding area with hay, concentrate and water, and (D) automatic feeder.Total experimental area 700 m 2 .

Figure 2 .
Figure 2. Observed frequencies of the duration of tactile stimulation (A), mean duration of tactile stimulation (brushing) (B) allowed by heifer calves over 24 days of the experiment according to the presence (EE) or absence (NE) of environmental enrichment.

Figure 3 .
Figure 3. Suckling speed (A), number of visits to the automatic feeder when food was not available (B), and total number of visits to the feeder (C) according to treatment (presence [EE] and absence [NE] of environmental enrichment).Means with lowercase letters differ significantly between each other on the same day (P < 0.05).

Table 1 .
Description of the structure of the dependent indicator variables of animal welfare and of the models and procedures used for the analyses.

Table 2 .
Estimated mean time (min) spent in each activity after tactile stimulation over 2 hours of continuous uninterrupted observation (first and second hour) in the presence (EE) or absence (NE) of environmental enrichment.
Means of each treatment followed by superscript letters in the same row differ significantly from each other by the Tukey test (5% probability).

Table 3 .
Least square means and respective standard errors of biochemical variables before and after tactile stimulation in the groups receiving (EE) or not (NE) environmental enrichment.
Means of each treatment followed by superscript letters in the same row differ significantly from each other by the Tukey test (5% probability).

Table 4 .
Paired analysis of the difference in least square means and respective standard errors of the biochemical variables before and after tactile stimulation in the presence (EE) or absence (NE) of environmental enrichment.