1. Introduction
The Australian beef industry is the most common and geographically dispersed agricultural activity in the country, with around 50% of all agricultural farms carrying beef cattle [
1]. Environmental features can vary greatly between regions, farms, and paddocks through differing climates, terrain, water sources, pasture composition, and the presence of other natural or artificial objects. This variation in surroundings likely results in differing degrees of environmental enrichment opportunities for pasture-based cattle, ranging from environmentally complex areas through to ‘barren’ pastures (i.e., devoid of natural or artificial objects beyond grassed fields). There has been a large amount of research into the use and benefits of environmental enrichment provision for dairy cattle welfare and production [
2], as well as enrichment provision for beef cattle in feedlots [
3,
4]. However, little is known about whether a barren pastured environment will impact beef cattle behaviour and welfare, and if there is utilisation of and preference for enrichments when provided to cattle kept in these conditions.
One major goal of environmental enrichment is to increase the range of normal, species-specific behavioural expression [
5]. Grazing, ruminating, and resting/lying occupy over 90–95% of beef cattle’s time budget when housed on varying types of pastured environments. Other behaviours that occupy the remaining 5–10% of their time budget include social behaviours, drinking, excreting, walking, and investigating or interacting with an object within the paddock [
6]. The degree of environmental complexity under which all these observations have been conducted is not certain, but common behaviours of grazing, resting, then walking in beef cattle across paddocks with some amount of vegetation have been reported [
7]. Studies on feedlot and dairy cattle can give some indication of preferred enrichments such as mechanical brushes [
8,
9], other grooming devices [
4], and hanging ropes [
9,
10]. However, these animals are generally housed with either no or restricted access to pasture [
2], which impacts their time budgets and may allow more time for enrichment use in place of grazing time. Barren pasture environments do allow for grazing, walking, and resting but may still be restrictive of a full cattle behavioural repertoire and could be optimised by enrichment provision. Lying time may be increased through the provision of alternative surface substrates, while additional behaviours that may be facilitated through enrichment for pasture-based beef cattle could include grooming and oral behaviours.
Grooming is seen in cattle at relatively low rates, at approximately 2% of the day, and includes self-grooming, allogrooming, and grooming on objects [
7]. It is thought to have multifaceted causes, including satisfying a physical requirement to stay clean through the removal of dirt, ectoparasites, and other substances, and also potentially as a mechanism to allow for de-arousal [
11]. Self-grooming and allogrooming were not impacted when pasture-based beef cattle were allowed access to trees for grooming [
12]; however, self-grooming, but not allogrooming, were increased when dairy calves were provided with a rotating brush [
13]. Therefore, a barren paddock with no trees or similar objects may deprive cattle of the opportunity to perform some grooming behaviours, which may not be substituted with self-grooming and allogrooming.
Oral manipulation of objects, particularly chewing, licking, and nibbling, are also seen in cattle, notably dairy calves and intensively managed cattle [
9,
10,
14]. When cattle are in a feedlot environment or on thick improved pastures (i.e., grazing lands not used for cropping), eating time is reduced and replaced with other oral behaviours when compared to cattle in paddocks with low pasture availability [
15]. Although some oral behaviours such as tongue rolling may be defined as stereotypic behaviours [
16], other exploratory behaviours may be rewarding to the cattle. Beneficial oral behaviours could be encouraged in both paddock-based and intensive systems through the provision of objects such as ropes, which can be chewed or suckled.
Lying is an important behavioural requirement of beef cattle, who have been observed to spend around 11 to 12 h a day lying at pasture [
17,
18]. Lying is often promoted for indoor-housed dairy cattle through the provision of bedding materials such as straw, wood shavings and chips, or sand [
19]. However, lying behaviour shows improvements when dairy cattle have access to pasture [
20,
21], dependent on pasture quality and grazing requirements [
22]. When provided with or without a woodchip area at pasture, lactating dairy cattle did not show consistent differences in lying time and most lying was on the grass [
23]. However, it is unknown if beef cattle would use additional bedding in barren pastures, or whether this resource would enhance time spent on this activity through either increased lying time and/or bouts, particularly if inclement weather resulted in muddy surfaces [
24,
25].
Overall, environmental enrichment has the potential to improve welfare by reducing frustration and boredom felt when an animal is housed in a barren or inappropriate environment and is unable to perform specific behaviours [
5,
26,
27]. Additionally, there is increasing interest in enrichment as a method to promote positive experiences. For example, grooming in cattle has been suggested to be a hedonistic behaviour, and is linked to positive affective states [
28,
29,
30].
The aim of the current study was to determine the preferred environmental enrichments of small groups of beef cattle housed in paddocks barren except for pasture, water, and the chosen enrichments; a rope, brush, stump, and woodchip pile. It was predicted that, over a period of 3 weeks, enrichment use would peak during the first week due to novelty, while the use of any highly valued enrichments would be maintained over time. In addition, high-value enrichments were expected to result in a greater number of displacements, as an indication of increased competition over the resource. Finally, the rope was expected to be least valued by the cattle, as comparable oral behaviours could potentially be displayed through grazing, whereas no comparable substitutes for the brush, stump, and woodchip pile would be available in the pasture.
2. Materials and Methods
2.1. Ethical Statement
The procedure of this study was approved by the CSIRO Agriculture Animal Ethics Committee (Armidale, ARA 21-09), under the New South Wales Animal Research Act 1985.
2.2. Facilities and Animals
The experiment was undertaken at CSIRO, FD McMaster Laboratory, Chiswick, Armidale, NSW from October to December 2021. Weather data were collected from an onsite MEA weather station (Green Brain, 41 Vine Street, Magill, SA, Australia). The average 24 h daily temperature was 14.3 °C (±0.4 °C), average daily rainfall was 4.0 mL (±1.0 mL, average humidity was 80.6% (±1.4%), and average wind speed was 13.2 km/h (±0.8 km/h). It was noticed that some areas of the paddocks became waterlogged and muddy due to rainfall, but this was not quantified.
Four neighbouring paddocks approximately 3 hectares each in size were used. Each contained a water trough and four enrichments, which were positioned 30–35 m apart in proximity of the observation area for both live data collection and video recording (
Figure 1), all of which were characterised by a similar mix of native and introduced pasture. Pasture data were not collected for this study, although experienced staff deemed the quantity sufficient for the number of animals grazing without any requirement to provide additional feed. No other objects, such as trees, were in the paddocks, and electric fences were used to prevent the use of fences for grooming. The four enrichment objects (
Figure 2) included were:
Cattle brush (Redpath, 16 Bounty Place, Kelvin Gove, Palmerston North 4414, New Zealand)
Tree stump placed on side with semi-intact root balls (found at the Chiswick site, approximately 1.6 m × 2.4 m × 1.5–2.5 m)
Sisal rope hanging from a slanted metal post, knotted at ends (10 mm diameter, two strands 1 m in length)
Woodchip mound (3 m in diameter, 1–1.5 m high)
Fifty-six Angus steers of approximately one year of age kept in groups of seven animals each across two cohorts were used in the study. Cattle from Cohort 1 were sourced directly from the Chiswick site, whilst Cohort 2 were purchased from a commercial producer, and arrived on site 3–4 weeks before they entered the testing paddocks. Prior to the current study, all cattle were housed in a range of standard commercial paddocks and were expected to have experience with objects such as trees and stumps, but were naïve to enrichments such as the brush, rope, and woodchip.
Cattle were weighed using walk-over weighing scales in a crush 1–2 weeks prior to testing commencing, and groups within cohorts were balanced for weight (average weight ± SE Cohort 1 = 303 ± 5 kg; Cohort 2 = 333 ± 6 kg). All cattle were retained on the Chiswick farm at completion of the study.
2.3. Study Design and Data Collection
Prior to cattle entering their enriched paddocks, individuals within each group were marked with numbers 1–7 using livestock paint (Leader Products Pty Ltd., Craigieburn, VIC, Australia). This was repainted once per week on a non-observation day, to ensure ease of reading. Cattle were walked into the centre of their test paddock and had free access to enrichments within the paddock for a period of up to 3 weeks. Group placement into the paddocks was staggered to facilitate ease of behavioural observations, with Day 1 considered as the day each group was introduced to their respective test paddock. Initially, Group 1 was continuously observed for 12 h (07:00 to 19:00) for four consecutive days to determine peak enrichment use times and refine the behavioural ethogram. After that, Groups 2–4 were placed into each paddock over 3 consecutive days, where they remained for up to 3 weeks. Observations were made on Days 1, 2, 8, 9, 15, and 16. The entirety of Cohort 1 completed their time in the test paddock before Cohort 2 was introduced to the paddocks.
On each observation day, the cattle were watched between 07:00 and 19:00 across Cohort 1 and from 06:30 to 19:45 across Cohort 2 due to lengthening daylight hours. Preliminary findings in Cohort 1 indicated the mornings and evenings to be the periods of peak enrichment use. All groups were observed live on their first two days, then a combination of live and video recordings was utilised for behavioural data collection on subsequent days (live from 15:00 to 19:00 for Cohort 1, and 16:00 to 19:45 for Cohort 2, with video recordings the remainder of time). Group 1 was an exception, with only video recordings used for days 9 and 16.
Live observations of behaviours were recorded by a minimum of two personnel at a time through an annotation application (CSIRO AnnoLOG v 1.0.23, St. Lucia, Brisbane, QLD, Australia) installed on a Samsung Galaxy Tab A 7.0 (Samsung, Seoul, Korea). All personnel were trained by a single individual (ED) who devised the behavioural ethogram (
Table 1). The time a relevant behaviour began, the duration, and the animal ID were recorded. Video recordings were taken on GoPros (HERO 7 and 8, San Mateo, CA, USA), set up in locations as per
Figure 1. Videos were replayed on “VLC media player” and time of relevant behaviours, duration, and Animal ID (when visible) were recorded.
Frequency of enrichment use was determined by counting the number of bouts. A bout ended when an animal moved over a body-length away from an enrichment, or had no interaction with an enrichment for 30 s or more.
2.4. Data and Statistical Analyses
Data from the AnnoLOG application were directly imported into Excel and combined with data from video recordings. All data analyses were performed in “R” [
31]. Main effects were considered significant at
p ≤ 0.05 and trends at 0.05 <
p ≤ 0.10.
2.4.1. Latency to Interact with Enrichments
The latency to first interact with an enrichment was obtained from the Day 1 data for all animals. This was analysed using the “survival” package [
32], with a stratified Cox proportional hazards model. An animal that failed to interact with a specific enrichment within 11 h was deemed a censored result. Each individual was treated as an experimental unit. To account for the effect of ‘Group’, it was fitted as a random effect.
2.4.2. Daily Enrichment Use
Summary statistics were calculated to compare the raw durations spent performing specific behaviours with enrichments, as a percentage of total time spent interacting with enrichments, over all groups (n = 8) and observation days (n = 6).
To examine any potential “queuing” effect for a specific enrichment, we compared the total duration interacting with an enrichment to the time spent “Near” (within one body-length of an enrichment with no other interaction).
Daily data were then separated into four Early/Late AM/PM time-period categories. Observation periods were each 3 h for Cohort 1, and 3–3.75 h for Cohort 2, to account for different hours of observation between Cohorts 1 and 2 (
Table A1). Due to technical issues with cameras resulting in a failure to record useable data for short periods, four time periods were excluded from the analysis (Group 1, Day 9, Early PM; Group 3, Day 9, Early AM; Group 5, Day 8, Late AM; Group 6, Day 9, Early PM).
The frequency of interactions with enrichments, displacements around enrichments, and daily weather data (temperature, rainfall, humidity, and wind speed) were then summarised by day number (n = 6), group (n = 8), enrichment (n = 4), and time-period category (n = 4). This resulted in a sample size of 752 unique events.
A principal components analysis (PCA) was performed using the “psych” package [
33], to reduce the dimensionality of correlated weather variables. The first two principal components (PC) had eigenvalues greater than 1 and were retained for use in later models, with their loadings compared to raw data used to infer relationships between extracted PCs and real-world weather conditions.
Models with all two-way interactions (Enrichment × Day, Enrichment × Time Period, and Enrichment × PCs) did not converge, and examination of raw data did not show obvious patterns, so these interactions were dropped to simplify models.
To analyse data examining enrichment use, the “lme4” package [
34] was used. Group was the experimental unit. A two-stage modelling approach was taken, as there were many time periods in which no interaction occurred for specific enrichments. Firstly, a model was fitted for the probability that a specific enrichment was used or not, and then, only in time periods where use had occurred, a model was fitted for the frequency of use and number of displacements.
For modelling the probability of an enrichment being used, ‘Use’ was considered as a binary outcome (i.e., whether or not an enrichment was used within the time period by any animal), and a generalised linear mixed model (GLMM) was fitted of the form:
where
is the probability of an enrichment being used within a time period (i.e., P(Use
t = 1). ‘Group’ is a random effect, and all other effects are fixed. Model residuals were checked for normality and homoscedasticity using a visual assessment of Q-Q and residuals relative to fitted values plots. Significance testing of fixed effects was conducted using Wald Chi-square tests from the “car” package [
35], and post hoc pairwise comparisons were made using contrast statements from the “emmeans” package [
36].
For the second stages of analysis, a Poisson GLMM was fitted for the frequency of enrichment use, and a negative binomial GLMM for displacement data, using the “MASS” package [
37], both generated using a binomial family link function. Model residuals were used as above to confirm adequate model fit, as well as significance testing of fixed effects and post hoc comparisons.
4. Discussion
The types of enrichments that beef cattle will utilise when housed in pastures devoid of natural or artificial objects beyond grassed fields are relatively unknown. Identification of cattle enrichment preferences will provide the first step in ultimately determining their value and potential positive impacts on animal welfare and production. The current study aimed to determine beef cattle preferences for a variety of enrichments presented simultaneously in a pasture environment. Cattle interacted with all provided enrichments and, as predicted, there was a general pattern of enrichment use decreasing over time, although the brush, stump, and woodchip maintained a higher level of use than the rope based on frequency of interactions and number of displacements around the enrichments.
A large amount of lying behaviour was observed on the woodchip pile in this study. Beef cattle are shown to prefer and will spend more time lying at pasture compared to a feedlot environment [
25,
38], and the provision of bedding materials to indoor-housed dairy cattle is a common occurrence [
2]. However, there have been few studies into the provision of bedding materials to cattle at pasture, which can potentially aid in cattle comfort and cleanliness [
39,
40,
41], and therefore reduce thermoregulatory challenges [
42]. Due to higher-than-expected rainfall, it was noted that some areas of the experimental paddocks became waterlogged and muddy; however, this was not quantified. Sub-optimal conditions may prevent or reduce lying time of cattle on pasture [
39,
43], which has negative health impacts such as disruption to the hypothalamic-pituitary-adrenal axis [
44], and increased risk of lameness [
45]. Considerable individual variation in woodchip area use when provided to dairy cows at pasture has been seen, but no differences in overall lying times between those with access to the woodchip and those without were observed [
23]. The relatively high level of use of the woodchip even after three weeks in the current study suggests that some aspect of the woodchip pile was viewed as favourable to the cattle, whether this was a more comfortable surface due to better drainage [
39] or improved thermoregulation by keeping the cattle out of mud [
46]. It is also possible that the woodchip acted as an insect repellent [
47], or there was some other unknown aspect such as the height that was attractive to the cattle. This indicates the potential that exists for further research on specific bedding or structures that facilitate lying behaviour in a pasture environment, especially under sub-optimal weather conditions; however, it is likely that the woodchip would need to be regularly replaced or refreshed, as quality may decline with time.
Both the brush and stump allowed for the performance of grooming behaviours, which have been extensively studied in feedlot and dairy cattle. Over 75% of cattle in the current study interacted with the grooming brush and the stump within 11 h, which is comparable to studies in dairy cattle in which 57–79% of animals interacted with a brush within one day [
8,
48]. Interestingly, daily frequency of use of the brush in the current study was higher than has previously been reported for cattle housed in a feedlot [
4]. Providing trees in a paddock setting to beef cattle allows for grooming that is not compensated for by other grooming behaviours, namely self- and allogrooming [
12]. Similarly, providing rotating brushes to dairy calves does not reduce the amount of self-grooming and allogrooming [
13]. Therefore, these enrichments facilitate the expression of grooming behaviours that would not be possible on an otherwise bare pasture, making them a potentially valuable inclusion in these environments, although further research is needed on the welfare impacts of the restriction of these behaviours.
Cattle were observed chewing woodchips and performing oral manipulations with both the rope and the stump, specifically chewing the rope and the bark and licking the root ball of the stump. Enrichments allowing for oral manipulation, specifically ropes, are provided to cattle when they may be deprived from fulfilling their natural oral behaviours of suckling and/or grazing, but mature cattle appear to interact with these ropes less; dairy: [
9,
10]; beef: [
14,
49,
50]. Bark stripping is an oral stereotype seen in cattle [
51,
52], and is associated with a lack of foraging [
16,
53], some dietary deficiencies such as lack of manganese or fibre [
54,
55,
56], or social learning [
51]. Considering oral behaviours were mainly concentrated on different enrichments between cohorts (woodchip vs. stump), it is possible that this was a socially learnt behaviour. As cattle in the current study were able to graze freely, this amount and variety of oral behaviours was not expected. However, the hanging rope appeared to be the least valued enrichment item, based on use and displacements occurring around it. Rope use may have increased if the stump and woodchip were not present. Overall, it appears that some amounts of oral behaviours other than grazing, perhaps focused on exploration, may be a natural and rewarding behaviour for cattle even when housed on pasture.
It was anecdotally noted that cattle did not use enrichments when it was raining throughout the current study, instead choosing to stand without grazing as observed in other studies [
57]; however, weather did not affect the likelihood of an enrichment being used, only the frequency of use given that at least one interaction occurred. It is likely that weather impacts enrichment use, for example brush use decreases under hot and humid conditions [
58], whilst use of a woodchip area increases with temperature but is not affected by rainfall [
23]. Our results may therefore reflect a rebound effect, in which cattle increased their use of enrichments following a brief period of rain, as weather data were based on the whole day rather than each time period. Additionally, pasture availability—whether due to seasonal or regional differences—may influence levels of enrichment use, as other behaviours such as time grazing tend to increase with decreasing pasture availability, whilst time resting decreases [
15,
59].
Early morning and late afternoon appeared to be the favoured times for enrichment use overall. This is in line with the activity and grazing peaks around sunrise and sunset seen in cattle housed on pasture [
7,
60,
61], and peak brush use time from other studies [
10,
48,
58,
62]. A limitation of the current study was that observations only occurred during daylight hours, and none at night. However, previous studies examining enrichment use of dairy calves found that enrichment objects are used approximately one third less during the night than day [
9], and cattle at pasture spend a large proportion of the night lying down [
38].
A greater number of displacements occurred in the late afternoon in the current study, further indicating that this may be a preferred time for enrichment use, evidenced by the increased competition. However, the current study used smaller group sizes and smaller paddocks than would likely be found under commercial extensive beef grazing. Having limited access to enrichments with increasing group size may create more agonistic encounters between cattle, and potentially negatively impact welfare for lower-ranking individuals. For example, displacements per dairy cow/hour at a brush were higher than those at a feed-bunk, even though the brush was free for the majority of the day [
63]. Dominant cattle have been observed to use a brush before [
48] and for longer [
64] than low-ranking animals at peak times. However, displacements performed and received at a brush do not differ between dominant and subordinate cows [
64], and therefore it is possible that another underlying factor associated with social rank, other than competition, has an impact on the performance of rewarding behaviours and enrichment use. For example, the potential for a generally increased social stress level in subordinate cows might play a role in these relationships. Similarly, brush use is also reduced under times of stress or impaired welfare [
65,
66,
67], highlighting that adequate physical health and nutrition should be met before benefits from enrichments may be seen. Further research is required to determine the optimal enrichment to cow ratio to minimise agonistic interactions, especially for large herds.
The cattle’s prior experience of enrichment objects was not well known, which may have impacted results. All animals were raised in commercial paddocks, and potentially had prior experience with trees and stumps but were naïve to the other objects. Repeated exposure of cattle to a novel visual object reduces the time spent exploring it and time taken to cross it [
68], while repeated exposure to novel objects in a paddock decreases consistency for approach [
69]. This neophobia may be reflected in the shorter latency to interact with the more familiar stump on Day 1 than the potentially less familiar woodchip pile. However, this time taken to approach may be driven by individual animals, and therefore future research may seek to determine effects of previous experiences on enrichment use. Similarly, it is unknown how enrichments may be utilised longer term in a paddock environment, with further studies across a longer period (e.g., months) required to quantify these relationships.
It was not possible to identify individual animals on each day of the study. This should be considered in future experiments, as there is considerable variation between individuals in their enrichment use [
10,
23,
70]. Cattle also did not interact with the stump or brush for around 50% of the time they were near the object (within one body-length). This could represent a ‘queuing’ effect, which was not observed to this extent at the woodchip pile, likely due to its higher surface area, or the rope, likely due to the reduced competition over this resource. Identifying individual animals may therefore allow a closer analysis of the social aspects surrounding enrichment use, including leader–follower relationships and social learning. Additionally, specific impacts on all aspects of welfare—behavioural, physiological, and affective states—are unknown, along with any effects on production. Finally, it is not known what effect a loss of enrichment has on these cattle, which is particularly relevant in Australia due to the large variation in landscape features and the common practice of paddock rotation, along with potential loss of enrichment (either natural or artificial) when finished in a feedlot.