Social Behaviour in Lambs (Ovis aries) Reared under an Intensive System during the Prepuberty Period

Abstract

The social behaviour of twenty-one lambs of three sheep breeds (Chios, Karagouniki, and Synthetic) was observed from the age of 2.5–7 months, divided into three equal periods (A, B, C) of 1.5 months each. The frequency of the performed agonistic behaviours was higher (p < 0.01) before four months of age (period A: 11 × 10⁻⁴ ± 2 × 10⁻⁴) compared to the other periods (B: 5 × 10⁻⁴ ± 1 × 10⁻⁴; C: 8 × 10⁻⁴ ± 1 × 10⁻⁴). The same was noticed for affiliative behaviours performed (A: 21 × 10⁻⁴ ± 3 × 10⁻⁴; B: 8 × 10⁻⁴ ± 1 × 10⁻⁴; C: 5 × 10⁻⁴ ± 1 × 10⁻⁴), and for agonistic and affiliative behaviours received (p < 0.001). The lambs had higher levels of wool cortisol (23.398 ± 5.344 pg/mg; p < 0.01) at period C compared to A (7.899 ± 1.19 pg/mg). Breed affected the affiliative behaviours. Karagouniki showed higher frequencies than Synthetic (16 × 10⁻⁴ ± 3 × 10⁻⁴ vs. 6 × 10⁻⁴ ± 1 × 10⁻⁴). Blood cortisol was higher (p < 0.05) in the Synthetic (4.789 ± 0.928 ng/mL) compared to the Chios (1.805 ± 0.417 ng/mL) breed. Overall, lambs’ behaviour displayed changes upon time, with four months of age being a step to a next developmental stage with fewer social interactions and higher levels of cortisol.

1. Introduction

Living in groups is common among many mammalian species. The groups include individuals of various ages, kinship statuses and sex, as well as having partially competing interests [1,2]. It has been well established that sociality plays a pivotal role in social species [3], providing benefits to individual group members. Social animals often engage in selective interactions within a complex social framework [4] that lead to stable social bonds [5] that are associated with fitness benefits. Such benefits could be greater longevity [6], and enhanced offspring survival (baboons) [7], as well as increased reproductive success (feral horses) [8].

The social environment can also affect the physiological responses of individuals [9]. The presence of familiar conspecifics narrows the effects of experimentally induced stress in species like rats, mice, goats, and monkeys [10,11]. Additionally, social interactions are linked to hormonal changes. In baboons, socially integrated males have lower levels of cortisol compared to non-socially integrated males [12]. In general, when mammals experience certain stressors, the hypothalamic–pituitary–adrenal axis responds with the secretion of cortisol or corticosterone, depending on the species [13,14]. As a result, the concentration of glucocorticoids in the blood increases within minutes [15] and decreases within hours, based on the intensity and duration of the stressors [14,16]. Consequently, measuring blood cortisol can indicate the acute stress response of a mammal to further associate it with sociability. Furthermore, analyses of long-term changes in cortisol levels gained more attention in the past few years as an indicator of chronically stressful conditions [17]. Glucocorticoids are incorporated into emerging biological matrices such as hair/wool and increase cumulatively during the development or growth of the material. These long-term changes reflect the stress levels of a specific period [18,19] and could be used for further association with other aspects of mammalian life.

Sociality includes both benefits and costs. Increased competition for resources and mating partners is linked to the costs [9,20]. Consequently, social interactions have a strong effect on the fitness of individuals. However, even though sociality is a very important fitness trait, scarce information exists about how it is developed [21]. The time between weaning and sexual maturity is fundamental for behavioural development and yet most of the research has been concentrated on adults and nursing offspring [21]. Detailed descriptions of the attributes and behaviour of growing animals could shed light on the ways that youngsters resolve challenges to construct biologically successful lives [21]. For example, calf welfare is improved when the calf is reared with the dam or with peers compared to individual rearing [22]. Mitigation of risk is a basic element of youngsters’ life. Accordingly, aspects of growth, behavioural development and even social organisation appear to have evolved to modulate the risk of growing animals [21,23].

Farm animal species are generally gregarious [24] and their welfare heavily depends on the social environment. Furthermore, their social behaviour is plastic and dynamic and allows them to adapt to varying environmental and social conditions within a confined group [24]. Domestic sheep display gregarious social behaviour, develop stable social relationships with other members of the flock [25,26], and perceive isolation as aversive [22]. Adult sheep can recognise their group members, and under free-ranging conditions, they avoid unfamiliar animals [27,28]. The sociality of sheep is influenced by a variety of factors, including breeding period [29], age of animals [1,30], and environmental and management factors such as group size [31,32], shelter type [33], weather conditions [30], and individual characteristics such as temperament or personality [30,31]. In addition, social interactions could increase the stress levels of individuals [30], suggesting the importance of sociality. The social behaviour of sheep consists of both positive (affiliative behaviours) and negative interactions (agonistic and submissive behaviours). A study of lambs demonstrated an increase in both agonistic and affiliative behaviours on the first day of a new group composition [34], possibly as a result of the unfamiliarity of the individuals [35]. Lambs also prefer individuals of their own breed as partners in affiliative behaviours [36]. An early study [37] highlighted play behaviour in lambs of bighorn sheep, but to our knowledge, there is no recent study focusing on lamb social behaviour and how it evolves from weaning until puberty under intensive rearing conditions. This study focuses on (1) analysing the social behaviour in prepuberty lambs using non-intervening behavioural observations and (2) investigating the effects of age, physical growth, and breed on the expression of social behaviour. It was expected that lamb behaviour would change according to age, possibly affected by the growth rates of the animals.

The first social bond that lambs establish is with their mother as newborn lambs. In intensive sheep farm conditions, this bond is disrupted when the lambs are weaned very early. This stressful experience may lead to reduced productivity [38]. Nevertheless, apart from the bond with their mother, lambs also create bonds with peers, which may play a crucial role in minimising the negative effects of the early weaning conducted in intensive farming conditions. Knowledge on lamb behaviour and interactions with peers could be very useful and provide guidelines for high-quality farming practices considering both productivity and animal welfare.

2. Materials and Methods

2.1. Animals

This study was conducted at the premises of the Agricultural University of Athens on 21 weaned female lambs (Ovis aries). The lambs were all born within the same month (December 2020) and they were reared together with their dams until weaning. The weaning took place around 55 days post-partum (average weight: 18.83 ± 0.6 kg). After weaning, the lambs were all housed together in one enclosure, including both inside (pen; 30 m²; 0.76 lambs/m²) and a fenced area on the outside (yard; 130 m²; 0.23 lambs/m²). The animals belonged to three dairy breeds as follows: (i) Chios breed (CH; N = 7 lambs, average adult weight = 52 ± 1 kg, average birth weight = 4.6 ± 0.4 kg), Karagouniki breed (KG; N = 7 lambs, average adult weight = 59 ± 2 kg, average birth weight = 5.5 ± 0.3 kg), and a Synthetic (crossbred) breed (ST; 50% Orino Epirus, 25% Chios breed, 25% East Friesian breed; N = 7 lambs, average adult weight = 48 ± 1.3 kg, average birth weight = 4.2 ± 0.2 kg). Feeding management was daily between 07:00–08:00 h and 16:00–17:00 h based on a concentrated diet and alfalfa hay according to the animals’ nutritional requirements. The lambs had free access to water in the yard area for ad libitum consumption. An experienced veterinarian confirmed that no health problems or injuries were noted in the lambs throughout this study.

2.2. Experimental Period, Measurements, and Samplings

The experimental period started 15 days after the grouping of the weaned lambs (55 days post-partum) and lasted for 4.5 months. It was divided into three equal periods (A, B, C), each lasting 1.5 months. The weight and shoulder height of the lambs were measured once per week during the whole experimental period. For the weight, a farm animal scale (accuracy: ±100 g) was used with each lamb entering alone. Since weighing forms a farm routine, the animals were very well adapted to the weighing process and entered the scale voluntarily. They were weighed at the same hour of the day (09:00 h) and by the same person. For the shoulder height, each lamb was carefully handled. One person restrained the animal, and a second person measured the height using a Lydtin’s rod. At the end of each experimental period (A, B, C), blood and wool samples were collected. Specifically, sampling always started at 09:00 h. For blood sampling, two experienced people carefully handled the lambs. The first person restrained the animal. The person silently entered the animal pen and moved slowly, held the lamb’s head horizontally with one hand under the lower jaw without touching the throat, and the other hand behind the ears. The second person entered the pen with the necessary sampling equipment, crouched in front of the animal and took the sample [39]. For the blood samples, the SARSTEDT S-Monovette^® capillary system (Sarstedt AG & Co. KG, Nümbrecht, Germany) for blood collection (5 mL) was used attached with S-Monovette^® needle (20G 11/2) (Sarstedt AG & Co. KG, Nümbrecht, Germany) immersed with heparin 16 IU/mL. Until sampling was complete, all samples were kept at 4 °C. At the end of sampling, all samples were centrifuged at 2000× g for 10 min at 4 °C to extract the plasma and kept at −20 °C until further analyses. Regarding wool sampling, a sample of wool was collected from each lamb (N = 21) from the back of the neck, an area identified as ideal due to low dirt concentration. The area was shaved before the start of the experiment to make sure that the cortisol concentration referred to each examined period. Each wool sample was wrapped inside aluminium foil and placed in a labelled Ziplock bag. They were kept at room temperature until further analysis.

2.3. Behavioural Observations

The behavioural observations lasted for the whole experimental period. The morphological characteristics of the lambs were used for individual identification. To avoid biases caused by the presence of staff, e.g. due to feeding practices, the observations were conducted from 08:30 to 14:00. An expert observer utilised the CyberTracker software v. 1.0.490 (www.cybertracker.co.za; accessed 15 January 2022) to record the behaviours. A camera was also placed in the recording area to validate the live scoring results. Observer’s bias was evaluated through an intra-observer reliability test, comparing live and video scoring. Ten videos were randomly sorted out and re-scored by the same observer. Kappa Coefficients for frequency data and the Wilcoxon Pairs Test for duration were performed for the result comparison [40]. Intra-observer biases were discounted (Kappa Coefficient: K = 0.612, p < 0.01; Wilcoxon Pairs Tests: p > 0.05).

Based on previous studies, two sampling methods were used, focal and ad libitum sampling [41]. The behaviours scored were described based on findings from previous studies ([42,43,44,45,46]

Table A1. Animal behaviours (ethogram) scored in the present study and their descriptions. The description of the behaviours was based on previously reported studies [39,40,41,42,43].

Behaviours	Description
Maintenance
standing	Standing still, sporadically moving the head
lying	Resting with eyes open or closed
head hanging	Standing quietly with head hanging
foraging	Head inside the feeder trough
walking	Shifting from one place to another
running	Going from one place to another on a faster pace
Agonistic
clash	Two individuals facing each other move backwards and then lunge forward and hit each other head to head
threats	Turns towards or approaches another individual with his head down and then lunges without making contact
butt	Uses the front of his head to make contact with another individual
push	Uses other parts of the body except the head to make contact with another individual
Submissive
avoidance	Actively moves away from another individual as a result of a previous agonistic interaction
Affiliative
grooming	Grooms another individual’s body using the teeth
resting chin	Puts the chin on another individual’s rump or back
face contact	Places the nose near a recipient’s facial region and holds the position while smelling or rubbing the recipient’s face or head

). Focal sampling for each animal took place with observations evenly distributed across time and in random order. Each focal sampling lasted for 15 min. The behaviours that were scored during focal sampling were maintenance behaviours, including standing, lying, head hanging, foraging, walking, and running. All social behaviours performed or received by an individual were also recorded ad libitum. Those were as follows: (a) agonistic interactions including clash, threat, butt, and push, (b) submissive interactions including avoidance and (c) affiliative interactions including grooming, resting chin, and face contact. For each animal, the duration of maintenance behaviour and the total number of times engaged in each of the social interactions was divided by the total observation time and per observation period.

2.4. Plasma Sample Preparation and Cortisol Analysis

The levels of cortisol in the plasma of each lamb were assessed after extraction of cortisol. To carry this out, 10 μL of plasma was transferred in a glass tube and 1 mL of diethyl ether was added. The tubes were covered with aluminium foil and placed at −80 °C for 2 min. The liquid phase was transferred to clean tubes and left to evaporate in a water bath at 45 °C. The cortisol concentration in each sample was determined using a commercially available cortisol ELISA assay (Neogen, Ayr, Scottland, UK) following the manufacturer’s instructions. Reconstitution occurred with 100 μL of the kit’s extraction buffer. The recovery of the extraction method was 98.6% ± 18.2 (n = 4), and the inter- and intra- assay CV % were 13.3% and 7.8% (n = 3), respectively.

2.5. Wool Sample Preparation and Cortisol Analysis

Cortisol was extracted from wool using published protocols [47,48], with few modifications. From each sample, 200–250 mg of wool was collected. The samples were washed using isopropanol and air-dried at room temperature for 8 days. Weighed sub-samples of wool (50 μg each) were cut into fine sections (around 2 mm) using a pair of scissors and pure methanol was added to each sample in a labelled 2 mL Eppendorf tube to enable steroid extraction to occur. The samples remained immersed in methanol for 18–20 h for the extraction to be completed. The samples were centrifuged at 7000 g for 2 min at room temperature to separate the wool from the liquid phase and 750 μL of the supernatant were collected, placed in clean tubes, and left to evaporate in a water bath for 1–2 h. The wool cortisol concentration in each sample was determined using a commercially available cortisol ELISA assay (Neogen, UK) following the manufacturer’s instructions. Reconstitution occurred with 200 μL of the kit’s extraction buffer. The recovery of the extraction method was 94.5% ± 11.4 (n = 4), the inter- and intra-assay CV % were 9.2% and 4.6% (n = 3), respectively, while dilution tests showed a high correlation coefficient, r = 0.732 ± 1.3 (n = 5).

2.6. Statistical Analysis

SPSS v.26 [49] was utilised for the statistical analysis of the data. The significance level was at 0.05 and two-tailed tests were conducted. Non-parametric tests were performed since the data were not normally distributed (Kolmogorov–Smirnov test). Friedman’s two-way ANOVA by rank was used to identify differences in the maintenance behaviour, social behaviour, growth, and cortisol levels between the three examined periods (A, B, C). One-way non-parametric ANOVA was performed to identify breed differences in the maintenance behaviour, social behaviour, growth, and cortisol levels of the lambs. Post hoc comparisons were conducted using the Bonferroni test. Running, one of the maintenance behaviours, was not included in the analysis since it was not observed in this study.

2.7. Ethics Approval

All experimental procedures involving animals were approved by the bioethical committee of the Agricultural University of Athens (no.: 5/14-1-2021) under the guidelines of “Council Directive 86/609/EEC regarding the protection of animals used for experimental and other scientific purposes”. The animals were minimally handled and kept in a pen with low animal density. They were observed in their home environment to prevent stress. No interventions occurred by the observer and the animals were familiarised with the presence of the observer at a distance that did not alter their behaviour. Considering the blood and wool collection, the lambs were carefully handled by two experienced people (one was the observer).

3. Results

To investigate the effect of age on the examined parameters, a comparison between the observation periods (A, B, C) was performed (Figure 1;

Table 1. Differences in the frequency of social behaviour, growth, and cortisol levels between the three age periods examined (A, B, C). Values are presented as LSM ± SEM. Different superscripts (a, b) within the same row indicate significant differences. NS: non-significant.

Variables	Period A (2.5–4 Months of Age) n = 21	Period B (4–5.5 Months of Age) n = 21	Period C (5.5–7 Months of Age) n = 21	Significance (p Value)
Social behaviour
Actor agonistic (frequency)	11 × 10−4 ± 2 × 10−4 a	5 × 10−4 ± 1 × 10−4 b	8 × 10−4± 1 × 10−4 b	p = 0.006
Actor affiliative (frequency)	21 × 10−4 ± 3 × 10−4 a	8 × 10−4 ± 1 × 10−4 b	5 × 10−4 ± 1 × 10−4 b	p < 0.001
Actor submissive (frequency)	6 × 10−5 ± 2 × 10−5	0 × 10−5 ± 0 × 10−5	2 × 10−5 ± 1 × 10−5	NS
Receiver agonistic (frequency)	836 × 10−4 ± 20 × 10−4 a	5 × 10−4 ± 0 × 10−4 b	9 × 10−4 ± 1 × 10−4 b	p < 0.001
Receiver affiliative (frequency)	150 × 10−4 ± 22 × 10−4 a	7 × 10−4 ± 1 × 10−4 b	5 × 10−4 ± 1 × 10−4 b	p < 0.001
Receiver submissive (frequency)	0 × 10−5 ± 0 × 10−5	0 × 10−5 ± 0 × 10−5	1 × 10−5 ± 1 × 10−5	NS
Growth rates
Average Daily Gain (ADG; kg/day)	0.137 ± 0.009 a	0.106 ± 0.008 a	0.064 ± 0.007 b	p = 0.001
Height growth/day (cm/day)	0.151 ± 0.010 a	0.051 ± 0.008 b	0.055 ± 0.010 b	p < 0.001
Cortisol levels
Blood cortisol (ng/mL)	3.391 ± 0.517	2.966 ± 0.626	3.089 ± 0.892	NS
Wool cortisol (pg/mg)	7.899 ± 1.195 a	15.831 ± 2.350 b	23.398 ± 5.344 b	p = 0.002

). No statistically significant differences were found between the examined periods for walking and standing (p > 0.05). Head hanging was significantly lower in periods A and B compared to period C (Friedman: N = 21, x² = 9.711, p = 0.008). Lying was significantly higher in periods A and B compared to period C (Friedman: N = 21, x² = 21.810, p = 0.001). Foraging was significantly lower in period A compared to periods B and C (Friedman: N = 21, x² = 14.000, p= 0.001). Actor agonistic behaviour was significantly higher in period A compared to periods B and C (Friedman: N = 21, x² = 10.289, p = 0.006). The same was noted for actor affiliative behaviour (Friedman: N = 21, x² = 22.571, p < 0.001), receiver agonistic behaviour (Friedman: N = 21, x² = 13.904, p <0.001) and receiver affiliative (Friedman: N = 21, x²= 26.000, p < 0.001). No statistically significant difference was found in actor submissive and receiver submissive (p > 0.05) behaviours. Considering the lamb’s growth, Figure 1 shows the average daily gain (ADG) and height increase in lambs during the examined periods. ADG was significantly higher in periods A and B compared to period C (Friedman: N = 21, x² = 18.727, p = 0.001), and height growth (cm/day) was significantly higher in period A compared to periods B and C (Friedman: N = 21, x²= 25.810, p < 0.001) (Table 1). Blood cortisol did not differ between the three periods (p > 0.05), while wool cortisol was significantly lower in period A compared to periods B and C (Friedman: N = 21, x²= 12.095, p = 0.002).

Considering the breed effect (

Table 2. Differences in maintenance behaviour, social behaviour, growth, and cortisol levels between the three examined breeds of lambs (CH: Chios breed, KG = Karagouniki breed, ST: Synthetic breed; values are presented as LSM ± SEM). Different superscripts (a, b) within the same row indicate significant differences. NS: non-significant; N/A: not observed.

Variables	CH n =21	KG n =21	ST n =21	Significance (p Value)
Maintenance behaviours
Standing (frequency)	0.031 ± 0.012 a	0.104 ± 0.018 b	0.072 ± 0.015 a,b	0.001
Lying (frequency)	0.260 ± 0.033	0.319 ± 0.059	0.261 ± 0.040	NS
Head hanging (frequency)	0.255 ± 0.035	0.162 ± 0.037	0.231 ± 0.029	NS
Foraging (frequency)	0.420 ± 0.029	0.378 ± 0.047	0.418 ± 0.041	NS
Walking (frequency)	0.022 ± 0.007	0.031 ± 0.009	0.014 ± 0.005	NS
Running (frequency)	N/A	N/A	N/A	N/A
Social behaviour
Actor agonistic (frequency)	9 × 10−4 ± 1 × 10−4	10 × 10−4 ± 2 × 10−4	6 × 10−4 ± 1 × 10−4	NS
Actor affiliative (frequency)	12 × 10−4 ± 1 × 10−4 a,b	16 × 10−4 ± 3 × 10−4 a	6 × 10−4 ± 1 × 10−4 b	0.011
Actor submissive (frequency)	2 × 10−5 ± 1 × 10−5	2 × 10−5 ± 1 × 10−5	3 × 10−5 ± 1 × 10−5	NS
Receiver agonistic (frequency)	457 × 10−4 ± 210 × 10−4	148 × 10−4 ± 76 × 10−4	246 × 10−4 ± 107 × 10−4	NS
Receiver affiliative (frequency)	61 × 10−4 ± 22 × 10−4	54 × 10−4 ± 20 × 10−4	47 × 10−4 ± 17 × 10−4	NS
Receiver submissive (frequency)	0 × 10−5 ± 0 × 10−5	0.6 × 10−5 ± 1 × 10−5	0.8 × 10−5 ± 1 × 10−5	NS
Growth rates
Average Daily Gain (ADG; kg/day)	0.108 ± 0.009 a	0.119 ± 0.011 a	0.080 ± 0.009 b	0.037
Height growth/day (cm/day)	0.092 ± 0.014	0.091 ± 0.015	0.074 ± 0.011	NS
Cortisol levels
Blood cortisol (ng/mL)	1.805 ± 0.417 a	2.851 ± 0.442 a,b	4.789 ± 0.928 b	0.029
Wool cortisol (pg/mg)	10.861 ± 2.385	17.328 ± 3.056	18.939 ± 4.966	NS

), no statistically significant (p > 0.05) difference was found in most social behaviours of the lambs, or in their height growth. Actor affiliative was the only social behaviour that differed between breeds (H_(2,21) = 9.072, r= 0.50, p = 0.011), with the KG breed performing more affiliative behaviours compared to ST (p = 0.010). From the maintenance behaviours, only standing appeared to differ between breeds (H_(2,21) = 13.099, r= 0.54, p = 0.001), with the lambs of the KG breed standing more than the CH lambs (p = 0.001). Average daily gain (ADG; kg/day) was found to differ significantly between breeds (H_(2,21) = 6.597, r = 0.50, p= 0.037), with the KG breed showing a higher rate compared to the ST breed (p = 0.016) and the CH breed showing a higher rate compared to the ST breed (p = 0.049). Wool cortisol did not differ significantly between the examined breeds (p > 0.05), while blood cortisol differed (H_(2,21) = 7.065, r = 0.50, p = 0.029). Higher values were noted in the ST breed compared to the values of the CH breed (p = 0.025).

4. Discussion

Sociality in social species is highly associated with fitness benefits [3]. The time between weaning and sexual maturity is fundamental for behavioural development and the formation of biologically successful lives [21]. Despite the importance of this field of study, most research has been concentrated on adults and nursing offspring, especially before weaning takes place, with scarce information existing on young farm animals reared under intensive conditions between the post-weaning period and puberty. Intensive farming conditions (intensive systems) could potentially raise issues in terms of low welfare and elimination of animals’ behavioural expression, and thus research on animals’ behaviour exploitation under intensive conditions in groups of animals with different ages is of the utmost importance [50]. This study provides information about the behaviour of lambs in different age periods, from weaning to prepuberty, to further expand the knowledge of this crucial period of the youngster’s life.

The first finding emerging from this study is that the age of prepubertal lambs affected their social behaviour. The lambs appeared to perform more agonistic and affiliative behaviours at a younger age, specifically from 2.5 months up to 4 months (end of period A), compared to an older age (4–5.5 months and 5.5–7 months; Periods B and C). These behaviours occur to prepare the animals for future social interactions, such as reproduction [28]. In our case, the pubertal age of CH lambs is 289 days and that of KG lambs is 301 days (almost 9–10 months of age in both breeds) [51]. The increased interactions were controlled for the recent formation of the group since the lambs were left for 15 days to interact and familiarise with each other before the observations began. Furthermore, in domestic ewes, the development of social familiarity is established very rapidly after the introduction of a novel and unfamiliar conspecific [28]. It is reported that 24 h is a sufficient period for two ewes to familiarise with each other [28].

Accordingly, one explanation for the extensive social interactions until 4 months of age might be play behaviour. Social play in sheep has been defined as unpredictable and often as shaking body movements that include components of rotational behaviours [37]. It also includes running activities which are considered play when occurring without attempting to avoid conspecifics or predators. This non-contact play, also defined as locomotor play, mostly occurs in younger lambs (younger than three weeks of age) [37] and was not observed in this study, as seen from the zero running frequencies of the lambs. Running is a behaviour that occurs more frequently in nursing lambs compared to weaned lambs or mature sheep as it reduces with age [37]. Furthermore, the intensive farming system does not promote behaviours of high activity. Locomotor play is usually replaced with contact playing in older lambs, which includes butting, pushing, touching heads and kicks [37], behaviours that can be characterised as agonistic. Newborn lambs that accompanied their ewes never played with resident lambs by using contact patterns [37,52]. Only locomotor play occurred at this young age. In contrast, young lambs that were weaned and associated with resident lambs used contact play [37,52], a fact that is also noted in the present study. Animals at different ages express play behaviour in different frequencies, with infant mammals playing less than juveniles, and juveniles playing more than adults [52,53]. Accordingly, the main period in which the animals play coincides with the main period of physical, hormonal, and social development [34], which is also supported by our results, as since the period between 2.5 and 4 months (Period A) was a period of reduced wool cortisol and increased height gain. The lower levels of wool cortisol measured at the end of period A (4th month of age) indicate lower stress levels compared to the other two periods, which may be associated with the more frequent interactions between the conspecifics. Finally, as the age increases, the breeds start to differentiate by means of weight and height. The KG breed had significantly higher weight compared to the other two breeds in the total period of observations (Figure A1). Therefore, the increased animal dimension (i.e., height) after the 4th month of their age could lead to a more stressful situation between individuals without similar morphological characteristics, and thus decreasing their interactions between them.

Another finding highlighted by the present study is that with the increase in age, specific maintenance behaviours (e.g., lying) decreased, and others (e.g., foraging and head hanging) increased. Foraging and head hanging are linked to each other since an increase in foraging behaviour could be associated with increased head hanging, as the latter facilitates rumination. In accordance with this finding, a previous study on goat youngsters also suggested that the time spent lying decreased and time spent feeding and ruminating increased with age [54]. Accordingly, the increase in foraging and head hanging after 4 months of age (end of Period A) could signal the full functionality of the digestive system and rumen, leading the lambs to greater foraging activity. However, further research is needed to clarify the assumption of a linkage between behavioural data and the development of rumen in sheep.

According to our findings, most of the studied behaviours were not affected by breed, which could be due to the fact that all examined breeds are of the dairy type. This is in contrast with previous studies where differences in maintenance and social behaviours have been observed in adult sheep that belong to different breeds [55,56]. Nevertheless, only standing and actor affiliative behaviours differed between the examined breeds largely as a result of their different genetic background. Individuals belonging to the KG breed flock more easily with individuals in short distance between them as the breed has been intended for semi-extensive farming in larger flocks. CH breed was intended for small-scale farming. The ST breed is medium-sized and was bred in semi-mountainous areas. It is a crossbreed of two local Greek breeds and a Friesian-type breed. Individuals of the ST breed tend to graze with distance between them without forming a tight flock compared to the other two breeds [47]. Accordingly, the higher values of standing and affiliative behaviours observed in the KG breed could be attributed to a higher tendency of KG individuals to flock. In addition, one possible explanation for the lack of any other behavioural differences could be the implemented rearing system (intensive), which may have influenced the expression of their natural behaviour [50]. Adaptation, defined as behavioural changes resulting from environmental pressures, has been previously documented in small ruminants [57]. Therefore, the limited available space of the intensive system compared to grazing conditions, together with the genetic differences, may also have altered the expression of affiliative behaviour between individuals of the KG breed.

Differences between breeds were also noted in some physiological parameters related to ADG (kg/day) and blood cortisol. The growth rate differences could be explained by the fact that the CH (adult average weight = 52 kg) and KG (adult average weight = 59 kg) breeds are heavier as adult sheep since they are intended for lowlands, while the ST breed (adult average weight = 48 kg), which is medium-sized, is bred in semi-mountainous areas. The semi-mountainous areas that the ST is bred for, require a smaller body to cope with mountain climbing and are usually linked to the limited presence of humans. This absence of familiarisation with human presence might also lead to higher blood cortisol due to human handling in the collection of samples. The observed physiological differences could also be explained by their different genetic background. Being bred in harsh and unfavourable environments requires specific adaptations that promote survival and are particularly important in parturition and juvenile periods when ewes and lambs are more vulnerable [55].

5. Conclusions

Age appeared to affect the behaviour of lambs as well as the height growth rate and cortisol concentration in wool. The noted differences in social behaviour may be associated with play. Four months of age appeared to be an important developmental threshold, a step to a following developmental stage with less play, less height gain, and higher levels of cortisol. The results of this study also highlighted that breed did not affect most behaviours, possibly due to behavioural adaptation. The KG breed presented more affiliative behaviours than the ST breed and spent more time standing compared to the CH breed. The ST breed demonstrated lower ADG compared to the CH and KG breeds and higher blood cortisol compared to the CH breed because of their different genetic background. As established by previous research, sociality has a strong effect on the fitness of individuals and detailed descriptions of lamb attributes and behaviour could shed light on ways that youngsters resolve challenges to form biologically successful lives. Future experiments in larger datasets, including male lambs, or focusing on other types of breeds (i.e., meat-type or wool oriented breeds) could also assist in further exploiting the knowledge of sheep social behaviour, especially during the postweaning period and before puberty, during which scarce information exists.