Social isolation of goats: significance of visual contact with conspecifics on behavioral and physiological responses

Abstract

Social isolation can increase distress in goats, particularly when they cannot maintain visual contact with conspecifics. This experiment was conducted to determine the behavioral and physiological responses in goats during isolation with or without visual contact with conspecifics. Male Spanish goats (uncastrated, 8 mo old, average weight 29.4 ± 0.59 kg) were randomly assigned to a control (CO) group with no isolation or to one of four isolation treatment (TRT) pens (1.5 × 1.5 m) with: 1) open grill panels but with no visual contact with conspecifics (IO), 2) covered grill to prevent visual contact (IC), 3) open grill with visual contact (IV), or 4) covered grill with a 30 × 30 cm window to allow visual contact (IW), for 90 min of social isolation (n = 12 goats per TRT). Blood samples were collected at 0, 30, 60, and 90 min (Time) from isolated and control goats. The experiment was repeated 1 wk later using the same animals, with each goat being subjected to the same isolation TRT the second time to study the effect of prior exposure to isolation. Friedman’s two-way analysis of variance by Ranks test in SAS showed that the median frequency of vocalization (rank score) in goats was high in the IO group, low in the IV and IW groups, and intermediate in the IC group (P < 0.01). The vocalization rank score was also higher (P < 0.01) during the first 30 min of isolation in goats. The median frequency of visual contact was higher in the IW group than in the IV group (P < 0.01). The frequency of climbing behavior was high in the IC and IO groups, low in the IV group, and intermediate in the IW group (P < 0.01). Repeated measures analysis using general linear models procedures in SAS revealed that plasma cortisol and glucose concentrations tended (P < 0.1) to be the highest in the IO group than in CO, IC, IV, and IW groups. Cortisol levels were also higher (Time; P < 0.05) at 0 and 90 min compared with 30 and 60 min. Norepinephrine concentrations decreased (P < 0.05) with Time, and plasma nonesterified fatty acid (NEFA) levels were affected by TRT × Time interaction (P < 0.01). Overall, epinephrine, norepinephrine, glucose, and NEFA concentrations were lower (P < 0.01) and cortisol concentrations and lymphocyte counts were higher (P < 0.01) when goats were exposed to isolation the second time. The results showed that goats with no visual contact with conspecifics during social isolation had greater physiological stress responses and spent more time vocalizing or trying to escape the pen, which may indicate distress.

Introduction

Social isolation of goats from their herd- or pen-mates is inevitable under commercial production situations in order to prevent the spread of contagious diseases, to provide veterinary care, or to prepare for slaughter when the animals are kept in single-file chutes prior to stunning. Isolation causes distress in goats and can remarkably affect their behavior and physiology (Carbonaro et al., 1992). When social isolation is combined with other stressors such as feed deprivation, the stress level could be higher (Kannan et al., 2002).

Goats vocalize more when socially isolated in an attempt to reestablish contact with conspecifics (Siebert et al., 2011; Briefer and McElligott, 2012; O′Bryan et al., 2019). Social isolation has been reported to increase plasma hormone and metabolites indicative of stress in small ruminants (Carbajal and Orihuela, 2001; Kannan et al., 2002). Stress is also greater when isolated goats cannot maintain visual contact with other animals compared with those that can see other animals (Kannan et al., 2002).

An animal’s previous experience may have a remarkable effect on its psychological stress level during handling procedures (Grandin, 1997). Although researchers have studied the benefits of using repeated exposure to stressors that are inevitable under practical situations to habituate livestock (Hemsworth et al., 1996) and poultry (Kannan and Mench, 1997), the results across species have not been consistent. Small ruminants do not appear to get habituated to isolation stress (Siebert et al., 2011) as repeated isolation evokes a greater increase in cortisol concentrations (Niezgoda et al., 1987), probably because the stressor is too intense (Hargreaves and Hutson, 1990).

The importance of visual contact during social isolation and repeated isolation on the emotional status of goats has not been systematically evaluated using behavioral and physiological measures to the best of our knowledge. Such data will provide valuable information in improving animal handing methods under commercial conditions by minimizing animal distress and improving health and productivity. The objectives of this experiment were to determine 1) the behavioral and physiological responses to different social isolation treatments (TRTs) with and without visual contact with conspecifics and to different durations of isolation and 2) the physiological responses to social isolation in goats previously exposed to the same isolation TRTs.

Methods

Animals

The protocols for this study were approved by the Animal Care and Use Committee (approval no. F-R-03-2019) at Fort Valley State University (Fort Valley, GA) following the ADSA-ASAS-PSA Guide for Care and Use of Agricultural Animals in Research and Teaching (Ag Guide, 2020). A total of 63 eight-month-old uncastrated male Spanish goats (average weight 29.4 ± 0.59 kg) were used for this experiment. Animals were purchased from a producer when they were 5 mo old and were allowed to graze on natural vegetation in a 10-acre pecan orchard for 3 mo. Goats were also fed a grain supplement and were given access to ad libitum hay and water. All animals were dewormed 2 wk prior to the isolation experiment. Three days prior to the beginning of the isolation experiment, all animals were weighed and blood sampled (pretrial sampling).

Isolation TRTs

Isolation testing was conducted over a period of 2 wk. The isolation pens were set up approximately 25 m from a single pen within a fenced, tarp-covered enclosure in the pecan orchard where all the goats were kept together on the days of the experiment before assigning them to TRT pens. During week 1, goats were randomly assigned to one of four isolation TRT pens (1.5 × 1.5 m) with metal grill panels: 1) open grill with no visual contact with conspecifics (IO), 2) covered grill (panels covered using blue-colored tarp sheets) to prevent visual contact with conspecifics (IC), 3) covered grill with a 30 × 30 cm window to allow visual contact with conspecifics (IW), or 4) open grill to allow visual contact with conspecifics in the adjacent pen (IV), for 90 min of social isolation. Goats from the holding area with no isolation TRT were time sampled as controls (CO). A separate open grill pen with three goats was placed such that the IV goats could maintain visual contact (VC). The VC goats were not used for data collection. Goats were able to maintain acoustic and olfactory contacts throughout the isolation period.

TRTs were applied to two sets of animals (five goats per set) simultaneously with pens arranged as shown in Figure 1. Five goats in each set were assigned to four different isolation TRTs and one CO (with no isolation TRT). TRTs were applied in this manner to 12 different sets of animals, all coming from the same fenced enclosure and the trial was conducted on two consecutive days in each week using a total of 60 goats (n = 12 goats per TRT). For each TRT, goats were randomly selected from the same fenced enclosure; therefore, individual animal was the experimental unit. Blood samples were collected at 0 (immediately after placing the goat in the pen), 30, 60, and 90 min (Time). Each animal was led into the isolation pen by an animal handler and was immediately sampled by a trained individual. Handling and leading an animal from the fenced enclosure into the respective isolation pen took about 60 s, but not exceeding 90 s. Blood sampling took only a few seconds (<30 s). Since the entire process was completed in less than 2 min for each goat, TRT would not have been applied at 0 min sampling. Time samplings at 30, 60, and 90 min involved just entering the pen and sampling (<30 s).

Figure 1.

Allotment of goats to TRTs and arrangement of isolation TRT pens.

To evaluate the effects of prior exposure to isolation (PEI), TRTs were applied to goats the first time (first exposure, FE) during week 1 as described above and the TRTs were applied to the same goats for the second time (previously exposed, PE) during week 2. The trial was repeated ensuring that the same TRTs were applied to the same goats to determine the PE effect on physiological stress responses. For example, if an animal was exposed to IC TRT during week 1, that animal was exposed to IC TRT in week 2 also, such that each goat was subjected to the same isolation TRT twice. In addition, sets of animals were maintained the same for both weeks in order to exactly repeat the trial.

Behavioral observations

During isolation TRTs, behavioral observations were conducted by four trained observers. Two observers recorded vocalization frequency continually during the isolation period and two observers recorded the frequencies of visual contact (only in IV and IW pens), standing (ST), lying (LY), and climbing (CL) behaviors at 5-min intervals throughout the isolation period. In addition, the spatial location of subjects in the isolation pens were also marked on observation charts with pre-drawn pen configurations. All four observers were positioned about 5 m from the isolation pens and the two observers who recorded the behaviors stepped forward deliberately and gingerly every 5 min to look into every pen in a sequence from above the panels without agitating the subjects. Two observers were assigned to one TRT set of four pens. The frequency of visual contact was recorded if the subject was looking at the conspecific(s) in the adjacent pen, regardless of whether or not there was reciprocation. CL behavior was recorded if the subject placed the forelimbs on a side panel in an attempt to climb over the panel. The spatial location of subjects recorded were 1) located close to and facing the corner of the pen (FC), 2) located close to and facing a side panel (FS), and 3) located in the middle of the pen regardless of the direction the subject was facing (MI). The frequencies of vocalization and other behaviors were grouped into the first 30-min period (P1), second 30-min period (P2), and third 30-min period (P3) to study the effect of isolation duration.

Blood sampling and analysis

Blood samples were collected by trained personnel as quickly as possible after the goats were caught in order to avoid confounding of the effect of blood sampling. All efforts were made not to agitate the goats, including avoiding loud noise and rough handling and minimizing metal sounds while opening and closing pen gates, to minimize the effect of the order of blood sampling. The order of placing goats in isolation pens and sequence of blood sampling were reversed for every two sets. Blood samples were collected by jugular venipuncture into disposable vacutainer tubes containing 81 µL of 15% ethylenediaminetetraacetic acid solution. The blood tubes were placed on ice until blood smears were made and plasma separated. The samples were centrifuged at 1,000 × g for 20 min, plasma pipetted into different aliquots, and then stored at −80 °C until analysis. Each animal was sampled four times (two from each side) during the 90-min TRT period with approximately 5 mL of blood collected each time.

Blood hormones and metabolites

Plasma cortisol concentrations were determined using the Cortisol ELISA Kit (Abnova, Taipei, Taiwan) according to the manufacturer’s instructions. Briefly, goat plasma samples (25 μL) were used to coat wells of 96-well microplates, and the cortisol enzyme conjugate solution (100 μL per sample) was added. The plates were incubated for 1 h at 37 °C, washed four times with the wash solution, and then the 3,3′,5,5′-tetramethylbenzidine (TMB) color reagent (100 μL) was added to each well. The reaction was stopped by adding 50 μL of the stop solution and then optical density was measured at a wavelength of 450 nm with a microwell reader (Synergy HTX Microplate Reader, Bio-Tek, Winooski, VT). The cortisol levels were read against a standard curve created using the standards provided by the manufacturer and according to the manufacturer’s instructions. Plasma epinephrine and norepinephrine concentrations were determined using the Epinephrine/Norepinephrine ELISA Kit (Abnova, Taipei, Taiwan). Epinephrine and norepinephrine were extracted using a cis-diol-specific affinity gel, acylated, and then converted enzymatically. The antigen was bound to the solid phase of the microtiter plate. The derivatized standards, controls and samples, and the solid phase bound analytes were allowed to compete for a fixed number of antibody-binding sites. After the system attained equilibrium, free antigen and free antigen–antibody complexes were removed by washing. The antibodies bound to the solid phase were detected by an anti-rabbit immunoglobulin G-peroxidase conjugate using TMB as a substrate. The reaction was monitored at 450 nm. Quantification of unknown samples was achieved by comparing their absorbances with a standard curve prepared with known standard concentrations. Epinephrine and norepinephrine concentrations were determined following the stepwise procedures, based on the above principle, provided by the manufacturer and using the reagents and plates provided by the manufacturer. The microtiter plates were read for absorbance values using the Synergy HTX Microplate Reader (Bio-Tek, Winooski, VT).

Glucose levels were determined using the Stanbio Glucose Liqui-UV (Hexokinase) Kit (Stanbio Laboratory, Boerne, TX). The principle for this procedure is based on the reactions: 1) glucose and adenosine triphosphate, catalyzed by hexokinase, form glucose-6-phosphate and adenosine diphosphate and 2) glucose-6-phosphate, in the presence of nicotinamide adenine dinucleotide, is then oxidized by glucose-6-phosphate dehydrogenase to form 6-phosphogluconate and nicotinamide adenine dinucleotide (NADH). The increase in NADH concentration is directly proportionate to the glucose concentration and is measured spectrometrically at 340 nm. The assay was performed according to the procedure outlined by the manufacturer and absorbances were determined using the Synergy HTX Microplate Reader (Bio-Tek, Winooski, VT). Plasma nonesterified fatty acid (NEFA) concentrations were determined by colorimetric assay with 96-well microtiter plates using the NEFA-HR (2) Kit (Fujifilm, Mountain View, CA) according to the manufacturer’s instructions. Briefly, plasma samples (5 μL) were added to the wells followed by 200 μL of color reagent A solution. Plates were then incubated for 5 min at 37 °C, and the first optical density measurement was done at a wavelength of 550 nm using a microplate reader (Synergy HTX Microplate Reader, Bio-Tek, Winooski, VT). Then, 100 μL of color reagent B solution was added to each well and the second measurement of optical density at 550 nm was performed. The difference between the optical density readings was used to estimate NEFA concentrations in each sample by reading against a standard curve created using the standards provided by manufacturer and according to manufacturer’s instructions.

Differential leukocyte counts

For studying differential leukocyte profiles on 90 min samples, two blood smears were made from each sample on microscope slide prior to the separation of plasma. The blood smears were dried at room temperature and manually stained with Wright’s–Giemsa solution. The different types of leukocytes were identified under the microscope, using a 100/1.25 oil immersion objective, and a total of 100 cells were counted per slide using the straight-edge method described by Schalm et al. (1971).

Statistical analysis

The behavior data were analyzed using Friedman’s two-way analysis of variance (ANOVA) by Ranks test in SAS. This nonparametric test, performed on rank score, was used since the behavior data did not meet the requirements for parametric analysis. The main effects of four isolation TRTs (IC, IO, IV, and IW) and three time periods (P1, P2, and P3) were tested for vocalization frequency, and two isolation TRTs (IV and IW) and three time periods were tested for visual contact frequency. When significant by Friedman’s test at P < 0.05, pairwise two-sided multiple comparison analysis was conducted using Dwass, Steel, Critchlow–Fligner method (Wilcoxon Z test) in SAS. However, original frequencies of behaviors or median frequencies with interquartile ranges were presented instead of rank scores for ease of interpretation.

Blood hormone and metabolite data were first examined for normality and homogeneity of variance by plotting residual vs. predicted values, Levene’s test, and Shapiro–Wilk’s test. Repeated measures analysis was then conducted using general linear models (GLM) procedures in SAS. For cortisol, epinephrine, and NEFA analysis, the respective pretrial (3 d prior to isolation experiment) concentrations were used as a covariate. The sphericity test was performed to check if orthogonal components were uncorrelated and had equal variances. When appropriate, the Greenhouse–Geisser test probability value was taken into account for within-subject effects. All plasma hormone and metabolite data required log transformation to meet the assumptions of ANOVA; however, the data were back-transformed to original scale before presenting. As differential leukocyte counts were determined only at 90-min with no repeated time sampling, the data were analyzed as a split-unit design using the GLM procedures in SAS with PEI as the whole-unit factor and TRT as the split-unit factor with error terms specified. When significant by ANOVA at P < 0.05, the means were compared using the pdiff procedure. Because the levels of time represented a quantitative factor, a separate repeated measures analysis was conducted using polynomial contrasts to assess the trends of cortisol, epinephrine, norepinephrine, glucose, and NEFA concentrations and presented to discern the time trends due to both PEI and TRT.

Results

The median frequency of vocalization (rank score) in goats was high in the IO group, low in the IV and IW groups, and intermediate in the IC group (Figure 2; P < 0.01). The vocalization rank score was also higher (P < 0.01) during P1 compared with P2 and P3. The median frequency of visual contact was higher in the IW group than in the IV group (P < 0.01), although visual contact was not influenced by isolation duration (Figure 3). The median frequency of ST was highest in the IO group, lowest in the IC group, and intermediate in the IW and IV groups (Table 1; P < 0.01). The frequency of LY was higher (P < 0.01) in the IC and IV groups compared with the IW and IO groups. The frequency of CL was high in the IC and IO groups, low in the IV group, and intermediate in the IW group (P < 0.01). With regard to the spatial location of goats (Table 1) in each isolation TRT pen, the frequency of FC was higher (P < 0.01) in IC group compared with the other TRT groups. The median frequency of FS was highest in the IW group and lowest in the IC group (P < 0.01). The median frequency of MI was high in the IO group, low in the IC group, and intermediate in the IV and IW groups (P < 0.01).

Figure 2.

Distribution of frequency of vocalization in goats subjected to isolation TRTs grouped by 30-min isolation time. ^a–cRank scores for isolation TRTs with different letters differ significantly by Dwass, Steel, Critchlow–Fligner test at P < 0.05. ***Rank score for P1 was significantly different from that of P2 and P3 by Dwass, Steel, Critchlow–Fligner test at P < 0.05.

Figure 3.

Distribution of frequency of visual contact in goats subjected to isolation TRTs grouped by 30-min isolation time. ^a,bRank scores for isolation TRTs with different letters differ significantly by Dwass, Steel, Critchlow–Fligner test at P < 0.05.

Table 1.

Effects of different social isolation TRTs and isolation time on the frequencies of behaviors and spatial location within pen and body orientation in Spanish goats

		Isolation time
		P1		P2		P3			Friedman’s test1P-value
Item	Isolation TRT	Median	IQR2	Median	IQR	Median	IQR	N	TRT	Time
Behavior
ST	IC	4.0	4.0	4.5	3.5	5.0	3.0	24	<0.001	0.706
	IO	6.0	1.5	6.0	0.0	6.0	1.0	24
	IV	5.0	2.5	6.0	3.0	5.0	2.0	24
	IW	6.0	3.0	6.0	1.0	6.0	2.0	24
LY	IC	0.0	1.5	0.5	3.5	1.0	3.5	24	<0.001	0.180
	IO	0.0	0.0	0.0	0.0	0.0	0.0	24
	IV	0.0	1.5	0.0	2.5	0.0	2.0	24
	IW	0.0	0.0	0.0	0.0	0.0	0.0	24
CL	IC	0.0	2.0	0.0	1.0	0.0	0.0	24	<0.001	0.182
	IO	0.0	0.0	0.0	0.0	0.0	0.0	24
	IV	0.0	0.0	0.0	0.0	0.0	0.0	24
	IW	0.0	0.0	0.0	0.0	0.0	0.0	24
Spatial location and body orientation
FC	IC	2.0	2.0	2.5	3.0	1.5	3.0	24	<0.001	0.525
	IO	1.0	2.0	0.0	2.0	1.0	2.0	24
	IV	0.5	2.0	0.0	1.5	1.0	2.0	24
	IW	0.5	1.5	0.0	1.5	0.0	0.5	24
FS	IC	3.0	2.5	2.0	2.5	3.0	3.5	24	<0.001	0.476
	IO	3.0	2.5	3.0	2.5	3.0	3.5	24
	IV	4.0	2.0	4.0	3.0	3.5	3.0	24
	IW	4.0	3.0	4.0	2.5	4.5	2.5	24
MI	IC	0.0	0.5	0.0	1.0	0.0	1.0	24	0.003	0.307
	IO	1.0	1.5	1.0	2.5	1.0	1.5	24
	IV	0.0	1.0	0.5	2.0	1.0	1.0	24
	IW	0.0	1.0	0.0	1.0	0.0	1.0	24

¹Friedman’s two-way ANOVA by Ranks test (nonparametric).

²IQR = interquartile range (Q3 − Q1, where Q1, first quartile; Q3, third quartile).

There was a trend toward a TRT effect on plasma cortisol concentrations (P < 0.1; Figure 4A). Overall cortisol concentrations were higher (P < 0.05) at 0 min (15.5 ng/mL) and 90 min (14.5 ng/mL) compared with 30 min (12.1 ng/mL) and 60 min (12.2 ng/mL; SEM = 1.04). Application of polynomial contrasts showed that the lines describing the TRT × Time relationship followed a significant quadratic trend (Figure 4B; P < 0.01). Plasma epinephrine concentrations were significantly influenced by Time (P < 0.05), and there was a significant cubic trend (P < 0.05) in the TRT × Time relationship (Figure 5A and B). Norepinephrine concentrations (Figure 6A) decreased (P < 0.05) with isolation time from 22.4 ± 1.11 ng/mL at 0 min to 17.4 ± 1.11 ng/mL at 90 min and followed a linear trend (Figure 6B; P < 0.05) with the slopes not differing due to PEI or TRT. There were significant effects of PEI (P < 0.01) and Time (P < 0.01) and a trend effect of TRT (P < 0.1) on plasma glucose concentrations (Figure 7A). Mean glucose concentrations were 91.2, 102.3, 104.7, 95.5, and 100.0 (SEM = 1.02) mg/dL in CO, IC, IO, IV, and IW TRT groups, respectively. Glucose concentrations gradually increased over isolation time and peaked at 60 min before decreasing again. The lines describing the TRT × Time relationship followed a significant quadratic trend (Figure 7B; P < 0.01), and the patterns were different between FE and PE sampling. Plasma NEFA concentrations were influenced by TRT (P < 0.05), Time (P < 0.01), and TRT × Time (Figure 8A; P < 0.01), with significant linear trend (P < 0.01) and the slope differing among TRTs (Figure 8B; P < 0.01). Overall, the epinephrine, norepinephrine, glucose, and NEFA concentrations were lower (P < 0.01) in PE compared with FE, while cortisol concentrations were higher (P < 0.01) in PE than in FE sampling.

Figure 4.

Effects of isolation TRT and Time (P < 0.05) on plasma cortisol concentrations in goats (A). ^a,bTime means within a cluster with different letters differ significantly (P < 0.05) by least significant difference (LSD) test. Polynomial contrasts showing a significant quadratic trend (P < 0.01) in cortisol concentrations over time in goats during FE to isolation and during subsequent isolation in PE goats (B).

Figure 5.

Effects of isolation TRT and Time (P < 0.05) on plasma epinephrine concentrations in goats (A). Polynomial contrasts showing a significant cubic trend (P < 0.05) in epinephrine concentrations over time in goats during FE to isolation and during subsequent isolation in PE goats (B).

Figure 6.

Effects of isolation TRT and Time (P < 0.05) on plasma norepinephrine concentrations in goats (A). ^a,bTime means within a cluster with different letters differ significantly (P < 0.05) by LSD test. Polynomial contrasts showing a significant linear trend (P < 0.05) in norepinephrine concentrations over time in goats during FE to isolation and during subsequent isolation in PE goats (B).

Figure 7.

Effects of isolation TRT (P < 0.05) and Time (P < 0.01) on plasma glucose concentrations in goats (A). ^a,bTime means within a cluster with different letters differ significantly (P < 0.05) by LSD test. Polynomial contrasts showing a significant quadratic trend (P < 0.01) in glucose concentrations over time in goats during FE to isolation and during subsequent isolation in PE goats (B).

Figure 8.

Effects of isolation TRT (P < 0.05), Time (P < 0.01), and TRT × Time (P < 0.01) on plasma NEFA concentrations in goats (A). ^a,bTime means within a cluster with different letters differ significantly (P < 0.05) by LSD test. Polynomial contrasts showing a significant linear trend (P < 0.01) in NEFA concentrations over time in goats during FE to isolation and during subsequent isolation in PE goats (B).

Isolation TRT did not have significant effects on neutrophil (N), basophil, eosinophil, lymphocyte (L), and monocyte counts, or N/L ratio; however, PEI had significant effects on basophil and L counts (Table 2; P < 0.01). The L counts were higher in PE compared with FE in the IC group, although a similar trend was noticed in all TRT groups. Basophil counts were higher in PE compared with FE sampling in the IV group.

Table 2.

Effects of different social isolation TRTs and PEI on differential leukocyte counts (%) in Spanish goats

		Isolation TRT							ANOVA P-value
Item	PEI	CO	IC	IO	IV	IW	SEM	N	PEI	TRT	PEI × TRT
N	FE	46.3	43.8	43.4	38.2	43.6	3.48	12	0.689	0.762	0.227
	PE	41.5	38.3	46.8	43.6	41.3	3.48	12
Basophil	FE	1.0	1.1	1.0	1.0b	1.0	0.09	12	0.001	0.559	0.409
	PE	1.2	1.3	1.3	1.5a	1.3	0.09	12
Eosinophil	FE	1.8	1.5	2.0	1.4	1.8	0.19	12	0.522	0.731	0.276
	PE	1.7	1.8	1.7	1.8	1.5	0.19	12
L	FE	56.1	53.2b	55.9	58.0	57.1	2.93	12	0.001	0.732	0.770
	PE	61.8	62.9a	58.6	61.1	60.1	2.93	12
Monocyte	FE	1.8	2.0	2.1	2.0	1.6	0.20	12	0.097	0.929	0.102
	PE	1.8	1.3	1.7	1.8	1.9	0.20	12
N/L ratio	FE	1.1	0.9	0.9	0.7	0.9	0.16	12	0.268	0.773	0.404
	PE	0.8	0.6	0.9	0.7	0.7	0.16	12

^a,bMeans within a leukocyte type with different superscripts differ significantly (P < 0.05).

Discussion

Goats are highly social in nature, and by remaining as a part of a herd, they get better security against predators and access to shelter, feed, and water (Dehn, 1990; O′Bryan et al., 2019). Several mammalian species, including goats (Carbonaro et al., 1992), vocalize when they are separated from their social groups that help them coordinate their behaviors. The frequency of vocalization was the highest in goats isolated in an open pen without visual contact with conspecifics and lowest in goats isolated with visual contact with other goats, although vocalization types were not categorized in the present experiment. Goats produce different types of vocalization that have been reported to provide the listening conspecifics a variety of information about the caller, including positive and negative emotions (Baciadonna et al., 2019); however, contact calls are their primary type of vocalization (Briefer and McElligott, 2012). Contact calls produced by goats when they are partially or fully socially isolated help them reestablish contact with their herds resulting in contraction of dispersed herds (Siebert et al., 2011; Briefer and McElligott, 2012; O′Bryan et al., 2019). Animals have the ability to differentiate delicate changes in contact calls based on the emotional valence felt by the call producer (Baciadonna et al., 2019). The frequency of vocalization was also higher during the first 30 min of isolation compared with the second and third 30-min periods that could indicate that fear and distress peaked immediately after exposure to social isolation and novel surrounding and decreased over time in isolation. Tölü et al. (2017) reported that novel environment during social isolation increases vocalization in goat kids.

The vocal cues from a distressed goat can be detected by its conspecifics and can affect its physiology and behavior (Baciadonna et al., 2019). Visual scanning of location and body orientation data showed that the higher frequency of vocalization in the IO group did not influence the behavior of the listening conspecifics in adjacent pens. Adult goats likely use cross-modal (visual and auditory) recognition of social partners and the caller is recognized by the receiver by looking in the direction of the source of the sound (Pitcher et al., 2017). In the present experiment, there is no evidence that vocalization by one subject during isolation testing influenced the behavior of conspecifics in adjacent pens as they did not spend time looking in the direction of the sound (IO pen). Furthermore, maintaining visual contact with conspecifics appeared to be more important than responding to contact call. Baldock and Sibly (1990) reported that visual isolation is more stressful than spatial isolation in sheep.

The frequency of visual contact was higher in goats allowed to maintain visual contact through a window compared with those that had free access to visual contact and the frequencies did not change over isolation time. This result, combined with the location and body orientation data (FS), showed that the IW goats spent most of their time during isolation trying to maintain visual contact with a conspecific in the adjacent pen. Carbajal and Orihuela (2001) observed that a single conspecific was sufficient to offset the distress due to social isolation and that additional companions did not yield any added advantage in sheep. These authors found that a 20-min isolation increased distress as indicated by vocalization, but not when two ewes were isolated and kept together. Limited visual access seems to be a strong motivation for goats to work continually to maintain visual contact. In addition, goats that are able to maintain visual contact with at least one group-mate tended to vocalize less than those that cannot see their conspecifics.

Acoustic and olfactory contacts were allowed among subjects during testing in order to focus on the effects of visual contact alone. Siebert et al. (2011) reported that in an attempt to reestablish social contact, rearing and jumping behavior rates were higher in goats isolated with acoustic and olfactory contacts maintained with conspecifics compared with those completely isolated. If goats are able to receive sensory feedback from conspecifics, though visually isolated, this can influence the behavior of goats. In the current experiment, the frequency of CL was higher in goats isolated in the IC pen compared with other TRT in an attempt to escape the pen and reestablish contact with conspecifics. Motivation to avoid a stressor or learning to escape is a characteristic sign of distress (Goldstein, 2010). The IC goats also spent more time lying down, a sign that they became fatigued sooner due to repeated attempts to escape the pen. The reason for this group spending more time facing the corner compared with the other TRT groups may indicate fearfulness due to a complete lack of visibility of surroundings. Although Carbonaro et al. (1992) observed other behaviors such as sniffing, trotting, and rearing in isolated goats, subjects seldom moved in the present experiment, which can be attributed to the smaller size of the isolation pens used.

The increase in plasma cortisol concentrations in response to social isolation in small ruminants is well documented (Parrott et al., 1988; Coppinger et al., 1991; Minton et al., 1992; Carbajal and Orihuela, 2001). Social isolation for 15 min can increase plasma cortisol concentrations in goats, and the stress level could further increase if isolation is combined with other stressors, such as feed deprivation (Kannan et al., 2002). In the present study, cortisol concentrations tended to be higher in goats subjected to social isolation with no visual contact compared with those that could maintain visual contact with conspecifics (IV and IW). In the latter groups, the cortisol concentrations were as low as that in the C group. This is consistent with our earlier finding that stress levels based on cortisol concentrations in goats isolated with visual contact with pen-mates were as low as those in non-isolated controls (Kannan et al., 2002). In sheep, the presence of a mirror during isolation is reported to be helpful in reducing the physiological stress responses to some extent, suggesting that visual contact with a conspecific may reduce distress (Parrott, 1990); however, such data are not available for goats to our knowledge.

The overall cortisol concentrations in the current experiment were high immediately after placing the goats in isolation pens, decreased and remained at a low level at 30 and 60 min, before increasing again at 90 min of isolation. This trend was consistent in all TRT groups both during FE and PE sampling. The unavoidable procedure of repeated entry of animal handlers into the fenced enclosure to catch animals and lead them to their respective isolation pens could have contributed to the initial increase in hormone concentration to some extent. Parrott et al. (1988) reported a biphasic cortisol response to isolation in sheep. It is not known, however, how the trends in cortisol would have been beyond the 90-min period, since the objective of this study was to determine the effects of short-term isolation typically encountered in commercial situations. Cortisol responses in goats were not attenuated by PEI in the present experiment. In fact, when the isolation TRTs were imposed again the following week, the cortisol concentrations were higher than when these animals were subjected to isolation the first time. Siebert et al. (2011) observed that goats do not get habituated to repeated isolation. Similarly, Niezgoda et al. (1987) found that sheep do not adapt to social isolation readily, as repeated imposition of isolation TRTs elicited greater increase in cortisol concentrations.

Isolation of lambs can activate both the sympathetic–adrenomedullary system as well as the pituitary–adrenocortical axis, as indicated by elevation in circulating epinephrine and cortisol concentrations (Apple et al., 1993, 1995). In the present experiment, plasma epinephrine concentrations tended to be the highest in the IO goats and lowest in the CO goats. Epinephrine concentrations were lower in PE compared with FE sampling. Epinephrine has a half-life of a few minutes and it is quickly broken down into products such as metanephrine and normetanephrine that appear to be more stable in circulation and, therefore, may be better measures of stress in goats (Batchu, 2020). Overall plasma norepinephrine concentrations were high immediately after placing the goats in isolation pens and decreased with isolation time, a pattern different from the cubic trend seen in epinephrine concentrations. Social isolation has been reported to cause a transient elevation in norepinephrine concentration in sheep (Parrott, 1990). Carbonaro et al. (1992) observed an increase in plasma norepinephrine concentration in response to isolation in dairy goats; however, they did not see any increase in plasma epinephrine or cortisol concentrations. The lack of agreement in the epinephrine and norepinephrine concentrations in response to stress is probably due to the fact that sympathetic nervous system and peripheral catecholamine systems may be governed by separate regulatory mechanisms (Kaufman et al., 1989).

Overall glucose concentrations gradually increased over Time and peaked at 60 min before decreasing again. This is in accordance with our previous reports that increase in plasma cortisol concentration is followed by an increase in glucose concentration in goats in response to a stressor (Kannan et al., 2000). Consistent with vocalization responses, the IO goats had the highest mean glucose concentrations. A similar trend was noticed in an earlier study when goats were isolated with no visual contact with pen-mates for 15 min (Kannan et al., 2002). Blood glucose concentrations have been reported to increase with increasing durations of restraint and isolation stress TRT in sheep (Apple et al., 1995). Plasma NEFA concentrations increased in isolated groups and peaked at 90 min of isolation, more prominently in the IO and IW groups. Pierzchała et al. (1985) observed a dramatic increase in circulating free fatty acids in sheep isolated from its flock. Apple et al. (1995) reported that free fatty acids increased and peaked at 1 h after isolation stress TRT in sheep (Apple et al., 1995). Since both plasma glucose and NEFA concentrations are induced by epinephrine release, these concentrations were lower when goats were subjected to isolation TRTs the second time compared with when they were exposed to isolation the first time.

Corticosteroids can alter the leukocyte profile; however, transient elevations in cortisol concentrations in response to acute stressors that are not intense or prolonged enough may not be sufficient to affect leukocyte profile in small ruminants (Minton and Blecha, 1990). Deitch and Bridges (1987) found in an in vitro study that cortisol did not affect N function, but suppressed L blastogenesis. Niezgoda et al. (1987) opined that repeated increases in serum cortisol may be essential to influence L function. Although isolation causes more intense emotional stress than handling and restraint (Parrott, 1990), there was no significant change in L counts due to isolation TRT; however, L counts were lower in FE compared with PE sampling in the IC group. Likely, there would have been more significant changes in L and N counts had the isolation TRTs lasted longer than 90 min. Basophils are a type of granulocytes and they play an important role in immune function and thus the counts can increase due to inflammation (Siracusa et al., 2013). The reason for higher basophil counts in the PE sampling is not clear in this experiment.

In conclusion, based on the behavioral and physiological changes seen in response to different social isolation TRTs in this study, lack of visual contact causes more distress than isolation with visual access to conspecifics. Goats appear to undergo a more intense level of emotional stress when they are kept in an open pen, yet not able to see any of the herd-mates than when they are kept in a closed isolation pen as indicated by the frequency of vocalization. This is also supported by physiological responses such as cortisol, epinephrine, glucose, and NEFA as the IO group had the highest overall concentrations. Distress was highest immediately after placing the goats in isolation pens due to the novelty of the environment as indicated by higher frequency of vocalization that decreased over isolation duration. Imposing isolation TRTs on goats previously exposed to isolation stress increased cortisol concentrations but decreased catecholamine, glucose, and NEFA concentrations, which may indicate that the emotional component of overall stress due to isolation is likely attenuated due to prior exposure. Short-term isolation stress for 90 min is not intense enough to cause changes in the leukocyte profile and possible negative effect on immune function. The results of this study showed that maintaining visual contact is very important in goats when socially isolated, such that they continually work to maintain visual contact even by trying to escape the pen. In practical situations where isolation of goats is essential, keeping the isolation duration to a minimum and allowing visual contact with herd- or pen-mates will greatly reduce their stress levels.

Glossary

Abbreviations

CL

climbing

CO

no isolation treatment

FC

spatial location close to and facing the corner of the pen

FE

first time exposure to social isolation

FS

spatial location close to and facing a side panel

GLM

general linear models

IC

covered grill to prevent visual contact with conspecifics

IO

open grill with no visual contact with conspecifics

IQR

interquartile range

IV

open grill to allow visual contact with conspecifics in the adjacent pen

IW

covered grill with a 30 × 30 cm window to allow visual contact with conspecifics

L

lymphocyte

LY

lying

MI

spatial location in the middle of the pen regardless of the direction the subject was facing

N

neutrophil

N/L ratio

neutrophil:lymphocyte ratio

NEFA

nonesterified fatty acid

P1

first 30-min period

P2

second 30-min period

P3

third 30-min period

PE

previously exposed to social isolation

PEI

prior exposure to isolation

Q1

first quartile

Q3

third quartile

ST

standing

TMB

3,3′,5,5′-tetramethylbenzidine

TRT

treatment

VC

an open paneled pen with three goats was placed such that the IV goats could maintain visual contact

Conflict of interest statement

The authors assure that there are no perceived or actual conflicts of interest that affect their ability to objectively present this research.