Behavioral and physiological responses of dairy goats to food thwarting

Behavioral and physiological responses of dairy goats to food thwarting

Physiology& Behavior,Vol. 51, pp. 303-308. © Pergamon Press plc, 1992. Printed in the U.S.A. 0031-9384/92 $5.00 + .00 Behavioral and Physiological R...

593KB Sizes 0 Downloads 69 Views

Physiology& Behavior,Vol. 51, pp. 303-308. © Pergamon Press plc, 1992. Printed in the U.S.A.

0031-9384/92 $5.00 + .00

Behavioral and Physiological Responses of Dairy Goats to Food Thwarting I D. A. CARBONARO, 2 T. H. FRIEND, G. R. D E L L M E I E R

Department of Animal Science, Texas A & M University, College Station, TX 77843 AND L. C. N U T I

International Dairy Goat Center, Texas A & M at Prairie View, Prairie View, TX 77446 R e c e i v e d 11 J u n e 1990 CARBONARO, D. A., T. H. FRIEND, G. R. DELLMEIER AND L. C. NUTI. Behavioral and physiological responses of dairy goats to food thwarting. PHYSIOL BEHAV 51(2) 303-308, 1992.--Eight Nubian dairy goat does in one experiment, and eight Alpine dairy goat does in a second experiment, were randomly allotted to food-thwarted or fed groups in a crossover experimental design. Food thwarting was hypothesized to produce an emotional state analogous to frustration. After a 1-week training period during which the goats of both breeds were conditioned to being simultaneously fed in adjacent feeding stalls, frustration was induced in half the goats by feeding only alternate does. Focal animal behavior was recorded for the initial five min after feeding frustration commenced. Blood samples were collected via a jugular cannula before, during and after frustration was induced for thyroxine ('1"4), triiodothyronine (T3), cortisol, norepinephrine (NOR), and epinephrine (EPI) determinations. Food thwarting was characterized by increased plasma concentrations of NOR, and increased incidences of pawing, head movements, mouthing of objects, behaviors directed toward neighboring does being fed, and rearing (p<0.01). When data were pooled across experiments, breed had a strong influence on cortisol (p<0.05), with Nubian does having higher concentrations regardless of treatment. Concentrations of hormones were not significantly correlated with behaviors. These findings suggest that frustration may elicit a discharge of NOR but not EPI. Goats





THE psychological components of environmental stimuli are powerful activators of endocrine responses that alter both physiological and psychological homeostasis; such alterations can be of a degree that are indicative of stress (5,8). Biologically, emotions can be defined as deviations of the normal mood state of an individual at rest (5) and characterized by the physiological responses that typically accompany them. The complex hormonal changes that occur in conscious individuals subjected to stressful situations, therefore, reflect the emotions the subjects experience (2, 5, 7, 20, 28). Increasing public concern about the well-being of domestic animals utilized for human consumption and other uses has created a particular need for a scientific method of evaluating the emotional states of livestock species. Hinde (18) describes frustration as occurring when animals engaged in a sequence of behaviors are unable to complete it because there are physical and psychological obstacles to doing so in the environment. Similar obstacles are likely to be encountered in conditions of intensive animal husbandry. There has been little systematic study of frustration and other behavioral states of farm animals other than in domestic fowl (28). Physio-

logical alterations are associated with emotional states, although a direct cause and effect relationship is difficult to demonstrate. However, the first response a stressed animal usually exhibits will be behavioral, i.e., moving away from a threatening stimulus (9,21). This study characterized the changes in behavior and physiology of goats subjected to the acute stressor of an abrupt denial of anticipated food. The objective of this and a companion study (6) was to contribute toward the development of techniques and methodology needed to objectively differentiate physiologically and behaviorally among various emotional states in livestock. METHOD A food-thwarting situation was used to produce what was hypothesized to be a state similar to "frustration" in the test animals. Food-thwarting situations, in which animals previously trained to obtain food are subjected to nonreinforcement or extinction procedures, are frequently used to produce this type of behavioral state (2).

1TechnicalArticle No. 25595 from the Texas Agricultural Experiment Station. 2present address: Division of Orthopedic Surgery, Children's Hospital of Los Angeles, Los Angeles, CA 90054-0700.




Experimental Animals and Their Management Eight Nubian does were randomly assigned to either fed or food-thwarted treatments in trial 1. In the second trial the treatments were reversed. The same procedure was followed for a second experiment using eight Alpine does. The goats of each breed were housed together and allowed to form an established group in a pen with continuous access to ten 1.27 × 0 . 5 3 m feeding stalls for one week before the trials. The stalls were constructed of horizontal 2 cm square tubing approximately 20 cm apart. They were fed a maintenance diet at 0800 h and 1500 h daily and water was available ad lib. At the beginning of each feeding each goat was locked into a stall and then offered 1 kg of feed. The goats remained locked in the stalls for 50 rain to habituate them to the duration of stall confinement required for data collection. On the test day the morning feeding was omitted, and at 1500 h only every other goat was fed (fed group) in sight of her unfed neighbors (thwarted); after 25 min the food-thwarted goats were then fed. After data collection, the goats were released and the group maintained as described for an additional week, after which the crossover occurred and the same experimental treatments were applied.


Defecation: Eat: Groom:

The behavioral data were recorded 2 weeks after the collection of blood samples because a pilot study found that the thwarted does would direct some of their vocalizations towards any humans in their vicinity, as if "demanding to be f e d . " Price and Thos (23) also reported a high incidence of interactions with observers in isolation studies with sheep and goats. A 2-week readjustment period was allowed because does may have learned that they would be fed after 25 min from the previous trial, and our goal was to evoke a behavioral reaction similar to the novel situation when the blood samples were obtained. Behavioral data were recorded in real time using a video camera, stored on tape and reviewed later as focal animal samples (1). Behavioral observations commenced with the presentation of feed to the fed goats and ended 5 min later. No humans were present in the test area during the recording of behavioral data. The following week the fed group became the thwarted group and vice-versa, so that each goat was thwarted once within a 4-week interval. Due to equipment problems, useable behavioral data could not be obtained for three fed and two thwarted does. The behaviors recorded are summarized in Table 1 along with the operational definition of each behavior quantified. Pawing, defecation, urination, and head movement were quantified as count data. Mouthing, sniffing, grooming, eating, standing, rearing, kneeling, recumbancy, and behaviors directed to neighboring goats were quantified as a percentage of time of the total test period. The group of behaviors operationally defined as " m o u t h i n g " included any behavior in which there was lip and/or jaw movement. A distinction could not be made between chewing and licking 100% of the time; therefore, these behaviors were placed in one category. Pilot observations revealed that head movements were more frequent and dramatic in the thwarted animals. Head movements were defined as any unidirectional displacement of the head greater than 90 ° and/or greater than 30.3 cm (1 foot).

RE~ (;F.~)i-.i :

Elimination of feces Mastication, swallowing and/ur tummatlou ,~J i'ee~ Licking, chewing or robbing of one "~, body parts

Head movement:

Change in head rotation :'-90~ on the *, ,~r z axis and/or elevation .>-90° on the y axis


Supporting the forequarters of the body with the knees while the hindfeet are in contact with the supporting surface


Manipulation of an object with the mouth


Touching neighbor with the nose or mouth: can include mouthing, licking or biting


Forward, striking motion made with foreleg; may or may not strike supporting surface


Jumping or standing up so that the body is supported only by the hindlimbs and a vertical surface


Behavioral Data Collection

DEFINITIONS OF BEHAVIORS ...........................

Lying laterally or sternally


Supporting the body with all four legs extended, and with no forward or backward locomotion


Movement of the nose within 10 em of a surface and/or object


Elimination of urine

nulae have been shown to reduce concentrations of plasma catecholamines in rats (22), for example, due to reduced stress from handling and restraint. The cannulae consisted of 18 gauge medical grade tubing and were held in place with adhesive tape. The tubing was filled with disodium ethylenediamine tetraacetate (EDTA) (20 mg/ml), to prevent clotting between blood samples. A nonpalatable agent (Chew Guard, Summit Hill Laboratories, Avalon, NJ 08202) was applied to the wrapping to discourage nibbling and eating of the wrapping by herdmates. The animals were allowed to rest in their groups for one hour following cannulation, after which the does were placed in the feeding stalls and 10 ml blood samples were drawn prior to feeding (time 0) and at 5, 10, 15, 25, 35, 45, and 55 min after feeding. The thwarted goats were fed after collection of the 25 min sample. The blood samples were drawn by people standing in the rear of vacant feeding stalls. One person could easily draw blood samples from the cannulae of two goats without actually touching the goats. The fed goats ignored people where as the thwarted does did occasionally direct vocalizations towards people obtaining the blood samples, as if requesting to be fed. Samples were obtained with a 10 cc syringe containing l0 mg ethylenediamine tetraacetacte (EDTA) as an anticoagulant. Samples were transferred to glass tubes and maintained in an ice bath until centrifugation. The plasma was drawn off and stored in duplicate polypropylene tubes. Equal volumes of plasma and perchloric acid (0.1 M) were added to one tube for use in catecholamine assays. The perchloric acid deproteinates the plasma, allowing the sample to be stablized for storage. The samples were capped with triple layers of parafilm and frozen until analyses could be completed.

Physiological Data Collection To minimize acute stress due to handling, a jugular vein cannula was established in each goat one hour prior to testing. Can-

Hormone Determinations The catecholamine extraction procedure and high pressure



TABLE 2 LEAST-SQUARESMEANS ( ---SE) OF BEHAVIORALDATA POOLED FOR BOTH NUBIANAND ALPINEDOES* Treatment Behavior Frequency (5 min): Pawing Head Movements (>90 °) Percentage of Time (5 min): Mouthing Neighbor Recumbent Standing Kneeling Rearing Eating~:

Fed (n = 13)

Thwarted (n = 14)

0.07 --- 0.09 4.99 --- 1.42

0.39 --- 0.09t 15.05 +-- 1.38t

0.05 0.13 0.06 101.35 0.06 -1.19 95.6

--- 0.93 --- 0.26 -+ 0.07 --- 1.61t - 0.07 --- 1.37 ± 1.13

7.08 0.99 -0.01 87.30 0.05 12.55

± 0.90t __+0.25~ ___ 0.07 ± 1.57 --- 0.06 --- 1.33t --

*Negative values are an artifact of least-squares means calculation. tHigher than other value in row (/9<0.01). :~Trial (breed) nested effect (p<0.01).

liquid chromotography (HLPC) assay were adapted and modified from an application note furnished by Bioanalytical Systems Inc. (West Lafayette, IN). Reference solutions of EPI, NOR and an internal standard, 3,4 dihydroxybenzylamine (DHBA), were prepared in 0.1 M HC104. Analytical reagent-grade and HPLCgrade solvents were used without further purification, Alumina oxide (Sigma Chemical Company, St. Louis, MO) with a neutral activity grade was acid washed. Plasma catecholamines were purified and concentrated by extraction with alumina. In a conical reaction vial all of the following were added: 2 ml of deproteinated plasma, 50 mg of alumina and 0.025 ml of DHBA. For every set of twelve sampies, two vials containing 50 mg of alumina, 2.0 ml of 0.1 M phosphate buffer (pH 7.0) and 0.05 ml of 75 ng/ml NOR and 25 ng/ml EPI were used to calibrate the instrument. In order to bind the catecholamines to the alumina, 1.0 ml of 1.5 tris buffer/EDTA (pH 8.6) was added to all samples and standards. Each vial was vortexed and reciprocally shaken for 5 min. After shaking, the alumina was allowed to settle, after which the supernatant was aspirated. The alumina was washed twice with distilled water and aspirated to near dryness after each wash. A slurry made by adding 0.5 to 1.0 ml distilled water to the alumina was then transferred to a microfilter with a 0.22 mm pore size nitrocellulose membrane (Model HP/HPX, Rainin Rabbit, Woburn, MA). The contents of the microfilter were spun to dryness at 1000 × g for 1 min. A new receiver tube was placed on the microfilter and 0.2 mi of 0.1 M perchloric acid were added to the microfilter compartment to unbind the catecholamines from the alumina. The microfilter was vortexed, allowed to stand for 5 min, vortexed again, and then centrifuged at 1000 x g for 1 min. The acidic extract in the receiver tube contained the catecholamines ready for injection on the HPLC. The catecholamines were assayed by high pressure liquid chromatography with electrochemical detection (Model HP/HPX, Rainin Rabbit, Woburn, MA), using an applied potential + 650 mV. The mobile phase consisted of 0.1 M citric acid, 0.5 mM EDTA, 0.2 mM octyl sodium sulfate adjusted to a pH of 3.0, along with 3% acetylnitrile, and filtered through a 0.45 mm pore

size nylon membrane. A flow rate of 1.0 ml min-1 was mainrained with a solvent delivery system (Bioanalytical Systems, Inc.) equipped with a pulse damper. Norepinephrine and EPI concentrations were obtained by comparing peak height ratios (relative to the internal standard DHBA) to those of the synthetic standards whose original concentrations were known. Twenty percent of the samples analyzed for NOR had concentrations below the lower detectable threshold (LDT) of 100 pg/ml, and were, therefore, assigned a default value of 100 pg/ml. Fiftyfive percent of the samples analyzed for EPI had concentrations below the LDT of 80 pg/ml, and were, therefore, assigned a value of 80 pg/ml. Using the LDT rather than a lower value is a relatively conservative approach and should be considered when interpreting the NOR and EPI data. Cortisol, T 3 and T4 concentrations were determined by double antibody iodinated kits (Pantex, Santa Monica, CA). All determinations were made on duplicate samples with a lower sensitivity of 0.3 ng/dl for cortisol, 25 ng/dl for T 3 and 0.4 ~g/dl for T 4. Determined with a pooled sample, the within- and between-assay coefficients of variation were 8.7 and 9.3% for cortisol; 2.1 and 2.3% for T3; and 3.3 and 4.8% for T 4, respectively. Hormone responses to the treatments were determined by integrating the area under the curve described by the plasma levels from 0 to 25 min (during thwarting) and from 35 to 55 min (during feeding of the thwarted does) by: (H s + Hs+l)/2 × I, where H is the hormone concentration in plasma of sample s and the successive sample, s + 1, with an interval of I hours between the samples (15). When blood samples for some time points were not obtained because of momentary cannulae failure, the missing samples were estimated by averaging two averaged values. The first value was an average across all sampling times within that goat and the second value was the average of that particular variable from goats of the same breed and treatment at that particular sampling time. There were not more than two missing samples for any animal with no more than five missing samples per treatment. Due to complete cannulae failure, hormone data from three fed and two thwarted does could not be obtained.

Statistical Analysis The statistical analyses were performed on an IBM-PC with SAS statistical package for personal computers (24). The General Linear Model (GLM), with appropriate error sums of squares was used to test for the main effects of treatment, breed, and their interaction, as well as for the nested effects of breed, trial, and goat. Two analyses were conducted. In one, the baseline plasma concentrations for each hormone (time 0) was used as a covariate. In the other analysis, no covariate was used. The only difference between the analyses was that the use of the covariate increased the level of significance by one or two units for the analyses in which the covariate was significant. Therefore, only the results from the analysis with the covariate are reported. Because a relatively high percentage of the NOR and EPI concentrations were below the lower sensitivity of the assay and the variation for those samples was artificially reduced by assigning the value of the lower detectable threshold (LDT), NOR and EPI treatment effects were determined using a t-test in which the variance was first calculated by assigning zero to all samples at the LDT. This inflated variance was then applied to a t-test in which LDT values were used to calculate the means and treatment effects. Least-squares means were used to approximate the means for a balanced design because data from some goats were not included in the analyses due to failure to establish or maintain





~,. &~,_



0--0 e--e ~--A A--&



,~I0" 8

Treatment Fed (n = 13)



Thwarted (n = 14)


Thyroxine, ixg/dl/h 0-25 min samples 35-55 min samples* Triiodothyronine, ng/ml/h 0-25 rain samples 35-55 min samples Cortisol, Ixg/dl/h 0-25 min samples*t 35-55 min samples Norepinephrine, pg/ml/h 0-25 min samples:~ 35-55 min samples Epinephrine, pg/ml/h 0-25 min samples*:~ 35-55 min samples

3.61 -4- 0.43 4.05 _+ 0.34 34.0 36.3

+-- 2.3 -+ 3.7

0.91 ± 0.09 1.04 _+ 0.28


1.19 + 0.09 1.71 _+ 0.26

- 10.7 +_ 23.9

234.9 240.6

- 11.2 # _+ 25.0

114.3 100.3

-+ 19.9 -+- 33.0

132.2 191.8

_+ 18.4 _+ 30.4














i.t ,hY2-h:

*Baseline sample was significant as a covariate (p<0.05). tBreed and trial (breed) nested effect (p<0.05) with Nubians higher across both treatments. :~Trial (breed) and goat (breed) nested effect (p<0.05). §Higher than other value in row (/3<0.05). '~Higher than other value in row and higher for Nubians across both treatments (p<0.01) using t-test with adjusted variance.

working cannulae, or problems with the video equipment.

: .... I


2.1 3.5§

164.4 233.4


~. 105

4.12 _+ 0.44 3.96 -+ 0.34 37.2 50.2








Behavioral Responses The dose subjected to the food-thwarting situation responded with an overall increase in activity. They had higher incidences of pawing and head movements and spent a greater percentage of time performing mouthing behaviors such as chewing, biting and/or licking that were directed toward their neighbors and surrounding inanimate objects (Table 2). In addition, they reared up in the feeding stalls more often than the fed goats, whereas the fed goats spent almost all of the test period (96%) standing still as they ate. There were no incidences of urination or defecation during the test period. These behavioral responses are similar to those reported for pigs required to walt for food delivery (19). The pigs vocalized frequently, became extremely active and bit the bars in their pens, and increased their performance of other activities such as ear and tail movements. It has been concluded that food thwarting was correlated with aggressive responses in poultry (11-13); however, an aggressive response was only elicited in the presence of a bird lower in the social hierarchy. Similar findings regarding aggressive responses were found to be true for pigs (2) and for goats (25). In this study, no attempt was made to determine the social hierarchy among the experimental animals. Several incidences of biting and chewing the back of an adjacent goat (who was eating) were observed, and if not restrained by the feeding stall, further aggression might have ensued. In this study, the goats were trained to expect feed during the period of time during which they were restrained in the

















i, 10




TIM[ (ml.)

FIG. I. Hormone profiles of Nubian and Alpine does subjected to food thwarting from time 0-25 min (shaded area). The thwarted does were fed immediately after the 25-min blood sample was obtained.

feeding stalls. However, food was withheld for 24 h prior to testing to increase motivation to eat. In a study using chickens (11), hens who were trained and food deprived the longest (most strongly motivated to respond) had a higher incidence of stereotyped movements that included pacing. Those that underwent the



least training and food deprivation and were, therefore, not as motivated, responded with an increase in preening. In this study, there were isolated incidences of mouthing of the feed bucket by the frustrated does. The feed bucket was also repeatedly pulled out of its holder and banged against the feeding stall walls. There were very few incidences of self-grooming behavior. However, attention to the neighboring (eating) goat was increased significantly, suggesting a high level of motivation to eat. There were no breed effects on any of the behavioral data quantified (p<0.10). Similarly, no strain differences in behavioral responses were found when White Leghorn hybrid and Rhode Island Red hens were subjected to food thwarting (3). Under many current husbandry systems, agricultural animals chronically kept under intensive conditions are likely to be behaviorally thwarted or under conflict (28). Ethologists have long considered the problem of thwarting of motivations and conflict of drives, and observed that under these conditions the most common types of responses include displacement activities and redirected activities, including aggression (27). The increases in activity of the thwarted goats in this study, including behaviors directed toward a fed neighbor, support these observations.

Physiological Responses The trend for does of a particular breed to respond in a similar manner regardless of the treatment can be readily seen in Fig. 1. Considerable variation in the concentrations of certain hormones from one blood sample to the next is also evident. A trend for T 3, T 4 and NOR concentrations in the Nubian does to return to base values during the 35-55 rain samples is evident. There were no significant treatment effects for the integrated T3, T,t, cortisol, and EPI concentrations during thwarting; however, they all tended to be greater in food-thwarted goats (Table 3). Thyroid hormone release may have been inhibited by the sympathetic activity (4, 7, 17). The slight trend towards increased concentrations of cortisol in the thwarted goats concurred with an increase in plasma cortisol concentrations (26) found in cattle allowed to view feed for 15 min before presentation. However, sympathetic activity was predominant in this study, with little or no adrenomedullary release. Nubians had a higher integrated plasma cortisol response than Alpines (1.31---0.14 txg/dl versus 0.80---0.14 v,g/dl, respectively; p < 0 . 0 5 ) , despite no differences between treatments. Baseline plasma concentrations of NOR were also higher for Nubians (p<0.05). Integrated NOR response was higher for Nubians subjected to food thwarting than thwarted Alpines (239 ---9.1 pg/ml versus 162---8.5 pg/ml, respectively; p<0.01). These dif-

ferences support the informal observations of dairy goat producers and the results of the companion study (6) that Nubians are more reactive to stimuli. Research in humans found that a physical stressor, such as running up stairs, stimulates a limited sympathetic response that results in a comparatively larger increase in NOR release than does a psychological stressor, such as public speaking, that stimulates the adrenal-medullary response and results mainly in an increase in EPI release (10). These and other findings (13) indicate that monitoring catecholamines, and possibly other neurotransmitters, may provide an effective means of objectively differentiating among various emotional states in nonhuman animals, i.e., fear, frustration, calm, etc. Although all of the goats were restrained in stalls, those who were food thwarted performed more physical activity than those who were fed, and this could have caused the increase in NOR. Thus, the physical activity of the goats was still confounded with the psychological component of food thwarting, something Dimsdale and Moss (10) were able to eliminate in their studies with humans by using public speaking as the psychological stressor.

Behavioral and Physiological Correlations Plasma concentrations of the hormones measured were not significantly correlated with any of the behavioral data. The lack of significant correlations was not expected and may have been due partly to the 2 week separation between collection of the behavioral and physiological data. Also, the performance of physical behaviors directed toward the fed may have masked or provided some type of coping that reduced the psychological stress expected with food thwarting. CONCLUSIONS Behaviorally, food thwarting (frustration) was characterized by increased frequency (or incidences) of pawing, head movements, mouthing, behaviors directed toward neighboring animals being fed and rearing. Physiologically, food thwarting (frustration) was characterized by an increase in NOR concentration. These findings suggest that there was a sympathetic discharge of NOR in response to the witholding of expected feed. It is highly probable that little of the NOR was secreted by the adrenal gland, since the medulla contains six times as much EPI as NOR, and its contribution to circulating levels is very small (16). It is also possible that the sympathetic discharge related to physical activity inhibited thyroid secretion, as was reviewed by Christy (7). Nubians would be the preferred breed to use in future stress research because of their greater physiological responsiveness to potential stressors when compared to Alpine does.

REFERENCES 1. Altmann, J. Observational study of behavior sampling methods. Behaviour 49:227-265; 1974. 2. Arnone, M.; Dantzer, R. Does frustration induce aggression in pigs? Appl. Anim. Ethol. 6:351-362; 1980. 3. Bareham, J. R. Effects of cages and semi-intensive deep litter pens on the behaviour, adrenal response and production in two swains of laying hens. Br. Vet. J. 128:153-163; 1972. 4. Brown-Grant, K.; Pethes, G. The response of the thyroid gland of the guinea-pig to stress. J. Physiol. 151:40-50; 1960. 5. Burchfield, S. R. The stress response: A new perspective. Psychosom. Med. 41:661-672; 1979. 6. Carbonaro, D. A.; Friend, T. H.; Dellmeier, G. R.; Nuti, L. C. Behavioral and physiological responses of dairy goats to isolation. Physiol. Behav. 51:297-301; 1992. 7. Christy, N. P. Adrenal cortex; catecholamines and sympathoadrenal system; thyroid storm or crisis. In: Werner, S. C.; Ingbar, S. H.,

8. 9. 10. 11. 12. 13. 14.

eds. The thyroid, 4th ed. New York: Harper & Row; 1978:223246. Dantzer, R.; Mormede, P. Stress in farm animals: A need for reevaluation. J. Anim. Sci. 57:6-18; 1983. Dellmeier, G. R. Motivation in relation to the welfare of enclosed livestock. Appl. Anim. Behav. Sci. 22:129-138; 1989. Dimesdale, J. E.; Moss, J. Plasma catecholamines in stress and exercise. JAMA 243:340-342; 1980. Duncan, I. J. H.; Wood-Gush, D. G. M. Frustration and aggression in the domestic fowl. Anim. Behav. 19:500-504; 1971. Duncan, I. J. H.; Wood-Gush, D. G. M. Thwarting of feeding behaviour in the domestic fowl. Anim. Behav. 20:a.A.~ ~51; 1972. Duncan, I. J. H.; Wood-Gush, D. G. M. An analysis of displacement preening in the domestic fowl. Anim. Behav. 20:68-71; 1972. Frankenhaeuser, M. Experimental approaches to the study of catecholamines and emotion. In: Levi, L., ed. Emotions--Their param-


eters and measurement. New York: Raven Press; 1975:209-234. 15. Friend, T. H.; Dellmeier, G. R.; Gbur, E. E. Comparison of four methods of calf confinement. I. Physiology. J. Anita. Sci. 60:10951101; 1985. 16. Goldstein, D. S.; McCarty, R.; Polinsky, R. J.; Kopin, 1. J. Relationship between plasma norepinephrine and sympathetic neural activity. Hypertension 5:552-559; 1983. 17. Gregerman, R. I.; Davis, P. J. Effects of intrinsic and extrinsic variables on thyroid hormone economy. In: Werener, S. C.; Ingbar, S. H., eds. The thyroid, 4th ed. New York: Harper & Row; 1978: 223-246. 18. Hinde, R. A. Animal behavior, 2nd ed. New York: McGraw-Hill; 1970:420. 19. Kiley, M. A comparative study of some displays in ungulates, canids and felids, with particular reference to their causation. Ph.D. Dissertation. England: University of Sussex; 1969. 20. Mason, J. W. A reevaluation of the concept of "non-specificity" in stress theory. J. Psychiatr. Res. 8:323-333; 1971. 21. Merck. Behavior. In: Siegmund, O. H., ed. The Merck veterinary

C A R B O N A R O . F R I E N D , D E L L M E I E R ~\NI) NU'I'i

manual. Rahway, NJ: Merck and Co., Inc.: 1979:145ti. 22. Popper, C. W.; Chieuh, C. C.; Kopin, I. J. Plasma catechoiammc concentrations in unanesthesized rats during sleep, waketidness, immobilization and after decapitation. J. Pharrnac~l E x p Ther. "?(/2: 144-148; 1977. 23. Price, E. O.; Thos, J. Behavioral responses to short-term social ~solation in sheep and goats. Appl. Anim. Ethol. 6:331--339:i980 24. SAS Institute, Inc. SAS guide for personal computers Carv. NC: Statistical Analysis Institute, Inc.: 1985. 25. Scott, J. P. Dominance and the frustration-aggression hypothesis. Physiol. Zool. 21:31-39: 1948. 26. Willett, L. B.; Erb, R. E. Short term changes in plasma corticoids in dairy cattle. J. Anim. Sci. 34:103-111; 1972. 27. Wood-Gush, D. G. M. Animal welfare in modem agriculture. Br. Vet. J. 129:167-174; 1973. 28. Wood-Gush, D. G. M.; Duncan, I. J. H.; Fraser, D. Social stress and welfare problems in agricultural animals. In: Hafez, E. S. E . ed. The behaviour of domestic animals, 3rd ed. London: Bailliere Tindall; 1975:182-200.