BREEDING AND GENETICS Behavior, Production, and Weil-Being of the Laying Hen. 2. Individual Variation and Relationships of Behavior to Production and Physical Condition A. B. WEBSTER 1 and J. F. HURNK Department of Animal and Poultry Science, University of Guelph, Guelph, Ontario, Canada NIG 2W1 (Received for publication February 26, 199) ABSTRACT Variation of behavior among hens and the relationships of behavior to measures of production and physical condition were investigated. The birds were 384 pullets from the mating of two stocks of males, obtained from a commercial breeder, to females from a third flock. The birds were housed as pairs in laying cages at 22 and 2 wk of age (Hatches 1 and 2, respectively). The laying phase was divided into 28-day periods. In Periods 1,3,5,7,9, and 11, samples of hens were video recorded for 8 h. In Period 13, direct visual observations were made of the behavior of individually identified hens. Eleven production-related variables were recorded throughout the laying phase. Feather scores were assessed in Periods 3,6,9, and 12. Body weight, lesions to the feet, and claw length were recorded in Periods 6 and 13. Spearman rank correlations were calculated between video-recorded behavioral variables and measures of production and physical condition. The data from direct visual observations were used for heritability estimates of behavioral traits. Eating and standing were positively correlated with egg production, whereas sitting and, for hens derived from male parental Stock 1, resting were negatively correlated with production. Inactivity also coincided with poorer plumage condition and higher body weight. For the offspring of male parental Stock 2, behavioral actions frequently performed in stereotyped manner, e.g., cage pecking and toe pecking, were positively associated with egg production. No significant additive genetic variation for behavior was evident among sires; however, for dams, fairly large heritability estimates occurred for a number of behavioral states. The apparent absence of additive genetic variability among sires for behavior may have been due to genetic fixation at gene loci which control behavior in the stocks acquired from the commercial breeder. (Key words: laying hen, heritability, behavior, egg production, physical condition) INTRODUCTION The ability of an animal to cope with an intensive production environment is influenced greatly by its behavior, as behavioral actions are essential to the modification of circumstances and exploitation of resources. If an animal's responses are not appropriate, its needs may go unsatisfied or it may cause harm to itself or other animals. On a cognitive level, failure of the consequences of actions to meet with expectations appears to result in suffering (Toates, 1987; Wiepkema, 1987). Insofar as production and well-being both require a constructive interaction between animal and environment, animals should be behaviorally adapted to the housing provided. One could 'Present address: Atlantic Poultry Research Institute, Nova Scoria Agricultural College, P.O. Box 55, Truro, NS, Canada B2N 5E3. 421 1991 Poultry Science 7:421-428 make environmental adjustments to accommodate the behavioral propensities of an animal; however, there is a limit to how much an environment can be modified without seriously compromising the economic efficiency of a production unit. Alternatively, one could select for behavioral characteristics that promote adaptation to intensive housing environments. The consensus in the literature is that species have been modified during domestication such that domestic animals are better able to cope with production environments than their wild progenitors (Hurnik, 198; Beilharz, 1982; Craig, 1982). Siegel (1979, 1989) has reviewed the subject of behavioral genetics in chickens. Despite domestication, it is not apparent that any breed or strain of chicken is ideally suited to intensive housing. In various genetic stocks, social dominance ability (Craig et al., 1965), color preference (Hurnik et al., 1977), mating behavior (Dunnington and Siegel, 1983), open-field activity (Faure, 1981; Faure and Folmer, 1975), and prelaying Downloaded from http://ps.oxfordjournals.org/ at Penn State University (Paterno Lib) on February 18, 216
422 WEBSTER AND HURNIK behavior (Mills et ah, 1985) have been altered through selection. Moreover, selection of a behavioral trait can, in some instances, influence a production-related variable, e.g., social dominance ability and age at sexual maturity (Craig and Toth, 1969). Selection of behavior to improve production or welfare, therefore, might be feasible provided suitable behavioral traits can be found upon which to exert selection pressure. Some relationships of behavior to production and well-being have been noted in the literature; for instance, hysteria has a negative impact on production and wellbeing (Hansen, 1976), and fearfulness generally has a negative association with feathering and egg production (Craig et ah, 1983, 1984; Hemsworth and Barnett, 1989). Even so, the contributions to production or well-being of many of the most prominant behavioral actions of chickens in intensive husbandry environments are not well understood, nor are their genetic variabilities well investigated. The present study focused on the relationships of behavioral actions constituting the ethogram of hens in battery cages to measures of production and physical condition. Heritabilities of these actions were estimated. The data are from a larger study of the effects of environment and genetic stock on the productivity and well-being of laying hens (Webster, 199). MATERIALS AND METHODS Pullets were derived from the mating of two White Leghorn-type stocks of males, acquired from Shaver Poultry Breeding Farms, Cambridge, Ontario, Canada, to females from a flock kept at the University of Guelph. The birds were housed at 22 and 2 wk of age (Hatches 1 and 2, respectively) in two adjacent three-tiered, semi-stairstep batteries of cages. Each cage was 45 cm wide, 32.5 cm deep, and 37.5 cm high. There were 13 and 12 males, respectively, from the two commercial stocks. Each male was bred to two females (different females for each male). The and designate the pullets derived from male parental Stocks 1 and 2, respectively. Rearing and management in dam family groups within hatches have been described by Webster and Humik (1989). Eight pullets were randomly selected from each dam family for housing in the laying batteries, and the assignment of these individuals to cages was randomized. The pullets were housed, two per cage, within blocks of four cages in each cage battery, being randomly assigned to two adjacent cages in a block and to the other two adjacent cages. A total of 384 pullets was housed in 192 cages. The batteries were in a closed, negative pressure, fan-ventilated room. Photoperiod was 14 h light: 1 h dark. Feed and water were provided ad libitum. The laying phase was divided into 13 28-day periods. On the first 4 days of Periods 1, 3, 5, 7, 9, and 11, video recordings were made of hens in 8 cages per day. For each videotape, beginning at 9 h, 6 records of the behavior of each hen were obtained at evenly spaced intervals covering 8 h. The data were pooled for hens within cages. A series of direct visual observations was conducted in Period 13 to gain information on individual hens. Eight cages per day were observed; the study comprised 24 days of observation over a 28-day period. The behavior of each hen in each cage was observed continuously for 2 min between 83 h and 12 h. To aid individual identification, a black felt marker was used the day before a given cage was to be observed to mark one cagemate on the back of the neck and the throat and the other on the sides of the neck. Hens paid no noticeable attention to these markings during the period of observation. Occurrences and durations of all behavioral actions were recorded directly onto microcomputer disk. Eleven variables related to production were recorded throughout the laying phase. Feather scores were assessed in Periods 3, 6, 9, and 12. Body weights, lesions to the feet, and length of the claws of the center front toe and rear toe of the right foot were recorded in Periods 6 and 13. Housing, management, experimental factors, procedures of observation, and dependent variables in the laying phase have been described in detail in Webster and Hurnik (199a,b). Spearman rank correlations were calculated between video recorded behavioral variables and the measures of production and physical condition. For the direct visual observations, a nested analysis of variance was used to test the significance of the sire and dam components of variance. The model equation was as follows: Yijkim = M. + C i + Rj + G fc + s(g)ki + d(s) klm + J^Oijkim-O ) + P2(Bijklm-B ) + eijklm where Y^um = an observation of an individual hen; \i = the grand mean; Q = the cage type Downloaded from http://ps.oxfordjournals.org/ at Penn State University (Paterno Lib) on February 18, 216
CORRELATION OF HEN BEHAVIOR AND PRODUCTION 423 effect (i = roost, conventional); Rj = the relationship of cage mates effect (j = siblings, nonsiblings); G^ = the male parental genetic stock effect (k = 1, 2); s(g)u = the sire-withinmale-parental stock effect (k = 1, 1 = 1, 2,..., 13) (k = 2, 1 = 1, 2 12); dcsw - the dam-within-sire effect (m = 1,2); P^OjjHnj O ) = the effect of order of observation within a day as a covariate; p^bytim - B ) = the effect of cage block as a covariate; and ejjkim = the error for an individual hen. Variation among days of observation was not sufficiently great to justify retaining it as an effect in the model. Before analysis, the proportions of time spent in the different behavioral states were transformed into arc sine square root values, and variables recorded as occurrences were transformed using the formula, (X + l)- 5 (Snedecor and Cochran, 198). Tests of the significance of sire components of variance were carried out separately for the two male parental stocks. RESULTS A large number of significant correlations existed between behavioral variables and measures of production and physical condition. Eating and standing were positively correlated with hen-housed egg production and egg mass, and eating also had a positive association with feed consumption (Table 1). In, cage pecking, toe pecking, bobbing, and physical displacement were positively correlated with measures of egg production. Sitting and resting (MPSl) were negatively correlated with egg production. Greater performances of head down and walking behavior (primarily MPSl) were associated with lower feather score (better feather coverage) and lower body weight and weight gain (Table 2). Increased levels of resting, immobility (still, MPSl), and sitting coincided with poorer plumage condition and higher body weight. Significant correlations between behavior in laying cages and foot lesions or claw length were so few as to be attributable to chance. There was no significant variation of behavior for sires within MPSl (Table 3). Li, only state of stillness had a significant sire component of variance. Li contrast, significant variability due to dams existed for a number of behavioral states. Heritability estimates for sires for the most part were small or zero. For dams, however, many of the behavioral variables had fairly large heritability estimates. DISCUSSION The most striking result was that behavioral activation (i.e., the tendency to manifest higher levels of activity rather than lower) was associated with higher egg production in battery cages. A similar phenomenon was apparent for intrastock relationships of behavior to production between the rearing and laying phases and for inter-stock comparisons (Webster, 199; Webster and Humik, 199b). Considering that various forms of activity in chickens are more energetically costly than inactivity (van Kampen, 1976a,b; MacLeod and Jewitt, 1985), one might have expected behavioral activation to draw energy away from egg production, leading to negative, rather than positive, associations between the two. It is not known if higher rate of production stimulated higher levels of activity, or if greater vigor led to higher production. In the present study, behavior indicative of higher activity tended to correspond to better feed conversion measures. However, Morrison and Leeson (1978) and Braastad and Katie (1989) reported that hens selected for feed efficiency were less active than feed-inefficient hens. El-Attar et al. (1983) noted that pullets that had a lower frequency of resting had better feed conversion. That eating was positively associated with production is not unexpected, and might appear not to be related to behavioral activation at all. However, hens spent an average of 2% of their time performing the behavior defined as eating (Webster and Hurnik, 199a), which is much more than actually necessary to consume an adequate amount of feed. Given the relatively low stimulus complexity of a battery cage environment, the feed trough is a major attraction and time spent manipulating feed probably reflects the degree of behavioral activation experienced by a hen. The relationship between eating and production in the present study arose from measures pooled for hens within cages, and so appears to be more general than that found by Cunningham and van Tienhoven (1984) in which hens of low social rank ate less frequently and had lower production in situations of feed restriction. Cage pecking, toe pecking, and bobbing often were performed in a stereotyped manner. Downloaded from http://ps.oxfordjournals.org/ at Penn State University (Paterno Lib) on February 18, 216
424 WEBSTER AND HURNK Because performance of stereotypic behavior may be a means of controlling sensory input and modifying motivational state in suboptimal environments (Broom, 1983), stereotypic behavior in caged hens could be a reflection of behavioral activation. The relationships of the above actions to egg production and feed conversion in, therefore, may be further indications of a link between behavioral activation and production. The connection between stereotypy and production evidently differs between stocks because the two sets of variables were much less interrelated in. It was not foreseen that walking would have a positive association and that measures of inactivity (resting, still, sitting) would have negative associations with plumage condition. Inactivity, particularly when recumbent, would expose a hen to feather damage caused by the claws of cage mates. Poorer physical condition Behavior 2 Hflik Eating Preening Hdwn Resting Peck Still Walking Bobbing Displ Standing Sitting Tpk Step TABLE 1. Spearman correlations between behavior and production-related variables Stock Hhprod.25*.33** -.27*.14 Eggwt.29*.17 -.31* -.6.32*.17 -.15 -.33*.19 -.29*.47****,34*.19.27*.34**.25* -.35** -.25*.1.34** -.33* -.46***.34**.46*** Eggmass Defrm.12.28* -.38** -.6.12.3*.18.3* -.18.3*.22.28* -.22 -.29*.3.3**.2.3* -.12 -.3* Production 3,4 Cxcol.28*.11 -.23*.9.23* -.9 Cxshk -.6 -.34** Soft -.32** -.26* -.9 -.25*.9.25* Feed 21*.43**** -.25*.1.29* -.11-3A* -.4 -.8.37** Fdconv.36** -.11 -.2 -.28* -.17-29** 'Data are presented only when the coefficient for at least one of the stocks differed significantly from. Fdoz -.26*.9 29* -.1 -.2 -.23* -.24* -.17 -.29* -.15.29*.14 -. -.3** -.18-21* Abbreviations for behavioral variables: Hflik = head flicking; Hdwn = head down; Peck = cage pecking; Displ = physical displacement; Tpk = toe pecking; Step = stepping on cagemate. Production data were from period closest to that in which behavior was observed, except egg cracks, which were averaged over all periods. Abbreviations for production variables: Hhprod = hen-housed production; Eggwt = egg weight; Eggmass = egg mass production; Defrm = egg shell deformation; Cxcol = cracks at collection; Cxshk = cracks after shaking; Soft = softshelled eggs; Feed = feed consumption; Fdconv = feed conversion; Fdoz = feed per dozen eggs. *P<5. **P<.1. ***P<1. ****P<.1. Downloaded from http://ps.oxfordjournals.org/ at Penn State University (Paterno Lib) on February 18, 216
CORRELATION OF HEN BEHAVIOR AND PRODUCTION 425 could have led to inactivity and plumage deterioration, but this was not investigated. The relationships of the above variables to body weight are consistent with die relative energy demands of higher versus lower activity. Head down behavior had a greater correspondence to behavioral activation man to inactivity (Webster and Hurnik, 199a), so it is not unreasonable that its relationships to feathering and body weight were similar to those of walking. The variability of behavior that was manifest in the correlations cannot, to any great extent, be attributed to sires within male parental stocks. Little variation due to sires was found during the rearing phase as well (Webster and Hurnik, 1987,1989), the greatest Behavior Head flicking Eating Preening Head down Resting Still Drinking Walking Bobbing Physical displacement Standing Sitting Toe pecking amount being apparent during a study of openfield behavior at 17 wk of age (Webster and Hurnik, 1989). The dam parental stock had sizable heritability estimates for a number of behavioral variables related to measures of production or physical condition. They are eating, standing, cage pecking, and physical displacement, which were positively associated with egg production. Resting and, to a lesser extent, sitting were negatively associated with egg production and plumage condition and positively associated with body weight. Walking and head down behavior were positively associated with feather condition and negatively associated with body weight. Furthermore, feather pecking, measures of aggression (pecking cage- TABLE 2. Spearman correlations between behavior and measures of physical condition^ Stock Feather score.14.29* -.31* -.24*.31**.33** 45****.13 -.27* -.23.21.23* -.7.25* Physical condition 2 Body weight -.31** -.1 -.42*** -.5****.4***.35**.32**.33** -.29* -.2 -.28* -.8.39*** -.35**.39***.36*.13 -.26*** Weight gain.11.31** -.32** -.11 -.33** -.34** -.26*.11 -.27*.1 -.3* -.7.9 -.15* *Data are presented only when the coefficient for at least one of the stocks differed significantly from. J Physical condition data were taken from the period closest to that in which behavior was observed. *P<.5. **P<1. ***P<1. ***P<.1. Downloaded from http://ps.oxfordjournals.org/ at Penn State University (Paterno Lib) on February 18, 216
426 WEBSTER AND HURNIK mate, pecking neighbor) and cage climbing (which appeared to be attempts to escape the cage) also had fairly large heritability estimates. Although it is not unreasonable to suspect that additive genetic variability exists in the female parental stock for many behavioral states, nonadditive genetic effects and effects of common environment also contribute to the dam component of variance (Falconer, 1981). The heritabilities for eating in and in the female parental stock, and for resting in the female parental stock, are comparable with those noted by Hurnik (1978). The heritability estimates in the present study for drinking and standing were less than those of Hurnik (1978). Prelaying behavior in the forms of pacing and sitting has been found to have a genetic basis (Mills et al., 1985). Bobbing in the present study occurred most frequently in the hour before oviposition (Webster and Hurnik, 199a), but there was no evidence of significant genetic variability for either sires or dams. If the relationship of behavioral activation to production and physical condition is confirmed by further research, it should be possible to develop genetic stocks with behavioral propensities better suited to the cage environment, at least with regard to the traits measured. Faure (198, 1981) has selected lines of chickens to have different levels of activity in a standardized environment. The female parental stock in the present study may have had sufficient genetic variability for selection of a number of actions related to behavioral activation. The apparent lack of genetic variability for sires within each male parental stock might have been due to genetic fixation at gene loci controlling the expression of behavior in the stocks acquired from the commercial breeder. However, the investigation of genetic variability in the present study involved a relatively small number of sires and a single set of observations of the behavior of each hen. In future, use of larger numbers of sires and TABLE 3. Components of variance and heritability estimates for behavior recorded by direct visual observation Behavior Head flicking Eating Preening Head down Resting Still Drinking Walking Bobbing Toe pecking Feather pecking Standing Sitting Cage pecking Pecking cagemate Stepping on cagemate Physical displacement Climbing cage Pecking neighbor U2 x 1 3. -.23 2.62 1.5 2.13 -.5.7.92.8 -.5.37 -.6-1.82-1.1 64.2-19.34-26 -4.17-2.78-2.45 Component of variance 1,2 Sires -.83-8.81 2.94.45 -. 2.53*.1 -.11.67. -.33-7.17-5.99-163.5 -.9 -.1-17.32-18.84-7.77 EPS Dams -.54 6.15*.28 3.5** 3.35** -.1 -.22.83**.4.11.29 6.58* 7.65* 153.56 15.16 -.23 53.72** 39.59** 14.6.26 ±.24.16 ±.2.5 ±.16.15 ±.2 21 ±.39.16 ±.19.7 ±.19.14 ±.2.1 ±.33 Heritability ± SE 3-4 Sires.24 ±.67 ±.3.4.2 ±.2.12 ±.25.6 ±.18 FPS Dams.35 ± 26.4 ±.18.45 ± 29.49 ±.3.47 ±.29.19 ± 22.57 ±.32.35 ±.26.29 ±.25.52 ±.3.42 ±.28.47 ±.29.48 ±.3.53 ±.48 2 = male parental Stock 1, df = 12; = male parental Stock 2, df = 11; FPS = female parental stock, df = 24. 3 Heritabilities were calculated from untransformed data. Heritability was considered zero when the sire or dam component of variance was negative. *P<.5. **P<1. Downloaded from http://ps.oxfordjournals.org/ at Penn State University (Paterno Lib) on February 18, 216
CORRELATION OF HEN BEHAVIOR AND PRODUCTION 427 repetition of measurement to improve the precision of behavioral assessment might help expose genetic variability for behavior in commercial stocks of chicken. ACKNOWLEDGMENTS Appreciation is expressed to the staff of the Arkell Poultry Research Station, University of Guelph for their cooperation and help and to the Ontario Ministry of Agriculture and Food for provision of facilities. Financial support for the research came from Natural Sciences and Engineering Research Council of Canada Grant Number A6784. REFERENCES Braastad, B. O., and J. Katie, 1989. Behavioral differences between laying hen populations selected for high and low efficiency of food utilisation. Br. Poult. Sci. 3: 533-544. Beilharz, R. G., 1982. Genetic adaptation in relation to animal welfare. Int. J. Study Anim. Prob. 3:117-124. Broom, D. M, 1983. Stereotypies as animal welfare indicators. Pages 81-87 in: Indicators Relevant to Farm Animal Welfare. D. Smidt, ed. Martinus Nijhoff Publishers, Boston, MA. Craig, J. V., 1982. Behavioral and genetic adaptation of laying hens to high-density environments. Bioscience 32:33-37. Craig, J. V., T. P. Craig, and A. D. Dayton, 1983. Fearful behavior by caged hens of two genetic stocks. Appl. Anim. Ethol. 1:263-273. Craig, J. V., S. K. Kujiyat, and A. D. Dayton, 1984. Tonic immobility responses of White Leghorn hens affected by induction techniques and genetic stock differences. Poultry Sci. 63:1-1. Craig, J. V., L. L. Ortman, and A. M. Guhl, 1965. Genetic selection for social dominance ability in chickens. Anim. Behav. 13:114-131. Craig, J. V., and A. Toth, 1969. Productivity of pullets influenced by genetic selection of social dominance ability and by stability of flock membership. Poultry Sci. 48:1729-1736. Cunningham, D. L., and A. van Tienhoven, 1984. The effects of management program and social rank on behavior and productivity of White Leghorn layers in cages. Poultry Sci. 63:25-3. Dunnington, E. A., and P. B. Siegel, 1983. Mating frequency in male chickens: Long-term selection. Theor. Appl. Genet. 64:317-323. El-Attar, M., P. Merat, and A. Bordas, 1983. Comportement de poulettes normales et naines (dw) en openfield et cages individuelles, en correlation avec la croissance et l'efficacite alimentaire. Arch. Geflugelkd. 47:9-95. Falconer, D. S., 1981. Introduction to Quantitative Genetics. 2nd ed. Longman Group Ltd., Essex, England. Faure, J. M., 198. To adapt the environment to the bird or the bird to the environment? Pages 19-42 in: The Laying Hen and its Environment. R. Moss, ed. Martinus Nijhoff Publishers, The Hague, The Netherlands; Boston, MA; and London, England. Faure, J. M, 1981. Bidirectional selection for open-field activity in young chicks. Behav. Genet. 11:135-144. Faure, J. M., and J. C. Folmer, 1975. Etude genetique de l'activite precoce en open-field du jeune poussin. Ann. Genet. Sel. Anim. 7:123-132. Hansen, R. S., 1976. Nervousness and hysteria of mature female chickens. Poultry Sci. 55:531-543. Hemsworth, P. H., and J. L. Bamett, 1989. Relationships between fear of humans, productivity and cage position of laying hens. Br. Poult. Sci. 3:55-518. Hurnik, J. F., 1978. Feed conversion and feeding behaviour in laying hens. Pages 7-85 in: Proc. 27th National Breeder's Roundtable, Kansas City, Missouri. Published by Poultry Breeders of America, Arlington, VA. Hurnik, J. F., 198. Animal welfare and modern agriculture. Anim. Regul. Stud. 2:145-154. Hurnik, J. F., F. N. Jerome, and I. McMillan, 1977. Hereditary effects of natural colour preference in domestic chickens. Biol. Behav. 2:23-212. MacLeod, M. G., and T. R. Jewitt, 1985. The energy cost of some behavior patterns in laying domestic fowl: simultaneous calorimetric, Doppler-radar and visual observations. Proc. Nutr. Soc. 44:34A. (Abstr.) Mills, A. D., D.G.M. Wood-Gush, and B. O. Hughes, 1985. Genetic analysis of strain differences in prelaying behaviour in battery cages. Br. Poult. Sci. 26: 187-197. Morrison, W. D., and S. Leeson, 1978. Relationship of feed efficiency to carcass composition and metabolic rate in laying birds. Poultry Sci. 57:735-739. Siegel, P. B., 1979. Behaviour genetics in chickens: A review. World's Poult. Sci. J. 35:9-19. Siegel, P. B., 1989. The genetic-behaviour interface and well-being of poultry. Br. Poult Sci. 3:3-13. Snedecor, G. W., and W. G. Cochran, 198. Statistical Methods. 7th ed. Iowa State Univ. Press, Ames, IA. Toates, F., 1987. The relevance of models of motivation and learning to animal welfare. Pages 153-186 in: Biology of Stress in Farm Animals: An Integrative Approach. P. R. Wiepkema and P.W.M Adrichem, ed. Martinus Nijhoff Publishers, Boston, MA. van Kampen, M., 1976a. Activity and energy expenditure in laying hens. 2. The energy cost of exercise. J. Agric. Sci., Camb. 87:81-M. van Kampen, M., 1976b. Activity and energy expenditure in laying hens. 3. The energy cost of eating and posture. J. Agric. Sci., Camb. 87:85-88, Webster, A. B., 199. A study of behavior, production and well-being in the White Leghorn-type chicken. PhJD. dissertation, Univ. Guelph, Guelph, Ontario, Canada. Webster, A. B., and J. F. Hurnik, 1987. Heritability of the peep vocalization in White Leghorn-type chicks. Appl. Anim. Behav. Sci. 19:157-168. Webster, A. B., and J. F. Hurnik, 1989. Genetic assessment of the behavior of White Leghorn type pullets in an open field. Poultry Sci. 68:335-343. Webster, A. B., and J. F. Hurnik, 199a. An ethogram of White Leghorn-type hens in battery cages. Can. J. Downloaded from http://ps.oxfordjournals.org/ at Penn State University (Paterno Lib) on February 18, 216
428 WEBSTER AND HURNIK Anim. Sci. 7:751-76. Webster, A. B and J. F. Humik, 199b. Behavior, production and well-being of the laying hen. I. Effects of movable roosts, relationship of cage mates and genetic stock. Poultry Sci. 69:2118-2127. Wiepkema, P. R., 1987. Behavioral aspects of stress. Pages 153-186 in: Biology of Stress in Farm Animals: An Integrative Approach. P. R. Wiepkema and P.W.M. Adrichem, ed. Martinus Nijhoff Publishers, Boston, MA. Downloaded from http://ps.oxfordjournals.org/ at Penn State University (Paterno Lib) on February 18, 216