Group Selection Theory: Lessons Learned
from Poultry with Implications to Swine Breeding
William Muir, Ph.D.
Over the past five decades the art of animal breeding has rapidly advanced into an exacting science with such developments as BLUP estimation of breeding values and REML estimates of variance components. These new methods promise much faster advances in genetic improvement than previously possible. However, actual responses have often fallen short of expectation and in some cases responses were worse than with previous methods. The reason for these disappointing results can only be due to assumptions inherent to the BLUP models used. The most commonly used and recognized assumption is that of an additive model, i.e. no dominance or higher order epistasis. However, a more important, and less recognized, assumption is that of non-interacting genotypes, i.e. genotypes do not compete. If higher producing animals tend to be more competitive; the effect of selection is to increase competition. Competition has the effect of lowering productivity of other animals that are in direct contention. As such, ignoring competitive interactions invalidates the BLUP model used and negates any advantages of this technology and could in fact make it a liability. This loss includes only costs due to reduced rate of gain and ignores the increased cost of building facilities, such as small group pens, needed to manage competitive interactions. In this paper I will review applications of group selection in poultry and examine potential importance in swine breeding program.
Group Selection Theory
Siegel (1989) considered adaptability to be an individual's ability to adapt to its environment. He concluded that individuals that adapt have a higher probability of contributing genes to subsequent generations than those that do not. This concept emphasizes the individual. What if an individual adapts to its environment by eating its cagemates? Survival of the individual is maximized, as is probably production, but what of that of the group?
There are numerous ways that performance of one individual can influence that of another. Accommodation for such interactions presents and insurmountable dilemma from the point of view of classical (non-interaction) quantitative genetic methodology. Griffing (1967) recognized that with competition, the usual gene model for a given genotype must be extended to include not only the direct effects of its own genes, but also the associate contributions from other genotypes in the group. The problem is to optimize production of a given genotype in a competitive environment. As a consequence of interacting genotypes, the same genotype can have different expressions in populations having different population structures.
Griffing (1967) extended selection theory to take into consideration interactions of genotypes. The conceptual biological model was extended to define the group and the usual model was extended to include not only direct effects of its own genes, but also associate contributions from other genotypes in the group.
In the presence of interacting genotypes the expected change in the mean from individual selection is
Where is the additive variance of the direct effects and is the additive covariance between direct and associative effects. If the covariance is negative, as occurs when there is competition for a limited resource, then selection based on individual performance can have a reverse effect on the mean, i.e. positive selection will reduce rather than increase the mean. This results because a gene which has a positive direct advantage for the individual has a negative associate effect on the group. The prediction of this theory, that individual selection will have a negative impact on the group, is evident from the previous section.
In contrast, if the group is defined as the unit of selection, then
where is the additive variance for associate effects in this case is always positive. Thus, transferring selection from the individual to the group ensures that the population mean will not decrease. As group size increases, associate effects take on an increasingly dominant role in determining the consequences of selection and implies that even for weakly competitive conditions, a negative response to selection can occur. This result was shown by Muir (1985) who demonstrated a significant genotype by group size interaction between single and 9-bird cages but not between single and 4-bird cages, holding density constant.
Griffing (1967) showed that the rate of progress with groups composed of random individuals is slow and decreases as the group size increases. However, if the group is composed of related individuals the efficiency is greatly increased, particularly as group size increases (Griffing, 1976). Consideration of the interaction of relatives is important in understanding the evolution of social behavior (Hamilton, 1964). However, individual selection is still possible by use of a progeny test consisting of sib offspring housed as a group. This procedure is particularly amenable to reciprocal recurrent selection (RRS) where a progeny test is already necessary.
It is also of special note that Griffing (1967) shows that selection only on associate components cannot guarantee a positive response to selection, i.e. selection for reduced aggression will not ensure that production will increase.
Experimental Examination Group Selection Theory
Group selection theory provides two testable hypothesis, the first is that individual selection will increase competitiveness and reduce group performance. The second is that group selection will do the reverse.
Individual selection. Results suggesting that individual selection can increase competitiveness has been demonstrated in poultry. Lee and Craig (1981a,b) who found that stocks which were selected for increased productivity had greater feather loss than its unselected control when kept in 3-bird cages. Craig et al. (1975) compared aggressive behavior among lines of chickens selected for part record egg production under competitive conditions and the non-selected control from which the selected lines were derived. Results generally showed that artificial selection had increased aggressiveness and social dominance in the adolescent period. Results from Lowry and Abplanalp (1970, 1972) showed that strains selected under floor flock conditions became socially dominant to both those selected in single bird cages and unselected controls. Craig et al. (1965) and Craig and Toth (1969) showed that hens of lines selected for social dominance had lower rates of lay than did hens of the same line selected for low social dominance. In addition, Craig (1970) showed that the high social dominance line withstood crowding less well than the low social dominance line. However, in single bird cages egg production of the high line was superior to that of the low. Biswas and Craig (1970) also showed that the high strain hens had much lower production than the low line in floor pens or multiple-bird cages but were equally productive in single-bird cages
Group selection. Results supporting benefits of groups selection have also been reported. . The first experiment was that of Goodnight (1985) who showed that leaf area of Arabidopsis thaliana would respond to group but not individual selection. The first experiment to use group selection with chickens was unsuccessful (Craig et al., 1982). Craig (1994) reflected that the reason for lack of response was because he had not provided an environment in which the hidden genetic variability could be expressed, i.e. he beak trimmed hens, density was low, and only a part record was used. The first experiment reported with chickens using kin selection to improve adaptability to social stress was unsuccessful (Craig et al. 1982). In retrospect, Craig (1994) concluded that the failure may have been due to the relatively benign environments in which the hens had been kept during selection, i.e. beak trimmed, relatively low density and part-record egg production. Craig (1982) states that with kin selection practices such as beak trimming, dim lighting, and declawing should be abandoned so that such tendencies toward feather and cannibalistic pecking and claw-inflicted injuries could be revealed.
A similar group selection experiment was initiated at Purdue University by the senior author in 1981 but with more stringent conditions (Muir and Liggett, 1995a), particularly with respect to duration of stress, beak trimming, and group size. In that experiment, females of each sire family were housed as a group in a multiple-bird cage and selected as a group. In the first two generations (G1 and G2), group size was 9 (413 cm2/bird) while in the next four (G3, G4, G5, and G6) group size was 12 (362 cm2/bird). In all generations, except G1, birds were not beak-trimmed and lights were at high intensity. Production was measured to at least 60 wk of age and in G1, G5, and G6 to 72 wk. The criterion of selection was initially egg mass (EM), which was computed as the product of eggs per hen housed (EHH) and egg weight (EWT). In later generations an index giving equal weight to eggs per hen per day (EHD) and days survival (DS) was used. A non-selected control, with approximately the same number of breeders as the selected line, was maintained for comparison and housed in 1-bird cages. Egg weights were collected weekly, biweekly, or monthly in different generations. Egg weights for missing weeks were found by regression. Performance beyond 60 wk of age in generations in which it was not measured was projected by linear regression of post-peak performance.
After four generations of selection, Kuo et al. (1991) and Craig and Muir (1991) compared performance of the selected and control lines in 6-bird cages with 387 cm2 floor space per bird from 16 to 36 wk of age with 0, 1/2, or 2/3 of the beak trimmed. Results showed a highly significant beak-treatment by genetic stock interaction for hen housed rate of lay, daily egg mass, and mortality. With intact beaks, the selected line had a significantly higher egg production, egg mass, and survival. With 2/3 of the beak removed, difference in egg production and egg mass remained significantly different but the magnitude of difference had declined. Further, mortality was not significantly different. Egg weights of the selected line were slightly higher than that of the control but not significantly so.
At the 6th generation, Muir and Liggett (1995a) summarized their results. Because birds were not beak-trimmed in G1, performance of that generation was not comparable to that in subsequent generations. Results based on generations 2 through 6, showed that rate of lay increased from 52 to 68% while percent mortality decreased from 30.6% to 8.8%. The combination of these factors resulted in an average increase in days survival from 160 to 348 and an increase in total eggs per hen housed from 91 to 237 eggs. In contrast, egg weights decreased from 59.1 to 56.0 g. However, the increase in total eggs more than offset the decline in egg size as egg mass per bird housed increased from 5.3 to 13.3 kg. Performance changes of the control over the same time period were in the opposite direction for mortality, increasing from 3.4 to 9.1%, and days survival which decrease from 357 to 348 days. These latter results may be due to cumulative inbreeding. The fact that days survival and mortality had improved over the generations in the selected line, housed in multiple-bird cages, to the point that livability was similar to that of the unselected control line, housed in single-bird cages, is dramatic evidence that group selection is effective in improving animal well-being in competitive environments
In the 7th generation, the selected and control lines were compared to a commercial line and were again housed in either single- or 12-bird cages (Muir and Liggett, 1995b) and management conditions were the same as in previous generations except birds which died were replaced with extra birds of the same line. Performance was measured from 20 to 58 wk of age. The residual record from 59 to 72 wk of age was again projected by linear regression. In general, annual performances (20 to 72 wk) in single-bird cages in terms of eggs per hen housed, eggs per hen per day, egg weight and egg mass were significantly greater for the commercial than for the selected line which was in turn greater than the unselected control and mortality from cannibalism was zero for all three lines. However, in 12-bird cages the reverse was seen with the selected line superior to the commercial line for eggs per hen housed, egg mass, and eggs per hen per day. The most remarkable difference was for mortality. The commercial line had an 89% mortality at 58 wk of age as compared to the selected line with 20% and the control at 54%. In this same study, Craig and Muir (1995) observed that feather scores did not differ in single bird cages among genetic stocks. However, in 12-bird cages, the selected line had significantly better feather score than the other lines.
At 36 wk of age, half of the birds of each line were subjected to cold stress (Hester et al. 1995a), after which, at 47 wk of age, the birds subjected to cold stress were further subjected to heat stress (Hester et al. 1995b). Blood samples were taken before, during, and after each stress period. Egg production was also summarized for each of those periods. Packed cell volume immediately after housing indicated that the selected line may have adapted to the new watering system more quickly than the other lines. During cold stress the commercial and control lines showed an increase in heterophil to lymphocyte ratio in 12-bird cages while the selected line did not. Egg production before, during, and after thermal stress indicated that the selected line withstood social, handling, and environmental stress better than the control and in some cases the commercial line. Similar observations with heat stress showed that the selected line withstood heat stress better as indicated by a lower mortality than the control or commercial lines. Egg production before, during, and after heat stress indicated that the selected line withstood social, handling, and environmental stress better than the control line and in come cases the commercial. Adrenal weights were larger in the selected line than the other lines suggesting that the line may have a greater capacity to respond to stress than the other lines. Hester et al. (1995c) also reported that the lines showed no differences in humoral immune response to sheep red blood cells after either cold or heat stress. These results showed evidence that the selected line was more resistant to stress than the other lines.
Taken as a whole, these results present conclusive evidence that group selection on the traits rate of lay and longevity is effective in improving well-being of layers in a relatively short period of time without sacrificing productivity. The way for commercial breeders to develop birds which do not need beak trimming is clear. Further, because group selection is shown to improve well-being in multiple-bird cages, alternatives such as redesigning cage environments, or housing such as floor pens or free ranges, may not be needed.
Is competition a problem in Swine?
Research conducted by Frank et al. (1997) clearly indicates the impact of competition on growth. In this experiment group size was increased from one pig per pen to multiple pigs in increment steps while keeping density (space per pig) and feeder space constant. Results of this experiment show that as group size increased, percent fat increased by about 12% (Figure 1) while average daily gain decreased 7% (Figure 2).
The cumulative effect of reduced gain and increased fat reduced feed conversion by about 6% Figure 3)
An even more dramatic effect was shown by Holck et al. (1997). In their study the effect of group size was confounded with rearing environment and density. The housed barrows in either a commercial grow-finish facility with 24 pigs/pen and 0.74 m2/pig space or 3 pigs/pen and 2.23 m2/pig. The average daily gain of pigs in pigs in group sizes of 3 was 42% greater than those in groups of 24 (Figure 4).
These two studies clearly show that competition has a major impact on growth in swine.
Competition has a major impact on growth in swine Group selection applied in poultry has been shown to overcome similar problems may be of tremendous benefit to the swine industry.
Appleby, M.C., B.O. Hughes, and H.A. Elson, 1992. Poultry Production Systems: Behavior,
Management and Welfare. C.A.B. International, Wallingford, Oxon, England
Appleby, M.C., and B.O. Hughes, 1991. Welfare of laying hens in cages and alternative systems:
environmental, physical and behavioral aspects. 47: 109-128.
Ash, M, 1990, A poultry producer’s view. Proceedings of a Symposium on Farm Animal Welfare in
Indiana- A Special Program on Animal Welfare Issues. Center for the Applied Ethology and Human-
Animal Interaction and School of Veternary Mecicene, Purdue University Agricultural Experiment
Station. W. Lafayette, IN. p. 10.
Bennett, G.L., G.E. Dickerson, and T.S. Kashyap, 1981. Effectiveness of progeny test index selection
for filed performance of strain-cross layers. II. Predicted and realized responses. Poultry Sci. 60:22-33.
Biswas, D.K., and J.V. Craig, 1970. Genotype-environment interactions in chickens selected for high
and low social dominance. Poultry Sci. 49:681-696.
Brambell, F.R, 1965. Report of the technical committee to inquire into the welfare of animals kept
under intensive livestock husbandry systems. Her Majesty’s Stationary Office, London.
Breward, J., and M.J. Gentle, 1985. Neuroma formation and abnormal afferent nerve discharges after
partial beak amputation (beak trimming) in poultry. Experientia 41:1132-1134.
Choudary, M.R., A.W. Adams, and J.V. Craig, 1972. Effects of strain, age of flock assembly, and cage
arrangement on behavior and productivity on White Leghorn type chickens. Poultry Sci. 51:1928-1937.
Consortium, 1988. Guide for the care and use of agricultural animals in agricultural research and
teaching. Association Headquarters, 309 W. Clark, Champaign, Illinois.
Craig, J.V., 1970. Interactions of genotype and housing environment in chickens selected for high and
low social dominance ability. Proc. 14th World’s Poultry Congress, Madrid 2:37-42.
Craig, J.V., 1982. Behavioral and genetic adaptation of laying hens to high-density environments.
Craig, J.V., 1992. Beak trimming benefits vary among egg-strain pullets of different genetic stocks.
Poultry Sci. 71:2007-2013.
Craig, J.V., 1994. Genetic influences on behavior associated with well-being and productivity in
livestock. Proc. 5th World Congress of Genetics Applied to Livestock Breeding. Guelph, Canada
Craig, J.V., M-L Jan, C.R. Polley, and A.L. Bhagwat, 1975. Changes in relative aggression and social
dominance associated with selection for early egg production in chickens. Poultry Sci. 54:1647-1658.
Craig, J.V., and A. Adams, 1984. Behavior and well-being of hens (Gallus domesticus) in alternative
housing environments. World’s Poultry Sci. 40:221-240.
Craig, J.V., T.P. Craig, and A.D. Dayton, 1983. Fearful behavior by caged hens of two genetic stocks.
Appl. Anim. Ethol. 10:263-273.
Craig, J.V., J.A. Craig, and G.A. Milliken, 1992. Beak trimming effects on beak length and feed usage
for growth and reproduction. Poultry Sci. 71:1830-1841.
Craig, J.V., A.D. Dayton, V.A. Garwood, and P.C. Lowe, 1982. Selection for egg mass in different
social environments. 4. Selection response Phase I. Poultry Sci 61:1786-1796.
Craig, J.V., and H. Y. Lee, 1990. Beak trimming and genetic stock effects on behavior and mortality
from cannibalism in White Leghorn-type pullets. Appl. Anim. Behav. Sci. 25:107-123.
Craig, J.V., and H.Y. Lee, 1989. Research note: Genetic stocks of White Leghorn type differ in relative
productively when beaks are intact versus trimmed. Poultry Sci. 68:1720-1723.
Craig, J.V., and W.M. Muir, 1989. Fearful and associated responses of caged White Leghorn hens:
Genetic parameter estimates. Poultry Sci. 68:1040-1046.
Craig, J.V., and W.M. Muir, 1991. Research note: Genetic adaptation to multiple-bird cage environment
is less evident with effective beak trimming. Poultry Sci. 70:2214-2217.
Craig, J.V., and W.M. Muir, 1993. Selection for reduction of beak-inflicted injuries among caged hens.
Poultry Sci. 72:411-420.
Craig, J.V., and W. M. Muir, 1995. Group selection for adaptation to multiple-hen cages: Cannibalism
and feather condition. Poultry Sci. 74 s1:101
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 J.C. Swanson, 1994. Review: Welfare pe rspectives on hens kept for egg production.
Poultry Sci. 73:921-938.
Craig, J.V., and A. Toth 1969. Productivity of pullets influenced by genetic selection for social
dominance ability and by stability of flock membership. Poultry Sci. 48:1729-1736.
Cunninghan, D.L, 1992. Beak trimming effects on performance, behavior and welfare of chickens: a
review. J. Appl. Poult. Res. 1:129-134.
Duncan, I.J.H. and J. H. Filshie, 1979. The use of radio telemetry devices to measure temperature and
heart rate in domestic fowl. In: C.J Amlaner. and D.W. Macdonald (Eds.) A Handbook on Telemetry
and Radio Tracking. Pergamon Press. Oxford, England.
Duncan, I.J.H., G.S. Slee, E. Seawright, and J. Breward, 1989. Behavioral consequences of partial beak
amputation (beak trimming) in poultry. Br. Poult Sci. 30:479-488.
Emsley, A., G.E. Dickerson and T.S. Kashyap, 1977. Genetic parameters in progeny-test selection for
filed performance of strain-cross layers. Poultry Sci. 56:121-146.
Eskeland, B., 1981. Effects of beak trimming. In: Y. Sorensen (Editor), First European Symposium on
Poultry Welfare, Danish Branch World’s Poultry Science Association., Vester Farimagsgade 1, DK-
1606, Copenhagen, pp193-200.
Falconer, D. S., 1981. Introduction to Quantitative Genetics, 2nd Edition, Longman, New York, N.Y.
Frank, G. R., M. E. Spurlock, S. G. Cornelius, G. M. Willis, and M. McComb. 1997a. Effects of group
size on performance, fat depth and serum concentrations of insulin-like growth factor-1 in grow-
finishing pigs. J. Anim. Sci. Vol. 75(Suppl. 1):37(abstr.).
Gentle, M.J., 1986. Beak trimming in poultry. World’s Poultry Sci. 42:268-275.
Gentle, M.J., B.H. Thorp, and B.O. Hughes, 1995. Anatomical consequences of partial beak amputation
(beak trimming) in turkeys. Res. in Vet. Sci. 58: 158-162.
Gentle, M.J., D. Waddington, L.N. Hunter, and R.B. Jones, 1990. Behavioral evidence for persistent
pain following partial beak amputation in chickens. Appl. Anim. Behav. Sci. 27:149-157.
Griffing, B., 1967. Selection in reference to biological groups. I. Individual and group selection
applied to populations of unordered groups. Aust. J. Biol. Sci., 10: 127-139.
Griffing, B., 1976. Selection in reference to biological groups. V. Analysis of full-sib groups. Genetics
Gross, W.B., A.E. Dunnington, and P.B. Siegel, 1984. Environmental effects on the well being of
chickens from lines selected for responses to social strife. Arch. Geflugelkd. 48:3-7.
Gross, W.B., and P.B. Siegel, 1965. The effect of social stress on resistance to infection with Eschericia
coli and Mycoplyasma gallisepticum. Poultry Sci. 44: 998-1001.
Gross, W.B., and P.B. Siegel, 1973. Effects of early stress and steroids on antibody production. Avian
Gross, W.B., and P.B. Siegel, 1980. Effects of early environmental stress on chicken body weight,
antibody response to RBC antigens, feed efficiency and response to fasting. Avian Diseases 24: 569-
Gross, W.B., and P.B. Siegel, 1985. Selective breeding of chickens for corticosterone response to social
stress. Poultry Sci. 64:2230-2233.
Hester, P.Y., W.M. Muir, J.V. Craig, and J.L. Albright, 1995a. Group selection for adaptation to
multiple-hen cages: Response to social and cold stress. Poultry Sci. 74: s1:102
Hester, P.Y., W.M. Muir, J.V. Craig, and J.L. Albright, 1995b. Group selection for adaptation to
multiple-hen cages: Response to social and heat stress. Poultry Sci. 74: s1:102
Hester, P.Y., W.M. Muir, and J.V. Craig, 1995c. Group selection for adaptation to multiple-hen cages:
Immunological response to social, cold, and heat stress. Poultry Sci. 74: s1:102
Holck, J. T., A. P. Schinckel, J. L. Coleman, V. M. Wilt, E. L. Thacker, M. Spurlock, A. L. Grant, P. V.
Malven, M. K. Senn, and B. J. Thacker. 1997. Environmental effects on the growth of finisher pigs. Vol.
Hill, J.A., 1983. Indicators of stress in poultry. World’s Poultry Sci. 39:24-31.
Hunton, P., 1990. Applied breeding and selection. Pages 985-1028 In: Poultry Breeding and Genetics.
R. D. Crawford, (ed.) Elsevier, New York, N. Y.
Hurnik, J. F., 1990. Animal Welfare: Ethical aspects and practical considerations. Poultry Sci. 69:1827-
Kashyap, T.S., G.E. Dickerson, and G.L. Bennett, 1981. Effectiveness of progeny test multiple trait
index selection for filed performance of strain-cross layers. I. Estimated responses. Poultry Sci. 60:1-
Kuo, F. L., J. V. Craig, and W. M. Muir, 1991. Selection and beak-trimming effects on behavior,
cannibalism, and short-term production traits in White Leghorn pullets. Poultry Sci. 70:1057-1068.
Lee, Y.-P., and J.V. Craig, 1981. Evaluation of egg-laying strains of chickens in different housing
environments: role of genotype by environment interactions. Poultry Sci. 60:1769-1781.
Lee, H.-Y., and J.V. Craig, 1990. Beak trimming effects on behavior and weight gain of floor-reared,
egg-strain pullets of three genetic stocks during the rearing period. Poultry Sci. 69:568-575.
Lee, H.J., and J.V. Craig, 1991. Beak trimming effects on behavior patterns, fearfulness, feathering, and
mortality among three stocks of White Leghorn pullets in cages and floor pens. Poultry Sci. 70:211-221.
Lowry, D.C., and H. Abplanalp, 1970. Genetic adaptation of White Leghorn hens to life in single cages.
Br. Poultry Sci. 11:117-131.
Lowry, D.C., and H. Abplanalp, 1972. Social dominance differences, given limited access to common food, between hens selected and unselected for increased egg production. Br. Poult. Sci. 13:365-376.
Mench, J.A., 1992. The welfare of poultry in modern production systems. Poultry Sci. Rev. 4:107-128.
Moberg, G. P., and J.A. Mench, 1993. Researchable problems and priorities related to animal well-
being. Proceedings of a Symposium on Farm Animal Welfare in Indiana- A Special Program on Animal
Welfare Issues. Center for the Applied Ethology and Human-Animal Interaction and School of
Veternary Mecicene, Purdue University Agricultural Experiment Station. W. Lafayette, IN. p. 101-105.
Muir, W.M., 1985. Relative efficiency of selection for performance of birds housed in colony cages
based on production in single bird cages. Poultry Sci. 64:2239-2247.
Muir, W.M., 1994. Poultry improvement: Integration of present and new genetic approaches for layers.
Proc. 5th World Congress of Genetics Applied to Livestock Breeding, Guelph, Canada. 20: 5-12
Muir, W.M., and D.L. Liggett, 1995a. Group selection for adaptation to multiple-hen cages: selection
program and responses. Poultry Sci. 74: s1:101
Muir, W.M., and D.L. Liggett, 1995b. Group selection for adaptation to multiple-hen cages:
Comparisons with a control and commercial lines. Poultry Sci. 74: s1:101
Muir W. M., W. E. Nyquist, and S. Xu, 1992 Alternative partitioning of the genotype-by-environment
interaction. Theor. Appl. Genet. 84:193-200.
North, M.O., and D.B. Bell, 1990. Commercial Chicken Production Manual, 4th ed. Van Norstrand
Reinhold, New York, N.Y.
Siegel, H.S., 1981. Adaptation of poultry to modern production practices. In: World Poultry Production
Where and How? Spelderholt Institute for Poultry Research, Beekbergeu, The Netherlands. Ed: C.W.
Scheele and C.H. Veerkamp.
Siegel, H.S., 1995. Stress, strains, and resistance. British Poultry Sci. 36:3-22.
Siegel, P.B., 1989. The genetic-behavior interface and well-being of poultry. British Poultry Sci. 30:3-
Siegel, P.B., 1993. Opportunities for science and technology to improve production systems to assure
animal well-being and economic viability. In: Proc. of Conference on Food Animal Well-Being. Purdue
University Office of Agric. Res. Prog. W. Lafayette, IN. pages79- 88
Wegner, R., 1990. Poultry welfare- problems and research to solve them. World’s Poultry Sci. 46:19-
Zulkifli, I., and P.B. Siegel, 1995. Is there a positive side to stress? World’s Poultry Sci. 51.