THE INTERACTIONS BETWEEN GENOTYPES AND HOUSING ENVIRONMENTS IN THE DOMESTIC HEN B Y G. M C B RIDE * Summary A study was made of the egg production and egg weight of a flock of Australorp pullets in which full sisters from each family were housed in each of three environments: cages, semiintensive and intensive housing. The estimates of heritability in these environments were 0.68, 0.50 and 0 64 respectively for egg weight and 0 8, 0 000 and 0.55 respectively for egg number in cages, semiintensive and intensive housing. No acceptable evidence for genotypeenvironment interactions was found, though one of the flocks used had been under selection for egg production in cages and one was a control flock. I. INTRODUCTION In the poultry industry in Australia, three main types of husbandry are in use: wiresided battery cages, intensive housing on deep litter, and a semiintensive system where the birds have access to range as well as a house. Since selective breeding is generally carried out either in cages or on deep litter, it is desirable to know whether birds bred in one environment are suitable for use in another; that is, whether interactions between genotypes and housing environments are present. Morley (956) and McBride (958) have described four types of genotypeenvironment interactions. The terminology of McBride ( 958) is used here to describe the classes of interaction. A number of studies have been made of interactions between cage and deep litter housing of type B (intrapopulation by macroenvironmental differences) and type D (interpopulation by macroenvironmental differences) in fowls. These were reviewed by McBride ( 958). No interactions of type B have been observed, though Gowe ( 956) observed significant interactions of type D for survivors egg production and March body weight. It was pointed out by McBride ( 960) that the design of battery cages influenced egg production, metalsided cages being superior to wiresided cages, and a single opening in the front being superior to openfronted cages. James and Foenander ( 96) have subsequently demonstrated social behaviour effects on part record production in wiresided cages with a single opening in the front. McBride ( 960) also presented evidence that the mean and variance of egg production in birds housed on the floor varied with the type of husbandry used. It would appear, then, that a detailed description of the type of cage and husbandry used is essential in any study of differences in performance of birds in different environments. * Department of Animal Husbandry, University of Queensland, Brisbane. 95
The aim of the present study is to describe differences observed in three flocks of birds housed in three environments, cages, intensive and semiintensive. Differences in means and variances, heritabilities of egg weight and egg number are presented as well as evidence concerning genotype environment interactions of types B and D. II. MATERIALS AND METHODS The birds used in this study were from the University of Queensland flock of Australorps described by James and McBride ( 958). The flock consisted of three groups: a random bred control, a flock selected for production index (Lerner, 950), and a number of families from dams with a right oviduct. The pullets used in this experiment were hatched in August and September 955, and housed in February 956. Full sister families of 6 pullets contributed two pullets, selected at random, to each of three housing environments. A number of smaller families were used and from these, one pullet was allocated to each of the houses. In all, 5 3 pullets from 04 full sister families were used. The three housing environm.ents were: ( a > threetiered, wiresided laying cages with a single opening in front; (b) an mtensive house, ft. by 33 ft., open to the east and containing deep litter of sawdust and shavings over cement; ( C > a semiintensive system with a deeplitter house as in (b) used in conjunction with a series of grassed yards of varying sizes. An adequate identical ration of laying mash was available to all birds at all times, and approximately four inches of hopper space and one halfinch of watering space were allowed for each pullet on the floor. The egg production of the birds on the floor was recorded by trap nests used five days a week and corrected to sevenday production. The average weight of ten eggs collected in September 956 was obtained for every bird in lay at this time. Heritability estimates were made within housing environments using the model put forward by Lerner ( 950) and the computational procedure of Williams ( 954). Standard errors were calculated using the approximate method of Fisher (948). The model of Lowry et nl ( 956) was adopted for the estimation of genotypeenvironment interactions. These analyses were carried out for each character in the three possible pairs of environments: (a) cages and intensive; (b) intensive and semiintensive; (c) cages and semiintensive. III. RESULTS (a) Egg Weight The frequency distributions and means of egg weight for all birds in lay are shown in Fig.. 96
InfensiVe X= 5 7. 3 3 S e m i I n t e n s i v e X= 55.80 40 4 5 50 55 60 65 7 0 Weig tit sf Eggs (grn.) Fig. l.distribution of egg weights in three environments The paternal halfsib estimates of heritability and the total phenotypic variances and means are shown in Table. Three analyses of variance of egg weight are shown in Table. Only one interaction significant at the 5 per cent. level was found, that between sires and houses for the caged and intensively housed birds.
. TABLE I HERITABIJJTY ESTIMATES AND PHENOTYPIC VARIANCES AND MEANS FOR EGG WEIGHT IN THREE hz CT P X Control x Selected x Right Oviduct. x All Flocks HOUSING ENVIRONMENTS Cages Semiintensive Intensive 0.675 ko. 0*66 60.3 59. 58.8 59.5 050 t 0.6 0*64t0=40.73 5 56. 5 7.4 55.4 56.9 56.5 58.6 55.8 57.3 Source of Variation Housing Flock Flock x Housing Sires (Flocks) Sires x Housing (Flocks) Dams (Sires, Flocks) Dams x Housing (Sires, Flocks) Full Sisters (Dams, Housing, Sires Flocks). TOTAL T ABLE ANALYSES OF VARIANCEF EGG WEIGHT Cage and Cage and Intensive SemiIntensive d.f. 69 60 9 Mean Sq. 34*5***.0 7e6 4*3*** 7 30*0* 7*8 d.f. 7 64 Mean Sq. Intensive and SemiIntensive d.f. 5.5 00 83 96 006*7*** 6.4 5. 54.0*** 7.7 e3 74 Mean Sq. 60*7*** 0*7 * 44.5 : 33.5 09 (b) Egg Production The frequency distributions of egg number in the three environments are shown in Fig.. Whereas the distribution approached normality in the cage environment, it was extremely skewed in the intensive and to an intermediate extent in the semiintensive house. (Both highly significant.) Since no adequate transformation could be found to remove the skew in these distributions, the analysis of variance was used, though its applicability to this character under these conditions is certainly open to question. The total variances, means, and halfsib heritability estimates of survivors egg number in the three environments are shown in Table 3. 98
Semi Intensive x= 07.9 Intensive jz= 9 Co es s X = 6.90 Number of Eggs Fig..Distribution of egg numbers in three environments
TABLE 3 HERITABILITY ESTIMATES AND PHENOTYPIC VARIANCES AND MEANS OF EGG NUMBER IN THREE HOUSING ENVIRONMENTS Cages Semiintensive Intensive h a P X Control x Selected x Right Oviduct x All Flocks 0*8t0*5 883 8 7 4 7 zero 0550t0.4 663 998 07 93 06 9 7 8 08 9 Similar analyses to those carried out for egg weight were carried out for egg number in the three pairs of environments. These are shown in Table 4. TABLE 4 Source of Variation Flocks Houses Flocks x Houses Sires (Flocks) Sires x Houses (Flocks) Dams (Sires, Flocks) Dams x Houses (Sires, Flocks) Individuals (Dams Sires, Flocks and Houses) ANALYSES OF VARIANCEOF EGG PRODUCTlON Cage and Intensive Intensive and SemiIntensive d.f. Mean Sq. d.f. Mean Sq. 74 64 406.5 4948504~ 9*4* 3959*4** 490.9 4.3 699. 73 63 47.5 005** 456.7 304*7* 763 00 360. 3.8 Cage and SemiIntensive d.f. 73 69 Mean Sq. 709 6437O 9678 490 6 8. 6~7~ 38.9 04 89.4 96 837.0 6 668~3.. TOTAL 303 086.3 93 897.8 39 8057 IV. DISCUSSION (a) Egg Weight Heritability estimates and phenotypic variances were not significantly different. Only one type B interaction, that between sires and cage vs. intensive housing, appeard to be significant. The amount of attention which should be given to 00
this interaction is open to question, as eighteen. interaction mean squares were tested for significance and this may well be the oneintwenty probability of error. However, the sire by housing term in the intensive and semiintensive analysis was high though not significant (0. > P > 0.05). This may support the suggestion that sire genotypes were ranked differently under intensive housing compared with the other two environments. The lack of a type D interaction between the housing environments and the three flocks is of interest, as one of the flocks had been subjected to selection in cages and another was a control flock. The estimates of heritability obtained are in general agreement with those of previous workers, and the suggestion of a higher heritability in cages than under intensive housing is in agreement with the findings of Lowry et al ( 956). (b) Egg Production Distributions of the type shown in Fig. have been previously reported for egg production (e.g. Blyth, 95; Lowry et al, 956). The presence of the secondary hump in the frequency distributions presents problems in the use of the analysis of variance. A case may be made for the elimination of the records of birds which did not lay and those which ceased laying early. Such an approach was used by Sheldon (956). Evidence will be presented in a later paper of the part played by the peck order in causing variation in egg production leading to the skewed distributions. It is thus of interest to observe that the reduction of the effect of social order in cages (James and Foenander, 96 ) does not bring about any genotypeenvironment interaction. Once again, the lack of a type D interaction between the selected and control flocks in the three environments suggests that selection for egg production in a cage environment may be adequate. The estimate of heritability obtained for caged birds is in agreement with those reported by Sheldon (956) and Lowry et nl (956). The latter found only slight evidence for a type B genotypeenvironment interaction for this character in birds housed intensively and in cages. (c) General The results of experiments such as this emphasise the need for more adequate information on the causes of variation under different husbandry conditions. In this case, the elimination of the negative skew in the frequency distribution seems likely to result in improved levels of production. The lack of genotypeenvironment interactions of type B and D provides evidence of the ability of different individuals to cope consistently with a fairly wide range of environments. It is probably necessary to go beyond environmental tolerance limits before different genotype responses to environmental change become apparent.. The absence of interactions of type D when a selected and control flock were placed in the three environments is encouraging, though selection had been practised for only two generations, and no significant differences between flocks were observed in any of the three environments. 0
V. ACKNOWLEDGEMENTS Acknowledgements are due to Miss J. B. Stephens, who assisted with all aspects of this work; and to the Rural Credit Development Fund of the Reserve Bank, which provided the necessary facilities. VI. REFERENCES B LYTH, J. S. S. (95).The correlation between egg number and egg weight; an investigation of its inconstancy. Poult. Sci. 3: 5468. FISHER, R. A. (948). Statistical methods for research workers. Oliver and Boyd. Edinburgh. GOWE, R. S. (956).Environment and poultry breeding problems. A comparison of the egg production of seven S.C. White Leghorn strains housed in laying batteries and floor pens. Poult. Sci. 35: 430435. J AMES, J. W., and FO ENANDER, F. ( 96).Social behaviour in domestic animals.. Birds in laying cages. Aust. Jr. Agric. Res. (96) :,39,5. JAMES, J. W., and MCBRIDE, G. (958).The spread of genes by natural and artificial selection in a closed poultry flock. Jr. Genet. 56: l8. LERNER, I. M. (950). Population genetics and animal improvement. Cambridge University Press. LOWRY, D. C. LERNER, I. M., and TAYLOR, L. W. (956).Intraflock genetic merit under floor and cage management. Poult. Sci. 35:,034.,043. MCBRIDE, G. ( 958).The environment and animal breeding problems. Anim. Breed. Abs. 6: 349358. MCBRIDE, G. (960).Poultry Husbandry and the peck order. Brit. Poult. Sci. : 6568. M ORLEY, F. H. N. (956).Selection for economic characters in Australian Merino sheep. VII. Interactions between genotype and plane of nutrition. Aust. J. Agric. Res. 7: 40. S HELDON, B. L. (956).Genetic paramaters associated with characters affecting egg production in the domestic fowl.. The Heritability of total egg production during the pullet year. Aust. J. Agric. Res. 7: 6569. WILLIAMS, E. J. ( 954). The estimation of components of variability. C.S.I.R.O. Div. Maths. Stats. Tech., Paper No..