Genetic analysis of production, immunity and behaviour in laying hens

The new regulations about the husbandry of laying hens and the so-called genomic revolution offer both opportunities and challenges for the breeding of layers. Hens are currently housed mainly in battery cages of 4 individuals each. Following recent developments of the communitarian legislation, many countries will soon adopt furnished cages or non-cage systems, which will lead to larger groups of hens. Also, beak-trimming will be prohibited in EU countries in the near future. Advancements in sequencing technology are making an always greater number of genetic markers available at increasingly cheaper prices, making genome-wide studies possible and helping geneticists to start unraveling the mystery of the genetic make-up of animals, which until a few years ago was considered a black-box. This thesis touches upon the impact of such innovations on the breeding of laying hens. Use of pooled data in the genetic evaluation of laying hens Hens are usually housed in cages and therefore pooled instead of individual egg records are often available: a pooled egg record is the total production of a cage, when the egg production of the individual hens is unknown. Current selection schemes are carried out in nucleus herds where hens are housed individually, so that egg production of individual birds can be recorded and used for genetic evaluations. Based on this information sires and dams are selected. Such a selection scheme based on individually housed hens introduces a discrepancy between the environment where hens are selected and the environment in which hens are kept for commercial egg production (group housing). Selecting animals in one environment and using them in a different environment might lead to genotype x environment interaction (Besbes and Ducroq, 2003), thereby reducing the realized response to selection. Future husbandry conditions, with larger groups of hens or hens housed in furnished cages might make this problem even worse. A method to use pooled data in the genetic evaluation of laying hens would therefore be of interest. In Chapters 2 and 3 of this thesis it is described how to use pooled records for the estimation of heritability and breeding values. In chapter 2 the use of individual and pooled observations is compared. Individual body weights of hens at different ages were available: these were then pooled by cage in order to create pooled records. Heritabilities estimated from pooled and individual data correlated well: the standard error of estimates based on pooled records was however about twice that of estimates based on individual records. The accuracy of EBVs from pooled data is lower than the accuracy of EBVs from individual data; in the case of sires with at least 10 offspring the reduction in accuracy was about 23%. This loss of precision in estimating genetic parameters and breeding values is understandable considering that pooled records are a less detailed of information. However, this lower accuracy should be interpreted in the context of direct vs indirect selection. The breeding goal is the trait under commercial conditions (group housing), and if testing is under individual housing, the genetic correlation between group and individual housing is relevant. The ratio of the selection response for direct and indirect selection is a function of the accuracies for both situations, the standard deviations of the traits and the genetic correlation between the traits (Falconer, 1989). Similarly, the ratio between accuracies based on pooled and individual data provides a threshold for the genetic correlation between individual and group housing below which pooled data would result in a greater selection response. In practical breeding also the costs of individual housing relative to the costs of group housing are relevant. Since group housing is cheaper than individual housing, more selection candidates could be tested for the same level of costs. This would in turn result in higher selection intensity and larger response to selection. In chapter 3 the method of analyzing pooled data developed in chapter 2 was compared with an approximation consisting in assigning cage means to each hen in a cage, then treating them as individual observations. Cross-validation was used to compare the two methods: the method developed in Chapter 2 performed consistently better than the approximate method in terms of predicting ability. In the general discussion, finally, it was described how to estimate genetic and phenotypic correlations from pooled data. Across-line association studies for immune response and feather pecking behaviour The great number of genetic markers available at increasingly lower prices has been fostering developments in genomic research. Association studies between genetic markers and phenotypes are typically conducted within populations (breeds, or lines): the amount of LD conserved in a population is exploited using high marker density, such as SNP chips, and markers relatively close to QTLs are expected to show significant effects in association studies. In this thesis we propose to take it one step further and perform association studies across lines. This requires higher marker density but increases the resolution. The amount of LD conserved across lines is expected to be lower than within lines and the phase of the marker-phenotype association might be different in the different lines. On the other hand markers that happen to show significant effects in an across-line association study are likely to be close to the QTL. These issues in conducting marker-phenotype association studies across populations were addressed in Chapters 4 and 5 of this thesis, where it was shown how to deal with multiple populations when analyzing hens from 9 different genetic lines of White Leghorn and Rhode Island Red origin genotyped for a panel of 1536 SNP (Single Nucleotide Polymorphism) markers. The traits analysed were immunological parameters and plumage damage due to feather pecking behaviour, two classes of traits for which, given that they have relatively low heritability and are difficult and expensive to measure, genomic information may be particularly valuable. Immunological parameters might be used in selection programmes aimed at improving disease resistance of laying hens, while information on the genetic background of feather pecking behaviour can be useful in reducing problems due to this behavioural disorder of layers. Under future husbandry conditions susceptibility to infectious diseases and feather pecking are expected to become more serious problems: both aspects of layer production are in fact related to the number of individuals that interact with each other, which will increase as a result of the application of the EU directive 1999/74/EC. In addition, the ban of beak-trimming will make it more difficult to control the consequences of feather pecking (plumage damage, cannibalism, mortality). Genetic selection might represent an appealing addition to the current control measures. The association studies identified several regions of interest. The gene for interleukin 17 (IL17), on chromosome 3, was found to be associated with natural and acquired antibody titres, and with the classical and alternative pathways of complement activation. The major histocompatibility complex (MHC) genes on chromosome 16 showed significant association with natural and acquired antibody titres and classical complement activity. The interleukin 12B gene (IL12B) on chromosome 13 was associated with natural antibody titres. As for feather pecking behaviour, a role of the gene for the serotonin receptor 2C (HTR2C) on chromosome 4 was found. This supports existing evidence of a prominent involvement of the serotonergic system in the modulation of this behavioural disorder in laying hens. The genes for IL9, IL4, CCL4 and NFKB were found to be associated to plumage condition, revealing relationships between the immune system and behaviour.