WP5 Use of traits in animal selection (Genetic parameter estimations, genetic model developments & evaluation of breeding schemes)
WP leader: Hélène Gilbert (INRA)
Involved partners: INRA, WUR, IRTA, TOPIGS, COBB, INCO, IFIP
The objective of WP5 is to explore new traits and models for estimating breeding values for feed efficiency, and to identify new strategies to use these in breeding programmes without impairing product quality, welfare and robustness considering the diversity of production environments and feed resources in the EU, and anticipating the effects of climate change on production systems. In this WP, we will:
- validate the potential of new on-farm measured heritable traits to predict breeding values for feed efficiency, robustness and welfare;
- propose criteria and methodologies to quantify variability in gut microbiota as a heritable trait influencing feed efficiency, robustness and welfare;
- produce new genetic models to better predict the individual variability of feed efficiency, robustness and welfare;
- develop new selection strategies for feed efficiency to account for crossbred performance and genomic information, accounting for correlated impacts on the animals’ sensitivity to environmental changes, product quality, environmental impact and welfare traits;
- validate the importance and use of social effects and crossbred data for improving selection for feed efficiency.
WP5 will benefit from new knowledge and data generated in the Feed-a-Gene project in terms of new feed resources for animals (WP1), new traits for innovative feeding and breeding (WP2) and model-derived traits of feed efficiency and robustness (WP3).
WP5 will revisit existing data and samples available from the partners, involving different breeding systems or climatic conditions. WP5 will carry out genetic analyses on trials designed to propose novel traits for feed efficiency (WP2). Specific trials will be run to generate new data, in particular including records on welfare, robustness and product quality associated to new feed resources (WP1), and for demonstration of selection strategies. Data will originate from purebred and crossbred pigs from conventional (TOPIGS, IFIP, and INRA) and alternative breeding systems (IRTA, INRA), from crossbred poultry from conventional populations (Cobb), and from experimental lines in pigs, poultry and rabbits (INRA and IRTA). Genetic parameters for new traits related to feed efficiency will be derived using conventional genetic approaches, and revisited models will be developed. New breeding strategies, including genomic information, crossbred data, and indirect indicators of feed efficiency will be explored to propose new selection indexes while accounting for the economic impact on welfare, robustness, and product quality. Results will be shared for the generation of new feeding strategies (WP4) and new models and tools of nutrient utilisation (WP3). The results will also be used for the sustainability assessment of new production systems (WP6).
Task 5.1: Genetics of components of feed efficiency and robustness indicators (M1 – M48)
Leader: Hélène Gilbert (INRA)
Involved partners: INRA, IRTA, TOPIGS, Cobb, IFIP
The genetic transmission of new feed efficiency traits generated in WP2, together with indicators of robustness, welfare, and product quality will be studied by variance component estimation techniques using classical models in quantitative genetics, and new models developed in Task 5.3. The underlying genetic determinism of these traits will be evaluated through high throughput methodologies to provide genomic and physiological indicators of feed efficiency and its components. When possible, nutrigenomics approaches will be applied. The genetics of the animals’ activity, competition and social interactions automatically assessed in WP2 will be evaluated as genetic components of feed efficiency in purebred and crossbred animals. Automatic feeders will be used in pigs (1000 Duroc and 2000 commercial pigs), 200 poultry and 600 rabbits, as well as video recordings in pigs. The genetic relationships between feed efficiency, activity and responses to stress will be assessed under classical and challenge diets in lines divergently selected for residual feed intake (100 layers and 50 pigs per line), for response to ACTH (50 pigs per line), and tested in commercialpigs, to identify indicators of sensitivity to stress related to feed efficiency. The genetic relationships between feed efficiency and robustness (quantified as longevity and performance homogeneity) will also be evaluated in reproductive females in two alternative pig breeds (1000 Duroc and 500 Iberian) and in a conventional breed (Large White). Two lines of mice selected for high and low homogeneity of birth weight during six generations will be used as a model for evaluating the correlated responses to selection. Finally, the genetics of digestive efficiency will be assessed using the methodology developed in WP2 under a classical and a challenge feed, together with their impact on feed efficiency and product quality in pigs (800 pigs per diet, from 100 sires) and in layers (50 hens per feed and line selected for residual feed intake). Metabolomic and genomic indicators of digestive efficiency identified in broiler lines (WP2) will be further tested as feed efficiency criteria for selection of commercial crossbred pigs and poultry. These new traits will be used to propose new breeding strategies in Task 5.4.
Task 5.2: Genetic relationships between the gut microbiota and feed efficiency (M1 – M48)
Leader: Hervé Garreau (INRA)
Involved partners: INRA, IRTA, TOPIGS
Variability in gut microbiota (characterised in WP2) and its genetic relationship with variability in direct genetic and maternal components of feed efficiency will be evaluated to propose criteria and methodologies to use variability in gut microbiota as a heritable trait affecting feed efficiency, robustness, and welfare. Records on mortality and gut disorders will be available, in addition to feed efficiency and production traits. The gut microbiota OTU based on the 16S sub-unit ribosomal gene, or combinations of these, will be used as new phenotypes for feed efficiency. The potential of individual OTU profiles to define a relationship matrix between individual records will be evaluated, to obtain (co)variance estimates of feed efficiency associated with the composition of the microbiota. The data structure will allow disentangling the maternal transmission of gut microbiota from the direct effect of the animal in a crossfostering trial between and within rabbit lines selected or not for feed efficiency (300 rabbits per line). The interaction between feeding regime (ad libitum or restricted) and the genetic control of the composition of gut microbiota will be studied in relation with growth rate and feed efficiency in a rabbit line selected for growth rate. For pigs, genetic relationships between stress, feed efficiency and gut microbiota will be examined by studying the variation in the composition of gut microbiota before and after a heat stress challenge, and the genetic ability of pigs to control this variation will be estimated (600 pigs per condition, using a cross between a heat-tolerant and a European line). In laying hens (200 animals), the sensitivity to variation in feed resources will be studied.
In the different species, SNPs chips already available or to be developed will be used to identify genomic regions of the animal associated with the composition of gut microbiota and their effects on feed efficiency. The impact of the composition of gut microbiota on sensitivity to sanitary and environmental conditions will be evaluated. These new parameters will be used to propose breeding strategies accounting for or selecting for the composition of gut microbiota in Task 5.4.
Task 5.3: Statistical-genetic modelling of feed efficiency and robustness features (M1 – M48)
Leader: Miriam Piles (IRTA)
Involved partners: IRTA, INRA, DLO, TOPIGS, IFIP, Cobb
In this task, partners with genetic modelling skills will work together to break the genetic modelling limits and allow accounting for indirect social and competition effects and genotypes x environment interactions, together with multiple repeated longitudinal data, in the estimation of variance components of traits of interest. The developed statistical models will be used in Tasks 5.1 and 5.2.
Different covariance structures between successive measurements will be tested in classical and Bayesian frameworks to properly fit repeated measurements on the variability of the nature and age-dependency of feed intake, body weight and feed efficiency records. Model goodness-of-fit and predictive ability will be used for comparing models on simulated data and on pig and rabbit data originating from Task 5.1. Social interaction models will be extended to incorporate extra phenotypic information, and to conduct unbiased and reliable genetic evaluations. The way to account for individual feeding behaviour (e.g., number of visits and time spent in the feeder) and aggressiveness will be studied using pig data from Task 5.1. Models accounting for the environmental sensitivity of individual animals, estimated as the animal’s genetic effect on the residual variance of the trait, will be developed and tested on feed efficiency to provide indicators for robustness. This implies improvement and extensions of existing models for growth and reproductive traits, and to account for multiple traits in the analyses. Results will be compared with direct indicators of robustness in reproductive females (e.g., health, mortality and longevity) in Task 5.1. The proposed models will be compared with those developed in WP3 and WP4 with the goal to include, social, and robustness factors in management tools.
Task 5.4: Selection strategies to account for crossbred and genomic data for a sustainable selection for feed efficiency (M12 – M60)
Leader: Mario Calus (DLO)
Involved partners: DLO, Cobb, TOPIGS, INRA, IFIP, IRTA
The results from the three previous tasks will be used as inputs to propose new selection strategies, accounting for genomic and crossbred information, social/competition effects, genotype x environment interactions, together with correlations with welfare, robustness, and meat quality traits.
The potential re-ranking between animals due to genotype by environment interactions will be evaluated on feed efficiency traits by the estimation of genetic correlations between purebred and crossbred breeding values in conventional pig breeds (i.e., Duroc datasets described in Task 5.1). The advantage of using crossbred data instead of purebred data in genomic evaluations for feed efficiency will be evaluated to propose new selection schemes for feed efficiency in monogastric animals. Methodologies to account for crossbred data in purebred selection schemes will be developed and tested on simulated data, and tested on pig crossbred data available in the project. The resulting selection indexes will be evaluated for their sensitivity to variation across breeding schemes. Economic weights for new traits related with feed efficiency and robustness across breeding systems will be estimated for different production systems and final products (i.e., Iberian, Duroc, and conventional). On the basis of these economic weights and variance components of feed efficiency and robustness, a net feed efficiency index (corrected for production, quality and robustness) will be evaluated as a selection criterion for feed efficiency and its sensitivity to variation across breeding systems will be tested. The developed genomic models to predict crossbred performance will also be tested on poultry data. Data produced in Task 5.5 will nourish the developments and decision making for selection strategies.
Task 5.5: Demonstration of the value of social interactions and crossbred information in selection to improve feed efficiency (M1-M60)
Leader: Egbert Knol (TOPIGS)
Involved partners: TOPIGS, IRTA, INRA, INCO
The concepts explored in the previous tasks will be demonstrated computation of genomic indexes for selecting purebred animals for crossbred performance in pigs, and for selection accounting for competition and social effects with various feeding regimes in rabbits.
In pigs, the value of genomic information for breeding crossbred animals will be empirically evaluated. Large groups of crossbred offspring tested for feed efficiency and production traits will be individually genotyped with a lowdensity SNP chip in the three first years of the project (1000/year). These will allow computing the Genomic Breeding Values (GEBV’s) of their genotyped parents with classical pedigree based EBV’s to predict the average outcome of their crossbred offspring. Genotyped crossbred offsprings give the opportunity to account for breed effects in GEBV computations. The validation of GEBV’s accuracy will be applied to feed efficiency traits but also to social interaction skills of the animals, by creating social and less social pens. Animals will also be sampled for blood and faeces (and potentially additional tissues), to validate the models and biomarkers of feed efficiency and robustness developed in Tasks 5.1 to 5.4 in commercial conditions. With these phenotypic, molecular and biochemical data, the responses to selection modelled in Task 5.4 will be consolidated on commercial data, and potential practical drawbacks will be identified.
The interaction between feeding regime and social effects will be demonstrated in two rabbit lines selected from the same initial population on feed efficiency with a high quality feed offered at a restricted level or a poorer feed quality offered ad libitum (obtained from WP1). In both lines, genetic evaluation will be conducted accounting for social effects. Individual feed intake will be recorded using an automatic feeder (WP2) and social effects will be considered by using the methodological developments from Task 5.3. Responses to selection will be compared after two generations of selection to the initial population (using frozen embryos) and to a line fed ad libitum and selected for growth rate. Direct and correlated responses to selection will be evaluated with or without accounting for social effects in animals fed both feeding regimes. These lines will also be compared under commercial conditions with an existing experimental rabbit line selected for feed efficiency and with a commercial rabbit line. In this commercial trial (2000 rabbits), the interaction between genetic types and feeding regimes will be studied.