Body Weight Deviations as Indicator for Resilience in Layer Chickens

Abstract

Resilience is the capacity of an animal to be minimally affected by disturbances or to rapidly return to the state pertained before exposure to a disturbance. Less resilient animals are expected to be more susceptible to environmental perturbations, such as diseases, and will consequently show more and/or greater fluctuations in production than more resilient animals. Natural antibodies (NAb) are antibodies recognizing antigens without previous exposure to these, and are hypothesized to be an indication of general disease resistance. The objective of this research was to investigate genetic parameters of resilience indicators based on standardized body weight (BW) deviations and to investigate its relation with immunity (i.e. NAb) and disease resistance. Keyhole limpet hemocyanin-binding NAb were measured in layer chickens, which were selectively bred for high and low keyhole limpet hemocyanin-binding NAb levels during six generations. In addition, BW data of these layers were collected on a four-weekly interval from 4 weeks of age until 32 weeks of age. Standardized deviations of BW from an individual were compared to lines' average BW (i.e. across individuals), and these were used to calculate resilience indicators: natural logarithm-transformed variance [ln(variance)], skewness, and lag-one autocorrelation of deviations (i.e. all within an individual). Heritabilities of resilience indicators were between 0.09 and 0.11. Genetic correlations between the three resilience indicators were between -0.20 and 0.40 (with high SE), which might suggest that the resilience indicators capture different aspects of resilience. Genetic correlations between resilience indicators and NAb were close to zero, which suggests that the resilience indicators and NAb capture different aspects of immunity. This might indicate that, in this dataset, environmental perturbations are only to a small extent affected by disease incidence, possibly due to a lack of disease occurrence. However, a lower estimated breeding value for ln(variance) was predictive for lower lesion scores after an avian pathogenic Escherichia coli inoculation and vice versa. In conclusion, this study shows that there is genetic variation in resilience indicators based on BW deviations in layer chickens, which opens up possibilities to improve resilience by means of selective breeding.

Keywords: body weight; deviation; disease resistance; immunity; micro-environment; natural antibody; production; resilience.

Figure 1

Examples of a more resilient sire-family (A and C; in blue) and a less resilient sire-family (B and D; in red) based on the sire's estimated breeding values for natural logarithm-transformed variance [ln(variance)]. A and B show body weights (BW) over time. C and D show standardized BW over time. C and D illustrate the difference in ln(variance) most easily: a more horizontal standardized BW line over time indicates a lower ln(variance) and thus a greater (hypothesized) resilience, and vice versa. The sire-families were obtained by mating the sire with two dams. The two families within the sire-family are indicated with different markers (round or square). Only BW of offspring with estimates for the resilience indicators (5 or more BW observations) are shown. The sire-families in this figure are families of generation 5 of the high line of the natural antibody (NAb)-selection experiment (i.e. the sires were generation 4).

Figure 2

Overview of the study populations. Body weight deviations and natural antibody (NAb) were measured during the selection experiment. Avian pathogenic Escherichia coli (APEC) mortality and APEC lesion scores were measured during the infection experiments. Base population is the unselected population (G0) and the NAb-selection lines are generation (G)1 to G6. High line and Low line are the selected lines for high and low total keyhole limpet hemocyanin-binding NAb titers at 16 weeks of age, respectively. G4⁺ and G6⁺ are the additionally bred chickens for the infection experiments and are similar to G4 and G6 of the selection experiment.

Figure 3

Average estimated breeding values (EBV) and standard errors of the two natural antibody (NAb)-selection lines (High line in blue and Low line in red) over six generations of selection for the three resilience indicators from the linear animal model (A: natural logarithm-transformed variance; B: skewness; and C: autocorrelation). Only EBV of females are shown. EBV of generation 0 and generation 1 are based on offspring phenotypes (i.e. no own phenotypes).