Estimation of heritability, phenotypic and genetic correlations for growth curve characteristics of Japanese quail

Introduction Growth traits of domestic animals and their economic significance have been always considered important by breeders and breeding specialist. Growth models, calculated non-linearly, are used to express growth rate and statistical ratio between age and body weight (Kum et al. 2010). One of the ways to draw and describe the growth curve is to use growth models. By selection, you can change the shape of the growth curve and increase body weight (Anthony et al. 1996), although selecting to increase body weight has adverse effects such as obesity, leg problems and sudden death syndrome in poultry. Growth curve parameters are suitable and efficient criterion allowing to change the relationship between age and body weight through selection (Narince et al. 2010). Many researchers have used linear and nonlinear regression models to model the growth phenomenon. These models are more reliable (Vuori et al. 2006) than linear models due to the limited number of parameters and have a better fit of the data (Lambe et al. 2006). Depending on the breed of livestock and population studied, a growth function presents different results, so it is necessary to model growth for each herd separately (Bahreini Behzadi et al. 2015). Though many have investigated the growth traits of quails, studies estimating the heritability of growth curve parameters, genetic correlations between these parameters and genetic correlation between growth curve parameters and other traits (e.g., growth traits, production, conversion ratio and carcass traits) are very limited. Therefore, the aim of this study was to estimate heritability, phenotypic and genetic correlations of growth curve parameters and different body weight traits in Japanese quail. Material and methods: Data were obtained from 242 male and 242 female quail hatched between 2017-2019 that belonged to the Khorasan Razavi Agricultural and Natural Resources Research and Education Research Center. Birds were weighed individually for four generations from hatching to 42 days of age over four generations. Pedigree information included bird number, sire number and cage number, year, month, and birthday of the bird and hatching times over 4 generations. The data were first edited with CFC software and prepared for analysis. The traits studied in this study included body weight records at different ages (1, 7, 14, 21, 28, 35 and 45 days of age). SAS software's linear model and GLM procedure were used to identify constant factors affecting traits. Nonlinear regression model was used to estimate the growth curve parameters. Covariance components and genetic parameters of growth and body weight trait at different ages in quail were estimated using animal model and restricted maximum likelihood method using multivariate analysis by DMU software. Results and discussion: In the present study, moderate to high heritability (0.21-0.72) was obtained for different traits of body weight. Estimation of heritability at 1 week (0.45 ± 0.06), 2 weeks (0.42 ± 0.05), 3 weeks (0.44 ± 0.06), 4 weeks (0.45 ± 0.07), 5 weeks (0.52 ± 0.07) and 6 weeks (0.72 ± 0.06) were higher than the values reported by Barbieri et al. (2015) yet less than the estimate of Narinc et al. (2010). Estimation of moderate heritability obtained for body weight at 1 and 4 weeks of age and high heritability at 5 and 6 weeks of age show that the response to selection is low at early ages and increases with age at 5 and 6 weeks. The strongest correlations were between 3-4 weeks weight (0.93) and 1-2 weeks (0.92). Except for genetic correlations between hatching weight and final ages, most correlations were positive. The highest phenotypic correlation was between 3-4 weeks (0.83) and the phenotypic correlation between hatch weight and 6 weeks (0.006) was the lowest estimate. The results showed that the Gompertz model with the highest coefficient of determination (0.998) and the lowest error variance (1.8262) was the best growth predictor model in Japanese quails. Estimation of growth curve parameters (a, b and k) were 298.58, 3.5 and 0.053, respectively. The small difference between the observed and predicted body weight values indicates that the Gampertz model was able to describe the growth of quails well. Heritability of growth curve parameters (a, b and k) were estimated to be 0.21, 0.48 and 0.22, respectively. Phenotypic and genetic correlations between parameters a and b with different traits of body weight were generally negative. The highest genetic correlation (0.86) and phenotypic (0.66) were obtained between k parameter and 3-week weight. Genetic correlation between mean growth rate (b) and various traits of body weight was 0.9-0.62. The highest genetic and phenotypic correlations were estimated between maturity rate (k) and 2, 3 and 4weeks weights. Genetic and phenotypic correlations between growth curve parameters were generally negative. The only correlation between asymptotic weight (a) and mean growth rate (b) was positive and high. These negative correlations indicate that if selection is used to increase puberty weight, we will have negative effects on growth rate and the growth rate will tend to decrease. Conclusion: Estimation of heritability of growth curve parameters and their correlation with body weights at different ages showed that increasing production efficiency by selection based on growth curve parameters rather than selection for body weight could be useful. Curve deformations are more important than other parameters for productive purposes, especially the parameters a and k that are related to growth rates. Choosing a higher growth rate at 0- to 14-day-old does not change maturity weight, but choosing a higher body weight, that is, the near-inflection point of the curve, changes the shape of the curve without any significant change in parameter a. Nonetheless, k would change without any change in parameter a when limiting selection index is applied. The heritability of the Gompertz growth curve parameters and their correlation with weekly weights suggest that the growth rate parameter may be useful in selecting animals that are of early growth