Estimation of heritability and genetic correlation of body weight gain ...

Open Access Asian-Australas J Anim Sci Vol. 00, No. 00:1-6 Month 2017 https://doi.org/10.5713/ajas.17.0179 pISSN 1011-2367 eISSN 1976-5517

Estimation of heritability and genetic correlation of body weight gain and growth curve parameters in Korean native chicken Prabuddha Manjula1,a, Hee-Bok Park2,a, Dongwon Seo1, Nuri Choi1, Shil Jin1, Sung Jin Ahn3, Kang Nyeong Heo4, Bo Seok Kang4, and Jun-Heon Lee1,*

*C orresponding Author: Jun-Heon Lee Tel: +82-42-821-5779, Fax: +82-42-825-9754, E-mail: [email protected] Division of animal and Dairy Science, Chungnam National University, Daejeon 34134, Korea National Institute of Animal Science, Jeju 63242, Korea 3 Department of Information Statistics, RINS, Gyeongsang National University, Jinju 52828, Korea 4 National Institute of Animal Science, Pyeongchang 25342, Korea 1

2

These authors equally contributed to this work. Submitted Mar 8, 2017; Revised May 3, 2017; Accepted Jun 7, 2017 a

Objective: This study estimated the genetic parameters for body weight gain and growth curve parameter traits in Korean native chicken (KNC). Methods: A total of 585 F1 chickens were used along with 88 of their F0 birds. Body weights were measured every 2 weeks from hatching to 20 weeks of age to measure weight gain at 2-week intervals. For each individual, a logistic growth curve model was fitted to the longitudinal growth dataset to obtain three growth curve parameters (α, asymptotic final body weight; β, inflection point; and γ, constant scale that was proportional to the overall growth rate). Genetic parameters were estimated based on the linear-mixed model using a restricted maximum likelihood method. Results: Heritability estimates of body weight gain traits were low to high (0.057 to 0.458). Heritability estimates for α, β, and γ were 0.211±0.08, 0.249±0.09, and 0.095±0.06, respectively. Both genetic and phenotypic correlations between weight gain traits ranged from –0.527 to 0.993. Genetic and phenotypic correlation between the growth curve parameters and weight gain traits ranged from –0.968 to 0.987. Conclusion: Based on the results of this study population, we suggest that the KNC could be used for selective breeding between 6 and 8 weeks of age to enhance the overall genetic improve ment of growth traits. After validation of these results in independent studies, these findings will be useful for further optimization of breeding programs for KNC. Keywords: Korean Native Chicken; Logistic Model; Selection; Genetic Parameter; Breeding

INTRODUCTION Chicken is one of the most popular poultry species and it accounts for 86.4% of the consumption of poultry meat worldwide because of rapid development in the broiler industry over the last 50 years [1]. There are numerous indigenous chicken breeds and their economic traits and genetic potential remain largely unknown. Recently, much attention has been focused on indigenous chickens as meat or layer strains because of increasing consumer demand and environmentally viable characteristics of local ecotypes. Elucidation of changes in chicken growth trajectory over time under a given management system is important for the improvement of poultry meat production. In fact, growth curves have been widely used to describe poultry growth trajectories and summarize it using a few biologically meaningful parameters [2-4]. Knowledge of growth curve parameters is used to determine the age at which to select birds and to design management procedures for breeding programs [5]. The use of body weight at a given time as selection criteria in breeding program and estimations for heritability of growth curve parameters for chicken has been reported previously [6-8]. Thus, they can provide a sound biological basis for designing a breeding program which improves productivity, efficiency, and capability throughout the entire growth period [9].

www.ajas.info

Copyright © 2017 by Asian-Australasian Journal of Animal Sciences This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

1

Manjula et al (2017) Asian-Australas J Anim Sci 00:1-6

Korean native chicken (KNC) population was established and maintained by the National Institute of Animal Science, Republic of Korea. As the result of project named “Restoration of Korean native chicken” launched in 1994 and the subsequent selective breeding program, five distinct chicken lines, based on their plumage color, were established: Heuk-saek Jaerae-Jong (black), Hoegalsaek Jaerae-jong (gray brown), Jeokgalsaek Jaerae-jong (red brown), Baeksaek Jaerae-jong (white), and Hwanggalsaek Jaerae-jong (yellow brown) [10]. Although KNCs have a unique meat flavor and superior nutrition content compared to commercial broilers [11], retarded growth performance and high variation in body weight are obstacles for their commercialization [12]. However, moderate to high heritabilities of body weight traits have been reported for KNC [13]. The objective of present study was to provide estimates of the genetic parameters of weight gain, growth curve parameters (i.e., α, asymptotic final body weight; β, inflection point at which 50% of the asymptotic body weight is realized; γ, constant scale that is proportional to the overall growth rate) in KNC.

MATERIALS AND METHODS Animal care All the practices and procedures in this study were strictly followed “The Guide for Care and Use of Laboratory Animals” published by the Institutional Animal Care and Use Committee of NIAS, Korea.

intervals. Body weight gain during each 2-week period was obtained from hatching to 20 weeks of age. (i.e., weight gain from hatch to 2 weeks of age [GR0-2], weight gain from 2 to 4 weeks of age [GR2-4], weight gain from 4 to 6 weeks of age [GR4-6], weight gain from 6 to 8 weeks of age [GR6-8], weight gain from 8 to 10 weeks of age [GR8-10] and so forth). The normality of body weight data and body weight gain traits were ascertained. If putative outliers were identified, they were excluded based on the Ryan-Joiner method, which was executed in Minitab software [14].

Growth curve data analysis For the KNC population, parameters of growth curves were estimated using Gompertz, von Bertalanffy, and logistic growth curve functions in the SAS NLIN procedure [15]. To determine the fit of the growth curves, an R2 value was computed by using the SAS NLIN procedure. In addition, convergence properties (i.e., number of chickens that the model did not converge) and convergence weight (i.e., number of chickens for which the converged weight at 20 weeks of age was 5 times heavier than their actual weight) were used as criteria for the fit of growth curves. 103 individual KNC, three paramete These two criteria were obtained using Minitab’s programming language [14]. Equations for the three104 growthweight curveismodels areand γ, a cons realized; given in Table 1. For each individual KNC, three parameters (i.e. 105 point extracted using50% Minitab’s progra α, asymptotic live body weight, β, inflection at which of the asymptotic weight is realized; and 106 γ, a constant scale that is proportional to the overall growth rate) from the best-fit equa107 Genetic parameter estimation tion were extracted using Minitab’s programming language.

Animals 108 Fixed effect of sex, line, and b The experimental animals included the F0 and F1 generations of Genetic parameter estimation KNC resource pedigree. The KNC nuclear pedigree consisted Fixed effect of sex, line, and batch were for significance 109 tested statistical software [14]. The (c of 83 F0 founders (15 males and 68 females) and 585 F1 progeny in each trait in a general linear model using Minitab statistical 110 univariate linear mixed-effects m (282 males, 303 females). Within-line mating was practiced in software [14]. The (co)variance components and heritability for which a total of three cockerels were mated with 4 to 5 hens to all traits were estimated with the following 111univariate linear mixedproduce F1 birds. The F1 progeny from two batches were categoeffects model using the ASReml program [16]. 112 Y = Xb+Za+e rized into the five lines: Red brown (R) (n = 135), White (W) (n = 122), Yellow brown (Y) (n = 130), Gray brown (G) (n = 110), Y = Xb+Za+e 113 Black (L) (n = 88) based on their plumage color. This resource 11 114 Where, Y is afor vector corresp population be made of 68 full-sib families with 3 to 20 birds (aver Where, Y is a vector corresponds to the phenotypic values 11 age 10.6). With regard to half-sib families, the population was weight gain traits and growth curve parameter traits; b is the vec115 is the vector of fixed effects inc 11 consisted of 15 half-sib families ranging from 28 to 59 birds torforofmodeling fixed effects including batch, lines, sex, and; a is a vector of 284 (averTable 1. Equations the growth of Korean native chicken 285 age 44.5). Dams were not used for hatching and brooding. All 1. Equations 116 to be random additive genetic effects, assumed bea~ a~NN(0, (0,A A 𝜎𝜎𝑎𝑎2 );); A is consi 284 Table for modeling the growth of Korean native chicken 286 285 animals were nurtured under standard breeding and management Model1. Equations for modeling the growth of Korean Equation 284 Table native chicken 287 286genetic variance; 117 is the additive 285 procedures implemented by the National Institute of Animal 288 Model Equation 287 286 289 Table 1. Equations for modelingEquation the growth of−𝛾𝛾𝑡𝑡Korean native chicken Science (NIAS) of Korea. The same environmental and feeding 288 Model 𝑊𝑊(𝑡𝑡) = 𝛼𝛼𝑒𝑒 −𝛽𝛽𝑒𝑒 118 Gompertz and 𝜎𝜎𝑎𝑎2 is the290 residual variance; a 287 289 −𝛾𝛾𝑡𝑡 −𝛽𝛽𝑒𝑒 regime was provided throughout the experimental period ofGompertz 20 Model Equation 288 𝑊𝑊(𝑡𝑡) = 𝛼𝛼𝑒𝑒 291 290 289 Genetic 292 correlation coefficie −𝛽𝛽𝑒𝑒 −𝛾𝛾𝑡𝑡 119 weeks. 𝛼𝛼𝑒𝑒(1 291 𝑊𝑊(𝑡𝑡) = 𝛼𝛼 − 𝛽𝛽𝑒𝑒 −𝛾𝛾𝑡𝑡 )3 Gompertz von BertalanffyGompertz

290 293 292 291 294 293 292 −𝛾𝛾𝑡𝑡 3 Growth data collection 295 −𝛾𝛾𝑡𝑡 = 𝛼𝛼𝛼𝛼 /(1 (1 −+𝛽𝛽𝑒𝑒 𝑊𝑊(𝑡𝑡) = 𝛽𝛽𝑒𝑒 )) von BertalanffyLogistic Logistic 294 293 296 Body weight traits were measured as a part of the breeding and 295 −𝛾𝛾𝑡𝑡 121 𝑊𝑊(𝑡𝑡) = 𝛼𝛼 /(1 + 𝛽𝛽𝑒𝑒 ) Logistic 294 297 W (t) is the corresponding weight at time t; α, asymptotic live body weight (gram); 296β, the 295 maintenance procedures. For each bird, body weight at298 a specific −𝛾𝛾𝑡𝑡 𝑊𝑊 = 𝛼𝛼 /(1 + 𝛽𝛽𝑒𝑒 ) Logistic (𝑡𝑡) asymptotic mature weight to be gain after log-function for the proportion of the 297 𝐲𝐲𝟏𝟏 𝐗𝐗birth 𝟎𝟎 𝐛𝐛𝟏𝟏 𝐙𝐙 299 [𝐲𝐲 (wk). ] = [296𝟏𝟏 ][ ]+ [ age was recorded from hatching to 20 weeks of age 298 at 2-week (wk); γ, a constant scale that is proportional to the122 overall growth rate 𝟎𝟎 𝐗𝐗 𝟐𝟐 𝐛𝐛𝟐𝟐 𝟎𝟎 297 𝟐𝟐 300 299 298 301 300 299 123 www.ajas.info 302 2 301 300 303 302 301 304 124 Where, y1 and y2 are vectors 303 302 von Bertalanffyvon Bertalanffy

𝑊𝑊(𝑡𝑡) = 𝛼𝛼 (1 − 𝛽𝛽𝑒𝑒 −𝛾𝛾𝑡𝑡 )3

120

[16]:

113

112

111

Y = Xb+Za+e

onds to the phenotypic values for weight gain traits and growth curve parameter traits; b 114Y = Xb+Za+e Where, Y113 is a112 vector corresponds to the phenotypic values for weight gain traits and growth curve parameter traits; b Y = Xb+Za+e uding batch, Manjula lines, sex, a Asian-Australas is a vector ofJ Anim random additive genetic effects, assumed to values for weight gain traits and growth curve parameter traits; b 114 Where, Y is Sci a vector corresponds to the phenotypic et al and; (2017) 00:1-6 115 is the vector of fixed 113 effects including batch, lines, sex, and; a is a vector of random additive genetic effects, assumed to

ered as additive genetic relationship matrix obtained from the nuclear F1 batch, pedigree andsex, 𝜎𝜎𝑎𝑎2 and; a is a vector of random additive genetic 2 isA the vector of fixed effects including lines, effects, assumed to 116Where, be Y Na (0, A115 𝜎𝜎corresponds is considered asisadditive genetic obtained from nuclear F1traits; pedigree 𝜎𝜎𝑎𝑎2 curve 𝑎𝑎114 vector toWhere, the phenotypic Y a vector values corresponds for relationship weight togain thematrix traits phenotypic and values curve forthe weight parameter gain traits and b and growth parameter traits; b Aa~ isisconsidered as);additive genetic relationship matrix obtained Wegrowth evaluated three widely used growth curve models (i.e., Gom2 2 e~ N (0, I𝜎𝜎𝑎𝑎 ); I is the identity matrixpertz, von Bertalanffy, is a vector offrom random residual effects assumed to be logistic to fit growth curve 116 a~ N (0,eand A 𝜎𝜎vector Aofisadditive considered as additive genetic relationship matrix from models) thematrix nuclear F1 the pedigree and 𝜎𝜎𝑎𝑎2 the nuclear Fincluding pedigree genetic vari𝑎𝑎 is);the is the genetic variance; is alines, random residual effects assumed to be e~ I𝜎𝜎𝑎𝑎2 );obtained Iand is the identity 1be is117 the vector ofadditive fixed effects 115 is the batch, vector of fixed sex, and; effects a isincluding a vector batch, of random lines,additive sex, and; genetic aNis(0, a effects, vector of assumed random toadditive genetic effects, assumed to ance; e is a vector of random residual effects assumed to be e~ N of the KNC population from hatching to 20 weeks of age (Table 2 nd X and Z are incidence matrices related to fixed and random effects, respectively. 2 2 117 is the additive genetic e is a matrices vector ofrelated random residual effects to respectively. be e~ N (0,models I𝜎𝜎𝑎𝑎 ); I performed is the identity matrix 2 2variance; 118 and the variance; and arerelationship fixed and random effects, 2). Inthe terms ofassumed RF21matrix values, the similarly be a~ N (0, A𝜎𝜎I𝑎𝑎𝜎𝜎is is 116 considered bematrix a~ as additive N (0,X Aand genetic 𝜎𝜎Z Aincidence isresidual considered matrix as additive obtainedto genetic from relationship nuclear pedigree obtained andthree 𝜎𝜎𝑎𝑎2from the nuclear F1 pedigree and 𝜎𝜎𝑎𝑎 (0, IA isresidual the identity and variance; 𝑎𝑎 );); 𝑎𝑎 is);the and Xby andthe118 Z following are incidence related to fixed and random in fitting the growth curve of KNCs. However, the logistic curves 2 matrices nts were estimated bivariate linear mixed-effects model using ASReml and 𝜎𝜎𝑎𝑎 is the residual variance;by and and Z are bivariate incidencelinear matrices random effects, 2 related to fixed 2 respectively. Genetic correlation estimated theX mixed-effects modeland using is119 the additive genetic variance; 117 ecoefficients isisa the vector additive ofwere random genetic residual variance; effects e following is assumed a vector of to random be e~ Nresidual (0, I𝜎𝜎 effects I is performance theassumed identity to matrix be e~ASReml N the (0, I𝜎𝜎 matrix 𝑎𝑎 );best 𝑎𝑎 ); I is the identity effects, respectively. showed the both in convergence properties

119 Genetic correlation coefficients were estimated by and the following bivariateweight linear mixed-effects model ASReml Genetic correlation estimated by the followthematrices convergence 2). Thus, the using logistic model 120𝜎𝜎𝑎𝑎2 is [16]: and the residual variance; 118 andcoefficients and X and 𝜎𝜎𝑎𝑎2 is Z the arewere residual incidence variance; matrices and related X and to Z fixed are and incidence random effects,related respectively. to fixed(Table and random effects, respectively.

ing bivariate linear mixed-effects model using ASReml [16]: chose to conduct the subsequent analyses. The descriptive statis[16]: 121Genetic correlation120 coefficients 119 were Genetic estimated correlation by the coefficients following bivariate were estimated linear mixed-effects by of thethe following model bivariate using ASReml linear mixed-effects model using ASReml tics growth curve parameter values (i.e., the asymptotic live 𝐞𝐞𝟏𝟏 𝟎𝟎 𝐚𝐚𝟏𝟏 body weight α [grams], the scaling parameter β [wk], and the 121 𝐞𝐞𝟏𝟏 𝐗𝐗 𝐙𝐙 𝟎𝟎 𝐛𝐛 𝟎𝟎 𝐚𝐚𝟏𝟏 ] [ ] + [𝐞𝐞𝐲𝐲𝟏𝟏] [16]: 𝐙𝐙𝟐𝟐 𝐚𝐚𝟐𝟐 122 [ 𝟐𝟐 ] = [ 𝟏𝟏 120] [ 𝟏𝟏[16]: ] + [ 𝟏𝟏 ] [ ] + [𝐞𝐞 ] 𝟎𝟎 𝐗𝐗 𝟐𝟐 𝐛𝐛𝟐𝟐 𝟎𝟎 𝐙𝐙𝟐𝟐 𝐚𝐚𝟐𝟐 intrinsic growth rate γ [wk]) estimated from the logistic growth 𝟐𝟐 𝐲𝐲𝟐𝟐 𝐲𝐲𝟏𝟏 𝐞𝐞𝟏𝟏 𝐗𝐗 𝐙𝐙 𝟎𝟎 𝐛𝐛𝟏𝟏 𝟎𝟎 𝐚𝐚𝟏𝟏 122121 [𝐲𝐲 ] = [ 𝟏𝟏 ] [ ] + [ 𝟏𝟏 ] [𝐚𝐚 ] + [𝐞𝐞 ] curve function are summarized in Table 3. The growth curves 𝟎𝟎 𝐗𝐗 𝟐𝟐 𝐛𝐛𝟐𝟐 𝟎𝟎 𝐙𝐙𝟐𝟐 𝟐𝟐 𝟐𝟐 𝟐𝟐 123 Where, y1 and y2 are vectors of measured phenotypes for the of KNC individuals were plotted from actual longitudinal body 𝐲𝐲𝟏𝟏 𝐞𝐞𝟏𝟏 𝟎𝟎 𝐛𝐛𝟏𝟏 b1 and 𝐞𝐞of 𝐗𝐗 𝟏𝟏 𝟎𝟎 𝐛𝐛 𝐙𝐙𝟏𝟏 traits 𝐗𝐗 𝐙𝐙𝟏𝟏 b2𝟎𝟎are 𝐚𝐚vectors 𝟎𝟎 𝐲𝐲𝐚𝐚under 𝟏𝟏𝟏𝟏 𝟏𝟏 𝟏𝟏 𝟏𝟏 the two of measured phenotypes for consideration; 123 [ Where, ] [ ]122 + ][[ b] ]=and +[[𝐞𝐞b2] are vectors ] [ ] +of[fixed ] [two] traits + [𝐞𝐞 under ]weight traits under consideration; effects data from hatching 20vectors weeks of of age (Supplementary 124[𝐲𝐲𝟐𝟐 ] =two y2[are consideration; b1 and b2toare 𝟎𝟎 𝐗𝐗 𝟐𝟐y1 𝐛𝐛and 𝟎𝟎 vectors 𝐙𝐙𝟐𝟐 𝐲𝐲𝐚𝐚𝟐𝟐𝟐𝟐of1 measured 𝟎𝟎𝟐𝟐 𝐗𝐗 𝟐𝟐phenotypes 𝟎𝟎for 𝐙𝐙the 𝐛𝐛𝟐𝟐 𝟐𝟐 𝟐𝟐 𝐚𝐚𝟐𝟐 𝟐𝟐 (i.e., sex, batch, and line) for the traits; a and a are vectors of Figure S1). 1 2 line) for the traits; a1 and a2 are vectors ofythe random additive of genetic effects for the for the two traits under consideration; b1 and b2 are vectors of Where, y2 are measured phenotypes 1 andfor 125 fixed effects 124 (i.e., sex, batch, and line) the vectors traits; of the random additive genetic effects for the 2 are vectors 123 the random additive genetic effects for the traits; aX1 1and andaX 2 are ce matrices relating measures of the traits to fixed effects; and Z2 are the incidence 1 and the incidence relating measures ofZthe traits toforfixed estimates of weight gain and growth curve 125 effects (i.e., sex, batch, line) traits; ato and a2 effects; are vectors of the additive genetic effects for the 1Heritability 126Where, traits; X1 yand X2matrices are fixed theofincidence measures ofthe the traits fixed Z1btraits and Zvectors are the 2 random y1 and 124 measured Where,matrices phenotypes y1 and yrelating forvectors the two oftraits measured under phenotypes consideration; for the b1 and two consideration; ofincidence b1 and b2 are vectors of 2 are vectors 2 are 2 are under effects; Z1 and Z2 are the incidence matrices relating phenotypic parameters servations with random 126 additivetraits; genetic e1 and e2 are matrices vectors of random X1 effects; and X2 and are the incidence relating measures of ethe traits effects; Z1ofand Z2 are theand incidence 127 effects matrices relating phenotypic with random additive genetic effects; and e2 to arefixed of random observations with random additive genetic effects; and e1 and Table 4and shows the heritability estimates weight gain growth fixed (i.e., sex, batch, 125 and line) fixedobservations for effects the traits; (i.e., sex, a1 and batch, a2 are and vectors line) for of the random traits; a1additive and a2 1are genetic vectors effects ofvectors the for random the additive genetic effects for the 6 are vectors of random residuals. The expectation and variance curve parameter traits in KNC. The heritability estimates of GR0-2 riance of theebivariate model were: 6 2 127 matrices relating phenotypic observations with random additive genetic effects; and e1 and e2 are vectors of random 128 X1 residuals. The and variance of bivariate were: traits; and X2 bivariate are theexpectation incidence 126 traits; matrices X1 and relating X2the are measures the incidence the traits matrices to fixed relating effects; measures Z1 and ofZthe traits the to incidence fixed effects; Z1 and Z2 areand theGR6-8 incidence 2 are 6ofmodel of the model were: were high estimates of 0.458, whereas GR2-4 6 with 128 residuals. The expectation and variance of the bivariate were modelmoderate were: 129 with estimates of 0.266 0.233 respectively. matrices relating phenotypic 127observations matrices with relating random phenotypic additiveobservations genetic effects; with and random e1 and additive e2 are genetic vectors effects; of random and e1 andand e2 are vectors of random The heritabilities of GR4-6, GR8-10, and GR10-12 were low with 𝐲𝐲 𝐗𝐗 𝐛𝐛 𝟎𝟎 𝟏𝟏 132And 𝟏𝟏 𝐲𝐲𝟏𝟏 𝐗𝐗 𝐛𝐛 𝟎𝟎of the EAnd [𝐲𝐲 ] = 129 [ 𝟏𝟏128 ] [ residuals. ] of theThe residuals. The bivariate expectation model were: model were: 𝟎𝟎 and𝐗𝐗variance 130 E [𝐲𝐲and ] =variance [ 𝟏𝟏 ] [ 𝟏𝟏bivariate ] heritability 𝟐𝟐 estimates of 0.161, 0.088, and 0.058 respectively. Heri𝟐𝟐 𝐛𝐛𝟐𝟐 expectation 𝟎𝟎 𝐗𝐗 𝟐𝟐 𝐛𝐛𝟐𝟐 𝟐𝟐 133 𝐲𝐲𝟏𝟏 tability 𝐗𝐗 estimates 𝟎𝟎 𝐛𝐛𝟏𝟏 of GR16-18 and GR18-20 were comparable 130129 E [𝐲𝐲 ] = [ 𝟏𝟏 ][ ] 𝟎𝟎 𝐗𝐗 𝟐𝟐 𝟐𝟐 𝐛𝐛𝟐𝟐for GR10-12. Moderate heritability for growth 131 And with the value 𝟐𝟐 𝟐𝟐 𝐚𝐚𝟏𝟏𝟐𝟐 𝐀𝐀𝛔𝛔𝟐𝟐𝒂𝒂𝒂𝒂𝒂𝒂 𝐀𝐀𝛔𝛔 𝟎𝟎 𝟎𝟎 𝐲𝐲 𝟎𝟎 𝐚𝐚𝟏𝟏 𝐀𝐀𝛔𝛔 𝐲𝐲𝐀𝐀𝛔𝛔 𝐗𝐗 𝟏𝟏 𝟎𝟎 traits 𝐛𝐛𝟏𝟏 α and β were observed with estimates of 𝟎𝟎 𝐛𝐛 𝟎𝟎 𝒂𝒂𝒂𝒂𝒂𝒂 𝒂𝒂𝒂𝒂𝒂𝒂 curve parameter 𝟏𝟏𝐚𝐚 𝒂𝒂𝒂𝒂𝒂𝒂 𝐗𝐗 𝟏𝟏 𝟏𝟏 𝟐𝟐 ] [ 𝟏𝟏𝟎𝟎 𝟐𝟐 𝟐𝟐 𝟎𝟎 𝟎𝟎 131 𝟐𝟐 130 E [ ] = [ ] 𝟎𝟎 𝟎𝟎𝟐𝟐 E [𝐲𝐲 ]𝟎𝟎𝟐𝟐= ][ 𝟎𝟎 𝐗𝐗 ] [𝐛𝐛 ] 𝟐𝟐 𝟐𝟐 𝐀𝐀𝛔𝛔 𝐀𝐀𝛔𝛔 𝐀𝐀𝛔𝛔 𝐚𝐚 𝟎𝟎 Var𝐲𝐲𝐀𝐀𝛔𝛔 𝐚𝐚𝟏𝟏 𝒂𝒂𝒂𝒂𝒂𝒂 [𝟐𝟐 𝟎𝟎𝟐𝟐𝒂𝒂𝒂𝒂𝒂𝒂 ] = 𝟎𝟎[𝐀𝐀𝛔𝛔𝟐𝟐𝒂𝒂𝒂𝒂𝒂𝒂 𝒂𝒂𝒂𝒂 134 𝒂𝒂𝒂𝒂𝒂𝒂 𝐀𝐀𝛔𝛔𝒂𝒂𝒂𝒂𝒂𝒂 𝐀𝐀𝛔𝛔Var 𝐗𝐗 𝟐𝟐 𝐛𝐛𝟐𝟐𝟐𝟐 𝒂𝒂𝒂𝒂𝒂𝒂 𝐈𝐈𝛔𝛔 𝟐𝟐 𝒂𝒂𝒂𝒂𝒂𝒂 [ 𝟐𝟐 ] = 𝟐𝟐 respectively. 𝟐𝟐 𝐈𝐈𝛔𝛔 [ ] 𝟐𝟐 0.211 and 0.249, The heritability estimate for γ was 𝐞𝐞 𝒆𝒆𝒆𝒆𝒆𝒆 𝒆𝒆𝒆𝒆𝒆𝒆 𝟎𝟎 𝟎𝟎 𝐈𝐈𝛔𝛔𝒆𝒆𝒆𝒆𝒆𝒆 𝐈𝐈𝛔𝛔𝒆𝒆𝒆𝒆𝒆𝒆 𝟎𝟎 𝟎𝟎𝟏𝟏 𝐞𝐞 𝟎𝟎 𝐚𝐚𝟐𝟐 𝟐𝟐 ] 𝐀𝐀𝛔𝛔𝟐𝟐 𝐀𝐀𝛔𝛔𝟐𝟐𝒂𝒂𝒂𝒂𝒂𝒂 𝟎𝟎𝟎𝟎 𝒂𝒂𝒂𝒂 𝟐𝟐 𝟐𝟐 𝐀𝐀𝛔𝛔𝟐𝟐𝒂𝒂𝒂𝒂𝒂𝒂 𝟏𝟏 𝟐𝟐 𝟐𝟐 𝐞𝐞𝟎𝟎𝟐𝟐𝟏𝟏 ] 𝒂𝒂𝒂𝒂𝒂𝒂 [ ] = [ 𝐈𝐈𝛔𝛔 𝐈𝐈𝛔𝛔 𝐈𝐈𝛔𝛔Var 𝐈𝐈𝛔𝛔 𝟐𝟐 𝟐𝟐 low (0.095). 𝟎𝟎 𝟎𝟎 𝐞𝐞 𝒆𝒆𝒆𝒆𝟏𝟏 𝒆𝒆𝒆𝒆𝒆𝒆 𝒆𝒆𝒆𝒆𝒆𝒆 𝐞𝐞 𝒆𝒆𝒆𝒆𝒆𝒆 𝐈𝐈𝛔𝛔 𝟏𝟏 𝒆𝒆𝒆𝒆𝒆𝒆 𝐈𝐈𝛔𝛔𝟎𝟎 𝐈𝐈𝛔𝛔𝒆𝒆𝒆𝒆𝟏𝟏 𝐈𝐈𝛔𝛔𝒆𝒆𝒆𝒆𝒆𝒆 𝒆𝒆𝒆𝒆𝒆𝒆 𝟎𝟎 𝟎𝟎 131 𝟏𝟏 𝟎𝟎 𝟎𝟎 𝟐𝟐 𝟐𝟐 𝟎𝟎 𝟏𝟏 𝒆𝒆𝒆𝒆𝒆𝒆 𝐈𝐈𝛔𝛔𝟐𝟐𝒆𝒆𝒆𝒆𝟏𝟏 𝐈𝐈𝛔𝛔𝟐𝟐𝒆𝒆𝒆𝒆𝒆𝒆 𝐈𝐈𝛔𝛔𝒆𝒆𝒆𝒆𝟏𝟏 𝐞𝐞𝐈𝐈𝛔𝛔 𝟎𝟎 𝟎𝟎 135 Table 3. Descriptive analysis of growth curve parameter related traits (α, β, γ)

2 2 2 2 136Where, Where, A:the theadditive additiverelationship geneticrelationship relationship matrix; A: genetic 2 2 𝜎𝜎𝑎𝑎11 2 , 𝜎𝜎𝑎𝑎12 2 , 𝜎𝜎𝑎𝑎21 , 𝜎𝜎𝑎𝑎22 are additive genetic (co)variances for the A:Where, the additive genetic matrix; 𝜎𝜎matrix; 𝑎𝑎11 , 𝜎𝜎𝑎𝑎12 , 𝜎𝜎𝑎𝑎21 , 𝜎𝜎𝑎𝑎22 are additive genetic (co)variances for the 2 2 2 2 Trait N Mean SE Min. 2 2 2 2 ; 𝜎𝜎𝑎𝑎11 ,genetic 𝜎𝜎𝑎𝑎12 , 𝜎𝜎𝑎𝑎21 , 𝜎𝜎𝑎𝑎22 arematrix; additive𝜎𝜎𝑎𝑎11 genetic for additive genetic (co)variances forthe the traits under ditive relationship , 𝜎𝜎2𝑎𝑎12 ,(co)variances 𝜎𝜎2𝑎𝑎21 , 𝜎𝜎2𝑎𝑎22 are 2 additive genetic (co)variances for the 137 traits under consideration; the traits; and I is the identity matrix. a and 2 2 𝜎𝜎𝑒𝑒11 2, 𝜎𝜎𝑒𝑒12 2 , 𝜎𝜎𝑒𝑒21 , 𝜎𝜎𝑒𝑒22 are residual (co)variances for UT traits under consideration; 𝜎𝜎𝑒𝑒11 , 𝜎𝜎𝑒𝑒12 , 𝜎𝜎𝑒𝑒21 , 𝜎𝜎𝑒𝑒22 are for the traits; and I is the identity matrix. a and consideration; areresidual residual(co)variances (co)variances 2 2 2 for the 2 dual𝜎𝜎(co)variances traits; and I (co)variances isidentity the identity matrix. a and α (g) 557 2,180.20 22.4 1,111.70 ion; , 𝜎𝜎the 𝜎𝜎𝑒𝑒22 are residual for the traits; and Iassumed is the identity matrix. a and 𝑒𝑒11 , 𝜎𝜎𝑒𝑒12 𝑒𝑒21 ,traits; for and I is the matrix. a and e were e were to assumed to be normally and (co)variances, as statedPhenotypic above. Phenotypic correlation e138 were assumed be normally distributeddistributed with zerowith meanzero andmean (co)variances, as stated above. correlation β (wk) 560 19.52 0.34 8.79 to be normally distributed with zero mean and (co)variances, as mean and (co)variances, stated above. correlation normally distributed withaszero mean and Phenotypic (co)variances, as stated above. Phenotypic correlation γ (wk) 567 0.23 0.01 0.14 139 coefficients were computed using theprogram Minitab program stated Phenotypic coefficients were[14]. computed coefficients wereabove. computed using thecorrelation Minitab [14]. NLT [14]. using the puted Minitab program program [14]. using the Minitab [14]. α (g) 567 7.66 0.01 7.01 140

141 RESULTS RESULTS RESULTS 142

Growth curve determination and descriptive statistics

143 curve Growth curve determination and descriptive Growth determination and descriptive statistics statistics

β (wk)

560

2.90

0.02

2.17

Max. 4,653.10 53.82 0.35 8.44 3.99

UT, untransformed data; α, asymptotic live body weight (g); β, inflection point at which 50% of the asymptotic weight is realized (wk); NLT, natural log transformed values; the inflection point age at t i and weight W(t i) are given by log e(β)/α and α/2, respectively.

cs mination andTable descriptive statistics 2. Estimated parameters of growth curve using three non-linear models in Korean native chicken 144evaluated We evaluated three widely used growth curve models (i.e., Gompertz, von Bertalanffy, andmodels) logistictomodels) to fit the growth We three widely used growth curve models (i.e., Gompertz, von Bertalanffy, and logistic fit the growth Parameter Convergence Convergence ., Gompertz, voncurve Bertalanffy, models) fit the growth dely used growth modelsand (i.e.,logistic Gompertz, von to Bertalanffy, and logistic models) to fit the growth Model R2 the three models performed 2 145 curve of the KNC population from hatching to 20 weeks of age (Table 2). In terms of R values, 2 property weight α±SE β±SE γ±SE (g) (wk) (wk) curve of the KNC population from hatching to 20 weeks of age (Table 2). In terms of R values, the three models performed ulation to 20ofweeks of agethe (Table In terms of R2 values, the three models performed of age from (Tablehatching 2). In terms R2 values, three2). models performed Gompertz 2,798.9 ± 76.2 3.562However, ± 0.04 0.102 ± curves 0.003 showed the 99.55 2 in the 48 146 similarly thecurve growth curve ofHowever, KNCs. thecurves logistic best performance both similarly in fitting in thefitting growth of KNCs. the logistic showed the best performance both in the von Bertalanffy 3,755.4 ± 168.7 0.750 ± 0.00 0.061 ± 0.003 99.57 30 138 growth curve of KNCs. However, the logistic curvesboth showed ver, the logistic curves showed the best performance in thethe best performance both in the Logistic 2,076.3 ± 27.5 16.09 ± 0.41 0.222the ± 0.004 99.45 to conduct the subsequent 1 3 147 convergence and the convergence weight logistic model chose convergence propertiesproperties and the convergence weight (Table 2).(Table Thus, 2). the Thus, logistic model chose to conduct the subsequent s and2).theThus, convergence weight 2).β, the logistic chose conductmature the subsequent able logistic chose toThus, conduct the for subsequent α, the asymptotic live model body (Table weight (g); the log-function themodel proportion of thetoasymptotic weight to be gain after birth (wk); γ, constant scale that is proportional to the overall 2 148 analyses. statistics of thecurve growth curveproperties parameter (i.e., asymptotic body weight [grams], Convergence weight can be growth rate The (wk); Rdescriptive is the coefficient of determination; Convergence bevalues defined as thethe number of chickens which theαmodel is notα converged; analyses. The descriptive statistics of the growth parameter valuescan(i.e., the asymptotic live body live weight [grams], ive statistics of the(i.e., growth curve parameter values (i.e., the asymptotic liveofbody α [grams], parameter values asymptotic live body weight α [grams], defined as thethe number of chickens for which the converged weight at 20 weeks age isweight 5 times heavier than actual weight. 149scalingtheparameter scaling parameter β [wk], and the growth intrinsicrate growth rateestimated γ [wk]) estimated from thegrowth logisticcurve growth curve are function are the β [wk], and the intrinsic γ [wk]) from the logistic function βate[wk], and estimated the intrinsic growth rate γ [wk]) estimated from theare logistic growth curve function are γ [wk]) from the logistic growth curve function 150 summarized in The Table 3. Thecurves growth of KNC individuals werefrom plotted from actual longitudinal bodydata weight data www.ajas.info 3 summarized in Table 3. growth of curves KNC individuals were plotted actual longitudinal body weight 3. The growth KNC individuals were plotted from actual dividuals were curves plottedoffrom actual longitudinal body weight data longitudinal body weight data 151 from hatching to 20 weeks of age (Supplementary Figure S1).


Table 4. Number of animals, mean and standard deviation, and heritability estimates for weight gain and growth curve parameter traits of Korean native chicken Trait1)

n

GR0-2 GR2-4 GR4-6 GR6-8 GR8-10 GR10-12 GR12-14 GR14-16 GR16-18 GR18-20 α β γ

578 584 584 583 581 582 582 576 582 583 575 568 565

Va

Ve

h ²±SE

102.77 179.90 164.87 406.03 352.10 336.88 246.74 584.16 326.57 258.98 0.005 0.010 6.16e-05

121.68 497.08 859.48 1,334.34 3,667.61 5,453.45 5,626.51 3,964.95 4,537.18 4,314.20 0.019 0.029 5.84e-04

0.458 ± 0.117 0.266 ± 0.098 0.161 ± 0.078 0.233 ± 0.090 0.088 ± 0.058 0.058 ± 0.056 0.042 ± 0.046 0.128 ± 0.067 0.067 ± 0.050 0.057 ± 0.052 0.211 ± 0.084 0.249 ± 0.091 0.095 ± 0.060

Mean±SD 105.36 ± 23.49 122.88 ± 48.84 161.22 ± 66.29 180.44 ± 71.82 160.32 ± 69.04 223.75 ± 93.35 188.05 ± 90.38 209.20 ± 77.20 206.44 ± 109.64 191.15 ± 72.89 2,178.9 ± 528.1 19.49 ± 8.09 0.230 ± 0.032

n, number of animals analyzed; Mean ± SD, mean with standard deviation; Va, additive genetic component; Ve, residual variance component; h² ± SE, heritability with standard error. 1) GR0-2, weight gain from hatch to 2 weeks of age; GR2-4, weight gain from 2 to 4 weeks of age; GR4-6, weight gain from 4 to 6 weeks of age; GR6-8, weight gain from 6 to 8 weeks of age; GR8-10, weight gain from 8 to 10 weeks of age; GR10-12, weight gain from 10 to 12 weeks of age; GR12-14, weight gain from 12 to 14 weeks of age; GR14-16, weight gain from 14 to 16 weeks of age; GR16-18, weight gain from 16 to 18 weeks of age; GR18-20, weight gain from 18 to 20 weeks of age; α, asymptotic live body weight (gram); β, inflection point at which 50% of the asymptotic weight is realized (wk), γ, a constant scale that is proportional to the overall growth rate (wk).

Genetic and phenotypic correlations Table 5 describe genetic and phenotypic correlation coefficients that were estimated from the growth curve parameters and weight gain traits. Most of the genetic correlation coefficients among the growth curve parameters and weight gain traits were higher compared to those for the phenotypic parameter correlation. The range of coefficients was between –0.968 and 0.993. The genetic correlations between early weight gain traits were high and positive, whereas it becomes low and negative at later weight gain. Genetic correlations between α and weight gain traits from GR10-12 to GR 18-20 were strong and positive. Similarly, genetic correlations between β and weight gain traits from GR10-12 to GR18-20 were positive and ranged between 0.214 to 0.723. Cor-

relation results between growth curve parameters are shown in Table 5. Genetic and phenotypic correlations between α and β were strong and positive (0.582 and 0.696, respectively). Negative phenotypic and genetic correlations were observed between β and γ (–0.264 and –0.370, respectively). Negative genetic and phenotypic correlations between α and γ were recorded.

DISCUSSION Growth can be characterized by size increment or gain at a given age or from an initial size. Therefore, weight gain can be defined as the difference between measurement at the beginning and end of the time interval. In our study, we measured body weight

Table 5. Genetic (above the diagonal), phenotypic (below the diagonal) correlation between weight gain traits and parameters of the growth curve (GR0-2 to GR18-20, α, β, γ)1) Trait2)

GR0-2

GR2-4

GR4-6

GR6-8

GR8-10

GR10-12 GR12-14 GR14-16 GR16-18 GR18-20

GR0-2 GR2-4 GR4-6 GR6-8 GR8-10 GR10-12 GR12-14 GR14-16 GR16-18 GR18-20 α β γ

0.808 0.754 0.743 0.213 0.401 –0.05 0.188 –0.298 0.115 0.399 –0.590 0.073

0.434 0.812 0.777 0.224 0.456 –0.03 0.192 –0.314 0.021 0.463 –0.665 0.021

0.723 0.742 0.799 0.216 0.498 –0.024 0.244 –0.365 0.099 0.576 –0.624 0.006

0.545 0.550 0.920 0.208 0.504 0.002 0.220 –0.371 0.099 0.581 –0.569 0.005

0.615 0.502 0.652 0.417 –0.011 0.190 0.280 0.223 0.091 0.251 0.096 0.297

0.717 0.706 0.939 0.971 0.981 0.079 0.224 –0.060 0.140 0.483 –0.071 0.219

1)

–0.258 0.295 0.993 0.083 0.507 0.730 0.228 0.367 0.150 0.031 0.475 0.544

–0.140 –0.300 0.270 0.208 0.143 0.605 0.780 0.194 0.193 0.014 0.231 0.542

0.050 –0.527 0.611 0.198 0.329 0.416 0.850 0.888 0.190 0.550 0.649 0.711

0.130 –0.241 –0.517 –0.228 –0.235 –0.385 –0.277 –0.271 0.021 –0.335 0.190 0.514

α

β

γ

0.063 –0.298 0.177 0.288 –0.277 0.687 0.936 0.772 0.930 0.584 0.696 –0.560

–0.667 –0.657 –0.393 –0.264 –0.485 0.214 0.369 0.723 0.311 0.528 –0.264

0.198 0.654 0.540 0.987 0.671 0.542 0.183 –0.210 –0.543 –0.968 –0.388 –0.370 -

Genetic correlations are above the diagonal and phenotypic correlations are below the diagonal, Traits: weight gain traits and growth curve parameter traits. GR0-2, weight gain from hatch to 2 weeks of age; GR2-4, weight gain from 2 to 4 weeks of age; GR4-6, weight gain from 4 to 6 weeks of age; GR6-8, weight gain from 6 to 8 weeks of age; GR8-10, weight gain from 8 to 10 weeks of age; GR10-12, weight gain from 10 to 12 weeks of age; GR12-14, weight gain from 12 to 14 weeks of age; GR14-16, weight gain from 14 to 16 weeks of age; GR16-18, weight gain from 16 to 18 weeks of age; GR18-20, weight gain from 18 to 20 weeks of age; α, asymptotic live body weight (g); β, inflection point at which 50% of the asymptotic weight is realized (wk); γ, constant scale that is proportional to the overall growth rate (wk). 2)

www.ajas.info 4


gain at 2-week intervals until 20 weeks of age. Similar to Aslam et al [5] and Ngeno et al [17], we considered three growth curve parameters (i.e., α, β, and γ) as a single trait to estimate genetic parameters in the moderate sized KNC population. In terms of the convergence property and weight, the logistic model described the growth pattern of KNC better than did the other two models (Table 1). Thus, the logistic growth curve function was selected to estimate growth curve parameters for further analyses. Heritability estimates for body weight of our study population were reported by Cahyadi et al [13] and were in agreement with those reported in previous studies [18-21]. Furthermore, the estimated heritability values for body weight gain traits were lower than those reported for body weight traits. The highest heritability estimate for weight gain was exhibited at GR0-2 weeks. The trend in heritability observed for weight gain for this KNC population was similar to that reported for body weight in previous studies. Heritability estimates for body weight related traits exhibited a decreasing trend with increasing age in the KNC population. Similar results were observed by Adeyinka et al [22] and Saatchi et al [23]. The α (asymptotic mature weight) showed a moderate heritability that was inconsistent with the results of MignonGrasteau et al [7], Aslam et al [5], and Narinc et al [24]. The moderate heritability for α indicated that some genetic gain through selection could be achieved. In addition, we found that the low heritability estimates for β and γ were consistent with the results of Aslam et al [5] and Narinc et al [24] for turkey and Japanese quail, respectively. Akbas and Yaylak [25] reported that birds with high α values had high hatch weights. However, we observed only a moderate level of genetic correlation between GR0-2 and α. The genetic correlations between body weight gain and β had different signs and/or magnitude. Strong negative genetic and phenotypic correlations were observed for GR0-2 to GR8-10, which become positive at later stages of growth, similar to that reported by Barbato [26] in mice and Mignon-Grasteau et al [7] in chickens. The positive correlation between α and β was in agreement with correlation estimates reported for Japanese quail growth using the Gompertz model [24]. These results indicate that individuals with high asymptotic body weight will longer to reach β. The parameters β and γ exhibited moderate negative phenotypic and genetic correlation in contrast with Aslam et al [5], where they obtained a positive correlation (0.64) between these para meters in turkey. The genetic correlation between α and γ was –0.388, whereas the phenotypic correlation was –0.560. The negative correlation between α and scaled γ indicated that if selection was used to increase asymptotic body weight there would be a negative effect on the scaling parameter γ. Heritability estimates of weight gain of the KNC population were high and moderate during the juvenile stage and decreased with age. Previously we reported the genetic parameters using cross-sectionally measured body weight data (e.g. body weight at 6 weeks of age, body weight at 8 weeks of age) in KNC [13].

Compared to the previous study, we analyzed weight gain data given two-week of period (e.g., GR6-8) in this study. As a result, we could clearly differentiate high or low heritability (Table 4). High heritabilities were observed from early stage of growth, which are likely to reflect genetic variability. On the contrary, low heritabilities were observed from late stage of growth, which are likely to reflect genetic invariability. For the genetic phenotypic correlation coefficients, analysis of weight gain data allowed us to detect both positive and negative signs which could not be seen the previous study [13]. The growth curve parameter was heritable with a low to moderate estimate, and low genetic and phenotypic correlation with weight gain, and consequently the alteration of the growth curve of KNC by selection. The results of estimation of heritability suggest that the KNC could be used for selective breeding at juvenile stages, between 6 and 8 weeks of age, to increase the genetic improvement of growth performance traits. The heritability for the period between 6 and 8 weeks of age was not highest (Table 4). In fact, the period between 2 and 4 weeks of age showed maximum heritability. However, the standard deviation of the period between 6 and 8 weeks of age is larger than that of the period between 2 and 4 weeks of age (Table 4). Thus, given the selection intensity, we can anticipate the maximum expected genetic gain based on the individual selection at the period between 6 and 8 weeks of age. The findings of this study could provide useful information for further optimization of breeding plans for KNC after verification of these results in independent studies.

CONFLICT OF INTEREST We certify that there is no conflict of interest with any financial organization regarding the material discussed in the manuscript.

ACKNOWLEDGMENTS This work was supported by Korea Institute of Planning and Evaluation for Technology in Food, Agriculture, Forestry and Fisheries (IPET) through Golden Seed Project, funded by Ministry of Agriculture, Food and Rural Affairs (MAFRA) (213010-05-1-SB250) and "Cooperative Research Program for Agriculture Science & Technology Development (Project No. PJ012820052017)" Rural Development Administration, Republic of Korea.

REFERENCES 1. FAO. FAO STAT database collection. Rome, Italy: FAO; 2009. 2. Ricklefs RE. Modification of growth and development of muscle of poultry. Poult Sci 1985;64:1563-7. 3. Laird AK. Post-natal growth of birds and mammals. Growth 1966;30: 349-63. 4. Aggey SE. Comparison of three nonlinear and spline regression models for describing chicken growth curves. Poult Sci 2002;73: www.ajas.info 5


1822-6. 5. Aslam ML, Bastiaansen JWM, Crooijmans RPMA, et al. Genetic variance, heritability and maternal effect on body weight, breast meat yield, meat quality and the shape of the growth curve in turkey birds. BMC Genet 2013;12:14. 6. Grossman M, Bohren BB. Logistic growth curve of chickens: heri tability of parameters. J Hered 1985;76:459-62. 7. Mignon–Grasteau S, Beaumont C, Le Bihan–Duval E, et al. Genetic parameter of growth curve parameter in male and female chickens. Br Poult Sci 1999;40:44-51. 8. Nestor KE, Anderson JW, Patterson RA. Genetics of growth and reproduction in turkey. Changes in genetic parameters over thirty generation of selection for increased body weight. Poult Sci 2000;79: 445-52. 9. Lambe NR, Navajas EA, Simm G, Bünger L. A genetic investigation of various growth models to describe growth of lambs of two con trasting breeds. J Anim Sci 2006;84:2642-54. 10. Jin S, Park HB, Seo DW, et al. Association of MCR1 genotype with shank color traits in Korean native chicken. Livest Sci 2014;170:1-7. 11. Jeon HJ, Choe JH, Jung YK, et al. Comparison of the chemical com position, textural characteristics, and sensory properties of North and South Korean native chickens and commercial broilers. Korean J Food Sci An 2010;30:171-8. 12. Cho ES, Chung WH, Choi JW, et al. Genome – wide copy number variation in a Korean native chicken breed. Korean J Poult Sci 2014; 41:305-11. 13. Cahyadi M, Park HB, Seo DW, et al. Genetic parameters for growthrelated traits in Korean native chicken. Korean J Poult Sci 2015;42: 285-9. 14. Minitab 15 User’s guide. Version 15. State College, PA, USA: Minitab. Inc.; 2007. 15. SAS/STAT User’s Guide, The SAS System Version 9.1, Cary, NC, USA: SAS Institute Inc.; 2004.

www.ajas.info 6

16. Gilmour AR, Gogel BJ, Cullis BR, Thompson R. ASReml user guide release 2.0, Hemel Hempstead, UK: VSN International Ltd; 2006. 17. Ngeno K, Magothe TM, Okeno TO, Bebe BO, Kahi AK. Heritability and correlation between body weight and Growth curve parameter of Indigenous Chicken Population reared intensively in Kenya. Res J Poult Sci 2013;6:43-52. 18. Chabault M, Baeza E, Gigaud V, et al. Analysis of a slow growing line reveals wide genetic variability of carcass and meat quality related traits. BMC Genet 2012;13:90-100. 19. Dana N, Vander Waaji EH, Van Arendonk JAM. Genetic and Pheno typic parameter estimate for body weight and egg production in Horro chicken of Ethiopia. Trop Anim Health Prod 2011;43:21-8. 20. Ghazikhani Shad A, Javaremi AN, Yeganeh HM. Animal model estimation of genetic parameter for most important economic traits in Iranian native fowls. Pakistan J Bio Sci 2007;10:2787-9. 21. Niknafs S, Nejati-Javaremi A, Mehrabani-Yeganeh H, Fatemi SA. Estimation of genetic parameters for body weight and egg produc tion traits in Mazandaran native chicken. Trop Anim Health Prod 2012;44:1437-43. 22. Adeyinka, IA, Oni OO, Nwagu BI, Adeyinka FD. Genetic parameter estimates for Body weights of Naked Neck broiler chickens. Int J Poult Sci 2006;5:589-92. 23. Saatchi M, Omed H, Dewi IA. Genetic parameter from univariate and bivariate analysis of egg and weight trait in Japanese quails. Poult Sci 2006;85:185-90. 24. Narinc D, Aksoy T, Karaman E. Genetic parameter of growth curve parameter and weekly body weight in Japanese quails (Coturnix coturnix japonica). J Anim Vet Adv 2010;9:501-7. 25. Akbas Y, Yaylak E. Heritability estimates of growth curve parameter and genetic correlation between the growth curve parameter and weights at different age of Japanese quail. Eur Poult Sci 2000;64:141-6. 26. Barbato GF. Genetic architecture of growth curve parameters in chickens. Theor Appl Genet 1991;83:24-32.