Deprtment of Animal Production, College of Agriculture, University of Duhok, Kurdistan, Iraq
Receive Date: 13 October 2014,
Revise Date: 12 January 2015,
Accept Date: 30 January 2015
The present studyinvestigated the impact of selection for body weight (BW) on productive performance, predicted BW and genetic improvement of related traits in Japanese quail. Base population, selected parent and F1 progeny were tested. Pens and cages were used for rearing and breeding. A total of 240 birds in two successive generations, 27 sires and 54 dams (as selected parents) were used. Body weight from multiple regression equation, covariance analysis and combining models were estimated, response to selection, realized heritability and genetic correlation were computed for BW, weight gain (WG), feed intake (FI) and feed conversion ratio (FCR). The results showed that the response to selection for BW, WG, FI and FCR were 11.48 g, 27.04 g, 37 g and -0.2, respectively. The estimated heritabilities for the same traits were 0.78, 0.67, 0.52 and 0.77, respectively. Predicted equation (DUHOK equation) for BW as dependent variable on the initial body weight and sex ratio was derived. The final response for the body weight was determined as 5.84 % of live BW.
In many European and Asian countries, Japanese quail is considered one of the main sources for meat production (Minvielle, 1998; Kayang et al. 2004; Ayasan, 2013). Thus body weight (BW) is a very important trait beside feed conversion ratio (FCR) to be improved from economical point of view (Kayang et al. 2004; Kaur et al. 2008; Baylan et al. 2009). It was indicated that selection for BW and FCR in Japanese quail improved the weight gain for the next generation (Kayang et al. 2004; Kaur et al. 2008; Baylan et al. 2009). Moreover, the accumulative response to selection for FCR was estimated to be 0.52 (Varkoohi et al. 2010), and the mass selection together with the differences in initial body weight resulted in an increase of body weight (Caron and Minvielle, 1990; Darden and Marks, 1989). Marks (1993a) concluded that the body weight differences ranged from (15 to 24%) at 4 weeks of age. Karaman et al. (2013) found that the live body weight of quail males at 1, 21 and 42 days old were 8.75, 111.42 and 178.13 gm., respectively and for females at the same previous ages were 9.07, 118.97 and 219.07 gm., respectively. The same authors derived 3 different fitted models for growthbynonlinear mixed effects modeling and they concluded that the best model has reduced variance by about 37%. In an experiment conducted by Muir et al. (2013) who compares multilevel selection for 43 days weight and survival with Japanese quail birds housed either in kin (K) or random (R) groups. They observed that multilevel selection has significantly increased weight and response in K more than R. In another study conducted by Vali (2009) to compare two strains of Japanese quail and their reciprocal crosses, it was found that the body weight at different ages (from 1 up to 63 days old) were highly significant between males and females for all studied groups. Also feed intake of just one hybrid significantly surpassed other groups, while feed efficiency did not differ significantly. Estimated heritability for body weight at 28 days of age and weight gain between 7 to 28 days of age were 0.22 and 0.28, respectively and the genetic correlation between FCR and each of WG and FI were -0.45 and 0.24, respectively (Varkoohi et al. 2011). The aim of the present work was to investigate the impact of selection for body weight on productive performance, predicted BW and genetic improvement of related traits in Japanese quail.
MATERIALS AND METHODS
The present work was conducted at poultry production farm of the Department of Animal Production, Faculty of Agriculture and Forestry, University of Duhok, Kurdistan Region, Iraq during 2014.
Rearing both generations and selection
A total of 240 Japanese quail birds that resulted from hatching fertile eggs obtained from commercial flock in Mosul city (NinevehProvince), Iraq were used as a base population (F0). The hatched one-day-old unsexed chicks were divided equally into 8 pens up to 21 days (30 birds/pen and was considered as a family). From 22 days old up to 42 days, the birds were re-divided into 30 replicates (8 birds each). Selected parent on the basis of their live body weight for both sexes were chosen, and the intensity of selection was determined as the percentage of individuals that had more than the average live body weight of each sex within the flock. Male to female ratio was 58 and 42 %, respectively, and the ratio of the selected males to females was 46 and 54%, respectively. Selection intensity was determined as 0.8 for females and 0.9 for males (0.85 as average for both sexes). After sexual maturity (42 days of age as average for females), sires and dams were identified, separated and re-divided randomly (54 females were housed individually in a cagesof 30×50×30 cm)for egg laying, while 58 males maintained in the pens and 27 of them were selected randomly to mate females twice a week intervals by putting each male with each female in a cage for reproduction. A total of 280 chicks were hatched as first generation (F1), 240 of them was randomly divided as same as the previous generation. Water and feed were submitted ad libitum. Light was provided 23 hours/day, vitamins (AD3K) was provided twice, ventilation was provided by using water coolers for both generations, and eggs were collected daily. Rations (combined from yellow corn, soya bean meal, wheat, barley, wheat bran, protein concentrate, vegetable oil, limestone, di-calcium phosphate, DL-methionine, L-lysine, phytase, salt and vit. Premix) fed to both flocks contained 2904 kcal. ME/kg diet with 26% CP; 2937 kcal. ME/kg diet with 21% CP and 2793 kcal. ME/kg diet with 18.3% crude protein (CP) for starter (1-21 d), grower (22-42 d) and layer or breeder (43-end of trial), respectively according to Lesson and Summers (2005).
Traits under consideration
Body weights (g) were individually recorded at 1, 21 and 42 days old for both generations. Weight gain (g) was computed as the difference between next and previous body weight. Feed conversion was calculated as the ratio of feed intake (g) to weight gain (g), at the same studied periods for both generations. The main estimated parameters for the studied traits were selection differentials, which were computed as the difference between selected parent mean and the base population mean; response to selection, was calculated as the difference between offspring F1- mean and the base population F0- mean; the response percentage was calculated as the ratio between response to selection value and F0 live body weight × 100; realized heritability was estimated from dividing response to selection on the selection differential for each trait and genetic correlation, was estimated from the covariance between two traits according to the equation of geometric method (Falconer and Mackay, 1996):
rg= √(cov Z2X1×cov Z1X2) / √(cov Z1X1×cov Z2X2)
1 and 2: characters.
Z: observations of selected parents.
X: observations of progeny.
The experiment was designed according to completely randomized design (CRD); the raw data were analyzed using SAS software (SAS, 2010). For all studied quantitative traits at different ages, general linear model (GLM) procedure was used as multivariate model to analyze the effect of selection, as following:
Y1ijY2ij Y3ij Y4ij= μ + Gi+ eij (Model 1)
Y1ij : observations of the BW.
Y2ij: observations of the WG.
Y3ij: observations of the FI.
Y4ij: observations of the FCR.
μ: overall mean.
Gi: effect of generation.
eij: experimental error.
For body weight trait only at 42 days old, the following three models (two way-analyses and multiple regression equation within GLM procedure and covariance analysis within MIXED procedure, respectively) were used in order to illustrate the interaction between the generation and sex (Model 2), to predict the final BW via initial BW and sex ratio (Model 3) and to eliminate the sex effect, as following:
Y42ijk= μ + Gi+ Sj + (G×S)ij + eijk (Model 2)
Y42ij: observations of the BW.
μ: overall mean.
Gi: effect of generation.
Sj: effect of sex.
(G×S)ij: interaction between generation and sex.
eij: experimental error.
Y42ijk= μ + W1i+ SRj+ eijk (Model 3)
Y42ijk: observations of the BW.
μ: overall mean.
W1i: observations of BW at 1 day old for males of both generations.
SRj: observations of sex ratio for each family (pen).
eij: experimental error.
The multiple prediction equation obtained from the previous analysis called male model and it was as following: Y42= b1 X1 + b2 X2 + a
fSexj: continuous fixed regression factor of generation on female sex ratio.
(Cov G×Sex)ij: covariance between generation and female sex ratio.
eij: experimental error.
The analysis of covariance (Cov G×Sex)ij was used to derive predicted body weight at 42 days old, in order to eliminate the sex effect, it is called female model. After applying the last model, regression analysis was used to estimate the predicted final BW from one-day old BW (modifying Model 3 for both sexes) via the following model:
Y42ijk= μ + W1i + eij (Model 5)
Y42ijk: predicted live body weight at 42 days old.
W1i: observed one-day old live body weight for both sexes and both generations.
Duncan’s multiple range test was used to compare means that resulted from GLM procedure (Duncan, 1955).
RESULTS AND DISCUSSION
Live body weights as affected by selection at different ages are given in Table 1. It can be noticed from the table that F1 surpassed significantly (P<0.01) the F0 generation and the difference between them amounted to average of 29.2%. Such result encourages applying selection for successive generations. This finding agreed with those obtained by Varkoohi et al. (2011), who concluded that the difference between base population (culled line) and selected line for 4 generations was (27 g) at 28 days of age. Similar results had been found by (Marks, 1993b; Grasteau and Minvielle, 2003).
Response to selection
The responses to selection for body weight of both males and females at 42 days of age are shown in Table 2. It appears from the table that the body weight of females are significantly (P<0.01) higher than the body weight of males for base population, selected parents and their progeny. The interaction between generation and sex was significant (P<0.05), where F1 females recorded the highest value. The response to selection in males was higher than females (7.06 vs. 4.87%). Such values reflect the important of selection for body weight in quail. Thus, the overall response to selection for body weight at 42 days of age was estimated to be 11.48 gm., which is equivalent to 5.84%. The selection differential for males was higher than females by 3.8 gm., which was due to the wide range in the body weights within selected female individuals.
Table 1 The effect of selection on live body weight (g) of Japanese quail at major ages (LSM±SE)
The means within the same row with at least one common letter, do not have significant difference (P>0.01).
F0: base population and F1: first generation.
LSM: least square mean and SE: standard error.
Table 2 Response to selection (R) for live body weight trait (g) of Japanese quail at 42 days old, for both sexes (LSM±SE)
The means within the same column with at least one common letter, do not have significant difference (P>0.01).
The interaction between generation and sex was significant (P<0.05).
LSM: least square mean and SE: standard error.
However, the overall selection differential was estimated as 16.6 gm. at 42 days of age. Response to selection for males was higher than that of females (11.98 vs. 10.89 gm.). Such values reflect the importance of selection in studied birds, because both previous values shared as increasing in live body weight by 7.06 and 4.87% for males and females, respectively. However, the overall response to selection was determined as genetic improvement by 11.48 g of the live body weight, which equal 5.84% increasing in overall live body weight of Japanese quail at 42 days of age (irrespective of the sex). This result disagree with that of Vali (2009), who indicated that the difference between males and females in the live body weight of Japanese quail recorded only 4.22 gm. at 42 days of age. On the other hand, the present result is in agreement with the results obtained by (Minvielle et al. 2000) for the significant difference between both sexes. Marks (1993b) concluded that the response to selection for body weight of J. quail for 40 generations achieved 110 g.
Body weight prediction
The prediction of the final live body weight (BW at 42 days of age) was computed depending on one day old and sex ratio. The following prediction equations were derived:
BW42(M)=5.86 BW1 – 0.63 MSR + 176.18 (Male model)
BW42(F)=1. 06 FSR + 176.18(Female model)
BW42(M)and BW42(F): BW at 42 days old for males and females, respectively.
BW1:initial body weight (one day old BW).
MSRand FSR:sex ratio for males and females, respectively in families.
It can be noticed that both equations have the same intercept (176.18) in spite that each was analyzed separately and in different methods; but both have a different regression coefficients for sex ratio % (-0.63 for males and 1.06 for females) and body weight at one day old variable was included for the male model only. So, both models have been joined (combined), to derive a new model called DUHOK equation (coming from DUHOK university and Determining Unsexed Homogeneity Offspring Kids), which could be used in Japanese quail. The final model used to eliminates the effect of sex (considering a sex ratio equal 50%) and predicting with average final live body weight at 42 days old was as follows:
BW42= 5.86× BW1 + 154.68 (DUHOK equation)
The final equation was applied, and its result (predicted values) are shown with the observed values (Figure 1). It could be noticed from the figure, that there is a high variation in the observed values. Predicted values tended to reduce such variation. So, it could be applying the selection program in the future on the basis of predicted values for body weight at hatching, by determining either the selection intensity or culling individuals depending on the special variation range. The statistical parameters for the three models are given above the curves, where the PW42 represent DUHOK equation. Karaman et al. (2013) found that the relationship between predicted and actual observations of Japanese quail's body weight was linear for all models. The growth as affected by some parameters was studied successfully via different functions to illustrate the optimum fit (Davis, 2002; Hall and Clutter, 2004; Sengul and Kiraz, 2005; Narinc et al. 2010).
The estimated realized heritabilities for BW, WG, FI and FCR at 42 days of age were 0.78, 0.67, 0.52 and 0.77, respectively (Table 3). Such high values encourage the use of selection for improving the body weight and FCR in Japanese quail.
Figure 1 Observed vs. Predicted live body weight at 42 days old
Table 3 Phenotypic correlation coefficients (above the diameter), genetic correlation (under the diameter) and realized heritability (on the diameter) for some productive traits in Japanese quail at 42 days old
BW: body weight; WG: weight gain; FI: feed intake and FCR: feed conversion ratio.
* (0.05) and ** (0.01).
Several authors have indicated that the heritability of FCR ranged between 0.2-0.8 (Chambers et al. 1984; Leenstra et al. 1986; Mielenz et al. 1994; N’Dri et al. 2006; Aggrey et al. 2010). Moreover, the genetic correlations between BW and each of WG and FI were high (0.99 and 0.92, respectively). A low values were found between WG and FI (0.45) and between FI and FCR (0.08). However, a negative genetic correlation exists between WG and FCR (-0.41). Similarly, Varkoohi et al. (2011) noticed that the genetic correlation between FCR and WG was negative. The phenotypic correlation coefficients between BW and WG; BW and FI; FI and WG were high (0.91-0.99), while FCR was correlated negatively with both BW and WG (-0.28 and -0.29), but it was correlated positively with FI. These results suggest that increased BW resulted in increase of both WG and FI. Similar findings were obtained by Varkoohi et al. (2011), for productive traits of Japanese quail at 28 days of age, these research workers concluded that a low FCR is genetically related to a high WG and low FI.
Weight gain comparison, between both generations and its linear curves at different periods, showed that F1 individuals surpassed F0 one, especially at initial and final periods (Figure 2). As it is shown from Figure 2, the accumulative response to selection for WG from 1-42 days old was estimated to be 27.04 gm. Also it could be noticed that both generations have a linear parallel curves, except the midpoint was almost the same for both generations. Two equations were described to predict WG as linear curve within each generation for the studied periods (1, 2 and 3). Varkoohi et al. (2010) stated that WG has a linear curve when selection was applied, during different periods of age for all generations.
Feed intake for F1 at initial and final periods was higher than F0 birds, except at the mid period where approximatly equal to F0 intake (Figure 3).
Figure 2 Weight gain and its linear curves for both generations of Japanese quail as affected by selection
* 1= WG during the period from 1-21 d; 2= WG from 22- 42 d and 3= WG from 1- 42 d
Figure 3 Feed Intake and its linear curves for both generations of Japanese quail as affected by selection
* 1= FI during the period from 1-21 d; 2= FI from 22-42 d and 3= FI from 1-42 d
Accumulative response to selction for FI trait was estimated to be 37 gm., this amount of feed was considered little than expected, which may be due to the sudden rise in environmental temperature (36 ˚C) for 3 days at period from 24 to 27 days old (June month). Both R2 for both linear curves are high (0.86 and 0.97), which indicate that the studied periods strongly affecting FI. This result is in agreement with the finding of Knizetova (1996) who found that selection for the live body weight in quail leads to a lower FI associated with higher WG and improved feed efficiency (WG:FI).
Feed conversion ratio
Feed conversion ratio (FCR) in F1 individuals was less and better than F0 flock during the study periods and the pooled response to selection for FCR is -0.2 (Figure 4). This response is considered valuable for one generation. But both R2 of equations for both linear curves are relatively small (0.25 and 0.32).
Figure 4 Feed conversion ratio and its linear curves for both generations of Japanese quail as affected by selection
* 1= FCR during the period from 1-21 d; 2= FCR from 22-42 d and 3= FCR from 1-42 d
This finding is in accordance with that obtained by Varkoohi et al. (2010), who concluded that response to selection for FCR at 28 days old was (-0.22) in the first generation.
From the results obtained in the current work, it can be concluded that selection for BW in Japanese quail resulted in an improvement in FCR.
We thank Dean of College Dr. M.S. Berwary, Head of department Dr. K.N. Dosky, Prof. J. Alkis and Mr. Renas H. Issa for the facilities and aid.
Aggrey S.E., Karnuah A.B., Sebastian B. and Anthony N.B. (2010). Genetic properties of feed efficiency parameters in meattype chickens. Genet. Sel. Evol.42, 25.
Ayasan T. ( 2013). Effects of dietary inclusion of protexin (probiotic) on hatchability of Japanese quails. Indian J. Anim. Sci. 83(1), 78-81.
Baylan M., Canogullari S., Sahin A., Copur G. and Baylan M. (2009). Effects of different selection methods for body weight on some genetic parameters. J. Anim. Vet. Adv. 8, 1385-1391.
Caron N. and Minvielle F. (1990). Mass selections for 45-day body weight in Japanese quail: selection response carcass composition, cooking properties and sensory characteristics. Poult. Sci.69, 1037-1045.
Chambers J.R., Bemon D.E. and Gavora J.S. (1984). Synthesis and parameters of new populations of meat-type chickens. Appl. Genet.69, 23–30.
Darden J.R. and Marks H.L. (1989). Divergent selection for growth in Japanese quail under split and complete nutritional environment. 3. Influence of selection for growth on heterotic effects for body weight, feed and water intake patterns, abdominal fat and carcass lipid characteristics. Poult. Sci.68, 37-45.
Davis C.S. (2002). Statistical Methods for the Analysis of Repeated Measurements. Normal-Theory Methods: Multivariate Analysis of Variance. Springer-Verlag, New York.
Duncan D.R.(1955). Multiple range and multiple F test. J. Biometrics.1, 1-42.
Falconer D.S. and Mackay T.F.C. (1996). Introduction to Quantitative Genetics. Longman, London, UK.
Grasteau S.M. and Minvielle F. (2003). Relation between tonic immobility and production estimated by factorial correspondence analysis in Japanese quail. Poul. Sci.82, 1839-1844.
Hall D.B. and Clutter M. (2004). Multivariate multilevel nonlinear mixed effects models for timber yield predictions. Biometrics.60, 16-24.
Karaman E., Narinc D., Firat M.Z. and Aksoy T. (2013). Nonlinear mixed effects modeling of growth in Japanese quail. Poult. Sci. 92, 1942-1948.
Kaur S., Mandal A.B., Singh K.B. and Kadam M.M. (2008). The response of Japanese quails (heavy body weight line) to dietary energy levels and graded essential amino acid levels on growth performance and immuno-competence. Livest. Sci. 117, 255-262.
Kayang B.B., Vignal A., Inoue-Murayama M., Miwa M., Monvoisin J.L., Ito S. and Minvielle F. (2004). A first generation micro satellite linkage map of the Japanese quail. Anim. Genet. 35, 195-200.
Knizetova H. (1996). Growth and carcass traits of Japanese quail. Zivoc. Vyroba.41, 225-233.
Leenstra F.R., Vereijken P.F.G. and Pit R. (1986). Fat deposition in a broiler sires strain. 1. Phenotypic and genetic variation in, and correlations between, abdominal fat, body weight and feed conversion. Poult. Sci.65, 1225-1235.
Lesson S. and Summers J.D. (2005). Commercial poultry nutrition. NottinghamUniversity Pres, Guelph, Ontario, Canada.
Marks H.L. (1993a). Carcass composition, feed intake and feed efficiency following long term selection four week body weight in Japanese quail. Poult. Sci.72, 1005-1011.
Marks H.L. (1993b). Divergent selection for growth in Japanese quail under split and complete nutritional environments. 6. Differential body weights in reciprocal crosses. Poult. Sci.72, 1847-1854.
Mielenz N., Groeneveld E., Muller J. and Spilke J. (1994). Simultaneous estimation of variances and covariance’s using REML and Henderson 3 in a selected population of white Leghorns. Br. Poult. Sci.35, 669-676.
Minvielle F. (1998). Genetics and breeding of Japanese quail for production around the world. Pp. 122-127 in 6th Proc. Asian Poult. Cong. World’s Poult. Sci. Assoc. Nagoya, Japan.
Minvielle F., Monvoisin J.L., Costa J. and Maeda Y. (2000). Long-term egg production and heterosis in quail lines after within-line or reciprocal recurrent selection for high early egg production. Br. Poult. Sci. 41, 150-157.
Muir M.W., Bijma P. and Schinckel A. (2013). Multilevel selection with kin and non-kin groups, experimental results with Japanese quail (Cotunix japonica). Evol. J. Spec. Sect. 2, 1-9.
Narinc D., Aksoy T. and Karaman E. (2010). Genetic parameters of growth curve parameters and weekly body weights in Japanese quail (Coturnix coturnix japonica). J. Anim. Vet. Adv.9, 501-507.
N’Dri A.L., Mignon-Grasteau S., Sellier N., Tixier-Boichard M. and Beaumont C. (2006). Genetic relationships between feed conversion ratio, growth curve and body composition in slow growing chickens. Br. Poult. Sci.47, 273-280.
SAS Institute. (2010). SAS®/STAT Software, Release 9.2. SAS Institute, Inc., Cary, NC. USA.
Sengul T. and Kiraz S. (2005). Non-linear models for growth curves in Large White turkeys. Turkish J. Vet. Anim. Sci.29, 331-337.
Vali N. (2009). Growth, feed consumption and carcass composition of Coturnix japonica, Coturnix ypsilophorus and their reciprocal crosses. Asian J. Poult. Sci.3, 132-137.
Varkoohi S., Moradi M., Pakdel A., Nejati A., zaghari M. and Kause A. (2010). Response to selection for feed conversion ratio in Japanese quail. Poult. Sci.89, 1590-1598.
Varkoohi S., Pakdel A., Moradi M., Nejati A., Kause A. and Zaghari M. (2011). Genetic parameters for feed utilization traits in Japanese quail. Poult. Sci.90, 42-47.