Breeding for the improvement of indigenous chickens of Bangladesh: evaluation of performance of first generation of indigenous chicken

Pure breeding is necessary for the conservation and improvement of indigenous chicken genetic resources. Present research is a part of the long-term selection program being undertaken to evaluate the performance and expected response to selection of first generation (G1) of three indigenous chicken genotypes under intensive management in Bangladesh. A total of 1439-day-old chicks comprising of 3 genotypes namely Naked Neck (NN), Hilly (H) and Non-descript Desi (ND) were hatched for this study. In first generation (G1), selection was practiced on body weight at 8 and 16 weeks of age, on the basis of their breeding value. At 40week of age, selection will be practiced on the basis of an index comprising the parameters of age at first egg (ASM), body weight (BW), egg production (EP) and egg weight (EW). At 8, 10 and 12 weeks of age, six birds from each genotype were slaughtered to analyze the meat yield traits. Data were analyzed in CRD by General Linear Model Univariate Procedure. Significantly (P<0.001) highest body weight of day-old chicks (28.65±0.12 g) and daily weight gain in all stages were found in H genotype than other two genotypes. Although there was significant (P<0.001) difference in live weight between ages at slaughter, dressing percent (65.87 66.89 %) of different ages was similar (P>0.05) but was affected (P<0.001) by genotype. Body weight at 8 weeks of age was expected to improve by 58.98 vs. 11.50; 81.56 vs. 40.91 and 53.81 vs. 15.82 g; respectively for ND, H and NN males and females. In terms of body weight and growth traits H genotype was superior and NN genotype was for dressing percentage. As a result of selection; chick weight, body weight at all stages increased and ASM reduced in first generation than foundation stock. These findings give an impetus for continuing the pure breeding research for more generations.


Introduction
Traditionally local chicken perform a variety of functions, e.g.laying eggs, hatching chicks, brooding and caring of them (Shahjahan et al., 2011).Indigenous chickens are generally better for disease resistance and could maintain higher level of performance under poor nutrition and high environmental temperatures compared to commercial strains under village systems (Horst, 1989).Indigenous chickens of Bangladesh are categorized as Non-descriptive Deshi (ND), Naked Neck (NN), Hilly (H), Aseel (AS) and Jungle fowl (Bhuiyan et al., 2005) in respect of the morphological variations as well as production performances.The genetic potential of indigenous poultry in Bangladesh is poor.Their productivity is low; as a result they are unable to meet the demand for eggs and meat in the country.This is considered as an important constrain to poultry development, which could be overcome through genetic improvement of indigenous stock by appropriate breeding as well as changing their productive environment.The improvement of productivity of native chicken is a long desire in the country.Latest data appeared in a national daily, Bangladesh is producing more than 400 million heads of broilers and 6600 million pieces of eggs from commercial sector which represent 60 to 70% of country's total poultry meat and egg production, the rest (30-40%) is being added from traditionally reared free range scavenging native chickens (Chowdhury, 2011).During the last three decades, the number of commercial hybrids has increased many folds, but the native chicken reared in the villages still remain the main chicken genetic resources for the rural farmers (Yamamoto et al., 2012).But this process has been reversing day by day.Now village farmers are interested to practice industrial poultry.Indigenous chickens are going to be disappearance.As a result, Bangladesh Livestock Research Institute (BLRI) since its inception did initiate programs for the conservation and development of indigenous chicken through several poultry development projects.It is a long-term vision.As a first step, foundation stock was established utilizing the existing stock of BLRI as well as by incorporating variation through screening of males/females/eggs from wider indigenous chicken gene pool of Bangladesh.A total of 32 males (ND=16, H=8, NN=8) and 160 females (ND=80, H=40, NN=40) were selected from foundation stock to produce first generation (G 1 ) and 1439 progenies (ND=628, H=475, NN=336) were produced for G 1 .However, the present study was undertaken to fulfill the objectives of i) to compare the productive and reproductive performances of first generation of 3 native genotypes under intensive management, ii) to select parental birds (males and females) and breed them in an assortative design for the production of second generation birds and iii) to improve the genetic potentiality of Indigenous chicken genotypes for selected economic trait(s) through successive generations of purebreeding

Location and agro-climate
The study was conducted at the Bangladesh Livestock Research Institute, under Poultry Production Research Division, Savar, Dhaka, Bangladesh, the largest delta in the world is situated between 88 0 10′ and 92 0 41′ East longitudes and between 20 0 34′ and 26 0 38′ North latitudes.The station is located at 23 0 53/N, 90 0 17/E at an altitude of 1 meter above the sea level.Agro-ecologically belongs to the Madhupur Tract (Agro ecological Zone 28) of Bangladesh, Red-Brown Terrace strong acidic (pH 4.5-5.5)soil with very little (<1.5%) organic matter (Brammer et al., 1988).Bangladesh has tropical and moderate climate with the summer, monsoon, autumn, late autumn, winter and spring seasons.Annual minimum temperature varies from 8.0 0 c to 13.4 0 c and maximum temperature 25.5 0 c to 36.8 0 c.Extreme temperatures range between about 4 0 c and 43 0 c, except on the coast (Bhuiyan et al., 2005).

Formation of first generation ( G1 ) from foundation stock (G 0 )
Foundation stock was established utilizing the existing stock of BLRI as well as by incorporating variation through screening of males/females/eggs from wider indigenous chicken gene pool of Bangladesh.A total of 4688 eggs (NN-1683, H-1546and ND-1459) were collected from different parts of Bangladesh to form foundation stock and 1585 day-old chicks were used comprising of 918 ND, 378 H and 289 NN.A total of 160 dams (ND=80 hens, H= 40 hens and NN= 40 hens) and 32 sires were selected from foundation stock to produce first generation (G 1 ).Selection was done using multi trait index selection and independent culling level at 40 weeks of age comprising the parameters of Age at first egg (days), Body weight (g) at 40 weeks of age, Egg production percentage (168-280 days) and Egg Weight (g) at 38 -40 weeks of age.In foundation stock, selected males and females were mated at the ratio of 1: 5 using artificial insemination.A total of 1439 progenies (ND=628, H= 475 and NN= 336) were hatched in a three batches for first generation (G 1 ).

Selection of first generation (G 1 ) birds
The selection objectives of the study were to improve the egg production, egg weight and / or growth rate of indigenous chicken depending on the genotype of birds.Improvement target of egg weight is to increase by 1 g and improvement target of egg production rate is to increase by 2% per generation.The main target was the mean body weight in Hilly genotype of chicken ND has to be gone up from the initial eight-week body weight of 378 g to 500 g at eight weeks of age after 3 generations of selective breeding.At 8 weeks of age, a total of 460 female birds (ND=230, H= 115 and NN= 115) and 95 male birds (ND=45, H= 25 and NN= 25) were selected according to breeding value on the basis of 8 week's body weight.At 16 weeks of age, a total of 400 female birds (ND=200, H= 100 and NN= 100) and 80 male birds (ND=40, H= 20 and NN= 20) were selected according to breeding value on the basis of 16 week's body weight.At 40 weeks of age, 16 males and females will be selected according to an index value.A total of 20 male birds (ND=10, H= 5 and NN= 5) and female birds (ND=20, H= 10 and NN= 10) will be kept as spare birds.

General flock management
Day-old chicks collected from the hatchery were weighed and leg banded individually.After that, all the chicks of three genotypes were transferred into the brooder, which were cleaned and disinfected earlier and 5% glucose solution was supplied for the first three days.After that, all the chicks of three genotypes were transferred into the brooder, which were cleaned and disinfected earlier.One-week later leg bands were removed and wing bands were provided all the experimental birds.Debeaking was performed after 10-12 days of age.The chicks were brooded and reared up to 16 weeks of age with individual wing band in a brooding and growing house with standard feeding and management.All chicks were vaccinated as per schedule given by veterinarian.After 16 weeks of age all female and male birds were transferred into individual cage.twice daily in the morning and evening during brooding, growing and laying period, respectively.Cool clean drinking water was supplied all the times.Water also was provided ad libitum twice daily in the morning and evening.Feeder and drinker were cleaned twice in a week.Refusals of the feed were measured everyday in the morning.

Vaccination
All chicks were vaccinated as per schedule given by veterinarian.The vaccination schedule that was practiced in the experiment is shown in the Table 1.

Lighting program
The photoperiod for brooding period was started at 24 hours/day reduced @ 1 hour/week.Depending on season and day length photoperiod was maintained for layer birds.All the birds were reared in a natural-ventilated poultry house and a 16h photoperiod with 12h sunlight and 4 h artificial lights.

Slaughtering and carcass characteristics data
At 8, 10 and 12 weeks of age, six chicks of each genotype were slaughtered to analyze the meat yield traits.The treatments were arranged in a 3 (genotype) × 3 (slaughter age) factorial experiment.All birds were kept off feed overnight before slaughtering but drinking water was provided ad libitum.Birds were slaughtered following 'halal' method (Singh et al., 2003) by severing the jugular vein allowed to bleed completely and then plucked and weighed to determine blood and feather losses (Kotula et al. 1960;Pandey and Shyamsunder, 1990).Preslaughter live weight, blood loss weight, eviscerated weight, breast meat weight, thigh plus drumstick weight etc. were recorded.All weight related to carcass characteristics were expressed as the percentage of live weight.
Carcasses were dissected according to Singh et al. (2003) except that birds were not scalded.

Data recording
All productive and reproductive parameters were recorded for first generation (G 1 ).Records were kept on dayold weight (g), fortnightly individual body weight up to 8 weeks, monthly weight up to 20 weeks, daily egg production, and egg weight at 40 weeks of age, temperature and humidity, growth rate, feed intake and feed conversion ration (FCR).Feed conversion ratio (FCR) was recorded for the whole period as total feed intake (kg) per kg weight gain.Temperature and humidity were recorded four times a day; (06:00h, 12:00h, 18:00h and 24:00h).

Statistical analysis
All recorded data were analyzed by Generalized Linear Model (GLM) procedure using SPSS 11.5 for Windows (SPSS, 1998).For all statistical purposes the theory of Snedecor and Cochran (1989) were followed.The present data used in the study were from three different genotypes and the structures of data were unbalanced.
The number of birds varied from class to class and subclass to subclass.Hence, it confirmed the characteristics of a non-orthogonal factorial experiment.To take this situation into account the data were analyzed by factorial arrangement in a CRD by General linear Model (GLM) Univariate Procedure in SPSS Computer Program.The treatments were arranged in a 3 (genotype) × 3 (slaughter age) factorial experiment.
The following general linear statistical models were used to analyze the different parameters: Where, Y ij is the dependent variable of the experiment; μ is the overall mean; g i is the effect of ith genotype (i=1-3); e ij is the error term specific to each record.ii) Y ijk = μ + g i +s j + (g×s) ij +e ijk, Where, Y ijk is the dependent variable of the experiment; μ is the overall mean; g i is the effect of ith genotype (i=1-3); s j is the effect of jth slaughter age (j=1-3); (g×s) ij is the effect of ith genotype (i=1-3) and jth slaughter age (j=1-3); e ijk is the error term specific to each record.

Estimation of heritability (h 2 )
Some dams either did not lay or had no chicks at hatching and records from chicks that lost their wing bands were not included.Only data on birds having proper identified number (Pedigree and performance) were used to estimate heritability for the considered trait.After data editing, a total of 1329 chicks of 30 sires and 129 dams were available for heritability estimation.Variance component of the body weight trait was estimated using Residual Maximum Likelihood (REML) approach by VCE4 computer program (Groeneveld, 1998).The animal model for 8 week body weight included the fixed effects of hatch number and sex of the chicken and birds itself as a random effect.The analysis was done in a single trait animal model.

Prediction of expected selection response
Expected selection response in three types of indigenous chicken for body weight at 8 weeks was estimated using the following equation (Falconer, 1981).

R = h 2 × S
Where, R = Expected response in mass selection h 2 = heritability, h 2 for BW at 8 weeks of age S = Selection differential

Results and Discussion
3.1.Body weight, weight gain and FCR Body weights, body weight gain, feed consumption and feed conversion ratio from day-old to 8 weeks, day-old to 10 weeks and day-old to 12 weeks of age are shown in Table 1.The average initial body weight of day-old chicks of ND, H and NN were 27.67, 28.65 and 27.89 g, respectively and the difference was significant (P<0.001).Faruque et al. (2011b) found that the body weight at hatch for ND, H and NN genotypes was 27.74, 28.00 and 24.96 g, respectively which were more or less similar to the present study.The body weight at hatch for ND, H and NN genotypes was 31.2, 30.5, 31.7 g, respectively under intensive management system which was much higher than the present study (Faruque et al., 2007).Significantly (P<0.001) the highest daily weight gain and total weight gain were found in H genotype at all stages.Mean daily weight gains and total gain between 0-8, 0-10, 0-12 weeks of age were 7. 68, 8.19, 8.19 g vs. 430.12, 573.34, 688.02; 8.32, 9.64, 9.66 g; vs. 465.77, 674.43, 811.53 g; 7.35, 8.23, 8.08 g vs. 411.88, 576.58, 679.46 g; respectively for ND, H and NN genotypes.The lowest and highest mean body weight gain per bird were recorded for ND (329.38±3.32g)and NN (351.56±5.08 g) genotypes, which indicated that there were an average daily growth rate of 5.88±0.05and 6.27±0.09g per bird per day at their 8 weeks growth phases, respectively (Faruque et al., 2011a).Daily weight gains between 0-4, 5-18, 0-18 weeks of age were 2.45 vs. 1.45; 8.65 vs. 6.75; 7.25 vs. 5.55, respectively for Normal Deshi (DN) and Desi dwarf (DD) genotypes (Yeasmin and Howlider, 1998).There was a nonsignificant (P>0.05)variation in FCR among the native chicken genotypes (Table 3).Similar result was found by Faruque et al. (2011a) who found that the FCR of indigenous chicken at 8 th week of age was 3.58, 3.45 and 3.34 respectively for ND, H and NN genotypes.Yeasmin (2000) conducted an experiment on normal feathered indigenous birds under farm condition and found that the daily feed intake was 33.95 g at 5-18 weeks of age and FCR was 3.95 at 5-18 weeks of age.
The results of this study show the heavier weight of H genotype than other two genotypes (Figure 1).The growth rates of H genotype sharply increase after 8 weeks of age.
Expected and observed response for 8 week body weight were found as 38.59 vs. 117.83g; 42.12 vs. 125.02g and 28.67 vs. 106.14g, respectively for ND, H and NN genotypes (Table 7)

Comparison between first generation (G 1 ) and foundation stock (G 0 )
Changes of body weight, age at sexual maturity (ASM), egg production (EP) and egg weight (EW) for first generation (G 1 ) are shown in Table 8.Weight gains in indigenous chicken at hatch, 8, 20 weeks of age and hen weight at maturity were -0.07, 0.65, 2.93 g; 115.82, 120.52, 110.39 g; 110.19, 228.66, 75.07 g; and 63.62, 224.67, 42.92 g respectively for ND, H and NN genotypes.It was observed that all genotypes came to sexual maturity as 8.82 -10.60 days earlier in the first generation (G 1 ) than in the foundation stock (G 0 ).Egg production and egg weight were recorded for G 1 up to 36 weeks of birds' age while EP and EW were recorded for G 0 up to 40 weeks of Age.So comparison between (G 1 ) and (G 0 ) were not explained.

Conclusions
In terms of body weight and growth responses at all stages of age, H genotype was superior and NN genotype was for dressing percentage.NN genotype was also found to have a higher survivability rate, reaching maturity earlier but attaining a lighter mature weight.Body weight increased at all stages in the first generation than in the foundation stock.All genotypes attained to sexual maturity on an average nine days earlier in the first generation than in the foundation stock.It is being expected that EP and EW will be increased if comparison is done at the same ages.These findings give an impetus for continuing the pure breeding research for more generations.

Figure 1 .
Figure 1.The average weight of first generation of indigenous chicken until the age of sixteen weeks.

0 200 400 600 800 1000 1200 1400 Day-old wt 4 wk wt 8 wk wt 10 wk wt 12 wk wt 16 wk wt Week weight (g)
Jaturasitha et al. (2008)turasitha et al. (2008)reported live weights at slaughter at the same age clearly differed (p<0.001)among genotypes, with a lower growth rate of the indigenous genotypes, especially Black boned chickens, compared with the imported.Dressing percentage did not differ (p>0.05)among genotypes.They also reported dressing percentage of 63.7, 65.9, 63.6 and 64.4 for Black boned, Thai, Bresse and Rhode Island Red genotypes.Breast meat weight as percentage of live weight and thigh plus drumstick weight as percentage of live weight were highly affected (P<0.001) by genotype but were not affected (P>0.05) by slaughter age.
genotype than ND (66.29) and H (64.50) genotypes.Similar result was found in NN (64.58) genotype than ND (60.26) and H

Heritability for 8 week body weight trait
The variance components and heritability estimates along with corresponding standard errors of 8 week body weight is presented in Table5.Calculated heritability for 8 week body weight of ND, H and NN were 0.499  0.034, 0.489  0.027 and 0.478  0.034 respectively.In case of ND, H and NN genotypes; 49.9%, 48.9% and of 47.80% variation of 8 week body weight due to heredity and rest is controlled by environment.Lwelamira et al.

Expected and observed response to selection
Genotype wise expected response and observed response to selection for 8 week body weight is shown in Table6 & Table 7.As a result of selection, body weight at 8 weeks of age was expected to improve by 58.98 vs.