Analysis of genetic diversity and conservation priorities in asian domestic chicken populations based on microsatellites

iv Analysis of genetic diversity and conservation priorities in Asian domestic chicken populations based on microsatellites In Chapter 3, the paper describes the important role of Taiw

NATIONAL CHUNG HSING UNIVERSITY

DEPARTMENT OF ANIMAL SCIENCE

July, 2013

i

ACKNOWLEDGEMENTS

It is impossible to complete my Ph.D program without the help and support from many people I owe my deep gratitude to those who made this Doctoral dissertation as much as possible

Firstly, I would sincerely thank my advisors, Prof Yen-Pai Lee, Prof Chih-Feng Chen and Dr Tran Xuan Hoan, for your daily supervision and great supports during my Ph.D program I hereby would like to give my special thanks to Prof Michèle Tixier-Boichard, Dr Cécile Berthouly-Salazar and Dr Richard Crooijmans for your critically scientific suggestions and linguistic revisions I would like to thank the Committee members for your useful comments to this Doctoral dissertation

I sincerely appreciate Poultry Breeding Lab members, teachers in Department of Animal Science, National Chung Hsing University, as well as many colleagues for your kind helps and sharing experiences, especially Mr Wen-Long Hsiao for Chinese translation of the abstract

I am very grateful to the members of Department of Animal Breeding and Genetics, Institute of Agricultural Sciences for Southern Vietnam and the recent Institute of Animal Sciences of Southern Vietnam, Ministry of Agriculture and Rural Development of Vietnam for your valuable supports within my study period

My special thanks are given to the colleagues of the National Institute of Animal Sciences in Hanoi and many Vietnamese private farmers for their help in collecting blood samples as well as breeding information of chicken populations

The project is not successfully done without the financial support from the National Science Council, Taiwan Personally, I gratefully acknowledge a grant from Taiwan Scholarship Program

I hereby would like to respect my all dear family members for your great support and encouragement I am never forgetting your kind contributions to me

Finally, I wish to express my deepest love to my wife, Pham Ha Phuong, and my little daughter, Pham Phuong Ngan This Doctoral dissertation is for your endless love and patience since the summer of 2009 Cám n gia đình, v và con r t nhi u!

iv

Analysis of genetic diversity and conservation priorities in Asian domestic chicken

populations based on microsatellites

In Chapter 3, the paper describes the important role of Taiwan commercial native chicken production and the genetic characterization within and between 10 Taiwan commercial native chicken (TCOM) populations, together with two exotic breeds and one RJF population sampled in Taiwan for comparison, using 22 microsatellites We used traditional methods to quantify the genetic diversity and the Bayesian clustering approach to quantify the admixture pattern Consequently, the TCOM exhibited the high genetic diversity within populations The TCOM breeds were more admixed than exotic breeds but they showed high production and suited to Taiwanese traditional cooking style and cultural values The results of this study suggest that the genetic pool of TCOM is well distributed among breeds and therefore there is a good potential for adaptation to new environmental conditions or markets

v

Chapter 4 aims at estimating genetic risk status and conservation priorities of eight Taiwan conserved chicken (TCS) populations using pedigree and molecular data Using three strategies for setting priorities in conservation, we estimated different effective population sizes and genetic contributions for eight chicken populations by pedigree-based analysis For molecular-based analysis, estimates of contemporary effective population size were computed from genotypic data of 22 microsatellites by linkage disequilibrium method In addition, we used three criteria to propose conservation priorities for eight populations The results showed that most TCS populations are considered to be in safe across pedigree- and molecular-based approaches Thus, this study of TCS populations shows how different types of data can

be combined to define conservation priorities considering risk, diversity, or utility of local breeds

The genetic structure among 32 Asian domestic chicken populations and Red Jungle Fowls using Bayesian clustering approaches is described in Chapter 5 VNN populations genetically exhibited higher degree of admixture than that of CNO, TCS, and TCOM populations due to gene flow among populations Little evidence of gene flow was observed between White Leghorn and remaining populations Bayesian clustering approaches revealed that CNO, TCS, TCOM populations and White Leghorn clustered together, which separated from VNN populations Interestingly, Taiwan Game bird and two Red Jungle Fowl populations were clustered to VNN populations BAPS (Bayesian Analysis of Population Structure) program provided a finer picture of admixture than STRUCTURE did

Chapter 6 discusses the final results of the whole thesis and the remaining issues that need to be improved Several suggestions are included for the future research

Key words: Bayesian clustering analysis, effective population sizes, genetic variation,

inbreeding, pedigree information, principal component analysis

vi

TABLE OF CONTENTS

Page

ACKNOWLEDGEMENTS i

ABSTRACT IN CHINESE ii

ABSTRACT iv

TABLE OF CONTENTS vi

LIST OF TABLES viii

LIST OF FIGURES ix

CHAPTER 1 1

General introduction 1

1.1 Introduction 2

1.1.1 Chicken domestication 2

1.1.2 The threats and opportunities to maintain local chickens 5

1.1.3 Conservation of chicken genetic resources 8

1.1.4 The strategies for setting priorities in conservation 8

1.2 Material and methods 10

1.2.1 Description of phenotypic characteristics and performance traits of chicken breeds from this study 10

1.2.2 Populations and study sites 19

1.2.3 Molecular-based analysis for monitoring genetic diversity 22

1.2.4 Pedigree-based analysis for monitoring genetic diversity 31

1.3 Scope of the thesis 35

CHAPTER 2 36

Genetic diversity of Vietnamese domestic chicken populations as decision-making support for conservation strategies 36

2.1 Abstract 37

2.2 Introduction 38

2.3 Material and methods 39

2.4 Results 45

2.5 Discussion 53

CHAPTER 3 69

vii

Genetic characterization of Taiwan commercial native chickens ascertained by

microsatellite markers 69

3.1 Abstract 70

3.2 Introduction 71

3.3 Material and methods 72

3.4 Results and discussion 74

CHAPTER 4 84

Monitoring of genetic diversity in Taiwan conserved chickens assessed by pedigree and molecular data 84

4.1 Abstract 85

4.2 Introduction 86

4.3 Material and methods 87

4.4 Results and discussion 89

CHAPTER 5 98

Genetic structure of 32 Asian domestic chicken populations and Red Jungle Fowls 98

5.1 Abstract 99

5.2 Introduction 100

5.3 Material and methods 101

5.4 Results and discussion 103

CHAPTER 6 116

General discussion, conclusions and prospective 116

6.1 General discussion 117

6.2 General conclusions 119

6.3 Prospective for further studies 119

REFERENCES 122

ANNEX 139

viii

LIST OF TABLES

Page

Table 1 Risk categories used for livestock breeds 9

Table 2 Production and reproduction data from Vietnamese domestic chickens 11

Table 3 Production and reproduction data from Taiwan country chickens 11

Table 4 List of chicken genes with visible effects 12

Table 5 List of complex chicken phenotypes 12

Table 6 Phenotypic characteristics of different chicken breeds used in the study 13

Table 7 History of different autosomal markers in domestic animals 23

Table 8 Summary information of 22 microsatellite markers used in this thesis 23

Table 9 Genetic diversity within 23 Vietnamese domestic chicken populations and one Red Jungle Fowl population 46

Table 10 Genetic diversity within 10 Taiwan commercial native populations, two exotic breeds and one Red Jungle Fowl population 75

Table 11 The proportion of membership of each of nine TCOM populations, two exotic breeds and one Red Jungle Fowl population (after removing Game bird population) in the 12 inferred clusters 79

Table 12 Pedigree information in the first generation of the eight 88

Table 13 Pedigree information in founders of the eight populations 90

Table 14 Estimates of effective population sizes for each population obtained by the pedigree and molecular data 91

Table 15 Genetic contributions of eight Taiwan conserved chicken populations 92

Table 16 Loss or gain of genetic diversity, contributions to allelic richness and aggregate diversity for each of the eight Taiwan conserved chicken breeds 95

Table 17 Genetic diversity within 31 Asian chicken populations, two Red Jungle Fowl populations and one White Leghorn population 105

ix

LIST OF FIGURES

Page

Figure 1 Phylogenetic tree of the Gallus species 2

Figure 2 Distribution of wild Jungle Fowl species 3

Figure 3 Risk status of the world's avian breeds in October 2010 6

Figure 4 Chicken plumage color patterns 18

Figure 5 The variations of chicken comb types 19

Figure 6 The variations of chicken shanks 19

Figure 7 Earlobe colors 19

Figure 8 Geographical origin of Vietnamese domestic chicken populations 20

Figure 9 Geographical origin of Taiwan country chicken populations, commercial lines and Red Jungle Fowl 21

Figure 10 Sequential substructuring pattern 49

Figure 11 Principal component analysis of the seven homogenous VNN populations and 25 flocks derived from nine VNN populations 51

Figure 12 Genetic structure of 10 Taiwan commercial native chicken populations, two exotic breeds and one Red Jungle Fowl population 78

Figure 13 Average pedigree completeness index for each year of birth in eight populations 89

Figure 14 Effective population sizes for each population obtained by the pedigree and molecular data 91

Figure 15 Genetic diversity (GD) loss in eight populations in the period between 2002 and 2008 93

Figure 16 Two-dimensional scaling representation of the FST pairwise genetic distances between 34 populations 106

Figure 17 Sequential substructuring pattern for the 34 populations 107

Figure 18 Histograms of genetic structure constructed from BAPS for the 34 populations 109

Figure 19 Gene flow network identified in the 34 populations 111

1

CHAPTER 1

General introduction

2

1.1 Introduction

1.1.1 Chicken domestication

1.1.1.1 Description of the Gallus species

The current domesticated chicken (G g domesticus) belongs to the Gallus species There are four wild species of Gallus known (Figures 1 and 2) such as, Gallus sonneratii, Gallus varius, Gallus lafayettei and Gallus gallus that can differentiate by their morphology and their geographical distribution in Asia (Tixier-Boichard et al., 2011)

Figure 1 Phylogenetic tree of the Gallus species

The Figure was adapted from Berthouly (2008) The pictures of Gallus sonneratii, Gallus lafayettei and Gallus varius were from Eriksson et al (2008) The pictures of Gallus g gallus, Gallus g spadiceus and Gallus g domesticus (Vietnamese Ri local chicken) were from our study

3

The G sonneratii (Grey Jungle Fowl) has a grey plumage and is found in the bamboo forests in South and Southwest of the Indian subcontinent The G varius (Green Jungle Fowl) distributes only on the Islands of Java, Bali, Komodo and Sumbawa and is characterized by several morphological peculiarities including a single three-colored wattle (i.e red, yellow, blue), the lack of indentations of the comb, two additional feathers on the tail and a greenish plumage color It is the most distant from domestic chickens in term of morphology (Figure 1) The G lafayettei (Ceylon Jungle Fowl) is found only in Sri Lanka and exhibits an orange-brown color of the breast with

a purple spot on the top of the neck and a yellow spot on the comb The G gallus (Red Jungle Fowl) is the closest to domestic chickens by its morphology and gives fertile offspring after crossing with domestic chickens, whereas crossing between domestic chickens and any of the three other wild species yields very poor hatchability and chick survival, G gallus always exhibits blue shanks This species inhabits in forests and covers a wide geographic area in Asia (Tixier-Boichard et al., 2011)

Figure 2 Distribution of wild Jungle Fowl species

A, Gallus sonneratii; B, Gallus lafayettei; C, Gallus g murghi; D, Gallus g spadiceus; E, Gallus g jabouillei;

F, Gallus g gallus; G, Gallus g bankiva; and H, Gallus varius Source: West and Zhou (1988).

It is assumed that the Gallus gallus subdivided into five subspecies based upon morphological differences and geographical distribution G g gallus is found in

4

Southeast Asia (South Vietnam, Cambodia, Thailand and Laos), with white earlobes G

g spadiceus distributes in Myanmar, Thailand, Malaysia and China (Yunnan province), with red earlobes (Nishida et al., 2000) G g murghi comes from Northeast of India, with white earlobes G g bankiva comes from Java and Sumatra, with red earlobes (Tixier-Boichard et al., 2011) G g jabouillei is from Southern China and North Vietnam, with red earlobes (Nishida et al., 2000) The wild ancestor of the domestic chicken was considered for a long time to be the G gallus species, on the basis of morphology but also on the basis of protein polymorphism (Tixier-Boichard et al., 2011)

1.1.1.2 Archeological evidences

There are at least two main regions for chicken domestication The first archeological evidence of domestication was found in China about 5400 BC, in geographically widespread sites such as Hebei Province, 5300 BC, Shandong Province,

5000 BC, and Shaanxi Province, 4300 BC (West and Zhou, 1988) This study proposed that these bones originated from domestic chickens, since Red Jungle Fowl was never reported to have inhabited North of China The second finding of domestication was identified at Mohenjo-Daro archeological site in the Indus valley about 2500 BC (West and Zhou, 1988) In Vietnam, archaeologists discovered many statues of chickens in Vinh Quang, Hoa Binh and Hanoi belonged to the Early Bronze Age, between 3000

BC, and in the Early Stone Age, 2000 BC (Vo, 1978) Domestication of chicken took place by human more than 8000 years ago directly from G gallus in Southeast Asia and became well established at Neolithic sites in Northern China and then in India at about

2000 BC (West and Zhou, 1988; Fumihito et al., 1994; Fumihito et al., 1996)

1.1.1.3 Ancestors of domestic chickens

Most molecular studies of evolutionary history of Gallus species were used the polymorphism of mitochondrial DNA (mtDNA) The first two studies used the D-loop regions of mtDNA from all Gallus species and three subspecies of G gallus, including

G g gallus, G g spadiceus and G g bankiva (Fumihito et al., 1994; Fumihito et al.,

5

1996) They concluded that G varius diverged first among four species, and G lafayettei and G sonneratii had diverged simultaneously from G gallus They also concluded that domesticated chickens descend from a single ancestor, G gallus (Fumihito et al., 1996) However, it was argued that neither samples from other two subspecies, G g murghi and G g jabouillei, nor any domestic breeds from other regions were used in their studies, so the results may not be strongly supported Nishibori et al (2005) compared sequences of mtDNA and five autosomal loci from domestic chickens and all four Gallus species The results revealed that two additional species of G sonneratii and G lafayettei might hybridize with the domestic chicken in the history In the study from China, Liu et al (2006) showed that three subspecies of G gallus (i.e G g gallus, G g spadiceus and G g jabouillei) were involved in the origin

of the current domestic chickens and they suggested two domesticated centers, one in Southern China together with Southeast Asia and one in Indian subcontinent According

to Oka et al (2007), game birds were developed in some areas in Southeast Asia and both non-game and game birds formed Japanese native chickens Kanginakudru et al (2008) found that Indian domestic birds come from G g spadiceus, G g gallus and G

g murghi, indicating that multiple domestication had occurred independently in different places of Asia including India and showed G g murghi as the most important contribution to Indian domestic chickens Eriksson et al (2008) found that the yellow-skinned gene of domestic chickens was inherited from the G sonneratii therefore indicating a past hybridization event during domestication The Sri Lankan indigenous chickens were more closely related to G gallus and G sonneratii than G lafayettei (Silva et al., 2009) Moreover, while domestication occurred long time ago, contemporary gene flow between wild and domestic chickens is still taking place in some regions (Berthouly et al., 2009)

1.1.2 The threats and opportunities to maintain local chickens

Diversity in chickens refers to genetic characteristics, which established many local breeds and strains with a great variety of plumage colors, morphological and performance traits (Tixier-Boichard et al., 2012) Nevertheless, many traditional and dual-purpose or fancy breeds face with extinction since the development of high

6

productive commercial lines This is aggravated by the fact that financial supports and infrastructures for maintaining the populations are always limited in many developing countries (Pham et al., 2013a) Local breeds make up most of the world’s poultry gene pool They currently play an important role in developing countries, where they represent up to 95 percent of the total poultry population (Besbes et al., 2008) FAO (2011b) reported that 32% of chicken breeds are currently being at risk status The percentage will be higher than this estimate because 42% of chicken breeds are mainly reared in developing countries with unknown status (Figure 3)

Figure 3 Risk status of the world's avian breeds in October 2010

Source: FAO (2011b)

1.1.2.1 Threats

A small number of exotic breeds are rapidly replacing the local breeds in developing countries in the attempt to improve productivity in the traditional poultry sector since the 1990s (Hall, 2004) For economic point of view, crossbreeding is very efficient for improvement of performance for many traits where the first crossbred generation exploits heterosis effect However, the erosion of chicken biodiversity due to indiscriminate crossbreeding may lead to loss of breed identity (Hall, 2004) Secondly, epidemics and disease potentially threatened chicken genetic resource The outbreak of highly pathogenic avian influenza (HPAI) spread across Vietnam About 43,170

Disease resistance within breeds are different between breeds due to ecological adaptation, but there are some conditions (parasite resistance), which differ between breeds (Hall, 2004) For instance, Schou et al (2010) showed that Vietnamese Ri local chicken was more resistant to A galli and S Enteritidis than commercial slow-growing Luong Phuong chicken, while it was susceptible to P multocida The major histocompatibility complex (MHC) gene is highly polymorphic and is able to cope with pathogens which themselves have the capacity to evolve Particular MHC alleles confer resistance to Marek’s disease (Kaufman, 2000) The association between disease challenge and specific alleles testifies to the lower plasticity of the avian MHC (Hall, 2004) In Taiwan conserved chickens, Hsin-Yi responded to the highest antibody levels from Infectious Bursal Disease (IBD) vaccine Quemoy showed a higher antibody response to low pathogenic avian influenza H6N1 virus, Newcastle Disease and IBD vaccines than that of others (Chang, 2011; Chang et al., 2012)

Therefore, it is important to understand and document chicken breeds in developing countries and design appropriate conservation strategies in order to conserve the present and future value of locally adapted breeds The effective management of

8

farm animal genetic resources requires comprehensive knowledge of the breeds’ characteristics, including data on population size and structure, geographical distribution, the environmental production, the historic and cultural values of the breeds, and within- and between-breed genetic diversity as well (Lenstra et al., 2012; Tixier-Boichard et al., 2012) Conserving the gene pools of local breeds also provides opportunities to commercial chicken industries as reservoirs of genes of special utility in their breeding programs (Dessie et al., 2012)

1.1.3 Conservation of chicken genetic resources

Conservation of chicken genetic resources is necessary The ideal conservation strategy is to maintain local breeds in their natural environment (in situ) because they perform better in the natural environment where they are evolving, while ex situ in vivo conservation is the conservation of live animal outside their normal habitat (i.e in a national park or zoo) and is recommended as a complementary strategy to conserve genetic diversity for the future The method of ex situ in vitro conservation uses to store genetic material (i.e semen or embryo) in liquid nitrogen (Hall, 2004; Oldenbroek, 2007) Prioritization of animals for conservation purposes in most cases is based on pedigree information (Goyache et al., 2011; Rizzi et al., 2011) Molecular markers, like microsatellite markers, are widely used for conservation decisions (Zanetti et al., 2010; Cuc et al., 2011; Wilkinson et al., 2012) A commonly used method to select animals for conservation purposes is the optimal contribution method (Toro et al., 2009; Zanetti et al., 2010; Wilkinson et al., 2012), which minimizes the average relationship between individuals, thereby maximizing the conserved genetic diversity By selecting those animals with the smallest pedigree relatedness, we assume that the maximum amount of genetic diversity has been conserved Until recently, the use of optimal contributions in conservation studies is based upon both pedigree information and SNP markers (Engelsma et al., 2012)

1.1.4 The strategies for setting priorities in conservation

9

Basically, there are three strategies for setting priorities in conservation: the maximum-risk strategy, the maximum-diversity strategy and the maximum-utility strategy, which were suggested by Bennewitz et al (2007)

1.1.4.1 The maximum-risk strategy

The maximum-risk strategy is defined as the choice of breeds for conservation based on their risk status The risk of extinction has been used as the initial criterion for conservation program Risk categories for livestock breeds (critical, endangered, and safe; Table 1) have been defined by FAO (2000) Risk status or degree of endangerment

of a breed is inferred from various criteria: number of breeding males and females (FAO, 2000), rate of inbreeding (∆F) (EAAP, 1998) and effective population sizes (Ne) (Meuwissen, 2009) This strategy takes into account only the degree of risk and can be used to prevent the case of short-term loss in 10 years of breeds at high risk of extinction (FAO, 2013)

Table 1 Risk categories used for livestock breeds

Females Males (in percentage) (individual) Critically endangered <100 <5 >40 <50

Not endangered >1000 >20 <5 50<N e <100

The Table was adapted from Bennewitz et al (2007)

Population size per se may not really reflect the risk status of breeds While the estimates of effective population size (Ne) make better predictions on population characteristics Ne can be estimated from the rate of inbreeding in the population and linked to the genetic diversity and therefore reflect the real risk status Rate of inbreeding is used as a criterion for the risk that a breed might become extinct This parameter is also account for the component of within-breed diversity and are calculated

by molecular data (Gizaw et al., 2011) The breeds with high-risk status might not contribute to the total diversity Contributions of a breed to the total genetic diversity of

a species depend on the within- and between-breeds components of diversity Therefore,

10

this strategy could not fully account for conserving the maximum genetic diversity for the future (Bennewitz et al., 2007; Gizaw et al., 2011)

1.1.4.2 The maximum-diversity strategy

The maximum-diversity strategy selects breeds that contribute in a significant way to the overall genetic diversity weighted by both between- and within-breed components of diversity Breeds can be ranked according to their contributions either to the present or to the expected future diversity (Bennewitz et al., 2007) This strategy may be optimal when the funds are available for conservation programs (FAO, 2013)

1.1.4.3 The maximum-utility strategy

This strategy is a potential method for a conservation program The strategy is applied when the extinction probability of breeds, contributions to genetic diversity, estimates for the relative economic values of neutral diversity, of the special traits and for specific values (i.e historical and cultural values) are all available The relative conservation priorities of the breeds used for this strategy will be changed in comparison with the rank by maximum-diversity strategy alone (FAO, 2013) This strategy may be applicable in government institutions in the case of breeds with high utilities

1.2 Material and methods

1.2.1 Description of phenotypic characteristics and performance traits of

chicken breeds from this study

Phenotypic data from chicken populations have proven to be useful for discrimination of local breeds (Tables 2 and 3) In reality, in many cases, either phenotypic data or pedigree information is not available (Tixier-Boichard et al., 2009) Acquiring this type of data is not always possible and it is often time-demanding and costly However, for some phenotypes in chickens, mainly coloration and plumages are linked to single genes with few alleles well described

11

Table 2 Production and reproduction data from Vietnamese domestic chickens

Chicken populations Body weight at 20 weeks (g) Age at first Number of Egg Fertilized Day-old

Vietnamese local chickens (VNN)

Chinese origin chickens (CNO)

Table 3 Production and reproduction data from Taiwan country chickens

Chicken breeds Matured body weight (g) Egg number to Egg weight

12

Therefore it becomes possible to establish the genotype for these characters (Berthouly, 2008) such as the one listed in Tables 4–6 and Figures 4–7

Table 4 List of chicken genes with visible effects

Single locus trait Name of trait Allele Recessive genotype Recessive phenotype

Mb Muffs and beards Mb; mb+ [mb+mb+] No muffs and beards

R and P Comb type R; r+ [r+ r+, p+ p+] Single comb

P; p+ [R-, p+ p+] Rose comb

[r+ r+, P-] Pea comb [R-, P-] Walnut comb

W and Id Shank color W+; w [id+ id+,ww] Green

Id; id+ [id+ id+,W+-] Blue grey

[id-,W+-] White [Id-, ww] Yellow

This Table was adapted from Berthouly (2008)

Table 5 List of complex chicken phenotypes

Comb color BlackCmb, CmbRB, CmbRv, CmbRc Black/Red-black/Dark Red/Light Red Earlobe color BlueO, RedO, WhiteO, WRO Blue/Red/White/White and Red

Shank without feathers TnF

This Table was adapted from Berthouly (2008).

13

Table 6 Phenotypic characteristics of different chicken breeds used in the study

Color of Chicken breeds Origin Comb shape Earlobes Skin Shanks Egg shell Plumage Characteristics Flock size Status

brown

Recessive white Recessive white

Traditional medicine, polydactyly

Traditional medicine Increasing

brown

Black, red, yellow Black, red, yellow

brown

Black-breasted red Yellow

Culture, polydactyly Endangered

brown

Black-breasted red Pale yellow

Culture, big shanks 2-230 Increasing

brown

Black-breasted red Pale brown

brown

Black-breasted red Pale yellow

brown

Black-breasted red Pale yellow

brown

Buff Columbian Buff Columbian

14

Table 6 (continued)

Color of Chicken breeds Origin Comb shape Earlobes Skin Shanks Egg shell Plumage Characteristics Flock size Status

brown

Black-breasted red Buff Columbian

brown

Buff Columbian

Buff Columbian

brown

Buff Columbian Buff Columbian

Dual purpose, dwarf Endangered

brown

Buff Columbian Buff Columbian

brown

Black-breasted red Buff Columbian

brown

Black-breasted white Pale brown

brown

Buff Columbian Buff Columbian

Dual purpose, small beard

Dual purpose, frizzle Endangered

brown

Wild type Wild type

15

Table 6 (continued)

Color of

Black and silkie

brown

Buff Columbian Buff Columbian

brown

Buff Columbian Buff Columbian

White and mottled

brown

Black and red Black and red

Dual purpose, dwarf Endangered

brown

Buff Columbian Buff Columbian

Commercial broiler Increasing

16

Table 6 (continued)

Color of Chicken breeds Origin Comb shape Earlobes Skin Shanks Egg shell Plumage Characteristics Flock size Status

brown

Buff Columbian Buff Columbian

Dam line, research 15-120 Stable

Meat, research 12-100 Stable

brown

Buff Columbian Buff Columbian

Selected for body weight

Stable

brown

Buff Columbian Buff Columbian

Selected for egg number

Stable

brown

Wild type Wild type

17

Table 6 (continued)

This Table was adapted from Lee (2006); Hoang and Vo (2010); and Chang (2011)

Color of Chicken breeds Origin Comb shape Earlobes Skin Shanks Egg shell Plumage Characteristics Flock size Status

brown

Black-breasted red Black-breasted red

brown

Black-breasted red Black-breasted red

Meat, strong shanks Stable

brown

Buff Columbian Buff Columbian

White

White

18

Figure 4 Chicken plumage color patterns

(a) White plumage; (b) black plumage; (c) silky plumage; (d) frizzle plumage; (e) red plumage; (f) buff Columbian; (g) the Dong Tao chicken flock in Hung Yen Province; (h) Choi chicken flock in Me Linh District, Hanoi; and (i) black

H’mong chicken flock in a research center, Hanoi

19

Figure 5 The variations of chicken comb types

(a) Single comb; (b) pea comb; (c) walnut comb; (d) butterfly comb; (e) rose comb; and (f) strawberry comb

Figure 6 The variations of chicken shanks

(a) Black shank; (b) blue shank; (c) green shank; (d) yellow shank; and (e) white and polydactyly shank

Figure 7 Earlobe colors

(a) Blue earlobe; (b) red earlobe; and (c) white earlobe

1.2.2 Populations and study sites

1.2.2.1 Vietnamese domestic chicken populations

In this study, Vietnamese domestic chicken populations have been conserved by the Vietnamese government since the 1990s (Hoang and Vo, 2010) These chicken

20

populations have been totally covered the current Vietnamese domestic chicken gene pool (Figure 8; Chapter 2) One chicken population may be conserved in several isolated places Unfortunately, as in many developing countries, the lack of pedigrees and infrastructure make impossible to estimate the relatedness of chicken, sampling was therefore done randomly

Figure 8 Geographical origin of Vietnamese domestic chicken populations

Vietnamese local chickens from Northwest (labeled in pink) is black H’mong; from Northeast (labeled in green) are Xuoc, Dan

Khao, Lien Minh, Tien Yen and Troi; from Red River Delta (labeled in black) are Dong Tao, Ho, Mia, Mong, Ri, Te, To and Tre; from the South central coast (labeled in purple) is Choi; and from Mekong River Delta (labeled in light blue) are Ac and Tau Vang Chicken breed from the Limestone Mountains (labeled in red) is Red Jungle Fowl Chicken populations of Chinese origin (labeled

in dark blue) are Luong Phuong from Hanoi; Hac Phong, Man and Quy Phi from Quang Ninh Province; and Huong Ke and Te Dong Bac from Lang Son Province.

1.2.2.2 Taiwan country chicken populations

The history of conservation of native chickens in Taiwan started from 1982, when native chickens were collected around the Islands and conserved at National

21

Chung Hsing University (NCHU) experimental farm Six Taiwan conserved chicken populations (TCS: Hsin-Yi, Hua-Tung, Ju-Chi, Nagoya, Quemoy and Shek-Ki) have been conserved with a range from 1 to 38 parents in the first generation (Figure 9; Chapter 4) In the same year, this university began to establish chicken lines for specialized purposes L2 and B strains were selected by NCHU from the same Taiwan native chicken population (Lee, 2006) Both strains were closed populations since their establishment in 1983, while B strain was a male line and a L2 female line for crossing

to produce commercial meat-type chicken The L2 and B strains have now been selected for 24 and 26 generations, respectively, and have been extensively used in research as well as in production (Chao and Lee, 2001; Chen et al., 2007) The L2 strain has higher egg production, but B strain is better for meat production after a long-term selection (Chen et al., 2007)

Figure 9 Geographical origin of Taiwan country chicken populations, commercial lines and Red Jungle

Fowl

The Taiwan commercial native chickens (TCOM; Chapter 3) have prevalently appeared to have a big erected comb, blue shanks, big legs, and a smaller breast with

22

less fat under skin than exotic broiler, and produce small eggs but with larger proportion

of yolk They grow faster than native chickens (i.e Hua-Tung, Hsin-Yi, Ju-Chi and Quemoy) and are preferred by Taiwanese consumers because its meat quality suits to the traditional cooking style and similar plumage color compared to Taiwan native chickens (Lee, 2006) Red feathered chicken was developed in the early 1980s through continuously introgression of imported exotic breed into the small Red feathered chicken and selection for early maturity and production efficiency The smaller body-size Black feathered chickens with white hackles were developed by farmers in Changhua County in the late 1980s In recent years, Golden chickens were produced by crossing Red feathered male with Black feathered female Simultaneously, Classical chickens with red-black plumage and strong shanks were also created Game bird was a fighting chicken and local farmers only used hens and capons for food consumption Hakka people preferred large chickens so their chickens had low egg production (Lee, 2006) Thus, NCHU used Hakka chicken sires crossed with L2 strain dams to establish the synthetic Hakka (NCHU-G101) strain in order to improve its egg performance since

2005

1.2.3 Molecular-based analysis for monitoring genetic diversity

Since the last two decades, several types of molecular markers have become available (Table 7) Among them, microsatellite markers have been developed and used for many chicken diversity studies (Schlötterer, 2004; Lenstra et al., 2012) Microsatellite markers (i.e single sequence repeats, SSRs) usually perform monitoring

of genetic diversity although studies with high density of single nucleotide polymorphism (SNP) are becoming more frequent (Gärke et al., 2012; Kranis et al., 2013) Microsatellites have been used as markers of choice for molecular genetics because of several natural advantages For instance, their total length including the flanking DNA is short enough (100–500 bps) to make them amenable to PCR amplification Moreover, typing for microsatellite loci can be automated, and they are highly polymorphic with a co-dominant mode of inheritance (Crooijmans et al., 1996; Pham, 2005) In addition, they are used to quantify genetic variability within and among livestock populations or breeds The multilocus microsatellite data can be used to assign

23

individuals to genetically similar groups at the population or breed levels (Berthouly, 2008) The standard panel of microsatellite markers (Table 8) within the AvianDiv project has been widely used for many chicken population studies (Hillel et al., 2003; Granevitze et al., 2007; Berthouly et al., 2008) and recommended for molecular characterization of farm animals (FAO, 2011a)

Table 7 History of different autosomal markers in domestic animals

Molecular markers No loci/chromosome No allele/marker Year

2 2005

This Table was adapted from Schlötterer (2004) and Lenstra et al (2012)

Table 8 Summary information of 22 microsatellite markers used in this thesis

Locus Chromosome

Genebank accession number Allele size

Fluorescent label

24

1.2.3.1 Measurement of genetic diversity within populations

Genetic diversity within population has been measured by several parameters

1.2.3.1.1 Number of alleles

The number of alleles can be determined by direct count The number of private alleles is equivalent to the number of alleles unique to a single population in the data set In this thesis, the number of alleles were calculated by GENALEX 6 (Peakall and Smouse, 2006)

1.2.3.1.3 Polymorphic informative content

The polymorphic informative content (PIC) suggested by Botstein et al (1980)

at both marker and population level is computed as

The expected heterozygosity (HE) or gene diversity was defined by Nei (1973)

as the probability that two alleles chosen at a randomly given locus within a population are different alleles The HE is calculated as the proportion of heterozygotes under the assumption of Hardy-Weinberg equilibrium (HWE) as E 1 i2

i

frequency of allele i in the population The level of heterozygosity to what we expect under HWE is compared to observed heterozygosity If the observed heterozygosity is lower than expected, the excess of homozygous invoke forces such as inbreeding or subpopulation structure, known as Wahlund effect (Wahlund, 1928) On the contrary, if observed heterozygosity is higher than expected, the excess of heterozygotes may suggest an isolate-breaking effect or nonrandom mating (Loywyck, 2007) The heterozygosity was computed using MOLKIN 3.0 (Gutiérrez et al., 2005) Tests for deviation from HWE were performed for each locus in each population: (i) for heterozygote deficiency and (ii) for heterozygote excess with GENEPOP 4.1.4 (Rousset, 2008) using the Fisher’s exact test and the Markov chain algorithm to calculate the P-values (Guo and Thompson, 1992) Test of significance was corrected with sequential Bonferroni correction on multilocus (Rice, 1989)

1.2.3.2 Measurement of genetic variability between populations

1.2.3.2.1 F-statistics and genetic distances between populations

26

The values of F-statistics estimated from genotypic frequencies analyzed from each defined subpopulations The deviations of genotypic frequencies from HWE in a subdivided population can be measured by the three parameters, FIS, FIT and FST

suggested by Wright (1951) FIS and FST are the correlations between the two uniting gametes relative to the subpopulation and relative to the total population, whereas FST is the correlation between two gametes drawn at random from each subpopulation as (1 –

FIT) = (1 – FIS)(1 – FST) The popular method of Weir and Cockerham (1984) have developed a variance based method for estimation of F-statistics This method takes into account multiple loci and unequal sample sizes across subpopulations (Lenstra et al., 2012) However, their estimations seemed to be exhibited bias at a small sample size of lower than 50 individuals (Ruzzante, 1998) The Weir and Cockerham (1984) estimators of F-statistics were computed by FSTAT 2.9.3.2 (Goudet, 2002), and tests them using randomization methods

Genetic distances explain the extent of the genetic differences between populations measured by allele frequencies Each kind of distance is based on different assumptions of selection, genetic drift and bottlenecks The DA genetic distances (Nei et

al., 1983) were computed by the formula A 1 1 ij ij,

by POPULATIONS 1.2.32 (Langella, 1999)

1.2.3.2.2 Bayesian genotypic clustering approaches

Two well-known Bayesian clustering methods have been implemented in available software packages such as STRUCTURE and BAPS (Bayesian Analysis of Population Structure) The individuals are assigned to inferred populations based on multilocus genotypes without a priori population descriptions if their genotypes indicate that they are admixed The methods assume that all markers are at HWE Both

27

STRUCTURE and BAPS software give the assignment probabilities of each individual

to each cluster, but differ in their approach to estimating admixture (Wilkinson et al., 2011) STRUCTURE 2.3.3 (Pritchard et al., 2000; Falush et al., 2003) used a Markov chain Monte Carlo (MCMC) method and estimates the natural logarithm of the probability (Pr) of the observed genotypic array (X), given a predefined number of clusters (K) in the data set (ln Pr(X|K)) The estimate of ln Pr(X|K) is a direct indicator

of the posterior probability of having K clusters, given the observed genotypic array Nevertheless, ln Pr(X|K) is really difficult to estimate (Pritchard et al., 2000) Evanno et

al (2005) calculated an ad hoc statistic, ∆K, to identify the optimal K value Basically, it

is based on the second order rate of change of Pr(X|K) with respect to K, the magnitude

of the estimated values indicates the strength of the subdivided population

The recent program of BAPS v5.3 (Corander and Marttinen, 2006) uses a

‘greedy stochastic optimization algorithm’ to directly estimate the most likely K and assign individuals to clusters (Corander et al., 2008) Either the value of K can be pre-defined to investigate the clustering solutions of populations with successive K values

or the algorithm searches for the most likely K value For each K value, BAPS searches for the optimal partitions, stores them internally and, after all K values have been processed, it merges the stored results according to log-likelihood values The estimating individual admixture is a two-tiered approach First, the clustering solutions

of populations were determined and then the admixture of genotypes was quantified by establishing the ancestral sources of alleles for each individual with respect to the determined clusters The evidence for admixture was considered significant for individuals with P-values < 0.05 (Corander and Marttinen, 2006)

1.2.3.2.3 Principal component analysis

The principal component analysis (PCA) is a mathematical procedure that uses

an orthogonal linear transformation that transformed the data to a new coordinate system, such that the first greatest variance by any projection data comes to lie on the first coordinate, the second greatest variance on the second coordinate, and so on (Berthouly, 2008) This method uses the eigenvectors of the covariance matrix and it finds only the independent axes of the data under Gaussian assumption Groups are then

28

created by forming composite axes that maximize the overall distance between data without a priori, because it does not take into account the class group information (Berthouly, 2008), in addition no HWE is assumed PCA is a powerful method for the analysis of admixed populations with big data sets (Gärke et al., 2012) This analysis was based on the function available in the ADE4 package (Chessel et al., 2005) implemented in R program (R Core Development Team, 2006)

1.2.3.2.4 Multidimensional scaling

Multidimensional scaling (MDS) was used to represent FST genetic distances in

a two-dimensional space using SPSS 19.0 (SPSS Inc., 2010) S-stress is a measure of fit ranging from 1 to 0 (perfect fit) For a good fit, the measure of stress is less than 0.15

An R2 value illustrates the percentage of variance in the model is explained by the dimensions (Young and Harris, 1993)

1.2.3.3 Quantification of contributions to diversity

The contribution of each analyzed population to the diversity of the whole data set can be assessed using the methods described by Caballero and Toro (2002) and Petit

et al (1998), and the Weitzman’s method (Weitzman, 1993) modified by Ollivier and Foulley (2005)

The method described by Caballero and Toro (2002) proposed setting priorities for conservation using as criterion the maintenance of the maximum overall Nei’s (1987) gene diversity (GD) in the preserved set of breeds This is equivalent to minimize the overall molecular coancestry ( f ) because GD =1− f The average molecular coancestry over an entire metapopulation ( f) consisting of n subpopulations, subpopulation i with Ni breeding individuals, fij is the average pairwise coancestry between individuals of subpopulations i and j, including all Ni×Nj pairs, and fii is the

average pairwise coancestry within subpopulation i: 1

1

,

n

ij j n

j i

ii

D NN