GENETIC CHARACTERIZATION OF INDIGENOUS 
GOAT POPULATIONS OF ETHIOPIA USING 
MICROSATELLITE DNA MARKERS 
  
 
 
 THESIS SUBMITTED TO THE 
 NATIONAL DAIRY RESEARCH INSTITUTE 
 (DEEMED UNIVERSITY) 
 KARNAL (HARYANA) 
 IN PARTIAL FULFILMENT OF THE REQUIREMENTS 
 FOR THE AWARD OF THE DEGREE OF 
 
 
 
DOCTOR OF PHILOSOPHY 
IN 
ANIMAL GENETICS & BREEDING 
 
By 
TESFAYE ALEMU TUCHO 
 
 
DIVISION OF DAIRY CATTLE BREEDING  
NATIONAL DAIRY RESEARCH INSTITUTE 
(ICAR) 
KARNAL ? 132 001 (HARYANA) INDIA 
2004 
 
REGN. No.1062k01
GENETIC CHARACTERIZATION OF INDIGENOUS  
GOAT POPULATIONS OF ETHIOPIA USING  
MICROSATELLITE DNA MARKERS 
 
By 
TESFAYE ALEMU TUCHO 
 
A THESIS SUBMITTED TO THE 
 NATIONAL DAIRY RESEARCH INSTITUTE 
 (DEEMED UNIVERSITY) 
 KARNAL (HARYANA) 
 IN PARTIAL FULFILMENT OF THE REQUIREMENTS 
 FOR THE AWARD OF THE DEGREE OF 
 
 DOCTOR OF PHILOSOPHY 
 IN 
 ANIMAL GENETICS & BREEDING 
 
 
 Approved by: 
 
 
(                               )                   (OLIVIER HANOTTE) ( B.R. YADAV ) 
EXTERNAL EXAMINER                       CO-GUIDE       MAJOR ADVISOR & CHAIRMAN 
(ADVISOR) (GUIDE) 
  
  
 
 Members, Advisory Committee 
 
1. Dr. V.K. Batish _______________________ 
Principal Scientist, DM Division 
 
2. Dr. A.K. Chakravarty _______________________ 
Senior Scientist, DCB Division 
 
3. Dr. A. K. Gupta  _______________________ 
Scientist, DCB Division 
 
4. Dr. T.K. Mohany _______________________ 
Senior Scientist, DCB Division 
 
5. Dr. Surjeet Singh  _______________________ 
Senior Scientist, Animal Biochemistry Division 
 
 
 ii
  
 
 
CERTIFICATE 
 
This is to certify that the thesis entitled, "GENETIC CHARACTERIZATION OF 
INDIGENOUS GOAT POPULATIONS OF ETHIOPIA USING 
MICROSATELLITE DNA MARKERS" submitted by Mr. TESFAYE ALEMU 
TUCHO towards the partial fulfilment of the award of the degree of DOCTOR 
OF PHILOSOPHY in ANIMAL GENETICS & BREEDING of the 
NATIONAL DAIRY RESEARCH INSTITUTE (DEEMED UNIVERSITY), 
Karnal (Haryana), India, is a bonafide research work carried out by him under 
our supervision, and no part of the thesis has been submitted for any other degree 
or diploma. 
 
 
 (OLIVIER HANOTTE) ( B.R. YADAV ) 
    CO-GUIDE                               MAJOR ADVISOR & CHAIRMAN 
    (ADVISOR)      (GUIDE) 
  
 Dated:____________                                            Dated:_______________ 
 iii
 ACKNOWLEDGEMENTS 
 
?Then Samuel took a stone and set it up between Mizpah and Shen, and called 
its name EBENEZER, saying ??Thus far the Lord has helped us?? (I Samuel 7:12).  I 
give thanks and all the glory to the Almighty God for seeing me, not only through four 
years of relentless effort to acquire acknowledge, but also a frantic race against time 
and resources. Thus far the Lord has brought me! I am grateful!! 
The work presented in this thesis would not have been accomplished without the 
support and assistance of many people.  It is impossible to adequately acknowledge 
everyone who contributed to this work, in one way or the other, within the space 
provided here.  Should I miss your name please do not consider it as an act of 
ingratitude. I consider this achievement yours too and I am grateful to all and sundry 
whose contribution(s) have enabled this work to be concluded fruitfully.  
I am grateful to my supervisors Dr. B. R. Yadav and Olivier Hanotte, for giving 
me an opportunity to work with them, for their firm support, for patiently teaching me 
many things I know today about characterization of Indigenous Animal Genetic 
Resources.  I would also like to thank Dr Edward Rege for accepting me as a graduate 
fellow at ILRI. 
I would like to acknowledge Dr. Han Jianlin for assisting me with the data 
analysis and for reading drafts of my thesis.  I am indebted to Russ Kruska and Michael 
Arunga for providing me with the geographic distance matrix for the Ethiopian goat 
populations.  Special thanks to Joel Ochieng, who ?got me started? and trained me on 
use of the various lab equipment for my work as well as providing me with invaluable 
technical support.  It has been a great pleasure to work with the members of the Animal 
Genetics Resources group.  
I am also thankful to the members of my Advisory Committee: Dr. V.K. Batish, 
Dr. A.K. Chakravarty, Dr. A.K. Gupta, Dr. T.K. Mohany and Dr. Surjeet Singh, for 
their suggestions and support during my study. 
To all my colleagues in Lab. 7; Paul Wafula, Moses Ogugo, Daniel Mwangi, 
Nemwel Nyamweya, John Wambugu, Samuel Njomo, Joel Mwakaya, Joseph Nthale, 
Henrie Gathuo, Sarah Waita, Sarah Nyongesa, Gerold Kierstein, Sonya Kierstein, 
Donald Rugira Kugonza, Batsuuri Lkhagua, Yeon Heun-Seong and Chang Yeon Cho, 
Suzan Musembi and Richard Kang?ethe. Thank you all so much for all your 
contributions into this thesis.  Special thanks also to Manish Thakur and Zenu who ?got 
 iv
 me started?? and trained me while I was at NDRI Livestock Genome Analysis 
laboratory at the beginning. 
I would like to acknowledge all the individuals who participated in collecting 
the goat samples for this project, especially Dr. Workneh Ayalew, Dr. Markos Tibo, 
Ato Girma Abebe, Ato Michal Temesgen, and Ato Eshetu Zerihun. 
To my colleagues and fellow PhD-students in the lab, especially David Odongo, 
Xuebin Qi, David Menge, Joseph Ng?ang?a, Lucy Muthoni and Mujibi Fidalis (Msc 
student), thanks all for your scientific, social and friendly discussions and ideas that we 
shared. 
I would like to acknowledge the contributions made by the following 
organizations: The Ethiopian Agricultural Research Organization (EARO) and Oromiya 
Agricultural Research Institute (OARI) for the financial assistance and study leave; 
International Livestock Research Institute (ILRI) and the Department for International 
Development (DFID) UK, for financial support, and all the field staff from the Ministry 
of Agriculture, who allowed the field team easy access to the site of interest. 
To my beloved wife, Meaza and our children, Elim (daughter), Yeabkal, 
Yididiya and Kebron (Sons), your prayer, sacrifice and support has been with me to the 
end, and I would like to thank you so much and say God bless you abundantly, I love 
you all from the bottom of my heart.  For the tremendous family prayer and support I 
would like to thank my Dad (Alemu Tucho), brothers, Hailu and Tsegaye, sisters, 
Asnakech (with her family), Meseret, Tayech and Tseganesh, my mother-in-law W/ro 
Ejigayehu and Father-in-law, Ato Workagegnehu, and all my brothers and sisters-in-
 law. 
 
Dated November, 2004     (Tesfaye Alemu Tucho) 
 v
 CONTENTS 
 
CHAPTER ONE .............................................................................................................1 
1.0. INTRODUCTION ...................................................................................................1 
 
CHAPTER TWO ............................................................................................................5 
2.0 LITERATURE REVIEW ........................................................................................5 
2.1. CLASSIFICATION OF GOATS .....................................................................................5 
2.2. DOMESTICATION OF GOATS.....................................................................................6 
2.3 GOAT GENETIC RESOURCE OF ETHIOPIA...................................................................8 
2.4 CONSERVATION OF DOMESTIC ANIMALS DIVERSITY...............................................24 
2.4.1 Meaning and importance of domestic animals diversity ...............................24 
2.4.2 Loss of Genetic Diversity in Domestic Animals ...........................................27 
2.4.3 Loss of Diversity within Population ..............................................................27 
2.4.4 Loss of Diversity among Populations............................................................29 
2.4.5 Approach to conservation of animal genetic diversity ..................................30 
2.5 MOLECULAR TECHNIQUES FOR THE ASSESSMENT OF ANIMAL GENETIC DIVERSITY 34 
2.5.1 Rationale for molecular genetic approach .....................................................34 
2.5.2 Molecular techniques in the study of animal genetic diversity .....................36 
2.5.3 Application of Microsatellite markers for the analysis of population ...........42 
2.6 METHODS FOR THE RECONSTRUCTION OF PHYLOGENETIC TREE .............................49 
2.6.1 Distance methods...........................................................................................49 
2.6.2 Maximum parsimony method........................................................................51 
2.6.3 Maximum likelihood......................................................................................52 
2.7. STATISTICAL ANALYSIS.........................................................................................53 
2.7.1. Hardy-Weinberg equilibrium (HWE)...........................................................53 
2.7.2. Linkage disequilibrium (LD) ........................................................................54 
2.7.3. Genetic diversity ...........................................................................................55 
2.7.4. Genetic relationship between the populations ..............................................55 
 
 
 vi
 CHAPTER THREE......................................................................................................62 
3.0. MATERIALS AND METHODS ..........................................................................62 
3.1. POPULATIONS/BREEDS AND SAMPLING STRATEGY................................................62 
3.1.1.  Animals........................................................................................................62 
3.1.2. Sampling protocol.........................................................................................63 
3.2. DNA EXTRACTION ................................................................................................66 
3.2.1. DNA extraction from peripheral blood lymphocytes (PBL) ........................66 
3.2.2. DNA extraction from FTA? blood cards (Whatman? Bioscience) ............66 
3.3. MICROSATELLITE MARKERS, PCR CONDITIONS AND GENOTYPING .......................67 
3.4. STATISTICAL ANALYSIS...................................ERROR! BOOKMARK NOT DEFINED. 
3.4.1. Hardy-Weinberg equilibrium (HWE)...........Error! Bookmark not defined. 
3.4.2. Linkage disequilibrium (LD) ........................Error! Bookmark not defined. 
3.4.3. Genetic diversity ...........................................Error! Bookmark not defined. 
3.4.4. Genetic relationship between the populations .............Error! Bookmark not 
defined. 
 
CHAPTER FOUR.........................................................................................................70 
4.0. RESULTS ...............................................................................................................79 
4.1. GENETIC DIVERSITY AND RELATIONSHIP BETWEEN ETHIOPIAN AND NON-
 ETHIOPIAN GOAT .........................................................................................................79 
4.1.1. Test for deviation from Hardy-Weinberg equilibrium (HWE).....................79 
4.1.2. Linkage disequilibrium.................................................................................83 
4.2.3. Inference of the recent bottleneck within populations..................................83 
4.3. THE EXTENT AND DISTRIBUTION OF THE GENETIC DIVERSITY WITHIN THE 20 
POPULATIONS...............................................................................................................89 
4.3.1. Mean number of alleles and heterozygosities...............................................89 
4.4. GENETIC RELATIONSHIP BETWEEN THE POPULATIONS...........................................99 
4.4.1. Nei?s standard genetic distance (DS).............................................................99 
4.4.2. Nei?s DA genetic distances............................................................................99 
4.4.3. Phylogenetic analysis..................................................................................102 
4.4.4. Principal component analysis (PCA)..........................................................106 
4.4.5. Inference of population structure................................................................110 
 vii
 4.5. GENETIC DIFFERENTIATION.................................................................................113 
4.5.1. GST and FST .................................................................................................113 
4.5.2. Analysis of molecular variance (AMOVA)................................................118 
4.6. INDIVIDUAL SPECIFIC ANALYSIS..........................................................................120 
4.7. GENETIC DIVERSITY AND RELATIONSHIP WITHIN AND BETWEEN ETHIOPIAN GOAT 
POPULATIONS.............................................................................................................124 
4.7.1. Hardy-Weinberg equilibrium (HWE).........................................................124 
4.7.2. Linkage disequilibrium...............................................................................127 
4.7.3. Analysis of inbreeding coefficient within the Ethiopian goat populations.128 
4.7.4. Analysis of Fixation Indices or F-statistics.................................................129 
4.8. THE EXTENT AND DISTRIBUTION OF THE GENETIC DIVERSITY WITHIN THE 
ETHIOPIAN POPULATIONS...........................................................................................132 
4.8.1. Mean number of alleles and heterozygosities.............................................132 
4.9. GENETIC RELATIONSHIP BETWEEN THE ETHIOPIAN POPULATIONS.......................136 
4.9.1. Nei?s standard genetic distance (DS)...........................................................136 
4.9.2. Nei?s DA genetic distances..........................................................................136 
4.9.3. Phylogenetic relationship among Ethiopian goat populations....................139 
4.9.4. Principal component analysis (PCA)..........................................................147 
4.9.5. Inference of population structure................................................................150 
4.9.6. Analysis of gene flow among populations..................................................152 
4.10. GENETIC DIFFERENTIATION...............................................................................154 
4.10.1. Test on allele size information content .....................................................154 
4.10.2. Genetic differentiation between populations ............................................156 
4.10.3. Analysis of Isolation by Distance (IBD)...................................................162 
4.10.4. Analysis of molecular variance (AMOVA)..............................................164 
4.10.5. Individual specific analysis.......................................................................164 
4.10.6. Marginal loss of genetic diversity.............................................................168 
 
CHAPTER FIVE ........................................................................................................170 
5.0. DISCUSSION.......................................................................................................170 
5.1. WITHIN POPULATION GENETIC DIVERSITY...........................................................170 
5.2. BETWEEN POPULATION GENETIC DIVERSITY .......................................................172 
 viii
 5.3. GENETIC RELATIONSHIP BETWEEN GOAT POPULATIONS ......................................175 
 
CHAPTER SIX ...........................................................................................................179 
6.0 SUMMARY AND CONCLUSIONS ...................................................................179 
6.1 SUMMARY............................................................................................................179 
6.2 CONCLUSIONS......................................................................................................188 
6.3. RECOMMENDATION.............................................................................................189 
 
REFERENCES............................................................................................................190 
 
APPENDICES.............................................................................................................215 
 
 ix
 LIST OF TABLES 
Table 1: The main differences between goats and sheep.................................................5 
Table 2: Populations/types used in the analysis of genetic diversity of Ethiopian  
goats and the reference Breeds/populations from outside Ethiopia..................65 
Table 3: Autosomal microsatellite loci used in the study and their characteristics. ......68 
Table 4: Results of Fisher?s exact test for Hardy-Weinberg equilibrium  
for 20 populations and 18 microsatellite markers. ...........................................81 
Table 5: Results of Fisher?s exact test for Hardy-Weinberg equilibrium  
for 20 populations and 15 microsatellite markers. ...........................................82 
Table 6: The Wilcoxon test of recent bottleneck effect for each population.................87 
Table 7: Observed number of alleles, mean number of alleles expected under  
SMM and IAM for each population and each locus.........................................94 
Table 8: List of specific alleles observed only in Ethiopian or reference populations. .97 
Table 9: Expected heterozygosity  observed heterozygosity and mean  
number of alleles of all 20 populations. ...........................................................98 
Table 10: Nei?s standard genetic distance for 11 Ethiopian and 9 reference 
populations/breeds genotyped, using 15 microsatellite markers. ...................100 
Table 11: Nei?s genetic distance for 11 Ethiopian and 9 reference populations/breeds 
genotyped, using 15 microsatellite markers. ..................................................101 
Table 12: Ranked principal component values of allele frequencies from 20 
populations using 15 microsatellite markers ..................................................108 
Table 13: Measure of genetic differentiation in Ethiopian and reference goat 
populations......................................................................................................115 
Table 14: Pair-wise tests of population differentiation between the 11 Ethiopian 
populations and 9 reference breeds.................................................................116 
Table 15: Summary of genotypic differentiation tests for all the 190  
population pairs at 15 loci. .............................................................................117 
Table 16:. Analyses of molecular variance for 20 goat populations using  
genotype data from 15 microsatellite loci. .....................................................119 
 x
 Table 17: Percent of correct individual-population assignment from 15 microsatellite 
loci using different methods. ..........................................................................121 
Table 18: Percent of correct individual-population assignment (for 20 populations) 
estimated for each microsatellite locus using the Bayesian method...............122 
Table 19: Frequency of correct assignment of individuals to the  
20 populations using genotype information from 15 microsatellite  
loci using the Bayesian method. .....................................................................123 
Table 20: Fisher?s exact test for Hardy-Weinberg equilibrium for all Ethiopian 
populations at 18 loci......................................................................................125 
Table 21: Results of Fisher?s exact test for Hardy-Weinberg equilibrium for all 
Ethiopian populations and 15 microsatellite markers.....................................126 
Table 22: Inbreeding coefficients (FIS estimator) estimated from 15 microsatellite loci 
for Ethiopian goat populations. ......................................................................128 
Table 23: Results of F-statistics (Weir and Cockerham, 1984) for 15 microsatellite loci 
by jackknifing across the 11 Ethiopian goat populations. ..............................131 
Table 24: Expected heterozygosity (HE), observed heterozygosity (HO) and mean 
number of alleles (MNA) of all 11 Ethiopian populations. ............................133 
Table 25: The expected heterozygosity (HE) and observed heterozygosity (HO) per 
locus and population for Ethiopian goats. ......................................................134 
Table 26: The expected heterozygosity (HE) and observed heterozygosity (HO) per 
locus and population for Ethiopian goats. ......................................................135 
Table 27: Nei?s standard genetic distance (DS) matrix for 11 Ethiopian populations and 
1 reference population from West Africa. ......................................................137 
Table 28: Nei?s DA genetic distance matrix for 11 Ethiopian populations and 1 
reference population from West Africa. .........................................................138 
Table 29: Ranked PC values of allele frequencies from 11 Ethiopian populations using 
15 microsatellite markers................................................................................148 
Table 30: The number of migrants per generation, (Wright, 1969) among 11 goat 
populations estimated from Weir and Cockerham?s (1984) FST estimator using 
the equation: FST = 1/ (1+4Nm) or Nm = (1-FST)/4FST. .................................153 
 xi
 Table 31: Differentiation among Ethiopian goat populations, estimated by global RST, 
mean pRST values per locus and for a multilocus average. ...........................156 
Table 32: Measure of population differentiation in Ethiopian goat populations. ........158 
Table 33: Pair wise tests of population differentiation between the 11 Ethiopian goat 
populations.  ...................................................................................................159 
Table 34: Pair wise test of population differentiation between the 11 Ethiopian and 1 
reference population.  . ...................................................................................160 
Table 35: Summary of genotypic differentiation tests for all the 55 population 
combination pairs of Ethiopian populations for all the loci. ..........................161 
Table 36: Analysis of molecular variance (AMOVA) for 11 Ethiopian goat 
populations, using genotype data from 15 microsatellite loci. .......................164 
Table 37: The accuracy of different methods to assign individuals to the source 
population. ......................................................................................................165 
Table 38: Percent of correct individual-population assignment (only for Ethiopian 
populations) estimated for each microsatellite locus using the Bayesian 
method. ...........................................................................................................166 
Table 39: Frequency of correct assignment of individuals to the 11 Ethiopian 
populations using genotype information from 15 microsatellite loci using the 
Bayesian method.............................................................................................167 
Table 40: Marginal loss of genetic diversity according to the Weitzman approach....169 
 
 xii
 LIST OF FIGURES 
Figure 1: Afar female goat.............................................................................................10 
Figure 2: Abergalle male goat .......................................................................................11 
Figure 3: Abergalle female goat ....................................................................................11 
Figure 4: Arsi-Bale male goat........................................................................................12 
Figure 5: Woyto-Guji Make goat ..................................................................................14 
Figure 6: Woyto-Guji female goat.................................................................................14 
Figure 7: Hararghe Highland female goat .....................................................................15 
Figure 8: Short-eared Somali male goat ........................................................................16 
Figure 9: Short-eared Somali female goat .....................................................................17 
Figure 10: Long-eared Somali female goat ...................................................................18 
Figure 11: Long-eared Somali male goat ......................................................................18 
Figure 12: Central Highland male goat .........................................................................19 
Figure 13: Western Highland male goat ........................................................................20 
Figure 14: Western Highland female goat.....................................................................21 
Figure 15: Western Lowland male goat.........................................................................22 
Figure 16: Western Lowland female goat .....................................................................22 
Figure 17: Keffa male goat ............................................................................................23 
Figure 18: Keffa female goat .........................................................................................24 
Figure 19a: Map of Ethiopia showing distribution of goat populations/types sampled.
 ..........................................................................................................................63 
Figure 19b - 19g: Gel images and scoring iages??????????????70-75 
Figure 20: Distributions of allele frequency from 15 microsatellite loci  
in 20 goat populations. .....................................................................................88 
Figure 21: Histogram showing the allele size range and their frequency detected in 776 
goat individuals of 20 populations at 15 microsatellite loci. ............................90 
 xiii
 Figure 22: Phylogenetic tree representing the genetic relationships between the 11 
Ethiopian populations and 9 reference populations........................................103 
Figure 23: Phylogenetic tree representing the genetic relationships between the 11 
Ethiopian populations and 9 reference populations........................................104 
Figure 24: Phylogenetic tree representing the genetic relationships between the 11 
Ethiopian populations and 9 reference populations........................................105 
Figure 25: Phylogenetic tree representing the genetic relationships between the 11 
Ethiopian populations and 9 reference populations........................................106 
Figure 26: Two-dimensional scatter plot representing the relationship  
between PC1 and PC2 using allele frequency data  
from 15 microsatellite loci typed in 20 populations. ......................................109 
Figure 27: Two-dimensional scatter plot representing the relationship  
between PC1 and PC3 using allele frequency data  
from 15 microsatellite loci typed in 20 populations. ......................................109 
Figure 28: Two-dimensional scatter plot representing the relationship  
between PC2 and PC3 using allele frequency data  
from 15 microsatellite loci typed in 20 populations. ......................................110 
Figure 29: Estimated population structure displayed with population Q-matrix. .......112 
Figure 30: Estimated population structure (graph)  
displayed with individual Q-matrix.. ..............................................................113 
Figure 31: Phylogenetic tree representing the genetic relationship between the 11 
Ethiopian populations with GUB as an out-group population. ......................140 
Figure 32: Phylogenetic tree representing the genetic relationship between the 11 
Ethiopian populations with GUB as an out-group population. ......................141 
Figure 33: Phylogenetic tree representing the genetic relationship between the 11 
Ethiopian populations. ....................................................................................142 
Figure 34: Phylogenetic tree representing the genetic relationship between the 11 
Ethiopian populations. ....................................................................................143 
Figure 35: UPGMA tree representing the genetic relationship between the 11 Ethiopian 
populations with GUB as an out-group population. .......................................144 
 xiv
 Figure 36: DA-UPGMA tree representing the genetic relationship between the 11 
Ethiopian populations with GUB as an out-group population. ......................145 
Figure 37: DS-UPGMA tree representing the genetic relationship between the 11 
Ethiopian populations. ....................................................................................146 
Figure 38: DA-UPGMA tree representing the genetic relationship between the 11 
Ethiopian populations. ....................................................................................147 
Figure 39: Two-dimensional graph representing the relationship between PC1 and PC2 
using allele frequency data from 15 microsatellite loci typed in 11 Ethiopian 
goat populations..............................................................................................149 
Figure 40: Two-dimensional graph representing the relationship between PC1 and PC3 
using allele frequency data from 15 microsatellite loci typed in 11 Ethiopian 
goat populations..............................................................................................149 
Figure 41: Two-dimensional graph representing the relationship between PC2 and PC3 
using allele frequency data from 15 microsatellite loci typed in 11 Ethiopian 
goat populations..............................................................................................150 
Figure 42: Estimated population structure displayed with population Q-matrix. .......151 
Figure 43: Estimated population structure displayed with individual Q-matrix. ........151 
Figure 44: Genetic differentiation among 11 Ethiopian goat populations. .................163 
 
 xv
 APPENDICES 
Appendix 1: Summary of linkage disequilibrium between all markers across all 20 
populations......................................................................................................216 
Appendix 2: Matrix of all pair-wise combinations of loci showing the linkage 
disequilibrium (for 20 populations). ...............................................................220 
Appendix 3: Allele frequency for locus BM1818 for all the 20 populations. .............221 
Appendix 4: Allele frequency for locus BMS1494 for all the 20 populations. ...........222 
Appendix 5: Allele frequency for locus ILSTS05 for all the 20 populations..............223 
Appendix 6: Allele frequency for locus ILSTS11 for all the 20 populations..............224 
Appendix 7: Allele frequency for locus ILSTS44 for all the 20 populations..............225 
Appendix 8: Allele frequency for locus ILSTS87 for all the 20 populations..............226 
Appendix 9: Allele frequency for locus INRA05 for all the 20 populations...............227 
Appendix 10: Allele frequency for locus INRA63 for all the 20 populations.............228 
Appendix 11: Allele frequency for locus INRA132 for all the 20 populations...........229 
Appendix 12: Allele frequency for locus MAF35 for all the 20 populations..............230 
Appendix 13: Allele frequency for locus MAF65 for all the 20 populations..............231 
Appendix 14: Allele frequency for locus MAF209 for all the 20 populations............232 
Appendix 15: Allele frequency for locus OarAE129 for all the 20 populations. ........233 
Appendix 16: Allele frequency for locus OarFCB304 for all the 20 populations.......234 
Appendix 17: Allele frequency for locus SRCRSP03 for all the 20 populations........235 
Appendix 18: Test for genetic differentiation between pairs of 20 populations  
for each locus..................................................................................................236 
Appendix 19: Summary of linkage disequilibrium between all markers across all 
Ethiopian populations. ....................................................................................244 
Appendix 20: Matrix of all pair-wise combinations of loci showing the linkage 
disequilibrium (for 11 Ethiopian populations). ..............................................247 
 xvi
 Appendix 21: Allele frequency for locus BM1818 for all the 11 Ethiopian populations.
 ........................................................................................................................248 
Appendix 22: Allele frequency for locus BMS1494 for all the 11 Ethiopian 
populations......................................................................................................248 
Appendix 23: Allele frequency for locus ILSTS05 for all the 11 Ethiopian populations.
 ........................................................................................................................249 
Appendix 24: Allele frequency for locus ILSTS11 for all the 11 Ethiopian populations.
 ........................................................................................................................249 
Appendix 25: Allele frequency for locus ILSTS44 for all the 11 Ethiopian populations.
 ........................................................................................................................250 
Appendix 26: Allele frequency for locus ILSTS87 for all the 11 Ethiopian populations.
 ........................................................................................................................250 
Appendix 27: Allele frequency for locus INRA05 for all the 11 Ethiopian populations.
 ........................................................................................................................251 
Appendix 28: Allele frequency for locus INRA63 for all the 11 Ethiopian populations.
 ........................................................................................................................251 
Appendix 29: Allele frequency for locus INRA132 for all the 11 Ethiopian 
populations......................................................................................................252 
Appendix 30: Allele frequency for locus MAF35 for all the 11 Ethiopian populations.
 ........................................................................................................................252 
Appendix 31: Allele frequency for locus MAF65 for all the 11 Ethiopian populations.
 ........................................................................................................................253 
Appendix 32: Allele frequency for locus MAF209 for all the 11 Ethiopian populations.
 ........................................................................................................................253 
Appendix 33: Allele frequency for locus OarAE129 for all the 11 Ethiopian 
populations......................................................................................................254 
Appendix 34: Allele frequency for locus OarFCB304 for all the 11 Ethiopian 
populations......................................................................................................255 
 xvii
 Appendix 35: Allele frequency for locus SRCRSP03 for all the 11 Ethiopian 
populations......................................................................................................256 
Appendix 36: Test for genetic differentiation between pairs of Ethiopia goat 
populations for each locus. .............................................................................257 
 xviii
 ABSTRACT 
 
Indigenous goats are an important resource for African farmers, providing meat, milk, 
manure, fibres and hides, and satisfying various cultural and religious functions.  In Africa, the 
survival of many indigenous goat populations is threatened by diseases, adverse climatic 
conditions, civil strife, pressure of economic development, abandonment of traditional farming 
practices, and more importantly through crossbreeding or replacement with animals from the 
developed world.  
Indigenous Ethiopian goat genetic resources have been classified phenotypically into 11 
types.  These are Abergalle, Arsi-Bale, Afar, Central Highland, Gumez, Hararghe Highland, 
Keffa, Long-eared Somali, North-West Highland, Short-eared Somali, and Woyto-Guji.  In the 
present investigations two main studies were carried out.  The first study dealt with the genetic 
diversity within and between the Ethiopian goats.  The second study dealt with the genetic 
diversity and relationship of the Ethiopian goat in relation to goat populations from other 
countries of African and outside African continents.  For the first study, the 11 indigenous 
Ethiopian populations were sampled.  In the second study, five populations from other African 
countries [(Boran-Galla and Small East African from Kenya), WAD from Guinea Bissau (West 
Africa), Tswana from Botswana and Baladie from Egypt] and four populations from outside 
African continent (Italian Alpine from Italy, Ardi from Saudi Arabia, Hair from Turkey and 
Zalaajinistiin from Mongolia) were sampled.  All the populations were assessed for genetic 
diversity using 15 microsatellite loci.  A wide range of statistical analyses were performed on 
the data to answer questions on population genetics diversity, relationship and differentiation. 
The genetic diversities observed in both the Ethiopian and non-Ethiopian populations were 
similar.  The analysis of molecular variance (AMOVA) indicated a greater proportion of the 
genetic diversity within the Ethiopian populations than between populations.  Genetic distances 
were observed to be relatively small for Ethiopian populations.  The high correlations observed 
between geographical distances and genetic distances for Ethiopian populations supported this 
result.  On the other hand, relatively large genetic distances were observed between Ethiopian 
and Non-Ethiopian populations with exception of Kenyan populations.  The results have shown 
that more than 60% of the goat populations showed significant inbreeding effect, which leads to 
loss of genetic diversity.  Phylogenetic trees, population structure and principal component 
analyses showed that all the Ethiopian populations are genetically distinct from all the reference 
populations except with two Kenyan populations.  However, they also indicated that all the 
Ethiopian goats populations are very closely related to each other.  Our results indicate that the 
11 Ethiopian populations can be grouped as eight distinct genetic entities: Arsi-Bale, Gumez, 
Keffa, Woyto-Guji, Abergalle, Afar, Highland goats (previously separated as Central and North 
West Highland) and the goats from the previously known Hararghe, Southeastern Bale and 
Southern Sidamo provinces (Hararghe Highland, Short-eared Somali and Long-eared Somali 
goats).  Weitzman analysis indicates that about 75% of the total genetic diversity of the 
Ethiopian goats is present in four breeds: Afar, Abergalle, Gumez and Keffa with marginal loss 
of diversity of 24.32%, 19.22%, 16.59% and 12.99%, respectively. 
Genetic differentiation tests showed that all the Ethiopian and non-Ethiopian 
populations were well differentiated.  Individuals from Ethiopian populations were assigned to 
their source populations with a low degree of accuracy indicating their genetic similarity.  
However, the proportion of correct assignment for the non-African population was higher than 
that for the Ethiopian populations.  Results support the theory of two separate entry points for 
the goats into the African continent: the Isthmus of Suez for the North African and the horn of 
Africa for the Eastern African goats.  Ethiopia hosts a large indigenous goat genetic resource 
adapted to the complexity agricultural production systems of the country; it represents a unique 
resource that has potential for further and future genetic improvement of its productivity. 
 1
 CHAPTER ONE 
 
1.0. INTRODUCTION 
Most food production systems depend heavily on utilisation of locally adapted 
animal species.  These animal genetic resources consist of species that are of 
agricultural, cultural and economic importance to man.  Different cultures and regions 
of the world bear importance to different domestic animals.  However, the common 
agricultural species kept in most regions include sheep, goats, cattle, horses, pigs, 
buffaloes and chicken (FAO, 1994; FAO, 1999).  In sub-Saharan Africa, indigenous 
breeds of sheep and goats are very important, in fact more important than cattle for the 
small-scale farmers since they are easier to acquire and to maintain.  These serve as a 
secure form of investment, a means of income, source of manure and for various 
religious and ceremonial functions (Chenyambuga, 2002).  These animals often provide 
the only practical means of utilising vast areas of natural grasslands in the areas where 
crop production is uneconomical (Rege et al., 2002).  Adapted to the local environment, 
the African sheep and goat represent a unique genetic resource for the farmer.  
However, the need for increased economic gains has led to crossbreeding these 
indigenous breeds with imported exotic breeds or directly replacing the indigenous 
genotypes (Wollny, 2003).  To compound the problem, indigenous breeds have not 
been characterised in many African countries and it would be tragic to lose these unique 
genetic resources that are the result of centuries of human and natural selection (Leak et 
al., 2002).  The rate of erosion of indigenous animal genetic resources therefore 
threatens the prospect of improving the livelihood of present and future human 
generations.  This is especially so when one considers that the demand for animal 
products is expected to double over the next 20 years as a consequence of urbanisation, 
population growth and increased income (FAO, 1999).  Developing countries host a 
vast majority of indigenous farm animal genetic resources (over 90 percent) that 
unfortunately are not being improved to respond to global food security needs (FAO, 
1999).  FAO estimates that 30 percent of livestock breeds are at risk of extinction and 
about six breeds are lost per month.  Moreover, more than half of these breeds are found 
in developing countries (FAO, 2000).  However, sustainable conservation and 
utilization of these indigenous goat resources cannot begin until their genetic diversity 
is understood and quantified.  Most indigenous goat breeds, despite being best suited 
 2
 for the local conditions, are not yet genetically classified and cannot thus be maximally 
exploited.  These indigenous genetic resources therefore need to be characterized before 
the conservation and utilization programs can begin.  
Ethiopia has the largest livestock population among all the countries in Africa.  
Ethiopia with its great variation in agro-ecological zones represents a potential reservoir 
of sheep and goat diversity.  They inhabit a wider range of environments extending 
from a tropical to cool temperate climates.  The sheep and goat population of Ethiopia, 
though rather variably estimated, ranks high both in Africa and in the world.  According 
to FAO (1994), there are approximately 21.7 and 16.7 million sheep and goats, 
respectively.  Sheep and goats are distributed in all agro-ecological zones of the country 
although the majority of the sheep population is concentrated in the highlands.  The 
majority of the goat population is found in large flocks in the arid and semi-arid 
lowlands where pastoralists in the south, east, and west keep them for milk production, 
for slaughter and for sale.  Goats in the highlands are widely distributed in the mixed 
crop-livestock production systems with very small flock size as a means of cash 
earnings and meat. Although there are severe environmental constraints to increase goat 
productivity, there is considerable potential for improved goat production in the 
country, where goat milk, meat, and skin are valued commodities.  
Ethiopia has been long back recognized as a source of genetic diversity in plants 
and animals.  It appears that Ethiopia has served as a gateway of domestic animal 
genetic materials from Asia to Africa, and its diverse ecology leads to further 
diversification and development of different genotypes.  It is believed that the first goats 
reached Egypt by 5000 BC and subsequently spread south and west of the African 
continent.  Remains of a small goat have been found in a site in Sudan thought to date 
before 3300 BC and the lop-eared Nubian type has been found in sites dating prior to 
2500 BC (Mason, 1984).  Therefore it is likely that the first wave of goats entered 
Ethiopia from the North of Africa sometime between 2000 and 3000 BC. 
The identification, description, classification and naming of livestock types are 
the first step in the assessment of a country?s livestock resources.  The physical 
description of types can only dimly reveal the genetic relationships between individuals, 
though it is the first step in classifying a diverse population into relatively homogeneous 
sub-populations.   
Work done in 1990s on physical description and management system at 
Alemaya University of Agriculture together with FARM ? Africa (DGDP - Dairy Goat 
 3
 Development Project) has shown the presence of 14 goat types in Ethiopia and Eritrea 
(Workneh, 1992; Alemayehu, 1993; Nigatu, 1994).  This survey was the very first step 
in breed characterization of Ethiopian goats through differentiating a heterogeneous 
population into relatively homogeneous sub-populations, which may then be named as 
breeds, types or subtypes.  However, it remains unclear how different at the genetic 
level are these 14 goat types.  The availability of new molecular techniques for large-
 scale genetic characterization is providing unique opportunities to tackle the above 
issues.  These techniques are able to reveal genetic distances between populations 
allowing a better understanding of how these populations evolved. 
In view of the above, it would be ideal to characterize at the molecular level the 
Ethiopian goats in order to determine their genetic diversity and describe the 
relationship among the different populations.  The information provided will allow 
formulation of conservation programs, maximizing diversity conserved and it may 
guide breeding improvement activities for meat and milk production.  
The primary measure should be to determine the magnitude of genetic 
differentiation and relationship among the populations.  One way of doing this is to 
determine genetic distances between pairs of breeds/populations.  The computation of 
genetic distances requires estimation of allele frequency data using either protein 
markers or DNA markers.  However, techniques for the determination of polymorphism 
of protein products lack the power to resolve the differences between closely related 
breeds since a great deal of genetic variation remains undetected by these methods 
(Meghen et al., 1994; Dowling et al., 1996). 
In recent years, a range of innovations in molecular genetics has been developed 
for the study of genetic variation and evolution of populations using DNA genotyping 
information.  The most utilized DNA marker for population genetics of livestock is 
microsatellite.  Microsatellite markers, also called short tandem repeats (STRs) or 
simple sequence repeats (SSRs), are a relatively new class of genetic marker.  Over a 
few years they have become a tool of choice to address population genetics and 
demographic questions (Sunnucks, 2000).  The application of microsatellite markers is 
currently considered to be useful in the analysis of genetic diversity as they are 
numerous, randomly distributed in the genome, highly polymorphic, and they show 
codominant mode of inheritance (Ellegren, 1993).  They allow the study of genetic 
diversity and differentiation of closely related populations.  Microsatellite, as genetic 
markers, have been applied successfully in the study of genetic variation of livestock 
 4
 including between and among European and African cattle breeds (Machugh et al., 
1997; Okoma et al., 1998), between Swiss goat breeds (Saitbekova et al., 1999), 
between Indian goat breeds (Ganai and Yadav, 2001) and between Chinese indigenous 
goat populations (Li et al., 2002).  Microsatellite DNA variation has also been 
employed to assess genetic distances between sheep breeds (Buchana et al., 1994; 
Arranz et al., 1998).  Unfortunately, for goats, throughout the world very limited work 
has been done to determine genetic variation among breeds/populations and in fact, 
information on microsatellite polymorphism of Ethiopian goats is completely lacking.  
Therefore, to utilize the Ethiopian goat genetic resources effectively, it is 
necessary to characterize Ethiopian indigenous goat populations genetically.  Such 
characterization would provide a comprehensive database of genetic variation among 
the goat populations in Ethiopia.  It would provide furthermore information as to which 
of the populations represent homogenous populations and which of them are genetically 
distinct.  It would also provide information on the risk status of the goat populations.  
The information generated would contribute to the understanding of the evolutionary 
history of goats in Ethiopia.  It will also contribute to the conservation and management 
of Ethiopian goat genetic resources. 
The overall goal of this project is to obtain detailed information on the genetic 
diversity and differentiation of the indigenous goat populations of Ethiopia with the 
following specific objectives: 
1. To quantify genetic diversity within and among Ethiopian goat populations. 
2. To clarify the evolutionary genetic relationships among the Ethiopian goat 
populations. 
3. To clarify the origin and relationships of Ethiopian goat populations through 
the study of non-Ethiopian breeds of references. 
4. To compare the morphological classification of the Ethiopian goat 
populations with the one based on molecular genetic characterization. 
 5
 CHAPTER TWO 
 
2.0 LITERATURE REVIEW 
 
2.1. Classification of goats 
Taxonomically the domestic goats (Capra hircus) belong to the order 
Artiodactyla, suborder Ruminantia, family Bovidae and tribe Caprini.  The tribe 
Caprini includes different members: - The goat-like sheep or sheep like goat: 
Ammotragus lervia (Barbary sheep) and Pseudois nayaur (Blue sheep), the goat like 
antelope: Oreamnos americanus (Mountain goat), and the sheep and goats. Table 1 
shows the main fifferences between goats and sheep. 
Table 1. The main differences between goats and sheep  
Goats Sheep 
60 chromosomes 54 chromosomes 
Presence of beard in most  Absence of beard 
Presence of face and caudal glands Absence of face and caudal glands 
Absence of foot glands Presence of foot glands 
Tail held up Tail hanging down 
(http://capra.iespana.es/capra/ingles/origen/origen.htm). 
 
Goats are divided into two genera ? Capra and Hemitragus.  The later is the 
tahr, which has three species (Ellerman and Morrison-Scott 1966): - H. jemlahicus, 
found in the Himalayan region from Kashmir to Sikkim (northern India); H. jayakari, 
found in Oman; and H. hylocrius, found in Nilgiri Hills and adjacent ranges in southern 
India.  
The genus Capra is divided into eight species (Alados 1985b; Ansell in Meester and 
Setzer 1977, Corbet 1978) (http://www.press.jhu.edu/books/walkers_mammals 
_of_the_world/artiodactyla/artiodactyla.bovidae.capra.html): 
(i) C. aegagrus (bezoar or wild goat), found in mountains from Asia Minor to 
Afghanistan and Pakistan, Oman, Crete, Aegean Islands;  
(ii) C. hircus (domestic goat), found worldwide in association with people, feral in 
many / some areas;  
 6
 (iii) C. ibex (ibex), found in European Alps, Palestine, Sinai, Arabian Peninsula, 
mountainous region from Afghanistan and northern India to Lake Baikal, Egypt 
and Sudan east of the Nile, northern Ethiopia;  
(iv) C. walie (Walia ibex), Simien Mountains of north-central Ethiopia;  
(v) C. caucasica (west Caucasian tur), western Caucasus;  
(vi) C. cylindricornis (east Caucasian tur), eastern Caucasus;  
(vii) C. pyrenaica (Spanish ibex), Spain and formerly southern France and Portugal;  
(viii) C. falconeri (markhor), mountains of southern Uzbek and Tadzhik (formerly 
Soviet Socialist Republic), Afghanistan, northern and central Pakistan, and Kashmir 
(India). 
Various authors have allegedly suggested that all these different species or 
subgenera have contributed to the domesticated stock (Epstein, 1971).  According to 
Mason (1984), the bezoar is the direct ancestor of the domestic goat and the markhor is 
said to have contributed to some breeds of Central Asia (Mason, 1981a).  However, for 
Mason (1984) it is less likely that the markhor has influenced some breeds.  The ibexes 
and turs have not been domesticated. 
 
2.2. Domestication of goats  
Domestic animals have played a key role in human history.  Despite their 
importance, the origin of most domestic species remains poorly understood.  The goat is 
the earliest domestic animal and probably the first ruminant livestock, after the wolf 
was domesticated (Hole, 1996; Uerpmann, 1996).  There are two reasons for this: 
Firstly, the wild goat was reported to be present in the regions of southwest Asia during 
the time when agriculture was developing.  Secondly, the goat is an extremely hardy 
animal, hence, could have withstood the rigours of being reduced to the state of 
domestication better than other ruminants.  Goats were domesticated at first for meat.  
However, as a dairy animal, the goat is regarded as the oldest, even older than cattle 
(Devendra and Burns, 1983). 
Domestic goats (Capra hircus) might have played a central role in the Neolithic 
agricultural revolution and the spread of human civilizations around the globe (Porter, 
1996; Pringle, 1998a).  The evidence for time and place for goat domestication is based 
on the identification of bones from archaeological sites.  Domestication of goats is 
considered to have occurred in the mountainous area of western Asia between the 7th 
 7
 and 9th millennium B.C. (Epstein, 1971; Devendra and Burns, 1983).  The origin of 
domestic goats remain uncertain and controversial, however, archaeological evidence 
suggests that they were probably first domesticated in the Fertile Crescent region of the 
Near East possibly first in the Zagros Mountains area 10,000 years ago (Zeder and 
Hesse, 2000).  However it has been suggested that the goats could have also been 
domesticated outside the Zagros Mountains range.  Some studies hint for a second 
domestication area in Sindhu valley civilization in northwest part of India (now in 
Pakistan) from which the cashmere breeds would have originated (Porter, 1996).  
Others suggest that at least two wild species of Capra (Shackleton, 1997) could have 
contributed to the gene pool of domestic goats (Clutton-Brock, 1981). 
Takada et al. (1997), studied the diversity of the cytochrome b gene of 
mitochondrial DNA (mtDNA) in Asian goats, and concluded that the strongest 
candidate for a matriarch ancestor of domestic goats is the bezoar or pasang (Capra 
aegagrus).  Furthermore, phylogeographic analysis of mtDNA revealed three highly 
divergent lineages (estimated divergence > 200,000 years ago) suggesting three 
separate maternal origins of domestic goats (Luikart et al., 2001).  Sultana et al. (2003) 
studied the mtDNA diversity of Pakistani goats, and reported four distinct mtDNA-
 lineages termed as A, B and C (previously reported) and a new lineage D.  The 
estimated divergence times between the most recently evolved mtDNA-lineages A and 
D were from 260 483 to 371 052 year ago.  They concluded that at least four different 
strains of wild Capra might have been the source of the modern domestic goats with the 
most likely wild ancestor being the bezoar (Harris 1961; Clutton-Brock 1981). Joshi et 
al. (2004) have undertaken the investigation of 363 goats belonging to 10 different 
breeds from different geographic regions of India using mtDNA sequence data from the 
hypervariable region I.  They found evidence for population structure and new 
mitochondrial DNA in Indian goats and cannot reconcile the genetic diversity found 
within the major lineage with domestication starting 10,000 years ago from a single 
mtDNA ancestor.  Thus, they proposed a more complex origin for domestic goats. 
The Bezoar (Capra aegagrus) is thought to be the progenitor of the domestic 
goat.  Bezoar means counter poison, as in the savage animals it is possible to find a 
concrete in the stomach which believed to be an antidote.  The main characteristic of 
the Bezoar is the long saber or scimitar shaped horns.  
It is believed that goats were probably present at Cay?n? in Southeast Anatolia 
about 7000 BC.  They were present at Jarmo in Iraqi Kurdistan where the site is dated 
 8
 from about 6500 BC to the end of the 6th millennium BC and at Sialk on the Iranian 
plateau as early as the 6th millennium BC.  Domestic goats with scimitar horns like the 
bezoar spread over a large part of the old World during the Neolithic period.  They were 
represented in vase paintings from ancient Mesopotamia in the 4th millennium BC.  The 
small short-eared sabre-horned generalized type was occasionally also illustrated in the 
Mesopotamia tomb painting of the same period.  It is believed that by the 5th 
millennium BC goats had reached Egypt and by about 3500 BC goats with spiral or 
corkscrew horns entered Egypt from the Middle East.  From Egypt the goats moved to 
the South and West part of the continent.  The dwarf goat is recorded at Shaheinab near 
Khartoum in the Sudan from 3300 BC while first evidence for the screw-horned lop-
 eared type is found in Nubia and Libya at site dated around 2650 BC.  The goats of sub-
 Saharan Africa are predominantly small, short-eared, short-haired and short-horned.  
Goats are noted to appear in East Africa by 4000 ? 3500 years BP (Marshall, 2000) and 
in southern Africa by 4th to 7th centuries AD (Smith, 2000).   
 
2.3 Goat genetic resource of Ethiopia  
Indigenous breeds constitute well over 95 percent of small ruminant populations 
in Africa (Rege, 1992).  These are well adapted to the environment and the ravages of 
various kinds such as drought, famine and civil wars that continually plague the 
continent.  Their adaptive features enable them to effectively cope with the stressful 
nature of marginal lands (Chenyambuga, 2002).  The population of goats in sub-
 Saharan Africa is estimated to be 163 million (Rege et al., 2002).  Migration to a new 
habitat and consequently the effect of natural and artificial selection have led to the 
evolution of breeds and types of goat, which differ in appearance and performance.  
Around 90 ?breeds? of African goats have been recognized using criteria as geographic 
distributions, ecotypes or communities-tribe ownership (Rege, 1992).  They presumably 
derive from goats that spread south from Egypt at an early date.  Generally, goats of 
sub-Saharan Africa are divided into three major types following their morphology; the 
long lop-eared type in north east and southern Africa, the small short-eared type 
dominant in eastern Africa and the dwarf short-eared type of West Africa.  
Intermediates morphological types are numerous. 
The majority of Ethiopian goat population is found in large flocks in the arid 
and semi-arid lowlands where pastoralists in the South, East, and West keep them for 
 9
 milk and meat production and for sale.  Goats in the highlands are widely distributed in 
the crop-livestock production systems with very small flock sizes as a means of cash 
earnings and meat.  Despite the huge resource potential, production and export 
opportunities, goat production in Ethiopia is relatively undeveloped.  Although there are 
severe environmental constraints to increase goat productivity, there is considerable 
potential for goat production in the country, where goat milk, meat, and skin are valued 
commodities.  
Information compiled on physical description and management system revealed 
that there are 14 goat types in Ethiopia and Eritrea (Farm-Africa, 1996).  Out of these 
eleven are found in today Ethiopia.  However, goats in Africa have traditionally been 
divided into three main families ? the Dwarf goats of West and Central Africa, the 
Savannah goats of sub-Saharan Africa and the Nubian type goat of Northern Africa 
(Epstein, 1971; Wilson, 1991).  Using a set of morphological characters and 
multivariate statistical analysis the goat of Ethiopia and Eritrea have been classified into 
four major families: - the Somali family (Short-eared Somali, Long-eared Somali, and 
Hararghe Highland), the Nubian family (Nubian and Barka), the small Rift valley 
family (Abergalle, Worre, Afar, Arsi-Bale, and Woyto-Guji) and finally the more 
heterogeneous Small East African family (Western Highland, Keffa, Central Highland 
and West Lowland).  The names given to the goat types of Ethiopia reflect mainly their 
geographical locations and to some extent also their ethnic affiliation.  These are:  
 
i) Afar  
The Afar goat is also known as Adal or Danakil.  It is descendant of the rift 
valley goat thought to have entered Ethiopia from Yemen and Saudi Arabia.  It is kept 
almost exclusively by the Afar-ethnic group in the Ethiopian and Eritrean rift valley, the 
Danakil depression; Gewane and northern and western Haraghe (see Fig. 1).  These 
goats are well adapted to arid environments, infrequent watering (every 3-4 days) and 
can travel long distances.  Where it is possible flocks may move to highland areas 
during the dry season.  They have concave facial profile, narrow face, prick-eared, 
leggy, long thin upward-pointing horns, and patchy coat colour.  The coat is very fine 
and short with variable colours and colour patterns with 48% of individuals white, 25% 
light brown, 27% black and flecks and patches are also common.  Afar pastoralists 
identify their goats by branding the ears or hindquarters, or by notching the ears.  Afar 
goats are extensively milked (hadore hana) for food, medicine and sale.  Udder size is 
 10
 relatively small with a circumference of 21 cm and length of 17 cm.  Meat may be 
consumed fresh or air-dried (teru). Kids less than one month old may be sold as a 
delicacy known as bekel (male) and bekelo (female).  Fresh blood is consumed to treat 
malaria and bullet wounds (Farm-Africa, 1996). 
 
Figure 1: Afar female goat 
ii) Abergale  
Abergale goats are descendents of the rift valley goat type from south-west 
Asia.  Their localized distribution is along the Tekeze River in southern Tigray, 
Northern Wollo and Eastern Gonder (see Fig. 1).  The Agew and Tigray ethnic groups 
keep the goats.  Abergalle goat milk is made into butter for home consumption and sale.  
Goat milk (seba) butter (t?esmi) is made and used for cooking and as a cosmetic (likai).  
Milk is also made into yogurt (rugeo) and cottage cheese (ajebo).  Skins are processed 
locally and used as apron, as container for grain, water, butter or honey, a small mat or 
shoulder cape, pages of religious books (brana), sandals, belts and edges of baskets.  
The manure is widely used as a soil fertilizer in Tigray (Farm-Africa, 1996).  
 11
  
Figure 2. Abergalle male goat 
 
Figure 3. Abergalle female goat 
iii) Arsi-Bale 
The Arsi-Bale goat is also part of the rift valley family (previously known as 
Galla-Sidamo in the literature).  They are distributed throughout Arsi and Bale regions, 
 12
 up to an altitude of 4000 meters.  They are also found in the higher altitudes of Sidamo 
and western Hararghe.  They occupy all the agro-pastoral lowlands within the rift 
valley, from Lake Abaya in the South to south Shoa in the North (Figure 1. ).  The Arsi-
 Bale goat is a relatively tall goat with a predominantly straight facial profile (98%).  
Males have curved (47%) and straight (41%) horns mainly pointing backwards (58%) 
with some pointed straight upward (28%).  The coat colour varies between any of seven 
colours (white, black, brown, fawn, gray, roan and red), with the first three being most 
dominant.  The ears are fairly long. About 10% have a semi-pendulous ear form, the 
rest being erect or horizontal.  About half of the females and 75% of the males in 
Sidamo are bearded, but in the agro-pastoral areas beards on females are less frequent 
(38.7%).  The goats in Sidamo appear to attain a greater number of average kidding for 
a given dentition class than other goat types in the region.  Arsi-Bale goats are 
extensively milked.  Among the Sidamo people goat milk is a highly valued source of 
nourishment for children.  The demand for goat milk is so high that the smallest amount 
of milk available is diluted with some water to make sure that as many children as 
possible drink some.  Goat meat is widely eaten, mostly during social and festive 
occasions.  This goat type is more prolific than other goat types in the region and is 
potentially a desirable meat producer in the highlands.  The skins of the hairy Arsi-Bale 
goats are highly valued as saddle covers (Farm-Africa, 1996).  
 
Figure 4. Arsi-Bale male goat 
 13
 iv) Woyto-Guji 
Woyto-Guji goats are also known as Woyto, Guji or Konso.  They are part of the 
rift valley family of goats.  They are distributed in north and south Omo, southern 
Sidamo and parts of Walayita (see Fig. 1).  These goats are mainly kept by pastoral 
ethnic groups (Tsemay, Malie, Hamer, Bena, Dasenatch, Bumie and Guji) and by a few 
agricultural groups (Konso and Gardula).  More notably this goat type inhabits those 
areas in Sidamo known to be endemic with trypanosomiasis, especially the Gelo valley 
to the south of Lake Abaya and the western Ghenale catchment?s area.  They are 
Brown, black or red colour with a shiny, smooth coat and small head with a straight or 
concave facial profile.  Coat colours are often marked with black or brown stripe along 
the back, on the underside or on the front of the legs.  They are medium-sized goats 
with a mainly straight (89%) to concave (11%) facial profile.  Straight horns occur in 
71% of the males, curved in 26%, with polled goats forming 3% of the population.  
Horns mainly point backwards (75%) or upwards (21%) and in a few cases laterally 
(2%).  The average number of kids born per breeding female is 4.3.  Goats may be 
milked once or twice a day for one to six months irrespective of the season.  Milk is 
sometimes made into butter and used in traditional healing treatment.  The Woyto-Guji 
goats are known to be good for meat.  Almost all pastoral groups in south Omo use goat 
skins to make traditional leather garments for women (Farm-Africa, 1996). 
 
 
 14
  
Figure 5. Woyto-Guji Make goat 
 
Figure 6. Woyto-Guji female goat 
v) Hararghe Highland  
The Hararghe Highland goats are most likely derived from the Somali goat type.  
They are distributed in the highlands of east and west Hararghe (see Fig. 1).  They are 
 15
 small, white, brown or black, commonly polled with a straight (60%) or concave (40%) 
facial profile.  Horned goats have short hair and straight (32%) or concave horns (29%).  
Beard is present on 72% of males but there are no ruffs.  Wattles are present on 14% of 
goats.  The average number of kids born per breeding female is 2.2.  They are 
commonly milked (anan ra?e) for home consumption and sale in some markets.  Goat 
meat is widely eaten and preferred to sheep meat.  Skins are used for beds and prayer 
mat.  They are also made into blacksmith?s bellows (buufa) (Farm-Africa, 1996). 
 
Figure 7. Hararghe Highland female got 
vi) Short-eared Somali  
The Short-eared Somali goats are also known as Denghier or Deghiyer.  They 
are probably related to the Arab goats in Somalia, which were introduced directly from 
Arabia.  They are distributed in the northern and eastern parts of Ogaden i.e. Jijiga, 
Degeh Bur and Werder (see Fig. 1) where they are maintained by the Isaaq and 
Mijertein Somali clans, and Dire Dawa, Issa and Gurgura communities.  They are 
medium sized; mainly have white, short hair.  The Short-eared Somali goat is smaller 
than the Long-eared Somali goat type.  It has a straight facial profile and males have 
straight (46%) and upward pointing (46%) horns.  Females have generally curved horns 
(50%), most of which point upwards (55%), but 27% are oriented backwards and 12% 
are lateral pointing.  Polled goats are found in 5% of males and 7% of females.  The 
 16
 average number of kids born per breeding female is 2.4.  Somali goats are widely 
milked and milk is consumed fresh and also made into butter.  The Somalis use butter 
for both food and medical purposes.  Meat may be eaten fresh or preserved by cutting 
into slices, frying it in butter or animal fat and keeping it in a container (Farm-Africa, 
1996). 
 
Figure 8. Short-eared Somali male goat 
 17
  
Figure 9. Short-eared Somali female goat 
 
vii) Long-eared Somali 
Long-eared Somali goats are also known as Large-white Somali, Degheir, 
Galla, Digodi, and Melebo.  They are probably related to the descent of the Arab goats 
in Somalia introduced directly from Arabia.  Long-eared Somali goats are distributed 
throughout the Ogaden, lowlands of Bale, Borana and southern Sidamo (see Fig. 1).  
They are kept by the Hawia, Ogaden, Rare Bare, Digodi clans of the Somali ethnic 
group and the Boran, Gabra and Geri ethnic groups.  The Long-eared Somali is a large 
white goat with a predominantly straight facial profile.  Horns are mainly curved (in 
41% males and in 46% females), and pointed backwards in 38% of males and upwards 
in 48% of females.  Horns in some 13% of both sexes have lateral orientation.  The ear 
is horizontal (in 46 % of females and in 66.8% of males) or semi-pendulous (in 22% of 
females and in 16% of males).  Vestigial ears, which are not found in other goat types, 
occurred concentrated in one site, Dollo, in Borana.  The average number of kids born 
per breeding female is 3.2.  Somali goats are extensively milked by the Somali and 
Boran pastoralists.  Goat meat is preferred to mutton in most areas where the Long-
 eared Somali goat is kept.  Meat may be eaten fresh or preserved by cutting into slices, 
frying it in butter or animal fat and keeping it in a container (Farm-Africa, 1996). 
 18
  
Figure 10. Long-eared Somali female goat 
 
Figure 11. Long-eared Somali male goat 
 
viii) Central Highland  
Central Highland goats are also known as brown goats.  They are highland type 
most likely derived from mixing of types in the past.  They are distributed in the central 
 19
 highlands west of the rift valley escarpment in central Tigray, Wollo, Gonder and Shoa 
(see Fig. 1).  They are owned by settled farmers of the Tigray, and Amhara ethnic 
groups.  Central Highland goats are medium sized, broad-faced with thick horns and 
reddish-brown colour.  The Central Highland goat has a predominantly straight (71%) 
facial profile; 29% of the goats have a concave profile.  Virtually all males have horns, 
82% being straight and pointing backwards, 13% curved and 5% spiral.  The coat type 
is short and smooth with 51% plain colour, 42% patchy and 7% spotted.  The 
predominant colour is red-brown (41%), the remaining goats being split between black, 
white and gray.  Most males (82%) have beard and ruff (99%).  Wattles are present on 
6% of males.  The average number of kids born per breeding female is 2.9.  Goats are 
expensively milked in Tigray.  The skins from central highland goats are an important 
export product.  The best quality skin ?Bati Genuine? is made from skins of brown/fawn 
coloured goats in the Bati area of Wollo (Farm-Africa, 1996). 
 
Figure 12. Central Highland male goat 
ix) Western Highland (North West Highland) 
The Western Highland goats are also known as Agew locally.  They are most 
likely derived from mixing of types in the past, distributed in the Highlands of south 
Gonder, Gojam, Wellega, and western Shoa (see Fig. 19a).  They are relatively tall with 
 20
 a concave facial profile (100%).  It has a relatively coarse long coat (82%); 12% of the 
animals have hair on their thighs.  The coat colour is mainly plain (51%), 42% have a 
patchy colour pattern, and 7% are spotted.  The main colours are white (42%) and fawn 
(42%), and combinations of these colors.  There are relatively a high proportion of 
polled goats in the western Highland population (14%).  The horned goats have straight 
(76%) horns directed backwards (73%).  A ruff is present on 99% of the males, and 
beard on 84%.  Wattles are present on 12% of the goats.  An unusual high proportion of 
polled and hermaphrodite goats were found around Lake Tana.  The goats showed 
hermaphroditism based on examination of external genitalia.  There is an established 
relationship between hermaphroditism and polledness in goats (Gall, 1981).  The 
average number of kids born per breeding female is 3.6 (Farm-Africa, 1996). 
 
 
Figure 13. Western Highland male goat 
 21
  
Figure 14. Western Highland female goat 
x) Western Lowland (Gumez) 
Western Lowland goats are also known as Shankela and Gumez locally.  They 
are derived from mixing of types in the past.  They are distributed in the western 
lowlands bordering Sudan in Gojam (Metekel), Wellega (Assosa) and Illubabor 
(Gambela) (see Fig. 1).  The Western Lowland goat is a relatively short goat with a 
straight facial profile (100%).  It has a predominantly short smooth coat (81%), with 
16% having a relatively coarse coat.  The main colours are white (42%) and fawn 
(38%), with some black (9%) and gray (11%), occurring mainly in patches (73%).  
Most male goats have straight horns (85%) oriented backwards (77%).  There are 12% 
polled males in the population.  A ruff is present in 96% of all males and beard in 70% 
of males.  Wattles are present in 12% of all male goats.  The average number of kids 
born per breeding female is 3.5.  Western Lowland goats are remarkably prolific with 
56% giving single births, 41% twin births and 3% triplets.  Quadruplets were also 
reported to occur.  Pastoral and agro-pastoral groups in the area of distribution use goat 
milk extensively.  Goat meat is widely eaten.  The horn is used as a musical instrument 
(zoombara) in Assosa (Farm-Africa, 1996).  
 22
  
Figure 15. Western Lowland male goat 
 
Figure 16. Western Lowland female goat 
xi) Keffa 
Keffa goats are derived from mixing of types in the past.  They are distributed in 
the highlands and lowlands of Keffa, and parts of south Shoa, Kambata and Hadiya  
 23
 (see Fig. 19a).  The Keffa goat is relatively short with a predominantly straight facial 
profile (92%).  Most males (83%) have straight horns pointing backwards (80%).  A 
small proportion, 14% have curved horns.  The incidence of polledness is low at 3%.  
Keffa goats have short pricked ears.  Most Keffa goats have a coarse (38%) to hairy 
(27%) coat type; some 16% have hair on their thighs.  Plain colors predominate (52%), 
with some patchy colour patterns (45%).  The main colours are black (30%) or brown 
(31%). Among males, 88% have beards, 97% have ruffs.  Wattles are present in 12% of 
all goats.  The average number of kids born per breeding female is 3.1.  The pastoral 
ethnic group, the Surmas, drink the milk and blood of goats.  Goat meat is widely eaten.  
The horn is used as a container for igniting a mixture of explosive minerals (Farm-
 Africa, 1996).  
 
Figure 17. Keffa male goat 
 24
  
Figure 18. Keffa female goat 
2.4 Conservation of domestic animals diversity 
2.4.1 Meaning and importance of domestic animals diversity 
Domestic animal diversity has been defined by Hammond (1993) as the 
spectrum of genetic variation existing among species, breeds and individuals of all 
animal species, which have been domesticated, and their immediate wild relatives.  This 
genetic variation has been developed during the millions of years of evolution to form 
and stabilize each species.  Globally, there are about 40 species of domestic livestock 
(FAO, 1995), which have been domesticated by humans over the last 12,000 years.  In 
the process of domestication, however, separate and genetically unique breeds and 
strains have developed within each species as a result of human development and 
occupation of new areas over the planet.  According to FAO (1995), there are about 
4,000 to 5,000 breeds and strains of domestic animals in the world, of which 300 are 
breeds of goats (Devendra and Burns, 1983).  These breeds and strains are referred to as 
the global animal genetic resources and the genetic variation both between and within 
the breeds are described as the diversity within the species of domestic animals.  The 
selection process, both environmental- and human-directed, has resulted in much of the 
diversity existing between the breeds.  Differences among the breeds have been created 
 25
 by reproductive isolation, often imposed by human through pursuit of different breeding 
objectives and physical separation for various length of time.  As a consequence of 
physical separation, each breed/strain has been adapted to particular ecological 
condition to suit the local climate and the requirements of the community.  Thus, in the 
conservation of domesticated species, it is the diversity between breeds, rather than the 
variation between species, which is of crucial importance (Hammond, 1993; Barker, 
1994). 
Intra-specific diversity has got both within-population and between-population 
genetic components.  Genetic variation at the population level consists of the 
differences in the types of alleles present and their frequencies across all members of a 
population considered together.  Genetic variation within population is caused by 
change of allele frequencies over time due to selection, random genetic drift and gene 
flow (immigration from or emigration to other population).  Genetic variation within 
population is important as it is related to heterozygosity, which is known to enhance 
fitness-related characteristics (Allendorf and Leary, 1986).  Variability between 
populations arises both from random processes (founder effects, demographic 
bottlenecks, genetic drift and mutations) and from local selection imposed by 
environment and humans (Hartl and Clark, 1997).  Variability between populations is 
also the result of adaptation of populations to their local conditions and it is important 
since locally adapted populations may have particular genes or gene combinations 
critical for viability in their local environment.  For conservation of animal diversity, it 
is therefore, important that both within and between populations variation be 
maintained.  This will enable better utilization of animals for production of food and 
agriculture to meet the current need of human kind while maintaining the potential to 
meet the range of future possible changes and needs (Hammond, 1993; FAO, 1995).  
This is because genetic diversity is the basis for a species evolutionary flexibility and 
responsiveness to environmental changes.  
The main reason to advocate conservation of genetic diversity is that the 
improvement of domestic animals in order to meet human needs, is dependent on 
genetic variation both within- and between-breeds/strains.  Loss of genetic diversity 
will likely decrease the ability of animals to respond to environmental change and will 
result in loss of genetic information potentially useful for breeding improvement 
(Primack, 1993; Hunter, 1996).  The earth?s climate varies widely and is continuously 
changing.  It is the diversity of species and breeds that will allow the environment to be 
 26
 utilized efficiently throughout the years both in temperate areas and in the tropics.  The 
urge of conservation, particularly in developing countries, is strengthened by the 
predictions of the scale of human population growth.  It has been predicted that, human 
needs for food will double over the next two generations with the demand for animal 
products increasing more rapidly than that of plants (Hammond, 1993; Hammond, 
1994; Degado et al., 1999).  With increasing population pressure, the quantity of food 
and other products must increase.  In developed countries consumer emphasis on 
product quality is increasing.  Changing the production level and product quality 
require different types and varieties.  In addition, changes in food and agriculture 
production will require new combinations of plant and animal species and lines (e.g. 
breeds and varieties) and changes of management and production strategies.  
Sustainability in this new production system will require different genetic types (FAO, 
1995). 
The second reason to advocate conservation of domestic animal diversity is that, 
we can reasonably expect that large differences observed today between geographic 
regions in human needs for food and agriculture and in production capability will 
persist in the future.  According to Hammond (1994) almost three quarters of the world 
agriculture will remain at the low- to medium-input levels where animal production 
environments incorporate combinations of stress such as feed shortage, drought and 
heat.  Breeds in these environments are reservoirs of genetic information responsible for 
the adaptations necessary for production under adverse environment.  Although they 
have been overlooked, the needs for those, which are able to produce in hard 
conditions, is making and will make in future important contribution to feeding humans 
(Simm, 1998). 
A third reason for conservation of domestic animal diversity is that sustainable 
crossbreeding schemes require at least two viable purebred populations and sometimes 
more.  Trends towards fewer breeds will reduce the opportunities for improvement of 
productivity through crossbreeding. 
The fourth reason is that the wide variety of livestock breeds and strains 
available today is part of our cultural heritage and it deserves protection.  At the breed 
level, unique genetic differences exist between the breeds of each domestic species, for 
example in goats, the Nubians have high milk yield, the Small East African are hardy 
animals and the West African Dwarf goats are trypanotolerant.  Thus, genetic diversity 
at the level of species can be exploited to satisfy human needs for varieties. 
 27
  
2.4.2 Loss of Genetic Diversity in Domestic Animals 
Reduction in genetic diversity has been expressed primarily in terms of loss of 
breeds and strains.  It is postulated that the current rate of extinction of species, breeds 
and strains is greater now than at any time in the past (Hammond, 1993).  It is estimated 
that at least 30% to 40% of all animal genetic resources are currently at high risk of 
extinction (Hammond, 1994).  However, adequate records do not exist to enable 
reliable estimates of either loss rates of the breeds or of domestic animal diversity itself.  
The existing data on the number of endangered breeds are likely to be underestimates of 
the magnitude of the problem.  The loss of genetic diversity is occurring both within 
populations and among populations. 
2.4.3 Loss of Diversity within Population 
The factors that diminish genetic diversity within populations are genetic 
bottlenecks, random genetic drift, inbreeding and human activities.  All the four factors 
are functions of population size (Lacy, 1987; Lande and Barrowclaugh, 1987).  These 
described as follows: 
2.4.3.1 Genetic Bottlenecks  
A demographic bottleneck occurs when a large population experiences a severe, 
temporary reduction in size due to environmental or demographic events.  For example, 
natural catastrophes, which occur at unpredictable intervals including events, like 
drought, disease outbreak and war.  These events may kill a certain percentage of a 
population and therefore reduce the effective population size.  The result is that the 
genetic variability of all subsequent generations is contained in the few individuals that 
survive the bottleneck and reproduce.  Hence, some genetic diversity is lost in the 
process.  The magnitude of the loss in diversity depends on the size of the bottleneck 
and the growth rate of the population afterward (Hunter, 1996).  Another demographic 
event which may lead to bottleneck effect is the founder event.  A founder event occurs 
when a few individuals of a population establish a new population.  The genetic 
constitution of the new population depends up on the genetics of the founder animals.  
The genetic diversity of the original larger populations is reduced because the sample of 
 28
 genes in the few founder animals is not likely to be representative of the original gene 
pool.   
Generally passing through a genetic bottleneck can create two problems 
(Carson, 1983): a loss of certain alleles, especially rare alleles, if no individuals 
possessing those alleles survive, and a reduction in the amount of variation in 
genetically determined characteristics due to the presence of fewer alleles and decline in 
heterozygosity.  Baker and Moeed (1987) have shown that the founder populations of 
several species of introduced birds have fewer alleles and lower heterozygosity than the 
source populations.  Bottlenecks have been shown to reduce allozyme heterozygosity.  
Leberg (1992) suggested that extremely low levels of allozyme heterozygosity in broad 
geographical surveys imply the occurrence of one or more recent severe bottlenecks.  
The overall effect of bottlenecks is the decline in fitness of the individuals in the 
population. 
 
2.4.3.2 Random genetic drift  
Genetic drift is a random change in gene frequencies in small populations, 
attributable to sampling error.  In small populations each generation retains just a 
proportion of the gene pool of the previous generation.  Subsequently after many 
generations of random genetic drift, small populations will usually result in loss of 
variability as a consequence of fixation of some alleles on loci and loss of other alleles.  
 
2.4.3.3 Inbreeding depression  
Inbreeding results from the mating of two closely related individuals.  However, 
the degree of relationship may vary.  ?Closebreeding? refers to the mating of very close 
relatives such as sibling to sibling or parent to offspring.  Its probability of occurrence 
increases in small populations if mating occurs at random.  Inbreeding allows the rare, 
harmful recessive alleles to become expressed in the homozygous form, with resulting 
harmful effects on the offspring (Selander, 1983; Ralls et al., 1988; and Charlesworth, 
1987; Lomker and Simon, 1994) such as reduction in fertility, fecundity, offspring size, 
growth and survival, and physical deformities.  Since inbreeding depresses reproductive 
fitness, it is assumed to increase the risk of extinction.  This presumption is supported 
by correlation between extinction and inbreeding on laboratory and domestic animals 
 29
 (Soul?, 1980).  However, gene flow from outside populations is beneficial in avoiding 
inbreeding and the erosion of genetic diversity (Miller and Waits, 2003). 
 
2.4.3.4 Human activities  
Extinction of species, breeds and strains/varieties through human activities 
represents the greatest threat to genetic diversity.  The major threats to genetic diversity 
that result from human activity are habitat destruction and degradation, pollution, 
introduction of exotic species, and over-exploitation (Frankham, 1994).  These threats 
are all caused by an ever-increasing use of the world?s natural resources due to 
expanding human population and development of market economy.  The growth of 
towns, factories and mines in developing countries creates a cash market for livestock 
products.  Consequently, the traditional farmers who formerly kept animals for their 
own needs begin to supply the cash market.  Genetic variation is being lost as farmers 
in developing countries abandon their local breeds in favour of high-yielding breeds for 
commercial production.  Even if human activities do not directly eliminate a breed or 
strain, loss of genetic variation is taking place as the number of individuals in 
populations is reduced.  In the long run the population size of a breed or strain may 
become so small that the bred/strain is no longer viable and may eventually go extinct.  
 
2.4.4 Loss of Diversity among Populations  
Loss of genetic diversity among-populations occurs when historically divergent 
and isolated populations experience an artificially high rate of gene flow from other 
populations.  In domestic animals this is typically happening, as traditional farmers mix 
and interbreed their populations by moving from one place to another when they set up 
new settlements.  It also occurs through crossbreeding by increasingly promoting 
universal use of very few superior breeds per species to upgrade local breeds.  The 
result is that the uniqueness of formerly distinct breeds/strains is diminished or lost.  
Moreover, there are concerns that, if there exist co-adapted groups of genes within each 
subpopulation, merging may result in the break-up of these and the loss of the particular 
adaptedness of each subpopulation to its environment (Templeton, 1986).  Lacy (1987) 
has shown that the effect of gene exchange between subpopulations is to increase the 
variance within groups, decrease the variance between groups and decrease the total 
 30
 variance.  According to Hammond (1994) about 50 % of the total variance for several 
quantitative traits in cattle, sheep and pig breeds is at the between-breed level, the 
remainder being common to all breeds.  Hence, a move to one breed would eliminate 
half of the total variation.  Loss of genotypic variance is irreversible as is the loss of 
alleles from a population.  This is because it is impossible to sort out the alleles into 
groups of genotypes resembling the original subpopulations (Nelson and Soul?, 1987). 
 
 
2.4.5 Approach to conservation of animal genetic diversity 
Conservation of genetic resources has become a major issue of public interest in 
the last decade or so.  This is due to increase awareness that the genetic diversity is the 
basis for a species evolutionary flexibility and responsiveness to environmental changes 
and that loss of genetic diversity will restrict the opportunities available to mould 
domestic animals to meet unpredictable future changes and requirements.  Hence, 
genetic diversity needs to be preserved.  The loss of variation within breeds is 
continually balanced by the introduction of new variation through new mutations; 
however, it is unwise to rely on new mutations as they are rare events and generally 
have deleterious effects on reproductive fitness (Frankham, 1994).  Moreover, the 
genetic variation present as differences among breeds cannot be regenerated.  Each 
breed or strain is a result of mutation, genetic drift as well as separate adaptation and 
evolution, often over many centuries following different selection pressures imposed by 
climate, endemic parasites and diseases, nutrition and human selection.  Each breed 
thus comprises a unique set of genes, which must be conserved.  
Conservation of domestic animal diversity has been defined as the total sum of 
all operations involved in the management of animal genetic resources so that the pool 
of genetic diversity is maintained over time (Hammond, 1993).  It encompasses 
management of human activities in such a way that animal genetic resources are best 
utilized and developed to meet immediate and short term human needs for future 
generations.  Most attention in conservation of animal diversity has been directed 
toward rare breeds.  However, in management of animal genetic resources, the 
fundamental problem is not the distinction between the breeds that are endangered and 
those that are not, but between those that are perceived to have little or no current utility 
 31
 and those which do have current utility or seem likely to have an immediate future use 
(Barker, 1994).  
The critical issues in conservation of domestic animal diversity according to 
Notter et al. (1994) includes: (i) defining utilization strategies for available breeds to 
optimize production efficiency and sustainable utilization throughout the global range 
of environments and production systems; (ii) livestock improvement strategies that 
make appropriate use of global domestic animal diversity; (iii) development of 
optimum strategies for assessment and preservation of the genetic diversity, which 
exists in breeds that have currently little or no value in commercial production.  FAO 
(1995) has recommended the following strategies for effective management of domestic 
animal diversity at global level and for each species: 
? Identifying and listing all breeds. 
? Describing and characterizing breeds in order to understand their unique 
qualities and potential contributions and to understand which breeds have the 
potential to make the greatest variety of future contributions. 
? Monitoring the population statistics for each breed and regularly reporting to the 
world the populations currently at risk of extinction.  
? Sorting adequate samples of as many breeds as possible, generally in the form of 
frozen semen, ova, and embryos, to enable the future regeneration of lost 
populations of animals. 
An overview of the historical background of animal genetic resources and the 
reasons for concern about their future, both in the developing and developed world, has 
been presented by Barker (2002).  He concluded that the primary focus in the 
conservation of domestic animal diversity is on conservation of breeds.  
In general, two methods are used for conservation of genetic resources to support 
future livestock improvement: in-situ conservation and ex-situ conservation.  
 
2.4.5.1 In-situ conservation  
This involves maintaining breeds in their farm environment.  In-situ 
conservation is primarily the active breeding of animal populations for food and 
agriculture, such that diversity is best utilized in the short term and maintained for the 
longer term (Hammond, 1994).  Operations pertaining to In-situ conservation include 
performance recording and development of breeding programmes.  It also includes 
 32
 ecosystem management for sustainable production of food and agriculture.  It has been 
suggested that the number of animals required maintaining a breed in western countries 
ranges from 150 to 1500 breeding females, depending on the species and reproductive 
rate.  However, in developing countries the number should not decline below 5000 
breeding females (Hodges, 1990).  The advantages of this approach are that the animals 
are still being utilized, the performance characteristics can be properly recorded and 
evaluated, and the breeds have the opportunity to evolve.  The disadvantages are that 
selection and genetic drift may result in unfavorable genetic changes if the population is 
small.  There is a risk of increasing inbreeding and hence homozygosity, which is 
associated with reduced fitness.  The animals are at risk from disease and other natural 
disasters. Also, they are likely to be less productive and so more costly to maintain. 
 
2.4.5.2 Ex-situ conservation 
In the context of conservation of domestic animal diversity, ex-situ conservation 
means storage.  Generally three approaches are used for Ex-situ conservation. 
 
2.4.5.2.1 Maintaining breeds in farm parks or other animal collections  
It involves the breeding animals of a sample of a breed outside its normal 
production environment or habitat.  Many of the pros and cons of this approach are 
similar to the in-situ conservation method.  However there is potentially more control 
over management of the population.  It has been suggested that 50 individuals should be 
the minimum number necessary to maintain genetic variability (Franklin, 1980).  With 
such population only 1 % of the genetic variability will be lost per generation.  This 
figure is on the safe side. Indeed it has been suggested that animal stocks can be 
maintained with a loss of 2?3% of the variability per generation (Primack, 1993).  
According to Notter et al. (1994) populations can be maintained at an inbreeding rate of 
1% per generation corresponding to an effective population size (Ne) of 50%, which 
can be achieved with 14 males and 100 females under random mating.  
 
2.4.5.2.2 Creating a conservation herd (gene pool)  
This involves crossing several rare breeds together, then breeding them to 
maintain genetic variability.  It is an effective way of conserving genetic variation for 
two or three breeds.  Maintenance of genetic diversity is almost better served by 
 33
 pooling five breeds in a conservation herd with Ne of 250 (Notter et al., 1994).  
However there is a greater risk of loosing useful genes when more populations are 
combined.  The disadvantage of this approach is that, although useful genes may be 
conserved, the identity of individual breeds is lost.  
 
2.4.5.2.3 Cryopreservation  
This involves frozen storage of rare breeds in the form of living semen, ova, 
embryos or tissues, which can be used to regenerate animals.  Cryopreservation of 
semen and embryos is a powerful tool for preservation of genetic diversity (Moore, 
1985).  In the situation of a critical threat with high probability that a breed will become 
extinct, preservation of genetic material of individual animals in the form of germ cells 
and embryos is necessary to ensure that an adequate genetic pool is retained for future 
improvement programs.  Breeding technologies as artificial insemination (AI) and 
embryo transfer (ET) may provide support for this approach.  The use of frozen semen 
in conservation program is particularly feasible where a tradition of use of AI is already 
strong.  Collection of semen of endangered local breeds should take place as part of the 
AI program.  According to Smith (1984) semen from 25 unrelated sires would provide 
an exceptionally good sample of the genetic diversity of a breed.  Also Notter et al. 
(1994) recommended cryopreservation of semen from 5 breeds with 10 sires from each 
breeds and argued that it will provide greater preserved diversity than frozen semen of 
25 sires of each of two breeds.  For loci with multiple alleles, the probability of 
retaining all alleles from the rare breed is higher for the conservation herd supported 
with frozen semen than for a conservation herd maintained by random mating alone 
(Lomker and Simon, 1994), however the ability to rapidly recreate living purebreds is a 
key issue in germplasm utilization.  If frozen semen is used as the only method of 
conservation, then several generations of backcrossing are needed to reinstate the breed 
concerned.  In contrast, breeds can be reinstated rapidly from frozen embryos.  Embryo 
preservation is particularly important when the reproductive rate is low or the 
generation interval is long.  Furthermore, maternal cytoplasmic genes (e.g. mtDNA) 
cannot be preserved from semen, but can be effectively conserved by maintenance of 
even a modest number of female lineages.  However, for all species, the likely cost per 
haploid genotype maintained or per animal sampled is much less for semen than for 
embryos (Notter et al., 1994).  In general, embryo preservation should be attempted 
 34
 only when ET is already in use for livestock improvement or when a breed has both 
clear potential value and a high likelihood of extinction. 
Frozen storage of semen and embryos are relatively expensive, but it has the 
advantage that after the initial investment, semen and embryos provide much greater 
flexibility for finding ways of enhancing the reproduction of individuals, which are 
refractory to natural breeding conditions (Moore, 1985).  In addition, the breed being 
conserved is free from unintended genetic changes (Simm, 1998).  During storage 
frozen genetic material is at less risk from disease and natural disasters than live 
animals, but obviously at risk from technological failures.  Splitting the material 
collected from a particular breed or population and storing it in/at different locations 
can minimize these risks.  The disadvantages are that reproductive technologies are not 
uniformly successful or presently available for all individuals or all species and the 
expertise is not always available in the places where it is needed most.  Also the breeds 
conserved in this way are not able to adapt to changes in the production environment or 
new disease challenges.  Semen or embryos frozen now may as well be unable to meet 
future requirements for improvement. 
 
 
2.5 Molecular techniques for the assessment of animal genetic diversity 
2.5.1 Rationale for molecular genetic approach  
Biological variation is of wide occurrence in nature. Morphological variations in 
type, traits and production characteristics have been known and characterized within 
and between species for many centuries.  Phenotypic characterization has traditionally 
been used in the description of breeds.  The characters vary from external features 
(horns, ears, coat colour), physical body measurements (heart girth, height, body 
length), production traits (body weight, milk yield, wool production, carcass traits), 
reproductive traits (fertility, age at first kidding, kidding interval), to survival traits 
(mortality, resistance to diseases, tolerance to heat and poor management and feeding) 
(Aboagye, 1992; Aboagye et al., 1994).  The description of breeds with respect to coat 
colour, horns, physical body measurements and productive traits is based up on 
observation of the phenotype.  An organism?s phenotype is principally a manifestation 
of its genotype.  Phenotypic characters have the advantages of being easily seen and 
measured and usually much lower costs are incurred during phenotypic characterization 
 35
 compared to genetic characterization (Minelli, 1993).  For these reasons, phenotypic 
characters have been used extensively for characterization and identification of breeds.  
One of the problems associated with phenotypic characterization is the difficulty to 
combining different measures in order to provide a useful tool for the description of a 
breed.  Most phenotypic characters are polygenetically inherited and many of them are 
influenced by environmental effects and with sometimes strong genotype-environment 
interaction.  Furthermore, phenotypic characters are affected by natural selection.  
Therefore, due to the influence of different environmental conditions and different 
selection pressure on different kinds of characters, interbreed phenotypic comparison is 
unlikely to give meaningful results. 
Earlier work to study genetic variability of individual animals and populations 
employed screening of protein variants by gel electrophoresis.  Polymorphism in gene 
products such as enzymes, blood group systems and leucocytes antigens have been used 
for investigating genetic diversity at the molecular level.  Using the above protein 
markers, numerous studies have estimated genetic variability, gene flow and 
phylogenetic relationships among populations (Barbancho et al., 1984; Nguyen, 1990; 
Pepin and Nguyen, 1994).  However, these techniques lack the power to resolve the 
differences between closely related breeds (Meghen et al., 1994; Dowling et al., 1996) 
since a great deal of genetic variation remains undetectable by using protein markers.  
Moreover, the genotype frequencies estimated from protein markers may be influenced 
by natural selection among alleles (Alexandrino et al., 1983; Pemberton et al., 1988; 
Mopper et al., 1991) making it difficult to interpret inter-population comparisons.  
The goal of breed characterization is to determine the genetic diversity within 
and between breeds.  The near ultimate description of an animal should be a description 
of the sequence of nucleotides that comprise its genome.  Describing differences and 
similarities in the DNA of two or more populations can provide the measure of relative 
genetic distances of such populations from each other.  The analysis of DNA has 
several significant advantages over protein markers for the study of molecular 
population genetics and systematics (Dowling et al., 1996): (i) the genotype rather than 
the phenotype is assayed; (ii) the expression of DNA markers is not influenced by 
development or by environmental factors;  (iii) one or more sequences appropriate to a 
problem can be selected on the basis of evolutionary rate or mode of inheritance; (iv) 
the methods are for most part general to any type of DNA and, (v) DNA can be 
prepared from small amounts of tissues and is relatively stable, even in non-
 36
 cryogenetically stored tissues.  The last attribute means that genetic information on rare 
or endangered species can be obtained without destructive sampling (Taberlet et al., 
1997) and it is possible to analyze DNA from extinct populations or species (Taylor et 
al., 1994).  More recently, molecular data from DNA markers have received particular 
attention in the study of population variability because of their possible use to 
determine the chronology of evolutionary events using neutral DNA markers.   
 
2.5.2 Molecular techniques in the study of animal genetic diversity 
DNA markers (molecular markers) are the most important tools added to the 
armory of the scientists to characterize, distinguish, or relate individuals within or 
between breeds, strains, or species.  The advances of the last two decades in techniques 
on DNA polymorphism and subsequent data analysis have greatly increased the ability 
to understand the genetic relationship among species at the molecular level.  DNA 
genetic markers have now been used for the molecular genetic characterization and 
genetic diversity studies in numerous species.  
The DNA polymorphism to be used widely for genome characterization and 
analysis was restriction fragment length polymorphism (RFLP) (Botstein et al., 1980).  
Large amount of data have been generated since the first demonstration of RFLP 
(Grodzicker et al., 1974) and RFLP have been used as markers in human genetics 
(Botstein et al., 1980), in genetic improvement of plant (Tanksley et al., 1989) and 
animals (Soller and Beckmann, 1982).  The introduction of PCR (Saiki et al., 1988) in 
conjunction with the constantly increasing availability of DNA sequence data also 
represents a milestone in this endeavor.  A variety of different molecular techniques are 
now being used in various laboratories for the study of inter-and intra-specific genetic 
variation at the DNA level.  The most widely used techniques are restriction fragment 
length polymorphism (Botstein et al., 1980) of nuclear DNA (Jeffreys and Morton, 
1987) and mitochondrial DNA (Loftus et al., 1994), minisatellites probing (Jeffreys et 
al., 1985; Jeffreys and Morton, 1987), microsatellites PCR amplification (Litt and Luty, 
1989; Weber and May, 1989), random amplified polymorphic DNA (RAPD) (Welsh 
and McClelland, 1990; Williams et al., 1990), amplified fragment length polymorphism 
(AFLP) (Vos et al., 1995) and sequencing of mitochondrial DNA fragments 
(Cunningham et al., 1994).  These techniques differ in the type of data they generate, in 
 37
 the way that they resolve genetic variations and in the taxonomic levels at which they 
may be most appropriate.   
 
2.5.2.1 Restriction fragment length polymorphism (RFLP) 
Polymorphism at the DNA sequence level is often detected as a variation in 
length of restriction fragments produced by cutting DNA with restriction enzyme 
(Botstein et al., 1980).  Such polymorphism is referred as RFLP and the first genetic 
map of the human genome was produced using this technique in the year 1980.  RFLP 
is the result of differences in the sequence of nucleotides in the DNA of individual 
animal.  These differences are the result of mutations in the genome occurring over time 
and are detected as variations (polymorphism) in length of restriction fragments.  In 
RFLP, DNA is digested with restriction enzymes that cut DNA at particular sites within 
a specific recognition sequence, typically of 4-6 base pairs (bp) long (Dowling et al., 
1996).  Since each enzyme cleaves DNA at a characteristic recognition sequence, the 
complete digestion of a particular DNA produces an array of fragments.  These 
fragments are separated by gel electrophoresis and blotted onto a filter and then probes 
are hybridized to the DNA.  Variations in fragment lengths between individuals, 
populations or species, are results of mutation, which either alter restriction sites or lead 
to insertions/deletions of cleavage sites.  Such polymorphism in a specific gene locus 
can be used to distinguish animal species, populations and individuals.  The advantages 
are that RFLPs give highly reproducible band patterns and are co-dominant markers, 
hence, heterozygotes are distinguishable.  The limitations of RFLPs are that, it is labor 
intensive and relatively expensive, as a good supply of probes is needed and the blotting 
and hybridization steps are time-consuming, and difficult to automate.  Also the extent 
of detectable polymorphism may be limited for intra-species comparison and sufficient 
quantities (e.g. 10 ?g per digestion) of good quality DNA are required.  Because of the 
last limitation, RFLPs are not applicable when very limited amount of source material 
or preserved tissues are available.  
 
2.5.2.2 Mitochondrial DNA (mtDNA) analysis  
Mitochondrial DNA (mtDNA) sequence differences among individuals or 
populations can be assayed indirectly through restriction site analysis (Loftus et al., 
 38
 1994) or directly through DNA sequencing (Cunningham et al., 1994; Bradley et al., 
1996; Kikkawa et al., 1997).  The former is often less expensive and allows rapid 
screening of large samples.  The latter provides often more information as all changes at 
the nucleotide level will be revealed.  In restriction site analysis, isolated mtDNA is 
digested with a restriction enzyme and the resulting fragments separated by gel 
electrophoresis.  Genotypes are comprised of the composite restriction fragment 
patterns following digestion with 10-20 restriction enzymes.  The pattern of restriction 
site loss or gain among genotypes for a panel of restriction enzymes can be used to 
estimate genetic differentiation and evolutionary history of populations or species 
(Avise, 1994).  In sequencing, mtDNA genes are directly sequenced using universal or 
species specific primers that allow amplification of sequences as much as several 
thousand base pairs in length.  Since different genes in the mtDNA genome evolved at 
different rates, rapidly evolving genes can be analyzed to determine the relationship of 
recently diverged populations, whereas slowly evolving genes can be used to answer 
systematic questions involving distantly related species.  
 
2.5.2.3 Random amplified polymorphic DNA (RAPD) 
RAPD markers were developed in the early 1990s (Williams et al., 1990).  The 
RAPD polymorphisms are based on mismatches in primer binding sites or 
insertion/deletion events and therefore usually result in the presence or absence of an 
amplified product from a single locus.  RAPD involves the use of a single and short 
arbitrary primer which can be synthesized or are available commercially.  Using of such 
primer in PCR amplification results in several discrete DNA band products.  Each 
product is derived from a region of the genome that contains two short segments in 
inverted orientation, on opposite strands that are complementary to the primer and 
sufficiently close together for amplification to work.  More particularly this technique 
utilizes random 10-base oligonucleotide as primer to amplify discrete fragments of 
genomic DNA through PCR (Williams et al., 1990) i.e. in contrast to using two specific 
primers as used in a conventional PCR for amplification of a specific region in DNA. 
RAPD employs a single primer (10 mer) to amplify DNA of any genome region that 
happens to be flanked in opposite orientation within 5000 bp of each other.  During the 
annealing step the primer anneals to the template at two sites on complementary strand 
of DNA template.  If these priming sites are within an amplifiable range of each other, a 
 39
 discrete DNA product is formed through the amplification.  In RAPD, the amplification 
products are separated on agarose gels in the presence of ethidium bromide and the 
resultant bands are visualized under ultraviolet light (Welsh and McClelland, 1990; 
Williams et al., 1990, 1993).  The presence or absence of bands can be scored and the 
data converted into similarity matrices for calculation of genetic distances (Gwakisa et 
al., 1994). 
RAPD method has opened a new area of genetic analysis due to its relative 
simplicity and speed.  It is being applied widely in livestock animals for various 
purposes such as breed characterization (Gwakisa et al., 1994; Kantanen et al., 1995; 
Teale et al., 1995; Yadav et al., 2001), species identification, parentage testing, and 
pedigree analysis (Chung et al., 1998), population studies (Parejo et al., 1998), and 
detection of genetic variations (Kantanen et al., 1995).  Luo et al. (2001) applied RAPD 
technique to evaluate genetic relatedness and diversity among five goat breeds and 
confirmed that RAPD marker analysis can be used to determine genetic diversity and 
relationship among and within goat breeds (http://www2.luresext.edu/goats/ 
library/Luo2001-1.html).  
The advantages of RAPD are that there is no requirement for DNA probes or 
sequence information for the design of primers, the procedure is quick, simple and can 
be automated, and very small amount of DNA (e.g. 10 ng per reaction) is required, 
moreover this technique does not require any restriction enzyme and radioactivity, and 
the technique is cheaper (Williams et al., 1993; Kantanen et al., 1995).  Disadvantages 
of RAPD are however many.  First, RAPD methods are very sensitive to the PCR 
reaction conditions, DNA quality and PCR temperature profiles (Karp et al., 1998).  It 
is absolutely critical to maintain strictly constant PCR conditions in order to achieve 
reproducible results.  Second, the markers are dominant, thus heterozygotes cannot be 
detected.  Third, in the absence of pedigree analysis, the identity of individual bands in 
the multi-band profiles is not known and there can be uncertainty in assigning markers 
to specific loci.  This makes it difficult to use RAPD in inter-population or inter-
 specific comparisons (Dowling et al., 1996).  Fourth, single bands on the gel can 
sometimes be comprised of several co-migrating amplification products.  
2.5.2.4 Amplified fragment length polymorphism (AFLP)  
The AFLP method is a more recent PCR based technique (Vos et al., 1995), 
which is essentially intermediate between RFLPs and RAPDs.  In AFLPs, the first step 
 40
 involves restriction digestion of the genomic DNA and the oligonucleotide adapters are 
ligated to the ends of the restricted fragments.  The second step involves selective 
amplification of sets of restriction fragments.  Either a pre-selection step is performed 
using magnetic beads followed by a round of selective PCR or two selective rounds of 
PCR amplification are applied (Vos et al., 1995).  In the last step, the amplified 
products are separated by polyacrylamide gel electrophoresis and can be visualized 
using radioactive or fluorescent labeling. 
Compared to RAPDs, AFLP results are much more easily reproducible (Vos et 
al., 1995).  Compared with RFLPs, AFLPs are faster, less labour intensive and provide 
more information (Powell et al., 1996).  Because of their large genome coverage (on 
average they give 100 bands per gel compared with 20 for RAPDs), AFLPs are 
particularly good for mapping, and also fingerprinting and genetic distances can be 
calculated between genotypes.  However, AFLPs share many of the limitations, with 
respect to band homologies and identities, as RAPDs, for example as RAPD bands, 
AFLP are codominant markers.  
 
2.5.2.5 Variable number of tandem repeats (VNTRs)  
Throughout the genome of individuals there are many regions comprised of 
tandemly repeated simple sequences.  These repeat sequences vary in number and 
hence in length and are generally called variable number of tandem repeats (VNTRs), 
although the terms minisatellites and microsatellite are used depending on the size of 
the sequence of base pairs which is repeated.  Jeffreys et al. (1986) first reported the 
term minisatellites.  Minisatellites region are characterized by randomly repeating 
oligonucleotide units ranging from 10 to 60 bp in length (Haberfeld et al., 1991, 
Armour and Jeffreys, 1992) whereas in microsatellite the repeated sequences are only 2 
to 5 bp long (Tauz, 1989; Weber and May, 1989; Queller et al., 1993).  Although 
variation in both systems is the result of change in copy number, the underlying 
mechanism of mutation and the chromosomal locations differ.  Microsatellite loci are 
randomly distributed and subject to replication slippage (Weber and May, 1989; 
Schl?tterer and Tautz, 1992) while minisatellites loci tend to be concentrated near 
telomeres and their length vary following intra-molecular or intra-allelic recombination 
and gene conversion mechanisms (Jeffreys et al., 1994). 
 41
 Minisatellites are assayed by digestion of DNA with restriction enzymes which 
do not cleave within the tandem repeats and Southern blot transfer and hybridization 
using minisatellites sequences or synthetic oligonucleotide probes.  Two strategies can 
be applied.  In one approach, the mixture of restriction fragments can be separated by 
size on agarose gel in the presence of an electric field and visualized by hybridization 
with a radioactively labeled probe based on the core sequence of the repeat (Jeffreys et 
al., 1985).  The result is complex patterns of bands that are usually unique to an 
individual.  The complex band patterns are widely known as DNA fingerprint.  DNA 
fingerprint has become an extremely powerful tool for testing parentage (Burke, 1989), 
relatedness of individuals as well as overall population variability (Wayne et al., 1991).  
A major advantage of this method is that many probes can be applied across a wide 
spectrum of plants and animals.  An apparent problem with DNA fingerprinting is that 
it is a multilocus approach, which reveals multiple fragments some of which 
segregating together.  Such linkage disequilibrium between bands can only be detected 
with the analysis of pedigrees.  Moreover, assigning specific fragments to a particular 
locus is difficult and hence, identifying alleles and determining genotypes is not 
possible in the absence of pedigree analysis. Last but not least co-migrating fragments 
might not be homologous.  This will lead to inflated band sharing coefficients between 
non-relatives (Lynch, 1988).  With second approach, specific minisatellites loci are 
studied using single locus probe rather than multilocus ones (Nakamura et al., 1987).  
The advantage of this method is that alleles can be assigned to specific loci and 
genotypes can be identified.  However, cloning of minisatellites loci is a relatively 
complex process.  This method requires more efforts than multilocus fingerprinting and 
the variation at individual loci tends to be taxonomically restricted (Hanotte et al., 
1991). 
Microsatellite loci are usually examined one at a time via PCR (several loci can 
be examined simultaneously by multiplexing).  The PCR amplification products can be 
visualized using radioactive or florescence methods.  Under radioactive methods, PCR 
amplified microsatellites can be detected either by direct incorporation of a single 
labelled dioxynucleotide triphosphate (dNTP) like (?-32P) dCTP during thermal cycling 
(internal labeling) or a single 5? (?-32P) dATP end-labelled primer in the PCR mix (end 
labelling).  The products are resolved on acrylamide gels, fixed, dried and 
autoradiographed.  Under the fluorescence methods, fluorescently labeled dNTPs are 
used for internal labeling in the PCR.  Alternatively and more commonly one of the 
 42
 PCR primers will be fluorescently labeled.   The PCR products, fluorescently labeled, 
are separated on polyacrylamide gel and detected when excited to fluorescence by a 
laser and analyzed by computer software.  Although the initial cost is high, this system 
offers considerable advantage in terms of size accuracy and enhances throughout.  An 
advantage of microsatellite over minisatellites is their smaller size and single loci can 
be amplified using PCR.  Thus, there is a great potential for studying allele frequencies 
at single hypervariable loci in contemporary and historic populations.  Moreover, co-
 dominant genotypes can be scored and exact allele sizes can be determined.  The 
technique has great potential for speed and accuracy once the appropriate PCR primers 
are known and a great deal of polymorphism can be determined.  The use of 
microsatellites over RFLPs and RAPDs is that their genetic basis of variability is 
readily apparent.  Unique primers amplify a genomic region including a well-defined 
repeat structure that is responsible for the observed variation.  Compared to 
minisatellites microsatellites are more conveniently spaced throughout the genome 
while minisatellites tend to be found more frequently near the ends of chromosomes.  
Also most minisatellites alleles are too long to be amplified easily by PCR.  
Microsatellites typically are less than 400-500 bp long, consisting of 10- 30 copies of a 
repeat that is four bp in length and so they are more amenable to PCR typing.  The 
disadvantage of microsatellite is that, the work required to develop primers for each 
new species examined and that only a few allelic states are possible, hence, increasing 
the chance of parallel evolution of a particular sequence repeat.  However, with the 
increasing number of microsatellite maps for economically important species and due to 
the fact that primers developed for one particular species can be sometimes applied 
across a wide range of related taxa (Moore et al., 1991).  Cloning of microsatellites for 
development of primers for new species will be unnecessary in near future. 
 
2.5.3 Application of Microsatellite markers for the analysis of population 
2.5.3.1 Microsatellite as genetic markers for population studies 
Microsatellites usually called Simple Tandem Repeats (STRs), are highly 
polymorphic class of genetic markers (Weber and May, 1989), consist of short 
sequence repeat motifs of tandemly repeated di, tri, tetra or penta nucleotide sequences 
that occur at large number of loci throughout the eukaryotic genome (Hamada et al., 
 43
 1984, Tauz, 1989).  The dinucleotide repeats cytosine ? adenine (CA) or Guanine ? 
thymine (GT) on the other strand is the most common in mammalian genomes (Meghen 
et al., 1994).  Polymorphism of microsatellite markers takes the form of variation in the 
number of repeats at any given locus.  Microsatellites with more than ten dinucleotide 
repeats tend to be highly informative (Weber, 1990).  For instance, in haploid human 
genome there are at least 35,000 CA repeats.  That is they occur every 100,000 bp 
(Weber, 1990).  Tri- and tetra-nucleotides have been shown to occur at a frequency of 1 
every 300 ? 500 kb on X-chromosome (Edwards et al., 1991).  Due to exceptionally 
high rate of mutation, the majority of microsatellite loci are highly polymorphic in most 
mammalian species (Jeffreys et al., 1988; Weber, 1990).  Alleles at microsatellite loci 
conform generally to Hardy-Weinberg principle and segregate in a Mendelian fashion.  
Most microsatellite loci are selectively neutral.  This makes them compatible with the 
assumption of the neutral theory of population genetics (Kimura and Crow, 1964).  
Microsatellite polymorphism has advantages over other techniques in that it is a very 
reliable, highly accurate and repeatable method and has the potential of being 
standardized all over the world and the other potentially very large advantage of 
microsatellite in population genetics is the fact that primers developed for a particular 
species can be sometimes applicable across a wide range of related taxa.  For example, 
some primers developed for cattle can be used for sheep, goats, yak, etc.  
Microsatellite markers are currently the markers of choice for a wide range of 
molecular genetic studies such as establishing genetic linkage maps (Kappes et al., 
1997), analysis of mating system and population structures (Bruford and Wayne, 1993; 
Queller et al., 1993; Schl?tterer and Pemberton, 1994) and reconstruction of 
phylogenetic relationships among populations (Bowcock et al., 1994; Roy et al., 1994, 
Forbes et al., 1995; Takezaki and Nei, 1996; MacHugh et al., 1997).  Their desirable 
characteristics are that: microsatellite loci are found in large numbers and are relatively 
evenly distributed throughout mammalian genomes, Mendelian co-dominant 
inheritance and relative ease of scoring through polymerase chain reaction.  
Microsatellites are conserved across related species.  Moreover microsatellite variation 
is independent of age, sex and environmental changes and hence can be detected at the 
early stage of development. 
The majority of microsatellites are found in non-coding regions of genome.  
However Morin et al. (1994) also reported the presence of microsatellite in protein 
coding region exhibiting regulatory role in gene expression and trinucleotide repeats 
 44
 have been used for linkage analysis in association with disease susceptibility genes in 
human (Richards and Sutherland, 1994).  Microsatellites have been found conserved 
between closely related species (Stallings et al., 1991) allowing primer sets to be used 
across species (Moore et al., 1991).  They have been extensively used for linkage 
mapping in diverse organisms from human to mosquitoes (Weissenbach et al., 1992; 
Crawford et al., 1995; Vaiman et al., 1996; Barendse et al., 1997) and their use has 
enabled the identification of quantitative trait loci in major livestock species (Georges et 
al., 1993).  Microsatellites have proven useful in the analysis of paternity and kinship 
(Queller et al., 1993) and statistical calculation of probability at the individual (Edward 
et al., 1992) and the population levels (Paetkau et al., 1995b). 
Microsatellite variation has been used to study the amount of hybridization 
between closely related species (Gottelli et al., 1994; Roy et al., 1994).  Spatial 
distribution of alleles has been used to study local genes flow and population sub-
 structure (Gottelli et al., 1994; Allen et al., 1995).  Effective population sizes and 
inbreeding have been estimated from microsatellite data (Edwards et al., 1992).  
Microsatellites have been increasingly used for the study of genetic variation between 
and within animal populations.  They have been successfully applied in the study of 
genetic variation in vertebrates (Kashi et al., 1990a, b) and more particularly in 
livestock species such as poultry, sheep, goats, buffaloes and bovines (Haberfeld et al., 
1991; Ganai and Yadav, 2001, Arora R et al., 2004, Li and Valentini 2004).  
Microsatellite DNA variation has been employed to assess genetic distances between 
strains of poultry (Kuhnlein et al., 1989) and within and among European and African 
cattle (MacHugh et al., 1994; MacHugh et al., 1997; Okomo et al., 1998) and sheep 
breeds (Buchanan et al., 1994; Arranz et al., 1998).  Microsatellite markers are well 
suited for forensic applications; population genetics studies etc. and consequently, have 
become the basis of genomic analysis in rodents, pigs, dogs, cattle and goats (Barendse 
et al., 1994; Ellegren, 1993; Roy et al., 1994; Serikawa et al., 1992; Ganai and Yadav, 
2001). 
 
2.5.3.2 Estimation of genetic distances from microsatellite data  
One of the most important problems in the conservation of genetic diversity in 
domestic animals is how to choose appropriate breed or population for conservation, 
among the many still available in role to maximize genetic diversity conserved.  One 
 45
 way could be to compute the genetic distance between pairs of breeds, and to choose 
breeds that show the widest genetic differences among them (Barker, 1994).  Genetic 
distances are used either for estimating divergent time or for construction of 
phylogenies, which in turn can be used to decide which breeds should be kept (Nei and 
Takezaki, 1994).  Many genetic distances have been developed, of which a few remain 
in regular use (Nei, 1987).  Each of these genetic distances has unique evolutionary and 
statistical properties, and evolutionary relationships inferred from each genetic distance 
can be quite different.  The merits and demerits of each measure have been investigated 
(e.g. Takezaki and Nei, 1996).  
Many of the earlier measures of genetic distances are geometric distances, 
which are based explicitly on geometric representations of gene frequencies rather than 
on any particular population genetic models.  These distance measures obey the axioms 
of Euclidean geometry.  The most commonly used of geometric distance is Cavalli-
 Sforza and Edwards?s (1967) chord distance (DC).  In this distance measure, 
populations are represented as points on a multidimensional hypersphere with 
dimension equal to the number of alleles.  DC is the chord distance between two points 
representing the gene frequencies in two different populations.  This distance 
incorporates an angular transformation of gene frequencies and the principle of triangle 
inequality is fulfilled.  The transformation has the effect of standardizing the distance 
with respect to random drift, so that the rate of increase in genetic distance under drift is 
nearly independent of the initial gene frequencies.  However, little attention is paid to 
the relationship between genetic distance and evolutionary change of populations.  The 
problem of geometric distance is that the absolute values of these measures do not have 
any particular biological meaning (Nei et al., 1983), and only the relative values are 
important for finding the genetic relationship among populations.  
Another widely used measure of genetic distance, which is close to DC, is the 
Nei et al.?s (1983) angular distance (DA).  This is a modified Cavalli-Sforza and 
Edwards distance measure.  It is based on genetic drift model and is more efficient in 
determining the true topology of phylogenetic trees especially for closely related 
populations.  It is the most widely used distance measure since it reportedly increase 
more slowly with time and maintains a linear relationship for longer periods of time 
(Nei, 1984).  
The most commonly used genetic distances are based on population genetic 
models and are supposed to be linearly proportional to time: either to the average time 
 46
 to a common ancestor (coalescence time) of pairs of genes or to the time since two 
populations diverged from one another.  There are two most important genetic models, 
which have been proposed to explain the mode of mutation as cause of evolutionary 
divergence between copies of homologous genes since the time of their common 
ancestor.  The models are infinite alleles model (IAM) and stepwise mutational model 
(SMM).  The IAM model assumes that every new mutation gives rise to an allele that 
does not already exist in the population while the SMM assumes that mutations increase 
or decrease allele size by single unit.  IAM seems to apply to classical genetic markers 
such as protein and blood polymorphisms (Nei, 1987) and SMM apply to microsatellite 
loci (Goldstein et al., 1995a) since alleles at microsatellites are thought to evolve by a 
stepwise mutation process.  However, Edwards et al. (1992) and Valdes et al. (1993) 
have shown that the pattern of mutation for microsatellite loci follow a kind of a 
stepwise mutation model which is close to the IAM of mutation when a relatively short 
evolutionary time is considered.  Thus microsatellite data can be analyzed in a manner 
similar to protein polymorphisms.  The most widely used distance measure based on 
IAM is Nei (1972) standard genetic distance (DS).  DS is intended to measure the 
number of codon substitutions per locus that have occurred after divergence between a 
pair of populations and is expected to increase linearly with time (Nei and Takezaki, 
1994).  DS assumes that the rate of gene substitution per locus is uniform across loci and 
lineages.  This assumption is violated under microsatellite loci.  According to Farris 
(1981) DS is not appropriate for making a phylogenetic tree because it is not a metric 
measurement and does not obey the triangle inequalities.  Another widely used distance 
measure under IAM is Rogers? (1972) distance (DR).  This measure has the virtues of 
simplicity and it satisfies the principle of triangle inequality.  However, DS is better than 
DR with respect to the linear relationship with time.  Both DS and DR have the 
undesirable property of being influenced by within-taxon heterozygosity.  That is the 
distance between two taxa that are fixed for alternate alleles exceeds that between two 
taxa in which one or both are heteroallelic but have no alleles in common.  
Although a large number of genetic distances could conceivably be applied to 
microsatellite, estimators of population parameters based on the IAM are unlikely to 
apply to microsatellite (Goldstein et al., 1995a).  This is because the majority of 
mutations at microsatellite loci are stepwise in nature, changing allelic sizes up or down 
by one or very few number of repeats.  Thus, distance measures, which apply to 
microsatellites generally, assume the SMM model.  Three distances have been 
 47
 developed specifically for application to microsatellite evolution assuming the SMM 
model: Goldstein et al.?s (1995a) genetic distance (??)2, Goldstein et al.?s (1995b) and 
Slatkin (1995) average square distance (ASD) and Shriver et al. (1995) stepwise 
distance (DSW).  The expected values of (??)2 and ASD increase linearly with time 
under the SMM model, but DSW seems to be non-linear with time.  Although these 
distances apply to microsatellite, their efficiencies for obtaining the correct topology of 
the phylogenetic tree are generally low (Takezaki and Nei, 1996).  This is because 
ASD, (??)2 and DSW have large sampling errors due to their dependence on the 
variation within population and on the sensitivity to fluctuation in effective population 
size (Takezaki and Nei, 1996).  
Other widely used measures of genetic distances are various estimators of 
population subdivision (FST) such as GST (Nei, 1973; Weir and Cockerham, 1984) and 
RST (Slatkin, 1995).   
Different measures of genetic distances are appropriate for different purposes.  
Nei and Takezaki (1994) and Forbes et al. (1995) have reported that classical distances 
based on the IAM model and the distances based upon multidimensional geometric 
considerations are more effective in obtaining the correct tree topology than the SMM-
 based genetic distances.  Nei and Takezaki (1994) and Takezaki and Nei (1996) have 
shown that DC and DA are more efficient in obtaining the correct tree topology than 
other distances whereas DS and (??)2 are more appropriate for estimating evolutionary 
time than other distances. 
 
 
2.5.3.3 Microsatellite genetic distances in Livestock 
Moazami-Goudarzi et al. (1997) used 17 microsatellite loci and 13 biochemical 
markers for studying the genetic variability between 10 cattle breeds.  They used Nei?s 
(1972) genetic distance and chord distance (DC) (Cavalli-Sforza and Edwards, 1967) for 
studying the genetic relationships between breeds.  Yadav et al. (1998) reported the use 
of 22 bovine microsatellite primer pairs in 36 animals of Murrah breed of buffalo.  All 
the 22 primer pairs amplified a product, and 14 of which were polymorphic.  Pandey 
(1999) also reported the conservation of cattle microsatellites in two breeds of Indian 
buffalo. 
 48
 Loftus et al. (1999) carried out a microsatellite survey on 8 humpless cattle 
breeds from the Near East, 3 from Europe, 1 from West Africa and 2 zebu breeds from 
India with 20 microsatellite loci.  They reported a large genetic divergence between the 
non-hybrid taurine and zebu groups.  Mommens et al. (1999) studied variation between 
an African and five European taurine using data on 31 microsatellites loci.  They 
calculated genetic distance and constructed Phylogenetic tree using neighbor joining 
(NJ) method.  They reported American bison as an out-group. Wilson and Strobeck 
(1999) have shown genetic relatedness among wood and plains bison populations by 
using 11 microsatellite loci.  Yang et al. (1999) studied the genetic relationship among 
five indigenous Chinese goat breeds with six microsatellite markers.  A neighbour-
 joining tree was constructed using Nei?s standard genetic distance.  In this tree, 
Neimonggol and Liaoning were grouped together, then with Taihang; while Tibetan and 
Matou each had their own branch. 
Hanslik et al. (2000) studied the introgression of new world Holstein-Friesian 
into European Holstein-Friesian populations.  A tree of individuals based on the 
proportion of shared alleles was produced; a clear distinction between old world and 
new world Holstein-Friesian populations was found.  Ritz et al. (2000) studied the 
phylogenetic analysis of the tribe bovine using 20 microsatellite loci.  They inferred that 
the genetic distances used for the study: (??)2 and DC are both considered being 
appropriate for phylogenetic studies among species and subgenus.  They concluded that 
Bos, Poephagus and Bibos each emerged as a subgenus within the genus Bos.  On the 
other hand, Bison, Bubalus and Syncerus each formed a separate discrete genus.  A 
suitable large panel of polymorphic markers for each species, especially those that are 
less common and less studied.  Cross-species utilization of microsatellite loci not only 
save time and efforts in the laboratory, but also enable the construction of comparative 
maps between related species. 
MacHugh et al. (1998) studied genetic demarcation among the breed of seven 
European cattle using 20 microsatellite markers.  They reported correct breed 
designation with accuracy approaching 100% using data from a panel of 10 
microsatellite loci.  Martin et al. (1999) studied genetic diversity in six Spanish native 
cattle breeds using 30 microsatellites markers.  They calculated allele frequencies and 
genetic distance for the characterization of breeds and construction of dendrogram 
respectively.  They found closest population those representing Austrian breeds, the 
most divergent being Menorquina and fighting bull. 
 49
 2.6 Methods for the reconstruction of phylogenetic tree 
Phylogenetic trees are graphical representations of nodes (taxonomic units) and 
branches (pathways connecting nodes) that summarize the evolutionary relationships 
among organisms (Avise, 1994).  Reconstruction of a phylogenic tree is considered as 
statistical inference of a true phylogeny, which is unknown (Nei, 1996).  A large 
number of methods have been described for phylogenetic tree building.  In principle 
there are two ways of building trees, either by clustering methods or by exhaustive 
search methods.  These methods have the advantage of being ease to perform, resulting 
in very fast computer programs (Saitou and Imanishi, 1989).  The clustering methods 
follow a set of steps (an algorithm) and leading to a tree.  The biggest limitation is that 
cluster methods do not provide a ranking criterion for evaluating trees.  Exhaustive 
search methods use optimality criteria for choosing the best tree among the set of all 
possible trees.  This criterion is used to assign to each tree a score or a rank, which is a 
function of the relationship between the tree and data. Exhaustive Search methods have 
the advantage that they provide a ranking criterion under which all trees can be 
evaluated (Page and Holmes, 1998).  However, they are computationally demanding, 
resulting in computer programs, which are much slower to perform and cannot give 
exact results when there are more than 12 taxa (Nei, 1996). 
The methods of phylogenetic inference currently used in molecular phylogenetic 
can be classified in to three major groups: distance methods, parsimony methods and 
likelihood methods.   
 
2.6.1 Distance methods  
In distance methods, an evolutionary distance is computed for all pairs of 
populations and a phylogenetic tree is constructed from the pairwise distances.  The 
most commonly used methods are the ?unweighted pair group method with arithmetic 
averages? (UPGMA) of Sneath and Sokal (1973) and the Neighbour-joining (NJ) of 
Saitou and Nei (1987).  
The UPGMA has gained popularity mostly because of its simplicity and because 
of its speed (though many other distance methods are as fast).  The UPGMA uses real 
(uncorrected) distance values and a sequential clustering algorithm.  This method of 
tree construction is not sensitive to differences in branch length or unequal rates of 
evolution.  The UPGMA is a clustering method, which operates as follows: A distance 
 50
 matrix is scanned for the smallest distance element and the corresponding operation 
taxonomic units (OTUs) are joined.  This distance element in the matrix is then 
discarded.  The matrix is again scanned for the smallest remaining distance element and 
the process is repeated until all OTUs are finished.  In each cycle of the clustering 
procedure, OTUs or previously formed clusters are grouped according to the smallest 
mean distance between the populations involved.  Each OTU contributes equally to the 
calculation of these mean distances.  UPGMA produces a dendrogram, which is rooted 
implicitly at the point where the last clusters join.  The major assumption of UPGMA 
clustering is the equal rate of evolution along all dendrogram branches (Nei, 1987).  
UPGMA is considered as a heuristic method for finding the least square ultra metric 
tree for a distance matrix.  UPGMA method performs well in recovering correct tree 
topologies (Tateno et al., 1982; Nei et al., 1983).  This is due to the fact that genetic 
distance estimates are subject to large stochastic error and the distance averaging 
aspects of UPGMA reduces the effects of this error (Nei, 1987).  
The NJ is a clustering method, which attempts to correct the UPGMA method 
for its strong assumption (frequently invalid) that the same rate of evolution applies to 
each branch.  Hence this method assumes unequal rates of molecular change among the 
branches (Nei, 1987) and yields an unrooted tree.  In the NJ method the raw data are 
provided as a distance matrix and the construction of a tree begins by assuming that the 
populations are all related to one another by a star phylogeny, thus the initial tree is a 
star tree.  The next step is to begin a procedure that groups certain populations together.  
In order to do this, first, branch lengths for tree are estimated by least square and the 
sum of the branch lengths for the entire tree is calculated.  Then pairs of populations 
that give the minimum sum of branch lengths are considered as neighbours and are 
joined and considered as single entity.  At each step of the analysis a transformed 
distance matrix is constructed in which the separation between each pair of nodes is 
adjusted on the basis of their average divergence from all other nodes.  The tree is 
constructed by linking the least distance pair of nodes as defined by the modified 
matrix.  When two nodes are linked, their common ancestral node is added to the tree 
and the terminal nodes with their respective branches are removed from the tree.  This 
pruning process converts the newly added common ancestor into terminal node on a 
tree of a reduced size.  The process of considering all possible pairs is repeated and 
completed when two nodes remain, separated by a single branch.  According to Nei 
(1996), NJ is a heuristic method for estimating the minimum evolution tree.  NJ 
 51
 produces a reasonably good phylogenetic tree when the extent of population differences 
is large and a large number of loci are examined (Rzhetsky and Nei, 1992).  The 
efficiency of the NJ method in obtaining the true tree is due to the fact that in each step 
of population clustering the principle of minimum evolution is applied, and according 
to Nei (1996) the repeated application of this principle reduces the effects of sampling 
errors in topology construction.  
 
2.6.2 Maximum parsimony method 
Parsimony implies that simpler hypotheses are preferable to more complicated 
ones.  Maximum parsimony (MP) is a character-based method that infers a 
phylogenetic tree by minimizing the total number of evolutionary steps required to 
explain a given set of data, or in other words by minimizing the total tree length.  
Maximum Parsimony methods fall under optimality criteria methods.  Under MP the 
best tree is the one that requires the smallest number of evolutionary changes to explain 
the observed differences among OTUs.  A number of related methods have been 
described for inferring phylogenies by parsimony approach, the most common 
employed are Fitch and Wagner parsimony (Fitch, 1971; Farris, 1970), Camin-Sokal 
parsimony (Farris, 1971), Dollo parsimony (Farris, 1981) and generalized parsimony 
(Swofford and Olsen, 1990).  These methods differ in their underlying evolutionary 
assumptions but they are united by the goal to reconstruct the evolution of OTUs on a 
tree based on the constraint of invoking the fewest possible evolutionary changes (Page 
and Holmes, 1998).  Among the advantages of parsimony methods are that the 
approach apparently makes few assumptions about the evolutionary process and it is 
based on implicit assumption about evolution that evolutionary changes are rare 
(Swofford et al., 1996).  The rarity of evolutionary changes implies that the tree that 
minimizes the change is likely to be the best estimate of the actual phylogeny (Pages 
and Holmes, 1998).  According to Swofford et al. (1996), if there are no multiple 
nucleotide substitutions at each site, MP is expected to generate the correct topology as 
long as enough parsimony-informative sites are examined.  The main disadvantage of 
MP is that under some models of evolution it is not consistent.  Indeed, when the true 
tree has a special type of topology and branch lengths, MP may generate an incorrect 
topology even if an infinite number of loci are examined (Nei, 1996).  Takezaki and Nei 
(1996) have shown that this can happen even if the rate of nucleotide substitution is 
 52
 constant for all evolutionary lineages.  Another disadvantage of MP is that in parsimony 
analysis it is difficult to treat the phylogenetic inference in a statistical framework 
because there is no natural way to compute the means and variances of minimum 
number of substitutions obtained by the parsimony procedure.  However, under certain 
circumstances, MP is quite efficient in obtaining the correct topology and it is the only 
method that can take care of insertions and deletions of nucleotides, which may give 
important phylogenetic information (Nei, 1996).  
 
2.6.3 Maximum likelihood  
Maximum Likelihood (ML) evaluates a hypothesis about evolutionary history in 
terms of the probability that the proposed model and the hypothesized history would 
give rise to the observed data set.  The supposition is that a history with a higher 
probability of reaching the observed state is preferred to a history with a lower 
probability.  The method searches for the tree with the highest probability or likelihood.  
The data are observed gene frequencies or nucleotide sequences; the unknowns are the 
branching order and branch lengths of the tree (Felsenstein, 1981).  In ML approach, a 
model of the evolutionary process that accounts for the conversion of one sequence or 
allele into another must first be specified.  The next step is to calculate the probability 
that the observed data would have been generated by a particular tree topology under 
that model.  The tree that makes the observed data the most probable evolutionary 
outcome (i.e. the tree that gives the highest likelihood) is chosen as the ML estimate of 
the phylogeny.  Maximum Likelihood is an appealing method of inference as it can 
incorporate explicit models of evolution and also allows statistical test of evolutionary 
hypothesis.  Hasegawa et al., 1991 reported that the maximum likelihood method is 
superior to distance methods in the efficiency to produce the correct true topology 
particularly when the evolutionary rate differs among lineages.  There advantages of 
maximum likelihood methods over other methods are: (i) they have often lower 
variance than other methods (i.e. it is frequently the estimation method least affected by 
sampling error); (ii) they tend to be robust to many violations of the assumptions in the 
evolutionary model; (iii) even with very short sequences they tend to outperform 
alternative methods such as parsimony or distance methods; (iv) the method is 
statistically well founded; (v)they evaluate different tree topologies, and they use all the 
sequence information.  The objection to likelihood methods is that the dependence on a 
 53
 model raises the question of which model to use as the pattern of nucleotide substitution 
varies from site to site and with evolutionary time (Tateno et al., 1982).  The biggest 
problem with ML is a technical difficulty for computing the likelihood itself.  This is 
computationally time-consuming, and recently it has been shown that more than one 
maximal likelihood values may exist for a given tree (Nei, 1996); making it difficult to 
guarantee that the likelihood value for that tree is actually maximal.  
 
2.7. Statistical analysis 
2.7.1. Hardy-Weinberg equilibrium (HWE)  
In a large random mating population with no selection, mutation or migration, 
the allelic or gene frequencies and the genotype frequencies remain constant from one 
generation to the next.  Thus a population with constant gene and genotype frequencies 
is said to be in Hardy-Weinberg equilibrium (HWE) (Falconer and Mackay, 1996).  
Factors that affect the HWE include: 
Non-random mating:  Non-random mating can occur in cases where related 
individuals have a greater probability of mating with each other than with other 
members of the population and where individuals that are geographically close are more 
likely to mate with each other than those that are not geographically close. 
Mutation:  This is the process that produces a gene or chromosome that is 
different from the wild type.  A mutation occurs when a DNA gene is damaged or 
changed in such a way as to alter the genetic message carried by that gene.  Mutation 
brings about genetic variation in a population by producing novel variants of genes. 
Migration:  The movements of genes caused by individuals moving, including 
new individuals entering (immigration) or leaving (emmigration) a population, 
introducing or removing genetic material and thereby changing allele frequencies. 
http://www.modernhumanorigins.com/m.thml).  Migration of genes into a population 
results in an increase in a population?s genetic variation and the migration of genes out 
of a population may result in a reduction of the genetic variation. 
Selection:  A natural process resulting in the evolution of organisms? best 
adapted to the environment (http://www.cogsci.princeton.edu/cgi-bin/webwn?stage= 
1&word =selection).  Only the individuals that are better adapted to the environment or 
 54
 able to mate successfully could pass their genes on to the next generation. Selection 
generally results in a reduction of genetic variation in a population. 
Random genetic drift i.e. the random sampling of alleles at every generation:  
Random genetic drift is a stochastic process (by definition).  Random genetic drift is the 
change in gene frequencies due to chance or sampling effects.  It is very dependent upon 
population size.  Genetic drift is fundamentally the result of a finite population size.  The 
effect of genetic drift is infinitely proportional to the population size.  Thus when the 
population size is small, e.g. due to strong bottleneck effects in the past, there are 
greater changes in gene frequency under genetic drift at every generation (Cavalli-
 Sforza et al., 1994).  Therefore, the smaller the population size, the greater the chances 
of sampling errors occurring. 
In the study of many populations, it is important to determine whether the loci 
and the populations genotyped were in HWE and whether there were any significant 
deviations from the HWE.  
 
2.7.2. Linkage disequilibrium (LD) 
Linkage disequilibrium is often termed ?allelic association? and is a measure of 
the degree of association between two alleles in a population.  Measures of linkage 
disequilibrium (such as D) quantify how frequently two alleles are found on the same 
chromosome in a certain population (http://dblab8.csie.ncu.edu.tw/Glossary.htm).  A 
population is said to be at linkage disequilibrium at a set of loci if the alleles are not 
randomly assorted in the next generation, but are inherited together as a unit.  Linkage 
disequilibrium can be generated by finite population size, random genetic drift, 
mutation, selection, non-random mating, and migration (http://darwin.cwru.edu/~conti/ 
notes/linkagedisequilibrium.htm).  It can have positive or negative values, the tendency 
for two 'alleles' to be present on the same chromosome (positive LD) or not to segregate 
together (negative LD).  LD is decreased by recombination.  Thus, it decreases every 
generation of random mating unless there are some processes opposing the approach to 
linkage 'equilibrium'.  
 
 55
 2.7.3. Genetic diversity 
Genetic diversity refers to the variation at the level of individual genes 
(polymorphism), and provides a mechanism for populations to adapt to their ever-
 changing environment.  The more variation the better the chance that at least some of 
the individuals will have an allelic variant that is suited for the new environment, and 
will produce offspring with the variant that will in turn reproduce and continue the 
population into subsequent generations (http://genetics.nbii.gov/GeneticDiversity.html).  
The mean number of alleles (MNA) detected in each population and the 
expected heterozygosities are good indicators of the genetic polymorphism within the 
populations.  The MNA is the average number of alleles observed in a population, while 
the expected heterozygosities are the proportion of heterozygote expected in a 
population.  Numbers of alleles per locus per population are obtained by direct 
counting.  Generally, the MNA is dependent on the sample size because of the presence 
of unique alleles that occur in low frequencies in populations and also because the 
number of observed alleles tend to increase with increases with population size (Nei, 
1987).  Therefore the comparison of the MNA between samples of different sizes may 
not be meaningful unless the sample sizes are more or less the same (Nei, 1987; Hart 
and Clark, 1989).  
 
2.7.4. Genetic relationship between the populations 
2.7.4.1. Genetic distance 
Genetic distance is a measurement of genetic relatedness of samples of 
populations (whereas genetic diversity represents diversity within a population).  The 
estimate is based on the number of allelic substitutions per locus that have occurred 
during the separate evolution of two populations.  This difference measured between 
two populations provides a good estimate of how divergent they are genetically (Avise, 
1994).  Thus when the genetic distance is large, the genetic similarity is low and the 
time they diverged from each other is greater, while when genetic distance is small the 
genetic similarity is high and the time they diverged from each other is smaller (Avise, 
1994).  Cavalli-Sforza and Edwards (1964; 1967) were the first to use gene frequency 
data to classify populations.  They used genetic distance data to construct a 
phylogenetic tree to infer the history of various human populations. 
 56
 One of the common measures of genetic distance in use today is Nei?s standard 
genetic distance (DS) (Nei, 1972, 1978).  The DS distance is based on the infinite alleles 
model (IAM) and is said to be more appropriate for estimating evolutionary time.  DS 
increases linearly with time from zero to infinity and will have a value proportional to 
the mutation rate (Kalinowski, 2002).  However, Nei et al. (1983) noted that the 
modified Cavalli-Sforza and Edwards distance measure (DA) usually known, as angular 
distance is more efficient in recovering the true topology of an evolutionary tree of 
closely related populations.  The DA distance has proven to be useful for reconstructing 
phylogenies (Takezaki and Nei, 1996).  DA has also been reported to increase more 
slowly with time and to maintain a linear relationship for longer periods of time (Nei et 
al., 1983).  
 
2.7.4.2. Construction of phylogeny  
The neighbor-joining (NJ) methodology (Saitou and Nei, 1987) implemented in 
the DISPAN program was used to construct the phylogenetic tree of population 
relationships.  The NJ is known to be more efficient than many other methods in 
obtaining the correct tree (Rzhetsky and Nei, 1992).  The neighbor-joining method does 
not assume an equal rate of evolution between lineages unlike the unweighted pair 
group-method with arithmetic mean (UPGMA), another method commonly used to 
construct phylogenetic trees (Nei, Tajima, and Tateno, 1983). 
The DISPAN program (Ota, 1993) was used to generate pair-wise lower triangle 
matrices of genetic distances between the populations.  A neighbor-joining and 
UPGMA trees were then constructed from these matrices using DS and DA (Ota, 1993).  
The reliability of the constructed trees was examined by a bootstrap test with 1000 
replicates re-sampling of loci with replacement (Felsenstein, 1985). 
 
 
2.7.4.3. Principal component analysis (PCA) 
Construction of phylogeny and calculation of genetic distances have the 
disadvantage of not being able to detect the effects of past bottleneck effects on 
populations and any admixture that has taken place between populations (Cavalli-
 Sforza et al., 1994).  An alternative method taking into account the admixture is 
 57
 multivariate analysis e.g. principal component analysis.  PCA can be used to reduce a 
large number of variables to a smaller number of variables without losing too much 
information (Alt, 1990; Cavalli-Sforza et al., 1994).  PCA involves a mathematical 
procedure that transforms a number of (possibly) correlated variables into a (smaller) 
number of uncorrelated variables called principal components.  The first principal 
component accounts for as much of the variability in the data as possible, and each 
succeeding component accounts for as much of the remaining variability as possible. 
http://www.fon.hum.uva.nl/praat/manual/Principal_component_analysis.html. 
PCA simply describes the variation of the multivariate data in terms of some 
linear combinations of the variables, in this case the observed microsatellite allele 
frequencies, which are resolved into smaller reference vectors.  The new variables are 
uncorrelated with each other.  The coefficients defining the linear transformation are 
chosen so as to minimize the variation of the transformed data measure along each new 
co-ordinate axis.  PCA was performed using the MVSP program (version 3.12d) from 
(http://www.kovcomp.com/mvsp/downl.html). 
 
2.7.4.4. Genetic differentiation of the populations 
The populations of most, if not all, species show some levels of genetic 
structure, which may be due to a variety of non-mutually exclusive agents.  
Environmental barriers, historical processes and life histories (e.g. mating system) may 
all, to some extent, shape the genetic structure of populations (e.g. Donnelly and 
Townson, 2000; Gerlach and Musolf, 2000; Palsson, 2000; Tiedemann et al., 2000).  
The understanding of the genetic structure of a population or differentiation between 
populations is of interest to population geneticists because it reflects the number of 
alleles exchanged between populations which influence the genetic composition of 
individuals within these populations (Balloux and Lugon-Moulin, 2002).  Gene flow 
between populations determines the effects of selection and genetic drift, generates new 
polymorphisms and increases the local effective population size (Balloux and Lugon-
 Moulin, 2002).  The simplest parameters for assessing diversity between breeds using 
microsatellite data are the genetic differentiation or fixation indices.  Several estimators 
have been proposed (e.g. FST, GST, ?), the most widely used being FST (Weir and 
Basten, 1990), which measures the degree of genetic differentiation among 
 58
 subpopulations through the calculation of the standardized variances in allele 
frequencies among populations. 
 
2.7.4.4.1. Standardized variance in allele frequencies among populations (FST)  
FST measures the effect of population subdivision, which is the reduction in 
heterozygosity in a subpopulation due to genetic drift.  FST is the most inclusive 
measure of population substructure and is most useful for examining the overall genetic 
divergence among subpopulations.  It is also called co-ancestry coefficient (?) [Weir & 
Cockerham, 1984] or 'fixation index' and is defined as correlation of gametes within 
subpopulations relative to gametes drawn at random from the entire population 
(Subpopulation within the Total population).  It is calculated using the subpopulation 
(average) heterozygosity and total population expected heterozygosity.  FST is always 
positive; it ranges between 0 = panmixis (no subdivision, random mating occurring, no 
genetic divergence within the population) and 1 = complete isolation (extreme 
subdivision).  FST values below 0.05 indicate negligible genetic differentiation whereas 
> 0.25 means very great genetic differentiation within the population analysed.  FST is 
usually calculated for different genes, and then averaged across all loci, and all 
populations (FST = (HT ? HS)/HS).  FST can also be used to estimate gene flow: Nm = 
0.25 (1-FST)/FST.  This highly versatile parameter is even used as a genetic distance 
measure between two populations instead of a fixation index among many populations 
(see Weir, Genetic Data Analysis II, 1996, and Kalinowski, 2002). 
FST estimates were calculated using the program FSTAT 2.9.3 (Goudet, 2001; 
http://www.unil.ch/izea/software/fstat.html) and GENEPOP version 3.4 (Raymond and 
Rousset, 1995). 
 
2.7.4.4.2 The coefficient of gene differentiation (GST) 
The coefficient of gene differentiation (GST) developed by Nei (1973) is an 
extension of the (Nei, 1972) genetic distance theory between a pair of populations. GST 
can be computed directly from allele frequencies in terms of expected heterozygosities 
within and between populations.  Unlike the FST, the estimation of heterozygosities in 
GST relies only on allele frequencies (Nei, 1987).  This method offers several 
advantages because it is not affected by the number of alleles at the locus and neither is 
affected by the evolutionary forces such as mutation, selection and migration, which 
may be taking place in the organism. GST can be defined as: 
 59
  
GST = DST/HT 
 
Where DST is average gene diversity between and within populations and HT is the 
expected total heterozygosity.  The GST estimates in this study were calculated using the 
DISPAN program (Ota, 1993). 
2.7.4.5 Analysis of molecular variance (AMOVA)  
Population differentiation and the distribution of genetic variation can also be 
determined by AMOVA (Excoffier et al., 1992).  AMOVA is built on the classical 
analysis of variance.  However, the data analyzed with AMOVA uses allele frequency 
data. The AMOVA analysis was done using the software Arlequin version 2.000 from 
(http://lgb.unige.ch/arlequin/). 
2.7.4.6 Assignment of individuals to populations 
In population genetics, individuals in a sample have to be classified into specific 
populations or breeds using a set of phenotypic criteria.  Population assignment 
methods assign individuals to populations using a clustering method. Here individuals 
that are genetically similar are assigned to the same cluster.  There are broadly two 
types of clustering methods that could be used; distance based and model based 
methods.  
Distance based methods first calculate a pair-wise distance matrix (for e.g. Nei?s 
distance or Cavalli-Sforza and Edwards Chord distance), then use them to assign the 
individual to the closest population.  These methods are very easy to apply but have the 
disadvantage of being too dependent on the distance measure and it is difficult to asses 
the confidence level that should be assigned to clusters obtained (Pritchard et al., 2000).  
Model based methods include the maximum likelihood based methods, which 
include the frequency and the Bayesian methods.  The frequency method, which was 
first described by Paetkau et al. (1995a, b) first computes the allele frequencies in the 
populations, then computes the likelihood of the individual?s genotype occurring in 
each population and then assigns an individual to the population in which the 
individual?s genotype is most likely to occur.  This method assumes that the populations 
are in HWE and there is no linkage disequilibrium.  However, the major drawback of 
this method is that when there are null or rare alleles in a population, the corresponding 
 60
 allele frequency is zero and hence a likelihood of zero leading to the elimination of the 
population from the analysis. 
The Bayesian method (Rannala and Mountain, 1997) is similar to the frequency 
method; here the sample population frequencies are used to derive the probability 
density of the population allele frequencies.  In this method, the problem of dealing 
with null allele is eliminated.  
Individual population assignments were done using two programs WHICHRUN 
4.1.0 (Banks and Eichert, 2000) and GeneClass (Cornuet et al., 1999).  Using the 
maximum likelihood approach that incorporated jackknife iterations, the WHICHRUN 
4.1.0 program used allele frequency distributions of the multilocus genotypic data to 
assign individuals to a population using the method of Paetkau et al. (1995a, b).  Here 
individuals are sampled one at a time and the allele frequency recalculated in the 
absence of each genotype before the individual is assigned to the most likely source 
population.  The program then computes a ratio between the likelihood of the most 
likely populations and the second most likely population.  The allocation stringency 
was of a log of odds (LOD) ratio greater than two or P < 0.001. 
Population assignments were also done using the GeneClass program, which 
allows for different estimation procedures.  The probability that the individual belongs 
to a population (P-value) was calculated by simulating 10,000 genotypes using Monte 
Carlo sampling of allele frequencies observed in a population.  The observed frequency 
of the tested individual?s genotype was computed in the population under investigation.  
An individual was considered correctly assigned to a population when it was excluded 
from all other populations with a high statistical significance of P < 0.001. 
 
2 7.4.7 Inference of population structure 
Admixture analysis at both the population and individual level and individual 
specific assignment to the populations was performed using the program STRUCTURE 
version 2.0 (Pritchard et al., 2000), which implements the Pritchard, Stephens and 
Donnelly (2000) method, a model based clustering algorithm that infers population 
structure using multilocus genotypes.  This Bayesian method assumes Hardy-Weinberg 
as well as linkage disequilibrium between loci within each population.  The method 
places individuals in K populations or clusters and individuals can have membership in 
 61
 multiple clusters, if their genotypes indicate that they are admixed.  These membership 
coefficients add up to one across the various clusters for each individual or population.  
Five runs were made at each four Ks, where K is the number of inferred clusters and a 
burn-in period of 100,000 steps as well as a Markov Chain Monte Carlo (MCMC) of 
500,000 being applied.  The program DISTRUCT offers various options for displaying 
these results, where an individual?s estimated membership coefficient are displayed as 
various segments partitioned into different colors, depending on the value of K chosen. 
 62
 CHAPTER THREE 
3.0. Materials and methods 
3.1. Populations/Breeds and sampling strategy 
3.1.1.  Animals 
The aim of this study was to quantify genetic diversity within and among 
Ethiopian goat populations/types and to clarify the genetic relationships between them.  
For this purpose, 11 indigenous goat populations/types of Ethiopia and nine reference 
breeds (four from other African countries viz. two from Kenya, one from Guinea Bissau 
and one from Botswana), and five from non-African countries (one breed from Turkey, 
Egypt, Italy, Mongolia and Saudi Arabia each) were included in this study.  The 
sampled animals of Ethiopian goats were selected to represent the type/population as 
much as possible based on the available information of phenotypic characteristics and 
geographical distribution from previous survey (physical description and management 
systems) by Farm-Africa (FARM ? Africa, 1996) (Fig. 19a).  The finally identified 
animals for the study were sampled from distantly located villages and sampling was 
made 1-5 animals per flock.  In order to avoid sampling of related individuals, farmers 
were asked about the source and familial relationships of the individuals.  For all 
populations/types/breeds, approximately equal number of females and males were 
sampled.  The breed, number of animals per breed, country of origin, and phenotypic 
characters/subtypes are shown in Table 2.  For this study, seven Ethiopian goat types 
(Woyto-Guji, Long-Ear Somali, Hararghe Highland, Central Highland, Keffa, Short-
 Ear Somali and Abergalle) were sampled fresh from the field, while for the project also 
used DNA or blood samples or FTA? cards already available at ILRI for remaining 
four populations of Ethiopia and all the reference breeds. 
 
 
 
 
 
 
 
 
 
 
 
 
 63
  
Figure 19a. Map of Ethiopia showing distribution of goat populations/types 
sampled. 
 
3.1.2. Sampling protocol 
Blood was collected from jugular vein by puncher using 10 ml capacity 
vacutainer tubes containing EDTA (100 ?l neat).  Every time a new needle was used for 
each animal to avoid cross contamination between samples.  The blood was mixed with 
EDTA by gently inverting the tube about three to five times.  Blood collected at the 
field site was brought to laboratory with all necessary precautions to avoid exposure to 
extreme temperature (hot or cold) and direct sunlight. In the laboratory blood samples 
were centrifuged at 2000 rpm for 15 ? 20 minutes the same day evening.  The plasma 
was discarded and the buffy-coat containing the peripheral blood lymphocytes (PBL) 
was transferred with a wide-mouth Pasteur pipette into a 2 ml cryotube.  Then 1 ml of 8 
M urea was added into the tube, the content of the tube mixed by gentle inversions five 
times.  The tube properly labeled and stored in upright position at room temperature, 
ready for DNA extraction.  
 64
 In the case of sampling using the FTA cards, the blood in the tube was placed on 
ice and transferred onto FTA classic cards (Whatman? Bioscience) on the same day.  
After drying, these cards were transferred into bags containing silica gel for storage and 
future use. 
 
 65
 Table 2 Populations/types used in the analysis of genetic diversity of Ethiopian goats and the reference Breeds/populations 
from outside Ethiopia. 
Population/Breed Number genotyped Origin Sub-type 
Abergale  46 Ethiopia Afar and Worre type 
Arsi-Bale  46 Ethiopia The Woyto-Guji goat is the closest relative 
Afar 44 Ethiopia Worre and Abergalle types  
Central Highland  40 Ethiopia Western Highland and Keffa goat types 
Western Lowland (Gumez) 43 Ethiopia Most closely related to the Central and Western Highland goats 
Hararghe Highland 40 Ethiopia Related to both the Short- and Long-eared Somali  
Keffa 38 Ethiopia Western Highland goat 
Long-eared Somali 48 Ethiopia Short-eared Somali 
North-West Highland 45 Ethiopia Central Highland and Keffa types 
Short-eared Somali 43 Ethiopia Long-eared Somali 
Woyto-Guji 46 Ethiopia Related to the Arsi-Bale 
Boran Galla 35 Kenya Short Ear Short Horn 
Small East Africa 39 Kenya Short Ear Short Horn 
Guinea Bissau 46 Guinea Bissau  
Tswana 29 Botswana Short Ear Short Horn 
Alpine 31 Italy  
Ardi 25 Saudi Arabia   
Baladie 22 Egypt  
Hair 31 Turkey  
Zalaajinistiin 39 Mongolia  
 66
 3.2. DNA extraction 
DNA was extracted either from peripheral blood lymphocytes (PBL) or FTA? 
blood cards (Whatman? Bioscience).  The extraction procedures applied are described in 
the following section. 
 
3.2.1. DNA extraction from peripheral blood lymphocytes (PBL) 
Buffy-coat of each sample (200 ?l) was put into 1.5 ?l-eppendorf tube and 500 ? 
700 ?l of digestion buffer was added and mixed by vortexing.  Then 40 ?l of proteinase-K 
(20mg/?l) + 100 ?l of 10% SDS (Sodium Dodecyl Sulphate) solution was added and mixed 
and incubated for 3 hours at 50oC.  Then equal volume of a mixture of phenol, chloroform, 
and Isoamylalcohol (25:24:1) was added, mixed by vortexing and centrifuged at rpm 
(13,000 to 14,000 rpm) for 10 minutes and the supernatant transferred to a new eppendorf 
tube.  To the supernatant, one volume of chloroform was added, mixed for 5 minutes by 
vortexing and centrifuged for 10 minutes at 13,000 to 14,000 rpm and the supernatant 
transferred to a new eppendorf tube.  To the supernatant, two volumes of absolute ethanol 
were added to precipitate DNA and kept for 10 minutes and spin/centrifuged at 13,000 to 
14,000 rpm for 3 minutes and the supernatant was discarded.  Then the pellet was washed 
twice in 70% ethanol and air-dried or oven-dried at 36.8oC for an hour and dissolved in 200 
?l of sterile triple distilled water.  The concentration of DNA was determined by 
spectrophotometer and then the DNA concentration was adjusted to 20 ng/?l by adding 
required amount of water. 
 
3.2.2. DNA extraction from FTA? blood cards (Whatman? Bioscience) 
Around thirty disks (1 mm) were punched out of the FTA? blood cards for each 
sample, from within the region well-stained by the blood smear.  These disks were put into 
an eppendorf tube. FTA purification reagent at the rate of 10 ?l for each disk i.e. 300 ?l was 
added.  The tubes were rotated up and down for 30 minutes and the spent solution was 
poured out carefully.  This process of rotation was repeated twice.  Tris-EDTA buffer (500 
?l) at pH 6.7 was added and the tubes rotated up and down for 30 minutes and the spent 
 67
 solution was poured out carefully.  This step was repeated once.  The FTA? blood card 
disks were dried overnight in an oven at 37?C.  Thus the disks were ready for use and kept 
at room temperature.  One of the disks was used for PCR reaction. 
 
3.3. Microsatellite markers, PCR conditions and genotyping  
Eighteen microsatellite loci were chosen from the goat genome database maintained 
at INRA (Institut national de Rechrch? Agronomique) (http://www.locus.jouy.inra.fr/) and 
from the sheep genome database maintained at Roslin Research Institute 
(http://www.ri.bbsrc.ac.uk/).  The characteristics of the 18 microsatellites are shown in 
Table 3. 
 68
            Table 3. Autosomal microsatellite loci used in the study and their characteristics. 
Locus Chromosome ABI (colour) Primer sequences 5? - 3? Annealing temperature oC 
BM 1818 23 Tet F-AGC TGG GAA TAT AAC CAA AGG R-AGT GCT TTC AAG GTC CAT GC 
55 
BMC 1222 13q12 Hex F-CCA ATT TTG CAG ATA AGA AAA CA R-CCT GAG TGT TCC TCC TGA GT 
55 
BMS 0357 12 Ovine Hex F-TCC AAA CAA GTC TTC TCT ATT TAC C R-CCA AAT AAT TGC TGG TCA GG 
58 
BMS 1494 21 Bovine Hex F-TCT GGA GCT TGC AAA AGA CC R-AAT GGA TGA CTC CTG GAT GG 
55 
ILSTS 0005 10 Fam F-GGAAGCAATGAAATCTATAGCC R-TGTTCTGTGAGTTTGTAAGC 
55 
ILSTS 0011 14 Hex F-GCT TGC TAC ATG GAA AGT GC R-CTA AAA TGC AGA GCC CTA CC 
55 
ILSTS 0044 1 Ovine Tet F-AGT CAC CCA AAA GTA ACT GG R-ACA TGT TGT ATT CCA AGT GC 
57 
ILSTS 0087 6 Ovine Tet F-AGC AGA CAT GAT GAC TCA GC R-CTG CCT CTT TTC TTG AGA GC 
58 
INRA 0005 12 Fam F-CAA TCT GCA TGA AGT ATA AAT AT R-CTT CAG GCA TAC CCT ACA CC 
52 
INRA 0063 18q22 Fam F-ATT TGC ACA AGC TAA ATC TAA CC R-AAA CCA CAG AAA TGC TTG GAA G 
55 
INRA 0132 23 Hex F-AAC ATT TCA GCT GAT GGT GGC R-TTC TGT TTT GAG TGG TAA GCT G 
57 
MAF 0035 23 Ovine Hex F-TCA AGA ATT TTG GAG CAC AAT TCT GG R-AGT TAC AAA TGC AAG CAT CAT ACC TG 
55 
MAF 0065  15 Tet F-AAA GGC CAG AGT ATG CAA TTA GGA G R-CCA CTC CTC TGA GAA TAT AAC ATG 
50 
MAF 0209 17 Fam F-GAT CAC AAA AAG TTG GAT ACA ACC GTG G R-TCA TGC ACT TAA GTA TGT AGG ATG CTG 
57 
OarAE 0129 7 Fam F-AAT CCA GTG TGT GAA AGA CTA ATC CAG R-GTA GAT CAA GAT ATA GAA TAT TTT TCA ACA CC 
60 
OarFCB 0304 19 Ovine Hex F-CCC TAG GAG CTT TCA ATA AAG AAT CGG R-CGC TGC TGT CAA CTG GGT CAG GG 
55 
SRCRSP 0003 10q36 Tet F-CGG GGA TCT GTT CTA TGA AC R-TGA TTA GCT GGC TGA ATG TCC 
55 
SRCRSP 0007 6 Fam F-TCT CAG CAC CTT AAT TGC TCT R-GTC AAC ACT CCA ATG GTG AG 
57 
 69
 The 18 microsatellite loci were analyzed at ILRI, Nairobi, Kenya.  The steps of PCR 
protocol applied was as follows:  all PCR amplifications were performed in a total volume 
of 10 ?l (except for some FTA? blood disks, where a total volume of 20 ?l was used) on 
GeneAMP?PCR system 9700 (Applied Biosystem) thermocycler.  Each PCR reaction 
contained 1.0 ?l of 20 ng template DNA (one FTA card disk), 1.0 ?l of PCR buffer (19.2 
mM MgCl2, 100 mM Tris-HCl (pH 8.3), 500 mM KCl, 0.01% gelatin, 0.25% Tween 20 and 
0.25% Nonidet P40), 0.5 ?l of 2.5 mM dNTP (i.e. mixture of dATP, dCTP, dGTP, and 
dTTP), 0.08 ?l of 200 mM each primer (forward and reverse), 0.04 ?l of Taq (Thermus 
aquaticus) DNA polymerase (Promega) and 7.3 ?l of sterile triple distilled water.  The PCR 
amplification for FTA? blood with a total volume of 20 ?l reaction contains 1 disc, 2.0 ?l 
of PCR buffer (19.2 mM MgCl2, 100 mM Tris-HCl (pH 8.3), 500 mM KCl, 0.01% gelatin 
and 0.25% Tween 20 and 0.25% Nonidet P40), 1.0 ?l of 2.5 mM dNTP (i.e. mixture of 
dATP, dCTP, dGTP, and dTTP), 0.16 ?l of 200 mM of each primer (forward and reverse), 
0.08 ?l of Taq (Thermus aquaticus) DNA polymerase (Promega) and 15.6 ?l of sterile triple 
distilled water.  All amplification conditions included an initial denaturing step of 5 min at 
95?C, followed by 35 cycles of 30 seconds at 95?C, 1 minute at an appropriate primer 
annealing temperature (Table 2), and 1 minute at 72?C.  Duration of final extensions was 
10 minutes at 72?C.  The PCR products were analyzed by electrophoresis on a 4.25% 
denaturing polyacrylamide gel using a 377 ABI automatic DNA sequencer (ABI PRISM 
377, Applied Biosystems).  For this purpose, the forward primer for each locus was labeled 
with a fluorescent dye, either FAM (blue) or TET (green) or HEX (yellow).  Two to four 
PRC products in different size ranges and labeling of the forward primers were co-loaded in 
a single lane of the gel.  In preparation for loading, the samples were usually prepared by 
mixing 0.5-1.0 ?l of each PCR products and 1.8 ?l of loading mix/buffer (made of 
GENESCANTM TAMRA 350, loading dye and formamide in the ratio of 1:1:4 with a little 
modification).  The mixture was heated to 95oC for 2 to 4 minutes and immediately put on 
ice and then 1.5-2.0 ?l was loaded in each lane of the gel.  The GeneScanTM 350-TAMRA 
served as an internal size standard in each lane.  A control sample was included in addition 
to the GeneScanTM 350-TAMRA for further standardization of fragment sizes.  Fragment 
lengths were identified using the GeneScan? software (version 3.1.2 Applied Biosystems), 
while allele sizes identifications were performed by GENOTYPER? software (version 2.0 
Applied Biosystems) using the third order least squares method for size calling (Fig. 19b ?
 g). 
 70
  
 
 
 
Figure 19b: Gel image of Abergalle breed shown by BMS357, ILSTS87 and BMC1222 
350
 300
 340
 250
 200
 150
  
 100 
75
 50
 35
 1 12 24 36
 71
 55 160 165 170 175 180 185 190 195 200 205
   Abergalle pop       1 Blue    ABG1 
1000
 2000
 3000
   (02)Abergalle pop   2 Blue    ABG2 
1000
 2000
 3000
   (03)Abergalle pop   3 Blue    ABG3 
1000
 2000
 3000
 4000
   (04)Abergalle pop   4 Blue    ABG4 
500
 1000
 1500
 2000
   (05)Abergalle pop   5 Blue    ABG5 
1000
 2000
   (06)Abergalle pop   6 Blue    ABG6 
1000
 2000
 3000
   (07)Abergalle pop   7 Blue    ABG7 
1000
 2000
 3000
  
Figure 19c: Allele scoring for Abergalle population using ILSTS05 
 72
  
       1                    12                24               36                    50   
 
Figure 19d: Gel image of Short-eared Somali population shown by OarFCB304, 
                       ILSTS11 and MAF65 
   350 
 
   340 
 
 
 
 
   300 
 
 
 
 
 
 
   250 
 
 
 
 
 
 
   200 
 
 
 
 
   160 
 
   150 
   139 
 
 
 
 
   100 
 
 
   75 
 
   50 
 
   35 
 73
 96 98 100102104106108110112114116118120122124126128130132134136138140142144146148150152154156158160162
 200
 400
 600
   (10)Shor?mali pop. 10 Red      
50
 100
 150
   (11)Shor?mali pop. 11 Green   MAF 65 
100
 200
 300
   (11)Shor?mali pop. 11 Red      
50
 100
 150
   (12)Shor?mali pop. 12 Green   MAF 65 
100
 200
 300
   (12)Shor?mali pop. 12 Red      
50
 100
 150
   (13)Shor?mali pop. 13 Green   MAF 65 
100
 200
 300
  
Figure 19e: Allele scoring for Short-eared Somali population using MAF065  
(Red color is internal standard). 
 74
  
 
 
Figure 19f: Gel image for Short-eared Somali population shown by INRA132-SRCRSP07 
 
1 12 24 36 45
  
 
 
 
 
  340 
350
   300 
  250 
  200 
160
 150
 139
 100
 75
 50
 35
 75
 95 100 105 110 115 120 125 130 135 140 145 150 155 160 165 170 175 180 185 190 195 200 205
 100
 200
 300
   (08)Shor?mali pop.  8 Red      
50
 100
   (09)Shor?mali pop.  9 Blue    OarAE 129 
500
 1000
   (09)Shor?mali pop.  9 Red      
30
 60
 90
   (10)Shor?mali pop. 10 Blue    OarAE 129 
1000
 2000
 3000
   (10)Shor?mali pop. 10 Red      
30
 60
 90
   (12)Shor?mali pop. 12 Blue    OarAE 129 
200
 400
  
Figure 19g: Allele scoring for Short-eared Somali population using Oarae129  
(Red color is internal standard). 
 76
  
3.4. Data analysis 
 
In this study, deviations of the locus/population combinations from HWE were 
determined using the program GENEPOP version 3.4 June 2003 (Raymond and Rousset, 
1995). This involved the use of the Markov chain method to obtain unbiased P estimates of 
Fisher?s exact test (Raymond and Rousset, 1995) for each locus/population combination 
through 30,000 iterations. In order to control type I error, i.e. accepting the hypothesis that 
the population are in HWE when it is not the case, a sequential Bonferroni correction (Rice, 
1989) was performed.  
 
Observed allele frequencies were computed using the excel macros (a program 
kindly provided by Stephen Park, (Spark@tcd.ie). The allele frequencies were used in the 
determination of both within- and between-population genetic variation. 
 
Linkage disequilibrium analysis among pairs of loci in each population was also 
done using the program GENPOP version 3.4 June 2003 using Fisher?s exact tests 
(Raymond and Rousset, 1995) with unbiased P values being derived by Markov chain 
method with 20,000 iterations per batch. A sequential Bonferroni correction (Rice, 1989) 
was performed to control type I error.  
 
Genetic variation was determined as the mean number of alleles (MNA) per locus 
and average observed and expected heterozygosities. Average observed and expected 
heterozygosities were calculated using Microsatellite Analyzer (MSA) version 3.12 
(Dieringer et al., 2003). The MNA and allele frequencies were calculated using the program 
MICROSATELLITE TOOLKIT version 3.1 (Park, 2001). To remove the sample bias, the 
MNA was calculated for all the individuals in a population and for random sample of 20 
individuals, with 250 re-sampling of 20 individuals with replacement using the program 
ALLELE SAMPLER version 1.2 (Mburu et al., 2003). 
 
Both DS and DA genetic distance measures were computed using the DISPAN 
program (Ota, 1993). Phylogenetic trees were constructed from the DS and DA genetic 
distance matrices generated using the Neighbour Joining (NJ) and the Unpaired Group with  
 77
 Arithmetic Mean (UPGMA) methods in the TREEVIEW program.   The reliability 
of the constructed trees was examined by a bootstrap test with 1000 replicates re-sampling 
of loci with replacement (Felsenstein, 1985). 
 
Principal component analysis, which simply describes the variation of the 
multivariate data in terms of some linear combinations of the variables was performed by 
Eigen analysis of covariance matrix calculated from allele frequencies data using the 
Multivariate Statistical Package (MVSP) program (version 3.12d) available from 
(http://www.kovcomp.com/mvsp/downl.html). 
 
FST estimates were calculated using the program FSTAT 2.9.3 Goudet, 2000 
(http://www.unil.ch/izea/software/fstat.html) and GENEPOP version 3.4 (Raymond and 
Rousset, 1995), while the GST estimates were calculated using the DISPAN program (Ota, 
1993). P-values were obtained after 55,000 permutations using standard Bonferroni 
corrections.  
 
Analysis of molecular variance, AMOVA (Excoffier et al., 1992) implemented in 
the program ARLEQUIN version 2.000 (Schneider et al., 2000) available from 
(http://lgb.unige.ch/arlequin/) was performed to quantify further the extent of population 
differentiation and the distribution of genetic variation in the sample populations.  
 
Individual population assignments were done using two programs WHICHRUN 
4.1.0 (Banks and Eichert, 2000) and GeneClass (Cornuet et al., 1999). 
 
Admixture analysis at both the population and individual level and individual 
specific assignment to the populations was performed using the program STRUCTURE 
version 2.0 (Pritchard, 2000), which implements the Pritchard, Stephens and Donnelly 
(2000) method, a model based clustering algorithm that infers population structure using 
multilocus genotypes. Five runs were made at each four Ks, where K is the number of 
inferred clusters and a burn-in period of 100,000 steps as well as a Markov Chain Monte 
Carlo (MCMC) of 500,000 being applied. The program DISTRUCT offers various options 
for displaying these results, where an individual?s estimated membership coefficient are 
displayed as various segments partitioned into different colors, depending on the value of K 
 78
 chosen. The program DISTRUCT offers various options for displaying these results, where 
an individual?s estimated membership coefficient are displayed as various segments 
partitioned into different colors, depending on the value of K chosen. 
 
The data was also tested for possible scoring errors due to stuttering or large allele 
drop-out and occurrence of null alleles using the program MICROCHECKER version 2.2 
(Van-Oosterhout et al.? 2003), where 1,000 Monte Carlo bootstrap simulations were used to 
calculate probabilities with a 99% confidence interval. 
 79
 CHAPTER FOUR 
4.0. RESULTS 
A number of analyses were performed using allelic variation obtained at 15 
microsatellite loci.  Allele frequencies at each locus were calculated for each population. 
Microsatellite variability such as the observed and expected numbers of alleles at each locus 
for each population, mean number of alleles per locus, observed heterozygosity and 
expected heterozygosity were calculated to quantify the genetic variation within the 
population.  Statistics such as fixation indices or F-statistics were performed to describe the 
population structure and genetic differentiation among populations.  Phylogenetic analysis 
and multivariate analysis such as principal component analysis were performed to infer the 
evolutionary relationship among populations.  A number of computer programs were used to 
perform all the analyses. 
 
4.1. Genetic diversity and relationship between Ethiopian and non-Ethiopian goat 
populations 
4.1.1. Test for deviation from Hardy-Weinberg equilibrium (HWE) 
Deviations from HWE proportions were determined for each of 360 locus-
 population combinations using Fisher?s exact test for contingency tables, which was first 
applied by Haldane (1954).  Exact tests are appropriate even when many rare alleles are 
present (Guo and Thompson, 1992; Chakraborty and Zhong, 1994), and should therefore be 
used for population genetic analysis of hypervariable markers such as microsatellite loci 
(Rousset and Raymond, 1995).  Three distinct tests were performed for testing of deviation 
from HWE proportions under the same null hypothesis (HO = random union of gametes): 
probability test, heterozygosity deficiency and heterozygosity excess.  An exact HW test 
(Haldane, 1954; Guo and Thompson, 1992) was applied for the probability test, while an 
exact score test (U-test) as described by Rousset and Raymond (1995) was applied to the 
heterozygosity deficiency and heterozygosity excess.  The probability of the observed 
sample has been used to define the reject zone, and the P-value of the test has been defined 
as a sum of the exact probabilities of all tables (with the same allelic counts) with the same 
or lower probability than the observed table.  Unbiased exact P-values were computed using 
 80
 the complete enumeration algorithm for loci with two or three alleles (Louis and Dempster, 
1987) or estimated by a Markov chain algorithm (Guo and Thompson, 1992) for loci with 
four or more alleles.  The tests were implemented in the program GENEPOP version 3.4 
(Raymond and Rousset, 1995) with the parameters as 30,000 steps of dememorization, 500 
batches and 30,000 iterations per batch. Significance levels for multiple tests were adjusted 
with the Bonferroni corrections (Rice, 1989). 
A total of 776 individuals from 20 different populations located in 9 countries 
(Ethiopian (11), Kenyan (2), Guinea Bissau (1), Mongolian (1), Saudi Arabian (1), Egypt 
(1), Italy (1), Botswana (1) and Turkey (1)) were genotyped using 18 microsatellite markers.  
After testing for linkage disequilibrium, null allele and HWE, 17, 18 and 14 populations 
were observed to have a significantly lower proportion of heterozygotes than that expected 
at HWE for BMC1222, BMS357 and SRCRSP07, respectively (Table 4). Consequently, 
these three markers were removed from further analysis. 
Single locus exact test for HWE was conducted for the 20 populations and the 
remaining 15 microsatellite loci.  Results indicated that 39 out of 300 i.e. 13% of the 
comparisons (P < 0.05; Table 5) deviated significantly from HWE, a figure higher than the 
5% that would be expected from type I error alone.  Application of sequential Bonferroni 
adjustment for multiple testing reduced this number of significant locus-population 
combinations to only three. 
Out of the 15 loci, there was no locus that deviated from HWE for all the populations 
and also no population was observed to deviate from HWE for all the loci.  However, as it is 
indicated in Table 5, there were some populations with several loci that deviate significantly 
from HWE.  On one extreme, five loci found to deviate from HWE in the Hararghe 
Highland population, while on the other extreme, the Gumez and Boran-Galla didn?t deviate 
from HWE at any locus.  Arsi-Bale, Afar, Short-eared Somali, Woyto-Guji, Small-east 
Africa, Guinea Bissau and Mongolia deviated from HWE at only one locus.  Most of these 
loci deviations from HWE corresponded to either heterozygote deficit or excess (Table 5). 
Five out of twenty populations had HWE deviation at loci ILSTS44, INRA05, and 
INRA63, while four out of the twenty populations had HWE deviation at loci NRA132 and 
SRCRSP03. 
 81
 Table 4. Results of Fisher?s exact test for Hardy-Weinberg equilibrium for 20 populations and 18 microsatellite markers. 
 
 
Population 
BM 
1818 
BMC 
1222 
BMS 
1494 
BMS 
357 
ILSTS 
05 
ILSTS 
11 
ILSTS 
44 
ILSTS 
87 
INRA 
05 
INRA 
63 
INRA 
132 
MAF 
35 
MAF 
65 
MAF 
209 
OarAE 
129 
OarFCB 
304 
SRCRSP 
03 
SRCRSP 
07 All loci 
ABG *1 **1 ns **1 ns ns ns ns ns *1 ns ns ns ns ns ns ns **1 ** 
ABA ns **1 ns **1 ns ns ns ns * ns ns ns ns ns ns ns ns ns ** 
AFA ns **1 ns **1 ns ns **1 ns ns ns ns ns ns ns ns ns ns **1 ** 
CEH ns **1 ns **1 ns ns ** ns ns **1 ns ns ns ns ns *1 ns **1 ** 
GUM ns **1 ns **1 ns ns ns ns ns ns ns ns ns ns ns ns ns **1 ** 
HAH ns **1 ns **1 ns ns *1 ns * ns *2 ns ns *1 ns ns **1 ns ** 
KEF ns **1 ns **1 ns *1 ns ns ns ns ns ns *1 *1 ns ns ns *1 ** 
LES ns **1 ns **1 ns ns ns ns ns ** ns ns ns ns ** ns * **1 ** 
NWH ns **1 ns **1 ns ns ns ns ns *1 ns ns ns ns **1 ns ns **1 ** 
SES *1 **1 ns **1 ns ns ns ns ns ns ns ns ns ns ns ns ns **1 ** 
WGJ ns **1 ns **1 ns ns ns ns *1 ns ns ns ns ns ns ns ns **1 ** 
BGA ns **1 ns **1 ns ns ns ns ns ns ns ns ns ns ns ns ns **1 ** 
SEA ns **1 ns **1 ns ns ** ns ns ns ns ns ns ns ns ns ns ns ** 
GUB ns **1 ns **1 ns ns ns ns ns ns ns **1 ns ns ns ns ns **1 ** 
TSW ns ns ns ns ns ns **1 ns ns *1 * ns ns ns *1 ns ns *1 ** 
ALP ns ns ns ns ns ns ns ns ** ns ns ns **1 ns ns ns *1 *1 * 
ARD ns ns **2 **1 ns ns ns ns ns ns ns *1 ns ns ns ns ns ns * 
BAL ns **1 ns **1 ns ns ns *1 * ns ns ns ns ns ns ns ns ns ** 
HAI ns **1 ns **1 ns ns ns ns ns ns *1 ns ns ns ns *1 *1 ns ** 
MON ns **1 ns **1 ns ns ns ns ns ns **1 ns ns ns ns ns ns **1 ** 
All 
population 
ns **1 ns **1 ns ns **1 ns * **1 **1 ns *1 ns ns ns *1 **1  
ns = not significant;  * = P < 0.05;  ** = P < 0.01;  1 = heterozygote deficit;  2 = heterozygote excess. 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA - Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF? Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared 
Somali, WGJ ? Woyto-Guji, BGA ? Boran Galla, SEA ? Small East African, GUB? Guinea Bissau, TSW? Tswana, ALP ? Italian Alpine, ARD - Ardi, BAL ? Baladie, HAI - Hair, MON ? Mongolia. 
 82
 Table 5. Results of Fisher?s exact test for Hardy-Weinberg equilibrium for 20 populations and 15 microsatellite markers. 
 
 
Population 
BM 
1818 
BMS 
1494 
ILSTS 
05 
ILSTS 
11 
ILSTS 
44 
ILSTS 
87 
INRA 
05 
INRA 
63 
INRA 
132 
MAF 
35 
MAF 
65 
MAF 
209 
OarAE 
129 
OarFCB 
304 
SRCRSP 
03 All loci 
ABG *1 ns ns ns ns ns ns * ns ns ns ns ns ns ns ns 
ABA ns ns ns ns ns ns * ns ns ns ns ns ns ns ns ns 
AFA ns ns ns ns **1 ns ns ns ns ns ns ns ns ns ns ns 
CEH ns ns ns ns ** ns ns **1 ns ns ns ns ns *1 ns ** 
GUM ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns 
HAH ns ns ns ns * ns * ns *2 ns ns *1 ns ns **1 ** 
KEF ns ns ns *1 ns ns ns ns ns ns *1 *1 ns ns ns ns 
LES ns ns ns ns ns ns ns ** ns ns ns ns ** ns * ** 
NWH ns ns ns ns ns ns ns *1 ns ns ns ns **1 ns ns ns 
SES *1 ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns 
WGJ ns ns ns ns ns ns *1 ns ns ns ns ns ns ns ns ns 
BGA ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns 
SEA ns ns ns ns ** ns ns ns ns ns ns ns ns ns ns ns 
GUB ns ns ns ns ns ns ns ns ns ** ns ns ns ns ns ns 
TSW ns ns ns ns **1 ns ns *1 * ns ns ns **1 ns ns ** 
ALP ns ns ns ns ns ns ** ns ns ns ** ns ns ns *1 ns 
ARD ns **2 ns ns ns ns ns ns ns *1 ns ns ns ns ns ns 
BAL ns ns ns ns ns *1 * ns ns ns ns ns ns ns ns ns 
HAI ns ns ns ns ns ns ns ns *1 ns ns ns ns *1 *1 * 
MON ns ns ns ns ns ns ns ns **1 ns ns ns ns ns ns * 
All population ns ns ns ns **1 ns * **1 **1 ns *1 ns ns ns *1 ns 
ns = not significant;  * = P < 0.05;  ** = P < 0.01;  1 = heterozygote deficit;  2 = heterozygote excess.  
ABG ? Abergalle, ABA ? Arsi-Bale, AFA - Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF? Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, 
SES ? Short-eared Somali, WGJ ? Woyto-Guji, BGA ? Boran Galla, SEA ? Small East African, GUB? Guinea Bissau, TSW? Tswana, ALP ? Italian Alpine, ARD - Ardi, BAL ? Baladie, HAI - 
Hair, MON ? Mongolia. 
 83
 4.1.2. Linkage disequilibrium 
The linkage disequilibrium tests within each of the 20 populations were performed 
for the 105 pair-wise combinations of the 15 loci.  Assuming a 5% type I error rate (P < 
0.05), about four to five pairs of loci are expected to be in linkage disequilibrium.  On 
average 6% of the locus pairs were in linkage disequilibrium within the populations.  A 
large number of locus pairs that were in linkage disequilibrium (P < 0.05) were observed in 
the Arsi-Bale and Guinea Bissau (11 pairs), followed by the Long-eared Somali, Italian 
Alpine and Mongolia (nine pairs) populations, Boran-Galla and Short-eared Somali (eight 
pairs) populations, Afar, Small-East Africa and Woyto-Guji (seven pairs) populations, 
Central Highland (six pairs) population, Hair, Gumez and Keffa (five pairs) populations, 
Abergalle, Hararghe Highland and North west Highland (four pairs) populations, Ardi and 
Baladie (three pairs) populations, and Tswana population which had only one out of 105 
pairs of loci in linkage disequilibrium (Appendix 1). 
The global exact test for linkage disequilibrium across all 20 populations (105 pair-
 wise comparisons) yielded significant deviations in the following comparisons: 
BM1818/ILSTS11 (P < 0.05); BM1818/ILSTS44 (P < 0.05); BM1818/MAF209 (P < 
0.05); BM1818/OarAE129 (P < 0.05); BM1494/ILSTS11 (P < 0.05), BM1494/ OarAE129 
(P < 0.05); ILSTS05/SRCRSP03 (P < 0.01); ILSTS87/INRA132 (P < 0.05); 
MAF209/OarFCB304 (P < 0.01) (Appendix 2).  Only one out of these nine tests remained 
significant following application of sequential Bonferroni corrections. 
 
4.2.3. Inference of the recent bottleneck within populations 
Many natural populations are suffering bottlenecks (i.e. severe reductions in size) as 
a result of habitat fragmentation and isolation.  Understanding the effect of population 
bottlenecks on genetic variation has become increasingly important in population genetics, 
speciation theory, and conservation biology.  In order to identify conservation objectives 
properly, it is important to identify populations that have lost genetic variability recently, as 
they may be more susceptible to demographic stochasticity (Lande, 1988; Mills and 
Smouse, 1994).  Unfortunately, it is often difficult to determine if a population has recently 
experienced a bottleneck since historical population size and levels of genetic variation are 
 84
 seldom known.  Identifying recently bottlenecked populations (populations severely 
reduced in size) is important as bottlenecks can increase demographic stochasticity, rate of 
inbreeding, loss of genetic variation, and fixation of deleterious alleles and, thereby reduce 
adaptive potential and increase the probability of population extinction (Luikart and 
Cornuet, 1998).  Cornuet and Luikart (1996) developed a statistical test (a sign test for 
heterozygosity excess) to detect recent historical bottlenecks using allele frequency data.  
The test requires no data on historical population size or levels of genetic variation; it 
requires only measurement of allele frequencies from 5 to 20 polymorphic loci in a sample 
of approximately 20 ? 30 individuals.  The test has reasonable statistical power when 
applied to allele frequency data sets generated by computer simulations (Cournet and 
Luikart, 1996). 
In natural populations, allele number and heterozygosity at selectively neutral loci 
result from an equilibrium between mutation and genetic drift.  The heterozygosity 
expected at a locus in an equilibrium population (Heq) can be calculated from the number 
of alleles observed and the sample size of individuals, assuming neutrality and mutation-
 drift equilibrium.  In non-bottlenecked population defined as not has been recently 
bottlenecked and is therefore likely to be near mutation-drift equilibrium, the expected 
heterozygosity (Heq) will equal the measured Hardy-Weinberg equilibrium heterozygosity 
(HE).  However, if a population has suffered a recent bottleneck, the mutation-drift 
equilibrium is transiently disrupted and the heterozygosity measured at a locus (HE) will 
exceed the heterozygosity (Heq) computed from the number of alleles? sampled (Luikart 
and Cornuet, 1998).  
Bottlenecks generate ?heterozygosity excess? as alleles are generally lost faster than 
heterozygosity during a bottleneck.  During bottleneck rare alleles are lost rapidly since 
they have little effect on heterozygosity (Hedrick et al., 1986).  Thus, many alleles can be 
lost without much reduction in heterozygosity in a bottlenecked population.  The 
bottleneck-induced heterozygosity excess is transient and is likely to be detected only for a 
short time, approximately 0.02 ? 4.0 Ne generations (Ne is the bottleneck effective size), 
until a new equilibrium between mutation and drift is reached at the new Ne (Cornuet and 
Luikart, 1996).  Thus, only bottlenecks that have occurred in the recent past (less than 4 Ne 
generations ago) are likely to be detectable by the sign test for heterozygosity excess.  This 
 85
 window of time is approximate and depends not only on Ne but also on factors such as the 
mutation rate and mutation model of the loci sampled (Cornuet and Luikart, 1996). 
In a non-bottlenecked, equilibrium population, approximately 50% of the loci 
sampled are expected to have a slight excess of heterozygosity (He > Heq), and 50% will 
have a slight deficiency of heterozygosity (He < Heq), resulting from genetic drift and 
sampling error.  However, in recently bottlenecked populations, a majority of loci will 
exhibit an excess of heterozygosity (Luikart and Cornuet, 1998).  In order to determine if a 
population exhibits a significant number of loci with heterozygosity excess, Piry et al. 
(2004) proposed three statistical tests:  Sign test, a standardized differences test (Cornuet 
and Luikart, 1996; Luikart and Cornuet, 1998), and a Wilcoxon?s signed rank test (Luikart 
et al., 1997).  They also proposed a graphical descriptor of the shape of the allele frequency 
distribution (?mode-shift? indicator), which can differentiate between bottlenecked and 
stable populations (Luikart et al., 1998).  The Wilcoxon?s test (although this test is 
analogous to the sign test) is generally the most useful of all the tests because it is the most 
powerful (along with the standardized differences test) and robust (like the sign test) when 
used with few (< 20) polymorphic loci.  In testing for bottlenecks, the null hypothesis of 
Wilcoxon?s test is no significant heterozygosity excess (on average across loci).  Thus the 
alternative hypothesis is significant heterozygosity excess (and thus evidence of a recent 
bottleneck).  This is a one tailed test that requires at least four polymorphic loci to have 
many possibility of obtaining a significant (P < 0.05) test result.  The sign test (Cornuet and 
Luikart, 1996) determines if a significant majority of loci in a population have a 
heterozygosity excess, and thus if a population appears to have been recently bottlenecked. 
Test for heterozygosity excess should not be confused with test for Hardy-Weinberg 
proportions.  Tests for Hardy-Weinberg proportions compare the observed proportion of 
heterozygosity (HO) to the heterozygosity expected (HE) when a population is in Hardy-
 Weinberg proportions.  The test for Heterozygosity excess compares HE to the 
heterozygosity (Heq) expected at mutation drift equilibrium in a sample that has the same 
size and the same number of alleles as the sample used to measure heterozygosity expected 
(HE). 
Alternatively, a quantitative graphical method (?mode-shift? indicator) can be used 
for detecting a bottleneck-induced distortion of allele frequency distributions (Luikart et al., 
 86
 1998).  This method involves comparing the distribution of allele frequencies observed in a 
population suspected to have been bottlenecked to the distribution expected in a non-
 bottlenecked population.  The graphical method consists of grouping alleles from a sample 
of many polymorphic loci (at least five loci) in to each of 10 allele frequency classes and 
then plotting a frequency histogram.  The graphical method concludes that a population has 
been recently bottlenecked if fewer alleles are found in the low frequency classes than in 
one or more intermediate frequency classes. 
The recent bottleneck effect was inferred for each population using a Wilcoxon?s 
signed rank test (Cornuet and Luikart, 1996), and the results are presented in Table 6 and 
Figure 20.  The calculations were implemented in the computer program BOTTLENECK 
version 1.2.02 (Cornuet and Luikart, 1996; http://www.montpellier.inra.fr/URLB/ 
bottleneck) based on 10,000 simulation replicates assuming that microsatellite alleles are 
evolved under the stepwise mutation model. Generally, there was no recent bottleneck 
effect detected in any population for both Wilcoxon?s signed rank test (Table 6) and 
quantitative graphical method (Figure 20).  Interestingly, ten populations out of the total 20 
populations showed a significant deficiency of heterozygosity.  This suggests that these 
populations are not at mutation-drift equilibrium but instead have experienced a recent 
expansion in population size or perhaps a recent entry of rare alleles from genetically 
distinct immigrants. 
 
 
 
 
 
 
 
 
 
 
 
 
 87
 Table 6. The Wilcoxon test of recent bottleneck effect for each population assuming 
that all the microsatellite alleles are evolved under stepwise mutation model.  The 
estimation is based on 10,000 simulation replicates. 
One-tail P-value 
Population Sample size 
Heterozygote deficit Heterozygote excess 
Abergale 46 0.0535 0.9527 
Arsi-Bale 46 0.12619 0.88535 
Afar 44 0.06769 0.93973 
Central Highland 40 0.0535 0.9527 
Gumez 43 0.00513 0.99582 
Hararghe Highland 40 0.01767 0.98492 
Keffa 38 0.03186 0.97232 
Long-eared Somali 48 0.01279 0.98923 
North-West Highland 45 0.12619 0.88535 
Short-eared Somali 43 0.24435 0.77286 
Woyto-Guji 46 0.11465 0.89612 
Boran-Galla 35 0.01279 0.98923 
Small East Africa 39 0.06027 0.94650 
Guinea Bissau 46 0.04730 0.95837 
Tswana 29 0.00903 0.99246 
Italian Alpine 31 0.10388 0.90619 
Ardi 25 0.00084 0.99934 
Baladie 22 0.01767 0.98492 
Hair 31 0.11465 0.89612 
Mongolia  39 0.01077 0.99097 
 
 88
 Adergalle
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
 Long-Ear Somali
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.7
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
 Tswana
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
  
Arsi-Bale
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
 North-West Highland
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
 Italian Alpine
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
  
Afar
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
 Short-Ear Somali
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
 Ardi
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
  
Central Highland
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
 Woyto-Guji
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
 Baladie
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
  
Gumez
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
 Boran-Galla
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
 Hair
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
  
Hararghe Highland
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
 Small-East Africa
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
 Mongolia
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
  
Keffa
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.7
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
 Guinea Bissau
 0
 0.1
 0.2
 0.3
 0.4
 0.5
 0.6
 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
  
Figure 20. Distributions of allele frequency from 15 microsatellite loci in 20 goat 
populations.  The bars represent the proportion of alleles expected in each of 10 allele 
frequency classes calculated from 10,000 bottleneck simulation replicates. 
 89
 4.3. The extent and distribution of the genetic diversity within the 20 populations 
4.3.1. Mean number of alleles and heterozygosities 
4.3.1.1. Analysis of allelic variation 
Genetic data for a population can be expressed as gene or allele frequency. The 
portion of a particular allele out of all the alleles at a locus observed in all individuals of a 
population is called the allele frequency.  It is a fundamental parameter in the study of 
population genetics, because the genetic change in a population is usually described by the 
changes in allele frequencies.  Microsatellite are co-dominant markers and all alleles are 
identifiable, so the allele frequency can be obtained by simply calculating the proportion of 
a particular allele over all alleles at a specific locus detected in a population.  The allelic 
frequencies were calculated using a computer program the EXCEL MICROSATELLITE 
TOOLKIT version 3.1 (available from http://oscar.gen.tcd.ie/~sdepark/ms-toolkit), and 
allelic frequencies for every locus and population are given in Appendices 3 to 17.  The 
alleles observed at a locus varied in length from two to 60 base pairs.  Figure 21 illustrates 
the allelic ranges and their frequencies detected at each of 15 microsatellite loci in 776 goat 
individuals of 20 populations. 
The number of alleles observed and expected under stepwise mutation model 
(SMM, Ohta and Kimura, 1973) and infinite allele model (IAM, Kimura and Crow, 1964) 
at each locus for each population were calculated using the computer program 
MICROSATELLITE ANALYSER (MSA) version 2.39 (Dieringer and Schl?tterer, 2003).  
In the SMM, an allele at a locus is assumed to mutate to an allele either with gain or loss of 
a repeat with an equal probability, while the IAM assumes that new mutant alleles are 
always different from the existing ones in the populations. 
 
 
 
 
 
 
 
 
 90
 BM1818
 (N=669)
 0.00
 10.00
 20.00
 30.00
 40.00
 50.00
 244 246 250 252 254 256 258 260 262 264 266 268  
BMS1494
 (N=771)
 0.0
 10.0
 20.0
 30.0
 40.0
 50.0
 60.0
 70.0
 244 246 256 264 270 276 280 282 284 288  
ILSTS05
 (N=776)
 0.0
 20.0
 40.0
 60.0
 80.0
 100.0
 178 180 182 184 186 188 190  
ILSTS11
 (N=774)
 0.0
 5.0
 10.0
 15.0
 20.0
 25.0
 30.0
 35.0
 40.0
 261 265 267 269 271 273 275 277 279 281 283 285  
ILSTS44
 (N=772)
 0.0
 10.0
 20.0
 30.0
 40.0
 50.0
 151 153 155 157 159 161 169 173  
ILSTS87
 (N=768)
 0.0
 10.0
 20.0
 30.0
 40.0
 50.0
 60.0
 70.0
 141 143 145 147 149 151 153 155 157 159 161 163  
INRA05
 (N=775)
 0.0
 10.0
 20.0
 30.0
 40.0
 50.0
 60.0
 131 139 141 143 145 149  
INRA63
 (N=775)
 0.0
 10.0
 20.0
 30.0
 40.0
 50.0
 157 161 163 165 167 169 171 173 175  
INRA132
 (N=769)
 0.0
 10.0
 20.0
 30.0
 40.0
 50.0
 60.0
 143 145 147 151 153 157  
MAF35
 (N=773)
 0.0
 20.0
 40.0
 60.0
 80.0
 100.0
 101 105 107 109 115  
MAF65
 (N=753)
 0.0
 5.0
 10.0
 15.0
 20.0
 25.0
 114 116 118 120 122 124 126 128 130 132 134 136 138 140 144 148  
MAF209
 (N=775)
 0.0
 10.0
 20.0
 30.0
 40.0
 50.0
 60.0
 70.0
 102 104 106 108  
OarAE129
 (N=751)
 0.0
 5.0
 10.0
 15.0
 20.0
 25.0
 30.0
 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 173 175 177  
OarFCB304
 (N=776)
 0.0
 10.0
 20.0
 30.0
 40.0
 50.0
 60.0
 12
 0
 12
 8
 13
 4
 13
 8
 14
 2
 14
 6
 15
 0
 15
 6
 16
 0
 16
 4
 16
 8
 17
 4
 18
 0
  
SRCRSP03
 (N=774)
 0
 10
 20
 30
 40
 50
 60
 111 115 117 119 121 123 125  
Figure 21. Histogram showing the allele size range and their frequency detected in 776 
goat individuals of 20 populations at 15 microsatellite loci. The horizontal axis 
represents the allele size (base pairs) and the vertical axis their frequency in percentage. 
 91
  
The observed number of alleles, mean number of alleles expected under SMM 
(MNASMM) and IAM (MNAIAM) are shown in Table 7.  The number of observed alleles is 
highly dependent on sample size, and thus comparison of this quantity between different 
populations is not meaningful unless the sample size is more or less the same.  The sample 
size dependence occurs as there are many low-frequency alleles in natural populations, and 
the number of observed alleles increases with the increase of the sample size.  The mean 
number of alleles per locus (MNA) is the average number of alleles observed in a 
population over all loci genotyped. It is an indicator of genetic variability due to allelic 
diversity with the conditions that the populations under investigation are at mutation-drift 
equilibrium and that sample size is more or less the same for each population.  To minimize 
the variation of MNA among populations affected by the sample size, we calculated the 
MNA for a random sample of 20 animals after 250 re-samplings with replacement for every 
population.  
A total of 159 alleles were detected at 15 microsatellite loci in 776 goat individuals 
of 20 populations.  Allele frequencies across all loci ranged from 1.04% to 98.91% 
(Appendix 3 to 17).  Ten of these alleles were specific to Ethiopian populations, being 
observed in a minimum of two of the 11 Ethiopian populations analysed (Table 8).  Private 
alleles meaning alleles that were unique to a single or observed in only one population were 
observed in the following Ethiopian populations: Arsi-Bale (one), Afar (one), Central 
Highland (two), Keffa (two), Long-eared Somali (two) and North-West Highland (two). 
However, the frequencies of these private alleles were less than 0.05 in all cases.  In 
addition to these, nine alleles were specific to non-Ethiopian reference populations, being 
observed in a minimum of two of the nine non-Ethiopian reference populations analysed 
(Table 8).  Private alleles, that were unique to a single population were observed in the 
following non-Ethiopian reference populations: Italian Alpine (one), Ardi (five), Boran-
 Galla (one), Small-East Africa (one), Guinea Bissau (one), Tswana (one), Baladie (two), 
Hair (two) and Mongolia (three).  However, the frequencies of these private alleles were 
also less than 0.05 in all cases. 
The most polymorphic marker was OarFCB304 with 25 alleles, while the least 
polymorphic marker is MAF209 with four alleles in the pooled population.  The total 
 92
 number of alleles sampled in a population across all loci ranged between 80 in Guinea 
Bissau population from West Africa and 116 in Long-eared Somali population from 
Ethiopia.  The MNAOBS across all populations was 5.46 ? 0.06, varying between 4.67 ? 
2.74 again in Guinea Bissau population and 6.47 ? 4.03 again in Long-eared Somali.  The 
MNAOBS for uniformed sample size with 20 animals was 4.88 ? 0.08, varying between 4.19 
? 2.26 again in Guinea Bissau and 5.74 ? 2.89 in Hair populations.  In general, the mean 
numbers of alleles were comparable among all the populations from the different 
geographic regions. 
The expected MNA across all populations was 4.60 ? 0.08 and 7.42 ? 0.16 for 
SMM and IAM, respectively, ranging between 4.05 ? 1.59 in Guinea Bissau and 5.77 ? 
2.70 in Hair for SMM, and between 6.39 ? 2.34 in Ardi and 9.29 ? 4.85 in Hair populations 
for IAM.  The results indicate that the stepwise mutation model (SMM) is more appropriate 
for this particular dataset and should be preferred in the model-based algorithms such as the 
inference of the recent population bottleneck. 
The expected heterozygosity (HE) and observed heterozygosity (HO) values 
averaged over loci showed an overall pattern similar to that observed for MNA per locus 
(Table 9).  The highest value for HE, around 64% was observed within the Hair (Turkey) 
population, while the lowest value around 52% was again for Guinea Bissau.  The highest 
value for HO, around 63% was observed within the Ardi population, while the lowest value 
around 52% was observed for Afar, Boran Galla and Tswana when all individuals in the 
populations are considered in both cases.  In all cases average observed heterozygosities 
were lower than that expected under HWE except for Guinea Bissau and Ardi populations. 
The null alleles were tested for every population at each locus using Brookfield null 
allele estimator 2 (Brookfield, 1996) which assumes that a proportion of the sample are not 
amplified because these are in fact null allele homozygotes.  The test, implemented in the 
computer program Micro-Checker version 2.2.1 (available from 
http://www.microchecker.hull.ac.uk), were carried out with 1000 Monte Carlo bootstrap 
simulations over the observed data to calculate probabilities and 99% confidence interval 
for both homozygote frequencies and heterozygote size differences.  Three loci (BMC1222, 
BMS357 and SRCRSP07) showed evidence for a null allele in 17, 18 and 14 populations, 
 93
 respectively.  Neither a single locus nor a single population showed evidence for null allele 
across all populations or across all loci. 
 94
 Table 7. Observed number of alleles, mean number of alleles expected under SMM and IAM for each population and each locus. 
BM1818 BMC1222 BMS1494 BMS357 ILSTS05 ILSTS11 ILSTS44 
Pop 
Obs SMM IAM Obs SMM IAM Obs SMM IAM Obs SMM IAM Obs SMM IAM Obs SMM IAM Obs SMM IAM 
ABG 8 5.2 9.3 8 4.2 6.7 4 3.7 5.6 8 5.4 9.8 3 1.6 1.7 7 5.2 9.2 4 4.7 8.1 
ABA 9 6.5 12.2 6 4.4 7.3 4 3.7 5.6 7 5.8 10.2 3 2.7 3.4 6 5.0 8.8 4 4.4 7.1 
AFA 10 3.9 5.8 7 5.6 10.1 4 3.2 4.3 9 6.8 12.7 5 2.9 3.8 5 4.4 7.2 4 5.0 8.6 
CEH  9 5.2 9.1 5 4.8 8.1 5 3.7 5.3 7 6.2 11.0 2 1.5 1.5 6 4.7 7.7 4 5.0 8.6 
GUM 8 6.4 11.8 5 4.5 7.4 4 2.5 2.9 9 7.6 13.8 3 1.8 2.0 5 4.4 7.0 5 4.5 7.5 
HAH 8 4.4 7.1 5 4.8 7.9 4 3.7 5.4 9 8.0 14.6 2 2.1 2.3 6 5.0 8.5 4 5.1 8.7 
KEF 7 3.6 5.2 4 2.9 3.7 5 4.2 6.5 9 4.8 8.0 2 2.5 3.0 7 5.0 8.6 4 4.8 8.0 
LES 12 5.7 10.7 7 4.9 8.6 5 3.4 4.8 8 6.1 11.5 3 2.7 3.3 6 5.0 8.8 5 4.7 8.0 
NWH 9 5.6 10.2 5 4.7 7.9 4 3.7 5.4 8 5.5 9.5 2 1.7 1.8 6 4.4 7.3 4 4.9 8.5 
SES 10 5.8 10.4 6 6.0 10.9 3 3.5 5.1 9 7.2 13.1 3 1.8 2.0 7 5.2 9.2 4 5.2 9.0 
WGJ 10 4.9 8.5 6 4.7 8.0 4 3.9 6.1 9 6.3 11.8 3 2.3 2.7 5 5.1 9.1 4 4.6 7.8 
BGA 10 6.2 10.9 6 3.6 5.2 3 3.3 4.4 9 5.5 9.1 2 1.5 1.6 4 4.6 7.4 4 4.7 7.5 
SEA 7 5.0 8.5 5 3.0 3.9 4 3.3 4.6 9 4.9 8.0 2 1.8 1.9 6 4.9 8.2 5 5.1 8.7 
GUB 9 6.7 12.8 3 3.1 4.1 4 3.7 5.6 4 4.1 6.2 2 2.2 2.5 4 4.1 6.4 3 2.3 2.6 
TSW 8 5.2 8.1 6 6.4 10.0 5 4.5 7.0 8 5.3 8.5 3 4.7 7.5 7 7.0 11.7 5 2.8 3.5 
ALP 6 6.0 10.2 5 4.0 5.8 4 3.8 5.4 8 4.3 6.3 3 4.0 5.9 5 5.7 9.5 5 2.6 3.1 
ARD 6 4.8 7.3 5 5.2 8.1 3 3.5 4.7 11 6.5 10.5 4 3.0 3.7 6 4.5 6.9 3 3.0 3.9 
BAL 9 7.4 11.6 5 5.5 8.4 3 3.6 5.0 11 9.3 13.9 4 2.9 3.6 6 5.8 9.1 4 3.3 4.2 
HAI 9 8.6 14.8 7 5.9 10.0 4 4.9 7.9 11 12.2 19.0 5 3.7 5.2 8 6.6 11.4 4 3.2 4.2 
MON 8 8.7 15.8 6 5.6 9.9 4 3.5 4.9 11 8.4 15.1 5 4.0 6.1 6 5.7 10.2 5 2.4 2.9 
Mean 8.6 5.8 10.0 5.6 4.7 7.6 4.0 3.7 5.3 8.7 6.5 11.1 3.1 2.6 3.3 5.9 5.1 8.6 4.2 4.1 6.5 
SD 1.5 1.4 2.7 1.1 1.0 2.1 0.6 0.5 1.0 1.7 1.9 3.2 1.1 0.9 1.7 1.0 0.7 1.4 0.6 1.0 2.4 
 95
 Cont. 
ILSTS87 INRA05 INRA63 INRA132 MAF035 MAF065 MAF209 
Pop 
Obs SMM IAM Obs SMM IAM Obs SMM IAM Obs SMM IAM Obs SMM IAM Obs SMM IAM Obs SMM IAM 
ABG 6 2.4 2.9 4 3.8 5.7 5 4.5 7.5 3 3.5 5.1 2 1.9 2.0 8 6.8 13.0 3 4.0 6.1 
ABA 6 3.4 4.9 4 4.4 7.3 5 4.1 6.4 3 3.7 5.6 3 2.1 2.4 9 9.6 18.3 3 4.4 7.3 
AFA 5 2.5 3.0 4 3.8 5.7 4 4.5 7.4 3 3.6 5.3 2 1.6 1.7 9 9.1 17.3 3 3.3 4.6 
CEH  5 3.4 4.7 3 3.9 5.9 5 4.0 6.0 3 3.5 5.0 2 1.6 1.6 8 9.0 16.3 3 4.0 6.2 
GUM 6 2.6 3.2 4 4.0 6.2 5 4.7 7.9 3 4.4 7.1 2 1.6 1.7 11 7.8 14.3 3 3.4 4.9 
HAH 6 3.0 3.8 4 4.4 7.0 6 4.3 7.0 3 3.7 5.4 3 1.6 1.6 11 10.7 19.4 4 3.9 5.9 
KEF 5 2.8 3.5 4 4.5 7.4 8 5.2 8.9 3 3.6 5.2 2 1.5 1.5 9 8.4 15.3 4 4.4 7.0 
LES 5 3.3 4.5 4 4.1 6.6 6 5.0 8.9 3 3.7 5.5 2 2.2 2.6 10 8.8 17.2 3 3.9 6.1 
NWH 7 3.0 3.9 4 4.2 6.6 5 4.7 7.9 2 3.5 5.0 2 1.9 2.1 9 8.2 15.7 3 3.7 5.6 
SES 7 3.5 4.9 4 4.2 6.6 5 5.0 8.6 2 3.6 5.1 2 1.7 1.8 9 9.2 17.5 3 3.4 4.7 
WGJ 7 2.3 2.6 4 4.0 6.2 5 4.3 7.0 3 3.8 5.7 2 1.9 2.1 10 9.1 17.6 3 4.4 7.3 
BGA 5 2.6 3.2 4 3.6 5.1 5 4.0 6.0 3 3.5 4.9 2 2.0 2.2 10 7.9 14.0 3 3.3 4.4 
SEA 5 2.2 2.6 4 3.0 3.8 5 5.6 9.8 3 3.4 4.7 2 1.6 1.7 10 8.8 16.0 3 4.3 6.7 
GUB 4 2.5 3.0 3 3.9 5.9 4 3.8 5.7 3 4.4 7.3 2 1.1 1.1 8 5.7 10.4 3 3.5 4.9 
TSW 6 3.3 4.4 4 4.1 6.1 5 5.4 8.8 5 3.8 5.5 3 1.9 2.1 8 5.8 9.6 3 3.5 4.9 
ALP 6 5.7 9.6 4 3.5 4.8 6 6.4 11.0 2 3.5 4.8 4 1.5 1.5 6 5.4 8.8 3 2.0 2.1 
ARD 6 4.4 6.7 4 4.3 6.4 5 4.4 6.5 4 3.5 4.8 2 2.0 2.1 9 7.2 11.5 3 3.7 5.1 
BAL 7 5.4 8.2 5 4.1 5.8 5 5.5 8.4 4 3.3 4.3 2 1.7 1.8 8 7.6 11.8 2 3.1 4.0 
HAI 10 8.4 14.4 4 4.5 7.0 5 6.4 10.9 2 3.1 4.1 2 1.7 1.8 11 9.2 15.6 3 2.7 3.4 
MON 8 5.8 10.2 3 3.0 4.0 6 5.2 8.9 4 3.4 4.7 2 1.7 1.8 8 5.7 9.9 2 2.6 3.1 
Mean 6.1 3.6 5.2 3.9 4.0 6.0 5.3 4.8 8.0 3.1 3.6 5.2 2.3 1.7 1.9 9.1 8.0 14.5 3.0 3.6 5.2 
SD 1.3 1.6 3.1 0.4 0.4 1.0 0.9 0.7 1.5 0.8 0.3 0.8 0.6 0.3 0.3 1.3 1.5 3.2 0.5 0.6 1.4 
 96
 cont. 
OarAE129 OarFCB304 SRCRSP03 SRCRSP07 Mean number of alleles per locus (SD) 
Pop 
Obs SMM IAM Obs SMM IAM Obs SMM IAM Obs SMM IAM 
Total 
OBS(all) OBS(20) SMM IAM 
ABG 8 6.9 13.2 11 5.8 10.8 4 5.7 10.4 3 3.8 5.7 99 5.33 (2.55) 4.65 (2.09) 4.40 (1.50) 7.37 (3.40) 
ABA 7 6.3 11.9 10 5.4 9.7 4 4.8 8.1 6 4.3 7.1 99 5.33 (2.41) 4.87 (2.28) 4.73 (1.64) 7.97 (3.65) 
AFA 11 9 17.1 9 5.2 9.2 4 4.6 7.6 5 4.7 7.9 103 5.47 (2.83) 4.77 (2.34) 4.64 (1.99) 7.74 (4.36) 
CEH  9 6.6 11.4 9 5.5 9.7 4 5.1 8.8 4 4.1 6.3 93 5.13 (2.53) 4.65 (2.15) 4.54 (1.75) 7.39 (3.53) 
GUM 11 8.2 15.6 9 4.3 7 4 4.4 7.2 5 4 6.2 102 5.53 (2.90) 4.71 (2.31) 4.51 (1.91) 7.42 (4.08) 
HAH 9 6.9 12.5 13 5.5 9.7 6 4.6 7.5 4 4 6 107 5.93 (3.13) 5.1 (2.53) 4.75 (2.11) 7.80 (4.29) 
KEF 12 7.4 13.5 8 5 8.5 4 4.1 6.2 5 5.4 9.3 102 5.60 (2.82) 5.05 (2.37) 4.45 (1.65) 7.17 (3.45) 
LES 14 8.3 16.2 14 4.2 6.7 5 4.5 7.5 4 4.3 7.1 116 6.47 (4.03) 5.28 (2.90) 4.72 (1.71) 8.02 (3.92) 
NWH 12 7.6 14.6 8 5 8.8 4 4.9 8.5 4 4.5 7.5 98 5.40 (3.02) 4.67 (2.26) 4.54 (1.64) 7.60 (3.62) 
SES 10 8.3 15.7 10 4.3 7 5 4.8 8.3 5 4.1 6.5 104 5.60 (3.00) 4.9 (2.41) 4.82 (1.98) 8.13 (4.26) 
WGJ 11 8 15.4 10 5.6 10.1 4 4.2 6.6 4 4.2 6.7 104 5.67 (3.09) 4.99 (2.57) 4.64 (1.81) 7.85 (4.04) 
BGA 10 8 13.6 8 4 6 5 4.7 7.6 4 3.8 5.5 97 5.20 (2.88) 4.63 (2.48) 4.26 (1.76) 6.59 (3.48) 
SEA 7 6.4 11.4 8 5.1 8.8 4 4.4 7.1 4 4.1 6.3 93 5.00 (2.27) 4.64 (2.10) 4.26 (1.74) 6.81 (3.63) 
GUB 7 5.6 10.1 11 7.4 14.3 3 4.2 6.8 3 4.6 7.8 80 4.67 (2.74) 4.19 (2.26) 4.05 (1.59) 6.53 (3.54) 
TSW 6 3.9 5.6 7 5.5 9.1 4 2.5 3 4 3.1 4.1 97 5.27 (1.71) 4.89 (1.63) 4.39 (1.40) 6.63 (2.71) 
ALP 7 5.4 8.9 6 5.1 8.4 4 3.2 4.3 3 4.7 7.5 87 4.73 (1.44) 4.4 (1.46) 4.25 (1.41) 6.54 (2.83) 
ARD 8 5.4 8.6 12 3.8 5.3 4 4 5.8 5 4.7 7.3 100 5.27 (2.69) 4.97 (2.41) 4.32 (1.25) 6.39 (2.34) 
BAL 8 6.3 9.8 12 8.8 13.4 4 3.9 5.5 4 4.1 5.9 103 5.53 (2.80) 5.45 (2.75) 5.08 (2.13) 7.48 (3.58) 
HAI 9 7 12 11 7.3 12.5 5 3.9 5.7 5 4.6 7.3 115 6.13 (3.20) 5.74 (2.89) 5.77 (2.70) 9.29 (4.85) 
MON 9 8.9 16.2 13 7.2 13.3 5 3.7 5.3 5 3 4 110 5.87 (2.95) 5.09 (2.51) 4.92 (2.25) 8.12 (4.70) 
Mean 9.3 7 12.6 10 5.5 9.4 4.3 4.3 6.9 4.3 4.2 6.6 100.5 5.46 4.88 4.60 7.42 
SD 2.1 1.3 3 2.2 1.3 2.5 0.7 0.7 1.7 0.8 0.5 1.3 8.5 0.09s.e. 0.08s.e. 0.08s.e. 0.16s.e. 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, LES ? Long-Ear Somali, NWH  ? North-West Highland, SES ? 
Small-Ear Somali, WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? Baladie, HAI- Hair, MON? 
Mongolia. 
 97
 Table 8. List of specific alleles observed only in Ethiopian or reference populations. 
Locus Allele size Type Populations observed in 
BM1818 246 Ethiopian specific Afar and Long-eared Somali 
BMS1494 288 Ethiopian specific Keffa and Long-eared Somali 
ILSTS05 186 Ethiopian specific Afar and Short-eared Somali 
ILSTS87 163 Ethiopian specific ABG, North-West Highland and Woyto-Guji 
INRA63 157 Ethiopian specific Gumez and Keffa 
MAF209 108 Ethiopian specific Hararghe Highland and Keffa 
OarAE129 145 Ethiopian specific Long-eared Somali and North West Highland 
OarAE129 175 Ethiopian specific ABG, CEH, Gumez, Keffa, LES, NWH, SES and WGuji 
OarFCB304 124 Ethiopian specific ABG, AFA, CEH, HAH and NWH 
OarFCB304 174 Ethiopian specific Long-eared Somali and Woyto-Guji 
ILSTS05 188 Reference population specific Hair and Mongolia 
ILSTS05 190 Reference population specific Ardi, Baladie and Mongolia 
ILSTS44 169 Reference population specific Tswana, Alpine, Ardi, Baladie, Hair and Mongolia 
ILSTS87 141 Reference population specific Hair and Mongolia 
ILSTS87 143 Reference population specific Ardi, Baladie, Hair and Mongolia 
ILSTS87 153 Reference population specific SEA, GUB Tswana Alpine, Baladie, Hair and Mongolia 
ILSTS87 157 Reference population specific Ardi and Hair 
INRA132 157 Reference population specific Tswana, Baladie and Mongolia 
MAF65 140 Reference population specific Guinea Bissau and Hair 
 
 98
 Table 9. Expected heterozygosity (HE), observed heterozygosity (HO) and mean 
number of alleles (MNA) of all 20 populations. 
Population 
Sample 
size 
MNA 
(All ind) 
MNA  
(20 ind) 
HE ? s.e. HO ? s.e. 
Abergalle 46 5.33 4.65 0.57 ? 0.056 0.54 ? 0.053 
Arsi-Bale 46 5.33 4.87 0.61 ? 0.043 0.58 ? 0.045 
Afar 44 5.47 4.77 0.56 ? 0.051 0.52 ? 0.049 
Central Highland  40 5.13 4.65 0.57 ? 0.056 0.53 ? 0.058 
Gumez 43 5.53 4.71 0.55 ? 0.057 0.55 ? 0.054 
Hararghe Highland 40 5.93 5.10 0.58 ? 0.052 0.55 ? 0.053 
Keffa 38 5.60 5.05 0.58 ? 0.050 0.56 ? 0.047 
Long-eared Somali 48 6.47 5.28 0.59 ? 0.044 0.56 ? 0.044 
North West Highland 45 5.40 4.67 0.57 ? 0.053 0.54 ? 0.049 
Small-eared Somali 43 5.60 4.90 0.58 ? 0.054 0.57 ? 0.055 
Woyto-Guji 46 5.67 4.99 0.58 ? 0.052 0.54 ? 0.049 
Ethiopian populations  5.59 4.88 0.58 ? 0.015 0.55 ? 0.015 
Boran-Galla 35 5.20 4.63 0.55 ? 0.055 0.52 ? 0.049 
Small East Africa 39 5.00 4.64 0.55 ? 0.058 0.53 ? 0.055 
Guinea Bissau 46 4.67 4.19 0.52 ? 0.058 0.53 ? 0.057 
Tswana 29 5.27 4.89 0.58 ? 0.045 0.52 ? 0.047 
Italian Alpine 31 4.73 4.40 0.56 ? 0.055 0.53 ?0.057 
Ardi 25 5.27 4.97 0.58 ? 0.037 0.63 ?0.050 
Baladie 22 5.53 5.45 0.62 ? 0.050 0.61 ? 0.055 
Hair 31 6.13 5.74 0.64 ? 0.055 0.62 ? 0.061 
Mongolia 39 5.87 5.09 0.58 ? 0.057 0.57 ? 0.061 
Reference population  5.30 4.89 0.58 ? 0.017 0.561 ? 0.018 
Mean of all populations  5.46 ? 0.158 4.88 ? 0.08 0.58 ? 0.011 0.56 ? 0.012 
 
 
 
 
 
 
 
 
 
 
 99
 4.4. Genetic relationship between the populations 
Two measures of genetic distances, Nei?s standard genetic distance (DS) and 
Nei?s DA genetic distance were used to estimate the genetic relationship for each pair of 
populations. 
 
4.4.1. Nei?s standard genetic distance (DS) 
The Nei?s standard genetic distance (DS) values for the 20 populations are 
shown in Table 10.  The smallest DS values were between Kenyan Boran Galla and 
Ethiopian Short-eared Somali (0.000 ? 0.002).  The next smallest DS values were 
observed between Ethiopian Short-eared Somali & Hararghe Highland and Hair (from 
Turkey) & Baladie (from Egypt) (0.001 ? 0.004 and 0.002 ? 0.005), respectively, while 
the Highest DS was observed between Mongolia and Guinea Bissau populations (0.427 
? 0.152). 
 
4.4.2. Nei?s DA genetic distances  
When the genetic distances of the populations in the study were determined 
using Nei?s DA index of genetic distance, the smallest DA was observed between 
Ethiopian Short-eared Somali and Hararghe Highland (0.024), the next smallest DA was 
between Ethiopian North-West Highland & Central Highland (0.026), followed by the 
next smallest DA between Boran-Galla of Kenya and Short-eared Somali of Ethiopia 
(0.028), while the highest DA was observed between Mongolia and Guinea Bissau 
populations (0.310) as in the case for the Nei?s DS genetic distance (Table 11). 
 
 100
  
Table 10. Nei?s standard genetic distance (DS) for 11 Ethiopian and 9 reference populations/breeds genotyped, using 15 
microsatellite markers (DS below the diagonal, standard errors above the diagonal). 
 ABG ABA AFA CEH GUM HAH KEF LES NWH SES WGJ BGA SEA GUB TSW ALP ARD BAL HAI MON 
ABG - 0.013 0.018 0.008 0.014 0.014 0.016 0.014 0.007 0.013 0.014 0.019 0.022 0.078 0.068 0.088 0.060 0.038 0.049 0.097 
ABA 0.034 - 0.018 0.006 0.009 0.007 0.011 0.005 0.006 0.007 0.007 0.013 0.017 0.080 0.046 0.071 0.044 0.025 0.036 0.073 
AFA 0.055 0.047 - 0.014 0.016 0.006 0.025 0.011 0.019 0.005 0.014 0.009 0.026 0.080 0.046 0.075 0.033 0.022 0.033 0.085 
CEH 0.010 0.019 0.035 - 0.012 0.005 0.007 0.008 0.004 0.006 0.013 0.014 0.026 0.085 0.056 0.066 0.047 0.028 0.035 0.075 
GUM 0.031 0.031 0.046 0.030 - 0.015 0.013 0.005 0.009 0.009 0.008 0.011 0.026 0.086 0.059 0.070 0.043 0.027 0.041 0.080 
HAH 0.031 0.014 0.010 0.008 0.033 - 0.010 0.005 0.008 0.004 0.012 0.008 0.026 0.077 0.043 0.067 0.037 0.023 0.029 0.074 
KEF 0.036 0.023 0.046 0.022 0.036 0.021 - 0.009 0.007 0.012 0.011 0.022 0.022 0.071 0.040 0.076 0.057 0.029 0.040 0.091 
LES 0.037 0.012 0.026 0.026 0.019 0.011 0.030 - 0.006 0.003 0.007 0.005 0.014 0.081 0.040 0.061 0.037 0.019 0.023 0.066 
NWH 0.014 0.011 0.035 0.004 0.022 0.012 0.012 0.016 - 0.007 0.008 0.015 0.019 0.085 0.046 0.068 0.050 0.025 0.032 0.074 
SES 0.032 0.019 0.014 0.017 0.024 0.001 0.030 0.005 0.007 - 0.006 0.002 0.019 0.078 0.042 0.059 0.037 0.022 0.023 0.062 
WGJ 0.034 0.012 0.033 0.036 0.021 0.020 0.022 0.011 0.019 0.015 - 0.006 0.008 0.063 0.042 0.080 0.046 0.025 0.035 0.090 
BGA 0.049 0.034 0.020 0.041 0.031 0.015 0.052 0.011 0.025 0.000 0.013 - 0.014 0.073 0.048 0.067 0.044 0.025 0.028 0.070 
SEA 0.057 0.055 0.055 0.065 0.058 0.053 0.056 0.037 0.045 0.044 0.020 0.034 - 0.058 0.053 0.086 0.059 0.041 0.047 0.093 
GUB 0.220 0.211 0.201 0.222 0.202 0.200 0.211 0.210 0.224 0.198 0.163 0.178 0.159 - 0.124 0.138 0.091 0.088 0.102 0.152 
TSW 0.212 0.176 0.179 0.193 0.206 0.152 0.164 0.169 0.167 0.152 0.163 0.177 0.203 0.358 - 0.059 0.073 0.043 0.033 0.072 
ALP 0.232 0.196 0.190 0.196 0.184 0.165 0.205 0.154 0.173 0.140 0.189 0.159 0.220 0.353 0.122 - 0.049 0.043 0.031 0.039 
ARD 0.178 0.133 0.106 0.128 0.137 0.102 0.158 0.106 0.141 0.109 0.152 0.132 0.190 0.251 0.206 0.139 - 0.036 0.026 0.069 
BAL 0.095 0.079 0.080 0.077 0.080 0.063 0.084 0.062 0.068 0.057 0.078 0.076 0.122 0.235 0.108 0.106 0.079 - 0.005 0.034 
HAI 0.148 0.123 0.115 0.119 0.143 0.096 0.138 0.096 0.109 0.081 0.129 0.109 0.170 0.272 0.084 0.070 0.082 0.002 - 0.032 
MON 0.239 0.205 0.206 0.199 0.206 0.187 0.232 0.174 0.186 0.159 0.225 0.183 0.252 0.427 0.193 0.089 0.172 0.077 0.070 - 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA - Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF? Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-
 eared Somali, WGJ ? Woyto-Guji, BGA ? Boran Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW? Tswana, ALP ? Italian Alpine, ARD - Ardi, BAL ? Baladie, HAI - Hair, MON ? Mongolia. 
 
 
 
 101
  
Table 11. Nei?s DA genetic distance for 11 Ethiopian and 9 reference populations/breeds genotyped, using 15 microsatellite 
markers. 
 ABG ABA AFA CEH GUM HAH KEF LES NWH SES WGJ BGA SEA GUB TSW ALP ARD BAL HAI MON 
ABG -                    
ABA 0.049 -                   
AFA 0.061 0.060 -                  
CEH 0.034 0.039 0.051 -                 
GUM 0.054 0.049 0.056 0.055 -                
HAH 0.048 0.040 0.038 0.040 0.049 -               
KEF 0.055 0.050 0.065 0.051 0.054 0.051 -              
LES 0.059 0.041 0.052 0.058 0.050 0.041 0.057 -             
NWH 0.034 0.044 0.051 0.026 0.048 0.044 0.044 0.052 -            
SES 0.048 0.037 0.033 0.038 0.052 0.024 0.051 0.033 0.033 -           
WGJ 0.051 0.030 0.043 0.050 0.044 0.041 0.046 0.035 0.041 0.033 -          
BGA 0.066 0.047 0.043 0.058 0.049 0.038 0.065 0.037 0.053 0.028 0.035 -         
SEA 0.069 0.066 0.060 0.067 0.074 0.059 0.065 0.050 0.058 0.053 0.042 0.055 -        
GUB 0.216 0.210 0.198 0.225 0.191 0.195 0.198 0.179 0.227 0.193 0.170 0.181 0.159 -       
TSW 0.152 0.137 0.143 0.156 0.138 0.122 0.147 0.121 0.137 0.125 0.127 0.151 0.143 0.234 -      
ALP 0.178 0.165 0.161 0.175 0.141 0.145 0.169 0.139 0.157 0.143 0.152 0.154 0.169 0.246 0.120 -     
ARD 0.171 0.144 0.123 0.164 0.133 0.128 0.178 0.128 0.164 0.137 0.144 0.140 0.175 0.222 0.180 0.158 -    
BAL 0.118 0.111 0.100 0.120 0.093 0.101 0.128 0.090 0.105 0.094 0.101 0.104 0.125 0.216 0.117 0.133 0.123 -   
HAI 0.131 0.127 0.122 0.133 0.128 0.114 0.141 0.094 0.119 0.099 0.112 0.116 0.131 0.221 0.106 0.119 0.129 0.061 -  
MON 0.198 0.182 0.183 0.195 0.172 0.171 0.215 0.160 0.183 0.160 0.190 0.176 0.203 0.310 0.178 0.149 0.203 0.113 0.104 - 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA - Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF? Keffa, LES ? Long-eared Somali, NWH ? North-West Highland,  
SES ? Short-eared Somali, WGJ ? Woyto-Guji, BGA ? Boran Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW? Tswana, ALP ? Italian Alpine, ARD - Ardi, BAL ? Baladie, HAI - Hair, 
MON ? Mongolia. 
 
 102
 4.4.3. Phylogenetic analysis  
The phylogenetic relationships between the populations in the study were 
determined using DS and DA genetic distances and their UPGMA and neighbour joining 
trees are shown in Fig. 22 to 25.  The DS neighbour joining phylogenetic tree (Figure 
22) for all populations revealed three clusters.  The first cluster comprised of six 
reference populations (Italian Alpine, Mongolia, Hair, Tswana, Baladie and Ardi), the 
second cluster was made up of Afar, Haraghe Highland, Short-eared Somali, Long-
 eared Somali, Gumez, Arsi-Bale, Keffa, North-west Highland, Abergalle, and Central 
Highland, and the third cluster was made up of Boran Galla, Small-east Africa (both 
from Kenya), Woyto-Guji (from Ethiopia), and Guinea Bissau (from West Africa).  The 
bootstrap values (number of times a node was observed in 1000 replicates of resample 
loci) ranged from 3% to 76%.  The DA neighbour joining phylogenetic tree (Figure 23) 
for all populations also revealed three, but more clear clusters.  The first cluster was 
made up of Abergalle, North-west Highland, Central Highland, Keffa, Gumez, Arsi-
 Bale, Woyto-Guji, Boran Galla, Short-eared Somali, Afar, Haraghe Highland, Long-
 eared Somali and Small-East Africa, the second cluster comprised of six reference 
populations (Italian Alpine, Mongolia, Hair, Tswana, Baladie and Ardi), and the third 
cluster was made up of only Guinea Bissau.  In general, the bootstrap values observed 
for the DA (Figure 23) were relatively lower (from 16% to 68%).  The UPGMA 
phylogenetic tree from DS genetic distances (Figure 24) for all populations also 
revealed three clear clusters similar to the DA phylogenetic tree with the bootstrap 
values relatively higher than for both DS and DA neighbour joining trees (from 8% to 
99%).  The UPGMA phylogenetic tree from DA genetic distances (Figure 25) for all 
populations also revealed three clusters very similar to the DA neighbour joining 
phylogenetic tree, but clustering the Ardi population as an intermediate fourth cluster 
between Guinea Bissau and the other five reference populations from which it is 
subdivided; with the bootstrap values relatively higher than for all the other three trees 
(from 17% to 99%).  In all cases, the Ethiopian goat populations cluster more closely 
together with the two Kenyan populations than to other reference populations, 
suggesting a common ancestry for them. 
 
 
 103
  
 
Figure 22. Phylogenetic tree representing the genetic relationships between the 11 
Ethiopian populations and 9 reference populations. The NJ tree was constructed 
using the genetic distance DS calculated from allele frequencies observed at 15 
microsatellite loci.  The numbers at the juncture of two branches are the percentage 
bootstrap values obtained after 1000 replications of re-sampled loci.  
(ABG ? Abergalle, ABA ? Arsi-Bale, AFA - Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe 
Highland, KEF? Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? 
Woyto-Guji, BGA ? Boran-Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW? Tswana, ALP ? Italian 
Alpine, ARD - Ardi, BAL ? Baladie, HAI - Hair, MON ? Mongolia) 
 
 
 
 
 
 
 
 
 104
  
 
Figure 23. Phylogenetic tree representing the genetic relationships between the 11 
Ethiopian populations and 9 reference populations. The NJ tree was constructed 
using the genetic distance DA calculated from allele frequencies observed at 15 
microsatellite loci.  The numbers at the juncture of two branches are the percentage 
bootstrap values obtained after 1000 replications of re-sampled loci. 
(ABG ? Abergalle, ABA ? Arsi-Bale, AFA - Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe 
Highland, KEF? Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? 
Woyto-Guji, BGA ? Boran-Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW? Tswana, ALP ? Italian 
Alpine, ARD - Ardi, BAL ? Baladie, HAI - Hair, MON ? Mongolia) 
 
 
 
 
 
 
 
 105
  
 
Figure 24. Phylogenetic tree representing the genetic relationships between the 11 
Ethiopian populations and 9 reference populations. The UPGMA tree was 
constructed using the genetic distance DS calculated from allele frequencies observed at 
15 microsatellite loci.  The numbers at the juncture of two branches are the percentage 
bootstrap values obtained after 1000 replications of re-sampled loci 
(ABG ? Abergalle, ABA ? Arsi-Bale, AFA - Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe 
Highland, KEF? Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? 
Woyto-Guji, BGA ? Boran-Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW? Tswana, ALP ? Italian 
Alpine, ARD - Ardi, BAL ? Baladie, HAI - Hair, MON ? Mongolia) 
 
 
 
 
 
 
 
 
 
 106
  
 
 
Figure 25. Phylogenetic tree representing the genetic relationships between the 11 
Ethiopian populations and 9 reference populations. The UPGMA tree was 
constructed using the genetic distance DA calculated from allele frequencies observed 
at 15 microsatellite loci.  The numbers at the juncture of two branches are the 
percentage bootstrap values obtained after 1000 replications of re-sampled loci. 
(ABG ? Abergalle, ABA ? Arsi-Bale, AFA - Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe 
Highland, KEF? Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? 
Woyto-Guji, BGA ? Boran-Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW? Tswana, ALP ? Italian 
Alpine, ARD - Ardi, BAL ? Baladie, HAI - Hair, MON ? Mongolia) 
 
4.4.4. Principal component analysis (PCA) 
The results of the PCA of the allele frequency data for the 11 Ethiopian and 9 
other reference populations using 15 microsatellite markers are shown in Table 12.  The 
variation accounted for by PC1, PC2 and PC3 was 31.81%, 18.04% and 10.38%, 
respectively, whish add up to 60.23%.  The PC1 separated the 11 Ethiopian, 2 Kenyan 
 107
 and 1 Guinea Bissau population from the other six (Mongolia, Italian Alpine, Hair, 
Tswana, Ardi and Baladie) populations on the other extreme.  The second principal 
component (PC2) ranked Guinea Bissau, Ardi, Small-east Africa, Hair and Italian 
Alpine on one side and the remaining 10 Ethiopian, one Egyptian, one Botswana, one 
Kenyan and one Mongolian on the other extreme leaving Woyto-Guji population just in 
the center of the axis.  The third principal component (PC3) ranked Ardi, Mongolia, 
Afar, Baladie, Central Highland, Gumez, Long-eared Somali, Hararghe Highland, 
Boran-Galla, Short-eared Somali, and Arsi-Bale on one side and Abergalle, Hair, North-
 West Highland, Guinea Bissau, Italian Alpine, Woyto-Guji, Keffa, Small-East Africa, 
and Tswana on the other extreme.  
Figure 26, 27 and 28 are two-dimensional plots of PC1 with PC2, PC1 with 
PC3 and PC2 with PC3 (in figure form), respectively, showing the same thing from the 
Table 12.  
 108
 Table 12. Ranked principal component (PC) values of allele frequencies from 20 
populations using 15 microsatellite markers (Log e). 
 
% variation accounted 31.81  18.04  10.38 
Cumulative % 31.81  49.85  60.23 
Breed PC1 Breed PC2 Breed PC3 
MON -0.179 GUB -0.257 ARD -0.118 
ALP -0.166 ARD -0.059 MON -0.05 
HAI -0.126 SEA -0.03 AFA -0.026 
TSW -0.118 HAI -0.009 BAL -0.02 
ARD -0.085 ALP -0.006 CEH -0.018 
BAL -0.081 WGJ 0 GUM -0.017 
SES 0.032 BAL 0.001 LES -0.013 
LES 0.035 BGA 0.005 HAH -0.011 
HAH 0.04 AFA 0.009 BGA -0.008 
Afar 0.044 TSW 0.009 SES -0.008 
BGA 0.046 GUM 0.02 ABA -0.001 
GUB 0.052 LES 0.023 ABG 0.001 
ABA 0.054 SES 0.026 HAI 0.001 
NWH 0.054 HAH 0.027 NWH 0.003 
GUM 0.057 KEF 0.031 GUB 0.016 
CEH 0.06 ABA 0.034 ALP 0.023 
WGJ 0.064 ABG 0.04 WGJ 0.025 
KEF 0.065 MON 0.041 KEF 0.028 
SEA 0.075 CEH 0.045 SEA 0.033 
ABG 0.078 NWH 0.05 TSW 0.161 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA - Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe 
Highland, KEF? Keffa, LES ? Long-Ear Somali, NWH ? North-West Highland, SES ? Short-Ear Somali, WGJ ? 
Woyto-Guji, BGA ? Boran-Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian 
Alpine, ARD - Ardi, BAL ? Baladie, HAI - Hair, MON ? Mongolia. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 109
  
Figure 26. Two-dimensional scatter plot representing the relationship between PC1 
and PC2 using allele frequency data from 15 microsatellite loci typed in 20 
populations. 
(ABG ? Abergalle, ABA ? Arsi-Bale, AFA - Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe 
Highland, KEF? Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? 
Woyto-Guji, BGA ? Boran-Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW? Tswana, ALP ? Italian 
Alpine, ARD - Ardi, BAL ? Baladie, HAI - Hair, MON ? Mongolia) 
 
 
 
Figure 27. Two-dimensional scatter plot representing the relationship between PC1 
and PC3 using allele frequency data from 15 microsatellite loci typed in 20 
populations. 
 110
 (ABG ? Abergalle, ABA ? Arsi-Bale, AFA - Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe 
Highland, KEF? Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? 
Woyto-Guji, BGA ? Boran-Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW? Tswana, ALP ? Italian 
Alpine, ARD - Ardi, BAL ? Baladie, HAI - Hair, MON ? Mongolia) 
 
 
 
 
Figure 28. Two-dimensional scatter plot representing the relationship between PC2 
and PC3 using allele frequency data from 15 microsatellite loci typed in 20 
populations. 
(ABG ? Abergalle, ABA ? Arsi-Bale, AFA - Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe 
Highland, KEF? Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? 
Woyto-Guji, BGA ? Boran-Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW? Tswana, ALP ? Italian 
Alpine, ARD - Ardi, BAL ? Baladie, HAI - Hair, MON ? Mongolia) 
 
4.4.5. Inference of population structure 
Pritchard et al. (2000) developed a model-based clustering approach to infer 
population structure and assign individuals to populations using multilocus genotype 
data, which identifies subgroups that have distinctive allele frequencies.  This model-
 based algorithm detailed in Pritchard et al. (2000) and Falush et al. (2003), places 
individuals in to K clusters. K is a parameter that is chosen in advance, but that can be 
varied in independent structure runs.  Individuals are given ?membership coefficient? for 
each cluster, such that the estimated membership coefficient of an individual sum to 1 
across the K clusters.  The matrix of membership coefficients, where the number of 
individuals is the number of rows and K is the number of columns, is referred to as the 
individual Q-matrix. For each population, membership coefficients for each cluster can 
 111
 be averaged across individuals to create a population Q-matrix.  The procedure was 
implemented using computer program STRUCTURE version 2.1 (available at 
http://pritch.bsd.uchicago.edu/software/structure2_1.html) with models assuming that 
individuals have mixed ancestry (admixture ancestry model; Pritchard et al., 2000) and 
allele frequencies are correlated among populations (Falush et al., 2003) with default 
parameters.  A burn-in period of 100,000 iterations of Markov chain Monte Carlo 
(MCMC) followed by 500,000 MCMC replications was used for running the structure 
program, and this parameter produced consistent results in five runs of each K.  The 
estimated individual membership coefficients were graphically visualized using a 
computer program DISTRUCT (Rosenberg, 2004; available from 
http://www.cmb.usc.edu/~noahr/distruct.html).  DISTRUCT is a program for the 
graphical display of population structure. Clusters are presented as colors, and 
individuals are depicted as bars partitioned into K colored segments that correspond to 
membership coefficient in K clusters. 
Figure 29 and 30 show the population structure, displayed with population Q-
 matrix and individual Q-matrix, respectively, inferred from allelic frequencies at 15 
microsatellite loci.  When K = 2 was applied, all the Ethiopian goat populations were 
clustered with the 2 Kenyan goat populations while all the remaining reference 
populations are clustered in the other cluster.  When K = 3 was applied, all the 
Ethiopian goat populations and the 2 Kenyan goat populations remain the same while 
the West African Guinea Bissau was separated from the remaining reference 
populations forming the 3rd cluster.  Then after K = 4 and K = 5 yielded no further 
population clustering. There was no clear population structure inferred from Ethiopian 
populations. 
 112
  
Figure 29. Estimated population structure displayed with population Q-matrix. 
Populations were partitioned in to K-coloured segments, which represent 
population?s estimated average membership fractions in the K clusters.  The black 
lines separate the different populations used for the study.  Populations are labeled 
below the figure with their country affiliations above the figure.  Five runs at each 
K produced nearly similar population membership coefficients.  
(ABG ? Abergalle, ABA ? Arsi-Bale, AFA - Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe 
Highland, KEF? Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? 
Woyto-Guji, BGA ? Boran-Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW? Tswana, ALP ? Italian 
Alpine, ARD - Ardi, BAL ? Baladie, HAI - Hair, MON ? Mongolia) 
 
 113
 Figure 30. Estimated population structure (graph) displayed with individual Q-
 matrix. Each individual was represented by a thin vertical line segment, which is 
partitioned in to K coloured segments, which represent individual?s estimated 
average membership fractions in the K clusters.  The black segments separate the 
individuals of different populations used for the study.  Population names are 
below the figure with their country affiliations above the figure.  Five runs at each 
K produced nearly similar individual membership coefficients.  
(ABG ? Abergalle, ABA ? Arsi-Bale, AFA - Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe 
Highland, KEF? Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? 
Woyto-Guji, BGA ? Boran-Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW? Tswana, ALP ? Italian 
Alpine, ARD - Ardi, BAL ? Baladie, HAI - Hair, MON ? Mongolia) 
 
4.5. Genetic differentiation 
4.5.1. GST and FST 
The GST estimates were slightly higher than the FST values obtained for all the 
15 loci (Table 13).  The GST values ranged from 0.014 for locus MAF35 to 0.162 for 
locus ILSTS87, while FST ranged from 0.003 for locus MAF35 to 0.152 for locus 
ILSTS87.  The overall GST and FST values indicate that 92.4% ? 93.6% of the total 
genetic differentiation was due to variation between individuals within populations 
when all populations used in the study are considered together. 
The FST values and their level of significance for pairs of populations are shown 
in Table 14.  The population pairs that were not genetically differentiated (P > 0.05) 
were the Afar/Hararghe Highland, Central Highland/Hararghe Highland, Hararghe 
Highland/Long-eared Somali, Central Highland/North West Highland, Afar/Short-eared 
Somali, Hararghe Highland/Short-eared Somali, Long-eared Somali/Short-eared 
Somali, North West Highland/Short-eared Somali, Arsi-Bale/Woyto-Guji, Long-eared 
Somali/Woyto-Guji, Hararghe Highland/Boran Galla, Long-eared Somali/Boran Galla, 
Short-eared Somali/Boran Galla, Woyto-Guji/ Boran Galla and Boran-Galla/Hair.  The 
genetic differentiation observed among the remaining population pairs were highly 
significant (P < 0.01) although the FST values were low except for two population pairs 
of Abergalle/Central Highland and Short-eared Somali/Woyto-Guji with significant 
levels at P < 0.05, indicating that there were significant genetic differentiation between 
these populations. 
 114
 The FST values were calculated for each locus to determine their ability to 
differentiate each pair of the populations.  The efficiency of the 15 microsatellite loci to 
differentiate between the 190 population pair combinations is shown in Table 15.  Out 
of the 15 loci, OarFCB304 was able to differentiate 85.26% of the population pair 
combinations (162 out of 190 population pair combinations) at P < 0.05 or lower, 
making it the most informative marker.  The next informative markers were ILSTS11 
and ILSTS87, which were able to differentiate 80% of the population pairs (152 out of 
190 population pair combinations) at P < 0.05 or lower.  In general, all the 15 loci were 
desirable at distinguishing between the population pairs, except for MAF35 which was 
able to differentiate only 13.16% of the population pairs (25 out of 190 population pair 
combinations) at P < 0.05 or lower, making it the least informative marker (Appendix 
18).  The genetic differentiation of the various goat populations studied is shown in the 
table below. 
 115
  
Table 13. Measure of genetic differentiation in Ethiopian and reference goat 
populations. 
Locus GST FST 
BM1818 0.057 0.044 
BMS1494 0.102 0.101 
ILSTS05 0.099 0.085 
ILSTS11 0.076 0.066 
ILSTS44 0.133 0.121 
ILSTS87 0.162 0.152 
INRA05 0.085 0.074 
INRA63 0.050 0.033 
INRA132 0.052 0.036 
MAF35 0.014 0.003 
MAF65 0.057 0.042 
MAF209 0.066 0.053 
OarAE129 0.070 0.052 
OarFCB304 0.053 0.040 
SRCRSP03 0.061 0.050 
Overall loci 0.076 0.064 
 116
 Table 14. Pair-wise tests of population differentiation between the 11 Ethiopian populations and 9 reference breeds.  FST values 
(below the diagonal) and their test of significance (above the diagonal) obtained after 190,000 permutations. 
 ABG ABA AFA CEH GUM HAH KEF LES NWH SES WGJ BGA SEA GUB TSW ALP ARD BAL HAI MON 
ABG - ** ** * ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** 
ABA 0.023 - ** ** ** ** ** ** ** ** ns ** ** ** ** ** ** ** ** ** 
AFA 0.04 0.033 - ** ** ns ** ** ** ns ** ** ** ** ** ** ** ** ** ** 
CEH  0.007 0.013 0.025 - ** ns ** ** ns ** ** ** ** ** ** ** ** ** ** ** 
GUM 0.024 0.023 0.034 0.022 - ** ** ** ** ** ** ** ** ** ** ** ** ** ** ** 
HAH 0.022 0.009 0.007 0.005 0.024 - ** ns ** ns ** ns ** ** ** ** ** ** ** ** 
KEF 0.025 0.015 0.033 0.015 0.027 0.014 - ** ** ** ** ** ** ** ** ** ** ** ** ** 
LES 0.026 0.008 0.018 0.018 0.015 0.007 0.02 - ** ns ns ns ** ** ** ** ** ** ** ** 
NWH 0.01 0.008 0.025 0.002 0.016 0.008 0.008 0.011 - ns ** ** ** ** ** ** ** ** ** ** 
SES 0.022 0.013 0.01 0.011 0.018 -0.001 0.021 0.003 0.005 - * ns ** ** ** ** ** ** ** ** 
WGJ 0.024 0.008 0.023 0.025 0.016 0.013 0.015 0.007 0.013 0.010 - ns ** ** ** ** ** ** ** ** 
BGA 0.036 0.025 0.015 0.03 0.024 0.011 0.038 0.008 0.019 0.000 0.009 - ** ** ** ** ** ** ** ** 
SEA 0.041 0.038 0.04 0.046 0.044 0.038 0.04 0.027 0.033 0.032 0.014 0.026 - ** ** ** ** ** ** ** 
GUB 0.142 0.13 0.134 0.143 0.137 0.13 0.135 0.132 0.142 0.129 0.11 0.125 0.113 - ** ** ** ** ** ** 
TSW 0.124 0.098 0.11 0.113 0.125 0.092 0.097 0.098 0.1 0.091 0.097 0.111 0.123 0.199 - ** ** ** ** ** 
ALP 0.137 0.112 0.119 0.119 0.118 0.102 0.121 0.094 0.107 0.088 0.114 0.105 0.135 0.202 0.078 - ** ** ** ** 
ARD 0.109 0.079 0.072 0.082 0.092 0.066 0.096 0.067 0.089 0.070 0.093 0.089 0.119 0.158 0.119 0.09 - ** ** ** 
BAL 0.06 0.046 0.053 0.049 0.056 0.04 0.052 0.038 0.044 0.037 0.049 0.053 0.078 0.143 0.063 0.066 0.05 - ns ** 
HAI 0.085 0.066 0.071 0.07 0.087 0.057 0.077 0.055 0.065 0.049 0.074 0.069 0.099 0.153 0.049 0.045 0.05 0.000 - ** 
MON 0.136 0.112 0.123 0.116 0.125 0.109 0.129 0.1 0.11 0.095 0.127 0.114 0.145 0.221 0.111 0.059 0.103 0.048 0.043 - 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Small-
 eared Somali, WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? Baladie, HAI- Hair, MON ? Mongolia. 
 117
 Table 15. Summary of genotypic differentiation tests for all the 190 pairs of 
population combinations (20 populations) for all the 15 loci. 
Locus 
Percentage of population pairs that could be distinguished by the 
marker, P < 0.05 
BM1818 63.68 
BMS1494 60.00 
ILSTS05 64.21 
ILSTS11 80.00 
ILSTS44 62.11 
ILSTS87 80.00 
INRA05 61.58 
INRA63 62.11 
INRA132 52.63 
MAF35 13.16 
MAF65 78.95 
MAF209 60.53 
OarAE129 73.16 
OarFCB304 85.26 
SRCRSP03 65.26 
Average for all population 64.18 
 
 118
 4.5.2. Analysis of molecular variance (AMOVA) 
In order to understand the partitioning of the levels of genetic diversity of the 20 
goat populations, an analysis of molecular variance (AMOVA) was conducted.  The 
result of the AMOVA (Table 16) revealed that 93.62% of the total genetic diversity 
existed among the individuals within populations and only 6.38% of the total genetic 
diversity accounted for differences among populations, which is exactly in accordance 
with the FST value. 
Subsequently three types of nested analyses were then carried out. The first nested 
analysis was carried out to assess the level of partitioning of genetic diversity between 
three different geographic regions: East Africa, West Africa and all the others (North 
Africa, Europe, Asia and Middle East).  The second nested analysis was carried out to 
assess the level of partitioning of genetic diversity between eight different geographic 
regions: Ethiopia, Kenya, Botswana, Egypt, Guinea Bissau, Turkey plus Italy (as 
Europe), Mongolia and Saudi Arabia.  The third nested analysis was carried out to assess 
the level of partitioning of genetic diversity between nine different geographic regions: 
Ethiopia, Kenya, Botswana, Egypt, Guinea Bissau, Turkey, Italy, Mongolia and Saudi 
Arabia.  The 20 populations were grouped into any of these regions depending on the 
country where they were sampled.  The results of the nested analyses (Table 16) show 
that differences between the three different geographic regions accounted for 7.20% of 
the variation, while differences among populations within these groups accounted for 
only 2.82% of the total variation.  The differences between the eight different geographic 
regions accounted for 7.07% of the total variation, while differences among populations 
within these groups accounted for only 1.78% of the variation.  The differences between 
the nine different geographic regions and among populations within these groups 
accounted for 7.32% and 1.59% of the total variation.  
 
 
 
 
 
 119
 Table 16. Analyses of molecular variance (AMOVA) for 20 goat populations using 
genotype data from 15 microsatellite loci. 
Source of variation Degree of  freedom 
Sum of  
squares 
Variance  
components 
Percentage  
of variation 
Among populations 19 504.66 0.2883 6.38 
Within populations 1532 6483.266 4.2319 93.62 
Nested analysis one     
Among 3A groups  2 258.765 0.33845 7.2 
Among populations within groups 17 245.895 13245 2.82 
Within populations 1532 6483.266 4.2319 89.99 
Nested analysis two     
Among 8B groups 7 370.761 0.32823 7.07 
Among populations within groups 12 133.899 0.0826 1.78 
Within populations 1532 6483.266 4.2319 91.15 
Nested analysis three     
Among 9C groups 8 388.14 0.33997 7.32 
Among populations within groups 11 116.52 0.0741 1.59 
Within populations 1532 6483.266 4.2319 91.09 
A Three groups are: Ethiopian plus Kenyan, Guinea Bissau, and the other 6 populations together (Botswana, Egypt, 
Turkey, Italy, Mongolia and Saudi Arabia). 
B Eight groups are: Ethiopia, Kenya, Botswana, Egypt, Guinea Bissau, Turkey plus Italy, Mongolia and Saudi Arabia. 
C Nine groups are: Ethiopia, Kenya, Botswana, Egypt, Guinea Bissau, Turkey, Italy, Mongolia and Saudi Arabia. 
 
 
 
 
 120
 4.6. Individual specific analysis 
Individuals were assigned to source populations using the GeneClass2 program 
(Piry et al., 2003) available at (http://www.montpellier.inra.fr/CBGP/softwares/), which 
allows different estimation procedures including the Bayesian (Rannala and Mountain, 
1997), frequency (Paetkau et al., 1995) as well as six distance methods; the Nei?s DA 
(Nei et al., 1983), Cavalli-Sforza and the Edwards chord distance (1967), the Nei 
Minimum distance (Nei, 1973), the Nei DS (Nei, 1972), the Jin and Chakraborty?s (1993) 
DAS and  the ??2 distance method of Goldstein et al. (1995). 
The efficiency of the different methods to assign individuals to their source 
population was tested and the results are shown in Table 17.  The Bayesian method 
computed using GeneClass2 program gave relatively the best result with 51.07% of 
individuals being assigned to their correct source population.  The frequency method and 
the maximum likelihood method of WHICHRUN also performed well with 49.66% and 
49.37% of individuals being correctly assigned to their source populations, respectively.  
The other methods are relatively less accurate with Goldstein et al. (1995) ??2 distance 
proving to be least accurate method with only 11.42% of the individuals being assigned 
to their correct source populations.  
The percentage of correct assignment of individuals to their source populations 
using a single locus was very low ranged from 6.44% for INRA132 to 15.98% for 
OarAE129 and OarFCB304 (Table 18).  However, when the six loci with the highest 
individuals i.e. BM1818, ILSTS11, ILSTS44, ILSTS87, OarAE129 and OarFCB304 
(Table 18) were used together, the percentage of individuals correctly assigned was 
increased to 39.82%.  Increasing the number of microsatellite loci to 15 resulted in 
correct assignment of 51.07% of the individuals.  
The results of the population assignment test obtained using frequency data from 
the 20 populations are shown in Table 19.  In the study 51.07% of the 776 individuals 
from the 20 populations (11 Ethiopian and 9 reference populations) were assigned to the 
corrected source population.  The population with the highest level of correct assignment 
was the Guinea Bissau sampled in West Africa (Guinea Bissau) with 97.8% of the 
individuals being correctly assigned, followed by the Mongolian population with 94.9% 
of the individuals being correctly assigned to the source population.  In general, all the 
 121
 Ethiopian and Kenyan populations had low levels of correct assignment, Hararghe 
Highland population having the lowest level of assignment (5%), however, higher 
assignments of 54.5%, 74.2%, 82.8%, 88.0%, 93.5%, 94.9% and 97.8% were observed in 
seven reference populations namely Baladie, Hair, Tswana, Ardi, Italian Alpine, 
Mongolian and Guinea Bissau, respectively. 
 
Table 17. Percent of correct individual-population assignment from 15 microsatellite 
loci using different methods. 
Methods Software % correct assignment 
Bayesian GeneClass 51.07 
Frequency GeneClass 49.66 
Maximum likelihood WHICHRUN 49.37 
The Nei et al. (1983) DA GeneClass 48.38 
Cavalli-Sforza and the Edwards chord distance GeneClass 44.14 
Nei minimum distance GeneClass 41.40 
The Nei standard distance (DS) GeneClass 41.39 
The Chakraborty and Jin DAS  GeneClass 39.37 
The Goldstein et al. (1995) (d?)2 GeneClass 11.42 
 
 122
 Table 18. Percent of correct individual-population assignment (for 20 populations) 
estimated for each microsatellite locus using the Bayesian method. 
Locus  % of correct assignment 
OarAE129* 15.98 
OarFCB304* 15.98 
ILSTS11* 15.08 
BM1818* 14.82 
ILSTS44* 14.56 
ILSTS87* 14.43 
MAF65 13.66 
INRA05 12.63 
BMS1494 12.24 
SRCRSP03 11.68 
INRA63 11.08 
MAF209 10.44 
ILSTS05 8.76 
MAF35 7.47 
INRA132 6.44 
*When these six loci taken together 39.82 
 123
 Table 19. Frequency of correct assignment of individuals to the 20 populations using genotype information from 15 
microsatellite loci using the Bayesian method. 
 No. ABG ABA AFA CEH GUM HAH KEF LES NWH SES WGJ BGA SEA GUB TSW ALP ARD BAL HAI MON 
ABG 46 45.7 10.9 0.0 6.5 0.0 2.2 4.3 4.3 15.2 2.2 8.7 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 
ABA 46 6.5 37.0 2.2 4.3 0.0 4.3 6.5 8.7 0.0 13.0 8.7 4.3 4.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 
AFA 44 4.5 0.0 38.6 2.3 4.5 18.2 0.0 2.3 0.0 11.4 4.5 6.8 4.5 0.0 0.0 0.0 0.0 2.3 0.0 0.0 
CEH  40 5.0 12.5 2.5 30.0 5.0 10.0 10.0 0.0 15.0 7.5 0.0 0.0 2.5 0.0 0.0 0.0 0.0 0.0 0.0 0.0 
GUM 43 2.3 2.3 0.0 4.7 48.8 2.3 7.0 2.3 9.3 0.0 9.3 11.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 
HAH 40 0.0 12.5 17.5 2.5 7.5 5.0 7.5 5.0 5.0 12.5 5.0 7.5 5.0 0.0 2.5 0.0 5.0 0.0 0.0 0.0 
KEF 38 2.6 0.0 2.6 2.6 10.5 2.6 44.7 5.3 10.5 5.3 7.9 0.0 5.3 0.0 0.0 0.0 0.0 0.0 0.0 0.0 
LES 48 6.3 4.2 4.2 6.3 4.2 4.2 2.1 25.0 0.0 18.8 6.3 14.6 2.1 0.0 0.0 0.0 0.0 0.0 0.0 2.1 
NWH 45 8.9 6.7 4.4 13.3 2.2 2.2 4.4 6.7 28.9 0.0 8.9 6.7 6.7 0.0 0.0 0.0 0.0 0.0 0.0 0.0 
SES 43 2.3 2.3 11.6 4.7 4.7 11.6 2.3 2.3 9.3 16.3 4.7 18.6 7.0 0.0 0.0 0.0 0.0 0.0 2.3 0.0 
WGJ 46 2.2 8.7 4.3 0.0 2.2 4.3 8.7 6.5 8.7 6.5 30.4 8.7 6.5 0.0 2.2 0.0 0.0 0.0 0.0 0.0 
BGA 35 2.9 2.9 5.7 0.0 5.7 11.4 2.9 17.1 0.0 11.4 8.6 31.4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 
SEA 39 5.1 2.6 7.7 0.0 0.0 5.1 0.0 0.0 0.0 10.3 10.3 2.6 53.8 0.0 0.0 0.0 0.0 0.0 2.6 0.0 
GUB 46 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 97.8 0.0 0.0 0.0 0.0 2.2 0.0 
TSW 29 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 82.8 3.4 0.0 6.9 6.9 0.0 
ALP 31 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 93.5 0.0 6.5 0.0 0.0 
ARD 25 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 4.0 0.0 0.0 0.0 4.0 88.0 4.0 0.0 0.0 
BAL 22 0.0 0.0 0.0 0.0 4.5 0.0 0.0 4.5 0.0 0.0 0.0 4.5 0.0 0.0 0.0 4.5 0.0 54.5 27.3 0.0 
HAI 31 0.0 0.0 3.2 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 22.6 74.2 0.0 
MON 39 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 2.6 2.6 94.9 
No. = number of animals. ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, LES ? Long-eared Somali, NWH ? 
North-West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? 
Baladie, HAI- Hair, MON? Mongolia. 
 124
 4.7. Genetic diversity and relationship within and between Ethiopian goat populations 
4.7.1. Hardy-Weinberg equilibrium (HWE) 
A total of 479 individuals, from 11 Ethiopian populations were genotyped using 18 
microsatellite markers.  After testing for linkage disequilibrium, null allele and HWE three 
markers BMC1222, BMS357 and SRCRSP07 were found to deviate significantly out of 
Hardy-Weinberg equilibrium across 11, 11 and 9 populations, respectively (Table 20).  
These populations (eleven, eleven and nine) were observed to have significantly lower 
proportions of heterozygotes than that expected at HWE for BMC1222, BMS357 and 
SRCRSP07 respectively (Table 20).  Consequently, these three markers were removed 
from further analysis.  
Then single-locus exact test for HWE were conducted for the 11 population and 
remaining 15 microsatellite loci.  Results indicated that 22 out of 165 or 13.33% of the 
comparisons (P < 0.05 or below; Table 21) deviated significantly from HWE, a figure 
higher than the 5% that would be expected from type I error alone.  Application of 
sequential Bonferroni adjustment for multiple testing reduced the number of significant 
locus-population combinations to one. 
Out of the 15 microsatellite loci used, no locus deviated from HWE for all the 
populations and also no population deviated from HWE for all the loci (Table 21).  
However, as it is indicated in Table 21, there were some populations with several loci 
deviate significantly from HWE. On one extreme, was found that five loci deviate from 
HWE in the Hararghe Highland population, while on the other extreme the Gumez 
population didn?t deviate from HWE at any locus.  Arsi-Bale, Afar, Short-eared Somali, 
and Woyto-Guji deviated from HWE at only one locus, while the remaining populations 
deviated from HWE at 2 to 3 loci. Most of these loci deviations from HWE corresponded to 
either heterozygote deficit or excess (Table 21). 
Four out of eleven populations had HWE deviation at the locus INRA63, while 
three out of the eleven populations had HWE deviation at the loci ILSTS44 and INRA05.  
 
 125
 Table 20. Results of Fisher?s exact test for Hardy-Weinberg equilibrium for all Ethiopian populations and 18 loci. 
Population 
BM 
1818 
BMC 
1222 
BMS 
1494 
BMS 
357 
ILSTS 
05 
ILSTS 
11 
ILSTS 
44 
ILSTS 
87 
INRA 
 05 
INRA 
63 
INRA 
132 
MAF 
35 
MAF 
65 
MAF 
209 
OarAE 
129 
OarFCB 
304 
SRCRSP 
03 
SRCRSP 
07 
All 
loci 
ABG *1 **1 ns **1 ns ns ns ns ns *1 ns ns ns ns ns ns ns **1 ** 
ABA ns **1 ns **1 ns ns ns ns * ns ns ns ns ns ns ns ns ns ** 
AFA ns **1 ns **1 ns ns **1 ns ns ns ns ns ns ns ns ns ns **1 ** 
CEH ns **1 ns **1 ns ns ** ns ns **1 ns ns ns ns ns *1 ns **1 ** 
GUM ns **1 ns **1 ns ns ns ns ns ns ns ns ns ns ns ns ns **1 ** 
HAH ns **1 ns **1 ns ns *1 ns * ns *2 ns ns *1 ns ns **1 ns ** 
KEF ns **1 ns **1 ns *1 ns ns ns ns ns ns *1 *1 ns ns ns *1 ** 
LES ns *1 ns **1 ns ns ns ns ns ** ns ns ns ns ** ns * **1 ** 
NWH ns **1 ns **1 ns ns ns ns ns *1 ns ns ns ns **1 ns ns **1 ** 
SES *1 **1 ns **1 ns ns ns ns ns ns ns ns ns ns ns ns ns **1 ** 
WGJ ns **1 ns **1 ns ns ns ns *1 ns ns ns ns ns ns ns ns **1 ** 
All 
population 
ns **1 ns **1 ns ns **1 ns * **1 ns ns ns ns * ns ns **1 ** 
ns = not significant; * = P < 0.05; ** = P < 0.01; 1 = heterozygote deficit; 2 = heterozygote excess. 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, LES ? Long-eared Somali, NWH ? North-
 West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
 
 126
  
Table 21. Results of Fisher?s exact test for Hardy-Weinberg equilibrium for all Ethiopian populations and 15 microsatellite 
markers. 
Population 
BM 
1818 
BMS 
1494 
ILSTS 
05 
ILSTS 
11 
ILSTS 
44 
ILSTS 
87 
INRA 
05 
INRA 
63 
INRA 
132 
MAF 
35 
MAF 
65 
MAF 
209 
OarAE 
129 
OarFCB 
304 
SRCRSP 
03 
All 
loci 
ABG *1 ns ns ns ns ns ns * ns ns ns ns ns ns ns ns 
ABA ns ns ns ns ns ns * ns ns ns ns ns ns ns ns ns 
AFA ns ns ns ns **1 ns ns ns ns ns ns ns ns ns ns ns 
CEH ns ns ns ns ** ns ns **1 ns ns ns ns ns *1 ns ** 
GUM ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns 
HAH ns ns ns ns * ns * ns *2 ns ns *1 ns ns **1 ** 
KEF ns ns ns *1 ns ns ns ns ns ns *1 *1 ns ns ns ns 
LES ns ns ns ns ns ns ns ** ns ns ns ns ** ns * ** 
NWH ns ns ns ns ns ns ns *1 ns ns ns ns **1 ns ns ns 
SES *1 ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns 
WGJ ns ns ns ns ns ns *1 ns ns ns ns ns ns ns ns ns 
All 
population 
ns ns ns ns **1 ns * **1 ns ns ns ns * ns ns ** 
ns = not significant; * = P < 0.05; ** = P < 0.01; 1 = heterozygote deficit; 2 = heterozygote excess. 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, LES ? Long-eared Somali, NWH ? North-
 West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
 
 127
 4.7.2. Linkage disequilibrium 
The linkage disequilibrium tests within each of the 11 populations were performed 
for the 105 pair-wise combinations of the 15 loci for each population.  Assuming a 5% type 
I error rate (P < 0.05), about four to six pairs of loci are expected to be in linkage 
disequilibrium.  On average 6% of the loci pairs were in linkage disequilibrium within the 
populations.  As indicated in Appendix 19, large number of loci pairs that were in linkage 
disequilibrium (P < 0.05) were observed in the Arsi-Bale (11 pairs), followed by the Long-
 eared Somali (nine pairs), Short-eared Somali (eight pairs), Afar and Woyto-Guji (seven 
pairs each), Central Highland (six pairs), Gumez and Keffa (five pairs each), Abergalle, 
Hararghe Highland and North-West Highland populations (four pairs each). 
As indicated in Appendix 20 the exact test for genotypic disequilibrium across all 
11 populations (105 pair-wise comparisons) yielded significant deviations in the following 
comparisons: BM1818/ILSTS11 (P < 0.01); BM1818/ILSTS44 (P < 0.05); BM1818/ 
MAF209 (P < 0.01); BM1818/OaeAE129 (P < 0.01); BM1494/OarAE129 (P < 0.05), 
BMS1494/OarFCB304 (P < 0.05); ILSTS11/INRA63 (P < 0.05); ILSTS87/INRA63 (P < 
0.05); and MAF209/OarFCB304 (P < 0.05).  These significant results observed are nine out 
of 105 pair-wise combinations (8.57%), a figure higher than the 5% that would be expected 
from type I error alone.  However, none of these nine tests remained significant following 
application of sequential Bonferroni corrections. 
 128
 4.7.3. Analysis of inbreeding coefficient within the Ethiopian goat populations 
 
The inbreeding was estimated for each population using Weir and Cockerham?s FIS 
estimator (Weir and Cockerham, 1984).  The calculations and tests were implemented in 
the computer program FSTAT version 2.9.3 (Goudet, 2001).  Significance levels were 
determined after 165,000 randomizations.  The results of inbreeding coefficient (FIS 
estimator) are given in Table 22. 
 
Table 22. Inbreeding coefficients (FIS estimator) estimated from 15 microsatellite loci 
for Ethiopian goat populations. 
Populations FIS 
Abergalle 0.044* 
Arsi-Bale 0.039ns 
Afar 0.076** 
Central Highland 0.077** 
Gumez -0.002ns 
Hararghe Highland 0.051* 
Keffa 0.03ns 
Long-eared Somali 0.051* 
North-West Highland 0.052* 
Short-eared Somali 0.016ns 
Woyto-Guji 0.059* 
ns not significant (P > 0.05), *P < 0.05, **P < 0.01. Significance levels were determined after 165,000 randomizations. 
 
Although the deficit (positive FIS value) was detected in 10 out of 11 populations, 
seven of these populations showed a significant inbreeding effect (P < 0.05 or below) 
(Abergalle, Afar, Central Highland, Hararghe Highland, Long-eared Somali, North-West 
Highland and Woyto-Guji populations).  Only one population (Gumez) showed 
heterozygosity excess (negative FIS value), which doesn?t show significant heterozygote 
excess (P > 0.05). 
 129
 4.7.4. Analysis of Fixation Indices or F-statistics 
The F statistics in population genetics has nothing to do with the F statistics 
evaluating differences in variances.  Here F stands for fixation index, fixation being 
increased homozygosity resulting from inbreeding. Population subdivision results in the 
loss of genetic variation (measured by heterozygosity) within subpopulations due to being 
small populations and genetic drift acting within each one of them.  This means that 
population subdivision would result in decreased heterozygosity relative to that expected 
heterozygosity under random mating as if the whole population was a single breeding unit.  
In the 1940s and 1950s, Sewall Wright developed three fixation indices to evaluate 
population subdivision: FIS (inter-individuals), FST (subpopulations), FIT (total population). 
FIS is a measure of the deviation of genotypic frequencies from panmixis 
frequencies in terms of heterozygous deficiency or excess.  It is known as the inbreeding 
coefficient (f), which is conventionally defined as the probability that two alleles in an 
individual are identical by descent (autozygous).  The technical description is the 
correlation of uniting gametes relative to gametes drawn at random from within a 
subpopulation (Individual within the Subpopulation) averaged over subpopulations.  It is 
calculated in a single population as: FIS = 1 - (HOBS / HEXP) [equal to (HEXP - HOBS / HEXP) 
where HOBS is the observed heterozygosity and HEXP is the expected heterozygosity 
calculated on the assumption of random mating.  It shows the degree to which 
heterozygosity is reduced below the expectation.  The value of FIS ranges between -1 and 
+1. Negative FIS values indicate heterozygote excess (out-breeding) and positive values 
indicate heterozygote deficiency (inbreeding) compared with HWE expectations. 
FST measures the effect of population subdivision, which is the reduction in 
heterozygosity in a subpopulation due to genetic drift.  FST is the most inclusive measure of 
population substructure and is most useful for examining the overall genetic divergence 
among subpopulations.  It is also called co-ancestry coefficient (q) [Weir & Cockerham, 
1984] or 'Fixation index' and is defined as correlation of gametes within subpopulations 
relative to gametes drawn at random from the entire population (Subpopulation within the 
total population).  It is calculated as using the subpopulation (average) heterozygosity and 
total population expected heterozygosity.  FST is always positive; it ranges between 0 = 
panmixis (no subdivision, random mating occurring, no genetic divergence within the 
 130
 population) and 1 = complete isolation (extreme subdivision).  FST is usually calculated for 
different loci, and then averaged across all loci, and all populations.  
FIT is rarely used. It is the overall inbreeding coefficient (F) of an individual relative 
to the total population (Individual within the total population). 
The unbiased estimators of Weir and Cockerham?s F-statistics (FIT, FST and FIS) 
(Weir and Cockerham, 1984) for each locus were calculated by jackknifing across 11 
Ethiopian goat populations using the computer program FSTAT version 2.9.3 (Goudet, 
2001).  Significance levels were determined after 10,000 randomizations of alleles within 
populations (FIS), of alleles in total populations (FIT), and after 15,000 randomizations of 
genotypes among populations assumed no random mating within populations (FST). Results 
of F-statistics and their significance levels are given in Table 23. 
The FST estimator was significantly different from zero (P < 0.01 or P < 0.001) for 
each of 15 loci (except for one MAF35, P = 0.4496) and for multi-locus, which means that 
the genetic differentiation has occurred among populations although the gene 
differentiation among populations was low (FST = 1.6%).  The FIS estimator was not 
significantly different from zero for single locus, but a significant positive value for six loci 
(BM1818 and MAF209, P<0.05; ILSTS05, P < 0.01; and ILSTS44, INRA63 and MAF65, 
P < 0.001), which means that an inbreeding effect within population was detected at these 
six loci.  The FIS estimator showed also non-significant negative values for three loci 
(ILSTS11, ILSTS87 and MAF35).  FIS was significantly different from zero for multi-locus 
(P < 0.001).  The values of FIT estimators were all positive except negative for two 
(ILSTS87, P = 0.72600 and MAF35, P = 0.82270) and significantly different from zero at 
P < 0.05 or P < 0.01 or P < 0.001 levels for all loci, but five (BMS1494, ILSTS11, 
INRA05, INRA132, and OarFCB304, P > 0.05).  FIT was significantly different from zero 
for multi-locus (P < 0.001).  As inbreeding effect within populations (FIS) was detected at 
BM1818, MAF209, ILSTS05, ILSTS44, INRA63, and MAF65, gene differentiation among 
populations has resulted in significant FIT estimators except for BM1818, MAF209, 
ILSTS05, ILSTS44, INRA63, and MAF65, which caused by both inbreeding within 
population plus gene differentiation among populations. 
 
 131
 Table 23. Results of F-statistics (Weir and Cockerham, 1984) for 15 microsatellite loci 
by jackknifing across the 11 Ethiopian goat populations. 
 
Locus F (FIT) ? (FST) f (FIS) 
BM1818 0.050 (0.024)* 0.010 (0.006)** 0.040 (0.024)* 
BMS1494 0.053 (0.034)ns 0.015 (0.009)*** 0.038 (0.036)ns 
ILSTS05 0.135 (0.038)** 0.021 (0.007)*** 0.116 (0.037)** 
ILSTS11 0.028 (0.028)ns 0.036 (0.013)*** -0.009 (0.019)ns 
ILSTS44 0.150 (0.032)*** 0.014 (0.014)** 0.138 (0.034)*** 
ILSTS87 -0.019 (0.033)ns 0.017 (0.006)*** -0.037 (0.037)ns 
INRA05 0.039 (0.049)ns 0.021 (0.009)*** 0.018 (0.044)ns 
INRA63 0.109 (0.043)*** 0.004 (0.004)*** 0.105 (0.042)*** 
INRA132 0.046 (0.047)ns 0.015 (0.007)*** 0.031 (0.046)ns 
MAF35 -0.043 (0.025)ns 0.000 (0.004)ns -0.043 (0.022)ns 
MAF65 0.084 (0.018)*** 0.017 (0.006)*** 0.068 (0.019)*** 
MAF209 0.083 (0.028)** 0.017 (0.007)*** 0.068 (0.030)* 
OarAE129 0.041 (0.017)* 0.009 (0.005)*** 0.033 (0.016)ns 
OarFCB304 0.025 (0.017)ns 0.009 (0.003)*** 0.016 (0.017)ns 
SRCRSP03 0.064 (0.031)* 0.030 (0.010)*** 0.035 (0.034)ns 
Multi-locus 0.061 (0.011)*** 0.016 (0.003)*** 0.045 (0.012)*** 
ns Not significant (P>0.05), * P < 0.05, ** P < 0.01, *** P < 0.001.  Significance levels were determined after 10,000 
randomizations of alleles within populations (FIS), of alleles in total populations (FIT), and after 15,000 randomizations of 
genotypes among populations assumed no random mating within populations (FST). Standard errors are given in 
parenthesis.  
 
 
 
 
 
 
 
 
 
 132
 4.8. The extent and distribution of the genetic diversity within the Ethiopian 
populations 
4.8.1. Mean number of alleles and heterozygosities 
A total of 133 alleles were observed in the 11 Ethiopian populations. Among the 
Ethiopian populations, the most polymorphic marker was OarFCB304 with 23 alleles, 
while the least polymorphic markers were INRA132 and MAF35 with 3 alleles.  Allele 
frequencies across all loci ranged from 1.04 to 95.00 (Appendix 21 to Appendix 35).  
Private alleles meaning alleles that were unique to a single or observed in only one 
population were observed in the following populations: Abergalle (one), Arsi-Bale (one), 
Afar (two), Central Highland (two), Gumez (two), Hararghe Highland (two), Keffa (three), 
Long-eared Somali (three), North-West Highland (four) and Short-eared Somali (one). 
However, the frequencies of these private alleles were less than 0.05 in all cases. 
The MNA ranged from 5.13 in Central Highland to 6.47 in Long-ear Somali when 
all the individuals in the population were considered (Table 24 to Table 26).  To remove 
any sample bias, the MNA for a random sample of 20 animals was calculated for each 
population with 250 replicate re-sampling of 20 individuals with replacement.  The 
Abergalle and Central Highland were the least diverse population both with MNA of 4.65, 
while the Long-eared Somali was again the most diverse population with MNA of 5.28. In 
general the mean numbers of alleles were moderate and were comparable among all the 
populations from the different geographic regions of Ethiopia in both cases. 
The expected heterozygosity (HE) and observed heterozygosity (HO) values 
averaged over loci showed an overall pattern similar to that observed for MNA per locus 
(Table 24 to Table 26).  The highest value for HE around 61% was observed within the 
Arsi-Bale population, while the lowest value around 55% was for Gumez population.  The 
highest vale for HO around 58% was again observed within the Arsi-Bale population, while 
the lowest value around 52% was observed for Afar.  In all cases average observed 
heterozygosities were lower than that expected under HWE.  
 
 
 
 
 
 133
  
Table 24. Expected heterozygosity (HE), observed heterozygosity (HO) and mean 
number of alleles (MNA) of all 11 Ethiopian populations. 
Population Sample size
 MNA 
(All ind) 
MNA 
(20 ind) 
HE ? s.e. HO ? s.e. 
Abergalle 46 5.33 4.65 0.57 ? 0.056 0.54 ? 0.053 
Arsi-Bale 46 5.33 4.87 0.61 ? 0.043 0.58 ? 0.045 
Afar 44 5.47 4.77 0.56 ? 0.051 0.52 ? 0.049 
Central Highland 40 5.13 4.65 0.57 ? 0.056 0.53 ? 0.058 
Gumez 43 5.53 4.71 0.55 ? 0.057 0.55 ? 0.054 
Hararghe Highland 40 5.93 5.10 0.58 ? 0.052 0.55 ? 0.053 
Keffa 38 5.60 5.05 0.58 ? 0.050 0.56 ? 0.047 
Long-eared Somali 48 6.47 5.28 0.59 ? 0.044 0.56 ? 0.044 
North West Highland 45 5.40 4.67 0.57 ? 0.053 0.54 ? 0.049 
Small-eared Somali 43 5.60 4.90 0.58 ? 0.054 0.57 ? 0.055 
Woyto-Guji 46 5.67 4.99 0.58 ? 0.052 0.54 ? 0.049 
Mean of all populations  5.59 4.88 0.58 ? 0.015 0.55 ? 0.015 
 134
 Table 25. The expected heterozygosity (HE) and observed heterozygosity (HO) per locus and population for Ethiopian goats.  
BM1818 BMS1494 ILSTS05 ILSTS11 ILSTS44 ILSTS87 INRA05 INRA63   
Pop  HO          HE  HO           HE  HO         HE  HO           HE  HO          HE  HO          HE  HO         HE  HO           HE 
ABG 0.630 0.705 0.522 0.532 0.130 0.124 0.696 0.701 0.609 0.661 0.304 0.292 0.522 0.542 0.522 0.634 
ABA 0.739 0.781 0.432 0.536 0.304 0.349 0.717 0.689 0.619 0.623 0.500 0.482 0.739 0.626 0.522 0.582 
AFA 0.523 0.553 0.523 0.439 0.295 0.390 0.614 0.627 0.477 0.685 0.273 0.306 0.432 0.543 0.705 0.633 
CEH 0.750 0.708 0.475 0.526 0.050 0.096 0.725 0.657 0.575 0.690 0.450 0.477 0.450 0.562 0.350 0.572 
GUM 0.767 0.777 0.349 0.302 0.186 0.173 0.605 0.620 0.558 0.641 0.333 0.336 0.628 0.574 0.698 0.658 
HAH 0.675 0.630 0.500 0.531 0.200 0.222 0.700 0.690 0.475 0.695 0.450 0.402 0.550 0.623 0.650 0.622 
KEF 0.500 0.520 0.500 0.605 0.289 0.321 0.605 0.694 0.579 0.672 0.421 0.369 0.684 0.646 0.711 0.705 
LES 0.771 0.742 0.438 0.473 0.271 0.339 0.667 0.684 0.583 0.653 0.563 0.449 0.625 0.590 0.542 0.687 
NWH 0.667 0.733 0.568 0.523 0.156 0.145 0.622 0.629 0.644 0.680 0.378 0.404 0.622 0.596 0.533 0.655 
SES 0.605 0.744 0.512 0.503 0.186 0.173 0.791 0.708 0.721 0.701 0.462 0.494 0.651 0.599 0.558 0.685 
WGJ 0.652 0.678 0.500 0.562 0.239 0.267 0.717 0.698 0.500 0.649 0.283 0.262 0.435 0.569 0.543 0.614 
Average 0.662 0.688 0.483 0.503 0.210 0.236 0.678 0.672 0.576 0.668 0.401 0.388 0.576 0.588 0.576 0.641 
SD 0.093 0.087 0.059 0.079 0.080 0.102 0.061 0.033 0.074 0.025 0.095 0.082 0.106 0.034 0.108 0.042 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, LES ? Long-eared Somali, NWH ? North-
 West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
 135
 Table 26. The expected heterozygosity (HE) and observed heterozygosity (HO) per locus and population for Ethiopian goats.  
INRA132 MAF35 MAF65 MAF209 OarAE129 OaeFCB304 SRCRSP03 Mean (SD) 
Pop 
HO          HE HO          HE HO          HE HO          HE HO          HE HO         HE HO         HE HO HE 
ABG 0.478 0.498 0.196 0.178 0.783 0.796 0.478 0.567 0.826 0.800 0.696 0.748 0.756 0.740 0.543 (0.053) 0.568 (0.056) 
ABA 0.478 0.534 0.217 0.232 0.844 0.869 0.543 0.626 0.727 0.776 0.696 0.716 0.652 0.663 0.582 (0.045) 0.606 (0.043) 
AFA 0.386 0.515 0.136 0.129 0.773 0.861 0.477 0.465 0.773 0.858 0.705 0.706 0.636 0.645 0.515 (0.049) 0.557 (0.051) 
CEH 0.425 0.500 0.125 0.119 0.757 0.860 0.650 0.578 0.781 0.791 0.650 0.728 0.700 0.698 0.528 (0.058) 0.571 (0.056) 
GUM 0.651 0.621 0.140 0.131 0.711 0.832 0.488 0.488 0.860 0.841 0.605 0.618 0.674 0.626 0.550 (0.054) 0.549 (0.057) 
HAH 0.700 0.530 0.125 0.120 0.800 0.887 0.425 0.562 0.842 0.800 0.700 0.728 0.462 0.651 0.550 (0.053) 0.579 (0.052) 
KEF 0.605 0.519 0.105 0.101 0.730 0.848 0.605 0.626 0.789 0.818 0.737 0.691 0.605 0.586 0.564 (0.047) 0.581 (0.050) 
LES 0.458 0.524 0.250 0.252 0.872 0.856 0.500 0.561 0.766 0.843 0.563 0.597 0.563 0.631 0.562 (0.044) 0.592 (0.044) 
NWH 0.467 0.497 0.200 0.182 0.795 0.841 0.467 0.533 0.800 0.825 0.689 0.691 0.556 0.677 0.544 (0.049) 0.574 (0.053) 
SES 0.442 0.506 0.163 0.151 0.884 0.864 0.442 0.472 0.810 0.845 0.628 0.616 0.744 0.674 0.573 (0.055) 0.582 (0.054) 
WGJ 0.543 0.538 0.200 0.182 0.761 0.861 0.630 0.628 0.783 0.834 0.783 0.731 0.587 0.594 0.544 (0.049) 0.578 (0.052) 
Mean 0.512 0.526 0.169 0.162 0.792 0.852 0.519 0.555 0.796 0.821 0.677 0.688 0.630 0.653 
SD 0.100 0.035 0.046 0.049 0.056 0.024 0.077 0.060 0.037 0.026 0.062 0.053 0.088 0.045 
0.551 
(0.015)s.e. 
0.576 
(0.015)s.e. 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar,  CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, KEF- Keffa,  LES ? Long-eared Somali,  NWH ? North-
 West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
 136
 4.9. Genetic relationship between the Ethiopian populations  
Two measures of genetic distances, Nei?s (1972) standard genetic distance (DS) and 
DA genetic distance (Nei et al., 1983) were used to estimate the genetic distances for each 
pair of populations.  
 
4.9.1. Nei?s standard genetic distance (DS)  
The Nei?s standard genetic distance (DS) values for Ethiopian populations with only 
one reference (as out-group) population are shown in Table 27.  The smallest DS for 
Ethiopian goat populations was -0.001 ? 0.004 between the Hararghe Highland and Short-
 ear Somali, followed by the value of 0.004 ? 0.004 between the Central Highland and 
North West Highland.  The highest DS for Ethiopian population was 0.055 ? 0.018 
observed between Abergalle and Afar populations.  Smaller DS values were observed for 
populations within neighbouring regions.  On the other hand the DS genetic distances 
observed between the populations in different geographic regions of the country were 
relatively higher.  
 
4.9.2. Nei?s DA genetic distances  
The genetic distances of the Ethiopian populations were determined using Nei?s DA 
index of genetic distance with one reference (as out-group) population, as in the case for the 
Nei?s DS genetic distance, the populations with smallest genetic distance were Hararghe 
Highland and Short-eared Somali with genetic distance of 0.024, followed by the value of 
0.026 again between the Central Highland and North West Highland (Table 28).  The 
highest DA among Ethiopian populations was between Keffa and Afar with genetic distance 
of 0.065.  As in the case for the Nei?s DS genetic distance, the DA genetic distances 
observed between populations from the neighbouring regions were small and those from 
relatively far apart regions of the country were relatively larger.  
 137
 Table 27. Nei?s standard genetic distance (DS) matrix for 11 Ethiopian populations and 1 reference population from West 
Africa. 
Pop ABG ABA AFA CEH GUM HAH KEF LES NWH SES WGJ GUB 
ABG - 0.013 0.018 0.008 0.014 0.014 0.016 0.014 0.007 0.013 0.014 0.078 
ABA 0.034 - 0.018 0.006 0.009 0.007 0.011 0.005 0.006 0.007 0.007 0.080 
AFA 0.055 0.047 - 0.014 0.016 0.006 0.025 0.011 0.019 0.005 0.014 0.080 
CEH 0.010 0.019 0.035 - 0.012 0.005 0.007 0.008 0.004 0.006 0.013 0.085 
GUM 0.031 0.031 0.046 0.030 - 0.015 0.013 0.005 0.009 0.009 0.008 0.086 
HAH 0.031 0.014 0.010 0.008 0.033 - 0.010 0.005 0.008 0.004 0.012 0.077 
KEF 0.036 0.023 0.046 0.022 0.036 0.021 - 0.009 0.007 0.012 0.011 0.071 
LES 0.037 0.012 0.026 0.026 0.019 0.011 0.030 - 0.006 0.003 0.007 0.081 
NWH 0.014 0.011 0.035 0.004 0.022 0.012 0.012 0.016 - 0.007 0.008 0.085 
SES 0.032 0.019 0.014 0.017 0.024 -0.001 0.030 0.005 0.007 - 0.006 0.078 
WGJ 0.034 0.012 0.033 0.036 0.021 0.020 0.022 0.011 0.019 0.015 - 0.063 
GUB 0.220 0.211 0.201 0.222 0.202 0.200 0.211 0.210 0.224 0.198 0.163 - 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar,  CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, KEF- Keffa,  LES ? Long-eared Somali, NWH ? North-
 West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji, GUB ? Guinea Bissau. 
 
 
 
 
 
 
 
 
 138
  
 
Table 28. Nei?s DA genetic distance matrix for 11 Ethiopian populations and 1 reference population from West Africa. 
Pop ABG ABA AFA CEH GUM HAH KEF LES NWH SES WGJ GUB 
ABG -            
ABA 0.049 -           
AFA 0.061 0.060 -          
CEH 0.034 0.039 0.051 -         
GUM 0.054 0.049 0.056 0.055 -        
HAH 0.048 0.040 0.038 0.040 0.049 -       
KEF 0.055 0.050 0.065 0.051 0.054 0.051 -      
LES 0.059 0.041 0.052 0.058 0.050 0.041 0.057 -     
NWH 0.034 0.044 0.051 0.026 0.048 0.044 0.044 0.052 -    
SES 0.048 0.037 0.033 0.038 0.052 0.024 0.051 0.033 0.033 -   
WGJ 0.051 0.030 0.043 0.050 0.044 0.041 0.046 0.035 0.041 0.033 -  
GUB 0.216 0.210 0.198 0.225 0.191 0.195 0.198 0.179 0.227 0.193 0.170 - 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, KEF- Keffa, LES ? Long-eared Somali, NWH ? North-West 
Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji, GUB ? Guinea Bissau. 
 
 139
 4.9.3. Phylogenetic relationship among Ethiopian goat populations  
The phylogenetic relationship between the 11 Ethiopian goat populations was 
determined using DS and DA genetic distances and their UPGMA with out reference 
population and with one reference population as out-group is shown in the form of NJ trees 
in (Figure 31 to Figure 38).  The DS phylogenetic tree with one reference population 
(Figure 31) shows nine relatively clear groups.  The first cluster being Abergalle and 
Central Highland with Abergalle far away from the Central Highland, the second cluster 
North-west Highland with very near branch to Central Highland, the third cluster Keffa, the 
fourth cluster Arsi-Bale, the fifth cluster Long-eared Somali, the sixth cluster Short-eared 
Somali with very close branch to the seventh cluster, which is Afar and Haraghe Highland, 
Afar being very far from the Haraghe Highland, the eighth cluster Gumez, and the ninth 
cluster Woyto-Guji with the bootstrap values ranging from 25% to76%.  Unlike the DS NJ 
tree the DA phylogenetic tree with one reference population (Figure 32) showed similar 
clustering except for the clustering of Central Highland together with North-west Highland 
instead of Central Highland with Abergalle and Afar together with Short-eared Somali 
instead of Afar with Haraghe Highland with bootstrap values ranging from 15 to 88%.  A 
similar topology is observed for the phylogenetic trees constructed from DS and DA genetic 
distances without any reference population (Figure 33 and Figure 34) with bootstrap 
values ranging from 6% to 79% and 33% to 89% respectively with a little difference for the 
DA.  On the other hand all the UPGMA from DS distance for with and without out-group 
showed consistent clustering of Central Highland with North-West Highland and Hararghe 
Highland with Short-eared Somali with Long-eared Somali very close to them (Figure 35 
and Figure 36) with bootstrap values ranging from 7% to 69% for with one reference 
population, while with bootstrap values ranging from 7% to 52% for with out reference 
population. The UPGMA from DA distance for with and without out-group showed 
consistent clustering of Central Highland with North-West Highland and Hararghe 
Highland with Short-eared Somali (Figure 37 and Figure 38) with bootstrap values 
ranging from 26% to 82% for with one reference population, while with bootstrap values 
ranging from 31% to 82% for with out reference population. 
 
 140
  
 
Figure 31. Phylogenetic tree representing the genetic relationship between the 11 
Ethiopian populations with GUB as an out-group population. The NJ tree was 
constructed using the genetic distance DS calculated from allele frequencies observed at 15 
microsatellite loci.  The numbers at the juncture of two branches are the percentage 
bootstrap values obtained after 1000 replications of re-sampled loci.  
(ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji, 
GUB ? Guinea Bissau) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 141
  
 
 
Figure 32. Phylogenetic tree representing the genetic relationship between the 11 
Ethiopian populations with GUB as an out-group population. The NJ tree was 
constructed using the genetic distance DA calculated from allele frequencies observed at 15 
microsatellite loci.  The numbers at the juncture of two branches are the percentage 
bootstrap values obtained after 1000 replications of re-sampled loci.  
(ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji, 
GUB ? Guinea Bissau) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 142
  
 
 
Figure 33. Phylogenetic tree representing the genetic relationship between the 11 
Ethiopian populations. The NJ tree was constructed using the genetic distance DS 
calculated from allele frequencies observed at 15 microsatellite loci.  The numbers at the 
juncture of two branches are the percentage bootstrap values obtained after 1000 
replications of re-sampled loci. 
(ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 143
  
 
Figure 34. Phylogenetic tree representing the genetic relationship between the 11 
Ethiopian populations. The NJ tree was constructed using the genetic distance DA 
calculated from allele frequencies observed at 15 microsatellite loci.  The numbers at the 
juncture of two branches are the percentage bootstrap values obtained after 1000 
replications of re-sampled loci. 
 (ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji) 
 
 
 
 
 
 
 
 
 
 
 144
  
 
Figure 35. UPGMA tree representing the genetic relationship between the 11 
Ethiopian populations with GUB as an out-group population. The NJ tree was 
constructed using the genetic distance DS calculated from allele frequencies observed at 15 
microsatellite loci.  The numbers at the juncture of two branches are the percentage 
bootstrap values obtained after 1000 replications of re-sampled loci. 
 (ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji, 
GUB ?Guinea Bissau) 
 
 
 
 
 145
  
 
Figure 36. DA-UPGMA tree representing the genetic relationship between the 11 
Ethiopian populations with GUB as an out-group population. The NJ tree was 
constructed using the genetic distance DA calculated from allele frequencies observed at 15 
microsatellite loci.  The numbers at the juncture of two branches are the percentage 
bootstrap values obtained after 1000 replications of re-sampled loci   
(ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, 
KEF ? Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji, 
GUB ?Guinea Bissau) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 146
  
Figure 37. DS-UPGMA tree representing the genetic relationship between the 11 
Ethiopian populations. The NJ tree was constructed using the genetic distance DS 
calculated from allele frequencies observed at 15 microsatellite loci.  The numbers at the 
juncture of two branches are the percentage bootstrap values obtained after 1000 
replications of re-sampled loci  
 (ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 147
  
 
Figure 38. DA-UPGMA tree representing the genetic relationship between the 11 
Ethiopian populations. The NJ tree was constructed using the genetic distance DA 
calculated from allele frequencies observed at 15 microsatellite loci.  The numbers at the 
juncture of two branches are the percentage bootstrap values obtained after 1000 
replications of re-sampled loci.  
(ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji) 
 
4.9.4. Principal component analysis (PCA) 
Table 29 shows the results of the PC analysis of the allele frequency data for the 11 
Ethiopian goat populations using 15 microsatellite markers.  The variation accounted for by 
PC1, PC2 and PC3 was 24.64%, 18.62% and 15.42%, respectively.  The PC1 separated the 
Abergalle, Central Highland, Keffa, North-west Highland, Gumez and Arsi-Bale on one 
end from Woyto-Guji, Long-eared Somali, Hararghe Highland, Short-eared Somali, and 
 148
 Afar populations on the other extreme.  PC2 ranked Woyto-Guji, Gumez, Long-eared 
Somali, Arsi-Bale, and Keffa on one end and Short-eared Somali, North-west Highland, 
Hararghe Highland, Abergalle, Afar, and Central Highland on the other extreme. PC3 
ranked Keffa, Arsi-Bale, Hararghe Highland, North-west Highland and Woyto-Guji on one 
end and Long-eared Somali, Short-eared Somali, Afar, Abergalle and Gumez, on the other 
extreme with Central Highland at the center. 
Figure 39 to Figure 41 is two-dimensional plots of PC1 with PC2, PC1 with PC3 
and PC2 with PC3, respectively, showing the same thing from the Table 26 in figure form. 
 
Table 29. Ranked PC values of allele frequencies from 11 Ethiopian populations using 
15 microsatellite markers. 
% Variation accounted 24.64  18.62  15.42 
Cumulative % 24.64  43.26  58.68 
Breed PC1 Breed PC2 Breed PC3 
ABG -0.122 WGJ -0.1 KEF -0.106 
CEH -0.065 GUM -0.085 ABA -0.071 
KEF -0.055 LES -0.06 HAH -0.034 
NWH -0.053 ABA -0.057 NWH -0.016 
GUM -0.031 KEF -0.018 WGJ -0.011 
ABA -0.027 SES 0.016 CEH 0 
WGJ 0.021 NWH 0.021 LES 0.009 
LES 0.051 HAH 0.055 SES 0.015 
HAH 0.061 ABG 0.062 AFA 0.027 
SES 0.066 AFA 0.069 ABG 0.065 
AFA 0.154 CEH 0.097 GUM 0.123 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji. 
 
 
 149
  
 
Figure 39. Two-dimensional graph representing the relationship between PC1 and 
PC2 using allele frequency data from 15 microsatellite loci typed in 11 Ethiopian goat 
populations. 
(ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji) 
 
 
Figure 40. Two-dimensional graph representing the relationship between PC1 and 
PC3 using allele frequency data from 15 microsatellite loci typed in 11 Ethiopian goat 
populations. 
 150
 (ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji) 
 
 
Figure 41. Two-dimensional graph representing the relationship between PC2 and 
PC3 using allele frequency data from 15 microsatellite loci typed in 11 Ethiopian goat 
populations. 
(ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji) 
 
4.9.5. Inference of population structure 
Figure 42 and Figure 43 show the population structure, displayed with population 
Q-matrix and individual Q-matrix, respectively, inferred from allelic frequencies at 15 
microsatellite loci.  In this case there was no clear population structure inferred from 
Ethiopian populations by applying any K values.  
 
 151
  
Figure 42. Estimated population structure displayed with population Q-matrix. 
Populations were partitioned in to K-coloured segments, which represent population?s 
estimated average membership fractions in the K clusters.  The black lines separate the 
different populations used for the study.  Populations are labeled below the figure with their 
country affiliations above the figure.  Five runs at each K produced nearly similar 
population membership coefficients 
(ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji) 
 
 
 
 
Figure 43. Estimated population structure displayed with individual Q-matrix.  Each 
individual was represented by a thin vertical line, which is partitioned in to K coloured 
 152
 segments, which represent individual?s estimated average membership fractions in the K 
clusters.  The black lines separate the individuals of different populations used for the 
study.  Populations are labeled below the figure with their country affiliations above the 
figure.  Five runs at each K produced nearly similar individual membership coefficients.  
 (ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji) 
 
4.9.6. Analysis of gene flow among populations 
Gene flow (also known as gene migration) is the transfer of genes from one 
population to another, resulted from movement of individuals or their gametes.  Migration 
into or out of a population may be responsible for a marked change in gene pool 
frequencies (the number of individual members with a particular trait).  Immigration may 
result in the addition of new genetic material to the established gene pool of a particular 
species or population, and conversely emigration results in the removal of genetic material.   
There are a number of factors that affect the rate of gene flow between different 
populations.  One of the most significant factors is mobility, and animals tend to be more 
mobile than plants.  Greater mobility of an individual tends to give it greater migratory 
potential. 
Gene flow is usually expressed as migration rate m, defined as the proportion of 
alleles of migrant origin in a population per generation.  Gene flow is particularly difficult 
to monitor directly, but usually is inferred from spatial distribution of genetic markers by 
statistical approaches.  Most of the approaches are based on equilibrium expectations 
derived from theoretical models of population structure under neutrality theory, e.g. the 
?island model? wherein a species is assumed to be subdivided into populations (demes or 
islands) of equal size N, all of which exchange alleles with equal probabilities (random 
migration between separate populations); or the ?stepping-stone? model wherein only 
adjacent demes (neighbouring populations) exchange alleles.  Allelic frequencies in finite 
populations are influenced by random genetic drift also, which is a function of effective 
population size.  The influence of drift and gene flow are difficult to tease apart and thus 
most statistical procedures applied to spatial genetic information only permit estimates of 
Nm, which can be interpreted as the absolute number of individuals exchanged between 
populations per generation.  Also Nm is of particular interest because under neutral theory, 
 153
 the level of divergence among populations at equilibrium is a function of the numbers of 
migrants rather than the proportions of individuals exchanged.  The most common 
approach to estimate Nm is from F-statistics defined as the following formula: 
FST = 1/ (1 + 4Nm) or Nm = (1-FST)/4FST (Wright, 1969). 
The number of migrants per generation (Nm), which is an indirect estimator of gene 
flow, was calculated for all pairs of populations using the above equation. The calculations 
were performed using the computer program GENETIX version 4.05 (available at 
http://www.univ-montp2.fr/~genetix/genetix/genetix.htm), and results are given in Table 
30.  For the 11 Ethiopian goat populations the highest Nm was estimated between two 
geographically adjacent populations: Short-eared Somali & Haraghe Highland (999999), 
followed by 120.88 between North-West Highland and Central Highland populations, 
followed by 82.19 between Short-eared Somali & Long-eared Somali, while the lowest Nm 
was estimated between Afar and Abergalle populations (6.06).  The level of gene flow was 
relatively low between geographically distant or isolated populations (for example the Nm 
was 6.06 between Abergalle and Afar, 7.3 between Arsi-Bale and Afar). 
 
Table 30, The number of migrants per generation, (Wright, 1969) among 11 goat populations 
estimated from Weir and Cockerham?s (1984) FST estimator using the equation: FST = 1/ 
(1+4Nm) or Nm = (1-FST)/4FST. 
 ABG ABA AFA CEH GUM HAH KEF LES NWH SES WGJ 
ABG            
ABA 10.65           
AFA 6.06 7.43          
CEH 35.79 19.2 9.63         
GUM 10.27 10.53 7.04 10.89        
HAH 11.19 27.04 35.26 48.98 9.99       
KEF 9.8 16.12 7.38 16.2 9.18 17.28      
LES 9.52 32.34 13.49 14 16.69 33.53 12.4     
NWH 24.55 31.95 9.6 120.88 14.99 30.95 30.51 22.32    
SES 10.98 19.21 24.32 21.67 13.43 999999.00 11.83 82.19 53   
WGJ 10.22 30.54 10.44 9.89 15.32 18.63 16.3 34.71 19.43 24.68  
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, KEF- 
Keffa, LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji. 
 154
 4.10. Genetic differentiation 
4.10.1. Test on allele size information content 
The mutation process at microsatellite loci typically occurs at high rates and with 
stepwise changes in allele sizes, features that may introduce bias when using classical 
measures of population differentiation based on allele identity (e.g., FST, Nei's DS genetic 
distance).  Allele size-based measures of differentiation, assuming a stepwise mutation 
process [e.g., Slatkin's RST (Slatkin, 1995), Goldstein et al.'s ??2 (Goldstein et al., 1995)] 
may better reflect differentiation at microsatellite loci, but they suffer high sampling 
variance (Hardy et al., 2003).  The relative efficiency of allele size- vs. allele identity-based 
statistics depends on the relative contributions of mutations vs. drift to population 
differentiation.  
Hardy et al. (2003) developed a method based on a randomization procedure of 
allele sizes to determine whether stepwise-like mutations contributed to genetic 
differentiation.  This test can be applied to any microsatellite data set designed to assess 
population differentiation and can be interpreted as testing whether FST = RST. 
The test indicates whether allele sizes provide information on population 
differentiation given a data set, that is, whether shifts in allele sizes resulting from 
stepwise-like mutations contribute to population differentiation.  Contribution of stepwise-
 like mutations to genetic differentiation requires (i) that the mutation process is at least 
partially SMM-like and (ii) that the mutation rate, ?, is large enough relative to the effect of 
drift and migration (e.g. ? ? m, otherwise new mutations are quickly spread beyond their 
native population by migration).  The principle of the test is based on obtaining a 
distribution of a statistic under the null hypothesis (HO) that differences in allele sizes do 
not contribute to population differentiation.  Therefore, a randomization procedure whereby 
the different allele sizes observed at a locus for a given data set is randomly permuted 
among allelic states.  Throughout the randomization procedure, genotypes are defined in 
terms of allelic states and are not modified, but allele sizes are randomly reassigned among 
allelic states.  After such a randomization, any two genes originally having the same allele 
size remain identical, although it can be for another allele size, whereas any two genes 
originally bearing different alleles of small size difference may bear alleles of large size 
 155
 difference, or reciprocally.  Hence, the allele identity information is kept intact but not the 
allele size information.  Under the null hypothesis, the randomization procedure should not 
affect the expectation of a measure of differentiation such as RST.  On the contrary, if allele 
sizes contribute to genetic differentiation, the RST computed after allele size permutation 
(hereafter called pRST) would depend solely on allele identity/nonidentity and hence have a 
smaller expectation than the value computed before randomization.  The test can thus be 
designed by comparing the observed RST value (before randomization) to the distribution of 
pRST values obtained for all possible configurations of allele size permutations.  From this 
comparison, a probability that the null hypothesis holds can be estimated as the proportion 
of pRST values larger than the observed RST (one-tailed test).  On a multi-locus RST estimate, 
permuting allele sizes within each locus can carry out the test.  It is noteworthy that the test 
makes no assumptions on the mutation model:  A significant result (RST significantly > 
pRST) suggests that mutations contributed to genetic differentiation, and that the mutation 
process follows at least partially a SMM and FST is likely to provide a biased estimate of 
gene flow parameters.  A non-significant result (RST not significantly different from pRST) 
would suggest that allele size is not informative for population differentiation, because the 
mutation process is not stepwise-like and/or because mutations had not contributed to 
differentiation.  In this case, allele identity-based statistics such as FST and Nei?s DS genetic 
distance should be preferential over allele size-based ones such as RST and ??2.  
The global RST statistics were computed and pRST values were obtained by applying 
the allele size permutation tests for our dataset. Significance levels of RST values were 
determined after 20,000 random permutations of the allele sizes.  The analyses were 
performed with computer program SPAGeDi (Hardy and Vekemans, 2002), and the results 
are in Table 31.  The RST was not significantly (P < 0.05) different from pRST values for 
either the RST or for the multi-locus test.  The tests indicate that mutations have not 
contributed to the genetic differentiations among Ethiopian goat populations, and genetic 
drift was responsible for the genetic differentiations among Ethiopian goat populations.  
Therefore, the allele identity based statistics such as FST (Weir and Cockerham, 1984), 
Nei?s standard genetic distance DS (Nei, 1972), DA genetic distance (Nei et al., 1983) and 
DC (Cavalli-Sforza and Edwards, 1967) could be used to explain genetic differentiations 
and genetic relationships between goat populations. 
 156
  
Table 31. Differentiation among Ethiopian goat populations, estimated by global RST, 
mean pRST values per locus and for a multilocus average. 
Locus No. of alleles RST PRST (95% C.I.) 
BM1818 12 0.009ns 2.56E-05 (-0.008, 0.012) 
BMS1494 9 0.033ns -2.27E-05 (-0.008, 0.012) 
ILSTS05 5 0.011ns -7.54E-05 (-0.008, 0.012) 
ILSTS11 10 0.049ns 4.67E-05 (-0.008, 0.013) 
ILSTS44 5 0.019ns -5.15E-05 (-0.008, 0.012) 
ILSTS87 8 0.025ns 2.82E-05 (-0.008, 0.013) 
INRA05 5 0.002ns 7.49E-06 (-0.008, 0.012) 
INRA63 8 0.014ns -5.47E-05 (-0.009, 0.014) 
INRA132 3 0.016ns 1.46E-05 (-0.008, 0.012) 
MAF35 3 0.000ns 1.32E-05 (-0.007, 0.011) 
MAF65 12 0.018ns -5.45E-05 (-0.009, 0.014) 
MAF209 4 0.019ns -5.09E-05 (-0.009, 0.013) 
OarAE129 20 0.030ns 1.47E-05 (-0.009, 0.013) 
OarFCB304 23 0.001ns 2.97E-05 (-0.008, 0.013) 
SRCRSP03 6 0.035ns 3.63E-06 (-0.008, 0.013) 
All loci 133 0.024ns 9.57E-06 (-0.005, 0.006) 
The 95% confidence interval given with pRST is the 95% central pRST values obtained after 20,000 random permutations 
of the allele sizes. P-values of allele size permutation tests on RST are denoted as follows: ns, non-significant (P > 0.05). 
 
4.10.2. Genetic differentiation between populations  
Table 32 shows two measures of population sub-division, the GST (Nei, 1973) and 
FST (Weir and Cockerham, 1984) values across populations.  The values for the two 
measures of population differentiation were similar with the GST values being slightly 
higher.  The GST values ranged from 0.010 for locus MAF35 to 0.044 for locus ILSTS11, 
while FST ranged from 0.000 to 0.037 for the same loci respectively.  The overall values 
were 0.026 for GST and 0.016 for FST, showing that 97.4 to 98.4% of the total genetic 
differentiation was due to differences between individuals within populations. 
 157
 Table 33 and Table 34 show the FST values and their level of significance for pairs 
of Ethiopian populations alone and with one out-group population respectively.  The least 
genetically differentiated population pairs were the Short-eared Somali/Hararghe Highland, 
North-West Highland/Central Highland, and Short-eared Somali/Long-eared Somali with 
FST value of -0.001, 0.002 and 0.003 respectively in both cases.  The population pair with 
the highest genetic differentiation was Afar/Abergalle with FST value of 0.040 in both 
cases.  The genetic differentiation observed among the population pairs were highly 
significant (P < 0.001) for all except for two pairs (P < 0.01) and, eight pairs (P < 0.05) 
even where low FST values were obtained, indicating that there was significant level of 
genetic differentiation between the Ethiopian goat populations. 
The FST values calculated for each population pair and each locus determine the 
ability of each locus to differentiate between population pairs.  Table 35 and Appendix 36 
show the efficiency of the 15 microsatellite loci to differentiate between the 55 population 
pair combinations.  Out of the 15 loci, ILSTS11 was able to differentiate 72.73% of the 
population pair combinations (40 out of 55 population pair combinations) at P < 0.05 or 
lower, making it the most informative marker.  The next informative marker was MAF65, 
which was able to differentiate 69.09% of the population pairs (38 out of 55 population pair 
combinations) at P < 0.05 or lower.  On the other hand both BM1818 and ILSTS44 were 
able to differentiate only 20% of the population pairs (11 out of 55 population pair 
combinations each) at P < 0.05 or lower followed by MAF35 which was able to 
differentiate only 3.6% of the population pairs (2 out of 55 population pair combinations) at 
P < 0.05 or lower, which makes it the least informative marker. 
 
 
 
 
 
 
 
 
 
 158
 Table 32. Measure of population differentiation in Ethiopian goat populations. 
Loci GST FST 
BM1818 0.021 0.011 
BMS1494 0.025 0.015 
ILSTS05 0.031 0.021 
ILSTS11 0.044 0.037 
ILSTS44 0.024 0.014 
ILSTS87 0.025 0.017 
INRA05 0.029 0.021 
INRA63 0.015 0.004 
INRA132 0.025 0.015 
MAF35 0.010 0.000 
MAF65 0.027 0.017 
MAF209 0.026 0.017 
OarAE129 0.019 0.009 
OarFCB304 0.019 0.009 
SRCRSP03 0.038 0.030 
Overall loci 0.026 0.016 
 
 
 159
 Table 33. Pair wise tests of population differentiation between the 11 Ethiopian goat populations.  FST values (below the 
diagonal) and their test of significance (above the diagonal) obtained after 55,000 permutations.  * = P < 0.05, ** = P < 0.01,  
*** = P < 0.001. 
Pop ABG ABA AFA CEH GUM HAH KEF LES NWH SES WGJ 
ABG - *** *** * *** *** *** *** *** *** *** 
ABA 0.023 - *** *** *** *** *** *** *** *** * 
AFA 0.040 0.033 - *** *** * *** *** *** * *** 
CEH 0.007 0.013 0.025 - *** * *** *** ns ** *** 
GUM 0.024 0.023 0.034 0.022 - *** *** *** *** *** *** 
HAH 0.022 0.009 0.007 0.005 0.024 - *** ** *** ns *** 
KEF 0.025 0.015 0.033 0.015 0.027 0.014 - *** *** *** *** 
LES 0.026 0.008 0.018 0.018 0.015 0.007 0.020 - *** ns * 
NWH 0.010 0.008 0.025 0.002 0.016 0.008 0.008 0.011 - * *** 
SES 0.022 0.013 0.010 0.011 0.018 -0.001 0.021 0.003 0.005 - * 
WGJ 0.024 0.008 0.023 0.025 0.016 0.013 0.015 0.007 0.013 0.010 - 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, KEF- Keffa, LES ? Long-eared Somali, NWH ? North-West 
Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji. 
 160
 Table 34. Pair wise test of population differentiation between the 11 Ethiopian and 1 reference population.  FST values (below 
the diagonal) and their test of significance (above the diagonal) obtained after 66,000 permutations. 
Pop ABG ABA AFA CEH GUM HAH KEF LES NWH SES WGJ ABG 
ABG - *** *** * *** *** *** *** ** *** *** *** 
ABA 0.023 - *** *** *** ** *** *** *** *** * *** 
AFA 0.040 0.033 - *** *** * *** *** *** * *** *** 
CEH 0.007 0.013 0.025 - *** * *** *** ns ** *** *** 
GUM 0.024 0.023 0.034 0.022 - *** *** *** *** *** *** *** 
HAH 0.022 0.009 0.007 0.005 0.024 - *** * *** ns ** *** 
KEF 0.025 0.015 0.033 0.015 0.027 0.014 - *** *** *** *** *** 
LES 0.026 0.008 0.018 0.018 0.015 0.007 0.020 0 *** ns * *** 
NWH 0.010 0.008 0.025 0.002 0.016 0.008 0.008 0.011 - * *** *** 
SES 0.022 0.013 0.010 0.011 0.018 -0.001 0.021 0.003 0.005 - * *** 
WGJ 0.024 0.008 0.023 0.025 0.016 0.013 0.015 0.007 0.013 0.010 - *** 
GUB 0.142 0.130 0.134 0.143 0.137 0.130 0.135 0.132 0.142 0.129 0.110 - 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, KEF- Keffa, LES ? Long-eared Somali, NWH ? North-West 
Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji,, GUB ? Guinea Bissau. 
 161
 Table 35. Summary of genotypic differentiation tests for all the 55 population 
combination pairs of Ethiopian populations for all the loci. 
Locus 
Percentage of population pairs that could be distinguished by the  
marker, P < 0.05 
BM1818 20.00 
BMS1494 50.91 
ILSTS05 36.36 
ILSTS11 72.73 
ILSTS44 20.00 
ILSTS87 50.91 
INRA05 40.00 
INRA63 29.09 
INRA132 34.55 
MAF35 3.60 
MAF65 69.09 
MAF209 34.55 
OarAE129 30.91 
OarFCB304 63.64 
SRCRSP03 60.00 
Average for all population 41.21 
 162
 4.10.3. Analysis of Isolation by Distance (IBD) 
IBD is a fast and simple application for analyzing (genetic) isolation by 
(geographic) distance.  The genetic similarity or differentiation among individuals or 
populations can be ascertained using a number of statistical techniques (re-viewed by 
Bohonak, 1999; Neigel, 1997; Roderick, 1996; Slatkin, 1985).  When populations can be 
defined a priori, one option is to analyze genetic ?isolation by distance? (Wright, 1943) by 
plotting the genetic similarity (or distance) among population pairs as a function of the 
geographic distance between those pairs. 
Qualitative and statistical analyses of isolation by distance can reveal much about 
population genetic structure.  The primary use for plots of (genetic) isolation by 
(geographic) distance is to assess whether more distant population pairs are more different 
genetically.  However, these plots can also be used to test the validity of simpler models of 
population structure (e.g. island or hierarchical island models).  Isolation by distance 
analyses may help separate the effects of population history from ongoing gene flow, and 
test the explanatory power of alternative dispersal pathways and the influence of 
geographic features or specific life-history traits on population differentiation can also be 
tested (Bohonak, 2002).  
Studies of isolation by distance typically seek to ascertain (i) whether there is a 
statistically significant relationship between genetic distance (similarity) and geographic 
distance, and (ii) the strength of this relationship.  Significance is usually assessed by 
asking whether the pair-wise genetic distance matrix is correlated with the pair-wise 
geographic distance matrix using a Mantel test (Manly, 1994).  For the genetic distance 
matrix A and the geographic distance matrix B, the test statistic is calculated as Z = 
Zi,jAijBij.  IBD also reports an alternative statistic, r, which provides a standardized Z that 
ranges from -1 to 1 (Manly, 1994).  Significance is assessed by comparing Z actual to a 
distribution of Z scores obtained by randomizing rows/columns of the B matrix and holding 
A constant.  The IBD application provides one-tailed P-values for this distribution.  
A logical way to quantify the strength of the isolation by distance relationship is to 
calculate the slope and intercept of genetic similarity or distance against geographic 
distance. Based on simulations, Hellberg (1994) suggested that reduced major axis (RMA) 
 163
 regression is more appropriate for this purpose than standard ordinary least squares (OLS) 
regression. 
The Ethiopian goat populations are found in a wide range of production systems all 
over the country (Figure 19) and the genetic differentiation may not have resulted from 
geographic isolation.  The correlation between genetic distance FST/(1-FST) of Rousset 
(1997) and geographic distance (km) among the Ethiopian goat populations was assessed 
using Mantel test (Manly, 1986, 1994) with 100,000 randomizations.  The test was 
implemented using a computer program IBD (Bohonak, 2002; available at 
http://www.bio.sdsu.edu/pub/andy/ibd.html).  FST distances were calculated with computer 
program FSTAT version 2.9.3 (Goudet, 2001) according to Weir and Cockerham (1984). 
 
0
 200
 400
 600
 800
 1000
 1200
 -0.01 0 0.01 0.02 0.03 0.04 0.05
 Gentic distance 
G
 eo
 gr
 ap
 hi
 c 
D
 is
 ta
 nc
 e 
(k
 m
  
Figure 44. Genetic differentiation among 11 Ethiopian goat populations.  Multi-locus 
estimates of pair-wise genetic distance (FST) is plotted against geographic distance (km).  
The regression is Y = 4723.4x + 431.05, Z = 475.97, r = 0.206, P < 0.1065.  FST was 
calculated according to Weir and Cockerham (1984). 
 
Figure 44 illustrates the correlation between genetic distance FST/(1?FST) and 
geographic distance (km) among Ethiopian goat populations for different geographic 
regions.  The geographic distance among Ethiopian goat populations ranges between 96.75 
km to 993.31 km, and no significant correlation was detected between genetic distance and 
geographic distance at the entire geographic regions. 
 164
 4.10.4. Analysis of molecular variance (AMOVA)  
In order to understand the partitioning levels of the genetic diversity of the 
Ethiopian goats, an analysis of molecular variance (AMOVA) was conducted.  The result 
of the AMOVA (Table 36) revealed that 98.36% of the total genetic diversity existed 
among the individuals within populations and only 1.64% of the total genetic diversity 
accounted for differences among populations.  
 
Table 36. Analysis of molecular variance (AMOVA) for 11 Ethiopian goat populations, 
using genotype data from 15 microsatellite loci.  
Source of variation Degree of freedom Sum of squares Variance  components 
Percentage  
variation 
Among populations 10 104.527 0.07109 1.64 
Within populations 947 4037.97 4.26396 98.36 
 
 
4.10.5. Individual specific analysis  
Individuals were assigned to source populations using the Cornuet et al. (1999) 
GeneClass2 program, which allows different estimation procedures including the Bayesian 
(Rannala and Mountain, 1997), and frequency (Paetkau et al., 1995) as well as six distance 
methods; the Nei?s DA (Nei et al., 1983), Cavalli-Sforza and the Edwards chord distance 
(1967), the Nei Minimum distance (Nei, 1973), the Nei DS (Nei, 1972), the  Jin and 
Chakraborty?s (1993) DAS and the ??2 distance method of Goldstein et al. (1995). 
The efficiency of the different methods to assign individuals to their source 
population was tested and the results are shown in Table 37.  In all cases the percentage of 
individuals being correctly assigned to their source populations ranged from the least 
13.57% with Goldstein et al. (1995) ??2 method to the best (in this case) 35.07% with the 
Bayesian method. 
The percentage of correct assignment of individuals to their source populations 
using a single locus was very low ranged from 9.19% for ILSTS05 to 15.87% for ILSTS87 
(Table 38).  However, when the five loci with the highest (in this case) individuals i.e. 
ILSTS87, MAF65, BMS1494, ILSTS11 and OarAE129, were used together the percentage 
 165
 of individuals correctly assigned was increased to 28.39% (Table 38).  Increasing the 
number of microsatellite loci to 15 resulted in correct assignment of only 35.07% of the 
individuals.  
Table 39 shows the results of the population assignment test obtained using 
frequency data from the 11 Ethiopian populations.  Only 35.07% of the 479 individuals 
were assigned to the corrected source population.  The population with the highest level of 
correct assignment was the Gumez with 55.81% of the individuals being correctly assigned, 
followed by the Keffa population with 47.37% of the individuals being correctly assigned 
to the source population.  In general all the Ethiopian populations had low levels of correct 
assignment with the Hararghe Highland population having the lowest level of assignment 
(7.50%). 
 
 
Table 37. The accuracy of different methods to assign individuals to the source 
population. 
Methods Software % Correct assignment 
Bayesian GeneClass 35.07 
Maximum likelihood WHICHRUN 33.90 
Frequency GeneClass 33.19 
The Nei et al. (1983) DA GeneClass 32.36 
Cavalli-Sforza and the Edwards chord distance GeneClass 31.52 
The Nei standard distance (DS) GeneClass 27.56 
Nei Minimum distance GeneClass 26.72 
The Chakraborty and Jin DAS  GeneClass 25.05 
The Goldstein et al. (1995) (d?)2 GeneClass 13.57 
 
 
 
 
 
 
 
 
 
 
 166
 Table 38. Percent of correct individual-population assignment (only for Ethiopian 
populations) estimated for each microsatellite locus using the Bayesian method.  
Locus Percent of correct assignment 
ILSTS87* 15.87 
MAF65* 15.66 
BMS1494* 15.45 
ILSTS11* 15.45 
OarAE129* 15.24 
INRA63 14.82 
ILSTS44 14.61 
OarFCB304 14.61 
INRA132 13.78 
SRCRSP03 13.78 
MAF209 13.57 
BM1818 12.94 
INRA05 9.81 
MAF35 9.81 
ILSTS05 9.19 
*When these five loci taken together 28.39 
 
 
 167
 Table 39. Frequency of correct assignment of individuals to the 11 Ethiopian populations using genotype information from 15 
microsatellite loci using the Bayesian method. 
Populations No. ABG ABA AFA CEH GUM HAH KEF LES NWH SES WGJ 
ABG 46 45.65 10.87 0.00 6.52 0.00 2.17 4.35 4.35 15.22 2.17 8.70 
ABA 46 6.52 36.96 2.17 4.35 0.00 4.35 6.52 10.87 2.17 15.22 10.87 
AFA 44 4.55 0.00 43.18 2.27 6.82 18.18 2.27 2.27 2.27 11.36 6.82 
CEH 40 5.00 12.50 2.50 32.50 5.00 10.00 10.00 0.00 15.00 7.50 0.00 
GUM 43 2.33 2.33 0.00 4.65 55.81 4.65 6.98 2.33 9.30 0.00 11.63 
HAH 40 2.50 12.50 20.00 2.50 7.50 7.50 7.50 7.50 5.00 20.00 7.50 
KEF 38 2.63 0.00 2.63 2.63 10.53 2.63 47.37 5.26 10.53 7.89 7.89 
LES 48 6.25 6.25 6.25 6.25 6.25 8.33 2.08 31.25 0.00 18.75 8.33 
NWH 45 13.33 6.67 6.67 13.33 2.22 2.22 4.44 8.89 28.89 4.44 8.89 
SES 43 2.33 4.65 11.63 4.65 4.65 11.63 4.65 11.63 11.63 23.26 9.30 
WGJ 46 2.17 10.87 4.35 0.00 2.17 8.70 10.87 8.70 8.70 10.87 32.61 
No. = number of animals; ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, KEF- Keffa, LES ? Long-eared 
Somali, NWH ? North-West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji. 
 
 
 
 
 
 
 168
 4.10.6. Marginal loss of genetic diversity 
Weitzman (1992, 1993) has suggested a method that uses genetic and non-
 genetic information to calculate a maximum-likelihood tree and the current diversity of 
a group of species and to Asses the expected change in diversity over a certain time 
horizon.  This method allows evaluation of the change in the group?s diversity when the 
degree of endangerment of one or several species is reduced (marginal diversity).  
Compared with other approaches (Crozier, 1992; Faith, 1992) mostly descriptive and 
illustrative, the Weitzman method provides a well-defined diversity metric, allowing 
quantitative assessment of the consequences of alternative conservation activities as an 
aid to decision-making in conservation.  An application to livestock breeds was 
presented by Thaon d?Arnoldi et al. (1998), and the approach was used in studies on 
European cattle breeds (Bremond, 2001; Ca?on et al., 2001), on African cattle breeds 
(Reist-Marti et al., 2003) and European pig breeds (Laval et al., 2001). 
Application of the Weitzman approach showed that removal of a random 
population from the total Ethiopian goat populations would result in a relative loss of 
genetic ?diversity? between 4 and 24%. 
The marginal loss of genetic diversity after a removal of Afar, Abergalle, Gumez 
and Keffa populations from the total set were 24.32%, 19.22%, 16.59%, and 12.99%, 
respectively (Table 40).  Removal of other populations had a lower impact being (Arsi-
 Bale; -6.23%; Woyto-Guji; -6.08%; Central Highland and North-West Highland 
combined; -5.33%; Hararghe Highland and Short-eared Somali combined; -4.73%; 
Long-eared Somali; -4.13% only). 
 
 
 
 
 
 
 
 
 
 
 
 169
 Table 40. Marginal loss of genetic diversity according to the Weitzman approach. 
FST V(S) = 0.133200 
i V(Si) ?V = V(Si) ? V(S) %V = ?V/V(S) 
AFA 0.1008 -0.0324 -24.32 
ABG 0.1076 -0.0256 -19.22 
GUM 0.1111 -0.0221 -16.59 
KEF 0.1159 -0.0173 -12.99 
ABA 0.1249 -0.0083 -6.23 
WGJ 0.1251 -0.0081 -6.08 
CEH&NWH 0.1261 -0.0071 -5.33 
HAH&SES 0.1269 -0.0063 -4.73 
LES 0.1277 -0.0055 -4.13 
i   population. 
V(S) Genetic diversity of the total set of populations. 
V(Si) Genetic diversity of the total set, after removal of population i. 
AFA- Afar, ABG- Abergalle, GUM- Gumez, KEF- Keffa, ABA- Arsi-Bale, WGJ- Woyto-Guji, CEH ? Central 
Highland, NWH ? North-West Highland, HAH ? Hararghe Highland, SES ? Short-eared Somali, LES- Long-eared 
Somali. 
 170
 CHAPTER FIVE 
5.0. DISCUSSION 
This is the first time that indigenous goat genetic resources of Ethiopia have 
been studied in details at molecular level using microsatellite loci.  These results 
provide important information for the future management, breed genetic improvement 
and conservation of Ethiopian goat genetic resources. 
The classification and characterization of livestock species/type is necessary 
before desired genetic materials can be preserved for future generations.  Much genetic 
variability, the raw material of breed genetic improvement has been lost in developed 
countries.  There is no doubt that preserving genetic diversity, for either immediate or 
future use, is important. 
 
5.1. Within population genetic diversity 
Genetic diversity is a fundamental component of biodiversity, forming the basis 
of species and ecosystem.  It represents all of the genetically determined differences that 
occur between individuals of a species in the expression of a particular trait or a set of 
traits.  Since species are composed of populations that exist somewhat independently 
each other, genetic diversity exists both within and among populations of species. 
A total of 159 alleles were detected at 15 microsatellite loci in 776 goat 
individuals of 20 populations sampled from 20 locations in Ethiopia, Kenya, Botswana, 
Guinea Bissau, Egypt, Turkey, Italy, Saudi Arabia and Mongolia.  The total number of 
alleles sampled in a population across all loci ranged between 80 in West African Dwarf 
(WAD) goat population from Guinea Bissau and 116 in Long-eared Somali goat 
population from Ethiopia.  The mean number of alleles per locus (MNA) across all 
populations was 5.46 ? 0.158, varying between 4.67 ? 2.74 in Guinea Bissau WAD goat 
population and 6.47? 4.03 in Long-eared Somali goat population.  The MNA for a 
uniformed sample size with 20 animals after random re-sampling procedure with 250 
replacements over individuals was 4.88 ? 0.08, varying between 4.19 ? 2.26 in Guinea 
Bissau population and 5.74 ? 2.89 in Hair goat population.  In general, the mean 
numbers of alleles were comparable among all the populations from different 
geographic regions.  The sample size of each population ranges from 22 individuals in 
Baladie population from Egypt to 48 individuals in Long-eared Somali population from 
Ethiopia.  The number of alleles at each locus in the Ethiopian populations ranged from 
 171
 4 to 23, which was slightly higher than that of the Swiss goat breeds (3 to 19) 
(Saitbekova, 1999), Asian goat populations (5 to 18) (Barker, 2001) and Chinese goat 
populations (4 to 19 and 4 to 13) (Li et al., 2002; Li and Valentini, 2004). 
The mean observed heterozygosity (HO) and expected heterozygosity (HE) for 
Ethiopian goats was 0.551 ? 0.015 and 0.576 ? 0.015, respectively, across all loci and 
all population.  The observed heterozygosity across all loci ranged between 0.515 ? 
0.047 in Ethiopian Afar population to 0.625 ? 0.050 in Saudi Arabian Ardi population.  
The highest expected heterozygosity (gene diversity) across all loci was observed in 
Hair goat population from Turkey (0.642 ? 0.055) and the lowest was 0.523 ? 0.058 in 
WAD (Guinea Bissau) population.  The differences of both values among the goat 
populations were not statistically significant (P > 0.05), and it also seemed that the 
values were not affected by sample size. 
The mean number of alleles per locus (MNA) and the expected heterozygosity 
(HE) are two fundamental parameters in the population genetics commonly used to 
quantify the genetic variability within populations.  The level of genetic diversity (HE = 
0.58 ? 0.015) obtained for the Ethiopian goat populations in the present study was 
similar to that previously reported in the Swiss goat breeds (HE = 0.59, Saitbekova et 
al., 1999), and Asian goats (HE = 0.52, Barker et al., 2001), but less than that reported 
for Chinese goats (HE = 0.716, Li et al., 2002; and HE = 0.707, Li and Alessio, 2004). 
The genetic diversity of the Ethiopian goat populations was not high, as 
indicated by the moderate MNA and expected heterozygosities observed for those 
populations.  The average number of alleles per locus was 8.87 (ranged from 4 to 23) 
for the 11 Ethiopian goat populations at the 15 microsatellite loci used in the study.  The 
results of the AMOVA showed that most of the diversity for the Ethiopian goat 
populations is found within populations, rather than between populations (Table 36). 
Seven Ethiopian goat populations showed a significant inbreeding coefficient 
(FIS) (P < 0.05 for Abergalle, Hararghe Highland, Long-eared Somali, North-West 
Highland and Woyto-Guji populations, and P < 0.01 for Afar and Central Highland 
populations), and none of the Ethiopian goat populations showed out-breeding effect.  
This inbreeding could be a consequence of the management systems (like tethering of 
the animals in the case of small flock size during crop season, herding of the flock by 
children and/or women, where the flock size is larger and some localized transhumance 
during the dry season), traditional/cultural practices (like preference for a specific coat 
colour at different cultural or religious ceremonies) and also ethnic/cultural barriers that 
 172
 could affect the between flock breeding-males(bucks) exchange.  Tethering minimises 
the number of different bucks that any particular female (doe) can mate with, leading to 
service by only males those preferred by the owner; while women and children herders 
usually avoid the mixing of different flocks or exchange of bucks from different areas 
even at the watering point.  Localised transhumance and ethnic/cultural barriers also 
lead to minimal encounters between different flocks.  In case each sampled 
populations/types have been subdivided into subpopulations, which contains fewer 
heterozygotes than predicted despite the fact that all subdivisions are in Hardy-
 Weinberg equilibrium, can also lead to inbreeding (Wahlund effect).  All these systems 
and events tend to increase the rate of inbreeding. 
 
5.2. Between population genetic diversity 
Genetic differentiation between the goat populations studied was mainly resulted 
from the genetic drift other than mutations at allele sizes among populations.  Wright?s 
fixation indices or F-statistics (Wright, 1969) are commonly used for describing genetic 
differentiations among populations.  The unbiased estimators of F-statistics (FIT, FST 
and FIS) were calculated for each locus across all populations with method of Weir and 
Cockerham (1984), and their significance levels were tested by permutation procedures.  
FIS statistic, an estimator for describing local inbreeding effect within subpopulations, 
was significantly different from zero at six out of 15 loci (BM1818 and MAF209 at P < 
0.05; ILSTS05 at P < 0.01; ILSTS44, INRA63 and MAF65 at P < 0.001) and across all 
loci at P < 0.001 level, indicating that the Ethiopian goat populations sampled for this 
study were generally at risk of inbreeding.  The FIT estimator, describing the inbreeding 
as well as the subdivision in the total population, was significantly different from zero at 
eight out of 15 loci (BM1818, OarAE129 and SRCRSP03 at P < 0.05; ILSTS05 and 
MAF209 at P < 0.01; ILSTS44, INRA63 and MAF65 at P < 0.001) and across all loci 
at P < 0.001 level.  As local inbreeding effect was detected only at six of 15 loci, 
significant FIT statistic at these loci was merely resulted from the subdivision among 
goat populations.  FST estimator is commonly called the coefficient of gene 
differentiation (analogue to GST of Nei, 1973) and used to measure the degree of genetic 
subdivision among populations.  FST estimator was significantly different from zero at 
14 out of 15 loci (BM1818 and ILSTS44 at P < 0.01; BMS1494, ILSTS05 ILSTS11, 
ILSTS87, INRA05, INRA63, INRA132, MAF65, MAF209 OarAE129, OarFCB304 
 173
 and SRCRSP03 at P < 0.001) and across all loci at P < 0.001, which means that 
genotype frequencies among populations are not randomly distributed, meaning that all 
the Ethiopian goat populations have been genetically subdivided even though the mean 
FST value across all loci was very low (0.016) compared to previously reported in Asian 
goats (FST = 0.143, Barker et al., 2001) and Chinese goats (FST = 0.105, Li et al., 2002; 
and FST = 0.054, Li and Valentini, 2004).  Similar value was obtained for the coefficient 
of gene differentiation using the method of Nei (1973) (GST = 0.026).  Both values of 
FST and GST indicate that the level of gene differentiation between Ethiopian goat 
populations was very low.  Only 1.6% to 2.6% of total variation was contributed by 
genetic subdivision among populations while 97.4% to 98.4% of total variation 
accounts for genetic variations among individuals within populations.  This figure went 
up to 6.4% (FST) and 7.4% (GST) when all the 20 goat populations are included in the 
analysis.  These differentiation values are much lower compared with those generally 
recorded in goats and other domestic animal populations [e.g. 9.9% between Spanish 
dog breeds (Jordan et al., 1992), 10% between European cattle (MacHugh et al., 1998), 
17% between Swiss goat breeds (Saitbekova et al., 1999), 8% between Spanish horses 
(Canon et al., 2000), 10.5% between Chinese indigenous goat populations (Li et al., 
2002), and as well as 7.7% between Chinese indigenous pig breeds (Yang et al., 2003)].   
Although the magnitude of gene differentiation among populations was low, the 
pair-wise genetic differentiation between the majorities (52 out of 55 pair-wise 
combinations) of Ethiopian goat populations was significant at P < 0.001, 0.01 or 0.05 
levels (Table 33 and Table 34).  There was no genetic differentiation (P > 0.05) 
between three pair-wise combinations of Ethiopian goat populations.  These populations 
(Central Highland/North-West Highland, Hararghe Highland/Short-eared Somali, and 
Long-eared Somali/Short-eared Somali) are located geographically neighbouring to 
each other (Figure 19), therefore they seem to have experienced frequent gene flow 
between each other.  The Ethiopian goats are a fairly agile species that has developed 
adaptations, e.g. long legs that enable them to walk long distances.  This could have also 
facilitated this genetic exchange between the goats from different locations within the 
country.  Also the large difference of FST values observed between the goat populations 
from Ethiopia compared to those from other countries may well be a consequence of the 
complexity and diversity of the eco-geographic and sociological barriers, and recent 
human artificial selection, which may differ in the case of different developed countries.  
 174
 The highest number of migrants per generation was estimated between Short-
 eared Somali and Hararghe Highland, where it shows that these two populations are not 
genetically different at all, followed by Nm = 120.88 between North-West Highland and 
Central Highland populations, and followed by Nm = 82.19 between Short-eared Somali 
and Long-eared Somali.  The lowest number of migrants was estimated between Afar 
and Abergalle populations (Nm = 6.06).  As expected, gene flow was low between 
populations located in distant geographic regions.  For instance, the Nm was 7.04 
between Gumez and Afar, 7.38 between Keffa and Afar and 9.52 between Long-eared 
Somali and Abergalle populations.  In the case of Afar and Abergalle populations, even 
though they are not geographically distant from each other, there is a highland topology 
between the two respective lowland areas, which keeps them isolated.  
A clear population structure was inferred in all goat populations considered in 
this study by using a multivariate analysis (PCA Figure 26 and Figure 27).  All 20 goat 
populations sampled in this study were largely classified into three major clusters.  All 
the Ethiopian plus two Kenyan populations form one cluster; Italian, Mongolian, Saudi 
Arabian, Egyptian and Botswana form another cluster, and the Guinea Bissau alone 
form the third cluster.  The genetic variation within and among these groups were 
further quantified by a hierarchical analysis of molecular variance (AMOVA), and the 
results showed that 89.99% to 91.15% of genetic variation accounted for variation 
within populations, 1.59% to 2.82% for variation among populations within groups, and 
7.20% to 7.32% for variations among groups (Table 16).  The genetic variation among 
populations within groups was the lowest when grouped according to their country of 
origin into nine populations (1.59%) and the highest when all the 20 goat populations 
were considered together as one group (6.38%).  The higher genetic variation among 
groups was observed between groups according to their country of origin (nine groups 
7.32%), and the lower was between group of three (all Ethiopian plus two Kenyan 
populations, Italian, Mongolian, Saudi Arabian, Egyptian and Botswana populations, 
and the Guinea Bissau populations 7.20%). 
When the AMOVA is conducted only for the Ethiopian goat populations, the 
results showed that 98.36% of genetic variation accounted for variation within 
populations and only 1.64% for variation among populations. 
 
 175
 5.3. Genetic relationship between goat populations 
The phylogenetic relationship between the goat populations were determined 
using Nei?s standard genetic distance (DS, Nei, 1972) and Nei et al.,?s DA genetic 
distance (Nei, 1983) (Table 10 and Table 11).  Phylogenetic reconstruction was 
performed using NJ (Saitou and Nei, 1987) and UPGMA method (Sneath and Sokl, 
1973) to summarize the genetic distances between goat populations.  The UPGMA 
method with both Nei?s standard genetic distance (DS, Figure 22) and Nei et al.?s DA 
genetic distance (Nei, 1983) (Figure 23); and the DA NJ basically classified the goat 
populations into three clear phylogenetic clusters concordant with the population 
structure inferred from the admixture analysis (STRUCTURE, Figure 29 and Figure 
30).  As expected that different populations from geographically distinct regions, all 
Ethiopian goat populations (together with the two neighboring Kenyan populations) and 
the other seven populations (from West Africa, South Africa, North Africa, Turkey, 
Saudi Arabia, Italy and Mongolia) clearly emerged as distinct lineages different from 
the West African (Guinea Bissau) standing completely alone (with relatively large 
genetic distance), suggesting different origins of ancestral populations.  These 
population structures correspond likely to their geographic distributions.  A very similar 
pattern of genetic relationship among the goat populations was observed by principal 
component analysis (PCA).  In general, all the Ethiopian goat populations were more 
closely related to each other than to non-Ethiopian counterparts.  The two exceptions are 
the Kenyan populations, which are in a cluster with the Ethiopian groups.  The two 
populations were sampled from the northern Kenya close to the boarder between 
Ethiopia and Kenya, suggesting a common ancestry.  All the reference populations 
(except the two Kenyan populations and the WAD goat population from Guinea Bissau) 
cluster closely together suggesting a different ancestry.  The close clustering of the 
Ethiopian populations with the two Kenyans ones is in accordance with the work of 
Chenyambuga (2002). (Nei, Tajima, and Tateno, 1983) 
When the Ethiopian goat populations are treated with one out-group (WAD, 
from Guinea Bissau) or alone for the above analysis, the UPGMA method with Nei?s 
standard genetic distance (DS,) basically classified the goat populations in to eight 
phylogenetic clusters (Figure 35 and Figure 37), while the DA genetic distance for both 
with and without out-group population classified them differently (Figure 36 and 
Figure 38).  Both the NJ trees from DS and DA distances for both with and without out-
 176
 group population were not consistent with their clustering (Figure 31 to Figure 36).  As 
one origin of evolution was expected for Ethiopian goats, it seemed that the UPGMA 
performs better and consistent compared to the NJ tree methods.  The UPGMA assumes 
an equal rate of evolution between lineages unlike the neighbor-joining method (Nei et 
al., 1983). 
Comparison of the genetic diversity of the Ethiopian goat populations with that 
of goats from West Africa, South Africa, Mongolian, Saudi Arabian, Egypt, Italy and 
Turkey indicated that the Ethiopian goat populations are less genetically diverse (Table 
10 and Table 11).  This could be the result of being geographically very close to each 
other, much animal movement across the regions (zones & districts) and in some cases 
the results of strong human selection pressure for certain economic traits.  
The Ethiopian goats, either as separated populations or when grouped as one, 
were found to have similar diversity with that of the non-Ethiopian populations in terms 
of MNA per locus and expected heterozygosity (Table 9). 
There were no significant deviations from HW equilibrium expectations in all 
the populations studied, also the 15 loci analyzed may be assumed to be unlinked, as 
linkage disequilibrium did not yield any significant comparison for locus pairs across all 
populations.  The slightly higher than expected (P < 0.05) deviation from linkage 
equilibrium observed in the Arsi-Bale, Guinea Bissau, Long-eared Somali, Italian 
Alpine Mongolia, Boran-Galla, Short-eared Somali, Afar, Small-East Africa and 
Woyto-Guji samples could indicate population subdivision, recent introduction of a 
non-random subset of genotypes or habitat-specific selection (King et al., 2001). 
The assumptions of the sign test, and the consequences of violating the 
assumptions, have been discussed by Cornuet and Luikart (1996).  It is worth reiterating 
here that, in test for bottlenecks, loci that are not in Hardy-Weinberg proportions should 
be used only with caution because they could bias the test results.  For example, a locus 
could deviate from Hardy-Weinberg proportions by having an excess of heterozygote 
due to strong over-dominance selection.  Such a locus might also have selection-
 induced heterozygosity excess and, therefore, should be used with caution in the sign 
test for heterozygosity excess.  For the data sets studied here, excluding loci that were 
not in Hardy-Weinberg proportions (Table 4 andTable 5) did not change the results of 
the sign test.  Generally, the Wilcoxon test didn?t detect any recent bottleneck effect in 
the studied Populations.  But if there was any undetected bottleneck effect, it should be 
either not very severe or not very recent. 
 177
 The present classification of Ethiopian goats is based on physical description 
and management system.  Results of the present study using autosomal microsatellites 
suggest a close relationship of Ethiopian goat populations.  It appears that all Ethiopian 
goat populations were seen to cluster more closely together, suggesting a common 
origin and that there is extensive gene flow between the North-West Highland/Central 
Highland, Short-eared Somali/Hararghe Highland and Long-eared Somali/Short-eared 
Somali populations.  The Tswana from Botswana was seen to cluster with the Italian 
Alpine, suggesting that crossbreeding between the local Tswana population and the 
exotic Alpine from Italy has been done or undergoing for improved milk production 
purpose.  The WAD population from Guinea Bissau was always clustered alone, 
suggesting a different origin from the Ethiopian populations or a differently developed 
population as trypanotolerant dwarf goats, which agrees with the suggestion of Wilson 
(1991) that the West African Dwarf (WAD) goats arose in the Fouta Djallon region of 
Guinea, as did the trypanotolerant N?Dama cattle.  The other breeds (Italian Alpine, 
Baladie, Hair, Mongolia and Ardi) cluster together, suggesting most probably a 
common ancestry for them. 
Given that no true wild goats in the strictest sense occur in Africa, even though 
Muzzolini (2000) asserts on the presence of goat progenitors in Africa which still needs 
to be demonstrated, it is generally agreed that the first goat came into Africa via Egypt 
(Epstein, 1971), this possibly through the Gulf of Suez (Close, 2002).  It is also believed 
that prior to 5th millennium BC goats had reached Egypt (Mason, 1984) and from there 
they dispersed southwards along the River Nile and westward to what is today the 
Sahara region.  The principal component analysis of our study (Figure 26 and Figure 
27) showed that the Ethiopian and Kenyan breeds cluster with the West African Guinea 
Bissau population, which might be an indication that the Eastern and Western Africa 
goats crossed the straits of Bab-el-Mandeb and entered the continent of Africa via the 
horn of Africa.  Here, the Baladie population from Egypt is consistently clustered with 
the European and Asian populations, which agrees with Epstein (1971).  Additional 
archaeological data, including more accurate dating of archaeological goat remains, is 
required to assess these theories. 
The total variation accounted for by the first, second and third principal 
components, following the principal component analysis of the Ethiopian goat 
populations, was 58.68%.  This figure increased to 60.23% when the reference 
populations were included in the analysis.  This variation (58.68%) was slightly lower 
 178
 compared to that observed among twelve Chinese indigenous goat populations (60%, Li 
et al., 2002) and for sub-Saharan Africa camel populations (72%, Mburu et al., 2003), 
but higher compared to that observed for sub-Saharan Africa cattle breeds (54%, 
Hanotte et al., 2002). 
Our microsatellite based molecular genetic study doesn?t support the distinction 
of eleven Ethiopian goat populations as indicated in the study published in 1996 using 
morphological traits (Farm-Africa, 1996).  Instead, the results indicate eight separate 
genetic entities: the Arsi-Bale, Gumez, Keffa, Woyto-Guji, Abergalle, Afar, Highland 
goats (previously separated as Central and North West Highland) and Eastern and 
Southeastern goats (previously separated as Hararghe Highland, Short-eared Somali and 
Long-eared Somali goats).  This shows that how the principal component analysis goes 
with the quantitative traits like body weight, height etc. but doesn?t go with the 
qualitative traits like horn shape, ear length.  This might be the reason that the results of 
molecular characterization are different from phenotypic characterization. 
Weitzman analysis indicated that about 75% of the total genetic diversity of the 
Ethiopian goats is present in four breeds; Afar, Abergalle, Gumez and Keffa, with 
marginal loss of diversity of 24.32%, 19.22%, 16.59% and 12.99%, respectively, while 
removal of other populations had only a lower impact (Arsi-Bale; -6.23%; Woyto-Guji; 
-6.08%; Central Highland and North-West Highland combined; -5.33%; Hararghe 
Highland and Short-eared Somali combined; -4.73%; Long-eared Somali; -4.13%). 
 179
 CHAPTER SIX 
 
6.0 SUMMARY AND CONCLUSIONS 
6.1 Summary 
Ethiopia has the largest livestock population among all the countries in Africa.  
Ethiopia with its great variation in agro-ecological zones represents a potential reservoir 
of sheep and goat diversity.  They inhabit a wider range of environments extending 
from a tropical to cool temperate climates.  According to FAO (1994), there are 
approximately 21.7 and 16.7 million sheep and goats, respectively.  Sheep and goats are 
distributed in all agro-ecological zones of the country although the majority of the sheep 
population is concentrated in the highlands.  The majority of the goat population is 
found in large flocks in the arid and semi-arid lowlands where pastoralists in the south, 
east, and west keep them for milk production, for slaughter and for sale.  Goats in the 
highlands are widely distributed in the mixed crop-livestock production systems with 
small flock size as a means of cash earnings and meat. Although there are severe 
environmental constraints to increase goat productivity, there is considerable potential 
for improved goat production in the country, where goat milk, meat, and skin are valued 
commodities.  
Ethiopia has been long back recognized as a source of genetic diversity in plants 
and animals.  It appears that Ethiopia has served as a gateway of domestic animal 
genetic materials from Asia to Africa, and its diverse ecology leads to further 
diversification and development of different genotypes. 
The identification, description, classification and naming of livestock types are 
the first step in the assessment of a country?s livestock resources.  
Work done in 1990s on physical description and management system has shown 
the presence of 14 goat types in Ethiopia and Eritrea (Workneh, 1992; Alemayehu, 
1993; Nigatu, 1994), out of which 11 are in Ethiopia.  This survey was the very first 
step in breed characterization of Ethiopian goats through differentiating a heterogeneous 
population into relatively homogeneous sub-populations, which may then be named as 
breeds, types or subtypes.  However, it remains unclear how different at the genetic 
level are these 11 goat types.  The availability of new molecular techniques for large-
 scale genetic characterization is providing unique opportunities to tackle the above 
 180
 issues.  These techniques are able to reveal genetic distances between populations 
allowing a better understanding of how these populations evolved. 
In recent years, a range of innovations in molecular genetics has been developed 
for the study of genetic variation and evolution of populations using DNA genotyping 
information.  The most utilized DNA marker for population genetics of livestock is 
microsatellite.  Microsatellite markers, also called short tandem repeats (STRs) or 
simple sequence repeats (SSRs), are a relatively new class of genetic marker.  Over a 
few years they have become a tool of choice to address population genetics and 
demographic questions (Sunnucks, 2000).  They allow the study of genetic diversity and 
differentiation of closely related populations.  Microsatellite, as genetic markers, have 
been applied successfully in the study of genetic variation of livestock including 
between and among European and African cattle breeds (Machugh et al., 1997; Okoma 
et al., 1998), between Swiss goat breeds (Saitbekova et al., 1999), between Indian goat 
breeds (Ganai and Yadav, 2001), between Chinese indigenous goat populations (Li et 
al., 2002, Li and Valentini 2004) and between Indian buffalo populations (Arora R et 
al., 2004,).  Unfortunately, for goats, throughout the world very limited work has been 
done to determine genetic variation among breeds/populations and in fact, information 
on microsatellite polymorphism of Ethiopian goats is completely lacking.  
Accordingly the overall goal of this project is to obtain detailed information on 
the genetic diversity and differentiation of the indigenous goat populations of Ethiopia 
with the following specific objectives: 
1. To quantify genetic diversity within and among Ethiopian goat populations. 
2. To clarify the evolutionary genetic relationships among the Ethiopian goat 
populations. 
3. To clarify the origin and relationships of Ethiopian goat populations through 
the study of non-Ethiopian breeds of references. 
4. To compare the morphological classification of the Ethiopian goat 
populations with the one based on molecular genetic characterization. 
For this purpose, 11 indigenous goat populations/types of Ethiopia and nine 
reference breeds (four from other African countries viz. two from Kenya, one from 
Guinea Bissau and one from Botswana), and five from non-African countries (one breed 
from Turkey, Egypt, Italy, Mongolia and Saudi Arabia each) were included in this 
study. 
 181
 The sampled animals of Ethiopian goats were selected to represent the 
type/population as much as possible based on the available information of phenotypic 
characteristics and geographical distribution from previous survey by Farm-Africa 
(FARM ? Africa, 1996).  The finally identified animals for the study were sampled 
from distantly located villages and sampling was made 1-5 animals per flock.  For all 
populations/types/breeds, approximately equal number of females and males were 
sampled. 
DNA was extracted either from peripheral blood lymphocytes (PBL) or FTA? 
blood cards (Whatman? Bioscience).  Blood was collected from jugular vein by 
puncher using 10 ml capacity vacutainer tubes containing EDTA (100 ?l neat).  Every 
time a new needle was used for each animal to avoid cross contamination between 
samples.  The blood was mixed with EDTA by gently inverting the tube about three to 
five times.  In the laboratory blood samples were centrifuged at 2000 rpm for 15 ? 20 
minutes the same day evening.  The plasma was discarded and the buffy-coat containing 
the peripheral blood lymphocytes (PBL) was transferred with a wide-mouth Pasteur 
pipette into a 2 ml cryotube.  Then 1 ml of 8 M urea was added into the tube, the 
content of the tube mixed by gentle inversions five times.  The tube properly labeled 
and stored in upright position at room temperature, ready for DNA extraction.  
In the case of sampling using the FTA cards, the blood in the tube was placed on 
ice and transferred onto FTA classic cards (Whatman? Bioscience) on the same day.  
After drying, these cards were transferred into bags containing silica gel for storage and 
future use. 
The 15 microsatellite loci (BM1818, BMS1494, ILSTS05, ILSTS11, ILSTS44, 
ILSTS87, INRA05, INRA63, NRA132, MAF35, MAF65, MAF 209, OarAE129, 
OarFCB304, and SRCRSP03) were analyzed at ILRI, Nairobi, Kenya.  All PCR 
amplifications were performed in a total volume of 10 ?l (except for some FTA? blood 
disks, where a total volume of 20 ?l was used) on GeneAMP?PCR system 9700 
(Applied Biosystem) thermocycler.  All amplification conditions included an initial 
denaturing step of 5 min at 95?C, followed by 35 cycles of 30 seconds at 95?C, 1 minute 
at an appropriate primer annealing temperature and 1 minute at 72?C.  Duration of final 
extensions was 10 minutes at 72?C.  The PCR products were analyzed by 
electrophoresis on a 4.25% denaturing polyacrylamide gel using a 377 ABI automatic 
DNA sequencer (ABI PRISM 377, Applied Biosystems).  The GeneScanTM 350-
 182
 TAMRA served as an internal size standard in each lane.  A control sample was 
included in addition to the GeneScanTM 350-TAMRA for further standardization of 
fragment sizes.  Fragment lengths were identified using the GeneScan? software 
(version 3.1.2 Applied Biosystems), while allele sizes identifications were performed by 
GENOTYPER? software (version 2.0 Applied Biosystems) using the third order least 
squares method for size calling. 
A number of analyses were performed using allelic variation obtained at 15 
microsatellite loci.  Allele frequencies at each locus were calculated for each population. 
Microsatellite variability such as the observed and expected number of alleles at each 
locus for each population, mean number of alleles per locus, observed heterozygosity 
and expected heterozygosity were calculated to quantify the genetic variation within the 
population.  Statistics such as fixation indices or F-statistics were performed to describe 
the population structure and genetic differentiation among populations.  Phylogenetic 
analysis and multivariate analysis such as principal component analysis were performed 
to infer the evolutionary relationship among populations.   
Results indicated that 39 out of 300 i.e.13% of the comparisons (P < 0.05;) 
deviated significantly from HWE, a figure higher than the 5% that would be expected 
from type I error alone.  Application of sequential Bonferroni adjustment for multiple 
testing reduced this number of significant locus-population combinations to only three. 
Out of the 15 loci, there was no locus that deviated from HWE for all the populations 
and also no population was observed to deviate from HWE for all the loci. 
The linkage disequilibrium tests within each of the 20 populations were 
performed for the 105 pair-wise combinations of the 15 loci.  Assuming a 5% type I 
error rate (P < 0.05), about 5 to 6 pairs of loci are expected to be in linkage 
disequilibrium.  On average 6% of the locus pairs were in linkage disequilibrium within 
the populations. 
The global exact test for linkage disequilibrium across all 20 populations (105 
pair-wise comparisons) yielded significant deviations in nine comparisons at P < 0.05 .  
Only one out of these nine tests remained significant following application of sequential 
Bonferroni corrections. 
The recent bottleneck effect was inferred for each population using a 
Wilcoxon?s signed rank test (Cornuet and Luikart, 1996).  Generally, there was no 
recent bottleneck effect detected in any population for both Wilcoxon?s signed rank test 
and quantitative graphical method.  Interestingly, ten populations out of the total 20 
 183
 populations showed a significant deficiency of heterozygosity.  This suggests that these 
populations are not at mutation-drift equilibrium but instead have experienced a recent 
expansion in population size or perhaps a recent entry of rare alleles from genetically 
distinct immigrants. 
A total of 159 alleles were detected at 15 microsatellite loci in 776 goat 
individuals of 20 populations sampled from 20 locations in Ethiopia, Kenya, Botswana, 
Guinea Bissau, Egypt, Turkey, Italy, Saudi Arabia and Mongolia. 
The expected MNA across all populations was 4.60 ? 0.08 and 7.42 ? 0.16 for 
SMM and IAM, respectively, ranging between 4.05 ? 1.59 in Guinea Bissau and 5.77 ? 
2.70 in Hair for SMM, and between 6.39 ? 2.34 in Ardi and 9.29 ? 4.85 in Hair 
populations for IAM.  The results indicate that the stepwise mutation model (SMM) is 
more appropriate for this particular dataset and should be preferred in the model-based 
algorithms such as the inference of the recent population bottleneck. 
The total number of alleles sampled in a population across all loci ranged 
between 80 in West African Dwarf (WAD) goat population from Guinea Bissau and 
116 in Long-eared Somali goat population from Ethiopia.  The mean number of alleles 
per locus (MNA) across all populations was 5.46 ? 0.09, varying between 4.67 ? 2.74 in 
Guinea Bissau WAD goat population and 6.47? 4.03 in Long-eared Somali goat 
population.  The MNA for a uniformed sample size with 20 animals after random re-
 sampling procedure with 250 replacements over individuals was 4.88 ? 0.08 varying 
between 4.19 ? 2.26 in Guinea Bissau population and 5.74 ? 2.89 in Hair goat 
population. 
The mean observed heterozygosity (HO) and expected heterozygosity (HE) for 
Ethiopian goats was 0.551 ? 0.015 and 0.576 ? 0.015, respectively, across all loci and 
all population.  The observed heterozygosity across all loci ranged between 0.515 ? 
0.047 in Ethiopian Afar population to 0.625 ? 0.050 in Saudi Arabian Ardi population.  
The highest expected heterozygosity (gene diversity) across all loci was observed in 
Hair goat population from Turkey (0.642 ? 0.055) and the lowest was 0.523 ? 0.058 in 
WAD (Guinea Bissau) population.  The differences of both values among the goat 
populations were not statistically significant (P > 0.05), and it also seemed that the 
values were not affected by sample size. 
The level of genetic diversity (HE = 0.58 ? 0.015) obtained for the Ethiopian 
goat populations in the present study was similar to that previously reported in Swiss 
goat breeds (HE = 0.59, Saitbekova et al., 1999), and Asian goats (HE = 0.52, Barker et 
 184
 al., 2001), but less than that reported for Chinese goats (HE = 0.716, Li et al., 2002; and 
HE = 0.707, Li and Alessio, 2004). 
Seven out of 11 populations showed a significant inbreeding coefficient (FIS) (P 
< 0.05).  This inbreeding could be a consequence of the management systems, 
traditional/cultural practices and also ethnic/cultural barriers that could affect the 
between flock breeding-males (backs) exchange.  In case each sampled 
populations/types have been subdivided into subpopulations, which contains fewer 
heterozygotes than predicted despite the fact that all subdivisions are in Hardy-
 Weinberg equilibrium, can also lead to inbreeding (Wahlund effect).  All these systems 
and events tend to increase the rate of inbreeding. 
FIS statistic, an estimator for describing local inbreeding effect within 
subpopulations, was significantly different from zero at six out of 15 loci and across all 
loci at P < 0.05 level, indicating that the Ethiopian goat populations sampled for this 
study were generally at risk of inbreeding.  FST estimator was significantly different 
from zero at 14 out of 15 loci at P < 0.01, which means that genotype frequencies 
among populations are not randomly distributed, meaning that all the Ethiopian goat 
populations have been genetically subdivided even though the mean FST value across all 
loci was very low (0.016) compared to previously reported in Asian goats (FST = 0.143, 
Barker et al., 2001) and Chinese goats (FST = 0.105, Li et al., 2002 and FST = 0.054, Li 
and Valentini, 2004).  Similar value was obtained for the coefficient of gene 
differentiation using the method of Nei (1973) (GST = 0.026).  Only 1.6% to 2.6% of 
total variation was contributed by genetic subdivision among populations while 97.4% 
to 98.4% of total variation accounts for genetic variations among individuals within 
populations.  This figure went up to 6.4% (FST) and 7.4% (GST) when all the 20 goat 
populations are included in the analysis.  These differentiation values are much lower 
compared with those generally recorded in goats and other domestic animal populations 
[e.g. 9.9% between Spanish dog breeds (Jordan et al., 1992), 10% between European 
cattle (MacHugh et al., 1998), 17% between Swiss goat breeds (Saitbekova et al., 
1999), 8% between Spanish horses (Canon et al., 2000), 10.5% between Chinese 
indigenous goat populations (Li et al., 2002), and as well as 7.7% between Chinese 
indigenous pig breeds (Yang et al., 2003)].   
The highest number of migrants per generation was estimated between Short-
 eared Somali and Hararghe Highland, where it shows that these two populations are not 
genetically different at all, followed by (Nm = 120.88) between North-West Highland 
 185
 and Central Highland populations, and followed by (Nm = 82.19) between Short-eared 
Somali and Long-eared Somali. 
A clear population structure was inferred in all goat populations considered in 
this study by using a multivariate analysis.  All 20 goat populations sampled in this 
study were largely classified into three major clusters.  All the Ethiopian plus two 
Kenyan populations form one cluster; Italian, Mongolian, Saudi Arabian, Egyptian and 
Botswana form another cluster, and the Guinea Bissau alone form the third cluster. 
The UPGMA method with both Nei?s standard genetic distance and Nei et al.?s 
DA genetic distance (Nei, 1983); and the DA NJ basically classified the goat populations 
into three clear phylogenetic clusters concordant with the population structure inferred 
from the Admixture analysis. 
As expected for different populations from geographically distinct regions, all 
Ethiopian goat populations (together with the two neighboring Kenyan populations) and 
the other 7 populations (from West Africa, South Africa, North Africa, Turkey, Saudi 
Arabia, Italy and Mongolia) clearly emerged as distinct lineages different from the West 
African (Guinea Bissau) standing completely alone as an out-group (with relatively 
large genetic distance), suggesting different origins of ancestral populations.  In general, 
all the Ethiopian goat populations were more closely related to each other than to non-
 Ethiopian counterparts.  All the reference populations (except the two Kenyan 
populations and the WAD goat population from Guinea Bissau) cluster closely together 
suggesting a different ancestry.  The close clustering of the Ethiopian populations with 
the two Kenyans ones is in accordance with the work of Chenyambuga (2002). (Nei, 
Tajima, and Tateno, 1983) 
When the Ethiopian goat populations are treated with one out-group (WAD, 
from Guinea Bissau) or alone for the above analysis, the UPGMA method with Nei?s 
standard genetic distance basically classified the goat populations in to eight 
phylogenetic clusters, while the DA genetic distance for both with and without out-
 group population classified them differently.  As one origin of evolution was expected 
for Ethiopian goats, it seemed that the UPGMA performs better and consistent 
compared to the NJ trees methods.   
Comparison of the genetic diversity of the Ethiopian goat populations with that 
of goats from West Africa, South Africa, Mongolian, Saudi Arabian, Egypt, Italy and 
Turkey indicated that the Ethiopian goat populations are less genetically diverse.  This 
could be the result of being geographically very close to each other, much animal 
 186
 movement across the regions (zones & districts) and in some cases the results of strong 
human selection pressure for certain economic traits.  
The Ethiopian goats, either as separated populations or when grouped as one, 
were found to have similar diversity with that of the non-Ethiopian populations in terms 
of MNA per locus and expected heterozygosity. 
The slightly higher than expected (P < 0.05) deviation from linkage equilibrium 
observed in some of the samples could indicate population subdivision, recent 
introduction of a non-random subset of genotypes or habitat-specific selection (King et 
al., 2001 
Results of the present study using autosomal microsatellites suggest a close 
relationship of Ethiopian goat populations.  It appears that all Ethiopian goat 
populations were seen to cluster more closely together, suggesting a common origin and 
that there is extensive gene flow especially between the North-West Highland/Central 
Highland, Short-eared Somali/Hararghe Highland and Long-eared Somali/Short-eared 
Somali populations. 
The WAD population from Guinea Bissau was always clustered alone as an out-
 group, suggesting a different origin from the Ethiopian populations or a differently 
developed population as trypanotolerant dwarf goats, which agrees with the suggestion 
of Wilson (1991) that the West African Dwarf (WAD) goats arose in the Fouta Djallon 
region of Guinea, as did trypanotolerant N?Dama cattle.  The other breeds (Italian 
Alpine, Baladie, Hair, Mongolia and Ardi) cluster together, suggesting most probably 
common ancestry for them. 
The principal component analysis of this study showed that the Ethiopian and 
Kenyan breeds cluster with the West African Guinea Bissau population, which might be 
an indication that the Eastern and Western Africa goats crossed the straits of Bab-el-
 Mandeb and entered the continent of Africa via the horn of Africa.  Here, the Baladie 
population from Egypt is consistently clustered with the European and Asian 
populations, which agrees with Epstein (1971).  Additional archaeological data, 
including more accurate dating of archaeological goat remains, is required to assess 
these theories. 
The total variation accounted for by the first, second and third principal 
components, following the principal component analysis of the Ethiopian goat 
populations, was 58.68%.  This figure increased to 60.23% when the reference 
populations were included in the analysis.  This variation (58.68%) was slightly lower 
 187
 compared to that observed among twelve Chinese indigenous goat populations (60%, Li 
et al., 2002) and for sub-Saharan Africa camel populations (72%, Mburu et al., 2003), 
but higher compared to that observed for sub-Saharan Africa cattle breeds (54%, 
Hanotte et al., 2002). 
Weitzman analysis indicated that about 75% of the total genetic diversity of the 
Ethiopian goats is present in four breeds; Afar, Abergalle, Gumez and Keffa, with 
marginal loss of diversity of 24.32%, 19.22%, 16.59% and 12.99% respectively, while 
removal of other populations had only a lower impact. 
Results presented in this study show that phenotypic characteristics subject to 
environmental influences could lead to a heterogeneity that is not reflected at the 
genetic level.  Accordingly our microsatellite based molecular genetic study does not 
support the distinction of eleven Ethiopian goat populations as indicated in the study 
published in 1996 using morphological traits (Farm-Africa, 1996).  Instead, following 
the analysis of 15 microsatellite loci, the result indicate eight separate genetic entities, 
the Arsi-Bale, Gumez, Keffa, Woyto-Guji, Abergalle, Afar, Highland Goats (previously 
separated as Central and North West Highland) and the goats from the previously 
known Hararghe, Southeastern Bale and Southern Sidamo provinces (Hararghe 
Highland, Short-eared Somali and Long-eared Somali goats).  It appears likely that all 
the Ethiopian goat populations have a common origin and that there is extensive gene 
flow between them.   
The present results show that the variation among Ethiopian goat populations is 
reasonable and therefore, within breed selection should be conducted to improve these 
eight breeds and lines with different characters should be established to conserve the 
genetic diversity in the future. 
The results from this study have also shown that more than 60% of the goat 
populations showed significant inbreeding effect, which leads to loss of genetic 
diversity. Therefore, measures like education (how to exchange the breeding males 
between flocks or introduction of new breeding-males and removing of old ones within 
a certain period of time) to the farming communities, pastoralists or goat owners are 
highly important.   
The results from this study have shown that the goat genetic resource of Ethiopia 
represents a substantial genetic resource and as Weitzman method provides a well-
 defined diversity metric, allowing quantitative assessment of the consequences of 
alternative conservation activities as an aid to decision-making in conservation, here it 
 188
 is also recommended to put more emphasis on the four (Afar, Abergalle, Gumez and 
Keffa) breeds in which more than 70% of the total genetic diversity of the Ethiopian 
goats is present; to maximizing diversity conserved and guide breeding improvement 
activities for meat and milk production. 
Other beneficial uses of present day populations will emerge as genome maps 
are completed and scientists learn to identify and manipulate genes involved in 
mechanisms of growth, reproduction and disease resistance tolerance.  Indigenous 
animal genetic resources that have evolved for 9,000-12,000 years are an integral 
component of our heritage and should be nurtured and preserved for future generations. 
 
6.2 Conclusions  
Phenotypic characterization of individual animals of particular breeds has been a 
routine process.  However, the results presented in this study show that phenotypic 
characteristics subject to environmental influences could lead to a heterogeneity that is 
not reflected at the genetic level.  Accordingly our microsatellite based molecular 
genetic study does not support the distinction of eleven Ethiopian goat populations.  
Instead, following the analysis of 15 microsatellite loci, the result indicate eight separate 
genetic entities: the Arsi-Bale, Gumez, Keffa, Woyto-Guji, Abergalle, Afar, Highland 
Goats (previously separated as Central and North West Highland) and the goats from 
the previously known Hararghe, Southeastern Bale and Southern Sidamo provinces 
(Hararghe Highland, Short-eared Somali and Long-eared Somali goats).  This study 
revealed that there is no genetic differentiation between Hararghe Highland and Short-
 eared Somali, and insignificant differentiation between Central Highland and Northwest 
Highland; as well as between Long-eared Somali and Short-eared Somali populations.  
It appears likely that all the Ethiopian goat populations have a common origin and that 
there is extensive gene flow between the North-West Highland/Central Highland, Short-
 eared Somali/Hararghe Highland and Long-eared Somali/Short-eared Somali 
populations. 
Furthermore, it can also be concluded from this study that microsatellite marker 
is a powerful genetic tool in estimation of genetic variation within and among closely 
related populations.  The present results show that the variation among Ethiopian goat 
populations is reasonable and therefore within breed selection should be conducted to 
 189
 improve these eight breeds and lines with different characters should be established to 
conserve the genetic diversity in the future 
6.3. Recommendation  
The results from this study have shown that more than 60% of the goat 
populations showed significant inbreeding effect, which leads to loss of genetic 
diversity.  Therefore, measures like education (how to exchange the breeding males 
between flocks or introduction of new breeding-males and removing of old ones within 
a certain period of time) to the farming communities, pastoralists or goat owners are 
highly vital.  The results from this study have shown that the goat genetic resources of 
Ethiopia represent a substantial genetic resource and as Weitzman method provides a 
well-defined diversity metric, allowing quantitative assessment of the consequences of 
alternative conservation activities as an aid to decision-making in conservation, here it 
is also recommended to put more emphasis on the four (Afar, Abergalle, Gumez and 
Keffa) breeds in which more than 70% of the total genetic diversity of the Ethiopian 
goats is present; to maximizing diversity conserved and guide breeding improvement 
activities for meat and milk production. 
Other beneficial uses of present day populations will emerge as genome maps 
are completed and scientists learn to identify and manipulate genes involved in 
mechanisms of growth, reproduction and disease resistance tolerance.  Indigenous 
animal genetic resources that have evolved for 9,000-12,000 years are an integral 
component of our heritage and should be nurtured and preserved for future generations. 
 190
 REFERENCES 
 
Aboagye, G. (1992). Characterization and evaluation Sanga and West African shorthorn 
cattle on the Accra plains. In: Africa Animal Genetic Resources: Their 
characterization, conservation and utilization. Proceedings of the research 
planning workshop held at ILCA, Addis Ababa, Ethiopia 19-21 February, 1992. 
ILCA (International Livestock Center for Arfica) Addis Ababa, Ethiopia. pp 73-
 75. 
Aboagye, G. S., Tawah, C. L., and Rege, J. E. O. (1994). Shorthorn cattle of West and 
Central Africa. III. Physical, adaptive and special genetic characteristics. World 
Animal Review 78 (1): 22-32. 
Akey, J. M., Zhang, K., Xiong, M., and Jin, L. (2003). The effect of single nucleotide 
polymorphism identification strategies on estimates of linkage disequilibrium. 
Mol Biol Evol 20(2), 232-42. 
Alados, C. L. (1985b). Distribution and status of the Spanish ibex (Capra pyrenaica 
Schinz). In Lovari (1985), pp. 204-11. 
Alemayehu Reda (1993). Characterization (Phenotypic) of indigenous goats and goat 
husbandry practices in east and south-eastern Ethiopia. M.Sc thesis, Alemaya 
University of Agriculture, Ethiopia. 135pp. 
Alexandrino, F. W., Knudson, K. L., and Leary, R. F. (1983). Adaptive significance of 
the differences in the tissue-specific expression of phosphoglucomutase gene in 
rainbow trout. In: Proceedings of theNational Academy of Science of the USA 
800: 1397-1400. 
Allen, P. J., Amos, W., Pomeroy, P. P., and Twiss, S. D. (1995). Microsatellite variation 
in grey seals (Halichoerus grypus) shows evidence of genetic differentiation 
between two British breeding colonies. Mol Ecol 4(6), 653-62. 
Allendorf, F. W., and Leary, R. F. (1986). Heterozygosity and fitness in natural 
populations of animals. In: Conservation Biology: The science of scarcity and 
diversity. (Edited by M.E. Soule) Sinauer, Sunderland, Massachusetts. pp 57-76. 
Armour, J. A. L., and Jeffreys, A. J. (1992). Biology and applications of human 
minisatellite loci. Current Opinion in Genetics and Development 2: 850-856. 
 
 191
 Arora, R., Lakhchaura, B. D., Prasad, R. B., M.S, T., and Vijh, R. K. (2004). Genetic 
diversity analysis of two buffalo populations of northern India using 
microsatellite markers. Anim. Breed. Genet. 121, 111-118. 
Arranz, J. J., Bayon, Y., and San Primitivo, F. (1998). Genetic relationships among 
Spanish sheep using microsatellites. Anim Genet 29(6), 435-40. 
Avis, J. C. (1994). Molecular markers, Natural history and evolution. Chapman & Hall. 
New York, London. pp 511. 
Ayalew, W. (1992). Preliminary survey of indigenous goat types and goat husbandery 
practices in Southern Ethiopia. MSc thesis, Alemaya University of Agriculture, 
Alemaya,  Ethiopia. 153pp. 
Baker, A. J., and Moeed, A. (1987). Rapid genetic differentiation and founder effect in 
colonizing populations of common Mynas (Acridotheres tristis). Evolution 
41:525-538. 
Balloux, F., and Lugon-Moulin, N. (2002). The estimation of population differentiation 
with microsatellite markers. Mol Ecol 11(2), 155-65. 
Banks, M. A., and Eichert, W. (2000). WHICHRUN (version 3.2): a computer program 
for population assignment of individuals based on multilocus genotype data. J 
Hered 91(1), 87-9. 
Barbancho, M., Llanes, D., Morera, L., Garzon, R., and Rodero, A. (1984). Genetic 
markers in the blood of Spanish goat breeds. Anim Blood Groups Biochem 
Genet 15(3), 207-12. 
Barendse, W., Armitage, S. M., Kossarek, L. M., Shalom, A., Kirkpatrick, B. W., Ryan, 
A. M., Clayton, D., Li, L., Neibergs, H. L., Zhang, N., and et al. (1994). A 
genetic linkage map of the bovine genome. Nat Genet 6(3), 227-35. 
Barendse, W., Vaiman, D., Kemp, S. J., Sugimoto, Y., Armitage, S. M., Williams, J. L., 
Sun, H. S., Eggen, A., Agaba, M., Aleyasin, S. A., Band, M., Bishop, M. D., 
Buitkamp, J., Byrne, K., Collins, F., Cooper, L., Coppettiers, W., Denys, B., 
Drinkwater, R. D., Easterday, K., Elduque, C., Ennis, S., Erhardt, G., Li, L., and 
et al. (1997). A medium-density genetic linkage map of the bovine genome. 
Mamm Genome 8(1), 21-8. 
 
 
 
 192
 Barker, J. S. F. (1994). A global protocol for determining genetic distances among 
domestic livestock breeds. In: Proceedings of the 5th world congress of genetic 
applied to livestock production. 7-12 August, 1994 Guelph, Canada. Vol 21: 
501-508." 
Barker, J. S. F. (2002). Relevance of animal genetic resources and differences to the 
plant sector. In: Groeneveld E. and Glodek P. (Eds), Animal Breeding and 
Animal Genetic Resources. Federal Agricultural Research Centre (FAL), 
Mariensee and Institute of Animal Genetics, Gottingen, Braunschweig, 
Germany, 15-21. 
Barker, J. S. F., Tan, S. G., Moore, S. S., Mukherjee, T. K., Matheson, J.-L., and 
Selvaraj, O. S. (2001). Genetic variation within and relationships among 
populations of Asian goats (Capra hircus). Journal of Animal Breeding and 
Genetics 118(4), 213-233. 
Beja-Pereira, A., Alexandrino, P., Bessa, I., Carretero, Y., Dunner, S., Ferrand, N., 
Jordana, J., Laloe, D., Moazami-Goudarzi, K., Sanchez, A., and Canon, J. 
(2003). Genetic characterization of southwestern European bovine breeds: a 
historical and biogeographical reassessment with a set of 16 microsatellites. J 
Hered 94(3), 243-50. 
Bermond, J., editor. (2001). Molekulargenetische Differenzierung verschiedener 
Rotviehpopulationen. Bundesministerium fur Verbraucherschutz, Ernahrung und 
Landwirtschaft, Munster-Hiltrup, Germany. 
Bishop, M. D., Kappes, S. M., Keele, J. W., Stone, R. T., Sunden, S. L., Hawkins, G. 
A., Toldo, S. S., Fries, R., Grosz, M. D., Yoo, J., and et al. (1994). A genetic 
linkage map for cattle. Genetics 136(2), 619-39. 
Bohonak, A. J. (1999). Dispersal, gene flow, and population structure. Q Rev Biol 74(1), 
21-45. 
Bohonak, A. J. (2002). IBD (Isolation by Distance): a program for analyses of isolation 
by distance. J Hered 93(2), 153-4. 
Botstein, D., White, R. L., Skolnick, M., and Davis, R. W. (1980). Construction of a 
genetic linkage map in man using restriction fragment length polymorphisms. 
Am J Hum Genet 32(3), 314-31. 
Bowcock, A. M., Ruiz-Linares, A., Tomfohrde, J., Minch, E., Kidd, J. R., and Cavalli-
 Sforza, L. L. (1994). High resolution of human evolutionary trees with 
polymorphic microsatellites. Nature 368(6470), 455-7. 
 193
 Bradley, D. G., MacHugh, D. E., Cunningham, P., and Loftus, R. T. (1996). 
Mitochondrial diversity and the origins of African and European cattle. Proc 
Natl Acad Sci U S A 93(10), 5131-5. 
Brookfield, J. F. (1996). A simple new method for estimating null allele frequency from 
heterozygote deficiency. Mol Ecol 5(3), 453-5. 
Bruford, M. W., and Wayne, R. K. (1993). Microsatellites and their application to 
population genetic studies. Curr Opin Genet Dev 3(6), 939-43. 
Buchanan, F. C., Adams, L. J., Littlejohn, R. P., Maddox, J. F., and Crawford, A. M. 
(1994). Determination of evolutionary relationships among sheep breeds using 
microsatellites. Genomics 22(2), 397-403. 
Burke, T. (1989). DNA fingerprinting and other methods for the study of mating 
success. Trends Ecol. Evol., 4:, 139-144. 
Canon, J., Alexandrino, P., Bessa, I., Carleos, C., Carretero, Y., Dunner, S., Ferran, N., 
Garcia, D., Jordana, J., Laloe, D., Pereira, A., Sanchez, A., and Moazami-
 Goudarzi, K. (2001). Genetic diversity measures of local European beef cattle 
breeds for conservation purposes. Genet Sel Evol 33(3), 311-32. 
Canon, J., Checa, M. L., Carleos, C., Vega-Pla, J. L., Vallejo, M., and Dunner, S. 
(2000). The genetic structure of Spanish Celtic horse breeds inferred from 
microsatellite data. Anim Genet 31(1), 39-48. 
Carson, H. L. (1983). The genetics of the founder effect. In: Genetics and Conservation: 
A reference for managing wild animal and plant populations. (Edited by C.M. 
Schonewald-Cox; S.M. Chambers; B.MacBryde and L.Thomas). 
Benjamin/Cummings, menlopark, CA. pp 189-200. 
Cavalli Sforza, L. L., Menozzi, P., and A, P. (1994). ''The history and geography of 
human genes,'' Princeton University Press, Princeton, New Jersey. 
Cavalli-Sforza, L. L., and Edwards, A. W. (1967). Phylogenetic analysis. Models and 
estimation procedures. Am J Hum Genet 19(3), Suppl 19:233+. 
Cavalli-Sforza, L. L., and Edwards, W. L. (1964). Analysis of human evolution. 
Proceedings 11th International Congress on Genetics, 923-933. 
Chakraborty, R., and Zhong, Y. (1994). Statistical power of an exact test of Hardy-
 Weinberg proportions of genotypic data at a multiallelic locus. Hum Hered 
44(1), 1-9. 
 
 
 194
 Charlesworth, D., Lyons E.E., Litchfield L.B., (1994). Inbreeding depression in two 
highly inbreeding populations of Leaven-worthia. Proc R Soc Lond B Biol Sci 
258: 209-214. 
Charlesworth, D., and Charlesworth, B. (1987). Inbreeding depression and its 
evolutionary consequences. Annual Review of Ecology and Systematics 18: 
237-268. 
Chenyambuga, S. (2002). Genetic characterisation of indigenous goat populations of 
sub-Saharan Africa using microsatelite DNA markers. PhD thesis, Department 
of Animal Science and Production, Sokoine University of Agriculture, Sokoine, 
Tanzania. 
Chung, E. R., Kim, W. T., Kim, Y. S., and Kin, D. K. (1998). Korean J. Dairy Sci., 21:, 
1-8. 
Ciampolini, R., Moazami-Goudarzi, K., Vaiman, D., Dillmann, C., Mazzanti, E., 
Foulley, J. L., Leveziel, H., and Cianci, D. (1995). Individual multilocus 
genotypes using microsatellite polymorphisms to permit the analysis of the 
genetic variability within and between Italian beef cattle breeds. J Anim Sci 
73(11), 3259-68. 
Close, A. E. (2002). Sinai, Sahara, Sahel: The Introduction of Domestic Caprines to 
Africa, In:Jennerstasse 8(Eds.) Tides of the desert, Africa Praehistorica 14, 
Heinrich Barth Institut, Koln, pp. 459-469. 
Cluctton-Brock, J. (1981). Domestic Animals from Early Times (Heinemann and Br. 
Mus. of Nat. Hist., London). 
Corbet, G. B. (1978). The mammals of the Palaearctic Region: a taxonomic review. 
British Museum. (National History), London, 314 pp. 
Cornuet, J. M., and Luikart, G. (1996). Description and power analysis of two tests for 
detecting recent population bottlenecks from allele frequency data. Genetics 
144(4), 2001-14. 
Cornuet, J. M., Piry, S., Luikart, G., Estoup, A., and Solignac, M. (1999). New methods 
employing multilocus genotypes to select or exclude populations as origins of 
individuals. Genetics 153(4), 1989-2000. 
Crawford, A. M., Dodds, K. G., Ede, A. J., Pierson, C. A., Montgomery, G. W., 
Garmonsway, H. G., Beattie, A. E., Davies, K., Maddox, J. F., Kappes, S. W., 
and et al. (1995). An autosomal genetic linkage map of the sheep genome. 
Genetics 140(2), 703-24. 
 195
 Crawford, A. M., Godds, K. G., and Pierson (1995). An autosomal genetic linkage of 
the sheep genome. Genetics 140, 703-724. 
Crozier R. H. (1992). Genetic diversity and the agony of choice. Biological 
Conservation 61, 11-15. 
Cunningham, E. P., Loftus, R. T., MacHugh, D. E., and Bradley, D. G. (1994). 
Molecular evolution of Africa, European and Asian cattle. In: Proceedings of the 
5th world congress of genetic applied to livestock production. 7-12 August, 
1994 Guelph, Canada. Vol 21: 86-89."". 
Delgado, C., Rosegrant, M., Steinfeld, H., Ehui, S., and Courbois, C. (1999). ?Livestock 
to 2020 ? The Next Food Revolution?. Food, Agriculture and the Environment 
Discussion Paper 28. IFPRI/FAO/ILRI. 
Devendra, C., and Burns, M. (1983). Goat production in the tropics. Commonwealth 
Agricultural Beauro. Farnham house, Farnham royal, Slough SL2 3BN UK. pp, 
183. 
Dieringer, D., and Schl?tterer, C. (2003). microsatellite analyser (MSA): a platform 
independent analysis tool for large microsatellite data sets. Molecular Ecology 
Notes 3 (1), 167-169. 
Donnelly, M. J., and Townson, H. (2000). Evidence for extensive genetic differentiation 
among populations of the malaria vector Anopheles arabiensis in Eastern Africa. 
Insect Mol Biol 9(4), 357-67. 
Dowling, T. E., Moritz, C., Palmer, J. D., and Rieseberg, L. H. (1996). Nucleic Acids 
III: Analysis of fragments and restriction sites. In: Molecular systematics 2nd 
edition (Edited by D.M. Hills; C. moritz, and B.K. Mable). Sinauer Associates, 
Inc. Publishers. Sunderland, Massachusetts, USA. pp 249-320. 
Edwards, A., Civitello, A., Hammond, H. A., and Caskey, C. T. (1991). DNA typing 
and genetic mapping with trimeric and tetrameric tandem repeats. Am J Hum 
Genet 49(4), 746-56. 
Edwards, A., Hammond, H. A., Jin, L., Caskey, C. T., and Chakraborty, R. (1992). 
Genetic variation at five trimeric and tetrameric tandem repeat loci in four 
human population groups. Genomics 12(2), 241-53. 
Ellegren, H. (1993). Genome analysis with microsatellite markers. PhD dissertation. 
Swidish University of Agricultural Science. 
Ellerman, J. R., and Morrison-Scott., T. C. S. (1966). Checklist of Palaearctic and 
Indian mammals. British Mus. (Nat. Hist.), London, 810 pp. 
 196
 Epstein, H. (1971). The origion of the domestic animals of Africa. Vol. II. African 
publishing corporation. New York, London, Munich. pp 719. 
Estoup, A., Garnery, L., Solignac, M., and Cornuet, J. M. (1995). Microsatellite 
variation in honey bee (Apis mellifera L.) populations: hierarchical genetic 
structure and test of the infinite allele and stepwise mutation models. Genetics 
140(2), 679-95. 
Excoffier, L., Smouse, P. E., and Quattro, J. M. (1992). Analysis of molecular variance 
inferred from metric distances among DNA haplotypes: application to human 
mitochondrial DNA restriction data. Genetics 131(2), 479-91. 
Faith, D. P. (1992). Conservation evaluation and phylogenetic diversity. Biological 
Conservation 61, 1-10. 
Falconer, D. S., and Mackey, T. F. C. (1996). "''Introduction to quantitative Genetics,'' 
Longman Group Ltd, Harlow." 
Falush, D., Stephens, M., and Pritchard, J. K. (2003). Inference of population structure 
using multilocus genotype data: linked loci and correlated allele frequencies. 
Genetics 164(4), 1567-87. 
Fan, B., Wang, Z. G., Li, Y. J., Zhao, X. L., Liu, B., Zhao, S. H., Yu, M., Li, M. H., 
Chen, S. L., Xiong, T. A., and Li, K. (2002). Genetic variation analysis within 
and among Chinese indigenous swine populations using microsatellite markers. 
Anim Genet 33(6), 422-7. 
FAO (1994). Production Yearbook. Food and Agriculture Organization. Rome, Italy. 
FAO (1995). World watch list for domestic animal diversity. 2nd edition. Food and 
Agricultural organization of the united nations, Rome, Italy. pp.769. 
FAO (1999). The global strategy of the management of Farm Animal Genetic 
Resources. Executive Brief., FAO, Rome. 
FAO (2000). Conserving and developing farm animal diversity. In State of the world 
animal genetic resources. Http://www.fao.org/News/2000/001201-e.htm. 
Farm-Africa (1996). Goat types of Ethiopia and Eritrea. Physical discription and 
management systems. Published jointly by FARM-Africa, London, UK, and 
ILRI (International Livestock Research Institute), Nairobi, Kenya. 76pp. 
Farris, J. S. (1970). Methods for computing Wagner trees. Systematic Zoology 34, 21-
 34. 
Farris, J. S. (1971). The hypothesis of non-specificity and taxonomic congruence. 
Annual Review of Ecology and Systematics 2, 227-302. 
 197
 Farris, J. S. (1981). Distance data in phylogenetic analysis. In: Advances in Cladistics. 
Proceedings of thr first meeting of the willi Henning Society (V.A. Funk and 
D.R. Brooks edition). New York Botanical garden, New York. pp 3-23. 
Felsenstein, J. (1981). Evolutionary trees from DNA sequences: a maximum likelihood 
approach. J Mol Evol 17(6), 368-76. 
Felsenstein, J. (1985). Confidence limits on phylogenies: An approach using the 
bootstrap. Evolution 39:783-791. 
Fitch, W. M. (1971). Toward defining the course of evolution. Minimum change for a 
specific tree topology. Systematic Zoology 20, 406-416. 
Forbes, S. H., Hogg, J. T., Buchanan, F. C., Crawford, A. M., and Allendorf, F. W. 
(1995). Microsatellite evbolution in congeneric mammals: domestic and bighorn 
sheep. Mol. Biol. & Evol., 12:, 1106-1113. 
Frankham, R. (1994). Conservation of genetic diversity for animal improvement. In: 
Proceeding of the 5th World Congress of Genetic Applied toLivestock 
Production. 7-12 August, 1994 Guelph, Canada. Vol 21: 385-392." 
Franklin, I. R. (1980). Evolutionary change in small populations. In: Conservation 
Biology: An evolutionary-Ecological perspective. (Edited by M.E. Soule and 
B.A. Wilcox). Sinauer Associates, Sunderland, MA. pp 135-149. 
Frantz, A. C., Pope, L. C., Carpenter, P. J., Roper, T. J., Wilson, G. J., Delahay, R. J., 
and Burke, T. (2003). Reliable microsatellite genotyping of the Eurasian badger 
(Meles meles) using faecal DNA. Mol Ecol 12(6), 1649-61. 
Gall, C. (1981). Goat Production. Academic Press Inc., New York, USA. 619 pp. 
Ganai, N. A., and Yaday, B. R. (2001). Genetic variation within and among three Indian 
breeds of goat using heterologous microsatellite markers. Anim Biotechnol. (2), 
121-36. 
Georges, M., Dietz, A. B., Mishra, A., Nielsen, D., Sargeant, L. S., Sorensen, A., Steele, 
M. R., Zhao, X., Leipold, H., Womack, J. E., and et al. (1993). Microsatellite 
mapping of the gene causing weaver disease in cattle will allow the study of an 
associated quantitative trait locus. Proc Natl Acad Sci U S A 90(3), 1058-62. 
Gerlach, G., and Musolf, K. F. (2000). Fragmentation of landscape as a cause for 
genetic subdivision in bank volves. Conservation biology 14, 1066-1074. 
Goldstein, D. B., A. Ruiz Linares, L.L. Cavalli-Sforza, and Feldman., M. W. (1995b). 
Genetic absolute dating based on microsatellites and the origin of modem 
humans. Proc. Natl. Acad Sci. USA 92:6723-6727. 
 198
 Goldstein, D. B., Ruiz Linares, A., Cavalli-Sforza, L. L., and Feldman, M. W. (1995a). 
An evaluation of genetic distances for use with microsatellite loci. Genetics 
139(1), 463-71. 
Gottelli, D., Sillero-Zubiri, C., Applebaum, G. D., Roy, M. S., Girman, D. J., Garcia-
 Moreno, J., Ostrander, E. A., and Wayne, R. K. (1994). Molecular genetics of 
the most endangered canid: the Ethiopian wolf Canis simensis. Mol Ecol 3(4), 
301-12. 
Goudet, J. (2001). FSTAT, a program to estimate and test gene diversities and fixation 
indices (version 2.9.3). Available at http://www.unil.ch/izea/software/fstat.htm. 
Grodzicker, T., Williams, J. L., Sharp, P., and Sambrook, J. (1974). Cold Spring Harbor 
Symp. Quant. Bio. 39: 439-446.           . 
Guo, S. W., and Thompson, E. A. (1992). Performing the exact test of Hardy-Weinberg 
proportion for multiple alleles. Biometrics 48(2), 361-72. 
Gwakisa, P. S., Kemp, S. J., and Teale, A. J. (1994). Characterization of Zebu cattle 
breeds in Tanzania using random amplified polymorphic DNA markers. Anim 
Genet 25(2), 89-94. 
Haberfeld, A., Cahaner, A., Yoffe, O., Plotsky, Y., and Hillel, J. (1991). DNA 
fingerprints of farm animals generated by microsatellite and minisatellite DNA 
probes. Anim Genet 22(3), 299-305. 
Haldane, J. B. S. (1954). An exact test for randomness of mating. Genetics 52, 631-635. 
Hamada, H., Seidman, M., Howard, B. H., and Gorman, C. M. (1984). Enhanced gene 
expression by the poly(dT-dG).poly(dC-dA) sequence. Mol Cell Biol 4(12), 
2622-30. 
Hammond, K. (1993). Why Conserve animal genetic resources? Giversity 9(3): 30-35. 
Hammond, K. (1994). Conservation of domestic animal diversity: Global overview. In: 
proceeding of the 5th world congress of genetic applied to livestock production. 
7-12 August, 1994 Guelph, Canada. Vol 21: 423-430. 
Hanotte, O., Bradley, D. G., Ochieng, J. W., Verjee, Y., Hill, E. W., and Rege, J. E. 
(2002). African pastoralism: genetic imprints of origins and migrations. Science 
296(5566), 336-9. 
Hanotte, O., Burke, T., Armour, J. A., and Jeffreys, A. J. (1991). Hypervariable 
minisatellite DNA sequences in the Indian peafowl Pavo cristatus. Genomics 
9(4), 587-97. 
 199
 Hanslik, S., Harr, B., Brem, G., and Schlotterer, C. (2000). Microsatellite analysis 
reveals substantial genetic differentiation between contemporary New World 
and Old World Holstein Friesian populations. Anim Genet 31(1), 31-8. 
Hardy, O. J. (2003). Estimation of pairwise relatedness between individuals and 
characterization of isolation-by-distance processes using dominant genetic 
markers. Mol Ecol 12(6), 1577-88. 
Hardy, O. J., Charbonnel, N., Freville, H., and Heuertz, M. (2003). Microsatellite allele 
sizes: a simple test to assess their significance on genetic differentiation. 
Genetics 163(4), 1467-82. 
Hardy, O. J., and Vekemans, X. (2002). SPAGeDi: a versatile computer program to 
analyse spatial genetic structure at the individual or population levels. Molecular 
Ecology Notes 2, 618-620. 
Harris, D. R. (1961). The distribution and ancestory of the domestic goat. The journal of 
Linnean Society of London 173, 79-91. 
Hartl, D. L., and Clark, A. G. (1989). ''Darwinian evolution in mendelian populations. 
Principles of population genetics; Sinauer Associates Inc, Saunderland, 
Massachusetts. 
Hartl, D. L., and Clark, A. G. (1997). Principles of population genetics, 3rd ed. Sinauer 
Associates. Saunderland, Massachusetts. pp 542. 
Hasegawa, M., Kishino, H., and Saitou, N. (1991). On the maximum likelihood method 
in molecular phylogenetics. J Mol Evol 32(5), 443-5. 
Hedrick, P. W., Brussard, P. F., Allendorf, F. W., Beardmore, J., and Orzack, S. H. 
(1986). Protein variation, fitness, and captive propagation. Zoo Biology 5, 91-99. 
Hellberg, M. E. (1994). Relationships between inferred levels of gene flow and 
geographic distance in a philopatric coral, Balanophyllia elegans. Evolution 48: 
1829-1854. 
Hills, D., Schlaepfer, J., Comincini, S., MacLean, I., Dolf, G., Ferretti, L., Olsaker, I., 
and Williams, J. L. (2003). Sequence variation in the bovine and ovine PRNP 
genes. Anim Genet 34(3), 183-90. 
Hodges, J. (1990). The global organization of animal genetic resources. In: Proceeding 
of the 4th World Congress on Genetics Applied to Livestock Production. Vol. 
XIV: 466-472. 
 200
 Hole, F. (1996). The Context of Caprine Domestication in the Zagaros region. In 
Harris,D. R. (Ed). The Origions and Spread of Agriculture and Pastoralism in 
EuroAsia. Smithsonian Inst., Washington,DC. pp 263-281., 390-412. 
Hunter, M. L. J. (1996). Fundamentals of conservation biology. Blackwell science, Inc., 
Cambridge, Massachusetts, U.S.A. pp 482. 
Iglesias, G. M., Soria, L. A., Goto, R. M., Jar, A. M., Miquel, M. C., Lopez, O. J., and 
Miller, M. M. (2003). Genotypic variability at the major histocompatibility 
complex (B and Rfp-Y) in Camperos broiler chickens. Anim Genet 34(2), 88-95. 
Jin, L. & Chakraborty, R. 1993.  Estimation of Genetic Distance and Coefficient of 
Gene Diversity from Single-Probe Multilocus DNA Fingerprinting Data. Mol. 
Biol. Evol. 11: 120-127. 
 
Jeffreys, A. J., and Morton, D. B. (1987). DNA fingerprints of dogs and cats. Anim 
Genet 18(1), 1-15. 
Jeffreys, A. J., Royle, N. J., Wilson, V., and Wong, Z. (1988). Spontaneous mutation 
rates to new length alleles at tandem-repetitive hypervariable loci in human 
DNA. Nature 332(6161), 278-81. 
Jeffreys, A. J., Tamaki, K., MacLeod, A., Monckton, D. G., Neil, D. L., and Armour, J. 
A. (1994). Complex gene conversion events in germline mutation at human 
minisatellites. Nat Genet 6(2), 136-45. 
Jeffreys, A. J., Wilson, V., and Thein, S. L. (1985). Hypervariable 'minisatellite' regions 
in human DNA. Nature 314(6006), 67-73. 
Jeffreys, A. J., Wilson, V., Thein, S. L., Weatherall, D. J., and Ponder, B. A. (1986). 
DNA "fingerprints" and segregation analysis of multiple markers in human 
pedigrees. Am J Hum Genet 39(1), 11-24. 
Jianlin, H., Mburu, D., Ochieng, J., Kaufmann, B., Rege, J. E., and Hanotte, O. (2000). 
Application of New World Camelidae microsatellite primers for amplification of 
polymorphic loci in Old World camelids. Anim Genet 31(6), 404-6. 
Joshi, M. B., Rout, P. K., Mandal, A. K., Tyler-Smith, C., Singh, L., and Thangaraj, K. 
(2004). Phylogeography and origin of Indian domestic goats. Mol Biol Evol 
21(3), 454-62. 
Kalinowski, S. T. (2002). Evolutionary and statistical properties of three genetic 
distances. Mol Ecol 11(8), 1263-73. 
Jordana J., Piedrafita J., Sanchez A., and Puig P. (1992). Comparative F statistics 
analysis of the genetic structure of ten Spanish dog breeds. Journal of Heredity 
83: 367-74. 
 201
  
Kanga, S. (2003). Development and Application of genetic markers linked to a QTL 
controlling resistance to Trypanosomosis on bovine chromosome 7. Ph.D. 
Thesis. Department of Biochemistry, Kenyatta University. 
Kantanen, J., Vilkki, J., Elo, K., and Maki-Tanila, A. (1995). Random amplified 
polymorphic DNA in cattle and sheep: application for detecting genetic 
variation. Anim Genet 26(5), 315-20. 
Kappes, S. M., Keele, J. W., Stone, R. T., McGraw, R. A., Sonstegard, T. S., Smith, T. 
P., Lopez-Corrales, N. L., and Beattie, C. W. (1997). A second-generation 
linkage map of the bovine genome. Genome Res 7(3), 235-49. 
Karp, A., Issac, P. G., and Ingram, D. S. (1998). Molecular tools for screening 
biodiversity. Plants and Animals. Chapman and Hall. London., Weinheim, New 
York, Tokyo, Melbourne, Madras. pp 498. 
Kashi, Y., Lipkin, E., Darvasi, A., Nave, A., Gruenbaum, Y., Beckmann, J. S., and 
Soller, M. (1990a). Parentage identification in the bovine using 
"deoxyribonucleic acid fingerprints". J Dairy Sci 73(11), 3306-11. 
Kashi, Y., Tikochinsky, Y., Genislav, E., Iraqi, F., Nave, A., Beckmann, J. S., 
Gruenbaum, Y., and Soller, M. (1990b). Large restriction fragments containing 
poly-TG are highly polymorphic in a variety of vertebrates. Nucleic Acids Res 
18(5), 1129-32. 
Kemp, S. J. (1992). The potential contribution of biotechnology in breed 
characterization. In: African animal genetic Resources:Proceedings of the 
research planning workshop held at ILCA, Addis Ababa, Ethiopia 19-21 
February, 1992. ILCA (International Livestock Center for Arfica) Addis Ababa, 
Ethiopia. pp 23-28." 
Kikkawa, Y., Yonekawa, H., Suzuki, H., and Amano, T. (1997). Analysis of genetic 
diversity of domestic water boffaloes and anoas based on variations in the 
metochondrial gene for cytochrome b. Animal Genetics 28, 195-201. 
Kimura, M., and Crow, J. F. (1964). The Number of Alleles That Can Be Maintained in 
a Finite Population. Genetics 49, 725-38. 
King, T. L., Kalinowski, S. T., Schill, W. B., Spidle, A. P., and Lubinski, B. A. (2001). 
Population structure of Atlantic salmon (Salmo salar L.): a range-wide 
perspective from microsatellite DNA variation. Mol Ecol 10(4), 807-21. 
 202
 Kuhnlein, U., Zadworny, D., Dawe, Y., Fairfull, R. W., and Gavora, J. S. (1989). DNA 
fingerprinting: A tool for determining genetic distances between strains of 
poultry. Theoretical and Applied Genetics 77, 669-672. 
Lacy, R. C. (1987). Loss of genetic variation from managed populations: Interacting 
effects of drift, mutation, immigration, selection and population subdivision. 
Conservation Biology 1: 143-158. 
Lande, R. (1988). Genetics and demography in biological conservation. Science 
241(4872), 1455-60. 
Lande, R., and Barrowclaugh, G. F. (1987). Effective population size, genetic variation, 
and their use in population management. In: Viable populations for conservation 
(Edited by M.E. Soule). University press. pp 87-123. 
Laval, G., Iannuccelli, N., Legault, C., Milan, D., Groenen, M. A., Giuffra, E., 
Andersson, L., Nissen, P. H., Jorgensen, C. B., Beeckmann, P., Geldermann, H., 
Foulley, J. L., Chevalet, C., and Ollivier, L. (2000). Genetic diversity of eleven 
European pig breeds. Genet Sel Evol 32(2), 187-203. 
Leak, S., Faye, D., and Geerts, S. (2002). Genetic resistance to vectors and vector borne 
diseases in indigenous West African livestock. In: Integrated Control of 
Pathogenic Trypanosomes and their Vectors (ICPTV) newsletter, 6: 36-38. 
Leberg, P. L. (1992). Effects of a population bottleneck on genetic variation as 
measured by alloyzme electrophoresis. evolution 46: 477-494. 
Li, M. H., Zhao, S. H., Bian, C., Wang, H. S., Wei, H., Liu, B., Yu, M., Fan, B., Chen, 
S. L., Zhu, M. J., Li, S. J., Xiong, T. A., and Li, K. (2002). Genetic relationships 
among twelve Chinese indigenous goat populations based on microsatellite 
analysis. Genet Sel Evol 34(6), 729-44. 
Li Xiang-Long, and Valentini, A. (2004). Genetic diversity of Chinese goat breeds 
based on microsatellite markers. J. Anim. Breed. Genet. 121, 350-355. 
Litt, M., and Luty, J. A. (1989). A hypervariable microsatellite revealed by in vitro 
amplification of a dinucleotide repeat within the cardiac muscle actin gene. Am J 
Hum Genet 44(3), 397-401. 
Loftus, R. T., Ertugrul, O., Harba, A. H., El-Barody, M. A., MacHugh, D. E., Park, S. 
D., and Bradley, D. G. (1999). A microsatellite survey of cattle from a centre of 
origin: the Near East. Mol Ecol 8(12), 2015-22. 
 203
 Loftus, R. T., MacHugh, D. E., Ngere, L. O., Balain, D. S., Badi, A. M., Bradley, D. G., 
and Cunningham, E. P. (1994). Mitochondrial genetic variation in European, 
African and Indian cattle populations. Anim Genet 25(4), 265-71. 
Lomker, R., and Simon, D. L. (1994). Costs of and inbreeding in conservation strategies 
for endangered breeds of cattle. In: Proceedings of the 5th World Congress on 
Genetics Applied to Livestock production, Vol. XIV: 434-442. 
Louis, E. J., and Dempster, E. R. (1987). An exact test for Hardy-Weinberg and 
multiple alleles. Biometrics 43(4), 805-11. 
Luikart, G. (1997). Usefulness of molecular markers for detecting population 
bottlenecks and monitoring genetic change (PhD dissertation). Missoula, 
Montana: University o Montana. 
Luikart, G., Allendorf, F. W., Cornuet, J. M., and Sherwin, W. B. (1998). Distortion of 
allele frequency distributions provides a test for recent population bottlenecks. J 
Hered 89(3), 238-47. 
Luikart, G., and Cornuet, J. M. (1998). Emperical evaluation of a test for identifying 
recently bottlenecked populations from allele frequency data. Conservation 
biology 12(1), 228-237. 
Luikart, G., Gielly, L., Excoffier, L., Vigne, J. D., Bouvet, J., and Taberlet, P. (2001). 
Multiple maternal origins and weak phylogeographic structure in domestic 
goats. Proc Natl Acad Sci U S A 98(10), 5927-32. 
Luo, J., Liu, Z. G., Yang, G. S., and Zhen, X. M. (2001). Evaluation of Genetic 
Relativeness and Diversity in Five Goat Breeds Using Randomly Amplified 
Polymorphic DNA (RAPD) Analysis. 
http://www.luresext.edu/goats/library/luo2001-1.html. 
Lynch, M. (1988). Estimation of relatedness by DNA fingerprinting. Mol Biol Evol 
5(5), 584-99. 
MacHugh, D. E., Loftus, R. T., Bradley, D. G., Sharp, P. M., and Cunningham, P. 
(1994). Microsatellite DNA variation within and among European cattle breeds. 
Proc R Soc Lond B Biol Sci 256(1345), 25-31. 
MacHugh, D. E., Loftus, R. T., Cunningham, P., and Bradley, D. G. (1998). Genetic 
structure of seven European cattle breeds assessed using 20 microsatellite 
markers. Anim Genet 29(5), 333-40. 
 
 204
 MacHugh, D. E., Shriver, M. D., Loftus, R. T., Cunningham, P., and Bradley, D. G. 
(1997). Microsatellite DNA variation and the evolution, domestication and 
phylogeography of taurine and zebu cattle (Bos taurus and Bos indicus). 
Genetics 146(3), 1071-86. 
Manly B.F.J. (1986). Randomization and regression methods for testing for associations 
with geographical, environmental and biological distances between populations. 
Res. Pop. Ecol. 28, 201-218. 
Manly B.F.J. (1994). Multivariate statistical methods: a primer, 2nd ed. New York: 
Chapman & Hall. 
Mannen, H., Nagata, Y., and Tsuji, S. (2001). Mitochondrial DNA reveal that domestic 
goat (Capra hircus) are genetically affected by two subspecies of bezoar (Capra 
aegagurus). Biochem Genet 39(5-6), 145-54. 
Marshall, F. (2000). The origion and spread of domestic animals in East Africa.  In: The 
origions and development of African livestock: archaeology, genetics, 
linguistics and ethinography (Edited by Blench, R.M. and MacDonald, K.C), 
UCL press. London, New York. pp 191-221. 
Martin-Burriel, I., Garcia-Muro, E., and Zaragoza, P. (1999). Genetic diversity analysis 
of six Spanish native cattle breeds using microsatellites. Anim Genet 30(3), 177-
 82. 
Mason, I. L. (1984). Goat. In: Mason I.L. (Ed), Evolution of domestic animals. 
Longman, Essex, England. pp 85-99. 
Mburu, D. N., Ochieng, J. W., Kuria, S. G., Jianlin, H., Kaufmann, B., Rege, J. E., and 
Hanotte, O. (2003). Genetic diversity and relationships of indigenous Kenyan 
camel (Camelus dromedarius) populations: implications for their classification. 
Anim Genet 34(1), 26-32. 
Meester, J., and Setzer., H. W. (1977). The mammals of Africa: an identification 
manual. Smithson. Inst. Press, Washington, D.C. 
Meghen, C., MacHugh, D. E., and Bradley, D. G. (1994). Genetic characterization and 
West Africa cattle. World Animal Review 78(1): 59-66. 
Miller, C. R., and Waits, L. P. (2003). The history of effective population size and 
genetic diversity in the Yellowstone grizzly (Ursus arctos): implications for 
conservation. Proc Natl Acad Sci U S A 100(7), 4334-9. 
Mason, I.L. 1981. Breeds: In: Goat production. Ed: by C Gall. Chapter 3, Academic  
press. 
 205
 Mills, S. L., and Smouse, P. E. (1994b). Demographic consequences of inbreeding in 
reminant populations. American Naturalist 144, 412-431. 
Minelli, A. (1993). Biological systematics. The state of the art. Capman & Hall. 
London, Glasgow, New York, Tokyo, Melbourne, Madras. pp 387. 
Moazami-Goudarzi, K., and Laloe, D. (2002). Is a multivariate consensus representation 
of genetic relationships among populations always meaningful? Genetics 162(1), 
473-84. 
Moazami-Goudarzi, K., Laloe, D., Furet, J. P., and Grosclaude, F. (1997). Analysis of 
genetic relationships between 10 cattle breeds with 17 microsatellites. Anim 
Genet 28(5), 338-45. 
Mommens, G., Peelman, L. J., Vanzeveren, A., G. D. 'leteren, and Wissocq, N. (1999). 
Microsatellite variation between an African and five European taurine breeds 
results in a geographical phylogenetic tree with a bison outgroup. J. Ani. Breed. 
Genet. 116:5, 325-330. 
Moore, H. D. M. (1985). Storage of gametes and embryos. In: Advances in Animal 
consevation. The proceedings of a symposium held at the Zoological Society of 
London on 31 May - 1 June, 1984. No.54: 137-148. 
Moore, S. S., Sargeant, L. L., King, T. J., Mattick, J. S., Georges, M., and Hetzel, D. J. 
(1991). The conservation of dinucleotide microsatellites among mammalian 
genomes allows the use of heterologous PCR primer pairs in closely related 
species. Genomics 10(3), 654-60. 
Mopper, S., Mitton, J. B., Whitham, T. J., Cobb, N. S., and Christensen, K. M. (1991). 
Genetic differentiation and heterozygosity in pinyon pine associated with 
resistance to herbivory and environmental stress. Evolution 45: 989-999. 
Morin, P. A., Moore, J. J., Chakraborty, R., Jin, L., Goodall, J., and Woodruff, D. S. 
(1994). Kin selection, social structure, gene flow, and the evolution of 
chimpanzees. Science 265(5176), 1193-201. 
Muzzolini, A. (2000). Livestock in Saharan art. In Blench, R.M.and Macdonald, K.C. 
(Eds.). The origion and development of African livestock, Archaeology, 
genetics, linguistics and ethnography. University College London press, New 
York. pp. 87-110. 
 
 
 206
 Nakamura, Y., Leppert, M., O'Connell, P., Wolff, R., Holm, T., Culver, M., Martin, C., 
Fujimoto, E., Hoff, M., Kumlin, E., and et al. (1987). Variable number of 
tandem repeat (VNTR) markers for human gene mapping. Science 235(4796), 
1616-22. 
Nei, M. (1972). Genetic distance between populations. American Naturalist 106, 283-
 292. 
Nei, M. (1973). Analysis of gene diversity in subdivided populations. Proc Natl Acad 
Sci U S A 70(12), 3321-3. 
Nei, M. (1978). Estimation of average heterozygosity and genetic distance from a small 
number of individuals. Genetics 89, 583-590. 
Nei, M. (1984). Molecular Evolutionary Genetics. Colombia University press/New 
York. 
Nei, M. (1987). Molecular Evolutionary Genetics. Columbia University press, New 
York. pp 506. 
Nei, M. (1996). Phylogenetic analysis in molecular evolutionary genetics. Annu Rev 
Genet 30, 371-403. 
Nei, M., Tajima, F., and Tateno, Y. (1983). Accuracy of estimated phylogenetic trees 
from molecular data. II. Gene frequency data. J Mol Evol 19(2), 153-70. 
Nei, M., and Takezaki, N. (1994). Estimation of genetic distances and phylogenetic 
trees fromDNA analysis. In: proceeding of the 5th world congress of genetic 
applied to livestock production. 7-12 August, 1994 Guelph, Canada. Vol 21: 
405-412. 
Neigel, J. E. (1997). A comparison of alternative strategies for estimating gene flow 
from genetic markers. Annu. Rev. Ecol. Syst. 28:, 105-128. 
Nelson, K., and Soule, M. E. (1987). Genetic conservation of exploited fishes. In: 
Population genetics and fishery management (Edited by Nils Ryman and Fred 
Utter). Washington press. Seattle and London. pp 345-368. 
Nguyen, T. C. (1990). Genetic systems of red cell blood groups in goats. Anim Genet 
21(3), 233-45. 
Nigatu Alemayehu (1994). Characterization of indigenous goat types of Eritrea , 
Northern and Western Ethiopia . MSc thesis, Alemaya University of 
Agriculture, Alemaya, Ethiopia. 136pp. 
 
 207
 Notter, D. R., Mariante, A. d. S., and Sheng, Z. (1994). Modern approach to active 
conservation of domestic animal diversity. In: Proceeding of the 5th World 
Congress of Genetic Applied toLivestock Production. 7-12 August, 1994 
Guelph, Canada. Vol 21: 509-516." 
Okoma, M. A., Rege, J. E. O., Teale, A., and Hanotte, O. (1998). Genetic 
characterization of indigenous East African cattle breeds using microsatellite 
DNA markers. In:. Proceeding of the 6th world congress on Genetics Applied to 
livestock production, Armidale, Australia, January 11-16, 1998. 
Oldenbroek, J. K. E. (1999). ''Genebanks and the conservation of farm animal genetic 
resources'' DLO, The Netherlands. 
Ohta T and Kimura M. (1973). A model of mutation appropriate to estimate the number 
of electrophoretically detectable alleles in a finite population. Genet. Res. Cambr. 
22: 201-204. 
Ota, T. (1993). ''Dispan: Genetic Distance and Phylogenetiv Analysis ''. Pennsylvania 
State University, Pennsylvania. 
Paetkau, D., Calvert, W., Stirling, I., and Strobeck, C. (1995). Microsatellite analysis of 
population structure in Canadian polar bears. Mol Ecol 4(3), 347-54. 
Paetkau, D., and Strobeck, C. (1995b). The molecular basis and evolutionary history of 
a microsatellite null allele in bears. Mol Ecol 4(4), 519-20. 
Page, R. D. M., and Holmes, E. C. (1998). Molecular evolution. A phylogenetic 
approach. Blackwell science. University press, Cambridge. pp 346. 
Palsson, S. (2000). Microsatellite variation in Daphnia pulex from both sides of the 
Baltic Sea. Mol Ecol 9(8), 1075-88. 
Pandey, A. K. (1999). Studies on DNA polymorphism in buffaloes using microsatellite 
markers. M.Sc. Thesis, NDRI, Deemed University, Karnal. 
Park, S. D. E. (2001). Trypanotolerance in West African Cattle and the Population 
Genetic Effects of Selection [ Ph.D. thesis ] University of Dublin. 
Pemberton, J. M., Albon, S. D., Guinness, F. E., Clutton-Brock, T. H., and Berry, R. J. 
(1988). Genetic variation and juvenile survival in red deer. Evolution 42: 921-
 934. 
Pepin, L., and Nguyen, T. C. (1994). Blood groups and protein polymorphisms in five 
goat breeds (Capra hircus). Anim Genet 25(5), 333-6. 
 
 208
 Piry S, Alapetite A, Cornuet, J.-M., Paetkau D, Baudouin, L., and Estoup, A. (2004). 
GeneClass2: a software for genetic assignment and first generation migrants 
detection. Journal of Heredity (accepted). 
Porter, V. (1996). Goats of the world (Farming press, Ipswich, U.K.). 
Powell, W., Morgante, M., Dndre, C., Hanafey, M., Vogel, J., Tingey, S. V., and 
Rafalski, J. A. (1996). The comparison of RFLP, RAPD, AFLP and SSR 
(Microsatellite) markers for germplasm analysis. Molecular Breeding 2:, 225-
 238. 
Pringle, H. (1998). NEOLITHIC AGRICULTURE:Reading the Signs of Ancient 
Animal Domestication. Science 282(5393), 1448. 
Pringle, H. (1998b). NEOLITHIC AGRICULTURE:Reading the Signs of Ancient 
Animal Domestication. Science 282(5393), 1448-. 
Pritchard, J. K., Stephens, M., and Donnelly, P. (2000). Inference of population 
structure using multilocus genotype data. Genetics 155(2), 945-59. 
Queller, D. C., Strassmann, J. E., and Hughes, C. R. (1993). Microsatellite and Kinship. 
Trends in Ecology and Evolution 8, 285-288. 
Ralls, K., Ballou, J. D., and Templeton, A. (1988). Estimation of lethal equivalents and 
the costs of inbreeding in mammals. Conservation Biology 2: 185-193. 
Rannala, B., and Mountain, J. L. (1997). Detection immigration by using multilocus 
genotypes. Proceeding of the National Accademy of Sciences USA 94: 9197-
 9221. 
Raymond, M., and Riousset, F. (1995). GENEPOP (version 1.2): Population genetics 
software for exact test and ecumenicsm. J.Heredity, 86, 248-249. 
Raymond, M., and Rousset, F. (1995). An exact test for population differentiation. 
Evolution 49, 1280-1283. 
Rege, J. E. O. (1992). Indigenous African small ruminants: a case for characterization 
and improvement. In: proceedings of the second Biennial conference of the 
African small ruminant research Network, 7 - december, 1992, AICC, Arusha, 
Tanzania. pp 205-211. 
Rosenberg, N. A. (2004), Distruct: a program for the graphical display of structure 
results. http://www.cmb.usc.edu/~noahr/distruct.html).   
 
 
 209
 Rege, J. E. O., Muigai, A. W. T., Chenyambuga, S., Kierstein, G., and Hanotte, O. 
(2002). Assesment of genetic diversity in African small ruminants: present status 
and future prospects. In: 53rd Annual meeting of the European association of 
animal production, Cairo, Egypt, September 1-4, 2002. 
Reist-Marti, S. B., Simianer, H., Gibson, J., Hanotte, O., and O., R. J. E. (2003). 
Weitzman's Approach and Conservation of Breed Diversity: an Application to 
African Cattle Breeds. Conservation Biology 17:5p, 1299-1311. 
Rice, W. R. (1989). Analysing tables of ststistical tests. Evolution 43:, 223-225. 
Richards, R. I., and Sutherland, G. R. (1994). Simple repeat DNA is not replicated 
simply. Nat Genet 6(2), 114-6. 
Ritz, L. R., Glowatzki-Mullis, M. L., MacHugh, D. E., and Gaillard, C. (2000). 
Phylogenetic analysis of the tribe Bovini using microsatellites. Anim Genet 
31(3), 178-85. 
Roderick GK (1997). Geographic structure of insect populations: gene flow, 
phylogeography, and their uses. Ann Rev Entomol 41, 325-352. 
Rogers, J. S. (1972). Measures of genetic similarity and genetic distances. Studies in 
Genetics, Univ. Texas Publ., 7213, 145?153. 
Rousset, F. (1997). Genetic differentiation and estimation of gene flow from F-statistics 
under isolation by distance. Genetics 145(4), 1219-28. 
Rousset, F., and Raymond, M. (1995). Testing heterozygote excess and deficiency. 
Genetics 140(4), 1413-9. 
Roy, M. S., Geffen, E., Smith, D., Ostrander, E. A., and Wayne, R. K. (1994). Patterns 
of differentiation and hybridization in North American wolflike canids, revealed 
by analysis of microsatellite loci. Mol Biol Evol 11(4), 553-70. 
Rzhetsky, A., and Nei, M. (1992). Statistical properties of the ordinary least-squares, 
generalized least-squares, and minimum-evolution methods of phylogenetic 
inference. J Mol Evol 35(4), 367-75. 
Saiki, R. K., Gelfand, D. H., Stoffel, S., Scharf, S. J., Higuchi, R., Horn GT, Mullis, K. 
B., and H.A., E. (1988). Primer-directed enzymatic amplification of DNA with a 
thermostable DNA polymerase. Science. 239, 487-91. 
Saitbekova, N., Gaillard, C., Obexer-Ruff, G., and Dolf, G. (1999). Genetic diversity in 
Swiss goat breeds based on microsatellite analysis. Anim Genet 30(1), 36-41. 
Saitou, N. (1988). Property and efficiency of the maximum likelihood method for 
molecular phylogeny. J Mol Evol 27(3), 261-73. 
 210
 Saitou, N., and Imanishi, T. (1989). Relative efficiencies of the Fitch-Margoliash, 
Maximum-Parsimony, Maximum-Likelihood, Minimum-Evolution and 
Neighbour-Joining methods of phylogenetic tree construction in obtaining the 
correct tree. Molecular Biology and Evolution 6, 514-525. 
Saitou, N., and Nei, M. (1987). The neighbor-joining method: a new method for 
reconstructing phylogenetic trees. Mol Biol Evol 4(4), 406-25. 
Schlotterer, C., and Pemberton, J. (1994). In molecular Ecology and Evolution: 
approaches and Applications (eds. B. Schierwater, R., Streit, G.P., Wagner and 
R. Desalle), pp.203-14. Bir Khauser Verlag Basel, Switzerland. 
Schlotterer, C., and Tautz, D. (1992). Slippage synthesis of simple sequence DNA. 
Nucleic Acids Res 20(2), 211-5. 
Schwartz, M. K., Mills, L. S., Ortega, Y., Ruggiero, L. F., and Allendorf, F. W. (2003). 
Landscape location affects genetic variation of Canada lynx (Lynx canadensis). 
Mol Ecol 12(7), 1807-16. 
Seehausen, O., Koetsier, E., Schneider, M. V., Chapman, L. J., Chapman, C. A., Knight, 
M. E., Turner, G. F., van Alphen, J. J., and Bills, R. (2003). Nuclear markers 
reveal unexpected genetic variation and a Congolese-Nilotic origin of the Lake 
Victoria cichlid species flock. Proc R Soc Lond B Biol Sci 270(1511), 129-37. 
Selander, R. K. (1983). Evolutionary consequences of inbreeding. In: Genetics and 
Conservation: A reference for managing wild animal and plant populations. 
(Edited by C.M. Schonewald-Cox; S.M. Chambers; B.MacBryde and 
L.Thomas). Benjamin/Cummings, menlopark, CA. pp 201-215.." 
Serikawa, T., Montagutelli, X., Simon-Chazottes, D., and Guenet, J. L. (1992). 
Polymorphisms revealed by PCR with single, short-sized, arbitrary primers are 
reliable markers for mouse and rat gene mapping. Mamm Genome 3(2), 65-72. 
Shackleton, D. M., ed. (1997). in Wild Sheep and Goats and their Relatives: Status 
Survey and Conservation Action Plan for Caprinae (IUCN, Gland, Swizerland). 
Shriver, M. D., Jin, L., Boerwinkle, E., Deka, R., Ferrell, R. E., and Chakraborty, R. 
(1995). A novel measure of genetic distance for highly polymorphic tandem 
repeat loci. Mol Biol Evol 12(5), 914-20. 
Simm, G. (1998). Genetic improvement of cattle and sheep. Farming press. London. pp 
433. 
 
 211
 Skalova, H., Pechackova, S., Suzuki, J., Herben, T., Hara, T., Hadincova, V., and 
Krahulec, F. (1997). Within population genetic differentiation in traits affecting 
clonal growth: <i>Festuca rubra</i> in a mountain grassland. J Evolution Biol 
10(3), 383-406. 
Slatkin, M. (1985). Gene flow in natural populations. Annu Rev Ecol Syst 16: 393-430. 
Slatkin, M. (1995). A measure of population subdivision based on microsatellite allele 
frequencies. Genetics 139(1), 457-62. 
Smith, A. B. (2000). "The origion and spread of domestic animals in Southern Africa.  
In: The origions and development of African livestock: archaeology, genetics, 
linguistics and ethinography (Edited by Blench, R.M. and MacDonald, K.C), 
UCL press. London, New York. pp 222-231." 
Smith, C. (1984). Estimated costs of genetic conservation in farm livestock. In: Animal 
Genetic Resources Conservation by Management, Data Banks and Training. 
FAO Animal Production and Health Paper 44/1: 21?30. 
Sneath, P. H. A., and Sokal, R. R. (1973). In: Numerical Taxonomy (pp; 230-234), 
W.H. Freeman and Copany, San Francisco, California, USA. 
Soller, M., and Beckmann, J. S. (1982). In: Proc. 2nd world Cong. Genet. Appl 
livestock prod. Madrid, 6: 396-404. 
Soule, M. E. (1980). Thresholds for survival: maintaining fitness and evolutionary 
potential. In: Conservation biology (Edited by M.E. Soule and B.A. Wilcox). 
Sinauer, Sunderland, Massachusetts, USA. pp 151-168. 
Southern, E. M. (1975). Detection of specific sequences among DNA fragments 
separated by gel electrophoresis. J Mol Biol 98(3), 503-17. 
Stallings, R. L., Ford, A. F., Nelson, D., Torney, D. C., Hildebrand, C. E., and Moyzis, 
R. K. (1991). Evolution and distribution of (GT)n repetitive sequences in 
mammalian genomes. Genomics 10(3), 807-15. 
Sultana, S., Mannen, H., and Tsuji, S. (2003). Mitochondrial DNA diversity of Pakistani 
goats. Animal Genetics 34(6), 417-421. 
Sunnucks, P. (2000). Efficient genetic markers for population biology. Trends in 
Ecology and Evolution 15(5), 199-203. 
Swofford, D. L., and Oslen, G. J. (1990). Phylogeny reconstruction. In: Molecular 
systematics (Edited by D.M. Hills and C. Mortiz). Sinauer, Sunderland, 
Massachusetts. pp 411-501. 
 212
 Swofford, D. L., Oslen, G. J., Waddell, P. J., and Hills, D. M. (1996). Phylogenetic 
inference.  In: Molecular systematics 2nd ed  (Edited by D.M. Hills and C. 
Mortiz). Sinauer, Sunderland, Massachusetts. pp 407-514. 
Taberlet, P., Camarra, J. J., Griffin, S., Uhres, E., Hanotte, O., Waits, L. P., Dubois-
 Paganon, C., Burke, T., and Bouvet, J. (1997). Noninvasive genetic tracking of 
the endangered Pyrenean brown bear population. Mol Ecol 6(9), 869-76. 
Takada, T., Kikkawa, Y., Yonekawa, H., Kawakami, S., and Amano, T. (1997). Bezoar 
(Capra aegagrus) is a matriarchal candidate for ancestor of domestic goat (Capra 
hircus): evidence from the mitochondrial DNA diversity. Biochem Genet 35(9-
 10), 315-26. 
Takezaki, N., and Nei, M. (1996). Genetic distances and reconstruction of phylogenetic 
trees from microsatellite DNA. Genetics 144(1), 389-99. 
Tanksley, S. D., Young, N. D., Peterson, A. H., and Bonierbale, M. W. (1989). RFLP 
mapping in plant breeding: new tools for an old science. Biotechnology 7, 257-
 264. 
Tateno, Y., Nei, M., and Tajima, F. (1982). Accuracy of estimated phylogenetic trees 
from molecular data. I. Distantly related species. J Mol Evol 18(6), 387-404. 
Tauz, D. (1989). Hypervariability of simple sequences as a general source of 
polymorphic DNA markers. Nucleic Acid Research 17, 6463-6471. 
Taylor, A. C., Sherwin, W. B., and Wayne, R. K. (1994). Genetic variation of 
microsatellite loci in a bottlenecked species: the northern hairy-nosed wombat 
Lasiorhinus krefftii. Mol Ecol 3(4), 277-90. 
Teale, A. J., Wambugu, J., Gwakisa, P. S., Stranzinger, G., Bradley, D., and Kemp, S. J. 
(1995). A polymorphism in randomly amplified DNA that differentiates the Y 
chromosomes of Bos indicus and Bos taurus. Anim Genet 26(4), 243-8. 
Templeton, A. R. (1986). Coadaptation and outbreeding depression. In: Conservation 
Biology. The Science of scarcity and diversity (Edited by M.E. Soule). Sinauer, 
Sunderland, Massachusetts. pp 105-116. 
Thaon d'Arnoldi, C., J. L. Foulley, and Ollivier., L. (1998). An overview of the 
Weitzman approach to diversity. Genetics Selection Evolution 30, 149-161. 
Tiedemann, R., Hardy, O., Vekemans, X., and Milinkovitch, M. C. (2000). Higher 
impact of female than male migration on population structure in large mammals. 
Mol Ecol 9(8), 1159-63. 
 213
 Uerpmann, H.-P. (1996). Animal Domestication-accident or intention? In: Harris, D.R. 
(Ed), The origions and spread of agriculture and pastoralism in Euroasia. 
Smithsonian press, Washington D.C pp 227-237. 
Vaiman, D., Imam-Ghali, M., Moazami-Goudarzi, K., Guerin, G., Grohs, C., Leveziel, 
H., and Saidi-Mehtar, N. (1994a). Conservation of a syntenic group of 
microsatellite loci between cattle and sheep. Mamm Genome 5(5), 310-4. 
Vaiman, D., Mercier, D., Moazami-Goudarzi, K., Eggen, A., Ciampolini, R., Lepingle, 
A., Velmala, R., Kaukinen, J., Varvio, S. L., Martin, P., and et al. (1994b). A set 
of 99 cattle microsatellites: characterization, synteny mapping, and 
polymorphism. Mamm Genome 5(5), 288-97. 
Vaiman, D., Schibler, L., Bourgeois, F., Oustry, A., Amigues, Y., and Cribiu, E. P. 
(1996). A genetic linkage map of the male goat genome. Genetics 144(1), 279-
 305. 
Valdes, A. M., Slatkin, M., and Freimer, N. B. (1993). Allele frequencies at 
microsatellite loci: the stepwise mutation model revisited. Genetics 133(3), 737-
 49. 
Van Oosterhout, C., Hutchinson, B., Wills, B., and Shipley, P. (2003). Micro-Checker: 
Microsatellite Data Checking Software (Last updated 4 August 2003). Available 
from http://www.microchecker.hull.ac.uk. 
Vos, P., Hogers, R., Bleeker, M., Reijans, M., van de Lee, T., Hornes, M., Frijters, A., 
Pot, J., Peleman, J., Kuiper, M., and et al. (1995). AFLP: a new technique for 
DNA fingerprinting. Nucleic Acids Res 23(21), 4407-14. 
Wayne, R. K., Gilbert, D., and Lehman, N. (1991). Conservation genetics of 
endangered Isle Royale gray wolf. Conservation Biology: 5, 41-51. 
Weber, J. L. (1990). Informativeness of human (dC-dA)n.(dG-dT)n polymorphisms. 
Genomics 7(4), 524-30. 
Weber, J. L., and May, P. E. (1989). Abundant class of human DNA polymorphisms 
which can be typed using the polymerase chain reaction. Am J Hum Genet 44(3), 
388-96. 
Weir, B. S. (1996). Genetic Data Analysis II: Methods for Discrete Population Genetic 
Data. 
Weir, B. S., and Basten, C. J. (1990). Sampling strategies for distances between DNA 
sequences. Biometrics 46(3), 551-82. 
 214
 Weir, B. S., and Cockerham, C. C. (1984). Estimating F-statistics for the analysis of 
population structure. Evolution 38, 1358-1370. 
Weissenbach, J., Gyapay, G., Dib, C., Vignal, A., Morissette, J., Millasseau, P., 
Vaysseix, G., and Lathrop, M. (1992). A second-generation linkage map of the 
human genome. Nature 359(6398), 794-801. 
Weitzman M.L. (1998). The Noah's ark problem, Econometrica 66, 1279-1298. 
Weitzman, M. L. (1992). On diversity, Quarterly Journal of Economics 107, 363-405. 
Weitzman, M. L. (1993). What to preserve? An application of diversity theory to crane 
conservation, Quarterly Journal of Economics 108, 157-183. 
Welsh, J., and McClelland, M. (1990). Fingerprinting genomes using PCR with 
arbitrary primers. Nucleic Acids Res 18(24), 7213-8. 
Williams, J. G., Hanafey, M. K., Rafalski, J. A., and Tingey, S. V. (1993). Genetic 
analysis using random amplified polymorphic DNA markers. Methods Enzymol 
218, 704-40. 
Williams, J. G., Kubelik, A. R., Livak, K. J., Rafalski, J. A., and Tingey, S. V. (1990). 
DNA polymorphisms amplified by arbitrary primers are useful as genetic 
markers. Nucleic Acids Res 18(22), 6531-5. 
Williams, J. G. K., Rafalski, J. A., and Tingey, S. V. (1993). In Wu R (ed), Methods in 
Enzymology 218:. Methods Enzymol. 86:, 704-740. 
Wilson, G. A., and Strobeck, C. (1999). Genetic variation within and relatedness among 
wood and plains bison populations. Genome 42(3), 483-96. 
Wilson, T. R. (1991). FAO animal health and production paper no. 88, pp. 231. 
Wollny, C. B. A. (2003). The need to conserve farm animal genetic resources in Africa: 
should policy makers be concerned? Ecological Economics 45: 341-351 
Workneh Ayalew (1992). Preliminary survey of indigenous goat types and goat 
husbandery practices in Southern Ethiopia. MSc thesis, Alemaya University of 
Agriculture, Alemaya, Ethiopia. 153pp. 
Wright, S. (1943). ISOLATION BY DISTANCE. Genetics 28(2), 114-138. 
Wright, S. (1969). Evolution and the genetics of populations. University of Chicago 
press, Chicago. 
Yadav, B. R., S.De, and Mitra, A. (1998). Bovine microsatellite primers reveal DNA 
polymorphism in river buffalo (Bubalus Bubalis). In: Indo-Brazilian Workshop 
on Agricultural Biotechnology held at NDRI, Karnal, April 27-29, 1998, p-26. 
 
 215
  
Yadav, B. R., Thakur, M. K., Singhal, A., and Singh R.P (2001). Genetic variation on 
three breeds of water buffalo analyzed by RAPD primers.  In: 5th ADNAT 
convention:  A two days symposium on perspectives in genome analysis held at 
CCMB, Hyderabad, Feb.23-24. Abst. P. 40. 
Yang, L. Zhao, S. H. Li, K. Peng, Z. Z. and Montgomery, G. W. (1999). 
Determination of genetic relationships among five indigenous Chinese goat 
breeds with six microsatellite markers. Anim. Genet 30(6): 452-5. 
Yang, S.L., Wang, Z.G., Liu, B., Zhang, G.X., Zhao, S.H., Yu, M., Fan, B., Li, 
M.H., Xiong, T.A., and Li, K. (2003). Genetic variation and relationships of 
eighteen Chinese indigenous pig breeds. Genet Sel. Evol. 35(6): 657-71. 
Zeder, M. A. and B. Hesse (2000). "The initial domestication of goats (Capra hircus) in 
the Zagros mountains 10,000 years ago. Science 287(5461): 2254-7. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
APPENDICES 
 216
 Appendix 1. Summary of linkage disequilibrium between all markers across all 20 populations. 
 Linkage between markers ABG ABA AFA CEH GUM HAH KEF LES NWH SES WGJ BGA SEA GUB TSW ALP ARD BAL HAI MON 
1 BM1818 & BMS1494          *   *        
2 BM1818 & ILSTS05              *       
3 BM1818 & ILSTS11   *  ** *     *          
4 BM1818 & ILSTS44     **         *       
5 BM1818 & ILSTS87                     
6 BM1818 & INRA05   *                  
7 BM1818 & INRA132       **    **        * * 
8 BM1818 & INRA63             *       * 
9 BM1818 & MAF209   **     *  *           
10 BM1818 & MAF35                     
11 BM1818 & MAF65                     
12 BM1818 & OarAE129  **         *          
13 BM1818 & OarFCB304           * *       *  
14 BM1818 & SRCRSP03            *         
15 BMS1494 & ILSTS05                *    * 
16 BMS1494 & ILSTS11    **      **           
17 BMS1494 & ILSTS44                *     
18 BMS1494 & ILSTS87         **            
19 BMS1494 & INRA05  *      *             
20 BMS1494 & INRA132                     
21 BMS1494 & INRA63                     
22 BMS1494 & MAF209  *  *                 
23 BMS1494 & MAF35        *        *     
24 BMS1494 & MAF65     *   *    *      *   
25 BMS1494 & OarAE129     *   *             
26 BMS1494 & OarFCB304  *           * *       
27 BMS1494 & SRCRSP03                     
28 ILSTS05 & ILSTS11                     
                      
 217
  Linkage between markers ABG ABA AFA CEH GUM HAH KEF LES NWH SES WGJ BGA SEA GUB TSW ALP ARD BAL HAI MON 
29 ILSTS05 & ILSTS44                     
30 ILSTS05 & ILSTS87                    ** 
31 ILSTS05 & INRA05              *  *     
32 ILSTS05 & INRA132                     
33 ILSTS05 & INRA63  *                  * 
34 ILSTS05 & MAF209          *           
35 ILSTS05 & MAF35    *                * 
36 ILSTS05 & MAF65                     
37 ILSTS05 & OarAE129  *             *      
38 ILSTS05 & OarFCB304                    ** 
39 ILSTS05 & SRCRSP03              **  * *    
40 ILSTS11 & ILSTS44                     
41 ILSTS11 & ILSTS87              *       
42 ILSTS11 & INRA05                     
43 ILSTS11 & INRA132                     
44 ILSTS11 & INRA63 *   *                 
45 ILSTS11 & MAF209          *           
46 ILSTS11 & MAF35                     
47 ILSTS11 & MAF65                     
48 ILSTS11 & OarAE129          *           
49 ILSTS11 & OarFCB304  *     **              
50 ILSTS11 & SRCRSP03          *           
51 ILSTS44 & ILSTS87  *          *         
52 ILSTS44 & INRA05                     
53 ILSTS44 & INRA132        **        *   * * 
54 ILSTS44 & INRA63              **       
55 ILSTS44 & MAF209              *       
56 ILSTS44 & MAF35           *          
57 ILSTS44 & MAF65           *          
58 ILSTS44 & OarAE129       *  *            
59 ILSTS44 & OarFCB304    *  *               
 218
  Linkage between markers ABG ABA AFA CEH GUM HAH KEF LES NWH SES WGJ BGA SEA GUB TSW ALP ARD BAL HAI MON 
60 ILSTS44 & SRCRSP03                     
61 ILSTS87 & INRA05    *  *               
62 ILSTS87 & INRA132             * *       
63 ILSTS87 & INRA63     *     *           
64 ILSTS87 & MAF209                     
65 ILSTS87 & MAF35                     
66 ILSTS87 & MAF65                     
67 ILSTS87 & OarAE129        *          *  * 
68 ILSTS87 & OarFCB304                     
69 ILSTS87 & SRCRSP03            **         
70 INRA05 & INRA132                     
71 INRA05 & INRA63             **        
72 INRA05 & MAF209             *     *   
73 INRA05 & MAF35  * *                  
74 INRA05 & MAF65                     
75 INRA05 & OarAE129   *                **  
76 INRA05 & OarFCB304  *                   
77 INRA05 & SRCRSP03                     
78 INRA132 & MAF209                     
79 INRA132 & MAF35            *         
80 INRA132 & MAF65   *           **  *     
81 INRA132 & OarAE129                *     
82 INRA132 & OarFCB304         *            
83 INRA132 & SRCRSP03                 *    
84 INRA63 & INRA132 *                    
85 INRA63 & MAF209              *       
86 INRA63 & MAF35                **     
87 INRA63 & MAF65   *    *              
88 INRA63 & OarAE129  *                   
89 INRA63 & OarFCB304                     
90 INRA63 & SRCRSP03                     
 219
  Linkage between markers ABG ABA AFA CEH GUM HAH KEF LES NWH SES WGJ BGA SEA GUB TSW ALP ARD BAL HAI MON 
91 MAF209 & OarAE129                     
92 MAF209 & OarFCB304 *       **    * *        
93 MAF209 & SRCRSP03                     
94 MAF35 & MAF209       *              
95 MAF35 & MAF65      *      *         
96 MAF35 & OarAE129                 *    
97 MAF35 & OarFCB304 **        *            
98 MAF35 & SRCRSP03           *          
99 MAF65 & MAF209                     
100 MAF65 & OarAE129                     
101 MAF65 & OarFCB304                     
102 MAF65 & SRCRSP03                   *  
103 OarAE12 & OarFCB304        *         *    
104 OarAE12 & SRCRSP03                     
105 OarFCB3 & SRCRSP03                     
* = P < 0.05; ** = P < 0.01. 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, LES ? Long-Ear Somali, NWH  ? North-West Highland, SES ? Small-Ear Somali, 
WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? Baladie, HAI- Hair, MON? Mongolia. 
 
 
 
 
 
 
 
 
 220
 Appendix 2. Matrix of all pair-wise combinations of loci showing the linkage disequilibrium (for 20 populations).  P values condensed 
across all populations.  For simplicity, the matrix only shows levels of significance. 
 BM 1818 
BM 
1494 
ILSTS 
05 
ILSTS 
11 
ILSTS 
44 
ILSTS 
87 
INRA 
05 
INRA 
63 
INRA 
132 
MAF 
35 
MAF 
65 
MAF 
209 
OarAE 
129 
OarFCB 
304 
SRCRSP 
03 
BM1818                
BMS1494                
ILSTS05                
ILSTS11 * *              
ILSTS44 *               
ILSTS87                
INRA05                
INRA63                
INRA132      *          
MAF35                
MAF65                
MAF209 *               
OarAE129 * *              
OarFCB304            **    
SRCRSP03   **             
* = P < 0.05; ** = P < 0.01. 
 
 
 221
 Appendix 3. Allele frequency for locus BM1818 for all the 20 populations. 
Allele 244 246 250 252 254 256 258 260 262 264 266 268 
ABG 1.09 - - 7.61 - 7.61 51.09 7.61 13.04 6.52 5.43 - 
ABA 3.26 - - 5.43 - 5.43 40.22 7.61 20.65 6.52 4.35 6.52 
AFA - 1.14 - 1.14 1.14 3.41 65.91 9.09 3.41 6.82 5.68 2.27 
CEH  5.00 - - 5.00 1.25 2.50 51.25 6.25 10.00 11.25 7.50 - 
GUM - - - 8.14 - 8.14 41.86 13.95 10.47 8.14 8.14 1.16 
HAH - - - 2.50 2.50 7.50 58.75 11.25 10.00 5.00 2.50 - 
KEF - - - 6.58 - 5.26 68.42 7.89 5.26 3.95 2.63 - 
LES 3.13 1.04 1.04 6.25 5.21 9.38 47.92 6.25 11.46 3.13 4.17 1.04 
NWH 1.11 - - 13.33 2.22 1.11 45.56 5.56 20.00 6.67 4.44 - 
SES 1.16 - - 6.98 4.65 4.65 46.51 10.47 15.12 6.98 2.33 1.16 
WGJ 1.09 - - 3.26 2.17 2.17 54.35 5.43 13.04 7.61 4.35 6.52 
BGA - - 1.43 2.86 8.57 18.57 41.43 12.86 7.14 4.29 1.43 1.43 
SEA - - - 5.13 12.82 6.41 52.56 7.69 8.97 - 6.41 - 
GUB 6.52 - - 2.17 6.52 2.17 36.96 8.70 - 9.78 22.83 4.35 
TSW - - - 2.08 6.25 27.08 45.83 4.17 6.25 4.17 4.17 - 
ALP - - - - - 35.48 30.65 6.45 9.68 11.29 6.45 - 
ARD - - - - - 12.00 52.00 18.00 12.00 4.00 2.00 - 
BAL - - 11.36 4.55 13.64 4.55 34.09 18.18 6.82 4.55 2.27 - 
HAI - - 9.68 9.68 14.52 8.06 25.81 19.35 9.68 1.61 1.61 - 
MON - - - 10.81 21.62 14.86 - 16.22 16.22 2.70 14.86 2.70 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, 
LES ? Long-Ear Somali, NWH  ? North-West Highland, SES ? Small-Ear Somali, WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? Small 
East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? Baladie, HAI- Hair, MON? Mongolia. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 222
 Appendix 4. Allele frequency for locus BMS1494 for all the 20 populations. 
Allele 244 246 256 264 270 276 280 282 284 288 
ABG - - 9.78 - - 65.22 18.48 - 6.52 - 
ABA - - 4.55 - - 57.95 36.36 - 1.14 - 
AFA - - - 1.14 - 71.59 22.73 - 4.55 - 
CEH  1.25 1.25 11.25 - - 65.00 21.25 - - - 
GUM - - 2.33 - - 82.56 13.95 - 1.16 - 
HAH - - 3.75 - - 60.00 33.75 - 2.50 - 
KEF - - 5.26 - - 57.89 21.05 - 14.47 1.32 
LES - - 2.08 - - 68.75 23.96 - 4.17 1.04 
NWH - - 6.82 - - 62.50 29.55 1.14 - - 
SES - - - - - 61.63 34.88 - 3.49 - 
WGJ - - 3.26 - - 58.70 30.43 - 7.61 - 
BGA - - - - - 65.71 32.86 - 1.43 - 
SEA - - 2.56 - - 70.51 12.82 - 14.10 - 
GUB - - 15.56 - - 11.11 7.78 - 65.56 - 
TSW - - 17.24 1.72 - 31.03 48.28 - 1.72 - 
ALP - - 3.33 - - 48.33 46.67 - 1.67 - 
ARD - - - - 4.00 60.00 36.00 - - - 
BAL - - 4.55 - - 52.27 43.18 - - - 
HAI - - 12.90 - - 37.10 40.32 - 9.68 - 
MON - - 3.85 - - 65.38 28.21 - 2.56 - 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, 
LES ? Long-Ear Somali, NWH  ? North-West Highland, SES ? Small-Ear Somali, WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? Small 
East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? Baladie, HAI- Hair, MON? Mongolia. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 223
 Appendix 5. Allele frequency for locus ILSTS05 for all the 20 populations. 
Allele 178 180 182 184 186 188 190 
ABG - 1.09 93.48 5.43 - - - 
ABA 1.09 - 78.26 20.65 - - - 
AFA 3.41 1.14 76.14 18.18 1.14 - - 
CEH  - - 95.00 5.00 - - - 
GUM 1.16 - 90.70 8.14 - - - 
HAH - - 87.50 12.50 - - - 
KEF - - 80.26 19.74 - - - 
LES 7.29 - 80.21 12.50 - - - 
NWH - - 92.22 7.78 - - - 
SES - - 90.70 8.14 1.16 - - 
WGJ 2.17 - 84.78 13.04 - - - 
BGA - - 94.29 5.71 - - - 
SEA - - 91.03 8.97 - - - 
GUB 14.13 - 85.87 - - - - 
TSW 34.48 - 37.93 27.59 - - - 
ALP 19.35 - 58.06 22.58 - - - 
ARD 18.00 - 74.00 6.00 - - 2.00 
BAL 6.82 - 75.00 13.64 - - 4.55 
HAI 17.74 1.61 64.52 14.52 - 1.61 - 
MON 16.67 - 61.54 14.10 - 6.41 1.28 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, 
LES ? Long-Ear Somali, NWH  ? North-West Highland, SES ? Small-Ear Somali, WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? Small 
East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? Baladie, HAI- Hair, MON? Mongolia. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 224
 Appendix 6. Allele frequency for locus ILSTS11 for all the 20 populations. 
Allele 261 265 267 269 271 273 275 277 279 281 283 285 
ABG - - - 42.39 13.04 3.26 2.17 3.26 32.61 3.26 - - 
ABA - - - 46.74 4.35 - - 10.87 28.26 7.61 2.17 - 
AFA - - - 15.91 6.82 - 4.55 15.91 56.82 - - - 
CEH  - - - 45.00 12.50 - 1.25 2.50 36.25 2.50 - - 
GUM - - - 37.21 3.49 - 1.16 9.30 48.84 - - - 
HAH - - - 32.50 2.50 - 1.25 17.50 42.50 3.75 - - 
KEF - 1.32 1.32 50.00 15.79 - - 6.58 17.11 7.89 - - 
LES - - 2.08 39.58 - - 1.04 21.88 34.38 1.04 - - 
NWH - - - 53.33 6.67 - 5.56 4.44 28.89 1.11 - - 
SES - - - 29.07 5.81 - 2.33 16.28 43.02 1.16 2.33 - 
WGJ - - - 32.61 3.26 - - 27.17 35.87 1.09 - - 
BGA - - - 21.43 2.86 - - 27.14 48.57 - - - 
SEA - - - 42.11 1.32 - 2.63 14.47 35.53 3.95 - - 
GUB - - 8.70 - - - - 5.43 57.61 28.26 - - 
TSW - - - 17.86 5.36 - 28.57 10.71 26.79 5.36 5.36 - 
ALP - - - 25.81 4.84 - - 19.35 38.71 11.29 - - 
ARD - - 34.00 2.00 4.00 - 6.00 6.00 48.00 - - - 
BAL - - - 27.27 4.55 - 4.55 27.27 31.82 4.55 - - 
HAI - - 4.84 19.35 8.06 3.23 4.84 25.81 32.26 - - 1.61 
MON 2.56 - - 26.92 2.56 7.69 - 34.62 25.64 - - - 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, 
LES ? Long-Ear Somali, NWH  ? North-West Highland, SES ? Small-Ear Somali, WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? Small 
East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? Baladie, HAI- Hair, MON? Mongolia. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 225
 Appendix 7. Allele frequency for locus ILSTS44 for all the 20 populations. 
Allele 151 153 155 157 159 161 169 173 
ABG - - 21.74 31.52 44.57 2.17 - - 
ABA - - 34.52 50.00 11.90 3.57 - - 
AFA - - 39.77 35.23 19.32 5.68 - - 
CEH  - - 33.75 33.75 30.00 2.50 - - 
GUM - 1.16 39.53 44.19 11.63 3.49 - - 
HAH - - 38.75 35.00 18.75 7.50 - - 
KEF - - 36.84 42.11 13.16 7.89 - - 
LES - 2.08 46.88 33.33 14.58 3.13 - - 
NWH - - 37.78 36.67 22.22 3.33 - - 
SES - - 40.70 30.23 20.93 8.14 - - 
WGJ - - 38.04 43.48 15.22 3.26 - - 
BGA - - 41.43 40.00 11.43 7.14 - - 
SEA 1.28 - 29.49 43.59 16.67 8.97 - - 
GUB - - 6.52 85.87 - 7.61 - - 
TSW - 3.45 77.59 10.34 - 1.72 6.90 - 
ALP - 9.68 80.65 4.84 1.61 - 3.23 - 
ARD - - 70.00 28.00 - - 2.00 - 
BAL - - 65.91 29.55 2.27 - 2.27 - 
HAI - - 72.58 11.29 9.68 - 6.45 - 
MON - - 83.33 7.69 1.28 - 1.28 6.41 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, 
LES ? Long-Ear Somali, NWH  ? North-West Highland, SES ? Small-Ear Somali, WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? Small 
East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? Baladie, HAI- Hair, MON? Mongolia. 
 
 226
 Appendix 8. Allele frequency for locus ILSTS87 for all the 20 populations. 
Allele 141 143 145 147 149 151 153 155 157 159 161 163 
ABG - - 9.78 83.70 1.09 1.09 - - - - 3.26 1.09 
ABA - - 9.78 70.65 6.52 - - 1.09 - 9.78 2.17 - 
AFA - - 4.55 82.95 2.27 7.95 - - - - 2.27 - 
CEH  - - 7.50 71.25 7.50 - - - - 8.75 5.00 - 
GUM - - 8.33 80.95 7.14 1.19 - 1.19 - 1.19 - - 
HAH - - 13.75 76.25 2.50 3.75 - 2.50 - 1.25 - - 
KEF - - 3.95 78.95 3.95 - - - - 5.26 7.89 - 
LES - - 22.92 70.83 3.13 - - 1.04 - - 2.08 - 
NWH - - 3.33 76.67 7.78 3.33 - - - 6.67 1.11 1.11 
SES - - 16.67 69.23 6.41 2.56 - 2.56 - 1.28 1.28 - 
WGJ - - 1.09 85.87 3.26 1.09 - - - 3.26 3.26 2.17 
BGA - - 14.29 80.00 1.43 - - 1.43 - 2.86 - - 
SEA - - - 85.90 6.41 1.28 1.28 - - - 5.13 - 
GUB - - 5.43 82.61 4.35 - 7.61 - - - - - 
TSW - - 9.26 70.37 1.85 14.81 1.85 1.85 - - - - 
ALP - - 4.84 8.06 1.61 20.97 27.42 37.10 - - - - 
ARD - 4.00 2.00 46.00 - 8.00 - 38.00 2.00 - - - 
BAL - 13.64 11.36 47.73 11.36 2.27 2.27 11.36 - - - - 
HAI 4.84 11.29 9.68 32.26 12.90 3.23 9.68 6.45 3.23 6.45 - - 
MON 3.95 10.53 18.42 9.21 1.32 7.89 3.95 44.74 - - - - 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, 
LES ? Long-Ear Somali, NWH  ? North-West Highland, SES ? Small-Ear Somali, WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? Small 
East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? Baladie, HAI- Hair, MON? Mongolia. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 227
 Appendix 9. Allele frequency for locus INRA05 for all the 20 populations. 
Allele 131 139 141 143 145 149 
ABG - 60.87 29.35 1.09 8.70 - 
ABA - 41.30 44.57 9.78 4.35 - 
AFA 1.14 42.05 53.41 3.41 - - 
CEH  - 43.75 50.00 6.25 - - 
GUM - 58.14 27.91 12.79 1.16 - 
HAH - 35.00 50.00 10.00 5.00 - 
KEF - 46.05 36.84 7.89 9.21 - 
LES - 46.88 43.75 6.25 3.13 - 
NWH - 51.11 37.78 6.67 4.44 - 
SES - 51.16 37.21 3.49 8.14 - 
WGJ - 60.87 22.83 10.87 5.43 - 
BGA - 61.43 32.86 4.29 1.43 - 
SEA - 75.64 14.10 8.97 1.28 - 
GUB - 50.00 - 44.57 5.43 - 
TSW - 58.93 8.93 14.29 17.86 - 
ALP - 67.74 11.29 17.74 3.23 - 
ARD - 14.00 48.00 36.00 2.00 - 
BAL - 54.55 31.82 6.82 4.55 2.27 
HAI - 46.77 35.48 12.90 4.84 - 
MON - 71.79 26.92 1.28 - - 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, 
LES ? Long-Ear Somali, NWH  ? North-West Highland, SES ? Small-Ear Somali, WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? Small 
East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? Baladie, HAI- Hair, MON? Mongolia. 
  
 
 
 
 
 
 
 
 
 
 
 
 
 228
 Appendix 10. Allele frequency for locus INRA63 for all the 20 populations. 
Allele 157 161 163 165 167 169 171 173 175 
ABG - - 6.52 8.70 54.35 25.00 5.43 - - 
ABA - - 3.26 5.43 57.61 29.35 4.35 - - 
AFA - - 2.27 14.77 47.73 35.23 - - - 
CEH  - - 7.50 7.50 62.50 17.50 5.00 - - 
GUM 9.30 - - 15.12 52.33 20.93 2.33 - - 
HAH - - 3.75 8.75 51.25 33.75 1.25 1.25 - 
KEF 6.58 1.32 3.95 11.84 44.74 28.95 1.32 1.32 - 
LES - - 10.42 8.33 50.00 20.83 9.38 1.04 - 
NWH - - 4.44 11.11 46.67 34.44 3.33 - - 
SES - - 8.14 8.14 43.02 34.88 5.81 - - 
WGJ - - 5.43 8.70 54.35 29.35 2.17 - - 
BGA - - 2.86 4.29 52.86 38.57 1.43 - - 
SEA - - 15.38 11.54 39.74 28.21 5.13 - - 
GUB - - 4.35 17.39 64.13 14.13 - - - 
TSW - - 10.71 28.57 12.50 41.07 7.14 - - 
ALP - 8.06 - 27.42 30.65 14.52 17.74 1.61 - 
ARD - - 2.00 36.00 48.00 12.00 - - 2.00 
BAL - - 11.36 27.27 38.64 20.45 - 2.27 - 
HAI - - 11.29 29.03 25.81 24.19 9.68 - - 
MON - 1.28 6.41 20.51 47.44 16.67 7.69 - - 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, 
LES ? Long-Ear Somali, NWH  ? North-West Highland, SES ? Small-Ear Somali, WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? Small 
East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? Baladie, HAI- Hair, MON? Mongolia. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 229
 Appendix 11. Allele frequency for locus INRA132 for all the 20 populations. 
Allele 143 145 147 151 153 157 
ABG 60.87 36.96 2.17 - - - 
ABA 52.17 44.57 3.26 - - - 
AFA 40.91 56.82 2.27 - - - 
CEH  63.75 31.25 5.00 - - - 
GUM 46.51 38.37 15.12 - - - 
HAH 50.00 47.50 2.50 - - - 
KEF 59.21 36.84 3.95 - - - 
LES 45.83 52.08 2.08 - - - 
NWH 56.67 43.33 - - - - 
SES 50.00 50.00 - - - - 
WGJ 41.30 54.35 4.35 - - - 
BGA 40.00 58.57 1.43 - - - 
SEA 30.77 65.38 3.85 - - - 
GUB 45.65 38.04 16.30 - - - 
TSW 61.11 24.07 9.26 1.85 - 3.70 
ALP 53.23 46.77 - - - - 
ARD 66.00 20.00 10.00 - 4.00 - 
BAL 68.18 22.73 4.55 - - 4.55 
HAI 67.74 32.26 - - - - 
MON 64.71 30.88 2.94 - - 1.47 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, 
LES ? Long-Ear Somali, NWH  ? North-West Highland, SES ? Small-Ear Somali, WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? Small 
East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? Baladie, HAI- Hair, MON? Mongolia. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 230
 Appendix 12. Allele frequency for locus MAF35 for all the 20 populations. 
Allele 101 105 107 109 115 
ABG - 90.22 9.78 - - 
ABA - 86.96 11.96 1.09 - 
AFA - 93.18 6.82 - - 
CEH  - 93.75 6.25 - - 
GUM - 93.02 6.98 - - 
HAH - 93.75 5.00 1.25 - 
KEF - 94.74 5.26 - - 
LES - 85.42 14.58 - - 
NWH - 90.00 10.00 - - 
SES - 91.86 8.14 - - 
WGJ - 90.00 10.00 - - 
BGA - 88.57 11.43 - - 
SEA - 93.59 6.41 - - 
GUB - 98.91 - - 1.09 
TSW - 89.29 8.93 1.79 - 
ALP 1.67 95.00 1.67 1.67 - 
ARD - 88.00 12.00 - - 
BAL - 90.91 9.09 - - 
HAI - 91.94 8.06 - - 
MON - 92.31 7.69 - - 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, 
LES ? Long-Ear Somali, NWH  ? North-West Highland, SES ? Small-Ear Somali, WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? Small 
East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? Baladie, HAI- Hair, MON? Mongolia. 
 
 231
 Appendix 13. Allele frequency for locus MAF65 for all the 20 populations. 
Allele 114 116 118 120 122 124 126 128 130 132 134 136 138 140 144 148 
ABG - 1.09 19.57 3.26 26.09 - 27.17 - - 2.17 17.39 3.26 - - - - 
ABA - 7.78 18.89 5.56 12.22 4.44 21.11 - 14.44 5.56 10.00 - - - - - 
AFA - 18.18 4.55 19.32 17.05 3.41 12.50 - 2.27 15.91 6.82 - - - - - 
CEH  - 13.51 9.46 6.76 21.62 - 18.92 - 4.05 16.22 9.46 - - - - - 
GUM - 1.32 11.84 2.63 27.63 1.32 22.37 - 1.32 17.11 5.26 6.58 2.63 - - - 
HAH - 5.00 13.75 5.00 18.75 7.50 11.25 - 10.00 16.25 10.00 1.25 1.25 - - - 
KEF - 1.35 4.05 18.92 14.86 - 17.57 - 5.41 24.32 9.46 4.05 - - - - 
LES - 5.32 3.19 5.32 26.60 4.26 19.15 - 11.70 11.70 9.57 3.19 - - - - 
NWH - 3.41 7.95 15.91 25.00 - 17.05 1.14 - 18.18 10.23 1.14 - - - - 
SES - 8.14 8.14 5.81 15.12 2.33 12.79 - 10.47 25.58 11.63 - - - - - 
WGJ - 3.26 7.61 4.35 15.22 5.43 23.91 - 7.61 19.57 10.87 - 2.17 - - - 
BGA - 4.41 4.41 1.47 19.12 1.47 14.71 - 10.29 29.41 13.24 1.47 - - - - 
SEA - 9.21 2.63 6.58 26.32 - 11.84 2.63 9.21 19.74 10.53 - 1.32 - - - 
GUB - - - - 27.17 - 3.26 - 1.09 32.61 29.35 4.35 1.09 1.09 - - 
TSW - - 14.81 5.56 3.70 1.85 44.44 - - 14.81 7.41 7.41 - - - - 
ALP - - 8.62 1.72 41.38 - - - - 22.41 24.14 1.72 - - - - 
ARD - 8.33 2.08 - 35.42 12.50 8.33 - 8.33 - 18.75 - - - 4.17 2.08 
BAL - 4.55 13.64 9.09 20.45 4.55 31.82 - - 6.82 9.09 - - - - - 
HAI - - 11.29 4.84 14.52 3.23 25.81 4.84 9.68 3.23 17.74 1.61 - 3.23 - - 
MON 2.86 - 24.29 42.86 10.00 - 7.14 1.43 4.29 7.14 - - - - - - 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, LES ? Long-Ear Somali, NWH  ? North-West Highland, SES 
? Small-Ear Somali, WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? Baladie, HAI- Hair, 
MON? Mongolia. 
 232
 Appendix 14. Allele frequency for locus MAF209 for all the 20 populations. 
Allele 102 104 106 108 
ABG 30.43 57.61 11.96 - 
ABA 22.83 51.09 26.09 - 
AFA 14.77 70.45 14.77 - 
CEH  27.50 57.50 15.00 - 
GUM 30.23 65.12 4.65 - 
HAH 26.25 60.00 12.50 1.25 
KEF 38.16 47.37 10.53 3.95 
LES 16.67 60.42 22.92 - 
NWH 23.33 63.33 13.33 - 
SES 17.44 69.77 12.79 - 
WGJ 20.65 50.00 29.35 - 
BGA 8.57 70.00 21.43 - 
SEA 14.10 37.18 48.72 - 
GUB 2.17 63.04 34.78 - 
TSW 13.79 65.52 20.69 - 
ALP 9.68 88.71 1.61 - 
ARD 20.83 62.50 16.67 - 
BAL 34.09 65.91 - - 
HAI 17.74 77.42 4.84 - 
MON 20.51 79.49 - - 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, 
LES ? Long-Ear Somali, NWH  ? North-West Highland, SES ? Small-Ear Somali, WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? 
Small East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? Baladie, HAI- Hair, MON? 
Mongolia. 
 
 233
 Appendix 15. Allele frequency for locus OarAE129 for all the 20 populations. 
Allele 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 173 175 177 
ABG 27.17 14.13 - - - - - - 14.13 29.35 8.70 - - - - 3.26 1.09 - 2.17 - 
ABA 35.23 20.45 - - - - - - 21.59 12.50 5.68 - - 1.14 - 3.41 - - - - 
AFA 12.50 9.09 - - - 1.14 - 3.41 18.18 25.00 10.23 - - 3.41 1.14 13.64 2.27 - - - 
CEH  35.94 15.63 - - - - - 1.56 7.81 20.31 14.06 - - - - 1.56 - 1.56 1.56 - 
GUM 30.23 13.95 - - - 2.33 - - 9.30 13.95 5.81 - - 15.12 2.33 4.65 - 1.16 1.16 - 
HAH 34.21 13.16 - - - - - - 13.16 22.37 7.89 - - 2.63 - 2.63 2.63 1.32 - - 
KEF 28.95 22.37 - - 2.63 - - 1.32 3.95 22.37 5.26 - 2.63 1.32 2.63 3.95 - - 2.63 - 
LES 22.34 13.83 - 1.06 - - 2.13 2.13 20.21 22.34 3.19 - 1.06 2.13 2.13 5.32 - - 1.06 1.06 
NWH 30.00 21.11 1.11 5.56 - - - 1.11 10.00 16.67 8.89 1.11 - - 1.11 1.11 - - 2.22 - 
SES 25.00 17.86 - - - 1.19 - 1.19 16.67 15.48 10.71 - - - 1.19 8.33 - - 2.38 - 
WGJ 27.17 17.39 - - - - - 3.26 17.39 18.48 2.17 - - 2.17 4.35 3.26 1.09 - 3.26 - 
BGA 17.24 22.41 - - - - - 1.72 13.79 25.86 5.17 - - 3.45 3.45 5.17 - 1.72 - - 
SEA 22.97 5.41 - - - - - 2.70 31.08 27.03 5.41 - - 5.41 - - - - - - 
GUB - - - - - - - - 40.91 26.14 5.68 - - 1.14 18.18 6.82 - 1.14 - - 
TSW 62.07 3.45 - - - - - 1.72 12.07 18.97 - - - - 1.72 - - - - - 
ALP 46.77 - - - - - - 11.29 17.74 6.45 - - 1.61 1.61 14.52 - - - - - 
ARD 4.00 2.00 - - - - - - 48.00 - 6.00 - - 4.00 14.00 14.00 8.00 - - - 
BAL 2.27 2.27 - - - 9.09 - - 34.09 27.27 - - - 4.55 18.18 2.27 - - - - 
HAI 12.90 1.61 - - - 3.23 - 6.45 30.65 17.74 1.61 - - - 24.19 1.61 - - - - 
MON 19.74 7.89 - - - 14.47 - - 22.37 9.21 3.95 15.79 - - 5.26 1.32 - - - - 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, LES ? Long-Ear Somali, NWH  ? North-West Highland, SES ? 
Small-Ear Somali, WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? Baladie, HAI- Hair, MON? 
Mongolia. 
 
 234
 Appendix 16. Allele frequency for locus OarFCB304 for all the 20 populations. 
Allele 120 124 128 132 134 136 138 140 142 144 146 148 150 154 156 158 160 162 164 166 168 170 174 178 180
 ABG - 2.17 - - - 2.17 - 45.65 9.78 5.43 14.13 - - 13.04 - 2.17 - 2.17 2.17 - 1.09 - - - - 
ABA 1.09 - - - - 3.26 - 50.00 14.13 3.26 7.61 - 3.26 6.52 - - 2.17 8.70 - - - - - - - 
AFA - 1.14 - - - 3.41 - 50.00 14.77 3.41 10.23 - - 12.50 - 3.41 - - - 1.14 - - - - - 
CEH  - 3.75 - - - 5.00 - 47.50 8.75 1.25 12.50 - 2.50 16.25 - 2.50 - - - - - - - - - 
GUM - - - - 1.16 3.49 1.16 59.30 13.95 4.65 1.16 - - 11.63 - - - 3.49 - - - - - - - 
HAH - 1.25 1.25 - - 3.75 - 48.75 5.00 1.25 13.75 - - 6.25 1.25 12.50 1.25 - - - - 2.50 - - 1.25
 KEF - - - - - - - 50.00 22.37 3.95 2.63 - 2.63 9.21 - 7.89 1.32 - - - - - - - - 
LES - - - - - 4.17 - 62.50 9.38 1.04 - - - 6.25 1.04 3.13 2.08 2.08 1.04 1.04 4.17 1.04 1.04 - - 
NWH - 1.11 - - - 2.22 - 50.00 11.11 8.89 3.33 - - 21.11 - 2.22 - - - - - - - - - 
SES - - - 1.16 - 4.65 - 60.47 9.30 1.16 5.81 - - 8.14 - 5.81 2.33 - - 1.16 - - - - - 
WGJ - - - - - 4.35 - 47.83 16.30 5.43 8.70 - - 8.70 - - 2.17 - 3.26 1.09 - - 2.17 - - 
BGA - - - - - - - 62.86 4.29 4.29 5.71 - - 14.29 - - 2.86 1.43 - 4.29 - - - - - 
SEA - - - - - 2.56  51.28 15.38 5.13 - - - 8.97 - 5.13 7.69 - 3.85 - - - - - - 
GUB - - - 6.52 - - 8.70 26.09 3.26 3.26 - - - - 3.26 10.87 30.43 4.35 1.09 2.17 - - - - - 
TSW - - - 3.45 - 32.76 1.72 36.21 18.97 - - - - - - - 3.45 - - - - 3.45 - - - 
ALP - - - - - 12.90 11.29 50.00 12.90 - - - - 9.68 3.23 - - - - - - - - - - 
ARD - - 2.00 4.00 4.00 2.00 2.00 66.00 8.00 2.00 2.00 - - 4.00 2.00 - - - - - 2.00 - - - - 
BAL - - - 4.55 - 22.73 6.82 27.27 6.82 4.55 - - - 11.36 - - - 2.27 4.55 2.27 2.27 - - 4.55 - 
HAI - - 3.23 8.06 - 20.97 1.61 35.48 8.06 - - 1.61 - 4.84 - - - - 4.84 8.06 3.23 - - - - 
MON - - 6.41 1.28 - 7.69 14.10 37.18 1.28 - - - 1.28 1.28 - - - 10.26 12.82 1.28 1.28 - - - 3.85
 ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, LES ? Long-Ear Somali, NWH  ? North-West Highland, SES ? 
Small-Ear Somali, WGJ ? Woyto-Guji, BGA? Boran Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, BAL ? Baladie, HAI- Hair, MON? 
Mongolia. 
 
 235
 Appendix 17. Allele frequency for locus SRCRSP03 for all the 20 populations. 
Allele 111 115 117 119 121 123 125 
ABG 27.78 - - 13.33 30.00 28.89 - 
ABA 38.04 - - 4.35 41.30 16.30 - 
AFA 19.32 - - 17.05 53.41 10.23 - 
CEH  18.75 - - 18.75 46.25 16.25 - 
GUM 52.33 - - 10.47 30.23 6.98 - 
HAH 19.23 - 2.56 15.38 53.85 7.69 1.28 
KEF 36.84 - - 1.32 52.63 9.21 - 
LES 40.63 - - 5.21 44.79 8.33 1.04 
NWH 27.78 - - 8.89 46.67 16.67 - 
SES 26.74 - 1.16 17.44 47.67 6.98 - 
WGJ 45.65 - - 3.26 44.57 6.52 - 
BGA 24.29 2.86 - 8.57 51.43 12.86 - 
SEA 20.51 - - 8.97 55.13 15.38 - 
GUB 31.52 - - 15.22 53.26 - - 
TSW 5.17 - 3.45 10.34 81.03 - - 
ALP 19.35 - - 6.45 70.97 3.23 - 
ARD 14.00 - - 16.00 60.00 10.00 - 
BAL 22.73 - - 13.64 59.09 4.55 - 
HAI 9.68 - - 16.13 62.90 6.45 4.84 
MON 1.28 - - 24.36 64.10 1.28 8.97 
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, 
KEF ? Keffa, LES ? Long-Ear Somali, NWH ? North-West Highland, SES ? Small-Ear Somali, WGJ ? Woyto-Guji, 
BGA? Boran Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW ? Tswana, ALP ? Italian Alpine, ARD- Ardi, 
BAL ? Baladie, HAI- Hair, MON? Mongolia. 
 236
 Appendix 18.: Test for genetic differentiation between pairs of 20 populations for each locus. 
Population pairs BM 1818 
BMS 
1494 
ILSTS 
05 
ILSTS 
11 
ILSTS 
44 
ILSTS 
87 
INRA 
05 
INRA 
63 
INRA 
132 
MAF 
35 
MAF 
65 
MAF 
209 
OarAE 
129 
OarFCB 
304 
SRCRSP 
03 
ABA & ABG ns * ** ** *** ** ** ns ns ns *** ns ns * * 
AFA&ABG ns ** ** *** ** ns *** ns * ns *** ns ** ns ** 
AFA &ABA ** * ns *** ns *** ns * ns ns *** * *** * ** 
CEH &ABG ns ns ns ns ns ** *** ns ns ns *** ns ns ns * 
CEH &ABA ns  ** * * ns ns ns ns ns * ns ns ** ** 
CEH &AFA ns *** ** *** ns *** ns * * ns ns ns ** ns ns 
GUM &ABG ns * ns ** *** ns ** ** ** ns * ns *** * *** 
GUM &ABA ns ** ns ** ns ns ns *** * ns *** *** ** ns * 
GUM & AFA * ns ns * ns ns ** ** ** ns *** ** *** * *** 
GUM & CEH ns * ns * ns * * ** * ns * ns * ** *** 
HAH & ABG ns ns ns ** ** ns ** ns ns ns *** ns ns * *** 
HAH & ABA * ns ns ns ns * ns ns ns ns ns ns ns *** ** 
HAH & AFA ns ns ns * ns ns ns ns ns ns ** ns * ns ns 
HAH & CEH ns * ns ** ns ** ns ns ns ns ns ns ns ns ns 
HAH & GUM ns * ns ns ns ns ** ** ** ns ns ns ns *** *** 
KEF & ABG ns ns ** ns *** * ns ns ns ns *** ns ns ** *** 
KEF & ABA ** ** ns * ns ns  * ns ns *** ** * ** ns 
KEF & AFA ns * ns *** ns ** ** ns ns ns * *** *** ns *** 
KEF & CEH ns ** * ns ns ns * * ns ns * ns ns ** *** 
KEF & GUM ns ** * *** ns * * ns * ns * * * ** ** 
KEF & HAH ns *  *** ns ** ns ns ns ns ns ns ns ** ** 
LES & ABG ns ns ** *** *** ns * ns ns ns *** * ns ** *** 
LES & ABA ns ns * ** ns ** ns ns ns ns * ns ns ** ns 
LES & AFA * ns ns *** ns *** ns *** ns ns ** ns ns * ** 
LES & CEH ns * ** *** ns ns ns ns ns ns ns ns ns *** ** 
LES & GUM ns ns ns * ns * ns *** ** ns * ** * ns ns 
LES & HAH ns ns * ns ns *** ns ns ns * ns ns ns ** * 
 237
 Population pairs BM 1818 
BMS 
1494 
ILSTS 
05 
ILSTS 
11 
ILSTS 
44 
ILSTS 
87 
INRA 
05 
INRA 
63 
INRA 
132 
MAF 
35 
MAF 
65 
MAF 
209 
OarAE 
129 
OarFCB 
304 
SRCRSP 
03 
LES & KEF ns ns * *** ns *** ns * ns ns * ** ns * ns 
NWH & ABG ns * ns ns * ** ns ns ns ns *** ns ns ns ns 
NWH & ABA ns  * ** ns ns ns ns ns ns *** ns ns *** ns 
NWH & AFA *** ** * *** ns ns * ns * ns * ns *** ns ns 
NWH & CEH ns ns ns ns ns ns ns ns * ns ns ns ns ns ns 
NWH & GUM * ** ns * ns ns ns ** *** ns ns ns ** ns ** 
NWH & HAH * ns ns ** ns * ns ns ns ns ** ns ns *** ns 
NWH & KEF * *** * * ns ns ns ns ns ns ns * ns * ns 
NWH & LES ns * ** *** ns *** ns ns ns ns *** ns ns ** ns 
SES & ABG ns *** ns ** ** ns ns ns ns ns *** ns ns * *** 
SES & ABA ns ns * * * ns ns ns ns ns * * ns ** ** 
SES & AFA ns ns ns ns ns * * * ns ns * ns ns ns ns 
SES & CEH ns *** ns ** ns * * ns * ns ns ns ns ns ns 
SES & GUM ns ** ns ns ns ns * *** *** ns ** ns ** * ** 
SES & HAH ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns 
SES & KEF ns ** * *** ns ** ns ns * ns * ** ns ns ** 
SES & LES ns ns * * ns ns ns ns ns ns ns ns ns ns * 
SES & NWH ns ** ns ** ns * ns ns ns ns ** ns ns * ns 
WGJ & ABG ns ns ns *** *** ns * ns ns ns *** * ns ns *** 
WGJ & ABA ns ns ns ** ns * * ns ns ns * ns ns * ns 
WGJ & AFA ns ns ns ** ns ns *** ns ns ns ** * * ns *** 
WGJ & CEH ns ** ns *** ns * ** ns * ns ns ns ns * *** 
WGJ & GUM * ** ns * ns ns ns ** * ns ns *** * * ns 
WGJ & HAH ns ns ns ns ns ** ** ns ns ns ns * ns *** ** 
WGJ & KEF * ns ns *** ns ns ns ns ns ns * ** ns ** ns 
WGJ & LES ns ns ns ns ns *** * ns ns ns ns ns ns ** ns 
 238
 Population pairs BM 1818 
BMS 
1494 
ILSTS 
05 
ILSTS 
11 
ILSTS 
44 
ILSTS 
87 
INRA 
05 
INRA 
63 
INRA 
132 
MAF 
35 
MAF 
65 
MAF 
209 
OarAE 
129 
OarFCB 
304 
SRCRSP 
03 
WGJ & NWH ns * ns *** ns ns ns ns * ns ** * ns * * 
WGJ & SES ns ns ns ns ns ** ns ns ns ns ns * ns ns ** 
BGA & ABG ** ** ns *** *** ns ns ns * ns *** ** ns * * 
BGA & ABA *** ns * *** ns ns ns ns ns ns ** * ns ** ns 
BGA & AFA ** ns * ns ns ** ns ns ns ns *** ns ns * ns 
BGA & CEH *** ** ns *** ns * ns ns ** ns ns ** ns ** ns 
BGA & GUM * * ns * ns ns ns *** *** ns ns *** ns * ** 
BGA & HAH ns ns ns ns ns ns * ns ns ns ns * ns *** ns 
BGA & KEF ** ** * *** ns * ns ns * ns ns *** ns *** ns 
BGA & LES ns ns * * ns ns ns * ns ns ns ns ns * ns 
BGA & NWH *** * ns *** ns * ns ns * ns ** * ns * ns 
BGA & SES ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns 
BGA & WGJ ** ns ns ns ns ** ns ns ns ns ns * ns * * 
SEA &ABG ** ns ns ** ** ** *** ns *** ns *** *** ** ** * 
SEA & ABA *** *** ns ns ns *** *** * * ns *** * ** *** ns 
SEA & AFA ** * * *** ns * *** ** ns ns * *** ** ** ns 
SEA & CEH *** *** ns ** ns ** *** ns *** ns ns *** *** *** ns 
SEA & GUM *** * ns ns ns **  *** ** ns ** *** *** * *** 
SEA & HAH ns ** ns ns ns *** *** ns ns ns ns *** * *** ns 
SEA & KEF ** ns ns ** ns * *** * ** ns * *** *** ns * 
SEA & LES ns ns * ns ns *** *** ns ns ns ns ** ns ns * 
SEA & NWH *** *** ns ns ns * ** ns ** ns * *** *** ** ns 
SEA & SES ns *** ns ns ns *** *** ns * ns ns *** *** ns ns 
SEA & WGJ ** ns ns ns ns ns ns ns ns ns ns * * * ** 
SEA & BGA * ** ns ** ns *** * * ns ns ns *** * ** ns 
 239
 Population pairs BM 1818 
BMS 
1494 
ILSTS 
05 
ILSTS 
11 
ILSTS 
44 
ILSTS 
87 
INRA 
05 
INRA 
63 
INRA 
132 
MAF 
35 
MAF 
65 
MAF 
209 
OarAE 
129 
OarFCB 
304 
SRCRSP 
03 
GUB & ABG *** *** *** *** *** ** *** * ** ** *** *** *** *** *** 
GUB & ABA *** *** *** *** *** *** *** ** ** *** *** *** *** *** *** 
GUB & AFA *** *** *** *** *** *** *** * ** * *** *** *** *** ** 
GUB & CEH ** *** *** *** *** *** *** ns * * *** *** *** *** *** 
GUB & GUM *** *** *** *** *** * *** *** ns * *** *** *** *** *** 
GUB & HAH *** *** *** *** *** ** *** * ** * *** *** *** *** * 
GUB & KEF *** *** *** *** *** *** *** ** * * *** *** *** *** *** 
GUB & LES *** *** *** *** *** *** *** *** *** *** *** ** *** *** ** 
GUB & NWH *** *** *** *** *** *** *** ** *** ** *** *** *** *** *** 
GUB & SES *** *** *** *** *** *** *** *** *** ** *** *** *** *** * 
GUB & WGJ *** *** *** *** *** *** *** * ** *** *** *** *** *** *** 
GUB & BGA *** *** *** *** *** ** *** *** *** *** *** ns *** *** *** 
GUB & SEA *** *** *** *** *** ** *** *** *** * *** *** *** *** *** 
TSW & ABG * *** *** *** *** ** *** *** * ns *** ns *** *** *** 
TSW & ABA *** ** *** *** *** *** *** *** * ns *** ns *** *** *** 
TSW & AFA ** *** *** *** *** ns *** *** ** ns *** ns *** *** *** 
TSW & CEH ** *** *** *** *** *** *** *** ns ns *** ns ** *** *** 
TSW & GUM ** *** *** *** *** * *** *** * ns ns ** *** *** *** 
TSW & HAH ns ** *** *** *** ns *** *** ** ns *** ns ** *** * 
TSW & KEF ** *** *** *** *** *** ** *** ns ns ** ** *** *** *** 
TSW & LES ns *** *** *** *** *** *** *** ** ns *** ns ** *** *** 
TSW & NWH *** ** *** *** *** ** *** ** *** ns *** ns *** *** *** 
TSW & SES ns *** *** *** *** ns *** ** *** ns *** ns *** *** *** 
TSW & WGJ ** *** *** *** *** *** * *** ** ns ** ns ** *** *** 
TSW & BGA ns *** *** *** *** ** *** *** *** ns *** ns *** *** *** 
 240
 Population pairs BM 1818 
BMS 
1494 
ILSTS 
05 
ILSTS 
11 
ILSTS 
44 
ILSTS 
87 
INRA 
05 
INRA 
63 
INRA 
132 
MAF 
35 
MAF 
65 
MAF 
209 
OarAE 
129 
OarFCB 
304 
SRCRSP 
03 
TSW & SEA * *** *** *** *** *** ** ** ***  *** ** *** *** *** 
TSW & GUB *** *** *** *** *** *** *** *** * ** *** * *** *** *** 
ALP & ABG *** ** *** *** *** *** *** *** ns * *** *** *** *** *** 
ALP & ABA *** ns *** ns *** *** *** *** ns * *** *** *** *** ** 
ALP & AFA *** ** ** ** *** *** *** *** ns ns *** ** *** ***  
ALP & CEH *** ** *** *** *** *** *** *** ns ns *** *** *** *** ** 
ALP & GUM *** *** *** ** *** *** ns *** *** ns *** ** *** ** *** 
ALP & HAH ***  *** ns *** *** *** *** ns ns *** ** *** *** ns 
ALP & KEF *** ** *** ** *** *** ** *** ns ns *** *** *** *** * 
ALP & LES *** * * ** *** *** *** *** ns ** *** *** *** ** * 
ALP & NWH *** ns *** *** *** *** ** *** ns * *** ** *** *** * 
ALP & SES *** ns *** Ns *** *** *** *** ns ns *** * *** *** * 
ALP & WGJ *** ns *** ns *** *** ns *** ns * *** *** *** *** * 
ALP & BGA ** ns *** * *** *** ** *** ns * *** ** *** *** ns 
ALP & SEA *** *** *** ns *** *** ns *** *  *** *** *** *** ns 
ALP & GUB *** *** *** *** *** *** *** *** *** ns * *** *** *** * 
ALP & TSW ns ns ns *** ns *** ns *** ** ns *** ** ** *** * 
ARD & ABG ns *** *** *** *** *** *** *** * ns *** ns *** ** ** 
ARD & ABA * ns *** *** *** *** *** *** ** ns *** ns *** ** ** 
ARD & AFA ns * ** *** *** *** *** ** *** ns *** ns *** * ns 
ARD & CEH ** ** *** *** *** *** *** ** ns ns *** ns *** ** ns 
ARD & GUM ns *** *** *** *** *** *** *** * ns *** ns *** ns *** 
ARD & HAH ns ns *** *** *** *** ** *** ** ns *** ns *** ** ns 
ARD & KEF ns *** *** *** *** *** *** ** ns ns *** ns *** ** ** 
ARD & LES ns * ns *** *** *** *** *** *** ns *** ns *** ns ** 
 241
 Population pairs BM 1818 
BMS 
1494 
ILSTS 
05 
ILSTS 
11 
ILSTS 
44 
ILSTS 
87 
INRA 
05 
INRA 
63 
INRA 
132 
MAF 
35 
MAF 
65 
MAF 
209 
OarAE 
129 
OarFCB 
304 
SRCRSP 
03 
ARD & NWH *** * *** *** *** *** *** *** *** ns *** ns *** ** ns 
ARD & SES ns ns *** *** *** *** *** *** *** ns *** ns *** ns ns 
ARD & WGJ * ** ** *** *** *** *** *** *** ns *** ns *** ** *** 
ARD & BGA ns ns *** *** *** *** *** *** *** ns *** ns *** * ns 
ARD & SEA ** *** *** *** *** *** *** *** *** ns *** *** *** ** ns 
ARD & GUB *** *** * *** *** *** *** ns * * *** *** *** *** ** 
ARD & TSW ns *** *** *** ns *** *** *** ns ns *** ns *** *** ** 
ARD & ALP ** ns * *** *** *** *** *** *** ns *** ** *** * ns 
BAL & ABG *** ** ** ** *** *** ns * ns ns ns * *** *** ** 
BAL & ABA *** ns ns * ** *** ns ** * ns ns *** *** *** * 
BAL & AFA *** ** ns * ** *** * * ** ns * *** *** *** ns 
BAL & CEH *** ns ** ** *** *** ns * ns ns ns ** *** *** ns 
BAL & GUM *** *** * * * *** ns *** ** ns ns ns *** *** ** 
BAL & HAH ** ns * ns ** *** ns * ** ns * * *** *** ns 
BAL & KEF *** ** * ** ** *** ns ns ns ns * * *** *** * 
BAL & LES ns ns ns ns * *** ns * ** ns * *** *** *** ns 
BAL & NWH *** ns ** ** *** *** ns * ** ns ns ** *** *** ns 
BAL & SES ns ns ** ns *** ** ns * *** ns * ** *** *** ns 
BAL & WGJ *** ns ns ns ** *** ns * ** ns ns *** *** *** * 
BAL & BGA ns ns ** ns * *** ns ** *** ns ** *** *** *** ns 
BAL & SEA * *** ** ns *** *** ns ns *** ns ** *** *** *** ns 
BAL & GUB *** *** *** *** *** *** *** ns ** ** *** *** * *** ns 
BAL  & TSW ** ns *** * ns ** ** * ns ns ns *** *** ** ** 
BAL & ALP *** ns * ns ** *** * *** ** ns *** ** *** ** ns 
BAL & ARD ** ns ns *** ns ** *** ns ns ns *** ** *** ** ns 
 242
 Population pairs BM 1818 
BMS 
1494 
ILSTS 
05 
ILSTS 
11 
ILSTS 
44 
ILSTS 
87 
INRA 
05 
INRA 
63 
INRA 
132 
MAF 
35 
MAF 
65 
MAF 
209 
OarAE 
129 
OarFCB 
304 
SRCRSP 
03 
HAI & ABG *** ** *** *** *** *** * ** ns ns ** * *** *** *** 
HAI & ABA *** ** *** *** *** *** ns *** ns ns * *** *** *** *** 
HAI & AFA *** *** * * *** *** * *** ** ns *** ns *** *** ns 
HAI & CEH *** *** *** *** *** *** ns *** ns ns *** * *** *** * 
HAI & GUM *** *** *** *** *** *** ns *** *** ns ** ns *** *** *** 
HAI & HAH *** * *** * *** *** ns *** * ns * ns *** *** ns 
HAI & KEF *** * *** *** *** *** ns ** ns ns *** ** *** *** *** 
HAI & LES ** ***  ** *** *** ns ** * ns ** ** *** *** *** 
HAI & NWH *** *** *** *** *** *** ns ** ns ns *** ns *** *** ** 
HAI & SES * *** *** * *** *** ns * ns ns *** ns *** *** * 
HAI & WGJ *** * ** ** *** *** ns *** ** ns * *** *** *** *** 
HAI & BGA ns *** *** * *** *** ns *** ** ns ** * *** *** * 
HAI & SEA ** *** *** ** *** *** ** ns *** ns *** *** *** *** * 
HAI & GUB *** *** *** *** *** *** *** *** *** ** *** *** *** *** *** 
HAI & TSW ** ns ** *** ns *** ** ns * ns ** * *** ** * 
HAI & ALP *** * ns ** * *** * ** ns ns *** ns ** *** ns 
HAI & ARD *** *** ns *** * *** *** ** * ns *** ns *** ** ns 
HAI & BAL ns * ns ns ns ns ns ns ns ns ns ns ns ns ns 
MON & ABG *** ns *** *** *** *** * ns ns ns *** *** *** *** *** 
MON & ABA *** ns *** *** *** *** *** * ns ns *** *** *** *** *** 
MON & AFA *** ns ** *** *** *** ** ** * ns *** *** *** *** *** 
MON & CEH *** ns *** *** *** *** ** ns ns ns *** *** *** *** *** 
MON & GUM *** ns *** *** *** *** ** ** * ns *** * *** *** *** 
MON & HAH *** ns *** ** *** *** *** * ns ns *** ** *** *** *** 
MON & KEF *** ns *** *** *** *** *** * ns ns *** *** *** *** *** 
MON & LES *** ns ** *** *** *** ** ns ns ns *** *** *** *** *** 
 243
 Population pairs BM 1818 
BMS 
1494 
ILSTS 
05 
ILSTS 
11 
ILSTS 
44 
ILSTS 
87 
INRA 
05 
INRA 
63 
INRA 
132 
MAF 
35 
MAF 
65 
MAF 
209 
OarAE 
129 
OarFCB 
304 
SRCRSP 
03 
MON & NWH *** ns *** *** *** *** ** ns ns ns *** *** *** *** *** 
MON & SES *** ns *** *** *** *** ** ns * ns *** ** *** *** *** 
MON & WGJ *** ns *** * *** *** ** * * ns *** *** *** *** *** 
MON & BGA *** ns *** ** *** ***  ** * ns *** *** *** *** *** 
MON & SEA *** * *** *** *** *** * ns ** ns *** *** *** *** *** 
MON & GUB *** *** *** *** *** *** *** * ** ** *** *** *** *** *** 
MON & TSW *** *** ** *** ns *** *** ** ns ns *** *** *** *** ** 
MON & ALP *** ns ns *** * *** *** * ns ns *** ns *** *** *** 
MON & ARD *** ns ns *** * *** *** * ns ns *** *** *** *** *** 
MON & BAL *** ns ns * * *** * ns ns ns *** ns *** ** *** 
MON & HAI *** ** ns * * *** *** ns ns ns *** ns *** *** ns 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, KEF- Keffa, LES ? Long-eared Somali, NWH ? North-
 West Highland, SES ? Small-eared Somali, WGJ? Woyto-Guji, BGA- Boran Galla, SEA ? Small East African, GUB ? Guinea Bissau, TSW- Tswana, ALP- Italian Alpine, ARD- 
Ardi, BAL- Baladie, HAI- Hair, MON- Mongolia. 
*=P<0.05; **=P<0.01; ***=P<0.001
 244
 Appendix 19. Summary of linkage disequilibrium between all markers across all 
Ethiopian populations. 
  Linkage between markers ABG ABA AFA CEH GUM HAH KEF LES NWH SES WGJ 
1 BM1818 & BMS1494                   *   
2 BM1818 & ILSTS05                       
3 BM1818 & ILSTS11     *   ** 
*         * 
4 BM1818 & ILSTS44         **             
5 BM1818 & ILSTS87                       
6 BM1818 & INRA05     *                 
7 BM1818 & INRA132             **       ** 
8 BM1818 & INRA63                       
9 BM1818 & MAF209     **         *   *   
10 BM1818 & MAF35                       
11 BM1818 & MAF65                       
12 BM1818 & OarAE129   **                 ** 
13 BM1818 & OarFCB304                     * 
14 BM1818 & SRCRSP03                       
15 BMS1494 & ILSTS05                       
16 BMS1494 & ILSTS11       **           **   
17 BMS1494 & ILSTS44                       
18 BMS1494 & ILSTS87                 **     
19 BMS1494 & INRA05   *           *       
20 BMS1494 & INRA132                       
21 BMS1494 & INRA63                       
22 BMS1494 & MAF209   *   *               
23 BMS1494 & MAF35               *       
24 BMS1494 & MAF65         *     *       
25 BMS1494 & OarAE129         *     *       
26 BMS1494 & OarFCB304   *                   
27 BMS1494 & SRCRSP03                       
28 ILSTS05 & ILSTS11                       
29 ILSTS05 & ILSTS44                       
30 ILSTS05 & ILSTS87                       
31 ILSTS05 & INRA05                       
32 ILSTS05 & INRA132                       
33 ILSTS05 & INRA63   *                   
34 ILSTS05 & MAF209                   *   
35 ILSTS05 & MAF35       *               
36 ILSTS05 & MAF65                       
37 ILSTS05 & OarAE129   *                   
38 ILSTS05 & OarFCB304                       
39 ILSTS05 & SRCRSP03                       
 245
  Linkage between markers ABG ABA AFA CEH GUM HAH KEF LES NWH SES WGJ 
40 ILSTS11 & ILSTS44                       
41 ILSTS11 & ILSTS87                       
42 ILSTS11 & INRA05                       
43 ILSTS11 & INRA132                       
44 ILSTS11 & INRA63 *     *               
45 ILSTS11 & MAF209                   *   
46 ILSTS11 & MAF35                       
47 ILSTS11 & MAF65                       
48 ILSTS11 & OarAE129                   *   
49 ILSTS11 & OarFCB304   *         **         
50 ILSTS11 & SRCRSP03                   *   
51 ILSTS44 & ILSTS87   *                   
52 ILSTS44 & INRA05                       
53 ILSTS44 & INRA132               **       
54 ILSTS44 & INRA63                       
55 ILSTS44 & MAF209                       
56 ILSTS44 & MAF35                     * 
57 ILSTS44 & MAF65                     * 
58 ILSTS44 & OarAE129             *   *     
59 ILSTS44 & OarFCB304       *   *           
60 ILSTS44 & SRCRSP03                       
61 ILSTS87 & INRA05           *           
62 ILSTS87 & INRA132       *               
63 ILSTS87 & INRA63         *         *   
64 ILSTS87 & MAF209                       
65 ILSTS87 & MAF35                       
66 ILSTS87 & MAF65                       
67 ILSTS87 & OarAE129               *       
68 ILSTS87 & OarFCB304                       
69 ILSTS87 & SRCRSP03                       
70 INRA05 & INRA132                       
71 INRA05 & INRA63                       
72 INRA05 & MAF209                       
73 INRA05 & MAF35   * *                 
74 INRA05 & MAF65                       
75 INRA05 & OarAE129     *                 
76 INRA05 & OarFCB304   *                   
77 INRA05 & SRCRSP03                       
78 INRA132 & MAF209                       
79 INRA132 & MAF35                       
80 INRA132 & MAF65     *                 
81 INRA132 & OarAE129                       
82 INRA132 & OarFCB304                 *     
 246
  Linkage between markers ABG ABA AFA CEH GUM HAH KEF LES NWH SES WGJ 
83 INRA132 & SRCRSP03                       
84 INRA63 & INRA132 *                     
85 INRA63 & MAF209                       
86 INRA63 & MAF35                       
87 INRA63 & MAF65     *       *         
88 INRA63 & OarAE129   *                   
89 INRA63 & OarFCB304                       
90 INRA63 & SRCRSP03                       
91 MAF209 & OarAE129                       
92 MAF209 & OarFCB304 *             **       
93 MAF209 & SRCRSP03                       
94 MAF35 & MAF209             *         
95 MAF35 & MAF65           *           
96 MAF35 & OarAE129                       
97 MAF35 & OarFCB304 **               *     
98 MAF35 & SRCRSP03                     * 
99 MAF65 & MAF209                       
100 MAF65 & OarAE129                       
101 MAF65 & OarFCB304                       
102 MAF65 & SRCRSP03                       
103 OarAE12 & OarFCB304               *       
104 OarAE12 & SRCRSP03                       
105 OarFCB3 & SRCRSP03                       
ABG ? Abergalle, ABA ? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM ? Gumez, HAH ? Hararghe Highland, KEF ? Keffa, 
LES ? Long-eared Somali, NWH ? North-West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
*=P<0.05; **=P<0.01 
 247
 Appendix 20. Matrix of all pair-wise combinations of loci showing the linkage disequilibrium (for 11 Ethiopian populations).  P 
values condensed across all populations.  For simplicity, the matrix shows only levels of significance. 
 
BM 
1818 
BM 
1494 
ILST 
S05 
ILST 
S11 
ILST 
S44 
ILST 
S87 
INRA 
05 
INRA 
63 
INRA 
132 
MAF 
35 
MAF 
65 
MAF 
209 
OarAE 
129 
OarFCB 
304 
SRCPSP 
03 
BM1818                
BMS1494                
ILSTS05                
ILSTS11 **               
ILSTS44 *               
ILSTS87                
INRA05                
INRA63    *  *          
INRA132                
MAF35                
MAF65                
MAF209 **               
OarAE129 ** *              
OFCB304  *          *    
SRCRS03                
? = P < 0.05; ** = P < 0.01. 
 
 
 248
 Appendix 21. Allele frequency for locus BM1818 for all the 11 Ethiopian populations. 
Allele 244 246 250 252 254 256 258 260 262 264 266 268 
ABG 1.09 - - 7.61 - 7.61 51.09 7.61 13.04 6.52 5.43 - 
ABA 3.26 - - 5.43 - 5.43 40.22 7.61 20.65 6.52 4.35 6.52 
AFA - 1.14 - 1.14 1.14 3.41 65.91 9.09 3.41 6.82 5.68 2.27 
CEH 5.00 - - 5.00 1.25 2.50 51.25 6.25 10.00 11.25 7.50 - 
GUM - - - 8.14 - 8.14 41.86 13.95 10.47 8.14 8.14 1.16 
HAH - - - 2.50 2.50 7.50 58.75 11.25 10.00 5.00 2.50 - 
KEF - - - 6.58 - 5.26 68.42 7.89 5.26 3.95 2.63 - 
LES 3.13 1.04 1.04 6.25 5.21 9.38 47.92 6.25 11.46 3.13 4.17 1.04 
NWH 1.11 - - 13.33 2.22 1.11 45.56 5.56 20.00 6.67 4.44 - 
SES 1.16 - - 6.98 4.65 4.65 46.51 10.47 15.12 6.98 2.33 1.16 
WGJ 1.09 - - 3.26 2.17 2.17 54.35 5.43 13.04 7.61 4.35 6.52 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar,  CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa,  LES ? Long-eared Somali,  NWH ? North-West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
 
 
 
Appendix 22. Allele frequency for locus BMS1494 for all the 11 Ethiopian populations. 
Allele 244 246 256 264 276 280 282 284 288 
ABG - - 9.78 - 65.22 18.48 - 6.52 - 
ABA - - 4.55 - 57.95 36.36 - 1.14 - 
AFA - - - 1.14 71.59 22.73 - 4.55 - 
CEH 1.25 1.25 11.25 - 65.00 21.25 - - - 
GUM - - 2.33 - 82.56 13.95 - 1.16 - 
HAH - - 3.75 - 60.00 33.75 - 2.50 - 
KEF - - 5.26 - 57.89 21.05 - 14.47 1.32 
LES - - 2.08 - 68.75 23.96 - 4.17 1.04 
NWH - - 6.82 - 62.50 29.55 1.14 - - 
SES - - - - 61.63 34.88 - 3.49 - 
WGJ - - 3.26 - 58.70 30.43 - 7.61 - 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar,  CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, KEF- 
Keffa,  LES ? Long-eared Somali,  NWH ? North-West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
 
 
 
 249
 Appendix 23. Allele frequency for locus ILSTS05 for all the 11 Ethiopian populations. 
Allele 178 180 182 184 186 
ABG - 1.09 93.48 5.43 - 
ABA 1.09 - 78.26 20.65 - 
AFA 3.41 1.14 76.14 18.18 1.14 
CEH - - 95.00 5.00 - 
GUM 1.16 - 90.70 8.14 - 
HAH - - 87.50 12.50 - 
KEF - - 80.26 19.74 - 
LES 7.29 - 80.21 12.50 - 
NWH - - 92.22 7.78 - 
SES - - 90.70 8.14 1.16 
WGJ 2.17 - 84.78 13.04 - 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar,  CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa,  LES ? Long-eared Somali,  NWH ? North-West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
 
 
 
Appendix 24. Allele frequency for locus ILSTS11 for all the 11 Ethiopian populations. 
Allele 265 267 269 271 273 275 277 279 281 283 
ABG - - 42.39 13.04 3.26 2.17 3.26 32.61 3.26 - 
ABA - - 46.74 4.35 - - 10.87 28.26 7.61 2.17 
AFA - - 15.91 6.82 - 4.55 15.91 56.82 - - 
CEH - - 45.00 12.50 - 1.25 2.50 36.25 2.50 - 
GUM - - 37.21 3.49 - 1.16 9.30 48.84 - - 
HAH - - 32.50 2.50 - 1.25 17.50 42.50 3.75 - 
KEF 1.32 1.32 50.00 15.79 - - 6.58 17.11 7.89 - 
LES - 2.08 39.58 - - 1.04 21.88 34.38 1.04 - 
NWH - - 53.33 6.67 - 5.56 4.44 28.89 1.11 - 
SES - - 29.07 5.81 - 2.33 16.28 43.02 1.16 2.33 
WGJ - - 32.61 3.26 - - 27.17 35.87 1.09 - 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar,  CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa,  LES ? Long-eared Somali,  NWH ? North-West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
 
 
 250
 Appendix 25. Allele frequency for locus ILSTS44 for all the 11 Ethiopian populations. 
Allele 153 155 157 159 161 
ABG - 21.74 31.52 44.57 2.17 
ABA - 34.52 50.00 11.90 3.57 
AFA - 39.77 35.23 19.32 5.68 
CEH - 33.75 33.75 30.00 2.50 
GUM 1.16 39.53 44.19 11.63 3.49 
HAH - 38.75 35.00 18.75 7.50 
KEF - 36.84 42.11 13.16 7.89 
LES 2.08 46.88 33.33 14.58 3.13 
NWH - 37.78 36.67 22.22 3.33 
SES - 40.70 30.23 20.93 8.14 
WGJ - 38.04 43.48 15.22 3.26 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar,  CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa,  LES ? Long-eared Somali,  NWH ? North-West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
 
 
 
Appendix 26. Allele frequency for locus ILSTS87 for all the 11 Ethiopian populations. 
Allele 145 147 149 151 155 159 161 163 
ABG 9.78 83.70 1.09 1.09 - - 3.26 1.09 
ABA 9.78 70.65 6.52 - 1.09 9.78 2.17 - 
AFA 4.55 82.95 2.27 7.95 - - 2.27 - 
CEH 7.50 71.25 7.50 - - 8.75 5.00 - 
GUM 8.33 80.95 7.14 1.19 1.19 1.19 - - 
HAH 13.75 76.25 2.50 3.75 2.50 1.25 - - 
KEF 3.95 78.95 3.95 - - 5.26 7.89 - 
LES 22.92 70.83 3.13 - 1.04 - 2.08 - 
NWH 3.33 76.67 7.78 3.33 - 6.67 1.11 1.11 
SES 16.67 69.23 6.41 2.56 2.56 1.28 1.28 - 
WGJ 1.09 85.87 3.26 1.09 - 3.26 3.26 2.17 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar,  CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa,  LES ? Long-eared Somali,  NWH ? North-West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
 
 
 
 
 251
 Appendix 27. Allele frequency for locus INRA05 for all the 11 Ethiopian populations. 
Allele 131 139 141 143 145 
ABG - 60.87 29.35 1.09 8.70 
ABA - 41.30 44.57 9.78 4.35 
AFA 1.14 42.05 53.41 3.41 - 
CEH - 43.75 50.00 6.25 - 
GUM - 58.14 27.91 12.79 1.16 
HAH - 35.00 50.00 10.00 5.00 
KEF - 46.05 36.84 7.89 9.21 
LES - 46.88 43.75 6.25 3.13 
NWH - 51.11 37.78 6.67 4.44 
SES - 51.16 37.21 3.49 8.14 
WGJ - 60.87 22.83 10.87 5.43 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar,  CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa,  LES ? Long-eared Somali,  NWH ? North-West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
 
 
 
Appendix 28. Allele frequency for locus INRA63 for all the 11 Ethiopian populations. 
Allele 157 161 163 165 167 169 171 173 
ABG - - 6.52 8.70 54.35 25.00 5.43 - 
ABA - - 3.26 5.43 57.61 29.35 4.35 - 
AFA - - 2.27 14.77 47.73 35.23 - - 
CEH - - 7.50 7.50 62.50 17.50 5.00 - 
GUM 9.30 - - 15.12 52.33 20.93 2.33 - 
HAH - - 3.75 8.75 51.25 33.75 1.25 1.25 
KEF 6.58 1.32 3.95 11.84 44.74 28.95 1.32 1.32 
LES - - 10.42 8.33 50.00 20.83 9.38 1.04 
NWH - - 4.44 11.11 46.67 34.44 3.33 - 
SES - - 8.14 8.14 43.02 34.88 5.81 - 
WGJ - - 5.43 8.70 54.35 29.35 2.17 - 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar,  CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa,  LES ? Long-eared Somali,  NWH ? North-West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
 
 
 
 252
 Appendix 29. Allele frequency for locus INRA132 for all the 11 Ethiopian populations. 
Allele 143 145 147 
ABG 60.87 36.96 2.17 
ABA 52.17 44.57 3.26 
AFA 40.91 56.82 2.27 
CEH 63.75 31.25 5.00 
GUM 46.51 38.37 15.12 
HAH 50.00 47.50 2.50 
KEF 59.21 36.84 3.95 
LES 45.83 52.08 2.08 
NWH 56.67 43.33 - 
SES 50.00 50.00 - 
WGJ 41.30 54.35 4.35 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar,  CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa,  LES ? Long-eared Somali,  NWH ? North-West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
 
 
 
Appendix 30. Allele frequency for locus MAF35 for all the 11 Ethiopian populations. 
Allele 105 107 109 
ABG 90.22 9.78 - 
ABA 86.96 11.96 1.09 
AFA 93.18 6.82 - 
CEH 93.75 6.25 - 
GUM 93.02 6.98 - 
HAH 93.75 5.00 1.25 
KEF 94.74 5.26 - 
LES 85.42 14.58 - 
NWH 90.00 10.00 - 
SES 91.86 8.14 - 
WGJ 90.00 10.00 - 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar,  CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa,  LES ? Long-eared Somali,  NWH ? North-West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
 
 
 253
 Appendix 31. Allele frequency for locus MAF65 for all the 11 Ethiopian populations. 
Allele 116 118 120 122 124 126 128 130 132 134 136 138 
ABG 1.09 19.57 3.26 26.09 - 27.17 - - 2.17 17.39 3.26 - 
ABA 7.78 18.89 5.56 12.22 4.44 21.11 - 14.44 5.56 10.00 - - 
AFA 18.18 4.55 19.32 17.05 3.41 12.50 - 2.27 15.91 6.82 - - 
CEH 13.51 9.46 6.76 21.62 - 18.92 - 4.05 16.22 9.46 - - 
GUM 1.32 11.84 2.63 27.63 1.32 22.37 - 1.32 17.11 5.26 6.58 2.63 
HAH 5.00 13.75 5.00 18.75 7.50 11.25 - 10.00 16.25 10.00 1.25 1.25 
KEF 1.35 4.05 18.92 14.86 - 17.57 - 5.41 24.32 9.46 4.05 - 
LES 5.32 3.19 5.32 26.60 4.26 19.15 - 11.70 11.70 9.57 3.19 - 
NWH 3.41 7.95 15.91 25.00 - 17.05 1.14 - 18.18 10.23 1.14 - 
SES 8.14 8.14 5.81 15.12 2.33 12.79 - 10.47 25.58 11.63 - - 
WGJ 3.26 7.61 4.35 15.22 5.43 23.91 - 7.61 19.57 10.87 - 2.17 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar,  CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa,  LES ? Long-eared Somali,  NWH ? North-West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
 
 
 
Appendix 32. Allele frequency for locus MAF209 for all the 11 Ethiopian populations. 
Allele 102 104 106 108 
ABG 30.43 57.61 11.96 - 
ABA 22.83 51.09 26.09 - 
AFA 14.77 70.45 14.77 - 
CEH 27.50 57.50 15.00 - 
GUM 30.23 65.12 4.65 - 
HAH 26.25 60.00 12.50 1.25 
KEF 38.16 47.37 10.53 3.95 
LES 16.67 60.42 22.92 - 
NWH 23.33 63.33 13.33 - 
SES 17.44 69.77 12.79 - 
WGJ 20.65 50.00 29.35 - 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar,  CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, 
KEF- Keffa,  LES ? Long-eared Somali,  NWH ? North-West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji.
 254
 Appendix 33. Allele frequency for locus OarAE129 for all the 11 Ethiopian populations. 
Allele 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 173 175 177 
ABG 27.17 14.13 - - - - - - 14.13 29.35 8.70 - - - - 3.26 1.09 - 2.17 - 
ABA 35.23 20.45 - - - - - - 21.59 12.50 5.68 - - 1.14 - 3.41 - - - - 
AFA 12.50 9.09 - - - 1.14 - 3.41 18.18 25.00 10.23 - - 3.41 1.14 13.64 2.27 - - - 
CEH 35.94 15.63 - - - - - 1.56 7.81 20.31 14.06 - - - - 1.56 - 1.56 1.56 - 
GUM 30.23 13.95 - - - 2.33 - - 9.30 13.95 5.81 - - 15.12 2.33 4.65 - 1.16 1.16 - 
HAH 34.21 13.16 - - - - - - 13.16 22.37 7.89 - - 2.63 - 2.63 2.63 1.32 - - 
KEF 28.95 22.37 - - 2.63 - - 1.32 3.95 22.37 5.26 - 2.63 1.32 2.63 3.95 - - 2.63 - 
LES 22.34 13.83 - 1.06 - - 2.13 2.13 20.21 22.34 3.19 - 1.06 2.13 2.13 5.32 - - 1.06 1.06 
NWH 30.00 21.11 1.11 5.56 - - - 1.11 10.00 16.67 8.89 1.11 - - 1.11 1.11 - - 2.22 - 
SES 25.00 17.86 - - - 1.19 - 1.19 16.67 15.48 10.71 - - - 1.19 8.33 - - 2.38 - 
WGJ 27.17 17.39 - - - - - 3.26 17.39 18.48 2.17 - - 2.17 4.35 3.26 1.09 - 3.26 - 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar,  CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, KEF- Keffa,  LES ? Long-eared Somali,  NWH ? North-
 West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
 
 
 
 
 
 
 
 
 
 
 
 255
 Appendix 34. Allele frequency for locus OarFCB304 for all the 11 Ethiopian populations. 
Allele 120 124 128 132 134 136 138 140 142 144 146 150 154 156 158 160 162 164 166 168 170 174 180
 ABG - 2.17 - - - 2.17 - 45.65 9.78 5.43 14.13 - 13.04 - 2.17 - 2.17 2.17 - 1.09 - - - 
ABA 1.09 - - - - 3.26 - 50.00 14.13 3.26 7.61 3.26 6.52 - - 2.17 8.70 - - - - - - 
AFA - 1.14 - - - 3.41 - 50.00 14.77 3.41 10.23 - 12.50 - 3.41 - - - 1.14 - - - - 
CEH - 3.75 - - - 5.00 - 47.50 8.75 1.25 12.50 2.50 16.25 - 2.50 - - - - - - - - 
GUM - - - - 1.16 3.49 1.16 59.30 13.95 4.65 1.16 - 11.63 - - - 3.49 - - - - - - 
HAH - 1.25 1.25 - - 3.75 - 48.75 5.00 1.25 13.75 - 6.25 1.25 12.50 1.25 - - - - 2.50 - 1.25
 KEF - - - - - - - 50.00 22.37 3.95 2.63 2.63 9.21 - 7.89 1.32 - - - - - - - 
LES - - - - - 4.17 - 62.50 9.38 1.04 - - 6.25 1.04 3.13 2.08 2.08 1.04 1.04 4.17 1.04 1.04 - 
NWH - 1.11 - - - 2.22 - 50.00 11.11 8.89 3.33 - 21.11 - 2.22 - - - - - - - - 
SES - - - 1.16 - 4.65 - 60.47 9.30 1.16 5.81 - 8.14 - 5.81 2.33 - - 1.16 - - - - 
WGJ - - - - - 4.35 - 47.83 16.30 5.43 8.70 - 8.70 - - 2.17 - 3.26 1.09 - - 2.17 - 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar,  CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, KEF- Keffa,  LES ? Long-eared Somali,  NWH ? North-
 West Highland, SES ? Small-eared Somali, WGJ ? Woyto-Guji. 
 256
 Appendix 35. Allele frequency for locus SRCRSP03 for all the 11 Ethiopian populations. 
Allele 111 117 119 121 123 125 
ABG 27.78 - 13.33 30.00 28.89 - 
ABA 38.04 - 4.35 41.30 16.30 - 
AFA 19.32 - 17.05 53.41 10.23 - 
CEH 18.75 - 18.75 46.25 16.25 - 
GUM 52.33 - 10.47 30.23 6.98 - 
HAH 19.23 2.56 15.38 53.85 7.69 1.28 
KEF 36.84 - 1.32 52.63 9.21 - 
LES 40.63 - 5.21 44.79 8.33 1.04 
NWH 27.78 - 8.89 46.67 16.67 - 
SES 26.74 1.16 17.44 47.67 6.98 - 
WGJ 45.65 - 3.26 44.57 6.52 - 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, KEF- Keffa, LES ? Long-eared Somali, NWH ? North-
 West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji. 
 
 
 
 
 
 
 
 
 
 
 
 257
 Appendix 36. Test for genetic differentiation between pairs of Ethiopia goat populations for each locus. 
Population 
pairs 
BM 
1818 
BMS 
1494 
ILSTS 
05 
ILSTS 
11 
ILSTS 
44 
ILSTS 
87 
INRA 
05 
INRA 
63 
INRA 
132 
MAF 
35 
MAF 
65 
MAF 
209 
OarAE 
129 
OarFCB 
304 
SRCRSP 
03 
ABA & ABG  ns * ** ** *** ** ** ns ns ns *** ns ns * * 
AFA &ABG  ns ** ** *** ** ns *** ns * ns *** ns ** ns ** 
AFA &ABA  ** * ns *** ns *** ns * ns ns *** * *** * ** 
CEH &ABG  ns ns ns ns ns ** *** ns ns ns *** ns ns ns * 
CEH &ABA  ns ns ** * * ns ns ns ns ns * ns ns * ** 
CEH &AFA ns *** ** *** ns *** ns * * ns ns ns *** ns ns 
GUM &ABG  ns * ns ** *** ns ** ** ** ns * ns *** * *** 
GUM &ABA  ns ** ns ** ns ns ns *** * ns *** *** ** ns * 
GUM & AFA * ns ns * ns ns ** ** ** ns *** ** *** * *** 
GUM & CEH ns * ns * ns * * ** * ns * ns * ** *** 
HAH & ABG ns ns ns ** ** ns ** ns ns ns *** ns ns * *** 
HAH & ABA * ns ns ns ns * ns ns ns ns ns ns ns *** ** 
HAH & AFA ns ns ns * ns ns ns ns ns ns ** ns * ns ns 
HAH & CEH ns * ns ** ns ** ns ns ns ns ns ns ns ns ns 
HAH & GUM ns * ns ns ns ns ** ** ** ns ns ns ns *** *** 
KEF & ABG ns ns * ns *** * ns ns ns ns *** ns ns ** *** 
KEF & ABA ** ** ns ns ns ns ns * ns ns *** ** * ** ns 
KEF & AFA ns * ns *** ns ** ** ns ns ns ** *** ***  *** 
KEF & CEH ns ** * ns ns ns * * ns ns *  ns ** *** 
KEF & GUM ns ** * *** ns * * ns * ns * * ns ** ** 
KEF & HAH ns * ns *** ns ** ns ns ns ns **  * ** ** 
LES & ABG ns ns ** *** *** ns * ns ns ns *** * ns ** *** 
LES & ABA ns ns * ** ns ** ns ns ns ns ** ns ns ** ns 
 258
  
Population 
pairs 
BM 
1818 
BMS 
1494 
ILSTS 
05 
ILSTS 
11 
ILSTS 
44 
ILSTS 
87 
INRA 
05 
INRA 
63 
INRA 
132 
MAF 
35 
MAF 
65 
MAF 
209 
OarAE 
129 
OarFC 
B304 
SRCRSP 
03 
LES & AFA * ns ns *** ns *** ns *** ns ns ** ns ns * ** 
LES & CEH ns * ** *** ns *** ns ns ns ns ns ns ns *** ** 
LES & GUM ns ns ns * ns * ns *** ** ns * ** * ns ns 
LES & HAH ns ns * ns ns ns ns ns ns * ns ns ns ** * 
LES & KEF ns ns * *** ns *** ns ** ns * * ** ns * ns 
NWH & ABG ns * ns ns * ** ns ns ns ns *** ns ns ns ns 
NWH & ABA ns ns * * ns ns ns ns ns ns *** ns ns *** ns 
NWH & AFA *** ** * *** ns ns * ns * ns * ns *** ns ns 
NWH & CEH ns ns ns ns ns ns ns ns * ns ns ns ns ns ns 
NWH & GUM * ** ns * ns ns ns ** *** ns ns ns ** ns ** 
NWH & HAH * ns ns ** ns * ns ns ns ns ** ns ns *** ns 
NWH & KEF * ** * * ns ns ns ns ns ns ns * ns * ns 
NWH & LES ns * ** *** ns *** ns ns ns ns ** ns ns ** ns 
SES & ABG ns ** ns ** ** ns ns ns ns ns *** ns ns * *** 
SES & ABA ns ns * * * ns ns ns ns ns * * ns ** ** 
SES & AFA ns ns * ns ns * * * ns ns * ns ns ns ns 
SES & CEH ns *** ns ** ns * * ns * ns ns ns ns ns ns 
SES & GUM ns ** ns ns ns ns * *** *** ns ** ns ** * ** 
SES & HAH ns ns ns ns ns ns ns ns ns ns ns ns ns ns ns 
SES & KEF ns ** * *** ns ** ns ns * ns * ** ns ns ** 
SES & LES ns ns * * ns ns ns ns ns ns ns ns ns ns * 
SES & NWH ns ** ns ** ns * ns ns ns ns ** ns ns * ns 
WGJ & ABG ns ns ns *** *** ns * ns * ns *** * ns ns *** 
WGJ & ABA ns ns ns ** ns * * ns ns ns * ns ns * ns 
WGJ & AFA ns ns ns ** ns ns *** ns ns ns ** * * ns *** 
 259
  
Population 
pairs 
BM 
1818 
BMS 
1494 
ILSTS 
05 
ILSTS 
11 
ILSTS 
44 
ILSTS 
87 
INRA 
05 
INRA 
63 
INRA 
132 
MAF 
35 
MAF 
65 
MAF 
209 
OarAE 
129 
OarFCB 
304 
SRCRSP 
03 
WGJ & CEH ns ** ns *** ns * ** ns * ns ns ns ns * *** 
WGJ & GUM * ** ns * ns ns ns ** * ns ns *** * * ns 
WGJ & HAH ns ns ns ns ns ** ** ns ns ns ns * ns *** ** 
WGJ & KEF * ns ns *** ns ns ns ns ns ns * ** ns ** ns 
WGJ & LES ns ns ns ns ns *** * ns ns ns ns ns ns ** ns 
WGJ & NWH ns * ns *** ns ns ns ns * ns ** * ns * * 
WGJ & SES ns ns ns ns ns ** ns ns ns ns ns * ns ns ** 
ABG ? Abergalle, ABA? Arsi-Bale, AFA- Afar, CEH ? Central Highland, GUM- Gumez, HAH ? Hararghe Highland, KEF- Keffa, LES ? Long-eared Somali, NWH ? North-
 West Highland, SES ? Short-eared Somali, WGJ ? Woyto-Guji. 
*=P<0.05; **=P<0.01; ***=P<0.001 
 
 
 
 
 
 
 
 
 
 
 
 
 
Kalinowski, S. T. (2002). Evolutionary and statistical properties of three genetic distances. Mol Ecol 11(8), 1263-73. 
Nei, M., Tajima, F., and Tateno, Y. (1983). Accuracy of estimated phylogenetic trees from molecular data. II. Gene frequency data. J 
Mol Evol 19(2), 153-70. 
Takezaki, N., and Nei, M. (1996). Genetic distances and reconstruction of phylogenetic trees from microsatellite DNA. Genetics 
144(1), 389-99. 
 260
 Weir, B. S., and Basten, C. J. (1990). Sampling strategies for distances between DNA sequences. Biometrics 46(3), 551-82.