IMPROVEMENT IN EGG PRODUCTION TRAITS IN THE LIGHT LOCAL CHICKEN ECOTYPE USING A SELECTION INDEX

IMPROVEMENT IN EGG PRODUCTION TRAITS IN THE LIGHT LOCAL CHICKEN ECOTYPE USING A SELECTION INDEX BY OLEFORUH-OKOLEH, VIVIAN UDUMMA (NEE EWA) PG/Ph.D/02/32927 DEPARTMENT OF ANIMAL SCIENCE FACULTY OF AGRICULTURE UNIVERSITY OF NIGERIA, NSUKKA MAY, 2010

IMPROVEMENT IN EGG PRODUCTION TRAITS IN THE LIGHT LOCAL CHICKEN ECOTYPE USING A SELECTION INDEX BY OLEFORUH-OKOLEH, VIVIAN UDUMMA (NEE EWA) PG/Ph.D/02/32927 SUPERVISORS: PROF. DR. C.C. NWOSU PROF. L.N. NWAKALOR DEPARTMENT OF ANIMAL SCIENCE FACULTY OF AGRICULTURE UNIVERSITY OF NIGERIA NSUKKA MAY, 2010

I IMPROVEMENT IN EGG PRODUCTION TRAITS IN THE LIGHT LOCAL CHICKEN ECOTYPE USING A SELECTION INDEX BY OLEFORUH-OKOLEH, VIVIAN UDUMMA (NEE EWA) PG/Ph.D/02/32927 A THESIS SUBMITTED TO THE DEPARTMENT OF ANIMAL SCIENCE, FACULTY OF AGRICULTURE, UNIVERSITY OF NIGERIA, NSUKKA, IN FULFILLMENT FOR THE REQUIREMENT OF THE AWARD OF THE DOCTOR OF PHILOSOPHY DEGREE (Ph.D) IN ANIMAL BREEDING AND GENETICS. SUPERVISORS: PROF. DR. C.C. NWOSU PROF. L.N. NWAKALOR DEPARTMENT OF ANIMAL SCIENCE FACULTY OF AGRICULTURE UNIVERSITY OF NIGERIA NSUKKA MAY, 2010

II CERTIFICATION OLEFORUH-OKOLEH, VIVIAN UDUMMA, a Postgraduate student in the Department of Animal Science, Faculty of Agriculture, University of Nigeria, Nsukka with registration number PG/Ph.D/02/32927 has satisfactorily completed the requirements for the research work for the degree of Doctor of Philosophy in Animal Breeding and Genetics. The work embodied in this Project Report is original and has not been submitted part or full for any other Diploma or Degree of this or any other University, to the best of my knowledge. --------------------------------- ------------------------------------ Professor (Dr) C. C. Nwosu Professor L. N. Nwakalor SUPERVISOR SUPERVISOR ---------------------------------- ------------------------------------ Dr. S. O. C. UGWU EXTERNAL EXAMINER HEAD OF DEPARTMENT

III DEDICATION This work is dedicated to the loving memories of Chief Paul Ewa Aja, Pa Daniel Oko Inya, Hori Daniel Idam Isu, Elizabeth Ugo Isu, Elizabeth Ugo Nwachi and Michael Oko Ezeali. You taught and made me understand that my gender should never deter or limit me from achieving my visions and missions. Your advice gave me the grace and strength to continue and finish this study.

IV ACKNOWLEDGEMENT First and foremost, I would like to thank my supervisors Prof (Dr) C.C. Nwosu and Prof. L.N. Nwakalor, for continuous guidance, inspiration, invaluable help, and patience with me throughout the course of this programme. They have sacrificed their time to make my thesis more readable and my arguments intelligible. They also constantly challenged me to expand my point of view as a geneticist/breeder. It is not possible to convey here the depth of the debt of gratitude I owe them. I am grateful to the Department of Animal Science, Ebonyi State University Abakaliki for allowing me to work with her facilities in the Poultry Teaching and Research Unit. I could not have completed all the studies contained in this thesis without the assistance of Profs, M.O. Ozoje, and C. Onwuka of University of Agriculture, Abeokuta; Prof. S.N. Ibe of Michael Okpara University of Agriculture Umudike, and Mr. Femi of EBSUTH, Abakaliki, especially in broadening my knowledge of statistics/genetics data analysis. I do in a most special way appreciate the kindness and assistance of Prof. G. C. Okeke of the Department of Animal Science, University of Nigeria, Nsukka. My parents - Sir and Lady G.A. Ewa, Vincent Agha-Ewah, L., all my siblings (especially Onyinye I love you), Adeolu Adewale, Uche Muoneke, Felicia Nwanchor and Innocent Omah without your help these project would not have been possible. I have to also appreciate Profs, I.I. Osakwe, Jonny Ogunji, Dr Cosmos Ogbu and my other colleagues (both graduate students at Dept of Animal Science, UNN and staff at EBSU, Abakaliki) for being there when I needed them most. Finally, I most sincerely thank my amiable husband Barr. Vincent Oleforuh Okoleh: without your support, love, patience and understanding finishing this programme would not have been possible. It was terrible to be apart and I am glad the storm is over. I can only love you more. To you Almighty God what can I say thank you for being the author and finisher of my faith! Oleforuh-Okoleh, V.U. (Nee Ewa) University of Nigeria Nsukka May, 2010

V TABLE OF CONTENT Title page i Certification ii Dedication iii Acknowledgement iv Table of contents v-vi List of Tables vii List of Figures vii Abstract ix CHAPTER ONE INTRODUCTION 1-4 1.1 Background of Study 1 1.2 Statement of Problem 2 1.3 Objectives of the Study 3 1.4 Justification 3 CHAPTER TWO LITERATURE REVIEW 5-16 2.1 Egg Production in Chicken 5 2.1.1 Factors Influencing egg Production 5 2.2 Principles of Selection. 5 2.3 Selection for Body Weight and Egg Production Traits in Chicken 9-13 2.4 Use of Selection Index in Chicken 13-14 2.5 Basis for Short-Term Egg selection in Poultry 14 2.6 The Nigerian Local Chicken 14-16 CHAPTER THREE MATERIALS AND METHOD 17-25 3.1 Experimental Site 17 3.2 Experimental Animals 17 3.3 Foundation Stock 17 3.4 Management of birds 18 3.5 Maintenance of a control population 20 3.6 Data Collection 20 3.7 Data Analysis 20 3.7.1 Evaluating the Performance Characteristics... 20

VI 3.7.2 Estimation of Genetic Parameters 21 3.7.3 Measurement of Selection Applied 23 3.8 Construction of Selection Index using the Selection Criteria 24 CHAPTER FOUR RESULTS AND DISCUSSION 27-48 4.0 Results 27-41 4.1 Mean Performance of the Various Egg Production Traits Studied 27-31 4.1.1 Age at First Egg 27 4.1.2 Body Weight at First Egg 27 4.1.3 Weight of First Egg 27 4.1.4 Average Egg Weight 27 4.1.5 Total Egg Number 28 4.2 Estimates of Genetic Parameters of the Selection Criterion Traits 32-36 4.2.1 Heritability Estimates 32 4.2.2 Phenotypic and Genetic Correlation between Traits 35 4.3 Measurement of Selection Applied 37-40 4.3.1 Selection Differential, Selection Intensity and Selection Response 37 4.4 Selection Indexes used for the Selection of LLCE Hens 40 4.5 Discussions 41-47 4.5.1 Mean Performance of the Various Egg Production Traits Studied 41 4.5.2 Heritability Estimates 43 4.5.3 Phenotypic and Genetic Correlation between Traits 46 4.5.4 Measurement of Selection Applied 46 CHAPTER FIVE CONCLUSION AND RECOMMENDATIONS 49-51 REFERENCES 52-65 APPENDICES 66-88

VII LIST OF TABLES Page Table 1: Population Size, Effective Population Size and Change in Inbreeding Coefficient over the Three Generations of Selection 18 Table 2: Proximate Composition of Commercial Diets 19 Table 3: Vaccination Schedule for the Birds 19 Table 4: Analysis of variance table 21 Table 5: Analysis of Covariance table 22 Table 6: Mean (±SE) by population for traits studied 28 Table 7: Table 8: Mean (± SE) of the Traits Performance by Generation and Population 1 29 Mean (± SE) performance and phenotypic regression coefficients in selected and control populations 30 Table 9: Table 10: Table 11: Table 12: Heritability (± SE) estimates of the three selection criteria by generation and population of the LLCE a 34 Genetic (r g ), Phenotypic (r p ) and Environmental (r e ) correlation by generation and population of LLCE.. 36 Selection differential, Selection Intensity, Expected Direct Response, Estimated Realized Response and Estimated Index Response over three generations a 38 Estimated Index Score (Selected I s and Whole I μ population), Selection Intensity Factor (ῑ), Heritability of Index (h 2 ) Genetic Gain in Aggregate (ΔH) and Correlation of Index and Aggregate Genotype r IH, Expected Annual Genetic Response (ΔG Ai ) and Generation Interval L i 41

VIII LIST OF FIGURES Page Figure 1: Phenotypic trend of body weight at first egg in 3 generation selection 31 Figure 2: Phenotypic trend of egg weight in 3 generation selection 31 Figure 3: Phenotypic trend of egg number in 3 generation selection 32 Figure 4: Regression of BWFE response on generation number 39 Figure 5: Regression of AEW response on generation number 40 Figure 6: Regression of TEN response on generation number 40

IX ABSTRACT Fifty hens and five cocks from a random mating population of light local chicken ecotype (LLCE) were mated and the fertile eggs hatched to obtain the parent generation (G 0 ) used for this study aimed at improving egg production traits in the LLCE using a selection index. The hens were monitored for short-term (90-days from first day of lay) egg production traits namely: Body Weight at First Egg (BWFE), Average Egg Weight (AEW) and Total Egg Number (TEN). Data obtained were subjected to statistical analysis using SPSS (2001) and paternal half-sib model with Harvey (1990) to estimate descriptive statistics and genetic parameters respectively. These were employed in constructing the selection index. Selection for all the selection criteria (BWFE, AEW and TEN) was in the positive direction. Selected parents were mated to produce next generations G 1 and G 2. Selection differentials, selection intensities and genetic response due to selection were also estimated. A control population which spanned for three generations (each generation had its own control population) was used to monitor environmental changes and to estimate the genetic changes due to selection. There were significant increases (P<0.05) in BWFE, AEW, and TEN in the selected populations over the three generations of study such significant increases (P<0.05) were not observed in the control population. Heritability estimates for all traits in all generations and populations were moderate to high. The heritability of the index was also moderate. Such moderate to high heritability estimates indicate high additive genetic variances, implying that these traits were most passed on from the parents to their offspring. Low to high positive genetic and phenotypic correlations was observed between BWFE and AEW in all populations of study. The genetic correlation and phenotypic correlation between BWFE and EN, and between AEW and EN, was moderate to highly negative for all generations and populations of study. A positive genetic correlation was observed between AEW and TEN in G 2 of the selected population. A cumulative selection differential of 269.38g, 1.58g and 3.88 eggs were obtained for BWFE, AEW and TEN respectively. Selection response for traits increased over the generations in a fairly linear manner. Realized response per generation was estimated to be 94.22g, 0.84g and 4.85eggs for BWFE, AEW and TEN respectively. It is evident that the simultaneous inclusion of BWFE, AEW, and TEN in a selection index generally improved the performance of selected birds over the generations in the Light Local Chicken Ecotype.

1 CHAPTER ONE INTRODUCTION 1.1 BACKGROUND OF STUDY The report of the 4 th Food and Agriculture Organization (1973) expert consultation on Animal Genetic Resources presented two important objections concerning the endeavour to improve and conserve the local chicken. The first was as to whether the local strains still possess genes, which are useful to the vastly improved exotic strains given the centuries of genetic screening, which the latter have undergone. The second objection pointed to the fact that the process of further screening of the local chicken will be long, laborious and very expensive. However, the results of recent research using local chicken (Ikeobi and Peters, 1996; Ayorinde et al., 2001; Udeh and Omeje, 2001) indicate that the local chicken is a repository of advantageous genes. Secondly, with molecular genetic techniques, genetic improvement of the local chicken is fast and less expensive. Incidentally, third world countries, which established breeding programmes based on the dilution of indigenous germplasm by extensive crossbreeding programme, suffered failures. Those failed efforts have made livestock breeders aware of the importance of indigenous breeds in overall food production systems because of their adaptation to the environmental stress of the tropics. In spite of the large number of livestock and poultry in the nation, the animal protein intake per caput per day still falls below the minimum requirement level recommended by UN/FAO (Ayodele and Ajani, 1999). This has been traced to the low production of animals, which could be due to genetic and/or environmental constraints. The above underscores the need to improve the level of animal protein production in Nigeria. Of greater importance is the improvement of the poultry sector since it has a number of advantages including short generation interval, and production of large number of offspring, due to its peculiar reproductive traits (Ibe, 2001). Furthermore, poultry meat is generally accepted by all religions and societies. In many countries the development of agriculture and breeding programmes has resulted in serious changes in poultry breeding stocks during the last decades. The establishment of breeding institutions has led to a pronounced supra-regional propagation of certain chicken breeds due to improvements in performance. As a consequence the local breeds have decreased continuously to the same extent as the preferred high performance breeds have expanded. For instance, it was in a bid to satisfy the need for increased production and profitability in intensive production systems to meet the increase in demand for animal protein by the populace that new high yielding and fast growing poultry breeds

2 were introduced into the existing poultry production systems in Nigeria since the late 1950 s (Obioha, 1992). Incidentally, such introduction has resulted in non-integration of the local breeds considered as low producers into large-scale poultry production. Nigeria has rich chicken genetic resources. A good number of workers have documented the characteristics of the local chicken, in terms of morphological, physiological, behavioral and production attributes. Nwosu (1990) gave a review of these. Ibe (1990a, b) identified some major genes of tropical relevance in Nigerian local chicken populations. Perhaps, the most distinguishing feature for physical characterization at present is the body weight of the local chicken found in the various ecological zones of the country (Olori and Soniaya 1992). Observations have shown that local chicken found within the swampy rainforest and guinea savanna regions are lighter in weight than those found within the highlands and sudan savanna regions (Nwosu, 1979). Such differences in body weight can be used to categorize the local chicken broadly as Light Ecotype those with lighter weight and Heavy Ecotype those with heavier weight. Research data on the local chicken in the past 50 years (Hill, 1954; Oluyemi, 1979; Omeje and Nwosu, 1982; Nwosu, 1987; Udeh and Omeje, 2001) indicate that the Nigerian local chicken has useful genetic attributes that can be harnessed in crossbreeding programmes for the development of egg-type and meat-type chickens. However, there exist limitations to the realization of total heterosis in such crosses with the exotic because - the local chicken is unpedigreed, unselected and unsegregated (Omeje, 1985) hence, unlike the exotics the local chicken cannot be considered a purebred. Furthermore, crossbreds from purebred parents show heterosis to the extent that their gene frequencies differ unlike hybrids from similar lines that manifest total heterosis, (Pirchner, 1983). In order to incorporate the local chicken as a parent breed to produce strains of chicken that are adaptable to the local environment as well as achieving the much desired goal of making Nigeria self-sufficient in the sourcing of poultry breeding stock and boosting her poultry industry, there is need for selective breeding. The practice of selective breeding among local strains has been found advantageous (El-Issawi, 1975). The concept underlying selective breeding is variation. For within a group of individuals there exist allelic variations that affect the outcomes of quantitative traits such as growth, egg production and egg quality traits. 1.2 STATEMENT OF PROBLEM Not much study has been done on selection of local chicken for meat or egg production; most of the studies have been on crossbreeding with exotic birds. Oluyemi

3 (1979) after seven generations of mass selection on 12-week body weight of the local chicken concluded that the local chicken is not a potential broiler strain. Although the local chicken has been termed a low producer with regards to egg production (40-80 eggs /bird/year under extensive management system), studies relating to the development of the local chicken as a potential layer have shown appreciable improvement in egg production traits of the birds under improved management system (Hill and Modebe, 1961; Nwosu et al, 1979, Omeje, 1985; Tule, 2005). Nwosu and Omeje (1985) further noted that the local chicken has a genetic potential of producing 128 eggs /bird/year. It should be noted that the results of these studies were from random-bred populations. It is quite possible that the local chicken subjected to selection and improved management can do better. This has prompted the present selection study of the local chicken using a selection index approach. 1.3 OBJECTIVES OF STUDY The objectives of this study are to: 1. Estimate genetic parameters, namely heritability of body weight at first egg, egg number and egg weight, and genetic and phenotypic correlations between these traits in the Light Local Chicken Ecotype (LLCE); 2. Develop an appropriate selection index (I) for the selection of LLCE using body weight at first egg, and short-term (90 days from first egg) egg production and egg weight; 3. Evaluate and summarize selection applied and response over three generations of selection. 1.4 JUSTIFICATION Over the last decades, poultry management techniques in Nigeria have improved significantly, with rapidly increasing production. However, due to the high cost of production inputs: such as feed and drugs, and the control of the market by the few livestock contract companies, many individuals and farmers could not compete with the companies and had to give up chicken meat/egg production. For these individual farmers, a streamlined production of local chicken could be an option for alternative income generation and for diversification of the agricultural production base. Akinwumi (1979) reported that 92% of poultry production in Nigeria was derived from indigenous poultry stock. Similar reports though from Asian survey carried out by Prawirokusmo (1988) stated that about 40% of the egg production and 30% meat production

4 in Indonesia was a contribution made by the local type of chicken. Local chickens may appear to produce less than highly specialized exotic breeds, but they are highly productive in their use of local resources and more sustainable over the long term. Products from local poultry stocks are widely preferred because of pigmentation, taste and leanness (Haitook et al, 2003; Horst, 1988). Local chicken can thrive with limited care and feeding and are often more tolerant or resistant to diseases. They are also better able to cope with drastic changes in food and water supplies as well as harsh, variable and extreme weather and climatic condition. By neglecting to develop locally adapted breeds for higher productivity, an opportunity is being missed to help the developing world feed its people. Barker (1982) argued that there are large phenotypic and possibly genetic variations existing within the indigenous/local breeds and varieties. He suggested that the application of genetics towards improving these stocks should be undertaken through proper evaluation and documentation of these breeds on a suitable selection procedure designed to provide an optimum genotype to the farmer. This implies that a breeding strategy, which recognizes the introduction and development of pure breeds and selection within local breeds, is beneficial. Ahmed and Hasnath (1983) described the usefulness of such strategy in native Delish chicken. Hence, continuous efforts to develop pure lines for meat and egg production locally may equal or excel in the future the best currently available in the country. No detailed examination of genetic parameter estimates/variance components for egg production traits in Nigerian local chicken have been reported in literature. Consequently, the little breeding experimental programmes on this bird rely heavily on estimates obtained from exotic populations. For effective genetic improvement, knowledge of genetic parameter estimates of the particular breed or population to be improved is essential. Thus this study is imperative to achieving the genetic improvement of the Nigerian light local chicken ecotype.

5 CHAPTER TWO LITERATURE REVIEW 2.1 EGG PRODUCTION IN CHICKENS The egg production of a chicken is a result of many genes acting on a large number of biochemical processes, which in turn control a range of anatomical and physiological traits. With appropriate environmental conditions (nutrition, light, ambient temperature, water, sound health, etc.), the many genes controlling all the processes associated with egg production can act to allow the chicken to express fully its genetic potentials (Fairfull and Gowe, 1990). 2.1.1 SOME FACTORS INFLUENCING EGG PRODUCTION 1. BODY WEIGHT Body weight is regarded as a function of framework or size of the animal and its condition (Phillip, 1970). One of the main factors influencing egg size is body size (Robinson and Sheridan, 1982). Variation in body weight within a flock can be attributed to genetic variation and environmental factors that impinge on individuals (Ayorinde and Oke, 1995). The poultry producer wants birds of minimum possible size and weights that will maximize production of standard sized eggs at an economic rate and still maintain market carcass value at the end of the production period (Oke, et al. 2004). Body weight in poultry is known to be moderate to highly heritable and hence the selection of heavier individuals in a population of Nigerian light local chicken ecotype for example, should result in genetic improvement of the trait. Though various factors are known to affect egg production, there are conflicting reports on the effect of body weight on egg production. Evidence obtained by Du Plessi and Erasmus (1972) indicated that larger hens within a bloodline laid larger eggs than those with smaller body weights. Telloni et al. (1973) found that hens carrying the dwarf allele (dw) laid fewer eggs than the normal hens (dw B ), even when hens of the same body size were compared. However, Quemenur et al. (1988), cited by Fairfull and Gowe (1990), reported that dwarf broiler females lay as well as normal broiler females. Various findings using broiler birds in egg production experiment tend to suggest the possibility that the lower egg production of dwarf chickens may be due to close linkage of the dwarfing gene and genes determining egg production on the sex chromosome. For the development of egg production strains within the local chicken, it is necessary to establish the nature of the relationship existing between body weights of the local chicken and egg production parameters.

6 2. AGE The age at which a hen begins to lay eggs affects the total egg production in its life cycle, all things being equal. Selection of laying hens is normally based on partial records; improvement in production occurs largely in the first part of the laying cycle. Khalil et al. (2004) found that selection of hens with lower age at first egg leads to improvement of performance of egg production. Nwagu et al. (2007b) obtained a positive response in a female line population which they attributed to reduced age at sexual maturity. Liljedahl and Weyde (1980) had reported that contribution of age at sexual maturity to response to selection lies between 50 and 80% over 4 generations of selection. Age influences egg production especially within the first laying cycle and over the subsequent laying cycles (Gowe and Fairfull, 1982) in each laying cycle, egg production (per hen housed or per live hen) quickly rises to a peak and declines slowly thereafter to the end of the cycle, usually terminating with a natural or induced molt. In successive cycles of egg production, the peak egg production of a flock is usually lower and the rate of decline of egg production is more rapid (Fairfull, 1982). In other words most traits deteriorate with advancing age. The decline in weekly egg production throughout the cycle is well studied (Gavora et al., 1982; McMillian et al., 1986; Yang et al. 1989). Thus, within egg production cycles, egg production declines with increasing age while its variation increases. 3. DISEASE Disease affects egg production through mortality and morbidity (sub-clinical and clinical disease, inadequate nutrition, toxic elements, etc.). Mortality reduces the number of layers available to lay eggs, and morbidity reduces the laying ability of affected hens (Fairfull and Gowe, 1990). However the effects of morbidity and mortality on egg production records depend upon the age of the hens when affected. For instance, if birds are affected towards the end of the production period, little is lost in terms of economic returns in egg production, however, infection or conditions of instability in hens before or about their peak period of egg production would greatly affect the overall egg production records. 2.2 PRINCIPLES OF SELECTION The purpose of applied poultry breeding is to improve production qualities of the domestic fowl. Although altering and improving the environment, or physiological situation or manipulation of the animals contribute immensely towards improvement of their production qualities, the possibility remains that variation nevertheless still exists after

7 optimum non-hereditary conditions have been established. This is because some of the variations in the economic traits are genetic in character and improvement brought about by heredity tends to be permanent. The ultimate goal of a breeding programme is genetic improvement of traits defined in the breeding objective for the animal population. The poultry breeder does this by ranking his animals, culling those with the poorest evaluation while selecting the best evaluations as replacements. With successful selection, the progeny generation will on the average be better than the average of the population from which the parents were chosen, resulting in a genetic progress being obtained. The principle of selection, thus, is an integral concept in animal breeding it is the basis of genetic improvement programme (Cameron, 1997). Selection means differences in reproductive rates in a population, whereby animals with some characteristics tend to have more offspring than animals without these characteristics. That is, selection refers to the practice of causing or permitting superior individuals to produce more offspring than inferior ones. More precisely, selection is essentially concerned with replacing an existing population with one that is genetically superior. It can be defined as choosing of animals of higher genetic merit than average, to be parents of the next generation. Thus, the genes of the favoured animals tend to become more abundant in the population and those of the less favoured animals less abundant (Lerner, 1950). So far as genes produce effects which are consistently desirable in all combinations with other genes, the changes produced by selection are permanent (unless and until equally effective counter-selection has taken place); but in so far as the effects of the genes are desirable in some combinations and undesirable in others, many of the changes which selection produces when it is first practiced are lost (Lush, 1965). Genetic improvement may, therefore, consequently be measured by the change in a population mean or gene frequency from generation to generation (Leymaster et al., 1979). Invariably, selection can be performed both between and within populations (e.g. breeds). To screen animal populations and thereafter use those that have characteristics in line with a desired breeding goal can be a way to get results quickly, assuming the populations can be properly compared. For continuous and long-lasting effects, it is necessary to conduct selection within populations (Strandberg and Malmfors, 2006). The genetic progress, improvement, or change that can be attained in a trait/ several traits is influenced by the following factors. These factors include the following heritability, selection differentials, generation interval, and genetic, phenotypic, and environmental correlations between traits. The driving force for genetic improvement is

8 heritability defined by the genetic superiority achieved by the selection of parents. Not only does it provide the breeder with a measure of genetic variation, and of the variations upon which all the possibilities of changing population by breeding methods depend, but also, its importance rests on its properties as a measurement of the accuracy with which a genotype can be identified from the phenotype of an individual or of a group of individuals (Adedeji et al., 2006). Heritability shows how important efforts to improve a trait through improved management or environmental conditions may be compared to genetic selection (Akbas et al., 2002). Although heritability estimation allows the prediction of the amount of gain expected from given amount of selection, the accuracy of heritability estimate is valid only for the particular generation of the specific population from which the data used in arriving at it was derived. The methods that have been most commonly used to estimate the heritability of various traits observed in chickens are: parental half-sib correlation, maternal half-sib correlation, full-sib correlation, parent-offspring regression, and realized heritability (Kinney, 1969). Heritability estimation from the full-sib correlation may be biased as the maternal effect; the common environmental effect and the dominance effect are included in the dam variance component (Falconer, 1981). Higher heritability estimates for dam component than those of sire component for traits show the existence of dominance deviations and/or maternal effects (Oni et al., 2000). The changes in allelic frequency under selection act as a fraction of the selective advantage of the desired genotype and of the reproductive potential of the population. Selection differential is the increment between the mean of the selected group and the population from which they were selected (Nordskog, 1981). The magnitude of the selection differential depends on two factors namely: the proportion of the population included among the selected group, and the phenotypic standard deviation of the character. As the proportion of animals selected to be parent s decreases, the mean predicted genetic merit of selected animals increases (Al-Murrani, 1974). Since it is the intention of the breeder to make as much genetic gain as possible, it is of fundamental importance to make selection differential as large as possible. Cumulative selection differential is often used to quantify total selection pressure applied (Frahm et al., 1985). The cumulative selection differential can be compared with total response to evaluate effectiveness of selection for the primary trait. Lush (1965) expressed selection intensity as the percentage of the population permitted to reproduce itself and is used to express the amount of selection applied in a breeding population. Reduced selection intensities invariably yield decreased selection

9 response per generation (Eisen et al., 1973). Selection response also referred to as the genetic gain and measured as fundamental improvement/year, is the most critical aspect of efficiency of a breeding plan (Falconer and Mackay, 1996). The expected genetic gain, ΔG or response to selection, R is the difference between the mean phenotypes of the progeny and parental generations, which can be predicted given the selection differential, ΔS, and the regression coefficient relating genotype to phenotype (Gjedrem and Thodesen, 2005). Genetic gain depends on: how well the animals are evaluated or the accuracy of prediction; the amount of selection or selection intensity; the magnitude of genetic differences among animals or standard deviation of additive genetic values and; how rapidly better younger animals replace their parents, known as generation interval. The generation interval is defined as the average age of parents when their offspring are born, where the offspring are parents of the next generation (Ibe, 1998). The length of the generation interval must be considered when evaluating alternative selection strategies (Cameron, 1997). Barria and Bradford (1981) noted that a breeder aims at reducing the generation interval at the same time increasing selection response. Chambers (1990) indicated that sexual maturity had much influence on egg production whereby early maturing birds have more opportunities to lay than those which mature later. Consequently, most breeders base a large part of their selection on part records of egg production. Lerner (1958) asserted that the reduction in generation interval made possible by selection of parents at approximately 40 weeks of age outweighs the added precision of a full year s record in attaining the goal of improved egg production over the first laying year. Hill et al. (1996) noted that breeding efficiency could be lowered seriously by postponing the first breeding to a needlessly late age Genetic relationships occur between various traits in chicken due to gene actions such as linkage and pleiotropy (Zhao et al., 2007). Genetic correlations have always been an important part of carefully constructed breeding programs (Cassell, 2001). Consequently, much more is known today about genetic correlations between economically important traits in different environments than what was known when only the animal model was used in genetic experiments.

10 2.3 SELECTION FOR BODY WEIGHT AND EGG PRODUCTION TRAITS IN CHICKEN The selection scheme used to change an animal s pattern of growth results in both short-term and long-term effects on other traits such as tissue growth patterns, the onset of sexual maturity, and overall reproductive efficiency. Genetic selection in various poultry species has resulted in increased body weight at various points along the growth curve (Anthony et al., 1991). Selection for body weight in poultry species and the resulting effects on egg production traits has been addressed in numerous studies (Dunnington and Siegel, 1984, Marks, 1987; Siegel and Dunnington, 1987). The common conclusion suggests that the onset of sexual maturity results from the interaction of body composition, age, and body weight. Soller et al. (1984) investigated the minimum weight for onset of sexual maturity in chickens. They estimated the heritability of minimum weight at sexual maturity and reported it to be.34 and.84 for the chicken populations they studied, in which minimum weight at sexual maturity was closely correlated to early growth rate. Data from Oruwari and Brody (1988) concluded that an interaction between chronological age, body weight, and body composition at the onset of sexual maturity are inseparable. Zelenka et al. (1984) suggests that multiple thresholds of minimum chronological age, body weight and body composition result in females reaching sexual maturity. Danbaro et al. (1995) noted that heritability estimates for body weight at 7 weeks and 30 weeks were between 0.10 and 0.34 for all lines used in their study of genetic parameter estimates from a selection experiment in broiler breeders. Chambers (1990) reported that heritability based on additive genetic effects were about 0.4 for growth traits in chicken. Besbes et al. (1992) reported estimates of heritability for 40-week body weight of egg-laying type chicken using Restricted Maximum Likelihood with animal model as 0.5. This is close to the range of 0.59 obtained by Wei and Van der Werf (1992). Hagger (1993) reported estimates of 0.732 and 0.790 for body weight in females and males respectively in a layer strain while Jorjani et al. (1993) obtained a range of 0.54 to 0.65 estimates for body weight. Momoh and Nwosu (2009) worked with a Nigerian heavy chicken ecotype and reported that values of heritability estimates of body weight of the heavy ecotype increased from 0.18 at 4 weeks to 0.43 at 8 weeks and thereafter declined to 0.16 at the 16 th week to rise again to 0.30 at the 20 th week. They concluded that on the average, the body weight of the heavy ecotype could be described as being lowly to moderately heritable and suggested that the heavy ecotype has dual potential to be selected either as a meat- type or egg- type bird.

11 Estimates of genetic parameters for egg production traits have been extensively reported in egg-type chickens. Kinney (1969) obtained an average heritability for early egg weight in light breeds using white leghorn, its classical and reciprocal with Rhode Island Red as 0.45, 0.53, 0.45 and 0.52. These were obtained respectively from sire, dam, and sire and dam covariance components, and daughter-dam regression and indicated that in light breeds, egg weight had high heritability. Oni et al. (2000) reported that estimates of heritability pooled over generations from sire, dam and sire and dam components of variance were 0.15, 0.2, and 0.18 for age at sexual maturity; 0.13, 0.16 and 0.15 for egg number for 280 days of age; 0.24, 0.2 and 0.24 for average egg weight and 0.07, 0.18 and 0.16 for 40- week body weight for a strain of Rhode Island chicken under selection. Edriss et al. (1999) cited by Vali (2008) examined heritability coefficient of laying characteristics of indigenous chickens of Iran using sire variance component and reported estimates of 0.26, 0.86, and 0.80 for age at sexual maturity, number of eggs at 34 weeks of age and egg weight from maturity age to 34 weeks respectively. Nordskog (1981) reported the following heritability estimates of egg production traits by age and breed of chickens: average short term rate of egg production, 0.8; sexual maturity, 0.38 (Light breeds); early egg weight, 0.45 (Light breeds); early egg weight, 0.57 (Heavy breeds); mature egg weight, 0.46 (Light breeds); mature egg weight, 0.58 (heavy breeds); fertility, 0.02; and hatchability, 0.19. The reported estimates of heritability for egg number varied from 0.11 to 0.53 (Francesh et al., 1997; Nurgiartiningsih et al., 2002, 2004; Szwaczkowski, 2003). Luo et al (2007) reported heritability estimates of cumulative egg numbers within the range of 0.16 to 0.54 in their study using broiler breeder strains and suggested that the result indicates a moderate to low additive genetic variance for egg production in broiler breeders. They further noted that the estimates of heritabilities were relatively low at the beginning of the laying period and attributed this to the significant physiological changes for hens commencing egg production. Soltan (1997) obtained a significant selection response or genetic progress for egg number (ninety-days from first lay) in Baladi fowl of 7.1 eggs in three generations of mass selection. Venkatramaiah et al. (1986) reported an average genetic change per generation of 2.16 eggs in egg number and 146g of egg mass in selected sublines of White Leghorn. Many studies have demonstrated that reproductive performance of hens decreases as birds become heavier and fatter this especially evident for broiler breeders (Applely et al., 1994; Chen et al., 2006). It has been shown that when selecting for traits besides reproductive fitness; the result is a negative correlation between the selected trait and reproductive fitness (Lerner, 1954;

12 Nestor, 1977; Falconer, 1981; Dunnington and Siegel, 1984; Marks, 1987). This negative correlation has shown to be true between growth rate and reproductive traits such as female fertility and egg traits (Goodman and Shealey, 1977). With female reproduction, there are some positive correlations between growth and reproduction but majority are reported to be negative. For example, there is a positive correlation between body and egg weight, yet a negative relationship has been shown between high pullet body weight and normal egg production (Siegel, 1963). There are factors such as breed, management, and nutrition that influence the onset of sexual maturity. Long-term selection for growth in chickens has also resulted in the decrease in reproductive traits (Siegel and Dunnington 1985). Data from Japanese quail selected for heavier body weight under different nutritional environments has shown decreases in hatchability, egg production, and increased abdominal and carcass fat (Marks 1991). Genetic correlation between different part records of egg production is an important parameter for describing the dynamics of egg production and designing an early selection program. Besbes et al. (1992) reported that the genetic correlation between egg production for 26 to 38 wk and 26 to 54 wk was 0.66. Luo et al. (2007) worked with a broiler breed and indicated that the genetic correlations between the cumulative eggs for production week 19 and the total cumulative eggs till week 40 were as high as 0.81, and the genetic correlation between the fifth monthly records (egg production from production week 16 to 20) and the total cumulative egg numbers was 0.95. They concluded that in a balanced consideration of selection response and generation interval, early selection based on the first 19 week of cumulative egg numbers could effectively improve annual egg production in the broiler dam line. Harms et al. (1982) reported a negative correlation between body weight and egg number and that egg weight increased with body weight in a linear fashion from onset of egg production. Everett and Olusanya (1985) indicated that larger hens within a bloodline laid larger eggs than those with smaller body weight. This is supported by Ricklefs (1983), who reported that large body size resulted in large egg length, width, and mass, all factors affecting egg weight. Akbas et al., (2002) showed that there is a medium to high negative correlation between age at first egg and egg number (for the first twelve weeks of lay) and noted that this could indicate that cost of production could be lowered by lowering this trait and increasing egg production without significant diverse effect on body weight and egg weight. Adedeji et al. (2006) obtained a negative and low genetic and phenotypic correlation

13 between egg number and egg weight. Similar report was given by Sabri et al. (1999) and Zieba and Lnkaszewicz (2003). Under subtropical environments, growth rate as well as egg production are generally depressed by high ambient temperature (Bordas and Merat, 1984; Deeb and Cahaner, 2001). A significant interaction between environments and lines was found for laying traits, which indicated that the selected line was more sensitive to environmental changes than the control line for production level (Chen et al., 2004). Chen et al. (2009) studied the performance comparison of dwarf laying hens segregating for the naked neck gene in two different environments in France and Taiwan. In the selected line, the estimated heritabilities exhibited higher values for performance recorded in France than in Taiwan. These they attributed to the large environmental variability in Taiwan and concluded that changes in environment affected genetic parameters to a larger extent. This affirms the premise that estimating genetic correlations between traits measured in two environments is one approach to reveal within-line genotype environment (G E) interactions (Mathur and Horst, 1994). 2.4 USE OF SELECTION INDEX IN CHICKEN The theory and methods of deriving selection indices have been adequately developed in the literature (Smith, 1937; Hazel, 1943; Robinson et al., 1951). Hazel and Lush (1942) established the utility of a linear function of traits as a basis for multiple trait selection. In their work, they described the fundamental principles of index construction. They demonstrated how selection indices could be arrived at using nexus of the characters (with characters correlated to each other). Selection index theory, thus, provides an objective method of selecting a linear function of several traits by maximizing a combination of all relevant information (including family and individual phenotypic values, genetic and phenotypic variances and covariance with relative economic weights) into a single value (I) that is ranked for selection (Nwagu et al., 2007b). The fundamentals in the understanding these relationships is based on Sewall Wright s (1934) path coefficient model. The use of a selection index can increase the efficiency of simultaneous improvement in multiple egg production traits more than when selection is directed towards a single trait. Johari et al. (1999) noted that index selection has been a very effective method for improving egg production in chicken. Sharma et al. (1983) showed that selections on the basis of an index (using data on body weight at 8 weeks, egg production to 300 days, and percent hatchability) was relatively more efficient than tandem selection or independent culling levels selection for all traits except hatchability. Comparisons between index selection and mass

14 selection for various traits such as body weight at 8 and 20 weeks of age, 35-week egg weight, age at sexual maturity, and egg production to 260 days of age in white leghorn strains were made by Verma et al. (1984). For aggregate genetic response, index selection was found to be 2.76, 3.33, 13.66, 1.32 and 1.53 times more efficient than direct selection for egg production, egg weight at 35-week of age, initial egg weight, 20-week body weight and 8- week body weight respectively. Singh et al. (1984) compared various selection indices to improve egg production in white leghorn flocks when 10% of the cockerels were selected in each population, selection of 40% of the females based on index selection was found to give better genetic response than when selection with different intensity was made on individual traits. Ayyagari et al. (1985) recorded similar observations in another white leghorn population and so did Makarechian et al. (1983) while working with indigenous chickens of southern Iran. 2.5 BASIS FOR SHORT TERM EGG SELECTION IN POULTRY Selection for early period part-records, generally up to 40 wk of age, is the usual approach for improving egg production in egg-type and meat-type chickens (Dempster and Lerner, 1947; Fairfull and Gowe, 1990). The first study on using partial egg production data as a selection criterion was according to Dickerson and Hazel (1944). In that study, they had reported that improvement was maximized when the interval between generations was kept in a minimum (Lowe and Garwood, 1980). Kinney (1969) observed that short-term records of egg production have positive correlation to full-time. Lowe and Garwood (1980) concluded that in poultry, selection on the early part of the egg record and improvement in animal production is a standard practice in poultry breeding. Bohren (1970) supported this concept by empirically showing that annual egg production is estimated with reasonable precision from an earlier part of the egg record. Nestor et al. (1996) and John et al. (2000) reported an average of 0.45 as genetic correlation between partial and residual egg number from Light Sussex and brown leghorn population. There are many advantages of selection based on partial records genetic advantage; these include to shorten the generation interval; reduce record keeping; achieve higher reproductive rates from layer; utilize high genetic correlation between part-time and full-time performance; achieve rapid selection response (Foster, 1981; Boukila et al., 1987; Nwosu, 1990). However, selection based on part-records has significant unfavorable effects on some important traits, including earlier age at first egg, poorer laying persistency after peak (Yang, 1994), and poor selection accuracy (Luo et al., 2007).

15 2.6 THE NIGERIAN LOCAL CHICKEN Nigeria is richly endowed with a large population of local chickens. According to RIM (1992) the local chicken population in Nigeria is about 103 million, 85% of which are found in the north and the rest in the southern part of the country. Local chickens are kept for various purposes, which include the provision of meat and eggs for home consumption, religious ceremonies and barter, as such they play a big role in rural as well as national economy (FAO, 2004). What is generally referred to as local chicken is a pool of heterogeneous individuals, which differ, in adult size, body weight, feather and plumage pattern (Fayeye et al., 2005). Several authors have reported on the unique adaptation features of the Nigerian local chicken (Adebambo et al., 1999; Ibe, 1993). These include their relatively small adult size, generally flighty nature, relatively thick eggshell, high tolerance to some tropical diseases and parasites and the presence of some major genes affecting feather structure and feather distribution for example naked neck and frizzle feathers. Peters et al. (2004) noted that the major genes of frizzling and naked neck are important as they enhance the themo-regulatory activities of the birds. There have been various genetic studies on the Nigerian local chicken round the ecological zones of the country in the east by Nwosu and Omeje (1985), Udeh and Omeje (2001), Tule (2005) and Momoh et al. (2007), in the west by Adebambo et al. (1999) and Peters et al. (2004) and in the north by Fayeye et al. (2005). Their main aims were mainly to characterize the Nigerian local chicken using its physical features and crossbreeding to monitor the performance of the progenies. Two major strategies have been offered to achieve new breed development: intense selection within the population followed by crossing (Osman and Robertson, 1968), and crossbreeding to create a heterogeneous broad-based population followed by intense selection (Pirchner, 1983). The first strategy requires extensive sampling of local chickens throughout the country to select desirable genetic material that will constitute the base population, which will subsequently be crossed either with themselves or with other populations. Success of this strategy will largely depend on the initial response to selection. Oluyemi (1979) did not obtain appreciable response following 7 years of selection for 12-week body weight within the Nigerian local chicken population. This is despite the belief that considerable additive genetic variation exists in the local chicken population since they have not been subjected to any conscious artificial selection (Ibe, 2001). The findings of Oluyemi (1979) tend to suggest that the Nigerian local chicken would not be suitable for development as a broiler breed. Incidentally, several works by researchers on the Nigerian