Schriftenreihe des Instituts für Tierzucht und Tierhaltung der Christian-Albrechts-Universität zu Kiel, Heft 225, 2019
©2019 Selbstverlag des Instituts für Tierzucht und Tierhaltung der Christian-Albrechts-Universität zu Kiel Olshausenstraße 40, 24098 Kiel Schriftleitung: Prof. Dr. J. Krieter ISSN: 0720-4272 Gedruckt mit Genehmigung des Dekans der Agrar- und Ernährungswissen- schaftlichen Fakultät der Christian-Albrechts-Universität zu Kiel Aus dem Institut für Tierzucht und Tierhaltung der Agrar- und Ernährungswissenschaftlichen Fakultät der Christian-Albrechts-Universität zu Kiel ______
Conservation Genetics and Management of Local Breeds
DISSERTATION
zur Erlangung des Doktorgrades Doctor scientiarum agrariarum (Dr. sc. agr.) der Agrar- und Ernährungswissenschaftlichen Fakultät der Christian-Albrechts-Universität zu Kiel
vorgelegt von
M.Sc. Jonas Schäler aus Potsdam Brandenburg
Kiel, 2018
Dekan: Prof. Dr. Dr. Christian Henning 1. Berichterstatter: Prof. Dr. Georg Thaller 2. Berichterstatter: Prof. Dr. Dirk Hinrichs Tag der mündlichen Prüfung: 23.01.2019
______Die Dissertation wurde mit dankenswerter Unterstützung des Ministeriums für Energie, Landwirtschaft, Umwelt, Natur und Digitalisierung im Rahmen der Europäischen Innovationsförderungspartnerschaft (EIP Agri) angefertigt.
Table of Contents ______
Table of Contents
General Introduction…………………….…………………………………………………………………..1
Chapter One Comparison of ancestral, partial, and genomic inbreeding in a local pig breed to achieve genetic diversity…………………………………………………………...………………………....5
Chapter Two Performance of a novel breeding value in context of breed conservation…...... 29
Chapter Three Genetic diversity and historic introgression in German Angler and Red Dual Purpose cattle and possibilities to reverse introgression…………..………………..…...37
Chapter Four The benefit of native genetic contribution in a local cattle breed……………………...... 59
Chapter Five Implementation of breed-specific traits for a local sheep breed……………………………...... 73
Chapter Six Exploration of conservation and development strategies for local cattle breeds in Northern Germany………………………………………………………………………………..……….97
General Discussion………………...... 123
General Summary………………………………..………………….………………………...…………..141
Allgemeine Zusammenfassung………………………………………………………………………...143
General Introduction ______
General Introduction
An impressive variety of breeds emerged since the domestication of livestock species due to a long history of migrations, selection, and adaptation (Groeneveld et al. 2010). Many well- defined breeds are utilized nowadays for a broad range of purposes and express distinct performance levels depending on specific local environments. Only few specialized and high- yielding breeds dominate the commercial sector, but still many local breeds can be found mostly for socioeconomic reasons (Tisdell 2003). However, advanced technology and efficient breeding programs further reduce the competitiveness of those local breeds. As a consequence, high-yielding breeds increasingly replace local breeds (Meuwissen 2009) causing negative impact on population genetic properties of local breeds (FAO 2010; Wang et al. 2017). Population sizes of numerous local breeds already decreased dramatically and some of them are threatened by extinction (Fernández et al. 2011). Conservation of genetic diversity is therefore of major importance (Frankham 1995; Caballero and Toro 2002) and there is international agree to maintain breeds and their genetic diversity as a unique genetic resource for the future (Boettcher et al. 2010).
In general, the genetic diversity of local populations is impacted by small effective population sizes and lack of pedigree recording which in addition enforces inbreeding. With this respect,
Chapter One deals with the estimation of relatedness and different inbreeding measurements based on pedigree and on the genomic level for a small local pig breed in Northern Germany.
Furthermore, correlations between inbreeding estimators were investigated. Another aspect is the fact that genetic material from high-yielding breeds has been introgressed into local breeds for decades in order to increase genetic gain and enhance profitability. As a consequence, genetic diversity per se increased whereas native genetic diversity decreased in populations.
Therefore, an appropriate instrument was developed in Chapter Two, which enables to select on native genetic contribution (NC) and thus, increase native genetic diversity demonstrated for two local cattle breeds, the German Angler and the Red Dual-Purpose cattle. The - 1 -
General Introduction ______population parameter of NC was considered as a trait and its genetic parameters were estimated with different linear mixed models. To find a balance between genetic gain, inbreeding, and the enhancement of NC or native uniqueness the advanced Optimum
Contribution Selection (OCS) accounting for historic introgression from Wellmann et al.
(2012) was applied for the two local cattle breeds in Chapter Three. Thereby, results for genetic gain of the conventional advanced OCS and of modified OCS scenarios, in which NC was included as a trait rather than a constraint, were compared in order to recover the native genetic background. In addition, population genetic parameters for both breeds were identified over the last generations. To date, there is no scientific justification for the utility of achieving NC or native genetic diversity within single breeds. In Chapter Four NC was used as a trait to prove correlations between NC and major traits on genomic level. Conventional breeding programs and intense management may cause negative effects regarding maintenance of special characteristics of local breeds. Chapter Five describes an approach how to identify and evaluate breed-specific characteristics for a typical environment in contrast to performance testing on station for a local sheep breed. Another important part beside conservation genetics is the aspect of breed management to maintain and support local breeds as far as possible. In Chapter Six a combination of qualitative and quantitative decision tools was adapted to identify strengths, weaknesses, opportunities, and threats. The output was transferred to the primary level in order to form the most objective conservation and development strategies for the two local cattle breeds.
REFERENCES
Boettcher PJ, Tixier-Boichard M, Toro MA, Simianer H, Eding H, Gandini G, Joost S, Garcia
D, Colli L, Ajmone-Marsan P, The GLOBALDIV Consortium (2010) Objectives,
criteria and methods for using molecular genetic data in priority setting for conservation
of animal genetic resources. Animal Genetics 41(1):64-77
- 2 -
General Introduction ______
Caballero A, Toro MA (2002) Analysis of genetic diversity for the management of conserved
subdivided populations. Conservation Genetics 3:289-299
FAO (2010) Breeding strategies for sustainable management of animal genetic resources.
FAO Animal Production and Health Guidelines. No. 3, Rome
Fernández J, Meuwissen THE, Toro MA, Mäki-Tanila A (2011) Management of genetic
diversity in small farm animal populations. Animal 5(11):1684-1698
Frankham R (1995) Conservation genetics. Annual Review of Genetics 29:305-327
Groeneveld LF, Lenstra JA, Eding H, Toro MA, Scherf B, Pilling D, Negrini R, Finlay EK,
Jianlin H, Groeneveld E, Weigend S, The GLOBALDIV Consortium (2010) Genetic
diversity in farm animals – a review. Animal Genetics 41(1):6-31
Meuwissen THE (2009) Genetic management of small populations: A review. Acta
Agriculturae Scandinavica, Section A – Animal Science 59(2):71-79
Tisdell C (2003) Socioeconomic causes of loss of animal genetic diversity: analysis and
assessment. Ecological Economics 45:365-376
Wang Y, Bennewitz J, Wellmann R (2017) Novel optimum contribution selection methods
accounting for conflicting objectives in breeding programs for livestock breeds with
historical migration. Genetics Selection Evolution 49:45
Wellmann R, Hartwig S, Bennewitz J (2012) Optimum contribution selection for conserved
populations with historic migration. Genetics Selection Evolution 44(1):34
- 3 -
Chapter One ______
Chapter One
Comparison of ancestral, partial, and genomic inbreeding in a local pig breed to achieve genetic diversity
J. Schäler1, B. Krüger2, G. Thaller1 & D. Hinrichs3
1Institute of Animal Breeding and Husbandry, Christian-Albrechts-University, Hermann- Rodewald-Straße 6, 24098 Kiel, Germany 2Institute of Zoology, Christian-Albrechts-University, Am Botanischen Garten 1-9, 24118 Kiel, Germany 3Faculty of Organic Agricultural Sciences, Department of Animal Breeding, University of Kassel, Nordbahnhofstraße 1a, 37213 Witzenhausen, Germany
Published in Conservation Genetics Resources (in press)
- 5 -
Chapter One ______
ABSTRACT
Inbreeding is omnipresent and unavoidable in population genetics. High levels of inbreeding result in a reduction of genetic diversity and inbreeding depression. The avoidance of inbreeding is a primary goal for the management of small populations and very important for the design of breeding programs in order to create productive progeny. The objective of this study was investigate how to maintain genetic diversity in the progeny of 76 selection candidates of a local pig breed by identifying individual relatedness and different inbreeding coefficients. In addition, valuable comparisons of ancestral, partial, and genomic inbreeding were examined. The data-set included the pedigree of 1,273 individuals born between 1980 and 2015 and genotypes of selection candidates born between 2004 and 2014. Classical, ancestral, and partial inbreeding coefficients were calculated based on pedigree information.
Genomic coefficients were calculated by using four approaches: (I) variance of additive genetic values, (II) SNP homozygosity, (III) uniting gametes, and (IV) runs of homozygosity.
Inbreeding levels of selection candidates were not high and just a few animals showed increased inbreeding coefficients. This result was in accordance with estimated relatedness among individuals. The lowest and highest pedigree inbreeding were estimated within partial
(0.004-0.006) and ancestral concepts (0.024-0.090). Genomic inbreeding was low and showed an obvious difference between different genomic coefficients (0.000-0.012). Correlations between pedigree and genomic inbreeding coefficients ranged from -0.44 to 0.57. Ballou’s concept of ancestral inbreeding was moderately correlated with genomic estimators of homozygosity (0.39), uniting gametes (0.24), and runs of homozygosity (0.49). Kalinowski’s concept of ancestral inbreeding was negatively correlated with genomic inbreeding measurements regarding the variance of additive genetic values (-0.44) and uniting gametes (-
0.29). Correlations of partial inbreeding varied between 0.22 and 0.48, where Kalinowski’s concept of ‘new’ inbreeding was positively correlated with all genomic inbreeding coefficients (0.25-0.48). However, Lacy’s concept of partial inbreeding correlated positively
- 6 -
Chapter One ______with genomic inbreeding regarding homozygosity (0.26) and runs of homozygosity (0.22).
Ancestral and partial inbreeding can have great importance concerning purging of deleterious alleles through individual mating to obtain genetic diversity especially for small populations.
Inbreeding estimators based on ROH may represent ancestral and partial inbreeding concepts better than genomic coefficients of additive genetic values, SNP homozygosity, or uniting gametes.
Keywords: genetic diversity, genomic inbreeding, local breed, ROH
- 7 -
Chapter One ______
INTRODUCTION
Inbreeding is unavoidable and changes genotype frequencies by increasing homozygosity in populations with finite size (Curik et al. 2017). An increase in homozygosity negatively affects heterozygosity, which is the term of an individual’s genetic diversity. Thus, high levels of inbreeding result in a reduction of genetic diversity and inbreeding depression (Haig et al.
1990; Zhang et al. 2015). Conservation of genetic diversity is a major issue (Frankham 1995;
Caballero and Toro 2002) and according to Boettcher et al. (2010), there exists a wide agreement to conserve breeds and their genetic diversity. Local breeds especially are threatened by the loss of genetic diversity as genetic resources due to socioeconomic aspects
(Tisdell 2003) and the causes of population genetics, e.g. a small population size, genetic drift through introgression, and defective management of inbreeding (FAO 2010; Wang et al.
2017). The measurement of inbreeding is very important due to the effects of inbreeding on individual fitness and population dynamics (Kardos et al. 2015). Hence, the avoidance of inbreeding is a primary goal in the management of small populations (Kalinowski et al.
2000). According to Zhang et al. (2015), information on inbreeding is crucial in the design of breeding programs to control the increase in inbreeding and inbreeding depression in the progeny. In the literature, several methods have been developed and extended to estimate classical inbreeding (Wright 1922; Meuwissen and Luo 1992; VanRaden 1992), ancestral inbreeding (Ballou 1997; Kalinowski et al. 2000), and partial inbreeding coefficients (Lacy
1989; Lacy et al. 1996; Lacy 1997) based on pedigree records. However, four different approaches have been developed so far in order to estimate genomic inbreeding coefficients
(Wright 1948; VanRaden 2008; McQuillan et al. 2008, Yang et al. 2010). Nevertheless, the potential of purging a population of its genetic load through management intervention, e.g. mating systems, to reduce disastrously high levels of inbreeding depression is also of interest besides the impact of inbreeding (Ballou 1997). Therefore, mating systems based on relatedness are preferred to may achieve purging effects between inbred but unrelated
- 8 -
Chapter One ______individuals. Research on inbreeding is of major importance to increase our knowledge of inbreeding and inbreeding depression, respectively (Curik et al. 2017). Ancestral inbreeding
(Ballou 1997; Kalinowski et al. 2000) especially is relevant in order to identify the inbreeding to which individual’s ancestors have been subjected. The idea behind it was that an inbred animal with inbred ancestry should be less susceptible to inbreeding depression than an inbred individual with non-inbred ancestors (Suwanlee et al. 2007). Purging effects are associated with ancestral inbreeding, which reduces the inbreeding depression that occurs in isolated populations of small size (Ballou 1997; Swindell and Bouzat 2006). However, partial inbreeding according to Lacy (1989; et al. 1996; 1997) examines whether the alleles contributing to inbreeding effects were distributed uniformly across founder genomes or were distributed from specific founders. Hence, an offspring or selection candidate with high founder distribution should receive necessary adjustments regarding the handling during individual mating.
The aim of this paper was the identification of individual relatedness and different inbreeding coefficients of selection candidates from a local pig breed in order to achieve genetic diversity within a mating system for the progeny. To date, no studies have been performed which simultaneously explore the comparison of ancestral, partial, and genomic inbreeding to extend our knowledge. Hence, worthwhile comparisons of ancestral, partial, and genomic inbreeding were investigated in case of a local breed.
MATERIALS AND METHODS
Animals
Data included pedigree and genomic information for the Angler saddleback, which is a local pig breed in northern Germany. The first data-set contained the pedigree information of 1273 individuals born between 1980 and 2015. The second data-set comprised the genomic information of 84 individuals (16 male and 68 female) born between 2004 and 2014. Animals
- 9 -
Chapter One ______were genotyped with the PorcineSNP60v2 BeadChip by illumina. All data were provided by the Institute of Zoology of the University of Kiel in Germany. The pedigree data was checked using the ‘completeness’ function (Fig.1) of the R-package ‘optiSel’ (Wellmann 2018) within the R-statistic software (R Core Team 2018).
Fig. 1 Pedigree completeness of the Angler saddleback
The mean index of pedigree completeness (MacCluer et al. 1983) was 0.96 and the mean of equivalent generations was 7.49 within the pedigree. Number of equivalent, complete generations is defined as the sum of the proportion of known ancestors over all generations traced. The mean numbers of full and maximal generations traced were 3.75 and 17.63. The genomic data quality check and filtering of the 84 genotypes was performed using the software package PLINK (Purcell et al. 2007). At first, five individuals were removed due to a missing genotype rate of more than 10 %. Secondly, all markers which had tested positive for a minor allele frequency below 5 % were also removed. Finally, markers which did not correspond to the Hardy-Weinberg equilibrium (P > 0.0001) were excluded from the dataset.
In addition, mapped markers on the X and Y chromosomes and markers with undefined map positions were excluded from the analysis. All genotyped animals which had not been listed in the herd book were also excluded from the analysis because these individuals had no
- 10 -
Chapter One ______breeding purpose. In total, 76 genotyped animals with marker information on 36,183 markers were available for further analysis.
Pedigree-based inbreeding coefficients
Different inbreeding coefficients (퐹) were calculated for 76 individuals based on their pedigree information. Classical inbreeding coefficients were computed using the GRAIN and
VANRAD programs in the FORTRAN77 software package PEDIG (Boichard 2002). The classical inbreeding coefficient according to Wright (1922) is defined as
′ 1 푛+푛 +1 퐹 = ∑ ( ) (1 + 푓 ) (1) 푊푟푖푔ℎ푡(푥) 2 푎 where 푥 is the individual, 푛 and 푛’ are the number of generations from sire and dam respectively to the ancestor in question, and 푓푎 is the coefficient of inbreeding worked out from the pedigree accessed through the coefficients of inbreeding of sire and dam respectively. The method proposed by VanRaden (1992) upgraded Wright’s concept of inbreeding by replacing unknown inbreeding coefficients with average inbreeding coefficients in the same generations. Inbreeding estimates were computed with the formula from Quaas
(1976) as
푖 2 퐴푖푖 = ∑푗=1 퐿푖푗퐷푗푗 (2)
푡ℎ where 퐴푖푖 is the individual of the 푖 diagonal element of the A matrix (additive genetic relationship matrix) and is equal to the inbreeding coefficient of the 푖푡ℎ individual aggregated with one. The lower triangular matrix 퐿 includes the fraction of genes which individuals derive from their ancestors, and 퐷 is a diagonal matrix containing the additive genetic variances within the family (Meuwissen and Luo 1992). Matrix elements 퐿푖푗 and 퐷푗푗 were calculated like the A matrix described by Meuwissen and Luo (1992). A detailed derivation of the computation of 퐴푖푖 is proposed by Meuwissen and Luo (1992). These estimates will hereafter referred to as 퐹푉푎푛푅푎푑푒푛(푥), where 푥 is the individual. Ancestral inbreeding
- 11 -
Chapter One ______coefficients were calculated using the algorithms of Ballou (1997) and Kalinowski et al.
(2000) from the GRAIN program. Ballou (1997) defines ancestral inbreeding as
[퐹퐵푎푙푙표푢(푠)+(1−퐹퐵푎푙푙표푢(푠))∗퐹푊푟푖𝑔ℎ푡(푠)+퐹퐵푎푙푙표푢(푑)+(1−퐹퐵푎푙푙표푢(푑))∗퐹푊푟푖𝑔ℎ푡(푑) ] 퐹 = (3) 퐵푎푙푙표푢(푥) 2 where 푠 and 푑 describe the sire and the dam of individual 푥. The ancestral inbreeding concept of Kalinowski et al. (2000) summarises the historical inbreeding component of 퐹푊푟푖푔ℎ푡(퐴) as
퐹퐾푎푙푖푛표푤푠푘푖(푥), where alleles are homozygous if they have met in the population history. This ancestral inbreeding coefficient only comprised the ancestral inbreeding of relationships from common ancestors on both sides of the pedigree. Thus, if 퐹푊푟푖푔ℎ푡(퐴) is zero then
퐹퐾푎푙푖푛표푤푠푘푖(푥) is also zero. The remaining component of 퐹푊푟푖푔ℎ푡(퐴) deducting the ancestral inbreeding 퐹퐾푎푙푖푛표푤푠푘푖(푥) is summarised as ‘new’ inbreeding 퐹푁푒푤(푥) from Kalinowski et al.
(2000) and can be defined as
퐹푁푒푤(푥) = 퐹푊푟푖푔ℎ푡(푥) − 퐹퐾푎푙푖푛표푤푠푘푖(푥). (4)
Hence, new inbreeding refers to that part of 퐹푊푟푖푔ℎ푡(푥) whereby alleles are homozygous and identical by descent but have not already occurred together in the pedigree. Equally to
퐹퐾푎푙푖푛표푤푠푘푖(푥), if 퐹푊푟푖푔ℎ푡(퐴) is zero then 퐹푁푒푤(푥) is also zero. The partial inbreeding coefficient was calculated using the algorithm of Lacy et al. (1996) and Lacy (1997), which were also implemented in the GRAIN program. Partial inbreeding is defined as the probability of an individual being homozygous for an allele transmitted by a certain ancestor.
Partial inbreeding coefficients accumulated across all ancestors determines the total inbreeding coefficient. The classical inbreeding coefficients computed using GRAIN and
VANRAD are hereafter referred to as FWright and FVanRaden, respectively. In the following, ancestral inbreeding according to Ballou (1997) and Kalinowski et al. (2000) is referred to as
FBallou and FKalinowski. The concept of ‘new’ inbreeding (Kalinowski et al. 2000) is hereafter referred to as FNew and partial inbreeding (Lacy et al. 1996; Lacy 1997) as FLacy. All pedigree- based coefficients were calculated with a base generation of 1980.
- 12 -
Chapter One ______
Genomic-based inbreeding coefficients
The computation of the first three genomic inbreeding coefficients (퐹) for 76 individuals was performed using the software package GCTA (version 1.23) from Yang et al. (2011). This software estimates the inbreeding coefficient from the SNPs by the three different methods described below (Formulae 6-8). The basis for the estimation of genomic inbreeding coefficients is the calculation of the genomic relationship matrix (GRM) developed by Yang et al. (2010). The GRM can be estimated between two individuals as
1 푁 (푥푖푗−2푝푖)∗(푥푖푘−2푝푖) 퐴푗푘 = ∑푖=1 (5) 푁 2푝푖(1−푝푖)
푡ℎ where 푗 and 푘 are the individuals, 푥푖푗 is the number of copies of the reference allele for the 푖
푡ℎ SNP of the 푗 individual, and 푝푖 is the frequency of the reference allele. The estimate of a genomic inbreeding coefficient defines the relationship between haplotypes within an
2 individual. The frequencies of the three genotypes of an SNP 푖 were assumed as: [i] (1-푝푖) +
2 푝푖(1-푝푖)*퐹, [ii] 2푝푖(1-푝푖)*(1-퐹), and [iii] 푝푖 + 푝푖(1-푝푖)*퐹, where 2푝푖(1-푝푖) = ℎ푖. The first estimate based on the variance of additive genetic values (diagonal of the GRM) can be written as
2 (푥푖−2푝푖) 퐹퐼 = − 1, (6) ℎ푖
푡ℎ where 푥푖 is the number of copies of the reference allele for the 푖 SNP, which is a specific case of equation (5) for a single SNP when 푗 = 푘. The second estimate of a genomic inbreeding coefficient based on SNP homozygosity following Wright (1948) can be described as
푥푖∗(2−푥푖) [O(#ℎ표푚)−퐸(#ℎ표푚)] 퐹퐼퐼 = 1 − = , (7) ℎ푖 [1−퐸(#ℎ표푚)] where O(#ℎ표푚) and 퐸(#ℎ표푚) are the observed and expected number of homozygous genotypes in the sample. The sampling variance of 퐹퐼 is dependent on the allele frequency.
- 13 -
Chapter One ______
Hence, an estimator based on the correlation of uniting gametes was implemented from Yang et al. (2010). The equation is defined as follows
2 2 [푥푖 −(1+2푝푖)∗푥푖+2푝푖 ] 퐹퐼퐼퐼 = . (8) ℎ푖
The fourth genomic inbreeding coefficient is defined as long stretches of homozygous genotypes, known as runs of homozygosity (ROH), and were computed with the software package PLINK by using the following adjustments of parameters: --homozyg-density 1000, -
-homozyg-window-het 1, --homozyg-kb 10, --homozyg-window-snp 20 (Purcell et al. 2007;
Bosse et al. 2012; Zhang et al. 2015). These settings were chosen to enable comparisons of results when using different inbreeding measurements. Genomic inbreeding coefficient based on ROH is defined as the total length of the genome covered by ROH divided by the overall length of the genome covered by SNP markers as follows (McQuillan et al. 2008):
퐿푅푂퐻 퐹퐼푉 = , (9) 퐿퐴푈푇푂 where 퐿푅푂퐻 summaries all ROH lengths and 퐿퐴푈푇푂 is the total autosomal length covered by reads. In this regard, four genomic inbreeding coefficients of 퐹퐼, 퐹퐼퐼, 퐹퐼퐼퐼, and 퐹퐼푉 are hereafter referred to as FGRM, FHOM, FUNI, and FROH according to the study of Zhang et al. (2015).
Correlation coefficients
Pearson’s correlations between different pedigree-based inbreeding coefficients (FWright,
FVanRaden, FBallou, FKalinowski, FNew, FLacy) and genomic estimates of inbreeding (FGRM, FHOM,
FUNI, FROH) were performed and tested for significant differences within the selection candidates. To do this, the ‘rcorr’ function of R-statistic software package (R Core Team
2018) was used.
- 14 -
Chapter One ______
RESULTS
Relatedness and inbreeding
Relations between selection candidates were identified with the additive genetic relationship matrix (Fig. 2a) based on the pedigree and by analyzing the genomic relationship matrix (Fig.
2b) based on genomic information. Selection candidates revealed different intensities of relatedness with each other, whereby some individuals also showed an enhanced relatedness.
Computed correlation between the additive genetic relationship (Fig. 2a) and the genomic relationship matrix (Fig. 2b) of all selection candidates was high with 0.65 (P < 0.05).
Fig. 2 Genetic relatedness of all selection candidates for a): pedigree-based (additive genetic relationship matrix) and b): genomic-based information (genomic relationship matrix)
Pedigree-based inbreeding of the selection candidates ranged from 0.030 to 0.033 for the classical, from 0.024 to 0.090 for the ancestral, and from 0.004 to 0.006 for the partial inbreeding coefficients. Calculated pedigree-based inbreeding coefficients of the selection candidates were 0.030 (FWright), 0.033 (FVanRaden), 0.024 (FBallou), 0.090 (FKalinowski), 0.004
(FNew), and 0.006 (FLacy). FNew and FKalinowski resulted in the lowest and highest inbreeding.
Estimated standard deviations of all coefficients varied between 0.014 (FLacy) and 0.053
(FKalinowski). The comparisons between the different pedigree-based inbreeding measurements were significant (P < 0.05), except the comparison of the inbreeding coefficients between
- 15 -
Chapter One ______
FWright and FBallou (P ≥ 0.05). Pedigree-based inbreeding coefficients showed a strong variation in different kinds of measurements (Fig. 3a). Investigated comparison of inbreeding coefficients between different sexes of the selection candidates showed no significant differences. Values of FGRM, FHOM, and FUNI ranged close to zero and showed significant difference by comparison with FROH (P < 0.05) with an average value of 0.012. Standard deviations between genomic coefficients varied between 0.110 (FGRM) and 0.059 (FHOM).
Genomic inbreeding coefficients showed an obvious difference between the four genomic estimates (Fig. 3b). There were also no significant differences measured according to average inbreeding between both sexes based on genomic estimators.
Fig. 3 Different coefficients from selection candidates for a): pedigree and b): genomic inbreeding
Comparison of inbreeding coefficients
Correlations between the pedigree-based significant inbreeding coefficients were all positive and ranged from 0.22 to 0.95 with p-values < 0.05, whereas correlations between FKalinowski and FNew as well as between FNew and FLacy were not significant (Table 1).
- 16 -
Chapter One ______
Table 1 Correlation between pedigree-based inbreeding coefficients of selection candidatesa
FVanRaden FBallou FKalinowski FNew FLacy FWright 0.69 (***) 0.95 (***) 0.34 (**) 0.53 (***) 0.68 (***) FVanRaden 0.67 (***) 0.25 (*) 0.64 (***) 0.42 (***) FBallou 0.22 (*) 0.59 (***) 0.42 (***) FKalinowski -0.19 (n.s.) 0.49 (***) FNew 0.13 (n.s.) aEstimates were tested for statistical significance: p-value ≥ 0.05 (n.s.), < 0.05 (*), < 0.01 (**), < 0.001 (***)
Significant correlations (P < 0.05) between genomic inbreeding coefficients varied between
0.41 and 0.95, whereas correlations of FGRM with FHOM and FGRM with FROH were not significant (Table 2).
Table 2 Correlation between genomic-based inbreeding coefficients of selection candidatesa
FHOM FUNI FROH FGRM -0.21 (n.s.) 0.80 (***) 0.15 (n.s.) FHOM 0.41 (***) 0.79 (***) FUNI 0.62 (***) aEstimates were tested for statistical significance: p-value ≥ 0.05 (n.s.), < 0.05 (*), < 0.01 (**), < 0.001 (***)
Correlations between pedigree and genomic inbreeding coefficients are summarised in Table
3. Significant correlations between pedigree and genomic inbreeding coefficients ranged from
-0.44 (FKalinowski and FGRM) to 0.57 (FVanRaden and FROH). There were no significant correlations between FGRM and pedigree-based inbreeding measurements except for FKalinowski and FNew.
FHOM was always significantly correlated with values from 0.22 to 0.57 except with FKalinowski.
This result was in accordance with FROH, whereby FROH was overall higher positive correlated.
FUNI was also mainly significantly correlated with pedigree inbreeding except for FLacy.
Table 3 Correlation between pedigree- and genomic-based inbreeding coefficients of selection candidatesa
FWright FVanRaden FBallou FKalinowski FNew FLacy
FGRM 0.19 (n.s.) -0.02 (n.s.) 0.00 (n.s.) -0.44 (***) 0.25 (*) -0.05 (n.s.)
FHOM 0.40 (***) 0.44 (***) 0.39 (***) 0.19 (n.s.) 0.29 (*) 0.26 (*) FUNI 0.23 (*) 0.45 (***) 0.24 (*) -0.29 (**) 0.41 (***) 0.11 (n.s.) FROH 0.48 (***) 0.57 (***) 0.49 (***) 0.00 (n.s.) 0.48 (***) 0.22 (*) aEstimates were tested for statistical significance: p-value ≥ 0.05 (n.s.), < 0.05 (*), < 0.01 (**), < 0.001 (***)
- 17 -
Chapter One ______
DISCUSSION
Relatedness and inbreeding
The main objective of this study was to identify individual relatedness and different inbreeding coefficients of selection candidates from a local pig breed in order to achieve genetic diversity for the progeny. Classical inbreeding was in accordance with each other’s inbreeding values and gave a first impression of inbreeding levels among the selection candidates (Fig. 3a). Inbreeding levels are not high in general and just a few animals show an increased classical inbreeding coefficient. In contrast, ancestral inbreeding of FKalinowski had the highest inbreeding estimates. This ancestral inbreeding coefficient only comprised the ancestral inbreeding of common ancestors on both sides of the pedigree (Kalinowski et al.
2000). However, the ancestral inbreeding coefficient of FBallou differed from FKalinowski. It can be assumed that historical inbreeding is much higher throughout the history of the population due to individuals having been frequently used as sires and dams. Due to the identification of ancestral inbreeding it is possible to select individuals with simultaneously high classical and ancestral inbreeding coefficients and mate them with unrelated animals (Fig. 2) with simultaneously high classical and ancestral inbreeding estimates in order to achieve purging effects (Ballou 1997; Suwanlee et al. 2007). Templeton and Read (1984) reported that inbreeding depression was lower in the offspring born to selected inbred parents than it was prior to a common selection program. In this way, deleterious alleles and inbreeding depressions may be averted. Contrarily, the partial inbreeding coefficients for FNew and FLacy were found to be very minor. Just a few individuals showed an influence by certain founders.
This indicates balanced mating of founders of the selection candidates and controlled inbreeding management of the applied mating system within the small population. Anyway, animals with increased partial inbreeding should be also strictly merged with unrelated individuals to prevent mating of the same founder. Otherwise, this would result in inbreeding and may cause inbreeding depression. Inbred selection candidates with simultaneously
- 18 -
Chapter One ______pedigree-based and genomic-based inbreeding estimates should be observed with special attention. In general, inbred selection candidates (Fig. 3) should be observed carefully during individual mating between male and female of the Angler saddleback in order to control the rate of inbreeding and inbreeding depression. Subsequently, all efforts should be taken to avoid inbreeding in populations of threatened species (Frankham et al., 2001). Therefore, special focus should be laid on ancestral (Ballou 1997; Swindell and Bouzat 2006) and partial inbreeding in order to may achieve purging effects through non-random mating of individuals.
Frankham et al. (2001) mentioned that natural selection can remove (purge) deleterious alleles from populations and should do this more efficiently under inbreeding than random mating. In addition, effectiveness of purging is an important issue in conservation genetics (Frankham et al., 2001). Nevertheless, the management of inbreeding in the population does not only depend on knowing the inbreeding coefficients of all selection candidates. It also depends on avoiding mating of related individuals (Fig. 2). Both relationship matrices showed a high correlation of 0.65. Currently, no studies regarding correlation values of pedigree-based and genome-based relationship matrices were published. It can be assumed that the correlation value has strong dependency on the individual pedigree data quality. In this study the high correlation value of 0.65 is in accordance with the sound pedigree data quality (Fig. 1).
Anyhow, genomic-based relationship matrix is more reliable due to the independency concerning pedigree traceability or completeness. In general, estimates based on genomic information exhibit more reliability than estimates based on pedigree information (Gorjanc et al., 2015; Zhang et al., 2015). For this purpose, individual mating of male and female should be realised based on preferably minor relatedness among them, especially in case of ancestral and partial inbreeding (Suwanlee et al. 2007). Additionally, special attention is required regarding mating of inbred individuals. Less inbred or non-inbred and unrelated animals should be preferred to maintain genetic diversity and decrease the risk of inbreeding depression. However, inbred animals should be preserved and mated carefully regarding
- 19 -
Chapter One ______relatedness as a gene reserve for the future population. Additionally, to achieve genetic diversity, the implementation of a constant or commercial breeding program with increased selection intensity for this small, local breed by focusing on genetic gain is difficult due to an increased rate of inbreeding. In that regard, purging is less effective with rapid inbreeding than with slow (Frankham et al., 2001).
Comparison of inbreeding coefficients
Pedigree-based measurements of correlations between classical inbreeding estimates were moderate (0.69) due to their related method of measurement. Classical inbreeding coefficients were correlated with ancestral inbreeding at a value between 0.25 and 0.95 where FKalinowski showed a lower correlation than FBallou with classical estimates. In the literature, Hinrichs and
Thaller (2013a) investigated a minor correlation of 0.14 between classical and ancestral inbreeding for a local cattle breed with historic introgression. In contrast, the same authors found a moderate correlation between classical and ancestral inbreeding of 0.58 in the case of a commercial German Fleckvieh population (Hinrichs and Thaller 2013b). A study by Mc
Parland et al. (2009) showed correlations between 0.36 and 0.99 for the relation between classical and ancestral inbreeding within a commercial Irish Holstein-Friesian population. The authors identified a high correlation of 0.99 between classical inbreeding estimates and
FKalinowski, whereas the correlation with FBallou was lower (0.36-0.40). The correlations show a lot of variation between the studies. It seems that population structure and introgression have an impact. Commercial breeds show a high positive correlation between classical inbreeding and FKalinowski, whereas breeds with introgression or local breeds have low correlations between classical inbreeding and FKalinowski. However, these presumptions rested upon very few studies and should not seduce to over-interpretation. Thus, more studies and further research on correlations is needed to validate such statements. Additionally, classical inbreeding estimates and FBallou is more highly correlated within local or introgressed breeds
- 20 -
Chapter One ______from our study (0.67-0.95). The correlations between ancestral inbreeding coefficients FBallou and FKalinowski were low at a value of 0.22. This is in accordance with Mc Parland et al. (2009), where correlations ranged between 0.28 and 0.38. Partial inbreeding estimates of FNew were moderately positive correlated with classical inbreeding (0.53-0.64) among the selection candidates. Mc Parland et al. (2009) found even higher correlations between 0.81 and 0.82 in the Irish Holstein-Friesian population. These moderate to high correlations were due to the strong relationship between FWright and FNew during the computation process (Kalinowski et al.
2000). Nevertheless, for this data-set of a local breed no significant correlations were observed for FKalinowski compared to FNew and FLacy. A deep well-recorded pedigree is an essential requirement for reliability of the inbreeding estimates. The pedigree completeness was high (Fig. 1) but the basis generation was chosen in 1980 and the traces within the pedigree were not comprehensive in the history. Pedigrees account only for inbreeding that has occurred since the beginning of pedigree recording. Thus, there might be an under estimation of inbreeding estimates. The reason for this is that the Angler saddleback is mostly preserved by hobby breeders and the breeding organisation is staged early. The realised IBD proportion of the genome is better predicted by a large number of genetic markers than by pedigrees (Keller et al. 2011; Hoffman et al. 2014; Kardos et al. 2015). Additionally, inbreeding estimates measured by genomic information is more informative than the inbreeding coefficient estimated from pedigree data (Zhang et al. 2015). Genomic inbreeding coefficients can be calculated based on the variance of additive genetic values (variance of diagonal elements of derived genomic relationship matrix), on the basis of SNP homozygosity, and based on ROH (ROH divided by the overall length of the genome covered by SNP markers). According to Zhang et al. (2015), genomic inbreeding is dependent on allele frequency within the base population. In theory, allele frequencies of the unselected base generation are necessary for the interpretation of allele frequency-dependent inbreeding estimates (e.g. FGRM and FHOM). However, these frequencies are often unavailable and most - 21 -
Chapter One ______studies use allele frequencies from the actual population. FGRM and FHOM did not account for base generation allele frequencies. The only stable genomic inbreeding measurement regarding allele frequencies is FUNI (Zhang et al. 2015). The authors found low correlations between FGRM and FHOM when the minor allele frequencies were extremely high with considerable variation. It was assumed that breeds with admixture have higher minor allele frequencies and hence, decreased inbreeding. In this study, FGRM and FHOM were strongly negatively correlated due to an assumed admixture. These results were in accordance with investigated correlations between different genomic inbreeding coefficients as well as the high and various minor allele frequencies within the selection candidates (Fig. 4).
Fig. 4 Minor allele frequency (MAF > 0.05) of the Angler saddleback selection candidates
Thus, we assume the same effects for this local breed due to historic introgression and admixture with other pig breeds (Hamphire, Pietràin, and Meishan). In our case, FHOM and
FUNI were even more highly correlated with each other (0.41) due to the extremely high variation of minor allele frequencies of the Angler saddleback candidates than in the previous study by Zhang et al. (2015) with a range between -0.44 and 0.24. Estimates of FROH showed high positive correlations with FHOM (0.79) and FUNI (0.62). Demonstrated results are in accordance with the results from Zhang et al. (2015) where the correlations with FHOM and
- 22 -
Chapter One ______
FUNI were also positive and ranged from 0.06 to 0.61 (FHOM) and 0.15 to 0.80 (FUNI) depending on different data-sets and distinct breeds. High correlation values of FROH regarding FHOM and FUNI indicates correct adjustments of parameters during the calculation of
ROH. Estimators based on ROH directly reflect homozygosity on the genome and have the advantage of not being affected by estimates of allele frequency or incompleteness of the pedigree (Zhang et al. 2015). Classical pedigree-based inbreeding estimates were moderately positive correlated with genome-based measurements of FHOM and FUNI. This is in accordance with Hinrichs and Thaller (2013b), who investigated correlations in a range between 0.09 and
0.45. Zhang et al. (2015) estimated correlations between the classical inbreeding coefficient
FVanRaden and all genomic inbreeding coefficients at between -0.31 and 0.45 within Danish Red
Cattle with strong historic admixture. This is also in accordance with our results regarding
FHOM (0.44) and FUNI (0.45). Estimates of inbreeding differed according to the approaches for calculation used. For coefficients estimated by methods using allele frequencies, such as the
FGRM, considerable variation was observed within the data-set of the selection candidates due to the high sensitivity regarding allele frequencies. In addition, classical inbreeding coefficients were moderately positive correlated with FROH. This result is also in accordance with Zhang et al. (2015), where investigated correlations ranged between 0.48 and 0.84.
Ancestral inbreeding of FBallou was moderately correlated with FHOM (0.39) and FUNI (0.24). In contrast, ancestral inbreeding of FKalinowski was negatively correlated with FGRM (-0.44) and
FUNI (-0.29). Hence, a high variation of correlations between the different inbreeding measurements of ancestral inbreeding with genomic inbreeding is given (from -044 to 0.39).
This result is not fully in accordance with Hinrichs and Thaller (2013b). Here, the authors identified correlations between ancestral and genomic inbreeding of 0.03 to 0.26. A reason for this may be the data-set for their estimates, which included a commercial German Fleckvieh population. German Fleckvieh is purebred, has no strong admixture, and allele frequencies have probably been harmonised throughout its history. Thus, there is no expectation for wide
- 23 -
Chapter One ______variation between correlations. This study is the first to compare ancestral and partial inbreeding estimates with the genomic inbreeding coefficient based on ROH. The estimates of
FBallou showed the highest positive correlation of 0.49 with FROH according to all other genomic estimates. Hence, it can be assumed that the FROH with its chosen parameter adjustments may describe ancestral inbreeding better than other genomic coefficients. Partial inbreeding ranged from 0.22 to 0.48, where FNew was moderately positive correlated for all genomic inbreeding measurements. However, the highest correlation with a value of 0.48 was investigated for FROH, whereas partial inbreeding of FLacy showed slightly correlations with
FHOM (0.26) and FROH (0.22). Low correlations could be explained by the different computation approaches for genomic and partial inbreeding. Genomic approaches did not focus on certain ancestors or ‘new’ inbreeding.
CONCLUSION
Low inbreeding and minor relatedness are essential especially for small, local breeds to keep their native genetic diversity and adapt an optimum inbreeding management for the progeny.
In this study, the authors identified individual relatedness and different inbreeding coefficients of selection candidates from a small local pig breed in order to achieve genetic diversity within the progeny. In case of small populations, ancestral and partial inbreeding reveals special importance regarding utilisation of purging effects and genetic diversity within mating systems. Additionally, worthwhile comparisons of ancestral, partial, and genomic inbreeding were investigated. Inbreeding estimators based on ROH may represent ancestral and partial inbreeding better than other genomic inbreeding coefficients in case of a local breed. Beyond, it would be interesting, to investigate genomic-based measurements of ancestral and partial inbreeding due to their importance of purging effects regarding conservation of small populations.
- 24 -
Chapter One ______
Acknowledgement: Financial support from the Ministry of Energy, Agriculture,
Environment, Nature, and Digitalization within the framework of the European Innovation
Partnership (EIP Agri) is gratefully acknowledged.
REFERENCES
Ballou JD (1997) Ancestral inbreeding only minimally affects inbreeding depression in
mammalian populations. Journal of Heredity 88:169-178
Boettcher PJ, Tixier-Boichard M, Toro MA, Simianer H, Eding H, Gandini G, Joost S, Garcia
D, Colli L, Ajmone-Marsan P, The GLOBALDIV Consortium (2010) Objectives,
criteria and methods for using molecular genetic data in priority setting for conservation
of animal genetic resources. Animal Genetics 41(1):64-77
Boichard D (2002) Pedig: a fortran package for pedigree analysis suited to large populations.
Comm. 28 - 13 in Proc 7th WCGALP, Montpellier, France. INRA, Montpellier, France
Bosse M, Megens HJ, Madsen O, Paudel Y, Frantz LA, Schook LB, et al. (2012) Regions of
homozygosity in the porcine genome: consequence of demography and the
recombination landscape. PLoS Genetics 8(11):e1003100
Caballero A, Toro MA (2002) Analysis of genetic diversity for the management of conserved
subdivided populations. Conservation Genetics 3:289-299
Curik I, Ferenčaković M, Sölkner J (2017) Genomic dissection of inbreeding depression: a
gate to new opportunities. Revista Brasileira de Zootecnia 46(9):773-782
FAO (2010) Breeding strategies for sustainable management of animal genetic resources.
FAO Animal Production and Health Guidelines. No. 3, Rome
Frankham R (1995) Conservation genetics. Annual Review of Genetics 29:305-327
Frankham R, Gilligan DM, Morris D, Briscoe DA (2001) Inbreeding and extinction: Effects
of purging. Conservation Genetics 2:279-285
- 25 -
Chapter One ______
Gorjanc G, Bijma P, Hickey JM (2015) Reliability of pedigree-based and genomic
evaluations in selected populations. Genetics Selection Evolution 47:65
Haig SM, Ballou JD, Derrickson SR (1990) Management Options for Preserving Genetic
Diversity: Reintroduction of Guam Rails to the Wild. Conservation Biology 4(3):290-
300
Hinrichs D, Thaller G (2013a) Inbreeding and impact of foreign dairy cattle breeds in the
German Angeln dairy cattle population. 64th EAAP Book of Abstracts No. 19:19
Hinrichs D, Thaller G (2013b) Analysis of pedigree and marker based inbreeding coefficients
in German Fleckvieh. 64th EAAP Book of Abstracts No. 19:603
Hoffman JI, Simpson F, David P, Rijks JM, Kuiken T, Thorne MA et al (2014) High-
throughput sequencing reveals inbreeding depression in a natural population.
Proceedings of the National Academy of Sciences USA 111:3775-3780
Kalinowski ST, Hedrick PW, Miller PS (2000) Inbreeding depression in the Speke's gazelle
captive breeding program. Conservation Biology 14:1375-1384
Kardos M, Luikart G, Allendorf FW (2015) Measuring individual inbreeding in the age of
genomics: marer-based measures are better than pedigrees. Journal of Heredity 115:63-
72
Keller MC, Visscher PM, Goddard ME (2011) Quantification of inbreeding due to distant
ancestors and its detection using dense single nucleotide polymorphism data. Genetics
189:237-249
Lacy RC (1989) Analysis of Founder Representation in Pedigrees: Founder Equivalents and
Founder Genome Equivalents. Zoo Biology 8:111-123
Lacy RC (1997) Importance of genetic variation to the viability of mammalian populations.
Journal of Mammalogy 78(2):320-335
Lacy RC, Alaks G, Walsh A (1996) Hierarchical analysis of inbreeding depression in
Peromyscus polionotus. Evolution 50:2187-2200
- 26 -
Chapter One ______
MacCluer JW, Boyce AJ, Dyke B, Weitkamp LR, Pfenning DW, Parsons CJ (1983)
Inbreeding and pedigree structure in Standardbred horses. Journal of Heredity
74(6):394-399
Mc Parland S, Kearney F, Berry DP (2009) Purging of inbreeding depression within the Irish
Holstein-Friesian population. Genetics Selection Evolution 41:16
McQuillan R, Leutenegger AL, Abdel-Rahman R, Franklin CS, Pericic M. Barac-Lauc L, et
al. (2008) Runs of homozygosity in European populations. American Journal of Human
Genetics 83(3):359-372
Meuwissen THE, Luo Z (1992) Computing inbreeding coefficients in large populations.
Genetics Selection Evolution 24(4):305-313
Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MJ, Sham PC (2007) PLINK: Atool
set for whole-genome association and population based linkage analysis. American
Journal of Human Genetics 81:559-575
Quaas RL (1976) Computing the Diagonal Elements and Inverse of a Large Numerator
Relationship Matrix. Biometrics 32(4):949-953
R Core Team (2018): R: A Language and Environment for Statistical Computing. R
Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org
Suwanlee S, Baumung R, Sölkner J, Curik I (2007) Evaluation of ancestral inbreeding
coefficients: Ballou’s formula versus gene dropping. Conservation Genetics 8:489-495
Swindell WR, Bouzat JL (2006) Ancestral inbreeding reduces the magnitude of inbreeding
depression in Drosophila melanogaster. Evolution 60(4):762-767
Templeton AR, Read B (1984) Factors eliminating inbreeding depression in a captive herd of
Speke’s Gazelle (Gazella spekei). Zoo Biology 3:177-199
Tisdell C (2003) Socioeconomic causes of loss of animal genetic diversity: analysis and
assessment. Ecological Economics 45:365-376
- 27 -
Chapter One ______
VanRaden PM (1992) Accounting for inbreeding and crossbreeding in genetic evaluation for
large populations. Journal of Diary Science 75:3136-3144
VanRaden PM (2008) Efficient methods to compute genomic predictions. Journal of Diary
Science 91(11):4414-4423
Wang Y, Bennewitz J, Wellmann R (2017) Novel optimum contribution selection methods
accounting for conflicting objectives in breeding programs for livestock breeds with
historical migration. Genetics Selection Evolution 49:45
Wellmann R, R package version 2.0. (2018) OptiSel: optimum contribution selection and
population genetics. https://cran.r-project.org/web/packages/optiSel/optiSel.pdf
Wright S (1922) Coefficients of inbreeding and relationship. American Naturalist 56:330-338
Wright S (1948) Genetics of populations. Encyclopedia Britannica 10:111-A-D-112
Yang et al (2010) Common SNPs explain a large proportion of the heritability for human
height. Nature Genetics 42(7):565-569
Yang J, Hong Lee S, Goddard ME, Visscher PM (2011) GCTA: A tool for genome wide
complex trait analysis. American Journal of Human Genetics 88:76-82
Zhang Q, Calus MPL, Guldbrandtsen B, Lund MS, Sahana G (2015) Estimation of inbreeding
using pedigree, 50k SNP chip genotypes and full sequence data in three cattle breeds.
BMC Genetics 16:88
- 28 -
Chapter Two ______
Chapter Two
Performance of a novel breeding value in context of breed conservation
J. Schäler1, R. Wellmann2, P. Stratz2, J. Bennewitz2, G. Thaller1 & D. Hinrichs3
1Christian-Albrechts-University of Kiel, Institute of Animal Breeding and Husbandry, Hermann-Rodewald-Straße 6, 24118 Kiel, Germany 2University of Hohenheim, Institute of Animal Science, Emil-Wolff-Straße 10, 70593 Stuttgart, Germany 3Humboldt-University of Berlin, Albrecht Daniel Thaer-Institute, Invalidenstraße 42, 10115 Berlin, Germany
Published in 11th World Congress of Genetics Applied to Livestock Production
- 29 -
Chapter Two ______
ABSTRACT For the maintenance of genetic diversity the improvement of approaches for breed conservation is important. Many breeds have low native genetic contribution (NC) due to introgression with high-yielding breeds. The aim of this work was to examine the performance of a breeding value for NC in local breeds. Hence, the NC was considered as a trait. Information on NC for two breeds was available. Genetic parameters were estimated by applying different linear mixed models. Additionally, correlations with estimated breeding values (EBV) of other traits were analysed. Estimated heritability for NC varies between 0.74 and 0.97. Furthermore, correlations between EBV for NC and reproduction traits were positive, whereas the correlation between EBV for NC and milk yield was negative.
Keywords: optiSel, ASReml, EBV_NC, linear mixed model, genetic conservation
- 30 -
Chapter Two ______
INTRODUCTION
To achieve conservation of local breeds different approaches have been proposed. Forming a synthetic breed by combining a non-endangered breed with one or two highly endangered local breeds as proposed by Bennewitz et al. (2008), may in fact result in continued introgression from the high-yielding breed, which reduces the genetic contribution from endangered local breeds. Optimum Contribution Selection (OCS) as proposed by Meuwissen
(1997) may further threaten the genetic material from the endangered local breeds if the native genetic contribution (NC) and total merit are negatively correlated. The OCS approach was extended to account for historical introgression by Wellmann et al. (2012) and Wang et al. (2017). The aim of this work was the investigation of an approach for conservation by computing estimated breeding values (EBV) for the phenotype NC for endangered local breeds with linear mixed models (LMM). Heritability (h²) and correlations (r) with EBV of other traits were calculated.
MATERIALS AND METHODS
Material
The data consisted of phenotypes and breeding values of 58,999 German Angler, 19,682 Red
Dual-Purpose cattle (RDP), and their pedigrees. Phenotype information included animal ID, sire, dam, sex, farm, breeding values for several traits, and the NC estimated from pedigree.
The NC were computed with function ‘pedBreedComp’ from R-package ‘optiSel’ from
Wellmann (2017a) in R-statistic software (R Core Team 2017). For every individual the genetic contributions from native founders and from other breeds were computed according to the pedigree. It was the fraction of genes that originate from the respective breed. Pedigree data was provided from vit (Vereinigte Informationssysteme Tierhaltung w.V., Verden,
Germany) for the single breeds. The data quality was checked with function ‘completeness’ from R-package ‘optiSel’, and individuals with an equivalent number of complete generations
- 31 -
Chapter Two ______smaller than 3 were removed from the analysis. All founders born after 1970 were considered non-native, whereas founders from the respective breed born before 1970 were considered native.
Statistical model
The linear mixed models A and B were fitted with R-package ‘asreml’ from Butler et al.
(2009). The LMM can be written as
풚 = 푿풃 + 풁푨풂 + ∑풌 풁풌풖풌 + 풆 (1) where y denotes the n-vector of NC, b is the vector of fixed effects, a is the vector of random
2 2 additive genetic effects of the animal distributed as 풂~푁(0, 휎푎 푨), where 휎푎 is the additive genetic variance, and 푨 is the additive relationship matrix. Vector 풖풌 of independent random
2 effects has distribution 풖풌~푁(0, 휎푢푘I), and e is the n-vector of independent residual errors
2 with 풆~푁(0, 휎푒 I). Matrices X, ZA, and Z are design matrices associating observations with the appropriate combination of effects. Fixed effects of birth, sex, sire, dam, and random effects of farm, sire, and dam were tested for significance with R-package ‘asremlPlus’ from
Brien (2016). Date of birth and sex were significant (P ≤ 0.001), thus they were included as fixed effects in all models. However, the random effects of farm, sire, and dam were not significant (P ≥ 0.01). The heritability was calculated as
흈ퟐ 풉² = 풂 . (2) 흈ퟐ +∑ 흈ퟐ + 흈ퟐ 풂 풌 풖풌 풆
- 32 -
Chapter Two ______
RESULTS
In order to avoid overestimations of h² two final model designs (Model A and B) with different random effects were compared for the two breeds. The trait was heritable in both models (P < 0.005).
² Table 1 Heritability (h²) with standard error (SE), additive genetic variance (σa), variance of ² ² herd effect (σu ), and residual variance (σe) for NC Na Model Ab Model Bc ² ² ² ² ² ² h² SE 휎푎 휎푢 휎푒 h² SE 휎푎 휎푢 휎푒
Nα 0.89 0.0093 0.0043 0.0005 <0.0001 0.93 <0.0001 0.0031 - 0.0002 Nβ 0.74 0.0036 0.0037 0.0013 <0.0001 0.97 <0.0001 0.0048 - 0.0001 a Nα = 58,999 German Angler, Nβ = 19,682 Red Dual-Purpose cattle (RDP) bModel A: (fixed effects = born, sex / random effects = farm) cModel B: (fixed effects = born, sex / random effects = none)
The results are given in Table 1. The h² ranged from 0.74 to 0.97 with standard errors between
<0.0001 to 0.0093 for the different sample sizes in all models and breeds. Hence, all models performed similarly, but the likelihood ratio test was not significant, thus the model B without farm effect provides a better fit for the data than the model A with farm effect. Figure 1 shows correlations (P < 0.05) with reliabilities over 90% between the EBV of NC (EBV_NC) and conventional EBVs.
Fig. 1 Correlations between EBV_NC and conventional EBVs for a): German Angler and b): Red Dual-Purpose cattle (RDP)
- 33 -
Chapter Two ______
For German Angler the EBV_NC is positively correlated with TMI (r = 0.53), which is a reproduction trait, and negatively correlated with EBV_M (r = -0.61), which is a milk yield trait, and TMI (r = -0.59), which is the total merit index. For RDP the EBV_NC is slightly positive correlated with EBV_N (r = 0.08), which is a longevity trait, and negatively correlated with EBV_R (r = -0.05) and EBV_M (r = -0.02). Surprisingly, the EBV_NC is also positively correlated with some fitness and conformation traits (EBV_S, EBV_MTY,
EBV_KOE, EBV_Fun, EBV_Eut, and EBV_E), which could be due to different breeding goals in the RDP and in the introgressed breeds.
DISCUSSION
Breeding values for NC were negatively correlated with milk yield and total merit index
(TMI), which shows that truncation selection or traditional OCS on TMI reduces the NC and thus threatens the native genetic background of local breeds. Hence, advanced OCS methods
(Wellmann 2017b) are needed for these breeds. The moderate correlations between EBV_NC and conventional functional traits show the connections between fitness traits and the native genetic background of these breeds. This indicates that the German Angler and the RDP carry valuable alleles that should be maintained. Both, the h² of the trait NC and the variance of the additive effects are high, which enables response to selection for high EBV_NC. However, if the EBV_NC estimated from pedigree is considered as a phenotype for the true NC, the model makes the wrong assumption that the residuals are uncorrelated, which results in an overestimation of the heritability. Moreover, the NC estimated from the pedigree does not account for Mendelian segregation, which would cause the variance of the estimated NC to decrease considerably after few generations. This can be avoided by replacing the EBV_NC estimated from pedigree with a EBV_NC estimated from genotypes after few generations of selection to achieve sustainable breeding progress.
- 34 -
Chapter Two ______
CONCLUSION
Favorable correlations between the EBV_NC and fertility traits were observed. The h² of the
EBV_NC and the variance of their effects are high, which enables response to selection for high EBV_NC. The negative correlation between EBV_NC and total merit shows the need to select for EBV_NC if the native genetic background should be preserved. This should be combined with advanced OCS to recover the genetic background and to maintain the diversity at native alleles. Investigations should be expanded to implement selection for high EBV_NC in breeding programs for endangered local breeds with historic introgression.
REFERENCES
Bennewitz J, Simianer H, Meuwissen THE (2008) Investigations on Merging Breeds in
Genetic Conservation Schemes. Journal of Diary Science 91:2512-2519
Brien C, R (2016) asremlPlus: Augments the Use of ‘ASReml-R’ in Fitting Mixed Models.
https://cran.r-project.org/web/packages/asremlPlus/asremlPlus.pdf
Butler DG, Cullis BR, Gilmour AR, Gogel BJ (2009) ASReml-R reference manual. State of
Queensland, Department of Agriculture, Fisheries and Forestry
Meuwissen THE (1997) Maximizing the response of selection with a predefined rate of
inbreeding. Journal of Animal Science 75:934-940
R Core Team (2017) R: A Language and Environment for Statistical Computing. R
Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org
Wang Y, Bennewitz J, Wellmann R (2017) Novel optimum contribution selection methods
accounting for conflicting objectives in breeding programs for livestock breeds with
historical migration. Genetics Selection Evolution 49(1):45
Wellmann R, Hartwig S, Bennewitz J (2012) Optimum contribution selection for conserved
populations with historic migration. Genetics Selection Evolution 44(1):34
- 35 -
Chapter Two ______
Wellmann R, R package version 0.9.1. (2017a) OptiSel: optimum contribution selection and
population genetics. https://cran.r-project.org/web/packages/optiSel/optiSel.pdf
Wellmann R (2017b). Optimum Contribution Selection and Mate Allocation for Breeding:
The R Package optiSel. Submitted to Genetics Selection Evolution
- 36 -
Chapter Three ______
Chapter Three
Genetic diversity and historic introgression in German Angler and Red Dual Purpose cattle and possibilities to reverse introgression
J. Schäler1, R. Wellmann2, J. Bennewitz2, G. Thaller1 & D. Hinrichs3
1Institute of Animal Breeding and Husbandry, Christian-Albrechts-University, D-24098 Kiel, Germany 2Institute of Animal Science, University of Hohenheim, D-70593 Stuttgart, Germany 3Department of Animal Breeding, University of Kassel, D-37213 Witzenhausen, Germany
Submitted for Publication - 37 -
Chapter Three ______
ABSTRACT
Recovering native uniqueness has major importance for breeds with historic introgression.
The aim of the study was to estimate population genetic parameters for two local red cattle breeds from Northern Germany and to study possibilities to reverse introgression.
Genealogical information was used for the analysis. The Pedigree data consisted of 90,783 individuals for German Angler (GA) and 187,255 individuals for Red Dual-Purpose cattle breed (RDP); with additional information on sex, born, breed, status, and conventional breeding values for 44,942 GA and 14,167 RDP. The estimated native effective population size was 67.3 for GA and 33.9 for RDP. The native contributions were on average 53 % in
GA, and 67 % in RDP, whereby founders with specified native breed names were considered to be purebred, independent from the year in which they were born. Genetic diversity at native alleles declined for both breeds due to lost breeding lines, and native contributions declined due to historic introgression. Two OCS scenarios were compared for the available selection candidates: (1) Minimizing kinship at native alleles and (2) Maximizing genetic gain. The native contributions were included as an additional trait in the total merit index and were estimated from pedigrees. Scenario (1) shows that the average native kinship in the population could be diminished within one year from 0.082 to 0.079 for GA and from 0.168 to 0.165 for
RDP. The results of scenario (2) show that increasing the weight of the native contributions in the TMI increases the native contribution and decreases the achievable genetic gain in the conventional total merit index, but this relationship was not linear. Therefore, different OCS scenarios with different weights for the NC should be taken into account in order to find the optimal weights in a given year. It is concluded that the native genetic contribution (NC) could be included as an additional trait in the total merit index in order to recover a part of the native genetic background. Native contributions should be estimated in the long-term from marker data in order to account for Mendelian sampling, but this is not necessary in the first years if the variance of the pedigree-based estimates is sufficiently large. The maintenance of
- 38 -
Chapter Three ______a sufficient genetic diversity of native alleles can be achieved by advanced optimum contribution selection (OCS) with appropriate constraints.
Keywords: genetic diversity, historic introgression, inbreeding, native genetic contribution, optimum contribution selection, red cattle
- 39 -
Chapter Three ______
INTRODUCTION
Conservation of local breeds is important due to social and cultural aspects. In addition, local breeds are relevant for society and humanity due to its function as reservoir of irreplaceable genetic diversity (Boettcher et al. 2010). During the last century, competitiveness of many local breeds was reduced as a consequence of high genetic progress in a few number of high yielding breeds. Hence, local breeds had a lower value for agricultural production and subsequently, genetic gain and effective population sizes declined for most of these breeds
(Meuwissen 2009; Fernández et al. 2011). In order to enhance genetic gain and increase profitability, breeding organizations upgraded with high-yielding breeds for decades. Due to this breeding history, genetic diversity increased whereas the native genetic contribution (NC) and the genetic diversity of native alleles decreased. Meuwissen (2009) pointed out that the best strategy to conserve local breeds is to make the breed profitable. Breeding programs, except conservation programs, are designed to maximize genetic gain. However, the management of inbreeding is important, and therefore, during the last decade of the last century, selection methods have been developed to manage the rate of inbreeding while maximizing genetic gain (Wray and Goddard 1994; Meuwissen 1997; Grundy et al. 1998).
Meuwissen (1997) and Grundy et al. (1998) restricted the increase of the average relationship to restrict the rate of inbreeding to a predefined value. Nowadays, these selection methods are known as optimum contribution selection (OCS). Sonesson et al. (2000) showed that OCS is also possible in breeding schemes with overlapping generations, e.g. dairy and beef cattle or pig breeding programs, which is important for the implementation of OCS. However, OCS may further decrease genetic uniqueness of local breeds if NC and economically important traits are negatively correlated. Therefore OCS was extended to account for historical introgression by Wellmann et al. (2012). This novel OCS approach enables to recover the genetic uniqueness of local breeds while restricting the rate of inbreeding and maximizing genetic gain. According to Wang et al. (2017a), breeding programs should focus on increasing
- 40 -
Chapter Three ______genetic gain and controlling the inbreeding, as well as maintaining the genetic originality via inclusion of migrant contribution and kinship at native alleles in the OCS procedure for local breeds with historical introgression. To maintain the genetic originality and recover the native genetic background an implementation of NC as an additional trait with appropriate weight in the total merit index may be also successful.
The objective of this study was to identify population genetic parameters for two local red cattle breeds from Northern Germany. Furthermore, a pedigree-based estimate for NC was incorporated in the total merit index during the OCS in order to recover the native genetic background. In addition, correlations of selection candidates’ optimum contributions between different OCS scenarios were calculated.
MATERIALS AND METHODS
Animal care and use committee approval was not obtained for this study because data were obtained from the existing data-sets described below.
Materials
All pedigree and additional information about sex, date of birth, and breed for German Angler
(GA) and Red Dual-Purpose cattle breed (RDP) were provided by the vit (Vereinigte
Informationssysteme Tierhaltung w.V., Verden, Germany), which is the computation center for the official breeding value estimation in Germany. For GA and RDP the pedigree file included information of 90,783 (10,115 male and 80,668 female) and 178,255 (15,935 male and 162,320 female) animals born between 1906 and 2017, respectively. This breed-specific pedigree file comprised for GA 70,136 (4,227 male and 65,909 female) animals (77.26 %) classified as purebred GA and for RDP 105,951 (3,783 male and 102,168 female) animals
(59.44 %) classified as purebred RDP. All other non-purebred breeds included in the GA pedigree file were mainly Red Holstein (RH; 12,136), Holstein Frisian (HF; 2,466), Rotes
Hoehenvieh (RHV; 821), Fleckvieh (FV; 709), RDP (573), Braunvieh (BV; 530), Jersey
- 41 -
Chapter Three ______
(JER; 489), and German black pied cattle (GBP; 239) and in the RDP pedigree file RH
(41,037), HF (22,048), FV (1,734), GA (1,802), GBP (600), and JER (490) in descending order. Breed-specific, total merit indices (TMI) for 52,989 (1,754 male and 51,235 female)
GA and 59,664 (1,591 male and 58,073 female) RDP were provided by the vit. Information about the status (alive or dead) of the purebred animals of GA and RDP was also provided by the vit. The base of the status was the breeding value estimation in August 2017, i.e. all living
GA and RDP animals in August 2017 were coded with one, whereas all other animals were coded with zero. For GA 23,261 (86 male and 23,175 female) animals and for RDP 43,148
(117 male and 43,031 female) animals were coded with one as alive.
Population genetic parameters
Pedigree completeness and population parameters were computed with functions
‘completeness’ and ‘summary’ in R-statistic software (R Core Team 2018) with R-package
‘optiSel’ from Wellmann et al. (2017b). The classical kinship (fA), kinship at native alleles
(fD), native effective population size (native Ne), native genome equivalents (native GE), and native genetic diversity (native GD) were computed.
The classical kinship 푓퐴(푖, 푗) between two individuals 푖 and 푗, which is also called genealogical coancestry, is the probability that two alleles chosen from both individuals from the same locus, are identical by descent (IBD). The kinship matrix 풇푨 is half the additive relationship matrix.
The native kinship 푓퐷(푖, 푗) between two individuals is defined as the conditional probability that two randomly chosen alleles are IBD, given that both originate from native founders. A native founder is an individual with unknown parents, which belongs to the native breed and was born before some base time 푡1. In this study, we used 푡1 = 1970 for GA and 푡1 = 1985 for RDP. That is, all animals born before this year are considered native, whereas founders
- 42 -
Chapter Three ______born thereafter were considered to be from another breed or from an unknown breed. The native kinship was computed as
푓퐼퐵퐷$푁(푖,푗) 푓퐷(푖, 푗) = , 푓푁(푖,푗) where matrix 풇푰푩푫&푵 contains for each pair of individuals the probability that two alleles randomly chosen from both individuals are IBD and native, whereas matrix 풇푵 contains the probabilities that they are native. The algorithm for estimating these probabilities from pedigrees can be found in Wellmann et al. (2012). These matrices are used to compute the average native kinship fD(푡) in the population at time 푡 with functions summary() and candes() from package optiSel.
The native gene diversity (native GD) of the birth year 푡 is the probability that two alleles chosen at random from the birth year are not IBD, given that both originate from native founders.
Since it can be computed as native GD(t) = 1 − fD(푡), this parameter provides the same information as parameter fD(푡).
Wellmann et al. (2012) defined native genome equivalents of the population (native GE) at time 푡 as the minimum number of unrelated individuals that would have been needed at time 푡 to establish a hypothetical new population that has the same genetic diversity at native alleles as the population under study. In this case, individuals born in year 푡0 = 1800 are assumed to be unrelated. Consequently, the individuals in the population at time 푡1 = 1970 are related, because the population had a limited historic effective size histNe between 1800 and 1970.
We assumed a historic effective size histNe of 150 in accordance with Villa-Angulo et al.
(2009). The native genome equivalents were computed as
1 native GE푡0(t) = , 2(1−condGD푡0(t))
where condGD푡0(t) is the native genetic diversity in year 푡 with respect to base year 푡0, computed as
- 43 -
Chapter Three ______
t1−푡0 1 1−fD(푡) I condGD푡0(t) = (1 − ) ∗ , 2 histNe 1−fD(푡1) for 푡 ≥ 푡0, where I is the generation interval. The left term accounts for the decrease of genetic diversity between times 푡0 and 푡1, whereas the second term accounts for the decrease of the native gene diversity between times 푡1 and 푡.
The native effective size (native Ne) is the size of an idealized random mating population for which the genetic diversity decreases as fast, as the native genetic diversity decreases in the true population. It is defined with respect to a time interval during which the decrease of the native genetic diversity is observed. One possibility to estimate this value is by computing the rate of increase in the mean native kinship during the last (say 6) generations. According to
Wellmann and Bennewitz (2011), another possibility is to estimate the native genetic diversity at each time point time 푡 as follows. The rate of increase in the mean native kinship is approximated by a smooth function, which enables to estimate the native effective size for arbitrarily short time intervals. The native effective size at time t is then defined as the limit that is obtained, when the length of the time interval goes to zero. That is,
native Ne(t) = limϵ→0nat. Ne([t − ϵ, t + ϵ]) for 휖 > 0. This increases the temporal resolution and can therefore reveal historic changes in the breed management. In this paper, the native effective size is estimated with both approaches.
Historic introgression was identified with the function pedBreedComp() in R-package
‘optiSel’ for both breeds. This function computes for every individual the genetic contribution from native founders and from other breeds as the fraction of genes that originate from the respective breed. Thus, native genetic contribution for each animal was calculated and expressed as the proportion of native breed alleles based on pedigree information. For this evaluation, we assumed in contrast to Wellmann et al. (2012), that all founders with specified
- 44 -
Chapter Three ______native breed names were purebred, independent from the year in which they have been born.
Additionally, cattle from other breeds were considered as non-native.
Total merit indices
The native contribution of an individual is the proportion of its genome originating from native founders. The pedigree-based estimate 푁퐶푖 for individual 푖 is the average native contribution of its parents. That is,
푁푃퐸퐷(푖) = 0.5(푁푃퐸퐷(푠푖) + 푁푃퐸퐷(푑푖)). where 푠푖 is the sire and 푑푖 is the dam of individual 푖. This estimate is not able to capture the
Mendelian sampling. The native contribution estimated from pedigrees is not a conventional trait because it has no Mendelian sampling variance. However, it can be used as a proxy for the native contribution 푁푆퐸퐺(푖) estimated from genotypes as done in this paper. The native contribution estimated from genotypes would be a trait with heritability close to one.
For each breed, the German total merit index (TMI) has mean 100 and a genetic standard deviation of 12. Since the pedigree-based native contribution is used as a proxy for a trait with heritability close to one, it was also standardized to obtain a relative native contribution
(RNC) with mean 100 and standard deviation 12. Both values were combined into a new breeding value for total merit as
휆 푇푀퐼푖 = 휆푅푁퐶푖 + (1 − 휆)푇푀퐼푖.
The larger the value of 휆 is (0 ≤ 휆 ≤ 1), the more emphasis is put on recovering the native genetic background of the breed.
Optimum contribution selection
Different optimum contribution selection scenarios are compared, which differ in their objective functions and constraints. In all scenarios, selection candidates were chosen among animals that were classified as alive purebred, have a conventional breeding value for total merit, and an equivalent number of complete generations ≥ 3.0. This resulted for GA in 58 - 45 -
Chapter Three ______male and 5,300 female selection candidates, while 37,099 animals were involved in the pedigree that included all selection candidates and their ancestors. For RDP, 28 male and
3,481 selection candidates were available and 23,903 animals were involved in the pedigree.
All selection candidates were born between 2000 and 2017 for both breeds.
The vector 풄 with optimum genetic contributions of selection candidates was computed based on pedigree information with R package ‘optiSel’ for a population with overlapping generations. This function optimizes the expected average value of a genetic parameter in the population in the next year, and restricts the expected average values of other parameters, depending on the scenario. The formulas for the objective functions and constraints can be found in Wellmann (2018). Contributions of age classes to the population were estimated with function agecont() such that the contribution of each class is proportional to the expected proportion of its offspring which is not yet born. All scenarios assumed that the total contributions of males and females are equal. That is,
c´s = 0.5, and c´d = 0.5, where vector s and d indicate the sex of candidates. Since the genetic contribution 푐푖 of an individual 푖 is the fraction of genes in the birth cohort that should originate from individual 푖, this value cannot be negative (푐푖 ≥ 0). All females within the same age cohort were assumed to have equal contributions, so the optimization was done only for the males.
The following scenarios were considered:
Scenario 1: In this scenario, the average native kinship in the population was minimized, while the average native contribution and the average total merit index were constrained not to decrease.
휆 Scenario 2(휆): In this scenario, the average total merit index 푇푀퐼푖 in the population was maximized, while the average native kinship was constrained to increase in accordance with
- 46 -
Chapter Three ______the desired effective size 푁푒 = 100 (as recommended by Meuwissen 2009). That is, the upper bound for the average native kinship in the population in the next year was
1 푢푏. 푓퐷 = 퐾̅ + (1 − 퐾̅) , 2푁푒퐿 where 퐾̅ is the average native kinship in the population in the current year, and 퐿 is the generation interval. The conventional kinship was not constrained because Wang et al.
(2017a) noted that constraining the increase in the native kinship automatically also constraints the increase in the conventional kinships because both kinships are correlated.
Constraining the native kinship instead of the conventional kinship also has the advantage that individuals with incomplete pedigrees are not favored for breeding. Solver ‘cccp’ from Pfaff
(2014) was called from the R package ‘optiSel’ and used to solve this optimization problem.
To identify the effects of incorporating the NC in the total merit index, this scenario was evaluated for different values of 휆: 0 (conventional TMI), 0.25, 0.5, 0.75, and 1. Note that
Scenario 2 with 휆 = 0 ignores completely the native contributions, whereas Scenario 2 with
휆 = 1 maximizes the native contributions but ignores completely the conventional total merit.
Optimized candidates’ contributions were compared between all scenarios with the function cor.test() from the R package ‘stats’ (R Core Team 2018).
RESULTS
Population genetic parameters
The pedigree completeness for GA was above 75% for both sexes, when 5 ancestral generations are taken into account (Figure 1). For RDP, the pedigree completeness depended on the sex. While males had a pedigree completeness of about 60%, the completeness in females was only 25%.
- 47 -
Chapter Three ______
Fig. 1 Completeness of pedigree for: a) German Angler (GA) and b) Red Dual-Purpose cattle breed (RDP)
Figure 2 shows the development of population parameters of fA, fD, native GD, native GE, and native Ne between 1970 and 2017 for GA (Figure 2a) and RDP (Figure 2b) based on the restrictive pedigree files without preselection. Both breeds exhibited an increased fA and fD since 1970, whereby the fA for GA (0.022) was higher compared to RDP (0.009). The opposite was observed for fD and the estimated final values were 0.128 and 0.083 for RDP and GA, respectively. The native GD decreased within both breeds. In the current population native GD was 0.917 for GA and 0.872 for RDP. The estimates of native GE for the actual population were 2.8 and 2.5 for GA and RDP, respectively. The native Ne decreased over the last six generations from 99.5 (1975) to 68.4 (2011) for GA and from 90.8 (1975) to 34.0
(2011) for RDP. Thus, estimated values for native Ne in 2017 were 67.3 for GA and 33.9 for
RDP.
- 48 -
Chapter Three ______
Fig. 2 Pedigree-based population parameters of kinship (fA), kinship at native alleles (fD), native genome equivalents (native GE), and diversity at native alleles (native GD) for: a) German Angler (GA) and b) Red Dual-Purpose cattle breed (RDP)
In Figure 3 breed compositions were shown from 1970 till 2017 for each target and respective introgressed breeds based on the respective pedigree files without preselection. In this figure, contributions with specified native breed names were purebred, independent from the year in which they have been born and thus, cattle from other breeds were considered as non-native.
Between 1972 and 1990 introgression in GA was moderate. Most foreign genetic contributions came from RHV and RH. After 1990 the contributions from RH increased considerably. Other breeds, i. e. FV, JER, and GBP, influenced GA not much. The current population of GA has 53 % native genetic contribution, whereby 40 % of the contribution comes originally from RH and HF, 2 % from RHV and 5 % from other breeds (e.g. FV, JER, and GBP). For RDP, the current population consist 67 % native genetic contribution, migrant contribution of 20 % from RH and 11 % from HF, and 2 % from other migrant breeds like FV and GBP. The RDP population was mostly influenced by RH since 1970.
- 49 -
Chapter Three ______
Fig. 3 Historic introgression for: a) German Angler (GA) and b) Red Dual-Purpose cattle breed (RDP). Bovine breeds analysed: RH (Red Holstein) and HF (Holstein Friesian), FV (Fleckvieh), and RHV (Rotes Hoehenvieh)
Optimum contribution selection
Average population genetic parameters of TMI, NC, fA, and fD were 101.828, 0.256, 0.029, and 0.082 in the current population of GA (Table 1). For GA, scenario 1 reduces the average native kinship in the population within one year from 0.082 to 0.079 while the average native contribution increased slightly from 0.256 to 0.265 and the average total merit index remained constant (Table 1).
Table 1 Genetic parameters for German Angler (GA) in the current population and after one year of selection with different OCS scenarios Objective TMIa NCb f c f d Constraint A D function
101.828 0.256 0.029 0.082 currente -
101.828 0.265 0.028 0.079 Min.fD TMI, NC
102.995 0.254 0.029 0.083 Max.TMI0 fD
102.874 0.258 0.030 0.083 Max.TMI0.25 fD
101.282 0.283 0.029 0.083 Max.TMI0.5 fD
101.171 0.283 0.029 0.083 Max.TMI0.75 fD
101.126 0.283 0.029 0.083 Max.TMI1 fD aconventional total merit, bnative contribution, cconventional kinship, dnative kinship, ecurrent population
- 50 -
Chapter Three ______
Scenario 2 restricted the native kinship 푓퐷 as desired and automatically also restricted the conventional kinship 푓퐴. The conventional total merit and the NC increased to different extent, depending on the weight 휆 given to the RNC (Figure 4). For Scenario 2 with 휆 ≥ 0.5 the TMI even decreased in GA because much weight was given to the NC, while the NC increased for 휆 ≥ 0.25. A high value for 휆 in the TMI negatively affects the conventional
TMI in case of the GA (Figure 4a) because the NC and the conventional TMI are negatively correlated with a value of -0.33 (***).
Table 2 Genetic parameters for Red-Dual-Purpose cattle (RDP) in the current population and after one year of selection with different OCS scenarios Object TMIa NCb f c f d Constraint A D function
100.604 0.145 0.041 0.168 currente -
100.797 0.145 0.040 0.165 Min.fD TMI, NC
101.449 0.142 0.040 0.166 Max.TMI0 fD
101.119 0.146 0.041 0.168 Max.TMI0.25 fD
100.870 0.147 0.041 0.168 Max.TMI0.5 fD
100.740 0.147 0.041 0.168 Max.TMI0.75 fD
100.698 0.147 0.042 0.168 Max.TMI1 fD aconventional total merit, bnative contribution, cconventional kinship, dnative kinship, ecurrent population
In the current population of RDP the estimated population genetic parameters were 100.604
(conventional TMI), 0.145 (NC), 0.041 (fA), and 0.168 (fD), as shown in Table 2. For RDP scenario 1 could reduce 푓퐷 (0.165) while TMI and NC were nearly unaltered. In scenario 2 the native kinship 푓퐷 was restricted to the desired values. The TMI increased for all values of
휆 while the NC increased for 휆 ≥ 0.25. As expected, the genetic gain in the conventional TMI was low when much weight was given to the NC. Figure 4b shows again that a large weight for the NC in the TMI negatively affects the genetic gain in the conventional TMI because of a negative correlation between the NC and the conventional TMI of -0.12 (***).
- 51 -
Chapter Three ______
Pearson’s correlations between optimized contributions (OC) for all available male candidates are shown in Table 3 and Table 4 for GA and RDP, respectively. Correlations of OC between different scenarios with 휆 ≥ 0.25 were positive in both breeds, while the correlations of OC between scenario 휆 = 0 and scenario 휆 = 1 were not significantly different from 0 in both breeds. The OC from scenario 1 were positively correlated with scenario 2 when 휆 ≥ 0.5.
This suggests that selection candidates with high native contributions carry rare native alleles.
Table 3 Pearson’s correlation between estimated optimum contributions of all available male selection candidates within different OCS scenarios for the German Angler (GA)a
Max.TMI0.25 Max.TMI0.5 Max.TMI0.75 Max.TMI1 Min.fD
Max.TMI0 0.96 (***) 0.19 (n.s.) 0.15 (n.s.) 0.14 (n.s.) 0.30 (*)
Max.TMI0.25 0.30 (*) 0.29 (*) 0.27 (*) 0.25 (n.s.)
Max.TMI0.5 0.98 (***) 0.95 (***) 0.72 (***)
Max.TMI0.75 0.99 (***) 0.67 (***)
Max.TMI1 0.64 (***) aEstimates were tested for statistical significance: p-value ≥ 0.05 (n.s.), < 0.05 (*), < 0.01 (**), < 0.001 (***)
Table 4 Pearson’s correlation between estimated optimum contributions of all available male selection candidates within different OCS scenarios for the Red Dual-Purpose cattle (RDP)a
Max.TMI0.25 Max.TMI0.50 Max.TMI0.75 Max.TMI1 Min.fD
Max.TMI0 0.49 (**) 0.07 (n.s.) -0.09 (n.s.) -0.12 (n.s.) 0.12 (n.s.)
Max.TMI0.25 0.75 (***) 0.56 (**) 0.48 (**) 0.56 (**)
Max.TMI0.50 0.95 (***) 0.91 (***) 0.85 (***)
Max.TMI0.75 0.99 (***) 0.91 (***)
Max.TMI1 0.91 (***) aEstimates were tested for statistical significance: p-value ≥ 0.05 (n.s.), < 0.05 (*), < 0.01 (**), < 0.001 (***)
- 52 -
Chapter Three ______
Fig. 4 History of the total merit index (TMI) and native contribution (NC) during the optimum contribution selection in scenario 2 for: a) German Angler (GA) and b) Red Dual- Purpose cattle breed (RDP)
DISCUSSION
Population genetic parameters
Aim of this work was to analyze breed-specific population genetic parameters for two local red cattle breeds from Northern Germany in order to get insights into the current population status regarding native kinship, native genetic diversity, native effective population size, and historic introgression. The fA was not high in both breeds due to the strong introgression with other breeds. The fA for GA (0.022) is nearly fully compliant with the fA estimated in the study of Wang et al. (2017a) of 0.020. The fA of GA was slightly higher than for RDP, which is also in accordance with the results of the classical inbreeding coefficients from Addo et al.
(2017). For GA and RDG, the fD increased as expected for a population with a native effective size between 50 and 100. Probably, the native effective size was lower than 100 because animals from foreign breeds were preferably mated to a small group of elite animals, which carried only a part of the native genetic diversity. Hence, the increase of fD for both breeds could have been caused by an increased selection intensity in order to enhance genetic gain. An increase of the inbreeding coefficients was avoided by the introduction of foreign genetic material. The native genome equivalent declined fast due to economic superiority of migrant breeds and the resulting loss of genetic diversity of native alleles. The native Ne for
GA was 67.3 and below the estimate of 86 from Wang et al. (2017a). Wang et al. (2017a) - 53 -
Chapter Three ______used samples from old GA cohorts with sires that had progeny born between 2005 and 2006, whereas we used the actual breeding population born between 2001 and 2017 without preselection in this study. The final value for the native Ne of RDP is 33.9, so it is under the recommended threshold of 50 for the classical Ne (Meuwissen 2009). A small native Ne over a long period of time would make it impossible to remove the foreign genetic material completely because this would result in high inbreeding coefficients. Consequently, RDP may be endangered or strongly threatened by extinction regarding native uniqueness. Note that the classical Ne is higher than the native Ne. According to Addo et al. (2017) the classical Ne is between 54 and 170 animals for RDP and between 50 and 156 for GA. In populations without migration events classical Ne and native Ne are equal, but in populations with steady gene flow from other populations, native Ne is smaller than the classical Ne because the gene flow causes the genetic diversity not to decrease (Wellmann et al. 2012).
Historic introgressions for both breeds were high (Figure 3) for the whole population per cohort without preselection, whereby the current GA population exhibits a minor native genetic uniqueness (53 %) than the RDP (67 %). Notice, that founders with specified native breed names were considered to be purebred, independent from the year in which they were born, during the estimation of NC. Wang et al. (2017a) observed a NC of 30.5 % for GA. This estimate of NC is lower because founders born after 1970 were considered to be non-native.
In addition, Wang et al. (2017a) used an older GA pedigree file, where animals were born between 1906 and 2015, whereas this study used an updated and revised GA pedigree file.
The considerable introgression with RH was due to farmers’ desires to enhance genetic gain.
This result is also in accordance with the study from Addo et al. (2017), where RH was investigated as major introgressor for both breeds. Local breeds began being unpopular because of the economically superior HF population. In order to combat the upcoming gap in genetic gain between local and modern breeds, HF sires were used as key founders. In addition, red colored and dual-purpose breeds with similar phenotypic characteristics, like the
- 54 -
Chapter Three ______
RHV, GBP, and FV, were used for upgrading. Strong historic introgression for GA, especially from RH individuals, was in accordance with genome-based results (data not shown) and with the studies of Wang et al. (2017a and 2017b).
Optimum contribution selection
An additional objective was to investigate the incorporation of NC as a trait in the total merit index during the OCS in order to recover a part of the native genetic background. Notice, NC differed between historic estimates (Figure 3) and current population estimates for OCS
(Table 3 and 4). Average NC for the OCS selection candidates were calculated such as in the studies from Wellmann et al. (2012) and Wang et al. (2017a), where founders born after
푡1 = 1970 for GA and 푡1 = 1985 for RDP were considered to be non-native or from an unknown breed. Hence, a critical NC (historic introgression) has been replaced by a qualified
NC for OCS in order to select with high reliability. Moreover correlations between candidates’ optimum contributions were compared among the OCS scenarios. Scenario 1 minimized the average native kinship but did not increase genetic merit in the population. In addition, the NC increased slightly for GA but was at a constant level for RDP. Scenario 1 could be suitable only in the first generation of selection for recovery of the native genetic background in RDP in order to increase the low genetic diversity at native alleles. Thereafter, it should be replaced by Scenario 2 with an appropriate value for weight 휆 given to the native contribution.
Meuwissen (2009) mentioned that breeds’ profitability has superior relevance regarding conservation of local breeds, which suggests to choose a moderately small value for 휆. On the other hand, the native contribution is already so low in many breeds that the breeds could be considered genetically extinct in short time if the weight given to the recovery of the native background is not sufficiently high. Native uniqueness is a functional reservoir of irreplaceable genetic diversity (Boettcher et al. 2010). TMI0.25 of scenario 2 would be suitable
- 55 -
Chapter Three ______for GA and RDP because the genetic gain increased, average native kinship was restricted
(not threatened), and NC slightly increased.
It was shown that the conventional TMI could even decrease when the weight given to the NC is high. An explanation could be that ancestors of animals with high conventional TMI originate from migrant breeds, which had a stronger selection intensity than local breeds over decades (RH or HF). In addition, results from the OC correlations suggest that animals with high NC have a simultaneously low average native kinship with the population. An explanation could be that animals with high NC belong subpopulations that had no large impact on the total population because of their lower breeding values.
CONCLUSION
Both cattle breeds, the German Angler and the Red Dual-Purpose cattle, showed a decline in the native genetic diversity due to lost breeding lines. Genetic uniqueness has also been negatively affected due to strong historic introgression from migrant breeds. For both breeds it can be recommended to include the native contribution as an additional trait in the total merit index and to set up an OCS program in accordance with scenario 2. Since the native genetic diversity of RDP is low, this can be preceded by one generation of selection according to scenario 1 in RDP in order to increase the native genetic diversity. An alternative to scenario 2 is to restrict the average NC by including an additional constraint in the OCS procedure. However, scenario 2 has the advantage that farmers are already familiar with the use of breeding values, but they are less familiar with OCS. After few years, the pedigree- based estimates for NC need to be replaced by marker-based estimates in order to account for
Mendelian sampling.
Acknowledgement: Financial support from the Ministry of Energy, Agriculture,
Environment, Nature, and Digitalization within the framework of the European Innovation
Partnership (EIP Agri) is gratefully acknowledged.
- 56 -
Chapter Three ______
REFERENCES
Addo S, Schäler J, Hinrichs D, Thaller G (2017) Genetic Diversity and Ancestral History of
the German Angler and Red-and-White Dual-Purpose Cattle Breeds Assessed through
Pedigree Analysis. Agricultural Science 8:1033-1047
Boettcher PJ, Tixier-Boichard M, Toro MA, Simianer H, Eding H, Gandini G, Joost S, Garcia
D, Colli L, Ajmone-Marsan P, The GLOBALDIV Consortium. 2010. Objectives,
criteria and methods for using molecular genetic data in priority setting for conservation
of animal genetic resources. Journal of Animal Genetics 41(1):64-77
Fernández J, Meuwissen THE, Toro MA, Mäki-Tanila A (2011) Management of genetic
diversity in small farm animal populations. Animal 5(11):1684-1698
Grundy B, Villanueva B, Woolliams JA (1998) Dynamic selection procedures for constrained
inbreeding and their consequences for pedigree development. Genetics Research
72:159-168
Meuwissen THE (1997) Maximizing the response of selection with a predefined rate of
inbreeding. Journal of Animal Science 75:934-940
Meuwissen THE, Hayes BJ, Goddard ME (2001) Prediction of total genetic value using
genome-wide dense marker maps. Genetics 157(4):1819-1829
Meuwissen T (2009) Genetic management of small populations: A review. Acta Agriculturae
Scandinavica, Section A - Animal Science 59(2):71-79
Pfaff B (2014) The R package cccp: design for solving cone constrained convex programs. R
Finance. Available at: http://www.pfaffikus.de/files/conf/rif/rif2014.pdf
R Core Team (2018) R: A Language and Environment for Statistical Computing. R
Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org
Sonesson AK, Grundy B, Woolliams JA, Meuwissen THE (2000) Selection with control of
inbreeding in populations with overlapping generations: a comparison of methods.
Animal Science 70:1-8
- 57 -
Chapter Three ______
Villa-Angulo R, Matukumalli LK, Gill CA, Choi J, Van Tassel CP, Grefenstette JJ (2009)
High-resolution haplotype block structure in the cattle genome. BMC Genetics 10:19
Wang Y, Bennewitz J, Wellmann R (2017a) Novel optimum contribution selection methods
accounting for conflicting objectives in breeding programs for livestock breeds with
historical migration. Genetics Selection Evolution 49(1):45
Wang Y, Segelke D, Emmerling R, Bennewitz J, Wellmann R (2017b) Long-Term Impact of
Optimum Contribution Selection Strategies on Local Livestock Breeds with Historical
Introgression Using the Example of German Angler Cattle. G3: Genes, Genomes,
Genetics 7(12):4009-4018
Wellmann R, Bennewitz J (2011) Identification and characterization of hierarchical structures
in dog breeding schemes, a novel method applied to the Norfolk Terrier. Journal of
Animal Science 89:3846-3858
Wellmann R, Hartwig S, Bennewitz J (2012) Optimum contribution selection for conserved
populations with historic migration. Genetics Selection Evolution 44(1):34
Wellmann, R. (2017a) Optimum Contribution Selection and Mate Allocation for Breeding:
The R Package optiSel. Submitted to Genetics Selection Evolution
Wellmann, R., R package version 0.9.1. (2017b) OptiSel: optimum contribution selection and
population genetics. https://cran.r-project.org/web/packages/optiSel/optiSel.pdf
Wray NR, Goddard ME (1994) Increasing long term response to selection. Genetics Selection
Evolution 26: 431-459
- 58 -
Chapter Four ______
Chapter Four
The benefit of native genetic contribution in a local cattle breed
J. Schäler & G. Thaller
Institute of Animal Breeding and Husbandry, Christian-Albrechts-University of Kiel, D-24098 Kiel, Germany
Manuscript in Preparation
- 59 -
Chapter Four ______
ABSTRACT
During last decades, native genetic contribution (NC) decreased in local species due to the introgression of high-yielding breeds. The aim of this study was to look at possible benefits of
NC in local cattle breed. Data contained a pedigree file of 178,255 Red Dual-Purpose cattle
(RDP) and genotypes of 809 RDP individuals with 3,581 genotypes from reference cattle breeds, like Red Holstein, Holstein Friesian, German Angler, German black pied cattle, Rotes
Hoehenvieh, and Fleckvieh, genotyped with the Illumina BovineSNP50 BeadChip (50K) version 1 and 3. Estimated traits for NC were correlated with breed-specific conventional traits of milk yield, longevity, exterior, somatic cells, fertility, and calving abilities. The study revealed that NC is positively related with the traits for longevity, exterior, and somatic cells which belong to the fitness, conformation, and health complex. Selection on NC could probably lead to an increased fitness and health of the RDP.
Keywords: exterior, local cattle, longevity, native genetic contribution
- 60 -
Chapter Four ______
INTRODUCTION
Conservation of breeds has a major importance for society and humanity as it is a reservoir of genes and genetic diversity (Boettcher et al. 2010). The domestication of livestock species has a long history and created an enormous variety of breeds due to migration, selection, and adaptation (Groeneveld et al. 2010). According to Boettcher et al. (2010), a wide agreement regarding conservation of livestock breeds and their genetic diversity due to the necessity for genetic issues exists. Native genetic diversity and native uniqueness of breeds harbor plenty of traits which will be irreplaceable for adaptation to changing environmental conditions.
Wang et al. (2017) hypothesised that the diversity at native alleles is also important for conservation. During the last decades, native genetic contribution (NC) and native genetic diversity decreased in local breeds due to the introgression of high-yielding breeds. In order to maintain this native diversity different approaches have been proposed. Bennewitz et al.
(2008) formed a synthetic breed by combining a non-endangered with one or two highly endangered local breeds. This could approach results in a continued or increased introgression from high-yielding breeds, which further reduces the native genetic contribution of the target breeds. Another approach from Wellmann et al. (2012) extended the traditional Optimum
Contribution Selection (OCS) as proposed by Meuwissen (1997) to account for historical introgression and migrant contributions (MC), respectively. This OCS approach enables to simultaneously recover the genetic uniqueness of breeds while restricting the rate of inbreeding and maximizing genetic gain. In addition, the extended OCS approach may show dependencies between selected parameters during the optimization process; especially if economically important traits or the total merit index are negatively correlated with NC.
Consequently, a powerful trait for NC is needed to may investigate breed-specific advantages and disadvantages of an implemented NC selection.
The objective of this study was to estimate a pedigree- and a genome-based trait for native genetic contribution (NC) for Red Dual-Purpose cattle (RDP) candidates. To proof the benefit
- 61 -
Chapter Four ______of NC, correlations between the trait for NC and breed-specific conventional traits were investigated. Representative status of chosen RDP candidates was tested by comparison of population genetic parameters between candidates and the restrictive RDP population based on genealogical and genomic data.
MATERIALS AND METHODS
Animals
In total genealogical and genomic information of 809 Red Dual-Purpose cattle (RDP) individuals (65 male and 744 female), born between 1994 and 2016, were available (Figure
1). The pedigree file of 178,255 RDP and genotypes were provided by vit (Vereinigte
Informationssysteme Tierhaltung w.V., Verden, Germany), which is the official computation center for the annual breeding value estimation in Germany. Genome-based datasets comprised genotypes from 4,390 animals regarding seven cattle breeds, i. e. 809 RDP, 21
German Angler (GA), 257 Red Holstein (RH), 2786 Holstein Friesian (HF), 477 Fleckvieh
(FV), 32 German Black Pied cattle (GBP), and 8 Rotes Hoehenvieh (RHV), provided by vit.
The animals were genotyped with the Illumina BovineSNP50 BeadChip (50K) version 1 and
3 with standard quality control parameters for missing genotype rate (MIND) < 0.1, minor allele frequency (MAF) < 0.05, and Hardy-Weinberg equilibrium (HWE) of P < 0.000001.
Analysis was performed within PLINK software (Purcell et al. 2007).
Fig. 1 Histogram of Red Dual-Purpose (RDP) candidates per cohort - 62 -
Chapter Four ______
Population genetic parameters
The classical inbreeding coefficient FPED was computed based on the function pedInbreeding() from R-package ‘optiSel’ (Wellmann et al. 2017) in R-statistic software
(2018). Genomic inbreeding FROH was computed using runs of homozygosity (ROH) in
PLINK with adjusted parameters of --homozyg-density 1000, --homozyg-window-het 1, -- homozyg-kb 10, and --homozyg-window-snp 20 (Purcell et al. 2007; Bosse et al. 2012; Zhang al. 2015). Genetic diversity (GD) of the birth year 푡 is the probability that two alleles chosen at random from the birth year are not IBD. Since it can be computed as GD(t) = 1 − fA(푡), whereas the classical kinship 푓퐴(푖, 푗) between two individuals 푖 and 푗, which is also called genealogical coancestry, is the probability that two alleles chosen from both individuals from the same locus, are identical by descent (IBD). For the pedigree-based GD (GDPED) the classical 푓퐴(푖, 푗) calculated with the pKin() function from R-package ‘optiSel’ (Wellmann et al. 2017) was used, whereas for the genome-based GD (GDGEN) the genomic 푓퐴(푖, 푗) calculated with the sKin() function from R-package ‘optiSel’ (Wellmann et al. 2017) was used. Classical multidimensional scaling (MDS), also known as principal coordinates analysis
(Gower 1966), of the relationship matrix was performed with the cmdscale() function from
‘stats’ R package (R Core Team 2018).
Estimation of native genetic contribution
The pedigree-based trait for NC (NCPED) was estimated as a proxy based on pedigree information with the function pedBreedComp() from ‘optiSel’ (Wellmann et al. 2017).
Genetic contributions from native founders and from other breeds were computed according to the pedigree for every individual. Therefore, fractions of genes that originate from the respective breed were considered as genetic contributions.
Genome-based estimation of the trait for NC (NCGEN) was computed with the function segBreedComp() in R-package ‘optiSel’ (Wellmann et al. 2017). This function computes the
- 63 -
Chapter Four ______breed composition for every individual, including the genetic contribution from native ancestors. This contribution can be explained as proportion of the genome belonging to segments whose frequency is smaller than a predefined value in all other breeds. The target breed for the phenotypic estimation was RDP. All other breeds, i. e. RH, HF, GA, FV, GBP, and RHV, were used as reference breeds in order to identify native haplotype segments of
RDP. For the estimation all markers with undefined map positions, mapped to the Y chromosome, and divergent minor and major alleles at the same marker position compared between the Illumina BovineSNP50 BeadChip (50K) version 1 and 3 were excluded from the analyses. The final dataset included 14,020 autosomal SNPs. Individual haplotype estimations were performed in breed-specific datasets with the software SHAPEIT (O’Connell et al.
2014).
Correlations
Conventional traits for milk yield (EBV_M), longevity (EBV_N), exterior (EBV_E), somatic cells (EBV_S), fertility (EBV_R), and maternal calving (EBV_Km) were provided by vit computation center for every individual. Pearson’s correlations between pedigree based
NCPED and genomic based NCGEN were calculated and tested for significance (P > 0.05) with respective conventional traits. Therefore, the cor.test() function from ‘stats’ implemented in
R-statistic software package (R Core Team 2018) was used.
RESULTS
Population genetic parameters
Candidates’ pedigree-based classical inbreeding coefficient FPED was higher (0.005-0.036) than the genomic inbreeding coefficient of FROH (0.001-0.008) and showed a higher variation over the years (Figure 2a). FPED of the restrictive RDP population showed even higher inbreeding but lower variation compared to FROH. However, candidates’ FROH is nearly in accordance with populations’ FPED compared to candidates’ FPED with populations’ FPED. - 64 -
Chapter Four ______
Inbreeding parameters of FPED(Population) and FROH(Candidates) slightly decreased over the years, whereas FPED(Candidates) increased. In total, inbreeding was low regarding both pedigree- and genome-based estimators.
Fig. 2 In a): Inbreeding coefficients (F) and in b): Genetic diversity (GD) for Red Dual- Purpose (RDP) candidates based on genealogical and genomic information per cohort
However, males (born between 1994 and 2001) were more inbred than females (born between
2000 and 2016) in all inbreeding estimates (Figure 2a). Candidates’ pedigree-based genetic diversity GDPED was lower (0.949-0.976) than the genomic genetic diversity of GDROH
(0.964-0.976) and showed a higher variation over the years (Figure 2b). GDPED of the restrictive RDP population showed even higher genetic diversity (0.997-0.998) but lower variation compared to FPED and FROH. Genetic diversity of the candidates (GDPED and GDROH) was lower than the pedigree-based GDPED of the restrictive population and showed also a lower variation over the years. History curves of candidates’ GDPED and GDROH showed accordance, whereas GDROH was higher than GDPED. In total, the variation of the candidates’ genetic diversity (GDPED and GDROH) was higher between 1994 and in the last decade. In general, the GD of the candidates is high. The MDS shows the genomic relatedness between chosen RDP candidates (Figure 3). Candidates matched into one interrelated population without any split subpopulations. Genetic diversity and relatedness between male and female candidates is nearly the same without any bias. However, the genetic diversity between chosen male and female individuals is widespread.
- 65 -
Chapter Four ______
Fig. 3 Genomic relatedness between Red Dual-Purpose (RDP) candidates
Estimation of native genetic contribution
Mean estimates of pedigree- and genome-based traits for NC are listed in Table 1. NCPED showed a mean value of 0.284, whereas NCGEN was higher with 0.786 across all individuals.
Table 1 Mean estimates of pedigree and genomic based native genetic contribution (NC) and respective migrant contribution (MC) RDP (N=809) Pedigree Genome Mean NC 0.284 0.786 Mean MCa 0.716 0.214 RH 0.497 0.039 HF 0.175 0.034 GA 0.001 0.041 GBP 0.005 0.053 RHV - 0.020 FV - 0.027 Other 0.038 - aMean MC for RH (Red Holstein), HF (Holstein Friesian), RDP (Red Dual-Purpose cattle), GBP (German Black Pied cattle), RHV (Rotes Hoehenvieh), FV (Fleckvieh), Other
In contrast, the mean MC was 0.716 for pedigree- and 0.214 for genome-based measurements.
For pedigree-based estimates, the introgressed RH breed had the highest impact on MC with - 66 -
Chapter Four ______
0.497, whereas migrant alleles from GBP breed had the highest influence on genome-based
MC with 0.053 (Table 1). NCPED ranged between 0.08 and 0.52, whereas NCGEN ranged from
0.48 to 0.87. Estimates for NCPED ranged from 0.12 to 0.53 and estimates for NCGEN ranged from 0.49 to 0.74, whereas both estimates showed a low accordance with a minor correlation of 0.12 (Figure 4). In general, NCGEN was higher than NCPED, whereas NCGEN showed lower variation of NC values (Figure 4).
Fig. 4 Pedigree- (NCPED) and genome-based (NCGEN) estimates for the trait of native genetic contribution (NC) for Red Dual-Purpose (RDP) candidates
Correlations
Correlation between NCPED and NCGEN was 0.12 (Table 1 and Figure 4). NCPED was positively correlated with EBV_N (0.19) and EBV_S (0.13), and negatively correlated with
EBV_R (-0.09) and EBV_Km (-0.10). NCGEN was positively correlated with EBV_N (0.16),
EBV_E (0.23), and EBV_S (0.08). NCGEN revealed no other significant correlations regarding conventional traits. Conventional traits were positively correlated between EBV_N and
EBV_E (0.09), EBV_N and EBV_S (0.40), EBV_E and EBV_S (0.10), and between EBV_R and EBV_Km (0.23). Negative correlations were calculated between EBV_M and EBV_R (-
0.11) and between EBV_E and EBV_R (-0.23). All computed correlations between estimated and conventional traits are shown in Table 2.
- 67 -
Chapter Four ______
DISCUSSION
To our knowledge only one study about correlations between NC and conventional traits exists (Schäler et al. 2018). However, many authors require the maintenance of native uniqueness and conservation of endangered breeds as gene reservoir (Boettcher et al. 2010 and Wang et al., 2017) even though there is no proof of the benefit for NC. The aim of this work was to identify possible benefits of NC in a local cattle breed.
Population genetic parameters
Results of the genome-based population genetic parameters were obviously more in accordance with the results of the restrictive RDP than compared to candidates’ pedigree- based estimations. Consequently, pedigree-based results have been not considered for further interpretations. Increased inbreeding coefficients FGEN for the individuals born between 1994 and 2004 were slightly higher (Figure 2) due to the frequency of animals (Figure 1), which was low between 1994 and 2004 and high between 2006 and 2015. It could be that male
(older) candidates were more inbred than female (younger) candidates due to genetic gain.
GDGEN showed a simultaneous history (Figure 2), whereas the variation was higher due to slight frequency of candidates at this time (Figure 1). However, GDGEN was high and not threatened which was also in accordance with the MDS plot (Figure 4). Although the frequency of candidates had a strong variation over cohort, older (male) and younger (female) animals revealed nearly the same relatedness and genetic diversity (Figure 4). Consequently, chosen animals were probably representatives of the restrictive population, which will make the value of the genome-based results and conclusions meaningful.
Estimation of native genetic contribution
The very slight compliance between NCPED and NCGEN estimates (Figure 4) strengthened the assumption of a non-reliable database in case of the candidates’ pedigree. The pedigree file showed large gaps and a high rate of incompleteness regarding dams as mentioned by Addo et
- 68 -
Chapter Four ______al. (2017). In this regard, animal selection comprised 744 females with incomplete pedigree data, which means 92 % of the whole dataset. It was not possible to get the same reliability for pedigree as for genomic information when the greater amount of selected individuals was female.
Correlations
Correlations between NCPED and NCGEN were slightly positive and not as high as expected
(Table 2). A selection on NCPED cannot be recommended due to low reliabilities of the pedigree database. However, NCGEN probably achieved higher reliabilities due to accordance of the genomic database with the restrictive population genetic parameters of F and GD. A selection on NCGEN would not cause a negative impact on fertility and calving traits, which are positively correlated with each other. In addition, the most important economic trait
EBV_M will not be affected by selecting on NCGEN. NCGEN showed positive relations with the traits for longevity and udder health. Thus, the trait of NC may be related to more robust and functional animals which are not as prone to diseases. In fact, these animals could be economically more efficient and socially sustainable in context of animal welfare. Therefore, it can be assumed to enhance the longevity by selecting for high NCGEN. Additionally, NCGEN revealed a positive correlation to EBV_E, which is related to the conformation. This correlation is not unsuspected due to the moderately positive correlation between EBV_N and
EBV_E. Generated results are in accordance with the study of Schäler et al. (2018) where NC was also positively correlated with udder health (EBV_S) and conformation traits (EBV_E).
These EBVs interact with each other in context of fitness, functional traits, and health.
CONCLUSION
A selection on NC would benefit in an improvement of longevity, confirmation, and udder health traits for the local Red Dual-Purpose cattle breed.
- 69 -
Chapter Four ______
Acknowledgement: Financial support from the Ministry of Energy, Agriculture,
Environment, Nature, and Digitalization within the framework of the European Innovation
Partnership (EIP Agri) is gratefully acknowledged.
REFERENCES
Addo S, Schäler J, Hinrichs D, Thaller G (2017) Genetic Diversity and Ancestral History of
the German Angler and Red-and-White Dual-Purpose Cattle Breeds Assessed through
Pedigree Analysis. Agricultural Sciences 8:1033-1047
Bennewitz J, Simianer H, Meuwissen THE (2008) Investigations on Merging Breeds in
Genetic Conservation Schemes. Journal of Diary Science 91:2512-2519
Boettcher P J, Tixier-Boichard M, Toro MA, Simianer H, Eding H, Gandini G, Joost S,
Garcia D, Colli L, Ajmone-Marsan P, The GLOBALDIV Consortium (2010)
Objectives, criteria and methods for using molecular genetic data in priority setting for
conservation of animal genetic resources. Journal of Animal Genetics 41(1):64-77
Bosse M, Megens HJ, Madsen O, Paudel Y, Frantz LA, Schook LB, et al. (2012) Regions of
homozygosity in the porcine genome: consequence of demography and the
recombination landscape. PLoS Genetics 8(11):e1003100
Groeneveld LF, Lenstra JA, Eding H, Toro MA, Scherf B, Pilling D, Negrini R, Finlay EK,
Jianlin H, Groeneveld E, Weigend S, The GLOBALDIV Consortium (2010) Genetic
diversity in farm animals – a review. Animal Genetics 41(1):6-31
Gower JC (1966) Some distance properties of latent root and vector methods used in
multivariate analysis. Biometrika 53:325-328
Meuwissen THE (1997) Maximizing the response of selection with a predefined rate of
inbreeding. Journal of Animal Science 75:934-940
- 70 -
Chapter Four ______
O’Connell J, Gurdasani D, Delaneau O, Pirastu N, Ulivi S, et al. (2014) A General Approach
for Haplotype Phasing across the Full Spectrum of Relatedness. PLoS Genetics 10(4):
e1004234. doi:10.1371/journal.pgen.1004234
Purcell S, Neale B, Todd-Brown K, Thomas L, Ferreira MAR, Bender D, Maller J, Sklar P,
De Bakker PIW, Daly MJ, Sham PC (2007) PLINK: a toolset for whole-genome
association and population-based linkage analysis. American Journal of Human
Genetics 81. http://zzz.bwh.harvard.edu/plink/contact.shtml
R Core Team (2018) R: A Language and Environment for Statistical Computing. R
Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org
Schäler J, Wellmann R, Stratz P, Bennewitz J, Thaller G, Hinrichs D (2018) Performance of a
novel breeding value in context of breed conservation. Proceedings of the World
Congress on Genetics Applied to Livestock Production, 11.22, Auckland, New Zealand
http://www.wcgalp.org/proceedings/2018/performance-novel-breeding-value-context-
breed-conservation
Wang Y, Segelke D, Emmerling R, Bennewitz J, Wellmann R (2017) Long-Term Impact of
Optimum Contribution Selection Strategies on Local Livestock Breeds with Historical
Introgression Using the Example of German Angler Cattle. G3: Genes, Genomes,
Genetics 7(12):4009-4018
Wellmann R, Hartwig S, Bennewitz J (2012) Optimum contribution selection for conserved
populations with historic migration. Genetics Selection Evolution 44(1):34
Wellmann R (2017) Optimum Contribution Selection and Mate Allocation for Breeding: The
R Package optiSel. Submitted to Genetics Selection Evolution
Zhang Q, Calus MPL, Guldbrandtsen B, Lund MS, Sahana G (2015) Estimation of inbreeding
using pedigree, 50k SNP chip genotypes and full sequence data in three cattle breeds.
BMC Genetics 16:88
- 71 -
Chapter Four ______
(***)
1
Purpose Purpose
-
0.10 (**) 0.10
EBV_Km -
0.00 (n.s.) 0.00 (n.s.) 0.05 (n.s.) 0.03 (n.s.) 0.01 (n.s.) 0.06
0.23 0.23
1
for for Red Dual
EBV_R
0.09 (**) 0.09 (**) 0.11
0.06 (n.s.) 0.06 (n.s.) 0.02
0.23 (***) 0.23
- -
0.01 (n.s.) 0.01
- -
-
traits
1
EBV_S
0.08 (*) 0.08
0.10 (**) 0.10
0.00 (n.s.) 0.00
0.13 (***) 0.13 (***) 0.40
(*), < 0.01 0.01 < < (***) 0.001 (**), (*),
1
EBV_E
0.09 (**) 0.09
0.01 (n.s.) 0.01
0.00 (.n.s) 0.00
0.23 (***) 0.23
-
1
contribution contribution (NC) and conventional
EBV_N
value ≥ 0.05 0.05 ≥ value 0.05 (n.s.), <
0.03 (n.s.) 0.03
0.19 (***) 0.19 (***) 0.16
-
genetic genetic
1
native
EBV_M
0.03 (n.s.) 0.03
0.00 (n.s.) 0.00
-
GEN
1
NC
0.12 (***) 0.12
PED
1
NC
a
Correlations Correlations between the trait for
Estimates were tested for statistical significance: p significance: statistical tested for were Estimates
a
PED GEN
Table 2 (RDP) cattle NC NC EBV_M EBV_N EBV_E EBV_S EBV_R EBV_Km
- 72 -
Chapter Five ______
Chapter Five
Implementation of breed-specific traits for a local sheep breed
J. Schäler1, G. Thaller1 & D. Hinrichs2
1Institute of Animal Breeding and Husbandry, Christian-Albrechts-University, Hermann- Rodewald-Straße 6, 24098 Kiel, Germany 2Department of Animal Breeding, University of Kassel, Nordbahnhofstraße 1a, 37213 Witzenhausen, Germany
Published in Agricultural Sciences 9(8): 958-973
- 73 -
Chapter Five ______
ABSTRACT
In recent decades, a considerable number of local breeds have been replaced by high-yielding breeds for reasons of profitability. Many local breeds are now threatened by extinction and the loss of their native genetic diversity. The need to conserve breeds and their genetic diversity has a major importance due to the necessity for genetic change within and between populations. Novel approaches have to be explored and extended to maintain this genetic diversity. The aim of this study was the identification and implementation of breed-specific traits for a small, local sheep breed in northern Germany. The data comprised pedigree information, estimated breeding values (EBVs) of several conventional traits, and phenotypic information from a field experiment for two novel traits: (1) average daily gain under extensive circumstances (ADGE) and (2) ultrasonic measurements of muscle-fat ratio
(UMFR). The experimental design included a dataset of 47 progeny from 14 pure-bred rams of German White-Headed Mutton (GWM). The methodical approach was divided into four parts: (I) the analysis of the breeding program, (II) the identification of breed-specific traits,
(III) the estimation and correlation of novel breeding values, and (IV) the consequences of implementing these novel traits. Genetic parameters and correlations were conducted by applying linear mixed models. The estimates for the heritability (repeatability) were between
0.70 and 0.83 (0.42 and 0.46). The genetic correlation was positive (0.61) and in accordance with the phenotypic correlation (0.62). Average daily gain under intensive circumstances
(ADGI) was moderately positive correlated with muscularity (0.60), as opposed to ADGE, which was moderately negative correlated with muscularity (-0.68). The EBV of ADGE was also moderately positive correlated with UMFR (0.64). Genetic response for ADGE enhanced to values of 481.09 g/day, 639.97 g/day, > 700 g/day and > 850 g/day for different selection intensity scenarios. Corresponding rates of inbreeding were 1.4 %, 2.7 %, 5.1 %, and 7.9 % after 10 years of selection. Genetic response for UMFR increased to 0.92, 1.34, 2.41, and >
2.75, whereas remaining rates of inbreeding increased to 1.1 %, 2.2 %, 5.1 %, and 7.9 %.
- 74 -
Chapter Five ______
ADGI and ADGE were tendentially negatively correlated (-0.11), which strengthen the assumption of a biased ADGI. ADGE has a positive influence on meat-quality aspects
(UMFR). Optimal use of reference sires with predefined selection intensity achieves genetic response for ADGE and UMFR with simultaneously acceptable rates of inbreeding.
Keywords: EBV, genetic gain, local sheep breed, new phenotypes, novel traits, selection intensity
- 75 -
Chapter Five ______
INTRODUCTION
The domestication of livestock species has created an enormous variety of breeds due to a long history of migrations, selection, and adaptation (Groeneveld et al. 2010). During the last few centuries, many well-defined breeds have peaked in numbers and have been used for a variety of purposes with different levels of performance depending on demographic and local environments. In recent decades, however, the reproductive technologies of artificial insemination and embryo transfer have become more widespread (Wang et al. 2017), and have facilitated the dissemination of genetic material. As a consequence, selection programs have become more efficient and have accelerated the genetic gain in a small number of breeds. These high-yielding breeds have replaced many local breeds, which has resulted in a high rate of loss of local breeds due to extinction (Meuwissen 2009). Thus, many populations of local breeds have dangerously decreased in number and are even threatened by extinction
(FAO 2007; FAO 2010; Fernández et al. 2011). According to Boettcher et al. (2010), there exists a wide agreement on the need to conserve breeds and their genetic diversity due to the necessity for genetic change within a population. In addition, local breeds with their native genetic diversity allow for the selection of special traits to increase productivity, competitiveness, and to adapt to changing environmental conditions. Different approaches and considerations were discussed by Meuwissen (2009) for the purpose of breed conservation. These methods consisted of aspects of optimum contribution selection
(Meuwissen 1997; Wellmann et al. 2012), integrating life and cryoconservation schemes
(Sonesson et al. 2002; Shepherd and Woolliams 2004), rotational breeding schemes (Colleau and Avon 2008), and the introduction of novel traits from conserved populations into commercial breeding populations (Hospital 2005; Odegard et al. 2009). The latter approach implies that traits from conserved breeds can be introduced into commercial breeds by introgression and genomic selection. For most local breeds, however, genomic selection is not implemented due to the small population sizes or costs. Thus, breeding progress is still based
- 76 -
Chapter Five ______on recording phenotypic information. However, most local breeds are implemented into conventional breeding programs, where the specific traits (e.g. fertility, meat quality, milk ingredients, and robustness) of local breeds attract no interest and may even be lost due to undefined negative correlations with positively selected conventional traits. Additionally, certain traits (e.g. disease resistance) of local breeds are not identified or phenotypically recorded in conventional breeding programs. Thus, usage, conservation, and breeder’s impact regarding these traits is complicated. It should be of major importance to identify and conserve these unique and worthwhile traits from conserved local breeds as long as they still exist.
The aim of the present study was to identify and implement breed-specific traits for a small, local sheep breed in Germany. Therefore, estimated breeding values (EBVs) for novel traits were computed based on collected phenotypic information from a field experiment. Further, correlations between novel and conventional EBVs were investigated and benefits of implementing these novel traits were clarified.
MATERIALS AND METHODS
Animals and phenotypes
For an effective recording of phenotypes and optimally statistical computation of novel
EBVs, preliminary considerations were carried out with regard to the experimental design dependent on the possibilities of local farmers and their local breed. Data comprised pedigree information, EBVs for several conventional traits, and phenotypic information on average daily gain under extensive circumstances (ADGE) and quantitative ultrasonic (QUS) measurements of muscle-fat ratio (UMFR) for 47 progeny from 14 pure-bred rams (reference sires) of German White-Headed Mutton (GWM), born between 2010 and 2014. Pedigree information was provided by LKV SH (Landeskontrollverband Schleswig-Holstein e.V., Kiel,
Germany). However, EBVs for conventional traits and phenotypic data on ADGE and UMFR
- 77 -
Chapter Five ______were provided by LV SH SZZ (Landesverband Schleswig-Holsteinischer Schaf- und
Ziegenzüchter e.V., Kiel, Germany).
Phenotypic data was collected and measured during a field experiment on one standardized farm, where 47 pure-bred male progeny of 14 GWM reference sires were fattened based on extensive feed without concentrates during a trial period of 100 days. During the field study ethical considerations of animal welfare along the lines of Putman (1995) were claimed. The animals had an average age of 99.4 days and were divided into two groups depending on their date of birth. One group included 24 animals born in the first half of January 2016. The other group contained 23 animals born during the second half of the month. The data collection consisted of measuring the rams’ weight at six different times at regular intervals over the trial period and the QUS measurements of muscle and the QUS measurements of fat depth separately at the end of the experiment. The trait of ADGE resulted from dividing the average weight gain, deducting the general birth weight of approximately 4.5 Kg, by the experimental time of the trial period per animal. The trait of UMFR was computed by dividing the QUS measurements of muscle by fat depth. Observed phenotypes for the trait of ADGE ranged between 373.0 and 243.3 g/day (Table 1). However, phenotypic observations for the trait of
UMFR were recorded between the maximum of 2.78 and the minimum of 1.32 (Table 1).
- 78 -
Chapter Five ______
Table 1 Observation averages of German White-Headed Mutton (GWM) reference sire´s progeny for the novel traits of average daily gain under extensive circumstances (ADGE) and ultrasonic muscle-fat ratio (UMFR) Observations for novel traitsa Reference sire ADGE (g/day) UMFR 1 ID 1 309.2 1.74 2 ID 2 333.0 1.53 3 ID 3 319.0 1.52 4 ID 4 329.0 1.53 5 ID 5 373.0 1.89 6 ID 6 317.0 1.40 7 ID 7 292.4 1.32 8 ID 8 347.0 1.67 9 ID 9 371.0 2.78 10 ID 10 287.0 1.65 11 ID 11 335.0 1.40 12 ID 12 288.4 1.36 13 ID 13 243.3 1.39 14 ID 14 314.0 1.59 a ADGE = average daily gain under extensive circumstances; UMFR = ultrasonic muscle-fat ratio
The inbreeding coefficients (F) were estimated with the function ‘pedInbreeding’ from the
‘optiSel’ R-package (Wellmann 2018), whereas the rates of inbreeding were calculated for each year as ∆퐹푖 = (퐹푖 − 퐹푖−1)/(1 − 퐹푖−1). The rate of inbreeding between year 푖 and 푗
(∆퐹푖−푗) was computed by the average of annual inbreeding rates (Lewis and Simm 2000).
The additive genetic relationship matrix was estimated with the function ‘makeA’ from R- package ‘optiSel’ (Wellmann 2018).
(I) Analysis of the breeding program
Important information regarding the breeding program was collected from the breeding organisation on demand. The breeding program was analysed by identifying breeding goals and conventional traits with their relative weights and analysing their influence on total merit index (TMI). The breeding goal for GWM is defined as a robust, muscled, and well-growing mutton, which is well-adapted to grazing in damp and maritime climates and various ground
- 79 -
Chapter Five ______conditions. The conventional TMI included the three traits of average daily gain under intensive circumstances (ADGI), muscularity (MUSC), and wool (WOL) with consistent relative weights of 0.33 for each trait.
(II) Identification of breed-specific traits
The breed and its breeding history were analysed to identify special and valuable traits. A comprehensive literature review was carried out and, in addition, face to face interviews with farmers and staff of the breeding organisation were conducted to emphasise breed-specific traits. The main purpose of the GWM breed is landscape conservation on the dykes of the northern coasts of Germany. Thereby, their job with their browsing is it to make the dyke slip- proof, densify the ground, and to ensure against flooding. In general, the animals were kept outside on the dykes with their progeny the whole year and only received extensive feed without supplemented concentrates. However, when young rams became licensed for breeding purposes, farmers began to fatten their rams additionally with supplements to increase weights in order to obtain a higher breeder valuation later. Thus, valuation results for
GWM rams were erroneously assumed to be correct for the trait of conventional ADG related to the constant feed environment. The trait of conventional ADG for GWM rams is biased, which has a major impact on the carcass value of the lambs. Currently, the main source of income from the GWM breed besides landscape conservation is to submit the lambs for slaughter. For the local sheep breed, a special feature of ADG based on extensive feed intake
(ADGE) was assumed due to the breeding history, demographic circumstances of husbandry, and economic value of this trait. In addition, it was assumed that the fattening period under extensive circumstances had an impact on the muscle and fat depth ratio (UMFR), which is an important meat-quality indicator.
- 80 -
Chapter Five ______
(III) Estimation and correlation of novel breeding values
To compute novel EBVs, linear mixed models (LMM) were applied using the R-package
‘asreml’ from Butler et al. (2009). The LMM can be written as
푦 = 푋푏 + 푍퐴푎 + ∑푘 푍푘푢푘 + 푒 (1) where y denotes the n-vector of phenotypic values, b is the vector of fixed effects, a is the
2 2 vector of random additive genetic effects of the animal distributed as 푎~푁(0, 휎푎 퐴), where 휎푎 is the additive genetic variance and 퐴 is the additive relationship matrix. Vector 푢푘 of
2 independent random effects has distribution 푢푘~푁(0, 휎푢푘I) and e is the n-vector of
2 independent residual errors with 푒~푁(0, 휎푒 I). Matrices X, ZA, and Z are design matrices associating observations with the appropriate combination of effects. Fixed and random effects of sire, dam, sex, date of birth, and breeder were tested for significance using the R- package ‘asremlPlus’ from Brien (2016). The statistical analyses were performed simultaneously in a bivariate analysis for each trait in order to estimate the repeatability (t) and the heritability (h2) for these traits. The repeatability following Lessells and Boag (1987) was computed as
휎2+∑ 휎2 푡 = 푎 푘 푢푘 (2) 휎2 +∑ 휎2 + 휎2 푎 푘 푢푘 푒 and the heritability following Falconer and Mackay (1996) was calculated as
휎2 ℎ² = 푎 . (3) 휎2 +∑ 휎2 + 휎2 푎 푘 푢푘 푒
The estimation of genetic (푟퐺̂ ) and phenotypic (푟푃̂ ) correlations between the traits were also conducted within the pairwise bivariate analysis by usage of the LMM. The genetic correlation was calculated as
2 2 푐표푣(휎푎(푋)휎푎(푌)) 푟퐺̂ (푋,푌) = (4) 2 2 √휎푎(푋)휎푎(푌) and the phenotypic correlation was computed as
- 81 -
Chapter Five ______
푐표푣(휎2 휎2 )+푐표푣(∑ ( 휎2 ) ∑ ( 휎2 ) ) 푎(푋) 푎(푌) 푘 푢푘 (푋) 푘 푢푘 (푌) 푟푃̂ (푋,푌) = , (5) √휎2 휎2 + ∑ ( 휎2 ) ∑ ( 휎2 ) 푎(푋) 푎(푌) √ 푘 푢푘 (푋) 푘 푢푘 (푌) where X and Y denote distinct traits. Correlations between novel and conventional EBVs were estimated with the ‘cor.test’ function from the R-package ‘stats’ (R Core Team 2018).
(IV) Implementation of breed-specific traits
Selection on breed-specific traits was carried out for 10 years with genetic evaluation of all animals once a year. The genetic evaluation was based on the estimation of genetic response with the formula from Rendel and Robertson (1950), calculated as
푖∗푟∗휎 ∆퐺 = 퐴, (6) 퐿 where 푖 is the selection intensity, 푟 is the correlation between true and estimated breeding value, 휎퐴 is the additive genetic standard deviation of the trait, and 퐿 is the generation interval of the species or breed. Four scenarios with different 푖 were considered based on the selected proportions (p%) of the reference sires. To translate p% to the corresponding 푖, different tables were used from Falconer and Mackay (1996). First scenario had a p% of 50 and was translated to an 푖1 of 0.798, which meant 7 out of 14 reference sires. Second, third, and fourth scenarios had different p% of 36, 21, and 14 and were compiled to corresponding 푖2 of 1.039,
푖3 of 1.372, and 푖4 of 1.590, which implemented 5, 3, and 2 out of 14 reference sires.
Correlation between true and estimated breeding value (accuracy of breeding value estimation) as well as the values for trait-specific 휎퐴 were calculated within the LMM.
Average generation interval for ewes (퐿푒) was defined as 3.5 years and the average generation interval for reference sires (퐿푠) was considered as 2.7 years (Lewis and Simm 2000). The true generation interval of 퐿 is the average of both parameters and had a value of 3.1 computed as
퐿 +퐿 퐿 = 푒 푠. Reference sires were chosen based on their phenotypic measurements for the goal 2 trait. Within a flock, 20 ewes were mated to one of the selected reference sires depending on different 푖-scenarios. For 푖1 in total 7 flocks with each of them 20 ewes were mated with 7 - 82 -
Chapter Five ______different reference sires over 10 years. For 푖2, 푖3, and 푖4 in total 5, 3, and 2 flocks with each of them 20 ewes were mated with 5, 3, and 2 different reference sires over 10 years. Average phenotypic measurements (퐺푖) was obtained for selected reference sires in the 푖th year and the annual rate of response between years 푗 and 푖 was computed as ∆퐺푖−푗 = (퐺푗 − 퐺푖)/(푗 −
푖), where 푗 > 푖 (Lewis and Simm 2000).
RESULTS
The number of GWM herd book animals tendentially declined until the year 2015. In 1970, the recording of the herd book started with a small number of animals and simultaneously the herd book animals began to increase up to a maximum of approximately 830 recorded animals between 1988 and 1989. After this peak, GWM herd book livestock steadily decreased to a minimum of 167 individuals in 2015 (Figure 1a).
Fig. 1 Indication of a): Frequency of herd book animals and b): Rates of inbreeding (ΔF) regarding past breed history of the German White-Headed Mutton (GWM)
The rate of inbreeding (∆F) was constant at a low level between 1970 and 1986. In 1987, ∆F increased from 0.07 % to 0.7 % till 2007. The time period between 2007 and 2015 showed a rapidly increased ∆F up to approximately 10 % (Figure 1b).
- 83 -
Chapter Five ______
Fig. 2 In a): Estimated inbreeding coefficients (F) and in b): Corresponding additive genetic relationship matrix of reference sires
The F of reference sires had an average value of 0.61 %, whereas 3 rams showed low coefficients between 0.78 % and 1.56 %, and 1 ram had a high inbreeding of 4.69 % (Figure
2a). In total, nearly all reference sires revealed relatedness (Figure 2b). Reference sires ID 2,
ID 3, ID 6, ID 7, ID 8, ID 10, ID 11, and ID 12 showed different levels of relatedness to one or more reference sires. Collected phenotypic observations for average daily gain under extensive circumstances and ultrasonic muscle-fat ratio (Table 1) were positively linearly related (Figure 3).
Fig. 3 Phenotypic relation between average daily gain under extensive circumstances (ADGE) and ultrasonic muscle-fat ratio (UMFR)
- 84 -
Chapter Five ______
Estimation and correlation of novel breeding values
The effects of sex and date of birth were significant (P ≤ 0.001) for both traits, thus they were included as fixed effects in the models. However, the random effect of dam was included in the genetic analysis as a random effect in order to avoid an over-estimation of heritability, an under-estimation of maternal effects, and to increase the reliability. The repeatability for the trait ADGE was 0.42 and 0.46 for the trait of UMFR. The corresponding heritability was 0.70 for ADGE and 0.83 for UMFR (Table 2).
Table 2 Results of bivariate analyses for heritability (ĥ2) and repeatability (t̂) of average daily gain under extensive circumstances (ADGE) and ultrasonic muscle-fat ratio (UMFR) and standard errors (SE) Linear mixed model (LMM) Traitsa Repeatability t̂ (SE) Heritability ĥ2 (SE)
ADGE 0.42 (0.31) 0.70 (0.95) UMFR 0.46 (0.46) 0.83 (0.59) a ADGE = average daily gain under extensive circumstances; UMFR = ultrasonic muscle-fat ratio
Phenotypic and genetic correlations are positive between both traits with 0.62 and 0.61 (Table
3). The correlation between true and estimated breeding value was fixed and had an accuracy of 0.725 for both traits. Investigated trait-specific parameters of 휎퐴 were 186.81 g for ADGE and 0.49 for UMFR.
Table 3 Results of bivariate analyses for phenotypic (above the diagonal) and genetic (below the diagonal) correlations (r̂G and r̂P) between average daily gain under extensive circumstances (ADGE) and ultrasonic muscle-fat ratio (UMFR) and standard errors (SE)
a Traits ADGE UMFR
ADGE - 0.62 (0.30)
UMFR 0.61 (0.29) - a ADGE = average daily gain under extensive circumstances; UMFR = ultrasonic muscle-fat ratio
Correlations between novel and conventional EBVs are shown in Table 4. ADGI was
- 85 -
Chapter Five ______moderately positive correlated with MUSC (0.60), whereas ADGE was moderately negative correlated with MUSC (-0.68). The EBV of ADGE was also moderately positive correlated with the EBV of UMFR (0.64). Other correlations between EBVs were not statistically significant.
Table 4 Correlation between novel estimated breeding values (EBVs) of ADGE and UMFR and conventional EBVs of ADGI, MUSC, and WOL a EBV ADGI MUSC WOL ADGE UMFR
ADGI 1 0.60 (*) -0.12 (n.s.) -0.11 (n.s.) 0.04 (n.s.) MUSC 1 0.06 (n.s.) -0.68 (**) -0.31 (n.s.) WOL 1 -0.40 (n.s.) -0.17 (n.s.)
ADGE 1 0.64 (*) UMFR 1 aEstimates were tested for statistical significance: p-value ≥ 0.05 (n.s.), < 0.05 (*), < 0.01
(**), < 0.001 (***); ADGI = average daily gain under intensive circumstances; MUSC = muscularity; WOL = wool; ADGE = average daily gain under extensive circumstances; UMFR = ultrasonic muscle-fat ratio
Implementation of breed-specific traits
Genetic response and corresponding ∆F were simulated for the implementation of the trait
ADGE regarding different 푖-scenarios over 10 years of selection (Figure 4). Genetic response of ADGE increased with an increasing 푖. In total, genetic response enhanced from the average of 318.45 g/day to values of 481.09 g/day for 푖1 and 639.97 g/day for 푖2 after 10 years of selection. However, genetic response for 푖3 and 푖4 reached values > 700 g/day and > 850 g/day after 7 years of selection (Figure 4a). Corresponding ∆F decreased from an average F of
2.22 at year 0 of selection to the first years of selection depending on different 푖, whereas ∆F started to increase rapidly with an increasing value of 푖 after some years of selection. After 10 years of selection, ∆F had values of 1.4 %, 2.7 %, 5.1 %, and 7.9 % for 푖1, 푖2, 푖3, and 푖4
(Figure 4b).
- 86 -
Chapter Five ______
Fig. 4 In a): Change in genetic gain for average daily gain under extensive circumstances
(ADGE) and in b): Corresponding rates of inbreeding (ΔF) over years with different selection intensities (i)
For UMFR the same simulations were performed and are shown in Figure 5. Genetic response increased from an average of 0.49 and reached values of 0.92, 1.34, 2.41, and > 2.75 for 푖1, 푖2,
푖3 and 푖4 after 10 years of selection (Figure 5a). Remaining ∆F increased to values of 1.1 %,
2.2 %, 5.1 %, and 7.9 % for 푖1, 푖2, 푖3 and 푖4 after 10 years (Figure 5b).
- 87 -
Chapter Five ______
Fig. 5 In a): Change in genetic gain for ultrasonic muscle-fat ratio (UMFR) and in b): Corresponding rates of inbreeding (ΔF) over years with different selection intensities (i)
DISCUSSION
The objective of this study was the identification and implementation of breed-specific traits for a small, local sheep breed. Therefore, phenotypic information from a field experiment were collected and used to estimate novel EBVs. In addition, correlations between these novel and conventional EBVs were investigated and benefits of implementing these novel traits were clarified.
Estimation and correlation of novel breeding values
The number of reference sires and their tested progeny in the experimental design was small due to the small number of living herd book rams of the GWM breed (Figure 1a). The field design depended mainly on the farmers’ motivation and conviction regarding this project.
Consequently, just a few farmers consented to participate and provide their young rams for the trial. Nevertheless, the reference sire lines were mostly widely selected to achieve genetic - 88 -
Chapter Five ______variance within the traits but showed still inbreeding and relatedness with each other (Figure
2). The phenotypic information was collected separately for each trait by one person to avoid bias. For the calculation of UMFR, QUS measurements for the muscle and fat depth were measured under predetermined conditions to achieve consistency. In general, repeatability set the upper limit to heritability as a very useful interpretation since the heritability of traits cannot often be obtained and there is a risk of overestimation (Dohm 2002). For ADGE, the repeatability and heritability were 0.42 and 0.70 (Table 2). In the studies of María et al.
(1993), Hassen et al. (2003), and Gowane et al. (2015), heritability estimated for conventional
ADG was low until moderate and varied between 0.15 and 0.26. It can be assumed that conventional ADG and ADGE are similar traits dependent on feeding as they show the same fixed and random effects. The repeatability and heritability estimated for UMFR were 0.46 and 0.83 (Table 2). Gilmour et al. (1994) estimated a heritability for muscle depth and fat depth of between 0.05 and 0.29. The UMFR can be assumed as a meat-quality trait and these traits exhibits a heritability < 0.18 (Larsgard et al. 1998). Safari et al. (2005) provided a heritability for meat-quality traits between 0.05 and 0.18. Hence, the heritability of ADGE and
UMFR with 0.70 and 0.83 was obviously over-estimated in this model. The true values for the heritability of these traits are limited by the repeatability of 0.42 for ADGE and 0.46 for
UMFR. This moderate heritability for both traits was also strengthened by the literature. The genetic correlation between ADGE and UMFR of 0.61 (Table 3) is probably over-estimated in comparison with the study by Safari et al. (2005), where the genetic correlations for live weight and muscle depth and live weight and fat depth were between 0.34 and 0.36. Strong overestimation of heritability and genetic correlation was due to the small number of animals used in the animal model. Probably, some fixed and random effects become significant with an increased number of animals and, hence, heritability may decrease and approximate reality.
The high standard error was due to the small number of individuals and limited reference sire lines within the model. Unrelated males and generally animal sizes are often the limiting
- 89 -
Chapter Five ______factors especially in context of local breeds’ small populations (Figure 1a). Thus, repeatability cannot be ignored in the interpretation of the heritability of these novel traits for the GWM breed. Investigated correlations in Table 4 shows, that ADGI was positively correlated with
MUSC (0.60), whereas ADGE was negatively correlated with MUSC (-0.68). This result strengthen the assumption of a biased trait observation regarding ADGI and ADGE, which were negatively (-0.11) but not statistically significant correlated (Table 4). Additionally, the unbiased trait of ADGE was positively correlated with the meat-quality trait of UMFR (0.64), which is in accordance with the investigated positive regression between collected phenotypes for both traits (Figure 3). Hence, a positive impact on meat-quality aspects regarding utilisation of ADGE can be assumed.
Implementation of breed-specific traits
The conventional breeding program of GWM was geared towards ADGI, MUSC, and WOL, although the measurement for the trait of ADGI is biased and WOL has now become a subsidiary income in Germany. Additionally, the sole trait of MUSC cannot provide any information concerning meat-quality aspects. It is simply an aspect of appearance without any profitability for the breed and shows a strong dependency on ADGI (Table 4). Furthermore, the conventional breeding program focused on biased ADGI and, hence, could have led to a reduction in the native genetic variance for the unbiased breed-specific ADGE trait and the loss of this native genetic diversity. The latter can cause long-term degradation of the GWM breed and may discount profitability potentials and competitiveness for the future developments in breed-specific environments. Gandini and Oldenbroek (2007) and
Meuwissen (2009) mentioned that the definition of relevant breeding goals, the improvement of breed genetics, and the enhancement of profitability are the best strategies to move from the conservation to utilisation of a local breed. Such improvements of genetic response were qualified depending on different 푖 in Figure 4 and 5 with special focus on ∆F. Especially in a
- 90 -
Chapter Five ______small, local breed such trait implementations and genetic improvements within a breeding program have consequences for ∆F depending on 푖. There exists a link between genetic response and inbreeding (Figure 4 and 5). An increased rate of genetic gain also led to higher rates of inbreeding (Toro and Silió 1990; Villanueva et al. 1995). In general, inbreeding results in biological risks of genetic variance reduction, inbreeding depression, and accumulation of deleterious alleles (Villanueva et al. 2000). In order to achieve a small population of a local breed with their native genetic diversity and fitness the mating of close relatives and an increased inbreeding should be avoided or restricted. Therefore, a maximum for ∆F of 1 % per generation was predefined from different authors (Meuwissen and Sonesson
2004; Nielsen et al. 2014), which means a maximum for ∆F of 0.3 % per annum in case of sheep. For both traits the inbreeding level of the first scenario 푖1 stayed under this threshold of
0.3 % per annum, all other 푖-scenarios had a ∆F > 0.3 % per annum (Figure 4b and 5b). The differences of ∆F regarding 푖1 and 푖2 in Figure 4 and 5 were due to different selected reference sires with distinct inbreeding coefficients between both traits. In addition, the whole population of GWM depicts a rapidly increased ∆F since 2007 (Figure 1b) and some selected reference sires exhibits inbreeding and relatedness among each other (Figure 2).
Consequently, a p% of 50 and 푖1 of 0.798 should be chosen for an optimal use of reference sires to achieve genetic response with simultaneously acceptable rates of inbreeding.
Nevertheless, with scenario 푖1 genetic response of 481.09 g/day for ADGE and 0.92 for
UMFR would be theoretically possible after 10 years of selection (Figure 4a and 5a). This implies a trait improvement of 51 % for ADGE and 87 % for UMFR. ADGI could be replaced by ADGE within the breeding program in order to achieve the special yield characteristics of the GWM breed. In addition, profitability and competitiveness could be positively influenced compared to conventional sheep breeds in the same environment due to an increased average daily gain based on extensive feed at the dykes. On extensive grassy landscapes, the GWM breed may have higher weight gain by comparison with conventional, intensive mutton breeds - 91 -
Chapter Five ______
(e.g. Texel and German Black-Headed Mutton). With an implemented trait for an unbiased average daily gain, the selection response for the true average daily gain (ADGE) could be processed and improved by breeders and, thus, the competitiveness and genetic gain of this local breed could probably increase. Furthermore, an implementation of ADGE would have an impact on meat-quality aspects due to positive correlations between ADGE and UMFR (Table
4 and Figure 3). Consequently, a positive selection response can be expected for both traits at the same time while implementing one or the other trait (ADGE or UMFR) within a breeding program. This implementation of a meat-quality trait (UMFR) may lead to enhanced profitability and a unique selling position due to the selection response of increased meat quality. However, the implementation of ADGE has a negative impact on the conventional trait of MUSC (Table 4). This conventional trait underlies high subjectivity during the live valuation of animals and has no objective data collection. Beyond, MUSC has a strong dependency on ADGI, as already mentioned. The trait of WOL has no economic value for sheep breeders in Germany. Subsequently, negative consequences for WOL would not sustainably threaten the local GWM breed and might be not fatal for farmers. Additionally, the monetary value for wool from land sheep breeds is minor due to the rough wool fibre.
CONCLUSION
The EBVs of ADGI and ADGE were tendentially negatively correlated, which strengthen the assumption of a biased ADGI. ADGE probably reflects the trait of average daily gain objectively and unbiased under normal environmental conditions in case of the GWM breed.
Furthermore, the utilization of ADGE has a positive impact on meat-quality aspects (UMFR).
With the optimal use of reference sires and predefined selection intensity it is possible to achieve genetic response for ADGE and UMFR with simultaneously acceptable rates of inbreeding. Implemented novel traits allow selection on breed-specific features, unique native genetic variance, and native genetic diversity. In addition, the profitability may increase due
- 92 -
Chapter Five ______to the selection of these economic and breed-specific traits, which enable increased competitiveness compared to common mutton breeds in the same environment.
Acknowledgement: Financial support from the ministry of Energy, Agriculture,
Environment, Nature, and Digitalization within the framework of the European Innovation
Partnership (EIP Agri) is gratefully acknowledged. In addition, the authors give special thanks to the farmers and breeding organization supporting this project faithfully.
REFERENCES
Boettcher PJ, Tixier-Boichard M, Toro MA, Simianer H, Eding H, Gandini G, Joost S, Garcia
D, Colli L, Ajmone-Marsan P, the GLOBALDIV Consortium (2010) Objectives,
criteria and methods for using molecular genetic data in priority setting for conservation
of animal genetic resources. Animal Genetics 41(1):64-77
Brien C, R (2016) asremlPlus: Augments the Use of ‘ASReml-R‘ in Fitting Mixed Models.
https://cran.r-project.org/web/packages/asremlPlus/asremlPlus.pdf
Butler DG, Cullis BR, Gilmour AR, Gogel BJ (2009) ASReml-R reference manual. State of
Queensland, Department of Agriculture, Fisheries and Forestry
Colleau JJ, Avon L (2008) Sustainable long-term conservation of rare cattle breeds using
rotational AI sires. Genetics Selection Evolution 40:415-432
Dohm MR (2002) Repeatability estimates do not always set an upper limit to heritability.
Functional Ecology 16:273-280
Falconer DS, Mackay TFC (1996) Introduction to Quantitative Genetics. Longman House,
Essex, UK
FAO (2007) The state of the World´s Animal genetic Resources for Food and Agriculture,
edited by B. Rischkowsky & D. Pilling. Rome (available at
www.fao.org/docrep/010/a1250e/a1250e00.html)
- 93 -
Chapter Five ______
FAO (2010) Breeding strategies for sustainable management of animal genetic resources.
FAO Animal Production and Health Guidelines. No. 3, Rome
Fernández J, Meuwissen THE, Toro MA, Mäki-Tanila A (2011) Management of genetic
diversity in small farm animal populations. Animal 5(11):1684-1698
Gandini G, Oldenbroek K (2007) Strategies for moving from conservation to utilisation. In:
K. Oldenbroek (ed.), Utilisation and Conservation of Farm Animal Genetic Resources,
pp. 29-54. Wageningen Academic Publishers, Wageningen
Gilmour AR, Luff AF, Fogarty NM, Banks R (1994) Genetic parameters for ultrasound fat
depth and eye muscle measurements in live Poll Dorset sheep. Australian Journal of
Agricultural Research 45(6):1281-1291
Gowane GR, Prince LLL, Lopes FB, Paswan C, Sharma RC (2015) Genetic and phenotypic
parameter estimates of live weight and daily gain traits in Malpura sheep using
Bayesian approach. Small Ruminant Research 128:10-18
Groeneveld LF, Lenstra JA, Eding H, Toro MA, Scherf B, Pilling D, Negrini R, Finlay EK,
Jianlin H, Groeneveld E, Weigend S, The GLOBALDIV Consortium (2010) Genetic
diversity in farm animals – a review. Animal Genetics 41(1):6-31
Hassen Y, Fuerst-Waltl B, Sölkner S (2003) Genetic parameter estimates for birth weight,
weaning weight and average daily gain in pure and crossbred sheep in Ethiopia. Journal
of Animal Breeding and Genetics 120(1):29-38
Hospital F (2005) Selection in backcross programmes. Transactions of the Royal Society B –
Biological Sciences 360:1503-1511
Larsgard AG, Olesen I (1998) Genetic parameters for direct and maternal effects on weights
and ultrasonic muscle and fat depth of lambs. Livestock Production Science 55:273-278
Lessells CM and Boag PT (1987) Unrepeatable Repeatabilities: A Common Mistake. Auk,
104(1):116-121
- 94 -
Chapter Five ______
Lewis RM, Simm G (2000) Selection strategies in sire referencing schemes in sheep.
Livestock Production Science 67:129-141
María GA, Boldman KG, Van Vleck LD (1993) Estimates of variances due to direct and
maternal effects for growth traits of Romanov sheep. Journal of Animal Science
71(1):845-849
Meuwissen THE (1997) Maximizing the response of selection with a predefined rate of
inbreeding. Journal of Animal Science 75:934-940
Meuwissen THE (2009) Genetic management of small populations: A review. Acta
Agriculturae Scandinavica, Section A – Animal Science 59(2):71-79
Meuwissen THE, Sonesson AK (2004) Genotype-assisted optimum contribution selection to
maximize selection response over a specified time period. Genet. Res. 84:109-116
Nielsen HM, Sonesson AK, Meuwissen THE (2014) Optimum contribution selection using
traditional best linear unbiased prediction and genomic breeding values in aquaculture
breeding schemes. Journal of Animal Science 89:630-638
Odegard J, Yazdi MH, Sonesson AK, Meuwissen THE (2009) Incorporating desirable genetic
characteristics from an inferior into a superior population using genomic selection.
Genetics 181:737-745
Putman RJ (1995) Ethical considerations and animal welfare in ecological field studies.
Biodiversity and Conservation 4:903-915
R Core Team (2018) R: A Language and Environment for Statistical Computing. R
Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org
Rendel JM, Robertson A (1950) Estimation of genetic gain in milk yield by selection in a
closed herd of diary cattle. Journal of Genetics 50(1):1-8
Safari E, Fogarty NM, Gilmour AR (2005) A review of genetic parameter estimates for wool,
growth, meat and reproduction traits in sheep. Livestock Production Science 92:271-
289
- 95 -
Chapter Five ______
Shepherd RK, Woolliams JA (2004) Minimising inbreeding in small populations by rotational
mating with frozen semen. Genetics Research 84:87-93
Sonesson AK, Goddard ME, Meuwissen THE (2002) The use of frozen semen to minimize
inbreeding in small populations. Genetical Research 80:27-30
Toro MA, Silió L, Woolliams JA (1990) Selection for lean meat production in sheep: a
simulation study. In: Proceedings 4th World Congress on Genetics Applied to Livestock
Production, Vol. 15:96-99
Villanueva B, Simm G, Woolliams JA (1995) Genetic progress and inbreeding for alternative
nucleus schemes for beef cattle. Animal Science 61:231-239
Villanueva B, Bijma P, Woolliams JA (2000) Optimal mass selection policies for schemes
with overlapping generations and restricted inbreeding. Genetics Selection Evolution
32:339-355
Wang Y, Bennewitz J, Wellmann R (2017) Novel optimum contribution selection methods
accounting for conflicting objectives in breeding programs for livestock breeds with
historical migration. Genetics Selection Evolution 49:45
Wellmann R, Hartwig S, Bennewitz J (2012) Optimum contribution selection for conserved
populations with historic migration. Genetics Selection Evolution 44(1):34
Wellmann R, R package version 2.0. (2018) OptiSel: optimum contribution selection and
population genetics. https://cran.r-project.org/web/packages/optiSel/optiSel.pdf
- 96 -
Chapter Six ______
Chapter Six
Exploration of conservation and development strategies for
local cattle breeds in Northern Germany
J. Schäler1, S. Addo1, G. Thaller1 & D. Hinrichs2
1Institute of Animal Breeding and Husbandry, Christian-Albrechts-University of Kiel, D-24098 Kiel, Germany 2Faculty of Organic Agricultural Sciences, Department of Animal Breeding, University of Kassel, Nordbahnhofstraße 1a, D-37213 Witzenhausen, Germany
Submitted for Publication - 97 -
Chapter Six ______
ABSTRACT
Many local breeds have become endangered due to their substitution by more high-yielding breeds. To conserve these local breeds, effective development strategies need to be investigated. The aim of this study was to explore conservation and development strategies based on quantified strengths, weaknesses, opportunities, and threats for two local cattle breeds from Northern German, namely the German Angler (GA) and Red Dual-Purpose cattle
(RDP). The data comprised 158 comprehensive questionnaires regarding both breeds’ strengths, weaknesses, opportunities, and threats, which were answered by 78 farmers of GA and 80 farmers of RDP. First, data were analysed using the SWOT-AHP method, which combines the qualitative strategic decision tool of SWOT (analysis of strengths, weaknesses, opportunities, and threats) and the quantitative tool of AHP (Analytic Hierarchy Process).
Second, prioritized SWOT factors were discussed with stakeholders in order to form final conservation and development strategies at breed level. For GA prioritized strengths were daily gain, meat quality, milk production, and the usage of new biotechnologies, weaknesses were genetic gain and inbreeding, opportunities were organic farming and breed-specific characteristics, and threats were milk prices and dependency regarding the dairy business.
Consequently, three conservation and development strategies were formed: 1) Changing relative weights and the relevant breeding goal to drift from milk to meat, 2) Increasing genetic gain and control the rate of inbreeding by the implementation of specific selection programs, and 3) Selection of unique and breed characteristic components on product level, i.e. milk-fat and fine muscle fibers. For RDP defined strengths were robustness, high adaptability for different housing systems, and a balanced dual-purpose of milk and meat, weaknesses were inbreeding, breed extinction, genomic selection with young bulls, and milk yield, opportunities were organic farming and dual-purpose aspects, and threats were milk and decreasing beef cattle prices. Thus, three conservation and development strategies were identified: 1) Adjust relative weights and the relevant breeding goal to balance milk and meat
- 98 -
Chapter Six ______yield, 2) Increasing genetic gain and avoid extinction by implementing targeted selection programs, and 3) Selection of unique and breed characteristic traits on breed level, i.e. environmental robustness. Quantified strengths, weaknesses, opportunities, and threats establish an ideal basis for the exploration of conservation and development strategies at breed level with less subjectivity. Explored strategies are objective even if the stakeholder approach was limited for small populations. Individual convenience of the farmers was better considered than a quantitative strategy decision tool on its own.
Keywords: conservation, local breed, strategy development, SWOT-AHP method
IMPLICATIONS
This study explores conservation and development strategies for local cattle breeds based on farmer surveys. Quantified strengths, weaknesses, opportunities, and threats can be used as an ideal basis to form objective strategies at breed level. This combined approach result in less subjective strategies even with a limited number of stakeholders. Described approach represents farmers’ individual convenience better than existing quantitative strategy decision tools on their own.
- 99 -
Chapter Six ______
INTRODUCTION
Many local breeds have been replaced by high-yielding breeds over the last few centuries, resulting in loss of local breeds (Meuwissen 2009; FAO 2010). Due to this breeding history, populations of many local breeds have dangerously decreased and some of them are even threatened by extinction (Fernández et al. 2011). By now two local cattle breeds from
Northern Germany are listed as endangered livestock breeds by ‘The Society for the
Conservation of Old and Endangered Livestock Breeds’ (GEH). Both breeds are red cattle breeds with a milk-emphasized dual purpose, namely the German Angler and the Red Dual-
Purpose cattle. According to Meuwissen (2009), the best conservation strategies are an increased profitability achieved by genetic improvements and the promotion of breed-specific products. The FAO (2010) suggests a guideline to develop breeding strategies for the sustainable management of animal genetic resources also based on the implementation of effective genetic gain programs. However, Martín-Collado et al. (2013) explores conservation and development strategies for local European cattle breeds, which focused production systems and the marketing of new products. Thus, considered aspects in case of effective conservation strategies may be more widespread than just prioritizing genetic gain.
Furthermore, strategies embracing many breeds should refer to general issues, whereas specific strategies and actions for single breeds should be identified at breed level (Martín-
Collado et al. 2013). Investigated actions at breed level may be not only convenient for single breeds, but also suggestive in case of a limited number of stakeholders during the strategy decision process especially for small endangered populations. The aim of the present study was to explore conservation and development strategies at breed level based on quantified strengths, weaknesses, opportunities, and threats for two local cattle breeds.
- 100 -
Chapter Six ______
MATERIALS AND METHODS
Data
Data were obtained from 158 comprehensive questionnaires completed by cattle breed farmers, divided into 78 farmers who rear German Angler (GA) and 80 farmers who rear Red
Dual-Purpose cattle breed (RDP). Both cattle breeds are local in Schleswig-Holstein, North
Germany. The design and sending of the questionnaires as well as the collection of completed surveys were done by the breeding organisation RSH (Rinderzucht Schleswig-Holstein e.G.,
Germany). Each questionnaire consisted of a total of 12 queries, open-ended and multiple- choice questions including information on sub-items, such as farm, herd, expectations, reproduction, traits, and difficulties as well as handwritten farmers’ personal opinions on the breeds’ strengths, weaknesses, opportunities, and threats (Supplementary Material S1).
Quantified strategy decision tool
To identify strategies using an organised approach, the FAO (2010) suggested employing the
SWOT analysis (analysis of strengths, weaknesses, opportunities, and threats) from Weihrich
(1982). A SWOT analysis is used to identify strengths, weaknesses, opportunities, and threats in order to enhance the profitability of an individual production system. This qualitative method has the disadvantage of high subjectivity during its application in decision-making
(Hill and Westbrook 1997; Pesonen et al. 2000; Martín-Collado et al. 2013). Therefore,
Kurttila et al. (2000) and Saaty and Vargas (2001) developed a combination of the AHP method (Analytic Hierarchy Process), which is a quantitative tool from Saaty (1980), and the
SWOT analysis to obtain more complex and reliable decisions. This hybrid method improves the quantitative information basis and forces the decision-maker to think over and analyse the situation more precisely and in more depth (Kurttila et al. 2000). SWOT-AHP methodology has been applied in the fields of forestry (Kangas et al. 2001), silvopasture adoption (Shreshta et al. 2004), and livestock (Wasike et al. 2011).
- 101 -
Chapter Six ______
In the current study the SWOT-AHP was used to quantify identified strengths, weaknesses, opportunities, and threats. Statistical analyses and computations for eigenvalues, priority vectors, and quality control parameters within the SWOT-AHP methodology were performed by using the R-software (R Core Team 2018). The SWOT-AHP approach was divided into five steps (Figure 1) and is described below:
I) Assignment of items
The farmers’ responses were implemented and sorted as items by an expert team into four single SWOT groups of strengths, weaknesses, opportunities, and threats. This expert team consisted of four scientists from two national, German animal breeding departments. All team members possessed comprehensive knowledge and professional experience in dealing with these local cattle breeds. The expert team focused on interactions and knowledge transfer with farmers, breeding organizations, and professional cattle breed associations. Experts discussed their attitudes and valuations as soon as they had reached consensus. The number of items and
SWOT factors implemented for each SWOT group by the team of experts exhibited substantial flexibility and was allowed to differ between both breeds.
II) Defining SWOT factors
The farmers’ items were defined as SWOT factors by the expert team (Table 1; Table 2).
Generally, the number of SWOT factors within each SWOT group can vary depending on the number of implemented items.
III) Computation of priority scores for SWOT factors
The computations of the factor priority vector (FPV), the group priority vector (GPV), and overall priority vector (OPV) were performed with AHP rankings by the named expert team above based on the frequency of SWOT factors among farmers. The fundamental scale from
Saaty (1980) was used for such AHP rankings (Table 3). This scale represented a transformation from verbal judgements into numerical judgements to determine the relative importance of each element. There were five verbal judgements, which were ranked by their
- 102 -
Chapter Six ______importance from equal to extreme (equal, moderate, strong, very strong, and extreme). These judgements were translated into the numeric values of: 1, 3, 5, 7, and 9 with four intermediate values (2, 4, 6, 8) for compromises in importance.
Table 3 Fundamental scale (Saaty, 1980; Yüksel and Dağdeviren, 2007)
Intensity of Definition Explanation importance 1 Equal importance Two criteria contribute equally to the objective
3 Moderate importance Experience and judgement slightly favour one over another 5 Strong importance Experience and judgement strongly favour one over another 7 Very strong importance Activity is strongly favoured and its dominance is demonstrated in practice 9 Extreme importance Importance of one over another affirmed at the highest possible order 2, 4, 6, 8 Intermediate values Represent compromise between the priorities listed above
In other words, the scale indicated how important or dominant one element was over another element (Saaty 2008). The elements for FPV were individual SWOT factors within each
SWOT group and for GPV the four single SWOT groups. The elements were ranked within
AHP matrices of different levels: Level 1 for the SWOT factors within each SWOT group and
Level 2 for the SWOT groups (Figure 1). Kahraman et al. (2007), Borajee and Yakchalie
(2011), and Görener et al. (2012) defined the set of elements as E = {Ej | j = 1,2, … , n}. The results of the n ranked elements were resumed in an evaluation matrix A (n ∗ n). Each element aij (i, j = 1,2, … , n) was the quotient of the weights of the elements wij (i, j =
1,2, … , n). In this matrix, the element aij = 1 when the weights of elements wi = wj. The ranked elements were also expressed via a square and reciprocal matrix:
1 푤1/푤2 … 푤1/푤푛 . . . . 푤 A = (a ) = 푤 /푤 1 ⋯ 푤 /푤 ; a = 푖 , a ≠ 0. (1) ij 2 1 2 푛 ij 푤 ij ⋮ ⋮ ⋱ ⋮ 푗 [푤푛/푤1 푤푛/푤2 ⋯ 1 ]
- 103 -
Chapter Six ______
Each matrix was normalised and their relative weights (A푤) were determined. The relative weights were given by the correct eigenvector (푤) corresponding to the largest eigenvalue
(λmax), as:
Aw = λmax ∗ 푤. (2)
The largest eigenvalue (λmax) was computed by forming the sum of the single normalised eigenvector (푤) per row multiplied by the computed priority vector per corresponding column:
∑ij 푤(column) λmax = ∑(∑ij 푤(row) ∗ ). (3) ∑(∑ij 푤(column))
If the pairwise comparisons were completely consistent, the matrix A had the rank 1 and
λmax = 푛 . In this case, weights could be obtained by normalising any of the rows or columns of A. The priority vectors of FPV and GPV were computed by dividing the single normalised eigenvectors (푤) per column by the sum of all single normalised eigenvectors per column for the single matrices:
∑ 푤 FPV / GPV = ij (column) . (4) ∑(∑ij 푤(column))
It should be noted that the quality of the results was related to the consistency of the comparison judgements. As a quality control, the inconsistency ratio (iCR) of the ranked elements was computed as a ratio between the consistency index (CI) and random index (RI):
CI iCR = . (5) RI
The iCR provided information on the inconsistency of the ranked elements. An iCR ≤ 0.1 was acceptable as a criterion (Kahraman et al. 2007; Borajee and Yakchali 2011). Thus, the ranking of the elements had to be repeated by the team of experts when the iCR exceeded this threshold. The consistency index (CI) was computed by using the subsequent formula:
λ −푛 CI = max . (6) 푛−1
- 104 -
Chapter Six ______
The random index (RI) was defined as an expected value of the CI depending on an individual number of ranked elements 푛. These fixed values were obtained from the study of Aguarón and Moreno-Jiménez (2003) for further computations of the CI parameter (Table 4). The authors simulated 100,000 matrices for several sizes of n and calculated standard values for
RI.
Table 4 Fixed values of random index (RI) dependent on the size of matrices (Aguarón and Moreno-Jiménez 2003) 푛 1 2 3 4 5 6 7 8 RI 0.00 0.00 0.525 0.882 1.115 1.252 1.341 1.404 푛 9 10 11 12 13 14 15 16 RI 1.452 1.484 1.513 1.535 1.555 1.570 1.583 1.595 풏, ranked elements RI, random index
IV) Calculation of group priority vector (퐺푃푉)
The GPV was computed for each corresponding SWOT group by the expert team based on the frequency of farmers’ items included in each SWOT group. Computation details for the GPV are described in step III) Computation of priority scores for SWOT factors.
V) Calculation of overall priority vector (푂푃푉)
However, the overall priority vectors (OPV) were calculated by multiplying the single FPVs of the single SWOT factors with the corresponding GPVs of each SWOT group:
OPV = GPV ∗ FPV. (7)
The number of computed OPVs should correspond with the number of computed FPVs.
- 105 -
Chapter Six ______
Fig. 1 Methodical implementation of Analytic Hierarchy Process (AHP) based on farmer surveys for local cattle breeds
Conservation and development strategies at breed level
In general, SWOT strategies embracing single breeds should be identified at breed level
(Martín-Collado et al. 2013) and not just at the team of expert’ level. Furthermore, the identification of SWOT factors is often unbalanced (Karppi et al. 2001) and a multi- stakeholder approach was preferred in several studies to avoid subjectivity (Impoinvil et al.
2007; Martín-Collado et al. 2013). Especially, in small endangered populations the number of stakeholders or experts is probably limited and a multi-stakeholder approach for single local breeds may be not achievable. Quantified strengths, weaknesses, opportunities, and threats have been identified by applying the quantified strategy decision tool of SWOT-AHP in the step before. To overcome subjectivity due to a limited stakeholder approach, quantified
SWOT factors with the three highest OPVs per SWOT group were comprehensively discussed face to face with farmers, breeders, and staff of the breeding organization (breeding board). Then discussions, ideas, and solutions were protocolled and collectively transformed together into less subjective conservation and development strategies at breed level.
- 106 -
Chapter Six ______
RESULTS
Quantified strategy decision tool
Farmer surveys were implemented by the expert team and resulted in a total of 87 items, whereby 42 belonged to GA farmers and 45 to RDP farmers. For GA, all 42 items were divided and 15 items were subsequently sorted into the SWOT group of strengths, 7 items into weaknesses, 12 into opportunities, and 8 into threats (Table 1; Figure 2a). For RDP, all
45 items were divided and assigned as amounts of 16, 6, 14, and 9 into strengths, weaknesses, opportunities, and threats, respectively (Table 2; Figure 2a). The strengths comprised the highest number of items, whereas the weaknesses contained the lowest number of items for both breeds.
Fig. 2 In a): Frequency of implemented items and b): computed group priority vectors (GPV) via SWOT group for German Angler (GA) and Red Dual-Purpose cattle breed (RDP)
The three highest FPVs for the strengths were 0.141 for SWOT factor number 8 (S8), 0.101
(S12), and 0.101 (S13) for GA (Table 1; Figure 3a) and 0.120 (S9), 0.110 (S14), and 0.100 (S11) for RDP. The three highest FPVs in weaknesses were 0.316 (W6), 0.226 (W7), and 0.129 (W4) for GA and 0.415 (W4), 0.175 (W5), and 0.153 (W6) for RDP. The three highest FPVs in opportunities were 0.142 (O2), 0.113 (O8), and 0.111 (O7) for GA and 0.110 (O3), 0.109 (O12), and 0.107 (O6) for RDP, and in threats 0.230 (T3), 0.170 (T4), and 0.132 (T5) for GA and
0.230 (T6), 0.164 (T3), and 0.146 (T5) for RDP (Table 2; Figure 3a). For GA, the computed
GPVs were 0.375 for strengths, 0.162 for weaknesses, 0.278 for opportunities, and 0.188 for threats (Table 1; Figure 2b). However, for RDP, the calculated GPVs were 0.360 for
- 107 -
Chapter Six ______strengths, 0.137 for weaknesses, 0.308 for opportunities, and 0.188 for threats (Table 2;
Figure 2b).
The three highest OPVs were 0.053 (S8), 0.038 (S12), and 0.038 (S13) for GA and 0.043 (S9),
0.040 (S14), and 0.036 (S11) in strengths for RDP. The three highest OPVs were 0.051 (W6),
0.036 (W7), and 0.021 (W4) for GA and 0.057 (W4), 0.024 (W5), and 0.021 (W6) for RDP in weaknesses, 0.039 (O2), 0.031 (O8), 0.031 (O9) for GA and 0.034 (O3), 0.034 (O12), and 0.033
(O6) for RDP in opportunities, and 0.043 (T3), 0.032 (T4), and 0.025 (T5) for GA and 0.045
(T6), 0.032 (T3), 0.028 (T5) for RDP in threats (Table 1; Table 2). The three highest OPVs within each SWOT group were highlighted in the graph for each local breed (Figure 3b).
Fig. 3 Computed a): factor priority vector (FPV) and b): overall priority vector (OPV) via SWOT factor for German Angler (GA) and Red Dual-Purpose cattle (RDP). SWOT factors with three highest OPVs are highlighted (red) for each SWOT group
The iCR values for GA and RDP are shown in Figure 4. The iCR of the ranked SWOT factors within the four single SWOT groups was 0.080 for GA in strengths (iCRS), 0.066 in weaknesses (iCRW), 0.078 in opportunities (iCRO), and 0.033 in threats (iCRT). The iCR of the ranking between the SWOT groups (iCRG) was 0.001. For RDP, the iCR within each group was 0.095 in strengths (iCRS), 0.043 in weaknesses (iCRW), 0.078 in opportunities - 108 -
Chapter Six ______
(iCRO), and 0.060 in threats (iCRT). The iCR of the ranked four individual SWOT groups
(iCRG) was 0.027.
Fig. 4 Quality control of comparison judgements via ranked elements. The threshold for the inconsistency ratio (iCR) is highlighted (red dotted line)
Conservation and development strategies at breed level
For GA prioritized strengths of daily gain, meat quality, milk production, and the usage of new biotechnologies were discussed. Furthermore, quantified weaknesses of genetic gain, especially for milk yield, and high rates of inbreeding have been comprehensively considered.
In addition, opportunities of organic farming and breed-specific characteristics as well as threats of low milk prices and a high dependency regarding the dairy business were debated in order to explore effective conservation and development strategies at breed level.
Three final conservation and development strategies have been formed for the GA:
1) Changing relative weights and the relevant breeding goal to drift from milk to meat,
2) Increasing genetic gain and control the rate of inbreeding by the implementation of
specific selection programs,
3) Selection of unique and breed characteristic components on product level, i.e. milk-
fat and fine muscle fibers.
However, for RDP defined strengths were robustness, high adaptability for different housing systems, and a balanced dual-purpose of milk and meat. Quantified weaknesses were - 109 -
Chapter Six ______inbreeding, breed extinction, genomic selection with young bulls, and milk yield.
Additionally, opportunities of organic farming and dual-purpose aspects as well as threats of low milk and decreasing beef cattle prices have been regarded during the strategy developments.
For RDP three final conservation and development strategies have been investigated:
1) Adjust relative weights and the relevant breeding goal to balance milk and meat
yield,
2) Increasing genetic gain and avoid extinction by implementing targeted selection
programs,
3) Selection of unique and breed characteristic traits on breed level, i.e. environmental
robustness.
DISCUSSION
Quantified strategy decision tool
For both breeds, the quality control parameters iCRG, iCRS, iCRW, iCRO, and iCRT were ≤ 0.1
(Figure 4). Hence, the quality of all comparison judgements of the ranked elements and the
SWOT-AHP analysis in general were consistent. Nevertheless, the ratio between the ranked elements and iCR showed a strong dependency. The inconsistency ratio was enhanced by an increased number of ranked elements. This suggests the iCR threshold may consider the number of ranked elements to overcome limitation problems of ranked elements in order to make the AHP method more comprehensive and flexible.
Conservation and development strategies at breed level
Emphasize on meat traits will be enhanced in the breeding goal for both breeds in order to drift from milk to meat in a balanced way. Therefore, economic and biological weights as well as correlations between these traits have to be analyzed in advance. Especially for the investigation of actual economic weights product prices have to be comprehensively collected
- 110 -
Chapter Six ______and compared on markets. However, measured additive genetic variance between commercial traits is of special importance in order to assess biological weights carefully. Thus, the identification of relative weights and the final breeding goal could be indicated by using the appropriate contingent value method from Davis (1963).
The adaptation of special breeding approaches like the Optimum Contribution Selection
(Meuwissen 1997; Wellmann et al. 2012) were prospectively proposed for these local breeds in order to increase genetic gain and simultaneously restrict the rate of inbreeding.
Consequently, inbreeding depression and breed extinction can be avoided through sustainable mating strategies and genetic improvement. Therefore, the breeding board suggests bulls for mating at farm level. At this, it has major relevance that farmers follow such suggestions and attend their mating records faithfully.
Focusing unique and breed characteristic components will result in high quality products, which can be offered on niche markets with higher earnings. Therefore, selection on such specific traits will be intensified and leads to an added value on product level (Verrier 2018;
Bernués et al. 2018; Martín-Collado et al. 2018; Hiemstra et al. 2018). In case of the GA, these added values on product level are milk fat (5.5 %) and fine muscle fibers as meat quality trait. However, for RDP the added value will be not on product level but rather on the production system level following the approach from Verrier (2014) and the study of Schäler et al. (2018), where unique traits were investigated within the respective natural production conditions. In case of the RDP, such added value on production system level is robustness in harsh environments. The competitiveness of the RDP may increase regarding pasture feeding and health traits compared to commercial breeds under the same environmental conditions, e.g. Holstein Friesian. Identified added values on product and production level will be also suitable especially for local or organic agriculture production schemes. Hence, dependency on commercial milk markets and prices is effectively reduced due to concentration of niche markets, the supply of high quality products, and a balanced milk/meat relation.
- 111 -
Chapter Six ______
An implementation of all conservation and development strategies is planned for each breed.
According to Gandini and Oldenbroek (2007) and Meuwissen (2009), it can be concluded that such explored strategies, which include the definition of relevant breeding goals, novel marketing products, and genetic improvements, are adequate to move further from conservation to utilisation. Explored conservation and development strategies indicate less subjective, more versatile, and more farmer- and consumer-oriented strategies at breed level compared to quantitative decision tools on their own, even when the stakeholder approach was limited.
CONCLUSION
Quantified strengths, weaknesses, opportunities, and threats establish an ideal basis for the exploration of conservation and development strategies at breed level. Therefore, quantified
SWOT factors have to be comprehensively discussed with a copious number of stakeholders in order to avoid subjectivity. This integrated approach result in strategies which are realistic, objective, and consider individual convenience of the farmers’ more than a quantitative strategy decision tool on its own.
Acknowledgement: Financial support from the Ministry of Energy, Agriculture,
Environment, Nature, and Digitalisation within the framework of the European Innovation
Partnership (EIP Agri) is gratefully acknowledged. In addition, the authors give special thanks to the large number of farmers faithfully supporting this project. We would like to extend our sincere appreciation to the breeding organisation (RSH e.G.) and to Arche Warder - Centre of rare livestock breeds, which is the leading partner in this EU project.
REFERENCES
Aguarón J, Moreno-Jiménez J (2003) The geometric consistency index: Approximated
thresholds. European Journal of Operational Research 147:137-145
- 112 -
Chapter Six ______
Bernués A (2018) Quantifying ecosystem services to add value in pasture-based livestock
systems. 69th EAAP Book of Abstracts No. 24:101
Boraje M, Yakchali SH (2011) Using the AHP-ELECTRE III integrated method in a
competitive profile matrix. In: IPEDR 11, International Conference on Financial
Management and Economics, pp. 68-72, IACSIT press, Singapore
Davis RK (1963) The value of outdoor recreation: an economic study of the Maine woods.
PhD Thesis, Harvard University
FAO (2010) Breeding strategies for sustainable management of animal genetic resources.
FAO Animal Production and Health Guidelines. No. 3, Rome
Fernández J, Meuwissen THE, Toro MA, Mäki-Tanila A (2011) Management of genetic
diversity in small farm animal populations. Animal 5(11):1684-1698
Gourdine JL, Sørensen AC, Rydhmer L (2012) There is room for selection in a small local pig
breed when using optimum contribution selection: A simulation study. Journal of
Animal Science 90:76-84
Görener A, Toker K, Uluçay K (2012) Application of Combined SWOT and AHP: A Case
Study for a Manufacturing Firm. Procedia - Social and Behavioural Science 58:1525-
1534
Gandini G, Oldenbroek K (2007) Strategies for moving from conservation to utilisation. In:
K. Oldenbroek (ed.), Utilisation and Conservation of Farm Animal Genetic Resources,
pp. 29-54. Wageningen Academic Publishers, Wageningen
Hiemstra SJ, Remjin N, Hoving-Bolink AH, Amat L, Traon D (2018) Valorization of locally
produced dairy and meat products from native livestock breeds. 69th EAAP Book of
Abstracts No. 24:102
Hill T, Westbrook R (1997) SWOT analysis: it’s time for a product recall. Long range
Planning 30:46-52
- 113 -
Chapter Six ______
I Impoinvil DE, Ahmad S, Troyo A, Keating J, Githeko AK, Mbogo CM, Kibe L, Githure
JI, Gad AM, Hassan AN, Orshan L, Warburg A, Calderón-Arguedas O, Sánchez-Loria
VM, Velit-Suarez R, Chadee DD, Novak RJ, Beier JC (2007) Comparison of
mosquito programs in seven urban sites in Africa, the Middle East and the Americas.
Health Policy 83:196–212
Kahraman C, Demirel NÇ, Demirel T (2007) Prioritization of e-Government strategies using
a SWOT-AHP analysis: the case of Turkey in Turkey. European Journal of Information
Systems 16:284-298
Kangas J, Pesonen M, Kurttila M, Kajanus M (2001) A’WOT: Integrating the AHP with
SWOT Analysis. In: Proceedings - 6th ISAHP 2001, International Symposium on the
Analytic Hierarchy Process, pp. 189-198. ISAHP 2001, Berne
Karppi I, Kokkonen M, Lähteenmäki-Smith K (2001) SWOT-analysis as a basis for regional
strategies. Working Paper 2001, 4. Nordregio-Nordic Centre for Spatial Development,
Stockholm, Sweden
Kurttila M, Pesonen M, Kangas J, Kajanus M (2000) Utilizing the analytic hierarchy process
(AHP) in SWOT analysis - a hybrid method and its application to a forest-certification
case. Forest Policy and Economics 1:41-52
Martín-Collado D, Díaz C, Mäki-Tanila A, Colinet F, Duclos D, Hiemstra SJ, Eureca
Consortium, Gandini G (2013) The use of SWOT analysis to explore and prioritize
conservation and development strategies for local cattle breeds. Animal 7(6):885-894
Martín-Collado D, Díaz C, Serrano M, Zanoli R, Naspeti S (2018) A critical perspective on
the current paradigm of high-quality products marketing strategy. 69th EAAP Book of
Abstracts No. 24:102
Meuwissen THE (1997) Maximizing the response of selection with a predefined rate of
inbreeding. Journal of Animal Science 75:934-940
- 114 -
Chapter Six ______
Meuwissen THE, Hayes BJ, Goddard ME (2001) Prediction of Total Genetic Value Using
Genome-Wide Dense Marker Maps. Genetics 157:1819-1829
Meuwissen THE (2009) Genetic management of small populations: A review. Acta
Agriculturae Scandinavica, Section A – Animal Science 59(2):71-79
Pesonen M, Kurttila M, Kangas J, Kajanus M, Heinonen P (2000) Assessing the priorities
using A´WOT among resource management strategies at the Finish Forest and Park
Service. Forest Science 47(4):534-541
R Core Team (2018) R: A Language and Environment for Statistical Computing. R
Foundation for Statistical Computing, Vienna, Austria. https://www.R-project.org
Saaty TL (1980) How to make a decision: The Analytic Hierarchy Process. European Journal
of Operational Research 48:9-26
Saaty TL (2008) Decision making with the analytic hierarchy process. International Journal of
Services Sciences 1(1):83-98
Saaty TL, Vargas LG (eds) (2001) Models, Methods, Concepts and Applications of the
Analytic Hierarchy Process. Kluwer Academic Publishers, Boston
Schäler J, Thaller G, Hinrichs D (2018) Implementation of breed-specific traits for a local
sheep breed. Agricultural Sciences 9(8):958-973
Shreshta RK, Alavalapati JRR, Kalmbacher RS (2004) Exploring the potential for
silvopasture adoption in south-central Florida: an application of SWOT-AHP method.
Agricultural Systems 81:185-199
Verrier E (2014) Some success factors for development and marketing of local breeds
products. Opportunities for conservation of local breeds – WIAS Seminar, Wageningen.
https://www.wur.nl/upload_mm/7/f/7/d779bd35-3b92-4d17-9261-
cd506db9f37d_Seminar_WIAS_Verrier_2014-12-09.pdf
- 115 -
Chapter Six ______
Wasike CB, Magothe TM, Kahi AK, Peters KJ (2011) Factors that influence the efficiency of
beef and dairy cattle recording system in Kenya: A SWOT-AHP analysis. Tropical
Animal Health and Production 43(1):141-152
Weihrich H (1982) The TOWS Matrix - A Tool for Situational Analysis. Long Range
Planning 15 (2):54-66
Wellmann R, Hartwig S, Bennewitz J (2012) Optimum contribution selection for conserved
populations with historic migration. Genetics Selection Evolution 44(1):34
Yüksel I, Dağdeviren M (2007) Using the analytic network process (ANP) in a SWOT
analysis - A case study for a textile firm. Information Science 177:3364-3382
- 116 -
Chapter Six ______
Table 1 Priority scores of ranked SWOT factors for German Angler (GA) SWOT group GPV SWOT factors FPV OPV 0.375 0.050 0.018 S1: the majority of farmers have a stable GA livestock 0.375 0.053 0.019 S2: over 90 % of farmers use GA as main source of income 0.375 0.056 0.021 S3: the majority of farmers produce their own forage for GA 0.375 0.059 0.022 S4: mortality rate of GA calves is low 0.375 0.027 0.010 S5: good adaptation of GA to marsh land 0.375 0.065 0.024 S6: claws and legs of GA are resilient
0.375 0.037 0.013 S7: GA has high resilience to climate 0.375 0.141 0.053 S8: GA is a milk-emphasized, dual-purpose breed with good daily gain
Strengths 0.375 0.041 0.015 S9: GA heifers are early-maturing 0.375 0.098 0.036 S10: GA has good fertility, udder health, foundation, and carcass traits 0.375 0.068 0.025 S11: the majority of farmers uses artificial insemination for GA reproduction 0.375 0.101 0.038 S12: the majority of farmers appreciates usage of new biotechnologies for GA 0.375 0.101 0.038 S13: good meat quality of GA with slight drip losses 0.375 0.040 0.015 S14: milk of GA contains enhanced ingredients 0.375 0.059 0.022 S15: milk of GA suits cheese production very well 0.162 0.058 0.009 W1: the majority of GA farmers are conventional farmers 0.162 0.095 0.015 W2: the majority of GA farmers have no ‘old’ GA cattle (breeding type) 0.162 0.048 0.007 W3: GA farmers do not fatten GA bulls 0.162 0.129 0.021 W4: milk yield of GA is not adequate 0.162 0.128 0.020
W5: longevity of GA is low Weaknesses 0.162 0.316 0.051 W6: low genetic gain and inbreeding of GA 0.162 0.226 0.036 W7: milk yield of GA is not competitive with high-yielding breeds 0.278 0.086 0.024 O1: consumers want beef of high quality 0.278 0.142 0.039 O2: local and organic agriculture are important to consumers 0.278 0.078 0.021 O3: funding of sustainable organic agriculture 0.278 0.076 0.021 O4: funding of local breeds by the government
0.278 0.058 0.016 O5: funding moved from conventional to organic farming 0.278 0.100 0.028 O6: higher income for farmers with local products 0.278 0.111 0.031 O7: enhance the economic ratio between feeding costs and milk yield for GA
Opportunities 0.278 0.113 0.031 O8: organic farming produces more benefit for farmers 0.278 0.040 0.011 O9: rearing GA in areas with strong climate variabilities 0.278 0.026 0.007 O10: rearing GA on marsh land 0.278 0.062 0.017 O11: increase milk protein of GA 0.278 0.108 0.030 O12: direct marketing of GA products 0.188 0.107 0.020 T1: brand awareness of GA breed related to other local breeds is low 0.188 0.107 0.020 T2: insufficient local awareness of GA breed 0.188 0.230 0.043 T3: strong dependence on milk prices
0.188 0.170 0.032 T4: decreasing milk prices 0.188 T : monopoly of dairy business 0.132 0.025 Threats 5 0.188 0.078 0.014 T6: establish specialized GA products on market 0.188 0.046 0.008 T7: some areas may be too sandy and dour for farmers’ own GA forage production 0.188 0.131 0.024 T8: consumers’ demand for breed-specific products is not high GPV = group priority vector; FPV = factor priority vector; OPV = overall priority vector
117
Chapter Six ______
Table 2 Priority scores of ranked SWOT factors for Red Dual-Purpose cattle (RDP) SWOT group GPV SWOT factors FPV OPV 0.360 0.040 0.014 S1: 81 % of farmers have a stable RDP livestock 0.360 0.030 0.011 S2: over 95 % of farmers use RDP as main source of income 0.360 0.050 0.018 S3: the majority of RDP farmers produce their own forage for RDP 0.360 0.030 0.011 S4: the majority of RDP farmers use conventional agriculture systems 0.360 0.080 0.029 S5: dual purpose feature of RDP is important for RDP farmers 0.360 0.060 0.022 S6: the majority of RDP farmers appreciate the project work 0.360 S : longevity of RDP is high 0.060 0.022
7 0.360 0.090 0.032 S8: carcass weight of RDP is good 0.360 0.120 0.043 S9: RDP show high robustness and adaptability to different housing systems Strengths 0.360 0.080 0.029 S10: foundation and udder quality traits of RDP are very good 0.360 0.100 0.036 S11: RDP show a balanced relation between daily gain and milk yield 0.360 0.030 0.011 S12: the majority of farmers use artificial insemination for RDP reproduction 0.360 0.010 0.004 S13: selection of RDP bulls for reproduction are made by farmers themselves 0.360 0.110 0.040 S14: RDP farmers achieve financial stability due to dual purpose 0.360 0.100 0.036 S15: carcass and meat quality of RDP is good 0.360 0.020 0.007 S16: milk of RDP contains enhanced milk protein 0.137 0.050 0.007 W1: 19 % of RDP farmers will quit their agriculture farms soon
0.137 0.105 0.014 W2: the majority of RDP farmers do not want to exchange information 0.137 0.100 0.014 W3: milk yield of RDP is too low 0.137 0.415 0.057 W4: inbreeding and breed extinction of RDP
Weaknesses 0.137 0.175 0.024 W5: farmers reject the use of genomic-tested young bulls of RDP 0.137 0.153 0.021 W6: milk yield of RDP is not competitive with high-yielding breeds 0.308 0.067 0.020 O1: consumers want beef of high quality 0.308 0.096 0.030 O2: local and organic agriculture are important to consumers 0.308 0.110 0.034 O3: higher income for farmers with local products 0.308 0.075 0.023 O4: funding of sustainable organic agriculture 0.308 0.043 0.013 O5: funding of local breeds by the government
0.308 0.107 0.033 O6: high flexibility of dual-purpose breeds on changing markets 0.308 0.037 0.011 O7: constant market prices for beef cattle 0.308 0.038 0.012 O8: rearing RDP in areas with strong climate variabilities 0.308 0.023 0.007 Opportunities O9: rearing RDP on marsh land 0.308 0.057 0.018 O10: enhance milk ingredients of RDP 0.308 0.053 0.016 O11: collaboration of RDP farmers 0.308 0.109 0.034 O12: search for organic sales markets for RDP products 0.308 0.090 0.028 O13: improve traits of fertility, udder health, and foundation of RDP 0.308 0.095 0.029 O14: increase the supply of RDP breeding bulls to the farmers 0.195 0.055 0.011 T1: brand awareness of RDP breed related to other local breeds is low 0.195 0.053 0.010 T2: insufficient local awareness of RDP breed 0.195 0.164 0.032 T3: milk quota disappeared 0.195 0.132 0.026
T4: decreasing prices for beef cattle 0.195 0.146 0.028 T5: decreasing milk prices
Threats 0.195 T6: decreasing prices for beef cattle due to a massive change from milk to beef cattle 0.230 0.045 production 0.195 0.025 0.005 T7: some areas may be too sandy and dour for farmers’ own RDP forage production 0.195 0.120 0.023 T8: no special market for breed-specific products
0.195 T9: breeding goals for RDP are not achievable 0.075 0.015 GPV = group priority vector; FPV = factor priority vector; OPV = overall priority vector
- 118 -
Chapter Six ______
SUPPLEMENTARY MATERIAL
Supplementary Material S1 Setup of questionnaires for German Angler (GA) and Red Dual-Purpose cattle breed (RDP)
- 119 -
Chapter Six ______
- 120 -
Chapter Six ______
- 121 -
General Discussion ______
General Discussion
The aim of the present study was to develop novel approaches for the maintenance and support of local breeds in Northern Germany. Thereby, concepts of conservation genetics and breed management have been comprehensively considered in the context of small livestock populations. Local breeds are often exposed to an extent of challenges, like small population sizes (Meuwissen 2009), introgression from migrant breeds (Wellmann 2012), mismanagement of breeds (FAO 2014), and wrong strategic directions (FAO 2010). Thus, special emphasis was given to the aspects of (I) relatedness and inbreeding, (II) native genetic contribution, (III) breed-specific traits, and (IV) breeding strategies.
(I) Relatedness and inbreeding
The potential of purging a population of its genetic load through management intervention to reduce high extents of inbreeding depression due to relatedness, is of interest besides the impact of inbreeding in general (Ballou 1997). Inbreeding has effects on individual fitness and population dynamics within populations (Kardos et al. 2015). Thus, the restriction of relatedness and inbreeding is crucial in the design of breeding programs to control the increase in inbreeding and to avoid inbreeding depression in the progeny (Zhang et al. 2015).
In Chapter One important relations between pedigree-based ancestral and partial inbreeding coefficients and genomic inbreeding coefficients were investigated in order to make the utilization of purging effects through mating possible. Especially in small local breeds inbreeding levels can be high due to bottlenecks over the years and thus, mating of inbred but unrelated animals became more important for a healthy and genetically diverse population with low extents of inbreeding depression. Ballou’s concept of inbreeding (1997) implements that inbred individuals with inbred ancestors are less prone to inbreeding depression than inbred individuals with non-inbred ancestors. A novel genome-based inbreeding approach is the concept of runs of homozygosity (ROH) from McQuillan et al. (2008) which allow
- 123 -
General Discussion ______analyzing old and new inbreeding based on short and long haplotype segments of selection candidates. It is assumed that long ROH segments, generated by recent inbreeding, caused rare deleterious variants to exist in homozygous form within populations (Szpiech et al.
2013). By contrast short ROH segments, generated by old inbreeding, comprised less deleterious variants due to purging a population of its deleterious genetic load over years.
Hence, if mating of inbred but unrelated individuals are planned, short ROH segments should be preferred compared to long ROH segments. Nowadays, ROH is considered as the most powerful inbreeding measurement (Mastrangelo et al. 2017) and makes it possible to distinguish between recent and ancient inbreeding (Keller et al. 2011). However, recent genome-based inbreeding methods are not able to determine a previous ancestor. Enhanced purging effects on genomic level may cause a decrease in inbreeding depression and an increase in genetic diversity within small populations sharing high rates of inbreeding. To identify the relatedness between individuals on the genomic level, the standard GRM (Yang et al. 2010) and a multidimensional scaling (MDS) plot analysis within two or three dimensions based on a symmetric dissimilarity matrix (De Leeuw and Mair 2009) can be used. MDS allow identifying relatedness not only between two individuals but rather among many individuals simultaneously. That could have importance in context of small populations with minor or restricted amounts of available parents, e.g. during identification of relatedness between rare sires or sire lines within an endangered local pig breed (Fig. 1). Identified MDS plots are more comprehensive and facilitate the maintenance of genetic diversity, avoidance of inbreeding depression, and utilization of purging effects in inbred populations.
Furthermore, a MDS plot can be helpful to investigate relatedness between local and introgressed breeds, e.g. to support results from pedigree analysis (Fig. 2). Comparison of pedigree-based and genome-based results for introgression events is relevant especially if pedigree data is erroneous (Addo et al. 2017) or even lacks like in our study from Chapter
One. For instance, relatedness between major and certain local breeds may give insights
- 124 -
General Discussion ______regarding genetic uniqueness or even crossbreeds, whereas the cohort of genotyped animals is crucial for interpreting results. Additionally, MDS plots reveal not only genetic diversity between breeds or lines but rather genetic diversity within breeds or groups of animals. The consideration of three dimensions during MDS analyses can probably give more detailed information about relatedness, which may necessary for small populations with high assumed levels of relatedness.
Fig. 1 Multidimensional scaling plot of genomic relatedness between different sire lines of the Angler Saddleback
- 125 -
General Discussion ______
Fig. 2 Multidimensional scaling plot for bovine populations based on 11,822 SNPs. Bovine breeds analysed: GA (German Angler), RDP (Red Dual-Purpose cattle), RH (Red Holstein) and HF (Holstein Friesian), FV (Fleckvieh), and GBP (German Black Pied cattle)
(II) Native genetic contribution
Monitoring characteristics of breeds and their biodiversity became more and more important over the last decades (FAO 2007; FAO 2010). Therefore, investigations on genetic diversity within and between breeds were performed across different livestock species (Egito et al.
2007; Berthouly et al. 2008; Leroy et al. 2009; Manunza et al. 2016; Deniskova et al. 2017).
Besides further aspects of breed uniqueness or origin has been considered in order to assign individual amounts to breeds which were stressed by migrant breeds during the last century.
Research regarding breed uniqueness in the context of native genetic contribution is of an early stage. Tapio et al. (2005) determined breed uniqueness, total gene diversity, and allelic richness based on microsatellites for 20 native and 12 modern Northern European sheep breeds. With respect to breed uniqueness the aspect of introgression came up. Therefore,
Wellmann et al. (2012) extended the traditional Optimum Contribution Selection (OCS)
- 126 -
General Discussion ______approach proposed by Meuwissen (1997) with an additional constraint for historic introgression. Based on this novel method it was possible to recover simultaneously the native genetic background of a breed while increasing genetic gain and restrict the rate of inbreeding for targeted populations. This approach was applied and extended by comparing different
OCS object functions from Wang et al. (2017a; 2017b). In general, genetic diversity of local breeds may be increased by high migrant contributions, which threaten the conservation of small local populations. Therefore, breeding objectives should not only focus to enhance genetic gain but also to achieve genetic uniqueness and genetic diversity at native alleles
(Wang et al. 2017a). So far, most studies with respect to added values of native genetic uniqueness or native genetic contribution based on phenotypic performances but not on the genomic level. Several authors assume that the key to conserve local breeds is the identification of their added value at the product level (Verrier 2014; Verrier et al. 2018;
Bernués et al. 2018; Martín-Collado et al. 2018; Hiemstra et al. 2018). In context of native genetic contribution one major aspect is the investigation of possible benefits of NC itself.
Benefits of NC may help to distinguish and to characterize added values of individual local breeds on the genomic level. Such knowledge may play an important role concerning the choice of selection candidates for conservation. Prospectively specific proportions of NC could be considered in addition. By now candidates were mostly selected based on high native contribution or high native uniqueness. To identify added values on the genomic level the use of sequence data are promising. Sequence data may help to explore rare native genetic variants more effective, which could have been already lost in conventional high-yielding breeds. Therefore, genome sequences of a certain target breed could be compared with genome sequences of several reference breeds, which have been probably introgressed.
Thereby, it is of major importance that chosen genome sequences will optimally reflect the corresponding populations. A first step to identify benefits of NC on genomic level may be the estimation of a trait for NC (Chapter Two, Three, and Four), which can be correlated
- 127 -
General Discussion ______with other traits (Chapter Four), following pedigree- and genome-based estimation methods from Wellmann et al. (2017). As a result pedigree-based as well as genome-based estimates for NC were directly treated as a trait of NC (Chapter Four). The assumption of using the measured NC as true breeding value (TBV) is only correct if and only if every individual has an estimate for NC. Local breeds are generally small and there is a poor financial support to genotype the whole population, which can be used for a reliable NC breeding value estimation. Hence, only few animals of the herd have genotypic information for NC and most animals not, although it is necessary for a sustainable selection for NC within the whole population. One consideration could be to collect genotypic information from a few animals and translate this information to the whole population. Therefore, a gEBV for NC could help based on existing methods, where genotypic and genealogical data were combined. At this, estimation parameters like reliability and accuracy have to be extensively proven. In Figure 3 the average estimated NC values from Chapter Four were compared based on four proportionally reduced genomic reference data scenarios (50K genotypes). In the first scenario all available reference breeds and corresponding reference animals (100%) were available. In the following scenarios the available reference breeds and corresponding reference animals have been proportionally reduced up to 67%, 50%, and 33% of the originally available reference data-set for the calculations of NC. It is shown that estimated values for NC increased when the available reference data has been reduced. The effect is stronger for a reduced number of reference breeds than for a reduced number of reference animals in case of the studied Red Dual-Purpose cattle genotypes. Thus, the number of available reference breeds and available animals per reference breed may have an effect of estimated NC values.
- 128 -
General Discussion ______
Fig. 3 Average estimated native contribution (NC) for the target breed of 809 Red-Dual purpose cattle (RDP) individuals based on four different available reference scenarios
A further relevant aspect will be the birth cohort of these selected animals. In case of local breeds, it is generally assumed that older animals have a high NC value, whereas younger animals have a low NC value due to migrant contributions from foreign breeds over time.
This assumption was supported by the study of Wellmann et al. (2012) and was also confirmed in Chapter Three (Fig. 3), where pedigree-based estimates for NC decreased over time of historic introgression. Figure 4 shows the expected generic profile of native genetic contribution for local breeds with different extents of migrant contribution. Migrant events started exemplarily around 1950 and increased due to the ongoing specialization of animal breeding. Local breeds with no introgression are very rare and only exist on islands with a closed nucleus. It can be generalized that nearly every modern or local breed has been exposed to migration events once in a while. Thus, even our oldest animals carry already migrant alleles. These old animals are rarely older than 20 years and pure-bred animals are already dead in current populations. Another very important aspect might be in which moment of time migration events occurred. It is assumed that old migration events cause small scattered migrant haplotype segments, which are very difficult to track and then to reverse, whereas new migration events cause long compact migrant haplotype segments
(Figure 4). - 129 -
General Discussion ______
Fig. 4 Generic native uniqueness for local breeds with different intensities of historic introgression over time
Due to these migration events over time, there exists the hypothesis that the native genetic variance of local breeds gets lost. To prove this purely hypothetical statement an experimental approach is needed. Therefore, old animals born before 1950 are assumed to be native and get analyzed on sequence level. Identified native genetic variance of these old and native animals could be compared to the native genetic variance of animals with introgression from 1975,
2000, and nowadays (Figure 5). By doing this, the hypothesis that native genetic variance get lost indeed through introgression may be verified.
Fig. 5 Pairings of native and migrant haplotype segments for a local cattle breed with historic introgression - 130 -
General Discussion ______
The relevance for the current animal breeding can be to identify, define, and conserve still existing native contributions, otherwise genome-based added values of local breeds might get lost. The relevance of NC for animal breeding with respect to genetic gain for local breeds can be disregarded because there seems to be a negative correlation between genetic gain and NC, and a positive correlation between genetic gain and migrant contribution (Wang et al. 2017a).
However, if rare benefits of NC will be proven, the selection on NC would make more sense for local breeds and the general importance for animal breeding is obvious.
(III) Breed-specific traits
Socio-economic aspects, climate change, and an expected rapid growth of human population are the most challenging objectives for agricultural livestock production systems (Tisdell
2003; Lee et al. 2018). Figure 5 shows the conflict between socio-economics and the need for efficiency. In developed countries, the demand for livestock products is stagnating due to socio-economic factors such as human health concerns or other changed socio-cultural values
(FAO 2017), while many production systems are increasing their efficiency and their environmental sustainability (Thornton 2010). In middle Europe the society becomes more and more sensible for the origin of their food and thus, the demand for a local food production may further grow. Consumers are willing to spend more money for high quality products with local origin. The taste of a product has priority towards food supply in general. Small local breeds are faced with a reduced economic profitability or competitiveness compared to conventional breeds, e.g. Holstein Friesian. As opposed to this, an efficient animal production also band on local breeds is necessary for food security especially in developing countries
(FAO 2017). However, the conservation of such threatened breeds is very important for ensuring our natural animal resource base for the future.
- 131 -
General Discussion ______
Fig. 6 Conflict between Socio-economic and efficiency aspects of livestock production
Breed-specific traits or breed-specific properties of local breeds (e.g. heat and disease resistance, digestibility of heavy feed, healthy ingredients) may play an important role giving the right answers in the context of fast changing environmental conditions due to climate change, changing socio-economic values, and global food security. Consequently, aspects of climate change and socio-cultural values may present future development potential. One approach to promote these breeds is the investigation of unique traits within the respective natural production condition (Verrier 2014). In Chapter Five a field experiment was used following the ideas of Verrier (2014) in order to detect added values in an example of a local sheep breed in Northern Germany. It has been confirmed that rams performed different under different conditions. Based on these results, it would be justified to compare the performances between local and conventional breeds in standardized and breed-specific environments. In ongoing studies the data-set of animals and phenotypes have to be more comprehensive to estimate more reliable genetic parameters. Breed-specific traits may be also identified on the genomic level, e.g. milk ingredients of the German Angler cattle following the association study of Sanders (2005).
(IV) Breeding strategies
Animal genetic resources, e.g. local breeds, comprise a major component of the basis for the global food security (FAO 2010). However, many local breeds have been replaced by high- yielding breeds due to profitability and are currently endangered (Meuwissen 2009). To - 132 -
General Discussion ______conserve endangered livestock breeds, efficient breeding strategies have to be identified, which defines relevant breeding goals and marketing products of added values to secure profitability (Gandini and Oldenbroek 2007). Increased breed profitability by genetic improvement and by promoting breed-specific products is also an efficient strategy
(Meuwissen 2009). For the development of such strategies, FAO (2010) suggested employing the SWOT analysis (Weihrich 1982), even though SWOT approach on its own are subjective
(Martín-Collado et al. 2013). In Chapter Six we used a less subjective strategy decision approach by accounting for breeders’ expertise at breed level in order to identify conservation and development strategies. Developed strategies focused the need for relevant breeding goals, novel marketing products, and genetic gain. More consumer-orientated strategies could offer far reaching opportunities. Furthermore, in Chapter Six a team of experts, which identified and ranked breeding strategies, was limited and may influenced the strategy output.
Choosing animal breeding specialists may be not the best approach. In contrast industrial players would be more fruitful as representatives. For small local breeds, especially, correspondence between farmers, breeders, breeding organizations, and further stakeholder has major importance in order to achieve promising, practicable, and sustainable breeding strategies (FAO 2010). Thus, the amount of stakeholders, who are willing to join discussions, has to be maximized during strategy identification in order to reveal most objective and comprehensive findings. For ongoing studies, chosen expert group should be more widespread regarding individual expertise and opinions. Additionally, a more comprehensive data-set of surveys may increase the outcome but has strong dependency on farmers’ motivation. Another important aspect was the design of surveys. The surveys should be harmonized between certain parties in advance to simplify evaluation and interpretation of results, and thus, maximize the outcome of the whole study. Moreover, interests of local society and consumers should be implemented in the analyses to improve the added value and acceptance of a breed-specific product. Besides, strategies may improve if aspects of
- 133 -
General Discussion ______environment, e.g. extensive land and climate change get higher emphasis. However, next step could be to identify valuable breeding goals based on identified key strategies (Figure 6).
Therefore, the contingent value method (Davis 1963) may be applied to identify total economic weights for individual local breeds according to Teegen et al. (2008).
Fig. 7 Derivation of the breeding goal for local breeds based on strategy decision tools
REFERENCES
Addo S, Schäler J, Hinrichs D, Thaller G (2017) Genetic Diversity and Ancestral History of
the German Angler and Red-and-White Dual-Purpose Cattle Breeds Assessed through
Pedigree Analysis. Agricultural Sciences 8:1033-1047
Ballou JD (1997) Ancestral inbreeding only minimally affects inbreeding depression in
mammalian populations. Journal of Heredity 88:169-178
Berthouly C, Bed’Hom B, Tixier-Boichard M, Chen CF, Lee YP, Laloë D, Legros H, Verrier
E, Rognon X (2008) Using molecular markers and multivariate methods to study the
genetic diversity of local European and Asien chicken breeds. Animal Genetics
39(2):121-129
- 134 -
General Discussion ______
Bernués A (2018) Quantifying ecosystem services to add value in pasture-based livestock
systems. 69th EAAP Book of Abstracts No. 24:101
Davis RK (1963) The value of outdoor recreation: an economic study of the Maine woods.
PhD Thesis, Harvard University
De Leeuw J, Mair P (2009) Multidimensional Scaling Using Majorization: SMACOF in R.
Journal of Statistical Software 31(3):1-30 http://www.jstatsoft.org/v31/i03/
Deniskova TE, Dotsev AV, Selionova MI, Kunz E, Medugorac I, Reyer H, Wimmers K,
Barbato M, Traspov AA, Brem G, Zinovieva NA (2017) Population structure and
genetic diversity of 25 Russian sheep breeds based on whole-genome genotyping.
Genetics Selection Evolution 50:29
Egito AA, Paiva SR, Albuquerque M, Mariante AS, Almeida LD, Castro SR, Grattapaglia D
(2007) Microsatellite based genetic diversity and relationships among ten Creole and
commercial cattle breeds raised in Brazil. BMC Genetics 8:83
FAO (2007) The state of the World’s Animal Genetic Resources for Food and Agriculture –
in brief, edited by Dafydd Pilling & Barbara Rischkowsky. Rome
FAO (2010) Breeding strategies for sustainable management of animal genetic resources.
FAO Animal Production and Health Guidelines. No. 3, Rome
FAO (2014) Characterization and value addition to local breeds and their products in the Near
East and North Africa - Regional Workshop, Rabat, Marocco, 19-21 November 2012.
Animal Production and Health Report No. 3, Rome
FAO (2017) The future of food and agriculture – Trends and challenges. Rome
Gandini G, Oldenbroek K (2007) Strategies for moving from conservation to utilisation. In:
K. Oldenbroek (ed.), Utilisation and Conservation of Farm Animal Genetic Resources,
pp. 29-54. Wageningen Academic Publishers, Wageningen.
- 135 -
General Discussion ______
Hiemstra SJ, Remjin N, Hoving-Bolink AH, Amat L, Traon D (2018) Valorization of locally
produced dairy and meat products from native livestock breeds. 69th EAAP Book of
Abstracts No. 24:102
Kardos M, Luikart G, Allendorf FW (2015) Measuring individual inbreeding in the age of
genomics: marer-based measures are better than pedigrees. Journal of Heredity 115:63-
72
Keller MC, Visscher PM, Goddard ME (2011) Quantification of inbreeding due to distant
ancestors and its detection using dense single nucleotide polymorphism data. Genetics
189:237-249
Lee MRF, Takahashi T, McAuliffe G, Tichit T (2018) Agricultural sustainability metrics
based on land required for production of essential human nutrients. 69th EAAP Book of
Abstracts No. 24:339
Leroy G, Verrier E, Meriaux JC, Rognon X (2009) Genetic diversity of dog breeds: between-
breed diversity, breed assignation and conservation approaches. Animal Genetics
40(3):333-343
Manunza A, Noce A, Serradilla JM, Goyache F, Martínez A, Capote J, Delgado JV, Jordana
J, Munoz E, Molina A, Landi V, Pons A, Balteanu V, Traorè A, Vidilla M, Sánchez-
Rodríguez M, Sánchez A, Cardoso TF, Amills M (2016) A genome-wide perspective
about the diversity and demographic history of seven Spanish goat breeds. Genetics
Selection Evolution 48:52.
Mastrangelo S, Tolone M, Sardina MT, Sottile G, Sutera AM, Di Gerlando R, Portolano B
(2017) Genome-wide scan for runs of homozygosity identifies potential candidate genes
associated with local adaptation in Valle del Belice sheep. Genetics Selection Evolution
49:84
- 136 -
General Discussion ______
Martín-Collado D, Diaz C, Mäki-Tanila A, Colinet F, Duclos D, Hiemstra SJ, EURECA
Consortium, Gandini G (2013) The use of SWOT analysis to explore and prioritize
conservation and development strategies for local cattle breeds. Animal 7(6):885-894
Martín-Collado D, Diaz C, Serrano M, Zanoli R, Naspeti S (2018) A critical perspective on
the current paradigm of high-quality products marketing strategy. 69th EAAP Book of
Abstracts No. 24:102
Meuwissen THE (1997) Maximizing the response of selection with a predefined rate of
inbreeding. Journal of Animal Science 75(4):934-940
Meuwissen THE (2009) Genetic management of small populations: A review. Acta
Agriculturae Scandinavica, Section A – Animal Science 59(2):71-79
McQuillan R, Leutenegger AL, Abdel-Rahman R, Franklin CS, Pericic M. Barac-Lauc L, et
al. (2008) Runs of homozygosity in European populations. American Journal of Human
Genetics 83(3):359-372
Sanders K (2005) Marker-assisted estimation of pedigree errors and QTL mapping in the
German Angeln dairy cattle population. Dissertation, Christian-Albrechts-University,
Kiel:42-58
Szpiech ZA, Xu J, Pemberton TJ, Peng W, Zöllner S, Rosenberg NA, Li JZ (2013) Long
Runs of Homozygosity Are Enriched for Deleterious Variation. American Journal of
Human Genetics 93:90-102
Tapio M, Tapio I, Grislis Z, Holm LE, Jeppsson S, Kantanen J, Miceikiene I, Olsaker I,
Viinalass H, Eythorsdottir E (2005) Native breeds demonstrate high contributions to the
molecular variation in northern European sheep. Molecular Ecology 14(13):3951-3963
Teegen R, Edel C, Thaller G (2008) Bewertung der Zuchtzielmerkmale des Trakehner
Verbandes mit Hilfe der kontingenten Befragungsmethode (“Contingent Valuation
Method”, CV). Züchtungskunde 80(2):99-103
- 137 -
General Discussion ______
Thornton PK (2010) Livestock production: recent trends, future prospects. Philosophical
Transactions of the Royal Society B 365:2853-2867
Tisdell C (2003) Socioeconomic causes of loss of animal genetic diversity: analysis and
assessment. Ecological Economics 45:365-376
Verrier E (2014) Some success factors for development and marketing of local breeds
products. Opportunities for conservation of local breeds - WIAS Seminar, Wageningen.
https://www.wur.nl/upload_mm/7/f/7/d779bd35-3b92-4d17-9261-
cd506db9f37d_Seminar_WIAS_Verrier_2014-12-09.pdf
Verrier E, Markey L, Lauvie A (2018) Specific products with added value for local breeds:
lessons from success and non-success stories. 69th EAAP Book of Abstracts No. 24:101
Wang Y, Bennewitz J, Wellmann R (2017a) Novel optimum contribution selection methods
accounting for conflicting objectives in breeding programs for livestock breeds with
historical migration. Genetics Selection Evolution 49(1):45
Wang Y, Segelke D, Emmerling R, Bennewitz J, Wellmann R (2017b) Long-Term Impact of
Optimum Contribution Selection Strategies on Local Livestock Breeds with Historical
Introgression Using the Example of German Angler Cattle. G3: Genes, Genomes,
Genetics 7(12):4009-4018
Weihrich H (1982) The TOWS Matrix - A Tool for Situational Analysis. Long Range
Planning 15(2):54-66
Wellmann R, Hartwig S, Bennewitz J (2012) Optimum contribution selection for conserved
populations with historic migration. Genetics Selection Evolution 44(1):34
Wellmann R (2017) Optimum Contribution Selection and Mate Allocation for Breeding: The
R Package optiSel. Submitted to Genetics Selection Evolution
Yang J, Hong Lee S, Goddard ME, Visscher PM (2011) GCTA: A tool for genome wide
complex trait analysis. American Journal of Human Genetics 88:76-82
- 138 -
General Discussion ______
Zhang Q, Calus MPL, Guldbrandtsen B, Lund MS, Sahana G (2015) Estimation of inbreeding
using pedigree, 50k SNP chip genotypes and full sequence data in three cattle breeds.
BMC Genetics 16:88
- 139 -
General Summary ______
General Summary
The main focus of this thesis was the investigation of conservation and management strategies of several local livestock breeds in Northern Germany. Therefore, numerous aspects in the field of conservation genetics and feasibility of implementation have been considered.
Avoidance of inbreeding is a major goal when managing small populations and challenging for the design of breeding programs. Chapter One investigates the genetic diversity within a local pig breed by estimating relatedness and different inbreeding coefficients. Inbreeding estimates were low in average but few animals showed high level of inbreeding. Ancestral and partial inbreeding coefficients were calculated to enable matings to purge deleterious alleles. Further it could be shown that inbreeding estimates based on runs of homozygosity represent ancestral and partial inbreeding concepts better than other genomic measurements.
Next to inbreeding the native genetic contribution (NC) is of some concern due to introgression of high-yielding breeds. Chapter Two examines properties of a concept to estimate a breeding value for NC applied to two local cattle breeds, namely the German
Angler and the Red Dual-Purpose breed. Genetic parameters were estimated by using appropriate linear mixed models. Heritabilities for the trait of NC vary between 0.74 and 0.97.
To control the rate of inbreeding and native uniqueness simultaneously in selected animals, a novel method, the advanced Optimum Contribution Selection (OCS), was applied. Chapter
Three compares the advanced OCS with an alternative implementation of NC as a trait, with respect to genetic gain and inbreeding. Using NC as a trait increased the native genetic background but decreased the genetic gain in the next generation. Further it could be shown that animals with high NC have a minor average kinship.
Native genetic diversity of breeds harbors plenty of traits which will be irreplaceable for the adaptation to changing environments. In Chapter Four population genetic parameters for NC were analyzed to quantify possible benefits of NC. The study revealed that NC is positively
- 141 -
General Summary ______related with longevity, some functional type traits, and somatic cells indicating a better fitness and health.
Chapter Five focuses on identification and evaluation of breed-specific traits for a local sheep breed. Genetic parameters and correlations for daily gain and ultrasonic muscle-fat ratio under extensive conditions were estimated with linear mixed models. Heritabilities and respective standard errors were generally high because of the low number of animals.
Correlations for these new traits and corresponding traits under commercial conditions were tendentially negative (-0.11), which generally supports to prove animals under the real production environment. These new traits also have a positive impact on meat-quality aspects.
Optimal use of reference sires with predefined selection intensities achieves genetic response for such new traits with simultaneously acceptable rates of inbreeding.
In addition to conservation genetics, effective management strategies need to be implemented to conserve local breeds. Chapter Six investigates conservation and development strategies out of the most promising SWOT (strengths, weaknesses, opportunities, threats) strategies based on farmer surveys. Developed strategies for the two local cattle breeds included the definition of relevant breeding goals, novel marketing products, and genetic improvements to secure breed profitability and conservation.
- 142 -
Allgemeine Zusammenfassung ______
Allgemeine Zusammenfassung
Das Hauptaugenmerk dieser Arbeit lag auf der Erhaltung und dem Management von lokalen
Populationen unterschiedlicher Nutztierrassen in Norddeutschland. Dabei wurden verschiedene Aspekte rund um den Erhalt der Genetik und die Entwicklung von
Management-Strategien betrachtet.
Die Vermeidung von Inzucht ist eines der Hauptziele im Management kleiner Populationen und besonders wichtig für die Ausgestaltung eines effektiven Zuchtprogrammes. Das Kapitel
1 beschäftigt sich mit der Identifikation genetischer Diversität einer lokalen Schweinerasse durch die Ermittlung von Verwandtschaftsbeziehungen und unterschiedlichen
Inzuchtparametern. Der Inzuchtgrad war im Durchschnitt gering, jedoch wiesen einige Tiere eine erhöhte Inzucht auf. Die Konzepte der ancestralen und partiellen Inzucht sind sehr bedeutsam, um die Population bei Folgeanpaarungen von schädlichen Allelen befreien zu können. Der genomische Inzuchtkoeffizient, ermittelt durch die Runs of Homozygosity, spiegelt die ancestrale und partielle Inzucht deutlich besser wider als andere genomische
Inzuchtparameter.
Neben dem Aspekt der Inzucht spielt die natürlich-genetische Eigenständigkeit (NC) eine wichtige Rolle, da über Jahre hinweg Hochleistungsrassen eingekreuzt wurden. Das Kapitel 2 setzt sich mit der Schätzung eines Zuchtwertes für das neue Merkmal NC am Beispiel zweier lokaler Rinderrassen, dem Angler Rind und dem Rotbunten Doppelnutzungsrind, auseinander.
Die genetischen Parameter wurden mit Hilfe unterschiedlicher gemischt-linearer Modelle geschätzt und verglichen. Die Erblichkeiten für NC ordneten sich dabei zwischen 0.74 und
0.97 ein.
Um die Inzuchtrate und die natürliche genetische Eigenständigkeit innerhalb der selektierten
Tiere zu kontrollieren, wurde die neue Methode der erweiterten Optimum Contribution
Selection (OCS) angewandt. Das Kapitel 3 vergleicht die erweiterte OCS mit einer alternativen Implementierung von NC als Merkmal. Durch die Betrachtung von NC als - 143 -
Allgemeine Zusammenfassung ______
Merkmal konnte die natürlich-genetische Herkunft erhöht werden bei gleichzeitiger Abnahme des Zuchtfortschrittes in der Folgegeneration. Weiterhin konnte gezeigt werden, dass Tiere mit erhöhter NC eine geringere durchschnittliche Verwandtschaft aufweisen.
Die natürlich-genetische Diversität von Rassen beherbergt eine Vielzahl an Merkmalen, die unabdingbar für eine Anpassung an plötzliche Umweltveränderungen sein könnten. In
Kapitel 4 wurden populationsgenetische Parameter für NC analysiert mit dem Ziel den möglichen Nutzen von NC quantifizieren zu können. Die Studie offenbarte, dass NC in einem positiven Zusammenhang mit der Nutzungsdauer, funktionellen Merkmalen und dem somatischen Zellzahlgehalt steht, was zu einer Verbesserung der Fitness und der Gesundheit beitragen könnte.
Das Kapitel 5 fokussiert die Identifizierung und Bewertung rassespezifischer Merkmale für eine lokale Schafpopulation. Die genetischen Parameter und Korrelationen für die
Tageszunahme und das Muskel/Fett-Verhältnis unter extensiven Haltungsbedingungen wurden mittels gemischt-linearer Modelle geschätzt. Die geschätzten Erblichkeiten und die dazugehörigen Standardfehler waren insgesamt hoch aufgrund der geringen Tierzahlen. Die
Korrelationen dieser neuen Merkmale und den dazugehörigen konventionellen Merkmalen waren tendenziell negativ miteinander korreliert (-0.11), was die Vermutung bestärkt die
Leistung der Tiere fortan unter extensiven Bedingungen zu prüfen. Die neuen Merkmale haben einen positiven Einfluss auf die Fleischqualität. Weiterhin ist es möglich, unter optimaler Nutzung der Selektionskandidaten mit vordefinierter Selektionsintensität einen
Zuchtfortschritt für die neuen Merkmale bei gleichzeitig akzeptablen Inzuchtraten zu generieren.
Um lokale Rassen erhalten zu können, müssen neben dem Aspekt der Genetik mitunter effektive Managementstrategien identifiziert werden. Im Kapitel 6 werden Erhaltungs- und
Entwicklungsstrategien aus zuvor abgeleiteten SWOT-Strategien (Stärken, Schwächen,
Chancen, Risiken) auf Basis von Züchterbefragungen abgeleitet. Die entwickelten Strategien
- 144 -
Allgemeine Zusammenfassung ______der zwei lokalen Rinderrassen beinhalteten die Definition neuer Zuchtziele, die Identifikation neuer Vermarktungsprodukte sowie die Verbesserung des Zuchtfortschrittes um die
Wettbewerbsfähigkeit und damit den Erhalt der Rassen sichern zu können.
- 145 -
Danksagung ______
Danksagung
Meinem Doktorvater, Herrn Prof. Dr. G. Thaller, danke ich ausdrücklich für die Überlassung des Themas und für die Unterstützung während der Anfertigung dieser Arbeit. Zudem möchte ich mich an dieser Stelle bei Ihnen für das mir entgegengebrachte Vertrauen innerhalb der
Wissenschaft, Lehre und Projektarbeit ganz herzlich bedanken. Ich habe während der
Promotionszeit sehr viel dazu lernen können.
Besonders hervorheben möchte ich die gute Zusammenarbeit mit Herrn Prof. Dr. Dirk
Hinrichs, der ganz wesentlich zu meiner noch jungen wissenschaftlichen Entwicklung beigesteuert hat.
Ein weiterer Dank richtet sich an die Herren Prof. Dr. Jörn Bennewitz und Dr. Dr. Robin
Wellmann für die fachliche Unterstützung und die institutionelle Gastfreundschaft in
Hohenheim.
Weiterhin möchte ich mich für die finanzielle Unterstützung beim Ministerium für Energie,
Landwirtschaft, Umwelt, Natur und Digitalisierung im Rahmen des Europäischen
Innovationsförderungsprojektes (EIP Agri) bedanken. Ein außerordentliches Dankeswort geht hierbei an die Arche Warder (Europas größter Tierpark für seltene und vom Aussterben bedrohte Haus- und Nutztierrassen) als Leadpartner und an Frau Stefanie Klingel als
Projektkoordinatorin, die den Überblick behielt und damit ganz maßgeblich zum Erfolg des
Projektes beigetragen hat. Die Zusammenarbeit mit Dir, liebe Stefanie, war sehr angenehm und ich werde Dich stets in positiver Erinnerung behalten.
Am Schluss möchte ich mich ganz herzlich bei meiner Familie bedanken, bei der ich zu jeder
Zeit Unterstützung sowie Rückhalt in jeglicher Form gefunden habe.
Lebenslauf ______
Lebenslauf ______
Persönliche Informationen Name Jonas Schäler Geburtsdatum 05. Februar 1990 Geburtsort Potsdam Staatsangehörigkeit Deutsch Familienstand Ledig ______Berufserfahrung 12/2018 - heute Fachgebiet Ökologische Agrarwissenschaften, Universität Kassel (wissenschaftlicher Assistent) 03/2016 - 11/2018 Institut für Tierzucht und Tierhaltung, Christian-Albrechts- Universität zu Kiel (wissenschaftlicher Mitarbeiter) 01/2016 - 02/2016 Landratsamt Ludwigsburg, Baden-Württemberg (Praktikum) 05/2015 - 07/2015 Harthof-Schweinzuchtbetrieb in Elsenfeld, Bayern (Praktikum) 02/2015 - 05/2015 Albrecht und Söhne GbR in Reichenbach-Steegen, Rheinland- Pfalz (Praktikum) 09/2014 - 10/2014 Landesamt für Ländliche Entwicklung, Landwirtschaft und Flurneuordnung, Brandenburg (Praktikum) 03/2013 - 09/2013 Lehr- und Versuchsanstalt für Tierzucht und Tierhaltung, Brandenburg (Praktikum) 03/2010 - 04/2010 Institut für Binnenfischerei, Brandenburg (Praktikum) ______Akademische Ausbildung 10/2013 - 02/2016 M.Sc. Agrarwissenschaften (Fachrichtung Tierwissenschaften) Universität Hohenheim 10/2010 - 11/2013 B.Sc. Agrarwissenschaften Humboldt-Universität zu Berlin ______Grundwehrdienst 07/2009 - 08/2010 Panzerbataillon 413 in Torgelow, Mecklenburg-Vorpommern ______Schulische Ausbildung 06/2009 Abitur am Wolkenberg-Gymnasium in Michendorf, Brandenburg