A consideration of the genetic aspects of some current practices in Thoroughbred horse breeding

B Langlois

Station de génétique quantitative et appliquée, Centre de recherche Inra de Jouy-en-Josas, 78352 Jouy-en-Josas cedex, France

(Received 16 November 1994; accepted 17 May 1995)

Summary — The genetic problems involved in Thoroughbred breeding are reviewed. First, an historical and sociological analysis highlights the following questions: Is performance selection, the great historical contribution of Thoroughbred breeding, still effective? Is breed purity a myth or a technically effective reality? What about inbreeding? What are the respective influences of the mare and stallion? Second, the answers put forth by the scientific community are examined. Finally, consideration is given to future prospects and to the progress required for genetic management in an industry where the con- cepts currently in use are very outdated. horse / Thoroughbred / racing ability / selection

Résumé — Considérations sur les aspects génétiques de quelques pratiques courantes dans l’élevage du cheval de pur sang. Les problèmes génétiques soulevés par l’élevage du pur-sang sont examinés. Dans un premier temps, l’analyse historique et sociologique décrit les questions posées. Dans une seconde partie, les réponses de la communauté scientifique à ces interrogations sont passées en revue. La sélection sur les performances, grande innovation historique de l’élevage du pur- sang, est-elle toujours efficace ? La race pure est-elle un concept mythique ou une réalité efficace ? Que penser de l’inbreeding? Quels sont les rôles respectifs de l’étalon et de la poulinière ? On conclut sur les perspectives ouvertes et en particulier sur les progrès à faire en matière de gestion génétique où les conceptions en vigueur sont très passéistes. cheval/pur-sang / aptitude à la course / sélection INTRODUCTION It was only later, when Darwinian thinking on evolution introduced the concept of selec- that the idea of the native Horse breeding has its historical and soci- tion, improving to counterbalance this ological dimension, in which the emergence population emerged and the desire for of the English Thoroughbred was a major effectively replace merely local The creation of the innovation, an event of prime importance conserving types. further for the development of the art and which English Thoroughbred reinforced these also the idea of was and continues to be a frequent focus ideas, imposing for rather than for of controversy. It therefore seems impor- selecting performance tant (Langlois, 1991) to situate the achieve- appearance. ments of Thoroughbred breeding histori- Thereafter, purebred management grad- cally, so as to highlight the main questions ually came to predominate over cross- raised. We shall then examine the answers breeding, culminating in the establishment of scientists currently propose, and conclude the Stud or Herd Books, a key feature of with a consideration of future prospects. animal production techniques in the ninetheenth century. These techniques required precise identification procedures THE RAISED QUESTIONS so that racing records achieved at different times and in different places could be related The concepts of blood, breed purity, selec- to the same horse. The Stud Books, by cer- tion and improvement have important his- tifying lineage, made it possible to link per- torical and sociological implications which, formance not only to the horse itself but also today, make any approach to the question of to its relatives. This opened the way to genetic improvement in the Thoroughbred a genealogical selection and progeny testing. delicate matter. The notion of ’improvement The respective roles of the mare and the of the equine species’ is therefore of very stallion could be examined from the records, great cultural importance. According to Mul- giving rise to a new notion, that of the ’dam- liez (1983), the expression dates back to sire’. The emphasis on the female line (dam the articles on the horse and donkey in Buf- to granddam, etc), whether justified or unjus- fon’s Histoire Naturelle of 1753. Despite the tified, also made it possible to reactivate the very strong imprint of the creationist dog- concept of crossing different lines in the mas of the time, these articles prefigured management of purebred stock. Efforts for evolutionist ideas in their use of the con- ’inbreeding’ and ’outbreeding’ triggered the cept of degeneration of a breed due to the same mechanisms. Since both were based influence of place or climate. These ideas on the belief that there is something to gain were adopted by the highly authoritative from the assortment of the mates that could Bourgelat, founder of the French veterinary not be obtained by random mating of colleges, in the cross-breeding doctrine he selected mares and stallions. imposed on the administration of fledgeling Until the dawn of the twentieth the national stud farms. To return to the century, the horse - with which men identified very ideal type of the Creation, as defined by his was the sole focus for famous canons, he recommended choos- closely - experimental testing of heredity theories. However, these ing a stallion of the opposite type to the dam, theories in line with the socio- to compensate for the divergence of her developed received. traits from the perfect ideal. This meant logical support they seeking out stallions in distant lands or While not wishing to advance any further regions. into the human sciences, we shall recapitulate the principal notions introduced, that apparent variability has become greater historically, by Thoroughbred breeding: than real variability because certain genetic

- effects are associated selection on performance within a pure systematically owing breed; to the generalized practice of assortative mating. This phenomenon, known as - the introduction of precise management homogamy, engenders linkage disequilib- of identity and lineage through the Stud rium, ie, a kind of due to Book; genetic variability associations of genes in addition to the com- - reflection on the roles of mare respective mon variability resulting from the effects of and stallion; the genes. The traditional methods of esti- - generalized use of planned mating. mating heritability should, in fact, be cor- Are the ideas and customs of Thor- rected for this phenomenon (Langlois, does not oughbred breeding circles in these areas 1981 Thus, genetic variability result from the of the still justified? What have the scientific com- simply variability genes munity to say about these questions today? at each locus, but also from covariance among the effects of these genes. It is likely that correlation between genetic VIEWPOINT OF THE SCIENTIFIC and environmental effects is a further source COMMUNITY of overestimation. How can one be certain that environmental effects on offspring are independent of the parental genotypes? Is performance selection, historically Surely it is the offspring of top-grade breed- the great innovation of Thoroughbred ing stock that will be predestined for a top- breeding, still effective? grade training establishment, improving their chances of a brilliant racing career? In our In 1988 this question hit the headlines in opinion, the great importance of the trainer effect revealed et Europe and America after a controversy by Schulze-Schleppinghoff between Gaffney and Cunningham, and Hill al (1985, 1987) and Preisinger et al (1989, in the April issue of Nature. The problem is 1990) is a reflection of this phenomenon. And the seems to be as follows: despite the intensity of selection yet general opinion carried out in the Thoroughbred, race times that all young Thoroughbreds are subjected have not improved any further for at least to the same training tasks each day, so that the past 50 years (Hamori and Halasz, 1959; they are all in a standard condition of phys- Cunningham, 1975). And yet the most com- ical preparation which some experts monly used selection criteria, handicap (d’Orsetti, 1993) consider too basic, although weights and prize money, generally have they acknowledge that it is hard to do better a high heritability rating (0.3 <2 h < 0.4) with young, growing horses. according to bibliographical reviews (Lan- With regard to these two phenomena - glois, 1975, 1980; Hintz, 1980; Tolley et al, nonrandom mating and the nonindepen- 1985). There is thus an apparent contra- dence of genetic and environmental effects diction between this selection for a heritable (also encountered in studies of the heri- trait and the failure to translate this into faster tability of Intellectual Quotient) - equine and track performance. There are two ways to human genetics are comparable. However, this study paradox: are they enough to explain the strong The first is fairly theoretical, involving a observed correlations between relatives in search for the reasons why heritability could the Thoroughbred population without includ- be overestimated, the main reason being ing an element of true heredity? How can one distinguish between genetic heritabil- being virtually zero. On the other hand, cri- ity and what the human geneticists call ’cul- teria derived from ranking, prize money (con- tural’ heritability? Some authors (Schulze- stant relative decrease according to the Schleppinghoff et al, 1985; Hill, 1988; rank, ie, the second horse receives half the James, 1990) seem to favour the predomi- prize money of the first, the third horse half nance of ’cultural’ effects. Others, including of the second and so on) or speed corrected ourselves in France and Gaffney and Cun- for race effect (within-race speed differential) ningham (1988) in Ireland and, more very probably have a heritability of more recently, Chico et al (1989) in Spain and than 0.10, a repeatability close to 0.25 and Preisinger et al (1990) in Germany, while a genetic correlation between them of 1. not denying the factors likely to lead to over- These different criteria therefore express estimation, favour a significant degree of the same thing and are in effect different heritability in racing ability. ways of grasping one and the same reality, It must be borne in mind that Thorough- ie, the ranking of the horses in the race, which is thus seen to be the essential cri- breds do not race against the clock but terion for against each other. For this reason in Euro- evaluating Thoroughbred perfor- mance. This the pean countries, professionals place very lit- fully justifies methodologi- cal Tavernier tle importance on speed achievements when approaches proposed by for this criterion. We would selecting animals; very often, they are not (1989) analysing even recorded. However, the few studies also point out that, with an average (n) of five the coefficient determination carried out on time performance (Dusek, races, = + of mon- 1965, 1977) show a strong relationship (R2 nir2/[1 (n-1 )r]) performance reaches 0.25 with a of between this and race category. The higher itoring heritability (h2) 0.10 and a close to grade the race, the faster the times. On the repeatability (r) 0.25, other hand, this criterion is only weakly her- which would seem to more than justify a itable (Artz, 1961; Bormann, 1966; Langlois, conventional genetic approach. 1975, 1978, 1980; Hintz, 1980; Tolley et al, In conclusion, it appears that, in the pre- 1985). Some exceptions can be encoun- sent state of research findings, racing per- tered but they generally concern short dis- formance in Thoroughbred horses is mod- tance races (Buttram et al, 1988; Moritsu et erately heritable, and is best expressed as al, 1994). It is therefore not surprising that the number of races won and ranks inde- indirect selection for a trait with weak heri- pendently of performance against the clock. tability should prove not very effective. All It is not surprising that, despite selection, depends on the genetic correlation between time performance is improving very little, the selection criteria and recorded speeds, because no one is trying to improve it. This and this has hardly been studied so far. in no way implies that there is no real However, Beatson (1989) showed that even progress in Thoroughbred populations, con- if heritability and genetic correlation are sistent with the heritability ratings and the strong, indirect selection on racing success intensity of selection. criteria would hardly improve speed at all, To the question asked in the introduc- because times achieved horses in a race by tion, we would answer: Yes, performance have a low variability. breeding is still effective in the Thorough- The study of the Spanish Thoroughbred bred, but it no longer effects speed as such. population (Chico and Langlois, 1990; It probably now affects only the physiologi- Chico, 1994) shows that racing speed is not cal and psychological traits connected with the best way to grasp Thoroughbred per- racing effort. Racing effort can be defined formance, its repeatability and heritability as an abrupt shift from a submaximum speed to a maximum speed, for a brief makes the figure converge towards zero moment. This sudden acceleration is twice as fast when heritability is 0.50 and prompted by challenges from other horses. three times as fast when heritability is 0.33. It is commanded by the horse’s psycholog- In all commonly found situations, therefore, ical makeup and involves the mobilization a three-generation pedigree seems suffi- of anaerobic muscular metabolism. The cient. number of such efforts a horse can make The advantage of knowing pedigrees is during a race is limited and depends on its that it makes it possible to control inbreeding and resistance fighting spirit. when selecting mating partners. In this con- nection, the work of Mahon and Cunning- ham (1982) and Galizzi-Vecchiotti (1977) on The pure breed: myth or reality? consanguinity in the Thoroughbred is of inter- est. The examination of pedigrees over five International rules for initially registering generations revealed a very low rate of con- in lower horses in the Stud Book require eight gen- sanguinity Thoroughbreds, generally than 5%. of blood mark- erations of successive Thoroughbred x However, analysis ers reveals a far in this Thoroughbred matings, that is, no more greater homogeneity breed than in others and than one part in 256 (4/1 000) of non-Thor- (Guérin Mériaux, Is this due to a ’foundation stock oughbred blood. That is a very high degree 1986). of breed purity indeed! Introduction of for- effect’? Mahon and Cunningham (1982) the of some brood mares eign genes into the population is thus offi- explored pedigrees cially extremely limited. And this purity is back to the foundation mares and stallions of further guaranteed by systematic parent- the breed and used a sampling technique to calculate their coefficients. age testing by blood typing tests which, inbreeding found values from 5 to since their introduction, have prevented any They ranging 20%, fraudulent input of non-Thoroughbred which seems to indicate that knowledge of over five which genes. pedigree generations, always gave very low ratings, did not adequately The of the horses can also be pedigree reflect the real situation. This phenomenon, obtained five readily generations back, pro- which has also been described in the Amer- the matter for numerous dis- viding subject ican Standardbred (Cothran et al, 1984), sertations. What have to geneticists say reflects the fact that there have been bottle- about this? necks in the population at some time in the Identity control and management using past, so that the genealogical tree widens blood markers makes the information that out for the first few generations back but can be deduced from lineage extremely reli- then narrows again, the distant ancestors able. However, the amount of genetic infor- proving to be always the same. If one dis- mation given by parentage diminishes counts the mutations that may have occurred rapidly with genealogical distance. A horse over the course of these many generations, and one of its parents have half their genes this can lead, in practice, to very wide dif- in common: a horse and one of its grand- ferences in the degree of consanguinity from parents have statistically one-fourth. The one mating to another. However, this seems figure for three generations is one-eighth, to be a very difficult problem to master: one and for n generations it is (1n. /2) For n = 8, would have to have reliable genealogical this is not far from zero. In practice, since data going right back to the origins of the only the phenotype is accessible, the coef- breed, and the means to process the data ficient of heritability of the traits being would require a very considerable computing selected must be taken into account. This capacity. To conclude, the emphasis on breed recommendation one can make is to avoid purity does not seem to be necessary in the practice of inbreeding, even though it genetic terms, especially in a structure still fascinates breeders at a subconscious where both parents are selected on the level. basis of their performance. It deprives the population of high-quality exogenous inputs and tends to reduce the available genetic Respective roles of dam and sire variability by maintaining a certain diffuse consanguinity that cannot be mastered with- It was in Thoroughbred breeding that the out considerable resources and data. notion of ’sire of dams’ or ’damsire’ first made its appearance, suggesting that it is not necessarily the same stallions that pro- Inbreeding versus outbreeding duce good race horses and good brood mares. Is this distinction well founded? And Whether to aim for consanguinity or to avoid what is the reason for the importance laid it is a common topic of conversation when it on maternal origin in this livestock branch? comes to Thoroughbred breeding. Expres- Is it a mystical belief dating back to the dawn sions such as &dquo;inbred 3 x 4 on Northern of time and the cult of the mother goddess? Dancer&dquo; have become a part of the jargon of Is it due to lack of confidence in the paternal the trade. This can be regarded as equiva- origin? Is it connected with variations in the lent to a calculation of the consanguinity ability of mares to bear and raise their off- coefficient, ie, the probability, conditioned spring in simple terms of high growth rates by the pedigree, that an individual will and good health? Is the training effect con- receive two identical allelic genes copied fused with a maternal effect? Is there a cyto- from one and the same gene in an ancestor. plasmic heredity factor, ie, genetic informa- When Thoroughbred breeding profes- tion that can only come from the dam? It sionals discuss this matter, they generally could also stem from a selection recom- mendation. that stallions are refer to a fairly distant blood relationship Knowing highly it be assumed that on this that will only carry a weak probability of fix- selected, may side is but on the ing the genes of the ancestor or ancestors in genetic variability low; dam be assumed to be question. In addition, by this process, genes side, variability may because dams have been bred less are arbitrarily fixed, whether or not the traits greater In which case it is recommended they carry are desirable. Independently of intensively. inbreeding depression, due to the loss of to pay greater attention to the choice of dam. genetic variability at each locus, this practice What can one conclude in the present also implies many failures in its chosen goal state of knowledge? The first question to of concentrating in one individual as many ask is: can one really demonstrate a differ- favorable genes from illustrious ancestors ence between the paternal and maternal as possible. pathways of hereditary transmission in Thor- The second It can be said, therefore, that inbreeding oughbred breeding? question is, at the rate that is usually practiced does not if such a difference does exist, what are the to account for it? enable much gene fixing and, at the much simplest hypotheses higher rates that were practiced previously, The answer to the first question is yes. As it fixed genes arbitrarily, with too high a fail- Langlois and Chico (1989) have shown, ure rate to be recommended in a species there is a clear difference (table I) between like the horse, which has a long generation the two parental transmission pathways, inn interval and a high cost per animal. The only favour of the maternal pathway. The same situation can also be observed in trotters does not solve the problem of where this (Langlois,1984). belief came from; it doubtless originated from some actual In our It is harder to give a conclusive answer to phenomenon. opin- it arose from ’maternal-breeder’ the second question. Perhaps, as Schulze- ion, effects, Schleppinghoff et al (1987) and Preisinger et borrowing an expression from Tavernier confusion between a al (1990) seem to assume, the most impor- (1987, 1988), ie, ’breeder confused with the tant environmental effects are trainer effects. effect’, always It is noteworthy, however, that the trainers brood mare effect, and the effect of variable that will make a mare more seek out products of the same origin and mothering ability attach great importance to the maternal lin- or less capable of bearing, suckling and her of the foal’s eage. A great part of the environmental raising foal, independently effects allowed to the horses will thus fol- genetic racing potential. low the genetic effects, and the greater the During the 48 months of a Thoroughbred importance placed on the maternal origins, foal’s growth, the most decisive stage is the the more this will be so on the female side. first third: the period of gestation and suck- As a result, belief in the importance of the ling, and this, of course, depends entirely female line would be self-sustained. But this on the dam. This stage has a very strong impact on the foal’s final development. usual economic operators, the national Therefore, the mother’s contribution is authorities ought to take an interest in the indeed not restricted to her genetic input issue, to avoid a decline in this profitable but includes part of the environmental influ- branch of international trade. ence: the that on the maternal part depends It therefore seems important, first of all, to environment she provides. Knowing that considerably improve the performance cri- these traits are themselves heritable, they teria used. Many statistical publications on the notion of a damsire and can justify gross prize money seem altogether out- the explain apparent asymmetry between dated and ought to be improved. The tech- the and maternal of paternal pathways niques for doing so have long been avail- heredity. able, are widely used in the saddle horse Finer phenomena such as cytoplasmic industry and are now also being applied to heredity or genomic imprinting (which means trotters. The new techniques could them- that only the alleles from the sire or only selves be improved by using ranks rather those from the dam are expressed in the than prize money, together with the system early stages of development of some tissue developed by Tavernier (1989, 1990, 1991). or other in the embryo and therefore strongly A second stage would be to establish influences the adult) are partly taken into precisely the degree of heritability of racing account in modelling the maternal effect. performance in the Thoroughbred. There is these into more Integrating phenomena much debate on this subject in scientific cir- detailed will an models be interesting cles, and the facts must be definitively estab- research prospect but, for the time being, lished. The credibility of present economic there is no evidence that suggesting they practices depends on it. Why should one are significant. pay more for this or that pedigree or high- performing sire if performance is not heri- table? It would be nonsense. However, this PROSPECTS problem is already largely solved. According to Preisinger et al (1990) and Schulze- The economics of Thoroughbred produc- Schleppinghoff et al (1987), it would seem tion is dominated by speculation on breed- that most of the hard-to-verify environmen- ing stock and depends to a much lesser tal effects are trainer effects. The German extent on the levies on bets that provide the analyses take into account the trainer effect, prize money. The industry therefore and this considerably reduces heritability depends on the postulate that heredity is of (table II). However, it must be stressed that, prime importance. And as breeding value in these analyses, genetic effects and trainer is difficult to establish clearly, this leads to a effects are assumed to be independent. In kind of poker game on a large scale, in other words, apart from statistical fluctua- which the operators have no interest in tions, the genetic level of every trainer’s foal openness or visibility. One should also take batches is assumed to be the same as that into account the national versus interna- of every other trainer. But this supposition tional aspect of this industry and not neglect does not seem to be justified. As a result, it certain purely technical aspects that could in seems that these analyses in fact provide the future make breeding methods backfire a lower limit for heritability: at present, for disastrously. The shrinking of a national annual total prize money, this is around 0.10. population’s gene pool is a problem often The values we have obtained for the mentioned. No one knows the situation in same variable, without correction for the this regard at present. Independently of the trainer effect, are around 0.25 (Langlois and Chico, 1989) and probably constitute an Thoroughbred, ie, a horse able to win on upper limit. Fine-tuning this approach is all 2 400 m. This is one way of reintroducing that is required for drawing a definite con- an outbreeding logic into a purebreeding clusion on this issue. scheme, aimed only at eliminating from the stock a number of males and As a third step, knowing the heritability breeding large as ratings will enable us to move on to a females, as accurately targeted possi- genetic evaluation of Thoroughbreds. The ble, so as to retain only the best to procreate BLUP animal model (Tavernier, 1987, 1988) the next generation. Whether mating is ran- dom or carried out has lit- seems particularly suitable since it incorpo- very thoughtfully rates a maternal effect and an environmental tle effect on the average quality of the next as a but this is the effect which is shared by all of a horse’s generation whole, major performances and which therefore includes preoccupation of individual breeders. To advise the breeder on this we shall the trainer effect. Furthermore, because a issue, horse is evaluated on the basis of its own therefore have to examine much more performance and that of all its family, the closely the different traits involved in racing BLUP animal model also corrects for the success. What we can say is that analysing quality of the mates bred to each breeding pedigree alone will provide no magic animal. Independently of the evaluation of answer to this issue. The various mating the animals, this procedure also makes it systems suggested so far have no biological whatever their possible to assess genetic trend over recent grounding and, protagonists decades. may claim (Roman, 1984), their efficacy has in no way been proven. There remains a need to pinpoint the rac- ing success factors that are selected. With this in view, distance will no doubt capacity CONCLUSION have to be studied more thoroughly. The emergence within the breed of more spe- cialized lines such as sprinters, milers and From our examination of all these questions, stayers diverges from the principal objec- it emerges that the genetic management of tive, which is the production of the classic Thoroughbred populations could be greatly improved by using methods that research performance in Standard bred horses. J Heredity has already virtually perfected. Unfortu- 75,220 Dusek J Evaluation of times reached Thor- nately, horse breeding is a very old-fash- (1965) by oughbreds in flat races at the state-owned race ioned convinced of the firm industry, ground- course in chuchle. V6deck6 prace vyzkumnd stanice ing of its archaic methods, and does not prochov koni ve Slatinanek 1965, 143-160 seem ready to adopt these proposals. The Dusek J (1977) Analyse der im Rennen bei ver- reason is doubtless that there are very few schiedenen Leistungsklassen erreichten Schnel- ligkeiten. Die Bodenkultur28, 208-2133 technical organizations responsible for coor- horse Breeders Gaffney B, Cunningham EP (1988) Estimation of genetic dinating breeding. regard trend in racing performance of Thoroughbred horses. their fellow breeders as competitors and Nature 332, 722-724 in can see no point cooperating for genetic Galizzi-Vecchiotti G (1977) Consanguinite et perfor- improvement. This restricts them to search- mances chez les pur-sang : etude sur les etalons ing for the formula for some mythical, ideal employes en Italie en 1971 et 1972. Rivista di Zootec- nia e Veterinaria 560-566 mating that will give them the champion they 5, Guerin G, M6riaux JC (1986) La distribution des mar- one to own. And with that hope day queurs sanguins dans les races equines. Analyse approach, obviously, dreams count for more sur 6chantillons Pur Sang, Trotteur FranQais et Selle than reality. FranQais. CEREOPA, 12e Journ dtud, Paris, France, 2-13 Hamori D, Halasz G (1959) Der Einfluss der Selektion auf die Entwicklungder Schnelligkeit des Pferdes. Z Tierz REFERENCES Zuchtungsbiol73, 47-59 Hill WG (1988) Why aren’t horses faster? Nature 332, 678 Artz W (1961) Ein Beitrag zur Aus wertung der Leis- tungsprufungen in der Vollblutzucht unter beson- Hintz RL (1980) Genetics of performance in the horse. derer Berucksichtigung der Rennleistungen einzel- J Anim Sci 51, 582-594 ner Giessener Hengstnachkommen-schaften. James JW (1990) Selection limits in Thoroughbred Schriftenr Tierz Hautiergenet 2, 1-62 horses. Proc 4th World Congress on genetics applied Beatson PR (1989) Scientific correspondence about to livestock production, Edinburgh, Scotland, 23-27 Thoroughbred breeding. Nature 337, 4144 July, vol XVI, 221-224 Bormann P (1966) Ein Vergleich zwischen Gener- Langlois B (1975) Analyse statistique et génétique des alausgleichsgewicht und Zeitmessung als Selek- gains des Pur-sang anglais de 3 ans dans les courses tionsmasstab in der Vollblutzucht. ZOchtungskunde plates francaises. Ann Genet Sél Anim 7, 387-408 38, 301-3100 Langlois B (1978) Aspects spécifiques de I’am6lioration Buttram ST, Wilson DE, Willham RL (1988) Genetics of génétique des chevaux de pur sang. 29th EAAP racing performance in the American Quarter Horse. Annual Meeting (horse comm), Stockholm, Sweden. 111. Estimation of variance components. JAnim Sci In: Le cheval, reproduction, selection, alimentation, 66, 2808-28166 exploitation (R Jarrige, W Martin-Rosset, eds) (1984), Inra, Paris, France, 387-408 Chico MD (1994) Genetic analysis of thoroughbred rac- ing performance in Spain. Ann Zootech 43, 393-397 Langlois B (1980) Heritability of racing ability in Thor- a review. Livest Prod Sci 7, 591-605 Chico MD, Langlois B (1990) Estimation des com- oughbreds - posantes de la variance des r6sultats en course des Langlois B (1981) Proposition d’extension du mod6le pur sang en Espagne. 4lth EAAP Annual Meeting g6n6tique additif classique a des situations non pan- (horse comm), Toulouse, France mictiques. Ann Genet Sel Anim 13, 151-164 Chico MD, Alenda R, Alonso A, Jurado JJ (1989) Thor- Langlois B (1984) H6ritabilit6 et correlations génétiques oughbred breeding in Spain. Genetic evaluation and des temps records et des gains établis par les trot- genetic trends using an animal model. 40th EAAP teurs Frangais de 2 a 6 ans. 35th EAAP Annual Annual Meeting (horse comm), Dublin, Ireland Meeting, The Hague, The Netherlands Cunningham EP (1975) Genetic studies in horse popu- Langlois B (1991) Analyse des probldmes gbn6tiques lations. In: Proc Int Symp on Genetics and Horse chez le pur sang : état des recherches en France et Breeding, Royal Dublin Society, 17-18 September a 1’6tranger. CEREOPA, 17e Journ dtud, Paris, 1975, 2-6 France, 74-83 Cothran EG, MacCluer JW, Weitkamp LR, Pfenning Langlois B, Chico MD (1989) Études frangaises sur 1’esti- DW, Boyce A J (1984) Inbreeding and reproductive mation de la valeur g6n6tique des chevaux de pur- sang. 40th EAAP Annual Meeting (horse comm), Bundesrepublik Deutschland. 36th EAAP Annual Dublin, Ireland Meeting (horse comm), Kallithea, Greece Mahon GAT, Cunningham EP (1982) Inbreeding and Schulze-Schleppinghoff W, Claus J, Kalm E (1987) Die the inheritance of fertility in the thoroughbred mare Erstellung eines BLUP-zuchtwertschdtz modells fur Livest Prod Sci 9, 743-754 die deutsche Vollblutzucht. ZOchtungskunde 59, 25- Moritsu Y, Funakoshi H, Ichikawa S (1994) Genetic eval- 30 uation of sires and environmental factors influencing Tavernier A (1987) Application of the animal model best racing times of Thoroughbred horses in Japan. methodology to horse genetic evaluation. 38th J Equine Sci 5, 53-58 EAAP Annual Meeting (Genet comm), Lisbon, Por- Mulliez J (1983) Les chevaux du royaume. Ed Montalba, tugal 385p Tavernier A (1988) Advantages of BLUP animal model d’Orsetti H (1993) Reflexions sur l’entrainement du for breeding value estimation in horses. Livest Prod galopeur. EquAthlon5, 12-13 Sci7, 591-605 Preisinger R, Bormann P, Kalm E (1989) The breeding Tavernier A (1989) Estimation of breeding value of programme of the Thoroughbred horses in the Fed- horses from their ranks in races or competitions. eral Republic of Germany. 40th EAAP Annual Meet- 40th EAAP Annual Meeting (horse comm), Dublin, ing (horse comm), Dublin, Ireland Ireland Preisinger R, Wilkens J, Kalm E (1990) Breeding val- Tavernier A (1990) Estimation of breeding value of jump- ues and estimation of genetic trends in German thor- ing horses from their rank. Livest Prod Sci 26, 277- oughbred horses. Proc 4th World Congress on genet- 290 ics applied to livestock production, Edinburgh, Tavernier A (1991) Genetic evaluation of horses based 23-27 vol 217-220 Scotland, July, XVI, on ranks in competitions. Genet Sel Evol23, 159- Roman SA (1984) An analysis of dosage. The Thor- 173 oughbred record, April 1979-1984 Tolley EA, Notter DR, Marlowe TJ (1985) A review of Schulze-Schleppinghoff W, Kalm E, Bormann P (1985) the inheritance of racing performance in horses. Analyse der Rennleistung bei Vollblutpferden in der Anim Breed Abstr 53, 163-185