Deep-Rooted Anomalies in Female Families


Pedigree Matters

Earliest Mares in the General Stud Book

Equine Genetic Genealogy

Mapping the Thoroughbred 
in the Equine Mitochondrial Tree

Deep Rooted Anomalies in Female Families Revealed by mtDNA Testing.

Mitochondrial DNA (mtDNA) is passed, with very rare exception, only from mothers to their offspring. The actual sequence of nucleotides changes very slowly over time because of spontaneous mutations that do not interfere with the normal functions. These two characteristics (inheritance only from the mother, & gradual creation of diverse sequences over time) make mtDNA testing potentially useful in identifying & validating Thoroughbred (TB) female families defined from pedigrees.

That mtDNA typing revealed multiple maternal lines of descent (matrilineages) in some Thoroughbred families (as defined from pedigrees) was first reported by Hill, et al (2002). The authors classified these discrepant lineages as modern, possible de novo mutation, and deep rooted anomalies. Deep rooted anomalies were identified in families numbered (in the Bruce Lowe system) 5, 6, 9 and confirmed by Bower, et al (2012b). An additional anomalous lineage in family 12 was reported by Bower, et al (2012a).

Bower, et al (2012b) also concluded that, despite the discrepancies reported, mtDNA test results were relatively consistent within branches (or sub-lineages) of families. Their proposed explanation for this phenomenon is that "errors in pedigrees must have occurred largely, though not exclusively, at sub-lineage foundation events, probably due to incomplete understanding of modes of inheritance in the past, where maternal sub-lineages were founded from individuals, related, but not by female descent."

Recent mtDNA testing of 2000 Thoroughbreds using methods that define the haplotypes more accurately, linked with detailed pedigree information (Rogers, 2017), has confirmed the presence of the deep-rooted and modern anomalies identified by Hill & Bower, but is contradictory of Bower's conclusions regarding when and why these "errors" largely occurred.

A more plausible explanation for the reported deep-rooted anomalies is suggested by the historical record. It should be noted that when Bruce Lowe numbered mare lines according to their production of winners of the English classics, he used as his pedigree authority the fifth edition (1891) of volume one of the General Stud Book (GSB). Three of these four deep-rooted anomalies can be traced to editorial changes made in the 5th edition where family lines were linked that had been separate in earlier editions. The fourth deep-rooted anomaly probably originated in the early practice of recording only the names of the stallions in pedigrees. Faced with two identical sequences of named sires, it was practically inevitable that later pedigree collectors would consolidate different matrilineages into the same family.

Changes in GSB

The 4th edition of Volume One of GSB in 1858 presented pedigrees collected and initially published in nine preceding editions, the earliest (An Introduction to a General Stud-book) in 1791. Few substantive changes to the roster of eighteenth century mares, beyond the addition of a section for Irish Mares, occurred after the 1808 edition, until 1891. In the 1891 edition, possibly reflecting increased interest in female lines of descent, as evidenced by the publications of Stamm-mütter des Englischen Vollblutpferdes by Hermann Goos in 1885 and Familien Tafeln des Englischen Vollbluts, by J D Frentzel in 1889, the earliest mare with a produce record in each "direct female line" was identified with a list of her descendants. Also added was an "index to principal stallions in the Stud Book up to the year 1890, shewing the original mares from which they descend in female line." Affirming this interest, another major editorial change was the presentation of a separate section devoted to the "earliest known mares, very few of which have been printed before as brood mares (though the substance has been given in the Miscellaneous Pedigrees at the end of earlier editions)…" Lists of descendants are also given for the "earliest recorded mares of their strains" whose entries occur in the "second alphabet" or in the addenda. Also added as an example was a "Table shewing descent in the female line from an original mare." This shows three branches of descent from "1st dam" Burton Barb mare to her descendants, Manilla (gr gr grandam of Blacklock) in the 12th generation, Electress (grandam of Teddington) in the 15th generation, Martha Lynn (dam of Voltigeur) and Peri (dam of Sir Hercules) in the 16th generation, and Bribery (dam of St Albans) in the 17th generation.

Families with deep-rooted anomalies

This table shows the affected Bruce Lowe figure families with the founder mare of each as defined by pedigrees in the 5th edition. Also shown is the name of the most recent common ancestor (MRCA) for those sharing a given haplotype and descent by pedigree. The standard nomenclature for mtDNA typing is followed here (Achilli, et al, 2012) and when necessary, other designations converted.
Family ("Founder Mare") Haplotype 1 (# tested) MRCA Haplotype 2 (# tested) MRCA
5 (Massey Mare) B1a (29) Ebony (1728) D1b (19) Miss West (1777)
6 (Old Bald Peg) L2a2 (4) Old Morocco Mare M2 (17) Horatia (1758)
9 (Old Vintner Mare) G2a (117) Curwen Spot Mare L1a (41) sister to Sloven (1728)
12 (Old Montague Mare) I2a2a (8) Mother Western (1731) L1a  (14) sister to Sampson (1723)
  Family 5 (Massey Mare)

29 descendants (according to GSB 1891 pedigrees) of Ebony (1728)  have haplotype B1a while 19 descendants of Miss West (1777) have D1b. The explanation appears to be that in the 5th edition, Miss West was made a descendant of the Massey Mare, whereas her pedigree in earlier editions had ended with an anonymous Basto mare.
GSB, i, 4th ed, 1858 GSB, i, 5th ed, 1891
(OLD) EBONY (black) - sister to Brown Betty,
Bred by the Duke of Rutland, in 1714, got by Basto, out of his Massey Mare, by Mr Massey's Black Barb.
❖ EBONY (or Ebony Mare),
Foaled in 1728, got by Flying Childers, her dam, Old Ebony, by Basto.
Bred by Lord Portmore, in 1742, got by Crab, out of the preceding mare [EBONY (1727)].
Bred by Mr Panton, her dam by Crab - Childers - Basto.
REGULUS MARE (sister to Favourite),
Bred by Mr Panton, her dam, Y Ebony, by Crab - Ebony, by Childers - Basto.
Bred by Mr Gee, in 1777, got by Matchem - Regulus - Crab - Childers - Basto. Died in 1802.
Bred by Mr Gee, in 1777, got by Matchem - Regulus - Y Ebony, by Crab - Childers - Basto.
❖ indicates new or substantially revised entries in the 1891 edition of GSB
  Miss West & her dam (a mare by Regulus) have had broodmare entries in GSB since 1791. Their pedigrees were given through the 1858 edition as:

Bred by Mr GEE, in 1777, got by Matchem - Regulus - Crab - Childers - Basto.
Bred by Mr PANTON, her dam by Crab - Childers - Basto.

In the 1891 edition, this lineage was revised to identify the Crab cross as Young Ebony, & the Childers cross as Ebony. An undated chesnut filly by Regulus attributed to Mr Panton was added to the produce of YOUNG EBONY, by Crab. There is no reference to any new records having come to light to justify the merging of these two lines. Instead, this would seem to be an example of what we see even today, of popular stallions being used in the same succession on mares descending from different founder mares.

Childers and Basto were stallions in the Duke of Devonshire's Stud at Chatsworth. Reference to the list of the Duke of Devonshire's horses at Chatsworth, 20 Jun 1729 reveals four mares sired by Basto, as well as numerous currently-unidentified mares (see C M Prior, The Royal Studs of the Sixteenth and Seventeenth Centuries, 1935; p 123-124). The known Basto daughters were Ebony (and foal by Childers), Brown Basto Mare (and foal by Childers), a mare by Basto out of Desdemona, and a mare by Basto, out of Ramsden. GSB itself includes three mares by Basto with produce by Childers:
(1) BASTO MARE (sister to Soreheels), dam of seven foals by Childers, including two fillies Bay Basto and Brown Basto;
(2) (OLD) EBONY (sister to Brown Betty), dam of four known foals, including Ebony, by Childers; and
(3) a mare by Basto, out of MAKELESS MARE - her daughter by Childers was dam of Gaul'em, by Starling.

Clearly there were candidates other than Young Ebony - Ebony that would have "fit" the sires-only version of Miss West's pedigree. In light of current mtDNA results, the particular Basto mare in Miss West's line is best regarded as unidentified, although she may very well also have come from the Duke of Devonshire's Stud.
  Family 6 (Old Bald Peg)

4 descendants of the Old Morocco Mare (according to GSB 1891 pedigrees) have haplotype L2a2 while 17 descendants of Horatia (1758) have M2. The explanation appears to be that in the 5th edition, the Spanker mare at the root of Horatia’s pedigree was assumed to be the same as the Spanker mare descended from the Old Morocco Mare.
GSB, i, 4th ed, 1858 GSB, i 5th ed, 1891
OLD MOROCCO MARE (sometimes called Old Peg),
Bred at Helmsley by Lord FAIRFAX, got by his Morocco Barb, out of Old Bald Peg, who was got by an Arabian, out of a Barb mare.
Bred by LORD DARCY, about 1690, out of The Old Morocco Mare.
* f Betty Percival, by [Leedes' Arabian] - Mr Leedes
* BETTY PERCIVAL had a f by The Paget Turk.
Bred by the Duke of RUTLAND, got by Grey Grantham – Paget Turk – Betty Percival, by Leedes’s Arabian – Spanker.
Bred by the Duke of RUTLAND, got by Grey Grantham - Paget (or Pigot) Turk - Betty Percival, by Leedes's Arabian - Spanker.
Bred by the Duke of DEVONSHIRE, out of Miss Belvoir, by Grantham – Paget Turk – Betty Percival, by Leedes’s Arabian – Spanker.
CHILDERS MARE (sister to Y Miss Belvoir),
Bred by the Duke of DEVONSHIRE, out of Miss Belvoir, by Grantham - Paget or Pigot Turk - Betty Percival, by Leedes's Arabian - Spanker.
HORATIA (also called Favourite)
Bred by Mr PANTON, in 1758, got by Blank, her dam by Childers – Miss Belvoir, by Grantham – Paget Turk, &c.
HORATIA (also called Favourite),
Bred by Mr PANTON, in 1758, got by Blank, her dam by Childers - Miss Belvoir, by Grantham - Paget Turk, etc.
  In editions before 1891, the founder mare of Horatia's line was an anonymous mare by Spanker, dam of Betty Percival, by the Leedes' Arabian. In 1891, Betty Percival was added to the produce of the Spanker Mare, dam of the LEEDES ARABIAN MARE (sister to Leeds, sometimes called Cream Cheeks). This assignment by the editor(s) of GSB is not supported by numerous surviving 18th century pedigrees which never extended Betty Percival’s pedigree any earlier than her dam. Current mtDNA results do not support the linkage introduced in 1891 & provide evidence that the Spanker Mare (dam of Betty Percival) was not the same as the SPANKER MARE (dam of Cream Cheeks).
  Family 9 (Old Vintner Mare)

117 descendants of the Curwen Spot mare (according to GSB 1891 pedigrees) have haplotype G2a while 41 descendants of the sister to the Bolton Sloven have haplotype L1a.

In this case, from the very earliest editions these two matrilineages have been considered to be the same. In editions of 1791, 1793, and 1808, the sister to Sloven’s pedigree was given as got by Bay Bolton - Curwen Bay Barb - Curwen Old Spot - White-legged Lowther Barb - Vintner Mare.

However, examining pedigrees that pre-date GSB provides evidence that distinguishes these two lineages. The Old Vintner Mare was evidently in use in Mr Henry Curwen’s stud; in fact, a pedigree in Mr Routh’s collection,* identified her sire as Mr Henry Curwen’s Arabian. Three more generations judging from the sire’s names were produced in the Curwen stud. In 1715, Mr Curwen advertised his whole stud for sale, including "30 Choice Brood Mares, 22 of which Mares are the Daughters of the famous bay Barbe" (Evening Post, 3 Sep 1715). It may have been at this time, that two sisters to Mixbury passed into the hands of other breeders. It is unlikely that all 22 of these daughters of the Curwen Bay Barb were descended from the Vintner Mare.
___ * C M Prior, Early Records of the Thoroughbred Horse, 1924.

The sister to Sloven (MRCA of the second lineage now included in family 9) was bred by the Duke of Bolton. 18th century pedigrees, notably those certified by Peter Hammond (identified as Stud-Groom, 1736-1739, and Steward, 1750, to the Duke of Bolton) never extended her pedigree to the Vintner Mare, ending instead with a cross of the White-legged Lowther Barb. An advertisement for Achilles (Newcastle Courant, 11 Jun 1757) is also informative. In this advertisement, the breeder, William Carr, of Etal, certified that he had purchased Achilles' dam "of Mr Hammond, out of the late Duke of Bolton’s Stud."

He gave the pedigree of his horse as
He was got by the Duke of Bolton’s Little John, his Dam, by Mogul, Son of the Godolphin Arabian; his Grand-dam, by Sweepstakes, Son of Lord Oxford’s Bloody Shoulder’d Arabian; his Great Grand-dam, by the famous Bay Bolton; his Great Great Grand-dam, by the Curwen Bay Barb, which got Old Partner’s Dam; his Great Great Great Grand-dam, by Old Spot, which got Partner’s Grand-dam; and his Great Great Great Great Grand-dam, by the Lowther White Legged Barb, which got Partner’s Great Grand-dam.
The pedigree of Achilles' great great grand-dam has the same sequence of sires as did Partner's dam (sister to Mixbury); that is Curwen Bay Barb - old Spot - Lowther White Legged Barb. However, although Carr obviously was pointing out the similarity in pedigree, he did not make the stronger claim that the mare in his horse's pedigree was Partner's dam, or a sister to Mixbury.

The Duke of Bolton is known to have owned a "Chesnut Curwen Mare" (Cheny 1728) whose pedigree was given as by Curwen's Bay Barb, dam by old Spot. The editor(s) of GSB had no original records of the early Bolton Stud available, the stud having been advertised for sale in 1755, after the death of the third Duke, and again in 1765, after the death of the fifth Duke. By that time it apparently consisted of only two mares, mother and daughter, and their produce. In fact, the first two editions of GSB gave Sloven’s pedigree as by Bay Bolton - Curwen Bay Barb - Selaby Turk. By the time pedigrees were being collected for GSB, it is also doubtful that the editor(s) knew how large the Curwen Stud had been with its potential for descendants of different founder mares to share the same sequences of sire names. Given the early practice of providing pedigrees with only the sires named along the female line of descent, it was practically inevitable that some separate matrilineages would have been consolidated into single families by later pedigree collectors.

  Family 12 (Old Montagu Mare)

8 descendants of the Old Montague Mare (according to GSB 1891 pedigrees) have haplotype I2a2a (as did Eclipse) while 15 descendants of the sister to Sampson (1723) have haplotype L1a.* In editions predating 1891, the pedigree of sister to Sampson (GREYHOUND MARE) ended with a Montagu Mare. In 1891, she was identified with the dam of Mother Western. The conglomeration of these 2 lines has been questioned in the past based on the estimated dates of the horses involved. It seems likely that this was an arbitrary decision based only on the occurrence of the name Montagu in both pedigrees.
___* This is also the haplotype found in descendants of the sister to Sloven (1728). That a line descended from the old Montagu Mare (dam of a Whiteshirt mare) was present in the Curwen Stud, as well as a line descended from the Vintner Mare, is evidenced by pedigree certificates signed by Mr Curwen (see C M Prior's Early Records of the Thoroughbred Horse, 1924; Mr Cuthbert Routh's Stud-book).

GSB, i, 4th ed, 1858 GSB, i, 5th ed, 1891
Bred by Lord DARCY, got by his (Old) Montagu, her dam by Hautboy – Brimmer – Royal Mare.
* f by Whiteshirt … Mr Smith
* This mare bred a f by Darcy Chesnut Arabian.
Got by the Bay Barb, her dam by the D’Arcy Chesnut Arabian – Whiteshirt – Montagu Mare.
❖ CURWEN BAY BARB MARE (sister to Little George),
Bred by Lord D’ARCY, got by the Bay Barb, her dam by the D’Arcy Chesnut Arabian – Whiteshirt – Lord D’Arcy’s Montagu Mare, by D’Arcy’s Old Montagu – Hautboy – Brimmer.
Bred by Mr CROFTS, in 1723, out of Sophonisba’s dam, by the Curwen Bay Barb, page 60. [Got by the Bay Barb, her dam by the D'Arcy Chesnut Arabian - Whiteshirt - Montagu Mare].
GREYHOUND MARE (sister to Sampson),
Bred by Mr CROFTS, in 1723, dam by the Curwen Bay Barb - D'Arcy's Chesnut Arabian.

Three of four deep-rooted anomalies in Thoroughbred families reported since 2002 can be traced to editorial changes made in the 5th edition of Volume One of the General Stud Book. The fourth is probably the result of confusion pre-dating GSB, arising from the absence of original records of the studs involved and the early practice of recording mare lines using only the names of the sires.

© 2018 A J Hibbard
Achilli A, Olivieri A, Soares P, Lancioni H, Kashani BH, Perego UA, Nergadze SG, Carossa V, Santagostino M, Capomaccio S, Felicetti M, Al-Achkar W, Penedo M C, Verini-Supplize A, Houshmand M, Woodward S R, Semino, O, Silvestrelli M, Giuliotto E, Pereira L, Bandelt H J, Torroni A: Mitochondrial genomes from modern horses reveal the major haplogroups that underwent domestication. PNAS USA. 2012 Feb 14;109(7):2449-54

Hill, EW, Bradley DG, Al-Barody M, Ertugrul O, Splan RK, Zakharov I, Cunningham EP. Animal Genetics, 33, 287-294 (2002). "History and integrity of thoroughbred dam lines revealed in equine mtDNA variation."

Bower, MA, Campana MG, Nisbet RER, Weller R, Whitten M, Edwards C J, Stock F, Barrett E, O'Connell TC, Hill EW, Wilson AM, Howe CJ, Barker G, Binns M. Archaeometry, first published online: 11 Apr 2012. "Truth in the bones: resolving the identity of the founding elite thoroughbred racehorses."

Bower, MA, Whitten M, Nisbet RER, Spencer M, Dominy KM, Murphy AM, Cassidy R, Barrett E, Hill EW, Binns M. J Anim Breed Genet (2012) 1-9. "Thoroughbred racehorse mitochondrial DNA demonstrates closer than expected links between maternal genetic history and pedigree records."

Rogers, Byron. CEO, Performance Genetics. Personal communications. 2017.