First return of Dexters to Ireland

[wagra dexters]

Beryl's early duns were registered as red in UK. Duns were being registered as dun by the time they went to Canada, such as Woodmagic Mourning Dove 2nd, dam of our first herd bull, Bindalee Celtic Chieftain, imported embryo.
For a long while US registered red/dun without differentiation. I don't know when differentiation began but even as late as 2002 at the 2nd Congress in Qld several of us were standing around a red bull by Elmwood Robson when a visiting American breeder said, "We call that dun." He was paler than the Outlaw red shown in the heifer above, but definitely not dun.

Edit: 2nd World Dexter Congress

[Tim Watson]

I believe it was during Beryl's trip to the US that she informed them that their 'red's' were actually duns.

Beryl definitely told me about this incident and how she felt very awkward doing so but felt she had to say something.

[JohnO]

It has been brought to my attention that Dr. Schmutz said that Redpoll was not tested for her Dun research paper. Since it is known that the Grinstead herd was crossed with Redpoll and that Grinstead Plover is believed to be the source of the Dun colour the research is fundamentally flawed and the conclusion that the dun colour in the Dexter breed is unique is entirely unsustainable.
In a previous posting I said that the only other breed that the owners of the Grinstead herd kept was Redpoll. This is incorrect. They also kept Jerseys.
Here is a link to footage of the Grinstead herd before the Redpoll introgression. Credit to Mike Sinnott for finding this.

http://www.britishpathe.com/record.php?id=56823

On the height issue the ADCA website has an article from the 1915 January Bulletin of the American Kerry and Dexter cattle club. In the section titled 'Important Dexter Characteristics' it states 'The Dexter is one of the very smallest breeds of cattle, a height of 38 or 40 inches being common. The bulls should not weigh over 900 pounds and the cows not over 800 pounds'. Again credit to Mike Sinnott for drawing my attention to this.

[JohnO]

The following is a complete list of the breeds tested before concluding that Dexter Dun is unique. The list is as published by John Potter who along with Carol Davidson and Sheila Schmutz conducted the project.

Angus, Belgian Blue, Blonde d’Aquitane, Braunvieh, Brown Swiss, Canadienne, Charolais, Flamande, Galloway, Gelbvieh, Guernsey, Hereford, Highland, Holstein, Jersey, Limousin, Shorthorn, Simmental, and Tarentaise.

[wagra dexters]

Well, still a few breeds to be tested then. It would be great if all our breed societies were financial enough to fund a mass genetic proofing of purity and exent of introgression. (Not a popular study I imagine). Why was the Dexter dun claimed to be unique then? I wonder if it was because it was expected to be the same as the jersey brown gene.
What does the wild red 'fawn' look like? Very interesting. Why were Dexters removed from the RBST list if only one or two herds are considered to be pure, or is the whole subject way too contraversial?
Most breeders don't want to know the complicated details, some can't read a pedigree and care less. They only want to enjoy their Dexters, and that is a one of the most legitimate reasons for Dexter ownership, but some of us are very keen on scientific studies to help protect what purity we do have left in the international Dexter herd.
Margaret

[monica a waltho]

Wasnt the Woodmagic herd tested when the breed was tested for introgression from other breeds
would this not have found Red poll or Jersey blood

[Carol K]

John Potter had a few red Polls after the Dun testing was finished. He tested them to see if they carried Dun and they did not.
That's not to say that some of them don't just thought it was interesting info.

Carol K.

[carol d]

Monica, I think Duncan would be the most experienced person to respond to that. I wonder if it would be possible to identify Red Poll genes from a (guessed at) illicit cross generations ago vs. the mishmash of genes found in the genetic soup that makes up Dexters?

After all, for 43 years, as long as it was solid coloured and a dwarf with horns, it was accepted into the registry as a Dexter. That has to have produced a lot of leeway in foundation genetics. Who's to say just when and how often one gene, visible or not, came along for the ride.
For the brown presentation in AU in 2002, I did exactly the same research John O. is referring to, and found the same connection. However, I didn’t come to the same conclusion because I know that early exports of Irish and English foundation and registered Dexters to the US between 1905 and 1915 included a number of dun animals. In fact, by 1985, there were only a couple of confirmed genetically red animals in the US: one in Florida and the other in Oregon. The overwhelming majority of non black Dexters, registered as red, were actually dun. I wonder if it’s safe to extrapolate that, since duns were part of the foundation genetics of registered Irish and English Dexters over a hundred years ago, and Suffolk Duns were considered acceptable into registries that were for red animals, this may point to a lack of distinction between the two colours in those days.
Any problems the Red Poll registry had with ‘sandy’ colouration is NOT related to Dexter ‘dun’. I can say that with some certainty because red in any form dominates Dexter brown. I agree that both Red Poll and Lincoln Red absorbed the Suffolk Dun and as a distinct breed it disappearedby 1906. A number of years ago, I contacted the breed secretaries of both Red Poll and Lincoln Red to see if their records could identify potential old bloodlines that might be based on the SD. In both cases I was told foundation animals were not recorded as to origin, per se. As well, there had been the foot and mouth epidemics which left large holes in their genetic fabric, so at this remove, the only way to be certain of the presence of the dun gene in any living Red Poll or Lincoln Red today would be to test every animal until either a dun gene was found, or the whole population had been proven exempt. That’s a pretty daunting task, and expensive, too and in the end, still wouldn’t prove anything.

It's so easy to jump to conclusions. We all tend to think that what we know must be the limits of knowledge. For years, dun, the creamy muzzle on reds, the pale tone of red, the black faces and legs on reds, were all attributed with confidence by Dexter experts to 'channel islands influence'. oops. Those proponents (and some were pretty rude about it) have now been shown to be wrong, and a lot of good animals and breeders maligned unfairly. (This is probably not the time or place to comment on damaging unsubstantiated gossip being given the same authenticity; another habit I deplore.)

I’m curious about the ‘faun’ coloration John O. refers to. Probably the best person to consult would be John Potter (USA) as he has bred hundreds of red and dun and produced a number of variations, including pale to dark red, cream muzzles, black faces and legs, changelings, and brindle. Because all his animals are colour tested, he can speak with certainty to the genetic make-up of those variations.
I think it possible but unlikely that Plover was a cross. Beryl always claimed Lady Loder’s passion for her Dexters would have made crossed registrations impossible; and John Tuckey told me she regularly crossed her Dexter cows with a Jersey bull if the cow was having too many bulldogs, as she thought this would ‘break the cycle’, but these were never registered. I’ve checked the entire Appendix registry from 1923 right up until the last purebred Grinstead was registered, and there is not one Grinstead animal listed. Not that it is proof in itself, but with a place to register them legitimately, one would think.....and, in Morley’s personal photo album during his tenure there, every photo, even the informal ones, shows only each breed by itself; never combined breeds in the same field together. At that time, all bulls were kept in separate bull pens and only had conjugal rights at managed times. All conjecture, of course, but I’m inclined to think poor old Plover and Lady Loder are more likely to be victims of selective research than dirty doings.

I would like to make a small correction: I was only peripherally involved in the DUN study. I do take credit for identifying the E+ gene as being present in Dexters, and as being expressed as red. That was in 1997. I faced some pretty stiff disbelief from the boffins, as at the time no other breed tested expressed E+ as red. I did mention to Sheila Schmutz that Dexters came in a 'dun' colour as well, which was recessive, and got much the same treatment. I gather this was because all the other known coat colour modifier genes were dominant, and 'everyone knows there is no brown gene in cattle'. it wasn't until two years later that John Potter, with his breeding records and photos was able to convince her to reconsider. He provided her with hair samples, she checked them with her dog brown study, and the rest is history. I was privileged to present their findings at the ’02 Congress in AU. I know Sheila tested all the known not red/not black coloured breeds, but didn’t have samples available to test all the other fringe breeds, too—and it would be a random chance to pick the right samples, one way or the other (think of taking six random samples of Dexters and expecting to find conclusive proof of dun).

I’d like to briefly address the issue raised here of height standards. In 1988, David Couper had a large herd of Dexters, and many were taller than 42 inches. That was before he did his big herd reduction in the early ‘90s.

The original DCS standard was defined solely for dwarfs, as nons were culled. Now that the nons are gaining in popularity, they need to be accommodated. If one doesn’t like to do that within a single standard, then a push needs to be made to set the standard for one type, and then say the other type’s limits are 4 or 5 inches smaller or taller. Otherwise, you have a standard set by human theory that is impossible to be met by cattle genetics.

Hindsight is never 20/20, all we can do is to fit the jigsaw together as best we can. I’ve often thought it might make a fun card game to set out all the known ‘facts’ concerning Dexters, divided into topics, like suits. Shuffle, deal, and the person who comes up with the best, most plausible explanation whilst using the most of the ‘fact’ cards in his/her hand, would win. Carol d.

[monica a waltho]

well thanks for that interesting reply- I love my dun red polls!! and im sticking to them whatever

[JohnO]

Carol,
I can assure you there is nothing selective about my research nor do I jump to conclusions based on an extremely limited series of genetic tests. My work would have progressed more quickly had access to some records and archives not been blocked or smoked.

Would you please name the dun dexters which were imported to America between 1905 and 1915 and provide any substantiation for this statement?

It would be useful if Duncan MacIntyre were to comment on the results of the Cardiff project, in particular as to whether Woodmagic genetics were found to be distinct from other dexters. He is one of the select few who have been given access to the results of this project which was paid for from the subscriptions of the members of the Dexter cattle society. It would also be useful if someone knowledgeable about Dexters and domestic stock genetics , for instance Mr. Sheppy, were to contribute to this conversation.

The UK Dexter register has more animals registered than all other registers combined. I am not aware of any animal registered as dun within this register which does not contain Plover blood. (save some errors of registration which are correct in the herd books).

It was Lady Loders son Sir Giles Loder who said that the Grinstead herd was bred to Red Poll.

Clearly from your comments you have access to the Grinstead archive. Where is it and will it be opened to other Dexter breeders, researchers? Have you photos of Plover. The photographs of Woodmagic Dormouse tell alot.

The following are pictures of a Suffolk Dun herd and the early Woodmagic herd. It was considerate of the photographer to include black animals in the Woodmagic photograph!





To my knowledge only red or black animals were accepted for registration in Ireland or Britain. The only time I am aware of this being changed was when Dun was allowed in. Both carriers and non carriers were always registerable. At some stage polling became accepted to accomodate the influx of Angus and other polled breeds. If you have evidence to the contrary please share.

All dun Dexters would be black if they did not have 2 copies of the dun gene. Being homozygous for dun causes black to be diluted and express as dun ( pale black).Sometimes Dun is so pale that even the eyes are a different colour. Red Dexters can carry the dun gene and not express it. This is why it should be relatively easy to find the Dexter Dun factor within the main Red Poll population. It is an individual gene within the Dexter breed, which has been declared unique to Dexters, that is being studied in order to determine where it entered the Dexter breed from.

Faun is the wild red of the Jersey breed and has often been confused for dun .

It is not only the colour dun that matches red poll origin the deep red dexters with a cream switch are tell tale. This forum has made mention of linear assessment of late note the skeletons of dexters containing high percentage Plover blood are also often distinct, the pelvis in particular. Early images of Dexters clearly show high hook bones not the low rounded ones so apparent today. The cow below has a high percentage plover.

[kansasdexters]

Perhaps I can be of help in answering some of your questions JohnO. First, I would like to direct you to the American Kerry and Dexter Cattle Herd Book, Volume 1, published in 1920. It is posted on the American Dexter Cattle Association website, www.dextercattle.org , on the Dexter Archives web page.

In this document, you will find descriptions of the original imported cattle and their progeny born and registered with the American Kerry and Dexter Cattle Club (I will refer to this registry as the AKDC) between 1911 and 1920.

Here are some of the animals that may interest you:

Belle of Hempstead House, AKDC #215, calved March 25, 1917
Body brown, switch black, udder some white
Sire: Castlegould Carson, AKDC #23
Dam: Castlegould Bluebell, AKDC#73

Deborah of Hempstead House, AKDC #212, calved January 4, 1917
Body red, switch black, udder white
Sire: Castlegould Carson, AKDC #23
Dam: Castlegould Dixie, AKDC #65

Castlegould Dixie was sired by Tom Thumb, AKDC #1, IRE #542 and out of Castlegould Dainty, AKDC #13. Here are their pedigrees:

Castlegould Dainty, AKDC #13, Calved November 28, 1907
Entirely black
Sire: Gort Tom, IRE #521
Dam: Summerhill Annie, IRE #2138
Breeder: Samuel Gibson, Dunmurry, Ireland

Tom Thumb, AKDC #1, IRE #542
Entirely black
SIre: Kenmare George, IRE #471
Dam: Kenmare Mary Ann, IRE #1883
Breeder: George Courtney, Kenmare, Ireland

When I read of a Dexter that is described as being red, but with a black tail switch, I don't think this animal was really red, I think that she was probably dun. When I see that Deborah of Hempstead House is a half sister to Belle of Hempstead House (same sire, different dams) and Belle is described as being brown, not red, I think that Deborah was probably reddish brown, and not really red. Note that both of these cows were the progeny of animals that were described as being entirely black.

Spend some time looking at the descriptions given in this old 1921 Herd Book and it may answer many of your questions.

Patti Adams

ADCA Historical Committee Chair

[JohnO]

Carol D,
I have seen all living Cornahir cows and all meet the 42 inch breed standard. I spoke with Prof. Couper regarding his herd and he said there has been no significant change in the size of the animals in his herd over the years and asked which animals you measured?

Kansasdexters,
I consulted with Mike Sinnott about the cow which was registered as brown.

Dun dexters don't have black switchs they are Dun so these animals were either variants of wild red or black/brown a dexter colour common in the UK to animals bred from the Doesmead line . These animals are often born brown with black points then gradually with age the body hair darkens . I have a cow, Moomin Nancy, which was born brown with black points and gradually turned black her first calf was born brown and as an adult would be described black.

Thank you for the information I was aware of your excellent archive facility well done to ADCA .

[Duncan MacIntyre]

Below is a pasted copy of a digest of the thesis produced for the Cardiff project. As you will see it looked at origins and inter and intra breed relations in general terms rather than specifics, so I don't think it really answers a lot of the detailed questions some of us would like. What it did do was establish the credibility of the Dexter as a breed more distinct from other breeds than many others are, with a good degree of genetic variation whilst not unduly affected by ingression from other breeds. This is really remarkable considering the bottleneck the breed came through in the 1960's and 70's.

I think it lets us look at the wider breed today with less concern over introgression etc than we might think, and thus accept totaly within the breed those individuals with small amounts of other breed in the background. If we look on the "original population" as the only "true" Dexters we are left with a breed so small as to have a very doubtful future. Each of us will have different priorities when breeding for our own herds. My main aim has been to preserve the small stature of the Dexter whilst avoiding the bulldog gene. In achieving what I have so far I have had to make a compromise and use a Dun bull extensively as he was the only bull I could get matching the size requirements. Purely for my own personal preference I would love to be able to select now against the dun and have a predominantly black herd of small size. When I say small size I mean a size which matches the old accounts of Dexter sizes and more particularly weights as weights are more often quoted in old breed descriptions than size.

As far as I am aware the results of the Cardiff project are contained entirely within the thesis for Tim Bray's PhD. I think the whole thing was available as a disc from the DCS office, and was available to any member who asked. I am not sure if there are still copies left, but if not the thesis will be a public document and should be available. I have posted here a very condensed report from it which does not include much of the technical genetics stuff which I have to admit I struggle to understand myself. I did enrol for a cattle genetics course at Glasgow Uni last year, but had my money refunded when it turned out I was the only applicant.

Report for Dexter Cattle Society
Report contents
This report concisely describes the findings of the genetics project. The report is in two sections, the first concerning a broad look at the Dexter among other breeds and the second describing the investigation of the Dexter in reference to its proposed ancestors. Much of the technical detail is absent from this document but is available in the thesis itself.
Dexter in a wider context
1.1. Summary
We looked at the Dexter in the context of a group of 27 distinct original traditional cattle breed populations, many of which have become minority breeds in recent times. Despite potentially high levels of gene flow between some breed populations, all maintain identifiable genetic distinctiveness. The Dexter was seen to be of high diversity but not the highest, and was seen to be of a unique genetic composition.
1.2. Introduction
The Dexter was investigated with respect to a number of other original population traditional breeds. In this way we intended to see whether there were strong relationships and even if the Dexter in particular did separate from the others, as it is possible that some breed populations have overlapping genetic compositions due to crossbreeding. We also looked at breed diversity measures.


Table 1.1. European breed genetic diversity (measured by expected heterozygosity)

Breed Population Diversity measure (HE)
Guernsey 0.71
German Black Pied 0.70
Red Devon 0.69
Aberdeen Angus 0.69
Limousin 0.68
Dexter 0.68
Milking Devons 0.68
Angeln 0.67
Welsh Black 0.66
Jutland 0.66
Berrenda 0.65
Shorthorn 0.65
Red Poll 0.65
Lincoln Red 0.63
Island Jersey 0.62
Kerry (UK) 0.62
Sussex 0.62
Mainland Jersey 0.61
Hungarian Grey 0.59
Belted Galloway 0.58
British White 0.58
Shetland 0.57
N'dama 0.57
Traditional Hereford 0.56
Gloucester 0.55
Irish Moiled 0.53
White Park 0.53

Figure 1.1. The position of the Dexter relative to the wider breed dataset as determined by the clustering application. The Dexter occupies a position among British breeds which is separate from a partition containing all of the European and several of the British breed populations.

1.3. Discussion
In this large number of breeds, the Dexter did not associate closely with any other single breed population. This suggested a closer look into the Dexter breed itself was necessary as well as into its documented ancestors. Almost all of the breeds here have high variability (table 1.1) and it is possible to see genetic distinction between European and British breeds. The genetic legacy of European ancestry is identifiable in some British breed populations but the Dexter was clearly grouped with other British breeds (figure 1.1). The consequences of maintenance of traditional breeds such as the Dexter at small and sometimes fluctuating population sizes does not seem to have affected their diversity although gene flow between some of the breeds has almost certainly artificially raised this diversity. The generally high levels of variation found suggest that these populations are likely to survive well provided the populations continue under careful management schemes. But the genetic status of breed populations is entirely dependent on the human interest in the breed, and it is imperative to promote this to maintain the diversity of breeds seen today.



The population genetic effects of ancestry and admixture in the Dexter breed
2.1. Summary
The genetic structure of the Dexter, a minority cattle breed with complex demographic history was investigated. Modern representatives of two putative ancestral populations, the Devon and Kerry (as the best modern representative of the original Irish black cattle), together with the different populations of the Dexter, which have experienced different demographic histories, were analysed. Distinct genetic groups were detected within the Dexter which could be attributed to known demographic events. Much lower diversity was identified in three isolated Dexter populations as well as higher differentiation of these populations from other subgroups. For one of these populations, where strong selection has taken place, we also found evidence of a demographic bottleneck. Methods for quantifying breed admixture were applied. High variation in results was shown, giving a consistent answer only in the case of a group consisting of selected Traditional Dexter animals, where the ancestor of the modern Kerry breed was determined as the greater parental contributor to the Dexter.
2.2. Introduction
The Dexter cattle breed provides a good example of the population dynamics found in a number of minority domestic breeds. The Dexter itself is thought to have been formed in Ireland from a Celtic black cattle ancestor, the most direct modern descendent of which is thought to be the Kerry. The Dexter is thought to have become gradually demographically separated from the Kerry as it became established in England, Wales and Scotland and is also reputed to have received genetic input from the old Devon breed shortly prior to the creation of the Dexter herd book in 1890 (Wilson, 1909). As is typical of many livestock breeds, the modern Dexter is divided into different populations or herds that may be separated from each other and have had their own, sometimes independent, demographic history.
This study investigates the origin and relationships of the different subgroups within the Dexter. We predicted that there would be a number of distinct groups within the Dexter due to the isolation of some herds. Genetic distinctiveness would be expected to be greater in isolated populations and for those under strong selective breeding programmes. The Dexter as a whole could be expected to retain high overall levels of genetic variation as we saw was present when it was seen with a larger number of breeds. The Kerry breed was predicted to be the most similar to the Dexter’s ancestor due to its close historical associations with the Dexter. The questions asked were; what subdivision can we find in the Dexter, and how does the breed relate to its proposed ancestors the Kerry and Devon breeds?
2.3. Samples used
Individuals representing the oldest form of the Dexter (not known to have been upgraded in herdbook records) were collected to form a sample of ‘Traditional’ Dexters. The oldest single herd represented in the breed today is the Ypsitty herd which was sampled as an example of a traditional yet demographically isolated population. American Dexters were sampled to represent a population established through exports prior to 1914. The Woodmagic herd was sampled as an additional isolated population receiving no genes from the rest of the breed for over 25 years and subject to strong selection. A larger group representing the wider Dexter breed as it exists today was also used. We used both existing Devon types which includes the milking Devon which is present only in North America today, and the British beef Devon. Also sampled were animals from the British Kerry breed.
2. 4. Methods used
Population structure was investigated and any groups of genetically similar individuals were grouped together. General genetic diversity estimates were made, as well as comparisons of the populations with each other, and an idea of whether populations were inbred or outbred. We also looked at whether the populations showed any sign of being down to very low numbers in their past, otherwise known as a genetic ‘bottleneck’.
In addition to other methods we looked at whether it was possible to determine the contributions of our proposed parent breeds, the Devon and Kerry. For this we used specific admixture analysis techniques, where we assumed the Kerry and Devon exchanged genetic material (admixed) to form the Dexter breed. We applied this approach to each subdivision of the Dexter.
2.5. Discussion
2.5.1 Diversity, subdivision, and population bottlenecks
The genetic information suggested that the Dexter was divided into five subpopulations when each showed a different genetic composition. Table 2.1 shows that genetic diversity varied widely across these populations. Diversity was lowest in the Dexter subpopulation groups that have been isolated for longer periods from the rest of the breed (Woodmagic, and Ypsitty, & American Dexter) compared with the remaining populations(Dexter breed population, Kerry, & Devon). The Traditional Dexter and main Dexter breed populations demonstrated extremely similar levels of diversity to a previous study of Wiener et al. (2004) on the Dexter breed using many of the same markers.


Table 2.1. Diversity measures (expected heterozygosity), and mean number of alleles per locus for all populations

Population HE Allele/locus
Dexter population 0.68 6.41
Kerry 0.64 5.00
Traditional Dexter 0.63 4.64
Devon 0.63 5.64
American Dexter 0.55 4.18
Ypsitty Dexter 0.54 3.68
Woodmagic Dexter 0.52 3.41



Figure 2.1 The process of partitioning breed groups into genetically distinct populations; each coloured vertical bar is an individual and each panel demonstrates a separate analysis. The different colours within a panel determine the membership of individuals to a genetic group. Where a genetic group is identified the group is removed and the remaining individuals reanalysed. Of particular interest here is the presence of two main genetic groups across the Traditional Dexter and Dexter Breed populations which can be seen as the blue and yellow colours in the bottom panel.

There was a high degree of similarity between the Traditional Dexter and the Dexter breed group, all other comparisons showed that populations were significantly different. As expected, the most divergent populations were also those that had the lowest levels of within-population variability. The large difference between the Woodmagic Dexter and all other populations, both Dexter and non-Dexter, is presumed to be a result of small founder number, extended isolation from the UK herd with a small population size, and the application of strong selection (Rutherford, 2005). The level of divergence of the Woodmagic from the rest of the Dexter is at least at the level of among-breed values seen for the other samples here, and for values observed among other commonly studied European breeds (Canon et al. 2001). Similarly, the Ypsitty population represents the oldest Dexter herd, older extant animals represent the closed nature of the population through unique genetic composition. Of particular interest here is the identification of two separate ‘Bayesian’ clusters within the Dexter breed and Traditional groups (figure 2.1). These are genetically distinct groups that have been recognised in the analysis. The existence of a split of this nature might suggest that there is some kind of historical population divide in the Dexter breed the legacy of which remains to be seen today.

The American Dexter underwent a bottleneck that took place when it was founded, and showed a greater similarity to both the Dexter breed and Traditional Dexter groups and the other breeds than the Woodmagic herd. This lower degree of differentiation from UK populations may be explained by a combination of cross-Atlantic gene flow, (at least eleven animals are documented as receiving export licenses for the United States between 1951-1988 (Dexter Cattle Society Herdbook; 1951, 1988)), chance retention of similar allelic spectra over time, or that the bottleneck was not as strong as originally believed. The latter suggestion is supported by the failure to detect the genetic signature of a bottleneck in this sample. However, uniquely within this dataset, the American sample analysed represents a low proportion of the extant total (approximately 6,000 individuals) and therefore inference may be improved by using an extended dataset.
There was weak evidence of a population bottleneck found in all of the populations except the American Dexter. There was also some evidence of a recent population expansion in the American Dexter, Kerry, and Devon breeds. The only consistent bottleneck signal present was detected in the Woodmagic herd which is known to have been founded by only five individuals (four females and one male) (Rutherford, 2005).
2.5.2. Parental contributions and admixture
The admixture methods did not converge on one ancestral population being the major contributor to all of the Dexter groups, although there was a high degree of agreement between two of the methods. There was consistency among all methods that the Kerry was the greater contributor in the Traditional Dexter group only, even considering the large error margins. One of the methods also suggested that the Dexter breed group has greater Kerry influence. Conversely, the American Dexter population appears to have a greater contribution from the Devon using all of the methods. The Traditional Dexter group was sampled from different pedigree lineages with no evidence of introgression and is therefore expected to reflect the historical links with the Kerry breed in the least biased way. That even this group has large variation in both parental proportions and within-method variance highlights the problematic determination of ancestry. The differing amount of genetic loss that is likely to have taken place in the history of the different groups makes these results difficult to interpret. For domestic breeds in general and for the Dexter in particular, due to the short timescale in generations and the relatively small effective population sizes involved, genetic loss (called ‘drift’) is more likely to have generated the observed patterns of differentiation than new variants arising (by normal mutation events).
Table 2.2. Relative parental contributions from Kerry (variance in parentheses) for each Dexter population as determined using the programs, ADMIX2.0, LEADMIX, and LEA

Hybrid population
Method Traditional Breed Ypsitty Woodmagic American
ADMIX2.0 0.67(0.10) 0.52(0.10) 0.50(0.10) 0.42(0.13) 0.22(0.07)
LEADMIX 0.86(0.32) 0.98(0.27) 0.58(0.40) 0.58(0.38) 0.25(0.22)
LEA 0.63(0.13) 0.46(0.11) 0.39(0.17) 0.41(0.15) 0.22(0.13)


The variation in contributions for the different hybrid populations could be explained by three non-exclusive factors. First, if, as seems likely, there was a large amount of genetic variation across the Dexter herd prior to the demographic bottleneck in the 1960s, sampling (founder) effects could explain differential contribution to lineages which became demographically isolated. Second, genetic influence from other breed sources (introgression) into the separate populations may have occurred during or shortly after foundation of some of the groups, though none is documented for any except the Dexter Breed group. Finally extreme genetic loss (drift) may by chance have obscured the true parental contributions of the different groups sampled. It should be stressed that if genetic drift is a major feature of the data, then most of the information related to the admixture (or any other) event will become lost as time increases from the point of admixture. Thus, in cases of extreme drift, it may simply have to be acknowledged that limitations to the methodology exist because information lost cannot be retrieved, at least with genetic data alone.
For management purposes it should be recognised that there are measurable differences between subpopulations of the breed, and that these may be partially due to the contributions of ancestral populations. If the Traditional Dexter sample is accepted as being representative of the original breed, the implication is that other subpopulations have diverged from this specific type. However, avoiding the widespread use of individuals from these subpopulations may risk the loss of potentially unique genes (when considering present-day populations as random samples from the diversity present in the Dexter at its formation). This concern is a potentially valid one provided introgression from other breeds has not contributed to this divergence of subpopulations. The issue of the main breed population being potentially divided into two groups is one that may not require management action. It may be that it is the consequence of a historical divide that no longer exists and is therefore not being perpetuated. If action were to be taken, one clear recommendation would be to avoid perpetuating the divide by ensuring that animals with high membership of the same group do not reproduce with one another.
What is clear from this study is that the admixture approaches used here limited our conclusions regarding the relative parental contributions. Developments such as the ability to account for many admixture events and to account for multiple parental populations would be useful in this context but whether this will increase the precision in admixture estimates needs investigation. Small population variation and the implications of genetic drift, particularly during and after population bottlenecks, still presents a major challenge. Genetic data such as those presented here therefore cannot by themselves answer all the questions relevant to the Dexter breed conservation.
2.6. Acknowledgements
We thank the Rare Breeds Survival Trust, Dexter Cattle Society, Instituto Gulbenkian de Ciência, Université Paul Sabatier and Cardiff University for funding and infrastructural support for this research, which forms part of TCB’s PhD. Part of this work was carried out during visits by TCB to Lisbon and Toulouse. Thanks go to A. Coutinho and B. Crouau-Roy. LEA calculations were performed using the High Performance Computing resource at the Instituto Gulbenkian de Ciência (IGC) with the help of P. Fernandes and using the Condor cluster at Cardiff University with the help of S. Adams (School of Biosciences) and J. Osborne (ARCCA). Some of this manuscript was written while L.C. was visiting IGC: the CNRS and B. Crouau-Roy are thanked for making this possible.

2.7. References
Canon J., Alexandrino P., Bessa I., Carleos C., Carretero Y., Dunner S., Ferran N., Garcia D., Jordana J., Laloe D., Pereira A., Sanchez A., & Moazami-Goudarzi K. (2001) Genetic diversity measures of local European beef cattle breeds for conservation purposes. Genetics, Selection, Evolution 33, 311-332.
Dexter Cattle Society Herd Book Volume LI (1951) Dexter Cattle Society
Dexter Cattle Society Herd Book Volume LXXXVIII (1988) Dexter Cattle Society
Rutherford B. (2005) My Love Affair With The Dexter (publisher Triple D Books, Wagga Wagga, NSW, Australia, ISBN 0 9756829 0 3
Wiener P., Burton D., & Williams J.L. (2004) Breed relationships and definition in British cattle: a genetic analysis. Heredity 93, 597-602.
Wilson J. (1909) The origin of the Dexter breed of cattle. Scientific Proceedings of the Royal Dublin Society 12(1) 1-17.

[Duncan MacIntyre]

Some of the tables and illustrations from the report have not copied well, or been missed altogether in the pasting, but the text is all there.

Duncan

[kansasdexters]

Hello JohnO,

The yearling heifer in the following photo is a DNA-tested (ED/ED, b/b), homozygous dun Dexter with a brown body and a dark blackish-brown switch:



We've had several purebred, registered dun Dexters with very dark tail switches calved on our place in the past 6 years. The switch does eventually become more brown than black, but it can take several seasons for this to happen. I can easily understand how a breeder in the early 1900's might have described such an animal as having a brown body and a black switch when they registered it as a young animal.

Anyone else have some dun Dexters with very dark tail switches, or this unique to the United States herd?

Patti