A genetic perspective on the recent demographic history of Ireland and Britain

[Capitalis]

This is a continuation of a post I made in the Genetic history of Cambridgeshire before and after the Black Death (2024) thread, but is more focused here on the findings of A genetic perspective on the recent demographic history of Ireland and Britain (2024 preprint).

Here's a map of Ireland for orientation.

This isn't a perfect overlay of this paper's UMAP plot of PoBI data with a PCA plot of G25 data, but it seems a reasonable effort, with caveats.


[...]

Some unknowns:

* What do the "non-core" G25 Irish samples heading towards Wales represent? More Irish G25 coordinates with known ancestry would be useful; I used to have Garvan's but have lost them.
* Are some of the G25 Irish and Scottish samples crypto-English?
* Are the G25 Welsh samples "old" or "new" Welsh (I suspect the latter).

I went fishing for Irish G25 coordinates and present my catch below.

To answer my first question, What do the "non-core" G25 Irish samples heading towards Wales represent? It seems that they represent South Leinster.



The following PCA plot of supplementary data from The Irish DNA Atlas: Revealing Fine-Scale Population Structure and History within Ireland (2017) recreates the Isles structure found in my G25-based PCA plot above but is far less compressed, more clearly separating Ireland from Scotland from England. The gap between Leinster and "old" Wales is also more apparent.



I checked the papers and supplements of The Irish DNA Atlas: Revealing Fine-Scale Population Structure and History within Ireland, Insular Celtic population structure and genomic footprints of migration (2018), and A genetic perspective on the recent demographic history of Ireland and Britain, and couldn't find evidence for a genetic connection between Leinster and "old" Wales, despite the known historical potential of Cambro-Norman influence, and despite PCA plot appearances.

I included modern Normandy on the first plot above to show that the "non-core" G25 Irish samples are heading beyond Normandy, rather than towards it. If we zoom out, ignoring Wales and taking a wider European perspective, it becomes clear that these samples are actually heading towards Atlantic oblivion.



In summary, there is structure within Leinster that, despite appearances, can't obviously be connected to Wales by any of the methods devised by the papers named above, including most recently IBD-sharing in A genetic perspective on the recent demographic history of Ireland and Britain. Some quotes from this paper follow.

Within the N.Munster and N.Kerry communities, we observe an enrichment of Welsh surnames. In Wexford, we find an increase of English, Anglo-Norman and Scandinavian surnames which possibly may reflect older Anglo-Norman settlements in these regions. When we extended this analysis to individual surnames (Supplemental Fig. 3), we observed an enrichment of “Walshes” in N.Kerry and W.Leinster [...].

A PCA plot of FST values shows that Wexford (located in South Leinster) is no closer to England, and North Kerry and West Leinster are no closer to Wales, than regions of Ireland without such surname enrichment. So surname enrichment may or may not be a useful metric.

Haplotype sharing patterns within Ireland and Britain can provide insight into the sizes and movements of populations, providing context for observed genetic structure. Irish communities show relatively high IBD-sharing with little variation in sharing between Irish communities (Supplemental Fig. 4) indicative of low effective population sizes (Fig. 4). Using IBDNe, we observe that this relatively homogenous sharing pattern is reflected in the similar Ne estimates of the Irish  communities across the island 100 generations ago (Supplemental Fig. 7a-c, Fig. 4). Two-thirds of the Irish communities show a reduction in Ne 40 generations ago, specifically the Wexford, N.Leinster, and S.Munster genetic communities which show 38-72% reduction in population size between 100 and 40 generations ago (Supplemental Table 7). We see a further reduction of 10% in the effective population size in the Wexford genetic community [...].

This is interesting, as the unusual Wexford / South Leinster PCA plot position visible in G25 may be the result of a repeatedly reduced population size producing a pseudo-shift towards "old" Wales.

To further contextualise the structure and signals of isolation in Ireland, the migration rate surfaces (Fig. 5) show that the only stable migration corridor that consistently existed between Ireland and Britain is between northeast Ireland and southwest Scotland, demonstrating the isolation of the island. The within island migrational cold spots however shifted across time from central-West Ireland in the [1-3cM) IBD bin to isolating Leinster from the rest of the island 15 generations ago (Fig. 5), separating Kerry from Galway, which may explain the signals of isolation in Connacht.

So here we have recent isolation in Leinster and Connacht, which just happen to produce the two most extreme outlying samples (with known ancestry) in my G25 PCA plot above, namely Atlantean from South Leinster and FionnSneachta from Connacht.

This is something to think about when running and interpreting G25 / Vahaduo models, as FionnSneachta may receive high Scandinavian values and we might interpret this as Norse influence dating to the Viking Age, whilst Atlantean may receive high Welsh / English / French values and we might interpret this as dating to the Norman Age, when in fact we are seeing the outcome of very recent population dynamics that push particular groups away from the national "core", sans immigration.

Note also the persistent migration barrier between Wales and Ireland. That's a big no to the Cambro element of the Cambro-Normans.

It would be interesting to see a fully Co. Donegal sample in G25, the most isolated region of Ireland as per The genetic landscape of Scotland and the Isles (2019).



Overall, I really enjoyed A genetic perspective on the recent demographic history of Ireland and Britain and recommend anyone interested to read it. Hopefully you enjoyed my summary, even if I avoided the thorny issue of IBD interpretation. :-)

[Mitchell-Atkins]

If it’s not too much trouble, could you add me to your plots with the other members?

Interested in the Leinster info as that is where my 1812 patrilineal brick wall was born (Wicklow)

[Nqp15hhu]

My mothers’ sample could probably be used as a “Donegal” proxy.

On her side, I have a lot of 100% or 95% Irish matches from the Northwest of Ireland.

[Capitalis]

My mothers’ sample could probably be used as a “Donegal” proxy.

On her side, I have a lot of 100% or 95% Irish matches from the Northwest of Ireland.

Very good.

[Capitalis]

If it’s not too much trouble, could you add me to your plots with the other members?

Interested in the Leinster info as that is where my 1812 patrilineal brick wall was born (Wicklow)

Yep, I'm going to churn out some more IBD plots from this paper, so I'll make another G25 PCA for you then.

[Nqp15hhu]

My mothers’ sample could probably be used as a “Donegal” proxy.

On her side, I have a lot of 100% or 95% Irish matches from the Northwest of Ireland.

Very good.

Just trying to help

[Capitalis]

My mothers’ sample could probably be used as a “Donegal” proxy.

On her side, I have a lot of 100% or 95% Irish matches from the Northwest of Ireland.

Very good.

Just trying to help

I know. I meant "good to know".

[Capitalis]

We chose to split ROH and IBD segments into the corresponding bins (a) 1 to 3cM [1,3cM), (b) 3 to 5cM [3,5cM), and © greater than or equal to 5cM (≥5cM). Using the methodology above, these length bins correspond to 100 generations ago, 40 generations ago and 15 generations ago respectively. An important caveat is that the estimates have very wide distributions, as well as the aforementioned assumption of population size.

If we take a generation as 28 years, then we have:

[1,3cM) ≈ 100 generations ago ≈ 776 BCE
[3,5cM) ≈ 40 generations ago ≈ 904 CE
(≥5cM) ≈ 15 generations ago ≈ 1604 CE

As an alternative to the PCA plots of IBD sharing in Supplemental Figure 6 which include all of the Isles, I've created PCA plots isolating the IBD sharing at 3-5 cM for,

The Isles excluding Orkney and Isle of Man


Ireland and England


Ireland and Wales


All data is from Supplementary Table 4. Note that there is no visible Wexford-Wales or Wexford-England connection to the exclusion of the other regions of Ireland.

[Capitalis]

To further contextualise the structure and signals of isolation in Ireland, the migration rate surfaces (Fig. 5) show that the only stable migration corridor that consistently existed between Ireland and Britain is between northeast Ireland and southwest Scotland, demonstrating the isolation of the island. The within island migrational cold spots however shifted across time from central-West Ireland in the [1-3cM) IBD bin to isolating Leinster from the rest of the island 15 generations ago (Fig. 5), separating Kerry from Galway, which may explain the signals of isolation in Connacht.

The left image of Figure 5 is labelled as 1-3 cM, but in the methods section the authors state,

We first limited the IBD sharing summary dataset to pairs of individuals both of whom had associated geographic data. It was further subset by IBD length to determine changes in migration surfaces over time: [1,5cM) for migration rates from approximately 100 generations ago and ≥5cM for migration rates from approximately 15 generations ago.

so I'm not sure which the left image represents. Ideally, I would like to see a 3-5 cM plot to hone in more on the Viking / Norman Age, but I'm 99% certain there would still be a Wales-Ireland migration barrier.

[Capitalis]

As an alternative to the PCA plots of IBD sharing in Supplemental Figure 13 which include all of the Isles, I've created PCA plots isolating the IBD sharing at 3-5 cM for,

The Isles (excluding Orkney and Isle of Man) and Europe


This plot is reminiscent of the PCA plot of supplementary data from The Irish DNA Atlas: Revealing Fine-Scale Population Structure and History within Ireland, which recreates the Isles structure found in my G25-based PCA plot.



The Isles and France


Note that there is no visible Wexford-France connection to the exclusion of the other regions of Ireland. In fact, the opposite is true.

The Isles and Belgium


The Isles and Germany / Poland


The Isles and Denmark / Sweden


The Isles and Norway


The Isles and Finland


The Isles and Italy / Spain


All data is from Supplementary Table 8.

[alanarchae]

i’m pretty sure Cassidy found that a minority of irish late Iron Age (which is roughly parallel with Roman period in Britain) pulled away from typical (pure bronze age Irish) towards western Britain.

[Capitalis]

i’m pretty sure Cassidy found that a minority of irish late Iron Age (which is roughly parallel with Roman period in Britain) pulled away from typical (pure bronze age Irish) towards western Britain.

Difficult for me to comment, as it is well-known that the lab won't share data with other researchers and they embargoed the PhD, if that's what you're referring to.

I criticised the PhD PCA plots at the time as difficult to interpret.

[Cejo]

I find it interesting that Connacht plots close to N. Leinster and S. Munster in 1-3 cM plots, but much closer to N. Munster in 3-5 cM plots. The latter would coincide temporally not only with the "viking age", but also with Dalcassian ascendance, when the Dal gCais would have projected influence from N. Munster into eastern Connacht and C. Ireland (via the Shannon). Is it possible that influence could be reflected here?

   

[Capitalis]

I find it interesting that Connacht plots close to N. Leinster and S. Munster in 1-3 cM plots, but much closer to N. Munster in 3-5 cM plots. The latter would coincide temporally not only with the "viking age", but also with Dalcassian ascendance, when the Dal gCais would have projected influence from N. Munster into eastern Connacht and C. Ireland (via the Shannon). Is it possible that influence could be reflected here?

Your image is Figure 6, a PCA plot of regional European IBD sharing in the Isles, not a PCA plot of within-Ireland IBD sharing.

Here is the within-Ireland IBD sharing at 1-3 cM and 3-5 cM. They look similar.





Here is ≥5 cM.



Data is from Supplementary Table 4.

[Cejo]

I find it interesting that Connacht plots close to N. Leinster and S. Munster in 1-3 cM plots, but much closer to N. Munster in 3-5 cM plots. The latter would coincide temporally not only with the "viking age", but also with Dalcassian ascendance, when the Dal gCais would have projected influence from N. Munster into eastern Connacht and C. Ireland (via the Shannon). Is it possible that influence could be reflected here?

Your image is Figure 6, a PCA plot of regional European IBD sharing in the Isles, not a PCA plot of within-Ireland IBD sharing.

Here is the within-Ireland IBD sharing at 1-3 cM and 3-5 cM. They look similar.





Data is from Supplementary Table 4.

Thank you for the clarification. That seems to answer the question!