Analysis‎ > ‎

Arbon Map (2013.09)

Notes on the Map of DNA samples in East Anglia,  both compiled by Stephen Arbon.

1. Objectives of the Map

To see if there were any significant patterns of the main haplogroups of DNA of project members and any correlation between them and geographical, historical or man-made features in East Anglia.

2. Study Area of East Anglia

East Anglia for the project includes:
  • Norfolk & Suffolk,
  • the eastern side of Cambridgeshire,
  • the northern side of Essex and
  • a south-western portion of Lincolnshire.
This more or less represents the greatest extent of the kingdom of the Anglii.
At one time the region was smaller, with the boundaries defined by several more defendable geographical features:
  • The Wash,
  • the eastern shore of the Fens,
  • a defensive dyke protecting the eastern [Roman] causeway,
  • the Devil’s Dyke stretching between the fens & the dense woodlands [thus protecting any approach from the south along Icknield Way],
  • the woodlands of the SW Suffolk and
  • the River Stour as the southern boundary.

3. Observations of Results

  • There are very few areas in common between sites of coin finds [of the British tribes] and dna samples [of any haplogroup].
  • In the rural parts of East Anglia, ‘R1b’ & ‘I’ haplogroups tend to be in separate bands.
  • There doesn’t appear to be any correlation between the distribution of dna samples and the main roads [as drawn in 1830], although there may be with some along the river valleys.
  • Project samples were generally from rural parishes, or Norwich & Yarmouth.
  • This may have been related to better soils for farming and hence the number of farmworker displaced in the 18th & 19th centuries.

• for ‘R1b’ haplogroup
  •  a cluster inland of the coast south of Cromer.
  •  a band stretching from Great Yarmouth SE thru to a just south of Cambridge.
  •  a band running west from Ipswich to Sudbury.
  •  a band running more or less in an arc about 30 km south & west of Norwich.
  •  a scattering of samples in the villages south of The Wash.
  •  most samples in Great Yarmouth are R1b.
  •  There doesn’t seem to be any pattern between the 2 main subgroups of R1b [although the numbers of each are still too few].
• for ‘I’ haplogroup
  • - a band along the valley of the River Stour, particularly towards the upper reaches.
  • - a thin band running N-S, well to the west of Norwich & the band of R1b.
• for other haplogroups [E1b1, G2a, J, Q & R1a]
  •  collectively, they seem to be well scattered and relatively few in number.

4. Preparation of the Information

4.1 The Plotting of Dna Samples on the Map

The location & haplogroup of the samples was copied from information listed on the spreadsheet from the East Anglia Project web-site [15 Feb 2009], and is based on what was provided by project participants, giving the baptism or birth [parish] of their earliest known male ancestor.
These locations were then plotted on a scanned copy of an AA 1:250 000 map.

4.2 Roads & Towns

Only the principal roads & towns, as they were in about 1830, are shown on the map .
These were copied from ‘A New and Comprehensive Gazetteer of England and Wales of 1834’ by James Bell, as printed in ‘Phillimores ATLAS and INDEX of Parish Registers’. [Although there’s a large network of local roads, paths & villages that haven’t been included].

The positions of markers should be close but may not be exact because:
  • of the small scale of the map.
  • and where there are a number of people from the same parish, these were grouped beside each other to give a better visual idea of both the relative strength and distribution of each haplogroup.
[In reality, the nominated location of samples would where the baptism was recorded, usually the family church; this may or may not be where the person was actually born.]

4.3 Rivers & Waterways

[yet to be completed]

5. Many Areas are Still Largely Under Represented

- Several populated areas which lie along the periphery of East Anglia, eg Cambridge & Colchester.
- Other areas may have only ever supported thin populations such as:

  • along the SE Suffolk coast,
  • low lying areas such as the Broads & the Fens and 
  • areas with poor quality soil such as the heathlands eg near Thetford.

- Some locations have no obvious reason for a lack of samples, eg Thetford, Claire, Haverhill etc. and may be just thru chance.

6. Sample Size and Probability

The number of dna markers so far submitted to the project only represent a tiny proportion of the population of East Anglia, ie a couple of hundred out of perhaps a half a million in the early 19th century. In addition, they are not a random sample. While this might not effect the accuracy of the overall results it certainly would if looking at any of the less populated areas.

6.1 Sample Size to Population

The sampling size needed for an accurate prediction of a population is related to the diversity in the parent population [ie the standard deviation], not its total size. Additionally, the minimum total sample size is recommended to be at least 50 points for statistical correctness, for a normal distribution of a population.

So in our case, the 140 participants would seem to give a reasonably size for Norfolk & Suffolk taken as a whole and assuming they was no inherent bias from the way they were supplied that might skew the results].

6.2 Randomness of the Samples

There are a number of factors that mean that the project samples aren’t truly random:

- Internal Movement & Migration

While the movement of people before the Tuder period is largely unknown, in later times, rural workers often moved in May within their ‘hundred’ depending on work on offer. Additionally there was a large drift out of Suffolk & Norfolk in the 18th & 19th centuries to other regions such as Yorkshire & London and to other countries such as Australia, Canada, & New Zealand. As well as the earlier movements of the religious non-conformers to the American colonies in the 16th-18th centuries.

If for instance, farm workers were more likely to have been descended from the Angles for instance, and they were the ones who migrated, than the results may be skewed in a particular way.

- Wealth

Those with spare finance [eg America] would be more likely to have had dna tests.

- Possible Effects of having the same Co-ordinator

As David Weston has been significant part of both East Anglia dna & an U106 dna projects, there may be a higher representation from that clad.

6.3 Effects of Chance

Based on the dna results sampled so far, one might expect for every 10 samples for a parish there should be:
  • 5 or 6 of R1b,
  • 2 or 3 of I haplogroup &
  • 2 from the other haplogroups.
But from the effects of chance, samples from any locality might vary.

7. Geographical Factors

7.1 Waterways

The river systems dominate the pattern of settlement in East Anglia, although in slightly varying ways to what would be the case in other regions:
  •  Because much of the region is prone to flooding, many settlements were built on higher ground [notably Ely].
  • On the higher land in West Suffolk, villages & roads were built on the chalky land above the heavy clay soils of the valleys.
  • Some settlements were sited using waterways as a natural defence.
  • Conversely rivers were also said to be the preferred route of entry by the Angles & Danes.
  • Inland waterways were the major transport system of bulk goods until the late 19th century.
  •  Many towns grew around convenient crossing/transhipment points.

7.2 Road Network and Towns
7.3 Elevation
7.4 Soil Types

[yet to be completed]

8. Geographical Distribution of Coin Finds of British Tribes

• Sites of coin finds of the Incenti & Trinovante tribes were scaled from maps from the ‘The Oxford Celtic Coin Index’ on

[A note from the website cautions that the maps display only a small proportion of coins found, ie only those with sufficient providence] and also the exact location may have been withheld to protect the site.

• Coin finds from Coritani & Cunobelin tribes were scaled up & then traced from maps in ‘’, sourced apparently from ‘Barry Cunliffe's Iron Age Communities in Britain’.

- The small scale of the original maps, mean that the positions of coin-finds are only very approximate, probably to a few km. Additionally, coins from each tribe were traced from separate maps which may mean that those shown close together could actually be from a common hoard.

9. Geographical Distribution of Roman Coin Finds

At this stage I haven’t been able to find any books or websites that include sufficient detail to accurately plot major Roman Coin Finds in East Anglia.

There is a web-site that showed significant areas of Roman coins in Britain, and I have traced the East Anglia section on to a separate copy of my map. But the very small scale of the original and the broad areas of Roman coin finds, mean that it is not all that useful. I’ve found a couple of books, but information that has been published on the geographical distribution of Roman coins in Britain seems minimal.

Observations on Coin Finds:
  • Roman coin finds tend to be spread more widely over East Anglia than British tribal coin finds.
  • The 2 largest concentrations of dna samples, at Norwich & Yarmouth, have no significant Roman coin finds.

Subpages (1): Arbon Data