Tuesday, November 18, 2014

DNA - Shellmiddens or Shenanighans?

This post is a gripe at a new paper (which I largely agree with) (*update 1* -I've changed my mind-I'll put in the next post) called "The Known Knowns and Known Unknowns" which briefly discusses the odd presence of mitochondrial "haplogroup H" among the people of the Mesolithic Portuguese shellmiddens.  I've finally reached the maximum vomit threshold with shellmidden DNA, so please bear with me.

The progress of ancient DNA research in Europe is beginning to show that these results are indeed outliers.  It was already a logical fallacy to suggest Haplogroup H was in an Iberian LGM refuge and remained there at an extremely high frequency but failed to expand to the rest of Europe and then, by-the-way, remained relatively hidden in the midst of fairly typical Neolithic farmer genomes within the region and finally went gang-busters to the Russian plain in the Early Bronze Age.  But looking closer at the "modern" results from the shellmiddens and a little background on the bones (that were probably licked by half the Portuguese population) the results become even more suspect.

Skeletons 7 & 8 Amoreira 1933 (Jackes,

Chandler, Sykes, Zilhão, 2005

This important study became confirmation for the retarded notion that modern Western Europeans are in large part maternally descended from indigenous, Mesolithic hunter-gatherers.

Chandler's results showed that Mesolithic Iberian maternal lineages were virtually indistinguishable from high frequency modern ones of Western Iberia and gave credence to the theory that Iberia was an important glacial refuge (for modern Europeans) during the Last Glacial Maxim.

I will offer a very respectful criticism here, especially since ten years have past, but this study becomes increasingly problematic the more you look at it and it is the foundation for so much wasted debate.  I will say that it is critical for those interested in the genetics of Western Europe to know the archaeology and conditions from these early studies.

This research used available remains excavated between 50 to 130 years ago.  No remains tested, as I understand, were excavated at least within the last thirty years.

Of the fifteen Mesolithic burial sites in Portugal, only half are "dated" and "of those" half are from shell middens.  Several are on the Rio Muge at the mouth of the Tagus, the Sado Valley and a single Atlantic site at Alentejo.  Most of these sites were excavated beginning more than a century ago, although intermittent work has continued in the last six decades.
From these, Chandler describes most being from the Sado Valley, including Amoreiras (which I will pick on):
"Mesolithic sites were primarily sampled from the Sado valley estuary [Alentejo], including Arapouco, Vale de Romeiras, Poças de São Bento, Cabeço de Pez and
Cabeço das Amoreiras (São Romão) (Arnaud 1989, Cunha and Umbelino 1995-1997). The two sites sampled outside the Sado valley were the early Mesolithic site of Toledo in coastal Estremadura (Araújo 1998) and the coastal Algarve site of Fiais (Araújo 1995-1997, Morales and Arnaud 1990)... "

Let's start by looking at the materials taken from Cabeço da Amoreira [on the Sado, not similarly named site on the Muge].  Keep in mind the quotes you are reading are concerning other anthropological studies, not this DNA study, but will help you with context.
"The Muge shellmiddens were found 150 years ago in 1863 by Carlos Ribeiro. Since then, several projects in the area excavated the various sites at the archaeological complex of the Mesolithic Muge shellmiddens. The result was the recovery of more than 300 human skeletons. However, most of those burials have insoluble problems of associated materials, provenience, stratigraphy and chronology.(Bicho et al, 2013)
Also this from Diniz:
"More recently, a 14C date from human bones put Amoreiras [in Alentejo] first (?) occupation in the beginning of the 6th millennium cal AC and our pottery analysis clearly show that beside Early Neolithic ceramic, Final Neolithic pottery is also present both recovered in the same artificial levels giving Amoreiras a chronological and cultural complexity not expected before." (Meiklejohn et al, 2009)
You will notice that Diniz refers to the levels within the middens as "artificial".  So in Amoreira [Alentejo] we have Early & Final Neolithic materials within a supposed Mesolithic level, which was man made to begin with and where intrusive burials continued for a great amount of time. 

Properly stratifying an un-contextualized find within a midden depends highly on carbon dating to verify its "authenticity".  The majority of radio carbon dates from the middens have come from bone, which by the way, come from individuals whose diet consisted of at least 50% shellfish.

Given the delta between charcoal dates and bone dates in this area, its easy to lose confidence in early C14 dates of particular materials quickly, not that much confidence would be had in the bones of a brackish marine life eater. 

Christopher Meiklejohn address the increasingly erratic radio dates of Mesolithic remains throughout the Portuguese middens: (Mesolithic Mescellany V20-9, 2009).  Keep this in mind as you read further.

Aside from the fact that Cardial folk also used the middens for burial (more clearly in Northeastern Iberia), the processing of the material itself may has left numerous problems for further scientific research.
"In the absence of commentary on the date and possible contamination, especially the issue of the use of paraffin in the preservation of the burials [1865], interpretation of this and the other direct dates on Sado midden burials should be made with considerable caution." (Meiklejohn)

Viability and Provenance of DNA extraction from Muge and Sado middens:

In general, the remains from Amoreira were neglected for anthropological study for an especially poor state of preservation (Cunha and Cardoso, 2001).   Because of this, they were contained in parafin, another issue.  The alkalinity of the middens is an issue.  Almost all of the material from Amoreira, Moita and Arruda were calcified in a dense matix.  (Cunha and Cardoso)

The remains from the Muge were spread across three museums, one of which caught on fire in 1974, causing multiple remains to become mixed.  (Cunha and Cardoso)  Some of the remains excavated at Amoreira were lost after excavtion, but some remains not from Amoreira somehow became associated with it.  Cunha and Cardoso also state (unrelated to the DNA study)
"Since some of the material lacked labels, some radiocarbon dates were performed in order to demonstrate their Mesolithic provenance (Cunha and Cardoso, 2002). The human bones retrieved at Amoreira were analised in this context."
Bear in mind that after the provenance of the materials being handled became questionable, the authenticity of the materials were reaffirmed through carbon dating.  They did exclude some skeletal material that came in a box labeled "Mesolithic" but later turned out to be skeletal material from the Iron Age that somehow became mixed in the Amoreira materials.  Those remains not calcified were easily removed from this association.

There are issues with how the bone material was cleaned in the months prior to DNA testing (unrelated to this study)
"Acetic acid seems to be efficient in removing the calcite, however, as secondary consequences are not fully known, we opted to study the material as it was."  (Cardoso in unrelated study)
**Update 2**  After re-reading Cardoso's comment it would appear that acetic acid was not used on remains after all, being as he said, the 'secondary consequenes not being fully known'.

There are other issues with these remains.  I don't want to come off like a chimpanzee on xanex, so I will leave it at that for others to study, but some of these early DNA studies from Southwest Europe and Italy need to looked at with a little more caution before we spend a lot of time debating European pre-history.


  1. "This important study became confirmation for the retarded notion that modern Western Europeans are in large part maternally descended from indigenous, Mesolithic hunter-gatherers."

    Although I accept your main point about the provenance your posts show there was at least a cultural distinction in Iberia between coastal trading settlements and inland settlements which might also suggest that the populations in the trading settlements were different or at least more cosmopolitan but also given how cities were generally population sinks being constantly replenished by migrants from the interior it's perfectly possible for the coastal cities to have been non-indigenous at some point but replaced over time from the interior.

    This doesn't necessarily apply outside Iberia (or Atlantic megalith generally) which seems to have this coastal trading colony pattern which to me looks like they may have originally been stopping points on the voyage to get amber (and later copper, gold, silver and later still tin). If they were primarily trading posts rather than colonies per se then it might not be surprising if there was more surviving HG dna in the interior of the Atlantic megalith culture.

    Also separate from that if the mdna in certain regions of Europe was largely indigenous from the mesolithic it doesn't necessarily mean it was indigenous to that region of Europe i.e.the indigenous mesolithic people of a particular region may have been replaced by farmers and then the farmers were replaced by people descended from mesolithic HGs who turned onto herders in an adjacent region i.e. continuity by cousin rather than continuity.

    So I don't think it's a retarded idea even if it's wrong.

    (Although having said all that personally I think mdna H is most likely from the Atlantic megalith people married into by BB males and it spread because it contained some advantageous farmer gene or other.)

    1. I think your phrase "continuity by cousin" makes a lot of sense in the scope of West Eurasia. This is probably especially true for many of the lineages and autosomal mixes reintroduced by later migrations such as the corded ware.
      A blogger at 'West Hunters' made a similar point http://westhunt.wordpress.com/2014/11/08/remix/

      My main complaint here is the shakiness of these Iberian studies. Essentially, I see little to suggest anything other than typical Mesolithic DNA but everyone else seems to be running with the herd.

    2. "My main complaint here is the shakiness of these Iberian studies."

      Fair comment,

    3. I share Grey's opinion that even if the studies are wrong, that their conclusions are not "retarded".

      I would critique the point that "mdna H is most likely from the Atlantic megalith people married into by BB males and it spread because it contained some advantageous farmer gene or other."

      mtDNA H is pretty definitively known to not have intrinsic fitness enhancing properties. mtDNA H may have ridden piggy back on fitness enhancing culture practices of people who happened to possession the fitness neutral ancestry informative mtDNA H haplogroup, or it may have been ancestry informative for a group of people who had some different fitness enhancing autosomals gene (like latose persistence) at a high frequency, but the only reason that mtDNA H is common in Europe is because it was common in population that was fit for reasons unrelated to their mtDNA itself.

  2. @bellbeakerblogger and Others
    Would You and Other guys like to take part in this Great Discussion?
    Good Day.

  3. I'm not sure you've understood the meaning of the term artificial as used by Diniz of the levels at Amoreiras. It doesn't mean that the deposit is artificial, i.e. man-made, but that the site was recorded by arbitrary levels (i.e. calling every 20 cms deeper a new level) rather than stratigraphically. Of course, that still means that skeletons need to be directly dated, and that those with potential problems of contamination should be avoided.

    1. Ok, thanks for pointing out the context. That makes sense, however it's still a midden. Also, I don't contest that these are mesolithic individuals (or rather pre-agriculturalists), it is decision to test remains that can be contested, especially when the results have the potential to be controversial.