New study on Salisbury Plain zircon-apatite fingerprinting

The chalklands of Salisbury Plain, showing the 4 studied sites.

There is an interesting new article by Clarke and Kirkland which has a bearing on the bluestone glacial / human transport debate.  I greatly welcome this -- all new research should be welcomed, especially when it involves a new and innovative technique.

Here are the details:

Detrital zircon–apatite fingerprinting challenges glacial transport of Stonehenge’s megaliths. 
Anthony J. I. Clarke & Christopher L. Kirkland
Nature Communications Earth & Environment | ( 2026) 7:54

https://doi.org/10.1038/s43247-025-03105-3

Abstract

How Stonehenge’s building blocks arrived on Salisbury Plain remains debated, with glacial and human transport mechanisms proposed. Here we test the possibility of Pleistocene glacial sediment input using grain-scale U–Pb fingerprinting of detrital zircon and apatite from modern stream sediments surrounding Stonehenge. Zircon ages span 3396–285 Ma, with age peaks at ~1090, 1690, and 1740 Ma, matching the Laurentian basement of northern Britain. Salisbury Plain detrital zircon ages match those of southern British rocks sourced from the London Basin, implying local sediment recycling rather than glaciogenic transport. Apatite ages of ~60Ma reflect post-depositional U–Pb resetting, consistent with the distal effects of the Alpine orogeny. Collectively, our data show Salisbury Plain remained unglaciated during the Pleistocene, making direct glacial transport of Stonehenge’s megaliths unlikely.

I haven't had a chance yet to suudy the article carefully, but I am immediately struck by certain limitations. For example, the lead researcher, Anthony Clarke, seems to have investigated only four sites -- in river sediments near Salisbury, Amesbury, Andover and Warminster. It appears that the samples were "bulk samples" of 1kg, dug from sand banks beneath water level without regard for lateral or vertical variations in sand composition. 

In the studies, "detrital" grains (eroded particles in river sand) were compared against bedrock "fingerprints" from the presumed source sites in Wales and Scotland. We do not know what variables or uncertainties might have been introduced in this process. There also appear to have been no western controls;  without sampling glacial sediments further to the west (where ice is known to have reached), comparing local Salisbury sand directly to distant bedrock may overlook the "missing link" of intermediate glacial deposits.

There is also a substantial risk in using high-precision mineral dating on a small number of physical sites.  This can lead to "interpretative inflation," where localized findings are over-emphasised, and used to dismiss a number of broad geological or glaciological possibilities.

I do not see anything reassuring in the study on sampling and sample processing bias that might have been introduced into the study.  It's widely acknowledged that the zircon and apatite cargo on any given landscape will involve variable rates and patterns of sedimentation; what guarantee do we have that certain horizons rich in zircons or apatites, subject to hydraulic or temporal sorting, have not been completely missed when Anthony Clarke took his 1 kg samples from his four wet sites?  Lab processes are described in the article, but again results can be skewed by manual or device-based means.

Quote:  Salisbury Plain lacks undisputed tills, erratics, or other diagnostic indicators of glacial

activity..........    That is a matter of opinion.  Erratics are present on Salisbury Plain -- that is not a matter of dispute.  And the faceted, abraded and heavily weathered bluestone boulders that are abundant in the "bluestone collection" are very strong indicators of glacial activity -- conveniently ignored by the proponents of the human transport theory.  And the alternative?  Where is the evidence of human transport of the bluestones?  There is none.  And that is not a matter of dispute.   Extraordinary theories need to be supported by extraordinary evidence..........so it is still the case that the most parsimonious explanation of the presence of bluestone boulders, slabs and pillars on Salisbury Plain is glacial transport.

Another conclusion which needs to be questioned:  "..........our data show Salisbury Plain remained unglaciated during the Pleistocene, making direct glacial transport of Stonehenge’s megaliths unlikely."  

The data may show that parts of the Plain were unglaciated, and that glacial transport of erratics all the way to Stonehenge might not have occurred.  That does not mean that glacial transport of the bluestones cannot have occurred over part, if not most, of the diatance between the source area and the site of the Stonehenge monument.

And there are other unanswered questions too.  Why are there so many rock types (including soft sedimentaries and igneous rocks) represented in the Stonehenge bluestone assemblage and in the on-site sediments?  If the bluestones were chosen and quarried, as argued by Parker Pearson, Bevins, Ixer and others, why do they carry the key diagnostic features of glacially transported erratics?  And if their transport had nothing to do with glacial processes, why have they all travelled from west to east, which happens to be the direction of predominant ice flow?  And what about the cobbles and pebbles in the Stonehenge area that have nothing to do with the monumental monoliths?  How did they get here?

Anyway, I will revisit all of this when I have had a chance to go through the article in much more detail.

====================

PS.  In my searches I have come across this article which seems to support some of my concerns:

Sourcing the sand: Accessory mineral fertility, analytical and other biases in detrital U-Pb provenance analysis.  2020.  David Chew , Gary O’Sullivan, Luca Caracciolo, Chris Mark, Shane Tyrrell

Earth-Science Reviews, Volume 202, March 2020, 103093

https://doi.org/10.1016/j.earscirev.2020.103093


Abstract

Interpreting the wealth of new data derived from the diverse suite of modern single-grain provenance approaches available to a sedimentologist requires a thorough understanding of the potential biases in the information recorded by each mineral-provenance system. This review focuses on the various possible mineral-specific biases in U-Pb accessory mineral provenance studies employing the minerals zircon, rutile, apatite, monazite and titanite, focussing on biases resulting from variations in source-rock mineralogy  (fertility). Fertility is intimately linked to the mineral petrogenesis of crystalline basement sources, which is another key aspect of this review. This petrogenetic information, which often resides in the specialist petrology literature, has great relevance to fertility studies (particularly those measuring mineral content in modern river sediment using confluence and along-trunk sampling) as trace-element abundances and/or elemental ratios in many accessory minerals can be linked to specific lithologies. Other mineral-specific biases in single-grain provenance analysis considered include physical and chemical modifications both before and after deposition, while the diverse suite of modern single-grain analytical approaches also requires understanding of potential methodological and laboratory induced-biases. A series of multi-proxy provenance studies are presented where fertility bias apparently plays a significant role. In magma-poor metamorphic belts (e.g. segments of the Himalayas and Caledonides-Appalachians), it is shown that zircon growth is limited, and monazite, apatite or rutile associated with the youngest tectonomagmatic events are significantly more fertile.  Such multi-proxy provenance studies will be greatly aided in the future by high-throughput, coupled U-Pb age – trace-element analyses integrated with automated heavy mineral determinations employing highly efficient sample preparation protocols.