Stigmaria and stumps of clubmoss trees II   I               

New insights
Hetherington et al (2016) examined the appendices using compression fossils and a large number of coal balls from England and they came up with surprising results. While it was generally assumed that the appendices branched off only sporadically, they saw that they could split up to four times. Also, the appendices turned out to be much longer than previously assumed: some became as long as 90 cm. Another striking observation was that the appendices became about 25 % thinner with each splitting and that the thickness of the part between two splits did not change.

The most important discovery is the presence of root hairs, which had not been observed before. This makes them look even more like real rootlets. These root hairs only occur on the last few branchings, which are known only from coal balls. On the basis of the number of splits they come to the conclusion that there were five to six times more appendices per unit length than when they where undivided. So the trees had a very dense network of appendices which enabled them to absorb nutrients and water from the soil very well. The Stigmaria axes with appendices attached to them formed a large root plate that could extend up to 12 m from the trunk. As a result, the giant trees were well anchored in the swamp.
In addition, the trees grew close together. We know this from, among other things, the discovery in 1887 of a fossil 'forest' of tree trunks in the Victoria Park in Glasgow. The 11 stumps uncovered there in a former quarry show that the trees grew quite close together so that the root plates were probably fused together (Figs. 11 and 12).

Interestingly, the only surviving relative of Sigillaria, the genus Isoëtes or quillwort, has appendices with the same characteristics: they fork four to five times and with each splitting they become about 25% thinner. They also remain equally thick between two splits. They also have root hairs, just like the clubmoss trees.

Fig. 11. Reconstruction of the Stigmaria system. A. The network of appendices, which could be up to 90 cm long.  B. The root plate and the forest. C. Appendix with root hair and vascular bundle. D. Ditto with splitting vascular bundle.  From Hetherington et al. (2016)

Fig. 12. Photo from 1887 of The Fossil Grove, the petrified 'forest' of stumps of clubmoss trees in Victoria Park in Glasgow. Nowadays part of it is covered and can be visited.

Complete root systems
More or less complete Stigmaria systems have relatively rarely survived. A well-known example is the one in museum Am Schölerberg in Osnabrück (Fig.13). It was discovered in 1886 in the Piesberg and put together in parts. The diameter of the system is about 8 m and it weighs three tons.

A very beautiful root system was (also) found in 1886 in a quarry in Clayton (Yorkshire) by Prof. Williamson. He had it excavated and set up at his own expense in the Manchester Museum (fig. 14). The original diameter was 9 m but due to lack of space it is limited to 6 m in the museum.

Fig. 13. The Piesberg root system in the museum Am Schölerberg in Osnabrück. Diameter about 8 m.  Photo A. Leipner.

Fig. 14. Stigmaria in the Manchester Museum, from a quarry in Clayton. Photo: Manchester Museum/University of Manchester.

Thomas & Seyfullah (2015) describe a fossil 'forest' found in a quarry in Brymbo, North Wales. It consists of 20 upright trunks up to 1.5 m in diameter and up to 2.5 m high. However, these don't have preserved Stigmaria. In a somewhat higher layer a fairly complete root system has been found which has been completely uncovered and from which they were able to draw conclusions about the way in which this system has been preserved.
The twenty tree trunks preserved without Stigmaria were found in a 2 m thick layer of shale/claystone that turned towards the top into the slightly less fine-grained siltstone. The trunk with the Stigmaria was found in a layer of sandstone. It has a diameter of about 5 m and the trunk is 50 cm thick and 1.7 m high. The root system has been completely excavated to preserve it (Fig. 15).  A loose Stigmaria axis of 8 m long was found in the same layer.

How does such a petrified 'forest' arise? It is generally assumed that the decaying parts of the trunk and root system are filled with sediment that subsequently hardens. However, the question is how the shape of the trunk and the underground parts are preserved in this process. In the Brymbo fossil a thin hard brown layer of iron oxide (FeO(OH)) was found on the outside. This mineral is also often found at the bottom of lakes and creeks and is formed when ferrous water comes into contact with oxygen from plant cells. This layer can ensure that the shape of trunk and Stigmaria remains intact when the interior decays and is replaced by sediment. The ends of the root systems are almost always missing, which is almost certainly due to the fact that they consisted of softer material and therefore decayed faster.
This process may also have occurred in some other systems.

Fig. 15. Stigmaria system of Brymbo. From Thomas & Seyfullah (2015). Photo: B. Thomas.

Fig. 16. The layers from the Carboniferous near Joggins (Nova Scotia, Canada). A group of excursionists look at a tree stump in the deposit. Photo: L. Nichol.

Joggins
At Joggins in Nova Scotia in the far north-east of Canada, the process has certainly been different. For almost two centuries, the coast near the town of Joggins has been a famous site for carboniferous clubmoss tree trunks and stumps (Fig. 16). Due to tidal differences of up to 12 meters and violent storms and rains, the cliffs there crumble regularly. New trunks constantly emerge, which disappear in an average of three years due to erosion. What has made the place especially famous is the fact that remains of carboniferous quadrupeds have been found in some of these trunks. The most famous is the Hylonomus lyelli, the oldest reptile found so far. The 20 cm long animal was discovered by Dawson in 1852, and described by him in 1860. Not until 100 years after the discovery it was discovered to be a reptile.
The famous geologist Lyell was collecting there together with Dawson.

Due to rising and falling sea levels in the Late Carboniferous, the marshes drowned from time to time and were re-formed. The flood plains were covered by enormous mudflows, with the present forests being embedded in sediment. That the sedimentation at Joggins was very fast is shown by the fact that the deposits of 900 meters were formed in one million years. After a short period of time, the mud hardened and encircled the lower few metres of the clubmoss trees. The soft material, which the trunks largely consisted of, rotted away and holes came into existence in the ground. It has always been assumed that animals fell in, died and were embedded in the next muddy stream. But a new theory states that the animals may have used the hollow trunk holes as hiding or living places.

The Joggins coast was placed on the World Heritage list in 2008.

Probably the formation of fossil 'forests' went in most cases comparably to Joggins, albeit in a more moderate form because there are no other known places where fossil animals have been found in the trunks.

Conclusion
Coal balls, Stigmaria systems and fossil 'forests' of tree stumps tell a lot about the structure and growth of the clubmoss trees in the Carboniferous. The trees stood quite close together, with extensive root plates anchored in the swampy ground. And although there was a lot of light on the ground,  undergrowth was only in open, often higher areas. Mudslides, caused by faster soil subsidence or rising sea levels, destroyed the forest but bedded the lower parts of the trunks and the root systems. In the hardening sediments the trunks and Stigmaria's rotted away, creating cavities that were later filled with new sediment. Because there was a discontinuity between the casts of the tree parts and the surrounding sediment, the forms could continue to exist for hundreds of millions of years.

Literature
Calder, J.H.; Gibling, M.R.; Scott, Andrew C.; Davies, S.J.; Herbert, B.L., 2006 A fossil lycopsid forest succession in the classic Joggins section of Nova Scotia: paleoecology of a disturbance-prone Pennsylvanian wetland. In Wetlands through Time. ed.  S.J. Greb; W.A. DiMichele. Vol. 399 Geological Society of America Special Publication, 169-194.

Falcon-Lang H.J., Gibling M.R., Grey M. ,2010. Classic localities explained 4: Joggins, Nova Scotia. Geol Today 26(3):108–114

Hetherington,  A.J.,  Berry, C.M. , Dolan, L., 2016. Networks of highly branched stigmarian rootlets developed on the first giant trees. Proc. Natl. Acad. Sc.i USA, 113(24): 6695–6700.

Stewart W. A., 1947. A comparative study of stigmarian appendages and Isoetes roots. Am J Bot. 34(6):315–324.

Taylor, T.N., Taylor E.L. & Krings, M., 2009. Paleobotany: The Biology and Evolution of Fossil Plants [2nd Ed]. New York: Academic Press.

Thomas, B, 2016. A Carboniferous Fossil Forest in North Wales: Problems and Potentials Associated with Developing and Conserving a ‘Soft-Rock’ Site. Geoheritage,  Vol. 8, Issue 4, 401–406.

Thomas, B, Seyfullah, L.J., 2015. Stigmaria Brongniart: a new specimen from Duckmantian (Lower Pennsylvanian) Brymbo (Wrexham, North Wales) together with a review of known casts and how they were preserved. Geol. Mag., 152(3): 858–870.