Vascular plants today constitute more than two-thirds of the global biomass, and include more than 380,000 different species. From the tiny Arabidopsis to the giant sequoias, all possess specialized water-conducting cells (tracheids and/or vessels) that form a tissue called the xylem. Water-conducting cells of the xylem are dead at maturity, forming a network of pipelines, but they are often associated with living parenchyma cells.
The oldest known fossils of vascular plants date back to the Silurian, more than 400 million years ago (see Fig. 1). During the Early Devonian, some of these plants grew larger and added to the primary water-conducting system a secondary xylem (also known as wood), produced by a lateral meristem.  But how many times did wood evolve? What were the water-conducting properties of different fossil woods? What interactions existed between the earliest woody plants and wood-degrading microorganisms? To answer these questions, we study anatomically preserved fossils: plant remains that have been infiltrated by water rich in minerals preserving their tissues in three dimensions. These fossils allow us to study very fine details of extinct plant anatomy, in many cases down to the cellular level. With this information at hand, we can better understand the evolution of plant vascular systems, not only from a structural but also from a functional perspective.  

One important feature that can be observed in certain well-preserved plant fossils is the presence of tyloses (Fig. 2) in the xylem. Tyloses are bubble-like projections of parenchyma cells that extend into a neighboring conducting cell. They can completely block the conducting cell and have a role in heartwood formation, protection against pathogens (e.g. fungi), and preventing the spread of embolism (i.e. air in the conducting cells). Thus, the fossil record of tyloses contributes to our understanding of the evolution of the vascular system, along with the mechanisms that protect it from biotic and abiotic stresses.

We recently described the occurrence of tyloses in Mississippian (early Carboniferous) woods from Australia and reviewed the documented evidence of tyloses in fossil plants. Tyloses in fact occur in most living and extinct groups of plants, but no evidence of these structures has been found to date in Devonian woods. This is somewhat puzzling because the oldest woody plants are early Devonian in age, and all prerequisites for tylosis formation appear to have existed, including vascular tissues with both conducting cells and closely associated parenchyma cells (rays), potential fungal pathogens, and physical stresses that could lead to embolism. This raises the question as to why no Devonian tyloses have been found to date? Maybe some physiological mechanism triggering the formation of these structures had not yet evolved. Maybe the Devonian fossils studied thus far were not well-enough preserved, or maybe they were prepared in a way that did not allow the observation of such delicate structures. It was this mystery that spurred us to look at newly discovered Late Devonian fossil plants which our group had collected in Ireland in fall 2021.

These new fossils are very interesting, but a bit tricky to prepare and examine due to their preservation in pyrite (Fig. 3). They first are embedded in epoxy (a resin) to prevent them from breaking apart. Then slices of 1–2 mm thickness are cut with a rock saw. To improve the contrast and enhance the contours of the fossil cells and tissues, these slices are then briefly put in warm nitric acid. After this, the slices of the fossils are mounted on a glass slide and ready for observation under the microscope.  

After several adjustments to the protocol, new slides of Irish Late Devonian fossils were ready for observation. The first fossils we looked at belong to two types of plants already reported from that locality, namely stems of the lycopsid Wexfordia and the archaeopteridalean progymnosperm Callixylon. Nothing new apparently, but one of the sections of Callixylon wood showed an area with what looked like many dark bubbles in the conducting cells (Fig. 4). On closer inspection, the bubbles turned out to be connected to ray cells. Their structure and the fact that they occurred only in one part of the wood ruled out the possibility that they were fossilisation artefacts. Rather, what we had finally found was the first evidence of tylosis formation in a Devonian wood!

In addition to being the oldest evidence of tylosis formation in a fossil wood, and the first from the Devonian, this discovery is also interesting because of the affinities of the plant that had produced the wood. Archaeopteridalean progymnosperms where among the first groups of plants that formed trees. They first appeared in the Middle Devonian, and by the Late Devonian they had become a major component of the vegetation worldwide. While they reproduced by spores (like ferns, for example), their vegetative body was quite comparable to that of trees we see today, with the presence of complex root systems, true leaves, and the production of a trunk with a large amount of wood. The fact that they produced tyloses is another "modern" aspect of their biology.

Our study illustrates how fossil plants can provide detailed "snapshots" of physiological processes even in (very!) deep time. This type of information allows us to understand fossil plants as once living organisms and reconstruct the evolutionary history of key biological processes that still exist and sustain plant life today.

Acknowledgements: This research would not have been possible without the support of many people, and we would like to thank all the colleagues who have generously shared their time and expertise, in the field and in the lab.

Decombeix AL., Harper C.J., Prestianni C., Durieux T., Ramel M. & Krings M. Fossil evidence of tylosis formation in Late Devonian plants. Nat. Plants (2023). https://www.nature.com/articles/s41477-023-01394-0