Caught between two mass extinctions: The rise and fall of Dicroidium

Print Friendly, PDF & Email

Chris Mays and Stephen McLoughlin (Sweden)

In the aftermath of Earth’s greatest biotic crisis 251.9 million years ago – the end-Permian mass extinction – a group of plants arose that would come to dominate the flora of the Southern Hemisphere. Recovery of the vegetation from the end-Permian crisis was slow; but steadily, one group of seed plants, typified by the leaf fossil Dicroidium, began to diversify and fill the dominant canopy-plant niches left vacant by the demise of the Permian glossopterid forests (Fielding et al., 2019). Eventually, Dicroidium re-established a rich peat-forming vegetation across Gondwana through the Late Triassic, dominating the flora between 30°S and the South Pole (Kustatscher et al., 2018). Indeed, few fossil plant assemblages of this age can be found in Gondwana that do not contain this plant.

The importance of Dicroidium is not just its role in showing biogeographic and tectonic linkages between southern lands or its value in determining the age of continental strata. Dicroidium and its associated plant groups were so successful that they were major contributors to the development of thick coal seams in the Late Triassic that are now mined to produce electricity.

Although Dicroidium is generally envisaged as a plant of cool temperate climates, the very first fossils that might belong to this group are from the Permian-Triassic transition of Jordan, located near the palaeoequator (Blomenkemper et al., 2018). Nevertheless, the distribution of Dicroidium soon shifted to high southern latitudes in the Early Triassic and they overwhelmingly dominated the southern vegetation by the Middle and Late Triassic.

The Dicroidium plant had leaves that were superficially fern-like (Fig. 1). However, unlike ferns, these plants produced pollen and seeds, and are commonly lumped into an ill-defined group called the ‘seed-ferns’. Dicroidium leaves are characterised by the presence of a distinctive fork near the base of the frond. Numerous types of Dicroidium are recognised, but one of the difficulties involved in separating species is their considerable variation in leaf form.

Many of the species established during the twentieth century have intergradational features. Moreover, some species may have hybridised, resulting in progeny that had different leaflet morphologies even on separate parts of the same frond (Fig. 1C). The world experts on this group, Heidi Anderson-Holmes and John Anderson, who have collected some 40,000 Triassic plant fossils during their careers, have recognised 23 species and 16 sub-species categories or ‘forma’ of Dicroidium (Anderson and Anderson, 1983).

Fig. 1. The same species of Dicroidium (D. odontopteroides) from three southern continents: (A) Birds River, South Africa; (B) Ipswich, Australia; (C) Transantarctic Mountains, Antarctica – this frond features both the typical rounded leaflets, but also elongated forms (arrow), which may indicate hybridisation.

Many of these species can be found all over southern Gondwana (Fig. 1). This suggests that these plants were part of a relatively monotonous flora that spanned the cool, high-latitude regions of the Southern Hemisphere (Fig. 2), much like the boreal conifer forests that dominate Canada, Scandinavia and northern Russia today.

Fig. 2. Map of Triassic Gondwana showing the distribution of Dicroidium.

But here, our modern analogues seem to break down. The climate of the Triassic was generally much warmer than today, with Dicroidium forests thriving at 80 to 85°S, and possibly stretching all the way to the South Pole. At these latitudes today, the south polar region is covered by ice sheets thousands of metres thick and no forests exist for thousands of kilometres in any direction. Moreover, the modern boreal forests are dominated by members of the pine family (Pinaceae), which have needle-like leaves adapted for cold temperatures and are mostly evergreen.

This contrasts with Dicroidium, whose leaves were broader and probably deciduous, being shed along with their fertile parts in the polar autumn. These features are like birch trees, which dominate some cool-temperate forests today. Polar forests, like those of Dicroidium during the Triassic, have been the norm since the first forest ecosystems of the Devonian, whereas forbidding polar ice caps have been the exception. Despite their common occurrence in Earth’s history, the specifics of how ancient polar forest ecosystems survived and thrived still elude us. The present is not always the key to the past, particularly when the present climate is exceptional.

Just how the leaves of these plants should be classified remains controversial. Some palaeobotanists have divided Dicroidium-type leaves into several additional genera (for example, Dicroidiopsis, Diplasiophyllum, Zuberia, Xylopteris, Johnstonia and Tetraptilon), but these all appear to be just morphological variants of the same basic leaf plan and can be comfortably accommodated in Dicroidium (Anderson and Anderson, 1983).

The practice of establishing all these redundant or synonymous genera based on very minor differences in leaf shape is problematic because it leads to ‘taxonomic inflation’. Thus, the disappearance of a single plant group in the fossil record might be mistaken for a mass-extinction of genera if all the redundant taxa are counted from the literature. At the other extreme, by adhering to general descriptions for a taxon, you may accumulate a broad range of different leaf forms under one generic banner, potentially hiding crucial differences between distinct natural plant groups. For example, this can result in species being recognised long after the rest of the genus has gone extinct, some of which are clear cases of convergent evolution, rather than survivors of an otherwise long-dead lineage. The ‘to-split-or-to-lump’ dilemma is particularly tricky when discussing plant leaves, such as Dicroidium, that have so few morphological characters.

Reconstructing Dicroidium

It is difficult to visualise the Dicroidium plant because few artistic reconstructions have been published (Retallack and Dilcher, 1988). As noted above, the Dicroidium plant seems to have been deciduous; the leaves are commonly preserved as stacked matted sheets, like thousands of autumnal forest floors frozen in time, with the leaflets blowing away in the wind once you crack open a rock along the bedding plane (Fig. 3).

Fig. 3. Paper shales from the Upper Triassic Leigh Creek Coal Measures. Like an autumn forest floor of the Triassic, compression and impression fossils of Dicroidium cover almost the entire area of the bedding surface.

The leaves have distinctive ‘odontopteroid’ venation – with several veins emerging from the base of the leaflets, and arching and branching towards the margins (Fig. 4).

Fig. 4. An impression of Dicroidium odontopteroides from New South Wales, Australia, with the characteristic venation pattern.

Fossil wood associated with the leaves has prominent growth rings indicating growth in a strongly seasonal environment. Some of the associated petrified logs can reach 10m long and over half a metre in diameter (Fig. 5) – so we can assume that these plants were probably large trees.

Fig. 5. A fossil log recovered from Dicroidium-rich Triassic strata at Leigh Creek, South

Palaeobotanists have also found the male, pollen-bearing, organs of these plants and given them the name Pteruchus (Figs. 6 and 7). The pollen grains contained inside these organs are winged, indicating dispersal by the wind and are given the name Falcisporites (Fig. 7). The female organs are given the name Umkomasia (Figs. 7 and 8) and consist of a stalk with numerous swan-like branches or ‘peduncles’ bearing rounded heads, each enclosing a single seed (Anderson and Anderson, 2003).

The seed-bearing heads of these organs commonly have a dimpled or wrinkled appearance (Fig. 7) suggesting they were fleshy or leathery in life, then splitting and shrivelling on releasing their seeds. The seeds were less than one centimetre long, pear-shaped and had narrow wings – again, like their pollen, for wind dispersal. This group of plants is of particular interest to palaeobotanists, as its possession of seeds tightly enclosed within a head or ‘cupule’ is similar to the architecture of the seeds of flowering plants (which have two tightly fused layers forming the seed wall).

Fig. 6. Pteruchus, the male pollen organ associated with the Dicroidium plant.
Fig. 7. A schematic reconstruction of the Dicroidium plant with different parts given separate names.
Fig. 8. Umkomasia (U. polycarpa), a male seedbearing
organ of the Dicroidium plant.

Dicroidium also has a role to play in understanding insect evolution. Since Triassic fossil deposits containing insects are few and far between, the feeding damage preserved on the leaves of this dominant plant helps us understand the guilds of insect herbivores that were active following the end-Permian extinction. Dicroidium leaves host almost all the major types of damage that we see in modern plants, for example, galling (Fig. 9A), hole feeding (Fig. 9B) margin feeding, leaf mining and surface feeding (Labandeira et al., 2018). Such damage enables the reconstruction of complex food webs from past ecosystems, even in the absence of the body fossils of the insects themselves.

Fig. 9. Insect damage trace fossils on Dicroidium reveal complex ecological interactions during the Triassic: (A) galling and (B) hole feeding.

Where can I find Dicroidium today?

Since Dicroidium was essentially a Gondwanan plant, most of the localities from which one can collect this fossil are in the Southern Hemisphere (Anderson and Anderson, 1983). The only exceptions to this general rule are those landmasses which have migrated far northwards since the Triassic (for example, the Indian subcontinent and the Arabian Peninsula). In Australia, good collecting sites are available around the old clay pits and coal mines at Ipswich, west of Brisbane, along the coastal cliff sections to the north of Sydney, and at the old Leigh Creek Coal Mine in South Australia (Fig. 10).

Fig. 10. Australian Triassic outcrops rich in Dicroidium and their associated fossils: (A)
Lower Triassic Terrigal Formation outcrop, north of Sydney – the arrow indicates a series of
fossiliferous horizons; and (B) Leigh Creek open-cut coal mine (now closed), South Australia,
exposing numerous thick, Upper Triassic coal seams.

In New Zealand, Dicroidium has been collected in Triassic exposures at Tank Gully, Long Gully and Benmore Dam in southern Canterbury and northern Otago. In South Africa, numerous collecting sites have been documented from exposures of the Molteno Formation in the Karoo Basin. In Argentina, well-studied Dicroidium assemblages have been collected from outcrops of the Ischigualasto Formation north of Mendoza. In India, Dicroidium is most commonly found in the Panchet, Parsora and Tiki formations of the South Rewa and Damodar basins in the northeast of the country. Some logistically challenging options are the excellent, windswept rock outcrops of Victoria Land in East Antarctica, from which several species of Dicroidium have been recovered.

So what happened to these plants?

Near the height of Dicroidium diversity at the end of the Triassic, the Earth faced another of its ‘big five’ mass extinctions – this event being commonly linked to volcanism in the Central Atlantic Magmatic Province. Millions of cubic kilometres of magma were extruded as ‘flood basalts’ or intruded into the crust as dikes and sills, the evidence for which is spread across much of modern-day western Africa, and the eastern regions of the Americas. The prevalence of Dicroidium in the vegetation of the Southern Hemisphere came to an abrupt end and the genus was thought to have been one of the major casualties of the end-Triassic crisis (Fig. 11).

Fig. 11. Triassic timeline showing the relative dominance of some of the main terrestrial groups and key extinction events.

However, recent findings suggest that Dicroidium may have lingered on in small populations into the earliest part of the Jurassic in Antarctica (Bomfleur et al., 2018). There is also a recent suggestion that this group of plants persisted into the Cretaceous in Mongolia, where reproductive structures similar to Umkomasia have been discovered. However, not all researchers agree with the identifications of these late Mesozoic fossils (Anderson et al., 2019) and it seems more likely that Dicroidium met its demise at, or just after, the end-Triassic crisis.

Thus, at the end of the Permian, a major mass extinction left an ecological vacuum for Dicroidium to fill, only for this group of plants to be vanquished by the next major global crisis. However, during the intervening 50 million years, Dicroidium reigned supreme as a true emblem of the Gondwanan Triassic.

A final consideration is to ponder just what caused the demise of this once hugely successful plant group. Flood volcanism in the Atlantic region, and associated doubling of greenhouse gas concentrations in the atmosphere, is commonly considered to have caused abrupt global warming at the end of the Triassic. Since plants are particularly responsive to climate, perhaps the polar environments of Antarctica provided the last refuge for these moisture-loving cool temperate plants in the earliest Jurassic.

In the modern world, atmospheric CO2 concentrations have maintained an accelerating upward trajectory from 280 parts per million (by volume; ppmv) during the pre-industrial era (ending about 1750) to 315 ppmv in 1960. This year, CO2 concentrations have reached 415 ppmv, the highest for millions of years. It is, therefore, pertinent to remember the lessons that can be learned from the fossil record about the ‘points of no return’ for major biomes under a changing climate.

About the authors

Dr Chris Mays (Research Fellow) and Professor Stephen McLoughlin (Senior Curator) are researchers at the Swedish Museum of Natural History. Both originally from Australia, they are presently investigating the patterns of vegetation turnover and recovery associated with the end-Permian and end-Triassic mass extinction events in palaeopolar regions. Further details of their research can be found at: and


Anderson, J.M. & Anderson, H.M. 1983. Palaeoflora of southern Africa: Molteno Formation (Triassic), Vol. 1: Part 1, Introduction, Part 2, Dicroidium. Balkema, Rotterdam. 227 pp.

Anderson, J.M. & Anderson, H.M. 2003. Heyday of the gymnosperms: systematics and biodiversity of the Late Triassic Molteno fructifications. Strelitzia 15. National Botanical Institute, Pretoria, 398 pp.

Anderson, H.M., Barbacka, M.K., Bamford, M.K., Holmes, W.B.K., Anderson, J.M. 2019. Umkomasia (megasporophyll): part 1 of a reassessment of Gondwana Triassic plant genera and a reclassification of some previously attributed. Alcheringa 43, 43–70.

Blomenkemper, P., Kerp, H., Hamad, A.A., DiMichele, W.A., Bomfleur, B. 2018. A hidden cradle of plant evolution in Permian tropical lowlands. Science 362, 1414–1416.

Bomfleur, B., Blomenkemper, P., Kerp, H. & McLoughlin, S. 2018. Polar regions of the Mesozoic–Paleogene greenhouse world as refugia for relict plant groups. In: Krings, M., Harper, C.J., Cúneo, N.R., Rothwell, G.W., eds, Transformative Paleobotany: Papers to Commemorate the Life and Legacy of Thomas N. Taylor. Elsevier, Amsterdam, 593–611.

Fielding, C.R., Frank, T.D., McLoughlin, S., Vajda, V., Mays, C., Tevyaw, A.P., Winguth, A., Winguth, C., Nicoll, R.S., Bocking, M., Crowley, J.L. 2019. Age and pattern of the southern high-latitude continental end-Permian extinction constrained by multiproxy analysis. Nature Communications 10:385.

Kustatscher, E., Ash, S.R., Karasev, E., Pott, C., Vajda, V., Yu, J. & McLoughlin, S. 2018. Flora of the Late Triassic. In: Tanner, L.H., ed., The Late Triassic World. Earth in a Time of Transition. Topics in Geobiology 46, Springer, Cham, Switzerland, 545–622.

Labandeira, C.C., Anderson, J.M. & Anderson, H.M. 2018. Expansion of arthropod herbivory in Late Triassic South Africa: The Molteno Biota, Aasvoëlberg 411 Site and developmental biology of a gall. In: Tanner, L.H., ed., The Late Triassic World. Earth in a Time of Transition. Topics in Geobiology 46, Springer, Cham, Switzerland, 623–719.

Retallack, G.J. & Dilcher, D.L. 1988. Reconstructions of selected seed ferns. Annals of the Missouri Botanical Garden 75, 1010–1057.

Leave a Reply