Australasian Science: Australia's authority on science since 1938

Ancient Rainforests or Burning Bush?

Heathland in the Stirling Ranges National Park

Heathland in the Stirling Ranges National Park in south-western Australia: is this more typical of Australia’s Cretaceous vegetation than rainforest?

By Ray Carpenter

New fossil evidence is forcing a rethink of whether rainforest or fire-prone shrubland prevailed in Australia during the age of the dinosaurs.

Picture Australia during the Late Cretaceous about 70 million years ago, and you may imagine yourself dodging a dinosaur in a lush Daintree-like jungle. Perhaps you know the story – Gondwana was covered in rainforest, and the heathlands and eucalypt woodlands came much later when the climate started to dry out and fires became more frequent.

However, recent plant fossil discoveries are challenging the “primeval rainforest” paradigm, revealing the immense anti­quity of heathland vegetation. You can now explore heathlands in Sydney’s Royal National Park or Blue Mountains and on a spectacular wildflower tour in the global biodiversity hotspot of south-western Australia.

As a palaeobotanist, my job is to study plant fossils to find out how our continent came to be clothed in such an interesting and beautiful flora. The fossil record can never be complete, and there are large gaps with few or no useful fossils.

One such Australian "black hole" for mammals and the leaves of flowering plants is from about 100–60 million years ago. This scarcity of fossils is a great shame, because it was during this interval that Australia began its isolation from Antarctica and the rest of Gondwana, and much of the distinctive identity of our biota must have emerged.

You can’t explore road or river cuttings for leaf fossils in the gap of 40 million years, because potentially useful surface exposures just don’t exist. Luckily, however, fossils buried deep underground can be brought to the surface during the search for minerals or water.

Enter Mike Macphail, one of Australia’s few palyn­ologists – an expert in the identification and dating of microfossils (including pollen grains and spores, and tiny aquatic animals). Mike’s work is valuable for resource explorers because he can help them pinpoint the right places to find what they’re looking for. His work is also great for palaeobotanists. Give him a microscope slide covered in thousands of microfossils extracted from a mining core, and he’ll quickly paint you a picture of a past environment – perhaps a placid lake surrounded by rainforest with lots of ferns.

So, I went looking through some mining core samples from near the arid centre of Australia, which Mike had dated at about 70 million years, and found the first useful bits of leaves from the “black hole”. These were not just any old leaves – some carried the tell-tale features of the Proteaceae, a plant family that now includes banksias, grevilleas, macadamias, proteas, waratahs and the like. It’s a fascinating group to study – species are found in most habitat types in Australia, and they have a lot say about how the Southern Hemisphere and its biota were put together.

The important leaf features I’m talking about are related to the amazing qualities of plant cuticle. This is a waxy layer that has evolved to protect the leaf from everything the environment can throw at it, including ultraviolet radiation, drying winds and insect attack. The palaeobotanical implications of this resistant layer are threefold.

  • First, cuticles are tough enough to withstand whatever chemical and physical forces they were subjected to for millions of years underground.
  • Second, the cuticle carries a fingerprint of its underlying cells, as well as hairs and glands, rather like what a skin peel taken from the back of your hand would show. These cells can have characteristic arrangements that make it possible to identify the plant from which the cuticle came.
  • Third, certain features, such as the number of stomata (the cells that regulate gas exchange for photosynthesis), can provide us with clues about past climates.

I found more and more Proteaceae cuticle types in the core samples – at least a dozen at last count, and Mike also found at least 35 different Proteaceae pollen types.

We weren’t the first to uncover the rich distant past of this family. Mary Dettmann, a pioneer of Proteaceae pollen research, described the family as the most successful flowering plant “invader” of the Mesozoic conifer, cycad and fern-dominated vegetation that prevailed across Gondwana during much of the dinosaur era. This Proteaceae “invasion” seemed to be first centred on the rift zone between Australia and Antarctica, but it spread pretty rapidly across Australia, and some Proteaceae even reached South America and South Africa by the end of the Cretaceous.

The Australo-Antarctic plants were envisaged as belonging to relatively open woodlands and scrubs on nutrient-poor or waterlogged soils. However, these ideas were not verifiable without any leaf fossils, and other scientists have had doubts related to the limited amount of identifying information that pollen grains carry. After all, almost all species of Proteaceae have a basic three-pored, triangular-looking grain, and you need to be an expert to recognise the more intricate details.

So, back to the leafy bits in the core samples. With serendipity and lots of hard work, we’d just published a paper showing that all species of Proteaceae that have stomata on both sides of their leaves grow in open habitats, such as sandplain heathlands, woodlands and sub-alpine thickets. That’s about 700 species, most of which are small- and tough-leaved shrubs. It was easy to tell that some of the new fossils had the same type of stomatal distribution because they were preserved as envelopes of cuticle, with both sides of the leaf intact. The fact that these leaves were also tiny – only about 1 mm wide – was pretty good supporting evidence that they’d come from open habitat plants. They were certainly a long way from the typically large, roundish leaves that rainforest plants have.

But wait, there’s more! We found lots of fossilised bits of charcoal in the core, indicating that fire was part of the Cretaceous landscape. We were onto something here – there was a neat fit with what evolutionary biologists such as Byron Lamont had said – some groups of Proteaceae that are now found in burnt, open vegetation were around a seriously long time ago.

In fact, the most recent common ancestor of one of the two largest subfamilies, Proteoideae, could well have had adaptations to survive fire in the Late Cretaceous. In south-western Australia the kwongkan (a Noongar Aboriginal word for sandy country with low scrubby vegetation) is now the best place on Earth to see the richest diversity of this subfamily. Three genera of Proteoideae are endemic to the region, while the smokebushes, woollybushes, coneflowers and drumsticks are most diverse there. Just like the best fossils we had, most species of these plants have tiny leaves, and almost all have stomata on both sides of their leaves.

“Found it!” yelled an excited Greg Jordan when he stumbled across a plant of one of our favourite Cretaceous candidates near Albany, Western Australia – a seemingly delicate shrub with curiously lanolin-scented flowers that belongs to Franklandia, one of the endemic genera of Proteoideae. Pollen grains just like those of Franklandia were well-represented on Mike’s microscope slides.

Several other truly fascinating kwongkan plants from different families have similar ancestries. A classic is the Albany pitcher plant (Cephalotus follicularis). This extraordinary plant is the sole member of a line that also dates to the Cretaceous. In this case there are no fossils – the evidence comes from its DNA. This plant and the better-known tropical pitcher plants (genus Nepenthes) showcase a textbook example of convergent evolution, where not even closely related organisms have independently evolved a way to solve a problem. In this case, the pitchers are extremely specialised insect traps.

Why did the Proteaceae take off in such a big way in the Cretaceous? We reckon that early on they hit on a winning formula to deal with the lack of nutrients in the Australian landscape, and that part of this formula was the ability to cope with, and perhaps benefit from, fire and its effects.

Most gardeners would be familiar with the advice that you should never apply phosphate-rich fertilisers to Proteaceae plants. Too much of a good thing will kill them because these plants have evolved superfine, high-surface area rootlets that are very adept at extracting scarce nutrients from the poor soils in which they usually live.

In terms of economics, there are a few fundamental, interlinked traits that apply when nutrients are scarce. Leafy assets are made small, tough and unpalatable because they’re expensive to produce. But these plants grow in open areas that are vulnerable to fire – favouring more evolutionary investment in traits such as woody fruits that don’t open until exposed to fiery heat, and underground lignotubers that can resprout.

Botanists have long been intrigued by the extreme richness of the flora of south-western Australia, and one of the most knowledgeable is Steve Hopper. He’s championed the concept of “old, climatically buffered, infertile landscapes” (OCBILs), and he’s proposed that a reason why kwongkan is so biologically diverse is because it hasn’t been inundated by the sea, or been covered in fertile soil from a volcano, or been blanketed by ice. The climate must have changed over time, but even in relatively severe times there’s always been enough rainfall to sustain at least the hardier lineages.

New discoveries will keep adding to the story, but what we found when we ventured into the fossil “black hole” of the Late Cretaceous was not jungle but burnt heathlands. We think that the current fossil evidence shows that rainforest like the Wet Tropics only became prominent in Australia during the global “greenhouse” period in the early Tertiary. Across that warm and wet time the older Proteaceae-rich heathlands persisted in areas that were too infertile to support rainforest, and one of these areas is likely to have been in south-western Australia.

Of course, things were still different back in the Cretaceous – we shouldn’t make the mistake of thinking that today’s kwongkan landscape is identical. For instance, our fossils show that there were exotic palms and hoop pines around, as well as several extinct gymnosperms that probably didn’t make it past the end-Cretaceous extinction event. Perhaps most intriguingly there were no eucalypts, which are now the most obvious plants in burnt open vegetation in Australia.

So, by all means, marvel at the wonders of Australian rainforests, but don’t forget the little battlers of the heathy scrublands – they just might have been around even longer. They certainly deserve our best conservation efforts. And, if you can, take a deeper look beyond the spring colours on your next wildflower tour. You might even spot the ghost of a hungry dinosaur.


Ray Carpenter is a Research Fellow at the University of Tasmania who works with Greg Jordan, Bob Hill (University of Adelaide) and Mike Macphail on an ARC-funded project exploring the origins of open vegetation in Australia.