Plant roots evolved at least twice, and step by step

Originally published at Lost Worlds Revisited for the Guardian Science Blog Network on 23 August 2018:

The discovery of a unique rooting anatomy from 407m years ago supports theory roots evolved at least twice, and step by step

 Exceptional cellular preservation of a meristem. Photograph: Sandy Hetherington

Most of us do not spend much time contemplating plant roots. Not only do they suffer from the wider issue of plant blindness, but they are also the bit of the plant that is not visible. In terms of getting people excited about plant science, it’s a tough gig. This is a shame, because plant roots are critical to all of our lives: no roots means no food. Roots provide anchorage, and allow plants to gain water and nutrients from the soil. They also form a key symbiotic relationship with mycorrhizal fungi, which provide minerals from the soil in return for a steady supply of carbs from the plant.

In the modern world, we can easily divide plants into the ones that have roots and the ones that don’t. The flowering plants (angiosperms) and the other vascular plant groups (conifers and other seed plants, ferns, horsetails and clubmosses) all have a recognisable root, defined by having a meristem of rapidly-dividing, undifferentiated cells, and by having a root cap, which protects the apex of the growing root and which is where the plant perceives gravity.

The other, non-vascular, land plants – mosses and liverworts – lack the plumbing to move water and food around their bodies, and also lack true roots. These plants do have rhizoids, which are simple, hair-like structures, analogous to the root hairs on the surface of a flowering plant root. Rhizoids are used for water transport, and some may be involved in nutrient transport, but they aren’t fully-functioning roots.

A new study by Alexander Hetherington and Liam Dolan, published in the Journal Nature, describes root meristems from the Rhynie Chert, belonging to a plant called Asteroxylon mackiei. Asteroxylon is a lycopsid, the group of land plants which includes small modern clubmosses such as Selaginella, and the extinct giant Lepidodendron forest trees of the Coal Measures. What Asteroxylon does not have is a root cap. The five apices of root-like structures that the researchers found (from examining 641 thin section slides from museum and university collections across the UK) are covered by a continuous layer of epidermis over the meristem and no tapering root cap, seen in all modern plants with roots, is present.

Given the exceptional preservation of everything else, the authors ruled out a preservational reason for its absence. They also showed that the cellular organisation within three of the apices indicated that they actively growing when they were preserved, so the root caps were not lost because growth had stopped. The orientations of cell walls, indicating cell division patterns within the meristem, also ruled out the formation of a root cap. Root hairs, present in modern true roots, are also completely absent in Asteroxylon.

Plant root.
A thin section slide of a plant root. Photograph: Sandy Hetherington

All of this evidence adds up to a startling conclusion: since modern clubmosses do have true roots with root caps, true roots have evolved (at least) twice: once, in a step-wise fashion, in the lycopsid branch, with Asteroxylon (with a meristem, but without root cap or root hairs) representing an intermediate stage, and independently in all the rest of the vascular land plants. The similarity in root anatomy between clubmosses and all other vascular plants is a case of convergent evolution.

One of the very first blog posts I wrote for Lost Worlds Revisited back in 2016 was also about fossil root meristems, and at the time I remember feeling really lucky to have found a platform to share palaeobotanical discoveries with a far wider audience than they would ever normally get. Sadly, with the imminent closure of the Guardian’s Science Blog Network, this is my last blog post. It’s been a wonderful experience, working with the finest team of palaeo bloggers (thanks Dave, Elsa, Hanneke and Mark!), and I will miss revisiting all of our lost worlds terribly.

The real palaeo diet: the nutritional value of dinosaur food

Originally published at Lost Worlds Revisited for the Guardian Science Blog Network on 18 July 2018:

Experiments on modern plants show that the nutrients which dinosaurs could get from plants varied with carbon dioxide levels

Our fascination with giant sauropod dinosaurs such as Diplodocus, Brachiosaurus and Brontosaurus stems from their colossal size. How could something 30 metres long, weighing 50 tonnes, function as a land animal? And how could something that big gain enough nutrition from plants?

We have little evidence for the diet of everyone’s favourite giant herbivores. Reports of fossilised stomach and gut contents have been contested, and coprolites (fossilised dung) are difficult to assign to their producer with any certainty. Indirect evidence from comparative morphology with giraffes and elephants, the largest modern analogues, suggests that sauropods would browse from the tree canopy, although researchers disagree about whether all sauropods held their heads high. Some have even suggested niche partitioning between the different sauropod groups, with the tallest brachiosaurs feeding from the top of the canopy, camerasaurs in the mid-canopy, feeding on seed-ferns and cycads, and diplodocids grazing on ferns and horsetails at ground level.


Indirect evidence from sauropod teeth is perhaps the strongest: many sauropods have well-worn teeth from abrasion by food. Wear patterns on the peg-like teeth of diplodicoids (the group including Diplodocus) have even been analysed to suggest a ‘grab and pull’ mode of feeding, known as unilateral branch stripping. Some titanosaurs (a globally-distributed group of sauropods which survived right up to the end-Cretaceous extinction event) may have had sharp beaks, perfect for cropping coarse vegetation.

It had originally been theorised that higher levels of carbon dioxide during the Mesozoic, increasing net primary productivity, had driven the rise of the sauropods. However, carbon dioxide levels alone do not dictate primary productivity, and all the other things that a plant needs – light, water and nutrients – would have been limiting factors. Even if primary productivity increased due to elevated carbon dioxide levels, it was thought that the low energy value of non-angiosperms would mean that sauropods would have to consume colossal amounts of plant material. Importantly, carbon dioxide levels also have indirect effects on plant digestibility and nitrogen content.

new study by Fiona Gill at the University of Leeds, and colleagues at Goettingen and Nottingham, investigated the energy nutritionally available to sauropods from plants grown in Mesozoic levels of atmospheric carbon dioxide, simulated in growth chambers. For canopy plants they used Metasequoia (the Dawn Redwood, a deciduous conifer), Araucaria (the Monkey puzzle, a scale-leaved conifer) and Ginkgo (everyone’s favourite living fossil seed plant). For understorey plants, species of the fern Polypodium and horsetail Equisetum were chosen, along with a buttercup, Ranunculus acris, as an early angiosperm analogue.

The plants were then fermented with cattle rumen fluid for 72 hours to mimic transit through a sauropod gut, and the gases produced analysed to give a measure for digestibility (a technique used in evaluation of agricultural feedstocks). The amount of metabolizable energy from each sample was calculated, and levels of carbon and nitrogen measured.

The study showed that the energy available to sauropods from the fern and Metasequoia were comparable to angiosperms, and that horsetails outperformed angiosperms in terms of their value as a food. This may be related to the low levels of the polymer lignin in horsetails, which use silica as a structural element instead.

Based on these results, Gill and colleagues calculated that a 30 tonne sauropod with a metabolic rate intermediate between modern reptiles and modern mammals would need to eat 110kg per day of Monkey Puzzle foliage grown at 2000 parts per million (ppm) of carbon dioxide (Mid-Triassic levels, five times higher than modern levels), whereas the same animal could get by on only 51kg per day of horsetails grown at only 1200 ppm of carbon dioxide (three times modern levels).

These findings challenge assumptions about both the nutritional quality of non-angiosperm plants such as horsetails, and any generalisations about plant food value at elevated carbon dioxide levels. Different plant species behave differently at higher carbon dioxide levels, and their nutritional values alter in surprising ways – something that our own species could do with considering.

Speculative biology: understanding the past and predicting our future

Originally published at Lost Worlds Revisited for the Guardian Science Blog Network on 30 May 2018:

A new edition of After Man by Dougal Dixon, a landmark piece of speculative biology which influenced a generation of palaeontologists, has been released

The Flooer as imagined by Dougal Dixon in his book After Man: A Zoology of the Future
 The Flooer as imagined by Dougal Dixon in his book After Man: A Zoology of the Future. Photograph: Breakdown Press

In 1981, a remarkable book was published: After Man: A Zoology of the Future, by Dougal Dixon. As a child of the eighties, growing up in a science fiction bubble where daleks, vogons and the fighting machines of the War of the Worlds were at least as concrete to me as anything happening in the real world, After Man presented a biologically-themed alternative world to lose myself in.

The premise of the book is simple: take the Earth today, remove the humans, and let evolution take its course for 50 million years. What new animals evolve? Of course, in other hands this approach could have resulted in a throwaway romp. In Dixon’s, it produced an incredibly detailed, thoughtful book, in which the principles of evolutionary theory and ecology are rigorously applied. Crypsis (adaptations to avoid being seen by either predators or prey) is a common theme, as is mimicry. And convergent evolution (the idea that unrelated organisms in similar ecological niches evolve similar adaptations) is everywhere. Each species has a scientific name which follows the conventions that taxonomists use, and the text describes their behaviours and inter-species interactions. The striking illustrations, with copious annotations, resemble a naturalist’s field notes.

The cover of After Man: A Zoology of the Future by Dougal Dixon.
 The cover of After Man: A Zoology of the Future Photograph: Breakdown Press

As well as being rigorous, After Man manages to be fun. On newly-formed volcanic islands in the Pacific, bats arrived before birds did, and radiated to occupy new niches. The Flooer, Florifacies mirabila, is one of these new species. The shape and colouring of its head mimics the appearance of a local flower, and its sweet-smelling saliva mimics nectar. It sits on the ground with its face upwards and its mouth open. It is, unsurprisingly, an insectivore.

Dougal Dixon, who studied geology at the University of St Andrews, followed up After Man with two similar speculative biology books. In 1988, The New Dinosaurs: An Alternative Evolution was published, and explored a world where the end-Cretaceous mass extinction hadn’t wiped out the non-avian dinosaurs. Man after Man: An Anthropology of the Future came out in 1990, and speculated on the evolution of our own over the next five million years. There’s a whole generation of evolutionary biologists and palaeontologists who, like me, were inspired by his work. Happily, a new edition of After Man, featuring an updated introduction from Dougal Dixon and a new cover, has just has been published by Breakdown Press, so a new generation can revel in this landmark piece of speculative biology.

The Reedstilt as imagined by Dougal Dixon.
The Reedstilt as imagined by Dougal Dixon. Photograph: Breakdown Press

There aren’t many areas of science where speculation is not only tolerated, but can be argued to be essential. Palaeontologists must take an incomplete body of evidence from fossils, and, like Dixon, apply their knowledge of how the natural world works, in order to understand the living organism as it once was. How far that speculation is allowed to expand beyond the existing physical evidence is a judgement call that the researcher, and often the palaeoartists they work closely with, must make. This theme is explored in All Yesterdays, by John Conway, CM Kosemen and Darren Naish, published in 2012. While the palaeontological reconstructions are fascinating, the most enlightening section of this book for me is the reconstructions of modern animals from their skeletons alone, some of which bear no resemblance to their real counterparts. I’ll never look at a baboon in quite the same way again.

Comparative morphology and trace fossils, such as trackways, can tell us a certain amount about behaviour, but without a time machine, interpreting the behaviour of extinct animals is speculative. With soft part preservation rare or absent for most fossils, careful, conservative speculation is essential in reconstructing the appearance of organisms. That’s not to say that bolder speculation isn’t great fun: the reconstruction of elasmosaurs engaging in a neck-swinging competition for mates in All Yesterdays is wholly speculative, but certainly makes you consider all the facets of animal behaviour in the past that we will never be able to know about. And where there are differences between researchers in the degree of speculation, coupled with interpretation of the physical evidence, it can result in ‘interesting’ debates: witness the recent scuffles over a revived theory for aquatic dinosaurs (the #FordvNaish hashtag provides a taster).

Speculative biology has been around ever since HG Wells tackled human evolution in The Time Machine in 1895, but its power to spark our imaginations, as well as inform our understanding of organisms through deep time, is undiminished. Careful, informed speculation is crucial to palaeobiology, and long may it continue.

Spore heroes: unlocking the life-cycle secrets of the earliest land plants

Originally published at Lost Worlds Revisited for the Guardian Science Blog Network on 25 May 2018:

Fossils from 432m years ago push back the origin of the alternating life cycle still seen in ferns today

Our world today is dominated by the flowering plants, or angiosperms, which appeared approximately 130m years ago and rapidly diversified to become the top dogs in most ecosystems. But there are plenty of other plants from more ancient lineages still around, doing deeply weird things in their life cycles, and doing them for much longer than we have realised.

Flowering plants themselves are a refinement of a much earlier innovation about 375m years ago: the seed. Seeds provide protection and a source of nutrition for the embryo, and a handy means of dispersal. A seed germinates and then grows into another plant which is more or less the same as its parents. This is not how spore-producing plants like ferns do things.

The plant that you think of as a fern – a tallish, green structure with well-defined leaves and roots – is just one half of its life cycle. This plant is known as the sporophyte and, like us, it has two copies of its chromosomes (it is diploid). The sporophyte produces spores, which contain only a single copy of the chromosomes. These spores disperse, and grow into a separate plant, called the gametophyte (which is haploid, like the spores). This gametophyte plant looks completely different from the sporophyte phase: it is a small, flat, amorphous blob. It is the gametophyte that produces eggs and sperm, and as long as there is plenty of water about to allow the motile sperm to swim to the egg, fertilisation occurs. A new, diploid sporophyte grows from this fertilised egg, and the cycle begins again … This process is known as alternation of generations, and has, ironically, confused generations of biology undergraduates. Nobody expects to see motile sperm in a plant lecture.

In angiosperms and other seed plants, that gametophyte phase is still there, but is neatly bundled away inside organs on the sporophyte, and the sporophyte phase is the only bit you ever see externally. In ferns, the sporophyte phase is dominant, but the gametophyte is still hanging in there as an independent plant. If you look at modern plants which branched off even lower down in the plant evolutionary tree, such as mosses and liverworts, the situation is reversed. In liverworts like Marchantia, which lives in most soggy places worldwide, the gametophyte is the dominant phase – the thing you would think of as the plant. The sporophyte phase is completely dependent on its gametophyte for photosynthesis, and does not live as a separate plant. Does this mean that the sporophyte phase of the earliest land plants was equally dependent on its gametophyte?

previous study had suggested a minimum size limit below which cooksonioid plants could not have functioned as photosynthetically independent plants. Cooksonia barrandei, with a stem diameter of more than a millimetre, is above this hypothesized size limit. As well as being the oldest known sporophyte, it may have also been fully independent of its gametophyte.

A reconstruction of Cooksonia, one of the earliest known land plants.
A reconstruction of Cooksonia, one of the earliest known land plants. Photograph: Matteo De Stefano/Muse via Wikimedia Commons

By around 30m years later, the plants preserved in the Rhynie Chert had unambiguous independent sporophyte and gametophyte forms, and alternation of generations was well established, without an obvious emphasis on either the gametophyte or the sporophyte. Cooksonia barrandei pushes the origins of this alternation of generations so much further back into the past that the commonly held assumption – that a gametophyte-dominant life cycle, as seen in modern liverworts, is ancestral – is being questioned. This new discovery allows the possibility that alternation of generations was established very early in, or even before, the evolution of earliest land plants. Approaches which combine evidence from the fossil record with developmental genetics might begin to explain just how this key switch, which led to the plants we rely on, was thrown so long ago.

Not one, but three Jurassic worlds, in new UK museum exhibition

Originally published at Lost Worlds Revisited for the Guardian Science Blog Network on 19 April 2018:

Yorkshire’s Jurassic World, a new exhibition in York, includes a pregnant ichthyosaur, a Mesozoic virtual reality experience and a dinosaur called Alan

Sir David Attenborough feeds a sauropod in the virtual reality exhibit at the Yorkshire’s Jurassic World exhibition in York
 Sir David Attenborough feeds a sauropod in the virtual reality exhibit at the Yorkshire’s Jurassic World exhibition in York. Photograph: Anthony Chappel-Ross/Yorkshire Museum

If you say the word Jurassic to people in the UK, the chances are that their first thoughts will be of a certain hugely successful film franchise. Most palaeontologists are fine with this, because it gives us an excuse to wheel out our well-honed “all the things that were wrong about the Jurassic Park film” material. If they mention anything else at all, it is likely to be the Jurassic Coast, a fantastic piece of tourism branding which ensures that Dorset seaside towns receive a steady stream of fossil-mad families on holiday every summer.

Other British Jurassic outcrops are available, however. In Scotland there are Jurassic outcrops up in the Highlands, and sites on the Inner Hebrides are yielding exciting new discoveries. The north coast of Somerset also has some productive areas, and just like the Jurassic coast rocks found at Lyme Regis, these are part of a swathe of Jurassic rocks running diagonally across the UK from the South coast all the way up to Yorkshire, where they are again seen in all their glory on the north-east coast. A new exhibition at the Yorkshire Museum in York, which was opened in March by David Attenborough, celebrates the fossils of the Yorkshire Jurassic.

A reconstruction of Yorkshire in the middle Jurassic era
A reconstruction of Yorkshire in the middle Jurassic era. Photograph: Yorkshire Museum

It is sobering to remember that the Jurassic period lasted more than 50 million years, an immense span of time over which geological processes could change the landscape and environment entirely. The exhibition at Yorkshire Museum presents the three distinct phases of Jurassic life preserved in Yorkshire’s rocks: early Jurassic oceans, life on land during a period of local uplift in the middle Jurassic, and the shallow coral seas of the late Jurassic.

A virtual theropod strides along a well-preserved trackway in the middle Jurassic gallery. Yorkshire has much evidence of dinosaurs in the middle Jurassic, but mostly in terms of footprints and trackways, with relatively few body fossils. One such rare body fossil on display is “Alan”, a sauropod known solely from a single vertebra, representing the earliest British sauropod. Since it is not possible to give the fossil a scientific name based solely on a single vertebrae, the specimen is informally named after the collector who found it, Alan Gurr.

A fossil of a single vertebra from Alan the sauropod.
A fossil of a single vertebra from Alan the sauropod. Photograph: Anthony Chappel-Ross/Yorkshire Museum

The mid-Jurassic rocks of Yorkshire are also famous in the world of palaeobotany for their beautifully preserved, classic Mesozoic plant remains, many of which are displayed. These show us an ecosystem dominated by seed plants such as conifers, ginkgoes, cycads and bennettites (an extinct group of plants superficially resembling modern cycads), accompanied by spore-producing ferns and horsetails. This is a world before the flowering plant revolution, but there are hints of things to come. The seed plant Caytonia has been linked to the flowering plants in some studies, but as an extinct group we have no molecular evidence to help us decide where it belongs in the plant evolutionary tree.

The highlight of the exhibition for me was a visit to the mid-Jurassic, with the help of a virtual reality helmet and wand. I suspect that most visitors will not be quite as excited as I was at picking branches of Caytonia, but will focus a bit more on the next stage: feeding the plant to a docile sauropod, who then hangs around for a meal.

The late Jurassic coral seas are in a gallery with a light, colourful reef display, with fossil ammonites, bivalves and other invertebrates displayed in large groupings, rather than as clinical, single specimens. You could almost be paddling over a Jurassic reef.

For anyone interested in a slice (or three) of life in the Mesozoic, this exhibition is well worth a visit. The exhibition is expected to be on display for at least the next two years, so plenty of time to plan your trip to the Jurassic.

How the earliest plants made our world muddy

Originally published at Lost Worlds Revisited for the Guardian Science Blog Network on 23 March 2018:

The first plants to make it on to land altered mud production and where it formed rocks, changing our planet forever

How and when the earliest plants made the first move on to land is always a hot topic for palaeobotanists. We know that early land plants likely evolved from freshwater algae, gaining a bunch of necessary adaptations in the process. Plants needed to support themselves, protect themselves from drying out and from the harmful effects of UV light, and gain water and nutrients from a finite supply on land. A study published last week by Mariusz Salamon and colleagues described fossils that push back the earliest evidence of land plants to around 445 million years ago.

The new fossils come from mudstones in central Poland, in beds that have been dated using other, much more common and cosmopolitan, fossils. The plant remains are tiny, branched fragments, up to about 3mm long. Some specimens appear to have spore-cases at the top of their branches, similar to those seen in younger, better-known early land plants such as Cooksonia. The preservation of the plants means details are hard to discern, but Salamon and colleagues present a single, tantalising stoma, or air pore, on one of the fragments as a key piece of evidence. If this plant had stomata for gas exchange, it was likely to have been living in land, a good 15 million years earlier than previously known plant fragments.

A reconstruction of Cooksonia, one of the earliest known land plants.A reconstruction of Cooksonia, one of the earliest known land plants. Photograph: Matteo De Stefano/MUSE via Wikimedia Commons

Of course, you can come at the problem from the other direction. What global changes would you see in the rock record that could be explained by the activities of plants? This is the same kind of approach taken by geologists looking for evidence of life on other planets. What changes would you see preserved in rocks that cannot be explained away by inorganic processes, and by land plants in particular?

William McMahon and Neil Davies took this approach in a recent study, looking at the relationship between the buildup of muddy sediments on land and colonisation by the earliest plants. It has long been assumed that until plants colonised the land, most fine-grained sediments eroding from the continents were washed by river systems into the sea, where they formed marine mudrock. Once land plants became established, changes to both how rocks weathered and how fine, muddy particles were trapped by plants meant that muddy sediments could accumulate on land in much greater quantities. This long-held assumption was crying out for some testing with hard data.

McMahon and Davies assembled data from more than a thousand published reports and made more than a hundred field investigations, looking at rocks formed by flowing water on land, known as alluvial formations. Their database comprised all 704 known alluvial formations, ranging from the Archean eon 3.5 billion years ago to the Carboniferous period, a mere 300 million years ago.

Their analysis of the proportion of mudrock (made of grains 0.063mm or smaller) in alluvial deposits agreed with long-held geological hunches. For the first 3 billion years of the rock record, mudrocks form an average 1% of sedimentary deposits on land. By around 300 million years ago, mudrocks had risen to an average of 26% of alluvial formations. They pinpoint an upsurge between the Late Ordovician period (458 Ma) and the Devonian period (359 Ma) – which agrees remarkably well with the fossil record for earliest colonisation by land plants.

No tectonic or other purely geological events seem to explain the upsurge in alluvial mudstone, but three mechanisms associated with the appearance of early land plants would have boosted muddy deposits on land. First, plants would promote the weathering of rocks to fine clay minerals, and as rooting systems evolved, and symbiotic relationships with microbes increased, the depth of the uppermost layer of rocky soil in which chemical weathering took place would increase. In these ways, plants were adding to the production of fine-grained sediments. Second, roots would have had a binding effect, helping to retain fine-grained sediments and preventing erosion. The root systems of the earliest land plants were limited, though, so this explanation cannot be the whole story for the Late Ordovician to Devonian muddy upsurge. The final mechanism that may well have been pivotal in creating a muddy planet is the baffling effect that even tiny early land plants would have produced above ground, trapping fine sediments between their stems, leaves and other organs. This effect can be seen today: in environments where liverworts and mosses are the only plants that can make a living by forming a low ground cover, fine-grained mud and silt are incorporated in their clumps.

Once they were established more than 400 million years ago, rates of mud production and retention were permanently altered and mud became a permanent fixture of the rock record on land, thanks to the work of tiny plants over deep time.

What fossils reveal about the spider family tree is far from horrifying

Originally published at Lost Worlds Revisited for the Guardian Science Blog Network on 14 February 2018:

Recent fossils in amber tell us how spiders evolved into their modern groups, but the fossil record for arachnids goes much deeper

An illustration of the Cretaceous arachnid Chimerarachne yingi found in Myanmar.
 An illustration of the Cretaceous arachnid Chimerarachne yingi found in Myanmar.
Illustration: Dinghua Yang/University of Kansas/Reuters

The discovery of a 100m-year-old spider ancestor with a whip-like tail, bearing a more than slight resemblance to everyone’s favourite parasitoid alien – the facehugger – gained a lot of media interest last week. Some arachnologists were upset by both the language of fear in the coverage (“creepy” and “horrifying” were popular descriptions) and by some folks expressing a desire to nuke it from orbit. It seems that despite (or perhaps because) of the intense responses that spiders evoke in people, there is always an interest in where and how they evolved.

The newly described species, Chimerarachne yingi, was based on two specimens found in amber of about 100 My old from Myanmar. Unusually, the new find was revealed in two simultaneously-published papers in the journal Nature Ecology & Evolution. The rules for the naming of species mean that only one of the papers, by Bo Wang and colleagues, gets to be the formal description and naming of the species (and new genus, the next level up in classifying organisms). Both Wang’s paper and that of Diying Huang and colleagues, aimed to place the new find in terms of the spider family tree. The new species has features of modern spiders, known as the Araneae: a male pedipalp (sensory appendage) modified for sperm transfer, and well-defined spinnerets for silk spinning. But, it also has its distinctive tail, a feature not found in modern spiders, but associated with an ancient grouping of “almost-spiders” known as the Uraraneida.

Further into the past, true spiders can be tracked all the way back to the Carboniferous Period, about 300 My ago. Paleothele montceauensis has the spinnerets required to place it in the modern Mesothelae group (which today line their burrows with silk), while another arachnid found in the same deposits in France, Idmonarachne braisieri, has true spider legs and jaws, but no spinnerets. It seems that different types of spider-like arachnids, with different combinations of features, have co-existed on Earth over deep time.

The trigonotarbid group includes heavily-armoured spiny forms, which would no doubt be the stuff of nightmares for the arachnophobic readers of media reporting on spider evolution. Whether we should fear these animals or not (short answer: most of the time, we really shouldn’t), spiders and their near relatives have been a part of the story of life on land since the very beginning, and they deserve our respect, if not our love, on Valentine’s Day.

Strangest things: fossils reveal how fungus shaped life on Earth

Originally published at Lost Worlds Revisited for the Guardian Science Blog Network on 16 January 2018:

Fossil fungi from over 400m years ago have altered our understanding of early life on land and climate change over deep time

Much of the weirdness depicted in the TV show Stranger Things is distinctly fungal. The massive organic underground network, the floating spores, and even the rotting pumpkin fields all capture the “otherness” of fungi: neither plants nor animals, often bizarre-looking, and associated with decay. As weird as they may seem to us, fungi are integral to the story of the evolution of our landscapes and climate.

Molecular studies show us that animals and fungi share a more recent common ancestor than either group does with plants, and that these groups had all diverged over a billion years ago. A sparse fossil record for fungi is not entirely surprising, given the low preservation potential of soft, microscopic threads, but we still have tantalising glimpses of their history. Recent work on the Rhynie Chert, a deposit formed in hydrothermal wetlands 407m years ago, preserving an early land ecosystem in exquisite detail, has helped to reveal the hidden history of fungi. All modern groups of fungi are abundant in Rhynie chert samples apart from the basidiomycota, the group which includes those most familiar of fungi: mushrooms. New findings have been published in a special volume of the Philosophical Transactions of the Royal Society B.

Prototaxites is an organism as odd as anything in Stranger Things. Bearing a superficial similarity to a fossil conifer trunk (and initially described and named as such in the 19th century), some specimens are up to a metre in diameter and more than eight metres long. It has long been debated, but most researchers have settled on a fungal affinity for Prototaxites. Until the appearance of the first forests around 375 Ma, the largest organism on land, towering over the landscape, was an enormous fungus.

The microstructure of Prototaxites. Tubes are around a twentieth of a millimetre in diameter.
The microstructure of Prototaxites. Tubes are around a twentieth of a millimetre in diameter. Photograph: k2727 via Wikimedia Commons

Rosmarie Honegger and colleagues studied permineralised fragments of Prototaxites from the Rhynie Chert, together with material preserved as fossilised charcoal from the Welsh Borderlands. The remarkable preservation of the outermost layer of these specimens allowed them to describe the reproductive structures of Prototaxites, enabling them to positively identify Prototaxites taiti as a basal member of the ascomycete group of fungi, which today includes brewer’s and baker’s yeasts, Penicillium, and thrush.

Fungi are experts in symbiosis, and arguably the most important fungi for humans are those living within the roots of the plants we eat. Mycorrhizal fungi are found in around 80% of modern plant species, and provide improved uptake of minerals from the soil in return for a steady supply of carbohydrates. Mycorrhizal relationships can be recognised in the exceptionally preserved early land plant remains in the Rhynie Chert, and seem to have been established as soon as plants moved onto land (fungi had probably been exploiting this new environment in another of their symbiotic relationships, as lichens, for much longer). Benjamin Mills, Sarah Batterman and Katie Field have taken the evidence of these symbiotic relationships, and combined it with experimental data, to model how they may have influenced climate change over deep time.

Arbuscular mycorrhiza seen under microscope. Flax root cortical cells containing paired arbuscules.
Arbuscular mycorrhiza seen under microscope. Flax root cortical cells containing paired arbuscules. Photograph: Msturmel via Wikimedia Commons /University of Manitoba, Plant Science Department

Carbon dioxide levels 400m years ago were much higher than today. For modern plants, high carbon dioxide levels can mean greater productivity, but for rootless early plants, growing on poor rocky soils, the acquisition of minerals, particularly phosphorus, would have limited their productivity. Mycorrhizal symbiosis would have helped plants to obtain phosphorus, enabling them to use more carbon dioxide to build more plant material, which would eventually be buried, leading to a drawdown of global carbon dioxide levels. Mycorrhizae would have also increased the weathering of silicate minerals in rocks, which increases the cycling of carbon from the atmosphere to ocean sediments. Over deep time, early plants and their fungal partners changed our atmosphere into a lower carbon dioxide, higher oxygen regime.

In their paper, Benjamin Mills and colleagues show that modern liverworts, the nearest living equivalents of the Rhynie Chert plants, behave differently in response to heightened carbon dioxide, depending on which particular mycorrhizal partnerships that they employ. By building this information on carbon and phosphorus cycling into existing climate models, they show that the efficiency of mycorrhizal uptake of phosphorus in the earliest ecosystems on land would have profound effects on global climate. More accurate climate modelling needs to take into account the nuances of our weird fungal neighbours, and more great work incorporating experimental and palaeontological approaches will help to bridge this gap.

Top fossil discoveries of 2017

An extract from an article originally published at Lost Worlds Revisited for the Guardian Science Blog Network on 28 December 2017:

First life on earth

Some of the smallest fossil finds of 2017 were among the most controversial. In March, Matthew Dodd and colleagues described tiny tubes and filaments composed of iron oxide in rocks from Quebec, Canada dated between 3.77bn and 4.28bn years old. They interpreted them as the remains of bacteria living around hydrothermal vents, pushing the earliest evidence of biological activity to more than 3.77bn years ago, and conceivably even a staggering 500m years earlier. In September, Takayuki Tashiro and colleagues analysed graphite particles from rocks 3.95bn years old from northern Labrador, Canada. They concluded from isotope ratios that the carbon was biologically produced, although this interpretation was not shared by all researchers.

An epoxy mount containing a sliver of a nearly 3.5 billion-year-old rock from the Apex chert deposit in Western Australia. Pictured at the Wisconsin Secondary Ion Mass Spectrometer Lab (WiscSIMS) at the University of Wisconsin-Madison.
An epoxy mount containing a sliver of a nearly 3.5 billion-year-old rock from the Apex chert deposit in Western Australia. Pictured at the Wisconsin Secondary Ion Mass Spectrometer Lab (WiscSIMS) at the University of Wisconsin-Madison. Photograph: Jeff Miller/William Graf, University of Wisconsin – Madison

Finally, in research published mid-December, Bill Schopf and colleagues used the carbon isotope composition of microfossils in the 3.46bn year old Apex Chert, from Western Australia, to confirm their previously-disputed biological origins and even work out which groups of microbes were represented. Two species were primitive bacterial photosynthesizers, one was a methane-producing Archaean microbe, and two others were bacterial methane consumers. This impressive study shows that methane-cycling microbial communities were already established by 3.5bn years ago.

There’s more than one way to build a tree, 374m-year-old fossils reveal

Originally published at Lost Worlds Revisited for the Guardian Science Blog Network on 15 November 2017:

Fossils from China show that evolution found an alternative – and ultimately overly-complicated – way to increase the size of the earliest tree trunks

A cross-section of the fossilised cladoxylopsid found in Xinjiang, China.
 A cross-section of the fossilised cladoxylopsid found in Xinjiang, China. Photograph: Chris Berry, Cardiff University

In the world of knee-high land plants 400m years ago, the battle to grow tall was won by plants which found biomechanical solutions to fight gravity. Vascular plants had already evolved a plumbing system, allowing them to transport water, and the food produced by photosynthesis, around the plant. The water-conducting cells in the xylem – dead, hollow and stiffened by the polymer lignin – also afforded them some structural support. But there are limits to the height that a plant can grow with a stem of fixed girth.

In modern trees, trunks grow outwards as well as upwards. Known as secondary thickening, a ring of dividing cells beneath the bark, called the vascular cambium, produces new xylem and phloem tissue. This is what wood is: secondary xylem, composed of dead lignified cells, now employed by trees as a building material to allow them to continue to grow tall.

Plants producing wood locked up carbon extracted from the atmosphere during photosynthesis and, when trees died, resulted in its burial in sediments. This storage over geological time as coal (which humans are so keen to dig up, burn and release the carbon from) changed how carbon cycled through our ecosystems. The first forests transformed our planet in other, less obvious, ways too. Tree root systems stabilised soil, changing the landscape and affecting how minerals in the sediments weather. These changes in weathering take carbon dioxide from the atmosphere, producing carbonic acid, which ends up in river systems, and ultimately puts the carbon in the ocean. The Earth’s carbon cycle, climate and the evolution of forests are inextricably linked.

A foreman uncovers a fossilised tree trunk, thought to be over 350 million years old, at a quarry in upstate New York in the 1920s.
A foreman uncovers a fossilised tree trunk, thought to be over 350 million years old, at a quarry in upstate New York in the 1920s. Photograph: Courtesy of Chris Berry, Cardiff University

Another group of early trees solved the structural problem of being a tree very differently. The gloriously-named cladoxylopsids first appeared around 390 Ma, and have been well-studied from sites in Germany, Scotland and USA. The fossil forest at Gilboa quarry in New York state, where tree stumps known as Eospermatopteris, preserved as sandstone casts, up to one metre in diameter, in life position, has been studied since the 1870s. These trees were reconstructed in 2007 as Wattiezaafter a fossil tree complete with a palm-like crown of leafy fronds was discovered nearby.

An artist’s impression of a fossil forest.
An artist’s impression of a fossil forest. Photograph: Peter Giesen

New discoveries in China from Hong-He Xu and colleagues, from Nanjing Institute of Geology and Palaeontology, Cardiff University and Binghamton University, have revealed the strange anatomy of the trunk of cladoxylopsid trees. Where the Gilboa Wattieza trees are preserved as sandstone casts with little detail, the new fossils dating from 374Ma, from Xinjiang, China, are silicified, preserving the cellular details of their wood. They show that rather than a simple ring producing secondary tissue, cladoxylopsids had many separate and distinct xylem strands around the outside of the trunk, each one producing its own thickening rings, almost like a mini tree. An intricate network of interconnecting xylem tissue joined up the strands throughout the trunk, which was otherwise hollow.

A cross-section of the fossilised cladoxylopsid found in Xinjiang, China.
A cross-section of the fossilised cladoxylopsid found in Xinjiang, China. Photograph: Chris Berry, Cardiff University

It is the “ordinary” cortical tissue between the xylem strands which appears to have driven girth increase in these trees, by having such a high rate of cell proliferation that it pushed the ever thicker mini-trees apart, ripping the connecting xylem tissues in the process. The tree was in a state of continual, controlled internal collapse, repairing its internal tears as it grew. This seems like an incredibly over-complicated way to be a tree. Some modern palm trees do increase their girth by primary growth but in a much less complex way. Perhaps the cost of this elaborate anatomy was a factor in the demise of the cladoxylopsids, which disappear from the fossil record soon after these Chinese finds. These findings are yet another demonstration how much we still do not know about the diversity of plants and their anatomy through deep time.


Xu, Hong-He, Berry, Christopher M., Stein, William E., Wang, Yi, Tang, Peng and Fu, Qiang 2017. Unique growth strategy in the Earth’s first trees revealed in silicified fossil trunks from China. Proceedings of the National Academy of Sciences of the United States of America 114 (45) , pp. 12009-12014.10.1073/pnas.1708241114