Introduction:

Development, the process by which a single egg cell transforms into a complex adult organism, has fascinated biologists for more than 200 years. In the mid-nineteenth century, before and during the time when Charles Darwin was uncovering the principles of natural selection, a number of biologists who wondered what caused evolutionary relationships among organisms looked to development for answers. The German zoologist Ernst Haeckel popularized the phrase “Ontogeny recapitulates phylogeny” — where ontogeny is an organism’s development and phylogeny is its evolutionary relationships. You may have seen a version of his famous diagram in biology textbooks (Fig. 1). Haeckel suggested that, during each successive stage of development, an animal would pass through a stage from its evolutionary history (for example, in the fourth week of development, human embryos have structures called pharyngeal arches, which look similar to gill slits in fishes). In the 1920s, marine biologist Walter Garstang contributed to discrediting Haeckel’s theory, instead proposing that ontogeny creates phylogeny. Garstang suggested that natural selection is able to modify the form of each larval stage independently, leading to complex life cycles in many marine organisms.

Figure 1 — A representation of Haeckel’s recapitulation theory, illustrating similarities in vertebrate embryos.  Source.

Even now, nobody knows the exact relationship between ontogeny and phylogeny. But modern biologists have made startling discoveries. The genetic ingredients used to build the bodies of all animals, from human limbs to fruit fly wings, are remarkably similar. These areas of research make up the emerging field of biology called evolutionary developmental biology, or evo-devo. This includes, but is not limited to: analyses of how embryonic development evolved; the origin and modification of anatomical features; the origin and modification of life-history stages; and the interaction of genetics (genotype) and observable characteristics (phenotype) throughout the life cycle. Ultimately, evo-devo seeks to answer the questions: why do organisms look the way they do and how did they get that way? Evo-devo has also been combined with palaeontology for increasingly powerful studies of evolution. Here I will introduce what fossils can tell us about development, as well as what development can tell us about the fossil record.

What can fossils tell us about development?

Fossil embryos have been described for a variety of animal groups from the Cambrian period (542 million to 488 million years ago). Some palaeontologists say that they have even found fossil embryos from the Ediacaran period (635 million to 542 million years ago), but these have mostly been discredited. The Cambrian embryos that have been assigned to groups that still have some living species —  such as the cnidarians, which include jellyfish — demonstrate that development has been conserved in some animals for more than 500 million years. Larval stages (which come in development after the embryo and are usually free-swimming) have also been preserved from a variety of animals, including vertebrates and arthropods. Fossil arthropod larvae are frequently preserved, because arthropods must completely shed their exoskeletons repeatedly during growth, a process known as moulting. Each moult has the opportunity to become fossilized, whether the animal is still living inside it or not. Some of the best-known fossil larvae come from trilobites. Thanks to preserved larvae, the patterns of segment addition in many trilobite species are well understood.

Lots of larval fossils are exceptionally preserved — that is, anatomical features that would normally decay and not be preserved are visible. Exceptional preservation can occur only under specific environmental conditions, and deposits that provide such conditions are rare but of tremendous scientific value. One group of deposits that bear fossil larvae, found worldwide and dating from about 525 million to 480 million years ago (in the Cambrian period and the Ordovician period), are known as Orsten. They include tiny fossils embedded in round limestone nodules, which can be revealed only by dissolving the rock in acid, and then painstakingly sorting through the remaining debris under a microscope. The resulting fossils are truly spectacular (Fig. 2).

Figure 2 — Representative Orsten fossil arthropod larvae. (a–c) Rehbachiella kinnekullensis, a possible relative of crustaceans such as sea monkeys. Image (c) is a magnification of the limbs of the fossil in (b), illustrating exquisitely preserved sensory hairs in orange (the scale bar is 30 micrometres, or a little smaller than the diameter of a human hair!). (d) Bredocaris admirabilis (discussed below). (e–f) The phosphatocopines Klausmuelleria salopensis and Hesslandona angustata, with their protective carapaces split open to reveal details of the limbs. What animal groups they belong to remains a source of controversy. (g–h) Yicaris dianensis, with a possible crustacean structure magnified in image (h) in purple. Sources.

Fossils are important for discovering evolutionary relationships, because scientists cannot always tell what anatomical features evolved from the same common ancestor, especially if those groups evolved a very long time ago. Fossils are closer in age to the time when particular groups split apart, and often have a mishmash of characters shared by each of the living related groups. Consider, for example, the famous fossil Archaeopteryx lithographica, which had both birdlike characters (feathers, wings, a wishbone) and reptilian characters (teeth, a long tail), and became key evidence for the discovery that birds evolved from dinosaurs.

Focusing on larvae, Orsten fossils are very important for understanding evolutionary relationships among arthropods, because they are some of the oldest members of living groups. My research looks at how larval stages affect the methods used to reconstruct phylogeny, and the Orsten fossils have been fundamental in my work. I have found that these larvae can make a difference in analyses of relationships between animals, and can extend the geological time range of living groups. For example, the Orsten fossil Bredocaris admirabilis is an early relative of barnacles, but it shows similarities to only larvae of living barnacles, rather than adults. The oldest adult barnacle fossil is from the Silurian period, and dates to about 425 million years ago, but B. admirabilis tells us that barnacles split from their common ancestor at least 75 million years earlier than was previously known — that’s a longer time than between the extinction of the dinosaurs and now!

Fossilized embryos may also provide information about the creature’s mode of life and behaviour. For example, embryos of ichthyosaurs and mosasaurs (marine reptiles from the Mesozoic era, 251 million to 65 million years ago) have been preserved. Certain specimens of these fossil embryos have been discovered lying in place in their mother’s womb, demonstrating that these animals gave birth to live young. Furthermore, a specimen of an armoured fish from the Devonian period of Australia preserved an embryo in place with its umbilical cord (Fig. 3), showing that many vertebrates gave live birth as far back as 380 million years ago!

The armoured fish Materpiscus attenboroughi (named after Sir David Attenborough). On the left is a reconstruction of the mother and its baby attached by an umbilical cord; on the right is the fossil itself. Source.

What can development tell us about fossils?

Unfortunately, the cloning of dinosaurs in Jurassic Park is pure fantasy: DNA sequences degrade after only hundreds of thousands of years. We cannot extract genetic information from the vast majority of fossils (although we can from the remains of more recently extinct species such as Neanderthals and mammoths!). The major body plans of animals evolved in the Cambrian, so what can the developmental biology of living animals tell us about extinct species? The answer is that we can study the genes that direct the processes of development, and make inferences about their evolutionary origins.

The overwhelming majority of knowledge about development in living animals comes from only a few species: the fruit fly (Drosphila melanogaster), the nematode or roundworm (Caenorhabditis elegans), the sea urchin (Strongylocentrotus purpuratus), the zebrafish (Danio rerio), the African clawed frog (Xenopus laevis), the chicken (Gallus gallus) and the mouse (Mus musculus). Four out of seven of these species are vertebrates, but vertebrates represent only 3% of described species diversity (Fig. 4). A significant goal of evo-devo is to expand this knowledge to other groups. It can be difficult to do experiments with many animals, either because there are problems with collecting them (for example, coelacanth fish live only in deep waters of the Indian Ocean) or with getting them to reproduce in the laboratory. An alternative approach is to use rapidly improving genomic technology. As recently as three years ago, it was difficult and very expensive to sequence a full genome. Now technology has become so efficient that we can sequence a full genome in a matter of days, for as little as US$8,000! Genomic information is incredibly useful for evolutionary study because scientists can mine the databases already created for other organisms to see how many genes are related, and what their function is (if it is known).

Figure 4 — The evolutionary relationships of developmental model species, showing the emphasis on vertebrates. M indicates common model species; m indicates species that have been studied (sometimes for only one or a few genes). Modified from: Jenner and Wills 2007.

Perhaps the most intriguing genomic data come from what is known as the developmental toolkit: the subset of genes required for development of the adult body plan. A large number of genes function solely to send signals to other genes, telling them to turn on or off (gene expression). Then those genes send signals to more genes, and on and on in a cascade of gene regulation, so the body is able to respond by, for example, creating proteins to build more cells to increase the size of the growing leg or heart or eye (see Video 1). The DNA sequences of the genes responsible for these signals are common across deep evolutionary time.

A particularly important and famous group of toolkit genes are known as homeotic (Hox) genes. These genes — whose sequences are similar between most animals, from humans to fruit flies — lay down the pattern of building blocks from the anterior (head) to posterior (rear) of the body (Fig. 5). These genes signal that a particular body segment should be made in that location — for example, in the fruit fly, they might signal that the second segment of the thorax (body) has wings whereas the third segment has a structure called halteres (used in flight).

Hox genes in the fruit fly. (a) The set of Hox genes are found in the same order on a chromosome as the order in which they are expressed from anterior to posterior in the embryonic and adult fly. (b) The effects of expressing the Antennapedia gene in a segment anterior to its natural position: wt is a normal fly, and antp is the mutant, with legs (found in the natural Antennapedia segment) growing from the head. (c) Expression of Ultrabithorax posterior to its normal position produces an extra pair of wings on the fly. Sources: A, B, C.

Mutations of these toolkit genes (natural or experimentally induced) provide clues as to their function (Fig. 5b, c). In this way, living embryos can be experimentally changed to develop features that look similar to those found only in fossils. For example, mutant chicken embryos have been given teeth, a characteristic lacking in modern birds, but found in crocodiles and dinosaurs. Homeotic shifts have also been induced in chicken wings so that they resemble morphologies found in dinosaur fingers. Of course, we cannot prove that these particular genes were responsible for patterning dinosaur teeth or fingers, but experiments do provide insight into potential pathways, especially in cases that we can also compare to the living sister group, such as crocodiles with birds and dinosaurs. Similarly, we cannot measure patterns of Hox gene expression in fossil arthropods and their relatives, but with some knowledge of the phylogenetic distribution of body patterning in living groups, and the relationship of fossils to those living species, it is possible to infer, for example, how the body is divided into different types of segments. Trilobites, as probable relatives of mandibulates (centipedes, millipedes, crustaceans, and insects) probably shared gene-expression patterns with those arthropods, especially centipedes and millipedes (which, like trilobites, have a series of repeated similar segments throughout the body).

Although most developmental genes were present in the common ancestor of cnidarians and bilaterian animals, the number of regulatory microRNA sequences in the genome increased dramatically as bilaterians evolved (Fig. 6). The function of microRNAs remains an active area of research, but we do know that they do not produce genes, yet they are involved in regulating when (during development) and where (in the body) other genes are expressed. It has been suggested that increases in the number and type of microRNAs contributed to the increase in complexity of animal forms during the Cambrian explosion. Note that in Fig. 6 the two largest increases are at the base of bilaterians (the first appearance of bilateral symmetry and three embryonic cell layers, as opposed to the two in cnidarians) and at the base of the vertebrates (with an internal skeleton and complex brain). Knowledge of the evolutionary distribution of developmental toolkit genes allows us to infer when in the fossil record complex body plans might have occurred, and provides some explanation for the rapid appearance of new body plans during the Cambrian explosion.

Figure 6 — Evolutionary acquisition of developmental toolkit genes. Most messenger RNAs (protein-coding genes) evolved in the common ancestor of cnidarians and bilaterians, but the set of microRNAs had massive increases at the base of bilaterians and again in vertebrates, indicating significant leaps in developmental complexity in those groups. Source: Erwin et al. 2011.

Summary:

Emerging research at the intersection of palaeontology and evo-devo is extremely exciting. As more and more fossil embryos and larvae are found, we learn about the development of extinct animals and gain new sources of data to understand their relationships with living groups. Complementary work on the genomics and functional development of living animals reveals shared aspects of body-plan organization throughout deep evolutionary time. A combination of these two approaches is needed to move towards understanding the centuries-old questions perplexing zoologists: why do organisms look the way they do and how did they get that way?

Suggestions for further reading:

Carroll, S. B. 2005. Endless Forms Most Beautiful: The New Science of Evo Devo. W.W. Norton. ISBN: 9780393060164. An episode of the show NOVA was partially based on this book. It can be watched online, and is excerpted in Video 1.

Erwin, D. H., Laflamme, M., Tweedt, S. M., Sperling, E. A., Pisani, D. & Peterson, K. J. 2011. The Cambrian conundrum: early divergence and later ecological success in the early history of animals. Science 334, 1091–1097. doi:10.1126/science.1206375

Hall, B. K. 2012. Evolutionary developmental biology (evo-devo): past, present, and future. Evolution: Education & Outreach 5, 184–193. doi:10.1007/s12052-012-0418-x

Hall, B. K. & Olson, W. M. 2003. Keywords and Concepts in Evolutionary Developmental Biology. Harvard University Press. ISBN: 0674009045.

Harris, M. P., Hasso, S. M., Ferguson, M. W. J. & Fallon, J. F. 2006. The development of archosaurian first-generation teeth in a chicken mutant. Current Biology 16, 371–377. doi:10.1016/j.cub.2005.12.047

Sánchez, M. R. 2012. Embryos in Deep Time: The Rock Record of Biological Development. University of California Press. ISBN: 9780520271937.

Shubin, N. 2009. Your Inner Fish: A Journey into the 3.5-Billion-Year History of the Human Body. Vintage Books. ISBN: 978-0-375-42447-2, The same NOVA episode was also partially based on this book.

Waloszek, D. & Maas, A. 1996–2013. Center of ‘Orsten’ Research & Exploration. A fantastic website with loads of information, pictures and links about Orsten fossils.

¹Division of Invertebrate Zoology & Sackler Institute for Comparative Genomics, American Museum of Natural History, Central Park West at 79^th Street, New York, NY 10024, USA.

Introduction:

Trilobites make up one of the most fascinating and diverse groups in the fossil record. Over the course of their long history — which dates back to near the beginning of the Cambrian period, around 520 million years ago — they have inhabited a wide range of marine environments, from reefs to abyssal depths. In addition, trilobites have evolved several different life strategies, from burrowing to swimming; these are reflected in their varied appearances, or morphologies (Fig. 1). Several species, famously those from the Devonian period of Morocco (about 420 million to 360 million years ago), developed a rich array of protective spines, which has made them a popular choice among fossil collectors and dealers.

Figure 1 — Key trilobite species discussed in the text. (a) Balcoracania dailyi (~10mm), (b) Paradoxides gracilis (110mm), (c) Olenellus (~20mm); (d) Acadoparadoxides briaerus (300mm) (e) Isotelus rex (720mm), (f) Parabarrandia bohemica (120mm), (g) Pricyclopyge binodosa binodosa (20mm), (h) Cybantyx anaglyptos (50mm), (i) Aulacopleura konincki (20mm); (j) Leonaspis (10mm), (k) Erbenochile erbeni (50mm), (l) Kathwaia capitorosa (~30mm), (m) Ditomopyge (~30mm). Sources: (a) Paterson and Edgecombe, 2006; (c,j) Levi-Setti, 1993; (e) Rudkin et al., 2003; (g) Horný & Bastl, 1970; (h) Whittington, 1997; (i) http://alturl.com/j58tk (k) Chatterton and Gibb, 2010; (l) Owen, 2003; (m) http://alturl.com/a8giq.

The earliest scientific report of a trilobite fossil was by Reverend Edward Lhwyd in 1698. This specimen was collected near Llandeilo in South Wales, and was originally described by Lhwyd as “some kind of flat fish” (Fig. 2a). Later, following more and more fossil discoveries, trilobites were recognized as arthropods, more closely related to crabs, spiders and lobsters than to fish. In the past 200 years, more than 17,000 species have been described, and trilobites have become increasingly important in the formulation of geological and evolutionary ideas. Historically trilobites were primarily used in the dating of rock strata and in the 19^th century geologists including Adam Sedgwick and Roderick Murchison used different species of trilobites (as well as other groups) to define sections of the Cambrian and Silurian respectively. Variation in the number of lens in the eyes of the genus Phacops was used by Niles Eldredge and Steven J. Gould as evidence for the theory of punctuated equilibrium, whereby species evolve with little net morphological change (stasis) which is occasionally interrupted by short periods of sudden change associated with speciation events (cladogenesis).

Figure 2 — (a) Reverend Edward Lhwyd’s “flat fish”. Source: http://alturl.com/9s5o9 (b) Trilobite bar in Prague. Credit: David Bressan. (c) Cambrian trilobites described by Joachim Barrande in 1852. Source: http://alturl.com/o7ck9.

In some countries, trilobites have even taken on importance as cultural icons. During the nineteenth century, the French geologist Joachim Barrande described in detail many fossils from what is now the Czech Republic, including assorted trilobites (Fig. 2c). The Czech city of Prague is very proud of its rich palaeontological heritage and trilobites can today be seen in many unusual places, including restaurants (Fig. 2b), and are also commonly used as the focus for the carvings and sculptures seen in walls and paving stones.

Trilobite diversification:

The oldest trilobites come from the lower Cambrian of North America, Siberia and Morocco. The fossil record shows that the number of trilobite species rapidly increased during the Cambrian, with the group reaching its peak diversity in the middle to upper Cambrian, around 500 million years ago (Fig. 3). Following this high point, trilobite diversity generally declined for about the uppermost Permian period, when the last remaining species went extinct. This fall in diversity was punctuated by two pulses of rapid diversification. The first of these was the ‘Great Ordovician Biodiversification Event’, which occurred from about 485 million to 460 million years ago and saw the innovation of forms that would be common for the rest of trilobite history (asaphids (Fig. 1e-g), phacopids (Fig. 1k) and proetids (Fig. 1i,l,m)), whereas the typical Cambrian (redlichiids and olenellids; Fig. 1a-d) forms became extinct. The second and last major rise in diversity occurred in the Lower to Middle Devonian (about 419 million to 382 million years ago) and was shorter-lived. This period was characterized by the evolution of very spiny forms (such as that in Fig. 1i-j); the development of these structures and the simultaneous increase in diversity could be the result of the trilobites being preyed on by other invertebrate groups, especially cephalopods such as orthocones, which are related to the modern nautilus. An increase in diversity due to predation pressure might sound counterintuitive, but is in fact very common in the natural world. Organisms develop adaptations to avoid being eaten, in this case through strengthening their defensive capabilities, which in turn drives a diversity increase in the group being predated upon; this evolutionary ‘arms race’ effect was named the escalation hypothesis by palaeontologist Geerat Vermeij in the 1980s.

Figure 3 — Species diversity of trilobites. Based on occurrences from the Paleobiology Database: Kiessling, W., M. Foote, A. I. Miller, M. E. Patzkowsky, P. J. Wagner, S. M. Holland, M. J. Hopkins, M. E. Clapham, and P. M. Novack-Gottshall. 2013. Taxonomic occurrences of Trilobita. Paleobiology Database. http://paleodb.org.

Morphology and ontogeny:

The name ‘trilobite’ (Latin for “three-lobed”) comes from the way that the exoskeleton is divided into a central axis with lobes (left and right pleural lobes) on either side (Fig. 4a). However, trilobites are also divided into three sections (called tagmata) from front to back. These are a head shield (the cephalon); a body made of articulated segments (the thorax); and a tail shield consisting of multiple fused segments (the pygidium) (Fig. 4b). From this seemingly simple arrangement comes an incredible plethora of form with species evolving a wide range of features from smooth, effaced exoskeletons (Fig. 1e-h), high spinosity (Fig. 1j,k) or smaller (or complete loss of) eyes (Fig. 1f). Trilobites show excellent examples of convergent evolution, whereby morphological characters such as these are independently evolved in unrelated species e.g. effaced exoskeletons occur in both the families Asaphidae (Fig 1. e) and Illaenidae (Fig 1. h). An interesting example of convergence is Aulacopleura konincki (Fig 1. i) from the Silurian of the Czech Republic that evolved features common to basal trilobites species including a narrow thoracic axis, large (and variable) number of thoracic segments and a small pygidium. This is likely associated with the adaptation towards life in low oxygen, high sulphur environment, known as the olenimorphic type; this was common in the Cambrian.

Figure 4 — Trilobite morphology. (a) upper (dorsal) and (b) lower (ventral) views. 1, cephalon; 2, thorax; 3, pygidium; 4, right pleural lobe; 5, axial lobe; 6, left pleural lobe; 7, facial sutures; 8 doublure. (c) Conterminant and (d) natant hypostome attachment condition; the hypostome is marked in blue and the outline of the glabella in red, (e) hypostome and doublure of Hydrocephalus carens. Sources: (a-d) Copyright Sam Gon III 1999–2005.

As with other arthropods, trilobites grew by moulting, passing through a series of distinct developmental stages or instars. During moulting, the exoskeleton broke across joints, or sutures, between the segments, such as the facial sutures, which join the free cheeks to the rest of the cephalon (Fig. 4a,b). After moulting, the individual took on water to swell its body size before growing a new exoskeleton.

Moulting was an important part of the trilobite life cycle. In the early stages of development, this process was used to increase the number of thoracic segments, as well as the overall size of the animal. Trilobite development can be broadly divided into three main stages: protaspid, meraspid and holaspid (Fig. 5). The protaspid and meraspid phases involve moulting followed by an increase in segment number (anamorphic growth pattern). In the protaspid stage is represented by a circular disc with facial sutures but with no division between body sections. The meraspid stage begins when the first segment between the cephalon and the pygidium develops. During the following moults of the meraspid stage, segments are released from the front of the pygidium into the thorax. This continues until a stable number of segments is reached. After that, the holaspid stage begins, and the individual increases only in size after each instar (epimorphic growth pattern). The stable number of thoracic segments varies across the group, ranging from two in agnostids to more than 100 in Balcoracania dailyi. This has led some researchers to suggest that some species continued to add segments throughout their entire lives.

Figure 5 — Trilobite ontogeny. Source: Hughes, 2007.

Trilobites have one of the largest size ranges of any Palaeozoic arthropod group — with the notable exception of the eurypterids. Adult, holaspid trilobites typically range from less than 1 centimetre up to around 10 centimetres long. The largest known complete trilobite, Isotelus rex, from the Upper Ordovician of Canada, measures 72 centimetres. Although large trilobite fossils are found in regions that were around the equator when trilobites were alive, they seem to be more abundant in areas that were located close to the South Pole, especially during the Cambrian and Ordovician. The trilobite species Acadoparadoxides briareus, which is found in rocks called the Jbel Warmast Formation in the Middle Cambrian of Morocco (Fig. 1d) typically reach around 30 centimetres. In addition, the Valongo Formation in the Middle Ordovician of Portugal contains a highly diverse variety of giant trilobite species including Hungioides bohemicus and Ogyginus forteyi, the latter of which is estimated to have reached around 90 centimetres.

Lifestyle and feeding strategies:

It is difficult to reconstruct the life habits or ecologies of wholly extinct groups such as trilobites, but we can make some inferences based on their exoskeletal morphology, comparisons with extant relatives and the types of rocks they are preserved in.

Trilobites can be broadly separated into two categories: predators and deposit feeders. One feature used to recognize these categories is how the hypostome, a plate on the lower surface of the exoskeleton, is attached (Fig. 4c,d). The hypostome was generally lined up with the back of the glabella, a dome that covered the stomach and mouthparts. In predatory trilobites, the hypostome was rigidly fixed to a rim around the underside of the trilobite, called the doublure (Fig. 4), allowing the trilobite to process prey such as worms and small crustaceans. This is called the conterminant condition By contrast, in deposit-feeding species the hypostome was less tightly fixed to the doublure (the natant condition), and would have been used mainly for sieving through mud on the sea floor for food particles. Predatory trilobites have other characteristics that further differentiate them from deposit feeders, including much greater variability in overall form, a more arched exoskeleton and a generally greater size range. These species appeared with the first trilobites in the lower Cambrian, but became extinct in a mass extinction during the Late Devonian period, with only the deposit-feeding proetids surviving until the end of the Permian.

Most trilobites are thought to have been benthic, living on or just below the sea floor; however, during the Ordovician, another life strategy became dominant. Several groups evolved morphologies consistent with an active, free-swimming lifestyle in the open ocean. The families Telephinidae and Cyclopygidae are typical of this life style, having developed a more streamlined body and a wider thoracic axis, presumably due to the need for stronger swimming muscles (Fig. 1f,g). They also developed greatly inflated eyes, which allowed for 360-degree vision. These species tend to be found in rock layers that formed in a deep marine environment that was penetrated by only a small amount of sunlight, and so usually contain blind trilobite species. This suggests that the swimming animals lived higher in the water column, falling into the depths only after their death.

Phylogeny:

Despite being one of the best-studied fossil groups, trilobites still cause a number of phylogenetic controversies. These mainly revolve around their relationships in the Euarthropoda (arthropods defined by a cephalon with conjoined segments, antennae and three pairs of limbs that each branch into two parts) and the relationships of groups within trilobites. Current evidence suggests that trilobites are closely related to chelicerates (a group containing spiders, scorpions, mites and horseshoe crabs), mandibulates (crustaceans, myriapods and insects) or both. Another issue concerns the relationships of the suborder Agnostina, which consists of small, blind species with only two thoracic segments. Many researchers place the group as a close relative of the eodiscid trilobite group, but others exclude it from the trilobites entirely, thinking instead that the organisms are crustaceans.

Summary:

Despite certain areas of uncertainty the study of trilobites shows no sign of abating, even after more than two centuries of collecting. New species are continually being discovered and described, illuminating previously unknown features and in turn increasing our understanding of the evolutionary history of this fantastic fossil group.

Suggestions for further reading:

Fortey, R. A. 2001 Trilobite! Eyewitness to Evolution. Flamingo. (ISBN:978-0006551386)

Fortey, R. A. & Owens, R. M. 1999. Feeding habits in trilobites. Palaeontology 42, 429–465. (doi:10.1111/1475-4983.00080)

Gutiérrez-Marco, J. C., Sá, A. A., García-Bellido, D. C., Rábano, I. & Valério, M. 2009. Giant trilobites and trilobite clusters from the Ordovician of Portugal. Geology 37, 443–446. (doi:10.1130/G25513A.1)

Hughes, N. C. 2007. The evolution of trilobite body patterning. Annual Review of Earth and Planetary Sciences 35, 401–434. (doi:10.1146/10.1002/bies.10270)

Hughes N.C., Chapman R.E., Adrain J.M. 1999 The stability of thoracic segmentation in trilobites: a case study in developmental and ecological constraints. Evolution and Development 1(1), 24—35. (doi:10.1046/j.1525-142x.1999.99005.x)

Levi-Setti R. 1993 Trilobites. Chicago, University of Chicago Press; 342 p. (ISBN:978-0226474526)

Rudkin, D. M., Young, G. A., Elias, R. J. & Dobrzanski, E. P. 2003. The world’s biggest trilobite—Isotelus rex new species from the Upper Ordovician of northern Manitoba, Canada. Journal of Paleontology 77, 99–112. (doi:10.1666/0022-3360(2003)077<0099:TWBTIR>2.0.CO;2)

www.trilobites.info — A fantastic website created and maintained by Sam Gon III, covering everything you always wanted to know about trilobites but were afraid to ask.

¹Department of Earth Sciences, University College London, Gower Street, London WC1E 6BT, United Kingdom.

Introduction:

Sitting in the sweltering heat of southern Japan, I’m faced with a conundrum. The limestone cliff in front of me preserves the boundary between the Permian and Triassic periods, a point in time around 250 million years ago that witnessed the greatest mass extinction of the Phanerozoic eon. I’m collecting rock and fossil samples from around this boundary to study how the make-up of fossil communities changed in response to this extinction event: this is palaeoecology. The boundary itself couldn’t be easier to spot — the lower (and older) part of the cliff is composed of a pale white-yellow limestone packed full of fossils of shelled marine invertebrates including brachiopods, bivalves and gastropods, as well as microscopic sea-floor-dwelling (benthic) creatures called foraminifera. Some of these foraminifera have been found elsewhere in the world and are dated to the Permian period. The younger, higher rocks are an ominous-looking black, with fine layering and a blotchy texture that you might otherwise associate with old blue cheese. Fossils in this dark, rotten-looking limestone are extremely rare and dominated by one or two species of mollusc, but researchers have found fossilized teeth belonging to an eel-like animal called a conodont, of the species Hindeodus parvus, which unequivocally dates the rocks as Triassic in age. Somewhere at the boundary between these white and black limestones, 95% of all marine organisms with skeletons became extinct in the geological blink of an eye — currently thought to be less than 200,000 years. The palaeontological story would seem to be extremely simple: a diverse Permian benthic marine community suffered a mass extinction and was replaced by a community composed almost entirely of one or two species (Fig. 1). This pattern is broadly the same all over the world during this transition. So where is the conundrum?

Figure 1 — Top: Sampling the Permian–Triassic boundary in the field, in Kamura, Kyushu (southern Japan). Dashed line indicates position of the boundary. Bottom: Schematic log section of the boundary (‘PT MEB’ – Permian-Triassic mass extinction boundary), with fossil diversity (i.e. total species) and evenness (relative abundance of different species) curves.

The problem comes with deciphering the striking colour change between the Permian and Triassic limestones. The shift from white to black actually has very little to do with the extinction itself, but instead records a dramatic environmental change. The white Permian limestone was laid down in a shallow marine lagoon. The dark and mottled Triassic limestones record something very different — an algal marsh along a shoreline, very similar to that forming on the modern-day Andros Island in the Bahamas. The fine layering and mottled texture were produced through complex interaction between fast-growing algae and sediment carried in by storms. The algae formed flat, sticky mats in low-lying areas protected from the wind, and with surprising adhesive properties. During storms, sediment (made up mostly of clumps of carbonate mud, foraminifera and gastropod shells) was stirred up into the water column, and then transported onshore as part of the storm surge. A thin layer of this sediment was trapped on top of the mat, and became fixed as the algae grew through and around it. This is what produced the layering and unsettling ‘blotchy’ texture of the limestone.

The fundamental environmental change that occurred here across the Permian–Triassic boundary highlights two issues that complicate the interpretation of these bodies of limestone: 1) they represent very different environments that probably hosted very different original communities; and 2) these two settings probably preserve very different components of the community (one might preserve small creatures and the other big ones, for example, or one might preserve those with hard shells and the other those with soft bodies). As a result, the fossils record the original living communities with varying accuracy.

Palaeontology helps us to deal with the first problem by comparing fossils found in rocks representing similar environments at different times, so that we know what sort of things are recorded in each type of rock: we compare apples with apples and blotchy oranges with blotchy oranges. Dealing with the second problem is slightly more complex. Processes such as being shifted by water currents, winnowing (whereby small and light material is swept elsewhere), selective predation (where certain species are destroyed or taken elsewhere) and disarticulation (creatures’ bodies breaking up after death) can strongly distort the appearance of the community and mask changes in community structure. Furthermore, the relative importance of these processes will vary between environments. As we go through different settings in the geological record, then, how do we know that the fossils that we find accurately represent the original make-up and ecologies of the living communities? Fortunately, live–dead studies conducted in modern environments offer a way to test the quality of the fossil record in a wide variety of sedimentary environments.

How do life–dead studies work?

On the face of it, live–dead studies are extremely simple. You choose a modern environment where sediment is being laid down and begin collecting members of the living and the dead communities (Fig. 2). In marine environments, the living community can typically be found: on or in the sediment (where you might find, for example, clams, sea urchins and soft-bodied worms); attached to blades of seagrass and other algae (many foraminifera and small gastropods); and at various heights in the water column (fish, squid and jellyfish, among thousands of others). Although some of these organisms may be rarer than others, and they may never interact, they all make up the living community in that environment, and in an ideal world would all enter the fossil record. The dead community, by contrast, is largely restricted to the sea floor, making up the sediment and organic debris scattered on and in the surface. This is the precursor or ‘sub-fossil’ record, and gives a good indication of what a palaeontologist might expect to see in the rock many millions of years later. Holding any handful of sediment under a microscope will reveal the typical contents: worn and broken shells, the broken up remains of sea urchin skeletons, and perhaps the withered cuticles of a few small arthropods.

Figure 2 — Top: Testing the quality of the fossil record using benthic foraminifera; a live–dead study on San Salvador Island in the Bahamas. Bottom: Some common benthic foraminifera from this setting. Live–dead agreement is in general very poor, with only about 20% of species in the death assemblage also found in the living community, possibly as a result of recent human impact. Species are: a. Psuedohauerina sp.; b. Borelis pulchra; c. Chysalidinella dimporpha; d. Reusella spinulosa; e. Triloculina bicarinata; f. Amphistegina gibbosa; g,l. Cymbaloporetta squamosa; h. Rosalina floridiana; i. Vertebralina mucronata; l. Vertebralina sp.; m. Pyrgo comata; n,o. Asterigerina carinata; p. Archaias ungulatus. Figure modified from Darroch (2012 – reference below).

How well these live and dead communities match (‘live–dead agreement’) is an effective measure of the potential quality of the fossil record in that environment. For clarity, palaeontologists refer to the living community and the death assemblage. The difference in terminology is due to the fact that the dead material is typically composed of biological remains both derived from the local environment and transported in from elsewhere (and potentially encompassing a large range of ages). Live–dead agreement can be calculated either on a presence/absence basis (who is there and who is missing?), or in terms of relative abundance (are the common species the most frequently preserved?). Both measures provide valuable information, and can be used to re-calibrate the fossil record in terms of how well the overall diversity and ecological make-up of original communities is being preserved.

History and recent advances:

Although the field of taphonomy, or fossil preservation, has enjoyed more than 70 years of study, the analysis of live–dead agreement as a way to interpret the past was first thrust into the limelight by the palaeobiologist Thomas Schopf in the late 1970s. Schopf undertook a comprehensive live–dead study of the organisms in the area between high tide and low tide in Friday Harbor in the US state of Washington across three environmental settings (muddy, sandy and rocky substrates). He looked at 169 genera in a wide range of animal groups. The principal findings were encouraging: a relatively large proportion of invertebrates visible with the naked eye (the groups typically considered in paleoecological studies) had better-than-expected frequencies of preservation. But the study also highlighted what many palaeontologists had suspected for years: not a single wholly soft-bodied group (such as marine worms, sea slugs or jellyfish) observed in the living community was found in the death assemblage. This is perhaps the most obvious taphonomic ‘megabias’ in the fossil record — soft-bodied animals are almost never preserved, so compilations of fossil diversity through time only really represent the diversity of biomineralized animals and plants, which is far from a complete picture. This observation also highlights why fossil deposits that preserve the remains of soft-bodied organisms are so important; they represent snapshots in time when live–dead agreement is much higher than usual, providing a much more complete picture of the palaeocommunity (Fig. 3).

Figure 3 — Potential ‘megabias’ in the fossil record when soft-bodied animals are not preserved. Top: Reconstruction of the Cambrian Burgess Shale deposit as it would look if only biomineralized (shelly) animals were fossilized, and not soft-bodied ones. Bottom: Reconstruction of the Burgess Shale as it may have actually appeared in life (i.e. with soft-bodied animals). Note that live–dead agreement is almost always higher when soft-bodied animals are fossilized. Both images from Fossils of the Burgess Shale, edited by Derek E.G. Briggs, Douglas H. Erwin, Frederick J. Collier and Chip Clark (1995).

Fortunately, however, palaeontologists can achieve a great deal by looking at biomineralized organisms alone, and over the past 30 years studies in live–dead agreement have made huge advances in calibrating the accuracy of fossil assemblages, as well as isolating and quantifying the relative impacts of specific processes in different environments. More and more careful live–dead studies are providing powerful ‘taphonomic vindication’ for the study of fossil communities. For example, it has been demonstrated that in modern communities of benthic molluscs, the abundance of species is more often than not well preserved in death assemblages. Put simply, species that dominate the living community tend to be more common in the piles of dead shells that accumulate on the sea floor; that may seem a trivial finding, but it is great news for palaeontologists! In addition, whereas a single sample of the living community will typically contain only species that happen to be there at that ecological instant, death assemblages represent the accumulation of dead material over time, and so typically contain more of the rare species that you might otherwise miss; this means that death assemblages actually paint a better and more complete picture of a given community on reasonable ecological timescales (weeks to years). Finally, even within transects drawn through a single living community of molluscs across an area of sea floor, researchers have shown that measures such as evenness (the relative abundance of different species – a metric beloved by palaeontologists studying mass extinctions) can be replicated faithfully in their corresponding death assemblages. In these cases, death assemblages (and the ‘sub-fossil’ record) provide extraordinary records of the composition and distribution of the original communities. These studies therefore show that when we find these environments in the fossil record, we can trust the fossils in them to be an accurate record of what was once living there.

Even when live–dead agreement in easily preserved organisms is shown to be poor, palaeontologists can turn it to their advantage. One of the most important reasons why living communities and death assemblages might show little agreement involves a hot-button term — human impact. It is no secret that humans are having a detrimental impact on the oceans; as the concentration of carbon dioxide in the atmosphere rises, more is absorbed by the oceans, making them more acidic. In coastal areas next to big cities, the water is being contaminated with everything from heavy metals and plastic to nitrates and organic fertilizers. Organic material will decay, using up oxygen in the process and leaving none for invertebrates such as molluscs and crustaceans. Other pollutants may act as outright poisons. In these settings, living communities tend to contain few organisms, and to be dominated by one or two hardy species, similar to the Triassic limestones described above. The death assemblage, however, may contain an accumulation of shell material dating back before the arrival of humans and pollutants — a species-rich and high-evenness assemblage that records the make-up of the community in its original pristine state. Here, then, the living community and the death assemblage are very different. Live–dead agreement (specifically, poor live–dead agreement) acts as an indirect measure of pollution and human impact, and so is an important tool in the emerging field of conservation palaeobiology, in which palaeontological data is used to provide information about important issues in ecology and conservation. Ecologists and palaeontologists alike can use live–dead studies to measure human impact and ecosystem health, and, if necessary, can use them to work out where and how to try to reverse any damage to the environment.

Wrapping up:

So where does that leave me, apart from sitting and staring at my cliff section (still sweltering, and now scratching irritably at some insect bites)? The palaeoecological data show a transition over the Permian–Triassic boundary, from an assemblage bursting with fossils to one containing almost nothing, save for a few lonely bivalves. The story is the same the world over, but are the fossils faithfully recording the living community? The pale Permian limestones probably represent deposition in a warm shallow-water lagoon; live–dead studies in equivalent modern settings suggest that the sea floor here also played host to a rich community of plants, arthropods, and countless soft-bodied organisms. None of these have been preserved as fossils, but the biomineralized groups at least should provide a reasonable record of both the overall diversity and relative abundance of immobile molluscs and brachiopods. The dark Triassic limestones, by contrast, record periodic deposition and algal growth in a shoreline algal marsh; the fossil bivalve shells were probably swept onto shore during storms, but nothing living in the marsh itself stood much chance of being preserved. In the marsh, live–dead agreement was almost certainly extremely low. Sadly, in this instance there is very little we can say about the rate or pattern of Permian–Triassic extinction and recovery, because the environments represented by these two units did not preserve their original communities with equal quality. There is an interesting story here, but it doesn’t involve pre- and post-extinction palaeoecology. Fortunately, not too far away there is another Permian–Triassic section composed of limestone from an area that was almost always submerged in water; there are still changes in the types of rock across the boundary, but they record palaeoenvironments that can be (and have been) studied in the context of their live–dead agreement. In the coming years, palaeontologists will attempt to calibrate all the settings we see in the fossil record, in terms of what is preserved and what isn’t, so that when it comes to studying the composition of fossil communities, we can compare apples with apples across the boundary, rather than apples with blotchy oranges.

Suggestions for further reading:

Darroch, S.A.F. 2012. Carbonate facies control on the fidelity of surface-subsurface agreement in benthic foraminiferal assemblages: implications for index-based paleoecology. Palaios 27, 137–150. (doi:10.2110/palo.2011.p11-027r)

Gould, S. J. 1984. The life and work of T. J. M. Schopf (1939–1984). Paleobiology 10, 280–285. (http://www.jstor.org/stable/2400401)

Kidwell, S. M. 2007. Discordance between living and death assemblages as evidence for anthropogenic ecological change. Proceedings of the National Academy of Sciences of the United States of America 104, 17701–17706. (doi:10.1073/pnas.0707194104)

Kidwell, S. M. & Bosence, D. W. J. 1991. Taphonomy and time-averaging of marine shelly faunas. In Taphonomy: Releasing the Data Locked in the Fossil Record (eds Allison, P. A. & Briggs, D. E. G.). 115–209. Plenum Press. (ISBN:9780306438769)

Olszewski, T. D. & Kidwell, S. M. 2007. The preservational fidelity of evenness in molluscan death assemblages. Paleobiology 33, 1–23. (doi:10.1666/05059.1)

Schopf, T. J. M. 1978. Fossilization potential of an intertidal fauna: Friday Harbor, Washington. Paleobiology 4, 261–270. (http://www.jstor.org/stable/2400205)

¹Department of Geology and Geophysics, Yale University, New Haven, Connecticut 06520-8109, USA.

Ladies and gentlemen, I give you tree-kangaroos. These wonderful animals can, in myriad ways,  demonstrate the power of evolutionary biology and geology in explaining the patterns we see in modern ecosystems. Here, I want to show how palaeontologists can piece together multiple lines of evidence to understand the evolutionary relationships of fossil and living organisms.

Introduction

First, a little introduction to the tree-kangaroos (genus Dendrolagus). These small, tree-dwelling (‘arboreal’) marsupials live in the rainforests of Australia and New Guinea, and belong to the macropod family of animals, which also includes ground-dwelling kangaroos and wallabies. They grow up to about 80 centimetres long, not including the tail, and mainly eat vegetation (see Fig. 1). More than 10 living species are known.

Figure 1 — Tree-kangaroos. Top: left, lithograph of Bennett’s tree-kangaroo (Dendrolagus bennettianus); right, photograph of Matschie’s tree-kangaroo (Dendrolagus matschiei). (Images from Wikimedia Commons). Bottom: outlines showing a kangaroo (left) and tree-kangaroo (right), redrawn from Martin (2005).

Macropod means ‘large foot’, and like all members of this group, tree-kangaroos do indeed have big feet (see Fig. 2). They have no hallux, or digit, corresponding to the first digit in other mammals, and their fourth and fifth digits are small and fused together (‘syndactylous’) to form a grooming claw. The feet are relatively shorter and broader than in terrestrial kangaroos, and have a large, continuous foot pad covered in small bumps that help them to grip. The toes have large, curved claws, rather than flat nails; the ankle joint is modified to allow the foot to be rotated so that the soles of the feet face inwards (ideal for gripping braches); and the tibia (shin bone) is relatively short. Tree-kangaroo arms are proportionally larger and more robust than the arms of other kangaroos, and the hands have pads and claws similar to those on the feet. This combination of longer arms and shorter legs is more like the proportions of a non-hopping mammal than of a terrestrial kangaroo. The tail is not as heavily built as in the hopping kangaroos, and is not prehensile, yet it may be very long: up to 15% longer than the body in some species. It may be used for balance in the tree canopy.

Figure 2 — Drawings of the hind feet of Bennett’s tree-kangaroo (left) and Doria’s tree-kangaroo (Dendrolagus dorianus; right), redrawn from Martine (2005)

Tree-kangaroos are said to be the only kangaroos that can move their hind feet independently of each other (although other kangaroos can do this while swimming). They have some ability to hop on two feet, but they can also move by bounding on four feet and walking on two, both along branches and on the ground.

The tree-dwelling possums are the closest known relatives to kangaroos, so one might imagine that terrestrial kangaroos evolved from tree-kangaroos in a ‘down from the trees’ scenario. However, evidence from DNA, anatomy and the fossil record shows that they evolved relatively recently (in the past 5 million to 7 million years), and that their arboreal features have been acquired through modification of the hopping-adapted anatomy of terrestrial kangaroos.

The tree-kangaroo fossil record

Tree-kangaroos are poorly known from the fossil record. Some tooth fragments are known from the Hamilton Fauna, a group of rocks in Victoria, Australia, dating back to the Pliocene epoch, 5.3 million to 2.5 million years ago, in an area thought once to have been temperate rainforest. A few other remains are known from cave deposits in New Guinea from the late Pleistocene epoch (2.5 million to 0.01 million years ago). Other evidence of Australian tree-kangaroos comes from the late Pliocene and Pleistocene remains of the giant tree-kangaroo Bohra, which has a body resembling that of the tree-kangaroos, but a skull more like that of the rock-wallabies (genus Petrogale). (Rock-wallabies are the sister group to tree-kangaroos; see Fig. 3.) With a body mass between 30 and 40 kilograms, Bohra was around three times the size of the largest living tree-kangaroos, and was probably not as arboreal in its habits, especially given that it is known from areas of southern Australia that were relatively dry during the Pleistocene, such as the Nullabor Plain.

Figure 3 — Evolutionary relationships of the Macropoidea (kangaroos and potoroids). Modified from Meredith et al. (2009); position of the extinct taxon Bohra from Prideaux & Warburton (2010). Note that the anatomy-based phylogeny of Prideaux and Warburton is slightly different from the DNA-based one, but this does not affect the relative position of the tree-kangaroos.

Tree-kangaroos as an example of evolution

So, how can tree-kangaroos be used as a superb example of evolution in action?

1. What better example could we have of a living ‘transitional form’ than the tree-kangaroo? It is not well adapted for its environment, so it is difficult to imagine it having been ‘designed’. Tree-kangaroos are clumsy climbers compared with animals such as monkeys, or even with other marsupials such as possums. Not only is their behaviour clumsy, but their anatomy seems largely unsuited to the arboreal lifestyle. Unlike other largish tree-dwelling mammals, they neither have a prehensile tail that helps them climb (like possums or howler monkeys), nor lack a tail, thus avoiding the problems that one brings (like koalas, sloths and apes). Tree-kangaroos’ tails dangle and can sometimes get in the way of their climbing. Their ankle joints do not resemble those of specialized arboreal mammals, but have clearly been modified from the highly specialized ankles of hopping kangaroos.

2. What could be a better illustration than one transitional form? An entire series of them! There are close to a dozen species (depending on how you count some subspecies) of tree-kangaroos: two in Australia and the rest in New Guinea. The evolutionary relationships of these animals show that the earliest-evolved forms (the two Australian species and one of the New Guinea ones) that have the least-derived foot anatomy, whereas a couple of the New Guinea species seem to have been modified back again for a more terrestrial existence. This diversity of living forms, showing a range of progressive adaptations, is counter to the Victorian notion that one species must go extinct after the generation of a newer, ‘better-adapted’ form.

3. How does such a transitional form develop in the first place? When a creature starts to exploit a new habitat, with no suitable competitors. There were no monkeys in the tropical forests of Australia and New Guinea when the tree-kangaroos first encountered them. There are monkeys on nearby islands to the west of New Guinea, but a boundary called the Wallace Line separates these populations.

The Wallace Line is named after Charles Darwin’s colleague Alfred Russel Wallace who was the first person to notice a strange anomaly: New Guinea is closer to many Asian islands than it is to Australia, but the island contains more organisms with close relations in Australia than with relations in Asia. In addition, there is a strange mix of Australian and Asian organisms in the islands between New Guinea and Malaysia, a region now known as Wallacea. One species of tree-kangaroo is found on a couple of the smaller islands to the east of the Wallace Line, but no monkeys are found there.

Evidence from plate tectonics shows that until a few million years ago, New Guinea was further away from Asia, so that immigration of Asian species would have been difficult. During the ice ages over the past few million years, lowered sea levels would have allowed free interchange between New Guinea and Australia, but not between New Guinea and the Asian mainland (because the sea is shallow between New Guinea and Australia, which are basically part of the same land mass, but not between New Guinea and the Asian islands).

DNA data suggests that the tree-kangaroos split away from rock wallabies between 5 million and 7 million years ago, which is about the same time that tree-kangaroos first appear in the fossil record. This date is also a perfect match for evidence from plate tectonics that shows when the northern movements of the Australasian plate brought New Guinea and northern Australia into the tropical zone, establishing new types of rainforest environments that the tree-kangaroos moved into, without competition.

4. Why are there are so many species of tree-kangaroos in such a relatively small area? This is exactly what one would predict as a result of frequent breaking apart and rejoining of forest habitats over the past couple of million years, when ice ages in the higher latitudes were accompanied by drying and loss of forest habitat in the tropics. Such fragmentation in the African and South American tropical forests is well documented as the driver of increased speciation during the Pleistocene, and the number of tree-kangaroo species in New Guinea echoes the diversity of African monkeys and Amazonian tropical birds.

5. Science relies heavily on consilience: it tests whether a result is accurate by checking whether multiple lines of evidence lead to the same result. In that spirit, a type of evidence completely separate from the fossil record throws light on the evolution of tree-kangaroos: genetic studies show that the closest relatives of the tree-kangaroos are the rock-wallabies (genus Petrogale). The most arboreal rock-wallaby today is the Proserpine rock-wallaby, Petrogale persephone. Now, surprisingly (or perhaps not), the genetic evidence shows that the Proserpine rock-wallaby is the most basal of the Petrogale group of a dozen or so species of rock-wallabies, so it is closest to the common ancestor of the tree-kangaroos and rock wallabies.

It can be tempting to think that palaeontologists and biologists might simply look at what fossils are found where and when, and deduce evolutionary history from that. But I have detailed here several separate, independent lines of evidence about tree-kangaroo evolution. Issues such as the divide of the fauna and flora across the Wallace Line (with the absence of monkeys in New Guinea) and the makeshift arboreal anatomy of tree-kangaroos are observations in today’s world that remain entirely independent of any evolutionary interpretations. Yet these, and other facts such as the evidence from plate tectonics, genetic studies and fossils, all fit together into a coherent evolutionary perspective on the distribution and anatomy of tree-kangaroos.

Further Reading

Aplin, K. A., Pasveer, J. M. & Boles, W. E. 1999. Late Quaternary vertebrates from the Bird’s Head Peninsula, Irian Jaya, Indonesia, including descriptions of two previously unknown marsupial species. Records of the Western Australian Museum. Supplement 57, 351–387.

Eldridge, M. D. B. & Close, R. L. 1997. Chromosomes and evolution in rock-wallabies Petrogale (Marsupialia: Macropodidae). Australian Mammalogy 19, 123–135. doi: 10.1071/ZO9890351

Flannery, T. F., Rich, T. H., Turnbull, W. D. & Lundelius, E. L. 1992. The Macropodidae (Marsupialia) of the early Pliocene Hamilton local fauna, Victoria. Fieldiana: Geology, New Series 25, 1–37.

Martin, R. 2005. Tree-kangaroos of Australia and New Guinea. CSIRO Publishing. ISBN: 9780643090729

Meredith, R. W., Westerman, M. & Springer, M. S. 2009. A phylogeny and timescale for the living genera of kangaroos and kin (Macropodiformes: Marsupialia) based on nuclear DNA sequences. Australian Journal of Zoology 56, 395–410. doi:10.1071/ZO08044

Meredith, R. W., Westerman, M., Case. J. A. & Springer, M. S. 2008. A phylogeny and timescale for marsupial evolution based on sequences for five nuclear genes. Journal of Mammalian Evolution 15, 1–36. doi:10.1007/s10914-007-9062-6

Prideaux, G. J. & Warburton, N. M. 2010. An osteology-based appraisal of the phylogeny and evolution of kangaroos and wallabies (Macropodidae: Marsupialia). Zoological Journal of the Linnean Society 159, 954–987. doi:10.1111/j.1096-3642.2009.00607.x

Introduction:

The size and shape of an organism is the product of genetics and environment. It is the raw material on which the process of natural selection (survival of particular animals over others) acts, and so is of central interest in studies of the evolution of ancient forms of life for which DNA information is not available. Fossil morphology, or shape, is the basis of most palaeontological studies, be they describing new species or making deductions about the animal’s lifestyle. Phylogenetic studies, those that place species in groups depending on how closely they are related to each other, are based on the presence and absence of particular features. This works on the theory that the more closely related two animals are, the more features they are likely to have in common. Fossil morphology also plays a major role in informing palaeontologists about the ecology of an animal, because form often reflects function. Details of diet, habitat, the way animals moved and the forces that parts of the body could withstand can all be investigated by studying morphology.

Figure 1 — Top left: sharp, serrated, recurved teeth of a carnivorous dinosaur (Tyrannosaurus, source, copyright Andrew A. Skolnick); bottom left: the short, peg-like teeth of a herbivorous dinosaur (Camarasaurus, source). Right: different mammalian limb shapes for (left to right) grasping, running, flying and digging (source).

Connected changes in the morphology of different parts of a biological structure can reflect developmental or evolutionary mechanisms that act on these parts together. Groups of parts that are affected together are often referred to as modules, and they give clues as to the limitations or freedoms of the mechanism that produced the animal diversity.

Collecting Shape Data:

Traditionally, morphology has either been described qualitatively or measured by collecting data sets of lengths, ratios and angles. However, in recent years, rapid advances in technology, computer power and analytical methods have led to the increased use of digital-imaging methods. These include photography, photogrammetry, microscribing, laser scanning and computed tomography (CT) scanning. Such techniques vary a great deal in terms of cost and speed, and how suitable they are depends on the research question.

Photography: Two-dimensional (2D) morphology can easily be — and has traditionally been — captured by photographing objects from the same relative plane of view. It is cheap (compared with other digital-imaging methods) and requires little expensive equipment — just a camera and usually a tripod. Additional lighting can help to highlight morphological features, and most museum collections have a photography stand. The background of the image is often important, particularly when it is necessary to see the boundary of the specimen clearly. One way of creating a non-reflective dark background is by placing a piece of black velvet under the specimen. The angle of the specimen relative to the camera lens is also important if the goal is to make meaningful comparisons of the shape of several specimens. The angle can be altered using modelling clay or wedges of foam plastic to support the object.

Photogrammetry: Within the last few years, a method has been developed for creating three-dimensional (3D) images by taking multiple photographs of an object from various angles. As long as the photographs overlap enough, so that each pixel of the image has been registered from at least three different views, software can combine these photos using triangulation methods (using the focal length and depth of field of the image) to transform the 2D pixel information from individual photographs into a 3D virtual model.

Figure 2 — A 3D image of a trilobite fossil, imaged using photogrammetry (Source: Falkingham 2012)

Microscribing: The microscribe is an articulated arm ending in a point, or stylus, attached to a heavy base. 3D Cartesian coordinates (distances along the x, y and — in the case of 3D data — z axes) are recorded by placing the stylus on chosen points of a shape and pressing the foot pedal to register the coordinates in a text file or spreadsheet. It is important that the object is kept still during this process to avoid mistakes. This tool is most frequently used to collect shape data in geometric morphometric studies, which simplify the shape to a configuration of landmark points (see ‘Statistical comparison of shapes’, below).

Figure 3 — A microscribe (left) and laser scanner (right).

Laser scanner: Portable laser scanners can be used to capture surface morphology of biological objects. They create point clouds of the surface topology of an object. Usually the specimen being imaged is rotated on a stage in front of the scanner, allowing it to capture the surface from different angles — although this doesn’t work with the largest samples. A complete image is achieved by compiling these views. The scanner may also be equipped with a camera that takes a photograph with each rotation of the object. This image is then laid over the model created from the laser to create realistic colouring in three dimensions.

CT scanning: CT scans are created using a combination of many X-ray images. If an X-ray has passed through a dense substance, the signal will be weaker than if it has passed through one which is less dense. This difference in X-ray strength is detected by a sensor behind the object and used to create an image. The X-ray beam and detector in a CT scanner both move in a circle around the object as the X-ray images are collected, unlike in standard X-ray images, for which they remain in a fixed position. The views from different angles can be used to create a cross-sectional ‘slice’ through the object. Many of these slices are then combined through digital processing to form a 3D images. CT scanners are most frequently used to look at internal morphology, because they create volume images based on differing densities of the material being scanned. These are composed of 3D pixels, known as voxels. This method is by far the most data-intensive in terms of the size of image file produced.

Figure 4 — A CT scan of the skull of a dinosaur (Eoraptor, source). Scale bar 1cm.

Before digital imaging, 3D fossils that could not be removed from their surrounding rock matrix were investigated by physically cutting the specimen into multiple slices to produce a ‘book’ of cross-sections from which morphology could be studied. Software specifically designed for transforming this cross-sectional information into a 3D image has been developed, and can also be used to process CT data. The benefit of CT imaging, however, is that it can produce 3D images of fossils enclosed in matrix without destroying the specimens — as long as there is a difference between the density of the fossil and the density of its surrounding matrix. CT scanning is also far less time-consuming than sectioning.

CT scans are often used as the geometric model when performing a technique called finite element analysis (FEA), which takes a complex shape and divides it into many smaller, simpler shapes to measure mechanical properties of the structure.

Figure 5 — FEA of a dinosaur skull (Allosaurus) modelling the distribution of stresses during biting. Source: Rayfield, E. J., Norman, D. B., Horner, C. C., Horner, J. R., Smith, P. M., Thomason, J. J. & Upchurch, P. 2001. Cranial design and function in a large theropod dinosaur. Nature 409, 1033–1037. doi:10.1038/35059070

Statistical Comparison of Shapes:

Geometric morphometrics, a method for comparing complex biological structures, has gained much popularity in palaeontological research. It uses the Cartesian coordinates of ‘landmark’ points to describe a shape.

Choosing Landmarks: Landmarks fall into three categories according to how precise or homologous they are considered to be:

Type I landmarks: boundaries between tissues. In mammal skulls, for example, these can be suture junctions, where one bone touches another.
Type II landmarks: the peaks and troughs of curves, such as the outside edge of the eye socket or the frontmost point of the hole where the spine enters the skull.
Type III landmarks: the extreme end points of a biological structure, such as the very front and back of the skull.

Figure 6 — Different types of landmarks on a mammal skull. Grey, type I; pink, type II; blue, type III.

Of these, type I landmarks are the most similar between life forms and can be placed with the most precision. However, the more distantly related two species are, the fewer truly homologous landmarks there are likely to be. The use of type II or III landmarks might be necessary to ensure full coverage of particular morphological features. How much landmarks correspond between specimens has phylogenetic, structural, developmental and biomechanical significance. But judging landmark correspondence is subjective and depends on the research question, as does the number and distribution of landmarks.

There are three major limitations with analysing the placement of landmarks one by one in specific locations. First, the whole form is not being described, so important, interesting and useful features that occur between landmarks — such as spaces, curves or other surfaces — can be overlooked. Second, the reliance on correspondence between the landmarks focuses on similarities, making large differences, such as the presence or absence of a feature, essentially impossible to capture. This potentially leads to an underestimation of variation. Third, some regions of the skull (with, for example, clear suture junctions or obvious peaks and troughs of curves) might be more suited to landmarking than others (such as large, smoothly curved surfaces) so some parts of the morphology can be overrepresented in comparison with others.

Other ways of describing shape have been developed, most notably the analysis of outlines and curves along which landmarks are evenly spaced. These are beyond the scope of this article, but more information can be found in the suggested reading.

Isolating shape data: Once landmark data have been collected from the biological sample, it is necessary to remove all size and position information, to leave only shape information. Several methods for doing so have been devised and argued for, but the most favoured method in  literature from the last fifteen years has been Procrustes analysis. This gets its name from Procrustes, a morbid character of Greek mythology, known for inviting people to lie down in his bed, then adjusting them to fit into it, either by stretching them or cutting off their limbs. Procrustes analysis is the process of superimposing one collection of landmark configurations on another by scaling, rotating and translating them so that the distances between corresponding points in each configuration are as small as possible, making a ‘consensus shape’.

Figure 7 — Diagram representing the different stages in a Procrustes superimposition of two sets of landmarks. Source: Mitteroecker, P. & Gunz, P. 2009. Advances in geometric morphometrics. Evolutionary Biology 36, 235–247. doi:10.1007/s11692-009-9055-x

Analysing Shape: The new configurations of landmarks can be used to estimate standard statistical parameters. For example, the difference between any two shapes — their ‘disparity’ — can be calculated by finding the distance between the two sets of Procrustes landmark coordinates that represent those shapes. This is the Procrustes distance. Shape variance (the amount of variety in the shape data) can be calculated by adding together the squares of the distances between Procrustes landmarks and their equivalent landmarks on the consensus shape for all the landmark configurations in a sample. The total of the squared distances can then be divided by the number of configurations minus 1 to normalize it.

Often, when many specimens with many landmarks are being compared, the number of variables makes further interpretation of the results tricky, because multivariate space is impossible to visualize. Principal coordinates analysis (PCA) is a common statistical method that finds the major axes of variation through multivariate space. Figure 8 demonstrates PCA for just two variables (X1 and X2) for the sake of simplicity, but the idea is exactly the same when finding the principal coordinates (PCs) for data sets of many variables. Figure 8a shows the distribution of data points according to their X1 and X2 values, and Fig. 8b is a simplification of the shape of this distribution. PCA finds the axes of maximum variation through the data: PC1 is the longest axis and PC2 the second longest, at right angles to PC1 (Fig. 8c). The more variables there are in the data set, the more PCs there can be. Figure 8d shows how the original data is rotated to show variation according to the major axes of variation.

Figure 8 — A simplified graphical representation of what PCA does to data. Rearranged from: Zelditch, M. L., Swiderski, D. L., Sheets, H. D. & Fink, W. L. 2004. Geometric morphometrics for biologists: a primer. Elsevier Academic Press.

Geometric morphometric data can be visually simplified by plotting the specimens against the PCs that explain the most variation in the data set. The area of this graph represents shape space, and the points are particular shapes within that space. In Fig. 9, the shapes are skulls.

Figure 9 — Principal Components Analysis of extant and extinct metatherian skulls.

Summary:

This article is a short summary of the big and exciting world of shape analysis — much more detailed descriptions of imaging and analysis techniques can be found in the suggested reading. The shape of fossils is central to a wide range of palaeontological studies, and there are many ways of collecting and analysing such data, as shown in this article. However, each method has particular limitations that need to be weighed up, including cost, resolution, computing power, destructiveness, portability and amount of data. Choosing the most appropriate method depends heavily on the research question, the available funds (many of these techniques require expensive equipment or software) and the nature of the fossils. Palaeontological research has come a long way from the descriptions written by the fossil hunters of yore, and although thorough qualitative descriptions of specimens are still an important part of the science, modern palaeontology calls much more on cutting-edge technology from the realms of mathematics, physics and computer science.

Suggestions for further reading:

Adams, D. C., Rohlf, F. J. & Slice, D. E. 2004. Geometric morphometrics : Ten years of progress following the ‘revolution’. Italian Journal of Zoology 71, 5–16. doi:1080/11250000409356545

Falkingham, P. L. 2012. Acquisition of high resolution three-dimensional models using free, open-source, photogrammetric software. Palaeontologia Electronica 15, 15.1.1T.

Geodetic Systems: The Basics of Photogrammetry

Klingenberg, C. 2010. Evolution and development of shape: integrating quantitative approaches. Nature Reviews. Genetics 11, 623–635. doi: 10.1038/nrg2829

Mitteroecker, P., & Gunz, P. 2009. Advances in Geometric Morphometrics. Evolutionary Biology 36, 235–247. doi:10.1007/s11692-009-9055-x

Zelditch, M. L., Swiderski, D. L., Sheets, H. D. & Fink, W. L. 2004. Geometric morphometrics for biologists: a primer. Elsevier Academic Press.

¹Department of Genetics, Evolution and Environment, University College London, London WC1E 6BT, UK

Introduction:

In an article on Palaeontology [online] last year, Sarah King explained how undertaking a PhD can help you to launch an academic career in palaeontology. Obtaining that PhD can be a frustrating yet ultimately rewarding experience, but it is only the beginning for many palaeontologists — and it is worth pointing out that a PhD isn’t a prerequisite for certain jobs in palaeontology (for example, dealing fossils). Here, I hope to give you a sense of what might happen after the PhD, and how this could lead to a wide range of new challenges and take you down previously unimagined paths. You didn’t think that getting a job in palaeontology would be straightforward, did you?!

As a PhD (Doctor of Philosophy) student, you are dedicated to working on your doctoral thesis: a single piece of independent research and thinking. After several years of work (in Europe, usually three or four), which could require copious amounts of blood, sweat and tears, you should finally complete your thesis; after successfully defending it (in an examination that is called a viva in the United Kingdom), you will obtain a doctorate that qualifies you as an expert in the field. In the rest of this article, I will summarize my experiences as a ‘postdoc’ (a postdoctoral researcher) in palaeontology. For me, becoming a postdoc provided the best opportunity to carry out the research that I enjoy and am good at — but it wasn’t without its difficulties. Ultimately, I hope that through this postdoc work I will be able to obtain a permanent position in a research institution that will allow me to continue to develop my ideas and keep working with fossilized and living plants. As an adviser once said to me: “A happy worker is a productive worker, and a productive worker is happy!”

Figure 1 — Leyla in a field of orchids in Germany. Credit: L. Seyfullah.

Time for a real job:

After I had finished celebrating getting my PhD, my family asked two (im)pertinent questions. Firstly, hadn’t I completed my studies yet? Secondly, when was I going to get a ‘real’ job? I was shocked that people might think that I was planning to stay a student forever (no way, living on a student budget is not easy!) and that what I wanted to do for the rest of my life — stay in academia — was not seen as a real job. Perhaps this perception was because I am the first person in my extended family to get a PhD, and also because I hadn’t explained very well what being a postdoc means. To combat this, I will here outline my experiences after completing my PhD, as a postdoc in palaeontology. I am most certainly no longer a student, and I definitely feel as if I have a proper job — I even pay taxes!

First, survive the PhD…

I thought long and hard when deciding on my PhD subject; I knew I would have to devote several intense years to working on living and fossil plants, and this would dictate what I could do next and how I would be able to work in the future. As a result, I really enjoyed most of the experience, although it was not always what I was expecting. During the last two years of my PhD I had to begin thinking about what to do and where to go next, because proposals for research funding can take many months or even years to devise, write up and get assessed. I started developing projects to follow on from my PhD work, and I also started getting involved in other people’s applications for funding. I was very optimistic! I breathed a huge sigh of relief when I passed the defence of my PhD at the end of 2008, and thought: OK, everything is going well on the application front, but I wonder how this disaster in the world economy is going to unfold? Well, indirectly it was to have a massive impact on my future.

Figure 2 — Cypress twig preserved in amber. Credit: L. Seyfullah.

The way I saw my choice:

The path I have taken is fairly typical of palaeontologists who want to continue working as researchers at a university or other academic institution. Being a postdoc is strange, liberating and exciting. It is the period of an academic career where you are starting to establish yourself in your field by developing your own research directions and collaborations (‘going solo’), but do not yet have too many teaching or administrative responsibilities. There is a major downside, however: you have only a fixed-term position. The postdoc phase can be a period of immense freedom because these temporary contracts provide the opportunity to work in a range of different labs, meet new working groups and test out ideas that you didn’t have time for in your PhD. If you are a postdoc in someone else’s research group, the principal investigator (your boss — normally a permanent member of academic staff at your university or institution) is frequently the person who won the research funding, so the responsibility for the project does not rest on your shoulders, although you can personalize the work to make it your own if you wish. The flip-side is the precariousness of this system, particularly if you have family considerations to take into account. Are you going to drag your young family around the world for a reasonably well-paid but temporary job? Are you going to tough it out long-distance for an unknown number of years? What if it all doesn’t work out? What happens between contracts? It can be a difficult personal decision.

Intermission:

When the economy went south in 2007, funding opportunities began to dry up as the UK government evaluated the country’s debt and realized how little money was left for publicly funded agencies such as research councils and museums. Few research grants were stopped outright, but the competition for funding greatly increased — finding money for research became much harder across the academic world, not just for postdocs but at all levels. I was allowed to stay on at the University of Birmingham, UK, after my PhD to work up papers and ideas for grants as an honorary (unpaid) postdoc. The advantage of this was that I could make use of the cutting-edge equipment in the lab, the extensive library collections and the support of all the staff around me. I also worked part-time as an administrator, helped out on student fieldtrips, took classes and processed samples — anything to pay the bills! Truthfully, it was not a stable or fulfilling time. Trying to juggle administrative work with grant applications, job-hunting and learning to write better scientific papers, along with the weekly search for temporary hours in offices or other university departments, was grindingly depressing. On top of this, I was continually trying to find small sources of funds to tap into so that I could participate in international conferences, network with other researchers and try to get a job elsewhere. I knew that it wasn’t just me; plenty of friends and colleagues had lost their jobs or couldn’t find anything. So I kept at it, and got by on the support of my family, department, collaborators and the international palaeobotanical community. My luck changed when I appealed (desperately, at this point!) to contacts abroad, because there was little available in the United Kingdom. In the end, as with buses, two offers came along at once. From nothing, I suddenly had the choice either to work on somebody else’s project in Switzerland or — rather more appealingly — to work in Germany on an independent project that I had proposed with a researcher whom I had met only once previously.

A new hope:

Having written it based on my own interests, I was particularly keen to work on the independent project, and duly accepted this offer. The project involved studying fossils preserved in amber to fill in some of the gaps in our knowledge of seed-plant evolution and to reconstruct ‘amber forest’ ecosystems. I was rather nervous about moving to a new place, but overwhelmingly relieved to have secured a paid position. Finally someone wanted to invest in me and my project — even if it was in a country I had never considered working in before, and where I didn’t speak the language. So, I left my home and said ‘Auf Wiedersehen’ to my partner, pets and plants, and set off with my laptop and suitcase. I arrived in Göttingen, Germany, armed with a guidebook (a thoughtful leaving present) and a phrasebook (an even better leaving present). By the end of the first week, I had a room to live in, an office with a working computer and a bank account, chiefly through the efforts of a very helpful undergraduate student. By the end of the month, I had my first salary instalment and I could nearly afford to be excited, except that I knew I had to pay off debts and start saving because I was contracted for only a couple of years. Putting all that aside, I focused on trying to learn German at night classes (a disaster and rather inefficient) and getting settled in the lab (much easier).

Figure 3 — Wilhelmsplatz, the offical centre of the University of Göttingen. Credit: L. Seyfullah.

When in Rome (or Göttingen):

My first few months in Germany were incredibly hard, especially because I spoke no German. I felt rather alone at times: I knew nobody, and the university is rather old-fashioned and somewhat conservative, located in what can feel like the middle of nowhere (especially in winter). The city is quaint and has lovely timbered buildings, but I quickly discovered that those can be drafty in the winter (as can old British buildings, I’m sure!). For three months, I had immense difficulty understanding the complex domestic recycling system, the obscure university bureaucracy and the general etiquette about basic things such as shaking hands, when to address people using the polite form ‘Sie’ or the informal ‘du’ and what gender a noun has. I reminded myself that at least now I had my own project and a salary, so I should make the best of it. I laughed the first time someone said that the best thing about my new university was that it was just a few hours from great cities like Berlin, Hanover or Frankfurt. After the fifth or sixth person said it, I was getting desperate; cabin fever was setting in, so I bought myself a travel card and started to spend one weekend every month investigating the German train system and the fabled places that people were telling me about. I threw myself into more of the terrible language classes, spoke awful German–English hybrids and found myself a lovely flat and a very understanding German flatmate.

After months of stressful negotiations, my supervisor and I finalized plans for the fieldwork that formed the core of my project, and set off on one of the most memorable experiences I have ever had. I spent several months carrying out fieldwork in the Southern Hemisphere, in the southern spring — so I avoided a wet German autumn, which is pretty much like a wet British one — all paid for from my fellowship. I was part of a truly international team of experts. For the first time, when I was introduced as the expert botanist and palaeobotanist, I thought to myself: Yes, I can live with this, and this is what it has all been about!

Sticky questions:

Fieldwork allowed me to develop ideas and collect material with which I could start to answer questions about the relationship between resin (the ‘sticky stuff’ in some living seed plants) and amber, which is fossilized resin.

Figure 4 — Millipedes trapped in fresh resin. Credit: L. Seyfullah.

I was particularly interested in answering the following questions:

1. What does and does not get preserved in resin and so in the amber?
2. Under what conditions do modern plants start to produce excessive amounts of resin? Some plants produce resin in small amounts (think of sticky bits on pine cones, or small oozing tacky blobs when you cut a pine branch), but amber in the fossil record can be found in large amounts (as well as tiny isolated drops), so what happened to make the trees weep so much resin that then got preserved as amber? There are several hypotheses that might explain this, including the evolution of new wood-damaging insects, diseases or ecological disasters such as hurricanes flattening large portions of forests. My aim was to try to distinguish between these possible explanations.
3. How could this massive resin outpouring be preserved as amber in the fossil record? Answering this meant looking at the environment and ecology of living trees to see how long resin lasts and how it might get buried and preserved (or not).
4. Can we tell apart the different reasons for resin production in the fossil record using chemical tests on ambers and using the modern resin material as a guide?

I could not possibly hope to answer all of these questions in a short field season or even in one three-year postdoc position, but I wanted to at least work out how I could go about answering the questions in the future. The research team that I joined also wanted to learn about newly discovered New Zealand amber deposits, and I was particularly keen to study the fossil plants surrounding them. All this would help us to figure out how these fossils are related to the resin-rich trees that still inhabit the Southern Hemisphere. This subject closely integrates both living and fossilized seed plants, and is strongly linked with their evolutionary history. It is the sort of work that I have always wanted to do.

Figure 5 — Fossils preserved in amber. Credit: L. Seyfullah.

Getting my hands dirty:

I spent several weeks in New Caledonia looking at the behaviour, ecology (including associated insects and fungi) and environment of the most resin-rich (and some of the rarest and definitely stickiest) trees in the world today. New Caledonia is a small, very isolated French territory in the southwest Pacific Ocean: a botanical sub-tropical paradise. It is made up of one large island and several small ones (including the Loyalty Islands) and is basically a massive outdoor laboratory, and an excellent example of island biogeography. Some of the strangest relict plants from the ancient land mass of Gondwana survive there. It is also the centre of diversity of the plant family Araucariaceae, which is thought to be the parent plant of most of the conifer-derived ambers. Many of the plants are extremely endangered, so it was vitally important that I get legal permissions to observe, record and collect them.

Figure 6 — Fieldwork in New Caledonia. Credit: L. Seyfullah.

On the North Island in New Zealand, I collected fresh and sub-fossil resins from endemic resin-rich tree species so that I could start to understand fossilization processes in resins. I also collected amber from lignite mines in the South Island. It was a messy, sticky, but amazing time, and I am incredibly lucky to have been able to go and do this as part of my job. A postdoc is not, however, just about fun in the field, and after the fieldwork the ‘real’ work of chemical analyses must begin; so, back to the lab in Germany.

Figure 7 — Fieldwork in New Caledonia. Credit: L. Seyfullah.

How do German universities work for academics:

I am part of a cross-disciplinary research group in Göttingen. This ‘geobiology’ group, which encompasses a range of fields such as molecular biology, evo-devo, microbiology, geology and palaeobotany, was founded with start-up money from the German Research Foundation specifically to promote the establishment of new groups that sit outside mainstream research fields. The foundation’s rationale was that continued fundamental reform of the German university structure is needed to bring in young, foreign scientists to work in new research areas that could not get a foothold under the traditional system. I have a 100% research contract, so I am not required to teach, but I have given some lectures and seminars in English. German is the preferred undergraduate teaching language, but some higher-level science courses are taught in English. Specifically, we teach palaeontology, geology and molecular biology to biology and geology students. I also demonstrate how to use microscopes and other useful but delicate (and expensive) machinery. I have supervised several student projects (in English), and will have even more this year (word must have got out that writing your thesis in English with the assistance of a native English speaker is a good thing for your career!).

During my day-to-day job, I get to work on fossils with amazingly helpful collaborators from all over the world, and also teach students about conifers, resins and ambers — all in my compact but well-equipped lab and office. This is more or less the best any academic palaeontologist could ask for, and it is just a shame it will not last for longer — but, anyway, on to the next challenge!

Where and what next?

This is the postdoc’s perennial question. I still miss my partner, pets and plants, but I have a system worked out with my host that allows me to come back to the United Kingdom every so often to get my fix of home — and to get a good curry! I do feel torn about leaving my partner in the United Kingdom, but there is really little other choice at the moment. My host and I have learned a lot from one another, and we have developed a great research group and ways of working. I might never find a good curry in Germany, but I can have fun there now and feel fairly comfortable. I am happy to stay in Germany, but like all postdoc positions, it is not permanent. So, I keep writing applications and looking for other posts. I would like to come home eventually, and set up my own research group in the United Kingdom; however, part of me is happy to look abroad, because I now realize that it can open up possibilities that I had not previously envisaged. Germany would not have been my first choice for a job after my PhD, but actually I am now quite fond of much of my German experience. I believe that part of this is luck, but much more is about what you make of it. Fingers crossed that plenty more papers and grant-writing lie ahead.

¹ Georg-August-Universität Göttingen, Courant Research Centre Geobiology, Göttingen, Germany

Introduction:

The results of scientific research can be of interest to experts and non-experts alike. This is perhaps especially true for palaeontology, which captures public interest — but obtaining access to this information is sometimes difficult, even for scientists. Taking a rather different tack from previous Palaeontology [online] articles, I’m going to provide a brief overview of how the Internet has changed and is significantly changing palaeontology and academia in general, helping to open up research for the greater benefit of science and society.

Figure 1 — Sir Tim Berners-Lee sends a message at the London 2012 Olympics.

When Sir Tim Berners-Lee helped to invent the World Wide Web more than 20 years ago, he did it ‘for everyone’ (see Fig. 1), and to this day he still campaigns to maintain open web standards. Had he patented his technology, or asserted restrictions on it, the world would certainly be a very different place and you probably wouldn’t be reading this website on a freely accessible network. The very fact that it is open and royalty-free is part and parcel of the tremendous success of the Internet.

For professional and amateur palaeontologists, access to information on the Internet is extremely valuable. Being able to see pictures of specimens, data from measurements and 3D scans saves significant effort being otherwise duplicated – not to mention time and travel costs given that fossils are scattered all over the world.

It is important to highlight the difference between free and Open. If digital content is ‘free’ to access, as is, for example, the website of the newspaper The Guardian, it can be viewed without paying money, but the viewer does not get any further legal permissions unless specifically stated. By contrast, with Open content or data, “anyone is free to use, reuse, and redistribute it — subject only, at most, to the requirement to attribute and/or share-alike” (see opendefinition.org).

The vast majority of Open content on the Internet, including Palaeontology [online], is now provided under a particular set of legal licences created by Creative Commons. These allow authors of digital works to clearly and simply make explicit the ways in which they allow their work to be shared, reused, remixed and redistributed. There are only three Open Creative Commons licences, each denoted by its own logo:

Creative Commons Zero (CCO)

The person who associated a work with this deed has dedicated the work to the public domain by waiving all of his or her rights to the work worldwide under copyright law, including all related and neighbouring rights, to the extent allowed by law. You can copy, modify, distribute and perform the work, even for commercial purposes, all without asking permission.

Creative Commons Attribution (CC BY, or CC-BY)

If you alter, transform, or build upon this work, you must attribute the work in the manner specified by the author or licensor (but not in any way that suggests that they endorse you or your use of the work).

Creative Commons Attribution Share-Alike (CC BY-SA)

If you alter, transform, or build upon this work, you may distribute the resulting work only under the same license to this one AND you must attribute the work in the manner specified by the author or licensor (but not in any way that suggests that they endorse you or your use of the work).

The most recent authoritative survey, in December 2010, estimated that there were more than 160 million different works on the Internet that used one of these three licences. Some websites specifically track the usage of CC licences; for example, more than 59 million photos have been made Open on the photo-sharing website Flickr.

Science has also been hugely transformed in the past decade by digital openness (see Video 1).

Creative Commons licensing has been an integral part of the success of open-access publishing under the CC BY license. “CC has provided a strong, consistent signal that you can use openly published research to do with what you want,” says Mark Patterson, director of publishing at the open-access publisher Public Library of Science (PLoS). “Because CC licenses are created by experts and have a solid legal foundation, they have become the gold standard in open access publishing” (adapted from The Power of Open 2011, made available under a Creative Commons Attribution License CC BY licence).

Open Access:

If you are unfamiliar with or unsure of what Open Access means in relation to scholarly works, then watch Video 2. It’s truly brilliant. It not only succinctly defines the issue, but also conveniently provides much of the historical background, politics, facts and figures.

I find it hugely embarrassing to tell you that even academics at institutions with very good libraries don’t have access to everything they need. Academic mailing lists such as DML are awash with requests from researchers looking for PDF files of academic papers, as are social-media networks such as Facebook and Twitter. Most of the time, these requests aren’t for ‘difficult to obtain’ ancient literature, but rather for twentieth or twenty-first century papers that have to be paid for. I won’t name names, but I vividly remember an example on Twitter recently, when a researcher at the University of Oxford asked if anyone had access to a particular paper. I, a researcher at the University of Bath, tried and failed; a researcher from the University of Cambridge tried and failed; so did a researcher from Bristol University. Only many hours later did a researcher from Royal Holloway, University of London, manage to get access to this cursed paper. Access to scientific literature is a hugely inefficient ‘postcode lottery’ and disadvantages academics at less wealthy institutions.

No wonder, then, that academics who choose to make their works Open Access receive more views, downloads and citations than those who publish in subscription-only journals, as has been shown in many different peer-reviewed studies (see suggestions for further reading).

Despite this, the majority of academic research is still made available on the Internet only behind a paywall. This applies as much to palaeontology as to any other discipline. For every Open Access palaeontology-publishing journal such as Acta Palaeontologia Polonica, Palaeontologia Electronica, PLoS ONE and ZooKeys, there are scores more profit-making subscription-access journals.

Much of the research in these journals is ultimately paid for by the public, through tax-fuelled government funding or charities. Yet for decades, subscription-access publishers have been monopolizing this public resource and charging increasingly exorbitant fees to access it (see Video 3).

With Open Access journals, everyone can gain the benefit and pleasure of reading every last word for free, forever. With subscription-access journals, unless you have a paid subscription or are affiliated with an institution that subscribes, it is likely that you will be denied access to most articles. But Open Access isn’t just about reading; crucially, it also enables royalty-free reuse. Teachers and educators can print off all or portions of as many Open Access papers as they wish without incurring legal costs. Subscription-access science effectively bars itself from classrooms and other educational settings — copy a 2-page paper for 50 students to read and it could cost you US$4,905! Even single-access, single-paper charges can be outrageous: who would pay US$113 plus tax to read a 5-page paper?

The situation in biomedical research is seriously unethical, and has had documented negative impacts on public health, although in recent years levels of access have improved. I would argue that the prevalence of paywall-restricted access also harms palaeontology.

It is hugely naive to assume that only academics benefit from access to academic works (as documented at WhoNeedsAccess?). Palaeontology has a long and undeniable history of ‘amateur academics’ making brilliant scientific contributions. For example, the nineteenth-century fossil collector Mary Anning not only found many outstanding fossils, but also contributed to the understanding of their anatomy: she was the first person to correctly interpret some of the strange objects that she had found as coprolites (see Fig. 2). There are thousands, if not millions, of potential amateur palaeontologists in the world, whose enthusiasm lies untapped and whose knowledge remains underdeveloped — and the situation is exacerbated because most of the world’s scientific literature is hidden behind paywalls on the Internet.

Figure 2 — Mary Anning knew what these were: fossil faeces! © electrobleme.

Examples of valuable ‘amateur’ contributions to science enabled and emboldened by Open Access include school students such as Jack Andraka, who this year discovered a novel method of cancer detection aided in part by access “to free online scientific research”. I believe that we would have more Andrakas in this world if cutting-edge science wasn’t so actively restricted.

A few academics (mostly more senior ones in my experience) seem fearful of Open Access; it represents a change in the way things are done. Academia is notorious for its conservatism. Many years ago, there were all sorts of strange criticisms levelled at Open Access — for example, that peer-review standards are lower in Open Access journals — but most if not all of these have been proven wrong over time. I refer interested readers to the Wikipedia section on that topic for more. It is of historical interest only.

There are many different and valid arguments for Open Access; in my opinion, the arguments work best when combined. Open Access provides many benefits so I am overjoyed that the United Kingdom has mandated that from 1 April 2013, all its publicly funded research output must be made Open Access. There are concerns over the cost, but when weighed against the global benefits, I am sure that it will be worth it. I could write my entire article on Open Access alone — some people have written whole books on the subject — but I think that it is worth moving on to other areas of digital openness in academia.

Open Data:

This year, the Open Knowledge Foundation granted me a Panton Fellowship to promote the Open Data movement, which is starting to gain momentum in science – as good scientific practice requires independent analysis and access to data. The Panton Principles were launched in 2010 as a guide for how scientists should publish their data online (see Video 4).

Open principles are even more important for data than for publication of results, because one of the major values of data is in its reuse, not mere presentation. The permission to reuse, reanalyse, resample and remix scientific data is vital to science. The trouble is that when research data does get shared online alongside an associated academic article, it is often in rather rudimentary ‘supplementary information’ files, which are not fit for purpose. They are often in formats such as Word documents and PDFs, which can’t be used for reanalysis: no statistical analysis software ever created accepts PDF files as an input, yet for the convenience of publishers this is how the supporting data for innumerable papers is made available online.

This regrettable oversight (which, to give them their due, many publishers are now fully aware of and working to change) creates a barrier to reuse that can prevent independent verification of the results and analyses presented in scientific papers, or make them much harder. Much of my PhD research involves trying to reuse and reanalyse previously published data in novel ways to gain new insight about the evolution of fossil animal groups. I have learnt from painful experience the inefficiencies how research data has been made available. I have tried to raise awareness of the problems and have helped to research and write papers encouraging best practice, but there is much still to do.

Independent verification is important:  it can uncover major problems with the original analysis. The peer-reviewed scientific literature is almost certainly hiding tens of thousands of errors and examples of misconduct.

Quote from Colin Macilwain published in Nature

Making data more immediately reusable and openly licensed would improve transparency and remove some of the barriers to independent reanalysis, for the greater good of science. At least one major publisher is now proposing Open Data as the default standard for data supporting a scientific publication. Exceptions from the Open norm could of course be made for justifiably sensitive data such as patient medical data, or information on the locations of rare animals, plants or fossils but the point is – these have to be justified as sensitive before they can be excepted from the (Open) norm.

Most reputable evolutionary-biology journals now require that the data behind research papers be archived on the Internet in repositories such as Dryad, MorphoBank and Figshare, to ensure unhindered and quick access for all readers. Palaeontology journals, on the whole, haven’t yet made this change, although the Journal of Vertebrate Paleontology did say in a 2011 editorial that it was “considering following other key journals in instituting a policy requiring that data supporting results presented in a publication be archived in a public repository”.

One might think that by politely emailing the authors of studies, one would be able to gain access to their research data. Alas, in my own and others’ experience, this system doesn’t always work well. Authors can be away on holiday, in the field or just plain unwilling to share (incidentally, willingness to share data is correlated with the quality of the results - ‘dodgy’ or ‘weak’ data are less likely to be shared), and email addresses change as authors move between institutions, making it hard to even contact them in the first place. Of course, authors can even be deceased — sometimes making it impossible to obtain the underlying research data from a published study. Emailing the author takes away precious time from both parties, and doesn’t work well if one is performing a meta-analysis and needs data from many different studies. For all these reasons and more, it is far more transparent, efficient and sensible to share research data as Open Data in a public archive for the benefit of everyone.

Open Source code:

Research presented in a publication can be broken down into three parts: the paper describing the work done; the data; and the statistical methods and manipulations applied to the data to provide the results and the basis of the conclusion. I have already talked about Open Access in relation to the first part and Open Data in relation to the second, but this third part is often the most perniciously non-Open.

Modern research is often very complex — many of the simple observations and discoveries have been made already. We need to use complex techniques and methods to make further groundbreaking insights. This is often done by computer, either in generalist computing environments such as R or MATLAB, or in specialist programs such as TNT, PAUP* or SPIERS. Yet many of these programs are not Open Source — it is not possible to check what the code that runs them is doing, so scientists have to take it on trust (sometimes with erroneous results).

Thus, in 2011, the Science Code Manifesto was released. It notes:

“Software is a cornerstone of science. Without software, twenty-first century science would be impossible. Without better software, science cannot progress.

But the culture and institutions of science have not yet adjusted to this reality. We need to reform them to address this challenge, by adopting these five principles:

Code All source code written specifically to process data for a published paper must be available to the reviewers and readers of the paper.

Copyright The copyright ownership and license of any released source code must be clearly stated.

Citation Researchers who use or adapt science source code in their research must credit the code’s creators in resulting publications.

Credit Software contributions must be included in systems of scientific assessment, credit, and recognition.

Curation Source code must remain available, linked to related materials, for the useful lifetime of the publication.”

To date, nearly 1,000 people have endorsed this manifesto and is still open for signatures. Regardless of numbers, the principles of this manifesto are vitally important to help to avoid controversies around the public perception and transparency of science.

Open Educational Resources and MOOCs:

In this last section, I would like to mention perhaps the most exciting impact of openness and the Internet: the ability to educate humanity en masse for a mere fraction of the cost of ‘traditional’ educational courses. Open educational resources (OERs) are freely accessible, openly formatted and openly licensed documents and media that are useful for teaching, learning, education, assessment and research purposes (see Video 5).

Aggregator sites such as OERcommons list more than 50,000 such resources, of which more than 20,000 are ‘Science & Technology’ related. This corpus is growing day by day. As well as classifying content by subject, it is also lists content by school age suitability (Primary, Secondary, Post-secondary) and type (Audio Lecture, Full Course, Lecture Notes, Activities). Examples of high-quality openly available content can be seen below.

Climate Literacy and Energy Awareness Network.

The Science Education Resource Center at Carleton College.

MIT OpenCourseWare.

UK universities are getting in on the act, with OER repositories of Open content now available online from the University of Oxford, the University of Leicester, University College London, the University of Bath, the University of Nottingham and more.

Audio and video resources and lecture slides are fantastic rich content to share openly. It takes a long time to prepare a good, high-quality lecture course, and sharing it online for others to use, adapt and remix could result in a lot of time saved. In addition, it could ensure that top-quality material can be delivered where it is needed most, for example to less-privileged institutions, or even at home as part of Open University-style distance-learning courses.

To put the emerging innovation of free, open online courses in a financial context:

• Americans collectively owe US$914 billion in student loans (source: Federal Reserve Bank of New York). Average tuition costs alone are $28,500 per year. Despite this, in autumn 2012, a record 21.6 million students are estimated to be attending US colleges and universities, an increase of about 6.2 million since autumn 2000.
• In the United Kingdom, many fees for domestic undergraduate courses are now set at £9,000 per year. At the moment, UK universities and colleges have only 2.5 million enrolled students, and these numbers seem to be relatively stable.

So there is serious money to be made in the higher-education market. OER-style online delivery of undergraduate courses has been touted as both a disruptive opportunity and a threat to some brick and mortar universities.

Online courses called MOOCs (massively online open courses) are attracting the interest of both for-profit and non-profit groups. Many aren’t strictly Open, but they do tend to have at least partly free content and do not require enrolment at a traditional university. Examples are Udacity (for-profit), Udemy (for-profit), Coursera (for-profit), EdX (non-profit) and Peer to Peer University (P2PU, non-profit).

For the moment, this content tends to be restricted to computer-heavy subjects such as maths, physics and computer science, which lend themselves to online teaching. However, there are courses coming online for a broader variety of subjects every week. Coursera has more than 1.9 million enrolled students this year, although this figure cannot be directly compared with traditional courses because MOOCs have high dropout and non-completion rates, with people tending to dip in and out quite freely. Traditional courses, with their high fees, are less lightly started and discarded!

With respect to palaeontology, it is obvious that online-only delivery couldn’t provide everything required — field trips and study of specimens are conceivably much richer experiences in person than in virtual learning environments. However, I think that before too long, we will see a full MOOC offered for highly theoretical palaeontological subjects such as phylogenetics and statistical analyses. The American Museum of Natural History in New York already offers MOOC-like courses, but they are not free – $495 + $25 registration fee if you’re interested, hence they’re neither MOOCs nor OERs.

Conclusions:

The Internet has forever changed the way academia operates — even palaeontology!

• Open Access has been mandated in the United Kingdom, with Europe soon to follow (not to mention that Latin America already largely embraces Open Access, and that Australia is getting there too).
• Open Data is beginning to be recognized as similarly important, with many funders such as the US National Science Foundation, the UK Biotechnology and Biological Sciences Research Council and the UK Natural Environment Research Council now actively tightening up their data-sharing rules.
• Open-source code and fully reproducible science are next on the agenda for wide-scale change going into 2013.
• Huge investment is currently being made to further develop OERs and MOOCs to enable higher education online for everyone. Very early days here, but keep watching.

My take on it is that we should all be grateful to Creative Commons for legally enabling free and open enterprise without the hassle of restrictive copyright terms that would prevent much societal good. The organization is especially brilliant considering its small size — even now, at perhaps its largest, it has only 21 full-time staff members. The power of Open should not be underestimated!

Suggestions for further reading:

On Creative Commons:

Creative Commons is nearing its tenth birthday (7–16 December) and has just elected a new scientific advisory board. This will ensure that it continues to serve and enable the needs of science as best it can.

On Open Access:

Gargouri, Y., Hajjem, C., Larivière, V., Gingras, Y., Carr, L., Brody, T. & Harnad, S. 2010. Self-Selected or Mandated, Open Access Increases Citation Impact for Higher Quality Research. PLoS ONE 5: e13636. doi:10.1371/journal.pone.0013636

Mike Taylor has an excellent set of tutorials explaining terms and concepts in Open Access starting here.

On Open Data:

The Open Knowledge Foundation has an excellent post that explains Open Data in the wider context, it’s much bigger than just academia  – financial, transport, environment, weather and cultural data too.

On Open Source Software:

Ince, D. C., Hatton, L., and Graham-Cumming, J. 2012. The case for open computer programs. Nature 482: 485–488. doi:10.1038/nature10836 — If you can’t access this paper, please drop me an email.

¹ PhD Candidate and Open Knowledge Foundation Panton Fellow | Biology and Biochemistry Department, University of Bath, Claverton Down, Bath, BA2 7PY, UK.

Introduction:

Breathe in. Breathe out. It’s a good bet that you’re currently sitting in front of a computer, reading; I’m going to go ahead and assume that you’re breathing, too. In, and out. You probably weren’t even thinking about breathing until I mentioned it, but all the same, it’s keeping you alive. Oxygen from the air is being transported into the cells of your body, which are using it to create energy. So far, so good. But what you may not realize is that the cellular machinery performing this process so integral to our existence (Fig. 1) has roots buried deep in the geological past. It’s a story that begins before the origin of organized cells, in an ancient, alien world. But if we’re going back that far, we might as well go all the way back, to the very beginning. After all, to be breathing, you have to be alive. How did that happen? How do we define ‘being alive’? Without further ado, let’s find out. Breathe in. Breathe out. And back to the origin of life.

Figure 1 — Two mitochondria, the structures within animal cells that are responsible for producing the cellular energy source ATP in the presence of oxygen. These two are from a mammal’s lung.

Origins:

In the beginning: The building blocks of life — as we know it, at least — are liquid water (a wonderful solvent that remains liquid at a large range of temperatures) and organic polymers that provide function and structure. Before we can have life, we need these raw materials. Water comes, in part, from Earth’s mantle, which in its early history would have contained lots of ‘hydrated’ minerals — those with molecules of water as part of their crystal structure. It’s a surprisingly soggy place. Water would have escaped from the mantle through volcanic eruptions, and got into the atmosphere. Another likely source of water is icy asteroids and comets that stuck the planet. The organic compounds on Earth originated both from Earth-based syntheses — in which they were made by elements reacting in the atmosphere — and from space. Recent research has shown that interplanetary dust particles, comets, asteroids and meteorites are all rich in organic compounds. These include amino acids (the building blocks of proteins) and nucleobases, which are integral to DNA.

Timing: Our little blue dot started forming 4.54 billion years ago (4.54 Ga) — that’s 4,540 million years. During accretion, the temperature would have been too high for water to have existed as a liquid, but current geochemical evidence suggests that from 4.4 to 4.0 Ga, extensive oceans could have existed for long periods, allowing simple organic compounds to accumulate. Perhaps it was during this quiet period that key steps in the origin of life occurred. An interplanetary stick in the spokes occurred at around 3.9 Ga: a period lasting between 20 million and 200 million years, called the late heavy bombardment (Fig. 2). During this period, material from space battered Earth, and could have killed any life that existed (we know about this event from craters on the Moon; no rocks on Earth are this old, thanks to the effects of plate tectonics). However, computer models suggest that the late heavy bombardment is unlikely to have sterilized Earth completely, so it remains a distinct possibility that life originated before 3.9 Ga.

Figure 2 — A diagram showing the number of impacts from space on an early Earth.

Abiognesis: If we’re looking for the origins of life, first we have to define it, which is where… ahem… life becomes tricky. At their most basic level, living organisms show self-sustaining biological processes, and the ability to replicate imperfectly (that imperfection providing the raw material for natural selection). It’s not a perfect definition, because there is something of a grey area between living and non-living — for example viruses, which replicate but can’t do many of the other things associated with life on their own — but it will have to do. Defining the point at which we can consider early molecules or systems life is similarly tricky, so we’ll just gloss over that and recognize that at some point, abiotic (non-living) chemistry must have acquired the characteristics of living systems, probably in a series of steps. This marks the origin of life: an event called abiogenesis.

We’ll cover here just two of the theories currently vying for attention regarding how abiogenesis may have occurred. The first — known as the prebiotic soup hypothesis — posits that life began in a relatively cool aquatic environment. Organic compounds would have accumulated in primordial oceans, and could have been concentrated by freezing or evaporation of the water. Further reactions could have led to increasingly complex molecules, including small polymers. All that would be required from that point is for one molecule, by chance, to develop the ability to catalyse its own replication, and an evolutionary cascade could begin. These molecules would become more and more abundant, and natural selection could mediate their changes.

An alternative model, the metabolist hypothesis, posits a hot and volcanic origin for life. This suggests that a self-sustaining chain of reactions could have evolved first, close to mineral-rich hydrothermal systems near the ocean floor (Fig. 3). If this was the case, the first ‘life’ would not have possessed informational molecules. Once the reactions had increased in complexity, though, these genetic molecules would be needed for modern biochemistry to develop. The two hypotheses aren’t entirely mutually exclusive. For example, self-sustaining reaction chains could have caused the prebiotic soup to become enriched in hard-to-synthesize or unstable molecules.

Figure 3 — Examples of hydrothermal vents on the ocean floor, such as that posited in the metabolist hypothesis. Left: a typical black smoker. Middle: a degassing event with bubbles of carbon dioxide; yellow sulphur is visible on the ocean floor. Top right: dendritic (branching) carbonate mineral growths, which develop when hot mineral-rich fluids hit colder water. Bottom right: a smoky plume found in the same location as the degassing event (middle) near the Northern Mariana Islands.

Early evolution:

So, that’s all very exciting: the first step in our, and everything else’s, evolution. By this point we’re probably somewhere before 3.8 Ga, and have all the ingredients for life: molecular entities capable of multiplication, heredity and variation. Good stuff. I hope you’re still reading and, for that matter, breathing. Again, I’ll assume you are and we can move swiftly on, because quite a lot happened over the next billion years.

Cells: Life as we know it involves cells: membrane-bound packets of life (again, we’re ignoring those pesky viruses!). Early genetic molecules probably needed some protection from the vagaries of the early oceans, and this may have come in the form of globules of fatty acids (Fig. 4). These long molecules are special in that they have a water-loving (hydrophilic) head, coupled with a water-hating (hydrophobic) tail. Because of this, when in water they can spontaneously join together to form balls with the hydrophilic end outside and the hydrophobic end inside, which can grow and divide, while retaining a portion of their contents. This may well be the origin of cell membranes.

Figure 4 — The probable origin of cell membranes: a vesicle. These small globules naturally form from fatty acids in water.

Early genetic molecules: Life as we know it relies on genetic molecules — long compounds found in cells, which store the information needed for life. In modern cells this is DNA. From DNA, proteins can be created by the molecule RNA, through a process known as protein synthesis (more information). DNA is, however, a horribly complex compound, and it is very unlikely that this was the first genetic molecule. We’re fairly sure that before DNA, RNA was the genetic molecule of choice. There could also have been a precursor to RNA — some form of polymerized self-replicating molecule, with the capacity to store and pass on information. The nucleobases and the backbone of any early genetic molecules may have been different from DNA and RNA.

RNA: RNA (Fig. 5) is an all-in-one molecule: it can both store information and catalyse reactions. When it acts as a catalyst we call it a ribozyme, and it has the capacity to carry out a wide range of important biochemical reactions that early life may have needed to survive. For example, it’s likely that by the time RNA-based life was established, there was no longer a ready supply of non-biological organic compounds. Because there were no raw materials to sustain life, simple metabolic-like pathways are likely to have appeared to provide the components needed for life. It is during this RNA world that protein synthesis may have become established: four of the basic reactions involved in protein biosynthesis are catalysed by ribozymes. It is possible that viruses are a hangover from an RNA world (although this is quite a can of worms and we probably shouldn’t open it here). Viruses don’t have cells — they hijack other cells’ molecular machinery for their own nefarious ends — but they are large RNA molecules.

Figure 5 — A comparison of the genetic molecules RNA and DNA. RNA is used during protein synthesis; DNA is stores the genetic information of an organism. Both have four bases and a molecular backbone, but RNA is often single-stranded, whereas DNA is double-stranded. Image credits.

DNA and protein world: In autumn — especially if you happen to be in regular contact with university students (trust me on this) — people start coming down with colds and flu. We don’t build immunity to these minor ailments because they mutate very quickly. That is because they are RNA-based, and RNA is relatively unstable compared to DNA. Thus RNA-based agents mutate quickly, which is not ideal for healthy self-replication (or indeed our autumnal/hibernal health). At some point, ribozymes that could catalyse the polymerization of DNA (Fig. 5) must have arisen, and genetic information was transferred to DNA — a much more stable molecule. This enhanced stability would have allowed molecules to get longer and store more genetic information, and to reproduce without as many mistakes. All of this would eventually have allowed more complex organisms to evolve. RNA would then have been demoted to its current role as a messenger and transcriber of DNA.

Milestones:

LUCA: Early in the history of life, probably before 2.5 Ga, the last universal common ancestor (LUCA) of all extant organisms lived. The tree of life seems to be rooted in hyperthermophilic organisms (those adapted to high temperatures), so it has usually been assumed that LUCA was just such a specialist. However, there is currently little conclusive proof of this — indeed, researchers remain uncertain that the tree of life is truly rooted here. Furthermore, it seems that the proteins used by these specialists are heat-adapted versions of those found in other organisms, so they probably didn’t arise first in a heat-loving organism. Thus — beyond the fact that it was probably some form of bacteria-like micro-organism that used DNA as its genetic molecules, we have relatively little idea what LUCA was like.

Prokaryotes: Life as we know it is split into two groups with different basic cell structures. One is the ‘prokaryotes’ — comprising the Archaea and Bacteria, which look similar (Fig. 6). They tend to be smaller than 10 micrometres (one micrometre is one-millionth of a metre) in size, and have no nucleus or internal membrane-bound structures. Their DNA is a single loop sitting freely inside the cell. Generally, Archaea and Bacteria are unicellular, reproduce by simple (asexual) splitting, or fission, and use horizontal gene transfer for genetic recombination. They obtain energy by a wide variety of means, meaning that they can ‘breathe’ all kinds of elements, from hydrogen sulphide to iron. Although the Archaea and Bacteria are superficially similar, they possess very different biochemistry, suggesting that they split fairly early in the history of life (see also Fig. 8 for a tree).

Figure 6 — The basic structure of a ‘prokaryote’ cell, such as a bacterium or archaean. There is actually notable morphological diversity with these single-celled organisms.

Eukaryotes: Organisms with more complex cells (such as fungi, plants, animals and amoebae), belong to a second group, known as the Eukaryota. Their cells tend to be larger (10–100 micrometres) and they possess organelles (Fig. 7) — membrane-bound structures in the cytoplasm (interior) of the cell. For example, mitochondria process oxygen to provide the cell with energy, and in some organisms chloroplasts are responsible for photosynthesis. The nucleus houses the DNA, which is found in long molecules that form chromosomes, and organisms are often multicellular with differentiated cells doing specialized jobs. Cell reproduction occurs by mitosis, with meiosis for sexual reproduction, which is the norm for genetic recombination.

Figure 7 — Examples of plant and animal eukaryote cells. A number of other organelles are omitted, but the difference between these and prokaryote cells is clear.

Endosymbiosis: One of the key theories for the origin of the complex structures in the cells of eukaryotes is an idea called endosymbiosis. It’s all kinds of awesome. Central to this idea is that the organelles come from the long-term cooperation, or symbiosis, of two prokaryotes. The earliest internal structure, and by far the hardest to tie down, is the nucleus. This could be the result of an endosymbiotic relationship, or it could have evolved without this process (an autogenous origin). However, the story for mitochondria and organelles such as chloroplasts (a type of plastid) has far less ambiguity. Mitochondria retain portions of their own DNA, and were originally bacteria with the ability to respire oxygen. At some point — perhaps owing to a failed attempt at predation — they started to live inside a larger organism. Current theories are split over whether this was a eukaryote, with nucleus already present, or another prokaryote. If the former, the mitochondria could have had a role in making oxygen less toxic for an anaerobic host. If the latter, the host may have been an archaen, within which primitive mitochondria could have produced hydrogen as a source of energy and electrons for the host cell. In both scenarios, over time, the organisms would have come to rely on each other totally, and mitochondria would have lost their cell walls and transferred some, but not all, of their genetic material to the hosts. So all the time you’ve been reading this you have been burning oxygen because of cellular heritage more than 1 billion years old: a lasting tryst between two early unicellular organisms. Neat, huh?

Figure 8 — A simplified tree of life. Many of the major branches are omitted, but a split between prokaryotes and eukaryotes is shown. The dark-grey arrow represents the endosymbiotic origin of eukaryote cells. Also marked on the tree are the ‘protists’, a grouping of single-celled eukaryotes such as amoebae and malaria (see below).

Plastids and malaria: Like mitochondria, plastids are membrane-bound organelles, but they are a little more independent of their host than mitochondria. Chloroplasts, which allow plants to photosynthesize, are one example. Plastids have evolved numerous times in the history of life, and often resemble cyanobacteria. In symbiosis, the cyanobacteria would provide carbon compounds to the host, and the host would provide mineral nutrients to the cyanobacteria. Another, independently evolved, plastid is that found in Plasmodium falciparum — the parasite responsible for the most virulent and prevalent form of malaria. This parasite is a nasty little unicellular eukaryote. It has a nucleus and organelles, so conventional antibiotics (which kill prokaryotic bacteria) can’t be used to fight it. In each malarial parasite there is a plastid that probably began life as a eukaryotic red alga. This — and the resulting plastid — was at one point photosynthetic. The organelles have four membranes, and have either lost their nuclei, or as we see in some lineages related to malaria, have a dramatically reduced remnant called a nucleomorph. These plastids are also no longer photosynthetic — ancestors of the group probably converted into parasitism early in the evolution of animals, more than 500 million years ago. Nevertheless, it seems that the plastids are integral to survival in a number of the life stages of the Plasmodium falciparum parasite, facilitating the biosynthesis of important compounds such as fatty acids. This, and the fact that they aren’t found in human cells, makes them a good target for drugs to fight malarial infection.

Fossils: So, this is Palaeontology [online] and not biology 101; it would be nice to look at some fossils. The earliest specimens that might be fossils are from around 3.4 Ga: one rock called the Apex Chert in Australia contains possible traces of life, in the form of carbon-rich structures shaped like hairs, called filaments (Fig. 9A–E). However, proving that such traces are biological in origin is very difficult, and these particular structures have been called into doubt, and the argument currently continues. The similarly aged Hoogenoeg Formation of South Africa also has possible traces of life, which could represent hyperthermophile prokaryotes. By around 3 Ga, stromatolites — layered rocks that, in many cases, may have been laid down by cyanobacteria — became common. By the time of the 1.88-Ga Gunflint Cherts, today found in Minnesota, USA, and Ontario, Canada, a wide range of prokaryotes existed; they are preserved as fossilized cells, colonies of small round bacteria (coccoidal colonies) and filaments (Fig. 9F–I). Possible multicellular fossils have been described from 2.1-Ga rocks in Gabon, but exactly where these fall on the tree of life is far from settled. The oldest unequivocal eukaryotes date from 1.5 Ga, and freshwater or terrestrial eukaryotes have been described from 1 Ga rocks in Northern Scotland (Fig. 9J–N).

Figure 9 — A–E: Examples of fossils from the Apex Chert. If they are biological in origin, they are contenders for the earliest known fossils. Modified from this source. F–I: The earliest universally accepted fossils, from the Gunflint Chert. Modified from this source. J–N: Cellular fossils from the Torridonian rocks of northern Scotland — which include the earliest preserved freshwater eukaryotes. Modified from this source. Scale bar in panel A: 100 micrometres. All others 10 micrometres.

Sex: Sex is a little bit weird — and not just the noises animals make when doing it (especially foxes). For any sexually reproducing species, there is a two-fold cost: only half the species can bear young, and males must be able to find females. Nevertheless, it is a common method of reproduction, especially in animals and plants, where it has evolved repeatedly. Thus, it must improve the fitness of any offspring. It seems that sex is preferable to asexuality when there is a threat that changes rapidly between generations, is sensitive to genetic variation and kills a large proportion of species populations. The most likely suspects to meet these criteria are parasites and disease-causing pathogens, which co-evolve with their hosts, changing rapidly between generations. Sex helps to fight this constant bombardment through increasing genetic variation, and helps to spread favourable traits quickly. This battle between organisms and disease is a race, very similar to that of the Red Queen in Lewis Carroll’s Through the Looking Glass, who said, “Now, here, you see, it takes all the running you can do, to keep in the same place. If you want to get somewhere else, you must run at least twice as fast as that!” So this idea is called the Red Queen hypothesis. An alternative (well, complementary) idea is that replication through sex is often imperfect, allowing beneficial mutations, which can then be spread. There are two copies of every gene, which minimizes the expression of harmful mutations. Sexual reproduction may date back to the origin of Eukaryotes, but there is an evolutionary radiation at 1.2 Ga that could also have resulted from the advent of sex.

Multicellularity: Building a body from multiple cells is another complicated and — I think — downright amazing adaptation. All the information needed to build every single cell in your body must be stored in the DNA that they all share. Furthermore, your entire body forms from a single cell: the genetic bottleneck that is a fertilized egg. Despite the ferocious complexity of this task, there are more than 20 independently evolved instances of multicellularity — including plants, fungi and animals, to name just three of the most familiar. Multicellularity allows cells of different types to form, and so labour is divided within an organism, encouraging increased specialization. When the system goes wrong and cell-growth gets out of control, cancer is the result. Multicellularity is most likely to have evolved through the symbiosis of unicellular organisms of the same species that work together, creating colonies with specialized roles for the different individuals. This process been observed numerous times in the living world, and the boundary between colonial organisms and a multicellular entity is rather diffuse. The first convincing evidence of multicellularity in the fossil record dates from about 1.7 Ga, with a possible contender at 2.1 Ga. The first reliable cellular differentiation is placed at about 1.2 Ga (the red algae Bangiomorpha pubescens, Fig. 10).

Figure 10 — The 1.2-Ga red algae Bangiomorpha pubescens, which displays cellular differentiation. Source.

Sponges, slime moulds and the Portuguese man o’ war: We need look no further than some of these special creatures to see how flexible, and sometimes inflexible, multicellular life can be. On the side of flexibility, can I introduce you to the common sponge? This is a creature made of layers of specialized cells, so it shows functional differentiation and a division of labour (Fig. 11A). However, if you put one through a sieve — breaking the cells apart and making them, in effect, unicellular — the cells show individual, amoeba-like behaviour. Eventually, they can group together to form cell agglomerations, and finally whole new sponges. If you do this with two different species, and mix the resulting mush, eventually the separated cells will mix only with their own species. This survival is unusual for multicellular creatures — normally, if you chop a bit off, or indeed push it through a fine sieve, the disaggregated bits die.

Inflexibility can be seen in the Portuguese man o’ war (Fig. 11B). This creature, which looks like a jellyfish, is in fact a colonial member of the same phylum, in a group called the siphonophores. We know from their anatomy and evolutionary relationships that Portuguese man o’ wars are colonies of individual zooids, and they share a unicellular common ancestor which is more closely related to them than it is to, say, true jellyfish. However, if you chop bits off, they cannot survive on their own — thus the Portuguese man o’ war is, to all intents and purposes, part of an independently evolved multicellular animal line. Awesome, huh?

A final example of just how creatively confusing life can be when it comes to multicellularity are the slime moulds. Dictyostelium, is a kind of slime mould. Members of this genus are singularly unprepossessing things that live in soils, most of the time as haploid social amoebae. They eat bacteria in soil, and divide asexually. However, if there is a lack of food, the single cells can do one of two things: in the first, two cells can fuse sexually, and attract other cells that are then eaten. Some of these (before being devoured) leave a protective barrier around a giant diploid cell that can, at a later date, hatch amoebae for the cycle to begin again. As if that wasn’t cool enough, the other option is that a social life cycle can kick off. The amoebae aggregate to form a small free-moving slug that acts like a single multicellular organism. This seeks out light, and eventually forms a fruiting body: some of the cells die to form a trunk-like extension, to lift up the remaining cells (Fig. 11C). These are then better placed to release spores that can be dispersed and hatch out into further amoebae in the correct conditions. Some colonies even farm bacteria, and carry these during spore dispersal to maximise their chances of survival. Aren’t they just the coolest?

Figure 11 — The freshwater sponge, Spongilla lacustris, courtesy of Kirt L. Onthank. B. The Portuguese man o’ war. C. The fruiting bodyof the slime mould Dictyostelium discoideum.

Conclusion:

In the last 20 minutes or so, we’ve covered a few of the major things that happened in the first 3 billion years of evolution (still breathing? Oh good!). It has been a wild ride from the origins of life, somewhere before 3.5 billion years ago, to the organization of cells through endosymbiosis, the advent of sexual reproduction and the development of multicellularity (speaking of, isn’t Fig. 12 awesome? More respect to the slime moulds!). By necessity, I have left quite a lot out, for which I can only apologize. I can’t pretend that the missing bits are not pertinent or interesting. Some of what we’re missing is factual, but I have also glossed over more than a little of the uncertainty and conflict inherent to palaeontology and life sciences in the murky depths of geological time. To compensate for these shortcomings, there are suggestions for further reading below, which outline aspects of the above in more detail. No doubt our understanding of much of this will change in the near future. Nevertheless, I hope that you have enjoyed reading as much as I enjoyed writing this.

No sponges were harmed in the creation of this article.

Further reading:

Coté, G. and De Tullio, M. 2010. Cell Origins and Metabolism. Nature Education 3(9). An excellent, freely available overview of a number of these topics from Nature Education’s online teaching/learning portal, Scitable. Good job, Nature!

Eldredge, N. & Eldredge, G. 2012 Introducing “The Origin of Life”. Evolution: Education and Outreach 5, 333. doi:10.1007/s12052-012-0451-9 A special issue of this excellent journal, aimed at a wide audience, exploring a number of these themes. Good job, all involved! Note: Until the end of 2012 access to this journal requires an institutional subscription, or is eye-wateringly expensive. If you can’t access the papers, please drop me an email.

Ruse, M. & Travis, J. 2011. Evolution: The First Four Billion Years. Harvard University Press. ISBN:0674062213 A lengthy but great overview of evolution in all its forms and guises. Any similarities to the title of this article are entirely coincidental. And all that jazz. 

Brusca, R. C. & Brusca, G. J. 2003. Invertebrates, 2nd edn. Sinauer. ISBN:0878930973 An introductory invertebrate-zoology textbook, which gives a clear picture of the context and biology of the animals mentioned above. 

And finally, some cool, more technical stuff:

Abramov, O. & Mojzsis, S. J. 2009. Microbial habitability of the Hadean Earth during the late heavy bombardment. Nature 459, 419–422. doi:10.1038/nature08015

Gargaud, M., Lopez-Garcia, P. & Martin, H. 2012. Origins and Evolution of Life: An Astrobiological Perspective. Cambridge University Press. ISBN:052176131X

Kalanon, M. & McFadden, G. I. 2010. Malaria, Plasmodium falciparum and its apicoplast. Biochemical Society Transactions 38, 775–782. doi:10.1042/BST0380775

Kumala, M. 2010. The never-ending story — the origin and diversification of life. Evolution: Education and Outreach doi:10.1007/s12052-010-0278-1

Strother, P. K., Battison, L., Brasier, M. D. & Wellman, C. H. 2011. Earth’s earliest non-marine eukaryotes. Nature 473, 505–509. doi:10.1038/nature09943

Figure 12 — Ernst Haeckel’s representation of the Mycetozoa, or slime moulds, from his 1904 work Kunstformen der Natur (Artforms of Nature).

¹ 1851 Royal Commission Research Fellow | School Of Materials / School of Earth, Atmospheric and Environmental Sciences, The University of Manchester, Oxford Rd., Manchester M13 9PL.

Introduction:

There are three groups of mammals alive today: the egg-laying monotremes (echidnas and platypuses); the marsupials (those with pouches); and the placentals (those that develop a placenta in the womb and give birth to comparatively developed young). Marsupials and placentals are sister groups, more closely related to each other than to monotremes. Along with their closest fossil ancestors, marsupials belong to the clade metatheria, whereas placentals belong to the clade eutheria. Together, metatheria and eutheria comprise the therian mammals. Marsupials are much less diverse than placental mammals in terms of numbers of different groups, range of lifestyles, range of body shapes and where they live. Why this is the case is still not well understood, and although research into the mystery of marsupial evolution is beginning to shed some light, there is still much work to be done.

Metatherian development:

One major physiological difference between marsupials and placental mammals is how they reproduce. Placental mammals give birth to well-developed young. By contrast, marsupials give birth only a few days after conception, to small and comparatively underdeveloped young. Some marsupial newborns are so tiny and at such an early stage of development that they breathe through their skin because their lungs have not yet properly formed. The marsupial newborn must find its own way into its mothers pouch and latch on to a nipple to survive. Those from different groups of marsupials manage this in different ways, but perhaps the most impressive is the long uphill climb of a small, sticky kangaroo newborn. About the size of a jelly-baby, the tiny joey has to crawl up the hairy abdomen of its mother using only its forelimbs (the hind limbs are not well developed; see Fig. 1, which shows not a joey, but a similar newborn) to reach the safety of the pouch. Because the pouch is a soft-tissue feature, it is currently impossible to tell from the fossil record whether early metatherians reproduced in the same way.

Figure 1 — Cleared-and-stained post-natal marsupial (grey short-tailed opossum, Monodelphis domestica; left) and pre-natal placental (four-striped grass mouse, Rhabdomys pumilio; right), showing differences in development at time of birth. Bony elements are highlighted in pink. (Image modified from Goswami, A., Weisbecker, V. and Sánchez-Villagra, M. R.. 2009. Developmental modularity and the marsupial-placental dichotomy. Journal of Experimental Zoology Part B, Molecular and Developmental Evolution 312B, 186–195. doi:10.1002/jez.b.21283)

The relationship between marsupials and placentals:

Despite having evolved from one common ancestor, groups within the therian mammals are not equally closely related. The closeness of their relationships depends on how much evolutionary history they share (Fig. 2). Evidence from DNA and anatomical features of both living and extinct therians has been used to work out these relationships. Although not everyone agrees about how this evidence should be interpreted, there is a general consensus on the major groups of living theria.

Figure 2 — Cladogram representing the relationships of the major orders of living eutherian and metatherian orders (positions of eutherian orders based on Madsen et al., 2001 and Murphy et al., 2001; metatherian orders based on Nilsson et al., 2010).

There are more than 4,000 species of placental mammals in 20 orders. They live on every continent, and occupy ecological niches covering almost all altitudes, latitudes, habitats and diets, and have specialized limbs for many different ways of moving about.

By contrast, there are only 331 species of marsupials, in seven modern orders (there are also three known extinct orders). They are very common in Australia, but apart from that, marsupials are found only in South America (about 90 species) and North America (one species). Living forms range from the single species of the mouse-like order Microbiotheria in South America to the Australian order Diprotodontia, which includes more than 100 species, from kangaroos to wombats.

There are no marsupials that are highly specialized runners and none that live in water or have powered flight. However, some fill very similar ecological niches to some placental mammals, and look superficially similar. For example, Notoryctes is called the marsupial mole; the sugar glider, Petaurus, glides between trees in the same way as the placental flying squirrel; the numbat, Myrmecobius, is a marsupial anteater; and the Thylacine, which sadly went extinct in the 1980s, is called the marsupial wolf. Despite the limited geographical range of modern metatherians, they are found in the fossil record on every modern continent.

Metatherian fossil record:

The earliest known metatherian, the opossum-like Sinodelphys szalayi (Fig. 3), comes from the early Cretaceous Yixian Formation of China (125 million to 121 million years old). Cretaceous metatherians have been found elsewhere in Asia and in Europe, but they are found much more often in North America. This is probably a reflection on the collection effort in different parts of the world.

Figure 3 — Sinodelphys szalayi from Luo, Z.-X., Ji, Q., Wible, J. R. and Yuan, C.-X. 2003. An Early Cretaceous tribosphenic mammal and metatherian evolution. Science 302, 1934–1940. doi:10.1126/science.1090718.

Most palaeontologists agree that Metatheria originated in Asia, diversified throughout the northern continents, dispersed into Gondwana (the modern Southern Hemisphere) around the beginning of the Palaeocene epoch (65 million years ago), then moved through Antarctica to Australia before the Australian continent was isolated by the opening of Drake Passage between South America and Antarctica. But the fine details of these dispersals and radiations are not well understood. Whether metatherians took a western route through Europe to North America or an eastern one through North America to Europe is undecided, mostly because the European fossil record is quite poor.

The early Metatherian fossil record is confined to the northern continents until the end Cretaceous period. Metatheria appear in South America in the Palaeocene then seem to diversify throughout Gondwana, but die out in Laurasia (the modern Northern Hemisphere) after the Eocene epoch (55 million to 34 million years ago).

European metatherians fall into two extinct, opossum-like families: the Herpetotheriidae and Peradectidae, both of which are also known from North America and Asia. The European and North American herpetotheriids are very similar, which is strong evidence that they travelled between the two continents during the late Cretaceous period (100 million to 65 million years ago). However, it is not clear for how long they would have been able to use this route and whether there was a preferential direction of travel; it is also possible that early metatherians travelled to both Europe and North America from Asia.

The only undoubted fossil metatherian from Africa is Peratherium africanum, a herpetotheriid known from the Oligocene of Egypt (34 million years ago to 23 million years ago), is thought to have originated from European stock.

North American metatheria have a good fossil record, possibly because preservation conditions were favourable, because there has been a large effort to collect fossils here or because there were in fact once very large numbers of marsupials on the continent. Most taxa belong to families of Didelphimorphia, or opossums.

The fossil record for metatherians in the Southern Hemisphere is far more substantial than that in the Northern Hemisphere. Palaeontologists know of several sites with fossils of many different kinds of animals.

The first South American Didelphimorph is Szalinia from the early Palaeocene site Tiupampa in Bolivia. More recent Didelphimorphs include the only living North American species, the Virginia opossum Didelphis virginiana, and are for the most part tree-dwelling (arboreal) insectivores, carnivores or omnivores. One genus, Chironectes, spends much time in the water and even has webbed feet, although it is not fully aquatic.

The extinct order Sparassodonta is known from the Palaeocene and Pliocene of South America (65 million to 3 million years ago). Sparassodonts were the largest South American metatherians, ranging from opossum-sized to bear-sized, and had ecologies ranging from carrion eaters to sabre-toothed predators, with molars specialized for cutting (Fig. 4). Their relationship to other marsupials is uncertain.
Only one genus of the South American order Microbiotheria survives: Dromiciops or ‘monito del monte’, a small, arboreal insectivore. The relationships of Microbiotheria are unclear; learning about them would help us to understand how metatherians evolved in Australia, particularly because there are almost no fossil marsupials in Australia until the Oligocene. Most morphological and DNA studies of metatherians place Microbiotheria with Australian forms. This would imply either that Microbiotheriids returned to South America after reaching Australia, or that Australian forms diversified before reaching Australia. Without fossil evidence, it is hard to tell for sure how species evolved.

Figure 4 — Artist’s impressions of the Sparassodont predators Thylacosmilus (left / source) and Borhyaena (right / source).

Antarctica, the ‘stepping stone’ between South America and Australia, potentially holds a wealth of significant fossils, but recovery of such evidence is limited by the difficult working conditions on the continent. However, a small number of metatherian fossils are known from the middle Eocene La Meseta Formation of Seymour Island, Antarctica, and include some from the orders Didelphimorphia and Microbiotheria. Species must have crossed into Australia from South America before the continents separated, between 43 million and 35 million years ago. It is possible that animals were still crossing through Antarctica well into the early Palaeocene; evidence for this is Chulpasia, an Eocene genus found at both Chulpas in Peru and Tingamarra in Queensland, Australia.

The Tingamarra site marks the beginning of the Australian fossil record for metatherians. There is then a significant gap in the Australian fossil record following this until the Riversleigh deposits of the boundary between the Oligocene and Miocene epochs (about 23 million years ago), by which point all extant orders of marsupials have made their appearance.

Most Australian fossil orders have recent representatives; the only extinct order is Yalkaparidontia, which appears in the Miocene deposits of Riversleigh. Specimens have molar teeth similar to those of Notoryctes, and skulls similar to those of bandicoots. It has been suggested that these animals could have been ‘marsupial woodpeckers’ that fed on wood-dwelling insect larvae, as does the living Indonesian Diprotodont Dactylopsila.

Diprotodontia are by far the most abundant and diverse Australian marsupials. They first appear in the late Oligocene, and were even more diverse in the Oligocene–Miocene period than they are now. Diprotodontidae, an extinct family, includes the largest known marsupials: the very largest, Diprotodon, was a fully quadrupedal form estimated to weigh more than one tonne. Members of the Palorchestidae, including the horse-sized Palorchestes, look similar to diprotodontids, and persisted until the late Pleistocene epoch (about 10,000 years ago). Thylacoleonidae, or the marsupial lions, Priscileo, Wakaleo and Thylacoleo, known from the boundary between the Pliocene and the Pleistocene epochs (about 2.5 million years ago), are carnivores with large cutting premolar teeth and an extremely powerful bite.

Living members of Diprotodontia fall into three major groups: the Phalangeriformes, the Macropodiformes and the Vombatiformes. Phalangeriformes comprises five arboreal families including the Trichosurus or brush-tailed possum; the Spilocuscus or common spotted cuscus; the nectar-feeding Tarsipes or honey possum; and the Petaurus or sugar glider. Macropodiformes comprise three families of bipedal bounders including the kangaroos. Vombatiformes include the quadrupedal, digging wombat family Vombatidae and the koalas (Phascolarctidae), which are now endangered because they eat only eucalyptus, which is disappearing.

The most abundant marsupial group in the late Oligocene and Miocene was Peramelemorphia, with around a dozen species of omnivorous hopping bipeds. The two extant families of this order are the Peramelidae (bandicoots) and the Thylacomyidae (bilbies). There is also one recently extinct family, the Chaeropodidae or pig-footed bandicoots.

The order Dasyuromorphia includes Myrmecobius, the numbat or marsupial ant-eater. Myrmecobius is interesting because no similar forms are known in the fossil record before the occurrence of still-living species in the Pleistocene.

Our knowledge of metatherian history fluctuates in volume and geographical location of fossils over time. Notable sites such as the Bolivian Tiupampa, the Brazilian Itaborai and Oligocene–Miocene deposits of Riversleigh in Australia provide a wealth of material. Yet time periods such as the Eocene in Australia and the early Palaeocene in South America are notably devoid of information. Significant gaps in the fossil record and large variations in the rigour of collection across the world also make an accurate interpretation of metatherian evolutionary history difficult despite the increasing popularity of techniques that investigate it through DNA and body shape.

Why marsupials are less diverse than placentals — the constraint hypothesis:

Palaeontologists sort fossils into species and families mostly on the basis of anatomical similarities and differences, but this classification system cannot explain much about why animals have different shapes. However, during the 150 years since Charles Darwin proposed his idea of evolution by modification through descent, our understanding of the mechanisms through which animals come to differ from each other has grown considerably. The concept of evolutionary constraints — limits on the nature of forms that can be evolved — has received much attention.

Biological form is produced during development: growth from the fertilized egg into an adult organism, as dictated by inherited genetic information. It has been recognized that early events in an animal’s development, such as the sequence or timing of the growth of individual body parts, have significant roles in the range of shapes that can be produced. If the mechanisms guiding the developmental process can change easily, diversity is promoted. By contrast, constraints in the developmental pathways can limit diversity.

The possibility that marsupials might be limited to certain shapes, owing to the need to crawl into the pouch after the short pre-natal development period, was first discussed in the 1970s as a way of explaining the lack of aquatic or flying marsupials. Study of shape changes during growth in the bones of the shoulder and a comparison of adult diversity in the shoulder blade and pelvis have found evidence for constraint in marsupial shoulder morphology.

Biomechanical demands on the skull change throughout growth, as the animal stops feeding by sucking and begins to use processes such as chewing. This is a key transition in skull growth for all therians. The sucking period is longer for marsupials than for placentals, and it is during this phase that the skull bones not present at birth in marsupials ossify (become bone rather than cartilage). It is possible that skull shape is also constrained in marsupials owing to these early mechanical demands. However, a recent study of morphological variation in the skulls of extant and extinct carnivorous metatherians and eutherians does not seem to support this hypothesis. Further study is under way to see whether skull shape is constrained for other metatherians.

Summary:

The difference in evolutionary history between metatherian and eutherian mammals is key to understanding the current contrasts in diversity between the marsupials and placentals. However, the fossil record is incomplete and biased, which potentially obscures our perception of true patterns. Work is under way to tease out true diversity patterns from the metatherian fossil record using statistical techniques. Patterns in species diversity and geography, however, tell only part of the story. The diversity of shape, appearance and ability, which must enable or limit ecological diversity of mammal groups, surely plays a major part in their evolutionary history. Physiological differences between placentals and marsupials such as those relating to the early mechanical demands on the newborn may well have caused metatherians to be limited where eutherians were not, and thus is a possible explanation for the greater diversity of today’s placental mammals. The combination of many lines of evidence and techniques from fields such as palaeontology, maths, molecular biology, developmental biology and even biomechanics is important for the comprehensive understanding of evolution. Such an interdisciplinary approach requires collaboration and communication between specialists in those fields but also a broad understanding of all areas.

Suggestions for further reading:

Tyndale-Biscoe, H. 2005. Life of Marsupials. Collingwood: CSIRO Publishing. ISBN:9780643062573
Nowak, R. M. & Dickman, C. R. 2005. Walker’s Marsupials of the World. Baltimore: Johns Hopkins University Press. ISBN: 9780801882227

Archer, A., Hand, S. & Godthelp, H. 2000. Australia’s Lost World: Riversleigh. Bloomington: Indiana University Press. ISBN:9780253339140

Kemp, T. S. 2004. The Origin and Evolution of Mammals. Oxford: Oxford University Press. ISBN:9780198507611

West-Eberhard, M. J. 2003. Developmental Plasticity and Evolution. Oxford: Oxford University Press. ISBN:9780195122350

Cox, C. B. & Moore, P. D. 2005. Biogeography: An Ecological and Evolutionary Approach. Hoboken: John Wiley & Sons. ISBN:9781405118989

Madsen, O., Scally, M., Douady, C. J., Kao, D. J., DeBry, R. W., Adkins, R., Amrine, H. M., Stanhope, M. J., De Jong, W. W. and Springer. M. S. 2001. Parallel adaptive radiations in two major clades of placental mammals. Nature 409, 610–614. doi:10.1038/35054544

Murphy, W. J., Eizirik, E., O’Brien, S. J., Madsen, O., Scally, M., Douady, C. J., Teeling, E., Ryder, O. A., Stanhope, M. J., De Jong, W. W. and Springer, M. S. 2001. Resolution of the early placental mammal radiation using Bayesian phylogenetics. Science 294, 2348–2351. doi:10.1126/science.1067179

Nilsson, M. A., Churakov, G., Sommer, M., Tran, N. V., Zemann, A., Brosius, J. and Schmitz, J. 2010. Tracking marsupial evolution using archaic genomic retroposon insertions. PLoS Biology 8, e1000436. doi:10.1371/journal.pbio.1000436

¹Department of Genetics, Evolution and Environment, University College London, London WC1E 6BT, UK

Introduction:

Pterosaurs are often mistakenly called flying dinosaurs, but they are a distinct, although related, lineage. They are an extinct group of reptiles from the Mesozoic era (251 million to 66 million years ago) and were the first vertebrates to evolve powered flight (Figs 1 and 2). Pterosaurs were first described as early as 1783 and recognized as flying reptiles shortly afterwards, and more than 150 species are now known. Fossil pterosaurs have been found around the world, with every continent yielding specimens.

Figure 1 — The holotype specimen of Pterodactylus, the first pterosaur known, described in 1783. Permission to use this photo was kindly granted by the Bavarian State Collection of Munich, Germany. Photograph taken by Georg Janssen.

Adult pterosaurs ranged in size from around 1 metre in wingspan to more than 10 metres; the largest species were the biggest flying animals of all time. They occupied the skies for much of the Mesozoic era and had the air to themselves until the birds first appeared in the middle to late Jurassic period (176 million to 146 million years ago). Pterosaurs died out along with the non-avian dinosaurs and many other groups 65 million years ago, in the great extinction at the end of the Cretaceous period.

Figure 2 — The ‘dark wing’ specimen of Rhamphorhynchus. This beautiful specimen is partially preserved in three dimensions and also has spectacularly detailed wing membranes. Top: the specimen under natural light. Bottom: under ultraviolet light, where extra details can be seen. Natural-light image by D. Hone, ultraviolet photograph kindly sent by Helmut Tischlinger.

The fossil record for pterosaurs is poor compared to that for many Mesozoic reptile groups, because their bones were fragile and so were not readily preserved. Until the past few years, there was little research dedicated to pterosaurs, and as a result many things relating to their biology are still either contentious or poorly understood. However, a recent resurgence in interest in this group and a raft of new finds are helping palaeontologists to get to grips with this important group, or clade.

Phylogeny:

The origins and the relationships of the pterosaurs have long been contentious, although a consensus is forming on both issues. Often confused with dinosaurs, pterosaurs are members of their own clade, but are close relatives of their more famous cousins.

Over the years, palaeontologists have hypothesized that pterosaurs originated from various parts of the reptile evolutionary tree. Very early researchers considered them to be the ancestors of birds or even bats, and for a long time it seemed that they were probably basal archosaurs (the clade that contains dinosaurs, birds, crocodilians and some other groups). More recently evidence has begun to stack up that they are a separate group to the dinosauromorphs (dinosaurs and their closest relatives) but that the two groups evolved from a common ancestor. Most researchers now support this position. This makes pterosaurs reasonably close relatives to birds, but they are not bird ancestors as is sometimes wrongly reported.

Pterosaurs are divided into two broad groups. The basal pterosaurs are called the rhamphorhynchoids and are characterized by a number of features of the skeleton, including: relatively small heads with a separate nostril and antorbital fenestra (an opening in the front of the skull between the eye and the nostril, also present in dinosaurs); short necks and large bodies; a short first bone in the fourth finger; a short pteroid bone (see below); a long fifth toe; and a long tail. The more derived pterosaurs have been grouped into the pterodactyloids and had the opposite set of characters: a long head with a combined (and often very large) nostril and antorbital fenestra forming one large opening in the skull; a long neck and short body; long fourth finger and pteroid bones; a short fifth toe; and a short tail (Fig. 3). (As an aside, the name pterodactyloid obviously derives from Pterodactylus, the genus of a type of pterosaur, although neither of these really means the same as the term pterodactyl, which is often misused in place of ‘pterosaur’).

Figure 3 — Skeletal outlines of Rhamphorhynchus (left) and Pterodactylus (right) showing off the basic body plans of the rhamphorhynchoids and pterodactyloids respectively. Note the different sizes of the heads and bodies, and the different proportions of the wings. Image by Edina Prondvai, based on an original by Peter Wellnhofer.

The rhamphorhynchoids and pterodactyloids remained really rather separate with a large anatomical gap between them, until the discovery of Darwinopterus in 2010. This animal is from the Middle Jurassic of China and has a mixture of traits: the large head, combined nasoantorbital fenestra and long neck of the pterodactyloids, but the long tail, short fourth finger bone, long fifth toe and other features otherwise seen only in the basal forms (Fig. 4). Darwinopterus (and several close relatives that have since been discovered) is a wonderful example of a transitional fossil showing in part how one group of animals evolved into another.

Figure 4 — Skeleton of Darwinopterus. Note the pterodactyloid-like head and long neck, but the rhamphorhynchoid body, wings and tail (compare with Fig. 3). Image kindly provided by Lü Junchang.

Anatomy:

Pterosaurs can be identified instantly by their highly modified arms. The first three fingers of the hand are small and would be used to move around when not in flight. The fifth finger of the hand is absent, but the fourth is both robust and massively elongated, and would have provided the main support for the wing membrane. Many of the bones of pterosaurs were thin-walled and hollow like those of birds and some dinosaurs, making the skeleton light overall.

Running from the tip of each wing finger to each ankle was the main wing membrane. This was not leathery, as is often stated, but in fact was a skin-like structure with layers of stiffening fibres, blood vessels and a sheet of muscle (Fig. 2). A smaller membrane sat in the crook of the elbow, supported by a modified wrist bone called the pteroid that was unique to pterosaurs. Finally, a membrane spanned the space between the legs. In rhamphorhynchoids, this was a single broad sheet and was anchored to the long fifth toe on each foot. In pterodactyloids, it was split into two smaller parts, with each half running from the ankle to the base of the tail. This arrangement freed the legs and allowed the reptiles to walk more easily on the ground. In addition to all this flight apparatus, the rhamphorhynchoids also had a vane, on the end of their long tails.

Pterosaurs were also ‘furry’. Their bodies were covered in thin, hair-like fibres termed pycnofibres. This was neither true fur as in mammals nor the simple feathers seen in early dinosaurs and baby birds, but probably evolved independently. It may have been linked to their ability to fly and there is a strong suggestion that pterosaurs were homeothermic (‘warm blooded’).

As might be expected of flying animals, in general the pterosaurs had rather conservative anatomy; that is to say that the restrictions on body shape imposed by flight meant that their overall shape was relatively similar between taxa. Over time there was a general trend for increasing size, with the earliest pterosaurs being rather small and the later ones being especially large. Early forms had lots of — often large — teeth, whereas the most derived forms from the late Cretaceous period were toothless. The most obvious deviation from conservatism was in the remarkable array of head crests that many members of the group sported. These had many different sizes and shapes and could be made of bone, soft tissues or a combination of both (Fig. 5).

Figure 5 — A wide variety of pterosaur heads showing off the different shapes and teeth, but especially the range of unusual head crests. (A) Dimorphodon, (B) Rhamphorhynchus, (C) Ornithocheirus, (D) Pteranodon, (E) Pterodactylus, (F) Pterodaustro, (G) Dsungaripterus, (H) Tupanadactylus, and (G) Thalassodromeus. Artwork kindly provided by Mark Witton.

Lifestyle:

Despite occasional reports, there is currently no evidence that any pterosaurs were flightless. Pterosaurs were not clumsy flappers or gliders as they have occasionally been portrayed, but were excellent fliers. It is likely that most pterosaurs hunted on the wing, and many lineages seem to have been well adapted to catching fish: some specimens have fish preserved in the stomach. However, other lineages were filter feeders, insect eaters, shellfish specialists or predators who hunted on land. Some species have been suggested to have fed mostly on fruits or seeds.

Pterosaurs laid thin-shelled eggs, which were probably buried in soil with vegetation to keep them moist. Several fossil eggs are known, including some preserved with intact embryos. Both the embryos and very young pterosaurs have remarkably well-formed bones, and it seems likely that even very young pterosaurs could fly.

Rhamphorhynchoids are generally thought to have had difficulty walking on the ground: no footprints have been found for them, and they would probably have stuck to the trees when not flying. The pterodactyloids were better adapted for life on the ground and numerous tracks are known for them (Fig. 6).

Figure 6 — Drawing of a pterodactyloid fossil trackway showing the large four-toed feet and the splayed three-fingered hands (the wing finger does not normally leave a mark; it is held up out of the way, as seen in Fig. 7). Drawing by Mark Witton.

At least some species lived in large colonies, and many may have been social animals. The head crests were probably some form of sexual adornment or signalling structure.

Fossil Record:

Pterosaur specimens are found spanning most of the Mesozoic era. Their fossil record is rather mixed — they are generally rare and often known only from fragments, but areas of exceptional preservation can produce superb specimens and some species are known from large numbers of fossils. The famous Pteranodon is known from more than 1,000 individuals, although most are fragmentary and in poor condition. Rhamphorhynchus is known from more than 100 specimens, most of which are more or less complete. Pterosaurs from areas of exceptional preservation are often preserved with soft tissues including wing membranes and head crests, but the bones are typically crushed flat.

The rhamphorhynchoids arose in the late Triassic period (around 200 million years ago) and go at end of Jurassic period. There are records of some in the early Cretaceous of China, but more recent studies suggest that these are the result of errors in fossil dating and the specimens are in fact older. Intermediate forms such as Darwinopterus date from the middle Jurassic; shortly afterwards, in the late Jurassic, the first pterodactyloids appear. Pterosaur footprints first appear in the late Jurassic alongside the origin of the pterodactyloids, and are found in many locations around the world.

Summary:

Pterosaurs were an important component of Mesozoic land and sea ecosystems. This group lived for more than 150 million years alongside the dinosaurs; they filled numerous ecological niches and included the largest flying animals of all time (Fig. 7). Well adapted for flight, these were not clumsy gliders as they are often unfairly portrayed, but were probably every bit as good as birds in the sky. In some ways they even may have been more agile. Pterosaur research and discoveries are currently booming, and palaeontologists are rapidly gaining a better understanding of the evolution and biology of these fascinating creatures.

Figure 7 — A full-sized azhdarchid pterosaur of around 10-metre in wingspan, standing next to a modern giraffe for scale. The giant azhdarchoids were the largest flying animals of all time. Image kindly provided by Mark Witton.

Suggestions for further reading:

http://www.pterosaur.net and http://www.pterosaur-net.blogspot.com.

Unwin, D.M. 2005. The Pterosaurs: From Deep Time. Pi Press. ISBN: 9780131463080.

Wellnhofer, P. 1991. The Illustrated Encyclopedia of Pterosaurs. Salamander Books. ISBN: 9780517037010. — Now out of print and a little dated, but still available.

Witton, M.P. 2012. Pterosaurs. Princeton University Press. — Not yet published.

¹ Department of Earth Science, University of Bristol, Bristol, BR8 1RJ, UK.

*This article has been modified from it’s original form.  The article originally stated “Fossil pterosaurs have been found around the world, with every continent except Antarctica (so far) yielding specimens.” This has been corrected to “Fossil pterosaurs have been found around the world, with every continent yielding specimens.”