Athena Review Vol. 5, no. 1
Records of Life: Fossils as Original Sources
5. Classification
Linnaean and cladistic approaches.
Studying ancient life requires learning terminology created by scientific researchers, whose formal written communications are often not very conducive to casual reading. The need for succinctness and accuracy underlies much of the terminology, along with an entrenched tradition of using Latin or Greek names for animals and plants.
One critical set of terms to understand are basic units of biological classification such as class, family, genus, and species, which provides a structure for relating all types of organisms. The second necessary set of terms are the geological periods and their date ranges. These need to become familiar as a kind of basic literacy in the subjects of geology and paleontology. The reader must also learn to put up with chronology shorthand such as “mya” for “million years ago.”
The terms of paleontology are more involved, not only because of the multiplicity of species and their various levels of grouping, but also because of innovations due to cladistics (from the Greek word “clade” for “branch”) whose terms and criteria are not likely to be at all familiar to most readers. Fortunately, many scientists in the field of early vertebrate paleontology are now writing books or articles aimed at the general reader, which translate the material into simple English. [An example is the book on tetrapod evolution entitled Gaining Ground, written by British paleontologist Jennifer Clack (2012).]
The traditional method of biologic classification is the Linnaean system, devised in the 18th century by the Swedish naturalist Carolus Linnaeus or Karl von Linne (1758; fig.1). This is really a system of organized cataloguing (called taxonomy), with a fixed hierarchy of levels of groupings (e.g., class, order, family, genus, species), each of which can be called a taxon (plural, taxa). The Linnaean system is not primarily designed to show ancestry or evolution, but placement in the Systema Naturae or the great scheme of nature. It is probably easier to grasp as a mode of learning species, as when trying to familiarize oneself with a wide range of animals or plants.
Fig.1: Karl von Linne (Carolus Linnaeus), in a painting ca.1738.
As already noted, a well-entrenched factor of all naturalist classification from Linnaeus onward is the use of descriptive Latin or Greek terms for the names of species and their families, as well as the more general groupings such as phyllum, class, and order. These names, however, when translated, are usually quite simple and descriptive. For example, the genus name of an early amphibian, Proterogyrinus, means “early tadpole.” Fortunately, name translations are being increasingly provided in scientific papers.
The cladistic method of classification, called phylogenetic systematics, was originally designed by the entomologist Willi Hennig (1966; fig.2) for classifying insects. It is intended to be used for deciphering and describing the closest evolutionary relations of species, both extant and extinct, though the extensive comparison of anatomical traits called characters. The goal of cladistics is to create phylogenies, or relationship trees showing presence or absence of common ancestors between species and groups of species called clades ("branches"). This approach (which is more frequently self-identified as "phylogenetic," rather than "cladistic"), has fundamentally different goals than the hierarchical ordering of Linnaean classification, or than a third approach to be described below, called evolutionary systematics.
Cladistics has its own terminology,
which is quite distinct from that of the Linnaean system (see Benton
2000 for a detailed analysis). The comparative methods used by
cladistics involve, first, the selection of various (sometimes
hundreds) of anatomical traits or characters thought to be diagnostic for the species
under consideration, and the listing of these in tables, for as many
species as can be relevantly compared. These trait tables are then
analyzed through computer programs such as PAUP, which perform complex
pattern-grouping and parsimony analysis to create “nearest
neighbor” tree diagrams. These are considered to reflect the probable
closest relations between taxa, which can then be expressed in
cladograms.Fig.2: Willi Hennig (photo: 1971)
The relationship trees or “cladograms” aim to show clades or branches called "sister groups," (fig.3) of which every member shares a common ancester, and avoids portraying long term groupings such as classes or orders. The goal of requiring every member of a clade to share the same common ancestor (called a “monophyletic” group) also, in practical terms, requires a constant mode of revisionism by those writing up these findings. Every time a significant new fossil is discovered, it may prompt reorganization of the entire “clade” along with all neighboring branches (Benton 2000). While the revisions may be radical and improvisational, however, the digital methodology can be rather rigid. A seemingly problematic aspect, related to the standardization sought via using computer grouping methods, is the dependence on discrete traits which are either present or absent. Perhaps, as Professor Jennifer Clack suggests (2012), evolution doesn't always work that way.
.
Genus, species, family, et al.
Most people are familiar with some basic level of Linnaean classification in living beings, such as that for different species of birds or plants, fish, or household pets. In the latter case, while there are many species of aquarium fish, all dog breeds are considered as domesticated varieties of the the same species, Canis familiaris, just as all household cat breeds are part of the same species, Felis cattus.This binomial naming of genus and species (the only categories which are italicized) is at the root
of the Linnaean system, along with a larger hierarchy of ranks. As Table 1 shows, the full Linnaean taxonomy of a dog includes a number of ranks: Kingdom (animals), Phylum (chordates), Subphylum (vertebrates), Class (mammals), Order (carnivores), Family (canidae), Genus (Canis), and Species (familiaris). The same ranks or categories used for extant species are also used for fossils. The Linnaean system, in paleontology as in zoology or botany, is built into all discussions of fossil evidence. Table 1 lists the basic structure of Linnaean classification for several well-known animals.
Table 1: Examples of Linnaean classification of living vertebrates.
As can be seen from Table 1, the taxonomy for the gray wolf (Canis lupus), considered the direct ancestor of domestic dogs, is almost the same as for dogs, except for the species name lupus (table 1). In this case, the dog has changed enough under domestication to be considered a distinct species. For other carnivores, such as the brown bear Ursus arctius, all levels above that of family (Ursidae, Latin for “bears”) remain the same as for dogs and wolves. This simply reflects what is quite well attested in the fossil record, that canids, bears, and all other canivores, have common mammalian ancestors (see also fig.3 below).
Grazing animals such as a white-tailed deer (Odocoileus virginianus, named for Virginia deer) belong to a different order, Artiodactylus (Greek for “even-toed”), and a different family (Cervidae, Latin for “deer”). Here the common ancestors of both deer and wolves, bears, and other carnivores represent much earlier mammals, dating from the Late Cretaceous through Eocene periods (about 75-55 mya), before their descendants specialized into the different orders of Carnivores and Artiodactyls.
Ingroups, Sister Groups, and Outgroups.
Presenting some of the same information as shown in the Linnaean grouping in Table 1 could also be done with a cladogram, using cladistic terms (fig.3).
In this case (again using the domestic dog as the subject taxon), the cladogram shows an "ingroup" consisting of the canidae family. This group of carnivores, in this example, includes the dog and its direct ancestor, the Gray Wolf, and another wolf from the Pleistocene, the Dire Wolf, which is now extinct. Needless to say, there are many other canids, such as coyotes and jackels or other types of wolves, that could also be included here.
A related carnivore group, the bear family (ursidae) is shown as a "sister group" to canidae. The two species of ursidae selected here, the Brown Bear and the American Black Bear (Ursus americanus) appear in the cladogram on the same level of ranking as the canids, in terms of Linnean classification. This contrasts with the "outgroup," as shown in fig.1.
Fig.3: Cladogram showing the relations of several modern mammal groupings, including canids, bears, and artiodactyls.
Placed in the outgroup status, in the case of this example, are two species of the ungulate order of Artiodactyla, the White-tailed Deer, and the Pronghorn Antelope (Antilocapra americana). The order of Artiodactyls split off from the ancestral mammal branch about 54 mya, about the same time that the carnivores also diverged. (This cladogram does not show details on the ancestral mammal groups). Given the split between orders, the families of this outgroup also do not appear in parallel with the ingroup and sister group. This illustrates an inherent difference between universal categories of the Linnean system, where classes, orders, and families are always shown on the same level, and the ingroup-oriented focus of a cladogram which only places sister taxa on the same level.
.
Amniotes and Synapsids.
Another, earlier split in vertebrate evolution, that between reptiles and mammals, has been subject to some thorough redefinition. As indicated on fig.3, fossil evidence shows the first appearance of mammals by the Late Triassic and Early Jurassic periods (about 220-180 mya).
The German anatomist Ernst Haeckel added to Linnean classification the evolutionary distinction between amniotic and non-amniotic eggs used in reproduction. This basic trait showed that mammals and reptiles share a more recent common ancestor than do mammals and amphibians.
Amniotic eggs were first formally defined by Haeckel in 1866 on the basis of the amnion, a protective sheath around the embryo (fig.4). Vertebrates which give birth through amniotic eggs which contain the embryo in a protective, amniotic sac. Examples are a chicken egg, a turtle or crocodile egg, or a mammalian foetus. Thus the group of amniotes includes the classes of birds, reptiles, and mammals, but excludes all members of the classes of fish and amphibians. The latter (called anamniotes), have eggs hatched in open water. and whose larvae or tadpoles develop unprotected.
Fig.4: Diagram of the amniotic egg of a reptile.
Haeckel (1866) was very involved in the study of developmental or embryonic simil
arities and differences between vertebrates, which
frequently yields important information, such as on the long-term
evolution of the mammalian ear. Such analysis of key developments in
fossil groups are an important part of what is now called evolutionary
systematics (see below).In the evolution of mammals, probably the single most notable advance in interpretation is the recognition of the Synapsid group or clade as the early ancestors of mammals. Based on the current fossil record, the origins of Synapsids, who descended from basal amniotes called Sauropsids, occurred by the Late Mississippian or early Pennsylvanian periods (335-320 mya). Both the sauropsids and the descendant early synapsid groups would previously have been called reptiles. The term Synapsid, however, has now largely replaced that of mammal-like reptile.
Fig.5: Ernst Haeckel (photo: 1860)
Mammals (who are both amniotes and the descendants of synapsids) are also divided from birds and reptiles by the number of holes or fenestrae in the back part of their skulls, behind the eyes, used as attachment slots for jaw muscles. As further discussed in another section of this report, mammals have one such hole on each side of their skull, and are thus called synapsids (“single arch”). Birds, crocodiles, and some other reptiles each have two skull openings, and are called diapsids (“two arches”). Still other reptiles, including turtles, have no such holes, and are called anapsids ("no arches"). The basal amniote ancestor of all these groups is called a Sauropsid ("lizard face"), a term devised in the 1860s by the English anatomist Thomas Henry Huxley (1862, 1878).
Equally much in the forefront of current debate is the still earlier scenario of the first evidence of land animals, called tetrapods (“four limbs”). The trait of having four legs (tetrapody) separates fish and all other early chordates from amphibians, reptiles, mammals, and birds (who descended from diapsid reptiles, as T.H. Huxley [1878] recognized). This process occurred in the Late Devonian period (about 385-360 mya), when lobe-finned fish (Sarcopterygians) developed limbs and adapted to land, giving rise to amphibians. An example of a transitional fish who was almost a tetrapod (called a tetrapodomorph) is Acanthostega gunnari, a species named for the "spiny roof" of its skull, and for its 1930’s discoverer in Greenland, Gunnar Save-Sodebergh (1932).
The traits of both tetrapody, and the number of skull fenestrae, are of key practical imporance in separating groups of vertebrates, because they are observable in the fossil record. The trait of being an amniote or anamniote, on the other hand, is not identifiable in the fossil record, and can only be inferred when the fossil has otherwise been identified through observable traits to be an a) an amphibian, or b) a sauropsid or other reptile. These and other key trait are shown in cladogram format in fig.6.
Fig.6: Cladogram of vertebrate groups.
Evolutionary systematics.
A third approach to the classification of fossil species is known as evolutionary systematics (Palaios 2014e). This viewpoint, attempting to synthesize Darwinian theories and new fossil findings with the Linnean system, began with late 19th century anatomists such as T.H. Huxley and Ernst Haeckel. A similar approach continued in the "new synthesis" of Darwinian theory with genetics in the 1920s-60s, as shown by the zoologist Ernst Mayr (1942), the geneticist Julian Huxley (1940), and the paleontologist George Gaylord Simpson (1961). Evolutionary systematics also developed over the same period in the writings of the anatomist Alfred S. Romer (1966; fig 7); and has more recently been practiced by paleontologists including R.L. Carroll (1988) and Michael Benton (2000, 2004).
Fig.7: Alfred Sherwood Romer (photo: 1927)
In recent paleontology, this approach has combined Linnaean classification with ancestral or evolutionary relationships, by emphasising the stratigraphic sequence in the fossil record, with paleontology playing an important role in providing criteria for organizing the interpretations of fossil taxa. The emphasis on stratigraphy links evolutionary systematics with the original "bedrock" concept of sequential interpretation used in geology and paleontology. Emphasis is placed on understanding more general levels of classification such as orders or classes (i.e., Placoderms or early armored fishes, and lobe-finned fishes; fig.8) , which are considered to be basal or stem groups from which later groups radiated. The methodology involves correlating horizontal or contemporary groupings of taxa with changes in their anatomy and environments over time, as seen in stratigraphic contexts.
A type of graphic represention often used by Romer and other evolutionary systematists are so-called spindle diagrams showing relative frequencies over time for selected taxa. An example is shown in fig.8 for the long-term occurrences of fish and tetrapod families (Benton 2004). Such graphs use lengthy time-spans of a succession of geological periods, to evaluate trends of ancestral groups with primitive anatomical traits.
Cladistic methodology seems little interested in charts of this kind,
since the classes or orders thus portayed extended a long way through
time (Palaios 2014e). Consequently, in cladistic terms, these are
paraphyletic groups (i.e., with mixed phylogenies or ancestry), and do
not illuminate the sister relationships of individual species Nor
is the cladistics approach very interested in stratigraphy in the
interpretation of fossil taxa. Instead, a "molecular clock" is relied
on, based on genetic differences between species to calculate the past
time of their divergence. There are sometimes, however, marked
discrepancies between dates obtained by stratigraphic means, as opposed
to a molecular clock (Benton 2000).Fig.8: Chart of relative frequences over time of fish groups and tetrapods (after Benton 2004).
Such matters seem remote in terms of the day-to-day, practical realities of reporting new fossil taxa. Optimal standard practice, at present, is that each time a new, well-preserved fossil is found and described (for example a new Silurian fish fossil from China, such as Entelognathus primordialis by Zhu et al. 2013), the anatomy of the fossil is presented in as much detail as the preservation permits, and its identification is done via cladistic methods, using the most detailed possible comparisons of the anatomical traits found in the new fossil. It may well be (as in the case of Entelegnathus) that the fossil also provides significant help in answering basic questions of interest to evolutionary systematics, such as when the first bony fish-type jaws appear in the fossil record (the answer, based on Entelegnathus. is 319 mya). Nothing further could be asked from a classification .
References:
Benton, M.J. 2000, Stems, nodes, crown clades, and rank-free lists: is Linnaeus dead? Biol. Rev. (2000), 75: pp. 633-648
Benton, M.J. 2004. Vertebrate Paleontology. Blackwell Publishers. xii–452
Carroll, R. L. 1988. Vertebrate Paleontology and Evolution. WH Freeman & Co.
Clack, J. 2012. Gaining Ground. University of Indiana Press.
Haeckel, E. 1866 Generelle Morphologie der Organismen : allgemeine Grundzüge der organischen Formen-Wissenschaft, mechanisch begründet durch die von C. Darwin reformirte Decendenz-Theorie. ) Berlin
Hennig, W. 1966. Phylogenetic systematics. University of Illinois Press, Urbana
Huxley, J. S. 1940. Towards the new systematics. In The new systematics (ed. J. S. Huxley), pp. 1-46. Clarendon Press, Oxford.
Huxley, T.H. 1863. The Structure and Classification of the Mammalia. Hunterian lectures, presented in Medical Times and Gazette.
Huxley, T.H. 1876. Lectures on Evolution. New York Tribune. Extra. no 36. In Collected Essays IV: pp 46-138
Laurin, M. and R.R. Reisz, 1995. A reevaluation of early amniote phylogeny. Zoological Journal of the Linnaean Society, 113: 165–223
Linnaeus, C. 1758. Systema naturae per regna tria naturae, secundum classes, ordines, genera, species, cum characteribus, differentiis, synonymis, locis, Editio decima, reformata, Tomus 1, Laurentii Salvii, Holmiae.
Mayr, E. 1942. Systematics and the origin of species from the viewpoint of a zoologist. Harvard University Press, Cambridge,
Palaios 2014e, "Evolutionary systematics." Website Palaios.org.
Palaios 2014f, "Cladistics." Website Palaios.org.
Romer, A.S. 1966. Vertebrate Paleontology (Third Ed.). University of Chicago Press.
Säve-Söderbergh, G. 1932. Preliminary note on Devonian stegocephalians from East Greenland. Meddelelser om Gr¢nland, 94(7), pp.1-107.
Simpson, G. G. 1961. Principles of animal taxonomy. Columbia University Press, New York
Zhu, M. et al. 2013. A Silurian placoderm with osteichthyan-like marginal jaw bones. Nature 50: pp.188–193
Glossary