How Can We Work Out The Relationships Between Animal Phyla? Why Might We Care?

The categorisation of all living matter is organised into the tree of life; this describes the pathway of evolution and how animals are related to each-other. The Kingdom Metazoa (animals) are but a comparatively tiny section of this, sitting within the subdomain Opisthokonta in the Eukaroytic domain. Phyla are the main branches on the tree of life, distinct from each other and each designating a new morphological body plan. There are between 33-35 recognised animal phyla (depending on ones source); 9 of most universal are: Poifera, Cnidaria, Arthropoda, Nematoda, Annelida, Mollusca, Platyhelminthes, Echinodermata and Chordata. (Holland, 2011)

There are three bodies of evidence to be examined when assigning animals their place in the evolutionary tree: morphological, molecular, and developmental. Before modern techniques, cladistics was a study of morphological characters and behaviours; whilst providing valuable evidence, the conclusions drawn from such observations can often be misleading. Molecular evidence is a far more reliable and comprehensive guide, however even this can overlook the importance of so called 'silent genes' that do not code for a physical trait or observable behaviour, and yet are part of a species' evolutionary history. This can be examined by the study of evolutionary developmental biology. (Carroll, 2005)

As phyla represent groups of organisms that are morphologically distinct, knowing to which phyla a subject organism belongs is essential when comparing it to other organisms and analysing how its morphological structure relates to function. Moreover, the relationships between phyla explain the evolutionary journey that led to the appearance of that phyla and the organisms within. This builds our understanding of the evolution of life. (Holland, 2011)

By compiling all available evidence we can determine the correct version of the tree of life. We can then use this blueprint of evolution to make comparisons and inferences about animal phyla.

Relationships between phyla are determined by the principal of parsimony; from an array of otherwise equivalent evolutionary paths, we should accept the path requiring fewest assumptions to be true. The number of changes required for that particular evolutionary pathway is termed the 'cost'; the model with the lowest cost is the most parsimonious. This principal can be applied to all routes of evidence, therefore when we discuss characteristics of organism this can be taken to mean morphological, genetic or developmental features.

Different paths will group organisms is different ways, this can include: monophyletic groups (clades) - all descendant organisms of a last common ancestor, and the ancestor itself; paraphyletic groups - some of the organisms from a common ancestor, but not all descendants included; and polyphyletic groups - groups of organisms with no single common ancestor. It is the assigning of monophyletic groups that is most pertinent as this shows the closest relationship; to do so we consider the characteristics on the organisms in question.

There are two types of shared characteristics to consider: synapomorphies and symplesiomorphies. Synapomorphies are shared derived characteristics (present in all members of a group, and not found outside the group) these are useful as they parsimoniously prefer one tree over another, indicating the correct route. Symplesiomorphies are shared primitive characteristics (present in all members of a group, but also found outside the group) these are not useful in determining relationships as all trees are parsimoniously equal. Determining the correct relationships is essential as a different tree implies a different route of character evolution.

Using morphological characters alone as a means of determining phylogenic relationships can be misleading. It is based on the assumption that features of the descendants are derived from features of common ancestor, (Dawkins, 2005) and that as we go 'up' the tree of life, organisms become more complex. Similar physical traits between animals are assumed to be homologous, and are used to define monophyletic groups. Aberrant features are accounted for by convergent evolution.

An example of this type of 'old' phylogeny is the use of the coelem as an homologous character; this creates a tree with diploplasts at the bottom (no mesoderm), leading to triploplasts (solid mesoderm). Triploplasts then branch off, starting with the least complex: Firstly acoeleomates (no coelem), then pseudocoelomates (rough cavity within mesoderm), and lastly coelomates (epithelium lines cavity). This tree structure using coelems as a synapamorphy dictates: segmentation as homologous in annelids and arthropods - as they are grouped together sharing a common ancestor, and spiral cleavage and ecdysis as not clade defining - as these characters are scattered throughout the tree.

Non-morphological characters can be analysed in the same way, as the behaviours organisms exhibit are themselves encoded by genes inherited from their ancestors. However, it is plausible and often the case that two distantly related species have independently developed the same, or similar morphology or behaviour in response to the same environmental pressure (convergent evolution). The conclusion derived from from a superficial observation of two species apparent similarities is therefore not always accurate. This is illustrated by the disproving of the 'old' phylogeny set out above, using molecular evidence.

Molecular evidence is a powerful tool in determining the relatedness of animals. Even very distantly related animals have a vast amount of overlap between their DNA sequences, such genes control the basic function of complex life (for example, Hox genes). The more related two species are, the more DNA they have in common. One can analyse the overlap of shared gene sequence; in theory if this was plotted from smallest to largest it would yield a scale of relatedness from most distant to closest. (Dawkins, 2005)

We can also use the 'Molecular Clock' method to determine closeness of relationship and infer a common ancestor. Mutations in DNA accumulate over time, therefore more closely related species have fewer discrepancies than more distant ones. Moreover, the number of molecular discrepancies between the species is proportional to the age of their common ancestor. (Dawkins, 2005) Indisputable fossil evidence is used to calibrate molecular clock data and assign dates to divergence of lineages. (Lee, 1999) In this way we can more fully appreciate the relationships between animal phyla by understanding how and when they arose. 1988 saw the first use of genetic evidence to examine relationships between animal phyla, and disprove a theory of relatedness based on morphology - the 'old' phylogeny described above was shown to be false.

A groundbreaking paper "Molecular phylogeny of the animal kingdom" used pairwise comparisons of 18S ribosomes from animals belonging to 22 separate classes, representing 10 animal phyla. The evolutionary distance between specimens was used to infer phylogenetic relationships. This method allowed for broad comparisons across the diversity of life, as commonly conserved sequences of cellular RNA were used; moreover, the results were "independent of morphological, biochemical and developmental traits" (Field, 1988, p748) therefore producing the most reliable and comprehensive version of the tree of life at this time, that was radically different from the accepted theory.

Analysis of the data showed some surprising results. The most influential discovery was perhaps that animals with and without coelems were intermingled throughout the tree, suggesting that it is not clade defining as previously thought - instead, spiral cleavage and ecdysis were shown to be homologous features. In the 'old' phylogeny, arthropods and annelids were grouped together under the term Articulata due to their assumed homologous character of segmentation - Field's data showed arthropods and annelids had very different ribosomal sequences and were therefore unlikely to be closely related. (Holland, 2011) Articulata is a polyphyletic group with no real significance and segmentation likely arose by convergent evolution.

However, morphological and genetic evidence can support each other as evidence for an evolutionary route. For example, it is suggested the unicellular choanoflagellates are the closest living relative of all animals. The morphological similarity between choanoflagellates and sponges has long been noted; choanocytes on sponges are almost identical in structure to the unicellular choanoflagellates and used for the same purpose (the flagellum creates a water flow through a 'basket' that catches micro-organisms used as a food source). This morphological similarity is supported by gene fusion evidence; the TK and EGF genes are separate in plants and fungi, but fused in choanos and metazoa. This is a synamomorphy of the clade containing choanoflagellates and metazoans and demonstrates a closeness of relationship.

Embryology provides another source of evidence; characters that may have been lost in the adult members of a species are often apparent in the developing embryo. As the mechanism of evolution is a change in the developmental pathway, then studying embryological development is key our understanding of how closely animals are related. Gastrulation is a highly complex process, the instructions for which are encoded within our genome. Those organisms therefore who have similar embryology, likely share a large percentage of gene sequence and are therefore closely related. It is often surprising how markedly similar the embryos of superficially very different species are; for example, a fish and a mouse embryo are near identical in the early stages of development as both species are Chordates.

In the 1980s, a massive development occurred in developmental biology: the genes controlling organisation of body plan in Drosphilla (Hox genes) were discovered. Sets of homologous Hox genes were later found in most other animals. This had huge repercussions for the tree of life as it had been thought that animals with distinct body plans had distinct sets of genes to control their development. (Carroll, 2011) Keeping in mind that phyla are groups of morphologically distinct organisms, this is now a factor that links all animal phyla and shows they have more in common than previously thought.

These examples highlight the importance of using all forms of available evidence when determining the relationship between animal phyla, so that our conclusion accurately reflects the reality.

The true phylogeny of model organisms has profound implications when interpreting the results of studies, as different versions of the tree of life imply differ routes of character evolution. The validity of cross species comparisons is determined by their phylogenetic closeness; we can be more confident in comparing organisms the closer related they are. Phylogenetically independent contrasts (known as the phylogenetic comparative method, or PCM) are commonly used for comparisons of traits between taxa (Quader, 2004) Organisms cannot be analysed in isolation as this leads to phylogenetic non-independence, (Purvis, 1994) which violates the statistical assumptions underpinning these analytical models. If the true phylogeny is known, this can be accounted and corrected for. An application of this is meta-analysis; a method to calculate the effect size of an experimental outcome for a general population, based on primary data. Comparing overall pooled effect sizes of analyses done with and without phylogenetic correction, gave a 47% difference. (Chamberlain, 2012) This clearly shows that knowing the relationships between animal phyla leads to more accurate experimental results and interpretations.

There are also profound benefits for research. For example, the green fluorescent protein (GFP) is found in the bioluminescent bell rim of Aequorea victoria. It was thought that any homologs of GFP would be found in organisms that are themselves bioluminescent, however the variant proteins were isolated from reef corals which are non-bioluminescent (Mikhail, 1999); the clue that they might contains fluorescent proteins is that both A. victoria and the reef corals are Cnidarians. Six homologous fluorescent proteins have now been produced, which are used as markers in various biochemical techniques. If the close relationship between the jellyfish and the reef corals was not known these proteins may never have been found.

Through understanding phylogeny we may also gain a deeper appreciation of the process of evolution. It is often easy to think of evolution as a gradual process of improvement towards the final goal of humanity; (Dawkins, 2011) the understanding of our place within the tree of life is a profound notion. Whilst a less scientific motivation, this is nevertheless a viable answer to the question of why we might care about the relationships between animal phyla. We are ourselves animals, and have an intrinsic curiosity to understand where we came from and where we sit within the vastness of the diversity of life.

This discussion has often focused on evolution and the relationships between animals at a species level. The question sought to answer is how we can work out the relationship between animal phyla, not animal species. However, it is the exact same process of evolution, but occurring over a larger timescale, that gives rise to differences between higher taxonomic levels such as the phyla as it does differences between individual organisms. (Carroll, 2011)

An accurate descri ption of the relationships between animal phyla is essential if we are to gain a full understanding of evolution. The tree of life provides an essential reference to accurately interpret data, that we may "compare anatomy, physiology, behaviour, ecology and development between animal species" (Holland, 2011, 13). Without which our research and experiments would be near meaningless.

References:

Carroll, S. (2011). Endless Forms Most Beautiful. London, England.: Weidenfeld & Nicolson.

Chamberlain, S. (2012). Does Phylogeny Matter? Assessing the impact of phylogenetic information in biological meta-analysis.. Ecology Letters. 15, p626-636.

Dawkins, R. (2005). The Ancestor`s Tale. London, England: Weidenfeld & Nicolson. Field, K. et al. (1988). Molecular Phylogeny of the Animal Kingdom. Science. 239, p748 - 753.

Holland, P. (2011). Animal Phyla. In: The Animal Kingdom: A Very Short Introduction. New York, USA: Oxford University Press Inc.

Lee, M. (1999). Molecular Clock Calibrations and Metazoan Divergence Dates. Journal of Molecular Evolution. 43, p385-39.

Mikhail, V. et al. (1999). Fluorescent Proteins from non-bioluminescent Anthozoa species. Nature Biotechnology. 17, p969-973.

Purvis, A, et al. (1994). Truth or Consequences: Effects of Phylogenetic Accuracy on Two Comparative Methods. Journal of Theoretical Biology. 167, p293-300.

Quader, S. (2004). Non linear relationships and phylogenetically independent contrasts. Journal of Evolutionary Biology. 17, p709-715.