Advances in Computational Research, Vol. 1, Issue 1, 2009, pp. 1-4.

Functional module analysis in metabolomics Chokes

Shaily Mehta and Somnath Tagore. Department of Biotechnology & Bioinformatics

Abstract: Since recent years the work on biological and metabolic network has been increasing due to the new biological discoveries and essential metabolites. Metabolomics being a urgeoning field, which produces voluminous data that, like other ‘omics’ data, should be seen as a resource that contributes specifically to the former half of an iterative cycle of hypothesis-generating and hypothesistesting phases. It is becoming increasingly apparent that our ability to generate large quantities of metabolomics or metabolic profiling data will help to open up many previously inaccessible areas of biology various high-throughput techniques are leading to an explosive growth in the size of biological databases and creating the opportunity to revolutionize our understanding of life and disease. Interpretation of these data remains, however, a major scientific challenge. With the study of enzymes and metabolites new pathways can be discovered, which can help in the analysis of the various process taking place in the organism. In order to identify potential drug targets the concept of choke points was used to find enzymes which uniquely consume or produce a particular metabolite. Hence the study of these choke are taken into consideration.

Keywords: Biological, metabolic network, metabolites, enzymes, choke points


The present scenario tells us that network analysis is essential for the analysis of genetic, proteomics and metabolomics data [1, 2].

The present discovery of enzymes and metabolites has made the study of metabolomics very much in need. Since past couple of decades we have understood the basic idea of the formation of metabolites. It all starts with the process of ingestion were the organism takes the material such as food into their bodies. These material become compounds and energy necessary for sustaining the activity of the organism by various chemical reactions. The whole of such chemical reaction taking place in the body is called metabolism. Here the substrate is converted to a product i.e. one compound is converted to another and chain of reactions is generated which forms a large scale network. This network is known as Metabolic Pathway.

With the increase in research work large no of metabolic pathways are continuously being discovered and there activity is studied. Due to which large no of enzymatic databases were build to store in these data. Also the activity of each of these enzymes taking part in the metabolic pathways is studied carefully. Various graph theories, mathematical, computational and programming aspects are taken into consideration for the verification of these metabolites and chemical reactions and to demonstrate the intrinsic hierarchical modularity of metabolic networks [3] and their robustness based on the shortest path analysis of the metabolic networks [4-6].

A typical metabolic network consists of reactions, metabolites and enzymes, which can be modeled using graph

theory [7-11]. These representations lead from a simple graph consisting of edges (reactions) and nodes (metabolites) or vice versa to a complex bipartite graph where two nodes (metabolites) share a common node (reaction/enzymes) [12].

Enzyme-centric networks can be created by joining enzymes that share a common metabolite in a path. The enzyme-centric view [13] simplifies the representation of the metabolic network by removing loose ends in the network (metabolites at the periphery of the network) and forming clusters of interacting enzymes. The gene-centric view has been successfully used in determining coregulated genes in the metabolic and regulatory networks [14-18].

Hence it becomes a necessity to understand the choke points which are enzymes which play a crucial role in the metabolic pathway. Understanding these enzymes will widen the scope of analyzing the pathway considerably. Choke points are those enzymes which uniquely consume and/or produce a certain metabolite. They are ranked by the number of k-shortest paths (in/out) passing through it and the load point (in/out) on it. Since it is a reasonable assumption that a large number of the biochemical reactions follow the shortest path, we assume that the shortest path count can good indicator of biochemical activities. Inactivation of choke points may lead to an organism’s failure to produce or consume particular metabolites which could cause serious problems for fitness or survival of the organism hence they are considered crirtical points in the metabolic networks.

Chokepoint analysis has several advantages. First, it allows us to test the consistency between experimental data and assumptions about the organization and regulation of the biochemical pathway and of its interdependencies with other processes. Second, it can be used to predict the consequences of various mutations or inhibitors.The concept of choke points was used in our study to find potential drug targets in the metabolic network of Bacillus anthracis Sterne. The metabolites and enzymes are further ranked on the basis of their loads in the given network. A comparative study was performed between the human metabolic network and pathogen choke points to discriminate human choke points from the pathogenic bacterial choke points. A homology search was performed against human genome to find non-homologus potential drug targets from the pathogen choke points.

A new method to analyse choke points by screening the entire metabolic network of pathogens and report the probable choke points in the network was discovered by group of scientists [19].

This extended graph theory model ranks the choke points according to the kshortest path passing through it and the load (in/out) on it. This ranking has a major advantage as this measure may help determine the biochemical essentiality of a metabolite/enzyme. For example, in Plasmodium falciparum—a parasite causing malaria in humans—a host cell enzyme (daminolevulinate dehydratase; ALAD) involved in hemebiosynthesis was suggested as an antimalarial target.

This enzyme is also a choke point enzyme and identifying such potential targets in the athogens can accelerate the drug discovery.

Also all three clinically validated drug targets for malaria are chokepoint enzymes. A total of 87.5% of proposed drug targets with biological evidence in the literature are chokepoint reactions.


In order to confer biological meaning to the graph-based approach of finding choke points, the present studies deal with the following steps:-

For building the biochemical network we used the LIGAND database from KEGG as this data model is the backbone for the Pathway Hunter Tool in addition to BRENDA. For the predicted choke points in the pathogen we performed a homology search against the human genome using BLAST.

Calculations of the top choke points which are ranked by number incoming shortest paths along with important load points are reported in the metabolic network of the organism. A network based comparative study of the important choke points between a model organism and Homo sapiens is performed using Pathway Hunter Tool (PHT). ‘+’ implies that a particular enzyme acts as a choke point in the human biochemical network as well as in the pathogen whereas ‘–’ indicates that this enzyme is only a choke point in the pathogen and not in the human biochemical network.

A homology search is performed between the human and model organism. Choke point enzymes using BLAST and chokepoints with a closest homologue with e-values

The above method was one of the techniques proposed by the scientist working in this field. There may be many more methods which have potential in identification of these choke points which are yet to be discovered or they are in the process to reach there out.


Chokepoint analysis can be implemented in various pathway analysis.

First, it allows us to test the consistency between experimental data and assumptions about the organization and regulation of the biochemical pathway and of its interdependencies with other processes.

Second, it can be used to predict the consequences of various mutations or inhibitors.

The targeting of metabolic pathways has several advantages on its own. Each step in the pathway is well validated as an essential function for pathogen growth. The target enzymes from the pathogen which are discarded and which share a similarity with the host proteins ensures that the targets have nothing in common with the host proteins, thereby, eliminating undesired host protein–drug interactions. Metabolic pathway analysis is becoming increasingly important for assessing inherent network properties in reconstructed biochemical reaction networks. As of now, identifying choke point reactions, identification of enzymes has been done that are essential to the parasite’s survival.

There is an enrichment of drug targets in chokepoints as compared with nonchokepoints.This leads to the conclusion that the classification of an enzyme as a chokepoint has some bearing on whether or not it would make a good drug target.

Another approach could be combining Choke Point analysis with chemogenomic profiling (micro-array data), Providing a complete and better annotation in vivo (thereby reducing the identification of false Choke Point enzymes and providing previously unreported Choke Point enzymes in the metabolic network) is one of the first steps in this direction.


Choke point analysis was successfully performed on large no of organisms to discover potential drug targets. e. g. Plasmodium falciparum, B.antracis, Corynebacterium glutamicum, E.histolytica.

If an enzyme catalyzes at least one chokepoint reaction, it is classified as a potential drug target. chokepoints and non chokepoints against proposed drug targets from the literature is compared to assess the usefulness of identifying chokepoint enzymes for proposing drug targets.

A complete literature search for proposed amoebiasis drug targets is attempted that were metabolic enzymes and met the criteria discussed above.

Chokepoints may not be essential. One reason could be that they create unique intermediates to an essential product which are not essential themselves and finally, there could be chokepoint reactions that are not essential due to other pathways that achieve the same metabolic goal within the organism. One example could be blocking the reaction that has no deleterious effects on the parasite. Due to the high percentage of enzymes identified as choke points, one additional criteria observed in addition to being a choke point enzyme for identifying potential metabolic drug targets is that an enzyme not having isozymes would make it more likely to be a good drug target.

An analysis of the top 10 choke points in B.anthracis, a pathogen, is presented. In a number of possible drug targets against infection of B.anthracis are identified. It was found that the enzymes tryptophan synthase (EC: and anthranilate phosphoribosyltransferase (EC: could be effective potential drug targets. Neither of these enzymes are choke points in the human metabolic network nor do they share a significant homology with the human genome. This means that blocking these enzymes might affect the pathogen but not the human as there exists an alternate pathway To identify potential drug targets, a chokepoint analysis of the metabolic network of E.histolytica is performed. A «chokepoint reaction» is a reaction that either uniquely consumes a specific substrate or uniquely produces a specific product in the Entamoeba metabolic network. It is expected that the inhibition of an enzyme that consumes a unique substrate result in the accumulation of the unique substrate which is potentially toxic to the cell and the inhibition of an enzyme that produces a unique product to result in the starvation of the unique product which potentially cripple essential cell functions [20]. Thus, it is believed that chokepoint enzymes may be essential to the parasite and are therefore potential drug targets.


While treating disease like diabetes, obesity, cancer, HIV etc it is very important that the drug enables target specific action. This includes the fact that the drug wold act directly on the metabolic pathway in whole or he enzyme which is responsible for spreading the disease.

Many important drug target specific metabolic reactions has been discovered. Drug target identification based on «omics» is a very promising approach that has only recently become possible. The concept of choke points in a given network contributes effectively in the identification of the lethality/bottleneck (here potential drug targets) in a network. Since a high load on a certain enzyme means that a large number of shortest paths go through it, therefore indicating a position in the central metabolism, ranking choke points on the basis of load will move enzymes with a higher probability of biochemical lethality to the top of the candidate list.

A comparative study of choke points with the human metabolic network is essential to identify possible interference of the drugs with the human metabolism which might lead to side effects. It has to be kept in mind though, that presently a large number of genes have unidentified functions which could lead to erroneous prediction of choke points. For example, often drug targets are identified by a unique pathogen-specific metabolic activity, as in the case of reverse transcriptase in the case of HIV.

Hence, the study of theses choke points is very much in demand and had potential to act on large no of targets thereby giving favorable results.


Choke points are important points in a reaction; they are reaction which consumes/produces certain metabolites which play important role in a given reaction. In absence of these choke points an organism cannot survive. The current analysis includes only the completely annotated enzymes in each organism. Including all the available enzymes for the organisms, such as putative enzymes, may complete the analysis of the metabolic network. The extended graph-based choke point concept can facilitate drug discovery andranking choke points based on their load values may be a likely pointer to the lethality level of such potential drug targets in the network. Further study and comparative analysis of various metabolic networks based on our network model can be beneficial for in vivo and in vitro studies.

There are further aspects on which the list of potential drug targets can be narrowed down. The drug should adversely affect the parasite but not the human host which means that if the drug target has a homologous enzyme in human, it should not be essential or have differential inhibition in human. In other way, it can be said that potential drug targets should be expressed in the human stages of the parasite. The rapid emergence of multi-drug resistant strains of these potentially lethal pathogens calls for the identification of new targets. The discovery of new targets with help of choke point analysis may lead to a drug formulation that would be able to counteract the resurgence of these diseases.

Also with further studies we can prove choke points to be helpful in the discovery of important regions in the pathway along with a better approach towards understanding the system well.

Its can be a key in the research of large metabolites and can provide with extremely important information which were hidden and needed to be discovered. The two most promising concepts for pathway analysis focused here are closely related. Assessing metabolic systems by the set of extreme pathways can, in general, give misleading results owing to the exclusion of possibly important routes. A full assessment of the proposed listed steps will require intense further effort. It is to be expected that some experiments may be significant which will stimulate the next phase of amendments and refinements. As it stands, it is hoped to serve the scientific community as a starting point for further data collection and experimentation in concert with, and based on, pathway analysis. One of the most important step is to reduce the Probability of identifying false choke points which can be done by undergoing annotation. Studies say that, Choke Point analysis with chemo-genomic profiling (microarray data), can Provide a complete and better annotation in vivo.


[1] Grigorov, M.G. (2005) Global properties of biological networks. Drug Discov. Today, 10, 365–372.

[2] Hartwell, L.H., et al. (1999) From molecular to modular cell biology.

Nature, 402, C47–52.

[3] Ravasz, E., et al. (2002) Hierarchical organization of modularity in metabolic networks. Science, 297, 1551–1555.

[4] Arita, M. (2004) The metabolic world of Escherichia coli is not small. Proc. Natl Acad. Sci. USA, 101, 1543–1547.

[5] Barabasi, A.L. and Oltvai, Z.N. (2004) Network biology: understanding the cell’s functional organization. Nat. Rev. Genet, . 5, 101–113.

[6] Papin, J.A., et al. (2004) Hierarchical thinking in network biology: the unbiased modularization of biochemical networks. Trends Biochem Sci, . 29, 641–647.

[7] Girvan, M. and Newman, M.E. (2002) Community structure in social and biological networks. Proc. Natl Acad. Sci. USA, 99, 7821–7826.

[8] Jeong, H., et al. (2000) The large-scale organization of metabolic networks. Nature, 407, 651–654.

[9] Ma, H. and Zeng, A.P. (2003) Reconstruction of metabolic networks from genome data and analysis of their global structure for various organisms. Bioinformatics, 19, 270–277.

[10] Oltvai, Z.N. and Barabasi, A.L. (2002) Systems biology. Life’s complexity pyramid. Science, 298, 763–764.

[11] Schuster, S., et al. (2000) A general definition of metabolic pathways useful for systematic organization and analysis of complex metabolic networks. Nat. Biotechnol, . 18, 326–332.

[12] Rahman, S.A., et al. (2005) Metabolic pathway analysis web service (Pathway Hunter Tool at CUBIC). Bioinformatics, 21, 1189–1193.

[13] Horne, A.B., et al. (2004) Constructing an enzyme-centric view of metabolism. Bioinformatics, 20, 2050–2055.

[14] Barrett, C.L., et al. (2005) The global transcriptional regulatory network for metabolism in Escherichia coli exhibits few dominant functional states. Proc. Natl Acad. Sci. USA, 102, 19103–19108.

[15] Covert, M.W., et al. (2004) Integrating high-throughput and computational data elucidates bacterial networks. Nature, 429, 92–96.

[16] Levchenko, A. (2003) Dynamical and integrative cell signaling: challenges for the new biology. Biotechnol Bioeng, 84, 773–782.

[17] Luscombe, N.M., et al. (2004) Genomic analysis of regulatory network dynamics reveals large topological changes. Nature, 431, 308–312.

[18] Ozbudak, E.M., et al. (2004) Multistability in the lactose utilization network of Escherichia coli. Nature, 427, 737–740.

[19] Yeh, I., et al. (2004) Computational analysis of Plasmodium falciparum metabolism: organizing genomic information to facilitate drug discovery. Genome Res, . 14, 917–924.

[20] I. Yeh, et al. (2004) Computational analysis of Plasmodium falciparum metabolism: organizing genomic information to facilitate drug discovery. Genome Res., 14: 917.

Annals of Botany. Vol.105, Issue 4, 2010. Pp. 505-511.

Structural colour and iridescence in plants: the poorly studied relations of pigment colour

Beverley J. Glover, Department of Plant Sciences, University of Cambridge, Downing Street, Cambridge CB2 3EA, UK

Heather M. Whitney, School of Biological Sciences, University of Bristol, Woodland Road, Bristol BS8 1UG, UK


Background Colour is a consequence of the optical properties of an object and the visual system of the animal perceiving it. Colour is produced through chemical and structural means, but structural colour has been relatively poorly studied in plants.

Scope This Botanical Briefing describes the mechanisms by which structures can produce colour. In plants, as in animals, the most common mechanisms are multilayers and diffraction gratings. The functions of structural colour are then discussed. In animals, these colours act primarily as signals between members of the same species, although they can also play roles in camouflaging animals from their predators. In plants, multilayers are found predominantly in shadeplant leaves, suggesting a role either in photoprotection or in optimizing capture of photosynthetically active light. Diffraction gratings may be a surprisingly common feature of petals, and recent work has shown that they can be used by bees as cues to identify rewarding flowers.

Conclusions Structural colour may be surprisingly frequent in the plant kingdom, playing important roles alongside pigment colour. Much remains to be discovered about its distribution, development and function.

Key words Diffraction grating, flower colour, interference, iridescence, multilayer, photoprotection, pollinator attraction, structural colour


The bright colours of flowers attract pollinating insects by making the floral tissue stand out against a background of vegetation. Analyses of insect visual acuity have shown that vegetation is visually very similar to bark, soil and stone from an insect’s point of view, because all these materials weakly reflect light across the whole range of the insect visual spectrum (Kevan et al., 1996).

Flowers are different — they appear as bright colours because they selectively reflect certain wavelengths of light, which are perceptible to pollinating animals, and, usually, to humans as well.

Colour is a property of both the coloured object and the perception of the animal observing it (Fig. 1). Light arriving at an object can be transmitted through it, absorbed by it or reflected back from it. If an object reflects or transmits all wavelengths of light equally, then it is perceived as white (Fig. 1, top). If an object strongly absorbs all wavelengths of light, then it is perceived as black (Fig. 1, centre). However, if it absorbs all light except one set of wavelengths, such as the red, which it instead reflects or transmits, then it can be said to have a colour. What that colour is depends on the visual system of an animal observing the object. If it has photoreceptors that are strongly activated by red light, as vertebrates do, then the object will appear red (Fig. 1, bottom left). If it has no photoreceptors that respond to red light, the object will appear black — to that animal the object is indistinguishable from an object that absorbs all wavelengths of light. Because photoreceptors are triggered by a curve of wavelengths the situation can be more complex. So, for insects that do not have red-light receptors but whose green-light receptors respond to a curve of wavelengths with the tail of the curve in the red part of the spectrum, the object in question would appear dull green.

Colour is a property of the light reflected by an object and the visual system of the animal observing it. If a flower reflects all wavelengths of light, it is perceived as white (top). If it absorbs all wavelengths then it appears black (centre). However, if it absorbs all wavelengths apart from one region of the spectrum, it has a colour. The flower shown in the bottom panel reflects red light. To the vertebrate eye, which has red-light receptors, the flower appears red. However, to the bee eye, which has no red-light receptors but whose green-light receptors are weakly stimulated by red light, the flower appears a dull green.

Plants, like animals, achieve colour in two main ways. First, they use chemicalor pigment-based colour. Pigments are compounds which absorb subsets of the visible spectrum, transmitting and reflecting back only what they do not absorb and causing the tissue to be perceived as the reflected colours. Chlorophyll absorbs light in both the red and the blue parts of the spectrum, reflecting only green light, and causes leaves to appear green to humans. Similarly, a flower that humans perceive as red contains pigments which absorb yellow, green and blue light, leaving red light as the only wavelength visible to us which is reflected. Plant pigments have been thoroughly studied from a biochemical perspective, and their synthesis and regulation have also been characterized by molecular genetics.

However, both plants and animals have also been shown to produce structural colours. A structural colour occurs when different wavelengths of light are selectively reflected from a substance, with the remaining wavelengths transmitted or absorbed. The famous blue butterflies of the genus Morpho have wing scales which selectively reflect a narrow bandwidth of blue light, allowing other wavelengths to be transmitted through the wing (Fig. 2A). The wings accordingly look intensely blue to humans, even though they contain no blue pigments (Vukusic et al., 1999). Structural colour has been well characterized in animals, but very little studied in plants.

Structural colour and iridescence.

(A) The intense blue colour of the Morpho butterfly is due to reflection of light by multilayers.

(B) Multilayers generate iridescence by reflecting different wavelengths of light at different angles at each boundary between layers.

(C) Diffraction gratings consist of ordered parallel grooves at particular frequencies, like the cuticular striations on this tulip petal.

(D) An iridescent beetle (rose chafer, Cetonia aurata) visits an artichoke flower.


Chemical and structural colours have several different properties. They differ first in the intensity of colour that they produce. Pigments are generally not very good at absorbing all but a very few wavelengths of light. Instead, they absorb most light of a number of wavelengths, but allow quite a broad range of wavelengths to be reflected or transmitted. This results in colours which can appear dull or muted, as they consist of a mixture of different colours of light. In contrast, structural colours can appear very intense, as reflective structures can be very precise in the bandwidths that they reflect.

Chemical and structural colours also differ in the patterns that they can produce. Chemical colours are diffuse, and look the same from all angles. To produce patterns of colour, different pigments must be localized to different cells or areas of a tissue. Commonly occurring pigment patterns in plants include different coloured venation on petals, and spots of dark pigment acting as targets at the bases of petals, near the nectaries. Structural colours have the potential to generate shifting patterns of colour as the viewer moves, rather than across different regions of the tissue. Reflective structures can reflect one particular peak wavelength of light at one angle, and another peak wavelength at a second angle. Thus, as an animal moves its position relative to the structure it will see the object change from the first colour to the second colour. The phenomenon of appearing different colours when viewed from different angles is called iridescence, and it is a unique attribute of structural colour. Iridescence can cover a few or many different colours, and can be in regions of the spectrum visible to a variety of animals, including in the ultraviolet (UV).


The mechanisms capable of producing structural colour in animals were described by both Hooke and Newton in the 17th and early 18th centuries, and a large body of literature has subsequently been produced, much of which is covered in several recent reviews (Parker, 2000; Vukusic and Sambles, 2003; Doucet and Meadows, 2009). A very brief overview shows that structural colour can be produced by either incoherent or coherent light scattering.

Incoherent light scattering takes place when individual light-scattering structures are randomly separated from one another by an average distance that is large when compared with the wavelength of the light. The light-scattering structures differentially scatter different visible wavelengths, but in such a way that the phase relationship of the scattered wavelengths is random. Although most structural colour in animals is produced by coherent light scattering, the blue colouration in many amphibians is attributed to incoherent scattering (Bagnara et al., 2007), as is the blue colour of the sky.

The majority of structural colour, and all iridescence, in animals is produced by coherent light scattering, which occurs when the distribution of lightscattering elements, and the resulting phase relationship of reflected light waves, is precisely ordered. An ordered distribution of light scatterers can result in either constructive or destructive interference. If the phase difference between two waves is a multiple of exactly one full wavelength then the two waves constructively interfere with each other and there is a strong reflection of light at that particular wavelength. By contrast, if the phase of the reflected waves differs by half a wavelength, or an odd multiple of half wavelengths, then destructive interference occurs such that reflection of this wavelength is weak or absent.

The simplest type of coherent light scattering is that of thin-film interference, which gives the colour to soap bubbles and oil-slicked puddles. Thin-film interference occurs when two transparent layers of materials with different optical densities meet. The optical density of a material determines the extent to which light waves are slowed down as they pass through it. Light is also reflected at each side of the boundary between the two materials — both before and after passing through each individual layer. Optical density, the thickness of the material layer, and the angle and wavelength of the light all help to determine if the light reflecting from the bottom of a layer is in phase or out of phase with the light reflected from the top of the layer, which will in turn determine whether constructive or destructive interference occurs for each wavelength. Constructive interference for one wavelength and destructive interference for others results in the reflected light being of one colour. Multilayer reflectors that produce structural colour consist of ordered layers of these pairs of thin films layered in series, producing even stronger constructive interference for specific wavelengths and resulting in very pure, intense colours (Fig. 2B). The classic example of multilayered structural colour in animals is shown by the blue Morpho rhetenor butterfly, in which the multilayered structure on its wing scales produces a vivid blue colour of such intensity that it is said to have a visibility of up to half a mile (Vukusic et al., 1999; Vukusic and Sambles, 2003).

A diffraction grating consists of a reflective surface over which runs a series of ordered and precisely spaced parallel grooves (Fig. 2C). Some of the light that hits the surface is reflected as normal, but light that hits the grooves is diffracted — split into its component wavelengths — and each wavelength is reflected at a different angle. Light with longer wavelengths has a higher diffraction angle than light with shorter wavelengths, so the light separates into its component parts, producing the rainbow effect that can be easily seen over the surface of a CD. Several beetle and spider species have been found to produce iridescence through this mechanism (Parker and Hegedus, 2003; Seago et al., 2009).

Iridescence can also result from the presence of photonic crystals, which are ordered three-dimensional structures. The classic example of a photonic crystal is opal, which consists of tiny spheres of silica packed together. The diffraction of light through opal is determined by the size and regularity of the spheres, which in turn determines the colours shown. Three-dimensional structures generating iridescence have been found in a wide range of animals, including comb-jellies, several butterfly species, the feathers of a number of bird species and in the annelid Aphrodita sp. (Parker et al., 2001; Vukusic and Sambles, 2003; reviewed in Welch and Vigneron, 2007). The spines of Aphrodita species show a multicoloured iridescence that is caused by a structure of holes ordered in hexagonal crystal structure within the spines (Parker et al., 2001). Biological photonic crystals can vary greatly in both form and method of function.


Structural colour and iridescence have arisen multiple times in the animal kingdom, so it is hardly surprising that they are also found in plants. All the general mechanisms used by animals to produce structural colour are also used by plants. Like animals, plants produce structural colour by both coherent and incoherent scattering. Incoherent ‘Rayleigh’ scattering (by particles smaller than the wavelength of light reflected) has been found in a number of plant species. The wax deposits on blue spruce (Picea pungens) and chalk dudleya (Dudleya brittonii) scatter shorter wavelengths of light preferentially, resulting in a blue colouration to the leaves (Vogelmann, 1993).

Iridescence has been shown to be produced by both multilayers (Fig. 2B) and diffraction gratings (Fig. 2C) in plants. The first example of multilayered iridescence in plants was found in the lycophyte Selaginella. Two species of Selaginella, S. willdenowii and S. uncinata, produce a vivid blue–green iridescence on their leaves when growing in shade. In the first detailed study into the mechanisms of plant iridescence, Hébant and Lee (1984) found that Selaginella leaves had two layers in the outer cell wall of their epidermal cells. These layers, visible under transmission electron microscopy, were each approx. 80 nm thick, the predicted thickness to cause multilayer interference that would result in the observed iridescence. These two layers were not found in ordinary green Selaginella leaves grown under higher light conditions and lacking iridescence (Hébant and Lee, 1984). Other plants with iridescent leaves are also found in low light environments, and all produce a similar blue–green iridescence. Although the multilayers in Selaginella appear to be relatively simple, with only a few layers producing the iridescence, other plant species produce more elaborate structures. The outer epidermal cell walls of the iridescent ferns Danaea nodosa, Diplazium tomentosum and Lindsaea lucida have many repeated dense layers alternating with arcs of cellulose microfibrils. The layers are of the correct thickness to cause iridescence through interference in the young iridescent leaves, but these layers are missing in the older leaves, which show no iridescence. The angle of the cellulose microfibrils changes gradually through the alternating layers up to a total 180 ° rotation (Graham et al., 1993; Gould and Lee, 1996; Lee, 2007). The resulting helicoidal structure is remarkably similar to the helical stack of chitin microfibrils found in some iridescent beetle species and may therefore be an example of convergent evolution (Lee, 2007; Seago et al., 2009). Leaf iridescence can also be caused by multilayers within the protoplast, not just within the cell wall. In the fern Trichomanes elegans and the angiosperms Phyllagathis rotundifolia and Begonia pavonina, specialized plastids called ‘iridoplasts’ are found in the iridescent leaves. These iridoplasts are much flatter than chloroplasts, and the thylakoid stacks within them are in such close contact that they form layers that cause the interference of light, resulting in the iridescent blue colouration (Graham et al., 1993; Gould and Lee, 1996; Lee, 2007).

Multilayers generating iridescence are also found in the fruits of Elaeocarpus angustifolius and Delarbrea michiana, in this case arising from a structure called an ‘iridosome’. This is secreted to the region outside the cell membrane of fruit epidermal cells, and consists of layers of cellulose that are of the predicted thickness to cause interference colouration (Lee, 1991; Lee et al., 2000).

Diffraction gratings were identified in plants more recently, with the first report of their presence on the petals of species including Tulipa sp., Hibiscus trionum (Fig. 3A) and Mentzelia lindleyi (Fig. 3E) published in 2009. In these species the petal epidermal cells are elongated and flat and the overlying cuticle produces a series of long, ordered ridges with a periodicity that acts as a diffraction grating and splits the light reflecting from the surface into component wavelengths (Fig. 3B, C; Whitney et al., 2009a). The iridescence produced is often predominantly in the UV wavelengths, which, although invisible to the human eye, are easily visible to many animal pollinators including bees and birds. The cuticular striations creating floral iridescence can also occur in patterns overlying those caused by pigment colour (Whitney et al., 2009a, b).

Plant iridescence. (A) The inner part of the Hibiscus trionum petal has an oily iridescence overlying red pigmentation. (B) Scanning electron microscopy of this region shows that the cells overlying the red pigment are covered with a diffraction grating made from cuticular striations, although the cells over the white region are smooth. (C) When petal diffraction gratings are replicated in transparent optical epoxy, light reflected from the epoxy is not white but shows a range of colours. (D) The iridescent labellum of Ophrys speculum is thought to mimic the wings of female pollinators. (E) Mentzelia lindleyi is iridescent as a result of diffraction gratings, but the iridescence is only detectable in the beevisible UV region of the spectrum.

Flowers are also the site of the one example of a three-dimensional photonic structure that has been found in plants. The elongated hairs that cover the attractive bracts surrounding edelweiss flowers (Leontopodium nivale subsp. alpinum) have an internal structure that acts as a photonic crystal (Vigneron et al., 2005). The hairs are hollow tubes with a series of parallel striations around the external surface. Through diffraction effects, the hairs absorb the majority of the UV light, effectively acting as an efficient sun-block. A variety of other epidermal cell morphologies are also known to influence light capture and reflection in petals (Kevan and Backhaus, 1998).


Iridescence appears to have as varied a range of functions as it does methods of production in the animal kingdom. The recent review by Doucet and Meadows (2009) gives a clear overview of the functions of animal iridescence. The most frequent role of animal iridescence appears to be in visual communication. Iridescence can relay information about the animal’s species (Silberglied and Taylor, 1978), about its age if iridescence changes or deteriorates over time (Kemp, 2006; Bitton and Dawson, 2008), about sex, as in many species only one sex has iridescence (Rutowski, 1977), and nutritional status, as individuals with poor nutrition may lack the resources to produce very vivid colouration (Kemp and Rutowski, 2007). Iridescence has also been found to play an important role in mate choice in birds, butterflies and fish (KodricBrown and Johnson, 2002; Sweeney et al., 2003; Kemp, 2007), while the depth of the blue structural colour on the testicles indicates the degree of dominance within the troop of a male vervet monkey (Prum and Torres, 2004).

As well as providing information for other animals, structural colour has also been implicated in helping animals avoid detection by their predators, either by mimicry or by camouflage. Colourful reef fish are well camouflaged against the equally colourful corals, while tiger beetles blend a range of structural colours together to produce a matt camouflage (Schultz, 1986; Schultz and Bernard, 1989; Seago et al., 2009).


As with animals, structural colour in plants is important in both display and defence. However, in plants the targets of the display are not other plants but pollinating insects, and the defence may be against potentially damaging levels of light as well as animal predators.

The primary function of flower and fruit iridescence is likely to be the attraction of animals, particularly those species whose visual systems are attuned to iridescence for animal–animal communication. The fruits of Elaeocarpus and Delarbrea michiana (Lee, 1991; Lee et al., 2000) have an iridescence that is thought to enhance animal attraction. Iridescence has also been shown to attract pollinating insects. It has been believed for some time that iridescence is used by pseudocopulatory flowers (such as species of Ophrys, Fig. 3D) to mimic female insects visually, but we were able to show that iridescence can act as an ordinary, learnable cue, in the same way that flower colour or shape might (Whitney et al., 2009a). Foraging bumblebees were trained that iridescent targets (generated by an artificial diffraction grating) contained a reward, whether they had a basic pigment colour of purple, blue or yellow, and that non-iridescent targets in the same pigment colours did not. The bees learned the iridescent cue, and were able to use it when presented with red targets to identify correctly the rewarding ones. The diffraction gratings generating floral iridescence often occur in patterns overlying those caused by pigment colour (Whitney et al., 2009b), suggesting that they might enhance pigment-based learnable cues.

The ability of structural colour to reflect strongly in specific wavelengths is thought to provide photoprotection to leaves. The Rayleigh scattering shown by Picea pungens and Dudleya brittonii is thought to result in enhanced reflection of shorter wavelengths, and thus to give protection against UV damage (Vogelmann, 1993). Protection against UV is also thought to be the primary function of the photonic crystal hairs overlying the surface of the edelweiss bracts, which protect the reproductive tissues against the potentially mutagenic UV levels found at the altitudes where this plant grows (Vigneron et al., 2005). Photoprotection may also be the function of the blue multilayer iridescence produced by understorey plants such as Danaea nodosa, Diplazium tomentosum, Lindsaea lucida and Begonia pavonina. These plants are all adapted to low light conditions, and so might be at risk of photodamage if they encountered sunflecks or other high-intensity light. The iridescent blue leaves of Begonia pavonina recovered significantly more rapidly from light exposure than green noniridescent leaves, although no difference was found between the iridescent and non-iridescent leaves of Diplazium tomentosum (Lee, 2007).

In contrast, it has been hypothesized that the iridescence of Selaginella species might aid the capture of photosynthetically active wavelengths in low light conditions because the leaf iridescence may act as a natural anti-reflective coating. Such coatings (on glasses and cameras) use thin film structures, analogous to those found in the iridescent Selaginella leaf, to produce constructive interference for certain wavelengths, increasing transmission of those wavelengths, but a side-effect is that the wavelengths not transmitted are strongly reflected because of destructive interference. In the same way, the iridescence in Selaginella could enhance blue-light reflection while enriching red-light absorption (Hébant and Lee, 1984).


Our understanding of plant structural colour and iridescence lags some way behind the work in animals, perhaps because plant pigment biochemistry has been studied so successfully or perhaps because animal structural colours are so striking. It is not surprising that similar mechanisms to generate structural colour have evolved in both plants and animals, but it will be important in the years to come to establish the molecular mechanisms underlying the development of these structures, which are likely to be very different in organisms with such basic differences in body architecture. The identification of structurally coloured plant species that are amenable to a genetic or transgenic dissection of candidate genes will be necessary to allow such work to progress rapidly. Preliminary studies suggest that some members of the Compositae, a number of petaloid monocots and certain species of Solanaceae might represent good targets for molecular and developmental analysis. It is also apparent that plant structural colour has evolved to mediate plant responses to both biotic and abiotic factors. A primary role is for communication with animals, and structures are therefore likely to target colours visible to pollinating or predatory species. One immediate challenge is to investigate how many species show structural colour (or iridescence) restricted to the UV region of the spectrum, and therefore invisible to the human eye. Investigation of the UV reflectance of flowers pollinated by insects that are themselves iridescent might be fruitful, as the visual acuity of such animals is already entrained to shifting colours, rather than to static ones. Such a study will also provide an understanding of the evolutionary lability of structural colour, and of the extent to which it appears to have co-evolved in response to interactions with particular groups of insect. Given that we do not currently have a good understanding of which plants produce structural colour, how they produce it and what they produce it for, one of the most exciting aspects of plant structural colour is the amount that still remains to be learned.


We thank Murphy Thomas, Sean Rands, Matthew Dorling, Matt Box and Rosie Bridge for help with figure production, and the Insect Collection in the Department of Zoology, University of Cambridge, for access to iridescent animals. The ideas explored in this article were developed through many helpful discussions with Ulli Steiner, Mathias Kolle and Lars Chittka. H.M.W. is in receipt of a Lloyd’s of London Tercentenary foundation fellowship.

Calibrating the Tree of Life: fossils, molecules and evolutionary timescales

Félix Forest Jodrell Laboratory, Royal Botanic Gardens, Kew, Richmond, Surrey TW9 3DS, UK July 9, 2009.


Background: Molecular dating has gained ever-increasing interest since the molecular clock hypothesis was proposed in the 1960s. Molecular dating provides detailed temporal frameworks for divergence events in phylogenetic trees, allowing diverse evolutionary questions to be addressed. The key aspect of the molecular clock hypothesis, namely that differences in DNA or protein sequence between two species are proportional to the time elapsed since they diverged, was soon shown to be untenable. Other approaches were proposed to take into account rate heterogeneity among lineages, but the calibration process, by which relative times are transformed into absolute ages, has received little attention until recently. New methods have now been proposed to resolve potential sources of error associated with the calibration of phylogenetic trees, particularly those involving use of the fossil record.

Scope and Conclusions: The use of the fossil record as a source of independent information in the calibration process is the main focus of this paper; other sources of calibration information are also discussed. Particularly error-prone aspects of fossil calibration are identified, such as fossil dating, the phylogenetic placement of the fossil and the incompleteness of the fossil record. Methods proposed to tackle one or more of these potential error sources are discussed (e.g. fossil cross-validation, prior distribution of calibration points and confidence intervals on the fossil record). In conclusion, the fossil record remains the most reliable source of information for the calibration of phylogenetic trees, although associated assumptions and potential bias must be taken into account.

Key words Calibration, fossil, incompleteness, molecular dating, rate heterogeneity, relaxed molecular clock, uncertainty


The use of DNA sequences to estimate divergence times on phylogenetic trees (molecular dating) has gained increasing interest in the field of evolutionary biology in the past decade. The abundance of publications on the subject, the numerous alternative methods proposed and the often heated debates on various aspects of the discipline demonstrate the interest it generates. The molecular clock hypothesis was first proposed by Zuckerkandl and Pauling (1965); they proposed that differences in DNA (or protein) sequences between two species are proportional to the time elapsed since the divergence from their most recent common ancestor.

The subsequent inclusion of temporal frameworks in many evolutionary studies has influenced the way results are interpreted and significantly modified the way in which conclusions are drawn from these findings. Linking the evolution of particular morphological characters or key ecological innovations to geological, climatic or biotic events is much improved in the light of an evolutionary timescale. The development of molecular dating tools became particularly valuable to the discipline of historical biogeography; it added a temporal gauge to the directionality of events demonstrated by the topology of phylogenetic trees. Inferences on observed distribution patterns were rendered significantly more plausible under a temporal framework, even if only descriptive. Furthermore, new methods of biogeographical reconstruction have been developed such as Lagrange, which uses a likelihood framework to infer the evolution of geographical ranges and incorporates divergence times as well as constraining the connections between areas to specific times (Ree and Smith, 2008).

The rationale of the molecular clock hypothesis, that evolutionary rates are constant, was shown to be invalid in the majority of examined cases; the clock does not tick regularly. The heterogeneity of substitution rates among different lineages in a phylogenetic tree explains this irregularity (Britten, 1986) and is a result of species-specific factors such as generation time, metabolic rate, effective population size and mutation rates (see Rutschmann, 2006). The extent of influence of some such factors, however, remains in dispute (e.g. Whittle and Johnston, 2003).

Rutschmann (2006) classified the most commonly employed methods for estimating divergence times into three categories depending on how they handle rate heterogeneity, namely (1) assuming a global substitution rate (standard molecular clock); (2) correcting for rate heterogeneity (e.g. by deleting branches or incorporating several rates categories before the dating procedure), and (3) incorporating rate heterogeneity (i.e. integrating rate heterogeneity into the dating procedure using rate change models; relaxed molecular clock). The four most commonly used methods in the literature all fall into the third category; these are non-parametric rate smoothing (NPRS; Sanderson, 1997), penalized likelihood (PL; Sanderson, 2002), the Bayesian method implemented in the Multidivtime package (Thorne et al., 1998) and Bayesian evolutionary analysis by sampling trees (BEAST; Drummond and Rambaut, 2007). The first three of these methods assume rate changes between ancestral and descendant lineages are autocorrelated, i.e. that substitution rates in descendant lineages are to an extent inherited from ancestral lineages; these methods differ in the way that rate autocorrelation is handled. BEAST does not assume rate autocorrelation; instead, it samples rates from a distribution. Additional flexibility is found in BEAST in its optional tree topology requirement that can incorporate phylogenetic uncertainty, and the possibility of assigning distributions to the calibration process a priori (see below). More details on these methods and several others are available elsewhere (Rutschmann, 2006, and references therein).

Two main topics have fuelled the controversy associated with molecular clocks: these are how to handle rate heterogeneity and calibration. At its outset, the field of molecular dating was focused on circumvention of rate heterogeneity among lineages. Meanwhile calibration, the process by which relative time is transformed into absolute age (e.g. million of years) using information independent of the phylogenetic tree and its underlying data, was somewhat trivialized. This situation has changed in recent years and many studies have now addressed the numerous difficulties associated with calibration. This paper is focused on molecular clock calibration (particularly based on palaeontological data), the potential problems and source of error associated with it, and the various methods proposed to incorporate these uncertainties in molecular estimates of divergence times.


Information used to calibrate a phylogenetic tree is obtained from three principal sources: (1) geological events; (2) estimates from independent molecular dating studies; and (3) the fossil record. Information from palaeoclimatic data has also been used to calibrate trees (e.g. Baldwin and Sanderson, 1998), but its use is limited and will not be discussed further here. The fossil record is the most commonly employed source of information to calibrate phylogenetic trees and will receive most attention here.

Plate tectonics, the formation of oceanic islands of volcanic origin and the rise of mountain chains are examples of geological events that can be used to calibrate phylogenetic trees. The assignation of such calibration points to a given node assumes that the divergence at this node is the result of this new geographical barrier, through either vicariance (e.g. continental split) or dispersal (e.g. oceanic islands) events. This type of calibration must be used with care in studies examining biogeographical patterns to avoid circular reasoning. Despite appearing less prone to imprecision than the use of the fossil record, geological events have their own suite of potential and often intractable problems. The timings of continental splits are often reported as unique values, but the actual separation of two continental plates occurred over millions of years (and is a continuous process). Furthermore, as two land masses drift apart, biological exchanges between them are likely to continue for several million years depending on the dispersal abilities of the organisms involved. These two points render the use of continental splits as calibration points a choice rather difficult to justify. Similar problems can be attributed to the rise of mountain chains; these phenomena take place over a long period of time (several tens of millions of years in the case of the Andes; e.g. Garzione et al., 2008) and exchanges between each side of a new geographical barrier will continue for some time.

Species endemic to oceanic islands of volcanic origin and therefore of known age can be used to apply a maximum age constraint on the divergence between the endemic species and their closest continental relatives. This approach accounts for the likelihood that the ancestor of the island endemic species arrived at an unspecified time after the formation of the oceanic island. Present-day oceanic islands, however, might only be the most recent element of a series of oceanic island formation over time in a particular region (Heads, 2005), which would invalidate their use as reliable calibration constraints. For example, molecular dating of Galapagos endemic iguanas shows that their dispersal to the archipelago pre-dates the age of the current islands (Rassmann, 1997). Submerged islands found in the vicinity of the present day islands suggest that the Galapagos archipelago is in fact much older (10–15 Mya to 80–90 Mya) than the extant islands (Hickman and Lipps, 1985; Christie et al., 1992). A similar situation is observed in Hawaii, leading some to describe evolutionary history on this archipelago as on a ‘volcanic conveyer belt’. Calibration based on species endemism to volcanic islands can be difficult to justify and caution is advised.

The use of estimates derived from independent molecular dating studies (also referred to as secondary or indirect calibration points) is the only source of calibration information for many groups, particularly for those in which the fossil record is scarce or non-existent. The primary problem with this approach is that sources of error generated by the first dating exercise remain and are propagated and likely to be magnified in subsequent analyses. The use of secondary calibration points should be a last resort and, when used, care should be taken to include error associated with the primary molecular estimate in the subsequent analysis (e.g. using confidence intervals or standard deviation as minimum and maximum values on a given node, or using a prior distribution; see below). Failure to take this error into account can result in estimates of divergence times with broader uncertainty, and thus of little use or scientific value. The use of substitution rates from independent studies to calibrate a phylogenetic tree also falls under this category of calibration (e.g. Richardson et al., 2001) and suffers drawbacks similar to those described above.

There is general consensus that the fossil record provides by far the best information with which to transform relative time estimates into absolute ages (e.g. Magallón, 2004). As with other sources of calibration information, the use of fossilized remains has disadvantages and is subject to various sources of errors. Nevertheless, promising methods recently proposed attempt to tackle these issues. The focus of the following discussion is on calibration using palaeontological data, but many of the aspects addressed below are also applicable to geological events and secondary calibration points.


Sources of error in molecular inference of divergence time are numerous, including phylogenetic uncertainty, substitution noise and saturation, rate heterogeneity (among lineages, over time and between DNA regions), incomplete taxon sampling and incorrect branch length optimization (e.g. Sanderson and Doyle, 2001; Magallón and Sanderson, 2005). The calibration process is not exempt from potential sources of error either; these include erroneous fossil age estimates, the incompleteness of the fossil record and the placement of fossils on phylogenetic trees. Although often difficult to circumvent, much progress has recently been made in mitigating these factors.

Generally, a taxon’s first appearance in the fossil record represents the time it became abundant rather than the time of its emergence (Magallón, 2004). Considering estimates from the fossil record as actual ages would underestimate the true age of the clade to which the fossil is assigned (Benton and Ayala, 2003; Conti et al., 2004). Older fossils assigned to a given group are likely to be discovered and to push back in time the earliest occurrence of a lineage; thus the age of a fossil is generally treated as a minimum constraint in calibration procedures (e.g. Benton and Ayala, 2003; Near et al., 2005). This means that the clade on which the constraint is applied cannot be younger than the fossil.

Fossil remains can be dated by use of stratigraphic correlations or radiometric dating. Uncertainty is introduced here as a result of any unreliability in the age assessment itself and the imprecision of the estimate when the fossil is assigned to a particular geological division (or stratum). For example, a fossil assigned to the Palaeocene can theoretically have any age between 55·8 and 65·5 Mya. Any time assigned to a calibration point within this epoch would be technically appropriate, but would result in significantly different estimates for the other nodes in the tree. To counter this, because the fossil represents a minimum age, it is preferable to use the upper boundary of the geological division (in this case 55·8 Mya) in a molecular dating study, once again as a minimum constraint. Some programs permit specification of a prior distribution on the age of a node which takes into account the uncertainty associated with the dating of a fossil (Drummond and Rambaut, 2007; see below).

The fragmentary nature of the fossil record and lineage extinction have important consequences for the accurate placement of fossil calibration points. Once a fossil has been accurately assigned to a group of extant taxa based on one or more synapomorphies, it is placed on the phylogenetic tree either with the stem group or with the crown group (Fig. 1). The crown group comprises all the extant taxa of a clade and their most recent common ancestor plus all the extinct taxa that diverged after the origin of the most recent common ancestor of the living taxa. The stem group comprises all the members of the crown group (extinct and extant) plus all the extinct taxa that diverged since the split of the crown group from its closest living relative (Fig. 1). In any rooted phylogenetic tree, all internal nodes are both stem group nodes and crown group nodes; the definition of stem and crown group nodes is relative to the other nodes in the tree (e.g. in Fig. 1, node 2 is the crown group node of clade B and the stem group node of clade A). Because the fossil record is fragmentary, one can never be certain that a given fossil will possess features that place it in the crown group rather than along the stem lineage leading to the crown group. Consequently, there can be large and difficult to quantify discrepancies between the time of divergence of a lineage, the time of appearance of a synapomorphy (a particular feature characterizing a clade) and the age of the oldest known fossil exhibiting this feature (Magallón, 2004; Fig. 1). This highlights the importance of taking the most conservative options (i.e. options resistant to subsequent changes that would invalidate the assumptions regarding the position of a fossil) in calibration by use of fossils as minimum constraints on the stem group node.

Ultrametric tree (in which branch lengths represent time) of extant taxa (solid branches) with the placement of fossil taxa (dashed branches). The designation of a node as either a crown group or stem group is relative; node 1 is the crown group node of clade A; node 2 is the stem group node of clade A, but also the crown group node of clade B; and node 3 is the stem group node of clade B. Fossil X has been assigned to the stem lineage leading to clade A based on one or more shared synapomorphies. If this fossil was to be used as calibration point, its age would be assigned as minimum constraint on the stem group node of clade A (node 2). Placing the calibration point on the crown group node of clade A (node 1) would result in an overestimation of molecular ages. The use of fossils as minimum constraints takes into consideration the incompleteness of the fossil record; should older fossils be discovered (such as fossil ‘?’), the minimum constraint placed on node 2 using this fossil would not be invalidated.

An exception to the rule of using fossils as minimum constraints can be applied to fossilized pollen grains. Pollen grains have a much higher fossilization potential than any other plant organs, but not all plant groups will have an extensive pollen fossil record or possess palynological features assigned with confidence to extant taxa. Tricolpate pollen grains (those with three apertures or colpi), for example, are unique to the eudicots in plants, and age estimates for these fossils place them in the Barremian and Aptian of the early Cretaceous (130–112 Mya; e.g. Doyle and Hotton, 1991); earlier occurrence is thought to be very unlikely. The abundance and widespread distribution of early tricolpate pollen fossils coupled to their easily identified features has led to their frequent use as a maximum constraint or fixed age in molecular dating of angiosperms (e.g. Anderson et al., 2005; Magallon and Castillo, 2009). It is only in such rare cases that fossils can be used as maximum constraints or fixed ages without serious risk of underestimating molecular ages.

The incompleteness of the fossil record also leads inevitably to the underestimation of node ages in a phylogenetic tree (Springer, 1995), presenting significant discrepancies between estimates obtained from the fossil record and molecular dating (e.g. Benton and Ayala, 2003). The selectivity of fossilization is largely responsible for this situation. Different plant groups (e.g. deciduous anemophilous trees are better represented in the fossil record than entomophilous/ zoophilous herbs) and structures (e.g. pollen is more easily preserved than flowers) have different preservation potential (Herendeen and Crane, 1995); thus the fossil record is biased towards groups and structures more conducive to fossilization.

Several methods have been developed by which to estimate the extent of incompleteness of the fossil record. Earlier studies proposed statistical approaches to calculate confidence intervals on stratigraphic ranges; the earliest occurrence of a given group in the fossil record is estimated by use of the number of known fossils and the number and size of gaps in the stratigraphic column (Strauss and Sadler, 1989; Marshall, 1990, 1994). Because these methods do not take into account the quality and density of the fossil record, Marshall (1997) proposed an additional function that allows for bias linked to collecting and preservation potential. Subsequently, Foote and colleagues (e.g. Foote, 1997; Foote et al., 1999) estimated rates of extinction, origination and preservation from the fossil record to produce a measure of completeness [see Magallón (2004) for more details on these methods]. Tavaré et al. (2002) proposed a method based on an estimate of the proportion of preserved species in the fossil record and the diversification patterns of the group. More recently, Marshall (2008) developed a quantitative approach to estimate maximum age constraints of lineages using uncalibrated ultrametric trees (i.e. with relative branch length optimization) and multiple fossil calibration points. Assessing the fossil record of a group using the procedures outlined above would theoretically produce a realistic age estimate for this group. Furthermore, the resulting estimate of earliest occurrence can be used as a fixed age or maximum constraint in subsequent molecular dating studies, minimizing the uncertainty associated with the age of a given fossil.


Many early molecular dating studies used a single fossil as a calibration point; this practice is now believed to lead to strong bias in molecular age estimates (e.g. Graur and Martin, 2004; Reisz and Müller, 2004). Where possible, it is currently advocated that multiple fossils should be used in the calibration process (e.g. Conroy and Van Tuinen, 2003; Graur and Martin, 2004; Forest et al., 2005; Near et al., 2005; Benton and Donoghue, 2007; Rutschmann et al., 2007), although an extensive and reliable fossil record is not always available. However, if the intrinsic accuracy of a fossil is questionable (i.e. doubtful age estimate assigned to an extant group or representing a lineage with a large gap between its divergence and the first appearance of remains in the fossil record), it is better excluded from the analysis (see above; Near et al., 2005). Near et al. (2005) proposed a fossil cross-validation procedure that allows potentially inaccurate fossils to be identified when multiple fossils are used to calibrate a phylogenetic tree. This method compares the molecular age estimates produced by the calibration of the phylogenetic tree with one of the fossils with the age estimates from the fossil record for the other nodes used in the calibration procedure. Individual fossils that produce age estimates inconsistent with the remainder of the fossils used as calibration points are removed (Near et al., 2005) and the analysis is repeated, but including only reliable fossils as calibration points. Rutschmann et al. (2007) built on the fossil cross-validation method of Near et al. (2005) to address another problem: the multiple potential positions of a given fossil on the phylogenetic tree. They assessed the effect of the alternative positions of each fossil on the consistency of the age estimates in a set of calibration points. This allows the selection of the best position on the phylogenetic tree for each fossil given an a priori selected set of assignment possibilities. By use of this approach, no calibration information is removed from the dating procedure (contrary to the method of Near et al., 2005, in which ambiguous fossils are removed), and the method provides more precise estimates because the most coherent calibration sets produce lower standard deviation (Rutschmann et al., 2007). Concern has been raised that such cross-validation methods might lead to the exclusion or repositioning of fossils that are not necessarily misleading, but rather misinterpreted (e.g. placement, dating) or victims of bias in the dating procedure itself (Hugall et al., 2007; Parham and Irmis, 2008; Lee et al., 2009).

Lee et al. (2009) recognize the usefulness of the two methods mentioned above, but note that the imprecision surrounding the phylogenetic position of a given calibration is not calculated in these methods. They propose a new method that integrates this uncertainty. This approach allows the inclusion of fossils in a combined matrix of morphological and molecular characters analysed under a Bayesian framework, and the assessment of estimates among sampled trees based on the position of the fossil in each particular tree as determined by the analysis (Lee et al., 2009). They demonstrate that the uncertainty associated with the phylogenetic position of a fossil used as calibration point can result in molecular age estimates with large confidence intervals (Lee et al., 2009).

While the three above methods deal with uncertainty associated with the phylogenetic position of fossil calibration points, another recent method implemented in the program BEAST (Drummond and Rambaut, 2007) allows the user to include in addition a level of uncertainty on the age of a given fossil using a prior distribution. The prior distribution of the age is assigned to the most recent common ancestor of a group of taxa circumscribed by the user (Drummond et al., 2006); these prior distributions are various (e.g. normal, lognormal or uniform). Applying a normal distribution would assume that the age of this node is equally likely to be older or younger than the fossil, how much so being determined by the standard deviation specified. A normal prior distribution is more appropriate for calibration points based on estimates from independent molecular studies (secondary calibrations) and geological events such as oceanic islands, in which it can be argued that uncertainty is equally distributed on either side of the age used (the mean of the distribution). For a calibration point based on fossil remains, a lognormal prior distribution covering a longer period of time towards the past is more appropriate, allowing for uncertainty of the age estimate of the fossil and for error associated with the incompleteness of the fossil record. The method implemented in the program BEAST provides a significant improvement over other methods, particularly as it considers uncertainty associated with tree topology and calibration.


In recent literature, some authors have voiced their concerns regarding molecular dating methods in general and the calibration procedure in particular (e.g. Graur and Martin, 2004; Heads, 2005; Pulquério and Nichols, 2007). The importance of a carefully designed calibration scheme in a molecular dating study cannot be overemphasized; it is one of the most fundamental aspects of the methodology. The identification of reliable fossils is a crucial step in this procedure, but finding unequivocal fossils may prove to be a tedious task in some plant groups. The lack of fossils in a given group may prevent the use of molecular dating completely. However, fossils from closely related groups may be used as calibration points if taxa representing them are included in the phylogenetic tree, but the further the calibration point is positioned in relation to the node(s) of interest, the greater will be the uncertainty of the resulting age estimates. A good starting point in the search for fossil calibration points is the Plant Fossil Record online database maintained by the International Organisation of Palaeobotany which contains several thousand extinct taxa from both modern and extinct genera.

The development of methods addressing the potential problems affecting calibration, particularly based on fossil data, has elegantly addressed some of the criticisms mentioned above and provided new opportunities and tools for more reliable calibration of phylogenetic trees. The program BEAST (Drummond and Rambaut, 2007) is one of the most promising methods on account of its flexibility regarding uncertainty in fossil age estimates, mainly due to the dating of the fossil and the incompleteness of the fossil record. The next phase in software development for molecular dating would include programs allowing better estimation of uncertainty by incorporating fossil cross-validation procedures, taking into consideration fossil abundance data and integrating the calculation of confidence intervals on the fossil record.

The assumptions and bias inherent to aspects of the methodology are not the only obstacles to reliable and plausible timescales; dating results must be viewed in light of the information that was used to obtain them, and uncertainty around resulting age estimates must be considered. Molecular dating is a powerful tool and its use continues unabated because it offers a tantalizing and otherwise unavailable glimpse into the evolutionary history of a group.


I thank E. Lucas and two anonymous assistants.

incial and Local Units. Bloomington, IN: Indiana University Press.