I've always found palaeontology fascinating, probably due to my love of dinosaurs from an early age. These days I have a different appreciation of palaeontology because of the fact that development evolves. I've met several people who believe that evo-devo (evolutionary developmental biology) is all about genomics and has little to do with fossils these days. I don't think that's quite right.
Evolutionary developmental biology (evo-devo) is a discipline concerned with evolution, development, and the interplay between the two. An adult-centric view of evolution leaves some people imagining that changes in the form of an adult ancestor leads to a different adult descendent. But adults do not directly evolve into different types of adults. Instead, evolution of adult form takes place when egg-to-adult development is altered within an evolving lineage. The fact that development itself evolves has implications for our understanding of development itself, but also evolutionary theory. Evolution of developmental processes results in new morphologies, introduces developmental constraints or biases in evolution, and creates complex problems for our understanding of potentially homologous characters.
Many of the most important discoveries in evo-devo have occurred in the last twenty years, thanks largely to advances in molecular biology. It is common to find evo-devo research focusing on the evolution of developmental genetics and working with extant model organisms, but throughout its history evo-devo has also relied on the study of embryology, morphology and palaeontology. It is often said that evo-devo truly begun in the 1980s with the discovery of the homeobox genes. New doors may have opened, but scientists had considered and debated the interplay between evolution and development for over a century.
In Darwin’s time there were arguments that changes to development could play a role in evolution. In The Origin of Species (1859), Darwin describes evolution taking place due to slight changes introduced at different stages of development and in different parts of the body over millions of years. This point of view was repeated by other early Darwinists, including Thomas H. Huxley in Evidence as to Man's Place in Nature (1863). Clearly the connection between development and evolution isn't a new idea. In an 1860 letter to American botanist Asa Gray, Darwin wrote:
"Embryology is to me by far the strongest single class of facts in favour of change of form, & not one, I think, of my reviewers has alluded to this."
Molecular biology was non-existent for most of this history. The most obvious evidence of morphology evolving could be found in the fossil record. Palaeontology has clearly been of vital importance to evo-devo long before the discovery of the homeobox genes and the development of powerful tools that reveal the mechanisms of developmental genetics. A quick glance at the most recent evo-devo papers could suggest that the genetic and genomic approach has usurped palaeontology in evo-devo; that it had its time and was the best we could work with before molecular genetics. In reality, palaeontology is vitally important in today's evo-devo research and even aids the molecular and genomic approaches. There are some questions that may never have been answered relying on palaeontology alone, but the same can be said for the molecular approach. Evo-devo is a multidisciplinary endeavour.
Despite being incomplete, the fossil record provides our only direct window to the body plans of extinct taxa. Without palaeontology we would not know of the existence of huge and diverse clades of organisms including the trilobites, the dinosaurs, and the bizarre body plans observed in the Cambrian strata. In rare cases palaeontology can actually provide genomic information too. A Neanderthal genome project is under way based on well-preserved fossil finds. Comparing the genomes of Homo sapiens and Pan troglodytes can reveal changes that have occurred on our lineage since the divergence from one another but that covers several million years of evolution. It's thought that information from the Neanderthal genome will identify which of these changes occurred since diverging from the Neanderthal lineage and may have played a role in making modern humans what they are today.
It is often claimed in the literature that preserved DNA will be impossible to recover in fossils older than approximately 100,000 years but the current record goes to a 1.12-times coverage draft genome from a horse bone dated to 560,000-780,000 years old. Unfortunately, this type of DNA preservation is unlikely to occur for extinct taxa that are of interest in many key evo-devo topics as they are usually hundreds of millions of years older. The remarkable ancient genomes being sequenced are very rare exceptions as the genomes of 99.9% of extinct organisms are forever lost. So if the fossil record can't provide many genomes for evo-devo scientists to work with, what can it provide instead? Is palaeontology simply the old-fashioned way to do evo-devo?
The fossil record can improve phylogenies by providing evidence of features along lineages of extant and extinct taxa. These fossils can fill gaps in phylogenies and reveal ancestral relationships that would otherwise by unknown. Phylogenies are concerned with the identification of speciation events and diverging lineages evolving independently but carrying homologous genes and other characters. The study of development (including its evolution) and phylogeny feed into each other. Our phylogenies inform our understanding of how development evolves, and discoveries in evo-devo can provide more informative data for use in phylogeny reconstruction.
The data for many modern phylogenies typically comes from extant taxa and statistical methods are used to predict tree topology and patterns of ancestry. But understanding the distribution of characters in extinct lineages leading to extant taxa allows us to polarise evolution and recognise patterns of evolution for features relevant to evo-devo. Understanding which lineages a character occurs in can inform our phylogenies and allow targeting of more informative extant taxa for analysis. Palaeontology can also help identify whether characters in extant taxa are homologous or convergently evolved, which is useful for phylogenies and evo-devo in general.
Molecular clock calibration
An important aspect of the interplay between development and evolution is the timing of evolutionary events including speciation events resulting in the origin of new taxa, the timing of characters gain and loss in extinct taxa, and extinction events. Some taxa have a tendency to be relatively well-preserved and are found in several strata. These are easier to date based on the rocks they are found in. But for taxa with incomplete fossil records, a molecular clock is required to date important times of divergence.
As useful as molecular clocks have been for taxa poorly represented in the fossil record, they still require palaeontological evidence as molecular clocks are calibrated using the divergence times of well-preserved fossil taxa (or sometimes geological events). There are four downsides to this approach:
We are unlikely to find the earliest member of a clade so statistical methods are used to infer the error limits of divergence.
This approach assumes that our phylogenies are correct. If tree topology is incorrect then nodes can shift forwards or backwards in time and ruin the calibration.
Fossils lie in sedimentary rocks, which cannot be radiometrically dated. Instead, the igneous rocks immediately above or below the sedimentary rocks are used. There is an uncertainty involved in correlating the sedimentary and igneous rocks that can result in over- or under-estimation of the divergence times.
The genes used must be orthologues, not paralogues, found in two taxa for which the fossil divergence time is estimated. The number of nucleotide changes that have occurred since divergence are calculated. This can be complicated by differences between taxa in the rates of nucleotide substitution and the possibility that the genes being compared are paralogues that evolved from orthologues after a divergence. An extremely careful choice of genes to be used is essential.
Some people assume that fossils and molecular data go hand-in-hand in other areas of evolutionary biology but not in evo-devo. In the fossil record, change over generations and between taxa is clear, but useful developmental evidence is rare. Nevertheless, palaeontology can be useful for evolutionary developmental biologists when constructing hypotheses using molecular data from living taxa as fossil evidence can constrain molecular hypotheses. Some studies on the development of digits in extant birds have suggested that their three digits are 1, 2 and 3 while others have suggested they are digits 2, 3 and 4. Fossils of extinct bird lineages and theropod dinosaurs constrain the molecular predictions to 1, 2 and 3. Fossils can help developmental biologists make sense of molecular data.
Despite developmental evidence being rare in the fossil record, it isn't non-existent. The development of individual fossil organisms is impossible to study but life histories have been described for some well-preserved fossil taxa including many trilobite species. Even fossils of ancient larvae and embryos are occasionally found, though these finds are typically limited to specific taxa and eras. With advanced imaging techniques, informative characters of these larvae and embryos may be discovered and be of use in evo-devo. New imaging techniques are uncovering fascinating details from adult fossils too. In 2012 Ma et al described the oldest tripartite brain in the fossil record, belonging to a stem-group arthropod that lived approximately 520 million years ago. Assuming reasonable preservation in the depth of a specimen, computed tomography or computed synchrotron and phase-contrast radiation X-ray tomography can be used to resolve soft tissue preservation in ancient fossils. In 2013, Tanaka et al were able to analyse the neuroanatomy of Alalcomenaeus sp., a Cambrian “great appendage” arthropod (see the image below). This research helps confirm Alalcomenaeus as a stem-group chelicerate. State-of-the-art imaging of Alalcomenaeus and Fuxianhuia protensa reveal that the brain configurations observed in the extant Chelicerata and Mandibulata had already evolved by the early Cambrian. With more advanced imaging techniques, important questions in evo-devo may be addressed by obtaining data from ancient nervous systems, larvae, and embryos.
Specific questions regarding origins
A layman's view might be that questions regarding ancient extinct taxa require answers from palaeontology and that developmental genetics and other molecular approaches are of more use when studying extant taxa. This is not always the case. Two of the most debated topics in evo-devo concern the origin of the metazoans and the description of the last bilaterian common ancestor (LBCA). These topics address truly ancient events and organisms but the fossil record is unfortunately incomplete and often uninformative on these topics. The best attempts so far at addressing these issues have involved reconstructing hypothetical ancestors using data from extant species by comparing genomes and developmental mechanisms, and using molecular data to reconstruct more accurate phylogenies.
Most reconstructions of features in important extinct basal taxa have relied heavily on the developmental genetics of living taxa. Comparisons of protostome and deuterostome genomes and development have generated most hypotheses regarding the LBCA. This is not a perfect approach, as there is a danger of mistaking homoplasy or homology. If a protostome and deuterostome possess homologous developmental genes used in patterning a shared structure, does that mean the structure is homologous? There is always the possibility that the genes themselves have been independently co-opted in both lineages.
The other approach, in the absence of desired fossils, is to use extant basal taxa as proxies when investigating major divergences in metazoan phylogenies. Sponges are often considered the most basal metazoans. Acoel flatworms are often considered the most basal bilaterians. But how representative are they? Sponges and acoel flatworms have been evolving along their own lineages for hundreds of millions of years. There is a risk that their genomes and morphologies may be too derived to be informative. If sponges possess genes or gene networks used by other metazoans for more complex development, it isn't clear whether or not the ancestor of all metazoans used the genes in a similar way to extant sponges. The development of extant sponges could be quite derived compared to the ancestor of all metazoans. Better fossil evidence could help determine how useful various extant basal taxa are as proxies.
Palaeontology often provides its most informative discoveries when highly unusual fossils are found. The Burgess Shale sponge Eiffelia globosa contains both hexaradiate and tetraradiate spicules, typical of calcarean and hexactinellid sponges, respectively. This information, combined with the molecular data suggesting that extant sponges are paraphyletic, allows us to learn more about sponge ancestry than was available through genomic information alone. Perhaps understanding ancient animal origins could be within our grasp if we were fortunate enough to find more unusually informative fossils. A single fossil can have a large effect on our view of ancient relationships and the timing of events. Another example of an unusually informative fossil is that of Kimberella, which appears to be a complex bilaterian existing in the Ediacaran. This suggests that the bilaterian radiation may have preceded the Cambrian explosion, which many have associated with evolution of the bilaterians.
Much like evo-devo itself isn't a threat or replacement for the "modern synthesis" or ecological evolutionary biology, molecular techniques add to evo-devo rather than replace the role of palaeontology in the discipline. There are many questions in evo-devo that have only been answered due to palaeontology, and the same can be said for molecular techniques. Both approaches have proven their worth and both are used to aid each other. The role of palaeontology in evo-devo is as important today as it ever was, or perhaps more so with the use of the fossil record in calibrating molecular clocks, constraining molecular hypotheses, and providing increased taxon sampling for phylogenetic analysis. The very nature of evo-devo makes it a multidisciplinary endeavour with many areas of research complimenting each other. The way forward for both palaeontological and molecular approaches in evo-devo is clearly together. Being obsessed with evo-devo, I'm as fascinated by fossils as I am genomes. I'm following the development of new imaging techniques as keenly as the next generation of sequencing technologies. Palaeontology isn't old-fashioned, it's vital.