As a phylogenetic systematist my primary research aim is the reconstruction of ancestral relationships among organisms using molecular and morphological character data. The resulting evolutionary trees allow us to trace the temporal trajectories of organismal traits and to identify key evolutionary developments: both of which enable us to reconstruct important aspects of the paleo-biosphere. This knowledge provides us with the context for the understanding of our place within the modern biosphere and to devised informed strategies to contend with future changes in our environment.

Since the ’60s the application of molecular methods to evolutionary systematics - especially DNA and protein sequences analysis - has had a profound effect on our understanding of the relationships among organisms. Today we can trace the evolutionary lineages linking bacteria to whales and we realistically speak about knowing, at one extreme, the relationships among the major groups in the Tree of Life, and at the end of the spectrum, about relationships among individuals and populations within a single species. However, the pace at which molecular sequence data can now be gathered (whole bacteria genomes take but minutes to sequence) has out-stripped our ability to effectively analyse the data due to computational constraints and often the difficulty is in determining what data are most appropriate to answer a question rather than what data are available. Moreover, the enormous growth in the availability of genetic data, and the increasing ease in which new data are obtainable in an era where genomic sequencing is routine, has meant that systematists are becoming increasingly reliant on bioinformatic frameworks to manage and manipulate the large amounts of data they employ.

Modeling of the molecular evolutionary process is an integral part of organismal systematics and increasing comptational power has enabled us to apply evermore complex models. In this regard, I am particularly interested in the modeling of different forms of substitution process heterogeneity in maximum likelihood and Bayesian analytical analyses. Although often ignored by phylogenetists, process heterogeneity is present at all levels of the Tree of Life and its accommodation in our models is having an increasingly profound impact on phylogenetic systematics. Indeed, it is becoming increasingly apparent that some of our long-standing and strongly supported ideas about the evolution of major groups may be wrong because of the use of poorly-fitting models that result in biases. I have focused much of my research on two examples that fall into this category: 1) the three-domains hypothesis for the tree of life, which states that there are 3 major lineages of life, namely bacteria, archaea, and eukaryotes, and 2) that vascular land plants (e.g. ferns and seeds plants) are derived from non-vascular bryophytes (e.g. mosses and liverworts). While both of these hypotheses routinely appear in biology textbooks, the work myself and colleagues have been doing show that they are both most likely wrong. Our analyses suggest that eukaryotes are derived from archaea and that therefore there are only 2 domains of Life (the so-called eocyte hypothesis), and that the vascular plants and bryophytes are both derived from the same ancestor and neither is the direct ancestor of the other.

Besides these over-arching themes of my research programme, I have worked on many diverse taxa from Candida (the fungus causing "thrush" infections) to Sphagnum peatmosses, the stuff used to mulch a garden, and used a wide variety of methods and desciplines from phylogenomics to population biology. In parallel with these research aims I also write and deploy software and manage a UNIX- based computational cluster that facilitates my research and that of my colleagues at CCMAR. The importance of bioinformatics skills as part of my research (as with many biologists today) cannot be overstated as it enables me to perform novel analyses that would otherwise be impossible.