The diversity and complexity of microorganisms is staggering. Their metabolic flexibility has enabled bacteria and archaea to live in places no other organism can tolerate. They are found in the sediment of the deepest oceans and within the rocks and ice of the highest mountains, and yet they also exist in the most commonplace of settings—in intimate association with our own bodies. Microbes can take the form of single spheres, long, bizarre corkscrews, or crowds of intercommunicating cells in a biofilm. Their history is far, far longer than our own, and their ancestral relationships are complicated by erratic combinations of asexual reproduction and promiscuous sharing of genes.
To make sense of this diversity, microbial taxonomists in the late 1800s devised a way of placing microorganisms in categories in which each organism was granted a “genus” and a “species” designation. Phenotypic properties were the only means of describing microorganisms at the time, so the taxonomic system of microbes was much like that used for plants and animals: it was based on the measurable and observable characteristics of the organisms, not on genetic traits. As the 20th century progressed, technological advances enabled more and more discoveries about microorganisms. Molecular biology uncovered the genetic relationships between microorganisms and some of the secrets of microbes that have yet to be cultured in the lab were revealed. Much of this new knowledge was incorporated into species descriptions, but difficulties in classification persisted and novel issues arose. Conflicts exist between phenotypic information and phylogenetic information, the means for classifying noncultured microbes are limited under the current paradigm, and microbial species do not always demonstrate the phenotypic or genotypic (genetic) cohesiveness expected of them. For these reasons and others, it has become clear that the species classification framework in use today is not capable of fully portraying and organizing microbial diversity.
The American Academy of Microbiology convened a colloquium in Washington, D.C., on September 27-28, 2006, to discuss problems in microbial taxonomy today. Participants with expertise in microbial taxonomy, systematics, ecology, physiology, and other areas described the history of microbial taxonomy, the state of the field today, and how work in the field should proceed in the future. This document is a record of their comments and recommendations.
The individuals who deal with microorganisms on a day-to-day basis, including scientists, medical professionals, and forensic scientists need a rigorous framework through which they can understand and communicate about these microbes. Today’s microbial classification system uses pragmatic, arbitrary, and artificial methods in a formalized polyphasic approach (one that incorporates multiple types of data) to assign species names to organisms. This system addresses some of the needs of the end users, but not all. The fixed rules and cut-offs in use today are quite useful but inadequate and inappropriate in some contexts. Genomics has great potential in the service of microbial taxonomy, but there are some caveats to applying genomics to classification, and these tools have yet to achieve widespread acceptance in taxonomy.
Genome sequence comparisons are ultimately the best reference for classification, but they have also revealed the limitations of the current concept of microbial species. Considering that microorganisms engage in intra-species genetic exchange, the significance of the microbial species concept is unclear.
In moving forward with microbial taxonomy, it is critical to determine whether microorganisms cluster in groups with meaningful commonalities or to determine what commonalities may be best used to cluster microorganisms into meaningful groups. The data gathered to date are ambiguous on this issue. Community genomics investigations are also needed to help refine our understanding of clustering, but these studies must be designed and carried out carefully to reduce biases. A number of promising techniques available in the near future will enable a greater understanding of microbial diversity.
It is clear that the current system for designating microbial species is somewhat functional, but inadequate in many ways. It is unclear whether this system should be replaced or renovated. Moreover, there is a lack of compelling data to support the deployment of either a theory-based model or an operationally-based model of microbial species. More population genetics and phylogenetics information is needed in order to determine which paradigm (or a blend thereof) is more appropriate for microorganisms.
A number of research recommendations were established by the participants, including the suggestion that the international microbial systematics community should coordinate an initiative to construct draft genome sequences of each of the roughly 6,500 type strains catalogued by the International Committee on the Systematics of Prokaryotes (ICSP).