On the surface, the definition for a species seems relatively straightforward. According to the Linnaean classification scheme, species means the most specific group. It consists of individual organisms which share similar characteristics, genetics, and are able to interbreed. But what about those organisms--like bacteria--that can swap genes at will, not only within species but between species? Does the definition of "species" as we know it still hold or is a bit of redefinition in order?
About a week and a half ago, I attended a seminar given by Claire M. Fraser-Liggett, the current president and lead investigator of The Institute of Genomic Research (TIGR). Fraser gave a talk about the impact of microbial genomics on the definition of microbial species that really challenged the entire concept of a species boundary. Although one wouldn't say that the bacterial mélange could possibly be lumped into one huge category, the promiscuous and fluid habits of microbes certainly make any attempt at definitive classification terribly difficult and complex.
Descriptions of bacteria began in the late 19th century, focusing mainly on gross and easily observable characteristics like morphology, physiology and ecology. It was only in the latter half of the 20th century that molecular approaches began to be used to classify microbes. In today's era, microbial taxonomy rests mainly on comparing the sequences of the gene for the 16s ribosomal RNA. But not only is sequencing one gene--or even a handful of genes--rather limiting, but there's also an arbitrary cut-off point for saying which sequences are similar or not similar. Who's to say what a species is or isn't based on purely artificial parameters?
Comparing sequences isn't the only problem. Other limits hamper current taxonomy as well. There are conflicts with phenotypic and phylogenetic information. Even within a species, bacteria do not exhibit the same phenotype. Metabolic reconstructions straight from the genome don't correlate with laboratory observations. And there isn't any good way to classify the bacteria that can't be grown by current culturing techniques let alone microbes that are nonculturable.
One approach that Fraser-Liggett and her team at TIGR is pursuing is that of strain sequencing. The goal was to sequence many strains, all of which are considered to belong to one species of bacteria, to compile a core set of genes that would belong to all strains and characterize a given species. The results, to say the least, question even the notion that a set of genes could be used to define a species. For example, if one compared three different strains of Escherichia coli--the laboratory strain K-12, an uropathic strain, and O157:H7 (the strain that has caused many outbreaks that have gotten into recent news)--they share less than 40% of the genes. Now if you compare that to the statistic that we humans share more than 98% of our genes with chimps, you'd have to start to worry what we can call a microbial species.
Lateral gene transfer and bacterial promiscuity aside, this gives rise to what Fraser-Liggett calls the "pan-genome" or a complete repertoire of genes that is associated with a species. This new term embraces the notion that there is significant genome diversity within a species. Not only is there a core set of genes belonging to all members of a species but there is a genetic "halo" of genes which belong to the set of genes for the species as a whole but actually is only strain specific. With these ideas in mind, there are perhaps two directions in which microbial taxonomy may take. One is the theory based model in which the notion of "species" is based on the mechanisms of speciation. The second is an operational based model in which species are defined as clusters or organisms that share common features. Unfortunately, the second model merely assumes that there are, in fact, species in the first place.
And there's the issue of whether it's even worthwhile to group organisms as species. Speciation as defined for sexually reproducing organisms tends to fall apart when applied to asexual organisms, which means defining species based on speciation will be sketchy.
Species concepts are useful for studying speciation in sexually reproducing organisms, but are they even necessary for bacteria? Are they simple another arbitrary taxonomic category?
