In this, the third in a four-part interview with Astrobiology Magazine, she explains why she believes that the notion of species doesn't apply to bacteria, and why she rejects the separation of Archaea into a different domain of life from Bacteria.

Astrobiology Magazine: You have argued that bacteria don't have species. I wonder if you could explain that idea.

Lynn Margulis: Bacteria are much more of a continuum. They drop their genes all the time. Like we say in What is Life?, it's like going swimming in a swimming pool, going in blue-eyed and coming out brown-eyed, just because you've gulped the water. Obviously, animals don't do that. But that's what bacteria do, all the time. They just pick up genes, they throw away genes, and they are very flexible about that.

Say you have a bacterium like Azotobacter. This is a nitrogen-fixing bacterium. It takes nitrogen out of the air and puts it into useable food. Nitrogen fixing is a big deal. It takes a lot of genes. If you put a little something like arsenium bromide in a test tube with these organisms, and put it in a refrigerator overnight, lo and behold, the next day the cells can't do this anymore, they can't fix nitrogen. So by definition you have to change them from one genus to another.

I'll give you another example: E. coli. It's a normal inhabitant of the human gut. If you put a particular plasmid into E. coli, all of a sudden you have Klebsiella and not E. coli. You've changed not only the species, but the genus. It's like changing a person to a chimpanzee. Can you imagine doing that, putting a chimpanzee in the refrigerator, and getting him out the next morning, and now he's a person?

Sorin Sonea, who was the chair of the microbiology department at the Universit� de Montreal, in Canada, has been saying for 25 or 30 years that you either have to consider all the bacteria on Earth as one species, or you have to consider them as no species at all. The criteria we use for species, which are good ones for animals and plants and fungi, do not apply, because bacteria can change overnight. You have all sorts of gradations, where adding or removing a few genes will change an organism's name, because those genes are what define the organism.

For example, Agrobacterium causes tumors on the crown of a plant, where the soil meets the base of the plant. If you get rid of just one plasmid, Agrobacterium can't make the tumor anymore. So it's now got a new name, by definition. So you've got this terrifically arbitrary situation.

But microbiology's mostly not science, it's practical art. That comes from its beginnings with Pasteur. Microbiologists are pragmatic, pious businessmen. There's no intellectual tradition in microbiology.

Bacteria do not have species. The rules for species naming do not apply to bacteria. But they need to have identity labels because of the practical importance in agriculture and health.

AM: So what would you call them, these different microbial organisms?

LM: You'd call them bacterial strains, or something like that. They definitely have some kind of identity. The genus name is more important than the species name. The genus name does correlate with a lot of common traits. But the reason the species were imposed is because they're socially so important, they're economically so important, that you can't have total chaos. You have to have a name, just like you have to have a name for diseases. How can you treat them if you don't have a name?

When you get to any eukaryotic organism, a species is a meaningful term. It represents the entire group. Species are discontinuous and distinct in the eukaryotes. But they're just not in bacteria. It's a continuum.

AM: Carl Woese in the late 1970s discovered a group of microorganisms that he named Archaea, and he drew a new tree of life, in which the major subdivisions were three domains - Archaea, Bacteria and Eukarya - rather than five kingdoms. But you like the five-kingdoms approach better. That was a surprise to me, because it was in part Woese's redrawing of the tree that helped bring the importance of microbes to more general awareness.

LM: Okay. I think that's a really crucial thing.

The first thing to say is that Woese made an unprecedented contribution, because he's got a single gene, namely the one for 16S RNA, which is present in all cells - because they don't work otherwise. And that gene can be compared in all organisms, whether they have feathers or teeth, or leaves, or spores, or whatever it is. That gives you what the 19th century people called a "partial phylogeny." It gives you a trait that can be traced through every cellular organism.

By tracing that gene he found that there are these two major groups of bacteria, and then most of the eukaryotes are relatively uniform with respect to this one trait. That's so far, so good. But he's worse than wrong about taking that trait as representing the live organism and the organism's history. That's where there's a horrible problem.

And the problem is that you've only got one gene. But an organism has 30,000 genes (if it's a eukaryote), and he's ignoring 29,999. (If it's a prokaryote, it typically has 5 or 6 thousand genes.) So you're not studying the lineage of live organisms when you study this gene. You're studying this gene.

Now this gene does give you an insight, and I would not deny that.

If Woese had not seen the differences between Archaebacteria and Eubacteria, I wouldn't have been able to see that all eukaryotes have both types in their ancestry, and therefore are products of symbiogenesis. It's a very valid point.

And being able to quickly assess what kind of bacteria an organism is with respect to this 16S RNA, all of that's very useful and very practical. And very informative. But what he's done is extrapolated it way beyond what it's really telling you. What amazes me is how fast people accepted this. Archaebacteria can exchange genes with Eubacteria. They are bacteria in every way.

Related Links
Darwin Today At TerraDaily.com