Divergence of bacilli and the purple bacteria

aroger at ac.dal.ca aroger at ac.dal.ca
Sat Feb 26 17:42:29 EST 1994


In article <1994Feb26.140431.21306 at dal1>, arlin at ac.dal.ca writes:

>  Other than the morphological
> similarity to cyanobacteria, there is no reason to suppose that the
> microfossils represent photosynthetic (as opposed to non-
> photosynthetic) organisms.

Actually there is a little bit of data to support that these organisms
were photosynthetic.  Bill Schopf says:
"The isotopic compositions of Early Archean organic and carbonate
carbon are indicative of photosynthetic CO2-fixation like that occurring
at relatively high CO2 concentrations in extant microbial populations."
He goes on to say that there is evidence that the banded iron 
formations contain concentrations of europium and cerium which are
consistent with models of early Archean ecosystems and suggest that
oxygen-evolving photosynthesis and aerobic respiration data from the
early Archean.

As for molecular clock estimates based on 16S genes, it is doubtful they
will give meaningful answers.  The reason why takes just a little 
inspection of these trees.  If mitochondrion-containing organisms
evolved 1.8 billion years ago (the earliest convincing evidence for
eukaryotes) then the distance on the rRNA trees from the common ancestor
of all of these organisms to the terminal nodes of the eukaryote tree
should be less than or equal to the length of the branches connecting
alpha-purple bacteria to other purple bacteria.  Instead, the 
former distance is 2-3 times as long as the alpha-purple bacterial
branch.  This is just a single example of the HUGE violations of
clock constancy in the rRNA tree.  It is very very difficult to get
around this problem.  
Contrary to what Woese repeatedly asserts, rRNA is not a very good
molecular clock and it is a downright useless "chronometer".  If
we used chronometers like this for the Olympic games, maybe the
Canadians and the Americans would win more medals....

The qualifications that I gave in my earlier posting betray my
skepticism.  I think that all my assumptions are shaky.  For instance,
cyanobacterial properties could be ancestral to all cells.  We
could evaluate this with parsimony if we could only root the tree of
extant life with certainty.  rRNA trees are useless in this respect.
The attempts to root the tree by Gogarten and Iwabe (using genes duplicated
prior to the divergence of all extant groups) have been promising.
However, they will only allow us to determine a root if all the 
RELEVANT taxa are included.  In Iwabe's paper only two representative
eubacteria are used.  How are we to know whether the root of the
tree falls within or outside of most known eubacterial groups
 if they do not appear on the tree?

Andrew Roger
aroger at ac.dal.ca



More information about the Mol-evol mailing list