A few days ago I posted on the discovery of the world’s earliest known fossils—sulfur-metabolizing bacteria from a site in western Australia, with the bacterial “microfossils” dated at 3.4 billion years old. (Earth is 4.54 billion years old, which means that life originated no later than 1 billion years after Earth’s formation). The paper, whose reference and link is below, was by Dave Wacey et al., with Martin Brasier as the last (i.e., “senior”) author.
I was a bit concerned, then, when I saw P. Z. Myers had a few concerns about this paper at Pharyngula, concerns that “gave him pause.”
P.Z.’s first worry was that the fossil cells seemed too large to be prokaryotes. The paper of Wacey et al. shows, as P. Z. said:
. . . lots of cells with 10-30µm diameters. And the authors come right out and report that:
‘The size range is also typical of such assemblages, with small spheres and ellipsoids 5-25 µm in diameter, rare examples (<10) of larger cellular envelopes up to 80 µm in diameter, and tubes 7-20 µm across (see ref. 24).”
How odd. When I poke into the nervous system of an embryonic insect or fish, those are the sizes of cells I often see (well, except there aren’t many tubes of that size!). When I poke into a culture or embryo contaminated with bacteria, they’re much, much smaller. So maybe paleoarchaean bacteria tended to be larger?”
P. Z.’s second concern was that “reference 24” cited by Wacey et al. in support of the size ranges of fossil bacterial assemblages was in fact a paper by J. William Schopf. This disturbed Myers because Schopf’s earlier claim to have found the oldest microfossils on earth (3.465 billion years old) has since been discredited; as P.Z. said, Schopf’s microfossils form “a data set that’s widely considered artifactual now.” And Schopf’s largest cells were from his oldest samples, with cells getting smaller as the samples came from successively younger strata.
These considerations gave the estimable Dr. Myers some reservations about Wacey et al.’s conclusions, and at the end of his report he asks “…isn’t this just a little bit strange? Maybe there are some micro people out there who can reassure me that this isn’t a surprising result.”
I was a bit distressed about this since I hadn’t noticed those “problems.” I then checked them out since I don’t want to report stuff that’s wrong or dicey. So I contacted Dave Wacey and Martin Brasier about P. Z.’s reservations. And I’m happy to report that there’s no problem with the Wacey et al. paper. What appears below is a bit technical, and I’m posting it to put the record straight, but if you’re following the early-fossil literature, you should read it.
Here’s what Martin Brasier wrote me, which I post with his permission:
1) Re the large size of the bacterial fossils shown:
The first point is that the larger microfossils proved to be the easiest to image and showed more convincing cellular features, so that there is a bias towards the illustration of large forms in our Figures 1 and 2. In fact, the majority of the microfossils from the Strelley Pool arenite actually fit into the same sort of size distribution patterns as seen in the younger Gunflint and Bitter Springs cherts. This is clearly shown within our Supplementary Figure 6.
Regarding the suggestion that bacterial cells tend to be small, this approach no longer looks safe. Sulphur bacteria have members that are particularly large today (12 to 160 μm). They reach great size because of cell vacuoles within them (e.g., AHMAD, A., BARRY, J.P. and NELSON, D.C. 1999. Phylogenetic Affinity of a Wide, Vacuolate, Nitrate-Accumulating Beggiatoa sp. from Monterey Canyon, California, with Thioploca spp. Appl. Environ Microbiol. 65, 270-277). Hence unsheathed Beggiatoa filaments can often reach a width of 65 to 85 μm (range about 20-140 μm). Individual cells can reach 20 by 70 μm. Multiple Thioploca filaments (2-43 μm) usually occupy a single sheath (the part that preserves). Coccoid Thiomargarita namibiensis can be 100–300 µm wide, but can sometimes reach 750 µm.
It would be unwise to speculate about the possibilities of a larger cell size in early bacteria. But we note that microfossils reported from the 3.2 Ga Moodies Group of South Africa can be almost ten times the diameter of those found in the Strelley Pool rocks. See JAVAUX, E., MARSHALL, C.P. and BEKKER, A. (2010). Organic-walled microfossils in 3.2-billion-year-old shallow-marine siliciclastic deposits. Nature 463, 934-939.
2) Re the concerns of PZ Myers about use of a Bill Schopf reference
Reference 24 in our paper is NOT the Schopf (1993) article about the Apex microfossils – it is Schopf’s 2006 Phil Trans R Soc paper that was a summary of ALL microfossil claims from units older than 2.5 Ga known at that time. There were actually 40 occurrences of Archaean microfossils described by Schopf in this paper. From younger rocks, he describes 10-28 µm tubular sheaths from the Ghaap Dolomite and 16-23 micron coccoids from the Sheba Formation. (See his Table 2.) Those records are what we were referencing.
It is true that our views on some of these records has been, and remains, cautious. See, for example, BRASIER, M.D., McLOUGHLIN, N., GREEN, O., WACEY, D. 2006. A fresh look at the fossil evidence for early Archaean cellular life. In Cavalier-Smith, T., Brasier, M.D. & Embley, T..M. (Eds) Major Steps in Cell Evolution: Palaeontological, Molecular and Cellular evidence of their timing and global effects.Philosophical Transactions of the Royal Society, Series B, volume 361, 887-902. See also the book by David Wacey (200) “Early Life on Earth. A Practical Guide”. Springer, Dordrecht.
For those who are interested, our re-interpretation of the 3.46 Ga Apex microfossils has been set out afresh in a new monograph: BRASIER, M.D., GREEN, O.R., LINDSAY, J. McLOUGHLIN, N., STOAKES, C., BRASIER, A. & WACEY, D. 2011. Earth’s Oldest Putative Fossil Assemblage from the 3.5 Ga Apex Chert, Chinaman Creek, Western Australia: A Field and Petrographic Guide. Records of the Geological Survey of Western Australia, Perth. This is now available on the GSWA website and on the Academia website. Out views on those follows has not changed from that we set out in 2002. We saw no need to cite the latter paper as its arguments are widely known.
17 thoughts on “More on the world’s earliest fossils: they’re real!”
And PZ is all OK now.
I’m no expert, but naively I’d expect primitive cells to be bigger than modern bacteria, for more or less the same reason that early cell phones were bigger than today’s phones: it just takes a while to figure out how to pack all the necessary parts efficiently into a small space.
I would expect the same, but for a different reason: at least some ecological niches currently filled by eukaryotes would probably have been filled by prokaryotes. This would likely encourage some prokaryotes to become larger. Nowadays the prokaryotes cannot compete with eukaryotes in these niches, but without any eukaryotes around they probably could fill some of them at least to some extent. Large prokaryotes may not be able to compete well with eukaryotes today, but they might have been able to compete with smaller prokaryotes back then.
I agreed, and thought along similar lines, with Greg’s comment.
One of the great things about Science is that even when most scientists are applauding the latest discovery, it is not only acceptable, but appreciated, when another scientist raises a good question that makes the rest think: “Good question, perhaps we are all wrong.”. Even better, as in this case, when they check, find there isn’t a problem, and the dissenting scientist happily changes their opinion.
I doubt we’ll see any creationists, eg. Kent Hovind, behaving similarly.
Forgive my ignorance on this, but if fossils this small can be found, doesn’t this mean that the finer details of larger organism that have fossilized can be found? Specifically, I am thinking the structure of the nervous system.
I think that’s a good question & I’d like to know the answer to (I’m not a biologist) ~ is anyone here in the know on this ?
My guess ~ most types of fossilization doesn’t capture the innards & soft parts. That said this pic is of a fossil Ottoia priapulid worm from the Burgess shale where soft-bodied preservation has occurred. The specimen was wetted and oriented to reflect the light, in order to show a delicate iridescent film which preserves details of muscle bands, the gut, and even the small hooks at one end of the worm ~ note the scale. I believe that most/all (?) Burgess fossils are preserved as a very thin film requiring special lighting to reveal detail.
I suppose that creatures preserved in amber might be a good target because they are not compressed. Also permafrosted beasties.
guessing also conditions would have to be anoxic, dry, cold or in some other way preservative. Are the remains found in peat, ice or amber, where the original organic material is not usually replaced, regarded as fossils ?
Rapidity of deposition is crucial. You need your organism to be covered over with sediments as quickly as possible to protect it from being eaten and general falling apart. Anoxic environments are helpful in general and in this specific case because you need this for the formation FeS. In rocks this old a major problem is that they tend to have been heated and compressed to the point where the minerals recrystalize and the whole thing is a completely new rock (metamorphosed). This destroys the fossils and is one of the main reasons fossils this old are so rare.
Cheers Marella ~ most of my remarks were an attempt to answer Chris’s question:
Thus I was thinking about much younger fossils. For example the oldest amber recovered dates to the Upper Carboniferous (320mya) & insects, spiders [+ webs!], annelids, frogs, crustaceans etc. have been found preserved in amber ~ although none older than 130mya I am told. I’m speculating that it should be possible to map the nervous system of an amber preserved 20mya amphibian ~ 3D magnetic resonance imaging techniques work well on organic material. I did a search, but I found little about this ~ perhaps my reasoning is wrong.
BTW I thought this was impressive [not applicable to the above]:
THIS is a fossil cyanobacterium (650mya) from Kazakhstan…
** Top = Standard optical image
** Middle = Confocal optical image of the same fossil
** Bottom left = Close up of a section of confocal optical image (equating to the red box on optical image)
** Bottom right = Raman 3-D chemical image of the same boxed region
The detail on this last pic is only roughly sub-mm, but a vast improvement on the others ~ just an example of imaging improvements to come
Aah, the fine art of scientists working together to come to a certain conclusion.
This was also covered over at the excellent blog Ediacaran.
I am glad Brasier et al found methods and results that are so much more testable than Schopf et al pattern search.
This is interesting from two other aspects.
– Strelley Pool formation, the find site, has earlier yielded stromatolites with micro-structures much as testable for bacterial origination. I don’t think they were effectively touched by Brasier’s criticism.
– And the old, not very constrained (putatively ~ 4.28 – 3.85 Ga bp), Nuvvuagittuq greenstone belt may have signs of precisely sulfur metabolism as well. At least you can find, unfortunately unpublished, texts to that effect on the web.
It would be nice if the Strelley Pool and the Nuvvuagittuq results coalesce on the same ancestral metabolism and push it further back in time, it would be a reasonable hope at this stage.
Confirming the age > 3.0 Ga bp as a hot spot for sulfur metabolism is work on gene families. Sulfur metabolic use can be traced early and the main evolution finishes in the beginning the Archaean Expansion (AE) of gene innovation, which itself is putatively dated to between ~ 2.8 – 3.3 Ga bp.
Speaking of that work on gene families, using it as a toy model clock allows for an interesting dating constraint on abiogenesis akin to Miller’s suggestion once. This toy model derives from the fairly steady gene family event rate (sum of birth, transfer, duplication, loss) shown there. The rate ranges from ~ 0.4 events/My to ~ 1.6 events/My.
David et al assume a date for the LUCA (Last Universal Common Ancestor), from the arguable trace record, of ~ 3.85 Ga. The model is self-consistent however, and accords with dates for end of Late Heavy Bombardment (AE) and atmosphere oxidation (electron transfer redox metabolism).
Moreover, before the AE there is a suggested steady gene family birth rate.
Let us assume that steady rate of ~ 0.4 events/My. Their LUCA is estimated to ~ 180 gene families. [Supplement fig 6.]
This gives a latest date for the first gene family of ~ (4.55 – 3.85)*103 – 180/0.4 ~ 240 My from Earth accretion. Or ~ 4.31 Ga.
Now, if the new data published elsewhere can pull the Earth-Moon impact event forward to an astounding 4.36 Ga bp, we have the following timeline for Earth:
~ 4.54 Ga before present: Planet formation starts.
~ 4.04 Ga bp: Latest formation of 1st crust (Jack Hill zircons).
~ 4.36 Ga bp: Earth-Moon impactor.
~ 4.35 Ga bp: Latest presence of liquid water (Jack Hill zircons).
~ 4.25 Ga bp: Earliest putative trace fossils (Jack Hill diamonds).
Since some researchers place Earth planet formation at ~ 30 Ma after the solar system starts to form, and Earth 1st crust formation at ~ 10 Ma, all these dates are possibly still consistent!
I have to assume that life survives the Late Heavy Bombardment. But later models predicts precisely that. Abramov et al found that prokaryotes proliferate and spread faster than any plausible LHB impact rate can sterilize the crust:
“Our analysis shows that there is no plausible situation in which the habitable zone was fully sterilized on Earth, at least since the termination of primary accretion of the planets and the postulated impact origin of the Moon.”
“Finally, the largest impactor in our baseline model (~ 300km in diameter) is insufficient to vaporize the oceans.”
In other words, life is a plague on a planet.
Putting the dates together may bound the time for pro- to protobiotic transition tightly to <= 40 Ma between 4.35 Ga bp (2nd crust formed, liquid water present) and 4.31 Ga bp (gene family clock).*
Not as tight as once Miller's bound of <= 12 Ma, but again getting into the comfort zone of an easy transition.
* It also means the Earth-Moon impactor, which demolished the crust and likely had no "Goldilocks survival zone" ~ 1 km deep in the crust, sterilized a first biosphere.
I don’t understand this part:
Can you explain ?
I refer to the part of abiogenesis where there is a transition between a set of probiotic “soups” (most likely) to populations of replicators (say, Szostak’s protocells).