Life on a Young Planet (11 page)

Figure 4.3.
Sedimentary features of Warrawoona rocks. (a) Gray-black gypsum crystals (now replaced by silica) that grew on the seafloor and were subsequently buried by thin beds of mud and sand (light layers). Scale bar in centimeters. (b) The stromatolite discovered by Roger Buick and colleagues in the early 1980s, now subject of much debate. Six-inch scale to left. (Photo (a) courtesy of Roger Buick)

Or did they? Early on, Roger Buick and his colleagues urged caution: the words “possible” and “probable” pervade a thoughtful essay published in 1983. It isn’t, they wrote, that the Warrawoona structures
can’t
have been built by bacteria, but rather that the actual evidence for biological accretion is iffy. And, in rocks this old, with so little independent evidence of biology, the interpretational stakes are high. We want to know whether or not life existed 3.5 billion years ago, and “maybe” is not a satisfying conclusion.

In 1990, Don Lowe retreated further. He explicitly reinterpreted Warrawoona stromatolites in terms of chemical processes that deposited layers of minerals on the seafloor and physical processes that deformed these layers into upturned folds, like carpets after a slip. In this view, the hand of life is nowhere to be seen.

Why this loss of confidence in a biological interpretation? Simply put, microbial mat processes can give rise to laminated structures, but they are not the
only
processes that can do so. Similar features can form without microbial mats if ambient waters are highly charged with dissolved minerals—for example, in Yellowstone Park, where silica-saturated waters spill episodically from thermal springs to form laminated structures that resemble stromatolites built by microorganisms.

We didn’t worry much about this in Spitsbergen, because 600–800 million years ago, the oceans weren’t sufficiently laden with calcium and carbonate (or silica) to drive this type of deposition. On the early Earth, however, ocean chemistry was different. John Grotzinger, an MIT geologist we’ll meet again, has shown that Archean oceans were charged with calcium and carbonate ions, so much so that the carbonate minerals calcite and aragonite commonly formed by direct precipitation on the seafloor. These deposits include giant crystal fans whose precipitated origins are obvious, but they also encompass flat laminated beds, domes, and small columns arrayed in parallel rows (see
chapter 7
). This being the case, it becomes important to examine the microscopic features of Warrawoona stromatolites to see whether mat-building fossils have been preserved. They have not, making it difficult to know to what extent organisms participated in the formation of these structures.

At present, the interpretation of Warrawoona stromatolites remains unresolved. We can accept them as biological only if we can rule out alternative physical explanations. Hans Hofmann and Kath Grey, two
experienced stromatolite hands, recently discovered new Warrawoona structures that they hope will decide the case in favor of biology. The buildups, formed layer after layer by the precipitation of thin crystalline veneers, are conical, a shape rarely generated on the seafloor in the absence of microorganisms.

But if the Warrawoona cones reflect life’s guiding hand, what
kind
of biology was at work? On the modern Earth, cyanobacteria are the best-known mat builders, and so there is a tendency to associate all ancient stromatolites with cyanobacterial mats. Other bacteria can form mats, however, and there is no a priori reason to assume that cyanobacteria existed 3.5 billion years ago. In our quest to find biological signatures in the ancient beds of Warrawonna, stromatolites provide only a suggestive scrawl—a tantalizing but ambiguous hint that life began in our planet’s infancy.

Spitsbergen cherts contain microfossils that are unmistakably cyanobacterial, increasing our confidence that these blue-green bacteria built contemporaneous stromatolites. Are the black cherts of Warrawoona similarly informative? For more than a decade, most people thought so, but recent reinvestigations have sown considerable doubt about Warrawoona micropaleontology.

In a preliminary draft of this chapter, I recounted the 1987 discovery of Warrawoona fossils in cherts from a knobby outcrop along the dry bed of Chinaman’s Creek, near North Pole. In these rocks, UCLA’s Bill Schopf and Bonnie Packer found tiny filaments 1 to 20 microns in diameter and up to a few hundred microns long (
figure 4.4
). The structures are rare and poorly preserved—the distorting effects of crystal growth are clear in published photos. Nonetheless, the photos, or at least the interpretative drawings that accompany them, look like simple cyanobacterial filaments. Equally, however, they resemble other types of bacteria, limiting taxonomic or metabolic interpretation.

In Spitsbergen, knowing where a microbe grew helped us to understand how it made its living. Geology colors our thoughts about Warrawoona biology, as well, but in a surprising way. Careful mapping by Australian stratigrapher Martin van Kranendonk shows that the cherty rocks at Chinaman’s Creek formed beneath the Warrawoona seafloor, not on it—these cherts originated in hydrothermal veins, as outlined earlier in this chapter (
figure 4.5
). Conceivably, cyanobacteria could have found their way into these cracks along with other sedimentary particles, but the environmental setting revealed by geology likely favored chemosynthetic growth over photosynthesis.

Figure 4.4.
Microstructures interpreted as bacterial fossils in Warrawoona chert. The alternative interpretation is that they are simply chains of crystals formed in a hydrothermal vein. (Photo courtesy of Martin Brasier)

From here, the problems of interpretation mount. Despite reservations about the specific interpretation of Warrawoona microstructures, I had no cause to doubt their biological origins as I drafted my chapter. After all, poorly preserved fossils with shapes obscured by mineral growth occur throughout the geological record. Why should the old cherts of Warrawoona be any different? Before committing my opinions to print, however, I wanted to see the Warrawoona material for myself.

Surprisingly, the place to view purported Warrawoona fossils is not Sydney or Perth, or even Los Angeles. It is London, in the collections of the Natural History Museum. In September 2000, I had to cross the ocean for a scientific meeting in Oxford and so made arrangements to spend a quiet day at the museum before returning home. I also planned a not-so-quiet day in Oxford visiting Martin Brasier, a well-known paleontologist and connoisseur of local pubs. Fortunately, I told Martin of my hope to study the Warrawoona rocks, and he, in turn, told me that the key samples were out on loan—to Oxford! In Martin’s lab in a quaint Edwardian building off Parks Road, we spent a revealing day examining thin sections of Warrawoona chert.

Figure 4.5.
Warrawoona chert at Marble Bar, Western Australia. The red (pigmented by iron oxides, gray in picture) and white bands accumulated on the seafloor. In contrast, the black bands cut across other beds and so are younger—and formed in a different way. Warrawoona microstructures interpreted as fossils come from these crosscutting cherts, interpreted as hydrothermal plumbing systems filled by silica.

In Spitsbergen cherts, microfossils are abundant. They have shapes similar to those of living microorganisms, but different from shapes made by purely physical and chemical processes. Most retain at least some of their original organic matter. And some even occur in environmental settings much like those of close living counterparts. No comparable claims can be made about the Warrawoona microstructures.
Observed through the microscope in Martin’s lab, the tiny filaments in Warrawoona rocks looked like minerals.

As a boy, I often spent idle summer afternoons gazing at clouds. Most were billowy masses, beautiful but shapeless. Every now and again, however, an unmistakable face appeared in the sky. Or a castle. Or a lion. For a moment, the shapes took striking form, but even as a youngster I was pretty sure that they were, in the end, just clouds. Are the Warrawoona microstructures “just clouds,” as well?

That question is difficult to answer based on a few pictures cut and cropped for publication. It requires
context
—the framework provided by the overall rock fabric seen in thin sections of Warrawoona chert. It was the rest of the clouds that revealed my air castles as watery illusions, and it is the overall fabric of Warrawoona cherts that casts doubt on those rare features that look biological. Martin and his colleagues painstakingly documented how volcanic and hydrothermal processes shaped the cherts from Chinaman’s Creek. They believe that physical processes can account for
all
microscopic features of the cherts, including those singled out as fossils. If this interpretation is correct, the Warrawoona microstructures cannot be cellular filaments, only stacked crystals that mimic but do not preserve a record of biology.

Are paleontology’s crown jewels, so old and rare, made of paste? In fairness, Bill Schopf disputes this reading. In a rebuttal to the claim by Brasier and colleagues, Bill and University of Alabama chemist Tom Wdowiak show that the disputed Warrawoona structures contain organic matter at their margins. This, of course, is consistent with the view that they are microfossils, but it doesn’t end the debate. Archean cherts commonly contain the ghosts of early formed minerals whose distinctive shapes are preserved by a veneer of organic matter. My own guess is that most Warrawoona structures are mineral chains draped by an organic film (that may, itself, have a biological origin). Continuing study may yet confirm the presence of fossils in these rocks—the debate is far from over—but I doubt that any such remains will teach us much about early ecosystems. Warrawoona microstructures, like Warrawoona stromatolites, can only suggest that something interesting and important lies just beyond our grasp.

Biomarker molecules are not retained in rocks from North Pole, but
isotopic signatures are; the carbon and sulfur isotopes of Warrawoona rocks provide our best indication of life’s deep history. As in Spitsbergen (and nearly everywhere else that sedimentary rocks were deposited in Precambrian times),
¹³
C/
¹²
C ratios in Warrawoona carbonates and organic matter differ by about 30 parts per thousand. This difference is most easily explained by photosynthesis, but given our experience with stromatolites and microfossils, we should ask once more whether physical processes can mimic the effects of biology. Some chemical reactions do form organic molecules depleted in
¹³
C. Only under carefully controlled experimental conditions, however, does nonbiological fractionation approach the levels recorded in Warrawoona rocks. For this reason, the
consistently
large fractionation measured in North Pole samples suggests the presence of an early biosphere.

Carbon isotopes in organic matter from the chert veins that cut through Warrawoona sediments and lavas could record chemosynthetic bacteria that lived in hydrothermal waters. But the widespread distribution of organic matter in sedimentary rocks deposited
on
the seafloor supports the hypothesis that photosynthesis fueled microbial life in the Warrawoona ocean. Whether primary producers were mostly cyanobacteria or other types of photosynthetic bacteria with a similar isotopic signature remains uncertain. Sulfur isotopes in sedimentary pyrite and barite likewise suggest that sulfate-reducing bacteria lived the Warrawoona lagoon, although this, too, has been questioned by geologists grown skeptical about early Archean biosignatures.