Life on a Young Planet (20 page)

Niklas’s models and Lenski’s experiments help to reconcile biological evidence for rapid bacterial evolution with paleontological observations of cyanobacterial stasis. For cyanobacteria, adaptive landscapes probably resemble Mt. Fuji rising out of the plains of Kansai. Early in Earth history, newly evolved blue-greens invaded tidal flats and other environments. In each new habitat, natural selection pushed cyanobacterial populations up a steep adaptive peak. Those that reached the summit were difficult to dislodge and unlikely to descend. If this picture is even approximately correct, cyanobacteria should adapt rapidly (on a
geological timescale) to new environments and then persist as long as the environment is present. That, of course, is just what we see in the Proterozoic fossil record.

More generally, this view suggests that on the timescale of Earth history, the tempo of bacterial evolution is determined by rates of environmental change. New habitats beget new adaptations, with the result that bacterial diversity has expanded as the range of habitable environments has grown. The evolutionary processes at work are Darwinian, and the pattern that results is reminiscent of Eldredge and Gould’s punctuated equilibrium—except that extinction is rare and diversity, therefore, accumulates. Of course, environments can be biological as well as physical; evolving plants and animals have simply provided the bacteria with new kingdoms to conquer.

The Great Wall fossils shed light on another riddle of Proterozoic rocks, this one posed by stromatolites, the laminated structures found in so many Proterozoic limestones and dolomites. In the 1950s, Russian geologists began the herculean task of mapping Precambrian rocks across the vast Siberian platform. Thick piles of Proterozoic sediments occur in many parts of this gargantuan landmass, mostly hidden beneath forests and swamps, but here and there excavated by rivers like the Kotuikan. Mapping requires an understanding of how scattered outcrops relate to one another, but the Proterozoic beds of Siberia don’t contain shelly fossils, the conventional guides to geological correlation. On the other hand, they are chockablock with stromatolites. Along the Kotuikan River, thick stromatolitic reefs can be followed for miles (
figure 7.3
), and similar features occur in Proterozoic carbonates throughout Siberia. Some of the stromatolites are club-shaped, others conical; some branch regularly, others not at all (
figure 7.4
). Details of lamination and microscopic texture vary, as well.

Russian geologists—including Misha Semikhatov, whom we met in
chapter 1
—described these features in exacting detail and, in the process, became convinced that stromatolites held an important key to the correlation of Proterozoic rocks. They were right. Middle and late Proterozoic stromatolites can easily be distinguished from one another, and early Proterozoic stromatolites are something else again. What’s more, the stratigraphic pattern seen in Siberia and the neighboring Ural Mountains is reprised in Proterozoic rocks around the world.

Figure 7.3.
Stromatolitic reefs in the 1.5-billion-year-old Bil’yakh Group, northern Siberia. The bunlike feature to the right of Misha Semikhatov (who, for scale, is precisely 2 meters tall when wearing a hat) is a small reef. Misha is standing on the curved upper surface of a second, larger reef. And the wall extending above him is part of still another reef, this one the size of a small office building.

Therein lies what Shakespeare called the rub. Stromatolites are built by cyanobacteria, but as we’ve already seen, cyanobacteria show little evidence of an evolutionary trajectory through the long Proterozoic Eon. Why, then, should stromatolites change through time?

Bil’yakh fossils suggest an answer. Dolomites in the Great Wall contain sedimentary features we’ve seen before: crinkly laminated mats, tepee structures, sheets of ooids, and low stromatolitic domes. Like the Spitsbergen rocks introduced in
chapter 3
, Great Wall beds accumulated along the edge of an ancient ocean, in tidal flats and adjacent coastal environments. Despite similarities in habitat, however, the fossils in Bil’yakh cherts differ substantially from those in the Akademikerbreen Group.
Eoentophysalis
, so abundant in the Great Wall, is rare in the younger Spitsbergen rocks. Conversely, the stalked remains of
Polybessurus
found in Spitsbergen cherts do not occur in the Great Wall assemblage. We can’t blame the differences on evolutionary change; the dominant cyanobacteria in Bil’yakh and Akademikerbreen cherts
all
have close counterparts among living blue-greens.

Figure 7.4.
Stromatolites in mid-Proterozoic carbonate rocks from Siberia. The largest column is about 4 inches wide.

The alternative is that the
environment
changed through time. The vertical tufts of blue-green filaments preserved in Great Wall cherts provide subtle but important support for this idea. Tufts are common in modern mats, but they don’t persist as recognizable fabrics in sediments, because the weight of overlying deposits flattens them. The Bil’yakh filaments remain vertical because they were stiffened by calcium carbonate cement
before
they were buried. We didn’t observe this in Spitsbergen. It suggests that the Great Wall tidal flat was a bit different from its younger counterparts, more prone to carbonate precipitation right at or just beneath mat surfaces.

Other observations support this conclusion. For example, in modern tidal-flat sediments (and in Spitsbergen cherts), accumulating sediments collapse buried cells, compressing them as they decay. On the Great Wall tidal flat, however, decaying cells left behind three-dimensional voids now filled by cement—when these cells decayed, surrounding sediments had already become rock (
plate 4b
). Research by Julie Bartley,
an alumna of my lab who is now at West Georgia University, indicates that the timescale for cyanobacterial decay is commonly days to weeks—those Great Wall carbonates hardened fast. A few beds in the Bil’yakh succession even display the finely layered seafloor precipitates that complicate our interpretation of stromatolites in much older rocks. Along the Great Wall coastline, then, carbonates accumulated like papier-mâché, entombing microorganisms and providing a distinctively firm seafloor for colonization. Stromatolite accretion reflects both life and environment. Thus, changing seawater chemistry helps us to understand why stromatolite forms vary through time.

In the early 1990s, I had the good fortune to study an invaluable collection of thin sections prepared from Siberian stromatolites by the late V. A. Komar, one of the resourceful geologists who pieced together the Proterozoic history of Siberia. Along with Misha Semikhatov, I spent many hours trying to understand the paleobiological messages encrypted in the microscopic fabrics of these rocks. We found that stromatolites built by microbial trapping and binding of fine-grained sediments generally exhibit the uninspiring (and uninformative) texture of mud—layered, uniform, and bland. In some samples, however, stromatolites formed principally by calcium carbonate precipitation contain carbonate-encrusted filaments that lie in tangles along laminae. Evidently, filamentous sheaths provided microscopic sites for the precipitation of fine calcium carbonate crystals.

Interestingly, encrusted filaments occur only in stromatolites younger than about 1.0 billion years. Mid-Proterozoic stromatolites formed by carbonate precipitation also display distinctive microfabrics, in this case vertical splays of crystals formed at or just beneath the seafloor. Going backward from late to middle Proterozoic, then, the fine calligraphy of encrusted filaments gives way to a coarser crystal-dominated texture that masks the signature of mat-building microorganisms. Of course, as we retreat still further back in time, we encounter greater volumes of
macroscopic
carbonate precipitates made of stacked crystal fans (
figure 7.5
). Stromatolite fabrics, thus, reprise a theme developed earlier: the older the carbonate, the greater the evidence for cement precipitation at or near the seafloor surface. Proterozoic stromatolites appear to be environmental dipsticks, recording episodic changes in the carbonate chemistry of seawater as atmospheric CO
₂
levels declined and O
₂
increased. Change through time in stromatolite form is fully consistent with evolutionary stasis in cyanobacterial mat builders.

Figure 7.5.
Fingerlike laminated structures, each 0.4 inches across, formed by calcium carbonate precipitation without any obvious participation by microbial mats. This specimen, collected by Linda Kah, comes from a 1.2-billion-year-old tidal-flat deposit on Baffin Island, northern Canada.

To the extent that evolution has influenced stromatolite history, it may have done so primarily in a rather different way. As seaweeds diversified in the late Proterozoic Eon, they began to form algal lawns where microbial mats once held sway. The subsequent diversification of animals further intensified competition for space on the seafloor, and introduced grazing as well. In consequence, for the past 500 million years, stromatolites have largely been limited to lakes and restricted coastlines where competition and predation are strongly limited—reclaiming something of their former glory only in the aftermath of mass extinctions, and then only transiently.

Returning one last time to the shores of the Kotuikan River, we can contemplate the extraordinary history of cyanobacteria. Remarkable metabolic innovators that made breathable air possible, these hardy microorganisms epitomize bacterial continuity in a world of constant change. But the lessons of Spitsbergen seep back into our thoughts,
prompting one last question. If Bil’yakh cherts are full of cyanobacteria, what might we find in associated shales formed on the open seafloor? In Spitsbergen, comparable rocks brim with the fossils of eukaryotic algae and protozoans. Do they occur in Bil’yakh shales, as well?

Alexei Veis, a colleague of Semikhatov and Sergeev in Moscow, has studied these shales thoroughly, finding cyanobacterial filaments along with hollow organic balls compressed into minute platters. We can’t be certain, but the large size of these balls (up to 500 microns) suggests that they are indeed the remains of eukaryotic cells. Rocks of similar age in Australia contain rare microfossils that are undoubtedly eukaryotic, their biological affinities given away by long, branching arms that ornament cell walls. Nonetheless, rocks of this age do not contain the exuberantly spiny spore coats, vase-shaped microfossils, or many-celled seaweeds found in Spitsberegn shales. By 1.5 billion years ago, then, the cyanobacterial revolution may have been complete, but a second revolution—the rise of eukaryotes to ecological prominence—was yet to come.

__________

¹
This does not mean that Great Wall cyanobacteria were identical in molecular detail to their modern counterparts. Without doubt, the nucleotide sequences for many genes have changed somewhat through time, and some enzymes in living populations are more efficient than those in long-dead ancestors. What I
do
mean to convey is that living cyanobacteria provide pretty specific guides to the functional biology of fossils preserved in Great Wall cherts.