An article in the New Scientist points to an important way to trace evolution through the sedimentary rock record.
In the development of life into multicellular organisms, there is some difference of opinion on the most significant step: whether it's from simple single-cellular organisms (prokaryotes) to complex ones with nuclei (eukaryotes), or the evolution of single-cellular life to multicellular. Most seem to favour the former.
Whereas stromatolites give a record of when the earliest lifeforms first emerged 3.5 billion years ago (according to Stephen Jay Gould this was just about as early as it could) there has been no way to trace back first emergence of eukaryotes - until now.
Peter Ward reports on chemical biomarkers in rock. Certain biological molecules break down under heating, cooling and pressure into highly stable organic compounds that could not be made by any known inorganic process. That last phrase is the kicker, which indicates that any such occurrence of the relevant compound would be an indicator of life. Moreover, some of the compounds are unique to particular groups of organisms. For example, C28 to c32 polyenoic fatty acids have been found to be unique biomarkers of sponges.
In the late 1990s, two Australians found steranes - biomarkers for eukaryotes - in Australian rock dating back to 2.7 billion years (but no older). This gives a good indication of the first emergence of lifeforms with cell nuclei.
By 800 mya, the biomarkers indicated multicellularity.
By 542 mya, the biomarkers indicated animal life - but we know this already as the Cambrian explosion.
In 2005, a Japanese team investigating the K-T extinction event (65 mya) they found, as expected, biomarker indicating a deluge of dead plant material. One study then found a decreased abundance of land plants for the next 7000 years.
Then to the Permian mass extinction event, the biggest of the lot. In 2005 a biomarker called isorenieratene was found at this point. These indicate green and purple sulpher bacteria, which cannot tolerate oxygen in water, in turn suggesting the oceans were [largely] devoid of oxygen and "saturated with hydrogen sulphide". A poisonous balance. A suggestion since (by Lee Kump) has been that hydrogen sulphide as the primary cause of the Permian extinction: so much in the ocean that it escaped into the atmosphere, poisoning land-based life and depleting the ozone layer.
The cause of this was in turn traced to global warming triggered by greenhouse gases from the Siberian Traps - the biggest ever volcanic eruption episode. The coinciding of the Permian extinction with the Siberian Traps has long been identified. although not all the mechanisms had been strung together: the global warming lessened the temperature differential between polar and tropic regions, in turn slowing ocean currents, causing stagnation (deoxygenation) then a buildup of anaerobic (oxygen-shunning) bacteria.
Via that biomarker isorenieratene, the same mechanism is seen to have happened - on a smaller scale, obviously - with the Devonian and Triassic extinctions. Ward comments that "it is beginning to look as if the K/T mass extinction was unique in having been caused by an impact [meteor]".
Ward's ultimate comment is that the planet is effectively a battleground between the early-emerging prokaryote life and the later-emergent eukaryotes (including us).
These periodic patterns seem only to serve to delay the emergence of tail-end complex life. And, as Gould would say, we just happened to get our chance due to that preceding history. Otherwise, it would have been down to the descendants of more advanced creatures than were our ancestors of the time.