Chapter I
1.2 Getting a Grip: Dinosaurs and Mass Extinction
Intelligent Design advocate Phillip Johnson wades into the extinction issue.
All this begs the question: while most organisms aren’t going extinct in mass events, they still are going extinct, eventually—why? The simplest answer is that living things go extinct because they finally confront an environment they can’t cope with. This happens to individuals all the time, of course. An antelope checks out because the lion caught up with it. But extinction in the sense we’re talking about here is all the members of the group failing to make the cut. That involves the interplay of the entire species in its total environment. For all of a group to go extinct things have to happen that causes the whole shebang to dip below the minimum threshold of survival (not enough of a population to actively sustain itself in the long term).
In the case of animals, they go extinct because conditions have changed but they haven’t. And the reason why that can happen at all is that organisms aren’t designed. They can’t be recalled by the manufacturer (like Toyota had to in 2010 over dangerously malfunctioning accelerator pedals) for a politic retrofit to keep them adapted to their altered environment. They can only run with the set of systems they were born with. If they or their ancestors didn’t have the luck of getting mutations that opened up new potential opportunities for them, that’s it—checkout time.
Hence the very existence of extinction as a phenomenon of the living world is a testament to the impact that chance processes have on how living systems can (or cannot) adapt to an environment where continents shift around, mountain ranges rise or fall, oceans appear or dry up, forests spread or recede, and every other animal (from predators to potential competitors for your own particular niche) is facing exactly the same dice rolling game. Can your species keep on going for the next round, based on what you have in your adaptive kit bag?
Knowing the finer points of what “extinction” means in real terms spread over Deep Time makes it all the more informative to see how the philosophical godfather of the Intelligent Design movement, Berkeley lawyer Phillip Johnson, has approached the subject.
As it happens, Phillip Johnson reviewed David Raup’s 1991 book Extinction: Bad Genes or Bad Luck? in the February 1992 issue of The Atlantic, conveniently reprised in Johnson (1998a, 41-47). Which means he had to have known about that background extinction rate and how most of that didn’t involve mass extinctions. And yet he has repeatedly invoked Raup’s work on mass extinctions as supposedly casting doubt on the prevalence of regular Darwinian processes rolling on during the remaining hundreds of millions of years when mass extinctions weren’t happening.
From the start Johnson (1991, 57) contended in Darwin on Trial that, “A record of extinction dominated by global catastrophes, in which the difference between survival and extinction may have been arbitrary, is as disappointing to Darwinist expectations as a record of sudden appearance followed by stasis” (we’ll get to the “stasis” issue in the next section 1.3 when we hit the Punctuated Equilibrium issue). In P. Johnson & Provine (1994), a debate at Stanford University with William Provine, Johnson reiterated this position (citing only Raup as his source) and by Reason in the Balance Johnson (1995a, 83) had tightened this conflation of background extinction and mass extinction into: “many authorities now attribute extinctions primarily to freakish catastrophes.” There were no references to any of these “many authorities” in the slim Johnson (1995a, 226-228) research notes, not even to his sock puppet Raup.
When Johnson appeared as a very congenial guest on Hank Hanegraaff’s Bible Answer Man radio show in December 2000 he had kneaded his misunderstanding into the blanket conviction that “the dinosaurs, and indeed perhaps all extinctions, were brought about by catastrophic event.”
Johnson’s behavior here is an important clue about what we will be discovering in terms of the tortucan mind. Even though the existence of a pervasive background extinction rate was clearly in evidence in the Raup work he had explicitly reviewed, he never saw that aspect of it, only the mass extinction spikes that seemed congenial to some allegedly non-Darwinian process whereby animals might be extinguished for other than their (designed?) adaptive perfection.
Interestingly, Johnson (1998a, 41) stressed that his review of Raup’s book back in 1992 had provoked letters to the editor that “were vehemently hostile, but Raup himself wrote to me privately and said I was right on target.” Raup has indeed been impressed with Johnson, as Witham (2002b, 69, 97-102) noted, and does believe that “impact-caused extinctions may actually dominate the extinction record” (personal communication, 2003). Insofar as Darwin abhorred the idea of mass extinction, in that sense Raup’s evolutionary views may be considered “anti-Darwinian.”
But it is instructive to take a look at those “vehemently hostile” letters to measure some of Johnson’s own gloss. There were six, of which only two emanated from scientists. None took aim at Raup’s position, but were all very doubtful about how Johnson framed the issue. Even the three pithiest remarks were fairly tame, though—suggesting Johnson wears rather a thin skin.
To wit: Robert Michael Pyle of Gray’s River, WA suggested in the May 1992 issue that Johnson was “a law professor slumming among scientists.” In June, McGill University genetics professor G. A. C. Bell likened Johnson to that “tawdry band” of literary outsiders who periodically announce the Death of Darwinism (such as George Bernard Shaw or Arthur Koestler)—cf. Peter Bowler (2002, 228) here.
When L. J. Marsh of Minneapolis described him in September as “pugnacious,” Johnson rejoined:
Rare catastrophes can be fit into a Darwinian framework if we assume that natural selection was at other times killing off the less fit and preserving the most fit. Suppose, however, that extinctions nearly always occurred in catastrophes, and that the victims were as proficient as the survivors at flying, seeing, reproducing, or whatever. That is what David Raup is suggesting. But if ancient species that were relatively unproficient at flying or seeing did not as a consequence dwindle and eventually die out, then what sense does it make to say that ‘natural selection’ produced improved capabilities in their successors?
A lot of abstract supposing here, all wonderfully divorced from specific example, which as we’ll see is the hallmark of Johnson’s apologetic Wedge approach to combating evolutionary naturalism.
Yet no matter how the fossil pie is sliced, Johnson’s recurrent supposition that “extinctions nearly always occurred in catastrophes” is tenable only if he restricts his attention to mass extinction events, and he can sustain his broader supposition that normal adaptive evolution hadn’t been going on the remaining 95% of the time only by paying no attention to any of the actual data. Worse, we (and Johnson) know of at least one very famous animal “relatively unproficient at flying” that apparently went extinct independent of any catastrophe: Archaeopteryx detailed in Chapter 4 of Downard (2003b) and Chapter 2 in Downard (2004). Unless of course Johnson has some Jurassic cataclysm hiding up his sleeve that he has yet to spring on the scientific literature.
Such rarified disdain for the body of available information makes Johnson’s concluding Atlantic reply sentence (p. 13) to Mr. Marsh of Indianapolis especially pompous: “Pressing awkward questions like this is not being ‘pugnacious’; it is being scientific.”
Well, let’s try being “scientific” shall we?
When you look at what was happening at the time of those decidedly rare mass extinctions it is clear something unusual (and therefore genuinely interesting) was going on, with the Permian crash being the most severe, and the earlier Ordovician event coming in second. Just how severe an extinct event is depends on what measure you’re using: an enormous number of species or genera can go extinct without necessarily removing all members of their family or class, as illustrated in Figure 2 below. Åžengör et al. (2008) and S. Wang & Bush (2008) provide some guidelines for assessing how severe a mass extinction is, and Mander et al. (2010) illustrate the challenges in identifying the level of disruption of plant diversity in the Triassic extinction.
But by all accounts the Permian event was the worst, decimating almost 90% of genera living then. For “well-skeletonized” marine families (thus better represented in the fossil record than soft bodied ones) illustrated in May (2012) the Permian stands out as a gigantic plunge in diversity that all but wiped out the previous 200 million years of general stability, followed by a resumed fairly steady climb afterward to much higher levels known today. Because most living things dwell in the sea, the odds were that marine life would tend to suffer more than their terrestrial cousins, though even at that around 70% of land species died out in the Permian event, hitting even the otherwise imperturbable insects, where eight entire orders went extinct, Stow (2010, 74).
The earlier Ordovician event tracks in second at around 60% of all genera checking out—sparing land life only because back then plants and animals hadn’t actually got out of the seas. See Finney et al. (1999) and Finnegan et al. (2012) on the Ordovician; A. Murphy et al. (2000) on the Devonian with Casier et al. (2002) illustrating how ostracod losses confirmed its global extent; and Erwin (1996), Hoffmann (2000), Jin et al. (2000) and M. Benton (2003) on the Permian event. How the big five extinctions fit in on the larger picture of earth history may be seen in Figure 1 in Downard (2003b, 15), and Eldredge (2014) offers an illustrated take on the issue for a general audience.
[FIGURE 002]Figure 2. Various estimates have been made for the severity of mass extinctions. The percentages here reflect the “Extinction Event” entry at wikipedia.org (accessed 5/3/2010) that drew primarily on a compilation by Baez (2006). There may have been two main pulses in the Ordovician extinction, indicated in the broader date ranges.
It is not unreasonable to expect something rare and distinctive lay at the root of these evident breaks in continuity, and several analyses have detected periodicities in extinctions: a 26 million year cycle identified by David Raup & John Sepkoski (1984) and favored by Davis & Muller et al. (1984)—later replaced by a much longer 62 million year pulse, Kirchner & Weil (2005) re Rohde & Muller (2005). Such leisurely cyclical elements have understandably prompted some to look skyward for their prime suspects. Least likely in this department is the “Nemesis star” theory promoted by Sepkoski and Raup, that the sun has a dark stellar companion that periodically orbits dangerously close. The chief problem with the Nemesis Star scenario is that advocates tried to force the known extinction data into a shoehorn of periodicity (or periodicities, since the initially proposed 26 my cycle would seem to suggest a longer 52 my rate, rather than the 62 my one Muller subsequently culled from Sepkoski’s data) dictated by what was itself a purely hypothetical astronomical cause, Pellegrino (1985), R. Ehrlich (2001, 102-121) and M. Benton (2003, 138-140).
Astronomical factors can affect life on earth, even at the scale of the solar system’s leisurely transit in and out of dust-laden spiral arms, A. Parker (2003, 293-295). More locally, geologist James Croll (1821-1890) studied the climate impact of variations in Earth’s orbital eccentricity in the mid-19th century, refined in the 1940s into the ice age cycling models of mathematician Milutin Milanković (1879-1978), Nield (2007, 109-112) and Hilgen (2010), with recent tweaking noted by Kerr (2013c). The Milankovich cycles not only constrain recent glaciation, Huybers (2011), but appear to modulate global warming periods too, Lourens et al. (2005). Oxygen and carbon isotopes vary in this way, Stow (2010, 18), and differing sunlight intensity by latitude alters oceanic temperatures in complex ways, Philander (2010) re Herbert et al. (2010) and Martin-Garcia et al. (2010). This climatological dance has been going on for a long time, as evidenced by Jurassic sediments, Sha et al. (2015), and others dating back 1.4 billion years, S. Zhang et al. (2015).
While Nemesis is problematic, the asteroid or comet impacts that inspire turgid movie plots (Armageddon pops into mind) are far less so. A devastating impact certainly occurred at Chicxulub in the Yucatan peninsula close to the time of the K-T event (Cretaceous-Tertiary, spelled with a K from the German), Carlisle (1995) and Smit (2008), which contributed a nice oil reserve to boot, Nishimura et al. (2000). Dingus & Rowe (1998, 11-104), Courtillot (1999, 119-134), Lubick (2001) and Palmer (2009, 182-187) place the evidence in larger context. Belcher et al. (2003) questioned just how severe the K-T fireball was, including the experimental check by Belcher et al. (2015), though the analysis of ejecta deposits by P. Schulte et al. (2010) and the accompanying “impact winter” detected by Vellekoop et al. (2014) support its role as an extinction trigger. Keller & Stinnesbeck et al. (2004) also suggested Chicxulub predated the K-T by some 300,000 years, but Pälike (2013) re Renne et al. (2013) confirmed a K-T correlation.
As for other mass extinctions, the giant Mancouagan crater in Canada dates to the Late Triassic, though its extinction effect may have been only localized, Walkden et al. (2002) and Onoue et al. (2012). A side issue concerns how much (or whether) the impact may have contributed to the rise of the dinosaurs, thus opening up adaptive opportunities for the survivors in exactly the way the likes of Phillip Johnson are loathe to imagine: Kerr (2002a) and Thulborn (2003) re Olsen et al. (2002; 2003), Kerr (2003b) on Basu et al. (2003), and more broadly by Fraser (2006, 243-256). More recently another impact contender has appeared: the smaller Rochechouart crater in France, which new dating also puts at the Triassic extinction boundary, Schmieder et al. (2010) with perspective by R. Smith (2011).
A Devonian impact (or even impacts) has also been proposed, though not without criticism, Racki (1996), Sandberg et al. (2002), Ellwood et al. (2003; 2004) contra Racki & Koeberl (2004). A modest impact in the Baltic around 455 Mya appears unconnected with the Ordovician extinction, Suuroja & Põldvere (2004). Perhaps most significantly there doesn’t appear to be a solid impact correlation for the most intense event in the Permian, covered by Kerr (2005a) and Marshall (2005) re P. Ward & Botha et al. (2005) and P. Ward et al. (2005), with the evidence considered still slim by the time French & Koeberl (2010) and Racki et al. (2011) assessed the forensic clues (and their limitations) whereby ancient impact events are identified (from shocked quartz and microtektites to larger crater features heavily eroded in Deep Time).
The snag regarding the role of impact events is well illustrated by the K-T itself, where support for the impact hypothesis grew rapidly in paleontology but was still not universal, Sabath (1996). Outright bolide skepticism such as Dorrik Stow (2010, 176-186) is even rarer these days, with general acceptance of the Chicxulub event of the Switek (2013b, 190-211) form common, but assessing the blast effect forensics (searing heat and vegetation fires, followed by global winter under a prolonged dust-shrouded darkness) trip on the peculiar range of victims and survivors.
Though even the smallest chicken-sized dinosaurs perished, the cold-blooded egg-laying frogs and most crocodiles made it past the Cretaceous, along with pollen-eating moths and light sensitive corals, Fortey (2009, 190-193). Stow (2010, 163-166) likewise noted a complex mix: the planktonic coccoliths (occupying the base of the Cretaceous marine food chain) were decimated, as were 75% of marsupials and birds and 25% of crocodiles, turtles and fish, but the majority of placental mammals survived, as well as lizards and snakes—though more recent work does suggest lizards and snakes suffered an 83% species-level hit after all, Longrich et al. (2012b). Other groups show similar variation: Late Cretaceous mollusc taxa underwent both gradual and abrupt extinction episodes, C. Marshall & Ward (1996), and while some foraminifera seemed to have emerged without much disruption, Alegret et al. (2003) and Alegret & Thomas (2004), the dramatic disruption of the ocean habitat did hit them overall, Keller et al. (2009) and Gallala et al. (2009).
Geerat Vermeij (2010, 64-65) spotted some patterns to the survivors, and not just in the K-T event. Animals that made it through easiest were ones that could hunker down in a crisis, going inert and able to tolerate some starvation until things settled down (crocodiles and turtles, for example), or at least isolate themselves from a stressful environment (clams can shut their shells tightly in a way the more vulnerable brachiopods can’t). Neil Shubin (2013, 137) noted the most common feature of the animals that survive a mass extinction (besides dumb luck) is their distribution range: taxa spread over wider areas stand a better chance of persisting after the crash than niche inhabitants—though again, as with Figure 2 above, it appears to matter at what taxonomical level the animals are viewed at (genera versus species for instance), as explored by Jablonski (2005).
Given these details it is improbable that any single event (intense though they may have been) “causes” a mass extinction. It is more likely multiple factors play a role, where an impact comes along just as the last straw to tip an otherwise unstable arrangement over the edge, as in the “press-pulse” model of mass extinction proposed by Arens & West (2008). It is relevant that the marine side of the Ordovician, Permian and Cretaceous extinctions took place against a background of overall increasing origination of new species, Bambach et al. (2004, 533-535), suggesting something out of the ordinary was stressing the system. Stow (2010, 172-176) recalled that the Cretaceous marine ecosystem was under pressure and undergoing gradual extinctions well before the mass event, and J. Mitchell et al. (2012) suggests much the same for the overspecialized terrestrial realm (such as a predator-prey mix in North America dominated by tyrannosaurs hunting ceratopsids).
Which brings us to another potentially serious culprit: magmatic plume breaches. Spun off by plate subduction, these form volcanic chains like the Cascades in my Pacific Northwest backyard, or as fixed “hot spots” can generate volcanoes along a conveyor belt like the Hawaiian Islands or the many extinct volcanic calderas leading up to the really massive and still dangerously active one in Yellowstone Park.
The dynamics of the Hawaiian hotspot turn out to be quite complex, shifting as the oceanic plate slides over it, Tarduno (2008), Tarduno et al. (2009) and Kerr (2009f) re Wolfe et al. (2009). See Courtillot (1999) on volcanism generally, Condie (2001) and Jackson & Carlson (2011) on mantel plumes, Bindeman (2006) on supervolcanoes, and Achenbach (2009) for the Yellowstone caldera. Such activity has been going on a very long time, of course, at least as far back as the Yellowstone-style eruptions in the Blake River Group in Canada 2.7 Ga, Pearson & Daigneault (2009), or the Warakurna large igneous province in Australia 1.0 Ga, Wingate et al. (2004), though back then only microbial sea life would have been the targets of any changes in oceanic conditions.
It looks far from coincidental, though, that there was massive volcanism during four of the five mass extinctions periods, sometimes trailing on for millions of years and significantly stressing the ecosystem. Oceanic volcanism (as a tectonic island arc collision eventually formed the Ural Mountains) disrupted the late Devonian, D. Chen et al. (2005), Brown et al. (2006) and Pravikova et al. (2008). The Siberian Traps hit the late Permian like a hammer, Browne (1998a-b), Wignall et al. (2009), Ogden & Sleep (2012), and twenty years of geophysical research has led to its acceptance as the main trigger for that extinction event, Kerr (2013e). The Triassic had the Central Atlantic Magmatic Province (CAMP)—recalling though that the “central Atlantic” was then nestled well inland and just opening up as the Pangea supercontinent began to fragment, Olsen (1999) re Marzoli et al. (1999), Rampino (2010) re Whiteside et al. (2010), Schoene et al. (2010), Schaller et al. (2011a-b) with caveats by Rampino & Caldeira (2011), Kerr (2012m), and S. Perkins (2013) re Blackburn et al. (2013) tracking work in this area. Finally, India’s Deccan Traps destabilized the environment as the subcontinent drifted north from its former location parked down by Antarctica, to cross an oceanic hotspot late in the Cretaceous before eventually slamming into Asia in the Eocene to push up the Himalayas, Kerr (2003d) re Ravizza & Peucker-Ehrenbrink (2003), Irving (2008) re Kent & Muttoni (2008), Keller et al. (2009), Kerr (2012i), and Stone (2014c) re Schoene et al. (2015).
The threshold for magma plume danger appears to be how much oceanic crust they recycle, Wignall (2011) re Sobolev et al. (2011). Knowing what modest plumes do beneath Hawaii, Yellowstone and the Cascades, consider the impact of far larger reserves nearer the surface. Besides basalt carpeting hundreds of thousands of square miles, such as Siberia’s Putorana Plateau illustrated by map in van de Schootbrugge (2005,37) and photographically by Palmer (2009, 108-109), their gas emissions (notably carbon dioxide and sulfur, but also chlorine and fluorine) don’t bother just terrestrial life—they hit the marine ecosystem that helps calibrate Earth’s climate.
Magma plume-driven extinctions play out over much longer time frames than splat asteroid impacts, though the biotic extinction event itself may still play out during a narrower window within an overall stressed ecosystem, as appears to have been the case for the Permian event constrained within a geologically very brief 60,000-year window, Erwin (2014) re Burgess et al. (2014). The Siberian Traps continued to spew carbon dioxide, affecting the climate for 5 million years after the extinction and possibly raising tropical ocean temperatures in the Early Triassic to an astonishing 40°C (104°F), Bottjer (2012) re Y. Sun et al. (2012)—though disputed by Goudemand et al. (2013), prompting a tart rejoinder by Sun et al. (2013).
Conditions grew so inhospitable for vertebrate sea life that they had to retreat toward the cooler poles, opening up a niche allowing heat tolerant stromatolite formations (layer cakes of bacteria normally devoured by grazing animals of many types) to make a brief comeback—a situation that appears to have occurred also after the earlier Ordovician extinction, Sheehan & Harris (2004). Over several hundred thousand years the CAMP eruptions ran the roller coaster again in the Late Triassic, with sulfurous clouds cooling the climate competing with CO2 warming it, fueling fire surges that significantly affected plant distribution, van de Schootbrugge (2010) re Belcher & Mander et al. (2010).
Or take Gastaldo et al. (2009) with perspective by Berardelli (2009), reevaluating the major Karoo Basin fossil beds in South Africa suggesting the Permian climate and faunal impact lasted over a hundred thousand years. Incautiously, both Answers in Genesis (2009b) and Brian Thomas (2009c) at the Institute for Creation Research pounced on this work to argue that the longer timeframe actually meant the complete opposite. As Thomas put it: “perhaps the layers that Gastaldo traced were formed from tidal oscillations that occurred while the earth was still underwater during the year-long Flood.”
Thomas didn’t stop to explain how such an event could have preserved the tiny crustacean burrows and fossil footprints the paleontologists found there. The burrows in particular could hardly have been formed along with the rock they are dug in, nor could they be filled in later unless there had been a “later” for that to happen, when the tiny tenant had moved on, Gastaldo & Rolerson (2008). The glib ease with which Young Earth Creationist authors invoke the Flood as a catchall explanation for deposits without examining the finer details are on display in Chapter 3 of Downard (2004).
Sea level fluctuations have also been correlated to marine extinctions, Peters (2008), with Stow (2010,75) noting the formation of the Pangea supercontinent dropped sea levels by 250 meters during the Permian, leading to substantial habitat loss as only 13% of continental shelves remained submerged. The Permian ocean underwent a complex shift as a burst of oceanic anoxia churned large amounts of carbon dioxide into the upper oceans, in turn affecting ocean acidification, Knoll et al. (1996), Payne et al. (2010) and Brennecka et al. (2011), spurred on by the Siberian Traps volcanic activity, Hand (2015c) re Carlson et al. (2015). Oceanic anoxia also occurred in the Cretaceous, though with more controversy about the forensics, Gibbs et al. (2011) contra Erba et al. (2010; 2011), and Higgins et al. (2012).
The release of methane hydrates trapped in ocean sediments may have been another factor, Berner (2002), M. Benton (2008b) and S. Shen & Crowley et al. (2011). Methane hydrates have been proposed for the mysterious Toarcian Oceanic Anoxic Event (TOAE) taking place early in the Jurassic (182 Mya), Hesselbo et al. (2000). Study of the phenomenon has been hampered by the facts of ocean sediment preservation: much early seafloor has simply disappeared due to plate movements in the scores of millions of years since. Tectonic activity overall appears to have been a contributing factor along with carbon cycling oscillations modulated by the broader astronomical Milankovich patterns that affect climate generally, Gröcke et al. (2011), Izumi et al. (2012), and Huang & Hesselbo (2014). Scientists take note of odd events like the TOAE because the concurrent extinction involving marine life may offer clues for assessing the impact of comparable ocean anoxia in modern ecosystems, van de Schootbrugge et al. (2005), Caswell et al. (2009), and Ullman et al. (2014).
Rothman et al. (2014) identified an even more intriguing line of tumbling Permian dominos: the Siberian volcanism increased nickel concentrations to the point where a group of new bacteria (Methanosarcina) that required that metal for their acetoclastic pathway to convert marine organic carbon to methane went on an ecologically disastrous methanogenic binge. Methane emissions appear to have played a similar role in the Triassic, Ruhl et al. (2011), as the overall changes in ocean chemistry reverberating after the Permian event prompted a major shift in photosynthetic phytoplankton from the green superfamily in favor of their biologically distinct red superfamily cousins, Quigg et al. (2003)—yet more instances of adaptive evolutionary changes invisible to antievolutionists ill-suited to climbing down from their doctrinal pedestal to take a closer look.
Peter Ward (2006) noted the further role of dissolved hydrogen sulfide during the Permian, Triassic and Cretaceous events, connected to oxygen depletion in the atmosphere. In that respect the Permian case was the most extreme, with oxygen levels plunging from 30% to only 12% in the early Triassic, Sheldon & Retallack (2002), M. Benton (2003), and Kerr (2005b) re Huey et al. (2005). To put this in perspective, Ward (2005b) calculated that any modern mountaineer time-traveling back to the Triassic (and used to our 21% oxygen level) would have been gasping for air at only 4.5 km altitude (around 14,000 feet). Later in the Cretaceous oxygen levels rebounded, Gale et al. (2001), putting a crimp on the C3 angiosperm plants just appearing then, while allowing the companion C4 plant groups to get a stem up, so to speak—still more adaptive evolution for antievolutionists not to observe.
Such broad fluctuations in atmospheric oxygen levels would have had a profound impact on flammability, Belcher & Yearsley et al. (2010). Below 16% fires are suppressed completely, remain still low until 18.5%, and substantially increased for 19-22%. High-risk periods were during the Carboniferous (350-300 Ma) and Cretaceous (145-64 Ma), intermediate levels during the Permian (299-251 Ma), Late Triassic (285-201 Ma) and Jurassic (201-145 Ma), and lowest Early-Middle Triassic (250-240 Ma). It is of interest that a depletion of oceanic oxygen levels occurred prior to the Devonian extinction (perhaps caused by nutrient-rich runoff from the proliferating land plants disrupting reef communities, more spin-offs from evolutionary adaptation), along with some massive mountain building in the Euramerica continent that formed during the Devonian: the Caledonide and Appalachian ranges, Prehistoric Life (2009, 110-111). Though highly eroded today, the Appalachians would rise to Himalaya heights, with presumably comparable impact on atmospheric circulation patterns.
Even larger climate cycles appear to be playing a role in mass extinctions too, as the seesaw from one mode to another (warm greenhouse to cold icehouse and back again) sets up the ecosystem for a fall. As illustrated in Fortey (2009, 42) four of the five mass extinctions took place at the beginning and ends of either an icehouse (Devonian and Permian) or greenhouse phase (Triassic and Cretaceous), and the fifth (the earlier Ordovician one) occurred during a sharp temperature drop and glaciation phase during what was otherwise a greenhouse period.
That “odd man out” Ordovician 450 Ma may reflect another player. Although there were massive pyroclastic events in the Ordovician, they don’t appear to have triggered the later extinction, Huff et al. (1992) and Botting (2002), but perhaps scientists have missed links that played out over a longer timeframe (remembering also that the farther back you go in time, the less available geological deposits there are to piece together the dynamic puzzle). With animals living in shallower depths being hit the hardest, though, Melott et al. (2004) suggested a rare but catastrophically dangerous gamma ray burst from some nearby star as the culprit, blasting away the protective ozone and explaining both the preferential extinction and concurrent climate change—though see Heim (2008) on the nuances of Ordovician extinction rates, and Vandenbroecke et al. (2010) and Finnegan et al. (2011) clarifying the feedback role and extent of the glaciation occurring then.
If you get the impression that trying to isolate all the many factors that can contribute to shifts in climate and trigger extinction events (mass or otherwise) is no easy trick, you’ve got it right. Genuine scientific reasoning involves exactly that level of caution, dealing with all the varied factors that might come into play. This may be seen from general presentations such as van de Schootbrugge (2005) at the college level to technical papers like Zachos et al. (2001), E. Thomas et al. (2006) and E. Thomas (2007) on the interlocking causes of the Cenozoic PETM, or Steven Stanley (2010) on the affect bacterial metabolism can have on carbon being released or trapped, again with relevance for working out the effects of global warming today. It can even be seen in the popular media as the multidisciplinary approach is reflected in the many competing factors (from impacts and volcanoes to faunal and disease interchanges due to shifting continents) offered in a History Channel science program on the extinction issue, The First Apocalypse (aired in January 2009).
How far a cry is this from creationists like Eric Lyons (2010a) of Apologetics Press, who didn’t even get as far as Thomas and AiG, though, finding it easier to dismiss an Internet account of the debate over the role of the K-T impact rather than dive into the technical literature and show they can making sense of the evidence relating mass extinctions vs. background extinction processes within their own model.
And how much farther still is this from the “scientific” Phillip Johnson over in Intelligent Design land, trying not to understand mass extinctions as a natural phenomenon at all, but bringing them up solely as a blunt instrument to beat back the perceived threat of Darwinian evolution?