The Permian-Triassic Mass Extinction

At the boundary between the Permian and Triassic systems, the diversity of life represented in the fossil record is reduced to a level not seen since the Cambrian. This remarkable event, dubbed 'The Mother of Mass Extinctions,' is by a wide margin the most severe in the history of life on earth, and drastically affected the composition and subsequent evolution of earth's biota. Some of the most important and intensively-studied Permian Triassic boundary sections occur in southern China, Japan, Pakistan, Iran, Greenland, Austria, Sicily, Italy, South Africa, and Antarctica. Systematic collections show that ~50% of families (267 out of 526, Erwin, p. 86), and perhaps as much as 90% of all species known from the late Permian disappear from the fossil record during the latest Permian (Sepkoski, 1986). Post-extinction recovery of plant and marine ecosystems was very slow compared to other exinction events, occuring over ~4-5 million years (Erwin, 2001; Retallack et al., 1996; Looy et al., 1999). Radiolarite deposition shut down in the oceans for several millions years. Coal and reefs are absent altogether from the earliest Triassic also, resulting in reef and coal 'gaps' that are unique in the Phanerozoic.

While the late Permian geologic record contains diverse and endemic faunas and floras, early Triassic faunas and floras are remarkably low-diversity and remarkably cosmopolitan. Life on the early Triassic earth was remarkably homogenous, as a handful of taxa which survived the extinction proliferated across the planet, for instance the brachiopod Lingula, the bivalves Claraia, Eomorphis, and Unionites, the weedy plant Isoetes, and the mammal-like reptile Lystrosaurus. Also among the inhabitants of the post-extinction world were peculiar gastropods and bivalves only a few mm or less in diameter. This essay summarizes some of the extensive research on this profound event.

Late Permian paleogeography. The giant V-shaped ocean in the east is the Tethys Sea. From the Paleogeographic Atlas Project

Marine Extinctions

Marine invertebrates were by the hardest-hit by the P-Tr extinction. In boundary sections preserving a record of the P-Tr transition, large numbers of species disappear over few meters of sediment or less. In the intensively-sampled south China boundary sections, for instance, 280 out of 329 marine invertebrate genera disappear within the final 2 conodont zones of the Permian (Jin et al., 2000). At the Meishan boundary stratotype, most of these genera disappear within ~50cm of the boundary (Bowring et al, 1998).

Stratigraphic ranges of 333 late Permian - early Triassic fossil species from the Meishan type section, plotted against 13Ccarb profile and lithostratigraphic position. A is plotted against thickness, and B is scaled with time, based on radiometric date from Bowring et al (1998). From Jin et al., 2000.

Systemic fossil collections from P-Tr sections worldwide show that the end-Permian extinction wiped out many taxa, including all the remaining trilobites, all of the fusulinid and 94% of the nonfusulinid foram genera, graptolites, all of the blastoids, acanthodians, rugose and tabulate corals, pelycosaurs, 98% of the crinozoans, including all of the inadunates and camerates, 96% of the anthozoans, virtually all of the radiolaria, 96% of the brachiopod genera, including all of the Orthids and productids, 85% of gastropods, 59% of the bivalves 79% of the bryozoans, including all of the fenestrates, trepostomes, and cryptostomes, 8 families of ostracods, 90% of gastropod genera and 3 of 16 gastropod families, 97% of the ammonoids, and others (McKinney, 1987; Lemon, 1993, p. 318; Erwin 1994; Hallam and Wignall, 1997; Yang et al, 1994).

Brachiopods were the dominant shelly benthic animals in late Paleozoic deposits. The diversity and sheer physical volume of brachiopod shells in the Paleozoic geologic record is astounding. Their dominance came to an abrupt end in the late Permian, when~90% of families and 95% of genera became extinct. Whereas hundreds of species are known from the mid-late Permian, only a few species of brachiopod occur in earliest Triassic deposits. One of these survivors, Lingula, briefly profilerated in the earliest Triassic (Rodland and Bottjer, 2001), and is still alive today. Although very rare today, Lingula has been around since the Cambrian, and is perhaps the oldest extant animal genus.

Bivalves, althogh present in the Paleozoic as inconspicuous elements, diversify following the P-Tr extinction. It is as if bivalves 'took over' the ecological niches vacated by the brachiopod extinctions. As with other taxa and other extinctions, the earliest Triassic bivalve fauna was cosmopolitan -- early Triassic bivalve faunas are dominated worldwide by the four genera Claraia, Eumorphotis, Unionites and Promylina, constituting what Hallam and Wignall call "the most cosmopolitan assemblage of all time" (p. 108). Hallam and Wignall describe this shift from brachiopod to bivalve shelly benthos a "fundamental legacy of the late Permian mass extinction" (p 105).

Rugose corals, which were fairly widespread during the late Permian, were wiped out. Scleractinian corals, on the other hand, first appear in the fossil record in the Triassic.

Over 33 genera of ammonoids disappear near the P-Tr boundary. In South China sections, an estimated 20 of 21 genera disappear during the latest Permian (Yang, 1993). On the other hand, nautiloid cephalopods were scarcely affected.

Radiolaria extinction was very nearly complete. Racki (1999) notes that "recent studies of sequences in Japan, Canada, China and Turkey document a crash in biosiliceous productivity, spectacularly recorded in a radiolarite gap across the Permian-Triassic boundary . . . All radiolarian morphologies disappeared from the record after the P-T boundary event, except simple spherical spumellarians," and that only 3 radiolarian species are known from the early Triassic (p. 112, refs omitted).

Recovery of marine ecosystems following the P-Tr event progressed very slowly. Reefs, which were abundant in the Permian (for instance, the massive late Permian reef at Laolongdong in Sichuan, China and the massive Capitan reef complex in Texas), disappear near the boundary and do not reappear until the middle Triassic (Flugel, 1994). At the same time, stromatolites spread into many "normal" marine environments during the early Triassic, for the first time since the Ordovician (Schubert and Bottjer, 1992). Stromatolites make a handy food source for benthic marine grazers, and their presence in early Triassic normal marine deposits was probably facilitated by the extinction of shallow marine fauna that would normally consume them.

Extinctions on Land

Tetrapod faunas were hard hit by the P-Tr extinction event. 21 tetrapod families (63%), disappear at or near the P-Tr boundary (Benton, 1989). The Karoo Basin in South Africa is one of the most intensively studied terrestrial P-Tr boundary sections. The P-Tr boundary in this basin can be defined by the last occurence of the dicynodont Dicynodon lacerticeps, which occurs just below the large end-Permian d13C excursion (see below). Sampling of the abundant vertebrate taxa here shows that only 6 of 44 reptilian taxa known from the late Permian Dicynodon zone are also present above the P-Tr boundary. For instance, the abundant late Permian taxa Pristerodon, Cyonosaurus, Diictodon, Rubidgea, Palanomodon, Theriognathus, and Cynognathus disappear within ~50m of the boundary. Above the boundary, within about 40m, the taxa Proterosuchus, Micropholis, Galesaurus, Thrinaxodon, Procolophon, and Lydekkerina appear.

Lystrosaurus actually first appears within the Dicynodon zone, but only becomes abundant post P-Tr. This shift to Lystrosaurus-dominated vertebrate faunas is seen in early Triassic sections worldwide, including Antarctica, India, China, Mongolia, Siberia and eastern Europe (Lozovsky, 1998; Lucas, 1998, 1999). In fact, the cosmopolitan distribution of Lystrasaurus was one of the pieces of evidence cited by Wegener, DuToit and others for the former existence of Pangea. Lystrosaurus is so abundant and widespread, that this genus alone constitues as much as 90% of some earliest tetrapod faunas (Hallam and Wignall, p. 111). This is another example of a low-diversity but widespread post-extinction fauna. The first archosaurs appear in the early Triassic as well, followed in the mid-triassic by the earliest dinosaurs.

"The zones of the Beaufort group are named after certain characteristic fossil reptiles, each of which is found only in that particular zone. There is recent evidence that the Endothiodon Zone and lower part of the Cistecephalus Zone should be united as a new Cistecephalus Zone, and that the upper part of the old Cistecephalus Zone should be regarded as a seperate Daptocephalus Zone." From The South African Museum.

Another interesting aspect of the Karoo PTB is that it is marked by a pronounced change in depositional conditions. Ward et al. (2000) note that the Karoo PTB is marked by "a rapid and apparently basin-wide change from meandering to braided river systems," which is argued to be a result of the rapid die-off of sediment-binding vegetation within the basin. Before the origination of land plants in the Silurian, braided fluvial systems were the norm. Similar changes are noted in other PTB sections from the South Devon and Sydney basins (p. 1742).

Plant Extinctions

Palynological data from Israel, China, the southern Alps and elsewhere show that many Permian pollen types disappear near the P-Tr boundary (Erwin, p.129; Retallack, 1996; Looy et al., 1999, 2001; Twitchett et al., 2001). Looy et al (1999) cite palynological evidence from Europe, Africa and Asia showing that the dominant Permian conifer taxa were replaced post-extinction by a few groups of surviving lycopsids, especially Pleuromeia. Lycopsids living today often function as pioneer plants, recolonizing disturbed areas (Pfefferkorn, 1999). Hallam and Wignall state that:

"The floral mass extinction selectively removed the dominant large plants of the Late Permian and left small, weedy survivors. One such weed, the quillwort Isoetes, a small pioneering plant of the oligotrophic lakes and ponds, formed the stem group for a moderately successful Triassic radiation. The seed ferns, typified by Dicrodium, similarly proliferated. Isoetes and Dicrodium can clearly be added to the lists of the Claraia and Lystrasaurus assemblages as Scythian disaster taxa. . ." (p. 112).

However, after the P-Tr extinction, these lycopsids remain the dominant component for several million years. Conifer-dominated forests return only ~4-5Ma year after the extinction. "Throughout Europe, plant megafossil information indicates that ecosystem recovery to precrisis levels of structure and function did not occur before the transition between the Early and Middle Triassic. A replacement of the Pleuromeia vegetation by coniferous forests is evidenced by the Voltzia-dominant communities that characterize the early part of the Middle Triassic" (p. 13857).

One consequence of the P-Tr floral extinction was a unique, early Triassic hiatus in coal formation, dubbed the "coal gap" (Veevers et al., 1994; Retallack et al., 1996). Whereas thick coals are widespread from the Carboniferous to the late Permian, and from the middle Triassic onwards, there are no coal seams at all in the early Triassic. Retallack et al. (1996) notes that extensive Permian coals "in Siberia, China, Australia, South Africa, and Antarctica disappeared along with their distinctive voynovskyalean and ruflorialean cordaites and glossopterid seed ferns," replaced in the early Triassic by a widespread, low-diversity conifer and lycopsid flora (p. 205). Thin coals reappear by the mid-Triassic, followed by coals comparable to those of the late Permian. Retallack et al. argue that the ~10Ma coal hiatus reflects the gap between the extinction of peat-forming plants at the P-Tr boundary, and the appearance of new plant groups in the mid-Triassic tolerant of the dysaerobic, acidic environments in which peats accumulate.

Another interesting aspect of the P-Tr in continental [and marine] sections is a short-lived abundance of fungal material. While fungi are present throughout the Phanerozoic, there is a very abrupt but short-lived enrichment of spores and fungal material. This "fungi" spike has been reported from marine and terrestrial P-Tr sections in Greenland, the Zechstein Basin, the southern Alps, Israel, Australia, the Karoo Basin in South Africa (Steiner, 2001), and other locations. Visscher et al. (1996) note that this fungal spike occurs "irrespective of depositional environment (marine, lacustrine, fluviatile), floral provinciality, and climate zonation. This fungal event can be considered to reflect excessive dieback of arboreous vegetation, effecting destabilization and subsequent collapse of terrestrial ecosystems with concomitant loss of standing biomass" (p. 2155). In the two sections described in detail by Visscher et al. (1996), one in Italy and one in Israel, fungal components increase abruptly to over 90% of the palynomorph assemblages right at the P-Tr boundary.

Biostratigraphic Correlation of end-Permian Marine Deposits

The P-Tr boundary in marine sediments is defined by the first occurence of the conodont Hindeodus parvus. The P-Tr boundary can also be defined by the first occurence of the bivalve Claraia or the ammonoid Otoceras woodwardi. Since Meishan, China section has been chosen as the P-Tr boundary stratotype, we will list the succession of conodont zones found there, and then see how other sections around the world correlate with Meishan (data from Lai et al., 2001; Yin and Tong, 1998).

Meishan I. isarcica H. parvus H. typicalis C. changxingensis C. deflecta C. subcarinata

Although conodonts are by virtue of their good preservation, nektonic habit, and wide distribution the best zonal fossils for global correlation of the P-Tr boundary, other fossil groups can also be used to mark the boundary. For instance, the condont zones H. parvus and I. isarcica are correlated with the bivalves Pseuodoclaraia wangi and C. griesbachi, followed by the first occerences of C. aurita, C. stachei, and E. multiformis. We find the same sequence, in whole or in part, just above the PTB in sections from the Alps, the western US and Canada, Kashmir, Siberia, Arctic Canada, Greenland, Spitzbergen, and elsewhere, in association with the same zonal conodonts (Yin, 1985). Yet another useful zonal fossil group is the foraminifera. For instance, the latest Permian sediments in south China contain the fusulinid genus Paleofusiella. Again, this same genus is found in latest Permian deposits in Japan, north Vietnam, Tibet, and the Carnic Alps, Austria (Kobayashi, 1999).

Stable Isotope Stratigraphy at the P-Tr Boundary

Stable isotope ratios can be an excellent tool for correlating deposits globally. Correlation by stable isotopes is possible because the isotopic ratios of O, C, S. Sr and other elements in seawater varies or "evolves" over time, as a result of several different processes. Since the shells of marine organisms and inorganic calcites are precitated from seawater, variations in the isotopic composition of ocean water are recorded as variations in the isotopic composition of inorganic and organic carbon deposited at that time. Boggs (1987 p. 688) writes:

Variations in the relative abundance of certain stable, nonradioactive isotopes in marine sediments and fossils can be used as a tool for chronostratigraphic correlations of marine sediments. Geochemical evidence shows that the isotopic composition of oxygen, carbon and sulpher in the ocean has undergone large fluctuations, or excursions, in the geological past -- fluctuations that have been recorded in marine sediments. Because the mixing time in the oceans is about 1000 years or less, marine isotopic excursions are considered to be essentially isochronous throughout the world. Variations in isotopic compositions of sediments or fossils allow geochemists to construct isotopic composition curves that can be used as stratigraphic markers for correlation purposes.

Carbon Isotopes

As with other extinctions, a major and abrupt 13C isotopic excursion has been documented in numerous PTB sections distributed across the entire earth, including Greenland, West Spitzbergen, the Carnic Alps Austria, Slovenia, northwest Iran, Armenia, Turkey, Nepal, Pakistan, New Zealand, several locations in China, and elsewhere (Baud et al. 1989; Magaritz et al. 1983; Dolenec 1996, 1998; Krull et al. 2000; Twitchett et al. 2001). Similar carbon isotope excursions have also recently been documented from P-Tr marine boundary sections in Slovenia (Dolenec et al., 2001), Japan (Musashi et al., 2001), and New Zealand (Krull et al., 2000). The shape of the curve in the most complete sections shows a relatively gradual decline beginning in the late Permian, with a sharp negative decline right at the boundary. Evidence presented by Rampino et al. (2000) from the Austrian GK-1 core indicate that the sharp drop at the P-Tr developed within less than 3-40ky, which is substantially shorter than the U-Pb estimate of Bowring et al. (1998) of less than 165k yr.

It is important to note that this excursion shows up in both carbonate carbon [d13C(carb)], such as brachiopod and conodont shells and corals, but also in organic carbon such as kerogen [d13C(org)], which can is extracted from carbonates or other sedimentary rocks by leaching away carbonate. As noted below, the negative excursion even shows up in the bone apatite of terrestrial vertebrates.

From Erwin, 1994, p. 195. del 13C curves from Kuh-e-Ali Bashi, northwest Iran, Vedi, Armenia, Emarat, Iran, Nammal Gorge, Pakistan, Thakkola, Nepal, and Shangsi, China. del 13C curves from many other PTB sections around the world document the same negative excursion. More recent work using more sophisticated analytic procedures produce curves with even less scatter -- compate for instance the curves from Austria and Meishan.

Although marine sections contain the most complete records of this carbon isotopic shift, it has also been documented in terrestrial PTB sections, for instance in paleosols and/or organic carbon from PTB sediments of the Karoo Basin, South Africa (Smith and Ward, 2001), Antartica (Krull and Retallack, 2000), and Australia (Retallack, 1999). One interesting aspect of the Karoo d13C profile is that two were constructed, one using the tusks of the mammal-like reptiles Dicynodon and Lystrosaurus, and another using paleosol carbonate nodules. The two resulting curves covary, with a large negative excursion occurring in the overlap zone of these two taxa. [In the portion of the section containing both Dicynodon and Lystrosaurus occur, they both yield similar values.] As a side note, one of the Antarctic early Triassic paleosols described by Krull and Retallack is a deeply weathered, well-developed Ultisol which is estimated to have formed over ~40-60Ky.

As Erwin notes, "the isotopic signatures are so similar from sections ranging from restricted basins to open marine that the only reasonable conclusion is that the major shifts are globally synchronous events" (p. 198). As noted above, this shift has now been documented in carbonate and organic carbon in terrestrial sediments as well.

What caused this negative carbon isotope excursion? Part of the excursion may have been caused by the release of volcanic CO2, or by the oxidation of buried organic matter as a result of the Changxingian regression. However, these alone do not explain the major spike seen at the P-Tr, since this spike developed very quickly (~>40k yrs; Rampino et al. 2000).

A shift of the magnitude and rapidity seen at the P-Tr would seem to require the addition of isotopically very light carbon to the ocean-atmosphere system. CO2 from volcanic gasses (about -5) would not produce such a large shift. The oxidation of buried sedimentary organic matter (-20 to -25) could produce the shift, but it doeesnt seem likely to explain the rapid spike seen at the P-Tr, since calcuations show that 6500-8400 gigatons would have to be oxidized and returned to the ocean-atmosphere system within less than a few hundred thousand years to produce the observed isotopic shift (Erwin, 1994, 200-204).

Methane carbon in the form of gas hydrates, on the other hand, is isotopically very 'light,' meaning that has a very low carbon C13/12 ratio. Gas hydrate is a "solid, ice-like substance, composed of rigid cages of water molecules that enclose molecules of gas, mainly methane" (Kevenholden). Massive amounts of methane hydrates (1x10^19 - 2x10^19g) exist in continental slope sediments today, at depths greater than ~400-1000m. It is assumed that a similar amount existed in the Permian. Methane has a delta 13C value of -60, and calculations show that the P-Tr carbon isotope changes could be produced by dissociation of between 10 and 25% of existing resevoirs.

The other explanatory advantage is that, unlike the buried organic carbon resevoir, the methane resevoir can be rapidly discharged into the ocean-atmosphere system, by oceanic warming. Converted to the gaseous state, the methane rapidly diffuses into the ocean and the atmosphere, is oxidized by methanotropic bacteria to form light CO2 and water [CH4+2O2->2H20+CO2]. The light CO2 is mixed with the oceans and atmosphere and incorporated into carbonates and organic matter, thus producing the observed carbon isotopic shift.

Methane and CO2 are both greenhouse gasses. Thus, once methane begins to dissociate, the ocean-atmosphere system warms even more, releasing even more methane, etc., and a positive feedback is initiated. Perhaps the warming was initiated by volcanism, and the coup de grace was delivered by methane dissociation.

Oxygen Isotopes

Oxygen isotopes (18O/16O) have also been studied from several P-Tr boundary sections (e.g. Holser et al., 191). These isotopic profiles also show a major shift towards negative values at the P-Tr. If the oxygen isotope excursion is interpreted solely as a temperate effect, this would indicate a rapid warming of about 6 degrees C near the equator. Paleobiological evidence also supports a brief "greenhouse" phase in the early Triassic. For instance, the disappearance of cold-adapted Glossopteris flora, and its replacement by floras typical of lower paleolatitudes (Dobruskina, 1987; Retallack, 1996; Ziegler, et al., 1991), the migration of calcerous algea to Boreal latitudes (Wignall, 2001), and the wide spread of latitudes over which Lystrasaurus and other early Triassic vertebrates are found. There is also paleosol evidence for warming. In Antarctica and Australia, late Permian coals are replaced in the earliest Triassic by intensely leached paleosols, characteristic of much lower and warmer paleolatitudes (Retallack, 1996, 1999; Retallack et al., 1996).

Strontium Isotopes

Strontium isotope stratigraphy of the P-Tr boundary also supports a very warm earliest Triassic period. Montanez et al. (2000) explains the controls on 87Sr/86Sr ratios in seawater:

Temporal variation in seawater ⁸⁷Sr/⁸⁶Sr is governed mainly by changes in the balance between Sr fluxes to the ocean from continental weathering and hydrothermal fluid-rock interaction at mid-ocean ridges (Palmer and Edmond, 1989). The ⁸⁷Sr/⁸⁶Sr composition of the volumetrically smaller seafloor hydrothermal flux is buffered at low values of ~0.703–0.705. Conversely, continental weathering, via riverine and groundwater fluxes, contributes Sr with high ⁸⁷Sr/⁸⁶Sr values (0.709–0.730) to the ocean. It is primarily the variation in the larger continental flux and its isotopic composition that has affected secular change in seawater ⁸⁷Sr/⁸⁶Sr values. The ⁸⁷Sr/⁸⁶Sr values of unaltered marine carbonate minerals directly record seawater ⁸⁷Sr/⁸⁶Sr values, given the negligible fractionation of Sr isotopes (Banner and Kaufman, 1994) and the homogeneity of Sr in seawater (DePaolo and Ingram, 1985).

Hallam and Wignall (1997) summarize Sr isotope changes across the P-Tr:

The 87Sr/86Sr trend during the Permian was one of long-term decline culminating in lowest values of 0.7066 late in the [early late Permian] followed by an increasingly rapid rise through the Changxingian and into the earliest Triassic. The rate of increase in the early Scythian is the highest known from the Phanerozoic and for once there is remarkable unanimity as to the ultimate cause of these changes. . . the rapid rise in of 87Sr/86Sr records a major increase in the global rate of continental weathering - a conclusion further supported by Nd isotopic data (p. 130).

This is consistent with a major increase in CO2 or temperature or both at the P-Tr boundary.

Timing and Tempo of the P-Tr Extinction

There are a few quality dates available to bracket the timing the P-Tr event. At the Meishan type section and other P-Tr sections in south China, several widespread volcanic ash beds occur near the paleontologically-defined boundary, allowing the boundary to be dated. An ash layer immediately above the boundary at Meishan yields a Rb/Sr sanidine age of 250 ± 6.0Ma, a U/Pb SHRIMP zircon age of 251.1 ± 3.4 Ma, and ⁴⁰Ar/³⁹Ar sanidine age of 249.9 ± 1.5 Ma (Bowring, 1998 and their refs 10,11, and 12). Bowring et al present additional U/Pb zircon dates from Meishan yielding an age of 251.4 ± 0.3. Dates published by Renne et al. (1995) indicate that the massive Siberian Traps flood basalt has an age indistinguishable from that of the ash bed at the P-Tr boundary in Meishan. This is consistent with biostratigraphic data also, which show that the sediments underlying the traps are late Permian in age.

Other bracketing dates include a U/Pb SHRIMP date of 253.4 ± 3.2 Ma for a tuff in the late Permian Ingelara Formation of central Queensland, Australia (Roberts et al., 1996), a U/Pb SHRIMP date of 265.3± 0.2Ma for an ash bed at the base of the Capitanian (early late Permian) in Texas (Bowring et al., 1998), a late Anisian tuff from New Zealand 40Ar/39Ar dated to 242.8 ± 0.6 (Retallack et al., 1993), a basal Ladinian tuff from northern Italy U/Pb dated to 241.0±0 (Mundil et al., 1996), and several Rb/Sr and 40Ar/39Ar dates from the the late Permian/early Triassic Ochoan series of Texas (Long et al, 1997; Sharp et al., 1997). Although these dates constrain the timing of the P-Tr extinction, the resolution is not high enough to to resolve events on, say, the level of 50,000 years.

U/Pb SHRIMP dates, d13C, and biostratigraphy at the Meishan section, south China. From Bowring et al. 1998.

A recent paper by Rampino et al. (2000) attempts to estimate the duration of the extinction event using Milankovitch cycles. Milankovitch cycles are cyclic variations in the amount of solar energy recieved by earth over time as a result of cyclic changes in the orbital parameters of eccentricity, obliquity, and precession. These cycles are recognized in numerous Quaternary climate proxy records (e.g. Hayes et al., 1976), and are increasingly being recognized in the pre-Quaternary sedimentary record (e.g. de Boer and Smith, 1994; Sageman et al., 1997; Olsen and Kent, 1999; Zachos et al., 2001).

Rampino et al. (2000) provide evidence for Milankovitch cycles in P-Tr marine sediments from the Carnic Alps, Austria, which they use to constrain the duration and tempo of the extinction event. They report cycles in the ratio ~40:10:4.7:2.3m, which they identify with the Milankovitch cycles of ~412:100:40:20Ky. Based on this assumption, the earliest Triassic (Griesbachian) would have been deposited over ~1.4-1.6Ma, which is consistent with estimates of ~1.6Ma for the Griesbachian based on radiometric dates (p. 645). Also notable is that the depositional rate inferred from the Milankovitch cycles, ~10cm/Ky, is consistent with measurements of depositional rates in modern shallow-water carbonate environments. Using this cyclostratigraphic framework, Rampino et al. (2000) estimate that the P-Tr faunal shift occurred in less than 60Ky, possibly much less, and that the major negative carbon isotopic excursion, seen worldwide in P-Tr boundary sections, persisted for ~480,000 years into the earliest Triassic.

Oceanic Anoxia During P-Tr Time

Several different pieces of evidence indicate that the oceans became anoxic or nearly so during the latest Permian and earliest Triassic. Twitchett (1999) notes that the "marine Permian-Triassic has been studied in South China, Greenland, Artic Canada, USA, Pakistan, and Japan amongst many other regions of the world. Every single studied section in each of these areas shows evidence of Griesbachian [earliest Triassic-ed] low oxygen conditions. . ." (p. 35). Anoxic or dysoxic conditions may have played a major role in the marine extinctions, and was surely a factor in very slow post-extinction recovery. Data presented by Isozaki and others demonstrates that this anoxia began up to several million years prior to the PTB, and persisted up to several million years after the PTB, returing to normal levels only by the early middle Triassic (e.g. Kato et al., 2002).

For instance, in deep-water marine P-Tr section in Japan, British Columbia (Isozaki, 1997; Kato et al., 2002) and elsewhere (Twitchett, 1999), early to late Permian and mid to late Triassic sediments consist of reddish, hematite-stained radiolarian cherts, while sediments deposited during the late Permian-early Triassic are grey-black carbonaceous mudstones which lack hematite completely and are almost completely devoid of radiolaria. Also present in this interval are framboidal pyrite grains, which are indicators of anoxic conditions.

Geochemical evidence also suggests that anoxic conditions developed in the oceans during P-Tr time. For instance, changes in Cerium fractionation in marine sediments. Erwin writes:

"Normally all the REE except for cerium exhibit the same chemical behavior. Cerium is normally depleted relative to lanthanum and neodymium in oxidizing seawaer and coprecipitates with metallic oxides. As the partial pressure of oxygen in the system drops, this cerium anamoly disappears and the relative abundance of cerium approaches that of the other REE. In the conventional notation for cerium abundances, the normal depletion of cerium produces a negative value; zero or positive values occur under anoxic to dysaerobic conditions" (p.205).

Analyses of REE in phosphatic microfossils from P-Tr sections suggests widespread anoxic conditions during P-Tr time (Wright et al., 1987; Wignall and Hallam, 1992). A similar phenomena occurs with Th/U ratios. According to Wilde et al, low Th/U ratios indicate anoxic depositional environment because "under anoxic conditions U is tetravalent and forms insoluble compounds which remain in the sediments while in oxic environment, U changes to hexavalent state and forms the soluble uranyl carbonate." In normally oxygenated marine depositional environments, the Th/U ratio will be ~2-7, while values less than 2 indicate low oxygen at the time of deposition. This ratio in P-Tr boundary marine sediments is often found to be very low (Wignall and Twitchett, 1996; Twitchett, 1999).

There is also paleobiological evidence for dysoxia, manifested as a dramatic proliferation of 'microbivalves' and 'microgastropods,' as well as in changes in trace fossil activity by these marine organisms. Many boundary sections contain an abundance of these tiny bivalves and gastropods, from 1 to a few mm(!) in diameter. Hallam and Wignall note that "the P-Tr boundary marks a curious, temporary change in the nature of the gastropod fossil record . . . microgastropod grainstones composed of mm-sized species are found throughout all equatorial Tethyan sections and extend into Perigondwanan sections of Pakistan . . . Gastropods returned to 'normal' in the Anisian as they once again became rather rare and attained sizes more typical of the group" (p. 107, see fig. 5.5).

This phenomenon is mirrored in the trace fossil record. At several P-Tr boundary sections (e.g. Italy, Greenland, China) a diverse assemblage of late Permian trace fossils is replaced near the PTB by a low diversity, low density assemblage, followed in the mid Triassic by a return to to late Permian levels of bioturbation. For instance, in the Dolomites, Italy (Twitchett, 1999), 5 ichnogenera are recognized in late Permian Bellerophon Formation - Zoophycos, Rhizocorallium, Skolithos, Diplocraterion, and Planolites. All but Planolites disappears near the boundary, and those which are present have a very smal burrow diameter (average ~1.5mm, vrs. average 10mm in late Permian). The same phenomena occurs at other boundary sections, for instance Jameson Land, east Greenland (Twitchett et al., 2001). The same relationship between burrow diameter/density and dissolved oxygen is seen in modern depositional environments as well (Savrda et al., 1984). [At both the Italy and Greenland sections, the marine fauna extinction level is succeeded by a zone containing large proportions of the foram Earlandia.]

A Possible Explanation?

In contrast to the K-T extinction, the cause(s) of the P-Tr extinction are not well understood. Bolide impact has been proposed, but the evidence for this is ambiguous, especially compared to the massive and unequivocal evidence for a major bolide impact at the K-T boundary. There is some indication, however, that large-scale volcanism may have played a role.

Several of the classic P-Tr sections in South China are capped by 3-6cm thick altered volcanic ash layers containing bypyramidal quartz, melt spherules, glass shards and isotopic ratios typical of siliciclastic rather than basaltic volcanism. The volume of these ash layers is estimated to be about 1000km3, equal to the volume of water flowing through the Yangtze River in one year (Zhou and Kyte, 1988). Analysis of 40Ar/39Ar data from two tuffs in southern China yield dates of 250.0 +/- 0.2 million years ago for the Permian-Triassic boundary, which is comparable to the inception of main stage Siberian volcanism at 250.0 +/- 0.3 million years ago (Renne et al., 1995). This is consistent with biostratigraphic data -- the basal tuffs of the Siberian volcanics are interbedded with midle late Permian fossils (Kozur, 1998).

The P-Tr boundary occurred at roughly the same time as the extrusion of the largest known Phanerozoic flood basalts, the Siberian Traps. Laser-heating 40Ar/39Ar plateau dating indicates that the bulk of these basalts was erupted over an extremely short time interval at mean eruption rates greater than 1.3 cubic kilometers per year (Renne et al. 1991). These volcanic flows presently cover an area of 337,000 square kilometers. They are estimated to have a volume of about 1.6-2.5 (estimates vary) million cubic kilometers of solidified basaltic lava. Spread evenly over the earth's entire surface, this volume of lava would produce a layer 10 feet thick.

Volcanism on this scale would release massive amounts of CO2 and SO2, as well as aerosols that would block a significant amount of sunlight. Initially, this would result in cooling. However, the SO2 would leave the atmosphere in the form of acidic rain, and within a few months most of the particulate matter would be gone from the atmosphere. This may have played some role in the extinctions on land. However, the CO2 would remain, and this would result in warming. Assuming a volume of 2 x 10^6 km3 of basalt, and release of 5 x 10^12g CO2 per km3 of basalt, then the Siberian Traps would have released 1 x 10^19g of CO2 (Wignall, 2001).

Warming as a result of volcanic CO2 may have resulted in the dissociation of methane hydrates, and the development of anoxic conditions in the oceans. Warming promotes anoxic is two ways. First, the solubility of O2 in water decreases with increasing water temperature. Second, warming can promote anoxia if the equator-to-pole temperature gradient is weakened, since this would weaken oceanic circulation (Hallam and Wignall, 1997, p. 141). Finally, warming via volcanic CO2 may have caused the dissociation of gas hydrates, which would cause even more warming.

Study of other mass extinction intervals shows that several are correlated with both flood basalts and anoxic episodes. Of 11 major flood basalts, 7 coincide with some form of extinction episode (Wignall, 2001). 2 of these extinction events, the Toarcian (Jurassic) and latest Paleocene, are similar in many ways to the P-Tr extinction. The Toarcian extinction occurred at the same time as the Karoo-Ferrar basalts in South Africa and Antarctica were extruded (183 +/-1Ma), and the late Paleocene 'event' occurred at the same time as the extrusion of the Brito-Arctic flood basalts (55Ma). Like the end-Permian, both events are associated with warming, marine anoxia, major carbon isotope excursions, and the preferential extinction of benthic marine organisms (Wignall, 2001). Methane release has also been proposed for both events (e.g. Hesselbo et al., 2000; Katz et al., 1999).

References

Benton, M.J., 1989. Mass extinctions among tetrapods and the quality of the fossil record. Philosophical Transactions of the Royal Society of London B325, pp. 369-386.

Bowring, S. A., Erwin, D. H., Jin, Y. G., Martin, M. W., Davidek, K., Wang, W., 1998. U/Pb Zircon Geochronology and Tempo of the End-Permian Mass Extinction. Science 280: 1039-1045.

De Boer, P.L. & Smith, D.G. (eds.), 1994. Orbital forcing and cyclic sequences. Int. Assoc. Sedim. Spec. Publ. 19, 559 pp.

Dobruskina, I. A., 1987. Phytogeography of Eurasia during the early Triassic. Palaeogeography, Palaeoclimatology, Palaeoecology 58, pp. 75-86.

Dolenec, T. & Ramovs, A., 1998. Isotopic changes at the Permian-Triassic boundary in the Idrijca Valley (W. Slovenia). Materiali in okolje 3–4:405–411.

Dolenec, T., Lojen, S., Ramovs, A., 2001. The Permian-Triassic boundary in Western Slovenia (Idrijca Valley section): magnetostratigraphy, stable isotopes, and elemental variations. Chemical Geology 175, pp. 175-190.

Erwin, D.H., 1993. The great Paleozoic crisis; Life and death in the Permian. Columbia University Press.

Erwin, D.H., 2001. Lessons from the past: Biotic recoveries from mass extinctions. Proceedings National Academy of Sciences 98: 5399-5403.

Eshet, Y, M.R. Rampino and H. Visscher (1995). Fungal event and palynological record of ecological crisis and recovery across the Permian-Triassic boundary. Geology, 23, 967-970.

Faure K de Wit MJ Willis JP 1995 Late Permian global coal hiatus linked to 13C depleted CO2 flux into the atmosphere during final consolidation of Pangea. Geology 23; 507-510.

Flugel, E., 1994. Pangean shelf carbonates: controls and paleoclimate significance of of Permian and Triassic reefs, in Klein, G. de V., Pangea: paleoclimate, tectonics and sedimentation during accretion, zenith, and breakup of a supercontinent. GSA Special Paper 288, pp. 247-266.

Hallam, A., and Wignall, P.B., 1997. Mass Extinctions and their Aftermath. Oxford University Press, Oxford.

Wignall, P.B., 2001. Large Igneous Provinces and mass extinctions. Earth-Science Reviews, v. 53, p. 1-33.

Hayes JD, Imbrie J, Shackleton NJ, 1976. Variations in the Earth's orbit: Pacemaker of the ice ages. Science 194:1121-1132.

Hesselbo S.P., Gröcke D.R., Jenkyns H.C., Bjerrum C.J., Farrimond P., Morgans Bell H.S. and Green O.R., 2000. Massive dissociation of gas hydrate during a Jurassic oceanic anoxic event. Nature 406, 392-395.

Holser, W.T., Schoenlaub, H.-P., Attrep, M., Boeckelmann, J., Klein, P., Magaritz, M., Orth, C.J., Fenninger, A., Jenny, C., Kralik, M., Mauritsch, H., Pak, E., Schramm, J.-M., Stattegger, K. & Schmöller, R., 1989. A unique geochemical record at the Permian/Triassic boundary. Nature, 337 (6202): 39-44.

Holser, W.T., Schönlaub, H.P., Boeckelmann, K. & Magaritz, M., 1991. The Permian-Triassic of the Gartnerkofel-1 core (Carnic Alps, Austria): synthesis and conclusions In Abh. Geol. B.-A. v. 45, pp. 213-232).

Isozaki, Y., 1997. Permo-Triassic boundary superanoxia and stratified superocean: records from lost deep sea. Science 276, pp. 235-238.

Jin, Y. G., Wang, Y., Wang, W., Shang, Q. H., Cao, C. Q., Erwin, D. H., 2000. Pattern of Marine Mass Extinction Near the Permian-Triassic Boundary in South China. Science 289: 432-436.

Kato, Y, Nakao, K., and Isozaki, Y., 2002. Geochemistry of Late Permian to Early Triassic pelagic cherts from southwest Japan: implications for an oceanic redox change. Chemical Geology 182:1, pp. 15-34.

Katz et al., 1999. The Source and Fate of Massive Carbon Input During the Latest Paleocene Thermal Maximum. Science 286, pp. 1531 - 1533.

Kobayashi, F., 1999. Tethyan uppermost Permian foraminiferal faunas and their paleogeographic and tectonic implications. Palaeogeography Palaeoclimatology Palaeoecology 150, p. 279-307.

Krull, S.J., and Retallack, J.R. 2000. 13C depth profiles from paleosols across the Permian-Triassic boundary: Evidence for methane release. GSA Bulletin, September.

Krull et al., 2000. d¹³C_org chemostratigraphy of the Permian-Triassic boundary in the Maitai Group, New Zealand: evidence for high-latitudinal methane release. New Zealand Journal of Geology & Geophysics, 2000, Vol. 43: 21-32.

Lai, X, Wignall, P., and Zhang, K., 2001. Paleoecology of the conodonts Hindeodus and Clarkina during the Permian-Triassic transitional period. Palaeogeography Palaeoclimatology Palaeoecology 171, pp. 63-72.

Long, L.E., Erwin, M.E., Fisher, R.S., 1997. Rb-Sr ages of diagenesis of Mg-rich clay in Permian sediments, Palo Duro Basin, Texas USA Panhandle. Journal Sedimentary Research 67, pp. 225-234.

Looy, C. V., Brugman, W. A., Dilcher, D. L., Visscher, H., 1999. The delayed resurgence of equatorial forests after the Permian-Triassic ecologic crisis. Proceedings National Academy of Sciences. 96: 13857-13862.

Looy, C.V., Richard J. Twitchett, David L. Dilcher, Johanna H. A. Van Konijnenburg-Van Cittert, and Henk Visscher, 2001. Life in the end-Permian dead zone. PNAS published June 26, 2001.

Lucas, S.G., 1998. Global Triassic tetrapod biostratigraphy and biochronology. Palaeogeography Palaeoclimatology Palaeoecology, 143: 345-382.

Lucas, S.G., 1999. A Tetrapod-based Triassic Timescale. Albertiana 23.

Magaritz, M., 1989. ¹³C minima follow extinction events: a clue to faunal radiation. Geology, v. 17, p. 337-340.

Montañez, I. P., Osleger, D. A., Banner, J. L., Mack, L. E., and Musgrove, M., 2000, Evolution of the Sr and C isotope composition of Cambrian oceans. GSA Today 10, no. 5 1-5.

Musashi, M., Isozaki, Y., Koike, T. and Kreulen, R., 2001. Stable carbon isotope signature in mid-Panthalassa shallow-water carbonates across the Permo-Triassic boundary: evidence for 13C-depleted ocean. Earth Planet. Sci. Lett., 193, 9-20.

Olsen, P. E. and D.V. Kent, 1999. Long-period Milankovitch cycles from the Late Triassic and Early Jurassic of eastern North America and their implications for the calibration of the early Mesozoic time scale and the long-term behavior of the planets. Trans. Roy. Soc. Lond A, 357, 1761-1786.

Pfefferkorn, H.W., 1999. Recuperation from mass extinctions. Proceedings National Academy of Sciences 96, pp. 13597-13599.

Racki, G., 1999. Silica-secreting biota and mass extinctions: survival processes and patterns. Palaeogeography Palaeoclimatology Palaeoecology, 154, pp. 107-132.

Rampino, M. R., A. Prokoph, and A. Adler, 2000. Tempo of the end-Permian event: high-resolution cyclostratigraphy at the Permian-Triassic boundary. Geology, 28: 643-646.

Renne, P. R. et al., 1991. Rapid eruption of the Siberian Traps flood basalts at the Permo-Triassic boundary. Science, 253: 176-179.

Renne, P.R. et al., 1995. Synchrony and Causal Relations Between Permian-Triassic Boundary Crises and Siberian Flood Volcanism, Science 269: 1413-1413.

Retallack, G.J., Krull, E.S., Robinson, S.E., 1996. Permian and Triassic paleosols and paleoenvironments of the central Transantarctic Mountains, Antarctica, Antarctic Journal of the United States Review 1996. (Marine and Terrestrial Geology and Geophysics).

Retallack, G.J., 1996. Paleoenvironmental change across the Permian-Triassic bopundary on land in southeastern Australia and Antarctica. Abstracts of the 30th International Geological Congress, Beijing, China, 2, p. 109.

Retallack, G.J., Veevers, J.J., Morante, R., 1996. Global coal gap between Permian-Triassic extinctions and middle Triassic recovery of peat forming plants. GSA Bulletin 108, 195-207.

Retallack, G.J. , 1999. Postapocalyptic greenhouse paleoclimate revealed by earliest Triassic paleosols in the Sydney Basin, Australia. GSA Bulletin 111, pp. 52-70.

Roberts, J, Claoue-Long, JC, Foster, CB., 1996. SHRIMP zircon dating of the Permian System of eastern Australia, Australian Journal of Earth Sciences 43: (4), p. 401-421.

Rodland, D.L., and Bottjer, D.J., 2001. Biotic Recovery from the End-Permian Mass Extinction: Behavior of the Inarticulate Brachiopod Lingula as a Disaster Taxon. PALAIOS 16:1, pp. 95–101

Sageman, B.B., Rich, J., Arthur, M.A., and Dean, W.E., and Birchfield, E.G., 1997. Evidence for Milankovitch periodicities in Cenomanian-Turonian lithologic and geochemical cycles, Western Interior U.S, Journal of Sedimentary Research, v. B67 (1), p. 286-302.

Savrda, C.E., Bottjer, D.J., and Gorsline, D.S., 1984. Development of a comprehensive oxygen-deficient marine biofacies model: evidence from Santa Monica, San Pedro, and Santa Barbara Basins, California continental borderland. AAPG Bulletin 68, pp. 1179-1192.

Sharp, W.D., Ludwig, K.R., Renne, P.R., Steiner, M.B., Fanning, C.M., 40Ar/30Ar and U/Pb SHRIMP dating of latest Permian tephras in the Midland Basin, Texas. Proc. South Am. Symp. Isot. Geol. (1), pp. 293-294.

Smith, R.M., and Ward, P.D., 2001. Pattern of vertebrate extinctions across an event bed at the Permian-Triassic boundary in the Karoo Basin of South Africa, Geology 29, pp. 1147-1150.

Steiner, M.B., 2001. Simultaneous Permo-Triassic boundary marine and terrestrial mass extinctions: the global fungal spike discovered in the Karoo Supergroup (South Africa). GSA Annual Meeting, November 5-8, Session No. 84.

Twitchett RJ, Looy CV, Morante R, Visscher H, Wignall PB, 2001. Rapid and synchronous collapse of marine and terrestrial ecosystems during the end-Permian biotic crisis. Geology 29 (4): 351-354.

Twitchett, R.J., and Wignall, P.B., 1996. Trace fossils and the aftermath of the Permo-Triassic mass extinction: evidence from northern Italy. Palaeogeography, Palaeoclimatology, Palaeoecology 124: 137-152.

Twitchett, R.J., 1999. Paleoenvironments and faunal recovery after the end-Permian mass extinction. Palaeogeography Palaeoclimatology Palaeoecology 154, pp. 27-37.

Visscher, H., H. Brinkhuis, D.L. Dilcher, W.C. Elsik, Y. Eshet, C.V. Looy, M.R. Rampino and A. Traverse (1996). The terminal Paleozoic fungal event: evidence of terrestrial ecosystem destabilization and collapse. Proceedings National Academy of Sciences USA, 93, 2155-2158.

Ward, P.D., Montgomery, D.R., and Smith, R., 2000. Altered river morphology in South Africa related to the Permian-Triassic extinction. Science 289, pp. 1740-1743.

Wignall, P.B., and Hallam, A., 1992. Anoxia as a cause of the Permian-Triassic extinction: facies evidence from northern Italy and the western United States. Palaeogeography Palaeoclimatology Palaeoecology 93, pp. 21-46.

Wilde et al., 1996. The whole-rock cerium anomaly: a potential indicator of eustatic sea-level changes in shales of the anoxic facies, Sedimentary Geology 101, p. 43-53.

Wright, J.H., Schrader, H., and Holser, W.T., 1987. Paleoredox variations in ancient oceans recorded by rare earth elements in fossil apatite. Geochimica et Cosmochimica Acta 5, pp. 632-644.

Yang, F., 1993. Biotic mass extinction and biotic alteration at the Permian-Triassic boundary. Ammonoids. In Permo-Triassic events of South China (ed. Z. Yang, S. Wu, H. Yin, G. Xu, K. Zhang and X. Bi), pp. 102-108. Geological Publishing House, Beijing.

Yang, Z.Y., et al, 1993. Permo-Triassic Events of South China. Geological Publishing House, Beijing.

Yin, H.F., 1985. Bivalves near the Permian-Triassic boundary in South China. Journal of Paleontology 59, p. 572-600.

Yin, H.F., and Tong, J.N., 1998. Multidisciplinary High-Resolution Correlation of the Permian-Triassic Boundary. Palaeogeography Palaeoclimatology Palaeoecology 143, pp 199-212.

Zachos, J.C., N.J. Shackleton, J. Revenaugh, H. Palike, and B.P. Flower, 2001. Climate response to orbital forcing across the Oligocene-Miocene boundary, Science, 292, p. 274-278.

Zhou, L., and Kyte, F.T., 1988. The Permian-Triassic Boundary Event: a geochemical study of three Chinese sections. Earth and Planetary Science Letters 90:411-421.

Ziegler, A.M., J.M. Parrish, J. Yao, E.D. Gyllenhaal, D.B. Rowley,J.T. Parrish, S. Nie, A. Bekker, and M.L. Hulver, 1993 . Early Mesozoic phytogeography and climate. In: Palaeoclimates and Their Modelling with Special Reference to the Mesozoic Era. Philosophical Transactions of the Royal Society of London, Ser. B, 341.