Background Task 1: Investigating Causes of Widespread Wildfire and Associated Reptile, Early Dinosaur, and Amphibian Deaths in the Upper Triassic

In the Colorado Plateau region of North America, the Late Carnian Shinarump and Cameron Members of the Chinle Formation formed during a humid but seasonal climate (Dubiel and Hasiotis, 1994).  Strata of the overlying Late Carnian to Middle Norian Petrified Forest Members indicate a strongly seasonal but possibly subhumid mega monsoonal climate (Therrien and Fastovsky, 2000).  The unconformably overlying Wingate Formation of Rhaetian age consists predominantly of playa mudstones and eolian sheet sandstones formed during more arid conditions (Lucas et al., 1987).  The faunal age of the Snyder Quarry horizon, within the Petrified Forest Formation of New Mexocp, is between 210-215 Ma (Zeigler, 2002).

Preliminary analyses of the main fossil bed indicate the 30-cm-thick gray mudstone sharply contacts an underlying 8-cm-thick calcite pelloidal conglomerate.  Petrographic examination of the well-indurated pellet conglomerate by the PI reveals round micritic fecal pellets, larger irregularly-shaped calcite nodules, clay rip-ups, and 0.2-0.5 mm long laths of apatite (microprobe analysis at ASU by the PI and Dante Lauretta indicates that laths of small, brown isotropic grains are likely fossil teeth and scales) cemented largely by calcite.  The size of the teeth and scales and their greater number in the pellet conglomerate suggest fish remains.  The unit lacks quartz lithics.  The round fecal pellets and lack of quartz grains suggest the conglomerate was formed in a shallow sediment-deprived lake, and clay rip-ups indicate reworking by currents (Tanner et al., 2003).  Carbon isotopic values on the pellet conglomerate range from -9.3 to -8.9, which are fairly light (depleted) values typical for pedogenic or groundwater cementation (Tanner et al., 2003).  The calcite nodules, light carbon values, and sharp upper contact indicate that the lake probably dried up and the remaining deposits may have been subject to some degree of pedogenesis, before later deposition of the overlying fossil bone and wood bed (Tanner et al., 2003).  Acidic fluids moving through a soil horizon often destroy material of calcite bones and shells, but tend to preserve apatitic teeth.

The overlying fossiliferous mudstone contains mudclasts (rip ups), calcite pellets, small fossil teeth, pyrite spherules (< 0.02 mm), partly oxidized pyrite grains (1.0-5.0 mm), blackened woody plant fragments, and black bone fragments and the matrix is micrite and cements are calcite, gypsum, and chalcedony (Chapman et al., 2002).  The quartz grains and mudstone rip-ups indicate fluvial input, scouring, and current action (Zeigler, 2002), as does the erosional contact with the underlying pellet conglomerate.  The fossil bones show no signs of abrasion indicating that they were not transported a long distance (Zeigler, 2002).  Location of bone in relation to blackened logs and bone alignment to NE suggests deposition by flowing water (Zeigler, 2002).  Skeletal fragments of single individuals are associated, but not fully articulated indicating that they animals were not fully decayed before being incorporated into the fluvial deposit (Zeigler, 2002).  Lack of weathering or scavaging of bones indicates that skeletons were rapidly buried and perhaps a lack of vertebrate scavengers in the area (Zeigler, 2002).  A wide range of skull sizes in the quarry indicate varied ages of animals and a catastrophic mortality event (Zeigler, 2002).

We know that the blackened wood was burnt and charred in a fire and not coalified because charred wood is resistant to biodegradation, which increases its preservation potential (Cope et al., 1985) and the Snyder samples are excellently preserved, resembling charcoal.  In addition, chemical analyses of fossil wood indicate 18.5% organic carbon, consistent with burning (Chapman et al., 2002).  Analyses of fossil bone indicate 0.36% organic carbon, 1.05% hydrogen, 0.05% nitrogen, and 1.0% manganese and are inconclusive, due to high manganese content (which can also blacken bone; Chapman et al., 2002).  Experimental studies of burnt modern bones demonstrate distinct changes in bone micromorphology with elevated stages of temperature range heating (Shipman et al., 1984).  To pursue whether the bone was blacken chemically or in the fire, SEM analyses were done.  SEM images of Snyder bone material appear to be consistent with Shipman et al.’s 1984 description of stage II (185-285°C) and stage III (285-440°C) burning morphology.  Humans burnt to death in fires show star-shaped radial fractures away from upraised pop-outs on their bones (Arizona State Coroner, Pers. Comm.).  This is due to additional heat provided by fat and flesh combustion.  The Snyder bones display identical star-shaped uplifted radial fractures.

Did an impact cause the fire?  The matrix and rip-up clasts contain occasional silt-sized angular quartz fragments that show highly undular extinction patterns and have numerous inclusions.  The PI identified several quartz grains that contain promising potential shock-induced Planar Deformation Features (PDFs).  However, all potential PDFs have to be confirmed by another party with another method to be viewed as legitimate.  The usual confirmation method requires a universal stage set-up on a microscope and a skill, which few people worldwide possess.  Ann Therriault of Canada is acknowledged as an expert and she agreed to the analysis.  Unfortunately, she found that the silt-sized Snyder grains were too small for PDFs to be confirmed with her universal stage, but agreed that they showed promise and suggested that Thermal Electron Microscope (TEM) analysis would be the method to use for confirmation (Ann Therriault, Pers. Comm.).  To suggest catastrophic deaths via impact related fires, the possible PDFs had to be confirmed.

In addition, west of Snyder Quarry, there are other possible burnt wood locales at approximately the same stratigraphic horizon in the Chinle Formation near Holbrook (Jones et al., 2002) and Cameron, Arizona.  [The PI has visited the Cameron site.  Jones et al. (2002) confirm charcoal and have sediment size descriptions and locations for the Holbrook sites.  These additional sites have sand-sized quartz material and did not require TEM for PDF confirmation.]  Although no bone material is present at these sites, this project suggested collecting samples in order to (1) search for diagnostic evidence of a fire at Cameron and (2) to look for petrographic evidence of bolide impact at all of the sites.  If these sites and Snyder Quarry all carry shocked quartz, a better case could be made for a widespread fire.

Task 1 Project Completion Details (all funding years)

(1)  To pursue PDF confirmation with TEM, the silt-sized grains of quartz had to be separated from the sample.  To this end, PI disaggregated a sample of the fossil bone bed using a heated sodium acetate buffer solution, then dried and sieved the sample to obtain the coarse-silt to fine-sand fraction.  A mineral density separation was then performed on this size fraction using a heavy liquid mixture of tetrabromethane diluted with ethanol.  All grains of the density of quartz and heavier were separated out with this method.  From the dried coarse-silt to fine-sand fraction of the quartz mineral density separation, pyrite, opaque grains, muscovite, biotite, and mutigranular quartz pellets were manually removed by the PI using a dissecting scope and eyelash probe (a painstakingly slow and time consuming process).

(2)  The separated material was sent on 9/10/08 to Dr. John Spray at the University of New Brunswick, Canada for confirmation of PDFs using TEM methods.  Dr. Spray is one of the few people worldwide that can use TEM to confirm these features in silt-sized quartz grains (the size present in the Synder Quarry sample).  Dr. Spray found one grain that showed apparent PDF structures with his TEM Method.  One grain was not enough to indicate shock—at least 3 confirmed grains are required before one goes public with a positive statement. 

(3)  The PI prepared another sample for Dr. Spray using the light TAR reserve sample.  This is the material that is determined to be lighter than quartz using heavy-liquid density separation techniques.  The TAR samples from the KT boundary were the ones that produced the confirmed coesite at that boundary (Dr. John McHone, Pers. Comm.).  Dr. Spray produced a TS mount of this material and no light quartz was present (communication 4/16/2009).

(4)  To expedite this investigation of a possible extraterrestrial cause for the faunal die-out and wide fire deposit, the PI sent (5/2009) samples of the Snyder Quarry burnt wood/bone horizon and underlying paleosol layer to Christian Koeberl (Dept. of Lithospheric Research Center for Earth Sciences, University of Vienna) who agreed (during LPSC meeting 3/2009) to do multiparameter gamma-gamma coincidence spectrometry on the materials to detect iridium.  No iridium was detected.  Dr. Frank Kyte also tested the material with the same results.  Tests for soot content/fullerines were negative, but such light particles would not be expected to survive in a fluvial deposit that post-dates a fire (Wendy Wolbach and LuAnn Becker, Pers. Comm.).  Cosmic Helium was reported in the paleosol pellet layer beneath the burned bone bed (LuAnn Becker, Pers. Comm.), however this claim was not confirmed and no iridium was detected in this bed either (Koeberl, Pers. Comm.).

(5)  Probe perfect thin-sections (TS) mounts of the burnt wood stratigraphic horizons in the Chinle Formation near Cameron Arizona were received (12/2008), PI petrographic analysis of these rocks found no unusual materials.  These samples were also later sent to Koeberl for Ir testing.  No iridium anomalies were observed.

(6)  To help better determine the age of the Synder Quarry samples relative to the TJB, the PI collaboratively sampled a rock section near Ghost Ranch, New Mexico (10/09), that shows an abrupt paleomagnetic reversal determined by Dr. Kate Ziegler (University New Mexico).  An abrupt reversal is characteristic of the TJB elsewhere and the stratigraphy between the Ghost Ranch section and Synder Quarry horizon (5 miles southwest) is well known.  These rocks were also sent to Christian Koeberl in late 2009 (Dept. of Lithospheric Research Center for Earth Sciences, University of Vienna) who agreed to do multiparameter gamma-gamma coincidence spectrometry on the materials to detect iridium.  Test results were received in 2010 and these show no iridium anomalies.  In addition, no metallic spherules or potential PDF grains were observed in TS.  The lack of these features does not negate that the abrupt paleomagnetic reversal marks the TJB, as these confirmation features may not be preserved in aeolian deposits due to difficulty of trapping fine grains of shocked material and cosmogenic dust in an active winnowing environments with rapid sediment input.

(7)  The P.I. travelled to and located the burnt wood stratigraphic horizons in the Chinle Formation near Holbrook and Cameron, Arizona.  Fossil wood from both sites was sent to a lab that returned chemical analyses in 2010 that indicated levels of organic carbon consistent with charred wood.  However, additional stratigraphic analysis indicates that neither of horizons are time correlative with Synder Quarry Samples.  No evidence of shock has been found in any thin sections from either site.

Background Task 2: Investigating the Triassic–Jurassic Boundary zone of the Lower Moenave in Nevada and Utah 

The TJB is contained within the equivalent Moenave-Wingate Formations in the southwest United States (Tanner et al, 2002), as we know the boundary lies below the early Jurassic dinosaur trackway in the lower Wingate Formation (Irby, 1995) and above the last known Phytosaur (diagnostic Triassic crocodile-like reptile) skull found in the Rock Point member of the Wingate Formation or the Moenave equivalent (Lucas et al., 1997).  The TJB was suggested to lie within a basal conglomerate (Unit A) of the undivided Moenave and underlying Kayenta Formations that uncomformably overlies the Petrified Forest Member of the Chinle Formation 50 km northeast of Las Vegas, in Nevada (Marzolf, 1990).  This calcite-cemented, 5-m-thick, conglomerate is unusual in that it consists of layers of pebble to cobble-sized, highly polished and faceted quartzite and chalcedony ventifacts; pedogenic limestone nodules; and limestone clasts mantled with up to 1-cm-thick iron oxide crusts within a matrix of fine sand.  A ventifact conglomerate suggests a dry climate and distal alluvial fan deposits accumulating in a playa setting with low sedimentation rates and active wind erosion (Marzolf, 1990).  Pedogenic nodules either formed in soil horizons during seasonal humidity or they and the limestone clasts result from erosion and incorporation of material from the underlying, older Owl Rock member of the Chinle (Chapman and Lauretta, 2004).

Thin section analysis of the unit ‘A’ matrix shows numerous spherules of different compositions (Chapman and Lauretta, 2004).  Preliminary microprobe work on one thin-section found 3 basic types of spherules: (1) iron, (2) pure sanidine, and (3) albite with orthoclase cores.  Albite is a common replacement mineral.  The feldspar types are micritic, and because their rounding/composition could be due to albite replacement at depth efforts were concentrated on the iron spherules.  The Unit A matrix has an unusually high concentration of very-fine to fine sand-sized iron spherules: 20 iron spherules were counted in a standard size 27 x 46 mm thin section.  In thin-section these spherules are opaque, but have a metallic sheen in reflected light that shows lath-like patterns. The laths are composed of two end member compositions.  Preliminary microprobe analysis of only one end member indicated a Wt% composition of 63.63 Fe, 0.75 Mn, 0.61 Al, 0.14 Mg, 0.12 Si, 0.1 Cr, 0.06 Ni, 0.01 K, and 0.01 Na (note the low amount of nickel and absence of cobalt).  Are these spherules cosmogenic dust?

Task 2 Project Completion Details (all funding years):

(1)  A 100 g whole rock sample was broken and dissolved by the PI for one hour in a heated mixture of 82 g anhydrous sodium acetate (NaOAc) and 27 g glacial acetic acid (HOAc) per liter distilled water.  The dried sample was sieved to recover the very-fine to fine sand fraction (between 0.06 and 0.25 mm).  This separate was then picked through under a dissecting microscope to separate out 82 of the partly oxidized, dull reddish to shiny brown spherules.  Most of the iron spherules are magnetic and 175-200 µm in diameter; the largest grains reach about 300 µm.  These sizes are compatible with that expected for cosmic dust.

(2)  SEM imaging shows spherules that have rough or vesicular surfaces, although one was observed to have remnants of a lath intergrowth texture similar to martite lamellae on the surface.  Some spherules have cavities on the surface.  They are mostly spherical, some are teardrop-shaped, a few are rod-shaped, and some have smaller spheres attached.  These types of characteristics are common to cosmogenic dust (Brownlee, 1987).  SEM compositional analysis by individual energy-dispersive spectrometer (EDS) and wave length dispersive spectrometer (WDS) analyses on a typical hematized spherule surface indicate a Wt% composition of 35.35 O, 33.68 Fe, 11.74 C, 7.99 Si, 4.89 Al, 2.39 Mn, 1.73 Mg,1.64 Ti, and 0.58 Ca. 

(3)  If the iron spherules in Nevada are cosmogenic the samples should show a modest anomaly of about 0.1 to 0.3 ppb (Alvarez et al., 1980).  Christian Koeberl (Dept. of Lithospheric Research Center for Earth Sciences, University of Vienna) agreed to do multiparameter gamma-gamma coincidence spectrometry on the Moenave samples to detect iridium.  Dr. Koeberl is well respected and his lab was the one that found the 0.285 ppb iridium anomaly at the TJB boundary in the Newark basin.  A complete section of Moenave from the Nevada site was sampled.  Whole rock samples from the base of each bed within the section were crushed and recently sent for testing. Dr Koeberl found a measureable Ir content of 0.41 ppb in one of the samples mid-section.  This amount matches that found at the TJB elsewhere and it confirms the location of the TJB in this section of Nevada rocks.  This is the first known confirmation of the TJB in western North America.  In addition, this value is within the expected level to support the cosmogenic origin of the materials.

(4)  The PI travelled to Utah and located and sampled material within drying cracks formed of the Wingate Formation at the TJB on the east side of the San Rafael Swell (approximately 17 miles NW of Moab).  These rock samples were sent for probe perfect thin-sections (TS) mounts and were received 2/09.  No unusual materials or spherules were detected within these samples.

Conclusions Tasks 1 and 2:

Aridity and low sedimentation rates favour the slow accumulation of cosmogenic (meteorite) dust in certain favorable environments (Alvarez et al., 1997).  Walter Alvarez originally suspected that the extinctions at the K-T boundary were due to widespread global aridity, an hence a high accumulation of cosmic dust, which he expected would show up as an anomalous concentration of 0.3 ppb iridium in the K-T boundary clay near Gubbio, Italy (Alvarez et al., 1980).  This was a guess based on a combination of background Earth value of 0.1 ppb and the typical chondritic meteorite value of .465 ppb.  Instead the Alvarez team reported a huge value of 3.0 ppb and based on this data, they suggest a large impact at the K-T boundary.  Chicxulub was discovered 10 years later.

An iridium anomaly of 0.285 ppb (.29 ng/g, 285 pg/g) occurs at the Tr-J boundary in Newark Basin of eastern North America (Olsen et al., 2000).  Multiple iridium anomalies, with a peak Ir of 0.450 ppb occur with 42 cm of T-J in Fundy basin, eastern Canada (Tanner et al, 2008).  Olsen et al. (2000) suggested an impact origin for the iridium and suggested Manicouagan.  Although there are 4 known impacts that occurred in the Upper Triassic, all four are smaller than that predicted to cause mass extinction (<150 km diameter), and although Manicouagan at 100 km diameter is the largest crater, the radiometrie age predates the boundary by about 15 million years.  Alternatively, Tanner et al. (2008) note that though they cannot exclude an impact origin, close association with CAMP basaltic rocks suggest volcanic rocks sourced iridium through post-eruptive fluid mobilization.  However, iridium is not considered a mobile element within sedimentary horizons and hence has been used as one of the best indicators of cosmic materials.

Iron-rich magnetic spherules have been reported to occur at the TJB in Hungary (Detre et al., 2000) and in Nevada (Chapman and Lauretta, 2004).  The spherules at the TJB in Nevada are consistent in size and shape with cosmogenic dust.  In addition, like the anomalies found in the east coast of north America, the results of this study show that a spherule-rich bed at the TJB in Nevada does have a similar lower iridium anomaly of 0.41 ppb.  All of these TJB anomaly values are well below that expected for a large bolide, but well within the lower value range expected for high concentrations of cosmic dust mixed with terrestrial sediments—dust that results from global aridity and resulting low sedimentation rates.  In addition, within Upper Triassic beds in the continental deposits of western America and in Europe there are numerous stratigraphic horizons containing blackened wood.  Three of these types of beds were studied in this project.  They appear to not be time correlative, but all have been determined to originate from wild fires.  One site at Synder Quarry, NM is a catastrophic kill bed with strong indications that animals perished in the fire.  This bed has a faunal age that may overlap the age of the Manicouagan impact.  The Synder Quarry bed and the other 2 burnt wood sites in Holbrook and Cameron, AZ show no evidence of shock whatsoever.  These fires appear to be merely evidence of aridity across the Upper Triassic mega-continent of Pangea.  In summation, the results of this study are more consistent with climate change occurring in the Upper Triassic that resulted in extreme aridity and such low sedimentation rates that cosmogenic dust accumulated globally to produce iridium anomalies in the range of 0.2-0.4 ppb.  One can envision animals severely stressed by this climate change to the point of mass extinction, and although the known Upper Triassic impacts did not appear to cause their demise, these events surely added to the stress.

(2) FY 09-11 International Space Science Institute (ISSI) Team Member for Mars Project “ILDs: What do they tell us about Mars Evolution” (‘09-‘11; Bern, Switzerland).

Attended 2010 telecon meeting in October to present preliminary results on a GIS based modeling study of possible origins of Ophir Chasma, one of the connected troughs of Valles Marineris.  My co-authors for this work are Dr. Lionel Wilson (Emitus at Lancaster Univ., UK) and Thomas Kneissl (HRSC Team, Univ. Freie, Berlin).  Research is ongoing.

(3) FY 06-Present Invited Science Team Member for MEX HRSC instrument

Attended a May HRSC Team meeting in Winthrop, WA.  Worked on sight at University Freie in Berlin from June 12 to July 12.  Submitted a September 20 NASA MDAP proposal on a research topic relevant to the Science Team titled “Dark material on plateaus surround Valles Marineris, Mars.”  The central hypothesis of this proposal is that eroded outcrops of dark material near some Valles Marineris chasmata may be the remains of a widespread, somewhat friable, internally layered rock unit that has been eroded to form dark aeolian dunes.  The objective of this proposal is to test this hypothesis by means of a detailed mapping study of the local dark materials outcrops on the Valles Marineris plateau using publicly available data from the MEX, MRO, MO, and MGS missions.  This topic will be discussed during a presention at the 41^st Lunar and Planetary Science Meeting in March, 2011.  Finally, three HRSC publications were produced or co-produced by Chapman in 2010 (titles shown below).