the record of miocene impacts in the argentine pampas€¦ · the record of miocene impacts in the...

Meteoritics amp Planetary Science 41 Nr 5 749ndash771 (2006)Abstract available online at httpmeteoriticsorg

749 copy The Meteoritical Society 2006 Printed in USA

The record of Miocene impacts in the Argentine Pampas

Peter H SCHULTZ1 Marcelo ZiexclRATE2 Willis E HAMES3 R Scott HARRIS1T E BUNCH4 Christian KOEBERL5 Paul RENNE6 and James WITTKE7

1Department of Geological Sciences Brown University Providence Rhode Island 02912ndash1846 USA2Facultad de Ciencias Exactas y Naturales Universidad Nacional de La Pampa

Avda Uruguay 151 6300 Santa Rosa La Pampa Argentina3Department of Geology Auburn University Auburn Alabama 36849 USA

4Department of Geology Northern Arizona University Flagstaff Arizona 86011 USA5Department of Geological Sciences University of Vienna Althanstrasse 14 A-1090 Vienna Austria

6Berkeley Geochronology Center 2455 Ridge Road Berkeley California 94709 USA7Department of Geology Northern Arizona University Flagstaff Arizona 86011 USA

Corresponding author E-mail peter_schultzbrownedu

(Received 02 March 2005 revision accepted 14 December 2005)

AbstractndashArgentine Pampean sediments represent a nearly continuous record of deposition since thelate Miocene (sim10 Ma) Previous studies described five localized concentrations of vesicular impactglasses from the Holocene to late Pliocene Two more occurrences from the late Miocene are reportedhere one near Chasicoacute (CH) with an 40Ar39Ar age of 924 plusmn 009 Ma and the other near Bahiacutea Blanca(BB) with an age of 528 plusmn 004 Ma In contrast with andesitic and dacitic impact glasses from otherlocalities in the Pampas the CH and BB glasses are more mafic They also exhibit higher degrees ofmelting with relatively few xenoycrysts but extensive quench crystals In addition to evidence forextreme heating (gt1700 degC) shock features are observed (eg planar deformation features [PDFs]and diaplectic quartz and feldspar) in impact glasses from both deposits Geochemical analyses revealunusually high levels of Ba (sim7700 ppm) in some samples which is consistent with an interpretationthat these impacts excavated marine sequences known to be at depth These two new impact glassoccurrences raise to seven the number of late Cenozoic impacts for which there is evidence preservedin the Pampean sediments This seemingly high number of significant impacts over a 106 km2 area ina time span of 10 Myr is consistent with the number of bolides larger than 100 m expected to enter theatmosphere but is contrary to calculated survival rates following atmospheric disruption ThePampean record suggests therefore that either atmospheric entry models need to be reconsidered orthat the Earth has received an enhanced flux of impactors during portions of the late CenozoicEvidence for the resulting collisions may be best preserved and revealed in rare dissected regions ofcontinuous low-energy deposition such as the Pampas Additionally the rare earth element (REE)concentrations of the target sediments and impact melts associated with the Chasicoacute event resemblethe HNaK australites of similar age This suggests the possibility that those enigmatic tektites couldhave originated as high-angle distal ejecta from an impact in Argentina thereby accounting for theirrarity and notable chemical and physical differences from other Australasian impact glasses

INTRODUCTION

Vast deposits of upper Tertiary to Holocene loess andloess-like (loessoid) sediments cover the Argentine PampasThese deposits began to develop during the late Miocene inresponse to the Andean orogeny and ensuing climatic changes(Zmiddotrate 2003) The late Quaternary loess record has receivedincreasing attention for interpreting the continentalpaleoclimates in this part of the southern hemisphere (Muhs

and Zmiddotrate 2001 Zinck and Sayago 1999 Orgeira et al 1998Iriondo 1997) The early history however is less extensivelyresearched due to the limited chronostratigraphic markers andthe very limited paleomagnetic studies

Enigmatic glasses within these sediments are locallycalled ldquoescoriasrdquo and often are associated with (and attachedto) reddish brick-like masses called ldquotierra cocidasrdquo Theterm ldquoescoriardquo can imply a wide range of vesicular glasses(eg volcanic or human-produced slag) Because they occur

750 P H Schultz et al

in large quantities at specific horizons within pre-Holocenesediments a different origin is required especially in theabsence of nearby volcanic sources Studies of such glasses inLate Pliocene to Holocene Pampean sediments haveestablished that they were formed as the result ofhypervelocity impact (Schultz et al 1994 1998 2004 Harriset al 2005a 2005b)

The identification and characterization of older Pampeanglasses would not only improve our understanding of possibleimpact products in the sedimentary record but also wouldprovide needed chronostratigraphic markers These oldersuccessions contain important fossil assemblages that wereproposed to track mammalian evolution and interchangesthrough time (Pascual et al 1996 Marshall et al 1983)Therefore radiometric dates of instantaneously formedimpact glasses provide unique benchmarks for re-evaluatingfaunal evolution

Here we report the discovery of two new glass depositsconcentrated along specific horizons in late Miocene strataand provide evidence for their impact origin The oldest glasslayer dates to about 92 Ma and provides a critical benchmarkfor important Miocene outcrops The second concentrationdates to about 53 Ma which is approximately the MiocenePliocene boundary (Berggren et al 1995) We first review thePliocene to Pleistocene impact record in the Pampas ofArgentina and then focus on the specific geologic setting ofthe two new localities Next we discuss the nature (includingevidence for impact origin) and age of the impact glasses

Lastly we consider the possible implications including thesize of the events comparison with other impact glasses inyounger sediments the production rate of glass-producingimpacts and the possible relation to the enigmatic AustralianHNaK tektites with an age similar to that of the ChasicUcircglasses

BACKGROUND

Argentine Impact Glasses

At least four impact glass deposits have been discoveredin post-Miocene Pampean sediments (Schultz et al 19941998 2002a 2002b 2004 Bland et al 2002) These impactglasses often are attached to or wrap around tierra cocidas thatare composed of oxidized and welded sediments that arebelieved to be lower-temperature impact products theequivalent of clastic breccias derived from unconsolidatedsediments (Schultz et al 1998 2004) The glasses containclear evidence for intense transient heating to temperaturesexceeding 1700 degC followed by rapid cooling and dynamicemplacement (Schultz et al 1994 1998 Harris et al 2005a)Shock indicators have been documented in the Pliocene andoldest Pleistocene glasses (Harris et al 2005a 2005b)

The two new glass occurrences described in this paperare exposed in sediments of the Southern Buenos Aires (SBA)Province (Fig 1) The older glasses occur in well-knownupper Miocene loessoid strata composed primarily of siltysands containing carbonate nodules and at certain levelscarbonate cements These generally unconsolidated depositscontain late Miocene fossil mammal assemblagescharacterized as the ldquoChasicUcircan assemblagesrdquo by Fidalgoet al (1978) and Bondesio et al (1980) While other materialsof ChasicUcircan age crop out far to the northwest (the central partof La Pampa Province San Luis San Juan and Catamarca)the exposures closer to the Atlantic coast are limited to a veryrestricted area surrounding Laguna de ChasicUcirc and SalinasChicas The ChasicUcirc (CH) glasses in this region are typically1ndash10 cm in diameter but have been found in tightly clusteredassemblages as large as 2 m across Although they do notform a continuous bed of melt they occur scattered withinspecific stratigraphic horizons that can be traced for morethan 5 km (discussed below)

The second glass layer is widespread in upper Miocene tolower Pliocene outcrops throughout the Southern BuenosAires (SBA) Province from BahIgravea Blanca (BB) east into theArroyo ClaromecUcirc Basin (east of the Sierra La Ventanamountain range comprised of Paleozoic basement rocks) andwest to near GuaminIgrave Multiple localities near BahIgravea Blancayield the largest concentrations of glasses (up to 10 cmacross) Our initial suspicion was that these glassesrepresented distal equivalents of the ChasicUcirc glasses initiallyemplaced on the Sierra La Ventana piedmont but transportedand reworked into much younger sediments through fluvial

Fig 1 A map showing the primary locations where Chasicoacute (CH)and Bahiacutea Blanca (BB) glasses have been investigated Relatedglasses occur in MioPliocene outcrops along a railroad cut (RR) nearBB and at other sites scattered throughout southern Buenos AiresProvince Pliocene and Pleistocene impact glasses occur in thevicinities of Mar del Plata (MDP) and Centinela del Mar (CdM)respectively

The record of Miocene impacts in the Argentine Pampas 751

reworking (Schultz et al 2000) As described belowhowever radiometric dates now clearly establish that the BBglasses represent a completely separate event generallyconsistent with their present stratigraphic position

In the following discussion the ChasicUcirc materials will bedescribed in greater detail because the stratigraphic setting ispresently better constrained by field studies paleomagneticprofiles and more continuous exposures

GEOLOGIC SETTING

The ChasicUcirc exposures represent the oldest outcroppingsof Miocene sediments in the southern Buenos Aires province(see review by Zmiddotrate [2003]) The sediments were depositedin response to enhanced transport processes created by upliftof the Cordilleras during the late Miocene Zmiddotrate et al(Forthcoming) provide greater details on the lithofacies anddetailed stratigraphy of the Arroyo ChasicUcirc Formation

Along the ChasicUcirc arroyo silty sands accumulatedduring the Miocene within a major region of subsidence of theColorado Basin Sediments were derived from the nearbypiedmont of the Sierra La Ventana range combined withfluvial and aeolian transport from the Andes far to the westBased on geophysical surveys the sediments extend to a totaldepth of 5 km with deposits similar to the exposedChasicUcircan-age (defined by the contained faunal assemblages)materials extending down 180 m (Zambrano 1980 Bonorinoet al 1987) The ChasicUcirc Arroyo cuts through and exposes theupper portion of these upper Miocene deposits as it dischargesinto Laguna ChasicUcirc Tertiary marine sequences occur belowthe ChasicUcircan sediments nearby

North and east of Laguna de ChasicUcirc upper Tertiarydeposits form an extensive and gentle plain descending fromnearly 400 m in the Sierra de la Ventana to elevations of about50 m in BahIgravea Blanca on the Atlantic coast and around 70ndash80 m at the northern margin of the ChasicUcirc depression A 15ndash2 m thick calcrete crust caps the deposits which are overlainby the upper Pleistocene-Holocene sandy aeolian mantle Thepresent drainage system carves into these upper Tertiarysequences which crop out discontinuously along riverbanksroad cuts and in quarries North and west of Sierra de laVentana the upper Tertiary deposits form a vast plainextending into La Pampa Province where they are groupedinto the Cerro Azul Formation

Folded and twisted glassy masses (along with tierracocidas) outcrop along a 5 km section near the lower reachesof the Arroyo ChasicUcirc Zmiddotrate et al (Forthcoming) provide amore detailed discussion of the revised stratigraphy of thesesequences There are three occurrences 1) a concentration oflarge masses just above a paleosol within sandy fluvial facies2) larger angular glass fragments within a lower diamicton(CH-LM) and 3) dispersed smaller and broken fragments in aseries of upper diamicton layers (CH-U) This pattern is very

similar to occurrences near Chapadmalal (Schultz et al 19982002a 2002b) and is interpreted as a lower primary glass-bearing layer that was subsequently covered by rapidlyreworked sediments from nearby exposures of the lowerprimary unit Although it is possible that the diamictondeposit could represent primary ejecta facies such aninterpretation cannot yet be tested

Glasses were also recovered from upper Tertiary depositsthat crop out in the surroundings of BahIgravea Blanca as well asfurther east in various isolated upper Miocene outcropsincluding railroad cuts (labeled ldquoRRrdquo) and exposures alongthe Arroyo ClaromecUcirc valley at distances of about 250 kmfrom BahIgravea Blanca Near BahIgravea Blanca the glasses occur influvialaeolian sequences that are part of a vast structuralplain descending from the Ventana Range to the AtlanticOcean One particular section is about 20 m thick Thelowermost 5 m are composed of massive light brownsiltstones which contain partially silicified calcareousnodules up to 70 cm long The upper 15 m of the exposure ismade up of light brown calcareous siltstones either massiveor displaying coarse stratification that suggests aqueousdeposition (Zmiddotrate et al Forthcoming) Siltstone intraclastsare scattered throughout the sedimentary matrix at somelevels with very common carbonate nodules Glass fragmentsare present throughout this upper section with the highestconcentrations and largest pieces at the lowermost part closeto the boundary with the underlying siltstones The entireexposure is capped by a thick calcrete crust up to 2 m thickwhich in turn is overlain by upper Pleistocene and Holoceneloess deposits Fossil remains of vertebrates belonging to theHuayquerian South American Land Mammal Age (SALMA)have been recovered from these sediments (Deschamps2003)

Glass fragments were also reported from Carhueacute northof Sierra de la Ventana Fossil remains exhumed from theoutcrops in the surroundings of BahIgravea Blanca suggest a lateMiocene Huayquerian age (Deschamps 2003 Deschampspersonal communication) Eastward at the Las Oscuras siterecovered vertebrate assemblages indicate a younger agebelonging to Montehermosan SALMA A late Tertiary agewas traditionally attributed to the sediments bearing glasses atArroyo ClaromecUcirc as well as the nearby locality of ArroyoLas Mostazas Fossil remains collected at several localitiesalong the Arroyo ClaromecUcirc valley (included in this study)were attributed to the ldquoIrenenserdquo a unit correlated to theMontehermosan (Fidalgo et al 1975)

CHARACTERISTICS AND COMPOSITION

ChasicUcirc Glasses

The vesicular impact glasses from ChasicUcirc resemble thePliocene and Pleistocene samples but there are significant


differences in the petrology and geochemistry consistent withtheir regional setting The CH glasses exhibit large twistedvesicles that occasionally contain seams of partially moltenstringers of loessoid materials and are similar to the RioCuarto Centinela and Chapadmalalan impact melt brecciasBut unlike the other glasses CH samples are much darkerreddish-brown to brown (in hand specimen) glassescontaining denser concentrations of quench crystals typicalof more proximal impact glasses

The glasses range from centimeter-size fragments to05 m diameter masses with some large clusters nearly 2 macross (Fig 2a) Thin sections were made of seven differentsamples representing contrasting sizes (large versus smallparent samples) and context (lower versus upper sequences)

The mineral modes of unmelted xenocrysts (point countingaverages from seven different thin sections) indicateplagioclase and K-feldspars as the most common (63 vol)followed by Ca-pyroxenes (29 vol) and quartz (28 vol)The Ca-pyroxenes comprise a much greater fraction ofunmelted minerals (factor of 2 to 5) than any other impactglasses from Argentina examined to date excepting some ofthe BB samples

The host sediments consist of both basaltic and andesiticrock fragments (40 vol) volcanic shards (15ndash20 vol)oligoclase-andesine plagioclase (25 vol) and K-feldspars(sim10 vol) Heavy minerals (2ndash3 vol) are represented byhornblende augite biotite and opaques (in decreasing order)The petrologic composition of the sediments is generallyconsistent with the xenocrysts observed in the glasses exceptfor the lower fraction of quartz in the host sediments(lt5 vol) relative to the glasses (28 vol) About half of thevolcanic shards are in a weathered state but have not yetconverted to clay The Chemical Index of Alteration (CIA) forthe bulk sediments is typically less than 50 (sim46) which isconsistent with a low degree of chemical alteration (Taylorand McLennan 1985)

Thin-sectioned glasses (CH-U CH-U5) from the upperreworked sediments exhibit little recrystallization with relictminerals comprising less than 5 vol Larger masses (CH-1CH-G) underwent extensive recrystallization and relativelyslow cooling rates but still contain few relict minerals(lt20 vol) From the samples examined glasses with thefewest mineral relicts and fastest cooling rates (based oncrystal lengths) also were the most vesicular

The bulk compositional range from XRF analyses(Table 1) indicates similar precursor materials (dacitesrhyolites andesites and trachytes) for some CH glasses Thiswould be expected for melting of bulk Pampean sedimentsderived from aeolian redistribution of fluvially transportedmaterials from Andean source regions Normativecalculations cannot be used to characterize the sourcematerials fully since impact glasses also contain foreigncomponents (xenocrysts melted clasts and welded masses)Nevertheless this strategy is useful for comparisons withother Pampean impact glasses and to explicitly identifypossible contributions from sediments at depth

The predominant glasses in certain samples (CH-1 CH-U5 and CH-U) have more basaltic compositions (basalticandesite and trachytic basalt) relative to the rest of the glasses(Table 2) The different CH samples in Table 2 refer todifferent locations within the suite CH-1 CH-W and CH-Gcorrespond to glasses from sediments lower in the section(CH-G and CH-1 from large samples) CH-U and CH-U5 areglasses from reworked sediments slightly higher in thesection and CH-Q is a sample from reworked sediments in aquarry about 50 km away from the primary exposures Allsamples were collected within about 2 m of the lowermostoccurrence in the section Examination of Table 2 reveals no

Fig 2 Examples of Chasicoacute impact glasses a) A photograph of anupper Miocene outcrop containing a large cluster of glasses b) Aclose-up photograph of an individual clast of folded glass thatcontains a trapped tierra cocida (tc)


positive correlation between Al and Ca Several Ca-bearingphases could account for the Ca pyroxenes (or apatite as verysmall acicular crystals) from excavated continental basaltsadditions from marine sequences or carbonate nodules orcements The high K2O is attributed in part to the melting ofK-feldspar-rich shards which are abundant in the finefractions

Table 3 focuses on different portions of the CH glassesthat are typified by Figs 3a and 3b These portions of theglass are normatively silica deficient and mafic enriched(diopside and hypersthene) Most CH samples containabundant basaltic melt clasts (especially CH-1 and CH-W)composed of plagioclase and Ca-pyroxene quench crystalsThese melt clasts have well-crystallized cores with grain sizes

Fig 3 a) A backscattered electron (BSE) image of basalt melt clast in ChasicUcirc glass sample (CH-1) Crystallization appears to have occurredwhile the host mass was still plastic eg undergoing aerodynamic flow (chilled margins elongation) b) A BSE image showing host CH-1glass with hyalophitic texture composed of skeletal (hollow and swallow-tail) plagioclase crystals with Mg-rich rims c) A plane-polarizedlight photomicrograph (PPL) showing a basaltic melt spherule trapped within ChasicUcirc impact glass (CH-1) d) A sketch illustrating thestructure of the melt spherule The dark core (regions 1 and 2) is composed of dense accumulations of very fine-grained randomly orientedacicular crystals of Ca-pyroxene and plagioclase Region 1 has a somewhat lighter appearance due to the presence of numerous tiny (and a fewlarger) feldspar xenocrysts A rind (region 3) consisting of relatively clear quenched glass and sub-radially oriented plagioclase crystalsseparates the spherule from the host glass The structure is similar to other xenocryst-bearing impact spherules found in the geologic record(eg Simonson 2003)


Table 1 Major elements of bulk loess and bulk impact glasses Host sed Glasses

CH-LM CH-1 CH-2 CH-A CH-C BB RR

SiO2 623 625 5966 6074 6258 6018 5944TiO2 070 099 094 079 077 081 079Al2O3 149 160 145 1501 1472 1655 1725Fe2O3 555 574 609 569 595 575 560MnO 0005 0004 005 002 002 011 010MgO 382 237 324 287 264 269 262CaO 555 418 697 527 492 558 671Na2O 432 34 371 424 379 288 266K2O 208 35 340 351 280 369 314P2O5 017 038 075 054 084 051 077

Total 9940 9956 9931 9868 9903 9874 9908(LOI) (1152) (431) (328) (479) (386) (220) (517)XRF data in wt by C K University of Vienna percentages exclude contributions from losses on ignition (LOI) given below Abbreviations correspond

to Chasico (CH) Bahiacutea Blanca (BB) and samples from a railroad cut (RR) near BB

Table 2 Representative analyses of interstitial CHa glasses and BBbCH-G CH-W CH-U CH-U5 CH-1 CH-Q BB

SiO2 649 608 623 649 621 658 620 (22)Al2O3 120 171 196 164 144 168 154 (14)TiO2 076 087 092 068 076 074 078 (008)FeO 86 665 50 365 46 44 45 (13)MnO 025 023 015 ndash 008 012 ndashMgO 232 315 220 145 321 227 23 (08)CaO 835 54 405 315 61 42 50 (19)Na2O 125 325 18 380 415 221 29 (05)K2O 212 179 189 30 30 280 64 (09)P2O5 058 067 045 013 029 033 066 (030)Sum 10013 9991 9836 9716 9874 9967 9994

BaO NiO Cr2O3 lt 002IUGS rock typec Andesite Dacite Dacite Trachyte Dacite Dacite

aElectron microprobe analyses at Northern Arizona University (T B and J W)bElectron microprobe analyses at the NSFKeck Electron Microprobe Facility at Brown University (P H S) with Na loss routine (Devine et al 1995)cAnalyses normalized to 100 and calculated into normative components according to the International Union of Geological Sciences (IUGS) rock types

Table 3 Microprobe broad beam analyses of CH series glasses (normalized to 100 wt) CH-1 CH-G CH-U5 CH-U CH-W CH-Q

Quench Dark Clear Dark Clear Flow

SiO2 627 532 605 600 623 601 579 606 700Al2O3 160 160 195 124 183 161 118 175 154TiO2 087 119 078 066 124 098 061 104 037FeO 365 51 289 292 273 45 336 405 186MnO 008 006 ndash ndash ndash 005 ndash 004 012MgO 236 34 17 298 207 260 43 263 091CaO 69 140 57 99 42 55 97 48 172BaO ndash ndash ndash 019 ndash ndash 025 003 012Na2O 475 368 50 55 63 342 38 50 381K2O 210 101 348 42 260 60 59 344 57P2O5 066 236 045 126 026 075 238 057 050

Total 10000 10000 10000 1000 10000 10000 1000 1000 1000Totals(before normalization)

9580 9884 9518 9630 9800 978 9531 9655 934

Electron microprobe data were performed at the Northern Arizona University (T B and J W) Low values before normalization are due to possible waterand holes


decreasing toward the chilled margins (Fig 3a) These clastsare surrounded by a melt with hyalophitic basalt textures(Fig 3b) Figures 3c and 3d provide an example of anotherbasaltic melt clast trapped in CH-1 Although most of thesemelt clasts have irregular to somewhat elongated shapes afew (eg Figs 3c and 3d) exhibit morphologies more typicalof classic impact spherules (Simonson 2003) Basaltictextures include ophitic to subophitic hyalophitic varioliticand intersertal Maximum grain size in all cases is lt75 micromwith most at lt4 microm The microcrystalline basalticcomponents shown in Fig 5 contain augite (Fs14Wo43) withmantles of hedenbergite (Fs49Wo51) plagioclase (An85) andresidual glass (FeO = 42 wt) In addition 048 wt P2O5was found in augite and 051 in the glass perhaps fromapatite in excavated basaltic materials or recently weatheredby-products Hedenbergite contains 013 wt BaO and theglass has 015 BaO Moreover BaO was also found in theflow glass in CH-U (025 wt) and the fractured glass in CH-U5 (019) Such enhancements may represent the effects ofhydrothermal processes

Bulk analyses (Table 4) of the example shown in Fig 3show that the ophiticsubophitic core has an alkali basaltcomposition the granular mantle picrobasalt (olivine-bearing) and the chilled margin andesite The textures andcompositional trends from core to chilled margin demonstraterapid in situ crystallization with compositionaldifferentiation over very short distances However thecrystallization rates are less rapid than those samples withoutsuch quench crystals as well as those samples that havequench crystals but little or no basaltic melt clasts Otheranalyzed melt clasts are alkalic basalts and basaltic andesites

The low totals in Table 3 are largely due to vesicles includedin the broad-beam analysis

The major element composition of the CH glassesgenerally resembles the post-Miocene Argentine impactglasses consistent with generally similar precursor materials(ie loess derived from Andean volcanics) but there are alsosignificant differences Figure 4 provides a plot of the

Fig 4 The ratio of different interstitial averaged compositions (CH-1 CH-2 and CH-G) to bulk Miocene sediments (CH-LM) For thesmaller Chasicoacute glasses there is little difference between theinterstitial glass and loess compositions This reflects nearlycomplete melting during formation CH-G is representative of largerCH masses which exhibit depletions consistent with incompletemelting of the target sediments

Fig 5 Photographs showing examples of the small melt fragment(arrow) within a late Miocene early Pliocene outcrop near BahIgraveaBlanca (a) Similar glass fragments occur throughout the upperMiocene to lower Pliocene sections within 500 km of BahIgravea BlancaExamples of extracted glasses are shown in (b)


interstitial clear and bulk glass composition from microprobeanalyses ratioed to the bulk sediment composition in whichthey occur While most of the major elements of the smaller(lt5 cm) glasses closely resemble the host sediments K2O andP2O5 levels are clearly enriched Analysis of a very large

(05 m in diameter) impact melt breccia (CH-G) howeverreveals significant depletions in TiO2 FeO MgO and CaOprobably reflecting incomplete melting of refractory oxideminerals Aqueous alteration possibly hydrothermal is likelyresponsible for the observed enrichment in K2O

Table 4 Microprobe broad beam analyses of inclusions in impact glass from Chasico (CH-1)Subophiticcore

Melt clast 1 (Fig 3a)granular mantle Chill margin Melt clast 2 Melt clast 3

SiO2 457 447 598 540 502Al2O3 128 133 164 170 187TiO2 10 096 097 10 078FeO 64 377 365 58 380MnO 026 010 005 005 007MgO 74 97 313 297 57CaO 228 243 92 127 175BaO ndash 010 013 ndash ndashNa2O 094 036 455 40 248K2O 018 008 150 111 027P2O5 252 263 057 042 055

Totals 1000 1000 1000 1000 1000Electron microprobe data were performed at the Northern Arizona University (T B and J W)

Fig 6 a) A reflected light photomicrograph of a literal impact melt breccia (ie composed of multiple distinct melt clasts in contrast to lithicclasts set in a melt matrix) from BB2 (polished thin section) b) A sketch of (a) showing the locations and relationships of different melts andloess stringers Melt A is black smooth and enriched in Ca Ti relative to Melt B and also contains fewer small vesicles and fewer clasts MeltB is red and contains abundant small vesicles and clasts Melt C is a sulfur-bearing melt that was originally a soil or similar material associatedwith loess and trapped along same seams as loess stringers Melt D is a K Fe-rich vesicular melt with abundant unmelted quartz and magnetitegrains Melt E is a silica melt probably a lechatelierite particle c) A BSE image of the boxed region in (b) showing the sharp seam betweenMelt A and Melt B d) A PPL photomicrograph of a melt-rich suevitic breccia from BB1 Distinct vesiculated or flow-banded glass clasts areset in a fine-grained clast-rich matrix


BahIgravea Blanca Glasses

The BahIgravea Blanca glasses (eg Fig 5) generally occur assmall (lt5 cm) pods of grayish to black materials some ofwhich do not resemble glasses at first glance except wherevesicles are conspicuous (eg two specimens from BB1 inFig 5) The BB-2 specimens tend to appear black and denseboth on weathered and polished surfaces (Fig 6) Asdiscussed above BB glasses occur in Miocene units coveringa wide region including the RR sample Tables 1 and 2provide data indicating that the BB and RR samples have bulkand interstitial glass compositions resembling the glassesfrom ChasicUcirc

More than twenty thin sections and polished SEMmounts were made from a dozen small glasses collected in thevicinity of BahIgravea Blanca The glasses exhibit tremendouscompositional and textural variability especially for the smallnumber of specimens collected from only a few sites Nearlyevery classification of impact melt is present including clast-poor and clast-rich melt-matrix breccias similar to theChasicUcirc glasses literal impact melt breccias (composed ofmultiple distinct melt-rich clasts) (Figs 6andashc) pieces ofsuevitic breccias (Fig 6d) and even some gradational formsBecause of this wide range a full description of thesematerials will be deferred to a later contribution Here thefocus is to provide a basic description and sufficientpetrography to establish their impact origin

Geochemistry

The CI-normalized distribution REE shown in Fig 7indicates that the CH and BB glasses are reduced in the lightREE relative to the Miocene sediments Certain minorelements in the CH glasses (Table 5) exhibit significantdifferences from the host sediments The most significant

difference is the order of magnitude increase in Ba from anominal level of less than 400 ppm (matrix) to as high as7600 ppm (CH-1 glass) As noted in the petrography andmicroprobe data BaO was found in the flow glass of twosamples (exhibiting evidence of hydration) and withinhedenbergite mantles around augitic cores The selectedelements that are more indicative of provenance also contrastwith the host sediments For example the impact meltsexhibit slightly more mafic affinities (V Sc La and Th) thanthe host sediments as well as reduced concentrations of REEOther ratios (KU ZrHf HfTa and ThU) also indicate thatthe impact glasses incorporated more mafic material thanthose in the host sediments Consequently the CH glassesindicate that the impact may have penetrated a basaltic flowandor was then subjected to hydrothermal alteration

SHOCK PETROGRAPHY

ChasicUcirc Glasses

Impacts into porous sedimentary targets such as the thickPampean sediments are not expected to efficiently transferthe energy of the shock wave to the mineral grains (egGrieve et al 1996) due to the low impedance of the porespaces between them Instead a significant portion of theenergy is consumed by collapsing pore-space andcompressing grain boundaries thereby producing copiousheat in the target materials (Kieffer 1971) Consequentlysedimentary targets typically experience extensive melting atmuch lower pressures than crystalline targets (Kieffer et al1976 Stˆffler 1984) Many grains therefore probably neverexperience sufficiently high pressures to either produceplanar deformation features (PDFs) or initiate other solid-state transformations The small number of grains that dorecord high pressures are likely melted soon thereafter(French 1998) The odds of finding shocked mineral grains in

Fig 7 Distribution of rare earth elements (REE) for average CHglasses (open circles) BB glass (filled circles) and host CHsediments (filled squares) The impact glasses exhibit a smaller Euanomaly consistent with a source more mafic than the hostsediments

Fig 8 A PPL photomicrograph of a polycrystalline quartz grain inCH glass that exhibits ballen texture


the melt products of impacts into the Pampean sediments arethen reduced even further by the very small size of thesurviving crystals Hence the chances of discovering shockedquartz are diminished by the extreme paucity of quartz inmany of the strata

The seven ChasicUcirc thin sections discussed above (alongwith numerous paired sections) were examined using opticalpetrography and scanning electron microscopy (SEM) for anyevidence of high pressurehigh strain rate deformation highpressurehigh temperature phase decomposition or very hightemperature melting that is either indicative or suggestive ofimpact processes While more complete results will bereported in future contributions the most significant findingsare presented here to establish the impact origin of the 92 MaChasicUcirc glasses

Initial observations revealed abundant tapered to droplet-

shaped polycrystalline grains (lt100 microm) interpreted to berecrystallized potassium feldspars and quartz Therecrystallized quartz grains (Fig 9) commonly exhibit theldquoballenrdquo texture first described by Carstens (1975) as areversion product from diaplectic glass When ballen quartzforms after lechatelierite it typically exhibits thepolycrystalline habit inhomogeneous extinction (Bischoffand Stˆffler 1984) and occasional vesiculation (Stˆffler andLangenhorst 1994) that is observed in the ChasicUcirc grains(Bischoff and Stˆffler 1984) Lechatelierite may form as aresult of shock pressures exceeding 50 GPa (Stˆffler andLangenhorst 1994) or temperatures exceeding 1700 degC Innaturally occurring rocks therefore lechatelierite isindicative only of impact metamorphism or lightning strikes(Montanari and Koeberl 2000) Given the morphology sizeapparent volume and geographic extent of the glasses

Table 5 Trace element geochemistry of Miocene impact glassesCH-1 CH-2 CH-A CH-C BB-A RR-1 CH-Sed MdP-Sed

V 120 119 110 107 128 114 93 116Sc 121 152 134 139 148 136 139 129Cr 287 239 341 307 466 883 303 291Co 758 905 917 101 135 116 842 105Ni 5 6 10 10 20 25 9 25Cu 3 lt2 lt2 6 27 36 lt2 31Zn 23 32 22 49 31 25 60 80Ga 30 60 5 20 20As 843 672 348 397 29 27 757 31Se 06 06 08 09 04 01 07 04Br 09 22 32 59 04 85 191 28Rb 665 439 713 701 581 548 692 734Sr 471 585 493 424 430 546 371 287Y 24 29 27 30 24 26 28 21Zr 176 157 162 163 164 178 159 220Nb 10 8 9 9 12 11 9 14Sb 026 029 008 031 037 031 034 062Cs 403 233 441 464 329 310 504 462Ba 7693 2391 1745 613 562 445 374 395La 201 204 188 205 194 187 215 227Ce 397 438 442 502 388 355 482 481Nd 211 259 254 261 204 195 251 264Sm 431 544 417 507 421 416 481 521Eu 108 140 126 141 115 119 135 125Gd 42 52 45 52 375 347 55 437Tb 067 079 076 092 063 065 074 076Tm 038 043 041 047 033 034 048 039Yb 239 294 246 269 225 229 317 276Lu 035 041 035 041 034 035 048 043Hf 455 435 482 452 411 418 469 538Ta 043 033 055 048 044 042 065 057W 08 15 05 06 07 09 15 11Ir (ppb) lt1 01 04 lt1 lt07 lt09 02 lt05Au (ppb) 5 8 3 lt8 11 05 lt5 02Th 642 532 718 723 612 575 741 790U 199 161 122 102 165 139 225 158C K University of Vienna XRF and INAA data trace elements in ppm except as noted CH stands for Chasico BBQ stands for Bahiacutea Blanca RR stands

for distal glass probably related to BB CH-Sed and MdP-Sed stand for host sediments for impact glasses at Chasico and Mar del Plata


(coupled with the additional evidence of impact processdiscussed below) it is most reasonable to conclude that thesegrains represent shock melted droplets of quartz andfeldsparthat became entrained in the glasses duringexcavation and ejection from the impact site

Evidence for a wide range of shock metamorphicconditions are found among the abundant unmeltedplagioclase grains in the CH thin sections In general shockindicators are observed in less than 20 grains per thin sectionThese numbers increase steadily as each section isreexamined as a consequence of searching grain of minutesize which are in close contact with arrays of tiny quenchcrystals Some plagioclase grains (Fig 9a) contain prominentkink banding corresponding to relatively low pressurehighstrain rate formation Although kink bands also can occur intectonically deformed rocks (though rarely) they arecommon in impactites (Bunch 1968) Smith and Brown

(1988) list their association with closely spaced isotropic slipbands as a characteristic of naturally shocked feldspars VonEngelhardt and Stˆffler (1968) refer to this type ofdeformation as ldquoa typical criterion for shock damagerdquo

Moderately high shock conditions are evidenced by theformation of PDFs in plagioclase The most obvious PDFs arethose that form the ldquoladder texturerdquo (Fig 9b) first described indetail by Stˆffler (1966) from twinned plagioclase grains inRies suevites Short incipient or relict PDFs with multipleorientations are also observed within the crystalline regions ofother plagioclase grains which have undergone a peculiar typeof transformation apparently unique to impact metamorphism(Fig 9b) ldquoAlternate twin meltingrdquo (Reimold et al 2004) orldquoasymmetric isotropizationrdquo (Dressler 1990) is a phenomenonalso first described by Stˆffler (1966) in plagioclase grainsfrom the Ries This style of deformation results in theformation of diaplectic feldspar glass but only within albite

Fig 9 a) A PPL photomicrograph of a kink banded plagioclase grain in CH glass showing narrow parallel isotropic slip bands trending NW-SE through the grain The slip bands are deflected across a set of kink bands oriented SW-NE b) A cross-polarized light (XPL) image revealsthe boundaries between each kink band In one kink band (center) the axis of inflection diverges from the band boundary which may indicatesome shearing following the rotation of the lattice c) An XPL photomicrograph of a partially altered shock-metamorphosed plagioclase grainwithin a CH glass showing the distinct ldquoladder texturerdquo (eg Stoumlffler 1966 French 1968) formed by sets of closely spaced PDFs stacked likerungs between albite twin planes (tp)


twins of a particular set (typically ever other twin) This isbelieved to occur because the lattice of each twin in a givenset is rotated some equal number of degrees relative to theintervening set thereby causing one set to have an orientationrelative to the shock wave more favorable for the disorderingof the crystal The ldquosurvivingrdquo twin set commonly willcontain PDFs (Stˆffler 1966) but the exact pattern canbecome quite complicated depending on the original rotationof each twin (Dressler 1990) The pattern illustrated by thegrain in (Fig 10a) might be the result of the originalcomplexities in lattice orientations but it also might haveresulted from perturbations in the shock wave induced by pre-existing clay minerals lining the twin planes of the weatheredfeldspar The reddish-brown coloration of the amorphous

twins (in the original section) is similar to the ldquoalteredalternate albite twinrdquo effect discussed by Short and Gold(1996) as a signature of shock metamorphism in rocks fromthe Manson structure

The most common shock features of CH plagioclasegrains include the irregular isotropic patches of reducedrefractive index interpreted as zones of diaplectic glass ormaskelynite (Fig 10b) Commonly these grains also containdistinct broad wedge-shaped domains that appear to invadefrom the crystal margins These features appear similar to theL2 isotropic lamellae characterized by White (1993) at theManicouagan structure Like the L2 lamellae such features inChasicUcirc plagioclase grains also seem to be associated withdense arrays of PDFs

Fig 10 a) A PPL photomicrograph of shock-metamorphosed plagioclase grain within a CH glass The grain exhibits a pattern similar to theamorphitization of alternate albite twins first noted by Stˆffler (1966) in shocked plagioclase grains from the Ries crater In the original sectionthe glass twins have a distinctive reddish-brown hue b) PPL and c) XPL photomicrographs of a shocked plagioclase grain within a CH glassIrregular isotropic patches occur where the grain has been partially converted to diaplectic glass Note that the relatively weak glass haspreferentially been removed during grinding leaving several holes A wedge-shaped zone of partially devitrified glassy material extends intothe top of the grain This zone may have been densely packed with PDFs the outlines of which appear to be preserved despite somerecrystallization


The search for shocked quartz thus far has identified sixgrains containing one or more sets of decorated PDFs andexhibiting the common ldquotoastedrdquo coloration as detailed byWhitehead et al (2002) Of the six five grains are part of thesame tiny (sim500 microm) angular quartzite fragmentUnfortunately the extinction pattern of only one grain(Fig 11) is regular enough to facilitate indexing the planesaccording to established techniques using the Universal stage(Engelhardt and Bertsch 1969) The grain contains decoratedPDFs along five crystallographic planes corresponding (plusmn 4degerror) to (0001) 10 3 10 1 11 1 and 10 0 If thequartzite clast containing this grain is derived from crystallinerocks at depth these PDF orientations are indicative of shockpressures between 12 and 20 GPa (Stˆffler and Langenhorst1994)


The same mineral shock indicators recognized in theChasicUcirc glasses are present in the BahIgravea Blanca glassestherefore only supplemental examples are provided hereFigure 12 shows a classic example of asymmetricisotropization or alternate twin melting which is indicativeof crystallographically controlled generation of diaplecticglass under shock pressure Lechatelierite (Fig 13a) isindicative of temperatures exceeding 1700 degC (eg Hudonet al 2002) or shock pressures greater than at least 50 GPa(eg Stˆffler and Langenhorst 1994) The partially moltenrutile grains (Fig 13b) indicate more extreme temperatures

(gt1847 degC) (El Goresy 1968) Additional consequences ofshock loading in BB grains include well-developed planarfracturingfaulting in quartz (Fig 14) as well as diaplecticquartz (Figs 15a and 15b) and diaplectic feldspar (Figs 15cand 15d) Quartz grains exhibiting pronounced isotropizationplanar fracturingfaulting or mosaicism typically showreduced refractive indices and birefringence and contain oneor more sets of generally faint often discontinuous planarfluid inclusion trails that probably represent decorated andpartially annealed PDFs Other likely shock productsobserved include ldquoladder-likerdquo PDFs in plagioclasecheckerboard K-feldspar microtwinning in ilmenite well-developed planar fracturing and microfaulting in olivine andintense microtwinning in clinopyroxene The details of thesefeatures will be reported in a future contribution

RADIOMETRIC AGE

Four different 40Ar39Ar age determinations for glassesfrom the Miocene loessoid sections from ChasicUcirc and BahIgraveaBlanca were made All samples were analyzed using laserincremental heating methods and noble gas massspectrometry at the Berkeley Geochronology Center asdescribed in Renne et al (1998) Data for these samples wereevaluated with the statistical program described by Sharp andDeino (1996) Glass samples were prepared for irradiation bygently crushing with a ceramic mortar and pestle followed byhand picking of about 100ndash200 mg with a petrographicmicroscope to minimize the effects of alteration xenocrysts

Fig 11 A close-up PPL photomicrograph of a shocked quartz grainone of five grains that comprise a shocked quartzite fragment foundin CH-W The grain contains at least five sets of decorated PDFs Thedecoration consists of extremely straight chains of minute fluidinclusions Four PDF sets are observable in the shown orientationthe fifth set parallel to 10 3 is seen with the U-stage upon tiltingthe thin section

1

1 1 2 1

Fig 12 An XPL photomicrograph of a shocked plagioclase grainembedded in impact melt breccia from BB2 Every other albite twinhas been converted to an amorphous isotropic phase Note thatwithin that set of diaplectic twins every other member of the setappears to have been compressed This same pattern is observed inshocked plagioclase grains from other established impact sites (egStˆffler 1966) The large isotropic twins also exhibit a reddish-brownhue in PPL


and vesicles Following irradiation about 10ndash20 mg ofmaterial was selected through a final inspection forincremental heating analysis

Two age determinations were made for different samplesfrom ChasicUcirc (Fig 16a) These samples were selected fromlower (CH-L paired sample with CH-1 glass) and upperreworked sediment layers (paired sample with CH-U) asdescribed in the section on geologic setting The glass foreach of the ChasicUcirc samples was dark brown and relativelyfresh similar to the glasses from BahIgravea Blanca Sample CH-Lyielded an incremental heating spectrum with initial releaseof non-atmospheric extraneous 40Ar in the first step

followed by three steps with ages of sim920 Ma and finally bydecreasing age in the highest temperature steps Thedecreasing age of the highest temperature steps appears to bethe result of some minor contamination from a calcic phase(carbonate or plagioclase) but it appears to have little effecton the overall age spectrum A plateau age for the sample of923 plusmn 009 Ma is defined by the first four heating incrementsand 87 of the 39Ar released Because the radiogenic yieldsfor the main steps of the heating analysis are uniformly about95 there is little spread of the data and the data do notdefine an isochron Regression of all the data for this sampleyields an age similar to that defined by the plateau and anldquoinitialrdquo 40Ar36Ar ratio of 314 plusmn 16 (that is controlled mainlyby the first analysis) The incremental heating data for sampleCH-U are very similar with a plateau age of 93 plusmn 03 MaThe radiogenic yields for heating steps of CH-U are generallylower than with CH-L thereby producing a higheruncertainty and suggesting a greater degree of glassalteration This interpretation is consistent with thestratigraphic context indicating deposition of reworkedsediments and glasses for the upper deposit

The glasses from ChasicUcirc are interpreted as the same ageDue to the statistically identical results and relativeuncertainties the preferred weighted average of 924 plusmn

Fig 13 Evidence for extremely high temperature melting in BahIgraveaBlanca glasses a) A BSE electron micrograph showing a vesiculatedgrain of fused quartz (lechatelierite) which is indicative oftemperatures exceeding 1700 degC Lighter gray portions of the grainare pure SiO2 while darker regions represent infiltration of ions fromthe surrounding melt along cracks in the silica glass These regionsare slightly enriched in magnesium sodium and aluminum (andlikely water) with lesser concentrations of potassium iron andcalcium b) A BSE image of an embayed partially melted rutile grainproviding evidence that BB impact melts locally achievedtemperatures exceeding sim1850 degC (El Goresy 1968)

Fig 14 An XPL photomicrograph showing extremely well-developed planar fracturing in quartz Note that the N-S trendingcleavages are displaced along those trending E-W (basal planes)indicating microfaulting that must have been synchronous withproduction of the host melt Otherwise the resulting sharp edgesalong the grain margin would have been weathered away


009 Ma is adopted as the timing of the ChasicUcirc glassformation An earlier publication cited a preliminary age ofabout 10 Ma (with larger uncertainties) for the ChasicUcircglasses from the same exposure (Schultz et al 1999) based onanalyses in another laboratory The glasses from ChasicUcircanalyzed in the present study however are superior in qualityand quantity to the material previously analyzed Moreoverthe techniques employed in this study are considered superiorfor the resolution of extraneous argon and thus the age of924 plusmn 009 is now preferred

Two glass samples were analyzed from exposures nearBahIgravea Blanca The glass for each sample was relatively freshand free of xenocrysts although vesicles were conspicuousthroughout The 40Ar39Ar incremental heating release spectraof these samples (Fig 16b) are characterized by release ofextraneous atmosphere-rich argon (low radiogenic yields) inthe initial heating steps followed by increments with ages of

sim52ndash53 Ma that have radiogenic yields of 80 or greaterPlateau ages for samples BB-1 and BB-2 are 532 plusmn 005 Maand 524 plusmn 005 Ma respectively Since there is considerablespread in their atmospheric argon contents each sampledefines an isochron the 40Ar39Ar ratio indicated for thetrapped argon in BB-1 is 2973 plusmn 16 and for BB-2 it is 2987plusmn 12 As these ratios are very near that of modernatmospheric argon and the higher-temperature release stepsare fairly radiogenic the isochron ages are identical to theplateau ages Our preferred assessment of the formation agefor glass from BahIgravea Blanca is 528 plusmn 004 Ma obtained byaveraging the plateau ages

DISCUSSION

The radiometric dates and stratigraphic placementestablish two different layers of impact glasses The 92 Ma

Fig 15 Shock effects in BahIgravea Blanca quartz and feldspar a) PPL and b) XPL photomicrographs of quartz grain containing broad regionscharacterized by reduced refractive indices and isotropy interpreted to be diaplectic glass c) PPL and d) XPL photomicrographs of feldspargrain showing extensive recrystallization inside well-preserved grain boundaries indicative of either diaplectic feldspar or fused feldspar thatquenched sufficiently quickly to prevent flow and deformation of the original grain boundary Either explanation requires impact shockprocesses (French 1998)


Fig 16 a) 40Ar39Ar incremental heating spectra for clast-free fractions two ChasicUcirc impact glasses Ch-U corresponds to material fromreworked deposits slightly higher in the section than CH-L The plateau and isochron age results are nearly the same The weighted averageyields an age of 924 plusmn 009 Ma


Fig 16 Continued b) 40Ar39Ar incremental heating spectra for clast-free fractions from two samples (BB-1 at top and BB-2 bottom) fromBahiacutea Blanca yielding an age of 528 plusmn 004 Ma as discussed in the text


ChasicUcirc impact glasses are primarily exposed along portionsof a river section that has incised upper Miocene sedimentswhereas the 52 Ma BahIgravea Blanca glasses occur in isolatedexposures and quarries higher in the stratigraphic sectionObvious primary ejecta facies have not yet been confidentlyidentified other than concentrations of large impact glassmasses In both cases the size of the impact glasses andclustered fields suggests a nearby parent crater

The CH and BB glass compositions from SBA locationsgenerally resemble each other but differ significantly fromthe Pliocene and Pleistocene glasses exposed near Mar delPlata These differences reflect contrasting source materialsboth in location and depth (time)

Ratios of bulk glass major elements for CH and BB to thehost sediments also indicate enrichment in K2O and P2O5relative to the host sediments but little difference in otherconstituents (Fig 4) Other differences are revealed byunmelted minerals and clasts normative calculations forcertain glasses quench crystals and trace elements

The much greater fraction of unmelted Ca-pyroxenes inthe CH and BB glasses indicate a mix of more pristine maficcomponents than sediments found elsewhere in SBA Therange in bulk compositions of the glasses and the quenchcrystals also indicate more mafic precursors Although highbarium levels (2000 ppm) could be attributed to entrainedmafic lasts the extremely high level of Ba (sim7600 ppm) inCH-1 is not associated with increases in other trace elements(Hf Eu Tb) as would be expected if it were from a purelymafic source Such a large enhancement may represent eitherexcavation of barium-rich evaporates (or hydrothermal

deposit) or incorporation of biogenic barite from the watercolumn (if a shallow marine impact) Alternativelyenrichment by post-impact hydrothermal processes couldhave played a role

This particular glass may have incorporated Tertiarymarine sequences which are known to occur at depth orhydrothermal alteration More detailed analyses will benecessary to understand the significance of suchcomponents

Comparisons of V Rb La Th Sc and other minor andtrace elements of both the bulk sediments and impact glassesfrom SBA reveal systematic changes over time (Table 5) Thedifferences can be understood by the increasing contributionsfrom Andean volcaniclastics (and transport by-products)associated with the late Miocene through PleistoceneCordilleran uplift (Schultz et al 2004) or evolving sourceregions in response to climate and dominant wind direction asnoted for the late Pleistocene (Smith et al 2003) Figure 17shows this trend in the decreasing value of ThSc withincreasing age The generally lower values of ThSc for someglasses relative to their host sediments suggest incorporationof materials from depth (eg CH-1) The BB samples bothhave ThSc levels lower than even the older ChasicUcircsediments More systematic studies may reveal trends relatedto the proximity to the source crater

Figure 18 is ternary plot comparing for ThSc LaTh andLaSc in the late Cenozoic impact glasses and associatedsediments Both the host sediments and impact glasses fromvarious localities become progressively more silicic with time(to the present) which reflects evolving source areas Three

Fig 17 Changes in ThSc values for sediments benchmarked by dated impact glasses from Rio Cuarto (RC) Centinela del Mar (CdM) Mardel Plata (MdP) and ChasicUcirc (CH) The increase in ThSc values since the Miocene reflect either evolving source regions through time frombasaltic andesitic toward andesitic or incorporation of a more mafic source region at depth Impact glasses from ChasicUcirc (CH) and BahIgraveaBlanca (BB) suggest even less silicic sources indicating different source materials perhaps due to incorporation of materials from depth


of the analyzed glasses (one from BB the others CH) indicatesource materials significantly less silicic

Consequently the impact glasses of ChasicUcirc generallyrepresent fused surface sediments reworked from volcanicfines and silts from the Miocene While they generallyresemble the composition of other post-Miocene glasses theMiocene glasses differ by being generally more mafic andhaving more extensive quench crystallization Althoughclassified as andesitic (based on major elements) the glassesalso contain mafic mineral clasts and a higher fraction ofunmelted pyroxenes Incorporation and decomposition ofcarbonate-rich claystones or mixed marine deposits alsocould contribute to locally higher CaO and lower silicacontents in the melt in certain samples The ThSc ratiosindicate that the BB and some of the CH impacts glassesincorporated less silicic materials (on average) from theTertiary prior to the late Miocene uplift responsible for theloessoid sediments comprising most of the Pampas Thegreater fraction of unmelted mafic xenocrysts within the CHglasses indicates materials either transported from Miocenemafic volcanics along the eastern Andes or excavated fromunits not yet sampled at depth

IMPLICATIONS

Cratering Flux Estimates

The seemingly large number of distinct ages of impactglasses could reflect at least in part a northern hemisphereperspective where repeated Pleistocene glaciations erodedburied or erased the true impact cratering record But it alsomay reflect either an incomplete understanding ofatmospheric entry or perhaps even an increased flux rate ofinterplanetary debris during the late Miocene

Disparities between predictions for terrestrial craterproduction rates depend on different approaches andassumptions above-atmosphere flux rates as a basis for thecratering production rate (eg Shoemaker et al 1990Chapman and Morrison 1994 Stuart 2001) downwardextrapolation of the terrestrial cratering record (eg Schultzet al 2004b [based on Shoemaker et al 1990]) rates ofobserved atmospheric bursts (eg Brown et al 2002) andtheoretical models of atmospheric disruption and survival(eg Bland and Artemeiva 2003 Bland et al 2004)

Each approach for assessing the likelihood of melt-

Fig 18 A ternary diagram illustrating the systematic differences between the compositions of Argentine impact glasses (and the sedimentsthey are found in) RC (Rio Cuarto) represents two events at sim6 ka and sim120 ka CdM (Centinela del Mar) represents one event at sim445 kaMdP (Mar del Plata) represents a late Pliocene impact at 33 Ma (Schultz et al 2004) CH (ChasicUcirc) and BB (BahIgravea Blanca) represent the lateMiocene events (sim53 Ma and sim92 Ma respectively) reported here Also plotted are the compositions of Holocene mafic ash (MA) deposits(MA) and Pleistocene silicic ash (SA) recovered from the sections The cutout shows values for glasses (filled symbols) relative to sedimentsat ChasicUcirc (open square) CH and BB glasses follow the same thread but are generally less silicic than the Miocene host sediments at ChasicUcircCH impact glasses are more mafic than the host sediments consequently the CH impact may have excavated unexposed units at depth Theshaded field indicates the range of compositions (from Chapman and Scheiber [1969]) for the sim10 Ma (Bottomley and Koeberl 1999) HNaK tektites


generating impacts has limitations The use of either the lunarcratering rates or observations of near-Earth crossingasteroids (NEAs) cannot be simply extrapolated to the surfaceof the Earth without model assumptions for the survivalduring atmospheric entry Estimates of the expectedproduction of craters subject to models of atmosphericdisruption (Bland and Artemeiva 2003) yield cratering rates afactor of 100times less than the above-atmosphere rate 20times lessthan predictions by Chapman and Morrison (1994) and stillabout 10times less than extrapolations from large to small craters

When specifically applied to the cratering rate in all ofSouth America Bland et al (2004) estimated that 18 craters gt10 km should have formed during the last 100 Myr Duringthe last 1 Myr 1 crater gt 1 km and sim200 Campo del Cieloevents should have occurred More specifically about 200small asteroids larger than 50 m across should haveencountered the upper atmosphere during the last 10 Myrover an area covered by the Argentine Pampean sediments(sim1106 km) based on rates in Bland and Artemeiva (2003)Their best-fit atmosphere-disruption model predictshowever that a 50 m object would strike the surface with onlya 1-in-5 chance in 10 Myr over 106 km2 Although 100 objectsgreater than 3ndash5 m across could have actually impacted over10 Myr most (95) of them would have been irons (possibleforming 100 m diameter craters) For comparison otherstudies of the observed flux of bolides (atmospheric bursts)from satellite data (eg Brown et al 2002) yield about 20objects gt 50 m in diameter (Tunguska-class) 6 objects gt75 m and 3 objects gt 100 m over an area of 106 km2 in the last10 Myr

If the observed production rate of larger craters (gt 20 km)is simply extrapolated downwards to smaller craters (1ndash10 km) then higher cratering rates since the Miocene inArgentina naturally result (eg Schultz et al 2004) Forexample more than 900 craters with diameters (D) greaterthan 10 km should have occurred globally over the last100 Myr based on the asteroid flux rates (Shoemaker et al1990) If the production rate depends on D-2 then thePampean sediments covering one million square kilometerscould contain sim20 craters larger than 10 km in diameter overthe last 10 Myr 7 craters over the last 25 Myr and perhaps 3in the last 10 Myr An even larger number of craters would beexpected if the dependence has a larger exponent (D-25)

It is possible that impact melts (glasses) can be generatedfrom craters as small as 05 km in diameter due to softsediments for example Aeouellel Monturaqui and Wabarcraters all have large quantities of impact glass Thisobservation however still does not address the role ofatmospheric disruption which is believed to prevent the mostlikely objects (small stony meteoroids) from contributing tothe crater population (Bland and Artemeiva 2003)

Consequently the discovery of at least six occurrences ofimpact melt preserved in relatively continuous deposits overthe last 10 Ma in Argentina may have implications for re-considering models of objects surviving entry to producecraters a better understanding of melt generation from softsediments or additional evidence for an enhanced flux rateover the last 10 Myr The last suggestion would be consistentwith evidence for increased flux of interplanetary dustConcentrations of 3He relative to 4He are found in ocean floor

Fig 19 A comparison of averaged rare earth elements for 92 Ma impact glasses from ChasicUcirc (solid squares) and 53 Ma glasses from BahIgraveaBlanca (solid circles) with averaged values for tektites with NaK ratios (from data given by Chapman and Schieber [1969]) shown by opencircles Similarity suggests that Argentine Miocene sediments could represent a viable source region Radiometric and fission-track ages forthe HNaK glasses also are approximately 10 Ma (Storzer 1985 Bottomley and Koeberl 1999) which is close to the well-constrainedradiometric date of the ChasicUcirc glasses


sediments between sim3 and 10 Ma (Farley 1995) The uniquedepositional record of Argentina may be preserving a landrecord of larger objects during this time as well

Possible Connection with HNaK Tektites

Radiometric and fission track ages 102 plusmn 05 Ma(Bottomley and Koeberl 1999) and 835 plusmn 090 Ma (1σ)(Storzer 1985) respectively indicate that the high sodiumpotassium (HNaK) Australian tektites are much older thanother Australasian tektites (07 Ma) that are found in the samearea However the age of this enigmatic tektite group is closeto the 924 plusmn 009 Ma (2σ) age of the ChasicUcirc glasses

In addition to their age and simple morphologies HNaKtektites also are compositionally distinct from members of theAustralasian strewn field (Chapman and Schieber 1969Compston and Chapman 1969 Taylor and Epstein 1969Montanari and Koeberl 2000) and are believed to haveoriginated from basaltic to andesitic target rocks Althoughthe CI-normalized REE abundances in HNaK tektites aresignificantly lower than in Australasian tektites their REEpatterns (Chapman and Schieber 1969) generally resemblethose for CH glasses (Fig 19)

Ternary plots such as Fig 18 also are commonly used todifferentiate the major tektite strewn fields and their sourceregions (eg Albin et al 2000) In the comparison of ThScLaTh LaSc the composition of HNaK tektites can beexplained as a mixture of predominantly mafic basaltic toandesitic rocks with a more Th-rich ldquocontinentalrdquo componentA nearly identical composition (see Fig 18) is found in theupper Miocene Pampean sediments (and the CH and BBimpact glasses produced from them) which representmixtures of early Andean volcanics and weatheredPatagonian basalts with lesser amounts of eroded basement

Although further work is necessary to demonstrate theconnection the age of the ChasicUcirc glasses and closesimilarities in important aspects of both major and trace-element chemistry suggest that a relationship with the HNaKtektites is plausible The possibility that Pampas is the sourceregion of HNaK tektites should be considered High-velocitymolten debris ejected at high angles (eg vapor entrainedmelts) at mid-latitudes result in a preferential re-entryconcentration to the west due to the Coriolis force (Wrobeland Schultz 2003) in this case over the Pacific and Australia

CONCLUDING REMARKS

The Argentine loessoid sequences are yielding anunprecedented archive of well-preserved impact materialsdating back to the late Miocene The 924 Ma age ChasicUcircimpact glasses may be associated with a source crater nearbybased on the size and distribution of the materials Althoughthe 528 Ma glasses are more widespread due to availableexposures a specific source has not yet been identified These

glasses closely correspond to the MiocenePliocene boundarybut a causal relationship cannot yet be inferred The REEpatterns for the ChasicUcirc glasses generally match both thecomposition and age of the enigmatic HNaK subgroup oftektites thereby opening the possibility for a source regionquite distinct from the rest of the Australasian strewnfield

Because the impact glasses establish a precise timelinethey can be used to establish unique chronostratigraphicbenchmarks for assessing mammalian evolution on-landrecord of paleoclimates and the evolving source regions Thegeochemistry of the bulk sediments reveals previouslyunrecognized differences related to changes in source regionssyngenetic with the late Miocene Andean orogeny And thedifferences in compositions between the impact glasses andhost sediments provide evidence that the impact incorporatedmaterials from depth in contrast with Rio Cuarto impactglasses that incorporated near-surface materials (Aldahanet al 1998 Schultz et al 2004) Impact glasses also providecritical tracers of processes that result in mixing of associatedfossil remains thereby potentially complicating inferredfaunal evolution The impact glass horizons and reworkedmaterials will contribute to better constrain interpretations ofthese fossil-rich Miocene sequences (eg Zmiddotrate et alForthcoming) while establishing a new reference for impactproducts in sedimentary settings

Editorial HandlingmdashDr John Spray

REFERENCES

Albin E F Norman M D and Roden M 2000 Major and traceelement compositions of georgiaites Clues to the source of NorthAmerican tektites Meteoritics amp Planetary Science 35795ndash206

Aldahan A A Koeberl C Possnert G and Schultz P 1997 10Be andchemistry of impactites and target materials from the Rio Cuartocrater field Argentina Evidence for surficial cratering andmelting GFF 11967ndash72

Berggren W A Kent DV Swisher CC III and Aubry M P 1995A revised Cenozoic geochronology and chronostratigraphy InGeochronology time scales and global stratigraphic correlationedited by Berggren W A Tulsa Oklahoma Society forSedimentary Geology pp 129ndash212

Bland P A de Souza Filho C R Tull A J T Kelley S P HoughR M Artemieva N A Pierazzo E Coniglio E Pinotti LEvers V and Kearsley A T 2002 A possible tektite strewnfieldin the Argentine pampas Science 2961109ndash1111

Bland P A and Artemieva N A 2003 Efficient disruption of smallasteroids by Earthrsquos atmosphere Nature 424288ndash290

Bland P A Artemieva N A and de Souza Filho C R 2004 Theproduction rate of small craters on Earth and the expected craterpopulation in South America (abstract) Meteoritics amp PlanetaryScience 39A16

Bischoff A and Stˆffler D 1984 Chemical and structural changesinduced by thermal annealing of shocked feldspar inclusions inimpact melt rocks from Lappaumljarvi crater Finland Journal ofGeophysical Research 89B645ndashB656

Bondesio P Laza J H Scillato YanEgrave G J Tonni E P andVucetich M 1980 Estado actual del conocimiento de losvertebrados de la FormaciUcircn Arroyo ChasicUcirc (plioceno


temprano) de la provincia de Buenos Aires Actas II CongresoArgentino de PaleontologIgravea y BioestratigrafIgravea y I Congresolatinoamericano de PaleontologIgravea Buenos Aires 1978 T III1980101ndash127

Bonorino A G Schillizzi R and Kostadinoff J 1987 InvestigaciUcircngeolUcircgica y geofIgravesica de la regiUcircn de BahIgravea Blanca Actas IIIJornadas Pampeanas de Ciencias Naturales UniversidadNacional de la Pampa Serie Suplementos 355ndash63

Brown P Spalding R E ReVelle D O Tagliiafieri E andWorden S P 2002 The flux of small Earth-crossing objectscolliding with the Earth Nature 420314ndash316

Bottomley R J and Koeberl C 1999 The age of a separateAustralasian tektite event Meteoritics amp Planetary Science 34A15

Bunch T E 1968 Some characteristics of selected minerals fromcraters In Shock metamorphism of natural materials edited byFrench B M and Short N M Baltimore Maryland Mono BookCorporation pp 413ndash432

Carstens H 1975 Thermal history of impact melt rocks in theFennoscandian Shield Contributions to Mineralogy andPetrology 50145ndash155

Catti M Freyria Fava F Zicovich C and Dovesi R 1999 High-pressure decomposition of MCr2O4 spinels (M=Mg Mn Zn) byab initio methods Physics and Chemistry of Minerals 26389ndash395

Chapman C R and Morrison D 1994 Impacts on the Earth byasteroids and comets Assessing the hazard Nature 36733ndash40

Chapman D R and Scheiber L C 1969 Chemical investigation ofAustralasian tektites Journal of Geophysical Research 746737ndash6776

Compston W and Chapman D R 1969 Sr isotope patterns withinthe Southeast Australasian strewn-field Geochimica etCosmochimica Acta 331023ndash1036

Deschamps C M 2003 EstratigrafIgravea y paleoambientes en elCenozoico del sur de la provincia de Buenos Aires El aporte delos vertebrados PhD thesis Facultad de Ciencias Naturales yMuseo Universidad Nacional de La Plata La Plata Argentina317 p

Devine J D Gardner J L Brack H P Layne G D andRutherford M J 1995 Comparison of microanalytical methodsfor estimating H2O contents of silicicvolcanic glasses AmericanMineralogist 80319ndash328

Dressler B 1990 Shock metamorphic features and their zoning andorientation in the Precambrian rocks of the ManicouaganStructure Quebec Canada Tectonophysics 171229ndash245

El Goresy A Fechtig H and Ottemann J 1968 The opaqueminerals in impact glasses In Shock metamorphism of naturalmaterials edited by French B M and Short N M BaltimoreMaryland Mono Book Corporation pp 531ndash553

Farley K A 1995 Cenozoic variations in the flux of interplanetarydust recorded by 3He in deep-sea sediment Nature 376153ndash156

Fidalgo F Laza J H Porro N and Tonni E P 1978 AlgunascaracterIgravesticas de la FormaciUcircn Arroyo ChasicUcirc y sus relacionesgeolUcircgicas VII Congreso GeolUcircgico Argentino Actas I pp 213ndash225

French B M 1968 Sudbury structure Ontario Some petrographicevidence for an origin by meteorite impact In Shockmetamorphism of natural materials edited by French B M andShort N M Baltimore Maryland Mono Book Corporation pp383ndash412

French B M Cordua W S and Plescia J B 2004 The Rock Elmmeteorite impact structure Wisconsin Geology and shock-metamorphic effects in quartz Geological Society of AmericaBulletin 116200ndash218

Grieve R A F Langenhorst F and Stˆffler D 1996 Shock

metamorphism of quartz in nature and experiment IISignificance in geoscience Meteoritics amp Planetary Science 316ndash35

Harris R S Schultz P H and Bunch T E 2005a Accessory phasesin Argentine impact breccias Implications for shock historyemplacement dynamics vapor composition and target lithologies(abstract 1952) 36th Lunar and Planetary Science ConferenceCD-ROM

Harris R S Schultz P H and Bunch T E 2005b Evidence forshocked feldspars and ballen quartz in 450000 year oldArgentine impact melt breccias (abstract 1966) 36th Lunar andPlanetary Science Conference CD-ROM

Hudon P Jung I and Baker D R 2002 Melting of β-quartz up to20 GPa and thermodynamic optimization of the silica liquidusup to 60 GPa Physics of the Earth and Planetary Interiors 130159ndash174

Iriondo M H 1997 Models of deposition of loess and loessoids inthe Upper Quaternary of South America Journal of SouthAmerican Earth Sciences 1071ndash79

Kieffer SW 1971 Shock metamorphism of the Coconino sandstoneat Meteor Crater Arizona Journal of Geophysical Research 765449ndash5473

Kieffer S W Phakey P P and Christie J M 1976 Shock processesin porous quartzite Transmission electron microscopeobservations and theory Contributions to Mineralogy andPetrology 5949ndash93

Marshall L G Hoffstetter R and Pascual R 1983 Mammals andstratigraphy Geochronology of the continental mammal-bearingTertiary of South America Palaeovertebrata MontpellierMeacutemoir Extr 1ndash93

Montanari A and Koeberl C 2000 Impact stratigraphy The Italianrecord BerlinSpringer-Verlag 364 p

Muhs D and Zmiddotrate M 2001 Late Quaternary aeolian records of theAmericas and their paleoclimatic significance InInterhemispheric climate linkages edited by Markgraf V SanDiego Academic Press pp 183ndash216

Orgeira M J Walther A M Vasquez C A dIgrave Tomasso IAlonso S Sherwood G H Hu Y G and Vilas J F A 1998Mineral magnetic record of paleoclimatic variation in loess andpaleosol from the Buenos Aires formation (Buenos AiresArgentina) Journal of South American Earth Sciences 11561ndash570

Pascual R Ortiz Jaureguizar E and Prado J 1996 Land mammalsParadigm for Cenozoic South American geobiotic evolution InContributions of southern South America to vertebratepaleontology edited by Arratia G Muumlnchen Pfeil pp 265ndash319

Reimold W U Kelley S P Sherlock S Henkel H and Koeberl C2004 Laser argon melting of melt breccias from the Siljanimpact structure SwedenmdashImplications for possible relations tolate Devonian extinction events (abstract 1480) 35th Lunar andPlanetary Science Conference CD-ROM

Renne R R Swisher C C Deino A L Karner D B Owens T andDePaolo D J 1998 Intercalibration of standards absolute agesand uncertainties in 40Ar39Ar dating Chemical Geology 145117ndash152

Schultz P H Koeberl C Bunch T E Grant J A and Collins W1994 Ground truth for oblique impact processes New insightfrom the Rio Cuarto Argentina crater field Geology 22889ndash892

Schultz P Zmiddotrate M Hames W CamiliUcircn C and King J 1998 A33 Ma impact in Argentina and possible consequences Science2822061ndash2063

Schultz P H Zmiddotrate MA and Hames W 1999 Three newArgentine impact sites Implications for Mars (abstract 1898)30th Lunar and Planetary Science Conference CD-ROM


Schultz P H Zmiddotrate M A Hames W King J Heil CKoeberl C Renne P R and Blasi A 2002a Argentine impactrecord Abstracts with programs Geological Society ofAmerica Boulder Colorado Geological Society of Americap 401

Schultz P H Zmiddotrate M A King J Blasi A Hames W 2002bFormation and evolution of impact craters in the ArgentinePampas Actas del XV Congreso Geologico Argentino pp 179ndash181

Schultz P H Zmiddotrate M A Hames W Koeberl C Bunch T ERenne P R and Wittke J 2004 The Quaternary impact recordfrom the Pampas Argentina Earth and Planetary ScienceLetters 219221ndash238

Sharp W and Deino A L 1996 Laser heating of samples for Ar-Argeochronology Eos 77F773

Shoemaker E M Wolfe R F and Shoemaker C S 1990 Asteroidsand comet flux in the neighborhood of Earth In Globalcatastrophes in Earth history An interdisciplinary conference onimpacts volcanism and mass mortality edited by Sharpton V Land Ward P D Boulder Colorado Geological Society ofAmerica pp 155ndash170

Short N M and Gold D P 1996 Petrography of shocked rocks fromthe central peak at the Manson impact structure In The Mansonimpact structure Iowa Anatomy of an impact crater edited byKoeberl C and Anderson R R pp 245ndash265

Simonson B M 2003 Petrographic criteria for recognizing certaintypes of impact spherules in well-preserved Precambriansuccessions Astrobiology 349ndash65

Smith J V and Brown W L 1988 Feldspar minerals Crystalstructures physical chemical and microtextural properties vol1 Berlin Springer-Verlag 828 p

Smith J Vance D Kemp R Archer C Toms P King M andZmiddotrate M 2003 Isotopic constraints on the source of Argentinianloess-with implications for atmospheric circulation and theprovenance of Antarctic dust during recent glacial maximaEarth and Planetary Science Letters 212181ndash196

Stˆffler D 1966 Zones of impact metamorphism in the crystallinerocks at the Nˆrdlinger Ries crater Contributions to Mineralogyand Petrology 1215ndash24

Stˆffler D 1984 Glasses formed by hypervelocity impact Journal ofNon-Crystalline Solids 67465ndash502

Stˆffler D and Langenhorst F 1994 Shock metamorphism of quartz

in nature and experiment Basic observation and theoryMeteoritics 29155ndash181

Storzer D 1985 The fission track age of high sodiumpotassiumaustralites revisited Meteoritics 20765ndash766

Stuart J S 2001 A near-Earth asteroid population estimate from theLINEAR Survey Science 2941691ndash1693

Taylor H P and Epstein S 1969 Correlations between 18O16O ratiosand chemical compositions of tektites Journal of GeophysicalResearch 746834ndash6844

Taylor S R and McLennan S M 1985 The continental crust Itscomposition and evolution Oxford Blackwell 312 p

von Engelhardt W and Stˆffler D 1968 Stages of shockmetamorphism in crystalline rocks of the Ries basin Germany InShock metamorphism of natural materials edited by FrenchB M and Short N M Baltimore Maryland Mono BookCorporation pp 159ndash168

von Engelhardt W and Bertsch W 1968 Shock induced planardeformation structures in quartz from the Ries craterContributions to Mineralogy and Petrology 20203ndash234

White J C 1993 Shock-induced amorphous textures in plagioclaseManicouagan Quebec Canada Contributions to Mineralogyand Petrology 113524ndash532

Whitehead J Spray J G and Grieve R A F 2002 The origin ofldquotoastedrdquo quartz in terrestrial impact structures Geology 30431ndash434

Wrobel K E and Schultz P H Accumulation of distal impact ejectaon Mars since the Hesperian (abstract 3242) Sixth InternationalConference on Mars CD-ROM

Zmiddotrate M 2003 The loess record of Southern South AmericaQuaternary Science Reviews 221987ndash2006

Zmiddotrate M Schultz P H Blasi A Heil C King J and Hames WForthcoming Geology and geochronology of the type ChasicUcircan(late Miocene) mammal-bearing deposits of Buenos AiresProvince (Argentina) Journal of South American EarthSciences

Zambrano J 1980 Comarca de la cuenca cretmiddotcica del ColoradoGeologIgravea Regional Argentina volumen II pp 1033ndash1070

Zinck J A and Sayago J M 1999 Loess-paleosol sequence of LaMesada in Tucummiddotn province northwest ArgentinaCharacterization and paleo-environmental interpretationJournal of South American Earth Sciences 12293ndash310


in large quantities at specific horizons within pre-Holocenesediments a different origin is required especially in theabsence of nearby volcanic sources Studies of such glasses inLate Pliocene to Holocene Pampean sediments haveestablished that they were formed as the result ofhypervelocity impact (Schultz et al 1994 1998 2004 Harriset al 2005a 2005b)

The identification and characterization of older Pampeanglasses would not only improve our understanding of possibleimpact products in the sedimentary record but also wouldprovide needed chronostratigraphic markers These oldersuccessions contain important fossil assemblages that wereproposed to track mammalian evolution and interchangesthrough time (Pascual et al 1996 Marshall et al 1983)Therefore radiometric dates of instantaneously formedimpact glasses provide unique benchmarks for re-evaluatingfaunal evolution

Here we report the discovery of two new glass depositsconcentrated along specific horizons in late Miocene strataand provide evidence for their impact origin The oldest glasslayer dates to about 92 Ma and provides a critical benchmarkfor important Miocene outcrops The second concentrationdates to about 53 Ma which is approximately the MiocenePliocene boundary (Berggren et al 1995) We first review thePliocene to Pleistocene impact record in the Pampas ofArgentina and then focus on the specific geologic setting ofthe two new localities Next we discuss the nature (includingevidence for impact origin) and age of the impact glasses

Lastly we consider the possible implications including thesize of the events comparison with other impact glasses inyounger sediments the production rate of glass-producingimpacts and the possible relation to the enigmatic AustralianHNaK tektites with an age similar to that of the ChasicUcircglasses

BACKGROUND

Argentine Impact Glasses

At least four impact glass deposits have been discoveredin post-Miocene Pampean sediments (Schultz et al 19941998 2002a 2002b 2004 Bland et al 2002) These impactglasses often are attached to or wrap around tierra cocidas thatare composed of oxidized and welded sediments that arebelieved to be lower-temperature impact products theequivalent of clastic breccias derived from unconsolidatedsediments (Schultz et al 1998 2004) The glasses containclear evidence for intense transient heating to temperaturesexceeding 1700 degC followed by rapid cooling and dynamicemplacement (Schultz et al 1994 1998 Harris et al 2005a)Shock indicators have been documented in the Pliocene andoldest Pleistocene glasses (Harris et al 2005a 2005b)

The two new glass occurrences described in this paperare exposed in sediments of the Southern Buenos Aires (SBA)Province (Fig 1) The older glasses occur in well-knownupper Miocene loessoid strata composed primarily of siltysands containing carbonate nodules and at certain levelscarbonate cements These generally unconsolidated depositscontain late Miocene fossil mammal assemblagescharacterized as the ldquoChasicUcircan assemblagesrdquo by Fidalgoet al (1978) and Bondesio et al (1980) While other materialsof ChasicUcircan age crop out far to the northwest (the central partof La Pampa Province San Luis San Juan and Catamarca)the exposures closer to the Atlantic coast are limited to a veryrestricted area surrounding Laguna de ChasicUcirc and SalinasChicas The ChasicUcirc (CH) glasses in this region are typically1ndash10 cm in diameter but have been found in tightly clusteredassemblages as large as 2 m across Although they do notform a continuous bed of melt they occur scattered withinspecific stratigraphic horizons that can be traced for morethan 5 km (discussed below)

The second glass layer is widespread in upper Miocene tolower Pliocene outcrops throughout the Southern BuenosAires (SBA) Province from BahIgravea Blanca (BB) east into theArroyo ClaromecUcirc Basin (east of the Sierra La Ventanamountain range comprised of Paleozoic basement rocks) andwest to near GuaminIgrave Multiple localities near BahIgravea Blancayield the largest concentrations of glasses (up to 10 cmacross) Our initial suspicion was that these glassesrepresented distal equivalents of the ChasicUcirc glasses initiallyemplaced on the Sierra La Ventana piedmont but transportedand reworked into much younger sediments through fluvial

Fig 1 A map showing the primary locations where Chasicoacute (CH)and Bahiacutea Blanca (BB) glasses have been investigated Relatedglasses occur in MioPliocene outcrops along a railroad cut (RR) nearBB and at other sites scattered throughout southern Buenos AiresProvince Pliocene and Pleistocene impact glasses occur in thevicinities of Mar del Plata (MDP) and Centinela del Mar (CdM)respectively




GEOLOGIC SETTING









ChasicUcirc Glasses





























9580 9884 9518 9630 9800 978 9531 9655 934





















Geochemistry



SHOCK PETROGRAPHY

ChasicUcirc Glasses




























RADIOMETRIC AGE



1

1 1 2 1













DISCUSSION




















IMPLICATIONS




















CONCLUDING REMARKS





REFERENCES












































































GEOLOGIC SETTING









ChasicUcirc Glasses





























9580 9884 9518 9630 9800 978 9531 9655 934





















Geochemistry



SHOCK PETROGRAPHY

ChasicUcirc Glasses




























RADIOMETRIC AGE



1

1 1 2 1













DISCUSSION




















IMPLICATIONS




















CONCLUDING REMARKS





REFERENCES




































































































9580 9884 9518 9630 9800 978 9531 9655 934





















Geochemistry



SHOCK PETROGRAPHY

ChasicUcirc Glasses




























RADIOMETRIC AGE



1

1 1 2 1













DISCUSSION




















IMPLICATIONS




















CONCLUDING REMARKS





REFERENCES




























































































Geochemistry



SHOCK PETROGRAPHY

ChasicUcirc Glasses




























RADIOMETRIC AGE



1

1 1 2 1













DISCUSSION




















IMPLICATIONS




















CONCLUDING REMARKS





REFERENCES

































































































RADIOMETRIC AGE



1

1 1 2 1













DISCUSSION




















IMPLICATIONS




















CONCLUDING REMARKS





REFERENCES




















































































DISCUSSION




















IMPLICATIONS




















CONCLUDING REMARKS





REFERENCES


























































































IMPLICATIONS




















CONCLUDING REMARKS





REFERENCES























































































CONCLUDING REMARKS





REFERENCES









































































the record of miocene impacts in the argentine pampas€¦ · the record of miocene impacts in the...

Documents