GERTA KELLER PUBLICATIONS

We report the results of stable oxygen and carbon isotope analysis of carbonate fine fraction and benthic and planktic foraminifers, %CaCO3 and TOC determinations of sediments across the Cretaceous/Tertiary (K/T) boundary at El Kef, Tunisia. We used grain-size analysis of the fine fraction carbonate to determine possible compositional effects and smearslide and scanning electron microscope examinations to evaluate possible diagenetic alteration and microfossil preservation. The data suggests the following sequence of events.

Upper Maastrichtian oxygen and carbon isotope signals indicate relatively warm temperatures, a decreasing thermal gradient possibly related to a shallowing sea and relatively high surface water productivity. A sharp cooling appears to begin just below the K/T boundary. The K/T boundary is characterized by a sudden 2‰ negative shift in fine fraction δ13C, a slight 0.6‰ enrichment in benthic δ13C and a reduction of about 40% in CaCO3 sedimentation. These data imply strongly reduced surface water productivity as also observed in numerous deep-sea sequences worldwide. The lower Danian (planktic foraminiferal Zones P0,P1a) is marked by generally low productivity and unstable environmental conditions as indicated by generally low but fluctuating fine fraction δ13C values and low (5–15%) carbonate deposition. At the base of Subzone P1b, %CaCO3, benthic and fine fraction δ13C values increase gradually, and reach near pre-K/T boundary levels in Subzone P1c indicating initial recovery after the K/T boundary event about 300,000–400,000 years after the K/T boundary. The prolonged low productivity episode after the K/T crisis and the pre-K/T boundary cooling associated with a major reduction in planktic foraminiferal diversity are difficult to explain by a single K/T boundary bolide impact.  PDF

High-resolution planktonic foraminiferal analysis of three Brazos River sections indicates a nearly continuous Cretaceous/Tertiary [K/T] boundary sedimentary record second only to the world's most complete record at El Kef, Tunisia. Species extinctions occur over an extended period of time and with two major extinction episodes. The first extinction episode with 46% of the species extinct occurs at and just below [10–15 cm] a short hiatus at the base of a sandy shell hash and clay-sand unit which was interpreted by Bourgeois et al. [1988] to represent a tsunami bed generated by the K/T boundary bolide impact. The top of this tsunami bed is about 17–20 cm below the K/T boundary as defined by the first appearance of Tertiary planktonic foraminifera. The second extinction phase with 45% of the species extinct occurs 25 cm above the K/T boundary [Zone P0/P1a boundary]. Of the remaining seven surviving Cretaceous species, six gradually disappear during planktonic foraminiferal Subzones P1a and basal P1b. No species extinctions or major faunal assemblage changes are directly associated with the K/T boundary. Iridium distribution is ambiguous, with one peak in the upper part of the tsunami bed and a second peak at the micropaleontologically defined K/T boundary. Relative abundances of dominant species are stable through the Late Maastrichtian, and only minor abundance changes coincide with the first extinction episode or the K/T boundary. The first major faunal change in the dominant species group coincides with the second extinction episode and leads to decline and eventual extinction of this group in Subzone P1a. Species disappearing at the two extinction episodes [46% and 45%] constitute only a small percentage [8% and 5%] of the individuals of the total planktonic foraminiferal population. This suggests that weakened species with low numbers of individuals and sensitive to relatively minor environmental changes were primarily affected by these extinction episodes. Magnetostratigraphy indicates that the first extinction phase began about 310,000 years before the K/T boundary, and the second extinction phase occurred 50,000 years after the K/T boundary. This stepped pattern of species extinctions suggests a progressively stressed ecosystem in continental shelf settings which may be related to an observed sea level regression and global cooling. The hypothesis of a global catastrophic mass extinction at the K/T boundary caused by a large extraterrestrial impact is not supported by the Brazos River planktonic foraminiferal data.  PDF

The Cretaceous/Tertiary boundary section exposed near El Kef, Tunisia is the most complete boundary sequence known to date. It contains nearly 1 m of black and gray boundary clay containing planktic foraminiferal Zone POa, b, 2 m of clayey shales of P1a (Globigerina eugubina) Zone and 4 m of shales and marls of P1b (G. taurica) Zone. Quantitative analysis of the benthic foraminiferal fauna suggests that deposition during the latest Cretaceous occurred in an upper slope to outer shelf environment which shallowed at the K/T boundary to an outer to middle shelf depth and shallowed further by P1b time to mid-shelf depth. A major reduction in benthic diversity occurred near the K/T boundary with about 50% of the fauna disappearing. Diversity remained an average of 37% lower during deposition of the first 3 m of sediment above the boundary (POa, b-P1a) and productivity was very low. Surviving and thriving foraminifers during this interval were primarily low oxygen tolerant epifaunal and infaunal species. A sharp decrease in the low oxygen tolerant fauna and appearance of a shallow mid-shelf fauna at about 4 m above the boundary (P1b Zone) signals a second regression, return to higher oxygen levels and higher productivity. Although the environmental effects of the K/T boundary event can be inferred from benthic faunas, the ultimate cause remains elusive. Faunal changes prior to and the long recovery period after the K/T boundary are difficult to explain by a single impact hypothesis.   PDF

Recent discoveries of microtektite and related crystal bearing microspherule layers in deep-sea sediments of the west equatorial Pacific DSDP Sites 292, 315A and 462, off-shore New Jersey in Site 612 and in southern Spain have confirmed the presence of at least three microspherule layers in Late Eocene sediments. Moreover, these discoveries have extended the North American strewn field from the Caribbean and Gulf of Mexico region to the northwest Atlantic, and have established a third strewn field in western equatorial Pacific and Indian Ocean which may extend to the Mediterranean.   PDF

If at least some mass extinctions are caused by impacts, why do they extend over intervals of one to three million years and have a partly stepwise character? The solution may be provided by multiple cometary impacts. Astronomical, geological and palaeontological evidence is consistent with a causal connection between comet showers, clusters of impact events and stepwise mass extinctions, but it is too early to tell how pervasive this relationship may be.  PDF

Comparison s o f microfossils fro m deep-sea cores, fro m sample s o f an explorator y drill hole, and fro m dredged rocks o f the Gulf o f Alaska, with coeval microfossil assemblage s on the Nort h America n continent, provide constraints on the northwar d migratio n o f the Yakuta t block and the Princ e William terran e during Tertiary time. The estimated paleolatitudes of microfauna and flora indicate that: (1 ) the Prince William terran e wa s attache d to North Americ a in its present position by middl e Eocen e time (40 to 42 Ma) , consistent with models derived fro m paleomagneti c data , and (2) the adjacent Yakuta t block was 30 ± 5 ° south o f its present position in earl y Eocen e (50 Ma) , 20 ± 5 ° south in middl e Eocen e (40 to 44 Ma) , and 15 ± 5 ° south in lat e Eocen e tim e (37 to 40 Ma) , thus requiring a northwar d motion o f about 30 ° since 50 Ma . Moreover , the Yakuta t block wa s at least 10° south o f the Princ e William terran e during Eocen e time . Thes e dat a ar e consistent with migratio n o f the Yakuta t block with the Pacifi c and Kul a plate s for a t least the last 50 Ma . T h e collision o f the Yakuta t block with Nort h Americ a resulted in subduction o f the block coincident with uplift o f the Kena i Mountains. Th e extension o f the Kena i Mountains into the Kodiak are a suggests tha t a southwest extension o f the Yakuta t block collided with the Kodiak margi n and was completely subducted. Th e subducted extension o f the Yakuta t block could have connected the now subducting head o f Zodiak deep-sea fan to a Nort h America n sourc e o f sediment during deposition o f the fan.  PDF

The Orca Group is a widespread, very thick, and complexly deformed accretionary sequence of flysch and tholeiitic basalt in the Prince William Sound area of Alaska (Winkler, 1976; Winkler and Plafker, 1981). Despite a number of extensive field studies of the Orca Group, reliable data on the age of the unit has been exceedingly elusive. On the basis of sparse paleontologic and radiometric data, the sequence was assigned a Pal eocene and early Eocene(?) age (Addicott and Plafker, 1971; Plafker and Lanphere 1974; Winkler, 1976, Winkler and Plafker, 1981). New paleontologic data suggest that some strata assigned to the Orca Group are of middle Eocene age and possibly as young as late Eocene or Oligocene. However, data suggesting an age younger than about 50 Ma appear to be incompatible with radiometricanydetermined ages for plutons that intrude the Orca. This paper summarizes the published and unpublished paleontologic data from the Orca Group, considers their implications for the age of the unit, and points out the problems in age assignments. PDF

Chronostratigraphic and paleoclimatic comparisons of microfossils from deep-sea cores, from samples of an exploratory drill hole, and from dredged rock of the Gulf of Alaska with coeval microfossil assemblages on the North American continent provide constraints on the northward migration of the Yakutat block, the Prince William terrane and the Pacific plate during Tertiary time. The comparative paleolatitudes of microfauna and flora provide three main constraints. (1) The Prince William terrane was in its present position with respect to North America (at high latitudes, 50° ± 5°N) by middle Eocene time (40–42 Ma), consistent with models derived from paleomagnetic data. (2) The adjacent Yakutat block was 30° ± 5° south of its present position in early Eocene (50 Ma), 20° ± 5° south in middle Eocene (40–44 Ma), and 15° ± 5° south in late Eocene time (37–40 Ma), thus requiring a northward motion of about 30° since 50 Ma. Moreover, the Yakutat block was at least 10° south of the Prince William terrane during Eocene time. These data are consistent with migration of the Yakutat block with the Pacific and Kula plates for at least the last 50 Ma. (3) site 192 on the Pacific plate was at about 15° ± 5°N latitude in the late Cretaceous (68 Ma), at 30° ± 5°N in early Eocene (50 Ma), at 40° ± 5°N in middle Eocene (40–44 Ma), at 45° ± 5°N in late Eocene (37–40 Ma), and north of 50° ± 5°N in latest Eocene to early Oligocene time (34–37 Ma). These paleolatitudes, based on planktonic foraminiferal assemblages, indicate northward drift consistent with the North America-Pacific plate reconstructions from about 68 Ma to 40 Ma (Engebretson, 1982). However, from Cretaceous to early Eocené time, faunal data indicate significantly lower latitudinal positions, and from Oligocwne to early Miocene time, significantly higher latitudinal positions. These discrepancies can be explained by the northward expansion of tropical faunas during the globally warm early Tertiary and southward expansion of cold subarctic faunas as a result of global cooling during Oligocene time.   PDF

Analysis of middle Eocene to early, Oligocene calcareous and siliceous microfossils shows gradual biotic changes with no massive extinction event across the Eocene/Oligocene boundary. Biotic changes in the late Paleogene appear to reflect changing paleoclimatic and paleoceanographic conditions and do not support suggestions of a catastrophic biotic event caused by a bolide impact at the Eocenel Oligocene boundary.

Keller et al. recently suggested (1) that there are several middle Eocene to middle Oligocene microtektite horizons and implied that these horizons indicate separate tektite events. Although there is no a priori reason why there could not be multiple tektite events during this period, Keller et al. do not provide any descriptive, petrographic, or compositional data to support their identification of microtektites from previously unreported stratigraphic layers. Furthermore, the lack of data on abundance versus depth and of compositional data does not allow the reader to decide if the microtektite occurrences are due to several events, as Keller et al. claim, or merely to one event with scattered younger occurrences attributable to reworking.  PDF

Planktic foraminiferal assemblages have been analyzed quantitatively in six DSDP sites in the Atlantic (Site 363), Pacific (Sites 292, 77B, 277), and Indian Ocean (Sites 219, 253) in order to determine the nature of the faunal turnover during Middle Eocene to Oligocene time. Biostratigraphic ranges of taxa and abundance distributions of dominant species are presented and illustrate striking similarities in faunal assemblages of low latitude regions in the Atlantic, Pacific and Indian oceans. A high resolution biochronology, based on dominant faunal characteristics and 55 datum events, permits correlation between all three oceans with a high degree of precision. Population studies provide a view of the global impact of the paleoclimatic and paleoceanographic changes occurring during Middle Eocene to Oligocene time.

Planktic foraminiferal assemblage changes indicate a general cooling trend between Middle Eocene to Oligocene time, consistent with previously published oxygen isotope data. Major faunal changes, indicating cooling episodes, occur, however, at discrete intervals: in the Middle Eocene 44-43 Ma (P13), the Middle/Late Eocene boundary 41-40 Ma (P14P15), the Late Eocene 39-38 Ma (P15P16), the Eocene/Oligocene boundary 37-36 Ma (P18), and the Late Oligocene 31-29 Ma (P20P21). With the exception of the E0 boundary, faunal changes occur abruptly during short stratigraphic intervals, and are characterized by major species extinctions and first appearances. The Eocene/Oligocene boundary cooling is marked primarily by increasing abundances of cool water species. This suggests that the E0boundary cooling, which marks a major event in the oxygen isotope record affected planktic faunas less than during other cooling episodes. Planktic foraminiferal faunas indicate that the E0 boundary event is part of a continued cooling trend which began during the Middle Eocene.

Two hiatus intervals are recognized in low and high latitude sections at the Middle/Late Eocene boundary and in the Late Eocene (P15P16). These hiatuses suggest that vigorous bottom water circulation began developing in the Middle Eocene, consistent with the onset of the faunal cooling trend, and well before the development of the psychrosphere at the E0boundary.   PDF

Microtektites have been recovered from three horizons in eight middle Eocene to middle Oligocene marine sediment sequences. Five of these occurrences are coeval and of latest Eocene age (37.5 to 38.0 million years ago); three are coeval and of early late Eocene age (38.5 to 39.5 million years ago); and three are of middle Oligocene age (31 to 32 million years ago). In addition, rare probable microtektites have been found in sediments with ages of about 36.0 to 36.5 million years. The microtektite horizon at 37.5 to 38.0 million years can be correlated with the North American tektite-strewn field, which has a fission track age (minimum) of 34 to 35 million years and a paleomagnetic age of 37.5 to 38.0 million years. There is no evidence for mass faunal extinctions at any of the microtektite horizons. Many of the distinct faunal changes that occurred in the middle Eocene to middle Oligocene can be related to the formation of the Antarctic ice sheet and the associated cooling phenomena and intensification of bottom currents that led to large-scale dissolution of calcium carbonate and erosion, which created areally extensive hiatuses in the deep-sea sediment records. The occurrence of microtektite horizons of several ages and the lack of evidence for faunal extinctions suggest that the effects of extraterrestrial bolide impacts may be unimportant in the biologic realm during middle Eocene to middle Oligocene time.