Introduction
The age of the Jurassic–Cretaceous (J–K) boundary remains one of the last
standing Phanerozoic system boundaries with a numerical age not tied by
adequate radioisotopic data. The numerical division of the geological
record is ultimately dependent on accurate and precise radioisotopic ages
of well-defined fossiliferous datums. Over the years the numerical age of
the J–K boundary has been difficult to measure due to the lack of datable
horizons close to boundary markers, which made it difficult to ascribe a
radioisotopic age directly to fossiliferous datums. Consequently, the
ill-defined age of the boundary has led to widely variable timescales for
the Late Jurassic to Early Cretaceous
(Channell
et al., 1995; Gradstein et al., 1995; Lowrie and Ogg, 1985; Malinverno et
al., 2012; Ogg, 2012; Ogg et al., 1991; Ogg and Lowrie, 1986; Pálfy,
2008; Pálfy et al., 2000). These various approaches attempted to
ascribe an age to the J–K boundary; nevertheless, the different estimates
for the age of the boundary lacked reproducibility, varying from 135 to 145 Ma, with a high degree of uncertainty and very little overlap. The most
recently used timescale of the Late Jurassic to Early Cretaceous is the
M-sequence model of Ogg (2012). The model relies
on the integration of data from a variety of fields such as M-sequence
magnetic anomalies from the northwest Pacific Ocean, magnetostratigraphy,
biostratigraphy, cyclostratigraphy, and scarce radioisotopic ages. The
model is based on the marine magnetic anomalies timescale of the
northwestern Pacific Ocean
(Channell
et al., 1995; Larson and Hilde, 1975; Tamaki and Larson, 1988). The interval
encompasses ∼1000 km of oceanic crust over a period of
∼35 Myr in the northwestern Pacific. The age of the polarity
changes in the northwestern Pacific was dated by key fossiliferous datums
from Mediterranean Tethys sedimentary sequences via the correlation with
magnetostratigraphy in these sequences
(Grabowski, 2011; Kent and
Gradstein, 1985; Ogg et al., 1991; Ogg and Lowrie, 1986). The duration of
the magnetic reversal changes are provided by cyclostratigraphic studies
(Huang et
al., 2010a, b) for some of the magneto-zone intervals and thus used to
calculate a decreasing spreading rate with the distance associated with the
magneto-anomalies in the Hawaiian spreading center. The numerical age of
stage boundaries from the Berriasian to Oxfordian were then back-calculated
from the age of the M0r at the base of the
Aptian. The age of the M0r used
was 126.3±0.4 Ma, which is the combination of the cycle duration of
the Albian stage (Huang et al.,
2010a) tied to a U–Pb age from the Aptian–Albian boundary of 113.1±0.3 Ma (Selby et
al., 2009). This linear fitting model is the basis for various Late Jurassic
and Early Cretaceous stage boundary numerical ages.
In the specific case of the J–K boundary the projected age of the M-sequence
age model was 145.0±0.8 Ma (Ogg, 2012),
which is almost identical to the radioisotopic age reported in Mahoney et al. (2005) of 145.5±0.8 Ma (recalculated by Gradstein, 2012) for the
sill intruded in Berriasian sediments in the Shatsky Rise with magnetization
M21-M20. Furthermore, the magnetization of borehole 1213B is reasonably
close to what has become a reliable secondary marker for the J–K boundary,
the M19.2n (Wimbledon, 2017, and references therein).
However, studies that obtained radioisotopic ages directly from sedimentary
sequences that spanned the J–K boundary reveal much younger ages for the
boundary
(Bralower
et al., 1990; López-Martínez et al., 2015; Vennari et al., 2014).
Recently, the base of the Calpionella Zone (the Alpina Subzone) has been
selected as a principal biostratigraphic marker for the base of the
Berriasian (Wimbledon, 2017). Nevertheless, its presence
alone is not sufficient to locate the boundary, and secondary markers such
as calcareous nannofossils and magnetostratigraphy are essential additional
constraints to aid the definition of the boundary, with the latter allowing
sections to be normalized against a global framework. The most complete
studies of the J–K boundary from a biostratigraphical and
magnetostratigraphic standpoint are located in Mediterranean Tethys.
Nevertheless, the radiometric age of the boundary is poorly defined in the
Mediterranean Tethys due to the absence of active volcanism close by during
the time of deposition of these sedimentary sequences. In this way, the
western Tethys (proto-Gulf of Mexico) and the Austral Basins (Neuquén Basin,
Argentina) offer a good opportunity to advance the study of the
radioisotopic age of the J–K boundary. Contrary to the Mediterranean
Tethys, the sedimentary sequences in the proto-gulf and Austral realms were
deposited close to active plate boundaries where significant volcanism took
place, which enabled the deposition of datable horizons suitable for U–Pb
geochronology. Recently, calpionellid biostratigraphy has been reported in
both regions
(López-Martínez et al.,
2013b, 2017), opening possibilities for better correlations with the
Mediterranean Tethys. It is worth noting that even if the calpionellid
biostratigraphy of the Neuquén Basin is still not complete and global
correlations are still tentative, for now they are the only known basins
with occurrences of calpionellid as markers around the J–K boundary in the
Austral realm with abundant datable horizons. A general definition of the
J–K boundary would, however, need to be of global validity and allow
for correlation with the Tethys realm.
In the present study, we date two independent sections, one in Mexico and
one in Argentina, using precise radioisotopic geochronological methods. We
present high-precision U–Pb age determinations using chemical abrasion–isotope dilution–thermal
ionization mass spectrometry (CA-ID-TIMS)
techniques to date zircon from interbedded volcanic ash layers in the Las
Loicas section, Neuquén Basin, Argentina, and the Mazatepec section,
Mexico. Such dates have proved to yield robust estimates for the timing of
the stratigraphic record, especially in combination with Bayesian age–depth
modeling (e.g.,
Ovtcharova et al., 2015;
Baresel et al., 2017; Wotzlaw et al., 2017). The coupling of high-precision
U–Pb geochronology and age–depth modeling allowed us to ascribe specific
numerical ages to key taxa in the early Berriasian and late Tithonian in the
studied sections. We also report new nannofossil data from the section in
Mexico such as the first occurrence (FO) of Nannoconus steinmannii steinmannii
and the FO of Nannoconus kamptneri minor (Fig. 2). Additionally,
we also present the first radioisotopic age in the early Tithonian at the
base of the Virgatosphinctes andesensis biozone in the La Yesera section, Neuquén Basin, close to
the Kimmeridgian–Tithonian boundary (KmTB)
(Riccardi, 2008, 2015; Vennari, 2016).
Lastly, our geochronological data allow us to reevaluate the numerical age
of the J–K boundary and discuss some complications with the currently
accepted age of ∼145 Ma.
Location of the studied sections and the general geological context of
each section.
Geological context and studied sections
To investigate the numerical age of the J–K boundary, we have selected two
sections where J–K boundary markers such as ammonites, calpionellids, and
calcareous nannoplankton have been recognized. The first section is Las
Loicas, exposed along the national road 145 (Argentina) from Bardas Blancas
to the international border at the Pehuenche Pass. It is located near the
Argentine–Chilean border, approximately 1 km to the southwest of
the settlement Las Loicas (Fig. 1). Geologically, the Las Loicas section
(Vennari et al., 2014)
is located in the Vaca Muerta Formation, Neuquén Basin, Argentina (Fig. 1). The Neuquén Basin in western Argentina accumulated an almost
continuous record of 7000 m of sediments from the Late Triassic to early
Cenozoic. The basin is located on the eastern side of the Andes in Argentina
between 32 and 40∘ S latitude (Fig. 1). The basin has a
triangular shape, covers an area of over 12002 km, and is bounded to
the west by the Andean magmatic arc on the active margin of the South
American Plate, to the northeast by the San Rafael Block, and to the
southeast by the North Patagonia Massif (Fig. 1). Two main regions
are commonly recognized in the basin: the Neuquén Andes to the west and
the Neuquén Embayment to the east (Fig. 1). The Neuquén Embayment is
relatively undeformed, in contrast to the Neuquén Andes where the late
Cretaceous–Cenozoic deformation has resulted in the development of a series
of N–S-oriented fold and thrust belts: Aconcagua, Malargüe, and Agrio,
where a substantial part of the Mesozoic sequence outcrops (Legarreta and
Uliana, 1991, 1996).
The Vaca Muerta Fm. is a 217 m thick sedimentary sequence of marine shales
and limestones, which spans an interval from the lower Tithonian
(Virgatosphinctes andesensis biozone) to the upper Berriasian (Spiticeras damesi biozone)
(Aguirre-Urreta
et al., 2005; Kietzmann et al., 2016; Riccardi, 2008, 2015). In Las Loicas,
the Substeueroceras koeneni and Argentiniceras noduliferum ammonite biozone and calcareous nannofossils have been described by
Vennari et al. (2014). Recently, López-Martínez et al. (2017) reported the occurrence of upper Tithonian to lower Berriasian
calpionellids, which is the only known section where the primary markers for
the J–K boundary occur together in the Argentinian Andes. The section contains
several ash beds, which allowed for precise age bracketing of the boundary using
high-precision U–Pb geochronology.
The La Yesera Section is located 50 km north of the town of Chos Malal in
the northern sector the Neuquén Basin (Fig. 1) and is exposed along the
national road 40. Geologically, the La Yesera section (Fig. 2c) represents a
distal portion of the basin farther from the magmatic arc than the Las
Loicas section. Tuff beds are less frequent than in the Las Loicas section
and generally thinner. The section has a total thickness over 400 m and is
one of the best continuous exposures of Tithonian ammonite zones, from the
early Tithonian Virgatosphinctes mendozanus to the Neocomites wichmanni (early Valanginian; Aguirre-Urreta et al., 2014). The section also has
one of the best-exposed contacts between the Vaca Muerta Fm. and the Tordillo
Fm.
Age correlation between the Las Loicas, Mazatepec, and La Yesera
sections. (a) Las Loicas section: ash beds in light blue with respective name
and U–Pb dates in black font; age–depth modeling ages are in red font next
to green stars (this study); ammonite and nannofossil zonation after Vennari et al. (2014);
calpionellid zonation after Lopez-Martinez et al. (2017).
(b) Mazatepec section: ash bed in light blue with respective name and U–Pb age
in black font, age calculated from sedimentation rate in red font (this study);
calcareous nannofossils (this study); calpionellid zonation after Lopez-Martinez
et al. (2013b). (c) La Yesera section: ash bed in light blue with U–Pb age
(Aguirre-Urreta et al., 2014).
The Mazatepec section is located in the Puebla State, Mexico, southeast of
Mexico City. Geologically, the Mazatepec section exposes the Pimienta and
the lower Tamaulipas formations of the Sierra Madre Oriental geological
province, Mexico (Fig. 1). The Sierra Madre Oriental is one of the many
tectonic terranes composed of Mesozoic volcano-sedimentary sequences
deformed during the Late Cretaceous and early Cenozoic during the Laramide
Orogeny in Mexico (Campa and Coney, 1983; Suter,
1980). A rift sequence characterizes the tectonic evolution of the
proto-gulf in the Late Triassic–Oxfordian due to the rifting of Pangea
characterized by continental sedimentation controlled by narrow grabens with
no marine sedimentation taking place (Salvador,
1987). The post-rift phase is characterized by ample marine carbonate
platforms of shallow waters. During the Tithonian to Early Cretaceous,
stable tectonic and climatic conditions prevailed with the sedimentation
being significantly slower with the development of shallow marine water
sedimentation, namely the deposition of the Pimienta Fm. (carbonates) and
Tamaulipas Fm. (argillaceous limestones, shales)
(Padilla and Sánchez, 2007). The Pimienta Fm. is
composed of darkish clayey limestones and the Tamaulipas Fm. is a gray
limestone (López-Martínez et
al., 2013b; Suter, 1980). The section has a dense occurrence of late Tithonian
Crassicollaria Zone (Colomi Subzone) and early Berriasian calpionellids from
the
Calpionella Zone (Alpina, Ferasini, and Elliptica subzones) to the
Calpionellopsis Zone (Oblonga Subzone). In the upper part of the section,
ash beds are scarce and occur at distinct levels. Ash bed MZT-81 is situated
within the Elliptica Subzone in the lower Tamaulipas Formation (Fig. 2b).
(a–h) Representative calcareous nannofossils from the Mazatepec
section, Mexico. (a) Conusphaera mexicana Trejo (BAFC-NP 4190) (2 m),
(b) Conusphaera mexicana Trejo
(BAFC-NP 4196) (11 m), (c) Hexalithus noeliae Loeblich and Tappan (BAFC-NP 4195)
(7.5 m), (d) Hexalithus geometricus
Casellato (BAFC-NP
4205) (25 m), (e) Nannoconus kamptneri minor Bralower (BAFC-NP 4201) (16 m),
(f) Nannoconus globulus Brönnimann
(BAFC-NP
4205) (25 m), (g–h) Nannoconus steinmannii subsp. steinmannii Kamptner
(BAFC-NP 4205) (25 m). Our suggestion is
to eliminate calcareous nannofossil images published previously from the Las
Loicas section in order to avoid more confusion with taxonomy.
Results
Calcareous nannofossil biostratigraphy in Mazatepec
Eighteen nannofossil species have been recognized in Mazatepec (Fig. S1).
The heterococcoliths are mostly represented by Watznaueriaceae including
Watznaueria barnesae, W. britannica, W. manivitae, Cyclagelosphaera margerelii,
and C. deflandrei; Zeugrhabdotus embergeri
is another frequent constituent. The nannoliths are represented by
Conusphaera mexicana, Polycostella senaria, Hexalithus noeliae, Nannoconus globulus, and N. kamptneri minor. These
nannofossils indicate a late Tithonian to early Berriasian age for
the Pimienta Formation and the lower part of the Tamaulipas Formation. The
assemblage composed of Conusphaera mexicana, Polycostella scenario, and Hexalithus noeliae indicates a late Tithonian age. The only useful
biological event recognized is the FO of N. kamptneri minor. An increase in the diversity of
nannofossils is identified with 11 species, among which the presence of N. steinmannii steinmannii
stands out (Fig. 2b).
U–Pb geochronology, age interpretations, age–depth modeling
A total of six ash beds were dated: four in the Las Loicas section, one in
the Mazatepec section, and one in the La Yesera section. In the Las Loicas
section, LL3 yielded an age of 139.238±0.049/0.061/0.16, LL9
139.956±0.063/0.072/0.17, LL10 140.338±0.083/0/091/0.18, and LL13 an age of 142.039±0.058/0.069/0.17 Ma. In La Yesera,
ash bed LY5 yielded an age of 147.112±0.078/0/088/0.18 Ma, and in
Mazatepec MZT-81 yielded an age of 140.512±0.031/0/048/0.16 Ma (Fig. 4). All zircons considered in the age distribution of the ash are
interpreted from ashfall deposits from nearby volcanic eruptions. The
final weighted mean ages are interpreted as a depositional age for each ash
bed. Uncertainties are reported as X/Y/Z where X includes analytical
uncertainty, Y includes additional tracer (ET2535) calibration uncertainty,
and Z includes additional 238U decay constant uncertainty. A full and
detailed description of the techniques, sample preparation, laboratory
procedures, data acquisition, and data treatment is provided in the
Supplement. The full U–Pb data set is reported in Table S1 in the Supplement.
Age–depth statistical modeling was performed by outputting a numerical age for
every meter of the Las Loicas sections, with a 95 % confidence precision
interval. The results with a meter-by-meter resolution are reported in Table TS.2.
Numerical age of faunal assemblages in studied sections
In Fig. 2a, the various markers and assemblages are indicated, as are
the ages of the ash beds. In Las Loicas,
López-Martínez
et al. (2017) reported on the late Tithonian Crassicollaria Zone and the Colomi Subzone
(upper Tithonian) based on the occurrence of Calpionella alpina Lorenz, Crassicollaria colomi Doben, Crassicollaria parvula Remane,
Crassicollaria massutiniana (Colom), Crassicollaria brevis Remane, Tintinnopsella remanei (Borza)
and Tintinnopsella carpathica (Murgeanu and Filipescu), and the
FO of the Umbria granulosa granulosa and Substeueroceras koeneni ammonite zone
(Vennari et al., 2014).
Our Bchron model age predicts an age of 141.31±0.56 Ma for the
faunal assemblage of Crassicollaria parvula and Crassicollaria colomi and the FO of Umbria granulosa granulosa Fig. 2b).
Another late Tithonian
marker in Las Loicas is the FO of Rhagodiscus asper, also within the Crassicollaria Zone, with a
Bchron age of 140.60± Ma (Fig. 2a).
In Las Loicas some early Berriasian markers are present. For instance, the
FOs of Nannoconus kamptneri minor (Figs. 2a, S1) and Nannoconus steinmannii minor are considered indicators of the early
Berriasian (Vennari et
al., 2014). Here they overlap with the base of the Argentiniceras noduliferum ammonite zone
(López-Martínez
et al., 2017; Vennari et al., 2014). The occurrence of the acme of Calpionella alpina (small and
spherical) and scarce specimens of Crassicollaria massutiniana, Tintinnopsella remanei,
and T. carpathica suggest an early Berriasian age
(López-Martínez et al., 2017) (Fig. 2a). These assemblages are
bracketed by ash beds LL9 (139.956±0.063 Ma) and LL10 (140.338±0.083 Ma) (Fig. 2a) and overlap with the FO of Nannoconus kampteri minor and
Nannoconus steinmannii minor, the base of
the Argentiniceras noduliferum zone, and the base of the Alpina Subzone (ca. 34 m of stratigraphic height)
(Fig. 2a). The Bchron model age for this assemblage is 140.22±0.13 Ma
(Fig. 2a).
In Mazatepec, ash bed MZT-81 is located within the Elliptica Subzone and has
an age of 140.512±0.031 Ma. (Fig. 4). Due to the lack of datable
horizons close to the Alpina Subzone in Mazatepec we have resorted to
assumed sedimentation rates to back-calculate the age of the base of the Alpina
Subzone. Here we assume the sedimentation rate to be 2.5 cm ka-1.
Although there
are no data on actual sedimentation rates in the Pimienta Fm., this rate is
realistic for similar coeval deposits (e.g., Grabowski et al., 2011) as well
as with the tectonic and environmental stability of the Sierra
Madre Oriental in the Tithonian–Berriasian stages (Padilla and
Sánchez, 2007). It is worth noting that our new data allow only a
confident numerical age for the Elliptica Subzone (Fig. 5).
Ash bed LY-5 was located below the contact, and it yielded an age of 147.112±0.078 Ma (Fig. 2C). The ash bed is located in the Tordillo Fm., 1.5 m
below the contact with the Vaca Muerta Formation and thus very close to the
base of the Virgatosphinctes andesensis zone.
Discussion
The chronostratigraphic and biostratigraphic framework of the studied
sections
In the past decade significant strides have been made in fixing the J–K
boundary by coupling calpionellids, calcareous nannofossils, ammonites, and
magnetostratigraphy (Wimbledon, 2017;
Wimbledon et al., 2011). Correlations between sections within the
Mediterranean Tethys have become consistent to the point of a trustworthy
correlation framework being developed for the various markers
(calpionellids, nannofossils, ammonites, and magnetostratigraphy) of the
J–K boundary (Wimbledon, 2017, references therein). Even
though important biostratigraphic studies have been carried out in other
regions outside of the Mediterranean Tethys, such as in the proto-gulf
and the Argentinian Andes, the correlation between these regions remains
uncertain. Notably, the lack of magnetostratigraphic data in studies from
the proto-gulf
(López-Martínez et al.,
2013a, b) and the Argentinian Andes
(López-Martínez
et al., 2017; Vennari et al., 2014) is a challenge and leaves room for
ambiguity in biochronostratigraphical correlations. Here we attempt to
describe the limitations of the biostratigraphical markers in the studied
sections.
Tentative correlation of the studied sections with the western Tethys correlation scheme of Wimbledon et al. (2017).
In Mazatepec, only two important calcareous nannofossil bioevents are
recognized, i.e., the FO of N. kamptneri minor and N. steinmannii steinmannii. In the Tethys realm, the former bioevent
occurs within the M19.2n, slightly above the base of the Alpina Subzone
(Bakhmutov et al., 2018), and it is
used as an upper limit to the base of the Alpina Subzone
(Wimbledon et al., 2013). In Mazatepec, the FO
of N. kamptneri minor occurs 5 m above the base of the Alpina Subzone; however, it is within the
lower Ferasini Subzone and thus slightly younger than in the Mediterranean
Tethys. Another bioevent in Mazatepec is the FO of N. steinmannii steinmannii, which occurs within
the Elliptica Subzone. This marker has been shown in the past to occur
within the Elliptica Subzone and coincident within the M17r
(Casellato, 2010), but has been found as low as the Alpina
Subzone, the base of M18r
(Bakhmutov
et al., 2018; Hoedemaeker et al., 2016; Lukeneder et al., 2010), or even
lower (Svobodová and
Košťák, 2016). Even though our new calcareous nannofossils from
Mazatepec are an addition to the biostratigraphic framework of the sections,
this is very preliminary and does not provide any definite constraints for the
J–K boundary or the base to the Alpina Subzone in the section. Valuable
markers such as N. steinmannii minor, N. wintereri, and H. strictus have not yet been reported. Furthermore, no calcareous
nannofossils have been reported below the base of the Alpina Subzone in
Mazatepec. Nevertheless, we feel that the FO of N. kamptneri minor so close to the base of the
Alpina Subzone in Mazatepec provides confidence for futures studies in the
section.
In the Mediterranean Tethys, important markers for the J–K boundary are the
first appearance datum (FAD) of N. kamptneri minor and N. wintereri. In the Tethys, these two markers usually occur in the middle
of the M19.2n, but in distinct stratigraphic horizons and commonly
bracketing the base of the Alpina Subzone
(Wimbledon, 2017; Wimbledon et al.,
2013). N. wintereri, for instance, occurs below the base of the Alpina Subzone
(Elbra
et al., 2018; Svobodová and Košťák, 2016; Wimbledon et al.,
2013) and in one occurrence as low as the M19r
(Lukeneder et al., 2010). In Las
Loicas, on the other hand, both occur virtually within the same
stratigraphic range
(Vennari et al., 2014).
The close FO of N. kamptneri minor, N. wintereri, C. deflandrei, and M. pemmatoide in Las Loicas (Vennari et al., 2014) is also
troublesome.
The most important secondary marker for the J–K boundary is the FAD of N. steinmannii
minor, which usually occurs in the vicinity of the Alpina Subzone
(Wimbledon, 2017), below it
(Bakhmutov et al., 2018), and
slightly above it
(Hoedemaeker
et al., 2016; Svobodová and Košťák, 2016). In Las Loicas,
the FO of N. steinmannii minor is present and occurs in the vicinity of the Alpina
Subzone;
however, it is limited to a single sample
(Vennari et al., 2014)
and not continuous. Furthermore, in Las Loicas the FOs of N. kamptneri minor and N. wintereri are
recorded below the FO of N. steinmannii minor. This order of occurrence in Las Loicas is
contradictory because the FO of N. steinmannii minor is considered older than the FO
of N. kamptneri minor and younger
than the FO of N. wintereri. These circumstances suggest that condensation and/or preservation
issues might be affecting the completeness and continuity of the calcareous
nannofossil biostratigraphy in Las Loicas, thus impeding a reliable
correlation between the Argentinian Andes and the Tethys.
Another possible issue with the biostratigraphy in Las Loicas pertains to a
couple of calpionellid assemblages that might seem unusual when compared to
the Mediterranean Tethys. The first is the presence of Tintinnopsella remanei in the upper part of the
Crassicollaria Zone. This is a nontypical appearance in the Mediterranean
Tethys, but it is usual in the western Tethys as discussed in
López-Martínez et al. (2017). The second is the
record of Crassicollaria massutiniana in the lowermost part of the Alpina Subzone. Even when it can be
unusual, the presence of this species in the lowermost Berriasian does not
affect the biozonation scheme as the Alpina Subzone is defined by the acme
of Calpionella alpina in small and globular form and not the last occurrence (LO) of any species. Therefore, the
Alpina Subzone is defined in Las Loicas in the same way as in the
Mediterranean Tethys and can be used as a reasonable marker for the base of
the Berriasian in Las Loicas.
In conclusion, there is still ambiguity in the biostratigraphic framework of
the studied sections with regards to the J–K boundary markers. The
incompleteness and frequency of key taxa call for further investigation and
improvements to the biostratigraphy; important elements are still
lacking for a definite and precise definition of the J–K boundary in both
sections and correlations are still troublesome.
Constraining the numerical age of the J–K boundary between the studied
sections
In Mazatepec, the middle of the Elliptica Subzone has an age of
140.512±0.031 Ma and consequently a numerical age in the lower
Berriasian (Figs. 2 and 4). Conversely, in Las Loicas, the Bchron age model
predicts that approximately the same age, i.e., 140.54±0.37 Ma (ca.
28.5 m; see TS.2), is found in the Crassicollaria Zone, 1 m above the
FO of R. asper and thus late Tithonian (Fig. 2a). In other words, the age of
∼140.5 Ma in one section is coincident with late Tithonian
fauna, and in the other it yields an age coincident with early Berriasian
fauna. We see no reason to question the accuracy of the radioisotopic
dates. It thus becomes apparent that both sections are offset in age, and
Mazatepec is older than Las Loicas. Therefore, our geochronology points to
limitations in the biochronostratigraphical correlation of these two sections.
Given the limitations of the biostratigraphy around the J–K boundary in both
sections, our ability to quote a single numerical age for the J–K boundary is
strongly hindered. Nevertheless, we feel that constraining, bracketing,
and/or creating an age confidence interval for the J–K boundary using the
biostratigraphical and geochronological constraints from both sections is a
reasonable alternative to circumventing these limitations. To constrain the
interval, we have tentatively chosen upper and lower limits to the interval
based on the available biostratigraphic markers and their estimated ages
that best bracket the J–K boundary. In Mazatepec, we suggest the FO of N. kamptneri as
the upper biostratigraphical marker for the J–K age interval. In this
section, the FO of N. kamptneri is close to the base of the Ferasini Subzone, and thus a
subzone normally associated with the upper Alpina Subzone
(Wimbledon, 2017, and references therein), the base of the
18r (Casellato, 2010), and M19n
(Wimbledon et al., 2013), although it has recently
been shown to be found at the base of the M19.2n
(Bakhmutov et al., 2018). We feel
that this could be used as a very conservative upper limit for the age of the
J–K boundary. Using the sedimentation rate of 2.5 cm ka-1 in Mazatepec, we
estimate the age of the FO of N. kamptneri and conceivably the base of the Ferasini
Subzone to be ∼140.7 Ma (Fig. 2b). This is a conservative
estimate for the upper age of the J–K boundary in Mazatepec and could very
likely be older since the FO of N. kamptneri is commonly older than the base of the
Ferasini Subzone (Wimbledon, 2017, and references
therein). The base of the Alpina Subzone in Mazatepec is estimated to be
∼140.9 Ma, although a bracketing of the Alpina Subzone was
not possible due to the absence of calcareous nannofossils commonly
occurring at the base of the Alpina Subzone, such as N. steinmannii minor, or older diagnostic
markers such as R. asper, N. erbae, and N. globulus. Therefore, a lower limit to the boundary in
Mazatepec cannot be delineated.
Conversely, in Las Loicas, a few late Tithonian calcareous nannofossils
occur in an assemblage with late Tithonian calpionellids such as the FO of R. asper, which is
within the upper Crassicollaria Zone, and close to the FO of U. granulosa
(Bralower et al., 1989; Casellato, 2010).
These markers in Las Loicas allow for a lower age limit for the J–K
boundary. Given these circumstances we suggest 1 m above the FO of R. asper as
the lower limit of the J–K interval in Las Loicas. The Bchron model provides
an age for the FO of R. asper at 140.60±0.4 Ma (ca. ∼27 m; see
TS.2), which allows for a small overlap between the estimated age of the base of
the Alpina Subzone in Mazatepec and the late Tithonian and early Berriasian
assemblages in Las Loicas.
In summary, we have attempted to constrain the age of the J–K boundary using
the biostratigraphical markers and our geochronology from Las Loicas and
Mazatepec. Ash bed MZT-81 (middle of the Eliptica Subzone) suggests a minimum
age. As a result, the age of the J–K boundary has to be older than
140.512±0.031 Ma, most likely older than ∼140.7 Ma
(FO of N. kamptneri and/or the base of the Ferasini Subzone in Fig. 5; base of the M18r, possibly within
M19.2n), but the latter age estimate derives from an approximate
sedimentation rate (2.5 cm ka-1) that carries some uncertainty. In Las
Loicas, the Bchron model age of the FO of R. asper (possibly the middle of the M19r) suggests a
maximum age of the J–K boundary at 140.60±0.4 Ma. Given
that the age of the Alpina Subzone in Mazatepec is estimated at
∼140.9 Ma, we suggest that the age of the J–K boundary be
bracketed between 140.7 and 141.0 Ma. This interval accounts for the age of
the boundary being slightly older than the base of the Alpina Subzone in
Mazatepec due to the lack of secondary markers below the subzone. Our
attempt to constrain the age of the J–K boundary is based only on the
diagnostic markers for the boundary reported in the studied sections and the
additional fact that we can calculate and estimate their ages, even if the chosen
upper and lower limit of the interval has been proven to lie distant to the
J–K boundary. Given the inherited uncertainties of the biostratigraphy and
geochronology, we consider this age bracket as our best estimate for the J–K
boundary interval.
The early Tithonian and the base of the Vaca Muerta Formation
The base of the Vaca Muerta Formation contains an early Tithonian ammonite
assemblage of the Virgatosphinctes andesensis zone (Riccardi, 2008,
2015; Vennari, 2016). The gradational contact between the Vaca Muerta and
the Tordillo formations is very well exposed in the La Yesera section and
contains ash beds very close to the contact (Fig. S2b). We dated ash bed
LY-5, and it yielded an age of 147.112±0.078 Ma (Fig. 2c). The ash
bed is located in the Tordillo Fm., 1.5 m below the contact with the Vaca
Muerta Formation and thus very close to the base of the Virgatosphinctes andesensis zone. This biozone is
mostly equivalent to the Darwini Zone of the Tethys Ocean, which is broadly
regarded as early Tithonian and widely distributed in various other regions
including Mexico and Tibet (Riccardi,
2008, 2015; see Vennari, 2016, for a thorough review of the subject).
Consequently, we suggest that the age of ash bed LY-5 (147.112±0.078 Ma)
can be regarded as an age in the early Tithonian. This result is in good
agreement with other studies that have dated the early Tithonian. For
instance, Malinverno et al. (2012) quote
an age of 147.95±1.95 Ma for the M22An magneto-zone, and
Muttoni et al. (2018) suggest that
the base of the Tethyan Tithonian (top Kimmeridgian) falls in the lower part
of M22n with an age of ∼146.5 Ma.
Assuming the age of ash bed LY-5 (147.112±0.078 Ma) in La Yesera
to be early Tithonian and coupling it with the estimated
bracketed age of the J–K boundary (140.7–141. Ma), we can calculate a
minimum duration for the Tithonian of ∼6–7 Myr (Fig. 2c). This
is in good agreement with the current full duration of the Tithonian
estimated at ∼7 Myr (145.5 to 152.1 Ma;
see Ogg et al., 2016b). Furthermore, the M-sequence geomagnetic polarity
timescale (MHTC12) of Malinverno et al. (2012) suggests a duration for the
Tithonian of 5.75±2.47 Myr (i.e., between magneto-zones M22An and
M19n.2n). Therefore, our new ages around the base of the Berriasian and close to
the earliest Tithonian are in good agreement with other independent timescale
estimates for the duration of the Tithonian. Incidentally, this result also
has direct implications for the age of the KmTB. Currently, the age of the
KmTB is 152.1±0.9 Ma according to the International Commission on
Stratigraphy (ICS) (see also Ogg et al., 2016b).
Admittedly, ash bed LY-5 is not at the KmTB, although it is close; therefore, we
acknowledge that the age of the KmTB would have to be older than bed LY-5.
Nevertheless, if the age of the KmTB is 152.1 Ma, it would imply that the
Virgatosphinctes ammonite zone itself lasts more than ∼5 Myr, resulting in a
total duration of ∼12 Myr for the Tithonian. It appears
reasonable that our results for the early Tithonian are in agreement with
other studies that dated the KmTB, and this also suggests that the current ICS
KmTB age may need revision.
Implications for the numerical age of the J–K boundary
As of now, the age of the J–K according to the ICS is ∼145 Ma, which is ∼4 Myr older than our ages around the J–K
boundary (Figs. 4 and 5). As we have explored in previous sections, the
level of detail of the biostratigraphy in the studied sections needs
improvement and fails to provide a precise constraint for the J–K boundary.
A significant offset of potentially ∼600 kyr outlines the
limitations of correlating biostratigraphy and geochronology between the two
sections. Nevertheless, the disparity between our ages presented here and
the current age of the J–K boundary is such that even with the
biostratigraphical limitations and the absence of magnetostratigraphy, it calls
for further attention to the numerical age of the J–K boundary. For
instance, in Las Loicas the assemblage of Crassicollaria parvula and
Crassicollaria colomi and the FO of
Umbria granulosa granulosa
have an age of 141.31±0.56 Ma (Fig. 2a), with the FO of R. asper at 140.60±0.4 Ma, which can be considered to lie within the late Tithonian and thus constrain
the approximate age of the boundary. Furthermore, our age in the Elliptica
Subzone in Mazatepec is at 140.512±0.031 Ma (early Berriasian).
Worthy of attention is the age of ash bed LY5 in the Virgatosphinctes andesensis biozone (early
Tithonian) at 147.112±0.078 Ma. These geochronological constraints
make it fairly difficult to reconcile the base of the Berriasian to be
∼145 Ma and also has important implications for the duration
of the Tithonian (see discussion above on the early Tithonian). From our new
geochronological data, ∼145 Ma would most likely be an age in
the middle of the Tithonian rather than the base of the Berriasian (Fig. 4).
Other recent geochronological studies on the age of the J–K boundary using
different geochronological approaches (e.g., Re–Os isochron ages from
shales or LA-ICP-MS U–Pb ages on zircons) and in the Early Cretaceous are
also at odds with the current age of the boundary.
López-Martínez
et al. (2015, 2017), Pálfy et al. (2000), and Tripathy et al. (2018)
have published geochronological results that overlap within uncertainty with
our age estimate of the J–K boundary (around 140–141 Ma). In summary, there
is growing evidence that the age of the J–K boundary is most likely younger,
although unequivocal evidence is still lacking.
The endurance of the numerical age of the boundary is mainly due to the
perfect overlap between the M-sequence age model of
Ogg (2012) and
Mahoney et al. (2005). The latter authors dated a
basaltic intrusion in lower Cretaceous (NK1) sedimentary rocks and argued
that the age of the basalt would be close to the age of the J–K. Their age
for the intruded basalt is 144.2±2.6 Ma (40Ar/39Ar). This
age was later corrected by Gradstein (2012)
and Ogg et al. (2012) to 145.5±0.8 Ma with the recalibrated
40K
decay constant of
Renne et al. (2010). The magnetization of drill core 1213B proved to be between
anomalies M19 and M20 (Sager, 2005), which was consistent at
that time with the working model for the base of the Berriasian placed
between M19 and M18 (now more precisely calibrated in the middle of the
M19.2n; Wimbledon, 2017). This overlap was also in
agreement with the numerical timescale of Gradstein et al. (1995). These facts have mainly been the anchors for the numerical age of the
J–K boundary in the past years. However, analytical and biostratigraphical
issues potentially reveal some inconsistencies in the numerical age for the
boundary in Mahoney et al. (2005). For instance,
the biostratigraphy of drill core 1213B poses problems.
Bown (2005) pointed out that the
sediments of this core are devoid of age-diagnostic NK1 nannofossils such as
Conusphaera and Nannoconus. Important markers such as the family Cretarhabdaceae are present but
in rare occurrences. Drill core 1213B is limited to the occurrence of
nannofossils considered secondary markers and lacked any primary markers for
the boundary. Even with the existing problems in the biostratigraphy of
drill core 1213B, the magnetization of the dated basalt is in reasonable
agreement with the magnetic timescale for the base of the Berriasian
(Wimbledon, 2017). More importantly, it is worth pointing out that
Mahoney et al. (2005) report the dated basalts to
be slightly altered, which could have consequences for the accuracy and
precision of their age.
The accuracy of the M-sequence age model of Ogg (2012) is ultimately dependent on the quality of available radioisotopic
ages and cyclostratigraphic data close to or around stage boundaries from the
Aptian to Oxfordian stages. New geochronological data from stage boundaries
from the Late Jurassic to Early Cretaceous suggest that the age of the
stage boundaries in this interval could be younger than used in the
M-sequence model of Ogg (2012). For instance, Zhang et al. (2018) provided magnetostratigraphic data for the U–Pb ages of
Midtkandal et al. (2016) in
the Svalbard cores, which suggest that the age of the M0 (base of the
Aptian) is 121–122 Ma rather than ∼126 Ma.
Aguirre-Urreta et al. (2015) presented a high-precision U–Pb age of 127.24±0.03 Ma in the late
Hauterivian (close to the base of the Barremian) in the Agrio Fm.,
Neuquén Basin, which
Martinez et al. (2015) used to anchor cyclostratigraphic studies in the in Río Argos,
Spain,
and calculated an age for the base of the Hauterivian at 131.96±1 Ma
and the base of the Barremian at 126.02±1 Ma.
Aguirre-Urreta et al. (2017) later reported
a U–Pb high-precision age at the early Hauterivian at 130.394±0.037 Ma,
which is fairly close to that of Martinez et al. (2015) for the base of the
Hauterivian. Therefore, new geochronological constraints in the Early
Cretaceous render an apparent systematic offset ∼3–4 Myr younger than those used and predicted by the M-sequence age model of
Ogg et al. (2012, 2016a). Incidentally,
the data we present here for the J–K boundary and close to the KmTB display
the same systematic offset (∼3–4 Myr) compared to the
M-sequence model age of Ogg (2012) and Ogg et al. (2012).
In summary, the M-sequence age model for the Late Jurassic to Early
Cretaceous stage boundaries is a creative solution to present numerical ages
for stage boundaries with a clear lack of reliable radioisotopic ages.
Nevertheless, recent geochronological developments in the Early Cretaceous
show that some of the ages used to anchor the model are likely younger than
previously accepted. Consequently, future updated versions of the M-sequence
model are bound to incorporate these newer age constraints, and the critical
overlap between the M-sequence model of Ogg (2012) and Mahoney et al. (2005) for the
age of the J–K boundary is likely to change. Be that as it
may, reliable radioisotopic ages for the J–K boundary with high-resolution
biostratigraphical markers and magnetostratigraphy in a single section are
still lacking, but growing evidence points to a younger age of the J–K
boundary as well as other stage boundaries in the Late Jurassic and Early
Cretaceous.