Case 1: The numerical age of the Jurassic-Cretaceous boundary

defined numerical age. Though there are radiometric ages available from several regions with Berriasian strata (N. California, Andes, Tibet, Mexico, Japan and others), the classical

sedimentary sequences through the Jurassic-Cretaceous (J-K) interval in Tethys are mostly without datable materials, as is the case with the Tithonian stage below and the Valanginian stage above. Assigning a numerical age to the J-K boundary has been difficult, and over the last 25 years has relied mainly on different approaches and some repeated assumptions, which have yielded a large range of ages, ranging from 135 to 145 Ma (Bralower et al., 1990; Gradstein et al., 1995; Liu et al., 2013; López-Martínez et al., 2015; Lowrie and Ogg, 1985; Mahoney et al., 2005; Vennari et al., 2014). Currently, the adopted age of the boundary by the International Commission on Stratigraphy (ICS) for the base of the Berriasian is 145.0±0.8 Ma. This age has been the most enduring one for the boundary, due to the agreement between the M-sequence model Ogg, (2012) and the Ar-Ar age of Mahoney et al. (2005).

The M-sequence model (Ogg et al., 2012) uses a spline-fitting regression that predicts the ages of stage boundaries from the Early Cretaceous (Aptian) to the Late Jurassic (Oxfordian). The model is a combination of astronomically calibration durations (Huang et al., 2010a) for some of the magnetozones and the distances between the magnetic anomalies in the NW Pacific (Channell et al., 1995; Larson and Hilde, 1975; Tamaki and Larson, 1988) with the aim of reaching a spreading rate for the NW Pacific. The ages of stages boundaries are back-calculated starting from the M0r, the base of the Aptian, estimated as 126.3±0.4 Ma (Huang et al., 2010b). In the case of the J-K boundary, the projected age was 145.0±0.8 Ma. Mahoney et al. (2005) measured an Ar-Ar age of 144.2±2.6 Ma from the M19-M20 anomalies in the Shatsky Rise, which was used to be a rough estimate for the base of the Cretaceous (Ogg and Lowrie, 1986; now more accurately placed in the middle of the M19.2n (Wimbledon 2017). This was later corrected to 145.5±0.8 Ma (Gradstein, 2012), because of the recalibrated ⁴⁰K decay constant (Renne et al., 2010). In short, both of these studies have taken completely different and independent approaches, but have yielded similar ages.

In contrast, when radio-isotopic dating has been applied directly to rocks closely associated with boundary markers, the results have been significantly different. Bralower et al. (1990) dated volcanic horizons in the Argentiniceras infracretacea (upper Berriasian) zone from Grindstone Creek, California, at circa 137.1 Ma (ID-TIMS). The authors back-calculated the age of the Berriasian base at 141.1 Ma, using an arbitrary duration of 4-6 Myr for the biozones and their presumed age in the upper Berriasian as an anchor. Liu et

al. (2013) dated interbedded bentonites in four sections in the Sangxiu Fm., Tibet. The ages cluster around 140-142 Ma (U-Pb, SIMS), but the ages violate stratigraphic superposition, and a single age for the boundary is not quoted, but rather left as a preferred interval (140-142 Ma). López-Martínez et al. (2015) dated a volcanic unit in the Pimienta Fm, Mexico, at 139.1 ± 2 Ma (LA-ICP-MS). The unit lies between the Colomi subzone (highest Jurassic) and the Elliptica subzone (lower Berriasian), that is, bracketing the J-K boundary, though the base of the Calpionella alpina subzone is not present so that a very precise constraint on the boundary age is not possible. Vennari et al. (2014) dated an ash bed 20 m above the J-K boundary at 139.55±0.03 Ma (CA-ID-TIMS), within the Andean upper A. noduliferum Zone, placing it in the NJK-D nannofossil zone. The authors then assumed a constant sedimentation rate of 4.28 cm/kyr to back-calculate the age of the boundary to ca. 140 Ma.

Pessagno et al., (2009) (after Mattinson et al., 2008), also report CA-ID-TIMS age of 143.734±0.060 Ma on the La Désirade igneous complex interbedded with the upper Tithonian radiolarian of Subzone 4 and the Buchia piochii Zone.

One of the drawbacks of geochronological studies dedicated to the J-K boundary is that the dated horizons are often stratigraphically distant from the boundary. This has forced the use of constant sedimentation rates, or durations of magnetozones and biozones to back-calculate the age of the boundary, which inevitably introduces unknown errors in the age.

Another issue is that geochronological data has seldom been coupled with magnetostratigraphy, and thus the connection to the middle of M19n.2 is elusive. As a result, the correlation between well-studied J-K sections in western Tethys and radio-isotopic ages around the boundary have been based on biostratigraphic data. However, in many of the dated sections, the presence of (primary and secondary) biotic boundary markers is rare. Lastly, as discussed before, the majority of the analytical techniques employed to date the boundary (e.g. SIMS, LA-ICP-MS) lack precision and accuracy, usually limited to 1-3% of their relative age, which in the case of the J-K transition (ca.

140-145 Ma) yields precisions in the order of 1-3 Myr, which is inadequate to calibrate the age of a stage boundary.

The comparison between studies that have attempted to calibrate the age of the base of the Berriasian using radio-isotopic dating and the M-sequence model (Ogg, 2012) is not a simple one. The absence of studies that combine radio-isotopic ages, biostratigraphy, and magnetostratigraphy is a major challenge for dateable global correlations.

Geochronological studies lacking any magnetostratigraphy effectively prevents accurate dating of Tithonian-Berriasian magnetozones, and thus an accurate assessment of the age of the base of the Berriasian (i.e. the base of the Alpina Subzone and the M19.2n) is not possible. As a result, the numerical age of the J-K boundary continues to be elusive despite recent advances. Nevertheless, it is worth pointing out that accuracy of the M-sequence age model is dependent on a complex web of correlations, and is ultimately dependent on the quality of available radio-isotopic ages and cyclostratigraphic data close to or around stage boundaries between the Oxfordian and the Aptian.

Recently, new high-precision U-Pb geochronological data from stage boundaries the Late Jurassic to Early Cretaceous suggest that the ages of the stage boundaries in this interval could be younger than those used in the M-sequence model. For instance, Zhang et al. (2018) provided magnetostratigraphic data allied to the U–Pb ages of Midtkandal et al. (2016) in the Svalbard cores, which suggest that the age of the M0 (Aptian base) 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 highest Hauterivian in the Agrio Fm. (Neuquén Basin, Argentina), which Martinez et al. (2015) used to anchor cyclostratigraphic studies at Río Argos, 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 for the lower Hauterivian at 130.394±0.037 Ma, which is fairly close to that of Martinez et al. (2015). Therefore, new geochronological constraints in the Early Cretaceous suggest a consistent offset to younger ages: that is, of ∼3–4 Myr less than those predicted by the M-sequence age model. Therefore, future updates of the M-sequence model are likely to predict younger ages for the stage boundaries from the Late Cretaceous to Early Cretaceous to younger ages and the crucial apparent agreement between the M-sequence model (Ogg, 2012) and the age of Mahoney et al. (2005) is likely to change.

Lastly, the age within the Tithonian are also at odds with the M-sequence model (Pessagno et al., 2009). In summary, U-Pb high-precision geochronology data in the Early Cretaceous (Hauterivian, Berremian), earliest Berriasian, late and early Tithonian contradict the currently agreed age for the boundary and indicate a younger age for the boundary, albeit the exact age is still contentious.

The issues surrounding the numerical age of the J-K boundary have been a challenge for decades. The clear lack of high-precision U-Pb ages around the J-K boundary precludes

and robust conclusion on the base of the boundary. Since newer high-precision U-Pb geochronology in the Hauteriavian, Barremian, and Barriasian, seem to indicate that the age of the boundary is likely younger, in this case study I aim to date the age of the boundary by age bracketing prominent markers for the boundary using high-precision U-Pb geochronology to test if the age of the boundary is in fact younger, all the while calibrating its age. Furthermore, I test the synchronicity of the First Occurrence of the primary markers in two distinction sections from different and disparate geological context as this is a primary condition for establishing a future GSSP.

1.4.2 Case 2: The late Pliensbachian (Early Jurassic) and the Karoo LIP