MAGNETOSTRATIGRAPHY

Butler et al. (1977) published the first magnetostratigraphy across the K/T boundary in the San Juan Basin and concluded that dinosaur extinction postdated the marine extinction at the end of the Cretaceous. This debate is well summarized by Fassett (2009), and all workers now agree that the interval of long normal polarity that corresponds to the Fruitland Formation and most of the Kirtland Formation is chron 33n (Figure 1). ⁴⁰Ar/³⁹Ar ages of ~75.56-73.04 Ma from this stratigraphic interval (Fassett and Steiner 1997), as well as all biostratigraphy, indicate a late Campanian age, and this constrains the magnetostratigraphic correlation (e.g., Lucas et al. 2006). The overlying reversal, which encompasses the uppermost Kirtland Formation, can be confidently correlated to the oldest part of chron 32r (Figure 1).

At the other end of the stratigraphic section considered here, the earliest Paleocene (Puercan) mammal assemblage from the lower Nacimiento Formation, which gradationally overlies the Kimbeto Member, is from a stratigraphic interval of normal polarity (Figure 1). This has long been correlated to chron 29n, and that correlation (Puercan = chron 29n) is also accepted in Texas, Utah and Montana, where the magnetostratigraphy of Puercan-mammal-bearing strata has also been determined (e.g., Lofgren et al. 2004, figure 3.2). Indeed, Fassett et al. (2007) reported an ⁴⁰Ar/³⁹Ar age of 64.4 ± 0.5 Ma for an ash bed stratigraphically low in the Nacimiento Formation that supports the correlation of the normal polarity interval in the lowest Nacimiento Formation to chron 29n, a correlation also advocated by Fassett (2009).

The problem is how to correlate the magnetic polarity stratigraphy between chron 29n (Puercan) and chron 32r (late Campanian) in the San Juan Basin. This is the stratigraphic interval of the Ojo Alamo Sandstone (Figure 1) and is either totally of reverse polarity (final conclusion of Butler and Lindsay 1985) or is mostly reverse polarity above and below a short normal polarity chron (conclusion of Fassett 2009). These differences result from Fassett's (2009) willingness to resurrect (as valid) a short normal polarity chron that was first recognized by Butler et al. (1977) and later rejected by them as a spurious normal overprint (Butler and Lindsay 1985; Lindsay et al. 1981). We are skeptical that the normal polarity chron in the Naashoibito Member should be recognized again, without any reanalysis of the paleomagnetism of the strata in question, especially because it was rejected by those who did the original analysis. Even if Fassett is correct, inclusion of a short normal chron within the interval of reverse polarity that corresponds to the Ojo Alamo Sandstone does not contradict assigning a Cretaceous age to the Naashoibito Member.

On the global polarity timescale, between chrons 33n and 29n, there are three normal polarity chrons – 32n, 31n and 30n – that should be present, and chron 32 is a composite chron that encompasses three normal chrons (e.g., Ogg et al. 2004). Polarity is mostly negative between the base of 31r (70.45Ma + 0.65: Ogg et al. 2004) and the top of chron 31n.1r (~68.5Ma: Lerbekmo 2009; top of chron 31n ~ 67.809 Ma: Ogg et al. 2004). During this ~1.95 m.y. interval there are as many as three short duration normal polarity subchrons (c31r.2n, c31r.1n: Lerbekmo and Braman 2002; c31n.2n: Lerbekmo 2009) that might not typically be detected, but any of which would most parsimoniously explain the normal polarity signal in the Naashoibito Member (Figure 1) if it is in fact genuine (sensu Fassett 2009). This means that between chrons 33n and 29n, there are seven normal polarity chrons (chrons 32n.1, 32n.2, 32n.3, 31r.1n, 31r.2n, 31n.2n and 30n). However, in the San Juan Basin section, there is at most one normal chron between 33n and 29n, so much of the magnetic polarity history must be missing. This is prima facie evidence of one or more substantial unconformities in the section. We agree with Fassett (2009) that there is an unconformity at the base of the Ojo Alamo Sandstone (base of the Naashoibito Member), but differ from him on the length of the lacuna represented by the unconformity (Figure 1). Thus, Fassett sees it as an ~8-million-year-long hiatus, an unconformity between late Campanian (~73 Ma) and earliest Paleocene (~65 Ma) rocks. We see it as an unconformity of shorter duration, about 4 million years, between rocks of late Campanian (~73 Ma) and early Maastrichtian (~69 Ma) age. Others, who regard the Alamo Wash local fauna as late Maastrichtian (Lancian) in age, see it as a slightly longer unconformity, ~6 Ma (Williamson and Weil 2008). The important point is that there is broad agreement on a substantial unconformity at the base of the Ojo Alamo Sandstone, so this explains at least some of the missing magnetochrons in the section (Figure 1).

There is also broad agreement that there is a profound unconformity at the base of the Kimbeto Member of the Ojo Alamo Sandstone, at the Cretaceous-Paleocene boundary of most workers (e.g., Baltz et al. 1966; Lehman 1981, 1985; Lucas 1981; Lucas et al. 1987; Newman 1987; Hunt and Lucas 1992; Sullivan and Lucas 2003, 2006; Sullivan et al. 2005b; Williamson and Weil 2008) (Figure 1). Fassett is an exception to this broad agreement; he does not recognize an unconformity within the Ojo Alamo Sandstone, even though the physical stratigraphic evidence of it has been well documented (e.g., Baltz et al. 1966; Powell 1973; Sikkink 1987). This unconformity represents a lacuna of between 2 and 4 million years depending on whether the Alamo Wash local fauna is assigned an early Maastrichtian (Edmontonian) or late Maastrichtian (Lancian) age (compare Sullivan and Lucas 2006 with Williamson and Weil 2008).

Thus, Fassett's (2009) correlation of the magnetostratigraphy recognizes the reversed-normal-reversed interval corresponding to the Ojo Alamo Sandstone as chrons 29r and 29n (Figure 1). He does so because he believes the fossils in the entire Ojo Alamo Sandstone are Paleocene, based primarily on the disputed palynomorph locality in the uppermost Kirtland Formation just discussed. However, the magnetostratigraphy of the Naashoibito Member alone does not demonstrate it is Paleocene in age. The magnetostratigraphy can only be correlated by reference to a datum – and in this case it has to be a biostratigraphic datum. Cretaceous dinosaurs in the Naashoibito Member would support one magnetostratigraphic correlation, while Paleocene dinosaurs would support another (Figure 1). Based on the vertebrate biostratigraphy, the fossils in the Naashoibito Member have long and widely been regarded as Cretaceous, so the chron(s) here should be older than chron 29. How the chron(s) are precisely correlated depends on whether you regard the unconformity-bounded Naashoibito Member as early Maastrichtian or late Maastrichtian in age.

Significantly, Fassett's (2009) reinstatement of the short normal polarity chron (of Butler et al. 1977) in the Naashoibito Member complicates his case for a Paleocene age for the entire Ojo Alamo Sandstone. Given that he accepted the normal chron in the lower Nacimiento Formation and upper Kimbeto Member of the Ojo Alamo Sandstone as chron 29n (= Puercan), he now is forced to include the reversed interval below it, and the short normal chron in the Naashoibito Member as the older part of chron 29n (Figure 1). This creates a composite chron 29n (normal–reversed–normal) not known in the global polarity time scale (cf. Luterbacher et al. 2004) or in other Paleocene magnetostratigraphic sections in western North America (cf. Lofgren et al. 2004). Like other workers (e.g., Butler and Lindsay 1985; Williamson 1996; Lofgren et al. 2004), we regard the reversed polarity interval below chron 29n in the lower Nacimiento Formation as chron 29r (Figure 1), and this is consistent with evidence that the Kimbeto Member is of Paleocene age. Correlation of the reversed chron (or reversed–normal-reversed chron) that corresponds to the Naashoibito Member remains problematic. This interval correlates to some part of chrons 32, 31 or 30 (our biostratigraphy suggests it is most likely part of chron 31r: Figure 1), but is an obviously incomplete record of that time interval and cannot be unambiguously correlated (Figure 1).

Fassett's assignment of the Naashoibito Member to chron 29n is dependent upon: (1) the disputed pollen locality near the top of the underlying Kirtland Formation, (2) the assumption of an unconformity-free, uninterupted magnetostratigraphic succession that spans the Naashoibito-Nacimiento interval and (3) assignment of a Paleocene age to the Naashoibito dinosaurs, which he then (circularly) uses to correlate the Naashoibito magnetostratigraphy to chron 29n. Fassett's (2009, p. 8) statement that "remnant magnetism of rock strata adjacent to the K-T interface in the San Juan Basin provides an objective geochronological tool for placement of the K-T interface and for estimating a more precise age for the base of the Ojo Alamo Sandstone" is thus questionable. We conclude that magnetostratigraphy does not provide definitive evidence of a Paleocene age for dinosaur fossils in the San Juan Basin.