GEOCHEMISTRY OF VERTEBRATE BONE SAMPLES

Original (2002) Study

Fassett et al. (2002) presented chemical analyses of 18 vertebrate-bone samples; nine from the Cretaceous Kirtland Formation and nine from the Paleocene Ojo Alamo Sandstone. Two Kirtland samples were turtle bones; the other 16 samples were dinosaur bones. The Fassett et al. (2002) study showed that there are distinct differences in the amounts of uranium (U), the sums of the rare-earth elements (REE), and the chondrite-normalized lanthanum/ytterbium ratios (La/Yb(n)) in Paleocene Ojo Alamo Sandstone bone samples compared to Cretaceous Kirtland Formation bone samples. Uranium abundances exhibited the largest differences between the two suites of samples. The mean value for U in Ojo Alamo bone samples (447 ppm) was found to be 18 times greater than the mean for Kirtland bone samples. REE exhibited less striking, albeit consistent differences. The mean sum of REE was 1,587 for the Ojo Alamo and 3,196 for the Kirtland. The mean La/Yb(n) ratio for Ojo Alamo samples was 6.2 whereas for the Kirtland samples, the mean was 16.1. In summary, the Fassett et al. (2002) study showed that uranium is greatly enriched in Ojo Alamo bone samples relative to Kirtland samples whereas REE are more abundant and relative abundances of REE are more fractionated in Kirtland bone samples.

New Samples

Sample Descriptions. Ten of 14 new samples were collected by the author from newly discovered dinosaur-bone localities. The other four dinosaur bone samples were provided by R.B. Sullivan (The State Museum of Pennsylvania) and by S.G. Lucas (New Mexico Museum of Natural History and Science), as discussed below and noted on Table 2. One of the samples provided by Lucas was from a hadrosaur scapula that was one of 34 bones in the bone assemblage of Hunt and Lucas (1991; and see Figure 38). This sample is number P-19147 (Table 2 and Table 3) collected from locality I of Figure 4. The stratigraphic level of the 34-bone assemblage was measured in the field by the author in 2003 and found to be 6.1 m above the base of the Ojo Alamo. A small lag deposit of vertebrate bone was discovered near this locality 4.6 m above the base of the Ojo Alamo Sandstone and a dinosaur bone fragment from that deposit was collected and analyzed (sample number 020103 of Table 2 and Table 3; locality H of Figure 4). Sample number 35957-LUC (Table 2 and Table 3) was from a large sauropod femur (Alamosaurus sanjuanensis) 4.6 m above the base of the Ojo Alamo at locality D, figure 11. This bone was collected by Lucas who provided a small sample for chemical analysis; Lucas stated (personal commun., 2004) that this bone "was much too massive and pristine to have been reworked from older strata."

Sample SMP VP-1625 was from a large A. sanjuanensis femur (Figure 37.2) collected by R.B. Sullivan who provided a small fragment for chemical analysis. This specimen came from locality O of Figure 4. The author determined that this locality was 4.9 m above the base of the Ojo Alamo; numerous other dinosaur bones in the Ojo Alamo were observed in the same area, including the fragmented bone shown on figure 37.3. Sample number 051504, a fragment of ceratopsian frill bone (locality N, Figure 4) was found in the same area 3.7 m above the base of the Ojo Alamo. Sample SMP VP-1494 was from a sauropod vertebra collected by Sullivan from locality G of Figure 4.

Two samples were collected from a lag deposit of abraded vertebrate-bone fragments found on the surface of the Kirtland Formation nearly 11 m below the base of the Ojo Alamo (locality C, Figure 4). At the time of collection, it was not known if these bone fragments had weathered out of the Kirtland Formation or had washed down from a higher level. The two samples analyzed from this assemblage were a fragment of turtle shell and a fragment of dinosaur bone. These samples were collected primarily to determine if there were significant differences between the chemistry of dinosaur bone and turtle shell.

New-Sample Analyses. The new set of 14 samples was prepared by R.A. Zielinski (USGS, Denver, Colorado) and analyzed by instrumental neutron activation by J.R. Budahn (USGS, Denver, Colorado) using the same procedures described in Fassett et al. (2002). The samples were taken, to the extent possible, from the outermost (cortical) surface of the bones. The chemical analyses of these new vertebrate bone samples (six from the Ojo Alamo Sandstone, six from the Kirtland Formation, and two provenance uncertain) are shown on Table 3. These samples yielded elemental concentrations similar to the original 18 samples discussed above: Ojo Alamo samples contained high levels of U and relatively low levels of REE; Kirtland samples contained low levels of U and relatively high levels of REE. Table 2 shows the concentrations of U and REE for all 32 bone samples (old and new); Figure 44 is a plot of the La/Yb(n) vs. U values for all samples.

As Table 2 and Table 3; and Figure 44 indicate, the small fragments of dinosaur bone (020203-B) and turtle shell (020203-T) collected from an erosion surface on the Kirtland Formation exhibit similar chemistry. Despite their stratigraphic position, these two samples were more like Ojo Alamo bone with somewhat high U and quite low La/Yb(n). On the other hand, these samples resembled Kirtland samples with very high values for their sums of REE (Table 2 and Table 3). Based on their overall chemistry, it might be inferred that these samples weathered out of the Ojo Alamo Sandstone and washed down on top of the Kirtland erosion surface where they were found. Because of the uncertain provenance of these samples, they are not included in the comparison of Cretaceous and Tertiary bone chemistry.

The fragmented dinosaur limb bone from which sample 110803-B was obtained was at the contact of the Kirtland Formation and the Ojo Alamo Sandstone at the Pot Mesa locality (Figure1, Figure 3). The sample was a fragment of a specimen that had been collected by S.G. Lucas in 1983 designated UNM-TOA-2. This specimen is discussed and illustrated in a photograph in Fassett et al. (1987, figure 11). (The stratigraphy of the Kirtland Formation and Ojo Alamo Sandstone in this area was discussed and the formation boundary between these rock units redefined in Fassett et al. 2002). This badly fragmented bone was originally thought to be in the lowermost Ojo Alamo Sandstone (Fassett et al. 1987, p. 29, figure 11; Fassett et al. 2002, p. 324, figure 17). Table 3 shows this bone to have a relatively low concentration of uranium and a relatively high sum of REE; chemical characteristics of a Cretaceous (Kirtland Formation) dinosaur bone. This bone was near the pre-Ojo-Alamo-Sandstone erosion surface when the Ojo Alamo was deposited on top of it. Despite its proximity to the overlying Ojo Alamo, this bone (labeled PM on Figure 44) has retained chemical characteristics of a Cretaceous bone sample; strong evidence that the U and REE concentrations were not appreciably modified when this bone was immersed in mineralizing fluids with a different chemistry nearly 8 m.y. later in Paleocene time.

The Figure 44 plot of La/Yb(n) vs. U for all of the bone samples analyzed for this study shows the significantly higher U content for Ojo Alamo samples vs. Kirtland samples. (Samples (020203-B, and 020203-T) found loose on the surface of the Kirtland Formation are shown in green.) The mean U content for all 15 Ojo Alamo samples is 422 ppm vs. 20 ppm for 15 Kirtland samples (Table 2); more than 20 times greater. The single anomalous U value of 33 ppm is from the large hadrosaur femur collected at the San Juan River locality (Figure 1). As Figure 44 shows, the U value for this sample FBHF on this figure overlaps U values for Kirtland Formation bone samples. This was the most northerly of the Ojo Alamo bone samples and is also the stratigraphically highest 15.2 m above the base of the formation. The northern setting could have affected this sample's anomalously low U content. Uranium mineralization has been documented in the Ojo Alamo Sandstone in the southeast part of the San Juan Basin at Mesa Portales (Fassett et al. 2002, p. 330) close to possible granitic source rocks in the incipient Nacimiento Mountains. Perhaps the U content of mineralizing fluids was less in the northern part of the San Juan Basin, farther from that possible source area, in Ojo Alamo time. One sample, however, provides insufficient data for more than speculation regarding this sample's low U content. The chemical analyses of dinosaur bones from the Animas Formation (when they are found) in the northern part of the San Juan Basin will provide a test of this hypothesis.

Except for the FBHF bone sample from the San Juan River site, the range of U concentrations within Ojo Alamo samples is 89 to 834 ppm. In stark contrast, U values for Kirtland samples range from 2 to 45 ppm (Table 4). Large variations in U concentrations within each of these bone populations may have resulted from: 1) variations in dissolved U concentrations in mineralizing fluids, 2) variations in host lithologies affecting permeabilities and thus volumes of mineralizing fluids the bones were exposed to, and 3) the progress of bone mineralization that influenced the accessibility of U-bearing ground water during fossilization.

Analytical data indicate generally higher sums of REE in Kirtland Formation bones relative to Ojo Alamo Sandstone bones (Table 2, Table 4). Three Ojo Alamo bone samples and one Kirtland sample contained particularly low "Sum REE" values (< 73 ppm, Table 2). These low abundances of REE may indicate samplings of bone further from the bone's cortical (outer) surface. When comparing samples from the outer surface with deeper-bone levels from 70 mm long cores within samples (051298-BB1 and FBHF), the sum of REE concentrations in deeper bone was lower by factors of 2-100, whereas uranium concentrations decreased only by factors < 2 ( Zielinski personal commun., 2007.) and Fassett et al. (2002, p. 329). Preferential concentration of REE in outermost layers is unexplained, but may indicate enhanced uptake of REE related to early diagenetic alteration or recrystallization of outermost fossilized bone. Such uptake is apparently more pronounced in Kirtland bones and must be of limited duration in order to preserve the apparent differences in REE patterns in the two suites of bones.

Figure 45 shows chondrite-normalized rare earth element patterns for the 21 vertebrate-bone samples that contained the greatest concentrations of REE. The newly analyzed samples (dashed lines) show the same subsets of patterns as the samples (solid lines) previously reported in Fassett et al. (2002); that is, more steeply sloped patterns for Kirtland Formation bones and flatter slopes for Ojo Alamo Sandstone bones. The steeper slope of REE patterns in Kirtland samples is primarily caused by a greater abundance of light REE (La, Ce, Nd). Steeper-sloped patterns are represented in the tables as a higher ratio of La/Yb(n).

The data plot for sample 110803-B in Figure 45.2 (Kirtland samples) is plotted in a different color because the slope for this sample is anomalous: it is noticeably flatter than the other Kirtland-sample plots and is more like an Ojo Alamo REE sample plot. As discussed above, sample 110803-B was collected less than 0.1 m below the Kirtland-Ojo Alamo contact and was originally thought to be from the Ojo Alamo Sandstone. Based on its very low U content (30 ppm) and relatively high Sum REE of 2705 (Table 2) this bone has a Kirtland geochemical signature. The anomalously flat slope of the REE plot for this bone (Figure 45.2) may be the result of its proximity to the Kirtland-Ojo Alamo contact that may have allowed for some slight alteration in REE content by Paleocene mineralizing fluids. If so, those fluids did not change the overall Kirtland chemical signature for this bone of low U and high REE (Figure 44). These data support the thesis that the U content of bone samples is fixed at the time of initial mineralization and is not subject to significant change by being immersed in mineralizing-fluids with different chemistry at a later time.

Sample 072598-6C also has a flatter slope on Figure 45 than other Kirtland bone samples and the reasons for this anomaly are unknown. This turtle-bone sample has a low U content (38 ppm) and very high Sum REE (Figure 2) clearly establishing it as a Kirtland bone sample.

Significance of Vertebrate-Fossil Geochemistry

The REE composition of fossil bone has been used to determine stratigraphic provenance and to identify reworked bone in the Triassic of southwest England (Trueman and Benton 1997) and in the Pleistocene in southern Kenya (Trueman et al. 2006). In this study, differences in REE and U contents between two suites of dinosaur bones (Table 2) are preserved, despite their close stratigraphic proximity, and despite their largely shared post-Cretaceous alteration history (Table 4). These data strongly suggest that the chemically distinct Ojo Alamo Sandstone dinosaur bones were fossilized in place during Ojo Alamo Sandstone (Paleocene) time and thus cannot be Cretaceous bones reworked from the underlying Kirtland Formation. These facts, coupled with independent documentation of the Paleocene age of the Ojo Alamo Sandstone presented elsewhere in this paper, indicate that some dinosaurs lived, died, and were preserved in earliest Paleocene time in the San Juan Basin area.