Ascent of Dinosaurs Linked to an Iridium Anomaly at the Triassic-Jurassic Boundary

See allHide authors and affiliations

Science  17 May 2002:
Vol. 296, Issue 5571, pp. 1305-1307
DOI: 10.1126/science.1065522


Analysis of tetrapod footprints and skeletal material from more than 70 localities in eastern North America shows that large theropod dinosaurs appeared less than 10,000 years after the Triassic-Jurassic boundary and less than 30,000 years after the last Triassic taxa, synchronous with a terrestrial mass extinction. This extraordinary turnover is associated with an iridium anomaly (up to 285 parts per trillion, with an average maximum of 141 parts per trillion) and a fern spore spike, suggesting that a bolide impact was the cause. Eastern North American dinosaurian diversity reached a stable maximum less than 100,000 years after the boundary, marking the establishment of dinosaur-dominated communities that prevailed for the next 135 million years.

One of the most striking events in the Mesozoic was the rise to dominance of dinosaurs in terrestrial ecosystems. The cause and timing of their early Mesozoic ascent have been debated (1–4), with difficulties in global correlation and low sampling density limiting the utility of global compilations and obscuring relations to possible forcing mechanisms. However, terrestrial vertebrate assemblages in eastern North America are temporally better constrained than elsewhere and provide high-resolution biological and geochemical data bearing on this issue. This region was within the tropics during the Triassic and contained rift valleys, which were formed during the incipient fragmentation of Pangea. These basins contain kilometer-thick sections of continental strata, termed the Newark Supergroup, which have recorded the rise of dinosaurs across 15° of paleolatitude (5). Milankovitch-type climate cycles permeate the lacustrine strata of these basins, and in conjunction with paleomagnetic reversal stratigraphy, all of the fossils can be placed within a high-resolution astronomically tuned time scale (6, 7) (Fig. 1).

Figure 1

Correlation of four key basins of the Newark Supergroup showing the temporal ranges of footprint ichnogenera and key osteological taxa binned into 1-My intervals showing the change in maximum theropod dinosaur footprint length (line drawn through maximum) and percent at each 1-My level of dinosaur taxa. Short, horizontal lines adjacent to stratigraphic sections show the position of assemblages, and the attached vertical lines indicate the uncertainty in stratigraphic position. Solid diamonds indicate samples of footprints, and open diamonds indicate samples with <10 footprints. Horizontal, dashed gray lines indicate the limits of sampling; thick gray line indicates trend in maximum size of theropod tracks; ?, age uncertain. Ichnotaxa are as follows: 1, Rhynchosauroides hyperbates; 2, unnamed dinosaurian genus 1; 3,Atreipus; 4, Chirotherium lulli; 5,Procolophonichnium; 6, Gwyneddichnium; 7,Apatopus; 8, Brachychirotherium parvum; 9, new taxon B (8); 10, Rhynchosauroides spp.; 11,Ameghinichnus; 12, “Grallator”; 13, “Anchisauripus”; 14, Batrachopus deweyii; 15, “Batrachopusgracilis; 16, Eubrontes giganteus; 17, Anomoepus scambus; and 18,Otozoum moodii. Stratigraphic and magnetostratigraphic columns and correlations are modified from (5). Details of vertebrate assemblages are given in supplemental data (9). Correlation with the other rift basin sequences is based on the larger scale magnetic polarity pattern, Milankovitch cycle stratigraphy, palynology, and basalt geochemistry (20). Ma, million years ago; Hett., Hettangian; Sine., Sinemurian.

Here, we focus on material from 80 localities in four Newark Supergroup basins, consisting of reptile footprints (8, 9), skeletal remains (2,10), and palynological material (11) keyed into the astronomically tuned time scale (Figs. 1 and2). The footprints are abundant, well-preserved, and diverse, and they offer a temporal sampling of terrestrial vertebrate communities that is better than the sampling from skeletal material around the Triassic-Jurassic boundary (4, 8). On the basis of comparisons between the reconstructed osteology of footprints and known skeletal remains, the ichnogenus level generally corresponds to an osteological family or higher taxonomic level (Table 1). However, footprints sample the terrestrial communities directly, and major changes in footprint assemblage composition probably represent important ecological changes (12). Even with uncertainty in the nature of the trackmakers, well-preserved footprints offer a useful independent proxy of faunal change (13), and the observed stratigraphic changes in the ichnological assemblages are consistent with the changes seen in osteological remains (Fig. 1).

Figure 2

Fine-scale correlation between the Ir anomaly and fern spike from the Jacksonwald syncline section of the Newark basin. The average Ir anomaly is based on four localities along strike within the Newark basin, each of which have an Ir anomaly in similar positions [details of data and averaging are included as supplemental material (9)]. The duration of the interval, as based on the linear extrapolation of accumulation rates, is derived from an astronomical calibration of the entire composite Jacsonwald syncline section (9). Percent spore data are averaged from three sections (11). Details of vertebrate assemblages and data averaging methods are given in the supplemental data (9). cly, claystone; md, mudstone; ss, sandstone.

Table 1

Correspondence between footprint and biological taxa. References and explanations are provided in the supplemental material (9).

View this table:

On the basis of compiled ranges tied to the time scale (Fig. 1), Newark Supergroup dinosaurian ichnotaxa show a slow increase in relative abundance and a stepped increase in maximum size below the Triassic-Jurassic boundary (9). The ornithischian dinosaurian ichnogenus Atreipus (14) is the most common dinosaurian form until its last appearance in the middle Rhaetian. Atreipus, the occurrence of a rare form (unnamed dinosaurian genus 1), and the early Norian first appearance of “Anchisauripus” spp. produce a Norian peak in dinosaurian ichnotaxonomic diversity. Nondinosaurian ichnological diversity tends to increase throughout the Late Triassic, with no apparent taxonomic manifestation of the Carnian-Norian boundary. Above the Triassic-Jurassic boundary, non-dinosaurian footprint diversity drops, and dinosaurian ichnogeneric diversity increases to a maximum. At the same time, the maximum size of theropod dinosaur tracks increases by ∼20% with the first appearance of Eubrontes giganteus (15). This pattern is also consistent with that seen at lower temporal resolution at other Triassic-Jurassic sections globally (16).

Skeletal remains are much less common than footprints in eastern North America, but the record parallels the ichnological data. Specifically, the last appearance of procolophonids and phytosaurs and the first appearance of protosuchians occur in the youngest known Triassic osteological assemblage, dated at ∼800,000 years (∼800 ky) before the palynological Triassic-Jurassic boundary (Fig. 1). The oldest Newark Supergroup Jurassic assemblages are from a variety of fluvial, aquatic, and eolian environments in Nova Scotia, dating to <100 ky after the boundary; Triassic forms, such as procolophonids and phytosaurs, are absent (2). This Jurassic osteological material is intimately associated with rich footprint assemblages containing Eubrontes giganteus but lacking Triassic-type footprints.

We found a modest Ir anomaly in the Newark rift basin at the palynologically identified Triassic-Jurassic boundary at the same sites producing much of the new Triassic vertebrate material (Fig. 2). The Ir anomaly of up to 285 parts per trillion (ppt), with an average maximum of 141 ppt (0.285 and 0.141 ng/g), is seen at four correlative sections; the anomaly is stratigraphically coincident with a transient, but large, increase in fern spore abundance and is substantially above background levels of ∼50 ppt (9,17, 18). The increase in Ir and the spike in fern spore abundance occur in a white clay layer that is ∼1 m (∼1 ky) above the last occurrence of Patinasporites densus and other typical Triassic pollen and spores and is ∼5 m (∼5 ky) below the first typical Hettangian assemblage, consisting mostly ofCorollina. Thus, within sampling resolution, the Ir anomaly and the “fern spike” are synchronous with the Triassic-Jurassic boundary. A few meters below the boundary is a thin zone of reversed magnetic polarity (chron E23r), identified at all three sections from which we have paleomagnetic analyses (7) (Fig. 2). Within 15 m above the boundary is the base of the oldest basalt flow in the Newark basin, the Orange Mountain Basalt, which is the oldest known North American part of the voluminous Central Atlantic Magmatic Province (CAMP) (19). On the basis of calibration by Milankovitch lake level cycles, the Triassic-Jurassic boundary occurs ∼20 ky before the extrusion of the oldest basalt and <20 ky after chron E23r (20).

The westernmost of the four sections examined for an Ir anomaly allows a test of the possibility that the apparent Ir anomaly was caused by diagenetic migration of Ir at a strong redox boundary such as the carbonaceous horizon present at the other three sections (9). The westernmost boundary section consists of red and minor light gray clastic rocks, lacking black shales and coals, but does contain both the fern spike and thin zone of reversed magnetic polarity characteristic of the boundary interval. Despite the absence of a redox boundary, the Ir anomaly is associated with the fern spike (9). In addition, there are no correlations between Ir and Al2O3 or other major elements that may be subject to diagenetic enrichment (9), although additional sampling for Ir and other platinum group metals is needed to further test the meaning of the observed anomaly.

The lower resolution [1 million years (1 My)] terrestrial vertebrate pattern at the Triassic-Jurassic boundary, based on data from the Culpeper, Fundy, Hartford, and Newark basins (Fig. 1), is mirrored at higher resolution (<20 ky) in the Newark basin (9) around the Ir anomaly and fern spike. On the basis of the Milankovitch cyclostratigraphy, the last occurrence ofBrachychirotherium parvum (a footprint form typical throughout the previous 20 My of the eastern North American Triassic) is <20 ky older than the palynologically dated Triassic-Jurassic boundary, whereas the first occurrence of Eubrontes giganteus is ∼10 ky younger. “Anchisauripus” and “Grallator” species occur with abundantBrachychirotherium parvum at a level that is ∼40 ky older than the boundary. Thus, the transition from Triassic- to Jurassic-type footprint assemblages occurred within a 50-ky interval that includes the boundary, fern spike, and Ir anomaly (9). These oldest examples of Eubrontes giganteus are as large as any in the rest of the Jurassic Newark Supergroup (35 cm).

These earliest known Jurassic track assemblages are of strikingly low diversity, despite identification of thousands of specimens. Only Rhynchosauroides sp., Batrachopus deweyii, “Grallator” spp., “Anchisauripus” spp., and Eubrontes giganteusare known from these beds. Herbivores (e.g., prosauropods and small ornithischians) appear within 100 ky after the Triassic-Jurassic boundary, represented by both tracks and bones (2). Apart from an apparent increase in the frequency of prosauropod tracks (Otozoum) and the size of the largest ornithischian tracks (Anomoepus), there is no obvious change in assemblage composition throughout the rest of the eastern North American record.

The association of an Ir anomaly, fern spike, and terrestrial mass extinction level is reminiscent of the pattern produced by bolide impact at the Cretaceous-Tertiary (K-T) boundary in western North America [e.g., (21)]. Shocked quartz, although not yet found in eastern North America (2,22), has been reported at the marine Triassic-Jurassic boundary in Italy (23). Further similarities include a negative δ13C shift in both marine organic matter (24) and terrestrial plants (25) at the Triassic-Jurassic boundary, suggesting major perturbations in the carbon cycle. These similarities in pattern suggest a similarity in cause (2), namely, an impact of one or more large extraterrestrial bodies (26) and/or the environmental effects of the emplacement of giant igneous provinces (e.g., CAMP) (19). However, unlike the extinction at the K-T boundary, non-avian dinosaurs survived the Triassic-Jurassic boundary, and theropod dinosaurs rapidly increased in size.

Two scenarios might explain the increase in size in theropod dinosaurs across the Triassic-Jurassic boundary: First, the appearance of the much larger theropods represents a dispersal event from some unknown location, or second, it represents an evolutionary event. We favor the second hypothesis and suggest that the increase in size may be explained by a rapid (thousands of years) evolutionary response by the theropod survivors (which may have been quite small) to ecological release (27). This response would be similar to that inferred for reptiles on modern islands lacking competitors (28). The evolutionary hypothesis could be falsified by the discovery of large (Dilophosaurus-sized) theropod bones or diagnosticEubrontes giganteus tracks in unquestionably Triassic strata. In any case, the evidence suggests that the dramatic decrease in nondinosaurian diversity was caused by an extrinsic environmental catastrophe. The resulting drop in competitive pressure was the trigger for the global spread of large theropods, and these, along with the surviving prosauropods and ornithischians, established the familiar global dinosaurian-dominated ecological pattern seen across the terrestrial world for the succeeding 135 My.


View Abstract

Navigate This Article