The footprints found in Kirmenjak quarry are, generally speaking, shallow for their dimensions (Supplementary Data 11), which implies that the carbonate mud in which they were formed was rather solid and hard, with greater imprint resistance. Despite their overall shallowness and un-pronounced morphology they are not preserved as undertracks. Clearly visible expulsion rims around the tracks verify their preservation as true tracks (Fig. 15C). Footprint depth varies greatly in relation to distance of the track from the waterline, as has been shown also in some recent examples (e.g., Cohen et al., 1993). Track depth is almost always related to the water saturation of the substrate. The deepest prints usually originate in substrate in which there is neither too much water nor too little (Cohen et al., 1993). The relatively shallow depth of Kirmenjak tracks does not necessarily mean that the substrate on which the dinosaurs walked was dry and exposed to the air; instead they could have been formed below water level where the sediment has greater water saturation. Footprint depths increase from the southern to the northern part of the outcrop in what could reflect different water saturation conditions in the substrate or a longer exposure to erosion. On the northern side of the outcrop footprints show a more pronounced morphology, which could indicate that the substrate was more pervious to the imprint of the foot. Obviously, all of the tracks at the outcrop were not formed at the same time. Those that were formed first are more eroded than the later-formed tracks.

To calculate the dimensions of sauropods from Kirmenjak we used the gleno-acetabular distance (GAD; Thulborn, 1990) and footprint length, and calculated hip height as 5.9 X foot length (FL) and 4 X FL (Alexander, 1989; Thulborn, 1990). The GAD was measured from the midpoint between a pair of hindfoot impressions to the midpoint between the next pair of forefoot impressions. There is a problem with this method

Sauropod Footprints
FIGURE 10—Heteropody of Kirmenjak footprints compared to other sauropod tracks (redrawn after Lockley et al., 1994a).

in that the distance is longer if the pace length becomes longer. Accordingly, a smaller dinosaur with longer steps would have a longer gleno-acetabular distance and thereby distort its dimensions. Applying the minimal (107 cm in trackway XII) and maximal (192 cm in trackway XV) measured values of the GAD (Supplementary Data 21) on the skeleton

Paluxy River Trackway
FIGURE 11—Trackway XIX. For position of the trackway, see Figure 7.

of Dicraeosaurus, we calculated the lengths of the animals as ranging from 7.5 m to 13.5 m. Using the calculation of hip height (Supplementary Data 21) from footprint length and comparing it to Dicraeosaurus, body length ranges from 8.5 m to 17.5 m (h = 5.9FL) and 6 m to 12 m (h = 4FL). Such discrepancies come from the fact that Kirmenjak tracks do not belong to a single individual but to a number of individuals of varying size, based on the variation in footprint lengths. According to the calculated parameters, the smallest sauropods on the site were about 7.5 m in length, while the largest individuals attained some 14.5 m in length. As noted above, the size of the sauropods from Kirmenjak quarry was smaller than other Late Jurassic sauropod titans established from skeletal remains. This could be explained by the presence of new sauropod taxa or simply by the presence of relatively smaller species within a genus of large-bodied types, keeping in mind that all the species in a single genus were not of the same size.

The speeds of Kirmenjak sauropods were estimated using Alexander's (1976) formula, v = 0.25g05 X SL167 X h-117, where v = meters/second, SL = stride length, and h = hip height (Supplementary Data 21). We did not use the average stride length and hip height of the single trackway; instead we calculated the speed for each segment of the trackway independently in order to obtain more relevant values. The speeds range from 0.5-2.5 km/h, which is comparable to the speeds estimated for sauropods in other localities (Thulborn, 1990; Lockley and Meyer, 1999). The average walking speed (AWS) for Kirmenjak sauropods was estimated (Supplementary Data 21) using Thulborn's formula v= 1.675h0129 (1990; his table 10.3). Values range from 3.21-3.50 km/h; an AWS of 3.50 km/h is similar to the predicted values for Diplodocus or Apatosau-rus (3.45 and 3.51 km/h respectively; Thulborn, 1990). Almost all the measured speeds of Kirmenjak sauropods are one half or one third of their AWS, suggesting that the preferred gait of Kirmenjak sauropods was a very slow walk. The relationship between stride length and footprint length (SL/FL) as well as stride length and hip height (SL/h) was also calculated (Supplementary Data 21). The average ratio of SL/FL for Kirmenjak sauropods is 3.24, significantly lower than the values obtained by Thulborn (7-8; 1990). Similarly, the SL/h ratio is 0.53, lower than the value predicted by Thulborn (0.95; 1990) for sauropods. Thus, Kirmenjak sauropods walked by taking relatively short strides. Furthermore, it is observed that the distance between the manus and pes prints in Kirmenjak trackways depends on the speed of the animal. The greater the speed, the longer the distance. In the case of the slow walk, overlap of prints occurs and the pes prints cover the manus prints. This

feature indicates that the animals move their legs in pairs; e.g., front left foot-hind left foot, front right foot-hind right foot, similar to modern elephants (see also Alexander, 1989). This is further proof that the gait of sauropods resembled that of modern quadrupeds more than the gait of reptiles.

feature indicates that the animals move their legs in pairs; e.g., front left foot-hind left foot, front right foot-hind right foot, similar to modern elephants (see also Alexander, 1989). This is further proof that the gait of sauropods resembled that of modern quadrupeds more than the gait of reptiles.

Sauropod Gaits

Lommiswil and Moutier sites in Switzerland (Meyer, 1993) are also Kim-meridgian in age; here the tracks represent wide-gauge, Brontopodus type. Although the Kirmenjak footprints lack distinctive morphologic features, their narrow-gauge, strong heteropody (small manus), and outwardly rotated manus are characteristic of the sauropod ichnogenus,

The orientation of the footprints on the outcrop usually depends upon some topographic or behavioral character. The direction of sauropod movement along the shoreline is a common case (Thulborn, 1990; Lock-ley and Meyer, 1999). Kirmenjak sauropods tended to walk north, northeast, and east, as indicated by measured trackway orientations (Fig. 17), so the former shoreline could have had a NE-SW direction. Analysis of pes print length (PL) and manus print width (MW) distribution allows interpretation of the composition of the ichnocoenosis. The distribution of the PL and MW among the Kirmenjak tracks indicates three distinct size categories (Fig. 18). The first category is the modal class of 28 cm pes length, second is the class of 33-35 cm, and third is the class of 40 cm upward. Such a grouping of footprint sizes suggests the presence of three sauropod taxa or three different age sizes of the same species (adult, subadult, and juvenile). They could indicate the presence of a sauropod herd composed of individuals of different ages (i.e., sizes), especially given the similar states of preservation of some trackways. The northern part of the site is especially interesting because four parallel trackways (XVII, XVIII, XIX, and XX; Fig. 7) and numerous individual footprints with the same orientation have been found (Fig. 15D). They most probably represent the footprints of the same generation because similar preservation suggests that they were produced at the same time. Because the trackways are so closely spaced and sometimes overlap, it could be assumed that those sauropods moved together, but not parallel to each other; instead they walked one behind the other.

Late Jurassic sauropod footprints are known in Europe from Portugal and Switzerland (Lockley and Meyer, 1999). The Cabo Espichel site in Portugal is similar to Kirmenjak in age and environment of deposition, but the footprints are of the wide-gauge type and have been ascribed to ichnogenus Brontopodus (Lockley et al., 1994b). The footprints at Avelino, Portugal, are of similar size, gauge, and type as the Kirmenjak tracks but are Kimmeridgian in age (Lockley and Santos, 1993). The

Lommiswil and Moutier sites in Switzerland (Meyer, 1993) are also Kim-meridgian in age; here the tracks represent wide-gauge, Brontopodus type. Although the Kirmenjak footprints lack distinctive morphologic features, their narrow-gauge, strong heteropody (small manus), and outwardly rotated manus are characteristic of the sauropod ichnogenus,

Sauropod Footprints
FIGURE 16—Kirmenjak quarry trackway compared to wide- and narrow-gauge types of sauropod trackways (redrawn after Lockley et al., 1994a).
FIGURE 18—Pes length and manus width distribution of Kirmenjak sauropods.

Parabrontopodus (Lockley et al., 1994a; Lockley and Meyer, 1999). Par-abrontopodus manus prints are wider than long, semicircular in shape, significantly smaller in size than the pes prints, and lack clearly visible digit impressions (Lockley et al., 1994a). Parabrontopodus pes prints are longer than wide with the longer axis rotated outward from the trackway midline (Lockley et al., 1994a). The Kirmenjak footprints most resemble those from the Courtedoux track site in Switzerland (Marty et al., 2003). These tracks are of the Parabrontopodus type, narrow gauge and similar dimensions, but are Kimmeridgian in age (Marty et al., 2003). Since the sauropod tracks from the Kirmenjak site are found in the carbonate environment of the ADCP, this ichnocoenosis could be assigned to Bron-topodus ichnofacies sensu Lockley et al. (1994c), which is defined as a sauropod footprint ichnocoenosis in carbonate facies related to a carbonate platform environment. This ichnocoenosis occurs in the Late Jurassic and Cretaceous from numerous localities (United States, South Korea, Switzerland, Portugal, Morocco, and Bolivia). Similar cases of sauropod tracks in a carbonate platform environment have been described from the Paluxy River site in the Glen Rose Formation, Texas (Albian; Pittman, 1989) and from the Cabo Espichel site in Portugal (Late Jurassic; Lockley and Meyer, 1999). The main differences between these two sites and Kirmenjak is the fact that the Glen Rose Formation was deposited on an epeiric platform (Pittman, 1989), and both of these sites show minor siliciclastic content indicating terrigenous influx. Kirmenjak sediments, on the other hand, were deposited on a more-or-less isolated carbonate platform with no terrigenous influx.

The African sauropod fauna from the Late Jurassic is more diverse than the European fauna (Weishampel et al., 2004), and the only find of sauropod bones on the ADCP (Lower Cretaceous of Bale, Istria) shows African affinities (Dalla Vecchia, 1998a). In Africa, a rich dinosaur fauna is known from the Mkoawa Mtwara locality in Tanzania (Weishampel et al., 2004). The locality belongs to the Kimmeridgian Tendaguru Formation and includes: Diplodocoidea: Dicraeosaurus hansemanni Janensch, D. sattleri Janensch, Tornieria africanus (Fraas), and Barosaurus gracilis Janensch; Brachiosauridae: Brachiosaurus brancai Janensch; Titanosau-ria: Janenschia robusta (Fraas); and Sauropoda incertae sedis: Tenda-guria tanzaniensis Bonaparte, Heinrich, & Wild (Weishampel et al., 2004). The sauropods from Kirmenjak quarry are more similar to Dicraeosaurus than to the other African taxa (Upchurch et al., 2004), in their smaller size (compared to the so-called titans of Late Jurassic age) and the diplodocoid nature of their tracks.

During the Late Jurassic the ADCP was one of the peri-Adriatic platforms oriented approximately NW-SE and was surrounded by deep marine basins on its western, northern, and eastern sides, partly analogous to the Bahaman archipelago in size, form, and internal structure (D'Argenio et al., 1975). The ADCP represented the northern tip of the Central Mediterranean Carbonate Platform (Vlahovic et al., 2005), a larger paleogeographic unit that was also elongated in a NW-SE direction. According to Dercourt et al. (1993) this unit was separated from Laurasia and Gondwana by deep marine trenches. Dalla Vecchia (1998b) suggested that the ADCP was connected with the Apulian, Apenninic, and Trento platforms during the Cretaceous and via those platforms with the African continent as well, although it was a hundred kilometers from the so-called true land. This would make the ADCP a unique example of dinosaurs living on an isolated carbonate platform. The other Mesozoic carbonate platforms with dinocoenoses (Texas and Portugal) represent periconti-nental platforms with carbonate prograding landward into siliciclastic shore and continental sediments. During the Late Jurassic the ADCP was characterized by shallow-water carbonate sedimentation (Vlahovic et al., 2005). The occurrences of Middle Jurassic-Lower Cretaceous ophiolite in the peri-Adriatic region indicate the separation of the ADCP from the Eurasian continent by an ocean (Bosellini, 2002); however there is no unequivocal evidence of the ADCP southern connection with the African continent. There is still no sedimentary record on the ADCP other than the one that implies continuous shallow-water carbonate deposition; terrestrial deposits have not been found (Vlahovic et al., 2005). The discoveries of dinosaurs within ADCP sediments are therefore extremely valuable, because they provide evidence of the existence of land and connection of the platform with the continent. All the available geological and sedimentological data reconstruct the ADCP as a flattened area surrounded by tidal flats and shallow lagoons. There is no direct evidence of any developed relief with forest cover and fresh water, crucial conditions for the survival of hundreds of large herbivores. Dinosaurs could not actually live on the carbonate platform itself, which was sensu stricto a few meters under water and characterized by tidal flats, channels, lagoons, and islets. There had to be a large terrestrial area nearby with lush vegetation and a well-developed river system in order to sustain such large animals. The entire ADCP area, as reconstructed earlier (e.g., Der-court et al., 1993; Blakey, 2004), does not represent an environment in which the sauropods could have had survived. Some interpretations, however, have envisaged the full terrestrialization of Istria during shorter or longer time periods of the Mesozoic (Maticec et al., 1996; Veseli, 1999). Comparing the ADCP with the Bahamas-type platform would be risky with regard to dinosaur finds, as the ADCP was not an isolated platform during the Late Jurassic, according to the presence of sauropod tracks. The fact that there is still no record of terrestrial sediments in the Late Jurassic of ADCP does not prove that there was no land. The lacustrine sediments in the Lower Cretaceous of Bale, Istria, are filled with dinosaur bones (Dalla Vecchia, 1998a) and are proof that there was a widespread terrestrial area, at least during the Early Cretaceous. The Kirmenjak tracks are preserved in shallow marine sediments, not of an isolated platform but of a more coastal environment. During the Late Jurassic the sauropods may have migrated from the African continent onto ADCP across the wide corridors of shallow water environments that were formed during emersions. During this time the ADCP may have been less like the Bahamas platform and more fully terrestrialized, even similar to the Florida peninsula (Bosellini, 2002).

0 0

Post a comment