DISCUSSION

There are several similarities in the isotopic patterns of Giganotosaurus and Tyrannosaurus with one major difference. In both dinosaurs, the vertebrae are the warmest (lowest ¹⁸O values) bones sampled. The cervicals and dorsal from Giganotosaurus centra show less variability than in the T. rex specimen. Previously, the ribs and dorsal vertebrae were considered as "core" bones representative of the warmest body temperatures. However, the mid section and distal end of ribs are in the outer shell of the body and may represent temperatures more similar to those of the extremities than the core body. In T. rex, values from the gastralia were similar to those from the femur (Barrick and Showers 1994). In both T. rex and Giganotosaurus, the ribs are 2-4°C cooler than the vertebra. The Giganotosaurus gastralia do have greater intrabone variability than in the T. rex individual. Both individuals display regional heterothermy in the limbs with cooler temperatures and greater isotopic variability occurring with greater distance from the core body. In both cases the femora are 3-4°C cooler than the vertebrae and the tibia 4-5°C cooler than the vertebrae. Only in Giganotosaurus were proximal and distal ends of these bones compared. No difference in mean temperature or variability was found along the femur while the tibia exhibited a trend of increasing temperature variability (1°C) and cooler mean temperature (1°C) distally along the shaft. Unfortunately there were no pedal elements preserved in Giganotosaurus. The consistency in trends of increasing isotopic variability and decreasing mean temperature distally along the limb elements between the Giganotosaurus and Tyrannosaurus suggest that the foot in Giganotosaurus was likely cooler and more variable than the tibia. Thus, the ¹⁸O values between the dorsal/cervical vertebrae, ribs and limbs of both animals are strikingly similar suggesting similar thermoregulatory patterns. However, the ¹⁸O values in the caudal vertebrae differ significantly. In T. rex, the caudal vertebrae exhibit similar intrabone patterns and gradually become more positive (~0.5-4°C cooler) with distance from the torso (Barrick and Showers 1994). In Giganotosaurus, however the caudal vertebrae exhibit the same mean temperature values as the dorsal and cervical vertebrae. The mean values from the proximal caudals resembled those of the dorsal vertebrae in both individuals. Yet in T. rex intrabone temperature variability in the 3 caudals analyzed (~2°C) was less than that in 4 Giganotosaurus caudals ranging from 2.5-5.5°C.

The features indicative of endothermy (defined as the maintenance a high and constant body temperature by metabolic means by Bennett and Ruben 1979) in Giganotosaurus include the homeothermy of the cervical and dorsal vertebra and regional heterothermy of the limbs. The pattern is even more suggestive than for T. rex as the core (vertebral) temperature range is nearly 1.5°C narrower in Giganotosaurus. Clarke and Jenkyns (1999) report maximum Albian –Cenomanian sea-surface temperatures of 16-18°C for southern latitudes of 49°-53°S. This strongly suggests that the terrestrial climate was temperate where Giganotosaurus roamed in the Neuquen basin. Such a climate provides evidence against the warm homogenous environment in which gigantothermy could provide stable core body temperatures. In order to maintain the thermoregulatory pattern seen in Giganototsaurus, an elevated metabolic rate is required. However, the pattern of high intrabone variability and no mean interbone differences in the tail is also typical of a mass homeotherm (Barrick et al. 1997). Core homeothermy and regional limb heterothermy indicate that both individuals maintained metabolic levels above those of modern reptiles. While core homeothermy within a temperate climate fits the definition of endothermy it is likely that these dinosaurs possessed intermediate metabolic levels rather than modern mammalian levels. A drop in mass-specific metabolic rates between juveniles and adults at this intermediate level would also explain the mass homeotherm-like isotopic pattern seen in the Giganotosaurus tail. In adults, bone turnover rates in vertebrae are much more rapid than in limb bones (Francillon-Vieillot et al. 1990) and reflect a narrower time of bone deposition. Thus, a change in thermoregulatory style would be presented in the vertebrae more rapidly than in the limbs. Endothermy has come to be synonymous with mammalian and avian tachymetabolism where metabolic rates are 10x greater than for ectotherms (e.g., Ruben 1995). Endogenous heat production through elevated metabolic rates is required for endothermy, but intermediate levels (e.g. 5x metabolic rate of ectotherms) may have been high enough for the maintenance of homeothermy in dinosaurs, thus making them at least intermediate endotherms.

At an adult body mass of 8000 kg, even if Giganotosaurus began life with a modern avian metabolic level, it would have had an intermediate metabolic rate the equivalent of a 1000 kg mammalian carnivore. At 6000 kg, the metabolic rate of T. rex would have been equivalent to an 800 kg mammal. The dinosaurian metabolic rate is calculated by projecting the mass vs. field metabolic rate regression of modern avians, mammals, and reptilians (Nagy 1987) to a mass of 6000 and 8000 kg (Fig. 4) and projecting a hypothesized dinosaur regression midway between the reptilian and eutherian mammal regression lines. In fact, at this weight (~7000 kg), the avian regression line falls midway between the projected eutherian mammal and reptilian regression lines, intersecting the hypothesized intermediate dinosaurian regression line. Theropods are more closely related to birds (Ji-Qiang et al. 1998) than other dinosaurs and a projected intermediate metabolic level for 6000-8000 kg birds, fits well with the interpretation of intermediate metabolic levels for these large theropods based on their p distribution. The thermoregulatory pattern in Giganotosaurus corroborates the interpretation based on the pattern seen in T. rex. If we assume that all adults of Giganotosaurus and Tyrannosaurus reflect the same thermoregulatory patterns as the two individuals sampled, they would have enjoyed the benefits of endothermy at metabolic levels intermediate between those of modern mammals and reptilians. An extension of this interpretation is that juvenile theropods also maintained intermediate if not high metabolic levels. Comparatively, an 8000 kg lizard or crocodile would have the same metabolic food requirements as a 200 kg carnivorous mammal (i.e., medium sized male African lion Panthera leo). Food requirements for various sized eutherian mammals can be calculated from the generalized equation y=ax^b where y is the food requirement in g/day, a=0.235, x is the mass of the individual in g, and b=0.822 (eq. 19 in Nagy 1987). By converting dinosaurian or reptilian metabolic rates to equivalent mammalian metabolic rates this single equation can be used to estimate food requirements for these dinosaurs. Using the intermediate metabolic rates for Giganotosaurus (8000 kg) and Tyrannosaurus (6000 kg), their food requirements would have been 20 and 17 kg/day respectively. This is the same food requirement for 3-4 large male tigers or African lions. Modern male lions (Panthera leo) reach only ~250 kg while large male tigers (Panthera tigris) may reach just over 300 kg (Walkers Mammal’s of the World, 1991). For comparative purposes, had Giganotosaurus maintained a strictly reptilian metabolic rate, it would have required 5.5 kg/day, whereas a Giganotosaurus individual with a eutherian metabolic rate would have required 111 kg/day of food. Thus, for under 4x rather than 20x the food requirements of a similarly sized bradymetabolic ectotherm or mass homeotherm, Giganotosaurus enjoyed the benefits of endothermy. These benefits include, core homeothermy 365 days a year, higher activity levels, and rapid growth rates. Concomitantly, these dinosaurs would also have kept some of the benefits typical of bradymetabolic ectotherms (e.g., lower food requirements, greater percentage of energy budget directed toward growth and reproduction). Table 1 [what is this???] records the food requirements of a tyrannosaur or giganotosaur at various body sizes throughout ontogeny assuming this intermediate metabolic rate. Each adult tyrannosaur or giganotosaur would have required the equivalent of 2 large ornithopods or 1 subadult sauropod (~7000 kg) per year. The amount of food required to sustain a metabolism high enough to maintain core homeothermy in these animals is much smaller than their stomach volume. Thus, these dinosaurs could have eaten much more than required for maintenance of their metabolism. A large male African lion (Panthera leo) weighing 250 kg may eat up to 40 kg of meat in one meal (Walker’s Mammals of the World, 1991) which is enough meat to support it for 7-8 days. 6000-8000 kg theropods could have held 10 times this amount of food in their stomachs. Thus, these dinosaurs could have stored 2-3 weeks food requirements in their stomachs with a single large meal. A truly bradymetabolic tyrannosaur or giganotosaur needing but 2000 kg of meat per year could have survived on 5 good meals/year. Evidence of relatively unaltered bone in a tyrannosaur coprolite indicates that residence time of food in these theropod digestive tracts was short (Chin et al. 1998; Bartlett et al. 1998) suggesting that these dinosaurs ate more often than once every 2-3 weeks. Evidence of soft tissue in dinosaur coprolites suggest even shorter residence times (Bertrand 1903; Chin 1999). Greater food intake would have supported increased activity levels and or growth rates. An intermediate metabolic level has important implications for growing juvenile theropods. For example, 100 to 1000 kg growing theropods would have needed 1-6 kg of food/day while 2000-5000 kg individuals would have needed 10-16 kg/day. Much greater amounts of food would have been available to these individuals and could have been held in their digestive systems. Energy from greater food acquisition greater than required by field metabolism could be put directly into growth, as the maintenance of homeothermy would have allowed continuous annual growth. Thus, greater amounts of energy were available for growth in these theropods than is available for mammalian style endotherms. More continuous, rapid growth due to maintenance of homeothermy would have been available for these dinosaurs than is available for bradymetabolic ectotherms. Curry (1998) and Ricqles et al. (1998) have suggested extremely rapid growth rates for Apatosaurus (=75% adult size in 7-10 yrs) and Maiasaura (adult size in 8 yrs or less) based upon histological analyses. Intermediate metabolic rates suggested here, allowing the maintenance of core homeothermy at all sizes, would have supported similarly rapid growth rates for these theropods as is suggested for Apatosaurus and Maiasaura. Intermediate metabolic rates are also consistent with the isotopic patterns of several herbivorous dinosaurs (Barrick et al. 1996, 1998). On the other hand, this intermediate metabolic strategy would have these individuals to survive longer periods during resource scarcity than true modern endotherms