Spearthrower wrote:You are quite right - there isn't one. I think the general habit has been to post palaeontological finds in the general Biological Sciences forum.

Spearthrower wrote:There's a whole new system of classification just begging to be explored there!

Stuff in the ground: Terraology - including; lava, worms, tunnels, and treasure.
Stuff in the air: Aeolology - including; birds, planes, giraffe's heads, and pinged rubber bands.
Stuff in the sea: Mareology - including; fish, dumped cars, concrete boots, and forgotten snorkellers.

Wroe et al, 2004 wrote:We show that at 2786 kg, the largest known marsupial, Diprotodon optatum, was much larger than has previously been suggested. Our results contradict the conclusion that the maximum attainable body mass of an Australian marsupial has been constrained by low productivity.

Wroe et al, 2004 wrote:1. INTRODUCTION

Body mass is fundamental to interpretations of biological patterns and its reliable prediction for fossil species has become increasingly important in the analysis of past ecosystems ( Janis 1990; Van Valkenburgh 1990; Alexander 1998; Fariña et al. 1998; Wroe et al. 2003a). Historically, estimates for body mass in fossil taxa have been determined subjectively, often producing widely divergent figures that can vary by an order of magnitude (Wroe 2002; Wroe et al. 2003a). Consequently, palaeoecologists have increasingly turned to quantitative approaches. Empirical methods, commonly based on regressions of craniodental and body mass data in extant taxa, are now available for many fossil placentals ( Janis 1990; Van Valkenburgh 1990; Christiansen 1999a). In recent years, predictive equations for fossil marsupials have also been developed (Myers 2001; Wroe et al. 2003a). However, because living marsupials do not exceed ca. 85 kg, the efficacy of methods derived from craniodental data is questionable where this figure is much surpassed (Wroe et al. 2003a). This is especially problematic regarding the largest known marsupial, Diprotodon optatum, and the body mass of this species is of particular significance because it strongly impacts on interpretations of Australian prehistory and ecology (Milewski & Diamond 2000; Burness et al. 2001; Wroe et al. 2003a,b).

A recent study wherein the body masses of the single largest herbivore and carnivore species on various landmasses were regressed against landmass area, found that D. optatum was unexpectedly small (Burness et al. 2001). This finding incorporated a subjectively determined estimate for mean body mass of 1175 kg. Burness et al. (2001) concluded that uniquely low productivity had probably constrained the maximal body mass attainable by an Australian marsupial, a proposition first made by Flannery (1994), who further posited that low productivity had generally limited mammalian body masses on the island continent. Body mass also correlates negatively with population size and fecundity, both of which impact on assessments of vulnerability to climatically or anthropogenically driven extinction (Johnson 2002).

Subjective inferences of body mass and comparisons of general morphology in D. optatum range widely, from comparisons with bullocks (Long et al. 2003) to rhinoceroses (Archer et al. 1994). A mounted Australian Museum specimen has a head–body length of 3.7 m (A. Musser, personal communication). In life, this animal would have exceeded 4 m because cartilaginous tissue, which is lost in fossils, amounts to ca. 20% of pre-sacral vertebral column length (Finch & Freedman 1986). However, even 3.7 m exceeds the head–body length of any extant bovid (Nowak & Paradiso 1983). Diprotodon was massive, and Hippopotamus or rhinoceros species are more appropriate analogues. Maximal head–body length and body mass in the hippopotamus (Hippopotamus amphibius) are 4.6 m and 4500 kg, respectively (Nowak & Paradiso 1983). For the largest rhinoceros (the white rhinoceros, Ceratotherium simum), these dimensions are 3.77 m and 3600 kg (Groves 1972). The mean body masses are 1405 kg for H. amphibius (Smithers 1983) and 2000 kg for C. serum (Bourlière 1965). With a mean body mass of 1000 kg, the black rhinoceros, Diceros bicornis, is closest to the estimate used by Burness et al. (2001). This smaller species has a head–body length of 2.80–2.90 m, and a maximum body mass of 1300 kg (Happold 1987; Hillman-Smith & Groves 1994).

2. MATERIAL AND METHODS

Body mass predictions founded on minimum mid-shaft circumferences of the femur and humerus (Ch+f) offer greater accuracy than those using craniodental data and are less constrained by phylogenetic differences (Anderson et al. 1985). To estimate mean body mass in D. optatum, we measured Ch+f in 18 quadrupedal marsupials of known body mass and combined these data with those taken for 32 placentals that ranged up to 5879 kg (Anderson et al. 1985; see electronic Appendix A available on The Royal Society’s Publications Web site). We then generated a predictive equation using Model I regression and holding body mass as the dependent variable (figure 1). A smearing estimate (SE) was calculated to correct log-transformed results for transformation bias (Smith 1993). To test for phylogenetic effects, we also compared the relationship between Ch+f and body mass in 17 quadrupedal marsupials with that of 15 placentals within the same size range (less than 44 kg). Slopes derived from the regressions of log-transformed marsupial and placental data were compared using Student’s t-test (Zar 1984). Our estimate of mean body mass in D. optatum (n = 17; see electronic Appendix B) was compared with predicted maximal mean body mass (MMBM) for endothermic herbivores based on landmass area, i.e. MMBM (endothermic herbivore) = 0.47 × (landmass area)0.52 (Burness et al. 2001).

3. RESULTS

The mean Ch+f for D. optatum was 530 mm (s.e.m. = 1.05). At 2786 kg, our resultant prediction of mean body mass greatly exceeds that of previous estimates (95% CI of 3417 kg to 2272 kg). The average Ch+f in D. optatum was much larger that that of the two closest individual extant placentals, an H. amphibius (Ch+f = 417 mm, body mass = 1950 kg) and an American bison (Bison bison) (Ch+f = 359 mm, body mass = 1179 kg). This is also much greater than mean Ch+f for C. simum (455 mm, n = 7; see Christiansen 1999b), an animal with an average adult body mass of 2000 kg.

Slopes were significantly different between marsupials and placentals with masses of less than 44 kg (t = 3.389, d.f. = 29, p < 0.005). Because the slope for marsupials was higher (3.32 versus 2.74), we infer that methods incorporating data from placentals, such as those presented here, may underestimate body mass in marsupials (figure 2).

The predicted MMBM for an Australian endothermic herbivore based on landmass area (7 682 000 km2) was 1788 kg. Marsupials consume ca. 20% less food than placentals of equal body mass (Burness et al. 2001). Correcting for this lower food intake gives a predicted MMBM of 2235 kg for Australian marsupials. After allowing for lower consumption, D. optatum is 25% larger than expected. Operating on the same premise, consideration of 95% confidence limits places the mean body mass in D. optatum at between 53% and 2% higher than predicted on the basis of landmass area.

4. DISCUSSION

We conclude that body mass in the Late Pleistocene giant, D. optatum, has previously been underestimated. These findings contradict the assertion that uniquely low productivity has constrained the MMBM attainable by Australian marsupials, but marginally strengthen the correlation between MMBM of endothermic herbivores and landmass area (r2 changes from 0.74 to 0.75).

The relationship between productivity and body mass is not necessarily simple or linear. Large body mass can be a response to highly seasonal, relatively unproductive conditions (Owen-Smith 1988). Similarly, the relationship between species richness and productivity can be linear, bimodal or unimodal (Wroe 2002). Discovering whether productivity or other uniquely Australian influences have more generally limited the body masses of the continent’s marsupials will require further empirical tests

Helgen et al, 2006 wrote:Abstract. A method, based on femoral circumference, allowed us to develop body mass estimates for 11 extinct Pleistocene megafaunal species of macropodids (Protemnodon anak, P. brehus, P. hopei, P. roechus, Procoptodon goliah, ‘P.’ gilli, Simosthenurus maddocki, S. occidentalis, Sthenurus andersoni, S. stirlingi and S. tindalei) and three fossil populations of the extant eastern grey kangaroo (Macropus giganteus). With the possible exception of P. goliah, the extinct taxa were browsers, among which sympatric, congeneric species sort into size classes separated by body mass increments of 20–75%. None show evidence of size variation through time, and only the smallest (‘P.’ gilli) exhibits evidence suggestive of marked sexual dimorphism. The largest surviving macropodids (five species of Macropus) are grazers which, although sympatric, do not differ greatly in body mass today, but at least one species (M. giganteus) fluctuated markedly in body size over the course of the Pleistocene. Sexual dimorphism in these species is marked, and may have varied through time. There is some mass overlap between the extinct and surviving macropodid taxa. With a mean estimated body mass of 232 kg, Procoptodon goliah was the largest hopping mammal ever to exist.

Helgen et al, 2006 wrote:On the basis of a regression of femoral circumference and body mass data derived from a large number of extant macropodids, here we estimate body mass for four Late Pleistocene species of Protemnodon, three of Sthenurus, two of Simosthenurus and two of Procoptodon (including P. goliah and ‘P.’ gilli, the largest and smallest of the megafaunal kangaroos, respectively), as well as multiple, temporally separated Quaternary samples of the eastern grey kangaroo (Macropus giganteus).

Materials and Methods

Previous estimates of body mass for megafaunal kangaroos generally derive from extrapolations based on univariate skull and dental measurements (Flannery 1980; Murray 1991), but these types of comparisons are potentially unreliable (Myers 2001; Wroe et al. 2003, 2004). Cranial metrics may be determined and/or tightly constrained by feeding requirements, and dental dimensions are particularly problematic for macropodids, in which molars erupt continually and are successively shed. In contrast, load-bearing long bones (such as the femur or humerus) exhibit similar scaling relationships across unrelated taxa and over a wide range of body sizes and are considered more appropriate for body mass extrapolations (Anderson et al. 1985; Anyonge 1993; Reynolds 2002). Predictions based on mid-shaft circumferences of load-bearing long bones are particularly powerful (Anderson et al. 1985; Wroe et al. 2004), and because kangaroos are bipedal, femoral dimensions are especially appropriate. For this study, we measured the least femoral circumference (c), i.e. the circumference of the femur immediately distal to the third trochanter (Fig. 1), in 107 wild-collected macropodid museum specimens with associated body mass data (see Appendix 1). These femoral and body mass data relate to adult, subadult and juvenile animals representing 26 extant species in nine genera, and span the full spectrum of lifestyle and body size variation among living macropodids (Appendix 1). Captive-living animals (potentially overweight and poorly mobile) and very young pouch young (those too young to hop) were excluded from our analyses. The maturity of specimens was evaluated by comparative size, degree of ephiphyseal fusion, and by comparing the ossification and dental development of associated crania when available.

We log10-transformed these data for extant macropodids to generate a regression equation that describes the relationship between actual body mass (Mobs) and c for our dataset of extant macropodid specimens (Fig. 2):

lg(Mobs) = 2.5932[lg(c)] – 3.2842. (1)

In retransforming these logarithmic values to actual values for body mass, we corrected for the effects of logarithmic transformation bias by multiplying all predicted values by a ‘smearing estimator’(Smith 1993), in this case corresponding to a value of 1.0146.We have used the resulting equation to derive estimated body masses (Mest) of megafaunal kangaroo specimens for which femora are available (Appendix 2):

Mest = 1.0146 × 10[2.5932 lg(c) – 3.2842].

In using this equation for extrapolative estimations, we have made the assumption that limb–body postures of extinct kangaroos were similar to those of extant kangaroos, and that cross-sectional geometry of the femur scales at a similar rate in all taxa examined.