Femoral metadiaphyseal and nutrient foramen perfusion suggests comparable maximal metabolic rates in a pterosaur and in a semi-aquatic maniraptoran dinosaur

Introduction

Maximal metabolic rate (MMR), also referred to as the maximal rate of oxygen consumption (VO₂max), represents the highest aerobic metabolic rate achieved during exercise and defines the upper limit of sustained metabolic performance (Bennett & Ruben, 1979; Weibel et al., 2004). In extant amniotes, MMR can be measured directly using a circuit respirometer during physical activity. For volant taxa, this involves measurements during forced flight (Hails, 1979; Norberg, 1996), while for terrestrial species, metabolic rates are assessed while running on a treadmill (Fedak, Pinshow & Schmidt-Nielsen, 1974; Taylor, Heglund & Maloiy, 1982; Seymour, Runciman & Baudinette, 2008). For swimming amniotes, such as ducks, penguins and sea turtles, MMR is measured in controlled-flow swim channels (Prange & Schmidt-Nielsen, 1970; Prange, 1976; Kooyman & Ponganis, 1994).

Reconstructing the maximal aerobic capacity of extinct archosaurs has posed significant challenges, leading to the development of an indirect method, referred to as the “foramen technique”; developed by Seymour et al. (2012), this technique has proven to be particularly effective (Knaus et al., 2021; Varela, Tambusso & Fariña, 2024). This approach estimates regional blood flow by quantifying the cross-sectional aperture area of vascular foramina in long bones. Long bones require blood perfusion for remodeling to e.g., repair microfractures induced by locomotion and weight-bearing stresses (Lieberman et al., 2003; Eriksen, 2010). Nutrient foramen size provides a useful proxy for bone perfusion and aerobic capacity (Seymour et al., 2012). A nutrient artery enters the femoral shaft through the nutrient foramen, typically accompanied by a vein (Currey, 2002). The size of the foramen correlates dynamically with the size of the vessels it contains. Thus, the MMR scales with the size of the nutrient artery and the corresponding blood flow supplying the bone (Seymour et al., 2019). In mature terrestrial vertebrates, femoral blood flow is known to be positively correlated with locomotor activity, with relatively larger nutrient foramina being detected in species with elevated metabolic rates during locomotion (Seymour et al., 2012; Allan et al., 2014; Newham et al., 2020). According to Seymour et al. (2012), the proximate causation of this relationship would be linked to the fact that tetrapods with high levels of activity encounter biomechanical constraints invoking microfractures that are repaired by secondary bone remodeling, thus necessitating high levels of oxygen consumption. The application of phylogenetic eigenvector maps (PEMs; Guénard, Legendre & Peres-Neto, 2013) to reconstruct metabolic rates in extinct amniotes has been adopted over the last decade (Legendre et al., 2016; Fleischle, Wintrich & Sander, 2018; Cubo et al., 2024). PEMs provide a framework to infer the metabolic profiles of extinct organisms by integrating phylogenetic and morphological data. In combination with the foramen technique, PEMs have been used to estimate mass-independent MMR in fossil archosaurs, such as extinct pseudosuchians (Cubo et al., 2024; Sena et al., 2023; Sena et al., 2025).

In this study, we used these methods by integrating a comprehensive phylogenetic amniote database with their corresponding nutrient artery blood flow measurements computed using the foramen technique and previously measured MMR values. This combined approach was applied to retrodict the mass-independent MMR of two extinct ornithodirans with extremely different lifestyles and locomotory strategies: Rhamphorhynchus and Halszkaraptor escuilliei. Rhamphorhynchus was a medium-sized, piscivorous pterosaur from the Late Jurassic Plattenkalks of southern Germany. Characterized by a prow-shaped lower jaw and forward-angled (procumbent) teeth, Rhamphorhynchus was adapted to forage extensively in aquatic environments (Voeten et al., 2018; Witton, 2018). H. escuilliei was late Cretaceous dromaeosaurid theropod hypothesized to possess an amphibious eco-morphology, potentially relying on neck hyper-elongation for predatory foraging (Cau et al., 2017; Cau, 2020). In this research, we aimed to elucidate the metabolic capacities of these two locomotory diverse ornithodirans.

Materials & Methods

Results

Based on their metadiaphyseal and nutrient foramen dimensions, the pterosaur Rhamphorhynchus sp. (MdC 20269891) and the theropod H. escuilliei (MPC D-102/109) had comparable calculated nutrient artery flow rates ($\dot{\mathrm{}Q}$) of 0.0017 ml s⁻¹ and 0.0010 ml s⁻¹, respectively (Fig. 1).

The retrodicted mass-independent MMR for the “long-tailed” pterosaur Rhamphorhynchus sp. is 5.55 mLO₂ h⁻¹ g^−0.87 (95% confidence interval CI [4.37–7.05] mLO₂ h⁻¹ g^−0.87) while that for the theropod H. escuilliei (MPC D-102/109) is 5.68 mLO₂ h⁻¹ g^−0.87 (95% CI [4.44–7.26] mLO₂ h⁻¹ g^−0.87). The predictive model included the phylogenetic eigenvectors 1, 2, 3, 4, 5, 6, 8, 9, 20, 22, 23, 27, 28, and 41, and the estimated nutrient artery $\dot{\mathrm{}Q}$ as the co-predictor (adjusted R² = 0.90; Akaike Information Criterion corrected (AICc) = 56.79; and p = 3.501e⁻¹²) (Fig. 2). The leave-one-out cross-validation test identified no significant difference between the predicted and observed values for the extant species (p = 0.7788).

Discussion

The dromaeosaurid H. escuilliei (specimen MPC D-102/109) is considered to be a mallard-sized sub-adult theropod, estimated to have a body mass of 1.5 kg (Cau et al., 2017). The Rhamphorhynchus specimen analyzed in this study weighed approximately 95 g and was identified as a juvenile based on its small size (Voeten et al., 2018) with a wingspan ranging from 45 to 50 cm, based on the work described previously by Prondvai et al. (2012). Despite their distinct eco-morphologies, these two ornithodirans exhibited a similar mass-independent MMR and femoral nutrient artery $\dot{\mathrm{}Q}$. These data suggest a strong influence of phylogenetic constraints on the outcomes of predictive energy modeling.

We here adopted the morphofunctional interpretation proposed by Cau et al. (2017), proposing that Halszkaraptor engaged in subaqueous foraging behavior. Certain pivotal anatomical adaptations of Halszkaraptor are comparable to those of extant sawbills, which are considered their ecological analogs (Cau, 2020). However, the ecological interpretation of Halszkaraptor remains debated. For instance, Brownstein (2019) reinterpreted its morphological features as indicative of a transitional form between non-paravian maniraptorans and more derived dromaeosaurids (but also consider Cau, 2020). Furthermore, Fabbri et al. (2022) argued that the lack of microanatomical specialization typically associated with semi-aquatic habits may suggest an ambiguous lifestyle for halszkaraptorines. Nevertheless, irrespective of proposed aquatic habits of Halszkaraptor, theropods in general (Legendre et al., 2016), and the dromaeosaur Velociraptor in particular (Pontzer, Allen & Hutchinson, 2009), are generally considered metabolically active taxa, reinforcing H. escuilliei as a valid model for suspected elevated dinosaur metabolic capacity.

Our mass-independent MMR estimations for H. escuilliei and Rhamphorhynchus align with those measured previously in Varanus gouldii, a predatory monitor lizard known for its enhanced oxygen transport capabilities, and certain avian taxa, such as emus (Dromaius novaehollandiae), gulls (Larus), and terns (Sterna). Monitor lizards achieve aerobic scopes and sustainable running speeds more than twice those of similar sized, through adaptations, and that include an increased lung surface area, blood buffers, and myoglobin-rich skeletal muscles (Bartholomew & Tucker, 1964; Bennett, 1972; Bennett, 1973). These features underline the importance of morphological investment in oxygen transport systems for enabling high aerobic performance (Bennett & Ruben, 1979).

As obligate cursors, flightless ratite emus possess increased hindlimb and heart mass relative to habitual cursorial and incidental burst flying phasianids to facilitate sustained locomotor activity (Hartman, 1961; Grubb, Jorgensen & Conner, 1983). The VO₂max of exercising emus is tenfold higher than their resting values (Grubb, Jorgensen & Conner, 1983). Conversely, non-migratory cursorial phasianids, such as guinea fowl (Numida meleagris) and jungle fowl (Gallus), are adapted for rapid escape flights, maximizing takeoff performance when avoiding predation (Witter, Cuthill & Bosner, 1994). These species exhibit lower endurance and aerobic performance due to the limited aerobic capacity of flight muscle (Kiessling, 1977; Ellerby et al., 2003; Askew & Marsh, 2001; Henry, Ellerby & Marsh, 2005). The contribution of flight muscles to organismal aerobic scope in these birds, even when combined with leg activity, is small (Hammond et al., 2000). Halszkaraptor, interpreted as an amphibious dromaeosaurid, likely combined vigorous hindlimb activity for terrestrial locomotion, similar to flightless ratites, with the use of its forelimbs for swimming. The plesiomorphic glenoid condition characteristic of paravians, which is also inferred for Halszkaraptor, may have served as a potential exaptation for a forelimb-assisted swimming style (Cau et al., 2017; Cau, 2020). However, its aerobic capacity appears to be lower than that of highly specialized swimmers, such as ducks (Anas spp.). This suggests that Halszkaraptor might be less adapted to sustain longer bouts of swimming compared to mallards.

Rapid growth rates in immature archosaurians have been observed from the Triassic period onward (Curry Rogers et al., 2024), which corroborated the active growing of the subadult Halszkaraptor individual showing a moderate growth rate (D’emic et al., 2023). It also highlights how the generally rapid juvenile growth across archosaurs, as exemplified by Rhamphorhynchus MdC 20269891, contributes to overestimation of mass-independent MMR (Cau et al., 2017; Prondvai et al., 2012; Araújo et al., 2023). Previous analyses of Rhamphorhynchus suggested that flight capability was likely achieved when individuals reached approximately 30–50% of the adult wingspan and 7–20% of the adult body mass (Prondvai et al., 2012). Recent findings indicate that the largest known specimen of Rhamphorhynchus had an estimated wingspan of 1.8 m and a body mass of approximately 3 kg (Hone & McDavid, 2025). Despite this, evidence of isometric growth and a piscivorous diet in this taxon supports the hypothesis of precocial flight capabilities (Voeten et al., 2018; Hone et al., 2021). The high mass-independent maximum metabolic rate (MMR) inferred by the present study for our small Rhamphorhynchus specimen, representing approximately 3% of the adult body mass, further suggests that powered flight may have been possible even at an early ontogenetic stage. Juveniles often exhibit relatively larger foramen areas due to the increased perfusion required for rapid growth. For example, growing kangaroos possess larger femoral nutrient foramina areas than adults, thus reflecting their higher energy demands for bone development (Hu et al., 2018). A similar pattern has been observed in egg-forming oviparous females, in which increased femoral blood flow supports the mobilization of calcium for eggshell formation (Hu, Nelson & Seymour, 2021). Thus, the high MMR identified for Rhamphorhynchus in this study is more likely associated with its growth demands. Furthermore, this juvenile Rhamphorhynchus could have been able to climb trees (Prondvai et al., 2012), thus suggesting a high level of ecological activity during early ontogeny, therefore reflecting the energetic demands associated with active exploratory behavior.

As the earliest archosaur group to evolve flapping flight, pterosaurs faced the energetic costs associated with aerial locomotion; this was particularly the case for pterodactyloids with a large body size, such as Quetzalcoatlus northropi. The flight energetics of large pterosaurs, such as Q. northropi, probably involved anaerobic metabolism during takeoff, followed by energy-efficient soaring or high-speed flight (Marden, 1994). To sustain flight aerobically, Quetzalcoatlus likely required an aerobic scope comparable to migratory swans (Marden, 1994). In contrast, smaller species of pterosaur, such as Rhamphorhynchus, might have had different metabolic requirements during takeoff, highlighting the need for further studies on adult specimens to elucidate their flight endurance and takeoff capabilities.

In volant birds, flying is generally more energetically efficient than running or swimming over equivalent distance due to the higher speeds achieved in flight (Fedak, Pinshow & Schmidt-Nielsen, 1974; Norberg, 1996; Butler, 2016). This aerodynamic efficiency is particularly pronounced in insectivore birds, which exhibit flight costs 49–73% lower than foraging costs measured for non-aerial species (Hails, 1979; Tucker, 1970). Consequently, the maximal aerobic capacity of adult Rhamphorhynchus is expected to be lower than of the juvenile specimen analyzed in this current study.

The mass-independent MMR of juvenile Rhamphorhynchus was comparable to that of migratory shorebirds, such as gulls and terns, although despite its elevated aerobic capacity, the juvenile pterosaur exhibits lower mass-specific power output for sustained vertical takeoff when compared to similarly sized migratory phasianids, such as Coturnix coturnix and Coturnix chinensis (Bishop, 1997; Askew & Marsh, 2001; Henry, Ellerby & Marsh, 2005). Further investigation of adult specimens is required to clarify ontogenetic influence on the flight performance of pterosaurs.

Although Rhamphorhynchus and H. escuilliei belonged to distinct groups and exhibit different lifestyles, parallels can be drawn with regards to their metabolism and energy efficiency in locomotion. Rhamphorhynchus would achieve a greater energy efficiency for flight in adults, although juvenile stages exhibited elevated metabolic rates due to rapid growth. On the other hand, H. escuilliei, with its semi-aquatic locomotion, would have adapted its metabolism for endurance rather than high bursts of energy, as seen in flight. Both groups demonstrated adaptations that reflect the need for metabolic capacity, albeit in different ways: while the aerobic energy costs of flight would be high, these would be offset by the flight efficiency in adult Rhamphorhynchus individuals, whereas the locomotion of Halszkaraptor would be more balanced, with lower energy expenditure during terrestrial and aquatic excursions.

Conclusions

Our analyses highlight the aerobic capacity of the theropod H. escuilliei and the pterosaur Rhamphorhynchus. Despite significant differences in lifestyle and ontogenetic state, both taxa exhibited similar mass-independent MMRs and femoral nutrient artery blood flow ($\dot{\mathrm{}Q}$). The mass-independent MMR of H. escuilliei aligned with that of the monitor lizard V. gouldii and certain birds emus and the Charadriiformes Larus and Sterna. However, the aerobic performance of H. escuilliei appears to fall below that of swimmers such as ducks and cursorial mammals. The juvenile Rhamphorhynchus analyzed herein likely exhibited elevated femoral blood flow when compared to mature individuals. While its endurance flight capabilities resemble those of migratory shorebirds, it is far from those seen in small migratory quails, Coturnix. In addition, adult Rhamphorhynchus, experiencing reduced growth-related demands, likely had lower mass-independent MMRs and operated under a more energetically efficient metabolic regime. Nevertheless, research is needed to clarify the ontogenetic influence on their metabolic demands and capacity.