Soaring styles of extinct giant birds and pterosaurs

Yusuke Goto, Ken Yoda, Henri Weimerskirch, Katsufumi Sato

Posted on: 26 November 2020 , updated on: 17 December 2020

Preprint posted on 17 November 2020

Article now published in PNAS Nexus at

From flying aces to soar losers: In their new preprint, Goto et al predict that not all giant extinct fliers were equally adept in the air.

Selected by Sophia Friesen

Background and context:

When picturing a pterosaur, one of the winged reptiles that lived throughout most of the Mesozoic Era, most people imagine it in majestic flight, soaring high above the ancient landscape. Yet recent research by Goto et al. suggests that one of the largest of these animals, Quetzalcoatlus northropi, spent most of its time on the ground, only flying for short distances in emergencies. Quetzalcoatlus’ 10-meter wingspan was about three times that of the longest-winged living bird, the wandering albatross, but despite this, the researchers have determined that it was a poor flier. However, the smaller pterosaur Pteranodon, as well as two giant extinct birds, Pelagornis sandersi and Argentavis magnificens, were most likely adept at soaring on thermal updrafts.

The researchers came to these conclusions through a physics-based analysis of soaring ability for these four giant extinct fliers. Using a model of “dynamic soaring”, which today’s seabirds use to travel long distances by extracting energy from wind gradients over the ocean, as well as a more comprehensive analysis of “thermal soaring”, in which birds like eagles and vultures ride thermal updrafts, the researchers were able to predict the soaring abilities and required wind conditions for the four extinct fliers.


Key findings:

  1. Quetzalcoatlus and Argentavis were poor dynamic soarers, and the dynamic soaring ability of Pteranodon and Pelagornis depended on uncertain morphological variables

Dynamic soaring harnesses differences in wind speed at different altitudes over the sea surface. Birds like albatrosses and petrels build up speed in high-velocity winds at higher altitude, then travel in a different direction in slower winds closer to the sea. By cycling between different wind speeds and adjusting their angles of flight, they can even fly upwind without flapping. The authors used a physics-based model of dynamic soaring, previously only applied to albatrosses (1) to predict how fast ancient fliers could fly using dynamic soaring, both upwind and downwind, and what wind conditions they would require.

Quetzalcoatlus and Argentavis required minimum wind speeds that were at least twice as fast as those for living dynamic soarers, and they were slower fliers under most wind conditions. The predicted soaring ability of Pteranodon varied depending on wing drag estimates: when a birdlike drag coefficient was used, Pteranodon’s thermal soaring speeds were predicted to meet or exceed modern seabirds’, although it was limited to a narrower set of wind conditions. However, wind-tunnel tests of reconstructed pterosaur wings have produced higher estimates of wing drag (2), and when the higher drag estimate was used, Pteranodon was predicted to be a poor dynamic soarer overall. Pelagornis’ predicted thermal soaring ability was strongly influenced by estimates of body mass; if it were heavier, it would have been a strong dynamic soarer in fast winds, but if it were lighter, it would be worse at dynamic soaring than living species.

  1. Except for Quetzalcoatlus, the ancient fliers were excellent thermal soarers

Modern birds such as hawks and condors use thermal soaring, circling in thermal updrafts to gain altitude, then gliding over and down to the next updraft. Thermal soaring ability includes two main aspects: how quickly a flier loses altitude as it glides in a straight line, and how quickly it loses altitude when circling (which can also be thought of as the upward wind speed required to keep the flier from sinking). All of the extinct fliers except for Quetzalcoatlus had circling abilities comparable to modern thermal soaring birds, and the straight-line gliding performance of Argentavis, Pelagornis, and Pteranodon was at least as good as living soarers. In fact, depending on uncertain estimates of wing drag and weight, Pelagornis and Pteranodon may have significantly exceeded modern birds in straight-line gliding ability.

Quetzalcoatlus, on the other hand, was predicted to lose altitude in circling flight much more quickly than the living birds tested, including the kori bustard, a mostly terrestrial bird which flies only rarely. This very poor circle-gliding ability would most likely prohibit Quetzalcoatlus from thermal soaring. Given that this extremely large pterosaur was likely unable to perform either type of soaring, and that its flapping flight would have been quite metabolically costly (3), the authors predict that Quetzalcoatlus spent most of its time on the ground.


Why I liked this paper:

As a non-paleontologist, my primary experience of extinct creatures is through the visually detailed reconstructions that are common in popular media. Especially in the era of CGI, it’s too easy to interact with these nearly photorealistic images as if they were live footage, taking for granted that they are accurate. I enjoyed this paper because it made me reconsider the massive amount of work that goes into a well-considered reconstruction. These creatures only exist on the earth as fossilized bones; everything we know about the way they actually lived is based in the kind of painstaking analysis exemplified by this paper. I also appreciate how the researchers went beyond previous studies in taking into account the physical effects of a wider variety of morphological variables that influence flight. This includes, for instance, how wing length would limit how closely a soaring flier could approach the ocean’s surface, as well as recent estimates of wing drag (2) and body weight (4). Changes in these interrelating variables lead to vastly different predictions of soaring ability.


Questions for the authors:

  1. It’s surprising to me that we still don’t know the shape of the wind gradient over the ocean. How is this usually measured?
  2. You mention that the model of dynamic soaring gives unrealistically high maximum soaring speeds for some wind speed values. Is this cause for concern that the model isn’t predictive of real flight performances? Are the wind speeds at which the simulation becomes problematic likely in real-world conditions?
  3. The more recent estimates of pterosaur body weight are around three times previous ones, which makes a huge difference to flight ability. Why are these newer estimates so dramatically different?



  1. Sachs G (2005). Minimum shear wind strength required for dynamic soaring of albatrosses. Ibis 147(1): 1-10.
  2. Palmer C (2011). Flight in slow motion: aerodynamics of the pterosaur wing. Proc R Soc B Biol Sci 278(1713): 1881-1885.
  3. Witton MP, Habib MB (2010). On the size and flight diversity of giant pterosaurs, the use of birds as pterosaur analogues and comments on pterosaur flightlessness. PLoS One 5(11): e13982.
  4. Henderson DM (2010). Pterosaur body mass estimates from three-dimensional mathematical slicing. J Vertebrate Paleontol 30(3): 768-785.



Read preprint (2 votes)

Author's response

Yusuke Goto shared

It’s surprising to me that we still don’t know the shape of the wind gradient over the ocean. How is this usually measured?

Under the assumption that the sea surface is flat, the theory proposes that the wind gradient should follow the logarithmic function. This relationship seems to prevail at altitudes greater than a few tens of meters above the sea surface. However, at sea level, where birds normally fly, the sea surface is not flat due to the presence of waves; hence, the assumption of a logarithmic wind gradient does is not applicable. The measurement of the wind gradient with waves was performed in the laboratory (1); to the best of my knowledge, there have been no field measurements of wind gradient at altitudes at which birds normally fly (a few meters above sea level).

You mention that the model of dynamic soaring gives unrealistically high maximum soaring speeds for some wind speed values. Is this cause for concern that the model isn’t predictive of real flight performances? Are the wind speeds at which the simulation becomes problematic likely in real-world conditions?

It is a very important question! In our model of dynamic soaring, the maximum speed the animal could achieve was very high. Depending on the shape of the wind speed gradient assumed in our model, the animals reached maximum speeds of over 100 km/h even at realistic wind speeds of 5–10 m/s (Fig. 4), however, the average speeds reported using GPS for albatrosses and shearwaters were about 30–55 km/h (2, 3). There are two potential reasons for the discrepancy between this model and reality.

The first reason is that the actual shape of the wind speed gradient is likely to be more gradual than that assumed in the present study. For example, it can be seen from our results in a sigmoidal wind gradient (Fig. 4 B and C) that the gentler the wind speed variation (i.e., the larger the δ), the slower the maximum travel speed. Although little is known about the shape of the wind gradients experienced by actual birds as mentioned in the previous paragraph, the actual wind speed gradient may be closer to a sigmoidal form, for example, with a larger value of  than assumed.

The second potential reason is that the cost of rolling was not taken into account. Dynamic soaring puts the wings at risk of breaking under the influence of the high moment of force. Accordingly, the wing strength may limit the speed of change in bank angle and the resulting flight speed of animals. This would be particularly detrimental for species with long wingspans, such as the wandering albatross and the extinct giants. Indeed, reinforcing wing strength is also an important issue when designing a UAV for dynamic soaring (4). In view of the above two points, the dynamic soaring performance and the required wind conditions obtained from the present numerical calculation should not be taken literally by the values themselves. However, since all species were evaluated under the same assumptions, the relative values are meaningful indicators for the purpose of estimating soaring style by inter-species comparison, as was done in this study. A more refined analysis incorporating constraints on body rolling speed will be an interesting challenge in the future. For this purpose, actual measurements of bank angles in dynamic soaring birds and assessment of wing bone strength in extant and extinct species will provide important information.

The more recent estimates of pterosaur body weight are around three times previous ones, which makes a huge difference to flight ability. Why are these newer estimates so dramatically different?

It is more likely that previous estimates of the body weight of pterosaurs were extremely low. Quetzalcoatlus, for example, was previously estimated to have a wingspan of approximately 10 m and a body mass of approximately 70 kg, which would make its body density 1/3 to 1/2 that of modern birds and bats. Such a low estimate was based on the fact that pterosaur skeletons have air-filled structures. However, it was later indicated that pneumatic skeletons are not necessarily related to light bone mass. As I am not a paleontologist, I have limited knowledge about this aspect. Please refer to chapters 4 and 6 of (5) for more detailed explanations by paleontologists.

More recent studies in the 2000s have used different approaches to the estimation of body weight, including investigating the relationship between dry skeletal mass and body mass that is consistent across birds and mammals (6), the allometric relationship between weight and wingspan obtained from seabirds (7), and the construction of the three-dimensional contours of pterosaurs (8). These body mass estimates showed body densities closer to those of extant birds than previous estimates.


  1. M. P. Buckley, F. Veron, Structure of the airflow above surface waves. J. Phys. Oceanogr. 46, 1377–1397 (2016).
  2. C. Péron, K. Delord, R. A. Phillips, Y. Charbonnier, C. Marteau, M. Louzao, H. Weimerskirch, Seasonal variation in oceanographic habitat and behaviour of white-chinned petrels Procellaria aequinoctialis from Kerguelen Island. Mar. Ecol. Prog. Ser. 416, 267–284 (2010).
  3. Y. Yonehara, Y. Goto, K. Yoda, Y. Watanuki, L. C. Young, H. Weimerskirch, K. Sato, Flight paths of seabirds soaring over the ocean surface enable measurement of fine-scale wind speed and direction. Proc. Natl. Acad. Sci. U. S. A. 113, 9039–9044 (2016).
  4. I. Mir, A. Maqsood, S. A. Eisa, H. Taha, S. Akhtar, Optimal morphing–augmented dynamic soaring maneuvers for unmanned air vehicle capable of span and sweep morphologies. Aerosp. Sci. Technol. 79, 17–36 (2018).
  5. M. P. Witton, Pterosaurs: natural history, evolution, anatomy (Princeton University Press, 2013).
  6. M. P. Witton, A new approach to determining pterosaur body mass and its implications for pterosaur flight. Zitteliana, 143–158 (2008).
  7. K. Sato, K. Q. Sakamoto, Y. Watanuki, A. Takahashi, N. Katsumata, C.-A. Bost, H. Weimerskirch, Scaling of soaring seabirds and implications for flight abilities of giant pterosaurs. PLoS One. 4, e5400 (2009).
  8. D. M. Henderson, Pterosaur body mass estimates from three-dimensional mathematical slicing. J. Vertebr. Paleontol. 30, 768–785 (2010).


Have your say

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Sign up to customise the site to your preferences and to receive alerts

Register here

Also in the animal behavior and cognition category:

Blue appendages and temperature acclimation increase survival during acute heat stress in the upside-down jellyfish, Cassiopea xamachana

Megan E. Maloney, Katherine M. Buckley, Marie E. Strader

Selected by 30 April 2024

Maitri Manjunath

Animal Behavior and Cognition

Experience-dependent plasticity of a highly specific olfactory circuit in Drosophila melanogaster

Benjamin Fabian, Veit Grabe, Rolf G. Beutel, et al.

Selected by 16 February 2024

T. W. Schwanitz


Behavioral screening of conserved RNA-binding proteins reveals CEY-1/YBX RNA-binding protein dysfunction leads to impairments in memory and cognition

Ashley N Hayden, Katie L Brandel, Paul R Merlau, et al.

Selected by 15 February 2024

Chee Kiang Ewe

Animal Behavior and Cognition

Also in the biophysics category:

Topology changes of the regenerating Hydra define actin nematic defects as mechanical organizers of morphogenesis

Yamini Ravichandran, Matthias Vogg, Karsten Kruse, et al.

Selected by 08 May 2024

Rachel Mckeown

Developmental Biology

Structural basis of respiratory complexes adaptation to cold temperatures

Young-Cheul Shin, Pedro Latorre-Muro, Amina Djurabekova, et al.

Selected by 10 April 2024

Pamela Ornelas


Actin polymerization drives lumen formation in a human epiblast model

Dhiraj Indana, Andrei Zakharov, Youngbin Lim, et al.

Selected by 05 April 2024

Megane Rayer, Rivka Shapiro


Also in the paleontology category:

Hidden limbs in the “limbless skink” Brachymeles lukbani: developmental observations

Daniel Smith-Paredes, Oliver Griffith, Matteo Fabbri, et al.

Selected by 03 February 2021

Alexa Sadier

Developmental Biology

Soaring styles of extinct giant birds and pterosaurs

Yusuke Goto, Ken Yoda, Henri Weimerskirch, et al.

Selected by 26 November 2020

Sophia Friesen