Catastrophe triggers diversification


An analysis of more than 2,000 species of bird provides insight into how the animals' diverse beak shapes evolved, and points to a single rare event as a trigger for the rapid initial divergence of avian lineages. See Letter p.344

In 1796, the comparative anatomist Georges Cuvier described a huge ground-dwelling sloth that he named Megatherium. Such creatures no longer existed, and Cuvier considered that there was an ancient biotic world — or perhaps a succession of worlds — that preceded the modern tapestry of life on Earth1. He conceived of Earth's history as being marked by a series of pivotal events or catastrophes. This catastrophism theory ultimately fell before the intellectually nuanced Darwinian idea that changes in the make-up of Earth's flora and fauna have been continuous and uniform2. But a large-scale study of bird beak evolution by Cooney et al.3, on page 344 highlights the importance of rare and even single events in the history of life.

The authors were interested in rates of evolution, and how the 3D shapes of beaks diverged into the wide variety of present-day morphologies (how they filled the 'shape space'), during both the initial radiation of the avian lineage and the later evolution of individual lineages such as land birds, shorebirds and songbirds. Typically, data gathered on such a massive scale consist of simple traits or measurements, or, more commonly, of genetic sequences (which, as part of the structure of an organism, have themselves simple, linear, micro-scale morphology).

To generate the huge amount of complex 3D data required for their analyses, Cooney et al. scanned the beak surfaces of more than 2,000 bird species representing the full range of present-day beak shapes, using museum collections (Fig. 1). They undertook an innovative citizen-science project (www.markmybird.org), in which members of the public helped to prepare the scanned beaks for analysis by marking out specific features, or landmarks, such as the tip and midline of the bill. The authors then used DNA-based evolutionary trees of birds to interpolate ancestral shapes, working backward from the extant taxa.

Figure 1: The diverse beak shapes of extant birds such as (from left) eagles, flamingos and toucans.

From Left: Getty; Klein & Hubert/Nature Picture Library; Juan Carlos Vindas/NIS/Minden Pictures/Getty

Cooney et al.3 analysed the beak shapes of more than 2,000 extant avian species to investigate how beak shape has diverged during the evolution of birds.

The researchers predicted that their data would be consistent with the concept of quantum evolution — the idea that an initial radiation involves rapid divergence into new forms and functions4. For instance, such rapid divergence occurred during the Cambrian explosion of lineages of animals that have bodies showing bilateral symmetry, which began about 541 million years ago and lasted for 20 million to 25 million years5. Under this model, rates of beak-shape diversification would be fastest during the initial avian radiation6. Indeed, the authors found that most of the shape space filled rapidly during this initial burst. Coupled with a comprehensive study7 of avian relationships, which indicated that the radiation was associated with the catastrophic end-Cretaceous mass extinction 66 million years ago, these results support the idea that evolution is highly contingent on chance occurrences, marrying a Darwinian and a Cuvierian world view.

““Large-scale evolution is contingent on both history and catastrophe.””

Cooney and colleagues' data show some regularity as well — after the initial burst, rates of evolution within lineages are steady and predictable over time, which the authors suggest represents fine-tuning within particular ecological niches. Also predictable is how individual lineages evolved to fill part of the shape space; this tends to mirror the way in which the entire avian radiation fills the whole space. Therefore, although a cataclysm was perhaps the main event that precipitated avian diversification, a certain uniformity followed.

As with any pioneering effort, Cooney and co-workers' study raises questions and would benefit from refinement. Despite covering more than 70 million years of evolution, for instance, the authors did not analyse any fossils, and thus did not explicitly take into account the historical record of avian evolution. As they point out, though, most fossil bird beaks are crushed. Moreover, the fossil record, at least superficially, supports the idea that a diversity of beaks was present soon after the end-Cretaceous extinction8.

There are exceptions: the parrot lineage diverged early in avian evolution, but parrot-like beaks do not appear as fossils until about 20 million years ago9. In addition, some extinct birds have extreme beak shapes that fall into a shape space unoccupied by any living taxa — the huge carnivorous terror birds, for example, and the bizarre dromornithids. No form of modelling or simulation based only on extant species can detect a past decrease in shape diversity.

The nature of the constraints on beak shape is a fundamental and unanswered question. Are they developmental? Gene-regulatory networks that control beak shape are slowly being elucidated10. Small modifications in early embryonic molecular patterning can cause dramatic shifts in beak form5 — are these types of change responsible for the jumps in form and function associated with quantum evolution?

Finally, perhaps the most powerful message of the current study is that large-scale evolution is contingent on both history and catastrophe. The exact extent to which single, rare events such as extinctions define the course of life on Earth remains to be determined. These events are problematic for statisticians and for many scientists, who are used to phenomena that follow laws, as in physics and chemistry. Perhaps the scientific community needs to become more open to including single, descriptive analyses in its studies. Imagine if we had waited for dozens more individuals to be found after the first discoveries of early human ancestors: the Taung Child, whose 2.8-million-year-old skull was discovered in 1924; the early finds of the palaeontologist Robert Broom in the 1930s and '40s; and Lucy, found in 1974. We would today still be unable to fully discuss the deep origins of humanity among the African apes.

Maybe even phenomena that seem from a distance to be regular or predictable are actually — when analysed at a finer grain and with insight from the fossil record — products of a series of smaller, less easily detectable, rare events. These phenomena (such as the small fluctuations in beak shape within lineages inferred by Cooney et al.) might turn out differently were the evolutionary tape replayed11. Indeed, the existence of humanity is probably a rare event, as are the various ways in which humans have modified Earth, from climate change to megafaunal extinction. Perhaps the future is predictable only between one cataclysm and the next.Footnote 1


  1. 1.

    See all news & views


  1. 1

    Rowland, S. M. Geol. Soc. Am. Mem. 203, 225–246 (2009).

    Google Scholar 

  2. 2

    Baker, V. R. in Lyell: The Past is the Key to the Present (eds Blundell, D. J. & Scott, A. C.) 171–182 (Geol. Soc. Lond. Spec. Publ., 1998).

    Google Scholar 

  3. 3

    Cooney, C. R. et al. Nature 542, 344–347 (2017).

    ADS  CAS  Article  Google Scholar 

  4. 4

    Simpson, G. G. Tempo and Mode in Evolution (Columbia Univ. Press, 1944).

    Google Scholar 

  5. 5

    Bhullar, B.-A. S. et al. Evolution 69, 1665–1677 (2015).

    Article  Google Scholar 

  6. 6

    MacArthur, R. Theor. Popul. Biol. 1, 1–11 (1970).

    CAS  Article  Google Scholar 

  7. 7

    Prum, R. O. et al. Nature 526, 569–573 (2015).

    ADS  CAS  Article  Google Scholar 

  8. 8

    Mayr, G. Paleogene Fossil Birds (Springer, 2009).

    Google Scholar 

  9. 9

    Mayr, G. Zool. J. Linn. Soc. 136, 715–729 (2002).

    Article  Google Scholar 

  10. 10

    Mallarino, R. et al. Proc. Natl Acad. Sci. USA 109, 16222–16227 (2012).

    ADS  CAS  Article  Google Scholar 

  11. 11

    Gould, S. J. Wonderful Life: The Burgess Shale and the Nature of History (Hutchinson Radius, 1990).

    Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Bhart-Anjan S. Bhullar.

Related audio

Shamini Bundell hears how thousands of bird beaks bolster a theory about the avian family tree

Related links

Related links

Related links in Nature Research

Evolutionary biology: Speciation undone

Evolution: An avian explosion

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Bhullar, B. Catastrophe triggers diversification. Nature 542, 304–305 (2017). https://doi.org/10.1038/nature21494

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.