Assessing the relative importance of biological interactions (biotic factors) and changes in the physical environment (abiotic factors) in driving evolution has proved difficult, partly owing to the lack of high-quality data sets that combine both fossil remains and associated ancient environmental data. The few studies that have quantified the relative contributions of these two drivers describe a complex relationship1. However, on page 92, Žliobaitėet al.2 report an unexpectedly simple pattern of driver action in peak evolutionary success.
Between a species originating and becoming extinct, its evolutionary success can be measured in a number of ways, such as the extent of its geographical range. Such metrics often form a ‘hat-shaped’ curve, with a rise towards a central peak, followed by a decline to extinction. Why this pattern occurs so often and the degree to which biotic and abiotic factors influence this trajectory is a matter of debate.
To assess the relative role of biotic factors, such as competition between organisms, and abiotic factors in evolutionary trajectories, Žliobaitė and colleagues analysed the fossil record of large herbivorous mammals. This grouping offers several advantages for this type of analysis. For instance, the authors could solve the problem of finding consistent regional ancient environmental data because the height of these mammals’ teeth correlates strongly with characteristics of their environment, including precipitation levels3 and the amount of plant material in the ecosystem4. Without this measure, the authors would have had to rely on standard global measurements of environmental change, an approach that can mask substantial regional-level variation.
Another advantage is that these animals all share a similar ecological niche, so the authors could use the average number of genera (groups of closely related species) per locality to measure competition intensity, a biotic factor that is otherwise difficult to quantify. In addition, the fossil record of large mammalian herbivores is sufficiently rich that the proportion of localities in which each genus was found is a good measure of the evolutionary success of each genus.
A fundamental innovation in Žliobaitė and colleagues’ study was their analysis of not only rates of origination and extinction, but also the rates at which genera reached peak evolutionary success. Their decision to do this was partly motivated by a wish to investigate the apparent inconsistency between the law of constant extinction proposed5 by the evolutionary biologist Leigh Van Valen, and the observed prevalence of hat-shaped evolutionary trajectories. The contradiction arises because, for such trajectories, the probability of extinction increases with time6, whereas the law of constant extinction states that the extinction probability is independent of how old a species is.
Van Valen put forward the Red Queen hypothesis5 to explain the law of constant extinction. The hypothesis is named after the character in Lewis Carroll’s book Through the Looking-Glass, who states that “it takes all the running you can do, to keep in the same place”. The law of constant extinction fits with this model, in which species must continually evolve to maintain their existence in the face of changing biotic interactions with other organisms, because the probability of extinction does not change with time for species able to keep up with this evolutionary race.
The key finding made by Žliobaitėet al. is that the rate at which the number of genera reached their peak representation in the fossil record (and thus their peak success) strongly correlates with just the biotic competition intensity (Fig. 1). For extinction, the situation is more complex, with biotic competition and abiotic environmental drivers both correlating with extinction rate, although competition has less effect than it does for peak success. Notably, neither competition nor environmental change correlates strongly with the rates at which species first appeared in the fossil record. Other studies1,6 have found similar patterns associated with origination and extinction rates, but none of these earlier studies analysed peak success. The dominance of competition during the peak expansion of genera found by Žliobaitė and colleagues is an unexpectedly simple result, given the anticipated complexity of the ecological and evolutionary dynamics of species and species richness.
As anticipated for hat-shaped trajectories, Žliobaitėet al. show that extinction events in the history of large mammalian herbivorous genera do violate the law of constant extinction; the probability of extinction increases the older the genus is. However, the probability of when a genus will stop expanding, that is, when it will reach its peak evolutionary success, was found to be independent of the age of the genus. Thus, Žliobaitė and colleagues have discovered a phenomenon that is equivalent to the law of constant extinction in the way that it describes an evolutionary-trajectory pattern that is time independent, but that pertains to when genera reach peak success, and which they call the law of constant peaking.
This key discovery resonates with another of Van Valen’s ideas: that only the expansion phase of an evolutionary trajectory, which equates to the acquisition of more resources, should be viewed as evolutionary success7. From this viewpoint, evolutionary failure begins at the inception of decline after the peak, well before the time of actual extinction. Thus, the authors argue that we must, at least conceptually, tip our hats to the law of constant extinction because the time at which a genus reaches its peak and starts the decline to extinction is independent of how old the genus is.
Žliobaitė and colleagues’ study was conducted at the level of genera, which is one taxonomic level above species, and both species and genera commonly exhibit hat-shaped evolutionary trajectories8. Yet such simple patterns are rare at the higher taxonomic levels that Van Valen studied to develop the law of constant extinction. At those levels, evolutionary trajectories of species richness (a measure of evolutionary success) often have multiple peaks instead, and sometimes have long-lasting, persistent ‘tails’ of reduced diversity that show no signs of extinction9. At lower taxonomic levels, complete ecological replacement of one species or genus with another is common — at higher taxonomic levels it is not.
This difference in the nature of evolutionary trajectories at different taxonomic levels highlights one of the challenges of studying evolution, which is the hierarchical structure of the evolutionary process10. It is worth noting too that, although hat-shaped trajectories are common at lower taxonomic levels, they are not universal. For example, the North American mammalian fossil record from the Cenozoic era fits the law of constant extinction with remarkable fidelity11,12.
The results of Žliobaitė and colleagues’ work also provide insight into the drivers of evolutionary innovation. The authors’ data for North America and Europe show that, although both biotic and abiotic factors contribute roughly equally to genus origination rates, neither contribution is statistically significant. As the authors note, this provides evidence that evolutionary innovation is not driven by biotic or abiotic external changes. Instead, the data support the idea that evolutionary innovation is influenced by intrinsic factors — the less-predictable origin of the ‘right’ variants at the right time, able to exploit either existing or new resources.
It will be necessary to fully scrutinize the approach used by Žliobaitė and colleagues, and especially the theoretical implications of their findings. More data sets need to be analysed using a range of measures of environmental change and competition. Other biotic factors, such as predator interaction, should also be included in future analyses. Nonetheless, this paper represents a next step towards the formulation of general laws and principles in palaeobiology13,14, building on the detailed description of fossil material14.
Sign up for the daily Nature Briefing email newsletter
Stay up to date with what matters in science and why, handpicked from Nature and other publications worldwide.
Ezard, T. H. G., Aze, T., Pearson, P. N. & Purvis, A. Science332, 349–351 (2011).
Žliobaitė, I., Fortelius, M. & Stenseth, N. C. Nature552, 92–95 (2017).
Fortelius, M. et al.Phil. Trans. R. Soc. B371, 20150232 (2016).
Žliobaitė, I. et al.Proc. Natl Acad. Sci. USA113, 12751–12756 (2016).
Van Valen, L. Evol. Theory1, 1–30 (1973).
Quental, T. B. & Marshall, C. R. Science341, 290–292 (2013).
Van Valen, L. Evol. Theory1, 179–229 (1976).
Foote, M. et al.Science318, 1131–1134 (2007).
Sepkoski, J. J. Jr Paleobiology7, 36–53 (1981).
Jablonski, D. Evol. Biol.44, 427–450 (2017).
Alroy, J. Quantitative Mammalian Biochronology and Biogeography of North America. PhD thesis (1994).
Foote, M. & Miller, A. I. Principles of Paleontology 185 (Fig. 7.11) (Freeman, 2006).