Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.


Different worlds

Patterns of species association reveal that terrestrial plant and animal communities today are structured differently from communities spanning the 300 million years that preceded large-scale human activity. See Letter p.80

The British author L. P. Hartley wrote in one of his best-known novels, The Go-Between, that “The past is a foreign country: they do things differently there.” This poignant imagery of remoteness from the past captures the essence of an emerging global consciousness. Human hegemony over nature has become so pervasive and profound that it is quite possible that we have created a world that has little or no precedent — in ecological parlance, it has no analogue. On page 80 of this issue, Lyons et al.1 detail a compelling case that this extraordinary situation is an undeniable reality for the rules that govern how plant and animal communities are structured.

The authors assembled data on the presence and absence of terrestrial plant and animal taxa for 80 fossil and modern assemblages in North America, Africa and Eurasia, spanning the past 300 million years. Using a statistical approach that was designed to compare occurrence data against a randomized 'null' assemblage, they quantified the fraction of species pairs in each assemblage that deviated from random expectations about where they should be found. Species pairs meeting this criterion provide valuable insight into the ecological processes that structure communities2,3.

In modern communities, most species pairs show random co-occurrence, but those that are non-randomly associated are typically segregated — that is, they tend to co-occur less frequently than would be expected by chance2 (Fig. 1). Lyons et al. wanted to know whether the fossil record is consistent with this pattern of species segregation. The headline finding is that the pattern of co-occurrence dominating modern communities departs sharply from that of the past. As in modern communities, co-occurrence of most pairs was random. But unlike in modern communities, the non-random associations were dominanted by aggregated species pairs, which co-occur more frequently than would be expected by chance2 (Fig. 1). This dominance of aggregated pairs persisted with little change for more than 300 million years on different contintents and across diverse taxa, until about 6,000 years ago, when the sharp transition to the segregated co-occurrence pattern began.

Figure 1: Species associations.

In a terrestrial ecological community, any two species may occur randomly (a) at locations in a landscape. Alternatively, species pairs may be non-randomly associated, in which case they can be either segregated (b), meaning that the two species co-occur less frequently than would be expected by chance, or aggregated (c), meaning that they co-occur more frequently than expected by chance. Among non-randomly associated pairs, Lyons et al.1 document a shift from a dominance of aggregated pairs before the expansion of human populations to the segregated pattern typically seen today. (Figure adapted from ref. 7.)

After running a battery of tests to ascertain that this temporal trend was not an artefact, the authors speculate that an expanding human population may explain why species co-occurrence patterns are so different today. The shift was most obvious in North American assemblages (where the most occurrence data were available) and coincided with the inexorable spread of agriculture in this region. The authors propose that habitat fragmentation and limitations on species dispersal associated with land use were probably the main engines driving the shift. The structure of plant and animal terrestrial communities would never be the same again.

This interpretation is sure to attract fervent debate and lead to further research to confirm the pattern and disentangle the proposed mechanisms involved. The tension between the distant past and the familiar present that the study highlights, however, has an underlying implication that may not be as obvious. If the past is different from the present (in this case not in the immanent processes that were operating, but in their frequency), its applicability to our current societal need to anticipate ecological changes and design adaptation measures — a goal that Lyons et al. acknowledge is a priority — is not immediately manifest. There is no easy way around this tension. At stake is whether we can reliably use the past as a guide to an uncertain, anthropogenically modified future.

A small cadre of voices argues that a human-dominated present limits the use of the past as a key to unlocking the future4. In this view, the world we live in today, and the immediate future that our grandchildren will inherit, has no analogue in the geological past. As a consequence, referencing 'natural experiments' in the distant past as a guide to predict what might happen, now or in the future, is a flawed strategy. Out is the use of uniformitarianism5 as a guiding principle, and in is a new kind of 'post-normal' science6. Lyons and colleagues' study of human impacts on community-assembly rules, at least as implicated by species co-occurrence patterns, seems to embody evidence for this no-analogue world.

A more optimistic view of this tension between the past and present — one that acknowledges that processes change and interact in complex ways over time, whether human action is involved or not — is that it poses a challenge for how we select analogues from the past to gain insight into future conditions. Lyons and colleagues' finding is a stark reminder that analogue selection often over-stresses likenesses at the expense of differences. However, small and unknowable differences in starting points may overwhelm the signal of the likenesses, making analogue selection a risky business. To use the past as a guide, we must select from the dense fabric of likenesses and differences that was its contingent state at a moment in time, and apply only those particular events and conditions relevant to our present needs.

Moving beyond this tension will require creative ways of thinking about how we use the distant past to improve our understanding of the present and our anticipation of the future, which may provide a ground for wiser action. Lyons and colleagues' study is an excellent entry point into thinking about this problem.Footnote 1


  1. 1.

    See all news & views


  1. 1

    Lyons, S. K. et al. Nature 529, 80–83 (2016).

    ADS  Article  Google Scholar 

  2. 2

    Gotelli, N. J. & Ulrich, W. Oecologia 162, 463–477 (2010).

    ADS  Article  Google Scholar 

  3. 3

    Connor, E. F., Collins, M. D. & Simberloff, D. Ecology 94, 2403–2414 (2013).

    Article  Google Scholar 

  4. 4

    Knight, J. & Harrison, S. Anthropocene 5, 71–75 (2014).

    Article  Google Scholar 

  5. 5

    Baker, V. R. Anthropocene 5, 76–79 (2014).

    Article  Google Scholar 

  6. 6

    Funtowicz, S. O. & Ravetz, J. R. Futures 25, 739–755 (1993).

    Article  Google Scholar 

  7. 7

    Blois, J. L. et al. Ecography 37, 1095–1108 (2014).

    Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Gregory P. Dietl.

Related links

Related links

Related links in Nature Research

Ecology: Global trends in plant naturalization

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Dietl, G. Different worlds. Nature 529, 29–30 (2016).

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.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing