Shared ancestry predicts disease levels

Ecological factors such as host density are important predictors of disease incidence. But another key determinant may be the evolutionary history and relatedness of the host community. See Letter p.542

Understanding the dynamics of disease-causing microorganisms is central to human medicine and to improving the health of domesticated animals and crops. Disease is also an increasingly recognized part of ecological research, because pathogens can influence the coexistence of host species, the potential for non-native species to invade and ecosystem productivity1,2. However, predicting disease levels in an ecosystem is challenging. On page 542 of this issue, Parker et al.3 argue that such predictions can be improved by understanding the evolutionary relationships of the host community — in ecological parlance, the assemblage of host species present at a location.

Communities of hosts and pathogens are complex. Starting simply, consider a single pathogenic species infecting a single host species. In this case, high host density is typically associated with increased disease — a relationship evident to any parent whose child brings a cold virus home from school. But what happens when there are multiple hosts? In some situations, increased host diversity might reduce disease. For example, in a process known as the dilution effect4, the risk of disease is lowered if the most favourable host for a pathogen is a small component of a diverse host community. Alternatively, pathogens may grow rapidly in certain hosts, and the resulting high numbers of diseased individuals might allow the pathogen to 'spill over' and contribute to increased infection of other hosts5. Although these topics are directly relevant to human and wildlife diseases, they are hard to study with mobile hosts. By contrast, Parker et al. show that such questions can be examined directly with plants.

The authors' field study combined observational and experimental approaches. The team first documented the average amount of diseased leaf tissue on 43 plant species (Fig. 1). Rarer species had lower disease levels, as expected given the typical positive association between host abundance and disease. The authors also determined the evolutionary relationships of the 43 species and found that species that had fewer relatives had lower disease levels. Next, they used a model based on an independent data set of 210 plant genera and 212 fungal pathogens to explore evolutionary relationships between host plants and pathogens. This model, combined with knowledge of host abundance, provided the best predictions of disease levels in the field.

Figure 1: Health amid disease.

Bruce Lyon, UC Santa Cruz

Parker et al.3 find that fungal infections are more likely to spread between closely related species, such as the grasses shown here in their field site, than to more distantly related species, such as this wild radish.

Finally, the team introduced 44 other plant species to the field sites. These were chosen for their evolutionary diversity: some were closely related to species at the field sites, whereas others were distantly related. The authors were able to successfully predict disease levels on these new species using their model of host–pathogen evolutionary relationships. In a nutshell, the message of this paper is that disease levels depend not only on ecological factors (host abundance) but also on the evolutionary relationships of the host community.

These results are perhaps not surprising — after all, many diseases are specific to plants of the same genus or family. However, Parker and colleagues' study has three main strengths: the authors quantified the relationship between disease levels and host evolutionary relationships in the context of natural disease transmission; they made explicit predictions and tested them experimentally; and the general nature of their model of host–pathogen evolutionary relationships means that it can be widely used by other researchers.

This work contributes to ongoing discussions about the diversity and composition of communities. For example, the authors' field sites were 'overdispersed', meaning that they contained fewer related plant species than would be expected by chance given the possible species in the region. This pattern has been thought to be a consequence of resource competition, because unrelated species may compete less with each other6. Parker and colleagues' findings suggest an alternative mechanism: that pathogens shape community structure if closely related host species are less successful owing to increased disease levels.

The study also has applied significance. For instance, it implies that less disease should occur in plant communities with combinations of species from distinct evolutionary lineages. This information is relevant to designs for intercropping (when two or more crops are grown in close proximity) and forest-tree mixtures7. Furthermore, the findings relate to the major challenge of predicting the fate of newly introduced species. Given a list of resident host species, one could use the authors' approach to explore whether a certain introduced species is likely to share pathogens with others, which may affect its invasion success. The abundances of individual host species (both native and non-native) will, of course, also contribute to determining disease levels.

Although Parker et al. focused on leaf-infecting fungi, evolutionary relationships are important for characterizing the host range for diverse natural enemies of plants, including other microbes and herbivores8. Future work should include mechanistic studies (for example, to assess whether there are certain plant traits that drive relationships between pathogens and host evolutionary history) and studies that explore the generality of these field results. For example, the nonlinear relationship in Figure 3 of the paper3 suggests that recent evolutionary history is particularly relevant to predicting disease levels. Would this type of relationship be expected across other plant communities and natural enemies? Future research should also explore the roles of other ecological factors in disease prediction, such as the mode and distance of pathogen dispersal, aggregation patterns of plant species and the spatial scale of studies. Finally, it is worth noting that the ecological interpretations of these studies assume an inverse relationship between disease expression and host fitness — but it is challenging to perform experiments that assess disease effects on fitness, especially in diverse field communities with many plants and pathogens.

Parker and colleagues' careful research is the latest in a suite of studies that emphasize the importance of evolutionary history in ecological processes6,7,9. This line of enquiry illustrates the continued relevance of a famous quote from biologist Theodosius Dobzhansky10: “Nothing in biology makes sense except in the light of evolution.”


  1. 1

    Mordecai, E. A. Ecol. Monogr. 81, 429–441 (2011).

    Article  Google Scholar 

  2. 2

    Maron, J. L., Marler, M., Klironomos, J. N. & Cleveland, C. C. Ecol. Lett. 14, 36–41 (2011).

    Article  PubMed  Google Scholar 

  3. 3

    Parker, I. M. et al. Nature 520, 542–544 (2015).

    ADS  CAS  Article  PubMed  Google Scholar 

  4. 4

    Keesing, F., Holt, R. D. & Ostfeld, R. S. Ecol. Lett. 9, 485–498 (2006).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  5. 5

    Power, A. G. & Mitchell, C. E. Am. Nat. 164, S79–S89 (2004).

    Article  PubMed  Google Scholar 

  6. 6

    Webb, C. O., Ackerly, D. D., McPeek, M. A. & Donoghue, M. J. Annu. Rev. Ecol. Syst. 33, 475–505 (2002).

    Article  Google Scholar 

  7. 7

    Castagneyrol, B., Jactel, H., Vacher, C., Brockerhoff, E. G. & Koricheva, J. J. Appl. Ecol. 51, 134–141 (2014).

    Article  Google Scholar 

  8. 8

    Gilbert, G. S., Magarey, R., Suiter, K. & Webb, C. O. Evol. Appl. 5, 869–878 (2012).

    Article  PubMed  PubMed Central  Google Scholar 

  9. 9

    Gilbert, G. S. & Webb, C. O. Proc. Natl Acad. Sci. USA 104, 4979–4983 (2007).

    ADS  CAS  Article  PubMed  Google Scholar 

  10. 10

    Dobzhansky, T. Am. Biol. Teach. 35, 125–129 (1973).

    Article  Google Scholar 

Download references

Author information



Corresponding author

Correspondence to Helen M. Alexander.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Alexander, H. Shared ancestry predicts disease levels. Nature 520, 446–447 (2015).

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.


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