Detecting convergence

Reading the footprints of evolution is one of the most fascinating pursuits in science, but the trails can become misleading. Whilst convergence provides some of the best evidence for evolution, it also reminds us that evolutionary analysis may not be as straightforward as is sometimes supposed. To the first approximation, the more similar the organism the more likely it is that they are closely related. The reality of convergence, however, does introduce (at least potentially) the risk of circularity. After all, are the organisms similar because they are indeed closely related, or because they have independently arrived at the only viable solution? In other words, are similarities caused by shared ancestry, or simply because there is no other choice? New discoveries about phylogenetic relationships between organisms and rare cases of ‘hidden’ convergence highlight the importance of answering this question correctly.

Revised phylogenies

Woodlouse isopoda imageIn recent years the techniques of molecular biology have not only swept the biological field, but in the case of phylogenetic analysis have frequently led to new proposals of relationships that are radically at odds with more traditional phylogenies that necessarily relied almost completely on morphology. Again and again as the new phylogenies are revealed the cry goes up as to how much convergence was lurking unobserved within the morphology of the group. Groups once thought to be closely related because they shared the same morphological characters (e.g. segmentation in arthropods and annelids) are now clearly seen to be much more distantly related. In such cases the inference is inescapable that life has repeatedly arrived at the same solution.

The recognition of convergences based on molecular phylogeny has become so common-place as to hardly merit comment, and indeed in the Map of Life these “routine” examples are simply too numerous for us to document. Nevertheless, selected cases of convergence are certainly worth drawing attention to, and for three reasons. First, some of these characters may represent key innovations and so their repeated emergence is part of the general argument for inevitabilities in evolution. Second, the repeated emergence of particular characters is potentially important in the analysis of how morphospaces are both defined and occupied. Third, and related to the second point, is the opportunity to identify instances of concerted convergence whereby characters are “unexpectedly” found to be in close association, thereby prompting a more detailed study of the functional and ecological context.

Unmasking hidden convergence

Chloroplasts imageRare instances exist in which until the convergence is properly recognized it will completely mask the true evolutionary story. In other words, in certain cases if the historical pathways are exceptionally constrained then the resulting degree of similarity is so close that it can trick the observer into thinking that the structure has evolved only once (and so is monophyletic), whereas in reality it had two (or more) separate origins. Such conceivably applies to the origin of chloroplasts. These have been widely regarded as having a single origin, but some evidence suggests they are convergent, evolving three times. However, in being “forced” to track the same pathway of gene loss and/or gene export (to the nucleus) so the chloroplasts effectively have obscured their evolutionary footprint, throwing dust in the eyes of the observer.

To reiterate, instances of convergence do not question, let alone negate, the fact of evolution, but instead they remind us that the process of evolution can be far more constrained than often imagined. Such instances confirm the reality of directionalities and hint at the possibility of a deeper structure that determines the patterns of constraint.

Go to the top of the page