Recently, scientific consider different "kind" of parrallel evolution. Recent review states that it can refer to phenotypic parrallel evolution (like a "convergence"), or genetic parrallel evolution (involving same locus or mutation). However, we can expected that different developmental changes could lead to similar phenotype. Therefore, should we talk about a third type of parrallel evolution: Developmental parrallel evolution?
Ok. Here is the article attached. Also read this reply to the article for another discussion.
http://www.sciencedirect.com/science/article/pii/S0169534708002206
I will point you to an article that discusses this very point.
http://www.sciencedirect.com/science/article/pii/S016953470700287X
If you cannot download it, I will post it up for you.
Ok. Here is the article attached. Also read this reply to the article for another discussion.
http://www.sciencedirect.com/science/article/pii/S0169534708002206
Parallel evolution should be restricted to situations where related species independently evolve the same features, usually adaptations in response to similar environmental challenges. This contrasts with convergent evolution where unrelated species, on widely divergent branches of the tree of life, independently achieve similar solutions in similar environments. In the first case a central question is whether the feature has arisen in parallel, or whether gene-flow has transferred adaptations among populations. I don't think it makes much difference whether the feature is genetic, developmental or phenotypic. That is really a secondary categorisation of how we are accessing the trait. I think that a developmental parallelism should really be defined as similar developmental changes, such as shifts in morphology at maturity (paedomorphosis), even if this results in quite different phenotypes in different groups. In any case, a parallel developmental shift can always be considered as a phenotypic parallelism. The contrast between parallel adaptations versus parallel non-adaptive changes, is more intriguing to me.
So, Michael, what is actually the fundamental difference between parallelism and convergence? Can't see any - in your view in both cases the same feature is developed independently. Species being ' closely' or 'distantly' related has little relevance as that is a most relative concept without any particular universality, rather derived from the point of view of each researcher's idea of relatedness on his/her focal group/study target. All species are related, and there is no border across which to claim when something is 'closely enough' related so that the term parallelism could be used instead of convergence, or vice versa. I might also speculate that the possibility of lateral gene transfer has no limits regarding the level of taxa's relatedness (probability though certainly increases the more close the taxa are). But, I do not advocate the use of these terms as different features, as I see them as one.
Thanks for all your reply!
To Lauri Kaila: I think that the intuitive limitation of parrallel evolution to "closely" species is based on the homolgy character concept. No? I mean that wings of birds and bats are convergence, not parrallelism. I suppose that when similar innovation in homologous trait occured in "distant" lineages, is parralelism.
To Michael: Yes, maybe parrallel development evolution it 's a kind of phenotypic parrallel evolution .... with a limit.
For example I work on lower mouth shape of chondrostoma (cyprinids). As previous authors considered two types of mouth, some species have arched mouth, other have straight mouth. Phylogeny showed that straight shape innovation occured three times independantly.
When I compared mouth shape between species using morphologic analysis, I see that in fact straight shape may be obtained by increasing width of the mouth development or decreasing height of the mouth. So, different kind of development may lead to same character state.
So, do consider that it is always a phenotypic parrallel evolution?
To Emmanuel: to you reply to me: problem is that the homology cannot necessarily be inferred a priori, based on educated guesses inferred from, e.g. closeness of taxa, but its assessment is a product derived from analysis (certainly there are ontogenetic or other ways as well, but they are not always feasible to study).
To your reply to Michael: what you did is exactly what I think should be done: getting discrepant results by different methods should lead the researcher to go back to the original data: it may well be, as in your example, that the originally thought apparent character is a result of different processes, which, indeed are the characters in cladistic analytical sense that should be considered instead of the resulting, at first glance apparent phenotype. Meaning that the same character state in your original data was found not relevant, observed by 'reciprocal illumination' that improved the quality of the data. So the understanding goes on. Well, this was perhaps a little beside the focal point of the parallelism/convergence discussion, where I realise that my opinion above was actually more or less repetition of the view expressed in the publication provided by Prem Subramaniam.
- Badges
- Science topic
- Similar topics
- Biological Science
- Evolutionary Biology