I have been interested on the theoretical foundations of the Systematics, especially when this area meets molecular biology and started to ignore other information sources as chromosomes or morphological traits.
This union was very strong and congruent with the neo darwinian definition of evolution as a process of change in the frequencies of alleles. But in my opinion the causal pluralism that defend the Developmental systems theory in the relation Genotype-Phenotype (and additional critics and proposal as Extended Inheritance) demand a new definition of evolution and by consequence a new systematics framework, that could make clear how build the history of the living things.
Molecular data is not better than morphology and morphology is no better than molecules.
Despite what Robert said, there is no guarantee that a molecular tree is closer to a "true tree". Also, different fragments may also disagree between each other. Should one discard one dataset? On what basis? There is no such a thing.
Robert, I am sorry, but I don't think your examples are really good. Morphology works really well in deeper levels of the hierarchy, although molecules may be more useful for the shallowest ones. Sharks and dolphins would never come out together in a properly done cladistic morphological analysis. One done with robust primary hypothesis of homology.
Remember that we found out that reptiles, dicotyledons, ferns and other groups are not monophyletic way back when nobody was using molecules.
Also, molecular data may be prone to convergence. Check the paper by Liu et al 2010 showing how genes related to echolocation places bats and dolphins as sister groups (https://www.researchgate.net/publication/41404012_Convergent_sequence_evolution_between_echolocating_bats_and_dolphins?ev=srch_pub&_sg=GKKuBnIDBAhKMuPVtKeprpt%2FviqCOofXG145ulIFUrUq1kOYvN0lvrhSC%2F1zuBpl_MUMTgjRZfM4SsyKzo9Z7ClLNnRRTOGxzUzvlfGPLO8URofV7ces2DeIfhOrNeO2g_nDvvkSNM4hzZ3hjQlOXrvZr1nkOlw8PicQgVEsQyJEUlp02vOYHv2ATCbO%2BjbCLe)
One thing that tells us that this is just molecular convergence is morphology. Bats and dolphins clearly do not form a monophyletic group.
Besides, you can have an alignment with more than 3000 bases without variation and that is, consequently, uninformative. Some colleagues are about to publish a paper showing that the whole chloroplast genome of Asteraceae has a very low level of variation and how the effort in sequencing is useless to build phylogenies for anything lower than tribes (and tribes in Asteraceae are kind of big). On the other hand, 30 or so morphological characters can have a lot of information.
I do think its is more than time for us to stop arguing what kind of data source is better and to join efforts to use all the good information we can get our hands on in total evidence analysis.
Of course "good" may mean different things for different people, but check that:
http://www.nature.com/nature/journal/v496/n7445/pdf/496295a.pdf
Cheers, you all,
Article Convergent sequence evolution between echolocating bats and dolphins
A very interesting question,
I think that although the definition of evolution as only changes of gene frequencies is indeed becoming outdated, because of it's incompleteness, those genetic changes are still part of the process, and should still considered informative, mainly because of neutral changes in the genome, that could tell us well how non-adaptive genetic changes happened in some clades.
But I hope that the new focus on development and evolution together will also promotes a resurgence of phenotypic (mainly morphological) systematics, and hopefully, the use of developmental data should also be adopted, in a perspective of total evidence.
IN my view if we consider systematics there is no room for classical hierarchical terms even though that systematics reflects cladistic. Molecular systematics just consider just clades and sister clades and taxon and there is a lot of research is to be done for well resolved phylogentic implication of the related taxon. still we are just on the anchor of the ship. thorough sampling is needed through out the geographical region around the world and then only we can resolve the thing and yes the with time for a particular taxon so thing for the whole life forms it would be going to messy and hectic but it will reveal the true phylogenetic implications. But for classical systematics people rely on the morphology and scientist who do molecular phylogeny need them for proper identification of taxon. Combined effort will lead to fruitful results.
Hi, I think we cannot judge the state or position of systematics on some ideas or minor trends. The field has very long tradition, and provided huge amount of very important information on the Earths biota. I work with systematics more than 15 years, using morphological, molecular and other data and think this methodological framework, including combination of different data, is very robust.
Hi Fedor you may right i am also doing classical taxonomy this is my point of view why every body is ignoring the classical experts they are indeed needed for the same. To describe a new species the morphology prior of molecular. plez have some suggestions. i dont know on phylogney what is better for a tree showing phylogenetic relation ship morphological or molecular data. as molecular guys saying that maikng a tree on morpho matrixtable is wastre of time. please suggest something or give your views.
I am agree with Fedor . The molecular phylogeny offers a framework to find answers to many process. Also combining molecular and morphological data charactres we could obtain a bettter stimation of the phylogeny.
But, I have a question if you have practiced an cladisitc analysis (morphologial data) and your results indicated some kind of homoplasy. It is worth combined thes two set of data??
Well i usually use morphology to describe species. For phylogeny and tree construction I prefer molecular data as they are, in my opinion, less biased by author's subjective view. Once I have a tree I try to discuss it with morphological data taken into account.
I think that any approach to systematics is valid, as long as it helps to establish a trusted classification. We often confuse the objective with the method. Molecular phylogenetics is not an objective, nor is it science per se, it is just a method. Currently it is arguably the best method to reconstruct phylogenetic relationships, and it is being continuously improved. That does not mean we should forget about everything else, on the contrary. A molecular phylogeny ONLY makes sense when we look at phenotypic evolution in its context. So, while phenotype characters might not be suitable to reconstruct phylogenetic relationships (because an often high degree of homoplasy), they become interesting in the context of the phylogeny, because they raise questions about macroevolution. So to me, all these methods and resources have to be integrated, but each at its level of merit.
To answer Clara: if molecular and morphological data give conflicting phylogenetic signal, and you can show that this is because of homoplasy in the morphological data, you should NOT combine both datasets to obtain the phylogeny, since by definition your resulting phylogeny will be further away from the (unknown) true tree. Rather, you should map morphological data a posteriori on the molecular phylogeny. In practice, often the addition of morphological data makes little difference, since the number of characters is low compared to DNA data. For example, if you have a 3-gene alignment with 3000 bases, a set of 30 morphological characters will probably not change anything in the phylogeny, but it may be very powerful and enlightening when instead mapped on the phylogeny.
Thank you Fedor and Robert . I have used map the characters on the tree, but it procedure was criticized, by its methodology. Thus, is it better characterize each clade obtained instead map the morphological characters? or perhaps is better use the both ways. In the particular case in the morphological phylogeny , was obtained a CI= 0.35 and the morphological characters numbers used were close 35
Hi Clara, here is a simple example: dolphins and sharks have several morphological features in common. However, other than both being vertebrates, they are not related. Now imagine you have a molecular and a morphological dataset. If you combine both, you might get a tree where dolphins and sharks appear more closely related than they actually are. In this particular case you know that this would be a biased result. If instead you just do a molecular phylogeny and then map morphological characters on the tree, you would get a tree closer to the truth and actually understand what is going on here: that the similar morphology is due to convergence, parallelism, homoplasy. This now gives you a powerful argument to compare such patterns with environmental constraints that may cause them. In many cases the situation is much more subtle and not as obvious, but nevertheless this example shows that it is generally better to first do a separate molecular phylogeny and map morphological characters on the resulting tree, before proceeding with a combined dataset.
Molecular data is not better than morphology and morphology is no better than molecules.
Despite what Robert said, there is no guarantee that a molecular tree is closer to a "true tree". Also, different fragments may also disagree between each other. Should one discard one dataset? On what basis? There is no such a thing.
Robert, I am sorry, but I don't think your examples are really good. Morphology works really well in deeper levels of the hierarchy, although molecules may be more useful for the shallowest ones. Sharks and dolphins would never come out together in a properly done cladistic morphological analysis. One done with robust primary hypothesis of homology.
Remember that we found out that reptiles, dicotyledons, ferns and other groups are not monophyletic way back when nobody was using molecules.
Also, molecular data may be prone to convergence. Check the paper by Liu et al 2010 showing how genes related to echolocation places bats and dolphins as sister groups (https://www.researchgate.net/publication/41404012_Convergent_sequence_evolution_between_echolocating_bats_and_dolphins?ev=srch_pub&_sg=GKKuBnIDBAhKMuPVtKeprpt%2FviqCOofXG145ulIFUrUq1kOYvN0lvrhSC%2F1zuBpl_MUMTgjRZfM4SsyKzo9Z7ClLNnRRTOGxzUzvlfGPLO8URofV7ces2DeIfhOrNeO2g_nDvvkSNM4hzZ3hjQlOXrvZr1nkOlw8PicQgVEsQyJEUlp02vOYHv2ATCbO%2BjbCLe)
One thing that tells us that this is just molecular convergence is morphology. Bats and dolphins clearly do not form a monophyletic group.
Besides, you can have an alignment with more than 3000 bases without variation and that is, consequently, uninformative. Some colleagues are about to publish a paper showing that the whole chloroplast genome of Asteraceae has a very low level of variation and how the effort in sequencing is useless to build phylogenies for anything lower than tribes (and tribes in Asteraceae are kind of big). On the other hand, 30 or so morphological characters can have a lot of information.
I do think its is more than time for us to stop arguing what kind of data source is better and to join efforts to use all the good information we can get our hands on in total evidence analysis.
Of course "good" may mean different things for different people, but check that:
http://www.nature.com/nature/journal/v496/n7445/pdf/496295a.pdf
Cheers, you all,
Article Convergent sequence evolution between echolocating bats and dolphins
- Badges
- Science topic
- Similar topics
- Biological Science
- Evolutionary Biology