"PRINCIPLES OF PHYLOGENETICS: ECOLOGY AND EVOLUTION" Integrative Biology 200B Spring 2011 University of California, Berkeley B.D. Mishler March 31, 2011. Reticulation,"Phylogeography," and Population Biology: Genealogy vs. Phylogenetics? We have discussed a number of assumptions that are used in phylogenetic methods. One fundamental shared assumption shared by all methods we have talked about so far is that divergence (splitting of lineages) occurs, not reticulation (joining of lineages). To address the latter possibility, following the general approach in this class, we need to first consider the principles of the topic before we can figure out how to study it: what, if anything, is reticulation? A. The fractal nature of the Tree of Life. The tree of life is inherently fractal-like in its complexity. Look closely at one lineage of a phylogeny (remember: defined as a diachronic connection between an ancestor and a descendent) and it dissolves into many smaller lineages, and so on, down to a very fine scale. Thus the nature of both the terminal units (OTUs -- the "twigs" of the tree in any particular analysis) and the characters (hypotheses of homology, markers that serve as evidence for the past existence of a lineage) change as one goes up and down this “fractal” scale. Ontologically speaking, larger-scale lineages are usually composed of smaller lineages nested inside them, and so on. Thus, epistemologically speaking, the choice of which lineage to represent in a particular analysis depends on the questions begin asked. Furthermore, the lineages at these different levels potentially have different histories; in other words the smaller lineages are not always proper subsets of the larger containing ones. This is sometimes called the gene tree / species tree distinction (Maddison and Maddison 1992), but that distinction is far too simplified; there are many nested levels of potentially incongruent lineages, not just two. Besides, there is no a priori "species tree" to compare gene trees to; on the contrary the "species tree" has to be inferred largely from gene trees! B. Is there a difference between genealogy and phylogenetics? or: Is there an important break at the "species level"? Rosenberg and Nordborg (2002) say that there is a difference, and many workers (primarily zoologists) do make a distinction between reconstructing trees at the population level (genealogies) and at the species level and above (phylogenies). The same workers would distinguish phylogeography from phylogenetic biogeography, and separate out study of coalescence ("gene trees") from branching evolution ("species trees"). Workers following this the distinction often make use of a concept of population, that is left relatively unexplored (see discussion by Millstein 2009). What is a population? A A B C c b ageographically circumscribed set of organisms in the same species? [Can you see the possible problems with a definition like this?] The goal in this approach is to look at patterns of genetic similarities and difference across these "populations," examining parameters such as effective population size, migration rates, divergence times, and population growth or decline (Knowles and Maddison 2002; see figure on last page). A bit if history: population genetics has always placed special emphasis on identity of genes by descent -- why? It goes back to the concept of replication we discussed previously. To study a process of natural selection, you need to know what the replicators are, as well as the interactors. So knowing identity of genes isn't enough, you need to know that they are related, Knowing how things are related is essential to testing functional hypotheses. Population geneticists developed important methods to study gene trees, independently from the methods developed by systematists to study species trees. But the two traditions have themselves exhibited hybrid vigor in the last decade! If it is acknowledged that branching and reticulation occur at many nested levels in the Tree of Life (even if the balance between the two processes shifts towards more reticulation as you go to finer scales), then we have to take into account the possibility of both processes at all levels, and it isn't useful to consider genealogy as different from phylogeny. It's all a matter of scale, and always a case of comparing trees. The appropriate methods for studying history remain phylogenetic (not distance-based) above and below the "species level." Questions about genetic distance, gene frequencies, etc. are of course important in population genetics, but should not be confused with questions about the history of genes and "populations." C. So what are the twigs on phylogenetic trees meant to represent? Thus, even if one wanted to try to avoid problems of hybridization by using only semaphoronts in a data matrix, one would still need to pay attention to the same issues of scale. On the theory side, one still needs to decide conceptually which lineages are being represented by what semaphoronts. Given the fractal nature of the problem as discussed above, there are semaphoronts nested inside of semaphoronts. On the empirical side, it is nearly impossible in practice to use single semaphoronts as terminals rather than compositely-coded OTUs that have data taken from a number of semaphoronts. For one thing, not all semaphoronts bear all the characters; there may be juvenile specializations or sexual dimorphism present in a lineage. Some specimens will be missing reproductive organs or other key features. Different genes will often be sequenced from different individuals. Furthermore, data are often taken from the literature, (e.g., a previously published ultrastructural analysis) or from a database (e.g., another lab’s gene sequence), in cases where no reference can be made to an original semaphoront (e.g., if no voucher specimen was deposited in a museum). Thus, data are virtually always compiled from studies of different individual organisms considered to represent the same terminal lineage. OTUs are nearly always composites in practice; their composition varying depending on the scale of analysis. This topic obviously touches on the species debate, which we have dealt with at length earlier. I just point out here that the fractal scaling of nested lineages includes those well belowthe traditional
View Full Document
Unlocking...