I am currently working on a project where I study the change of a continuous morphological variable with body size (measured as body weight) across diverse taxa (insects, spiders, lizards, frogs). In order to account for the non-independence of the data, I built an approximate tree based on published phylogenies. I have a reasonable branching pattern down to family or genus level for most taxa, but lack statistically supported branch length.

I find a strong change of the scaling coefficient when using phylogenetic least squares (with a covariance matrix based on my tree) vs a non-corrected least-squares approach, suggesting that most of the change observed in my continuous variable is explained by evolutionary history, rather than by body mass as such. I would now like to investigate how early on in the tree these shifts occur.

Intuitively, I could start collapsing my tree into polytomies down to arbitrary taxonomic levels, eg genus, family, order etc, and re-do my phylogenetic least squares. I would expect a stepwise approximation of the pgls to the uncorrected result (in the extreme case, I would end up with a star phylogeny, which should give an identical result to my uncorrected regression). Such an approach may allow to say something about whether most of the change occurs between genera, families etc, but it requires the use of somewhat arbitrary (and controversial) taxonomic levels. I would be grateful if someone would have an alternative suggestions on how to perform such an analysis, with arbitrary branch length and non-ultrametric trees.

Thanks Thriceguy

  • $\begingroup$ Quick question - do you have fossils for your trait and group? $\endgroup$ – C_Z_ Apr 7 '15 at 20:50
  • $\begingroup$ I have no fossils. The non-contemptuous tree tips (I assume your question relates to them?) arise because different parts of the tree are resolved to different levels, and I set all branch length to 1 in my analysis. I suppose I should be able to ultrametrize the tree, however, if this would be helpful/better/required. $\endgroup$ – David Apr 7 '15 at 21:21
  • 1
    $\begingroup$ In case anyone encounters a similar problem, this blog entry: blog.phytools.org/2012/11/…, and, in more detail, this paper onlinelibrary.wiley.com/doi/10.1111/2041-210X.12066/full may provide a solution. In short, one can estimate ancestral states of continuous or discrete variables and plot them on an existing tree. In case of scaling studies, phylogenetic residuals from a regression of the trait against mass may be the most meaningful parameter, clearly visualising the contribution of common history independent of size. $\endgroup$ – David Apr 9 '15 at 12:41
  • $\begingroup$ Do you want to test this formally or just evaluate whether the variance seems to lie closer to the root of the tips of the tree? To me, your basic idea to test "how early on in the tree these shifts occur" seems to imply a model with different rates of evolution (e.g. stasis-shift-stasis). If that is the case, I think you need to fit evolutionary models with different assumptions (bursts, brownian motions etc) to your dataset, to see which one fits best. See e.g. Uyeda et al (2011) for inspiration. Or am I misunderstanding you? $\endgroup$ – fileunderwater Apr 12 '15 at 22:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.