The Unsolved Challenge to Phylogenetic Correlation Tests for Categorical Characters When comparative biologists observe that animal species living in caves also tend to have reduced eyes, they may see such correlation as evidence that the

I should note the figure is from this excellent paper by Wayne Maddison and Rich FitzJohn. academic.oup.com/sysbio/artic...

When @omearabrian.bsky.social saw the manuscript, he described it as: ‘What if this figure could be an entire paper?’

I choose to interpret that as high praise.

Now accepted at AmNat: The geometry of macroevolution (with Dan Rabosky). www.journals.uchicago.edu/doi/10.1086/...

Opinion | The Anti-D.E.I. Crusader Who Wants to Dismantle the Department of Education (Gift Article) Christopher Rufo’s mission to make universities feel “existential terror.”

www.nytimes.com/2025/03/07/o...
Worth a read in order to better know who wants to attack higher ed. 1/4

Eight herbarium sheets displaying various specimens of the blueberry genus, each with detailed labels and color calibration bars. The text above provides code for searching and downloading specimen metadata and images. Image credit: OUHC and BRIT herbaria.

mvh: An R tool to assemble and organize virtual #herbaria from openly available specimen images

New in #AppsPlantSci by @tvasconcelos.bsky.social & @jboyko.bsky.social

bsapubs.onlinelibrary.wiley.com/doi/full/10.... @r-foundation.bsky.social @gbif.org #botany #biodiversity #iamabotanist

Aronoff Lab at The Ohio State University

The Pease Lab in the Dept of Evolution, Ecology, and Organismal Biology at The Ohio State University is looking for a Postdoc interested in genotype-phenotype-environment evolution in plant and animal genomes. Details at osu.wd1.myworkdayjobs.com/en-US/OSUCar... and more info at www.peaselab.org

"nasty consequences FOR*..."

honestly, a spelling error in my first post really just confirms my identity on bsky.

a potential nasty consequence. these are profile likelihoods for a default model (b/d) and a more model (a/c). (b/d) is what you'd get fitting a standard Pagel model with ARD. You get these horrible likelihood ridges for parameter estimates (e.g., b. theta4). But for the same dataset, there is a slightly non-standard model (non-standard in the sense of it's not ER/SYM/ARD) that has really nice statistical behavior

I wrote this about finding the optimal discrete character model structure. I hope this will be useful for empiricists. The main point is that relying only on the package default models is bad and can have nasty consequences or parameter estimates and ASR.

www.biorxiv.org/content/10.1...