AI Reads Darwin

A special two-host discussion of four papers on cleaner-wrasse coloration and reef-fish color signaling — Cacela-Rodrigues et al. (2025) on clients preferring more blue-saturated cleaners; Cheney, Grutter, Blomberg & Marshall (2009) on the blue-and-yellow cleaning signal; Cheney & Marshall (2009) on the accuracy of mimicry and how deception improves with depth; and Tea et al. (2026) revising the genus Labroides and naming two new, color-coded species. The through-line: color as an honest signal, how client fish perceive it, and how signal and perception shift with depth and light — building toward concrete experiments a Cubert H5 snapshot hyperspectral camera could run on live cleaner fish in the lab (whole-body honesty maps, hyperspectral video of cleaning interactions, receiver-view reconstruction at depth, spectra on the new species mapped to their UCE phylogeny, the UV channel, and catching the switching fangblenny mimic in the act). Prepared following a lab visit and discussion with Jesse Freeze (Mehta Lab, UC Santa Cruz). Companion page with all papers, summaries, and research directions: https://michaelalfaro.github.io/alfaro-lab/cleaner-wrasse-color.html