Assesses vocalization differences in bird species across geographies, indicating vocal divergence patterns.
We apply WCM-PL's polytope vertex-occupancy machinery (Cell24Sensor, 24 vertices, dim=4), previouslyused to characterize baseline structure in epileptic EEG, to a new domain: 1,655 short field recordings of sixJapanese resident bird species (Zosterops japonicus, Cettia diphone, Hypsipetes amaurotis, Parus minor,Poecile varius, Emberiza cioides) across up to 10 sites spanning Hokkaido to the Ogasawara and Ryukyuisland chains. For each recording we compute a per-file occupancy Shannon entropy from three independentspectral features (centroid, flux, loudness) and test for cross-site differences. Four of six species showstatistically significant, cross-feature-replicating site differences. A literature search confirms this generalphenomenon (geographic song variation in these species) is already documented by a three-decade researchprogram (Hamao and colleagues, National Museum of Nature and Science, Tokyo) -- so the core contributionhere is methodological, not a biological discovery. Restricting the comparison to mainland-only sites reveals agenuine species-level divergence not explicit in the literature we could access: one species' (Parus minor) siteeffect is entirely island-driven and vanishes on the mainland, matching the general Japanese-ornithology claimthat geographic song variation is peripheral-island-specific, while another (Cettia diphone) retains an extremelystrong site effect (p as low as 6.7×10-19) using ONLY four purely mainland sites. A cross-species control rulesout a shared recording-habitat confound. All findings are screening-level (single embedding scale, omnibustests only, no independent dataset replication) and are reported with this status explicitly. ※ The License for the code file is AGPL 3.0-or-Later
No takes yet. Share an insight, caveat, or question.
Watabe et al. (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: