Jake Berv

@jakeberv

Schmidt AI in Science Postdoctoral Fellow | evolution | systematic biology | natural history | barbeque enthusiast

613 Siguiendo468 Seguidores289 Publicaciones

Bluesky

¿Es tu perfil?Reclámalo y jakeberv.dirty.social es tuyo.Reclamar perfil

Jake Berv@jakeberv· 4d

It's been real @umich.edu 11/11/19-08/09/26 I'm out.

Jake Berv@jakeberv· 19d

Jake Berv@jakeberv· 22d

However, grid-cell means can obscure compositional patterns, so we dug deeper and studied compositional enrichment for globally fast or slow lineages (across space). It turns out that high-latitude grid cells are compositionally enriched for globally fast lineages (Supplementary Figure 17).

$Supplementary Figure 17. Quantile occupancy and tail-composition diagnostics. (A–B) Quantile occupancy maps showing the number of species per H3 grid cell belonging to the global fast25 (top quartile of lineage rates; A) and slow25 (bottom quartile; B) sets. Each species contributes +1 to every cell it occupies. (C–D) Corresponding compositional maps showing, for each cell, the fraction of the local assemblage comprised of fast25 (C) or slow25 (D) lineages (0–1 scale). Mean lineage rates are positively correlated with the fast25 fraction (E) and negatively correlated with the slow25 fraction (F); dashed vertical lines indicate the global expectation under the quartile definition (0.25). (G–H) Cells in the highest decile of mean lineage rates are enriched for fast25 lineages (G) and depleted of slow25 lineages (H), visualized as violin plots with embedded boxplots; dashed horizontal lines indicate the 0.25 expectation. Some tropical regions, including the Amazon basin, show substantial representation from both fast- and slow-rate tails, whereas others, including parts of central Africa, appear less strongly enriched for either extreme; compositional maps also suggest a transition across Wallacea. Because these summaries are based on year-round resident distributions, such contrasts may reflect the combined effects of climatic history, dispersal barriers, lineage turnover, and sampling structure.$

Jake Berv@jakeberv· 22d

In addition to temporal patterns, we investigated macroecological patterns across space and discovered a strong statistical signal of an inverse latitudinal gradient in mean rates of phenotypic evolution (after accounting for many potential covariates). Figure 3 (and supp info section 3 and 4).

Spatial predictors of phenotypic evolution across global assemblages. a, Predicted grid-cell mean multivariate phenotypic rates (inverse range
weighted; Methods), from an aggregated SAR-Lag model including key
predictors: temperature seasonality (BioClim4, °C × 100), species richness, and latitude, as well as additional controls for sampling and speciation rates (Methods). Estimated (Fig. 1a and Supplementary Figs. 8 and 9) and predicted mean rates are typically higher at higher latitudes and are well approximated by a simple quadratic relationship with latitude (Supplementary Fig. 13). b, Temperature seasonality (BioClim4) shows spatial patterns similar to a. c, Aggregated SAR-Lag model (in a) residuals projected across space, showing minimal residual structure after accounting for spatial autocorrelation.

Jake Berv@jakeberv· 22d

Here is Supplementary Figure 18, which better showcases the beautiful illustrations we commissioned from @umich.edu undergrad Olivia Stein. These illustrated species represent clades where our method identifies significant increases in the rate of body-shape evolution. Also see Supp Table 7.

Supplementary Figure 18. from Berv et al 2026. Figure 1 is reproduced, with enlarged illustrations that are approximately to scale relative to one another and suitable for printing. Illustrated species
names are noted in Supplementary Table 7.

Jake Berv@jakeberv· 29d

RIP Sam Neill. His portrayal of Dr. Alan Grant contributed significantly to inspiring a whole generation of paleontologists, including myself. Can't be overstated, really.

Matt Ferguson, Jurassic Park, They Do Move in Herds

Jake Berv@jakeberv· Jul 1

Don’t forget to RSVP for Sunset Cocktails #WidowsBay

Jake Berv@jakeberv· Jun 26

Jake Berv@jakeberv· Jun 23

Perfect

Jake Berv@jakeberv· Jun 22

#Evol2026 I’m launching a new research group at UNC Charlotte in January 2027. Prospective students, postdocs, and collaborators are welcome to get in touch: jacob.berv@charlotte.edu More at jakeberv.com, lab website coming soon.