henhaobio
Sign in

Proceedings of the Royal Society B Biological SciencesPaleontologyPeer-reviewed journal article

Diversification dynamics of the Palaeozoic actinopterygian radiation

Joseph T. Flannery‐SutherlandiD, Struan HendersoniD, Sophie A. FaseyiD, Sam GilesiD

PDFDOIPreprintCode
In-depth full-text digest available · Jump to digest ↓

Abstract

Ray-finned fish are the most speciose vertebrate group today, but the dynamics of their early diversification are contentious. Their fossil record suggests a first radiation in the Carboniferous following the Late Devonian mass extinction events. Conversely, recent phylogenetic hypotheses imply a radiation originating in the Late Devonian, but lack the taxonomic breadth required to robustly test this. This necessitates phylogeny-free inference of actinopterygian diversification rates from fossil occurrences, itself challenging due to complex systematics, incomplete occurrence databases, and severe spatiotemporal sampling biases. Here, we analyse a comprehensive dataset of Palaeozoic actinopterygian genera and species using approaches that accommodate spatial and temporal sampling variation. We detect elevated actinopterygian diversification in the Late Devonian, with substantially greater lineage survival across the Hangenberg extinction event than indicated by the raw fossil record. Surprisingly, we detect no positive shifts in origination across the event, refuting previous hypotheses of explosive actinopterygian radiation in its wake. Instead, cryptic survival of diversified lineages appears responsible for the robust signal of increased diversity across different geographic scales in the Carboniferous. Nonetheless, these trends are overwhelmingly driven by low palaeolatitude Euramerican fossil assemblages, highlighting the ongoing spatial limitations of the actinopterygian fossil record.

In-depth summary

Claude Opus 5.5 · xhigh effort · 8 Nov 2026 · from the full text of version 2 (24 Oct 2025) · CC BY 4.0

Birth-death-sampling models of Palaeozoic ray-finned fish occurrences detect no origination increase after the Hangenberg event; cryptic Late Devonian diversification and survival likely explain Carboniferous diversity gains.

Raw fossil counts imply an explosive Carboniferous radiation of ray-finned fish after the Late Devonian extinctions, whereas a recent phylogenetic finding hints at Devonian roots, yet no prior study used a full taxon compilation or quantified their origination and extinction rates. This work pairs an expanded occurrence database with spatially standardised, phylogeny-free diversity and birth-death-sampling rate estimates.

Key results

  1. ShownNeither genus- nor species-level PyRate analyses detect a significant rise in net diversification across the Devonian-Carboniferous boundary, with or without spatial standardisation; rates instead dip modestly at the Carboniferous onset as origination falls significantly.

    PyRate: no statistically significant diversification increase at the boundary for genera or species in total or MST-standardised data; genus origination (λg) drops sharply in the early Tournaisian (total data) or across the boundary (standardised data); only weak evidence of raised Famennian diversification. · 3.3 Palaeozoic actinopterygian diversification…; 4.2 Actinopterygian diversification through mass…

  2. SuggestedRevBayes relaxes the assumption that every lineage was sampled. It infers lower sampling rates than PyRate and a significant Famennian increase in genus and species origination, which fits cryptic diversification before the Hangenberg event.

    Bayes Factor tests flagged few RevBayes shifts as significant, but Famennian genus and species origination increases were significant with and without MST standardisation; genus net diversification was strongly positive in the Middle Devonian, Famennian, Viséan and Kungurian, and near zero in the Tournaisian. · 3.3 Palaeozoic actinopterygian diversification…; 4.2 Actinopterygian diversification through mass…

  3. SuggestedPyRate's sampling-corrected stratigraphic ranges carry many known taxa across the Hangenberg event, implying substantial cryptic survival, although some boundary crossers may be artefacts of polyphyletic wastebasket genera.

    At least 33 known actinopterygian taxa may have true ranges spanning the Devonian-Carboniferous boundary; Kentuckia and Gonatodus are recovered as boundary crossers, but comparing their Devonian and Carboniferous species suggests both genera are polyphyletic. · 4.2 Actinopterygian diversification through mass…; Figure legends

  4. ShownGenus richness rose by about an order of magnitude across the Devonian-Carboniferous boundary. The rise persists after coverage-based rarefaction and Squares corrections, after MST standardisation, and when averaged over local equal-area grid cells.

    Global RTT diversity correlates with CBR (Spearman 0.82, p < 0.0001) and SQR (Spearman 0.97, p < 0.0001), with little change after MST standardisation; mean grid-cell SIB, CBR and SQR all show a shift from low to high local diversity across the boundary. · 3.2 Palaeozoic actinopterygian diversity trends; 4.1 Signals of sampling versus macroecology

  5. ShownObserved diversity tracks sampling intensity and, more weakly, spatial extent. MST subsampling dampens fluctuations in extent but barely changes Devonian-Carboniferous trends; the authors attribute this to the dominance of low-palaeolatitude Euramerican assemblages.

    SIB diversity vs occurrence counts: Spearman 0.90, p < 0.0001; vs MST extent: Spearman 0.34, p < 0.05. Subsampling to an 18000 km target reduced the MST length SD from 8942 km to 6801 km, with most bins within ±10% of target from the Serpukhovian onward. · 3.1 Palaeozoic actinopterygian sampling trends; 4.1 Signals of sampling versus macroecology

Methods, data and reproducibility

Paleobiology Database (PBDB) occurrences were binned by formation and spatially standardised by minimum spanning tree (MST) subsampling and equal-area hexagonal grids. They were analysed with sampled-in-bin, range-through-time, coverage-based rarefaction and Squares richness, plus PyRate and RevBayes birth-death-sampling rate models.

  • Early Devonian to Late Permian genus-level actinopterygian occurrences from PBDB (taxa described up to February 2024), 1644 resolved to species, in stage- to substage-constrained collections · ray-finned fish (Actinopterygii, fossil) · n = 2142 occurrences; 907 collections
  • Early Triassic actinopterygian occurrences added to reduce edge effects at the end of the study window · ray-finned fish (Actinopterygii, fossil) · n = 212 occurrences; 62 collections
  • Analysis scripts and data deposited in the paper's electronic supplement on figshare · 10.6084/m9.figshare.28789472
Methods & tools
PyRate (hierarchical birth-death-sampling model), RevBayes (skyline fossilised birth-death range model), divDyn R package (SIB and RTT richness), iNEXT R package estimateD (coverage-based rarefaction, quorum 0.5), Squares richness extrapolation (Alroy equation), icosa R package (equal-area hexagonal grids), MST spatial subsampling (Flannery-Sutherland et al. method), Formation binning (Dean et al. approach), Paleobiology Database (PBDB), Bayes Factor tests for rate shifts
Reproducibility gaps
Author scripts and data are in a figshare supplement given only as DOI 10.6084/m9.figshare.28789472 (no URL); PBDB occurrences were compiled to February 2024, and cleaning, binning and rate-model settings appear only in the supplementary information.
Limitations (6)
  • The record is heavily skewed toward intensely sampled, low-palaeolatitude Euramerican assemblages, so truly global trends cannot be identified and the mostly Euramerican signal has to be assumed representative. · stated by the authors
  • Wastebasket genera and over-splitting (e.g., Rhadinichthys, Elonichthys) distort species-level counts. They can create artefactual boundary crossers such as Kentuckia and Gonatodus and may drive the abnormal Early Permian global CBR and SQR spikes. · stated by the authors
  • Euramerican Lagerstätten (e.g., Glencartholm, Bear Gulch, Mazon Creek, Montceau Les Mines) coincide with high-diversity intervals and may inflate richness even after correcting for sampling effort. · stated by the authors
  • The 18000 km MST target was not reached in the Early and Middle Devonian or the Viséan. Early shortfalls may reflect genuine geographic restriction, but later deficient bins likely reflect poor sampling, so the authors avoid interpreting them biologically. · stated by the authors
  • Frasnian-Famennian inferences are fragile: MST subsampling may drop Gogo Formation taxa, and resolving the roughly 0.1 Ma Kellwasser pulses would need section-level stratigraphic analysis. · stated by the authors
  • The cryptic Late Devonian diversification scenario rests mainly on RevBayes's incomplete-sampling assumption, since PyRate gives only weak Famennian support; no phylogenetic analysis tests the proposed early divergences. · noted in this summary

Read this if you study early vertebrate macroevolution or recovery from mass extinctions, or need to estimate diversification rates without a phylogeny from spatially biased fossil occurrence data.

Citations
2
FWCI
1.40
Field percentile
top 19%

Reproducibility ledger

Nobody has reported trying to reproduce this paper yet. Did you replicate an experiment or run the code? Your report saves the next lab a week.

Sign in to file a reproducibility report.

Related papers

Similarity Map
SelectedSame fieldCross-field
THISGiles '21Henderson '22Peoples '26Flannery‐Su '26Brownstein '24Santos '26
84% semantic matchPaleontology
S. Giles, K. Feilich, R. Warnock+2 more·bioRxiv·Sep 2, 2021