Fossil-informed biogeographic analysis suggests Eurasian regionalization in crown Squamata during the early Jurassic

Abstract

Background: Squamata (lizards, snakes, and amphisbaenians) is a Triassic lineage with an extensive and complex biogeographic history, yet no large-scale study has reconstructed the ancestral range of early squamate lineages. The fossil record indicates a broadly Pangaean distribution by the end- Cretaceous, though many lineages (e.g., Paramacellodidae, Mosasauria, Polyglyphanodontia) subsequently went extinct. Thus, the origin and occupancy of extant radiations is unclear and may have been localized within Pangaea to specific plates, with potential regionalization to distinct Laurasian and Gondwanan landmasses during the Mesozoic in some groups.

Methods: We used recent tectonic models to code extant and fossil squamate distributions occurring on nine discrete plates for 9,755 species, with Jurassic and Cretaceous fossil constraints from three extinct lineages. We modeled ancestral ranges for crown Squamata from an extant-only molecular phylogeny using a suite of biogeographic models accommodating different evolutionary processes and fossil-based node constraints from known Jurassic and Cretaceous localities. We hypothesized that the best-fit models would not support a full Pangaean distribution (i.e., including all areas) for the origin of crown Squamata, but would instead show regionalization to specific areas within the fragmenting supercontinent, likely in the Northern Hemisphere where most early squamate fossils have been found.

Results: Incorporating fossil data reconstructs a localized origin within Pangaea, with early regionalization of extant lineages to Eurasia and Laurasia, while Gondwanan regionalization did not occur until the middle Cretaceous for Alethinophidia, Scolecophidia, and some crown Gekkotan lineages. While the Mesozoic history of extant squamate biogeography can be summarized as a Eurasian origin with dispersal out of Laurasia into Gondwana, their Cenozoic history is complex with multiple events (including secondary and tertiary recolonizations) in several directions. As noted by previous authors, squamates have likely utilized over-land range expansion, land-bridge colonization, and trans-oceanic dispersal. Tropical Gondwana and Eurasia hold more ancient lineages than the Holarctic (Rhineuridae being a major exception), and some asymmetries in colonization (e.g., to North America from Eurasia during the Cenozoic through Beringia) deserve additional study. Future studies that incorporate fossil branches, rather than as node constraints, into the reconstruction can be used to explore this history further.

Keywords: Biogeography; DEC; Fossils; Gondwana; Jurassic; Laurasia; Lizards; Snakes; Squamata.

Figure 1. Fossil localities of Jurassic squamates.

Map of earth with major plate boundaries adapted from Bird (2003). Red stars indicate the present-day localities of Jurassic squamate fossils (see Table 1), blue represent earliest Cretaceous records (Bittencourt et al., 2020), and purple indicate records across both horizons (Sigogneau-Russell, Monbaron & Russell, 1988; Lasseron et al., 2020).

Figure 2. The abbreviated geographic regions in the Mesozoic.

Map of Jurassic and Cretaceous paleocontinents from Scotese (2016) drawn at the (A) Early Jurassic (Toarcian, 180Ma), with the three areas of Laurasia (yellow) and Gondwana (purple), Northern Pangaea (green; the boundary of which is Laurasia plus the area designated by the dashed line), and Pangaea (blue) indicated with their transition boundaries; (B) Jurassic/Cretaceous boundary (145 Ma), at which time fossil squamates are known from all three combined areas (Fig. 1; Table 1; Evans, 2003); and (C) K-Pg boundary (66 Ma), after which we see a significant decrease in relative dispersal probabilities between areas (Table 2; Fig. 4).

Figure 3. Ancestral state reconstruction of Mesozoic squamates.

Reduced representation of the squamate backbone tree (Tonini et al., 2016), showing the best-fit estimates from the DEC+J model (see Appendix S1 for full results and uncertainty) for the geographic origins of early squamate lineages. Named clades of particular interest are discussed in the Results section. The nodes that were constrained are highlighted in red. The color scheme for the major combined areas is consistent throughout the rest of the article.

Figure 4. Ancestral range estimation of Squamata into the Cenozoic.

A circle tree (Jetz et al., 2012) showing the complete phylogeny of Squamata (Tonini et al., 2016) onto which the ancestral range estimations are mapped (A), which expands the temporal range from the K-Pg in Fig. 3 into the current time. (B) A network analysis of dispersals between the different biogeographic regions. The size of the circles corresponds to extant diversity. The width and density of the arrows correspond to the number of dispersals from one area to another in the direction the arrow is pointing. Red arrows indicate extinction.