Cymbiodyta, Bedel, 1881

Toussaint, Emmanuel F. A. & Short, Andrew E. Z., 2019, Historical Biogeography of Holarctic Cymbiodyta Water Scavenger Beetles in the Times of Cenozoic Land Bridge Dispersal Routes, Insect Systematics and Diversity (AIFB) 3 (5), No. 8, pp. 1-7 : 5-6

Evolution of Cymbiodyta

Our BEAST dating analyses converged well and the one with two clocks and a birth–death tree model (A2, see Table 2) was preferred based on comparison of marginal likelihoods ( Table 2). The chronogram resulting from this analysis is presented in Fig. 2 . This dating analysis indicates that the divergence between C. marginella and the remaining species of Cymbiodyta (including Helocombus ) occurred in the Upper Cretaceous ca. 96 Ma (95% confidence interval: 69.59–130.14 Ma; Table 3). This estimate is slightly younger than the ones from Toussaint et al. (2016) and Toussaint and Short (2018), likely due to the use of secondary calibrations in this study and of multiple fossil calibrations in the two other ones. However, these differences in divergence time estimates are unlikely to change the biogeographic implications within Cymbiodyta (see below). The DEC analyses significantly supported a Nearctic origin of Cymbiodyta (log-likelihood [LnL] = −12.56) when compared with an Oriental (LnL = −17.83) or Palearctic (LnL = −15.39) origin. The biogeographic pattern within the phylogeny is extremely conserved with all nodes within the phylogeny having an ancestral range in the Nearctic region ( Fig. 2 ).

Median age estimates with 95% credibility interval in millions of years.

Our dating and biogeographic analyses suggest an origin of Cymbiodyta in the Nearctic region about 100 Ma, when the Nearctic and western Palearctic regions were connected by land within Laurasia (Seton et al. 2012). Although our data recover the Nearctic region as the most likely ancestral range with significant likelihood difference, additional sampling in particular among Enochrinae will be needed to test this hypothesis in a more comprehensive framework. An origin in the Palearctic region would also be plausible but is not preferred by our data. The range evolution of the genus is fully conserved throughout geological times ( Fig. 2 ), indicating a geographic stasis in the Nearctic region with ‘in situ’ diversification most likely caused by fine-scale allopatric speciation events that cannot be investigated without a denser sampling across Nearctic Cymbiodyta . We infer a comparatively recent dispersal toward the Oriental region in the past 30 Ma. This dating is compatible with landmass configuration at that time because the Oriental and Nearctic regions were geographically close albeit separated by a water corridor (Seton et al. 2012). However, our divergence time estimates are incompatible with a Beringian route and associated geodispersal colonization that would require a split between the Nearctic and Oriental lineages ca. 58 or 65.5 Ma ( Brikiatis 2014). The credibility interval we estimate for this particular split does not overlap with the existence of a Beringian land bridge in the Eocene, and therefore oversea dispersal is the most likely scenario based on our data ( Fig. 2 ). The presence of C. marginella in the Palearctic is probably the result of dispersal from the Nearctic region; however, considering that this lineage is placed on a long branch and as sister to the rest of the genus, it is not possible to discuss in more detail the possible timing of this event. It is possible that the mechanisms of lineage dispersal from Nearctic to Palearctic are related to land bridges throughout the Cenozoic, albeit it is not possible to disentangle which between the De Geer and Thulean routes would be the most likely explanation for the colonization of this new geographic range.

Figures

Fig. 2. Bayesian median divergence time estimates of Cymbiodyta. Maximum clade credibility chronogram resulting from the best BEAST analysis using secondary calibrations with outgroups pruned.The credibility intervals are indicated with gray horizontal boxes.The best ancestral range estimation scheme from the DEC analyses is presented with the most likely ancestral range for each node along the phylogeny of the genus.Color codes follow the inserted caption.