Cenozoic climatic changes drive evolution and dispersal of coastal benthic foraminifera in the Southern Ocean

Abstract

The Antarctic coastal fauna is characterized by high endemism related to the progressive cooling of Antarctic waters and their isolation by the Antarctic Circumpolar Current. The origin of the Antarctic coastal fauna could involve either colonization from adjoining deep-sea areas or migration through the Drake Passage from sub-Antarctic areas. Here, we tested these hypotheses by comparing the morphology and genetics of benthic foraminifera collected from Antarctica, sub-Antarctic coastal settings in South Georgia, the Falkland Islands and Patagonian fjords. We analyzed four genera (Cassidulina, Globocassidulina, Cassidulinoides, Ehrenbergina) of the family Cassidulinidae that are represented by at least nine species in our samples. Focusing on the genera Globocassidulina and Cassidulinoides, our results showed that the first split between sub-Antarctic and Antarctic lineages took place during the mid-Miocene climate reorganization, probably about 20 to 17 million years ago (Ma). It was followed by a divergence between Antarctic species ~ 10 Ma, probably related to the cooling of deep water and vertical structuring of the water-column, as well as broadening and deepening of the continental shelf. The gene flow across the Drake Passage, as well as between South America and South Georgia, seems to have occurred from the Late Miocene to the Early Pliocene. It appears that climate warming during 7-5 Ma and the migration of the Polar Front breached biogeographic barriers and facilitated inter-species hybridization. The latest radiation coincided with glacial intensification (~ 2 Ma), which accelerated geographic fragmentation of populations, demographic changes, and genetic diversification in Antarctic species. Our results show that the evolution of Antarctic and sub-Antarctic coastal benthic foraminifera was linked to the tectonic and climatic history of the area, but their evolutionary response was not uniform and reflected species-specific ecological adaptations that influenced the dispersal patterns and biogeography of each species in different ways.

Figure 1

Maps showing sampling localities (white circles) and, in colour, schematically marked ranges of different species of Globocassidulina (left map) and Cassidulinoides (right map), based only on data presented in this paper. For Cassidulinoides, ranges of morphotypes are indicated by different hatching. Map data from https://freevectormaps.com. Location of the Subantarctic (SAF), Polar (PF), and southern Antarctic Circumpolar Current (SACCF) fronts after.

Figure 2

Molecular operational taxonomical units (MOTUs) or species, marked by shades of different colors, based on six different species delimitation analyses. MOTUs with single sequences not indicated. Bayesian ultrametric tree is based on haplotype data reconstructed in BEAST 2.5. Sequence numbers are colored according to sampling location.

Figure 3

SEM images of Cassidulinidae from the Drake Passage and the Scotia Sea sector of South Atlantic, including West Antarctica: (1–3) Globocassidulina aff. rossensis from southern Patagonia, the Falklands, and South Georgia; (4–5) Globocassidulina biora from South Shetlands, immature form with bifurcated aperture and mature one with doubled aperture; (6) Globocassidulina aff. subglobosa from the Ross Sea; (7) Cassidulinoides parkerianus s.s. from southern Patagonia; (8, 13) Cassidulinoides aff. parkerianus, the smoothly-walled conical morphotype from South Georgia; (9–12) Cassidulinoides aff. parkerianus, the porous morphotype from South Georgia (two specimenes), South Shetlands, and the Falklands; (14–16) Cassidulinides parvus, granular morphotype from South Shetlands, megalospheric and microsphelic variants of the smoothly-walled tubular morphotype from the Ross Sea. For more images showing full morphological diversity, see Appendix 3.

Figure 4

Haplotype networks of different species of Southern Ocean Globocassidulina based on matrices constructed from partial SSU rDNA sequences. The area of the circles is proportional to haplotype frequency. Different colors represent different locations.

Figure 5

Haplotype networks of different species of Cassidulinoides, based on matrices constructed from partial SSU rDNA sequences. The area of the circles is proportional to haplotype frequency. Different colors represent different locations.

Figure 6

Historical demographic trends of population size constructed using Bayesian skyline plot approach based on SSU rDNA of Antarctic and sub-Antarctic Cassidulinidae. The y-axis (population size) is on a log scale;the x-axis is the time in 10³ (left column) and 10⁶ (right column) years before present. Results of neutrality Tajima's D and Fu's Fs tests are indicated in Appendix 6.

Figure 7

Bayesian time calibrated phylogeny reconstruction using foraminiferal SSU rDNA, showing major divergences in Southern Ocean Cassidulinindae. The posterior probabilities provided above the nodes and supplemented with ML bootstrap values. Sequence ids from Antarctica (RS Ross Sea, ROT Rothera, Margarite Bay, ADM Admiralty Bay, South Shetlands) are marked in blue, from southern Patagonia (PAT), Falkland Islands (FK) and South Georgia (SG) in green, and other in black. Id of sequences used to calibrate the molecular clock are in black; A and B—calibration points. Red circles show divergence events used for testing of the calibration; green bars show age-ranges for these events suggested by the fossil record. Gray bars on nodes show 95% Highest Posterior Density intervals for discussed divergence events; − not recovered in ML phylogenetic reconstruction shown in Appendix 7. Paleoceanographic events marked on the time scale (warm events in grey): E/O the Eocene/Oligocene cooling, MMCO the mid-Miocene Climatic Optimum, MMCT the Middle Miocene Climate Transition, LMTZ the Late Miocene strengthening of thermal bathymetric zonation, LMPW the Late Miocene to Pliocene warming, LPPGI the Late Pliocene to Pleistocene glacial intensification, MW the modern warming.