Early evolution of the angiosperm clade Asteraceae in the Cretaceous of Antarctica

Abstract

The Asteraceae (sunflowers and daisies) are the most diverse family of flowering plants. Despite their prominent role in extant terrestrial ecosystems, the early evolutionary history of this family remains poorly understood. Here we report the discovery of a number of fossil pollen grains preserved in dinosaur-bearing deposits from the Late Cretaceous of Antarctica that drastically pushes back the timing of assumed origin of the family. Reliably dated to ∼76-66 Mya, these specimens are about 20 million years older than previously known records for the Asteraceae. Using a phylogenetic approach, we interpreted these fossil specimens as members of an extinct early diverging clade of the family, associated with subfamily Barnadesioideae. Based on a molecular phylogenetic tree calibrated using fossils, including the ones reported here, we estimated that the most recent common ancestor of the family lived at least 80 Mya in Gondwana, well before the thermal and biogeographical isolation of Antarctica. Most of the early diverging lineages of the family originated in a narrow time interval after the K/P boundary, 60-50 Mya, coinciding with a pronounced climatic warming during the Late Paleocene and Early Eocene, and the scene of a dramatic rise in flowering plant diversity. Our age estimates reduce earlier discrepancies between the age of the fossil record and previous molecular estimates for the origin of the family, bearing important implications in the evolution of flowering plants in general.

Keywords: Antarctica; Asteraceae; evolution; fossil; phylogenetics.

Fig. 1.

Map showing distribution of Upper Cretaceous rocks of the Snow Hill Island and López de Bertodano Formations. The studied sections in Brandy Bay–Santa Marta Cove (James Ross Island) and Cape Lamb (Vega Island) are also indicated. Adapted from Olivero (7).

Fig. S1.

Stratigraphic sections of the Upper Cretaceous Snow Hill Island and López de Bertodano Formations. (A) Stratigraphic section of the Snow Hill Island Formation at Santa Marta Cove, James Ross Island with the situation of the studied samples and Ammonite Assemblages [Assemblages 8–9, adapted from Olivero (7)]. (B) Stratigraphic section of the Snow Hill Island and López de Bertodano Formations at Cape Lamb, Vega Island with the situation of the studied samples. To highlight the stratigraphic continuity of the samples, the lower 100 m of the section includes the Gamma Member of the Snow Hill Island Formation exposed on Humps Island, which bear the same Ammonite Assemblages 8-2 and 9 (Assemblages 8-2 and 9) recorded in Santa Marta Cove area (see A).

Fig. S2.

Extant species of Campanulaceae and Ranunculaceae that bear superficial similarities with the fossil T. lilliei type A. (A) Canarina canariensis (L.) Vatke shows a general resemblance to T. lilliei. (B) Clematis montevidensis Spreng., family Ranunculaceae, has been previously considered related to T. lilliei. Both Canarina and Clematis, show marked differences in exine structure and sculpture. Note the clearly columellate exine structure (see Supporting Data, Systematic Remarks). (Scale bars, 5 µm.)

Fig. S3.

Details (SEM) of the fossil T. lilliei type A from the Late Cretaceous of Antarctica and extant representatives of Asteraceae, Campanulaceae, and Ranunculaceae. (A and D) Specimens of Tubuliflorides lilliei type A (blue frames). (A) Poorly defined intercolpal depression (arrowhead); (D) detail of sculpture, note microspine (arrowhead) and bacula. (B and E) Extant Dasyphyllum inerme (Barnadesioideae subfamily, Asteraceae, pink frames). (B) Well-defined intercolpal depression (arrowhead); (E) microspines and baculum (arrowhead). (C) Canarina canariensis (L.) Vatke (Campanulaceae, green square), verrucate-perforate sculpture. (F) Clematis montevidensis (Ranunculaceae, green square) details of the microechinate- microgranulate-punctate sculpture. (Scale bars, 2 µm.)

Fig. S4.

Specimens of Tubulifloridites lilliei (Couper) Farabee & Canright from the Late Cretaceous of New Zealand that bear strong similarities with T. lilliei type A. (Supporting Data, Systematic Remarks). Specimens on slide L5664/3 (Paparoa Coal Measures, Westland Plate VI, figures 17–18 in ref. 8). (A) Specimens in equatorial view with a poorly defined intercolpal depression (arrowhead), L40(1). (B) Specimen in subpolar view, Q40(0). (Scale bars, 5 µm.)

Fig. 2.

Phylogenetic analyses of the fossil taxa. Branching positions of the fossil T. lilliei type A mapped onto a backbone tree derived from a molecular analysis of Beaulieu et al. (4), with some asteracean taxa added, following a recent comprehensive analyses of Panero et al. (24). Thicker black lines indicate the most parsimonious (MP), one step less parsimonious (MP + 1), and two steps less parsimonious (MP + 2) positions for T. lilliei type A. Letters indicate the nodes used to calibrate alternative scenarios, A: Fig. 5; B–E: Fig. S5 and Table S2.

Fig. 3.

Fossil and extant representatives of Asteraceae observed by light microscopy. (A, B, D, E, G, H) Specimens of Tubulifloridites lilliei type A from the Late Cretaceous of Antarctica (blue frames). Specimens on slide BAPal. ex CIRGEO Palin 963b; (A and B) N42(4); (D and E) L36(0); (G and H) P57(1). (A, B, D, E) Equatorial view. (G and H) Subpolar view. (A and B) Exine thickened at the poles (arrowhead). (A, E, and H) Microechinate-baculate sculpture. (D and E) Thickened exine at apertures level (arrowhead). (B, G, and H) Poorly defined intercolpal depressions (arrowhead). (G and H) Rounded colpi ends. (C, F, and I) Pollen of extant species for comparison (pink frames). (C and I) Extant Dasyphyllum inerme (Rusby) Cabrera, with well–defined intercolpal depressions and rounded colpi ends comparable to those of T. lilliei type A (arrowheads). (F) Extant Dasyphyllum velutinum (Baker) Cabrera, with microechinate-baculate exine surface similar to that of T. lilliei type A. (Scale bars, 5 µm.)

Fig. 4.

Fossil and extant representatives of Asteraceae observed by scanning electron microscopy. (A, D, E, G) Specimens of Tubulifloridites lilliei type A from the Late Cretaceous of Antarctica (blue frames). (A) Subpolar view showing details of sculpture and poorly defined depressions (arrowhead); note the microgranulate apertural membrane. (D) Subequatorial view showing a poorly defined depression (arrowhead). (E) Polar view with small apocolpium and thickened colpi margins. (G) Equatorial view showing the microechinate-baculate-verrucate sculpture. (B) Specimen of Quilembaypollis tayuoides Barreda and Palazzesi from the Miocene of Patagonia (light blue frame) that shares morphological features with both the Cretaceous and extant asteraceous specimens; note the microechinate-baculate sculpture. (C, F, H, I) Extant species of Dasyphyllum (pink frames) showing variations in the development and number of intercolpal depressions. (C and H) Dasyphyllum inerme (Rusby) Cabrera. (F) Dasyphyllum latifolium (Gardner) Cabrera. (I) Dasyphyllum leptacanthum (Gardner) Cabrera. (Scale bars, 5 µm.)

Fig. S5.

Timing of diversification of Asterales using different calibration scenarios. Chronograms (A–E, scale at the bottom in Mya) estimated using a Bayesian relaxed clock calibrated with a previously described fossil inflorescence and pollen from the Eocene (crown Asteraceae, except Barnadesioideae and Famatinanthoideae) and the oldest eudicot records (ref. , and references therein) from the Cretaceous (see Table S4). Our newly discovered specimens from the Cretaceous of Antarctica (red arrow) were used to calibrate alternative nodes according to the results of our sensitivity analysis (see SI Materials and Methods, Estimation of Divergence Times, and Table S2).

Fig. 5.

Evolutionary timescale of the diversification of Asteraceae. Chronogram (scale on the right in Mya) estimated using a Bayesian relaxed clock calibrated with a previously described fossil inflorescence from the Eocene (“B”) and our newly discovered specimens from the Cretaceous of Antarctica (“A”). We assume that this Cretaceous species (T. lilliei type A) represents an extinct branch nested within Dasyphyllum (crown representative). Other possible calibration scenarios are illustrated in Fig. S5. Light-blue bars at nodes represent 95% credibility intervals on estimates of divergence times. Orange horizontal lines indicate the timing of the K–P extinction event and the Cenozoic’s warmest interval. Most subfamilies of Asteraceae diverged during the Paleogene, but the earliest divergence occurred in the Late Cretaceous. B., Barnadesia; Barn., Barnadesioideae (91 species); Card., Carduoideae (2,500+ species); D., Dasyphyllum; F., Famatinanthoideae (1 species); Goc., Gochnatioideae (90 species); H., Hecastocleidoideae (1 species); Mut., Mutisieae (254 species); Nass., Nassauvieae (313 species); On., Onoserideae (52 species); S., Schlechtendalia; Stiff., Stifftioideae (44 species); Wund., Wunderlichioideae (41 species); rest of Asteraceae (19,600 + species).

Fig. S6.

Placement of the fossils used in the calibration scenario 1. Fossils A (T. lilliei type A) is an extinct species of Dasyphyllum in the Crown Group 3 and hence we used the age of this Fossil A to calibrate the split between Dasyphyllum and its sister genus Barnadesia. Fossils B (Raiguenrayun cura + Mutisiapollis telleriae) are extinct taxa that we identified as stem relatives of Crown Group 1. We used its age to calibrate the split between Crown Group 1 (Stifftioideae; Wunderlichioideae; Gochnatioideae; Hecastocleidoideae; Carduoideae subfamilies and rest of Asteraceae) and Crown Group 2 (Onoserideae; Mutisioideae; Nassauvieae subfamilies). Barn., Barnadesioideae; Card., Carduoideae; F., Famatinanthoideae; Goch., Gochnatioideae; H., Hecastocleidoideae; Mut., Mutisieae; Nass., Nassauvieae; On., Onoserideae; Stiff., Stifftioideae; Wund., Wunderlichioideae.