A perfect flower in a mid-Cretaceous (early Cenomanian) Myanmar amber is described as Lijinganthus revoluta gen. et sp. nov. The fossil flower is actinomorphic and pentamerous, including calyx, corolla, stamens, and gynoecium. The sepals are tiny, while the petals are large and revolute. The stamens are dorsifixed, filamentous, and each has a longitudinally dehiscing bisporangiate anther. The gynoecium is in the centre of the flower, composed of three fused carpels with a stout style. Lijinganthus revoluta gen. et sp. nov. demonstrates a great resemblance to the flowers of Pentapetalae (Eudicots), adding new information to the enigmatic early evolutionary history of Pentapetalae and Eudicots.
An increasing number of insects and plants have been reported in a mid-Cretaceous Myanmar ambers1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23. Among them, fossil flowers have shed light on the diversification of angiosperms during this important radiating period for angiosperms2,18,19,20,21,24. Core Eudicots comprise a major portion of the species diversity in extant angiosperms, and they underwent a rapid increase in diversity and abundance at the transition between the Early and Late Cretaceous19,25,26,27,28. Hitherto, the earliest record of a flower with distinct sepals and petals is approximately 94 Ma (the Cenomanian)29,30 (However, according to the latest study (Manchester et al.27), some fruit specimens belonging to the same taxon may be dated back to 105 Ma (the Albian)). Various molecular clocks indicate that angiosperms and Eudicots have a significantly earlier origin than the earliest fossil record indicates31,32,33. The gap between these estimations and the fossil record makes many conclusions in angiosperm systematics tentative. Here, we describe a new flower, Lijinganthus revoluta gen. et sp. nov., from an earliest Cenomanian-latest Albian (98.79 Ma)34 amber, which was collected from Noije Bum 2001 Summit Site, Hukawng Valley, Kachin, Myanmar (26°20′N, 96°36′E) (Fig. 1). This bisexual, pentamerous, actinomorphic flower with distinct sepals and petals, filamentous bisporangiate stamens, and trimerous gynoecium with superior ovary and axile placentation is unique, demonstrating a great resemblance to Pentapetalae (Core Eudicots) and thus shedding a new light on the evolution of Core Eudicots. Lijinganthus, as a representative of Pentapetalae, is among the first Core Eudicots. Together with contemporary fossil finds19,25,26,27,28, Lijinganthus suggests a Core Eudicot Boom at the transition from the Early to Late Cretaceous and helps to narrow the gap between fossil record and molecular clock estimates. The presence of a nectary disk in Lijinganthus suggests the possibility of insect-mediated pollination.
Family incertae sedis
Lijinganthus gen. nov.
Type species: Lijinganthus revoluta gen. et sp. nov.
Diagnosis: Flower actinomorphic, bisexual, with one whorl of five sepals, one whorl of five petals, androecium, and gynoecium. The calyx consisting of five distinct, small sepals with entire margins and rounded apices. The corolla consisting of five distinct, large, elongated oval, entire-margined revolute petals with rounded apices, alternating with the sepals. Androecium diplostemonous, filamentous. The anthers bisporangiate, introrse, dorsifixed, dehiscing by longitudinal slits. Nectary disk at the base of the gynoecium. The gynoecium tricarpous, lacking an obvious style and probably with axile placentation.
Lijinganthus revoluta gen. et sp. nov.
Description: The flower (including the pedicel) is 6.5 mm long and 4.8 mm wide. The flower includes a pedicel, sepals, petals, stamens, and a gynoecium (Figs 2a–c, 3h and 4a). The pedicle is 0.2 mm in diameter and 3.8 mm long, slightly tapering distally. There are five distinct sepals, each 0.4-0.5 mm long and 0.32-0.33 mm wide, with an obtuse or round tip (Figs 2d, 3f,h and 4a). Alternating with the sepals are five distinct petals, each approximately 1.8 mm long and 0.67 mm wide, revolute (Figs 2a–d, 3a,b and 4a). At least eight distinct stamens are seen inserted on the nectary disk at the base of the ovary, each including a slender filament and a dorsifixed bisporangiate anther (Figs 2a–d, 3c–e,h and 4a,b). The filament is approximately 43 μm wide, 1.3 mm long, tapering distally (Figs 2a–d, 3c–e and 4a,b). The anther dehisces longitudinally, 0.5 mm wide and 0.5 mm long, without in situ pollen grains (Figs 3c–e,h and 4b). Several clumps of a single type of pollen grains are present in the same amber block and closely associated with the flower (Figs 3i,m and S1a–q). Pollen grains are tricolpate, 14–21 μm long and 10–12 μm in diameter (Fig. 3i–l). Nectary surrounds the base of the ovary (Fig. 3g–h). The gynoecium is situated centrally, including an ovary that tapers distally with no obvious style (Figs 3g,h and 4a,b). The ovary appears to be composed of three fused carpels, probably with axile placentation (Fig. 3g). The ovary is approximately 0.77 mm in diameter, with its tip only 0.13 mm in diameter, and round triangular in cross view (Figs 2b,c, 3g,h, 4b and 5).
Etymology: The generic name, Lijinganthus, is dedicated to poetess Ms. Jing Li (1967–2015) for her talent and poetry, and “anthus” from the Latin “anthos” (flower). The specific epithet revoluta is for the revolute form of the petals.
Type species: Lijinganthus revoluta gen. et sp. nov.
Deposition: The Nanjing Institute of Geology and Palaeontology, Chines Academy of Sciences.
Type locality: Noije Bum 2001 Summit Site, Hukawng, Kachin, Myanmar (26°21′33.41″N, 96°43′11.88″E).
Horizon: the earliest Cenomanian-latest Albian, Cretaceous (98.79 Ma).
Remarks: Although only eight stamens are preserved in the flower, the spatial relationship between stamens and petals (Fig. 2d) and the presence of five petals (Fig. 2a–c) suggest that the total number of original stamens in Lijinganthus should be ten. Further research need to be done to confirm this.
The character combination of the fossil does not allow us to assign it to any known fossil or extant genus of angiosperms, thereby justifying a new genus.
Comparison with extant angiosperms
The occurrence of over 800 single type of tricolpate pollen grains closely associated with Lijinganthus revoluta gen. et sp. nov. (preserved in the same block, and their distances from the flower range from 2.5 to 0 mm (Figs 3i,m and S1a–q)) strongly suggests a eudicot affinity for Lijinganthus, as tricolpate pollen grains are a characteristic feature of Eudicots, which are frequently termed Tricolpates. Among Eudicots, the occurrence of distinct calyx and corolla in Lijinganthus distinguishes it from the basal eudicots (with undifferentiated perianth)30 and Gunnerales (lacking perianth)35, suggesting that Lijinganthus belongs to the Pentapetalae in Core Eudicots. Several features of Lijinganthus are also seen in Crassulaceae (Saxifragales), but the latter has a basifixed anther, >3 more or less free carpels, decurrent stigma, and parietal placentation36,37 and is thus distinct from Lijinganthus with dorsifixed anther, fused carpels, capitate stigma, and axile placentation. Although the presence of distinct sepals and petals suggest that Lijinganthus is most likely related to the Superrosids, and its tricarpous gynoecium suggests a possible affinity to Malpighiales, we think it is premature to assign Lijinganthus to any group within Pentapetalae, at least for the time being.
Comparison with the Early Flowers of Core Eudicots
Previously, the earliest record of a flower with distinct calyx and corolla was marked by a fossil flower named “Rose Creek flower” from the Albian-Cenomanian30, which was recognized by Basinger and Dilcher in 198429 and reworked on and renamed as Dakotanthus cordiformis by Manhcester et al. (2018) with additional specimens (especially of fruits) using CT technology27. Lijinganthus is similar to Dakotanthus cordiformis in distinct pentamerous symmetry (5 sepals and 5 petals) and bisexuality, but it differs from the latter in long (rather than ovate) petals, a slender (rather than stout) filament, a bisporangiate (rather than tetrasporangiate) anther, tricolpate (rather than tricolporate) pollen grains, a trimerous (rather than pentamerous) gynoecium, and 1 (rather than 5) style27,29. Dakotanthus cordiformis is initially interpreted as approximately 94 Ma old29, but recent study, with more specimens (especially of fruits) from strata other than the original locality (Rose Creek), suggests that the age of Dakotanthus cordiformis may be extended to 105 Ma27. Dakotanthus cordiformis was considered to be “the first fossils with unequivocal features of core eudicots” from the Albian-Cenomanian30. The age of 98.79 Ma of Lijinganthus places our flower near the Albian/Cenomanian boundary (97.2 Ma to100.5 Ma by various authors). The concentrated occurrence of various flowers of different lineages (Lachnociona terriae (Brunelliaceae/Cunoniaceae (Oxalidaels) + Rosids + Saxifragles), Tropidogyne pikei and T. pentaptera (Cunoniaceae), Eoëpigynia burmensis (Cornaceae), Dakotanthus cordiformis (Quillajaceae)) and Lijinganthus at about the same time19,25,26,27,28 seems to suggest the Core Eudicots underwent a rapid diversification (“Core Eudicot Boom”) at the very beginning of the Late Cretaceous (Table 1). It is intriguing to investigate whether there is a coupling between this important plant event and the Upper Albian OAE 1d event (including rapid CO2 concentration rising)38 as well as the decline of Gnetales and Bennettitales.
The presence of nectary disk in Lijinganthus suggests that insects may have begun interacting with flowers by the Cenomanian. Whether such an interaction is a major driving force for the diversification of Core Eudicots is apparently a question deserving further investigation.
Various molecular clock studies have indicated that Eudicots and angiosperms originated much earlier than formerly assumed31,32,33. The discovery of Lijinganthus from the late Albian-Early Cenomanian (98.79 Ma) adds to the diversity and abundance of early Core Eudicots and helps to reconcile the conflicts between different schools and studies. Compatible with the molecular studies, the early age of Lijinganthus, together with recently found earlier-than-recorded fossils of Poaceae39 and Solanaceae40 as well as various contemporaneous Core Eudicots19,25,26,27,28 points to a cryptic and unexpectedly longer history of angiosperms and Eudicots. It is noteworthy that this conclusion is compatible with previously documented pre-Cretaceous traces of angiosperms41,42,43,44,45,46,47,48.
Lijinganthus revoluta gen. et sp. nov. and other contemporaneous fossil flowers suggest a Core Eudicot Boom at the very beginning of the Late Cretaceous. Increasing number of reports of early fossil flowers seem to converge earlier origins of various lineages that have been predicted by molecular clocks.
The specimen was collected from Noije Bum 2001 Summit Site, Hukawng Valley, Kachin, Myanmar (26°20′N, 96°36′E) (Fig. 1). Paleontological studies indicate that the specimen belongs to the earliest Cenomanian-latest Albian, Early Cretaceous (98.79 Ma)34, which is generally agreed on by Poinar et al.28 and Xing et al.49. Two parallel planes were made on the amber sample before observations. Observations and photographs were made with a Nikon SMZ1500 stereoscopic microscope at the Nanjing Institute of Geology and Palaeontology, Nanjing, China. Micro-CT was performed using a Zeiss Xradia 520 versa X-ray microscope at the Nanjing Institute of Geology and Palaeontology, Nanjing, China. The 3D reconstruction and virtual sections were generated using VGStudio MAX 3.0. All figures were organized for publication using Photoshop 7.0.
The holotype (PB22841) is accessible in the palaeobotanical collection of the Nanjing Institute of Geology and Palaeontology, Chines Academy of Sciences, 39 Beijing Dong Road, Nanjing 210008, China.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
We appreciate Dr. Zongjun Yin and Ms. Suping Wu for help with Micro-CT, Mr. Yan Fang for help with the microscopy, and Ms. Lijun Chen for drawing the reconstruction. We appreciate Dr. Walter S. Judd for his suggestion on the possible affinity of this fossil taxon. This research is supported by the Strategic Priority Research Program (B) of Chinese Academy of Sciences (XDPB26000000), the National Natural Science Foundation of China (41688103, 91514302, 91114201) awarded to X.W. We thank the support from the State Key Laboratory of Palaeobiology and Stratigraphy, Nanjing Institute of Geology and Palaeontology. We appreciate the constructive and detailed suggestions from two anonymous reviewers that made this paper much improved.