1. Introduction

Ferns comprise a group that is second only to the flowering plants, with a diversity greater than that of gymnosperms (Mcelwain 2011). The fossil record of ferns extends back to the Devonian and their origin is inferred to have been in the mid–late Silurian (Testo & Sundue 2016) in the Palaeozoic over 400 Mya, thus revealing their ecological success over time. They continued to prosper and were able to maintain their diversity during the rise of the angiosperms that made their appearance in the Early Cretaceous (Friis & Endress 1990).

The mid-Cretaceous was an extremely important period in the development of ferns and flowering plants. Both groups were evolving and competing for space, sunlight and nutrients with various established gymnosperms. Representatives of five fern families (Dennstaedtiaceae, Cystodiaceae, Lindsaeacea, Thyrsopteridaceae and Pteridaceae) are currently known to have been part of the flora of the Burmese amber forest (Table 1). Both ferns and angiosperms had to contend with invertebrate and vertebrate herbivores. Ferns had a much longer period to adjust to insect herbivores than angiosperms and that may be why today, there are fewer insect groups attacking ferns than angiosperms.

The purpose of the present paper is to describe a new fern of the family Dennstaedtiaceae in Burmese amber, to compare it with extant ferns as well as those previously described from this source, to discuss the presence of a possible herbivore and to suggest interactions that probably occurred between mid-Cretaceous ferns and evolving angiosperms.

2. Material and methods

The specimen originates from the Noije Bum 2001 Summit Site mine first excavated in the Hukawng Valley in 2001 and located southwest of Maingkhwan in Kachin State (26°20′N, 96°36′E) in Myanmar. Based on palaeontological evidence this site was dated to the late Albian of the Early Cretaceous (Cruickshank & Ko 2003: fig. 1), placing the age at 97–110 Ma. A later study using U-Pb zircon dating determined the age to be 98.79 ± 0.62 Ma, at the Albian/Cenomanian boundary (Shi et al. 2012). A more recent zircon U-Pb and trace element analyses of amber from different locations in northern Myanmar confirmed an age of around 100 Ma for amber from the Hukawng Valley as well as an age range of 72–110 Ma for amber from other sites in northern Myanmar (Xing & Qui 2020: fig. 1). Nuclear magnetic resonance (NMR) spectra and the presence of araucaroid wood fibers in amber samples from the Noije Bum 2001 Summit Site indicate an araucarian tree source of the amber (Poinar et al. 2007).

Observations and photographs were made with a Nikon SMZ-10 R stereoscopic microscope and Nikon Optiphot compound microscope with magnifications up to 800 X. Helicon Focus Pro X64 was used to stack photos for better depth of field.

3. Systematic description

The fossil represents two terminal fragments of a fertile pinna, which together contain some 25 sori with developing sporangia and spores. One of the fragments is shown in Fig. 1.

Syninclusions include the remains of an immature blattoid.

Table 1.

Ferns described from Burmese amber.

Table 2.

Characters of other genera in the Dennstaedtiaceae that differ from Cladarastega burmanica gen. et sp. nov. (after YaÑez et al. 2014; Brownsey 1998; Smith et al. 2006).

Fig. 1.

Terminal fragment of frond with sori of Cladarastega burmanica gen. et sp. nov. in Burmese amber. Bar = 1.2 mm.

4. Discussion

Presently, seven genera of ferns from five families have been described from Burmese amber (Table 1). Cretacifilix fungiformis Poinar & Buckley (2008) possessed oval-shaped monolete spores, reniform indusia, embossed sori, and sporangia with a short annulus, which distinguishes it from Cladarastega. The sori of Krameropteris resinatus Schneider, Schmidt & Hendriks (2016) in the family Dennstaedtiaceae lack indusia, which differs from Cladarastega. Cystodium sorbifolioides Regalado, Schmidt, Schneid, Krings & Hendrichs (Regalado et al. 2017a), in the family Cystodiaceae lacks scales, which separates it from Cladarastega. Thyrsopteris cretacea Li & Moran (Li et al. 2019) in the family Thyrsopteridaceae possesses terminal sori with cyathiform indusia, which distinguishes it from Cladarastega. Heinrichsia cheilanthoides Regalado, Schmidt, Krings & Schneider (Regalado et al. 2019) in the family Pteridaceae has a pseudoindusium that separates it from the present fossil. A member of the Lindsaeaceae was also reported from Burmese amber (Regalado et al. 2017b). The sori follow a continuous line in the upper margin of the segments, which distinguishes it from Cladarastega. Another member of the same family was described as Proodontosoria myanmarensis Li & Moran (Li et al. 2020). It possesses linear sori, which separates it from Cladarastega.

The mid-Cretaceous age of Burmese amber (∼100 Mya) approaches the 113.993 Mya estimated molecular age of the dennstaedtioides as determined by Schneider et al. (2004). Divergence time estimates for the Dennstaedtiaceae based on rbcL DNA sequence data and analysis carried out in BEAST shows the clade appearing some 90 MYA with the genus Microlepia appearing between 30 and 40 Mya. Using relaxed clock estimates and the assignment of the Krameropteris fossil to the split between the Monachosorum clade and Hypolepidioideae clade using Yule parameter and other analyses (further details provided by Schneider et al. 2016), average ranges of 95 Mya to 137 Mya were presented for the appearance of the family Dennstaedtiaceae (Schneider et al. 2016).

Fig. 2.

Leaf appendages of Cladarastega burmanica gen. et sp. nov. in Burmese amber. A – Elongate multi-segmented epidermal hair. Bar = 90 µm. Insert shows detail of short portion of hair. Bar= 25 µm. B – Trichomes adjacent to pinnule. Bar= 0.6 mm. C –Simple setae on veinlet. Bar = 80 µm. D – Epidermal gland. Bar = 63 µm.

Fig. 3.

Indusia of Cladarastega burmanica gen. et sp. nov. in Burmese amber. A – Disintegrtion of the inner indusium (arrows). Bar = 170 µm. B – Outer indusium (left arrow) and inner indusium (right arrow). Bar = 180 µm. C – Partial outer indusium (right arrow) and inner indusium (left arrow). Bar = 200 µm. D – Detail of partial outer indusium (arrow). Arrowheads show paraphyses. Bar = 10 µm.

Fig. 4.

A – Sporangium of Cladarastega burmanica gen. et sp. nov. in Burmese amber. Arrow shows papillate structures lining inside wall of lip portion of sporangium. Bar = 48 µm. B – Detail of papillate structures (arrowheads). Bar = 8 µm. C – Two sporangia of Cladarastega burmanica in Burmese amber. Arrowheads show spores. st= stoma; lc= lip cell. Bar = 103 µm. D – Two peltate scales on epidermal surface of pinna of Cladarastega burmanica in Burmese amber. Bar = 33 µm.

Fig. 5.

A – Polar view of perispore 1 of Cladarastega burmanica gen. et sp. nov. Bar = 19 µm. B – Polar view of perispore 2 of Cladarastega burmanica. Bar = 19 µm. C – Lateral view of perispore 3 of Cladarastega burmanica. Bar = 12 µm. D – Lateral view of perispore 4 of Cladarastega burmanica. Bar = 13 µm. E – Partial remains of immature blattoid adjacent to pinnule of Cladarastega burmanica. Bar = 0.6 mm.

Ferns and angiosperms were competing for space, sunlight and nutrients in the Burmese Amber forest. It has been suggested that the spread of flowering plants in the terrain resulted in the formation of new niches in forests that could be occupied by leptosporangiate ferns, allowing them to further diversify (Schuettpelz & Pryer 2009). Both ferns and angiosperms had to contend with herbivores. Many flowers recovered from Burmese amber show definite signs of insect damage (Chambers & Poinar 2020) and in some cases, the actual insect herbivore is entombed with the flowers (Poinar & Chambers 2018). Very little is known about insect herbivores of fossil ferns. That is why the partial remains of an immature blattoid adjacent to the pinna of Cladarastega is interesting (Fig. 5E). What relationship this insect had with Cladarastega is not known, however, Late Palaeozoic cockroaches were speculated to have fed on fern spores (Scott & Taylor 1983) and Blaberus giganteus (L.) was attracted to the axillary nectaries of extant bracken ferns in North America (Douglas 1983).

Certainly, some of the present day fern herbivores, such as sawflies (Tenthredinidae: Hymenoptera), gall gnats (Ceccidiomyiidae: Diptera) and aphids (Aphididae: Hemiptera) (Poinar 2016) that have a fossil record extending back at least to the Early Cretaceous (Rasnitsyn & Quicke 2002) could have fed on Burmese amber ferns.

Aside from herbivorous insects, herbivorous dinosaurs also probably influenced the evolution of various fern lineages. Early Cretaceous sauropods were considered to be the dominant herbivores that fed on ferns as well as angiosperms (Ryan 1997). Tiffney (1997) suggested that large sauropods could have fed in “fern prairies” that provided a disturbance-tolerant and high quality food source. Ferns are fascinating plants and those found in Burmese amber represent some of the most detailed fossil representatives available for study.