A new species and new generic synonymy in the family Vietnamellidae (Insecta: Ephemeroptera) from mid-Cretaceous Burmese amber with notes on ancient dispersal across East Gondwana

Material

The well-preserved specimen of the holotype of B. inconspicua sp. nov. is embedded in mid-Cretaceous Burmese amber. The stone originates from a mine situated in Northern Myanmar, Kachin State, Tanai Township, Hukawng Valley, SW of Tanai City, which corresponds to the findings of all other material of fossil Vietnamellidae (Godunko, Martynov & Staniczek, 2021), as well as many other Mesozoic mayfly taxa from mid-Cretaceous Burmese amber (Chen & Zheng, 2023).

Based on UePb zircon dating, the Hukawng amber can be assigned to the Early Cretaceous, Upper Albian, with a maximum age of 98.79 ± 0.62 Ma (Shi et al., 2012), which is equivalent to the earliest Cenomanian (Gradstein, Ogg & Smith, 2004).

The female imaginal holotype of B. inconspicua sp. nov. belonged to the Patrick Müller collection (Käshofen, Germany) (former inventory number BUB–3325) and was kindly donated to the State Museum of Natural History Stuttgart (SMNS) and filed under the inventory number SMNS BU–385. The male imaginal holotype of Burmella paucivenosa (inventory number SMNS BU-179) and holotype of female imago of B. clypeata (inventory number SMNS BU-321) are also both housed at SMNS (Godunko, Martynov & Staniczek, 2021). The male imaginal holotype of Burmaheptagenia zhouchangfai is deposited at the Insect Collection of Jiangsu University of Science and Technology, China (inventory number CZT-EPH-MA5).

Specimen processing and imaging

The fossil material and comparative extant material was studied under a Leica M205 C (Leica Corporation, Wetzlar, Germany) and an Olympus SZX7 (Olympus Corporation, Tokyo, Japan) stereomicroscope. We used a Leica Z16 APO Macroscope equipped with a Leica DFC450 Digital Camera, using Leica Application Suite v. 3.1.8. to obtain stacked series of photographs with different levels of focus. Photo stacks were processed with Helicon Focus Pro 8.3.0 to obtain combined photographs with extended depth of field. Photographs were sharpened and adjusted in contrast and tonality in Adobe Photoshop™ (Adobe Systems Incorporated, San Jose, CA, USA). Drawings were made using a camera lucida on a Leica M205 C stereo microscope. Morphological characters of the holotype of Burmaheptagenia zhouchangfai were studied and interpreted based on images from the original description (Chen & Zheng, 2023). We processed the images using graphic tools embedded in Windows10 and 11.

Terminology and taxonomic notes

General anatomical terminology is based on Kluge (2004) and Bauernfeind & Soldán (2012). Abbreviations for wing veins used throughout the text follow Kluge (2004), with some modifications by Bauernfeind & Soldán (2012), Staniczek, Godunko & Kluge (2018), and Godunko, Martynov & Staniczek (2021). Morphological terms and abbreviations to describe thorax morphology used herein follow Kluge (2004).

Taxonomic notes on previously described fossil species of Vietnamellidae are based on examination of the type material in the State Museum of Natural History Stuttgart (SMNS) and published data (see above).

Nomenclatural acts

The electronic version of this article conforms to the requirements of the International Code of Zoological Nomenclature, and hence the new names contained herein are available under that Code from the electronic edition of this article. This published work and the nomenclatural acts it contains have been registered in ZooBank, the online registration system for the ICZN. The ZooBank LSIDs (Life Science Identifiers) can be resolved, and the associated information viewed through any standard web browser by appending the LSID to the prefix “http://zoobank.org/”. The LSID for this publication is urn:lsid:zoobank.org:pub:B1FAF587-9EB2-4CDB-BD71-5AE0C9C6AD78. The new species is registered under urn:lsid:zoobank.org:act:6B45EB0E-8DF7-49D8-B317-E564009C00DC.

The online version of this work is archived and available from the following digital repositories: PeerJ, PubMed Central SCIE and CLOCKSS.

Systematic palaeontology

Type species. Burmella paucivenosa Godunko, Martynov & Staniczek, 2021; ibid.: 101 (original designation)

Species composition. Burmella paucivenosa Godunko, Martynov & Staniczek, 2021 (male imago, SMNS BU–179); Burmaheptagenia zhouchangfai (Chen & Zheng, 2023) comb. nov. (male imago, CZT-EPH-MA5); Burmella clypeata Godunko, Martynov & Staniczek, 2021 (female imago, SMNS BU–321); Burmella inconspicua Godunko & Staniczek sp. nov. (female imago, SMNS BU–385).

Locality and horizon. Hukawng Valley, Kachin State, Myanmar (Burma); Cenomanian, mid-Cretaceous.

Amended diagnosis of adults (modified from Godunko, Martynov & Staniczek, 2021; see also Tables 1 and 2). Based on the study of the extinct taxa, as well as comparative analysis of extant Vietnamella, we provide here additional generic characters.

Measurements (i) Body length 5.75–7.00 mm; forewing length 4.64–5.45 mm; hind wing length 0.45–0.68 mm; hind wing very small, as long as 0.08–0.14 of forewing length;

Male head (ii) lower portion of compound eyes small and narrow;

Female head (iii) anterior part of eyes covered or uncovered by anterolaterally expanded clypeal shield;

Thorax (iv) mesonotum relatively narrow; medioscutum markedly extended distally; basisternum of mesothorax nearly rectangular or slightly trapezoidal, slightly elongated;

Forewings (v) with small number of cross veins; pterostigma with less than nine simple veins, not anastomosed; MP sector with small number of marginal intercalaries; only one intercalary vein between MP₂ and CuA; two secondary bifurcate veins in cubital sector, mostly not connected by additional cross veins; CuP smoothly curved towards wing base; free marginal intercalaries along ventral margin are distributed in all vein sectors;

Hind wings (vi) strongly rounded; distinct reduction of the number of cross veins; RS triad present; MA, MP and secondary CuA and CuP triads absent; costal process developed, rounded apically, situated centrally or slightly distally;

Abdomen (vii) with vestigial gill sockets recognizable at least on segments II–VI; no trace of paracercus;

Female genitalia (viii) subgenital plate well developed, distinctly convex, widely rounded; subanal plate triangular-shaped, rounded apically;

Male genitalia (ix) with large median projection of styliger plate, widely rounded apically; three distal segments of forceps strongly elongated and slender; segment II longest, 5–6 × as long as segment III; segments III and IV approximately of equal length or slightly longer; segment IV elongated; penis lobes widely separated by V-shaped cleft, tube-like or trapezoidal.

Material examined. Holotype. Female imago in mid-Cretaceous Burmese amber, housed at SMNS under inventory number BU–385. Well preserved specimen, embedded in dorsoventral aspect (Figs. 1A, 1B). Wings almost entirely preserved, except of distal parts of veins C, Sc, RA and RSa in the left forewing (Figs. 1A, 1B, 2A, 2B). Dorsal side of thorax and abdominal tergites I–VII are damaged or lost. Left forewing is twisted and bent approximately at half of its length (Figs. 1A, 1C). Tarsomere 5 of left middle tarsus is damaged. Abdominal segments I–VII are complete; sterna VIII and IX are damaged; segment X and cerci not preserved (Figs. 1A, 1B).

Derivation of name. The species epithet is derived from the Latin adjective inconspicuus for its inconspicuous habitus.

Diagnosis. Female imago. Body length 5.62 mm (preserved part); eyes rounded, widely separated medially, not covered by a clypeal shield; distance between eyes 0.57 × of head width; forewings with four small, basally free marginal intercalaries in MP and Cu sectors, and between CuP and A₁; no small free marginal intercalaries in R sector; hind wing rounded, small, 0.10× of forewing length; three cross veins between C–Sc; three cross veins between Sc–RA; no RS and RA–MA forks; subgenital plate approximately 0.15 × as wide as long, strongly convex, outer margin widely rounded.

Description. General colouration of body pale, light brown to brown; legs darker, brown to dark brown; body covered by irregular blackish maculation (Figs. 1A–1C).

Measurements: body length 5.62 mm (preserved part); forewing length up to 5.45 mm. Maximum forewing width 0.37 × maximum length. For complete measurements see Table 1.

Head. General colour brown; intensively brown maculation on eyes and facial keel. Clypeus not expanded, small; clypeal shield not developed. Antennae brown, slightly longer than head; flagellum slightly paler than scape and pedicel. Eyes brown, moderately rounded, widely separated medially; distance between eyes 0.57 × of head width; facets of eyes hexagonal. Ocelli large, without conspicuous colouration (Figs. 3A, 3C).

Thorax. General colouration of ventral cuticle light brown, lateral sides of terga paler; dorsal side of thorax missing. Prosternum relatively wide, light brown. Lateral aspect of thorax hardly visible. Mesosternum complete, light brown coloured; basisternum slightly elongated, nearly trapezoidal; furcasternal protuberances distinctly separated, with parallel inner margins. Metasternum relatively short (Fig. 1C).

Forewings. Hyaline, translucent, relatively narrow; venation well recognizable on right forewing and less visible on left forewing due to wing damage and deformation; three cross veins between Sc and RA, cross venation in other sectors well developed; no naturally pigmented vein sections between C and RA; no pigmented clouds around cross veins; no jagged edge along costal, tornoapical and basitornal margins. Pterostigma with three simple veins (Table 2; Figs. 2A, 2B).

General pattern of forewing venation typical for Burmella: veins C and Sc light brown to brown; RS forked close near base, after 0.15 of its length; iRS long, connected with RSp by eight cross veins, approximated to RSa₁; MA fork strongly asymmetrical, forked after 0.55 of its length; iMA relatively short, MA₁ and MA₂ connected with iMA by 2–three cross veins; MP asymmetrical, forked after 0.40 of its length, MP₁ and MP₂ basally connected by a single cross vein; iMP relatively short, connected with MP₁ and MP₂ by three cross veins from each side; CuA closely approximated to CuP base; CuP smoothly curved toward wing base; cua-cup cross vein and cup-a ₁ cross vein present; cua-cup located distally from cup-a ₁; two secondarily forked veins iCu₁₊₂ and iCu₃₊₄ arising from CuA to basitornal margin of forewing (i.e., in the cubital sector four veins iCu₁–iCu₄ each reaching basitornal margin); iCu₁ connected to CuA by additional cross vein. Vein A₂ arising from A₁; no cross veins in anal sector. Several intercalaries (i.e., iRSa, iRSa₂, iMA, iMP) connected to longitudinal veins by 2–5 cross veins; four small, basally free marginal intercalaries in MP (two) and Cu (one) sectors, and between CuP and A₁ (one); one basally free marginal intercalary vein in cubital sector and one such vein between CuP and MA₁ (Table 2; Figs. 2A, 2B).

Hind wings. Hyaline, translucent, rounded, small, as long as 0.10 × of forewing length (Fig. 2B); venation light brown to brown; venation significantly reduced; no jagged edge along dorsal and ventral margins. Three cross veins between C–Sc and three cross veins between Sc–RA; no RS and RA–MA forks; RSa, RSp and MA short, not forked and not connected to other veins; MP longer than MA, not forked; no forks of cubital veins; CuA relatively short; CuP absent; no free small marginal intercalaries. Costal process rounded apically, protruding above anterior wing margin, situated slightly distally at middle of hind wing length (Table 2; Figs. 2C, 2D).

Legs. Almost preserved, except for damaged apical part of tarsomere V of left middle tarsus. Legs darker than body, brown to dark brown; no visible strong spines or setae on margins of leg segments. Tibiopatellar suture present on middle and hind legs, absent on forelegs (Figs. 4A–4C). First tarsomere of middle and hind legs fused with tibia. Forelegs: length ratio of femur/tibia/tarsus =1/1.58/0.98; length ratio of tarsomeres: 1/0.79/0.74/0.84/0.90 (1 > 2 < 3 < 4 < 5). Middle legs: length ratio of femur/tibia/tarsus =1/0.97/0.39; length ratio of tarsomeres: 1/0.95/0.95/1.11/1.58 (1 > 2 = 3 < 4 < 5) (Fig. 4C).

Hind legs: length ratio of femur/tibia/tarsus =1/0.76/0.35 (proportion of tarsomere V calculated based on right leg only); length ratio of tarsomeres: 1/1.10/1.10/1.14/1.38 (1 > 2 = 3 < 4 < 5) (proportion of tarsomere V calculated based on right leg only). For details of measurements of tarsomeres see Table 1. Pretarsal claws dissimilar (one pad-like, the other one apically hooked) (Figs. 4A–4C).

Abdomen. Apical part of tergite X and most part of abdominal terga I–VII; preserved abdominal segments light brown to brown. Vestigial gill sockets hardly recognizable on segments V and VI; the other segments with gill sockets not distinguishable; no prominent posterolateral projections on preserved abdominal segments. Subgenital plate strongly convex, with presumably straight outer margin, relatively narrow, approximately 0.15 × as wide as long. Subanal plate and cerci not preserved (Table 2; Figs. 4D, 4E).

Affinities. With the establishment of B. inconspicua sp. nov., currently only two species of the genus Burmella are known based on adult females. The marked differences between these species are summarized in Table 2 and can be found especially in the head. While the eyes of B. inconspicua sp. nov. are large, widely separated, and anteriorly not covered by the clypeus, in B. clypeata the eyes are elongated, and the clypeus is expanded anterolaterally, covering the anterior part of eyes (Fig. 3B). As we have noted earlier (Godunko, Martynov & Staniczek, 2021), in contrast to all other known extant and fossil representatives of Vietnamellidae, only in B. clypeata the clypeus has such a characteristic shape (Fig. 3; compare with e.g., Auychinda, Sartori & Boonsoong, 2020: p. 30, fig. 9A; Auychinda et al., 2020: p. 9, figs. 5A–5E).

While the venation pattern of the forewings is similar to other Burmella species, there are some minor differences in the number and location of free marginal intercalaries. In B. inconspicua sp. nov., these are mainly distributed in MP and CuA sectors, whereas in B. clypeata there are more abundant free intercalary veins in R sector and between RSp and MA₁ (Table 2; Godunko, Martynov & Staniczek, 2021: p. 112, figs. 8D–8F).

Furthermore, in Burmella, the female forewings are slightly narrower than those of males, the width/length ratio of the well-preserved right forewing of B. inconspicua sp. nov. is 0.36. In males of B. paucivenosa and B. zhouchangfai the ratios are 0.40 and 0.41, respectively (Table 2; Figs. 5A, 5B). The female forewings of B. clypeata are partially damaged and deformed, but the left forewing, which is better preserved also appears relatively narrow, with a width/length ratio of 0.30 (as preserved) (for comparison see Table 2; Godunko, Martynov & Staniczek, 2021: p. 112, figs. 8C–8E; Chen & Zheng, 2023: p. 4: figs. 3A, 3B).

More differences can be found in the hind wings. The shape of hind wings of B. clypeata is poorly preserved due to fossilisation. Except for B. zhouchangfai, which has elongated hind wings bearing a basal costal process, the hind wings of other fossil Vietnamellidae are markedly rounded, with a costal process located at mid-length, as in the case of B. inconspicua sp. nov. slightly in the distal half. Additionally, in contrast to all extinct species, in B. inconspicua sp. nov. a R–MA furcation is absent, and the venation is generally very simplified.

Finally, the proportions of the subgenital plate also differ, it is slightly wider in B. clypeata compared to the narrower plate in B. inconspicua sp. nov. (Table 2; Figs. 4D–4F).

As with B. clypeata, we cannot exclude that the female of B. inconspicua sp. nov. is conspecific to the male of B. paucivenosa or B. zhouchangfai. The preserved body part of the female holotype of B. inconspicua sp. nov. is 5.62 mm in length, with the tenth segment of the abdomen lacking. Nevertheless, despite its incompleteness, the body length of this species is generally comparable to other representatives of the genus Burmella, which are distinctly smaller than extant Vietnamellidae. On the other hand, clear differences in the shape and venation of hind wings between all discovered fossil specimens rather point to the presence of four separate species. At least, as shown above, there is no doubt that females of B. inconspicua sp. nov. and B. clypeata belong to different species.

In order to maintain nomenclatural stability, we thus follow the approach to describe males and females of the same genus as different species, unless specimens of different sex are syninclusions and fossilised in mating position (see e.g., Staniczek & Godunko, 2016; Godunko, Neumann & Staniczek, 2019; Godunko, Martynov & Staniczek, 2021).

Differentiation between B. paucivenosa and B. zhouchangfai

By establishing synonymy between Burmella and Burmaheptagenia, we can also confirm that B. paucivenosa is well separated from B. zhouchangfai. The male of B. zhouchangfai is well distinguished from B. paucivenosa by: (i) the structure of compound eyes and size of facial keel; (ii) forewing venation with larger number of cross veins in pterostigma; location of basal end of A₂, and presence of cross veins in anal sector; number of marginal intercalaries; (iii) shape, proportions and venation of hind wings; and (iv) structure of gonostyli and penis lobes (Table 2).

Amended diagnosis of male imaginal Burmella paucivenosa (modified from Godunko, Martynov & Staniczek, 2021); see also Table 2; Figs. 6B, 6G, 5B, 5D, 8B, 8C): body length 5.75 mm; forewing length 4.64–4.68; compound eyes contiguous medially; facial keel relatively small; forewings with 4–5 cross veins in pterostigma; 3–4 marginal intercalaries connected with longitudinal veins, and two free small marginal intercalaries; A₂ arises from A₁; no cross veins in anal field; hind wings strongly rounded, small, 0.94 × as wide as long; costal process situated nearly at half-length of hind wing; three cross veins between C and Sc; RSa free, not arising from RSp; no marginal intercalaries; gonostylus segment III nearly equal to segment II in length; penis lobes obliquely truncate and moderately rounded apically, nearly tube-like; tip of penis lobes slightly bent dorsally; strong apical tooth on outer margin of lobes.

Amended diagnosis of male imaginal Burmella zhouchangfai (modified from Chen & Zheng, 2023); see also Table 2; Figs. 6A, 6C–6F, 5A, 5C, 8A): body length 6.00 mm; forewing length 7.00; compound eyes separated medially; facial keel relatively large; forewings with at least eight cross veins in pterostigma; up to two marginal intercalaries connected with longitudinal veins, and four free small marginal intercalaries; A₂ closely approximated to A₁; one cross vein in anal field; hind wings slightly elongated, small, at least 0.65 × as wide as long; costal process situated near to the wing base; no cross veins between C and Sc; RSa arises from RSp; at least two marginal intercalaries; gonostylus segment III shorter than segment II; penis lobes convergent apically, trapezoidal; tip of penis lobes obtuse, distinctly bent dorsally without any apical teeth or spines.

Systematic placement of the genus Burmella

Godunko, Martynov & Staniczek (2021) noted that the general shape of male genitalia in Burmella, as well as the presence of a divided mesothoracic furcasternum in combination with two secondarily forked veins arising from CuA to the basitornal margin of forewing, resemble the combination of characters also found in Heptageniidae. This obviously led to the initial erroneous placement of B. zhouchangfai in the family Heptageniidae (Chen & Zheng, 2023). However, the species here assigned to the genus Burmella also share the most important apomorphic characters of Ephemerelloidea (see also Kluge, 2004), namely in the forewing the presence of a transverse vein cua-cup localised more distally than cup–a ₁; more than one bifurcated vein in the cubital sector of the forewing; the specific arrangement of thoracic sutures as discussed in detail above, and well separated mesothoracic furcasternal protuberances. The assignment of Burmella to Vietnamellidae within Ephemerelloidea is more difficult, as the family is mostly defined by larval apomorphies. However, it presents rounded hind wing with shallow costal process, which is not found in any other Ephemerelloidea, and also not in any other mayfly taxa except of Baetiscidae, which are neither closely related to Ephemerelloidea, Caenoidea, Leptophlebioidea nor Ephemeroidea.

Within Vietnamellidae, the shape of male genitalia is a reliable generic character. Here, Burmella significantly differs from Vietnamella, and this is also the main reason for our attribution of B. zhouchangfai to the former genus. The penis lobes in Burmella are largely separated and thus retain a plesiomorphic condition, while in Vietnamella the penis lobes are medially almost entirely fused. There are no traces of paracercus in the three species of Burmella with preserved tip of abdomen, in contrast to the plesiomorphic character in Vietnamella and Austremerellidae, in which the paracercus is well-developed (see Table 2; Riek, 1963; McCafferty & Wang, 2000; Jacobus & McCafferty, 2006; Suter & Mynott, 2013).

Further differences between Burmella and Vietnamella can be found in the wing venation: The small marginal intercalaries in the forewing of Vietnamella are basally connected to the main longitudinal veins. In Burmella, there are also some basally free small intercalaries present, which we regard as autapomorphy of this genus. In Vietnamella and other Ephemerelloidea, as well as in the majority of most other families of Ephemeroptera, small marginal intercalaries are usually basally attached to the longitudinal veins. Burmella is plesiomorphic compared to Vietnamella and Austremerellidae regarding the forewing pterostigma, which is composed of simple cross veins in Burmella, in contrast to a “cellular” structure of the pterostigma in Vietnamella and Austremerellidae. Further wing characters of unclear phylogenetic information concern the course of CuP, which is smoothly curved towards wing base in Burmella, whereas in extant Vietnamellidae and Austremerellidae, CuP is sharply curved inwards. Also, the hind wing venation of Burmella is distinctly simplified to a few cross and longitudinal veins forming a single triad only (see Table 2), thus resembling the condition in Austremerella picta Riek, 1963, in contrast to Vietnamella with abundant venation (Riek, 1963). As the female imago of Burmella inconspicua sp. nov. is incompletely preserved without tip of abdomen, we cannot conclude on the female genitalia, but otherwise it correlates well with the female of B. clypeata.

Phylogenetic position of Vietnamellidae within Ephemerelloidea

McCafferty & Wang (2000) placed the “Austremerella group” within Pantricorythinis. Kluge (2004), excluding the Ephemerellinae and Timpanoginae lineages. Later, these two last lineages were strongly supported as monophyletic in a combined molecular and morphological analysis by Ogden et al. (2009a). Even earlier, McCafferty & Wang (1997) placed Oriental Vietnamella and Australian Austremerella in the subfamily Austremerellinae (McCafferty & Wang, 1997) within Teloganodidae McCafferty & Wang, 1997.

In a cladistic analysis of Ephemerelloidea proposed by Jacobus & McCafferty (2006), Vietnamellidae and Austremerellidae are clustered as basal clades to the remaining Ephemerelloidea. Most basally, the fossil Caenozoic monotypic genus Philolimnias sinica Hong, 1979 (see Hong, 1979) was included in this cladistic analysis as sister group to all the other Ephemerelloidea (including Vietnamellidae and Austremerellidae). Hong (1979) and Jacobus & McCafferty (2006) attributed Philolimnias to Ephemerellidae, while Kluge (2004) and Hubbard (1987) regarded it as Euephemeroptera incertae sedis. Lastly, Staniczek, Godunko & Kluge (2018) concluded that the placement of Philolimnias in Ephemerelloidea is doubtful and its systematic position needs thorough re-evaluation.

Recently, Auychinda, Sartori & Boonsoong (2020) and Auychinda et al. (2020) in a molecular study of extant species confirmed the monophyly of Vietnamellidae based on well supported genetic distances between Vietnamellidae, Teloganodidae, and the genera Dudgeodes Sartori, 2008, Teloganella Ulmer, 1939, and Teloganopsis Ulmer, 1939. Unfortunately, representatives of Ephemerellinae were not included in these analyses.

Investigating six complete mitochondrial genomes of different species of Ephemerellinae, Xu et al. (2020) concluded that Ephemerellidae was the sister group to Vietnamellidae. The same result was confirmed by Tong et al. (2022) and Guo et al. (2024). However, these both aforementioned studies should be treated with great caution regarding phylogenetic analyses, since some of the results show significant differences with most part of the phylogenetic reconstructions earlier proposed for Ephemeroptera (see above).

In summary, most molecular or combined phylogenies consider Ephemerellidae as sister group to Pantricorythini sensu Kluge (2004). Also, Vietnamellidae and Austremerellidae are generally considered the most basal clades of Pantricorythini (see McCafferty, 1991; Kluge, 2004; Ogden & Whiting, 2005; Ogden et al., 2009a; Ogden et al., 2009; Ogden et al., 2019). According to these contributions, Vietnamellidae is the sister group of Austremerellidae, which is represented by the single species Austremerella picta (Riek, 1963), known from adults and larvae found in the Lamington National Park in Queensland, Australia (Riek, 1963; Suter & Mynott, 2013). Considering their pattern of distribution, McCafferty & Wang (2000) suggested Vietnamellidae and Austremerellidae as ancient Gondwanan taxa. Moreover, Austremerella was considered as mostly “generalized genus” in comparison to Vietnamella, according to McCafferty & Wang (2000) probably representing “the extant taxon closest to the hypothetical precursor of the entire superfamily Ephemerelloidea”. Unfortunately, Vietnamellidae and Austremerellidae were not included in the otherwise most comprehensive phylogenetic study on mayflies by Ogden et al. (2009a), so their position within mayflies and Ephemerelloidea awaits further clarification. This in turn, together with the accumulation of data on fossil representatives of the superfamily, may lead to refinement of biogeographic scenarios.

Palaeoecological considerations

The larvae of extant Vietnamellidae inhabit small and middle-sized rivers and streams of moderate to fast flowing current, where they can be found on cobbles and pebbles, sometimes also on sandy substrate. Here, they co-occur with nymphs of other mayflies, particularly other Ephemerelloidea (Sinha et al., 2018; Auychinda, Sartori & Boonsoong, 2020; Auychinda et al., 2020). The biology of adults is still poorly known.

The paleohabitats of the Burmese mayfly fauna most probably represented humid tropical forests (Poinar Jr, Lambert & Wu, 2007; Xing et al., 2019; Bolotov et al., 2021; Yu, Neubauer & Jochum, 2021) with dense nets of waterflows. The rich mayfly fauna documented for Burmese amber mostly represents extant families with numerous rheophilous taxa, namely Heptageniidae (Yang et al., 2023), Baetidae (Sinitshenkova, 2000; Poinar Jr, 2011), Prosopistomatidae (Lin et al., 2018a), and Leptophlebiidae (Chen & Zheng, 2022; Chen & Zheng, 2024). Two other mayfly taxa reported from this paleoecosystem, namely Australiphemeridae (McCafferty, 1991) and Hexagenitidae Lameere, 1917 (Lin et al., 2018b; Zheng & Chen, 2023), most probably also had rheophilous larvae, possibly distributed in lower (deltaic) sections of the waterflows with hyporhithral to epipotamal conditions. Similarly, presumably lotic or deltaic ecological niches are assumed for taxa of these families recorded from Jurassic Limestone of Solnhofen, Germany (see Bechly, 2015) and from the Lower Cretaceous Crato Formation in Brazil (McCafferty, 1990; Storari et al., 2021; Storari et al., 2022).

In addition to mayflies, the Kachin amber is rich in other groups of freshwater insects (Rasnitsyn et al., 2016; Staniczek, 2019; Jarzembowski, Coram & Song, 2023). Based on these data and following Bolotov et al. (2021), we assume that the Kachin amber-producing forest extended from the deltaic (estuarine) areas of tropical rivers and streams to their upper sections, and also covered different types of freshwater and estuarine habitats (lakes, estuaries, coastal rivers and their river deltas). The taxonomic composition of the mayflies from Burmese amber rather indicates that their fossilisation took place along the middle and lower reaches of the waterflows, which does not exclude the presence of strong rheophilic taxa typical for the upper epirhitral portions of the streams and rivers. In that case, swarming of adults of rheophilous taxa of Ephemeroptera in lower river sections was compensated by upstream flight of fertilised females. Such upstream flight compensation (compensatory flight of lotic aquatic insects) is well described and discussed for several mayfly groups (see Hubbard, 1991; Jacobus, Macadam & Sartori, 2019; Domínguez et al., 2023). The rheophilic character of the Burmese fossil mayfly fauna is also well supported by the co-occurrence and dominance of the stonefly family Perlidae (Sroka, Staniczek & Kondratieff, 2018; Chen, 2018; Jouault et al., 2022), whose recent larvae also prefer rheophilic conditions.

Palaeobiogeographical considerations

The current distribution of Vietnamellidae is restricted to Eastern India, Thailand, Central and Southern China. In Thailand, most species have been recorded close to the border with Myanmar (Auychinda, Sartori & Boonsoong, 2020; Auychinda et al., 2020). All fossil Vietnamellidae have been found in mid-Cretaceous Burmese amber only, among them are also species that are still undescribed (Figs. 9A–9C). Thus, both fossil and recent taxa of this family are known exclusively from Southeast Asia.

The extant worldwide distribution of Ephemerelloidea is in sharp contrast to its almost complete absence in the fossil record, and only a few extinct taxa including Burmella can be assigned to this group without doubt. For instance, Staniczek, Godunko & Kluge (2018) reinvestigated the holotype of Ephemerella viscata Demoulin, 1968 from Eocene Baltic amber, and transferred it to the genus Eurylophella Tiensuu, 1935 (Ephemerellidae: Timpanoginae Allen, 1984). In the same publication, the first confirmed record of the genus Ephemerella Walsh, 1862 s.l. (Ephemerellinae Lestage, 1917) was reported based on adults of both sexes from Baltic amber. In contrast to the sparse fossil record of Ephemerellidae, the recent fauna of this family includes approximately 150 species in 22 genera in the Holarctic and Oriental regions, with clear dominance in the Palaearctic and Nearctic (Barber-James et al., 2008; Sartori & Brittain, 2015).

The fossil record of Ephemerellidae and the restricted distribution of extant species in the Northern Hemisphere well reflects their Laurasian origin. According to McCafferty & Wang (2000), the first major branching within the Ephemerelloidea resulted in a Laurasian clade (i.e., Ephemerellidae), and all the other lineages, which constitute a Gondwanan clade (i.e., Pantricorythini sensu Kluge, 2004), with only recent invasions into the Holarctic, which are for instance well documented for the Nearctic Leptohyphidae (McCafferty, 1998; Barber-James et al., 2008; Sartori & Brittain, 2015).

However, in contrast to the scenario of McCafferty & Wang (2000), suggesting the Palaeozoic origin of Ephemerelloidea, a different theory, initially put forward by Pescador et al. (2009), seems more plausible: The separation of Ephemerellidae and Pantricorythini could have resulted from a vicariant event in the Jurassic at the breakup of Pangea, where the ancestors of these two lineages most likely existed even before the main continental breakout, dispersed, and initially diversified on several modern continents no later than the Early Jurassic (200–175 My). This may indicate an Early Mesozoic age for Ephemerelloidea but seems in no way older than the P–Tr extinction event. A similar scenario was also proposed for another mayfly family with Pangean origin, namely Baetiscidae (Godunko & Sroka, 2024).

A high probability of a Pangaean origin can also be assumed for the extinct mayfly families Hexagenitidae and Australiphemeridae, as well as for Baetidae, Leptophlebiidae and possibly Heptageniidae, which are also recorded from Burmese amber. On the other hand, among mayflies and other freshwater invertebrates in Burmese amber, there are also groups that are nowadays typical of the Southern Hemisphere (Sroka, Staniczek & Kondratieff, 2018; Staniczek, 2019; Yu, Neubauer & Jochum, 2021; Bolotov et al., 2022). Without any doubt, the ephemerelloid families Vietnamellidae and Austremerellidae also belong to these groups with possible Gondwanan origin.

The primordial supercontinent of Gondwana in the Southern Hemisphere included landmasses of South America, Africa, India, Madagascar, Antarctica, Australia, New Zealand, and New Guinea. Up to the Middle Jurassic (about 170–150 My), the active exchange of both terrestrial and aquatic biota between these landmasses was undoubtedly possible. With regards to eastern Gondwana, a biotic transfer within the Indian Plate, Burma Terrane, Greater India, Sri Lanka and Madagascar on the one hand and Australia (plus Antarctica) on the other hand, with high probability was still possible in the Early Cretaceous, but evidently not later than 135 My ago (Wan, 2011; Wan, 2020; Wan & Zhu, 2011; Bolotov et al., 2022).

The early separation of Vietnamellidae and Austremerellidae from the remaining Pantricorythini correlates well with the paleogeography of the East Gondwana. Considering that the breakup of East Gondwana started around 133–130 My, when the assembly of circum-Indian blocks and continents moved northwest away from Australia–Antarctica (Acharyya, 1998; Acharyya, 2000; Williams et al., 2013; Yoshida & Hamano, 2015; Wan, 2020), the ancestors of these families should have arisen even earlier, possibly during the Middle-Late Jurassic interval, with Austremerellidae being perhaps the easternmost group, while the ancient Vietnamellidae were very early separated and isolated by the drift of the Burma Terrane.

In this way, the ancestral Austremerellidae became isolated on the Australian continent. Fossil finds of mayflies in Australia are scarce. Only a few taxa have been described from the Early Cretaceous Koonwarra Fossil Bed in Victoria and the Pliocene Vegetable Creek tin-mining field in New South Wales (Etheridge & Olliff, 1890; Riek, 1954; Jell & Duncan, 1986; Jell, 2004; Poropat et al., 2018). Furthermore, there are still undescribed fossil mayflies from the recently discovered Middle Miocene McGraths Flat Konservat–Lagerstätte in New South Wales (McCurry et al., 2022; Baranov et al., 2024; Godunko & Sroka, 2024). However, within all the listed fossils, taxa of Ephemerelloidea are neither recorded nor described. On the other hand, Cretaceous mayfly specimens from Koonwarra were mostly described as Siphlonuridae (e.g., the monotypic genus Australurus plexus Jell & Duncan, 1986), but may have rather affinities within Leptophleboidea or even Ephemerelloidea (see also Godunko & Sroka, 2024). The recent Australian fauna of Ephemerelloidea is restricted to Austremerellidae as earlier noted (see Chessman & Boulton, 1999; Dean & Suter, 1996; Suter & Mynott, 2013; Sartori & Brittain, 2015). The only other finding of Ephemerelloidea within the Australasian Realm refers to Dudgeodes celebensis Sartori, 2008 from Sulawesi (Sartori, Peters & Hubbard, 2008).

Little is known about the drift of smaller elements of the Indian plate after the breakup of Gondwana, and there is no consensus on the spatial and temporal dynamics of this process. Among these smaller plate elements detached from Gondwana and migrating northwards, the Burma Terrane (also known as the Indo-Burma-Andaman block, Myanmar Microplate, Burmese Platelet, West Myanmar Terrane or West Burmese Plate) is particularly interesting in regard of the mid-Cretaceous Burmese fauna and its impact on the evolution of recent Oriental biota including freshwater insects.

It is still disputable at what time the Burma Terrane separated from East Gondwana (Staniczek, 2019). Poinar Jr (2019) discussed two hypotheses of its origin. It is either considered to have split from Australia and successively shifted to its current location in southeast Asia (Metcalfe, 1996; Metcalfe, 2017; Hall, 2012; Seton et al., 2012), or it became attached to the east of the Great Indian Plate after the break-up of Gondwana (see below).

According to the latter hypothesis, the initial separation of the Burma Terrane, together with the Indian Subcontinent and surrounding blocks from Gondwana, took place less than 135 My ago, with a subsequent final separation of the Burma Terrane from the Indian Plate, probably by a partial submergence of Greater India at around 75 My (Bolotov et al., 2022). Other authors have suggested that the West Burma block separated from Northern Gondwana already at the end of the Jurassic period about 160 My ago together with the Indian Subcontinent, shifting northwards through the Tethys Sea and reaching Asia around 100–90 My ago, or even later in the Santonian or Campanian (86.3–83.6 My ago) (Heine, Müller & Gaina, 2004; Khin, Zaw & Aung, 2017; Bolotov et al., 2022). On the other hand, Westerweel et al. (2020) assumed that, approximately 95 My ago, the Burma Terrane was a part of the Trans-Tethyan island arc located on a near-equatorial southern latitude, “suggesting island endemism for the Burmese amber biota”. A significant clockwise rotation of the Burma Terrane would have occurred between 80–50 My, before the collision of the Late Eocene Burma Terrane–Sibumasu Terrane in Asia around 45–40 My ago (Licht et al., 2020; Bandopadhyay et al., 2022; Bolotov et al., 2022; Min et al., 2022). Finally, there is also a hypothesis that assumes a separation of the Western Burma block as early as in the Devonian (418–361 My ago) (Metcalfe, 2017), merging with the Asian plate already in Early Cretaceous (app. 120 My) (Hall, Cottam & Wilson, 2011; Wan, 2020).

The scenario proposed by Bolotov et al. (2022), which is linked to the time-calibrated phylogeny and to statistical biogeographic models for freshwater mussels of the subfamily Parreysiinae, may also apply to other groups of freshwater insects, including mayflies. Not only does the early separation of the Vietnamellidae and Austremerellidae from the rest of Pantricorythini correspond well with this scenario, but also the diverse fossil fauna of Vietnamellidae found in the mid-Cretaceous Burmese amber, as well as the presence of extant species in the Oriental region. From there, the family would have expanded northwards to China and India in post-Eocene times. Moreover, this scenario corresponds well with the assumption of a Gondwanan origin of other groups of Pantricorythini, which now also has palaeontological evidence based on the report of the first fossil finding of the family Teloganodidae from India (Agnihotri et al., 2020) (Fig. 9D). Teloganella gurhaensis Agnihotri et al., 2020 was described as a compressed single larva from the Palana Formation (Gurha lignite mine, Rajasthan, India). Judging by its morphological characters, the placement within Teloganodidae is indeed quite possible. The geological age of the Palana Formation is indicated as Late Paleocene-Early Eocene by Agnihotri et al. (2020), while Farooq et al. (2019) and Mathews et al. (2020) rather indicated an Early Palaeocene age (app. 66–56 My). At that time, the Indian Plate and Greater India were still drifting towards Asia, which allows us to consider this fossil taxon as a part of the more ancient East Gondwanan fauna.

The model proposed by Metcalfe (1996), according to which the West Burma Block was separated from Australia in the Late Jurassic and docked with SE Asia in the Early Cretaceous (see also Hall, 2012; Poinar Jr, 2019; Jouault et al., 2022), correlates well with the initial evolution of Pantricorythini, implying an early separation and isolation of the ancestors of the basal lineages represented by Vietnamellidae and Austremerellidae.

Also, the recent and fossil mayfly fauna of Australia has noticeably less similarities with the palaeofauna preserved in Burmese amber and is also not closely related to the extant and fossil fauna of Palearctic and Oriental Asia. Therefore, models of the tectonic evolution of the East Gondwana postulating the initial separation of Australia and Antarctica (isolation of the family Austremerellidae) and their subsequent detachment from the common landmass comprising the Indian Subcontinent and Burma Terrane (the “biotic ferry” for Vietnamellidae and part of Teloganodidae), correlate well with the evolution and current biogeography of these basal groups of Pantricorythini.

There is however no doubt that the colonisation of the Burma Terrane by groups of Laurasian origin was only possible at a time, when this block was approaching the Asian part of Laurasia and thus, the influence of Northern faunal elements increased during the Late Cretaceous. New groups of aquatic insects then probably migrated southwards to the Burma Terrane by passive transfer of winged stages with the winds or even by drift. However, due to the calibration of Burmese amber to 99 My based on UePb zircon dating, there is no doubt at all that the Burma Terrane already at that time must have already been reachable for Laurasian elements. As the time frame for the tectonic shift suggested by Bolotov et al. (2022) implies a location of the Burma Terrane far south of the equator 100 My ago, either the calculated time of the continental shift would be wrong or those groups to be considered of Laurasian origin already at that time had migrated south towards Gondwana or in fact were of Pangean distribution.

This dispersal scenario may not only apply for recent families with Laurasian origins (e.g., Prosopistomatidae, Heptageniidae, Baetidae: Palaeocloeoninae), but also for fossil Pangaean groups (Australiphemeridae and Hexagenitidae), nevertheless it does not explain the spatial and temporal collisions described here for basal Pantricorythini. The alternative would be a wider distribution of Pantricorythini in the northern hemisphere before the Cretaceous, which is not yet supported by the fossil evidence of the Ephemerelloidea analysed in detail above. As for other insect groups (including aquatic and terrestrial), early faunal elements are supposed to have Gondwanan origin with later influence of Laurasian taxa as suggested by Poinar Jr (2019), Staniczek (2019), Jouault et al. (2022) and Wichard (2023).

The discovery of new fossil sites from the Oriental region and the Indian subcontinent may fill the gaps in our present knowledge and may help in a more accurate reconstruction of spatial and temporal dispersal of biota during the Cretaceous. For example, Zheng et al. (2018) reported on a unique amber biota from the Upper Cretaceous of Tilin in central Myanmar, dated to the uppermost Campanian (app. 72.1 My). Likewise, Xing & Qiu (2020) pointed to a little known but diverse arthropod fauna from the mid-Cretaceous Hkamti amber in northern Myanmar with an approximate age of 110 My. In both fossil sites, undescribed adults of aquatic insect groups (e.g., Chironomidae and Ceratopogonidae, and some Trichoptera) were listed, so the discovery of mayflies in these sites may seem well possible.