Ferns and fern allies in the Rhaetian ﬂ ora of Wüstenwelsberg, Bavaria, Germany

Twelve of and are of and and excellently two lycophyte and ) and one sphenophyte Equisetites ) identi ed. eight families; Osmundaceae with one Todites and two Cladophlebis species; Matoniaceae with two Phlebopteris species, and Dipteridaceae with Clathropteris , Dictyophyllum and Thaumatopteris with one species each. Curled fern fronds have been attributed to the fossil-genus Spiropteris . BesidesseveralkeyRhaetiantaxa,twokeytaxaforHettangian ﬂ oras, Phlebopterisangustiloba and Thaumatopteris brauniana , are present in Wüstenwelsberg, albeit not in large numbers. The comparison of the ﬂ ora from Wüstenwelsberg with adjacent Rhaetian ﬂ oras revealed distinct local differences in the respective ﬂ oras, which are discussed in the light of paleogeography involving dispersal patterns or mechanisms and adaptations of the plants.

The ecology and environment of the entire flora are discussed as well as its habitat, with a focus on the cryptogams; in addition, we compare the composition of the Rhaetian flora of Franconia with the Rhaetian and Hettangian floras of East Greenland (Jameson Land), Sweden (Scania), Poland, Ukraine (Donets Basin) and Iran, and discuss potential relationships, biogeography and dispersal patterns.

The Wüstenwelsberg quarry
The studied section is located in a sandstone quarry near the village of Wüstenwelsberg, approximately 20 km SW of Coburg, Germany (Fig. 1). The sediments were deposited in the Germanic Basin and are characterized by an alternation of clay and sandstone layers (for details see Bonis et al., 2010;Pott et al., 2016;Van Konijnenburg-van Cittert et al., 2018b). The plant fossils come from clay layers, one of which is the so-called "Hauptton" that can be up to 10 m thick. Most of the specimens originate from this horizon (level 3 in Bonis et al., 2010). Almost all layers in the section are Rhaetian in age, only the uppermost one (without any macrofossil remains but with palynomorphs) might be Hettangian in age (Bonis et al., 2010).

Description of the fossil material
The fossil leaf material used in this study originates from fieldtrips by some of the authors (SS, GD, JHAvKvC). The fossils are stored in the collections of the Laboratory of Palaeobotany and Palynology, University of Utrecht (The Netherlands; UU numbers) and in the private collections of Stefan Schmeißner (Kulmbach, Germany; numbers preceded by Q) and Günter Dütsch (Untersteinach, Germany; numbers containing the acronym wü). The plant fossil remains are mainly compression fossils of relatively small size, giving only information on the macromorphology. Some fertile fern specimens yielded in situ spores and so contributed to our knowledge of this Rhaetian flora.

Methods
In situ spores were prepared by picking sporangia from fertile specimens. These were macerated according to the standard procedure using Schulze's reagent (30% HNO 3 with a few crystals of KClO 3 ) and subsequently treated with 5-10% ammonia (NH 4 OH) or potassium hydroxide (KOH). Macerated sporangia were rinsed with water and dehydrated in glycerine. Then they were separated with needles so that separate spores could be seen, or spore clusters in the case of immature sporangia. These were embedded in glycerine jelly and sealed with transparent nail polish or paraplast. The slides are stored in the collection of the Laboratory of Palaeobotany and Palynology, Utrecht University, and in the private collections of SS and GD. Slides and specimens of the latter two collections will be donated to a publicly available collection after the research on the Wüstenwelsberg flora has been completed.
The macrofossil specimens were photographed with a Nikon D80/ Nikkor AF-S Mikro 60-mm 1:2.8G ED system digital camera and partly with a Panasonic DMC-FZ1000 with a Leica DC Vario-Elmarit 1:2.8-4.0/9.1-146 lens. Oblique lightning and polarizing filters in front of the camera lenses and the lights were used to enhance contrast and fine details. Spores were analyzed with an Olympus BH2 light microscope.
Lepacyclotes sp. Plate I, 1-2 Description: One specimen (20wü04) in the Wüstenwelsberg flora yields lycophyte remains, albeit not too well preserved (Plate I, 1). The the cryptogam flora from Wüstenwelsberg, we figure two good specimens here that so far have not been figured (Plate I, 3-4).

Equisetites laevis
Remarks: The specimens from Wüstenwelsberg agree perfectly with this typically Rhaetian species as described and figured by, e.g., Halle (1908), Harris (1926) and Schweitzer et al. (1997). The type material comes from the Rhaetian of Bjuv (Sweden) (Halle, 1908), but the species is also found at other Rhaetian localities in the area such as Billesholm (Lundblad, 1950) and Rögla (Pott and McLoughlin, 2011). The latter is probably a diaphragm of E. laevis, even if left unassigned as Equisetites sp. by the authors. A diaphragm that might also be from E. laevis is recorded from the coeval and close-by locality of Heilgersdorf (Kelber and Van Konijnenburg-van Cittert, 1997). Equisetites laevis has also been recorded from the classical Rhaetian localities of Jameson Land (mainly as nodal diaphragms; Harris, 1926 and Iran (Schweitzer et al., 1997). The latter authors recorded larger stems, some of which with attached leaves.
A species very similar to Equisetites laevis is E. muensteri, which is the generic type. Equisetites muensteri is commonly found in Hettangian deposits, and occasionally has been recorded from the same Rhaetian localities as E. laevis (Jameson Land; , but we question this identification because the remains he figured are far too fragmentary to make a specific assignment; this material should be referred to as Equisetites sp. Equisetites muensteri is distinguished from E. laevis by narrower stems, by fewer and narrower leaves per node, the latter even elongate and with acute apices (Harris, 1926Schweitzer et al., 1997). Barth et al. (2014) wrote in their review of the Norian-Rhaetian flora from Seinstedt (Germany) that they never found the specimen of E. muensteri in the collections mentioned by Jüngst (1928). Pacyna (2014) mentioned Equisetum chalubinskii Raciborski, 1890, from the Rhaetian of Poland (Tatra mountains), commenting that the species is very similar to Equisetites muensteri and should properly be referred to the genus Equisetites. He also stated that E. muensteri shoots have been recorded from Norian Polish sediments, however, without reference.
Remarks: In the Rhaetian-Hettangian of Franconia, two quite similar osmundaceous ferns occurred at the same time, viz. T. roessertii and T. goeppertianus (Münster in Göppert) Krasser, 1922. Both names have been used for the same type of fern foliage and consequently, some authors consider them conspecific, while others claim that the difference lies in the venation: T. goeppertianus is regarded to have more of a neuropterid-type venation (similar to that of the Middle Jurassic type Plate III. Macroremains of Osmundaceae from the Rhaetian of Wüstenwelsberg with specimen numbers. 1. Todites roessertii, Q941/13. 2. Todites roessertii, 68wü08. 3. Todites roessertii, fertile specimen, Q501/07. 4. Todites roessertii, fertile specimen showing venation, Q502/07. 5. Cladophlebis scoresbyensis, with clear venation, Q670/08. 6. Cladophlebis scoresbyensis, 92wü08. 7. Cladophlebis scoresbyensis, 29wü10. 8. Cladophlebis scoresbyensis, small marginal dentations with two veinlets ending there, Q970/14. Scale bars 1-3, 5-6: 10 mm; 4, 7: 5 mm. species T. williamsonii), while T. roessertii has more of a pecopterid-type venation. However, both species are often reported from the same layers in a locality or area. Schenk (1865Schenk ( -1867 described and figured Acrostichites goeppertianus (Münster) Schenk from the Hettangian of Theta in Franconia, and Asplenites roesserti (Göppert) Schenk from three localities around Nuremberg, including Theta. From the given description and figures, a range of leaf morphologies is obvious: compare the transitional forms of Schenk's pl. 2, figs. 5, 5a (Acrostichites goeppertianus) via pl. 7, fig. 2 (Asplenites roesserti) to pl. 10, figs. 1-4 (also A. roesserti but with a much larger variability in pinnule shape, size and venation). Gothan (1914) considered the two species as conspecific and assigned all specimens from the Liassic of Franconia to Todites roessertii as that name has priority over T. goeppertianus, which was later agreed on by Weber (1968). Harris (1926) described material from the Rhaetian of Jameson Land (Greenland) as Cladophlebis roessertii (Schenk non Presl) Saporta and Todites cf. williamsonii. Later,  transferred the latter to Todites goeppertianus and placed the specimens earlier assigned to C. roessertii in the new species Cladophlebis scariosa , that is distinguished from C. roessertii in having narrower, parallel-sided pinnules with a simpler venation and more delicate lamina. This differentiation was later confirmed (Harris, 1937), when evaluating other reported specimens. Harris (1937) recommended that many specimens identified as T. roessertii should be assigned to T. goeppertianus, which also is the case for the original specimens of Alethopteris roessertii. Cladophlebis scariosa is difficult to be clearly distinguished from what others described as C. roessertii and C. nebbensis (Harris, 1937, p. 17).
Remarks: The material from Wüstenwelsberg agrees in all aspects with the specimens from Jameson Land described by Harris (1926), including the variability in pinnule shape and size, and the character that, when present, small dentations occur only in the more apical part of the pinnules with two veins ending in a single tooth. Later, Harris (1931) described additional material including fertile frond portions, and transferred, therefore, the species to Todites, the genus used when fertile specimens are known in which the complete lower side of the pinnules is covered with sporangia. The material from Wüstenwelsberg yields only sterile fragments, thus necessitating the specimens to be assigned to Cladophlebis. Records of Cladophlebis scoresbyensis from outside Greenland or Jameson Land are rare. Lundblad (1950) reported some sterile leaf fragments from the Rhaetian of Sweden. In a catalogue of material from Alborz (Iran) Cladophlebis scoresbyensis appears (Sadovnikov, 1983, p. 10, pl. 11, figs. 3, 4), but the specimens are not described but only figured; due to the poor quality of the figures, we cannot say whether the material belongs to C. scoresbyensis. Even Schweitzer et al. (1997) refrained from a definite identification of those specimens.
The by far most similar species is Cladophlebis nebbensis (Brongniart) Nathorst, 1876. Harris (1926) stated the differences as (1) the greater size of the pinnules in C. scoresbyensis and (2) the termination of two veins in a single tooth in C. scoresbyensis. Schweitzer et al. (1997) added the thinner rachis in C. nebbensis (c. 1 mm) and the obtuse pinnule apex. Bodor and Barbacka (2008) compared Cladophlebis/Todites specimens from the Hettangian of Hungary to T. scoresbyensis, which differ in the arising angle of the secondary veins and the morphology of the margin. Another similar species is Cladophlebis denticulata (Brongniart) Nathorst, 1876, but this species differs by the distinctly dentate margin of the pinnules (see e.g., Harris, 1961).
Cladophlebis sp. Plate IV, 1-3, Plate VI, 10-14 Description: Three specimens yield small sterile fragments of a Cladophlebis/Todites-type frond morphology (Q783/09, 03wü05, 27wü10), but with a clear neuropterid venation that is different from that of the specimens assigned to T. roessertii and C. scoresbyensis as described above. The apical portion of Q783/09 (Plate IV, 1) is only 14 mm long and up to 9 mm wide (3.9 mm at its apical part). It consists of six pairs of oppositely arranged pinnules, attached perpendicularly and almost with their whole base to a 0.8-mm-wide rachis. Pinnules vary in length between 4.8 (most basal one) and 2.5 mm (most apical one) but have an almost uniform width of 2.0-2.5 mm; apices are obtuse. The neuropterid venation is best visible in the lowermost pinnules (Plate IV, 1) and consists of a central vein arising at c. 45°from the basiscopic pinnule angle, and half way bending towards a horizontal plane proceeding to the apex. Secondary veins arise in a fan-shaped manner at regular distances from the central vein; they occasionally bifurcate once. Specimen 03wü05 (Plate IV, 2) yields a 36-mm-long and 12-mm-wide fragment, consisting of a small number of oppositely arranged pinnules that are equal in size throughout the fragment, 8 mm long and 6 mm wide. Venation is clear as described above, with a central but less obvious vein arising from the basiscopic angle giving off fanshaped secondary veins that commonly bifurcate once. Specimen 27wü10 (Plate IV, 3) provides the least preserved fragment with seven pairs of oppositely inserted pinnules, similar in size, shape and venation pattern to the other two specimens. Another specimen (part and counterpart Q529/07, Q530/07) is too poorly preserved to be assigned to this taxon without any doubt.
The fragment on Q783/09 (Plate IV, 1) unexpectedly yielded a thin cuticle providing some rare glimpses into the epidermal structure of this species: The upper cuticle (Plate VI, 11) is thicker than the lower one (Plate VI, 10, 14). The epidermal cells both on the upper and the lower cuticle are more or less isodiametric (Plate VI, 11, 12), commonly with straight anticlinal cell walls that occasionally might express slight sinuosity (Plate VI, 14); veins are indicated by more elongate epidermal cells. Trichomes and trichome bases are present, especially frequent on the adaxial cuticle (Plate VI, 12). Stomata occur sparsely on the abaxial cuticle only (Plate VI, 11; they consist of two slightly thickened guard Phlebopteris angustiloba, 05wü04. 6. Phlebopteris angustiloba, Q181/02. Scale bars 2, 4, 6: 10 mm; 1, 3, 5: 5 mm. cells surrounded by a number of unspecialized subsidiary cells (Plate VI, 13).
Remarks: These specimens are small pinna fragments with tiny pinnules that exhibit typical characters such as an opposite and perpendicular insertion of the pinnules on the rachis and the typical neuropterid, fan-shaped venation. Similar specimens have occasionally been described as Todites cf. williamsonii or Todites/Cladophlebis goeppertianus. However, due to the limited amount of material we keep it unassigned in Cladophlebis sp.
Antevs (1919, pl. 1, figs. 20-22) reported specimens from the Hettangian of Sweden as Todites williamsonii with bipinnate fronds with oppositely and perpendicularly inserted pinnules with a neuropterid venation consisting of a weak midrib and once-bifurcate secondary veins arising in a fan-shaped manner. The specimen of his pl. 1, fig. 22 is very similar to our Cladophlebis sp. specimens, both in size and shape. Harris (1926, p. 55, text- fig. 2F) described a similar specimen from the Rhaetian of Jameson Land as Todites cf. williamsonii, which the author later included in Todites goeppertianus. We consider this species conspecific with T. roessertii (Harris, , 1937. However, this particular specimen is more similar to our Cladophlebis sp. specimens in shape and venation than to the other T. roessertii-material from Jameson Land (see below). Pott and McLoughlin (2011, p. 1029, text- fig. 3F) reported a specimen from the Rhaetian of Rögla as Todites sp. cf. T. williamsonii. This specimen is also very similar in shape, size and venation to our Cladophlebis sp.

Comparison of the osmundaceous species from Wüstenwelsberg
The three osmundaceous species from Wüstenwelsberg mainly differ in pinnule size and shape, and type of venation. Cladophlebis scoresbyensis has the largest pinnules (15-23 mm long have been found) and a definite pecopterid venation, with two veinlets ending in one small marginal dentation. Todites roessertii commonly has smaller pinnules (6-12 mm long), no marginal dentations and a venation that is intermediate between pecopterid and neuropterid. Finally, Cladophlebis sp. has even smaller pinnules (2.5-8 mm), with a completely fan-shaped, neuropterid venation, more of the type found in Todites/Cladophlebis goeppertianus.
However, some authors consider Todites/Cladophlebis goeppertianus and Todites/Cladophlebis roessertii as conspecific (e.g., Gothan, 1914; as there are intermediates between the two species. Poorly preserved specimens of C. scoresbyensis are also quite similar to those of Todites/Cladophlebis roessertii, especially those in which no marginal dentations have been preserved. Hence, it is possible that all the fragmentary specimens from Wüstenwelsberg in fact belong to one species only, which should be named Todites roessertii as that name has priority. However, as we do not have many intermediates between the three taxa described here, we prefer to separate them at the moment.  fig. 65. Description: A few specimens assignable to Phlebopteris angustiloba have been found in Wüstenwelsberg. All of them appear to be fertile portions, however, without any preserved sori or sporangia apart from an indication of the receptaculum (Plate IV, 5). Most fragments of primary segments are 3-4.5 cm long (66wü02, 18wü04, 05wü04, Q181/ 02; Plate IV, 4-6), with a distinct but thin rachis (b500 μm wide) and up to 18 secondary segments ("pinnules") preserved. Pinnules are attached at angles of 70°-90°to the rachis and are densely spaced (Plate IV, 4-6). None of the pinnules is complete and the largest reaches up to 4.5 mm in length by 2 mm in width (Plate IV, 4). Pinnules are tapering towards the [designated] apex, but apices are not preserved in any pinnule. The distinct pinnule central vein is thin and gives rise to several lateral veins at intervals of approximately 1 mm, thus producing a mesh consisting of hexagonal to roundish depressions (Plate IV, 5). These depressions represent the areas where sori were attached but details of sori or sporangia are not preserved; solely, an indication of the annuli is ascertainable but too poorly preserved to be illustrated.
Remarks: The material from Wüstenwelsberg assignable to Phlebopteris angustiloba is only fragmentary, but there is no doubt that specimens represent this species because of the typical mattress-like appearance of the pinnules, a feature that has not been recorded from any other species in Phlebopteris. All specimens from Wüstenwelsberg represent fertile frond portions, which is the case for many of the records worldwide (see e.g., Tralau, 1965), but sterile specimens occur occasionally in, e.g., Jameson Land  and Hungary (Barbacka et al., 2019. The secondary veins are often obscured, and sporangia are rarely preserved, thus in situ spores are known from a few specimens only (e.g., Van Konijnenburg-van Cittert, 1993). Spores of Phlebopteris angustiloba have further been described by Lundblad (1950) and Tralau (1965). All those descriptions have been based on light microscopy only; there has never been enough material preserved to study the spores under SEM and TEM.
16-30 mm in length. (e.g., 147wü02, 59wü03; Plate V, 2). Adjacent secondary segments are connected through a narrow wing along the rachis; for most of their length, secondary segments keep the same width; solely at the apex, they taper rapidly forming a rounded apex. The venation is often difficult to recognise, but the central vein of the secondary segments is clearly ascertainable, and secondary veins arise at 70°-90°bifurcating at least once (Q302/03, Plate V, 3).
A number of isolated fertile primary-segment fragments have been also found (e.g., Q115/02, Q875/11, Q960/14, Plate V, 5, 6). These resemble the sterile fragments in size and shape. Complete secondary segments are up to 52 mm long (Q960/14), although a length of c. 15 mm is more common (Q875/11). On both sides of the central vein is a row of sori, with receptacula c. 1 mm apart. The sori reach up to 1 mm in diameter and consist of 6-8 sporangia (100wü02; Plate V, 6), but these have often fallen off and as a consequence, only the receptaculum is left/preserved (Plate V, 5). Although coaly material has been preserved occasionally, no in situ spores could be isolated from the material at hand.
Remarks: Phlebopteris muensteri was first described as Laccopteris muensteri by Schenk (1865Schenk ( -1867 from the Hettangian flora of Bavaria (from the type locality Theta) in detail including in situ spores. The latter is remarkable because in situ spores are commonly not preserved in any fossils of this species; in fact, they are only known from a few German specimens (Van Konijnenburg-van Cittert, 1993). Gothan (1914) described sterile and fertile material from a number of Hettangian Bavarian localities as Laccopteris sp.; Weber (1968) reported specimens from a number of additional Hettangian localities from Germany; Weber (1968) also reported descriptions and illustrations of complete juvenile fronds (Weber, 1968, pl. 7, figs. 60, 61). Hirmer and Hörhammer (1936) gave the most elaborate description of Phlebopteris muensteri in their revision of the Matoniaceae. Spores were first described in detail by Van Konijnenburg-van Cittert (1993). Hirmer and Hörhammer (1936) included specimens in Phlebopteris muensteri that have later been recognized as a different species, viz. Phlebopteris lunzensis by Pott et al. (2018).
Harris (1931,1980) considered Phlebopteris muensteri and Phlebopteris braunii (Göppert) Hirmer et Hörhammer as conspecific, as he was of the opinion that they were sun (P. muensteri) and shade (P. braunii) fronds of the same species, as both species were usually reported as occurring together in the same localities. Harris (1980) merged them into a single species and named it Matonia braunii Harris, 1980. Barbacka et al. (2016 followed Harris' (1980) hypothesis when describing specimens from the Jurassic of Poland. Van Konijnenburg-van Cittert (1993) in contrast separated the two species, because the sori in Phlebopteris muensteri are larger (diameter c. 1 mm), covering a prominent part of the lamina, than those of Phlebopteris (Matonia) braunii (diameter c. 500 μm). In addition, the spores of both species differ considerably in size (60-70 μm in P. muensteri and c. 50 μm in P. (M.) braunii). The latter view is supported by localities were only one of the two species has been reported, such as Wüstenwelsberg and a number of Iranian/Afghan localities (Schweitzer et al., 2009). Barbacka et al. (2019) described Phlebopteris kirchneri from the Hettangian of Hungary. On first view, it appears to be similar to P. muensteri but in well-preserved specimens, the delicate secondary venation is visible, which consists of oval to hexagonal meshes. This distinct feature would be unique amongst Phlebopteris species but is usually very common in Thaumatopteris species; therefore, this species might belong to the latter genus instead.
Another species that is similar to Phlebopteris muensteri is Phlebopteris lunzensis (Stur ex Krasser) Pott et Bomfleur, 2018, from the Carnian of Lunz (Austria) . Hirmer and Hörhammer (1936) considered P. lunzensis to be conspecific with P. muensteri, but Pott et al. (2018) regarded it to be a separate species based on the possible occurrence of an indumentum on the primary-segment bases and the secondary-segment veins, the wider distances between the secondary segments and the widely separated individual sori. Pott et al. (2018) also briefly described and figured a specimen tentatively assigned to Phlebopteris sp., which resembles specimens of Phlebopteris muensteri, figured by Schweitzer et al. (2009) from Iran (their pl. 4, figs. 1, 2).
Remarks: Although some authors (e.g., Seward and Dale, 1901;Herbst, 1992) consider Clathropteris as a subgenus of Dictyophyllum, Choo and Escapa (2018) in their phylogenetic study of the Dipteridaceae, found that Clathropteris always appeared as a highly distinctive monophylic clade, recognizing that Clathropteris as separate genus is justified.
The earliest records are reported from the Ladinian of Thale, Germany (Kustatscher and Van Konijnenburg-van Cittert, 2011, p. 226), but the attribution of those specimens to Clathropteris meniscioides is questionable as only based on "typical venation," which is not a reliable character here (see Choo et al., 2016;Pott et al., 2018), and segment margins are not preserved. An assignment to C. reticulata is consequently suggested after the re-investigation of the original specimens by Pott et al. (2018). Carnian records are, e.g., from Germany (Frentzen, 1922) and Malaysia (Kon'no, 1972). Clathropteris platyphylla (Göppert) Brongniart, 1849 and C. muensteriana Schenk, 1865-1867 are nowadays considered to be conspecific with C. meniscioides (see, e.g., Pott et al., 2018).
More than 20 Clathropteris species have been described in the past, but ambiguous features, incomplete preservation and high morphological variability render the delimitation of many species difficult (Choo et al., 2016). As a result, many species have been merged into one of the two common and broadly defined species, viz. C. meniscioides and C. obovata Ôishi, 1932. Clathropteris obovata is distinguished from C. meniscioides by its typical (and smaller) obovate primary segments with sub-acutely and deeply lobed margins and secondary veins arising at lower angles of c. 45° (Harris, 1961;Schweitzer et al., 2009). Moreover, the primary vein is often much wider in C. meniscioides than in C. obovata (c. 2.0 mm versus 0.5-1.0 mm) (Schweitzer et al., 2009).
Another rather common species is the Ladinian-Carnian Clathropteris reticulata Kurr ex Heer, 1877 . Clathropteris reticulata is distinguished from C. meniscioides by its symmetrical teeth with evenly rounded to slightly acute tips, while in C. meniscioides teeth are asymmetrical and acutely pointed . Additionally, venation of C. reticulata appears to produce even more regularly square areoles than C. meniscioides where the areoles are more often polygonal than regularly square (Choo et al., 2016;Pott et al., 2018).  fig. 9.
Description: Dictyophyllum exile is one of the most common ferns in the flora. Some 90 specimens have been found so far. Most of them are primary segment fragments, but a few yield primary segments attached to the rachial arm (64wü08, 195wü08, Q615/08; Plate VII, 5, 7). Q615/08 shows at least eight attached primary segments (Plate VII, 5) that appear to be connate for less than 1 cm. Both 64wü08 (Plate VII, 7) and 195wü08 have at least six primary segments and especially in 195wü08, these are clearly free almost to their base and partly overlap each other. Primary segments are usually 18-28 mm wide in their presumed middle portion (see e.g., Q616/08, 97wü08). Specimen 98wü02 yields an apical primary segment fragment where the width tapers from 20 to 10 mm without the exact apex preserved. The central primary vein is distinct and 1 mm wide. Segment margins are strongly lobed (almost dentate) with sometimes almost falcate lobes. The lobes are predominantly sub-oppositely arranged, with a prominent secondary vein entering from the primary segment that gives off a complex reticulum of tertiary and quaternary veins (Plate VII, 7). The apices of the lobes are rather variable, but are usually acutely rounded (e.g., Q616/08, Plate VII, 6).
Remarks: In their phylogenetic study on the Dipteridaceae, Choo and Escapa (2018) did not recognize Dictyophyllum as a separate clade, but assigned some Dictyophyllum species (including, e.g., D. exile and D. nathorstii) to the new genus Sewardalea Choo et Escapa, 2018, that also included many Camptopteris species. The main difference to other fossil genera in the Dipteridaceae lies in the number of primary segments attached to a single rachial arm (N 12, but up to 100). However, we doubt if this is a feature diagnostic for a genus and prefer to identify the present material as Dictyophyllum exile. The remaining Dictyophyllum spp. were kept by Choo and Escapa (2018) in the unresolved group "Dictyophyllum" together with Kenderlykia (Turutanova- Ketova, 1962).
The primary segments in the specimens from Wüstenwelsberg are always basally connate to a maximum length of 1 cm, hence we attribute this material to D. exile. Webb (1982) mentioned that sori in D. exile are round, arranged very crowded over the whole lower surface and c. 0.5 mm in diameter, while those of D. nathorstii are more variable in outline, less densely scattered on the lower surface but more concentrated near the veins, and smaller, only up to 0.2 mm in diameter.
Plate VIII. Macroremains of Dipteridaceae and incertae sedis from the Rhaetian of Wüstenwelsberg with specimen numbers. 1. Thaumatopteris brauniana, long secondary segments showing venation, Q428/06. 2. Thaumatopteris brauniana, venation, 02wü04. 3. Thaumatopteris brauniana, fertile specimen, Q230/02. 4. Spiropteris sp., Q426/05. Scale bars 1, 3: 10 mm; 2, 4: 5 mm. Description: Thaumatopteris brauniana is the third representative of the Dipteridaceae in the Wüstenwelsberg flora with 16 specimens found so far. These are only small primary segment fragments up to 5.2 cm long (Q230/02) or detached fragments. The central primary vein is 1.5-2.5 mm wide and longitudinally striate (Plate VIII, 3). Lobes (or secondary segments) are attached at angles of 80°-90°and are almost free up to the base with a narrow wing along the central vein ("rachis") connecting neighbouring secondary segments. The secondary segments are inserted oppositely to sub-oppositely (Q230/02; Plate VIII, 3) and, although, none is complete, they probably reached a considerable length of at least 4.1 cm (Q428/06; Plate VIII, 1); secondary segments taper in width from basally 11 mm to 4 mm distally. The secondary-segment margin is basally almost straight while it is strongly lobed distally (Plate VIII, 1); no apex was preserved. The secondary segment's central vein (secondary veins) is conspicuous but the tertiary (and quaternary) veins are commonly indistinct (Plate VIII, 1). When visible (Q428/06; Plate VIII, 2), they form a network of irregular, often elongate, hexagonal meshes. All specimens represent sterile fragments, except for one (Q230/02) that appears fertile, but no details of sori, sporangia or spores are visible, only some imprints of possible sporangia could be observed.
Remarks: Thaumatopteris was published by Göppert (1841Göppert ( -1846 with T. muensteri as the only and consequently type species. However, Thaumatopteris muensteri has been allocated to Dictyophyllum (Nathorst, 1876;, thus rendering Thaumatopteris illegitimate. The genus was, however, continuously in use and new species were added (e.g., Schweitzer, 1978;Schweitzer et al., 2009). Nathorst (1907 identified the two most important differences between Thaumatopteris and Dictyophyllum (see below) and defined an updated generic diagnosis with T. brauniana as type species. Nomenclatorially, this is illegitimate, but the genus has been continuously used in this way until to date (e.g., , as index fossil for the Thaumatopteris zone in Jameson Land). Zijlstra and Van Konijnenburg-van Cittert (2019) submitted a proposal to conserve the generic name Thaumatopteris with the conserved type T. brauniana considering it distinct from Dictyophyllum. The characters distinguishing Thaumatopteris from Dictyophyllum given by Nathorst (1907) are the almost perpendicular insertion of the basally constricted secondary segments to the central vein ('rachis') of the primary segments in Thaumatopteris versus the more obliquely and basally broadly inserting secondary segments in Dictyophyllum. Schweitzer (1978) added differences in sori (fewer but larger in Thaumatopteris than in Dictyophyllum) and sporangial size (smaller in Dictyophyllum than in Thaumatopteris). Choo and Escapa (2018) recognized Thaumatopteris as a separate clade in the Dipteridaceae based on laminal dissection, arrangement of primary segments and sporangial diameter.
Thaumatopteris first occurred in the Anisian of Argentina (T. barrealensis Bodnar et al., 2018) and late Ladinian of the Dolomites (Thaumatopteris sp.; Kustatscher et al., 2014). It became more abundant during the Carnian of Australia (T. shirleyi Herbst, 1979) and Lunz, Austria (T. lunzensis Stur ex Krasser, 1909 [attributed to Dictyophyllum lunzensis by Pott et al., 2018]) and extended up to the Hettangian with its heyday in the latter period (e.g., the indexed Thaumatopteris zone in Jameson Land; Harris, , 1937. Other Hettangian occurrences include Poland (Pacyna, 2013), Hungary (Barbacka et al., 2019), Romania (Popa et al., 2003) and Iran (Kilpper, 1964;Schweitzer, 1978). Although in Germany, the abundance of Thaumatopteris is much higher in the Hettangian (Schenk, 1865(Schenk, -1867Gothan, 1914;Weber, 1968), the species occurs, as documented here, also in the Rhaetian. Nathorst (1878) described Thaumatopteris schenkii Nathorst, 1878, from the Hettangian flora of Scania. The species resembled T. brauniana closely, differing only in the almost straight to slightly sinuous secondary segment margin in T. brauniana and the more lobed margin in T. schenkii. Nathorst (1907) included part of the Bavarian material described by Schenk (1865Schenk ( -1867 as T. brauniana in T. schenkii. Schenk (1867, pl. 18, fig. 1) figured both secondary-segment types in the discussion of T. brauniana, stating that the secondary-segment margin is entire near the primary-segment rachis and more lobed near the apex. Schenk (1865Schenk ( -1867 was of the opinion that these morphologies belonged to the natural variability of one species. After Nathorst's work, several scholars distinguished between both species, thereby agreeing with Nathorst (1907), even from the same localities (e.g., Lundblad, 1950;Weber, 1968), while others used one specific name for particular material (e.g., Gothan, 1914used T. schenkii, while Popa et al., 2003. Next to T. schenkii, Kilpper (1964) described Thaumatopteris bipinnata Kilpper, 1964 from the Hettangian of Iran differing only in the fact that a few secondary segments were so deeply lobed in their apical part that they were almost bipinnate. However, nowadays, most authors (e.g., Schweitzer et al., 2009, and references therein) consider the three species conspecific and use the name T. brauniana that deserves priority, as we also do with the specimens from Wüstenwelsberg. Stanislavski (1976, pl. 3, text- fig. 6) reported T. variabilis Stanislavski (1976), from the Upper Triassic of the Donets Basin, which shows all three secondary-segment shapes in one frond), comparing it with T. brauniana, T. schenkii and T. bipinnata, and with a few of Ôishi's (1932) Rhaetian species such as T. elongata Ôishi, 1932 (with relatively long secondary segments) and T. nipponica Ôishi, 1932 (with relatively short secondary segments). In our opinion, it is very likely that all these species fall within the natural variability of one species, viz. T. brauniana.
Remarks: Circinnate vernation is typical for most if not all fern fronds. At this stage of development of the fern frond, it is impossible to say, which fern species a leaf belongs to if found isolated. This applies especially to fossil fronds and therefore, these fronds are commonly assigned to the fossil-genus Spiropteris. Consequently, it is impossible to assign these specimens to any species. Therefore, we keep the specimens unassigned as Spiropteris sp.

Composition of the flora
The Rhaetian flora from Wüstenwelsberg is currently under detailed study by the authors (see Van Konijnenburg-van Cittert et al., 2018b, and references therein). Here, we discuss the ferns and fern allies (sphenophytes and lycophytes) of this assemblage. Especially the ferns also constitute an abundant portion of the flora, just as the seed ferns, the cycads and the Bennettitales Van Konijnenburg-van Cittert et al., 2018a, 2018b. As the entire composition of the flora is not yet entirely known, we can only compare the fern portion of the flora with that of contemporary and adjacent floras of the Northern Hemisphere.
Lycophytes are rare; two species of two different orders, Isoetales and Selaginellales, have been recorded. One specimen of the isoetalean Lepacyclotes sp. has been found, a genus that is commonly recorded from Upper Triassic-Lower Jurassic outcrops in the Germanic Basin, but records form the Rhaetian were unknown until to date. The spikemoss Selaginellites coburgensis is quite well known from Wüstenwelsberg (Van ; both sterile and fertile material with in situ spores has been described in detail previously. Equisetites laevis is the only horsetail recorded so far from the assemblage. Ferns are represented by three families. The Osmundaceae are rare with three species of Cladophlebis/Todites, all of which occur only with small fragments in low numbers. Matoniaceae occur in much higher numbers of fossil remains and thus were more common; two species of Phlebopteris have been identified, of which especially P. muensteri is very abundant and reflected by the presence of more or less complete fronds. Dipteridaceae also represent a large portion among the cryptogams from Wüstenwelsberg, especially in terms of abundance of specimens. Three species have been identified, one each of the genera Clathropteris, Dictyophyllum and Thaumatopteris. Thaumatopteris is the least common, while Clathropteris and Dictyophyllum are abundant. It is interesting to note that no marattialean species has so far been recorded from the flora, while representatives of these families have been recorded from other Rhaetian-Hettangian floras (see below).

Comparison with other Rhaetian floras from the Northern Hemisphere
The flora from Wüstenwelsberg is a typical Rhaetian flora yielding key pteridophyte taxa such as Equisetites laevis, Phlebopteris angustiloba, Phlebopteris muensteri, Dictyophyllum exile, Clathropteris meniscioides and Thaumatopteris brauniana, although some of these taxa also extend into Jurassic floras. These species place the Wüstenwelsberg flora in line with the renowned Rhaetian floras from Jameson Land (Greenland, Harris, 1926, 1937), Scania (Sweden, Nathorst, 1876, 1878, 1907Lundblad, 1950;Pott and McLoughlin, 2011), southern Poland (Pacyna, 2014;Barbacka et al., 2014aBarbacka et al., , 2014b, the Donets Basin (e.g., Stanislavski, 1971) and Alborz, Iran (Schweitzer et al., 1997(Schweitzer et al., , 2009. Especially the floras from Jameson Land, Scania, Franconia and, to a lesser degree, southern Poland have several of the mentioned key Rhaetian taxa in common (Table 1). Floras further to the east such as those from the Donets Basin and Alborz in Iran share less taxa with the central European Rhaetian floras (Table 1), although the Alborz floras have 5 (or possibly 6) taxa in common with the Rhaetian flora from Franconia (Schweitzer et al., 1997(Schweitzer et al., , 2009. The flora from the Donets Basin (Stanislavski, 1971) shares only one species with all the other floras, viz. Clathropteris meniscioides. The Rhaetian-Hettangian flora from the Cheliabinsk Basin (eastern Ural) does not even have one species in common with the Wüstenwelsberg flora (Kryshtofovich and Prinada, 1933).
Clathropteris meniscioides is the only species that apparently occurred in all Rhaetian floras. Most other species occur in at least three or four floras, viz. Equisetites laevis, Todites roessertii, Phlebopteris muensteri, P. angustiloba, Dictyophyllum exile and Thaumatopteris brauniana (Table 1). Solely, Todites (Cladophlebis) scoresbyensis has been recorded from Franconia, Jameson Land and Scania only. Consequently, the fern flora of Wüstenwelsberg lacks any species with remarkable features, except for pinna fragments assigned to Cladophlebis sp. that yield small fragments of cuticle that are described here, a feature that is very rare in ferns.

Comparison with the Hettangian flora of Franconia
The Rhaetian flora from Wüstenwelsberg shows some differences with the Hettangian flora from adjacent areas in Franconia (see Van Konijnenburg-van Cittert et al., 2014, 2018bPott et al., 2016, and references therein). All major plant groups are present, but the species and even genera within the two floras vary considerably; for details on other groups than the ferns and fern allies, see Pott et al. (2016) and Van Konijnenburg-van Cittert et al. (2018b). In the Hettangian flora of Franconia, a Lepacyclotes species occurs as well, viz. L. kirchneri, although it is rare and only known from one locality (Bauer et al., 2015) while spikemosses (Selaginellales) have not been recorded so far in the Hettangian flora of Franconia. The number of equisetalean genera and species is higher in the Hettangian flora; not only is there a different Equisetites species: E. muensteri instead of E. laevis in the Rhaetian flora. Representatives of two additional genera occur in the Hettangian flora, viz. Neocalamites lehmannianus (Göppert, 1841(Göppert, -1846 Weber, 1968, and Schizoneura carcinoides (Harris) Weber, 1968(Weber, 1968. Ferns are also more abundant and more diverse in the Hettangian than in the Rhaetian flora of Franconia (see e.g., Schenk, 1865Schenk, -1867Gothan, 1914;Weber, 1968). Representatives of several This study Pott and McLoughlin (2011), Lundblad (1950) Pacyna (2014), Barbacka et al. (2014b) Stanislavski Schweitzer et al. (1997Schweitzer et al. ( , 2009 families, not yet present in the Rhaetian, occur in the Hettangian flora, such as the marattialean Marattia intermedia (Münster) Kilpper, 1964, and Gothan, 1914 is the more abundant species in the latter flora.
The two matoniaceous species from Wüstenwelsberg, Phlebopteris muensteri and P. angustiloba, occur in the Hettangian flora as well, but with Selenocarpus muensterianus (Presl in Sternberg) Schenk, 1866, there is an additional but rare Hettangian species. This species was long thought to be endemic to Franconia (Harris', 1961 record from Yorkshire was a misidentification), but Czeir (1999) described the species from the Liassic of Romania and this was confirmed by Popa and Van Konijnenburg-van Cittert (2006). In contrast to its very abundant occurrence in the Rhaetian flora, the dipteridaceous Clathropteris meniscioides is comparatively rare in Hettangian floras, but Thaumatopteris brauniana, in turn, is again more common in the Hettangian than in the Rhaetian. The fern and fern ally genera found in the Wüstenwelsberg flora are all common genera found throughout most of the Mesozoic. Equisetites and Cladophlebis/Todites have been recorded from all over the world with large numbers of species (Tidwell and Ash, 1994;Collinson, 1996;Skog, 2001;Kustatscher et al., 2018). Phlebopteris has been known from the Late Triassic onwards and became widespread during the Jurassic but has only a few Cretaceous representatives (Van Konijnenburg-van Cittert, 1993;Tidwell and Ash, 1994;Collinson, 1996). It was especially widely distributed in the Northern Hemisphere but taxa from, e.g., South America and Australia are known as well (Tidwell and Ash, 1994). From the Late Jurassic onwards, Phlebopteris tends to disappear from the northern regions, and to date the family only occurs in the Malesian Archipelago. The same applies for the representatives of the Dipteridaceae in the Wüstenwelsberg flora. All three genera have their first occurrences in the Middle Triassic to early Late Triassic. Thaumatopteris is mainly known from Late Triassic-Early Jurassic localities, while Clathropteris and Dictyophyllum are well-known in the Middle Jurassic but decline during the Late Jurassic and only one species of Dictyophyllum is still known from the Wealden (Skog, 2001).
For Jameson Land,  established the Lepidopteris zone for the Rhaetian beds (with Lepidopteris ottonis as index fossil) and the Thaumatopteris zone for the Hettangian beds (with Thaumatopteris brauniana and Phlebopteris angustiloba as index fossils). Although these zones have been in general use since then, Thaumatopteris brauniana and Phlebopteris anugustiloba occur in Rhaetian sediments as well, albeit in much lower numbers than in Hettangian floras (Table 1; see e.g., Lundblad, 1959;Pacyna, 2014). The characterization of the zones thus should be only used carefully nowadays.

Paleoecological and paleogeographical implications
Climate conditions during the Rhaetian in Europe are generally reconstructed as hot and arid (Preto et al., 2010), but more humid conditions may have prevailed locally and for short periods of time (Bonis et al., 2010). In Wüstenwelsberg, this hypothesis is supported in the palynomorphs by a spike in horsetail, lycophyte and fern spores, and remains of aquatic algae (e.g., representatives of the genera Botryococcus, Cymatiosphaera and Tasmanites) in some of the layers (Bonis et al., 2010), indicating that bodies of stagnant or slowly running water existed in the Wüstenwelsberg area during the latest Rhaetian (Van Konijnenburg-van Cittert et al., 2014, 2016, 2018b).
The diverse and abundant flora, rich in hygrophytic elements, which in many cases are dependent on the presence of water for their reproduction cycles, supports this. Selaginellites coburgensis, a small and delicate lycophyte, would have grown near water bodies in the more humid understorey, just as the sphenophyte Equisetites laevis and possibly the lycophyte Lepacyclotes sp.; the latter might also have grown on more open and disturbed habitats (see e.g., Kustatscher et al., 2010). The osmundaceous ferns Todites and Cladophlebis could have been small arborescent plants with slender stems (Schenk, 1865(Schenk, -1867Schweitzer, 1978;Taylor et al., 2009;Barbacka et al., 2019) similar somewhat to modern tree fernspreferring warm and humid environments, such as riverbanks, lake shores, freshwater marshes, or brackish near-coast environments (Harris, 1961;Vakhrameev, 1991;Van Konijnenburg-van Cittert and Van der Burg, 1996;Deng, 2002;Van Konijnenburg-van Cittert, 2002;Wang, 2002;Sun et al., 2010). They could have lived also in slightly more disturbed and wetter environments although these plants probably had high adaptation to moderately disturbed and relatively dry environments (Barbacka, 2011;Bodor and Barbacka, 2012).
The ecology of matoniaceous ferns during the Mesozoic is variable. They are known both as arborescent ferns as with short stems and an extended rhizome system (Schweitzer, 1978). Phlebopteris angustiloba and Phlebopteris muensteri could have had a similar morphology. They are considered herbaceous plants with large fronds that grew in humid environments under low light conditions (understorey; Schweitzer, 1978;Wang, 2002;Bomfleur and Kerp, 2010). They could have been also pioneer plants that colonized disturbed, short-lived, moderately wet areas formed by alluvial deposits (Barbacka et al., 2010(Barbacka et al., , 2019Barbacka, 2011).
Dipterid ferns occupied during the Mesozoic mainly humid and warm-temperate to subtropical climate zones. They are considered opportunistic plants colonizing disturbed habitats like riverbanks, exposed ridges and clearings (Cantrill, 1995;Van Konijnenburg-van Cittert, 2002;Bomfleur and Kerp, 2010;Pott et al., 2018;Barbacka et al., 2019). Representatives of Clathropteris, Dictyophyllum and Thaumatopteris are commonly considered herbaceous plants (Schweitzer, 1978;Wang, 2002;Bomfleur and Kerp, 2010). Wang (2002) suggested that Dictyophyllum species with large fronds could be dwellers in humid environments under reduced light conditions (understorey). Thaumatopteris brauniana has been reconstructed with several metres long, horizontally growing rhizomes (Schweitzer, 1978) based on its resemblance with the modern analog Dipteris Reinwardt, 1825. Barbacka (2011) considered this species a plant colonizing highly disturbed and moderately wet habitats, whereas later, Barbacka et al. (2019) assigned it to the wettest and most disturbed habitats. Clathropteris meniscioides apparently formed also large monotypic stands in large areas along floodplains in environments, where light and water availability were not limiting factors for growth and thriving (Choo et al., 2016). In Antarctica, the plants were reconstructed as herbaceous members of open vegetation dominated by bennettitaleans that became a dominant element during the colonization phase of disturbed sites after catastrophic volcanic events (Bomfleur and Kerp, 2010).
The fern remains were collected from fossil-rich levels yielding also other abundant plant remains, including seed ferns, bennettitaleans and conifers. Considering that most plant remains co-occur in the same horizons, this suggests that both the xerophytic (such as some seed ferns and conifers) and hygrophytic (such as the ferns and fern allies) forms lived during the same period of time. This would suggest that the plants lived together in the same area and/or environment, but in different microhabitats. This would enforce the hypothesis that this area represented a complex environment with highly disturbed and rapidly changing conditions Van Konijnenburg-van Cittert et al., 2018b), such as perhaps due to the rise and fall of the sea level, which also could explain locally very abundant algae in the succession.

Conclusions
The fern remains from the Rhaetian of Wüstenwelsberg, Franconia, southern Germany, show a high diversity: two lycophytes (Lepacyclotes sp. and Selaginellites coburgensis) and one sphenophyte species (Equisetites laevis) and nine fern species have been identified, all with sterile and fertile fragments. Three fern families represent the ferns. For Osmundaceae, three species (Todites roessertii, Cladophlebis scoresbyensis, Cladophlebis sp.) with rare occurrences in the assemblage are reported; fossils of Matoniaceae are more abundant, representing two species (Phlebopteris muensteri, Phlebopteris angustiloba). Dipterid remains are very common, assignable to three species of each another genus (Clathropteris meniscioides, Dictyophyllum exile, Thaumatopteris brauniana). Frond fragments preserved in the developing stage showing circinnate venation (Spiropteris sp.) could not be assigned to any of the species.
The cryptogams of Wüstenwelsberg include key Rhaetian taxa such as Equisetites laevis, Phlebopteris muensteri, Dictyophyllum exile, and Clathropteris meniscioides. These are represented in most coeval Rhaetian floras of Central Europe such as Jameson Land, Scania, Franconia and southern Poland. However, taxa that are considered as key taxa for Hettangian floras, e.g., Phlebopteris angustiloba and Thaumatopteris brauniana, are present in Wüstenwelsberg, albeit not in large numbers. The comparison with the Jurassic floras of Germany shows, on the other hand, that the latter plant assemblages are much more diverse.
The lycophytes and sphenophytes from Wüstenwelsberg were probably part of the more humid or wetter environments of the understoreys. The ferns too possibly colonized the understorey of humid environments but were also pioneer plants of disturbed, shortliving, wet areas in the alluvial plains, just as the lycophyte Lepacyclotes sp. might have been.

Author declaration
None.

Declaration of Competing Interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.