Megaspores from the Late Triassic‒Early Jurassic of southern Scandinavia: taxonomic and biostratigraphic implications

ABSTRACT Here we investigate megaspores from 10 Triassic‒Jurassic localities of southern Sweden and Bornholm, Denmark, based on collections housed in the Swedish Museum of Natural History. We identify and describe 19 megaspore taxa belonging to three stratigraphically constrained assemblages, representing the Rhaetian, Hettangian and Pliensbachian, respectively. Megaspores are abundant and diverse (12 taxa) in the Rhaetian assemblage. Diversity markedly decreases across the Triassic–Jurassic boundary (TJB), with species richness reducing from 12 to two taxa. The Hettangian assemblage is well-preserved but depauperate, and is overwhelmingly dominated by Nathorstisporites hopliticus. A subsequent recovery of lycopsid diversity followed, recorded by an increase in richness to six taxa in the Pliensbachian assemblage. The disappearance of the hygrophilous and diverse heterosporous lycophyte communities across the TJB, suggests a shift to drier conditions in the earliest Jurassic. This is supported by lithological changes from coal-forming environments in the Rhaetian to sandstone-dominated fluvial-estuarine facies in the Hettangian. Throughout this study, we analysed the megaspores using fluorescence microscopy, which revealed detailed morphological features on specimens that were otherwise opaque under visible light. This non-destructive technique is particularly useful for examining opaque megaspores embedded in permanent mounting media, such as epoxy resin, and may provide new insights into historical megaspore collections elsewhere.

Here we aim to document and describe Triassic and Jurassic megaspores housed in the collections of the NRM. We assess the biostratigraphic relevance of these megaspores and apply UV fluorescence microscopy for morphological analysis of dispersed megaspores.

Geological setting
During the Triassic and Jurassic, basin formation in Scandinavia was controlled by continental rifting along the Sorgenfrei-Tornquist Zone. This is a continuous zone of deformation stretching from southeastern Europe to the central North Sea (Sorgenfrei & Buch 1964), and is a part of the Fennoscandian Border Zone, constituting a prominent component of the landscape in Skåne (Bergström 1984;Norling & Bergström 1987;Larsson et al. 2000). Swedish Upper Triassic and Lower Jurassic sedimentary successions are restricted to the southernmost province, Skåne. These successions were deposited in alluvial, deltaic, coastal plain and shallow marine shelf environments in several sedimentary basins along the Fennoscandian Border Zone ( Fig. 1; Ahlberg et al. 2003;Lindström et al. 2017).
In northwestern Skåne, uppermost Triassic (Rhaetian) to Lower Jurassic successions are represented by the Höganäs and Rya formations (Figs. 1-3). The Rhaetian-Hettangian Höganäs Formation incorporates three subdivisions: the Vallåkra, Bjuv and Helsingborg members (in stratigraphic order). The TJB lies within the Höganäs Formation, approximately at the transition between the Bjuv and Helsingborg members, and the facies spanning this interval comprise plantbearing mudstones, siltstones, sandstones and coals, characteristic of continental and deltaic environments. The overlying Rya Formation (Sinemurian-lower Aalenian) is subdivided into four subunits: the Döshult, Pankarp, Katslösa and Rydebäck members (Troedsson 1951;Norling 1968;Pieńkowski 1991;Vajda & Wigforss-Lange 2009;Lindström et al. 2017). A major part of the Lower Jurassic succession of NW Skåne (the three upper members of the Rya Formation) is marine in origin (Kruger et al. 2021, this issue), represented by sandstones, siltstone and ferruginous oolitic sandstones in the topmost Rydebäck Member. These successions yield diverse marine fossil assemblages, including bivalves, ammonites, ostracods and foraminifera, but also contain rich spore and pollen assemblages (Norling et al. 1993 and references therein).

Material
Megaspores studied here are from historical collections of the Department of Palaeobiology, NRM. The megaspores include specimens mounted in Canada balsam on glass slides, and unmounted specimens, stored in microfossil sorting trays (cardboard slides with glass covers). A portion of the material examined was collected by A.G. Nathorst and B. Lundblad from various localities across Skåne and Bornholm, and some specimens have been published (Nathorst 1902(Nathorst , 1908aLundblad 1950aLundblad , 1950bLundblad , 1956. Previously unstudied material is also investigated here from the Kävlinge drillcore (BH-928, the Norling foraminifera collection), the Höllviken II drillcore (the Brotzen foraminifera collection), and the Munkerup beds of Bornholm (collected by B. Lundblad). Details of each locality are provided below.

Helsingborg Harbour (Hettangian)
Megaspores mounted on glass slides embedded in Canada balsam by A.G. Nathorst (slide numbers: S064343, S064347, S064348, S064744, S064747) were studied. These derive from an outcrop section discovered during an expansion of Helsingborg Harbour in the late 1800s and dated to Hettangian (Lundblad 1956).

Drillcore Stabbarp BH-74, 5 m (Hettangian)
The Stabbarp borehole, No. 74 is situated immediately north of the old coal mine of Stabbarp, near to the town of Eslöv (Fig. 1). The sample containing the studied megaspore assemblage (slide number: S075612) is from a thin coal bed from a depth of ca. 5 m in the drillcore. These strata are also referred to as the "Gamla övre gruvan" (The old upper mine), which comprise the coal beds "John Ericsson" and "Jean Molin" and are Hettangian in age (Lundblad 1956).

Munkerup, Bornholm, Denmark (Hettangian)
Megaspores were examined from one sample from the Munkerup Member from coastal outcrops in southwestern Bornholm, Denmark. The material was accessioned into the collections of the NRM in 1961, possibly by B. Lundblad. The clay beds of Munkerup host abundant plant fossils characteristic of the Early Jurassic Thaumatopteris flora (Harris 1937), and the succession has been dated as early Hettangian based on these macrofossils and palynology (Koppelhus 1991).

Reflected light microscopy
Unmounted specimens were initially investigated using an Olympus SZX10 reflected light microscope at ×6.3 to ×63 magnifications. Due to the relatively low level of magnification of reflected light microscopy, selected specimens were subsequently examined using transmitted light, UV light and scanning electron microscopy.

Transmitted light microscopy
Specimens that were mounted on glass slides and embedded in Canada balsam (hampering SEM investigation) were examined using an Olympus BX51 transmitted light microscope, and photographed with an Olympus DP71 camera. Objective lenses of ×5 and ×10 were used to locate specimens, and ×20 was used for imaging. Where possible, slides were inverted and examined from both sides to assess proximal and distal surfaces of the megaspores.

Fluorescence microscopy
Megaspores that were mounted in Canada balsam were studied and imaged with UV light using an Olympus BX51 microscope and photographed using an Olympus DP71 camera. Various filters were used to examine megaspores under UV light, among which the WIB (blue filter) was optimal for imaging the studied material.

Scanning electron microscopy (SEM)
Megaspores from the drillcore Kävlinge BH-928, Billesholm and Munkerup (Bornholm) (Fig. 1) were examined using scanning electron microscopy. Thirteen specimens were mounted onto SEM stubs (with carbon tape) and gold coated for 20-50 seconds using an Agar high-resolution sputter coater. Specimens were studied using an ESEM FEI Quanta FEG 650 scanning electron microscope at the NRM.

Systematic palaeontology
Nineteen species belonging to eight genera were identified and described. Several taxa are presented in open nomenclature due to insufficient specimens, poor preservation or a lack of SEM observations for specimens sealed in glass slides. We follow the descriptive terminology of Punt et al. (2007).
Genus Banksisporites Dettmann, 1961 Remarks Two separate layers of the exine were first recognized as a taxonomically significant character for dispersed fossil megaspores by Høeg et al. (1955), and a new genus, Duosporites was erected. Likewise, Dettmann (1961) established Banksisporites based on the presence of a cavate exine; this consists of an intexine (homogeneous or composed of granules) and an exoexine (smooth or granular), which are fused near the laesurae on the proximal surface. However, Potonié (1966) rejected Banksisporites because the detachment of two exine layers was difficult to observe on the type species originally described by Harris (1935). Fuglewicz (1973, p. 419) also disagreed with the erection of Banksisporites and transferred B. sinuosus to Trileites. Batten and Kovach (1990) recorded the transfer of Fuglewicz (1973), but stated that the type species B. pinguis must remain unless this genus is regarded as a junior synonym of Trileites. Glasspool (2000) considered the presence of an inner body as a character of limited taxonomic importance, and did not incorporate this feature into the emended diagnosis of Singhisporites. Glasspool (2003) later emended the diagnosis of Banksisporites, using surface ornamentation as a taxonomic character, which differs from the original description by Dettmann (1961).

Remarks
The studied specimens are consistent with the type species, except that 35 of the 230 grains examined here, are larger than those described by Dettmann (1961), which range between 210 and 530 μm in equatorial diameter.

Description
Trilete megaspore, amb generally circular to subcircular, rarely rounded triangular. Laesurae straight to slightly sinuous, almost reaching the equator, labra strongly raised, or in a few cases weakly developed, tapering gradually in width. Curvaturae absent. Exine two-layered, cavate, adherent around the laesurae on the proximal face. Surface smooth as viewed with fluorescence, showing prominent folds mainly peripherally, and around the labra.

Remarks
The Skåne specimens are similar to those described by Scott & Playford (1985) in possessing prominent peripheral folds.

Description
Trilete megaspore, circular in outline. Laesurae straight, never extending to the equator, labra uniformly high with flat crests. Curvaturae distinct, perfectly developed, delimiting the contact area. Spore surface smooth, strongly folded on the distal face. Exine two-layered, inner layer indistinct on the single specimen examined.

Remarks
This specimen is similar to the type material of Banksisporites dettmanniae in having perfectly developed curvaturae near the equator, but the inner layer of the studied specimen is less distinct compared to those described by Banerji et al. (1978).

Remarks
Spores attributed to this species from Skåne have distinctly sinuous laesurae consistent with Banksisporites sinuosus specimens described from the Upper Triassic of Australia by Dettmann (1961). However, the Swedish examples differ in having longer laesurae that almost reach the equator. Fuglewicz (1973) transferred Banksisporites sinuosus to Trileites, suggesting the "inner body structure" was not considered a valid taxonomic feature for the erection of Banksisporites. However, Fuglewicz (1973) provided only SEM images of the external spore morphology, and transmitted light micrographs that may have revealed the inner body structure were lacking. The Polish specimens described by Fuglewicz (1973) are also markedly older (Early Triassic) than the Australian type material (Dettmann 1961). Thus morphological variation regarding the presence or absence of an inner body among these assemblages would not be particularly surprising given that they are separated by considerable time and distance. Here we retain the use of Banksisporites, as the inner body may prove to be a useful distinguishing feature among morphologically similar megaspore taxa in future studies, and may indeed shed light on potential producers.
Spore surface covered with dense short gemmae, verrucae, coni, spinae, bacula with sharp, rounded or blunt apices. These sculptural elements are in some cases united into rugulae.

Remarks
The ornamentation of our specimens is consistent with examples of Pusulosporites populosus described by Fuglewicz (1973, p. 425, pl Remarks Banerji et al. (1984) emended the generic diagnosis to encompass species with rounded bacula, and those with bacula confined to one face. However, Potonié (1956) already included those species with rounded bacula in the generic diagnosis to read: " . . . Exine covered on all sides with truncated or rounded bacula . . . " (translated from German, Potonié 1956, p. 35) Bacutriletes tylotus (Harris, 1935) Potonié, 1956 Description Trilete megaspore, more or less globular. Laesurae straight and narrow, usually spanning half the radius of the spore, never reaching the equator, labra weakly developed, up to 30 μm high on some specimens. Curvature absent. Entire spore surface covered by cylindrical bacula typically up to 50 μm long, and 30 μm wide at the base. Bacula smaller towards the proximal pole, bearing transversely truncate ends. Bacula straight or in a few cases bent, unbranched, generally occurring in groups of two or three, where they may be fused at their apices. Smaller bacula or grana weakly developed between larger sculptural elements.

Remarks
Our specimens are similar to those described by Harris (1935) in bearing diagnostic apically truncated and usually clustered bacula. Subordinate sculptural elements between bacula were faintly visible on specimens mounted in Canada balsam (Fig.  7B, D, F). The type of clustering and fusion of sculptural elements is similar to the ribbon-like architecture of Singhisporites hystrix (Slater et al. 2011, pl. IV); however, the specimens studied here lack the multi-branched and reticulate flared sculptural elements of S. hystrix, which are also relatively uniform compared to the more varied sculptural elements of Bacutriletes tylotus.

Description
Trilete megaspore, amb circular. Laesurae straight, labra well developed, narrow, generally tapering to the top. Curvaturae distinct as low ridges, narrow, imperfect. Contact area covered with more or less evenly distributed bacula, usually flattened at their apices and broadened and united basally to form a delicate reticulum.

Remarks
The single specimen recorded was heavily corroded so that only one interradial area possessed ornamentation. The more or less reticulate pattern formed by the short baculate appendages is similar to the description of Bacutriltes asaphus by Fuglewicz (1973). Specimens described as Bacutriltes reticuliferus by Bertelsen & Michelsen (1970) differs from Bacutriletes sp. cf. B. asaphus in having shorter laesurae, lacking curvaturae, and bearing densely spaced blunt bacula over the whole spore surface.

Remarks
This specimen is comparable with that of Bacutriletes srivastavae in bearing dense bacula that are not confluent at the base; and the configuration of the bacula on the spore outline under transmitted light (Fig. 9G) is identical to that of B. srivastavae (Banerji et al. 1984, pl. 4, fig. 5). However, on the studied specimen, bacula are smaller on the proximal face compared to the Indian specimens.

Description
Trilete megaspore, amb circular. Laesurae conspicuous, short, extending to half of the spore radius. Curvaturae absent. Varied coni, gemmae and short clavae present on contact area. Remainder of spore surface densely but irregularly covered with baculate appendages, up to 85 μm long and ca. 25 μm wide, cylindrical or slightly tapering to rounded or blunt apices. Bacula usually straight, or slightly bent near the apices, in some cases fused at the base and bifurcated at the top.

Remarks
The long bacula in the studied specimen are similar to those of Bacutriletes arnoldii (Miner, 1932) Potonié, 1956, B. clavatus Marcinkiewicz, 1960 corynactis (Harris, 1961) Marcinkiewicz, 1971, B. spicatus (Marcinkiewicz, 1962) Marcinkiewicz, 1971, B. onodios (Harris, 1961) Hopkins et Sweet, 1976and B. corynactiformis Fuglewicz, 1977 However, Bacutriletes arnoldii differs from the studied specimen in having vermiculate bacula; B. clavatus has clavae that are narrow at the base but widen at the top; B. corynactis and B. spicatus have baculate appendages that are usually curved, or even recurved; and B. corynactiformis has smooth contact areas and clavate appendages on the distal surface. The studied specimen is very similar to Bacutriletes onodios with respect to its straight and almost cylindrical bacula, however, the labra on our specimen are less prominent than those described by Harris (1961), which are 40-70 μm high.

Description
Trilete megaspore, amb subtriangular, laesurae distinct, straight or strongly curved near the proximal pole, labra well developed, with small gemmae and bacula. Curvaturae faintly conspicuous or lacking. Spore surface covered with numerous, densely distributed coni, verrucae, bacula, clavae and/or spinae, and a few short fimbriae. Sculpture more prominent on the distal face. Appendages usually curved or bent backwards, rarely fused at the base.

Remarks
These specimens are characterized by their variable and elongate surface ornamentation. We tentatively identify these specimens as Bacutriletes? sp. because the bacula are more developed than other appendages. The contact area is faintly delineated by denser ornament (Fig. 9E), suggesting that these specimens could alternatively be assigned to Narkisporites Kannegieser et Kozur, 1972; however, the sculptured curvaturae, which are typical of Narkisporites, are absent on one of the two specimens here assigned to Bacutriletes? sp. (Fig. 9A-C).

Description
Trilete megaspores, amb circular. Laesurae distinct. Labra straight and raised, up to 175 μm long and 65 μm high, rounded at the top, extended to curvaturae. Curvaturae conspicuous, bearing various sizes of coni, spinae and blunt bacula. Surface of contact area smooth, slightly swollen in the centre. Distal surface covered with evenly distributed bacula, slightly tapering to the top, blunt at the apices, not forming a reticulum.

Remarks
This specimen was attributed to Narkisporites Kannegieser et Kozur, 1972 based on the presence of a distinct curvaturae sculptured with coni, spinae and bacula, although sculptures on the curvaturae are heavily corroded, making comparison with other material difficult.

Description
Trilete megaspore, amb circular. Laesurae straight or slightly curved, narrow, almost extending to the equator, labra tapering to the top. Curvaturae developed perfectly, consisting of low ridges. Spore surface covered with reticula, lumina irregularly polygonal, ca. 5-20 μm, muri low and narrow, developed on both the proximal and distal surfaces. Finer reticula are present within the coarse reticula, representing the construction of the exoexine (Fig. 10D).

Remarks
The two specimens of Horstisporites planatus recorded here are consistent with those described by Marcinkiewicz (1971, pl. XII, figs. 1-9) from Poland. These specimens all have a thin exine, which allows the observation of reticulum under transmitted light; this feature was not visible under reflected light. The presence of a similar delicate reticulum described by Marcinkiewicz (1960Marcinkiewicz ( , 1971; areolae diameters of 4-30 μm) and clear curvaturae further support our identification.
Horstisporites compositus Li, Batten, Li et Peng (in press) is similar to the H. planatus specimens described here in that both have narrow curvaturae and a composite reticulate sculpture (a fine reticulum occurring within the lumina of a coarse reticulum). Li et al. (in press) note that the fine reticulation within the larger areolae is a manifestation of the open construction of the exoexine, thus its value for morphological classification should be carefully considered. For example, the fine reticulation is also present in H. minutus Schultz et Noll, 1987and H. alveolatus Schultz et Noll, 1987(Schultz & Noll 1987; however, H. minutus is smaller (spore diameter is consistently around 225 µm) and the muri of its reticulum are arranged radially; H. alveolatus shows subdivision of the coarse mesh of the reticulum into much finer segments (Schultz & Noll 1987, pl. 4, fig. 8) and areolae tend to be smaller towards the distal surface. Reticula on our specimens do not show evidence of radial arrangement, and are more or less evenly developed across the whole surface.

Description
Trilete megaspores, amb circular, laesurae narrow, curved, labra thin and short, ca. 160 μm long from the curvaturae to proximal pole, extending to two-thirds of spore radius. Curvaturae imperfect, narrow, ca. 75 μm long. Spore surface densely covered with irregular ridges, ca. 3 μm wide, tapering towards the rounded top, never sharply spinose at the crests, forming reticulate to slightly rugulate surface patterns. In the contact area, ridges coalesce to form tiny circular or polygonal reticula, lumina ca. 3 μm, but generally less than 5 μm, in diameter. Outside contact areas, ridges are arranged to form rugulae or imperfect reticula. Spore outline includes short projections formed by ridged sculpture.

Remarks
This specimen is similar to Horstisporites microlumenus Dettmann, 1961 in having very small lumina co-occurring with rugulae. However, the lumina of the studied specimen are even smaller (ca. 3 μm) than those (7-18 μm) described by Dettmann (1961). Furthermore, the curvaturae are variably absent to well developed on the specimens of Dettmann (1961, pl. 2, fig. 9) or perfectly developed (Dettmann 1961, pl. 2, fig. 8). The variation in ornamentation of the Swedish specimens, from a proximal reticulum to distal rugulae, is uncertain due to corrosion.

Remarks
Similarly to Horstisporites sp. described here, Horstisporites bertelsenii Fuglewicz, 1977 has distinct, strongly elevated trilete laesurae and lacks reticula in the contact areas. However, corrosion of the contact area of Horstisporites sp. and a lack of information on the distal surface hamper a more resolved identification.

Remarks
These specimens are similar to Horstisporites sulcatus Fuglewicz, 1973; both have well-developed laesurae and relatively high labra on the convex contact area, distinct curvaturae, and rugulate ridges on the distal surface. However, there are differences between our specimens and those described by Fuglewicz (1973); the specimens studied here possess wider (10-20 μm) rugulate ridges on the distal surface, which are narrow on the illustrated specimens (10-12 μm, Fuglewicz 1973, pl. XXX, fig. 3 and 4). The specimens of Fuglewicz (1973) have rugulate ornamentation that is different from the "alveolar to reticulate patterns" of Horstisporites Potonié, 1956, thus our specimens, and those described by Fuglewicz (1973), may need to be transferred to another genus in the future following more thorough examination. Our specimens are similar to Biharisporites scarber Marcinkiewicz, 1960, with respect to the occurrence of distinct curvature and "larger clusters of low appendages that covered the entire surface of the spore" (translated from Polish, Marcinkiewicz 1960, p. 719), and to illustrations of B. scarber in subsequent publications (e.g. Marcinkiewicz 1962, pl. IX, fig. 5, 1971, pl. VIII, fig. 2, 1981. However, in a later description of B. scarber, Marcinkiewicz (1981) states "the distal surface of the sizable area is decorated with small conical processes, irregularly shaped nodules appear on the contact surfaces" (translated from Polish, Marcinkiewicz 1981, p. 83). This agrees with the generic diagnosis of Biharisporites Potonié, 1956emend. Bharadwaj et Tiwari, 1970 ("exine ornamented with coni, setae or spinae of various shapes . . . ", Bharadwaj & Tiwari 1970, p. 38), but is different to the description of Marcinkiewicz (1960) and that provided here, therefore we do not assign the specimens recorded here to Biharisporites.

Description
Trilete megaspore, elliptical in equatorial view. Laesurae distinct, characterized by raised membraneous labra that are

Description
Trilete megaspore, slightly elliptical in lateral view. Laesurae distinct, usually with strongly developed membraneous labra, up to ca. 75 μm high, dissected and slightly convoluted, subtended by elongate and membraneous appendages ca. 30 μm high, spinae occur near appendages, up to ca. 25 μm long. Curvaturae distinct, narrow, raised to form a narrow equatorial zona, merged with labra and membraneous appendages. Both surfaces covered with a course irregular polygonal reticulate ornament.

Remarks
This specimen differs from: Paxillitriletes ales (Harris, 1935) Batten et Koppelhus, 1993 in bearing reticula; from P. fairlightensis (Batten, 1969) Hall et Nicolson, 1973 in its larger size; and from P. granulatus (Mädler, 1954) Hall & Nicolson, 1973 in lacking fine granules and small warts on the distal surface. Similarly to Paxillitriletes reticulatus, P. arcticus, P. phyllicus and P. thorenensis, Paxillitriletes sp. from Kävlinge has reticulate ornamentation. Paxillitriletes arcticus possesses varied meshes, and based on the micrographs of Bose (1961), the reticulum is imperfectly developed. Paxillitriletes phyllicus and P. thorenensis have spinae and warts, respectively, on the muri. Paxillitriletes sp. is most similar to P. reticulatus, the specimen is slightly corroded and the spinae near the labra on the proximal surface are poorly preserved (Fig. 12E), making further identification difficult.
Genus Nathorstisporites Jung, 1958 Nathorstisporites hopliticus Jung, 1958 (Figs. 13 and 14) Description Trilete megaspore, amb circular. Laesurae distinct, straight or curved, bordered by highly raised membranous labra that are jagged at the crests, partly consisting of strongly sinuous capilli (long hair-like appendages), up to 170 μm long. Capilli grow on numerous rounded verrucae in the contact area and are in some cases bifurcate at the apices, and small coni/tips occur on the top of rounded verrucae on the distal surface. Curvaturae distinct, formed by densely distributed circular verrucae near the equator, merged with the laesurae. Circular verrucae generally ca. 50 μm wide, intervening spaces are ca. 10-50 μm wide. Smaller gemmae, coni and verrucae occur on larger verrucae and over the entire spore surface.

Remarks
The specimens recorded here are larger than those described by Jung (1958;size range: 375-535 μm). Nathorstisporites hopliticus differs from: N. cornutus Fuglewicz, 1980 in lacking horn-like appendages near the laesurae; from N. flagellulatus Dettmann, 1961 in lacking thick and large spine-like processes on the distal surface. Nathorstisporites imprimus Reinhardt et Fricke, 1969 has bacula and verrucae on the distal surface that differ from the coni of N. hopliticus. Nathorstisporites muricatus Scott et Playford, 1985 has complex and long proximal capilli relative to its small size. Nathorstisporites nammalensis Sah et Jain, 1968 andN. reticulatus Dettmann, 1961 are both distinguished from N. hopliticus by their reticulate surface. Nathorstisporites pulcherrima Helby, 1966 has varied spine-like appendages on the distal surface. Nathorstisporites peltasticus has shorter lips, rarely bifurcate proximal capilli and blunt coni.

Discussion
Based on the occurrence of megaspores from the various study sites, we recognize three stratigraphically constrained assemblages, corresponding to the Rhaetian, Hettangian and Pliensbachian stages, respectively (Fig. 3).   (Kovach & Batten 1989).

The Hettangian assemblage
This assemblage comprises two taxa, Nathorstisporites hopliticus and Bacutriletes? sp. With the exception of Bacutriletes? sp., all megaspore taxa of the Rhaetian assemblage disappear and are replaced by N. hopliticus, indicating a major loss in diversity across the TJB. Although N. hopliticus is restricted to the Hettangian assemblage, it is recorded from several sites and is generally abundant, but the record of Bacutriletes? sp. is based on a single specimen. Nathorstisporites hopliticus has previously been found in situ within the reproductive structure of the Pleuromeian lycophyte, Lycostrobus scottii (Nathorst 1908a;Lundblad 1956;Balme 1995). The loss of most of the Rhaetian megaspore taxa recorded here, and the dominance of N. hopliticus in the Hettangian indicate a major change in heterosporous lycopsid parent vegetation throughout the Triassic-Jurassic interval in Scandinavia. A decline in megaspore diversity across this interval has also been observed from various localities on a global scale, from Greenland (Harris 1937), Sweden (Lundblad 1956(Lundblad , 1959, this study), Germany (Jung 1960), Poland (Marcinkiewicz 1971), China (Li et al. in press) and Norway (Morris et al. 2009). The disappearance of Sellaginellales and Isoetales from the Rhaetian to the Hettangian assemblage likely represents a shift to drier conditions, which is supported by the lithological change from coaland mudstone-dominated, to sandstone-dominated facies across the Triassic-Jurassic transition.

The Pliensbachian assemblage
This assemblage was identified from material collected by E. Norling during his work on foraminifera from Kävlinge BH-928 (Norling 1968). Six megaspore taxa were recognized from four drillcore samples (Fig. 3 (Marcinkiewicz 1971). The presence of this species in sample 34.75-33.75 m, therefore, suggests an age younger than late Sinemurian; this agrees with dating based on foraminifera that indicates interval 54-28.75 m is Pliensbachian (Norling 1968).
The producers of several taxa recorded in the Pliensbachian assemblage remain unresolved, for example, Horstisporites has been described with various wall structures (Kovach 1994), and these different forms may have been produced by different parent plants. However, Paxillitriletes sp., probably indicates the presence of isoetalean lycopsids at this time (Kovach 1994).

Fluorescence microscopy
A common problem that arises when examining megaspores is that specimens are commonly opaque under transmitted light microscopy (e.g. Figs. 6E, 7A,C, 10C,E and 13A). This may inhibit identification of the megaspores and is difficult to overcome if the specimens are embedded in a permanent mounting medium, such as epoxy resin or Canada balsam. Reflected light microscopy and SEM may provide a solution, but only if the specimens have not been embedded in this way. Here we used UV fluorescence microscopy to examine opaque megaspores that were embedded in a fixed mounting medium. Prior to the use of this technique, we were only able to assess the overall shape and size of the opaque specimens (e.g. Fig. 10E). Fluorescence microscopy is used frequently in the analysis of plant cuticle remains (e.g. Bomfleur et al. 2014), and to a lesser extent for palynology and palynofacies analysis (e.g. Rodrigues et al. 2020). This method provided exceptional results: UV revealed fine-scale details, such as surface folds (Fig. 4F), raised trilete labra (Fig. 5B), and spore wall architecture (Figs. 6,7,9,13,14) in specimens that were opaque under transmitted light. This technique permits identification of opaque megaspores in standard palynological slides, and assessment of the distal and proximal faces by inverting the slide and/or adjusting the focal plane (Fig. 9J,K).

Conclusions
Based on fossil material housed in the collections of the NRM, we described 19 megaspore taxa from Triassic-Jurassic successions of Scandinavia. We recognized three distinct megaspore assemblages from the Rhaetian, Hettangian and Pliensbachian stages, and note a decrease in megaspore diversity across the TJB. Our application of fluorescence microscopy to the analysis of opaque megaspores provided exceptional results, suggesting that this non-destructive method may be useful for the re-assessment of historical collections elsewhere.