Neutron tomography, fluorescence and transmitted light microscopy reveal new insect damage, fungi and plant organ associations in the Late Cretaceous floras of Sweden

ABSTRACT Neutron tomographic reconstructions, macrophotography, transmitted light microscopy and fluorescence microscopy are employed to assess the quality of organic preservation, determine organ associations, identify insect damage, and document fungal interactions with selected Santonian–lower Campanian plant fossils from the northern Kristianstad Basin, southern Sweden. Fricia nathorstii (Conwentz) comb. nov., is proposed for a composite fossil comprising an anatomically preserved (permineralized) cupressacean conifer cone and its subtending, concealed, leafy axis (preserved as a mould) in the Ryedal Sandstone. Several other impressions of conifer and angiosperm leaf-bearing axes and isolated leaves are described under open nomenclature. Three cuticle types are described from the non-marine plant-bearing beds in the basal part of the succession exposed at Åsen, but these are only assigned to informal morphotypes pending a comprehensive review of the extensive fossil cuticle flora. Two species of ascomycote epiphyllous fungi from Åsen are established: Stomiopeltites ivoeensis sp. nov. (Micropeltidales) and Meliolinites scanicus sp. nov. (Meliolales). The latter provides an important calibration point for dating the divergence of Meliolales, being the first pre-Cenozoic representative of the order. Various additional fungal remains, including thyriothecia, scolecospores, chlamydospores, putative germlings, and hyphae, are described from the cuticular surfaces of conifer and angiosperm leaves from Åsen. Insect herbivory is expressed in the form of both margin-feeding and piercing-and-sucking damage on angiosperm leaves. The Santonian–early Campanian vegetation is inferred to have grown in strongly humid, mid-latitude, coastal plain settings based on the depositional context of the assemblages, leaf morphology, and the pervasive distribution of epiphyllous fungi.


Introduction
The southern Swedish province of Skåne (Scania) and portions of adjacent Blekinge are the only regions of the Fennoscandian Shield hosting exposed Upper Cretaceous strata containing plant macrofossils. Fennoscandia was largely an emergent landmass through most of the Mesozoic and Cenozoic but significant coastal and marine deposits accumulated on the southern margin of this region in the Late Cretaceous (Fig.  1A). The plant assemblages preserved in the Swedish deposits represent the northernmost Late Cretaceous floras in Europe (Vakhrameev 1991) and are crucial for reconstructing the palaeovegetation of this region. Cretaceous rocks of southernmost Sweden are preserved in three major geological depressions, the Vomb Trough in the south, the Båstad Basin in the northwest, and the Kristianstad Basin in the northeast (Fig.  1B). The basal claystone-, siltstone-, and sandstone-dominated portion of the Cretaceous succession in the Kristianstad Basin has traditionally been assigned to the Ryedal Sandstone or Holma Sandstone (Holst 1888;De Geer 1889;Magnusson 1958) and is locally rich in plant remains. Relatively few fossil plant taxa have been thoroughly documented from these Cretaceous strata (Table S1). Nevertheless, the Late Cretaceous represents a time of dramatic radiation in angiosperm diversity globally (Lidgard & Crane 1990;McLoughlin et al. 2010;Friis et al. 2011;Halamski et al. 2020), and rocks of southern Sweden potentially host important fossil evidence of the diversification of this group.
Upper Cretaceous plant fossils (then assumed to be Cenozoic in age) were first reported from Köpinge in the Vomb Trough by Nilsson (1824Nilsson ( , 1832. Nathorst (1876) initially reported fossil plants from Upper Cretaceous strata in the northern Kristianstad Basin and Conwentz (1892) subsequently documented a range of woods, cones and fungi preserved as adpressions and siliceous permineralizations from the Ryedal Sandstone and its stratigraphic equivalents. Felix (1894) later described the saprotrophic fungi found in association with the woods reported by Conwentz (1892). Sporadic reports of fossil plants, especially lignitized and permineralized wood and palynomorphs, continued through the early and middle twentieth century (e.g., Linnell 1937;Regnéll 1940;Ross 1949;Nykvist 1957), mostly associated with geological mapping and kaolin quarrying in the region (Table S1).
In the last three decades, work on the Åsen fossil flora has declined as investigations on early angiosperm flowers have largely shifted to older deposits elsewhere in the world (e.g., Eklund et al. 1997;Friis et al. 2011). However, as part of a campaign to document Late Cretaceous adpression floras through central and northern Europe, the Campanian mega-and palynoflora of the Vomb Trough was described recently by . Fossil plant axes were also documented by McLoughlin et al. (2018) from associated marine deposits where they acted as substrates for Late Cretaceous sessile marine invertebrates. Relatively few other records of Cretaceous terrestrial fossils from the Kristianstad Basin have been published over the past 200 years (Table S1), and these are not investigated further in this study.
The present study has four main objectives: 1, to describe the Santonian-lower Campanian impression floras of the Ryedal Sandstone; 2, to re-evaluate the structure and systematics of the permineralized cone and associated axis impression initially assigned to Pinus nathorstii by Conwentz (1892) from this unit; 3, to test the quality of cuticular preservation of angiosperm and conifer leaves from the Åsen deposit for future systematic and palaeoecological studies; and 4, to provide an initial survey of fossil fungi and insect interactions with plant remains from the Kristianstad Basin. In addition to traditional macrophotography and transmitted light microscopy, we employ techniques that have not been applied previously to the Late Cretaceous plants of Sweden, e.g., neutron tomography and incident-light fluorescence microscopy, in order to gain additional anatomical and palaeoecological information from the fossil flora. These aspects, together with data compiled from previous studies, are linked to provide some initial insights into the palaeocology of the Late Cretaceous vegetation of the Kristianstad Basin. A systematic re-evaluation of the Upper Cretaceous silicified woods and their associated fungi and invertebrate borings from southern Sweden is intended for a later study.  2015), modified after data from Chatziemmanouil (1982), Surlyk in Voigt & Wagreich (2008), Halamski (2013) and ]. B. Bedrock geological map of Skåne (Scania) and adjacent regions indicating fossil localities and basins hosting Upper Cretaceous strata (after Christensen 1986;Koistinen et al. 2001;Vajda & Gravesen 2008).

Geological setting
The Kristianstad Basin, represents the onshore extension of the Hanö Bay Basin (Kumpas 1980) and is situated in northeastern Skåne and western Blekinge provinces around the city of Kristianstad in the southernmost part of the Fennoscandian Shield (Fig. 1B). The basin hosts Barremian to Maastrichtian sedimentary rocks overlying a Precambrian granitic-gneissic basement (Norling & Skoglund 1977;Bergström & Sundquist 1978;Kumpas 1980;Norling 1981;Lindström et al. 1991), but most of the exposed strata are dated to the Santonian-Campanian (Christensen 1975). The Cretaceous sedimentary succession reaches 250 m thick in the central Kristianstad Basin (Erlström & Gabrielson 1992) but increases to about 700 m thick in the offshore Hanö Bay Basin (Sopher et al. 2016). The Kristianstad Basin's northern boundary is erosional; numerous outliers of Cretaceous strata occur on adjacent basement rocks. Surficial Quaternary glaciofluvial deposits up to 30 m thick blanket much of the basin. A few natural exposures of Cretaceous strata occur along the shores of Lake Ivö (Lundegren 1934), and artificial exposures up to 20 m thick have been generated by kaolin quarrying, e.g., at Åsen. The oldest Cretaceous plant fossils are palynological assemblages from subsurface Albian strata recovered from a borehole near Österslöv, northern Skåne (Guy-Ohlson 1984; Table S1). A formal lithostratigraphic scheme for the basin has never been established and, traditionally, local geological correlations have employed biostratigraphic subdivisions of the strata (summarized by Einarsson 2018). Accessible strata range from lower or lower middle Santonian (Gonioteuthis westfalica westfalica Zone, sensu Christensen 1997) to uppermost Campanian (Belemnella lanceolata Zone; Thibault et al. 2012;Voigt et al. 2012) and were deposited at palaeolatitudes of 47-49°N (Kent & Irving 2010;Van Hinsbergen et al. 2015).
In the Late Cretaceous, Skåne was located in the border zone between the Fennosarmatian landmass (which, in this region, is essentially equivalent to the exposed portion of the Fennoscandian Shield) and the epeiric Chalk Sea (Fig. 1A). At that time, the margin of the Kristianstad Basin consisted of an archipelago of granitic islands and elongate peninsulas (Surlyk in Voigt & Wagreich 2008). Palaeoshorelines of the northeastern coast of the basin are recorded from the famous locality of Ivö Klack, where basement rocks are encrusted by Cretaceous (Campanian) epizoan brachiopods, mollusks, corals and other invertebrates (c. 200 shell-bearing invertebrate species in total; Surlyk & Christensen 1974;Surlyk & Sørensen 2010;Sørensen et al. 2012). Shoreline and shallow embayment deposits also occur at Åsen, where Campanian oysters are interpreted to have attached to transported woody debris from terrestrial arborescent plants (McLoughlin et al. 2018). The deposits at Ryedal and Åsen yielding plant fossils for this study (Fig. 1B) were laid down in coastal plain settings situated at the Santonian-lower Campanian shoreline or in the immediate hinterland (Surlyk & Sørensen 2010, fig. 3 & 4). Some permineralized axes from Ryedal contain marine molluscan borings (Conwentz 1892) indicative of extended immersion in normal marine waters. Their preservation in uniform-grained quartz sandstones suggests accumulation in strandline deposits after moderate transport via fluvial systems to the coast (Rees 1999). The plant remains at Åsen are well preserved in matted accumulations and probably experienced negligible transport in coastal swamps.

Ryedal
Medium-to coarse-grained quartzose sandstones exposed at Ryedal (56°08ʹN 14°38ʹE) in Blekinge are assigned to the Ryedal Sandstone (Holst 1888), although this unit is probably a lateral equivalent of lithologically similar Holma Sandstone exposed sporadically a short distance to the west in Skåne (De Geer 1889;Magnusson 1958). These deposits have yielded a few plant macrofossils, including permineralized woods and a cone that were documented by Conwentz (1892). In addition to associated fungal fossils (Felix 1894), some of these woods contain traces of invertebrate borers suggesting a foreshore depositional environment. Two conifer twigs and a single angiosperm leaf from this unit are described below. Conwentz (1892) also reported various other indeterminate stems and roots from the deposit at Ryedal, which he considered to be part of the Holma Sandstone. Inspection of several outcrops of the Holma Sandstone around the shores of Lake Ivö by the lead author in 2018 failed to yield any additional identifiable plant remains. The ages of the Ryedal and Holma sandstones are not well resolved but are inferred to be Santonian to lower Campanian based on their stratigraphic position and close association with basal marine deposits in the basin.

Åsen
The kaolin quarry at Åsen (56°9ʹN, 14°30ʹE), formerly belonging to Höganäs AB, is especially famous for mesofossils of charcoalified angiosperm flowers (Friis & Skarby 1981). About 20 m of unconsolidated sands and clays were previously exposed in a quarry that is now largely infilled (Sørensen et al. 2013). The basal plant-bearing non-marine succession is divided into two units by a distinctive weathered horizon. These strata were deposited in a NNE-trending palaeovalley, opening to the south and flanked by Proterozoic igneous rocks (Lundegren 1934;Siverson et al. 2016;see fig. 3B of Surlyk & Sørensen 2010). The lower unit is dominated by finely laminated lacustrine clays, silts and sands. The upper unit consists of cross-bedded or laminated sands, silts, and clays deposited in fluvial settings (Koppelhus & Batten 1989). The material studied herein derives from the lower part of the non-marine succession at Åsen.
The non-marine succession as a whole is assigned to the upper Santonian-lower Campanian on palaeontological and palaeomagnetic data (Mörner 1983;Friis et al. 2011 and references therein), and there are clear floristic differences between the upper and lower parts of the section, but as yet, no finer age controls are available for the two units. Besides plant microspores, pollen and angiosperm mesofossils, quarries at Åsen and nearby Axeltorp yielded conifer (Nykvist 1957) and angiosperm woods (Herendeen 1991), conifer leaves and cones (Srinivasan & Friis 1989), seeds (Kunzmann & Friis 1999), and lycopsid megaspores (Koppelhus & Batten 1989). Overlying the continental deposits at Åsen are marine marls and carbonates of latest early to earliest late Campanian age (Iqbal 2013;Einarsson et al. 2016;Siversson et al. 2016) that have yielded a rich invertebrate and vertebrate fauna (Sørensen et al. 2013;Einarsson 2018;McLoughlin et al. 2018, and references therein).

Material and methods
All specimens examined and illustrated in this study are held in the collections of the Swedish Museum of Natural History, Stockholm (registration numbers prefixed S for plants and NRM Mo for plant impressions on encrusted oyster shells). Some of the impression and permineralized fossils from Ryedal studied by Conwentz (1892) were originally illustrated using lithograph drawings. However, several of these specimens could not be located in the museum collections. The new combination and selected lectotype are registered with unique PFN numbers in the Plant Fossil Names Registry (https:// www.plantfossilnames.org/), hosted by the National Museum, Prague, for the International Organisation of Palaeobotany (IOP). New fungal taxa are registered in the Mycobank database (https://www.mycobank.org/) hosted by the International Mycological Association.

Leaf impressions and compressions
The angiosperm leaf and conifer leafy twigs from Ryedal are preserved as impressions in medium-grained sandstone, so they are described following the procedure advocated for poorly preserved leaf remains (Halamski & Kvaček 2015, pp. 102-103;Halamski et al. 2018, pp. 128-129). In contrast, fossil plants from Åsen are preserved as lignitized remains retaining cuticles, and as charcoalified material. Accordingly, several bulk samples (c. 250 g) from the lower clay bed at Åsen that are rich in cupressacean (=taxodiacean) conifers (Srinivasan & Friis 1989) were tested for leaf and twig extraction via bulk HFmaceration; the recovered cuticles were cleared in Schultze reagent and then immersed in glycerine. Selected cleared fossil leaves were manually split along the margins with a fine needle to separate the abaxial and adaxial cuticles then mounted and sealed on glass slides in glycerine jelly. Individual leaves from the same sample are differentiated by capital letter suffixes affixed to the sample number (e.g., S084386A, S084386B, etc.). Photo-micrographs of plant cuticles and fungi in transmitted white light (brightfield) and incident-ultraviolet light (fluorescence: blue light excitation at c. 460-490 nm) were taken using an Olympus BX51 microscope with an Olympus DP71 digital camera. Whole leaves, cones and axes were photographed with either an Olympus SZX10 stereomicroscope equipped with a Sony Exmoor E3CMOS digital camera or a Canon EOS 40D digital camera. Final images were obtained by merging up to thirty photographs taken at different focal depths using Adobe Photoshop CC and Helicon Focus software using the "autoalign" and "auto-stack" functions.

Permineralized remains
A siliceous permineralized cone (S085156) illustrated by Conwentz (1892) is embedded in a large block of quartzose sandstone. Rather than use destructive thin-sectioning for its anatomical analysis, we trialed non-destructive tomographic reconstruction of this fossil. A fission reactor neutron source was chosen to provide the desired combination of: 1, high penetration of the large (20 × 20 × 10 cm) block; and 2, differentiation of organic (hydrogen-rich) and inorganic (hydrogenpoor) fossil components (Sutton 2008). This study utilized the Open-Pool Australian Lightwater reactor at the Australian Nuclear Science and Technology Organisation (ANSTO), Lucas Heights, New South Wales, Australia. Data collection and reconstruction was conducted using the DINGO imaging and neutron tomography facility at ANSTO in January 2017. Neutron tomographs were reconstructed from a compilation of 1600 projections across a total rotation of 360°, each with an exposure length of 23 seconds. Since DINGO employs a quasiparallel collimated beam, the spatial resolution is determined by the following factors: 1, the collimation ratio (L/D, where L is the neutron aperture-to-sample length and D is the neutron aperture diameter); 2, the thickness of the scintillation screen; and 3, the pixel size of the detector (Garbe et al. 2011). Images were collected on DINGO's high-resolution collimator setting, which has an L/D value of 1000. A 100-μm-thick ZnS/6LiF scintillator screen was employed, with a 200 × 200 mm field-of-view. The detector system was a liquid-cooled, 16-bit Andor IKON-L CCD camera fitted with a 50 mm lens. The resultant pixel width for these projections was approximately 95.5 μm. All additional NT experimental setup details not outlined herein follow Mays et al. (2018). Tomographic reconstruction was performed by CM, J.J. Bevitt, M.-A. Harvey and A. Langendam using filtered backprojection with Octopus Reconstruction v.8.8 (Inside Matters NV). Volume rendering and segmentation were performed by CM using Avizo v.9.5.0 (FEI Company). The specific visualization techniques applied (volume or surface rendering) varied based on the different preservation styles within the same specimen. Specifically, a surface rendering was conducted for the cavity-forming mouldic preservation of the foliage-bearing axis, whereas a volume rendering was produced for the silicified seed cone portion of the fossil (and surrounding matrix). The reasons for the employed visualization techniques are discussed further in the systematic palaeontology section ("Neutron tomographic reconstruction of the cone and axis").

Systematic palaeontology
Both Nathorst (1891) and Schuster (1930) reported examples of the fern Weichselia in isolated sandstone boulders within glacial till located in Germany that were putatively derived from the Ryedal or Holma sandstones of Sweden (see Edwards 1933;Alvin 1971). No equivalent material is available in the collections of the Swedish Museum of Natural History (Stockholm) for evaluation, and these specimens are not discussed further. The following descriptions are restricted to material available in the Stockholm collections.  Harris 1979 Geinitzia sp.

Description
Both specimens are preserved as impressions. S166194 is a leafy twig fragment c. 2.5 cm long. The axis is longitudinally striate, c. 1 mm thick. Leaves are helically arranged, awl-shaped, erect, weakly incurved adaxially, quadrangular in cross-section, 3.5-4.5 mm long, with slightly expanded bases (1 mm thick). S083900 is a leafy twig 18 mm long and 4 mm wide that lacks details of the axis but consists of spirally arranged, awl-shaped, leaf imprints, each c. 1 mm wide and 2.5 mm long.

Remarks
These specimens are attributed to the fossil-genus Geinitzia based on their quadrangular, slightly decurrent leaves. They are broadly similar in size and leaf shape to Geintizia reichenbachii (Geinitz 1842) Hollick & Jeffrey (1909, a widely reported (and possibly heterogeneous) taxon through the Upper Cretaceous of central and western Europe (Bosma et al. 2009;Halamski 2013;Halamski et al. 2018Halamski et al. , 2020Płachno et al. 2018) but the lack of cuticular details or attached cones inhibits further comparisons.

Description
Axis fragments 10 mm wide and up to 90 mm long, bearing rhombic, appressed leaves in tight helices. Leaves are typically 5-7 mm wide (forming a broad rhombic basal cushion) and 4-7 mm long, although bases are commonly concealed by overlapping leaves of the preceding spiral. Leaf apices are obtusely pointed and slightly reflexed.

Remarks
In the absence of cuticular details, these specimens are attributed to Brachyphyllum based on their consistently short, bluntly tapered leaves with lengths being roughly equivalent to their basal widths (Harris 1979). The leaves are less narrowly attenuated than those of Geinitzia sp. described above. A few short-leafed examples of Pagiophyllum sp. from Campanian strata of Köpinge   fig. 3D) approach the shape and size of these Santonian specimens. Although the precise source of the illustrated specimens can not be determined, they likely derive from the basal plant-bearing units of the northern Kristianstad Basin (e.g., Ryedal or Holma sandstones or laterally equivalent units).

Remarks
Various leafless axis impressions and compressions have been reported from the Ryedal Sandstone (Conwentz 1892), the basal carbonaceous succession at Åsen, lignites at nearby Axeltorp (Nykvist 1957), and as impressions on encrusting oysters from the overlying carbonate succession at Åsen (McLoughlin et al. 2018). Some examples bearing dense, helically arranged, leaf scars ( Fig. 3C, E, F) probably represent distal branches of conifers, but they lack the morphological details necessary for identification to any particular family. A few axes have longitudinally ribbed textures and sporadic branch scars ( Fig. 3D) that are not diagnostic for any specific woody plant group. Although the frequency of these woody remains (some with invertebrate borings) is important for ascribing coastal to nearshore depositional environments to the host strata (Nykvist 1957;McLoughlin et al. 2018), they can not be identified with precision and are not described further herein. Several permineralized axes documented by Conwentz (1892) will be redescribed in a separate study.

Remarks
This species was fully described by Srinivasan and Friis (1989), together with several other cupressacean scaleleafed twigs and cones (Table S1). We illustrate three examples of Quasisequoia florinii leaves here simply to highlight the potential use of fluorescent light microscopy of coniferous macrofossils from Åsen to obtain epidermal     fig. 1G).

Plant Fossil Names Registry Number for lectotype designation
PFN002284.

Emended diagnosis
Ovuliferous cone ovoid, borne terminally on leafy twig. Conescale complexes cylindrical to peltate or obovate, arranged helically, relatively massive, with rounded apices. Vascular bundle of cone scale complexes positioned centrally at base and dividing distally into several diffuse veins. Cone scales flattened rhombic in transverse section basally. Distal portions of cone scales bearing a shallow central adaxial depression, rhombic in cross-section with rounded apex. Seeds large, curved, of Seletya type, borne on adaxial side of cone scale. Subtending twig bearing helically arranged imbricate leaves of Pagiophyllum-or Geinitzia-type.

Description of surficial features
The lectotype (S085156) was selected by us from an assortment of other fossils (twigs, wood, leaves) that were considered by Conwentz (1892) to belong to the same taxon. The ovuliferous cone represents a unique permineralized specimen that Conwentz (1892) considered to be attached to the imprint of a thick branch lying adjacent on the rock surface and bearing numerous weakly defined, helically arranged, rhombic leaf scars (Fig. 3C). However, our investigation negated this assumption. Instead, neutron tomographic analysis revealed that the cone is attached to another, completely concealed, twig bearing Pagiophyllum/Geinitzia-like leaves ( Fig. 4A-C). The ovuliferous cone is split longitudinally, and is 36 mm long and 25 mm wide (Fig. 3A). The other half of the fossil cone was apparently never collected. The longitudinal section of the ovuliferous cone exposed on the rock surface reveals a cone axis c. 4 mm in diameter with a distinct dark inner pith and whitish outer woody zone (Fig. 3A). Cone scales emerge at c. 80° to the axis basally, reducing to lesser angles apically. Ovuliferous cone scales are roughly obovate in plan view, reaching 12 mm long and 7 mm wide. Cone scales have a narrow basal attachment (rhombic in cross-section) to the cone axis. They are roughly cylindrical near the cone apex but are clavate to peltate elsewhere, enlarging into a 3.2-mm-thick head distally. The terminus is rounded and rhombic in cross-section (Fig. 4D). There is no obvious differentiation of the cone scale into separate bract and ovuliferous scales, but the vascular trace of some examples appears to split into a lower and upper strand that probably fed the fused bract and ovuliferous scale, respectively ( Fig. 3B: arrowed). At least one slightly curved but otherwise spindleshaped seed is borne on the adaxial surface of each cone scale and is 1 mm thick and 3.2 mm long (Fig. 3A, B).

Neutron tomographic reconstruction of the cone and axis
The fossil specimen was preserved in a medium-grained quartz arenite, and the high silicon and oxygen content of the host rock permitted neutrons to penetrate the large (20 × 20 × 10 cm) block with only minor impedance. However, the size of this block required a large field of view during data collection, thus limiting the spatial resolution for the tomographic reconstruction and providing a relatively coarse pixel size for each projection (c. 96 μm). The neutron tomographic reconstruction revealed substantial variability in preservation both within the specimen and within the matrix. Organic remains were evident by regions of high relative neutron attenuation (RNA), whereas the sedimentary silicate grains, cement and permineralised regions of the fossil all had low attenuation, and pore spaces had negligible attenuation (Fig. 4A). The relatively high RNA in organic remains results from the higher hydrogen density (Sutton 2008), and this can be employed for distinguishing, and "virtually extracting", organically preserved plant remains from a hydrogen-poor silicate matrix (Mays et al. 2017a). Hydrogenous (presumably organic-rich) regions were identified both within the seed cone, and dispersed within the matrix, the latter likely indicating clastic plant debris in the sediment. In the present study, however, some parts of the preserved cone (e.g., at least one of the subsurface cone-scale complexes) lacked enough contrast to be distinguished from the matrix, indicating a nearcomplete replacement of the organic material with silicate minerals. Furthermore, the subtending axis and attached foliage was nearly entirely weathered away, leaving only a cavity in the sedimentary matrix. This variability in preservation precluded a consistent visualisation technique for all components. Firstly, because the mould of the leaves and axis was entirely encased within the sediment, the interface between the matrix and the mould facilitated a strong neutron attenuation contrast, and a high fidelity surface rendering could be produced ( Fig. 4B; Supplementary Video 1). Secondly, the partially preserved organic matter in the seed cone generally provided a good contrast with the siliciclastic matrix, except where it was replaced by silicate minerals. Thus, a volume rendering was produced, illustrating regions of differing neutron attenuation within the seed cone and matrix (Fig. 4C, D); Supplementary Video 2).
The neutron tomographic approach enabled nondestructive study of the unique specimen. This technique revealed anatomical details that either supported our findings from surficial features, or were not observed by other means. Up to 41 individual leaves were identified on the axis bearing the ovuliferous cone (Fig. 4B). These leaves are partly appressed to the axis, and have helical phyllotaxis. The leaves are short (3-4 mm long), 1-2 mm wide and thick, elliptical to rhombic in cross-section, weakly keeled, have entire margins and acute and straight apices. Neutron tomography revealed that the ovuliferous cone preserves remnants of about 20 cone-scale complexes ( Fig. 4C-G). It also confirms that the cone-scale complexes are arranged helically, as clearly demarcated by their well-defined insertion areas in the volume rendering (Fig. 4C, E-G; Supplementary Videos 2 and 3). The volume rendering indicates that the cone-scale complexes are obovate in outline, with pedicellate bases, and peltate heads that are rhombic in distal view (see Fig. 4D, E, H). The cone scales appear to contain several parallel veins or resin canals and one centrally placed vascular bundle (see upper left scale of Fig. 4D).

Remarks
The original material of Pinus nathorstii consists of an ovuliferous cone, various axis impressions and leaves; it is uncertain whether they belong to the same species. The cone (S085156) is designated herein as the lectotype, whereas the leaves and leafless axes (several of which could not be located in the collections) are removed from the species.
Kvaček 2013 described from the Cenomanian of the Bohemian Cretaceous Basin (Czech Republic, Velenovsky 1885; Kvaček 2013). The Ryedal cone is particularly similar to Fricia in the shape of its cone-scale complexes and curved seeds of Seletyatype (Dorofeev 1979 Kunzmann & Friis 1999. It differs from Fricia nobilis in having seeds on the adaxial side of cone-scale complexes and having a rhombic rather than polygonal escutcheon with a shallow depression. Fricia nathorstii has some similarities in general form to the ovuliferous cones described by Kunzmann (1999) as Geinitzia formosa Heer 1871. Fricia nathorstii differs from G. formosa in having cones that are ovoid and less elongate. Cone-scale complexes of F. nathorstii have rhombic termini in contrast to the hexagonal escutcheons typical of G. formosa. Similar ovuliferous cones attributed to G. schlotheimii by Kunzmann et al. (2003) from the Santonian of Aachen (Germany) differ from F. nathorstii in having elongate ovuliferous cones of much smaller size. Geinitzia schlotheimii also differs from F. nathorstii in having much longer leaves on the subtending twigs. Lignitized/charcoalified ovuliferous cones from the Santonian/Campanian of Åsen associated with Quasisequoia florinii Srinivasan and Friis (1989) are significantly different from the Ryedal F. nathorstii specimen in being much smaller and having ovuliferous cone-scales that are markedly peltate with a slightly convex or slightly depressed escutcheon, and in bearing winged seeds. Material S166195 (collected by Ture Hemming in 1905); S166196 is the counterpart but shows almost no venation details; Ryedal Sandstone, Ryedal, Blekinge, Sweden; ?upper Santonian-lower Campanian.

Description
The only available specimen is a relatively large (preserved length 95 mm, preserved width 70 mm; estimated length c. 120 mm, estimated width c. 100 mm) leaf fragment lacking base, apex, or margin. Moderate venation details are retained on the impression in medium-grained sandstone. The venation pattern may have been palmately pinnate or pinnate. The midvein follows a gently zigzagged course with a change of direction at each departure of a secondary vein, the changes becoming more pronounced distally. The midvein bifurcates near the eroded apical margin. Secondaries (three preserved with departure and two without) emerge at an acute angle (30-50°) from the midvein. Tertiary veins are percurrent and V-shaped between secondary veins.

Remarks
This venation pattern is characteristic of the platanoid form group. The absence of the base, apex, and margin precludes any more resolved identification, which should be based, in particular, on the basal vs suprabasal departure of the first pair of secondaries. However, the zigzagged midvein is a notable feature distinguishing the Ryedal platanoid leaf from other Late Cretaceous members of the form group, such as Ettingshausenia sp. from the Campanian of southern Scania ), E. lublinensis from the Campanian of southern Poland (Halamski 2013), or E. onomasta from the Coniacian of the Sudetes, Czech Republic .

LEAVES WITH CUTICLE FORM GROUP UNRESOLVED
Here we illustrate the cuticles of three selected leaf forms preserved as lignitized compressions from the well-known Åsen (Santonian-lower Campanian) plant fossil assemblage. Previous studies have documented a range of angiosperm reproductive structures, palynomorphs, and lignitized wood from this deposit (Table S1). Our intention in the descriptions below is not to provide a systematic account of the extensive Åsen angiosperm leaf flora, but to illustrate the preservational quality of the fossil foliage and to highlight the potential of this assemblage for investigations of Santonian-Campanian angiosperm diversity, plant-insect interactions, and palaeoenvironmental analysis.

Description
No complete leaf available. Venation with a single straight midvein and two thinner lateral veins (Fig. 5B). Distal parts (Fig. 5A) linear, 4-9 times longer than wide; apex probably emarginate. Margin with irregularly spaced (

Remarks
Anomocytic stomata are the most common type of angiosperm stomata; they are present in basal angiosperms (Nymphaeaceae), basal eudicots (Ranunculaceae, Lardizabalaceae), but also in more derived groups, such as Aceraceae, Rosaceae and even Campanulaceae (143 families in total; Metcalfe & Chalk 1950, pp. 1331-1332. However, combined with the presence of glandular cusps, a chloranthaceous affinity is suggested for this leaf. Glandular teeth are known from various extant representatives of Chloranthaceae (Todzia & Keating 1991), although in most cases they are more densely spaced than in the Åsen fossil. The presence of chloranthaceous leaves in the Åsen deposits comports with the mesofossil assemblages, which have yielded floral structures of this group (Crane et al. 1989;Eklund et al. 1997). The quality of the cuticle morphology is outstanding. If calibrated with its nearest living relative, the stomatal indices of this chloranthaceous leaf could prove useful in future palaeoclimatic analyses as a proxy for pCO 2 in the late Mesozoic (compare with, e.g., Steinthorsdottir & Vajda 2015;Steinthorsdottir et al. 2016). Moreover, cuticle of this type is host to a range of epiphyllous fungi ( Fig. 5G; and descriptions below) that attest to complex plant-fungal interactions in the Late Cretaceous and growth in a humid climate.

Remarks
We tentatively suggest a platanaceous affinity for this cuticle type. The occurrence of this group is supported by the presence of fossil reproductive structures of platanaceous affinity from the same deposits (Friis et al. 1988). Similar material was described by Golovneva (2011) as Ettingshausenia cuneifolia from the Cenomanian of Siberia. The cuticle type is also extensively overgrown by epiphyllous fungal hyphae, suggesting growth in a moist habitat.

Description
The available material consists of two cuticle fragments (the larger is 13 mm long and 4 mm wide) with subparallel sides (Fig. 6D); hence, it is inferred that leaves were originally linear. Both available cuticle fragments bear stomata; they may represent either abaxial cuticles or the leaves may have been amphistomatic. Cells subrectangular, arranged in rows parallel to the leaf margins, 50-100 µm long, c. 50 µm wide in the median region, 20-25 µm wide near the margins; pairs of longitudinally arranged cells represent sister cells having originated via bipartition of a mother cell. Trichomes and secretory structures absent. Stomata rare, irregularly disposed, cyclocytic (Fig. 6E, F), with 10-12 surrounding cells, 40-75 µm long and 30-50 µm wide, greatly to slightly longer than wide, rounded at poles, aperture c. 20 µm long and 15-20 µm wide. Weak striae are locally preserved on cells (Fig. 6F).

Remarks
Monocots have a fossil record extending back to the Early Cretaceous (Friis et al. 2004(Friis et al. , 2011Doyle et al. 2008;Coiffard et al. 2013). Although their fossil record is patchy through the Cretaceous, this group clearly diversified and increased in abundance through the latter part of the period (e.g., Kvaček & Herman 2004), such that monocots became locally dominant components of some vegetation types by the Maastrichtian (Upchurch 1995;Herman & Kvaček 2010). Pole (2007) outlined some of the problems associated with confidently identifying isolated fossil monocot cuticle fragments and assigning them to constituent families. Generally, monocot cuticle preserves longitudinally oriented files of rectangular epidermal cells and stomatal complexes, typically demarcated into long costal and intercostal bands. The stomatal complexes  group, especially within Orchidaceae (Stern 2014

Etymology
After Lake Ivö (Ivösjön), located adjacent to the fossil locality.

Description
Epiphyllous fungus consisting of circular discoid to low domal plectenchymatous thyriothecium, 80-200 μm in diameter, with a 5-20 μm central ostiole and weakly thickened collar ( Fig. 7A-F). Margin of thyriothecium relatively smooth and entire or with sporadic radial clefts (Fig. 7A, C). Plectenchymatous wall consisting of densely and irregularly interwoven hyphae c. 3-4 μm wide, becoming concentrically arranged at the thyriothecium margin (Fig. 7B, F). In only a few cases, tightly contorted septate hyphae, 3-11 μm wide, emerge from the margin of the thyriothecium and extend irregularly across the host cuticle surface (Fig. 7D), giving off sporadic simple short hyphopodia. The thyriothecia occur on various angiosperm cuticle types either in isolation or in relatively dense clusters (Fig. 7A). They are especially common on the stomatiferous (presumably abaxial) surfaces of chloranthaceous leaves (Fig. 7E) and are positioned over stomata ( Fig. 7A-C). They rarely occur over major veins. Samarakoon et al. (2019) noted that, although the fossil record of shield-like epiphyllous fungi is quite extensive, many are described from incomplete specimens that lack clear characters resulting in morphological confusion and uncertain systematic placements. The small circular ostiole, tangled hyphae constituting the thyriothecium, and the relatively sharp boundary of the latter are features consistent with Micropeltidaceae (Zeng et al. 2019). The history of higher classification of Micropeltidaceae and segregation of its constituent genera is complex, as outlined by Phipps and Rember (2004). Classification below family level typically requires details of the ascospores, which are not available for the specimens described here. Nevertheless, the small ostiole, complex intertwined hyphae, and sharply defined margins with radial clefts is consistent with the features of Stomiopeltites (Alvin & Muir 1970). Although Alvin and Muir (1970) claimed that this taxon lacked hyphopodia, their illustrations do not appear to capture the full morphology of their studied fossils, and Phipps and Rember (2004, fig. 9) illustrated similar forms with adventitious hyphae bearing sporadic lateral extensions that probably represent hyphopodia. The type species, Stomiopeltites cretacea Alvin & Muir 1970, from the Wealden of the Isle of Wight, differs from the new species by its narrower hyphae (1.7-3 μm wide) and slighly larger thyrothecia (up to 250 μm wide). Stomiopeltites amorphos Phipps & Rember 2004, from the Miocene of Idaho, differs in its more prominently thickened collar around the ostiole and generally less tightly interwoven hyphae. The shield-like Stomiopeltites fossils and other forms described below can be separated from most commonly reported examples of fossil epiphyllous fungi affiliated with Microtheriaceae by the latter having much more regimented and regularly septate rays of hyphae constituting the thyriothecium (Dilcher 1965;Wu et al. 2011;Du et al. 2012;Worobiec & Worobiec 2013).

Description
Epiphyllous fungi consisting of prominently domed (Fig.  7J) plectenchymatous thyriothecia, 160-450 μm in diameter ( Fig. 7G-M), each with an apparent central cap or operculum (Fig. 7K) c. one-third of the thyriothecium diameter (in most cases detached: Fig. 7L), and with an 8-μm-wide central ostiole. Margin of thyriothecium ragged with irregular short (up to 30 μm) hyphal extensions and sporadic radial clefts (Fig. 7L, M). Plectenchymatous wall consisting of dense, contorted and partially interwoven, c. 3-4 μm wide, hyphae that generally become radially arrayed and laterally fused towards the margin forming a flat stellate skirt (Fig. 7I, K-M). Plectenchyma strongly thickened around outer margin of operculum and inner part of the outer thyriothecium. Localized thinning of the plectenchyma apparently provides a weakness in the thyriothecium wall for the detachment of the operculum (Fig. 7K), the apical part of the thyriothecium also opening by starshaped splits after operculum detachment (Fig. 7J). The thyriothecia occur on various angiosperm and conifer cuticle types, typically in isolation or weakly aggregated (Fig. 7G, H). They occur on both stomatiferous and nonstomatiferous leaf surfaces but appear to be significantly more common on the latter.

Remarks
We did not detect any adventitious hyphae emerging from the thyriothecia but further searches of the rich Åsen fossil assemblage are likely to yield additional details of the morphology of this fungus. In its generally ragged fringe, radial arrangement of slender hyphae around the margin of the thyriothecium, and stellate apical opening, this fossil form is similar to various modern Asterinaceae taxa, including Halbania cyathearum, and to Petrakina mirabilis, an ambiguously placed dothideomycete (see Hongsanan et al. 2014, figs 15d, 45f), but the extant examples lack a line of thinning to produce a broad, operculum-like, apex. Similar ring-like thinnings or ruptures are evident around the apices of some microthyrialean thyriothecia, such as Dictyopeltis applanata (see Gallo et al. 2018, fig. 2E), but such examples tend to lack the more markedly radial arrangement of hyphae. The Åsen form is also similar to some Cenozoic fossils, e.g., Asterilla kosciuskensis Selkirk 1975, in its fine radial hyphae, but the presence of an apical, ring-like, thinning or stellate clefts is distinctive to the Swedish taxon. Based on the dominant characters, we tentatively assign these remains to Asterinales; confirmation and further taxonomic resolution will require additional details of the hyphae, ascocarps and ascospores.

Etymology
After the province of Skåne (Scania), southern Sweden.

Remarks
The arcuate contortions evident in the hyphae near some septa are superficially reminiscent of clamp connections among Basidiomycota but instead appear to be short, lobate appressoria (hyphopodia) of an ascomycote fungus. Similar short curved cells were illustrated on hyphae marginal to ascomycete thyriothecia by Phipps and Rember (2004, fig. 19). This taxon is similar, in the general form of its perithecia, Y-branched and cross-connected adventitious hyphae with apparently short hyphopodia, to a range of species attributed to Meliolinites (Meliolales). Meliolales has a sparse fossil record with occurrences scattered through the Cenozoic (Köck 1939;Dilcher 1965;Selkirk 1975;Mandal et al. 2011;Taylor et al. 2015;Wang et al. 2017). The present record from Santonian-Campanian strata at Åsen supplies a new calibration point for dating the evolutionary diversification of this order that is consistent with the estimated divergence of crown group Meliolales at 177-93 Ma (late Early Jurassic -mid-Cretaceous) based on molecular data (Hongsanan et al. 2016).
At least ten species of this fossil genus have been established but, in many cases, the morphological features distinguishing these taxa are very subtle. In most cases, it is the characters of the spores, appressoria, and other hyphal extensions (e.g., mycelial setae) that are used to distinguish the fossil species, but these are either absent from the Åsen specimens or (with respect to appressoria) simple and ill-preserved. Based on the relatively simple appressoria and poorly ordered perithecia, Meliolinites scanicus appears to be a more archaic form within the genus. In general, other species of the genus, e.g., Meliolinites spinksii (Dilcher) Selkirk 1975, M. nivalis Selkirk 1975, M. dilcheri Daghlian 1978

Description
Thickened, translucent, stellate epicuticular structures occurring on both stomatiferous and non-stomatiferous surfaces of angiosperm leaves and centred on stomata or epidermal cell junctions (Fig. 8G-I). Larger examples are roughly circular (up to 200 μm in diameter: Fig. 8J), but smaller specimens form irregularly stellate thickenings (c. 20-40 μm in diameter: Fig.  8H) at epidermal cell junctions. These features lack clearly defined internal structures or adventitious hyphae but the thickenings show radiating extensions along epidermal cell boundaries (Fig. 8H, I).

Remarks
Various dothideomycete fungal germlings (e.g., Bianchinotti et al. 2020, pl. 1, fig. 1-3) can have a stellate appearance superficially similar to these fossils. Such germlings may become established at weaknesses in the leaf cuticle substrate -either at epidermal cell junctions, loci of physical damage, or around stomata. In some respects, the Åsen examples are similar to diminutive examples of insect mucivory (piercing-and sucking-) damage (see examples described below) but they tend to occur away from veins, are strongly variable in size, and not all form around ruptures in the cuticle. The identity of these features on the Åsen angiosperm leaves is far from clear but their abundance, variable size and stellate form leads us to provisionally consider them to be fungal germlings. (Figs. 8L-O, 9A, B) Material S084294-04, S084294-05, S084298-06, S084299-01, S084386-03, S084413-01, S084413-02, S084413-05, S084413-05, S084468-12 Description Weakly (Fig. 8N) to strongly (Fig. 8M) sinuous hyphae with few or rare septa, variable branching patterns, and simple rounded termini. Hyphae are smooth, typically 5-8 μm wide, hundreds of micrometres long, with walls 1-2 μm thick, and lacking obvious hyphopodia, reproductive structures or other adornments. These hyphae typically skirt stomata (Figs. 8N, O, 9B) and are generally arranged irregularly across the leaf surface. In a few cases, they appear to penetrate stomata (Fig. 9A). Where hyphae are very abundant, they commonly aggregate in bands and follow epidermal cell walls, especially along veins (Fig. 8L).

Remarks
A very large number of relatively featureless fungal hyphae unassociated with reproductive structures occur on both angiosperm and conifer leaves in the collection. Owing to the dearth of morphological characters and lack of attached reproductive bodies, the affinities of these hyphae are unclear. Some may be affiliated with the various epiphyllous fungi described above, since similar crenulate hyphae are known to be associated with a wide range of micropeltidacean thyriothecia (Maslova et al. 2020). Others may represent saprotrophic fungi that grew over the dead leaf surface before burial. We illustrate a range of examples (Figs. 8L-O, 9A-B) to highlight the potential for additional discoveries of fossil (Santonian-Campanian) mycobiota in the Kristianstad Basin.

Description
Featureless, gently curved, aseptate hyphae with simple blunt termini in tangled masses on degraded cuticle.

Remarks
One example of this hyphal type is available. It lacks the marked crenulations or sinuosity of the isolated hyphae described above. Septa are not obvious but overlapping of the hyphae and the degraded nature of the attached angiosperm cuticle make these difficult to detect. In the absence of attached reproductive features we can not assign these remains to any systematic or palaeoecological grouping of fungi.

Remarks
Although broadly similar simple, laevigate, spherical-globose reproductive structures are produced by various fungi, including Chytridiomycota (Krings et al. 2009), zygomycetes (Krings & Taylor 2012) and Glomeromycota (Walker 1983), we regard these remains as probable ascomycote chlamydospores. These reproductive structures are borne on hyphae that resemble some examples of those described above as "sinuous hyphae", although they are generally less contorted. The putative chlamydospores are relatively featureless and lack preserved contents. We draw attention to the similarity of these structures to thick-walled chlamydospores borne on short lateral hyphal branches of modern ascomycotes, such as Fusarium (Pérez-Vicente et al. 2014) and Verticillium (Grum-Grzhimaylo et al. 2016). The lack of any additional distinctive morphological characters detracts from assignment to any particular group of Ascomycota. Chlamydospores are asexual reproductive structures that tend to be produced during unfavourable environmental conditions (e.g., excessive heat or drought) as a resting stage in the fungal life cycle (Lin & Heitman 2005).

Description
Scolecospores 60-163 μm long, 8-18 μm in maximum width, spindle-shaped to linear with tapered base. The spores are typically attached to an indistinct hypha c. 4 μm in diameter on the cuticle surface by a short, tapered, strongly translucent or darkened pedicel (Fig. 9I, J). Scolecospores divided by transverse septa into 8-36 cells (Fig. 9G, H). The apex is generally tapered but, in rare cases, may be capped by an enlarged rounded cell (Fig. 9L). Where attached to cuticle, the scolecospores are typically positioned over or adjacent to stomatal apertures. They occur scattered over the stomatiferous surfaces of angiosperm leaves and, in a few cases, occur in dense aggregations (Fig. 9K).

Remarks
Scolecospores are common among representatives of extant Phyllachoraceae (Ascomycota, Sordariomycetes), which are obligate parasites on living plant hosts. Isolated scolecospores may vary considerably within an individual population in size, gross shape, terminus shape and number of septa ( Fig. 9G-L).
Owing to the dearth of other morphological characters, they are difficult to assign to any species with consistency. The Åsen examples are especially similar to septate ascospores of Scolecopeltidium hormosporum Stevens & Manter 1925(Wu & Hyde 2013. Similar scolecospores (filamentous phragmospores) illustrated as "Scolecospore Fungal multicellate spore" have also been documented from the Bathonian-Tithonian of Libya (Thusu & Vigran 1985), and a range of other comparable spindle-shaped septate spores have been recorded from Upper Cretaceous and Paleogene strata globally (Kalgutkar & Braman 2008;Saxena & Tripathi 2011). These remains are also superficially similar to some examples of isolated elongate dothideomycete conidia (Hyde et al. 2017, figs. 12j-m, 33i-q;Crous et al. 2007, fig. 6B-H), but the Åsen examples are never aggregated into sporodochia or disarticulating chains. Kalgutkar and Jansonius (2000) summarized the few species formally established within Scolecosporites. Of these, the Åsen specimens appear to be most similar to Scolecosporites scalaris (Kalgutkar) Kalgutkar & Jansonius 2000 and S. modicus Kalgutkar & Jansonius 2000 in their gross shape, size and degree of septation. However, the Swedish population seems to encompass the full range of characters expressed by all four fossil species recognized by Kalgutkar and Jansonius (2000).

Description
One chloranthaceous leaf from Åsen (leaf morphotype 1) bears at least two examples of margin feeding. The feeding traces (Fig. 5B) extend c. 4 mm along the leaf margin and penetrate 2 mm into the lamina (i.e., reaching but not transecting the midvein). The traces are represented by roughly semicircular scallops, although the lower example in Fig. 5B appears to show a secondary scallop that extends from the initial damage area. The damaged areas are flanked by a reaction rim consisting of thickened tissue, 100-200 µm wide, with an increased density of fungal hyphae (Fig. 9M).

Remarks
This style of leaf-margin feeding is broadly similar to the damage category DT14 illustrated by Labandeira et al. (2007). Because of extensive convergence in mouthpart architecture and foraging behaviour in leaf-margin-feeding insects, only rare cases of this damage style can be attributed to specific animals. Diverse larval and adult insects, especially among Coleoptera (beetles), Orthoptera (grasshoppers and their relatives), Lepidoptera (moths and butterflies), Phasmatodea (phasmids), and Hymenoptera (ants, bees and their relatives), are known to produce simple semicircular scallops on leaf margins (e.g., Carvalho et al. 2014;Sohn et al. 2017) similar to the examples illustrated here. It is notable that the scalloped damage occurs along parts of the lamina margin between the blunt glandular cusps that are characteristic of this leaf form. Trichomes of various morphologies, including glandular hairs and secretory cells are common in a broad range of flowering plants, including some of the earliest diverging angiosperm clades, and typically represent structures producing chemical defenses against herbivory (Fahn 1979;Agrawal & Fishbein 2006;Chin et al. 2013). However, some glandular cusps (in the form of extrafloral nectaries) produce insect attractants (Elias 1983) and others aid the reduction of water loss via cuticular transpiration (Gonzalez & Tarragó 2009). In extant Chloranthaceae, glandular cusps, in the form of hydathodal glands, are known to aid water regulation via guttation (Todzia & Keating 1991;Feild et al. 2005;Feild & Arens 2007). Marginal cusps also provide physical obstacles that are disruptive to regular margin feeding by insects (Brown et al. 1991;McLoughlin et al. 2015) and the apically orientated teeth of angiosperm leaf type 1 may have directed herbivore traffic distally, and ultimately off the end of the leaf (Vermeij 2015). It is likely that the Åsen chloranthaceous leaf was employing glandular cusps as both physical anti-herbivory and water-regulatory devices, and that the insect was actively avoiding marginal teeth in an ongoing "arms-race" of herbivory versus plant defence at a time of rapid diversification of both angiosperms (Magallón et al. 2019) and at least some groups of insects (Condamine et al. 2016). The record of chemical defences in fossil plants is scant but does extend back to the late Palaeozoic (Krings et al. 2002). The preservation of glandular structures on Cretaceous leaves at Åsen offers one line of investigation for tracking the development of induced (chemical) defence mechanisms against herbivores in early angiosperms.

Description
Prominent circular to elliptical openings in cuticle, typically c. 40-80 μm wide and 70-120 μm long, surrounded by a zone (c. 30 μm wide) of thickened (darkened) cuticle or necrotic tissue (Fig. 9N). Damage features are typically positioned on veins and, in some cases, several damage scars are arrayed in a row at least 300 μm apart (Fig. 9O).

Remarks
Most previous records of piercing-and-sucking damage attributable to a specific group relate to shield scars left by scale insects (Labandeira et al. 2007;Wappler & Ben-Dov 2008;Wilf et al. 2017). Only a small number of puncture damage features on leaves have been illustrated at high resolution from fossil cuticles or mummified leaves, and some of these occur aligned in rows above thickened veins or midribs (Tosolini & Pole 2010;Labandeira et al. 2014). The relative scarcity of clearly defined fossil records attributable to this category of herbivory probably relates to the diminutive size of the damage marks and the difficulty to differentiate these from other forms of fungal, bacterial or physical damage on carbonized leaf compressions or impressions. Nevertheless, mucivory has a patchy fossil record extending back to at least the Early Devonian and is among the oldest styles of herbivory documented in the fossil record (Labandeira 2013).

Discussion and conclusions
The Santonian-early Campanian floras of the northern Kristianstad Basin are apparently the northernmost plant fossil assemblages of this age from Europe. Future studies of the Swedish assemblages could potentially provide insights into a range of palaeobotanical and palaeoecological questions. These include: 1, What was the full floristic diversity and composition of the Åsen flora?; 2, What were its palaeophytogeographic relationships with Central European (warmtemperate) and Siberian (cool-climate) floras?; 3, What anatomical information can be gained from these plants using advanced cuticular analysis, palynology, thin-sectioning, and tomographic approaches to aid whole-plant reconstructions?; 4, Can palynostratigraphy of the host strata provide better age resolution of these units?; 5, Can phylogenetic analysis of the wealth of plant fossils from the Åsen deposit provide a better understanding of the changes in diversity and abundance of major conifer and angiosperm clades at a time when flowering plants were undergoing an explosive radiation, globally?; and 6, Can the Kristianstad Basin Cretaceous floras provide new calibration points for clade divergence in the context of biome re-structuring after the rise of angiosperms to dominance (e.g., Schneider et al. 2004;Le Renard et al. 2020)?

Palaeofloral diversity
Our reconnaissance sampling of the Åsen fossil leaf assemblage (Figs. 5,6), and initial scanning of rock sample surfaces using fluorescence microscopy ( Fig. 2I-N) suggests that the palaeoflora is extremely well preserved and quite diverse. The extensive bulk samples already registered in the museum collections host leaf cuticles amenable to study by transmitted light and fluorescence microscopy, and preserve a diverse array of epidermal and cuticular ornamentation (Figs. 2L-N, 5B-G, 6A-F). To date, described angiosperm remains from these beds are limited to reproductive structures of chloranthaceous, platanaceous, saxifragalean, ericalean or ebenalean, hamamelidacean and fagalean (Normapolles complex) plants, together with mention of undescribed material of thealean affinity (Eklund et al. 1997; Table S1). The few palynological studies of this deposit have documented a considerably greater diversity of some groups (e.g., fern spores and angiosperm pollen: Table S1) than represented, thus far, by mesofossils. We suggest that a dedicated survey of fossil angiosperm leaf cuticles from this deposit would be highly productive and should greatly advance broader phytogeographic reconstructions of the European Late Cretaceous. Moreover, recovery of data on the types of seeds and other disseminules in these deposits will have implications for understanding the roles of plant dispersal mechanisms in the Cretaceous vegetation (McLoughlin & Pott 2019), and potentially provide insights into herbivory that may supplement the meagre terrestrial vertebrate fossil record from the Kristianstad Basin (Table S1). The Ryedal and Holma sandstones host a low-diversity flora of conifer and angiosperm remains. The few fossils available from these units suggest that the flora is, nevertheless, similar in composition to other Late Cretaceous assemblages from central Europe. In general terms, the palaeocommunities best represented in the megafossil record of Central Europe are riparian forests with platanoids and floodbasin conifer forests (e.g., Kvaček et al. 2015;Halamski et al. 2020;Heřmanová et al. 2020). Presumably, the single platanoid leaf and the two Geinitzia twigs identified in this study derive from similar riparian and floodbasin forests of coastal plains flanking the Kristianstad Basin. In that respect, the megafossil assemblages from northern Skåne described herein differ from the Campanian assemblage of Köpinge, southern Skåne , which is dominated by Dewalquea haldemiana, a species with coriaceous leaves tentatively interpreted as a dune-dweller (Halamski et al. 2020).
The Ryedal and Holma sandstones, and the laterally equivalent non-marine organic-rich deposits at Åsen and Axeltorp, are poorly exposed and have discontinuous distributions along the northern margin of the Kristianstad Basin. Future significant fossil discoveries from these units are likely to become available only through quarrying for clay, sand or sandstone resources. Nevertheless, rare permineralized remains of plants recovered from these units offer significant potential for insights into Late Cretaceous plant anatomy and interactions with fungi.

Palaeobotanical applications of neutron tomography
We have shown that neutron tomography (NT) has great potential for recovering anatomical data for the reconstruction of permineralized cones preserved in coarse-grained siliceous facies, such as the Ryedal Sandstone. Neutrons provide a stronger attenuation contrast between organic and inorganic components than X-ray techniques (Dawson et al. 2014). The fission neutron source utilized herein has a high neutron flux (Garbe et al. 2011), thus providing excellent penetration through voluminous siliceous sedimentary rock. Hence, the high-flux neutron tomography technique is particularly promising for the analysis of large permineralized plant remains in general, and particularly permineralized peats (Slater et al. 2015) or root mantles (McLoughlin & Bomfleur 2016) that encompass a diverse array of organic remains and plantanimal-fungal interactions.
Neutron tomography enabled linkages between the cone and the attached leafy axis mould entombed within the sedimentary rock matrix. Although neutron reconstruction of mouldic plant fossils has been conducted by Dawson et al. (2014), here we demonstrate the utility of NT for the virtual extraction of plant organs of different preservation within the same specimen: the mouldic leafy axis and the permineralized ovulate cone. It is common to find plant fossils with differential preservation within a single specimen, which likely reflects underlying anatomy, e.g., fleshy vs woody organs (this study), lignitized wood vs resin (Mays et al. 2017b(Mays et al. , 2018, vascular vs ground tissue (Herrera et al. 2020). Despite the relatively course spatial resolution, the ability of NT to differentiate a wide range of preservation styles may be a critical consideration for future fossil visualization studies. Conwentz (1892) illustrated, using lithographic sketches, various, apparently saprotrophic, fungi associated with woods from the Holma Sandstone. Soon after, Felix (1894) provided a short description (without illustration) of one taxon of probable saprotrophic ascomycete from this collection. These woods and contained fungi will be re-analysed in a forthcoming study.

Fossil fungi
Numerous studies have documented fossil epiphyllous fungi from various deposits around the world. However, there have been few studies devoted to a thorough systematic evaluation of the palaeomycoflora from any one succession, or an evaluation of the stratigraphic or geographic distributions of fungal taxa. Our reconnaissance survey of angiosperm and conifer cuticles from the Åsen deposit indicates that fungal remains are ubiquitous on these plant remains. This suggests that epiphyllous fungi are likely to be commonplace in wetland deposits, especially of late Mesozoic and Cenozoic age. Our survey identified the first occurrence of putative Meliolales in the Cretaceous. Thorough surveys of the fossil mycofloras from such deposits offer the potential for acquiring important temporal calibration points for fungal phylogenies, for documenting the development of plant-fungal interactions through deep time, and for understanding the evolution of Earth's mycofloral diversity and turnover in general. For example, Trichopeltinites, considered to have gone extinct in North America at the K-Pg boundary, has been shown to have survived in the Southern Hemisphere based on the study of a cuticle assemblage (Upchurch & Askin 1989).

Palaeoecology and palaeoenvironments
Epiphyllous fungi may be saprotrophic, obtaining nutrients on the leaf surface from the decay of material in the surrounding forest or from leaf exudates (Cooke & Rayner 1984). Others are biotrophic (parasitic on living hosts), some obligate to specific plant taxa, obtaining nutients via haustoria that penetrate the cells of the leaf substrate (Bannister et al. 2016;Suzuki & Sasaki 2019). High diversities and abundances of epiphyllous fungi are generally signals of warm, ever-wet climates (Wang 1991;Bannister et al. 2016) but higher latitude settings may also host similar fungal biotas under consistently humid conditions (e.g., McLoughlin et al. 2002;Lücking et al. 2009).
The presence of over a dozen taxa of fern, lycopsid and bryophyte spores in the Campanian palynofloras of the Kritianstad Basin (Table S1) and 18 taxa of these groups from the Vomb Trough , together with abundant and diverse Tetraporina (Lindgren 1980), a zygnematacean or sphaeroplealean freshwater alga (Mays et al. 2021), also suggest that a humid climate prevailed across southern Sweden in the Late Cretaceous. Growth features of fossil woods from the Kristianstad Basin have not yet been analysed for palaeoclimatic signals apart from the illustration by Conwentz (1892, pl. 8, fig. 2) of indistinct growth rings in Sequoites holstii wood from the Holma Sandstone that suggest relatively subdued seasonality.
Another signal of humid conditions is the glandular cusps of the chloranthaceous leaves. If the glandular teeth of the chloranthaceous angiosperm leaf type 1 (Figs. 2M, 5C, F) are primarily hydathodal, then this may have been an adaptation for removal of excess water from the leaf in a consistently moist mid-storey or understorey environment where internal flooding of mesophyll intercellular spaces may otherwise have reduced CO 2 diffusion and inhibited photosynthesis (Feild et al. 2005(Feild et al. , 2009). Our study shows that a moderate diversity and great abundance of epiphyllous parasitic Ascomycota, and potentially some generalist saprotrophs were present at Åsen in this humid-climate setting. Chloranthaceae are also known to produce chemicals inhibitive of germination and mycelial growth of fungal (e.g., Botrytis) pathogens (Jacometti et al. 2010). With further study, the interactions between chloranthaceous leaves and fungi may provide insights into the parasite and host-defence mechanisms during the early stages of angiosperm diversification. The generally fine preservation of fungal remains and their great abundance in the Åsen deposit also offers opportunities for expanding the fossil record of other fungal and fungi-like groups, such as Basidiomycota, Chytridiomycota, Glomeromycota, and Peronosporomycetes.
Late Cretaceous plant assemblages belong to the modern, angiosperm-dominated flora (Cenophytic Evolutionary Flora sensu Cleal & Cascales-Miñana 2014). Similarly, the Late Cretaceous entomofauna is also considered "modern" (Grimaldi & Engel 2005). In other words, the broad-scale taxonomic composition of the insect fauna is similar to that of the extant fauna (Szwedo & Nel 2015). Moreover, the ecological relationships between plants and insects were of an equivalent nature and complexity as in the modern biosphere (Labandeira 2006). This broad similarity is related to both insects and plants having suffered from the K-Pg event at lower taxonomic levels, but their diversity being significantly less affected at higher taxonomic levels (Labandeira et al. 2002;Nichols & Johnson 2008), a response markedly different from that of vertebrates (e.g., Kielan-Jaworowska et al. 2004). The insect fossil record is generally sparse and unequal in space and time. Compared to better-studied Early Cretaceous-Cenomanian insects, Turonian-Maastrichtian entomofaunas are less well known owing to the scarcity of exposed strata bearing rich fossil insect assemblages (Szwedo & Nel 2015). In the absence of insect macrofossils, feeding damage on foliage can provide some insights into the insect herbivores and herbivory strategies of Late Cretaceous.
External foliage feeding and piercing-and-sucking represent the oldest kinds of insect-plant interactions dating back to at least the Devonian (Labandeira 2006 and references therein). In the modern fauna, piercing-and-sucking is especially characteristic of true bugs (Hemiptera: Yoshizawa & Lienhard 2016), although similar feeding habits also occur in thrips (Thysanoptera) and spider mites (Tetranychidae, Acari). The modern fauna probably contains several tens of thousands of plant-feeding hemipteran species and the Cretaceous was a time of strong taxonomic turnover and extensive radiations within this group (Szwedo & Nel 2015 and references therein). Although we can not confirm that the mucivory damage was caused by hemipterans, they are strong candidates for this form of herbivory in the Åsen flora. Similarly, the margin-feeding damage on chloranthaceous leaves (Figs. 5B, 9M) can not be attributed definitively to any one group but coleopterans, orthopterans, lepidopterans, Phasma-todea, and possibly even Hymenoptera, are all potential candidates for this type of herbivory.
A recent study that proposed a detailed vegetation reconstruction of a Central European Late Cretaceous flora was achieved only through the assessment of mega-, meso-, and microfossil records (Halamski et al. 2020). Similarly, our preliminary analysis shows that applying multiple analytical approaches to the study of Santonian-Lower Campanian plant remains from the Kristianstad Basin yields information on plant organ associations, palaeophytodiversity and palaeoecology that can not be obtained from any single investigative method. We note that the Åsen flora, in particular, represents one of the largest resources of Santonian-Campanian fossil plants in northern Europe and has great potential to expand upon the known Late Cretaceous terrestrial biota of Sweden (Table S1). We urge a major investigation of the Kristianstad Basin plant fossil assemblages (cuticular, lignitized, charcoalified, permineralized and palynofloral remains) using a broad battery of methodologies to fully document the extensive upper Santonian-lower Campanian floras and place them within a more robust palaebiogeographic, phylogenetic, and palaeoecological context.