Komlopteris : A persistent lineage of post-Triassic corystosperms in Gondwana

Komlopteris is a genus that includes the youngest representative of the so-called ‘seed ferns ’ , an informal group of gymnosperms that were prevalent during the Mesozoic but largely went extinct at the end of the Cretaceous. New fossil material, morphological data, and an extensive literature review allowed us to clarify the systematics of Gondwanan post-Triassic leaves of the Komlopteris lineage that were formerly assigned to diverse genera. Trends in diversity and distribution were identified. Ten species of Komlopteris were recognized in Jurassic to Eocene strata across Gondwana. Earliest records of the genus derive from South America soon after the end-Triassic extinction event. The genus reached its peak diversity and range in the Late Jurassic and Early Cretaceous and declined markedly after the Aptian. Its youngest representation is in mid-Cretaceous to Eocene deposits of southeastern Gondwana. Although never dominant, Komlopteris represents an important subsidiary component of austral Jurassic – Paleogene plant fossil assemblages. Striking morphological similarities to Dicroidium and Kurtziana , and co-occurrence with low concentrations of Alisporites / Falcisporites -type pollen, suggest that Kom-lopteris was a gymnosperm belonging to Umkomasiales ( = Corystospermales) that survived the end-Triassic and end-Cretaceous biotic crises in climatically buffered humid habitats of high southern latitudes. Arthropod induced leaf damage was rare on Komlopteris foliage and recognized only in two Late Jurassic – Early Cretaceous species.


Introduction
Various orders of seed plants with superficially fern-like leaves (commonly called 'seed ferns' or 'pteridosperms') were prominent globally during the Mesozoic.Among these, Umkomasiales (=Corystospermales) were overwhelmingly dominant in the Southern Hemisphere vegetation during the Triassic.Fossil remains of this group contributed greatly to Middle and Late Triassic peats that now constitute economic coal deposits.Foliage, wood, and reproductive structures of Umkomasiales were studied extensively from Gondwanan Triassic assemblages, and biological linkages between the dispersed organs are becoming progressively well understood based on physical attachments and shared cuticular characters (Anderson andAnderson, 1983, 2003;Unverfärth et al., 2022).Pollen recovered from Triassic umkomasialean microsporangiate organs is bisaccate and is typically referred to Alisporites or Falcisporites when found dispersed (Balme, 1995).
Although umkomasialean (and related matatiellalean) foliage typical of the early Mesozoic (e.g., Dicroidium, Kurtziana) disappeared during or soon after the end-Triassic extinction event (Bomfleur et al., 2018), pollen characteristic of this group persists in Jurassic, Cretaceous and even Paleogene strata across the region (de Jersey and Paten, 1964;Harris, 1965;Burger, 1973;McKellar, 1974;Dettmann, 1986).The continuity of such pollen in the geological record is unlikely to be the result only of reworking and suggests that umkomasialean lineages persisted across the Southern Hemisphere to an extent that has not so far been fully appreciated.A similar record, indicating the persistence of corystosperm-like lineages (variously attributed to Umkomasiales or Doyleales) in the Northern Hemisphere through the Jurassic and Cretaceous, has been revealed recently (Popa, 1997;Barbacka and van Konijnenburg-van Cittert, 1998;Shi et al., 2016Shi et al., , 2019)).
In searching for examples of post-Triassic umkomasialean (or mattatialean) fossils, we identified a range of leaf forms in Southern Hemisphere Jurassic to Eocene assemblages that have features strikingly similar in leaf architecture and venation to Dicroidium and Kurtziana.Some of these have remained undescribed, a few were attributed to Komlopteris, and others were assigned to Thinnfeldia or Alicurana.Doludenko, 1969Doludenko, , 1971Doludenko, , 1974 demonstrated that the type species of Thinnfeldia (T.rhomboidalis Ettingshausen) has features indistinguishable from representatives of Pachypteris.Consequently, Komlopteris was erected by Barbacka (1994) to accommodate leaves previously attributed to Thinnfeldia, having generally large, lanceolate pinnae that do not fall within the circumscription of Pachypteris.Both Komlopteris and Pachypteris are widely represented in Northern Hemisphere Upper Triassic and Jurassic strata.Five species of Komlopteris are currently recognized: K. nordenskioeldii (Nathorst) Barbacka, K. rotundata (Nathorst) Barbacka and K. speciosa (Ettingshausen, 1952) Cleal and Rees from the uppermost Triassic to Jurassic of Europe, and K. indica (Feistmantel) Barbacka and K. cenozoicus McLoughlin, Carpenter, Jordan and Hill from the Jurassic to Cretaceous and Eocene, respectively, of southern Gondwana.The morphology and cuticular details of the Northern Hemisphere forms were redescribed by Barbacka (1994), Popa (1997) and Cleal and Rees (2003), enabling clear species segregation.Gondwanan forms incorporate a broad range of leaf, pinnae and cuticular morphologies that have not been evaluated in detail for species delimitation.
Here, we review the representatives of Komlopteris from Gondwana, emend the genus, establish three new species, and propose five new combinations based on macro-morphological traits, in some cases supported by micromorphological details.We also provide new morphological observations of the youngest known seed fern, Komlopteris cenozoicus, and emend its diagnosis.We document the geographic and temporal range of Komlopteris species to assess their diversification and decline through the Mesozoic and Paleogene.We also evaluate general patterns of foliar variation within and between species through this interval.In addition, arthropod damage on Komlopteris foliage is noted.
Koonwarra Fossil Bed, Gippsland Basin, Victoria, Australia.The extensive fossil biota of the Koonwarra Fossil Bed within the Eumeralla Formation (=Wonthaggi Formation) of the Gippsland Basin, southeastern Australia, was described in detail by Drinnan and Chambers (1986), Dettmann (1986), and Jell and Duncan (1986).The fossils are preserved primarily as impressions in fine siltstone or shale.Fissiontrack and palynological dating favor an Aptian age for these fossiliferous strata (Dettmann, 1986;Wagstaff et al., 2020).
Gingin, Perth Basin, Western Australia.A single specimen of Komlopteris is preserved as an impression in strongly ferruginous siltstones from the Leederville Formation exposed in Macintyre Gully near Gingin in the northern Perth Basin, Western Australia.The associated macrofloral assemblage was described by Walkom (1944) and McLoughlin (1996) and is considered to be of Hauterivian-Barremian age based on the stratigraphic position of the host beds and palynostratigraphic data (Backhouse, 1988).However, the age of the base of the unit is not well resolved in some areas, and it may extend down to the Berriasian based on the inferred ages of underlying units.
Stewarts Creek, Stanwell Basin, Queensland, Australia.Sparse Komlopteris fossils are preserved as impressions in pink siltstones from the Stanwell Coal Measures at Stewarts Creek, near Rockhampton, central Queensland, Australia.These deposits have a broadly Valanginian-Barremian age based on their assignment to the Phyllopteroides laevis macrofloral Zone (Cantrill and Webb, 1987).
Talbragar Fossil Fish Bed, Surat Basin, New South Wales, Australia.The Talbragar Fossil Fish Bed has been studied extensively for its rich assemblage of vertebrate, invertebrate and plant fossils (Walkom, 1921;White, 1981;Turner et al., 2009;Beattie and Avery, 2012).The deposit represents an outlier of the Surat Basin and is generally correlated with the Purlawaugh Formation.The fossils are preserved as kaolinite-and opal-coated ferruginous and silicified volcaniclastic siltstones laid down in a small lacustrine setting.SHRIMP (Sensitive High Resolution Ion Microprobe) dating of zircons from volcanic ash in these beds yields an age (for the youngest zircon population) of 151.55 ± 4.27 Ma, i.e., latest Oxfordian-Tithonian (Bean, 2006).
Hope Bay, Antarctic Peninsula.The Hope Bay fossil flora, initially described by Halle (1913) and later revised by Gee (1989), Rees and Cleal (2004) and Birkenmajer and Ociepa (2008) derives from the upper Mount Flora Formation of the Botany Bay Group exposed near the tip of the Antarctic Peninsula (see Rees andCleal, 2004 andHunter et al., 2005 for details of the stratigraphy and geological setting).Zircon crystals from tuff layers about 170 m stratigraphically below the plant-bearing beds of the upper Mount Flora Formation were dated to 163.5 Ma indicating an Oxfordian (or potentially younger) age for this plant assemblage (Scasso et al., 2022).
Eumamurrin, Surat Basin, Queensland, Australia.Eumamurrin (26 • 10′39″S 148 • 49′25″E) is 40 km north of Roma, southern Queensland, Australia.This appears to be the same locality reported as "fossil site SB266, north of Lyndon Homestead" by White in Exon et al. (1967).The fossils are preserved as iron-stained or silica-coated impressions in argillaceous to silicified siltstones that were mapped as belonging to the lower part of the undifferentiated Injune Creek Group by Milligan et al. (1971).Recent palynostratigraphic correlations and U-Pb CA-IDTIMS dating of zircons from tuffs intercalated with these strata indicate a Callovian-Oxfordian age (Wainman et al., 2015(Wainman et al., , 2018)).

Fossil preparation
The Tasmanian leaf compressions were prepared using the dégagement method (Fairon-Demaret et al., 1999) and stored permanently refrigerated at 4 • C to reduce desiccation.Leaf fragments were oxidized with 10% aqueous H 2 O 2 for up to 48 h to recover cuticles, and some specimens were subsequently stained using safranin.Cuticles were mounted on microscope slides in glycerine jelly.Samples for scanning electron microscopy (SEM) were mounted on stubs using carbon tabs and coated in ~10 nm of platinum.

Imaging
Macromorphological features were photographed using a Canon EOS 700D camera, a Canon EOS 5D Mark III and a Nikon SMZ25 dissection microscope with NIS-Elements AR imaging software.Ultraviolet (UV) fluorescence macroscopic images were obtained in a dark room using high-energy UV torches (peak wavelength 365 nm).The images were manipulated in Adobe Photoshop 23.5.1 with the noise filter "dust and scratches" to remove fluorescing dust on the matrix, while fossils were spared.Transmitted light and UV fluorescence micrographs of cuticles were obtained using Olympus AX70 and Olympus BX51 optical microscopes (with a U-RFL-T light source and U-MBV2 filter with excitation wavelength(λ) = 400-440 nm and emission λ = 455 nm), respectively, and employing cellSens software.A Quanta 450 SEM and xT Microscope control software were used for SEM imaging of the cuticle.

Systematic paleontology
The Gondwanan species of Komlopteris are described below from the youngest to the oldest, in order to establish a foundation for comparison based on the best preserved and most confidentally assigned fossils, and to aid understanding of temporal trends in foliage architecture.To differentiate the species in this paper, each is compared to those mentioned subsequently, usually omitting those mentioned previously to avoid repetition.Where taxa were adequately described and illustrated in previous studies, we refer to the original reports and include only additional observations.Comparative overview tables of the diagnostic macro-and micromorphological characters of the Komlopteris species are provided in Supplementary Materials I and II.Silhouettes of the frond architecture and line drawings of the venation styles of all species are provided in Supplementary Materials III and IV, respectively.Ages and localities of the fossils are summarized in Fig. I.
Remarks: The original generic diagnosis provided by Barbacka (1994) is emended to encompass the broader morphological variation evident in the Australian Komlopteris species.Pachypteris (Brongniart, 1828) Harris, 1964 has a close relationship to Komlopteris as inferred from several shared features.The genera were originally distinguished based on four characters evident in Northern Hemispheric specimens (Barbacka, 1994;table 1).According to Barbacka (1994), Pachypteris comprises pinnate to bipinnate leaves with generally smaller (<30 mm length) pinnae, entire to dissected pinna margins, and evenly distributed stomata that are cyclocytic.
We argue that the Southern Hemisphere species discussed in this paper are consistent with Komlopteris Barbacka based on their pinnae being larger than typical for Pachypteris (>30 mm maximum length) and, where observable, their stomata are restricted to interveinal areas and have amphicyclic architecture.Although Barbacka (1994) defined Komlopteris as exclusively pinnate with entire (or "occasionally slightly undulate") pinna margins, some Gondwanan species described in this study incorporate a broader intraspecific (and even intra-leaf) spectrum of pinna forms, with some having entire to crenate, or greatly dissected margins that result in bipinnate architecture.
We identified additional characters to distinguish Komlopteris from Pachypteris.The lamina of Pachypteris is usually narrowly elongated with typically L:W proportions >3:1 and it is commonly bipinnate, in contrast to the typically 1:1-3:1 proportion of the commonly oncepinnate Komlopteris.Further, Pachypteris pinnae tend to be broadly attached to the rachis across the entire leaf, whereas in Southern Hemisphere Komlopteris they are constricted, or they vary greatly from narrowly constricted or petiolulate bases proximally, to broadly attached pinnae distally in the leaf.Overall pinna shape varies in both genera, but Pachypteris normally has small, ovate ultimate segments, whereas Komlopteris has larger, mostly lanceolate to falcate pinnae and, where bipinnate, pinnules that are short and triangular to rhombic.Finally, the rachis of Pachypteris commonly bears pustular ornament in contrast to the smooth or longitudinally grooved rachis surface of Komlopteris.
Another similar umkomasialean taxon, Dicroidium (and its distinctive ovuliferous and polleniferous organs Umkomasia and Pteruchus, respectively) has a well-supported stratigraphic range from the uppermost Permian to uppermost Triassic (with one known possible extension into the earliest Jurassic: Bomfleur et al., 2018).The Dicroidium rachis typically features a single fork in the proximal portion of the leaf, which is key to distinguishing this genus from foliage of Kurtziana and Komlopteris.Like Komlopteris, the degree of pinna dissection and venation style in Dicroidium is extremely variable (Anderson and Anderson, 1983).Kurtziana Frenguelli, 1942, was erected for relatively large pinnate leaves with alethopteroid venation from lower Mesozoic strata of Gondwana and it is morphologically similar to Komlopteris (Herbst and Gnaedinger, 2002).Anderson and Anderson (2003) recognized about 16 Kurtziana species, distributed across the Southern Hemisphere, that appear to be restricted to the Triassic (Pattemore et al., 2015).Kurtziana species can be distinguished from Komlopteris by their consistently entire-margined pinnae and typically higher-angled secondary veins.However, there is overlap between the diagnostic characters of these genera, and cuticular or reproductive features will ultimately be required to clarify their relationships.Cuticular morphology has not been reported for Kurtziana leaves.
Huncocladus was established by Andruchow-Colombo et al. (2019) as a new Podocarpaceae genus related to Phyllocladus based on a specimen from Argentina (MPEF-Pb 8028) of similar age to Komlopteris cenozoicus.However, doubt has been raised about its affinity owing to the cuticle being atypical for Podocarpaceae and its cycad-like stomata (Dörken et al., 2021).We conclude that the genus may be related to Komlopteris based on characters concerning both macro-and micromorphology.The principal traits shared by Huncocladus and Komlopteris are the bipinnate leaf, decurrent and falcate pinnae (see Andruchow-Colombo et al., 2019, figs. 3, 4, 5), alethopteroid venation with secondary veins that bifurcate 1-2 times (see fig. 5), and amphicyclic stomatal complexes with rings of 3-5 cells (see figs.6F-I).Further traits illustrated by the authors that are similar to those of K. cenozoicus (see section Species-Komlopteris cenozoicus) include possibly two size ranges of stomata (figs.6C, D), potential resin bodies (fig.6C), and thickened cuticular rings around the stomatal pit (figs.F-J).However, the apparent presence of resin canals in Huncocladus which have so far not been observed in other Komlopteris species and the crowded preservation of the Argentinian fossil that obscures the pinna arrangement, hinders definitive systematic placement of this fossil.
Alicurana Herbst and Gnaedinger, 2002 was established to accommodate Early Jurassic pinnate leaves with elongate pinnae and alethopteroid venation that were previously included in Kurtziana (Arrondo and Petriella, 1980;Petriella and Arrondo, 1982;Artabe et al., 1991) but differed in having cordate and petiolate pinna bases.The cuticle of the type species, Alicurana artabeae, shares several features (e. g., prominent polar extensions on guard cells, and dicyclic stomata with 6-7 subsidiary cells) with Komlopteris cenozoicus.Although the cordate bases of the pinnae are distinctive in A. artarbeae, a second species from the same locality (A.nestarensis, Herbst and Gnaedinger, 2002), has simple contracted bases that are indistinguishable from Komlopteris species.Based on their similarities in both macro-and micromorphology, we consider Alicurana to be a junior synonym of Komlopteris.
Pachydermophyllum Thomas and Bose, 1955 leaves are differentiated from Komlopteris by being once-pinnate with leathery lanceolate pinnae bearing a midrib with simple or rarely forked secondary veins.The leaves are weakly amphistomatic, with stomata consisting of sunken guard cells rimmed by subsidiary cells with papillae overhanging the stomatal pit.Although Harris (1964) transferred the Yorkshire Jurassic Pachydermophyllum leaves to Pachypteris on the basis of insufficient and inconsistent differentiation of key characters, many subsequent authors have continued to recognize Pachydermophyllum, predominantly in Southern Hemisphere Mesozoic assemblages, based on its more elongate leaflets and strongly verrucate rachis (Retallack, 1977(Retallack, , 1981;;McLoughlin et al., 2002).There remain uncertainties about the differentiation of Kurtziana and Pachydermophyllum, since their diagnoses are similar, except that the latter incorporates cuticular features not available in the former (Anderson and Anderson, 2003;Bomfleur et al., 2011).Elgorriaga et al. (2019) noted that the cuticular features of several Pachydermophyllum species are strikingly similar to those of Lepidopteris, indicating a possible peltaspermalean affinity.
Archangelskya Herbst is distinguished from Komlopteris by its forked rachis and, generally, by its more deeply dissected pinnae (Herbst, 1964).In many respects, Archangelskya (ranging from Lower Jurassic to Lower Cretaceous) is similar to Dicroidium (which ranges through the preceding Triassic), but reproductive structures have not been linked to Archangelskya, so comparison with the fructifications of Dicroidium is not yet possible.
Description: Komlopteris cenozoicus was originally described by McLoughlin et al. (2008) based on leaf fragments.Here, we provide a comprehensive description based on additional morphological details found in many newly collected specimens from the type locality.
Preservation: Komlopteris cenozoicus is preserved as compression fossils forming leaf mats.Preserved leaves include cuticular micromorphology.
Leaf architecture: Leaves are once-pinnate, typically imparipinnate, with L:W proportions of ~1:1-2:1.Maximum leaf length and width are 103 mm and 71 mm, respectively.Petioles are up to 20 mm long.Petioles and rachises bear weak longitudinal striations and reach maximum widths of ~2 mm.The rachis grades into the midvein of the terminal pinna.Pinnae have opposite to subopposite arrangements on the rachis with lateral to slightly dorsal attachment.
The maximum pinna length is 58 mm, and maximum width is 13 mm.Margins are mostly entire and, rarely, slightly sinuous.Pinna morphology changes gradually throughout the lamina.Proximal pinnules are smaller, shorter, and broadly lanceolate.Distal pinnules are larger and lanceolate to falcate, or commonly rhombic.Pinnae are sessile and have acute to acutely truncate bases.Proximally, pinnae have equally constricted bases, whereas distally they become decurrent to broadly attached.The distalmost pinnae are typically conjoined.Bases are otherwise acute to acutely truncate with angles of 42-90 • .Apices are bluntly acute with angles of 33-72 • .
Venation: Veins are raised.The midvein is proximally distinct, narrowing distally, and commonly evanescent near the pinna apex.Venation is alethopteroid.Secondary veins diverge from the midvein at high angles and then taper towards the margin.Angles between the midvein and secondary veins are up to ~45 • , but most commonly ~20-30 • .Secondary veins bifurcate usually 1-3 (mostly 2) times before terminating at the leaf margin.The maximum distance between secondary veins is ~0.8 mm.
Micromorphology: Leaves are typically hypostomatic.Stomata are equally distributed between veins (which lack stomata).Stomata present on the petioles and rachises, but at lower density.Stomatal orientation is variable on pinnules (Plate III 2-4) and parallel to the longitudinal axes on the petiole and rachis (Plate III 10).Guard cells are reniform, deeply sunken and possess a thickened rim surrounding the stomatal pit (Plate III 2).Guard cells are surrounded by a deep groove on the inner side of the cuticle (Plate III 5-11).Stomata are typically dicyclic with 4-7 subsidiary cells per ring.On the inner cuticular surface, a thin anticlinal wall is present, which divides the subsidiary cells into two rings (Plate III 5-6, 11).On the petiole and rachis, polar subsidiary cells are elongated (Plate III 10).
One incorporates large guard cells (23-41 μm long) with variable morphology, incorporating pore ledges and outer stomatal ledges (sensu Ash et al., 1999, fig. 55.1) and in some cases with polar extensions.The second has smaller guard cells and are present on pinnules but absent on the petiole and rachis.Around these stomata, guard cells are 13-17 μm long, variably shaped, but typically reniform to square, with a smooth surface or having multiple ledges.Epidermal cells are variable (Plate III 1) but are typically square or triangular to slightly undulate (sensu Nishida and Christophel, 1999, fig. 5g).Cell lengths are ~60-250 μm and widths are ~40-150 μm.
Epidermal cells on the veins and rachis/petiole are elongate rectangular, ~130-350 μm long and ~40-70 μm wide.Cuticular impressions over anticlinal cell walls are shallow, and more pronounced on the adaxial than on the abaxial cuticle.Under UV light, small (~200 μm diameter) interveinal circular bodies are visible in some specimens (Plate I 4).
Comparison and remarks: The high-paleolatitude Lowana Road fossils represent by far the youngest evidence of the umkomasialean lineage, extending about 14 Myrs into the Cenozoic.Despite this temporal isolation, this foliage retains the diagnostic micro-and macromorphological characteristics of Mesozoic Komlopteris species (McLoughlin et al., 2008).Circular features originally interpreted as probable trichome bases, are here reinterpreted as aberrant or incompletely developed stomata (Plate II 2, 11), since they are rimmed by subsidiary cells and variably develop a central pore.No trichomes were detected on these leaves.The small, circular, opaque interveinal structures are interpreted to be resin bodies based on the corresponding size and position of resin bodies in various Dicroidium leaves (Supplementary Material V 7; for discussion see Section 4.3 Resin bodies).

(caption on next page)
M. Slodownik et al. indica can be distinguished from K. cenozoicus by its typically more slender pinnae, paripinnate leaf architecture and amphistomatic habit with stomatal apertures orientated parallel to the veins.Komlopteris constricta is separated by its generally slender pinnae in alternate arrangement, lower secondary vein angle and lesser vein bifurcation.Komlopteris artabeae and K. nestarensis are distinct in pinna shape (auriculate bases and/or obtuse to rounded apices).Furthermore, K. artabeae is weakly amphistomatic with parallel stomata and bears hairs or clavate papillae on both adaxial and abaxial surfaces-features absent in K. cenozoicus.
Distribution: Known only from the type locality.
Diagnosis: Pinnate leaves with typically opposite to subopposite, or, in a few cases, alternating pinna arrangement; lamina L:W ratio up to ~3:1.Pinnae lanceolate to slightly falcate; margins entire to lobed, up to ~50% incised (from the outermost leaf margin towards the midvein).Pinnae sessile or petiolulate; base acute to obtuse, narrowing equally to midvein or decurrent, attachments typically broaden towards leaf apex.Apices bluntly acute to rounded.Secondary veins sub-parallel, bifurcating 1-3 (mostly 2) times, curving away from midvein to angles of < 70 • .
Comparison and remarks: A detailed description of this species was provided by Drinnan and Chambers (1986).All specimens are preserved as impressions without epidermal micromorphology.The material was originally assigned to Thinnfeldia sp.cf.T. indica Feistmantel, 1876 (=Komlopteris indica) by Drinnan and Chambers (1986).The Koonwarra fossils can be differentiated from K. indica because the pinnae of the former have a smaller L:W ratio, lobed margins on the proximal pinnae, and more truncated and sharply contracted bases and are, therefore, assigned to a new species.
Komlopteris victoriensis can be separated from K. boolensis, K. tiruchirapalliense and K. purlawaughensis mainly by the absence of deeply dissected pinna margins.Komlopteris khatangiensis is distinguished by its larger lamina and smaller pinna L:W ratio and exclusively odontopteroid venation.The distinguishing features of K. victoriensis compared to K. constricta are its lower L:W pinna ratios, generally higher secondary vein angles and greater number of secondary vein bifurcations.Komlopteris artabeae and K. nestarensis are distinguished by their larger basal angles and entire pinna margins.
A specimen originally assigned to Dicroidium acuta (Gothan) Walkom by Medwell (1954) appears to be consistent with K. victoriensis in all respects.This specimen was found near Jumbunna, 20 km northwest of the type locality, and probably derives from the same stratigraphic unit.
A single specimen with contracted pinnule bases registered in the GSQ collections as Sphenopteris flabellifolia from the mid-Albian Styx Coal Measures of central eastern Queensland (Walkom, 1919) may also belong to K. victoriensis (Supplementary Material V 8), but represents only an apical fragment lacking the full spectrum of leaf characters.Other specimens attributed by Walkom (1919, pl. 3, figs 9, 10) to S. flabellifolia (Tenison-Woods) from the Burrum Coal Measures (Albian) of the same region have slender serrate-margined pinnae and clearly do not belong to K. victoriensis.
Thinnfeldia sp.illustrated by Daniel (1989, pl. 10, fig. F) and mentioned by Daniel et al. (1990, p. 28) from the upper Albian to Cenomanian Warder Formation in the Clarence Valley, South Island, New Zealand (Browne and Reay, 1993), appears to be conspecific with K. victoriensis, sharing such characters as sharply constricted (truncate) pinna bases, slightly lobed margins and bluntly acute to rounded apices.
Distribution: Confidently identified specimens are known only from the Aptian type unit in south-eastern Australia, but the species possibly extends to the Albian of north-eastern Australia and the Albian-Cenomanian of New Zealand.

Synonymy:
1969 Thinnfeldia cf.T. chunakalensis Sah and Dev;Douglas,pp. 52,53,pl. 16, Diagnosis: Leaves pinnate to bipinnate, L:W ratio up to ~3:1.Pinnae mostly opposite to subopposite; lanceolate to slightly falcate.Margins entire, lobed, or (rarely) incised to the midvein forming pinnules, incision depth reducing distally.Pinnae sessile to petiolulate, commonly decurrent.Pinna bases usually constricted near leaf base, but widening in distal part, acute to obtuse angles.Apices bluntly acute.Alethopteroid or odontopteroid venation in pinnae with entire and dissected margins, respectively.Secondary veins emerging at low angles, and arching to intersect margin at <70 • , typically bifurcating 1-3 (mostly 2) times.The following pinnate to bipinnate species can be separated from K. boolensis: K. tirichuparaliense and K. purlawaughensis have rounded, only slightly angular pinnules and generally larger pinna bases and, in the latter species, larger apical angles; K. constricta has typically alternate pinnae that are generally narrower, with secondary veins at smaller angles that typically bifurcate once.The remaining species are strictly once-pinnate and can be distinguished by their greater leaf or pinna L:W proportions.A single bipinnate specimen (Supplementary Material V 4-5; GSQF718) from the Stewarts Creek Formation in central Queensland (Stanwell Basin) is here assigned to Komlopteris boolensis based on its apparently identical leaf morphology.The Stewarts Creek Formation may be coeval with the Rintoul Creek Formation (yielding the K. boolensis holotype) based on the shared Valanginian-Barremian stratigraphic index taxon Phyllopteroides laevis (Cantrill and Webb, 1987).
As noted by Douglas (1969), Thinnfeldia chunakhalensis Sah and Dev, 1957 has a superficially similar leaf architecture and venation to K. boolensis, but the rachis of the former is very slender and the pinnae are elliptical with bases that contract to a small petiolule.That species was considered a synonym of the osmundaceous fern Phyllopteroides laevis Cantrill andWebb, 1987 by Banerji (1992).
Distribution: The species is known from the Valanginian type locality in Victoria, Australia, and from Valanginian-Barremian strata in central Queensland.
Thinnfeldia nirmalensis Roy, 1973 was established based on a single leaf fragment with five incomplete pinnules from Tetria in the Rajmahal Hills, India.This specimen appears to represent the basal part of a leaf with relatively squat pinnae, but the lamina and venation features are otherwise consistent with K. indica specimens from the same region.Some non-Indian fossils that were previously assigned to K. indica (Drinnan and Chambers, 1986;McLoughlin et al., 2002;Rees and Cleal, 2004;Birkenmajer and Ociepa, 2008) are here assigned to other species based on differences in pinna form and cuticular micromorphology.
Komlopteris indica is differentiated from other species in the genus by its typically entire-margined, slender pinnae, with gently tapered bases.However, some intraspecific variation in pinna shape is revealed in a specimen illustrated by Feistmantel (1877, pl. 39, fig. 1, 1a) that has partially rhombic pinnae with more sharply constricted bases and undulate to slightly lobed leaf margins.The venation of K. constricta (Plate VI) is most similar to K. indica, having secondary veins that typically bifurcate twice.However, the secondary veins of K. constricta are generally more acute and bifurcate close to the midvein-features that appear to be more variable within K. indica (e.g., Zeba-Bano et al., 1979; text-fig.2).Furthermore, the pinna arrangement of K. constricta is typically alternate, in contrast to the more variable leaflet insertion in K. indica.
As noted by Maheshwari (1986), a specimen assigned to "Pecopteris (?) salicifolia Oldham and Morris" by Feistmantel, 1877;pl. 26, fig. 2, is possibly an apical fragment of K. indica but the line drawing lacks sufficient details for confident identification.Pal et al. (2009) discussed this specimen in detail, noting profound dissimilarities of the illustration to features of Thinnfeldia indica (=Komlopteris indica).Consequently, they reassigned the specimen to Cladophlebis salicifolia Morris.
Distribution: Feistmantel (1876Feistmantel ( , 1877) ) recorded this species from the Rajmahal Group (Lower Cretaceous) at several localities in the Rajmahal Hills of northeastern India.Subsequent authors have identified it from the same unit at additional localities in the Rajmahal Hills (Zeba-Bano et al., 1979;Banerji and Jana, 2000), and also from the 'Utatur plant beds' of the Tiruchirapalli District in the Cauvery Basin, Tamil Nadu, southeast India (Maheshwari, 1986) (1988, p. 191).
Comparison and remarks: Sukh-Dev and Rajanikanth (1988) established Sphenopteris tiruchirapalliense based on three incomplete leaves from Lower Cretaceous strata of Tamil Nadu, southern India.The type specimen conforms to the emended definition of Komlopteris (based on its lanceolate to falcate pinnae with dissected margins) and we here reassign the species to that genus.Chinnappa et al. (2015) established "Thinnfeldia vemavaramensis" (although failed to designate a registered holotype) for a single incomplete illustrated specimen from Lower Cretaceous strata of southern India and regarded the earlier-published Sphenopteris tiruchirapalliense as a synonym.Additional specimens illustrated by Jain (1968 pl. 2, figs 11, 12) as "Dicroidium sp." are also referred to K. tiruchirapalliense and appear to confirm the presence of moderately dissected pinnae throughout the leaf in that species.Pinna fragments illustrated by Baksi (1968, pl. 2, figs 11a, b) as "Dicroidium sp." and by Sharma et al. (1972) as "Thinnfeldia cf.T. feistmanteli" have similar venation but are too incomplete or poorly preserved for unequivocal assignment to K. tiruchirapalliense.Chinnappa et al. (2015) included material illustrated by Feistmantel (1879, p. 203, pl. 1, figs. 1, 1a, 6, 6a, 7, 7a) as Thinnfeldia subtrigona in this species.However, Feistmantel's specimens have pointed segment apices and are too fragmentary to confidently assign to any established species.
Komlopteris tiruchirapalliense is most similar to the bipinnate variants of K. purlawaughensis in having broadly equivalent dissected pinnae, rounded to slightly angular (rhombic) segments, and odontopteroid secondary venation.However, the pinnae of K. tiruchirapalliense are markedly dissected throughout the leaf (including the terminal pinna), whereas K. purlawaughensis has more or less entire pinnae near the leaf apex.Owing to the scarcity of K. tiruchirapalliense specimens, and, particularly, the lack of basal portions of leaves, it is unclear whether this species encompasses a strong degree of intraspecific variation.Thus, we provisionally retain this form as a species separate from K. purlawaughensis, but more complete material will be needed to assess the full scope of its intraspecific variability, and coalified compressions are required to evaluate its micromorphological details.
Two specimens previously assigned to Thinnfeldia from the Leederville and Bullsbrook formations (Lower Cretaceous) of the Perth Basin by Walkom (1944) and McLoughlin (1996) are here assigned to Komlopteris sp.cf.K. tiruchirapalliense (Supplementary Material V 6).The most complete leaf (from the Leederville Formation) is distinctly petiolate but is preserved as an impression in ferruginous siltstone and lacks clear details of the venation and epidermal features.In gross morphology, it shares the subopposite, relatively short, broad, moderately lobed and rounded pinnae with K. tiruchirapalliense from southeast India, but the lack of venation details hinders unequivocal assignment to that species.The fragmentary pinna (from the Bullsbrook Formation) shares the short rounded rhombic pinnules and odontopteroid venation of K. tiruchirapalliense but is too incomplete for unequivocal identification to that species.
Distribution: This species is known exclusively from Lower Cretaceous strata in southeast India.The morphologically similar specimens from Gingin and Bullsbrook in the Perth Basin, Western Australia, are of approximately Hauterivian-Barremian age.
Diagnosis and description: See Sengupta, 1988, pp. 75-77.Comparison and remarks: Komlopteris khatangiensis differs from all other species by its narrow once-pinnate leaf with L:W > 3 and having alternately arranged, ovate to triangular pinnae with odontopteroid venation and weakly and irregularly lobed margins.For example, other Komlopteris species have generally lower L:W ratios and other oncepinnate species have exclusively alethopteroid venation.The pinna bases are decurrent and distinctly contracted (especially acroscopically)-features common to this genus but absent in superficially similar foliage of Phyllopteroides and Cladophlebis.The overall morphology of K. khatangiensis is considered to be sufficiently similar to be included into the genus at present, but further study is necessary owing to the scarcity of material and its incomplete state.
Distribution: Sengupta (1988) recorded this species only from the upper part of the Dubrajpur Formation in assemblages assigned to the Ptilophyllum acutifolium-Gleichenites gleichenoides Zone, purportedly of Early Jurassic age.However, Tripathi and Ray (2006) indicated that the upper parts of the Dubrajpur Formation may have been deposited later in the Jurassic or be as young as Early Cretaceous based on revised palynological data.
Diagnosis: Leaves pinnate to bipinnate.Pinnae opposite to alternate; margins variably entire, crenate, lobed or dissected to midvein.Pinnae sessile to petiolulate.Pinna bases constricted acroscopically and constricted to slightly decurrent.Distal pinnae of entire forms broadly attached.Basal angles acute to reflexed; apices typically rounded, obtuse or bluntly acute.Pinnules broadly attached, slightly rounded rhombic.Secondary venation of entire and slightly lobed pinnae is alethopteriod at angles <60 • , bifurcating once or twice; and of dissected pinnae is odontopterioid at angles <120 • , bifurcating up to five times.

Description:
Preservation.Komlopteris purlawaughensis is preserved as isolated leaf impressions (some with kaolinite coatings) lacking cuticular details.
Leaf architecture.Leaves are paripinnate to imparipinnate, generally having a pattern of strongly lobed pinnae proximally, to entire pinnae distally.In large, bipinnate forms, a strong degree of lobing is generally consistent throughout the leaf.Length to width proportions of fully preserved leaves are ~2:1.Complete leaves in the Talbragar population are up to 126 mm long and 74 mm wide.Bipinnate leaves from the Injune Creek Group have maximum lengths of >125 mm and widths of >110 mm; pinnate leaves are generally smaller (reaching >42 mm long and > 40 mm wide).The rachis is up to 5 mm wide with weak longitudinal striations.Maximum free-rachis length is 25 mm.Leaves have opposite to subopposite pinnae and pinnules (rarely fully alternate).
Pinnae have a lateral or slightly dorsal attachment to the rachis.The longest pinnae occur in the medial part of the leaf and decrease in size proximally and distally.Maximum pinna length and width are ~68 mm and ~ 18 mm, respectively.Most pinnae are crenate, slightly lobed, or incised to the midvein forming discrete pinnules.Some pinnae, mainly near the leaf apex, may be entire.Pinna shape varies from lanceolate and slightly falcate to elongate elliptical.Pinnae are sessile to petiolulate with basal angles that are acute to obtuse (~72-100 • ), and commonly truncate.The attachment of typical pinnae is acroscopically contracted to the midvein and basiscopically contracted or decurrent along the rachis.Attachments of most distal pinnae are broadly decurrent.Pinna bases are acute or, more commonly, obtuse (~60-240 • ).Pinna apices are bluntly acute to obtuse, rounded or sporadically reflexed (~43-263 • ).The number of pinnules per pinna is variable, but with a maximum count of 10.Pinnules are variably semicircular to rhombic, ~10 mm long and ~ 7 mm wide (decreasing towards the pinna apex) and with a wide base.
Venation.Pinna midveins are distinct but evanesce distally.Secondary veins terminate at the leaf margin.Near the margin, the maximal distance between secondary veins reaches ~0.7 mm.Venation pattern varies between entire and dissected pinnae.Entire pinnules have alethopteroid venation with secondary veins emerging at low angles from the midvein then arching to angles <55 • in relation to the midvein.Secondary veins are mostly sub-parallel, generally bifurcating once near the midvein and again medially or near the margin.Pinnules of deeply dissected pinnae have odontopteroid venation (fan-like arrangement without an apparent midvein) incorporating veins that bifurcate up to five times and arch to a maximum of ~118 • .
Comparison and remarks: Townrow (1965, p. 505) regarded pinnate to bipinnate leaves assigned to Thinnfeldia talbragarensis by Walkom (1921) to be a synonym of Pachypteris crassa (Halle) Townrow.However, we conclude that the specimens illustrated by Walkom (1921) represent two distinct species.The more complete specimens (Walkom, 1921, pl. 1, figs. 1, 2) with oblong and commonly distorted pinnae with entire or weakly undulate margins and high-angle secondary veins, are best assigned to Pachypteris (in agreement with Townrow, 1965).Walkom's (1921, pl. 2, figs. 7-9) fragmentary specimens have ovate pinnae with lobed margins and more acute secondary veins and represent a distinct species.Since Walkom (1921) did not select a holotype, we apply the name T. talbragarensis to the most complete specimens, hence, they are synonymous with P. crassa.We establish the new name Komlopteris purlawaughensis for Walkom's mostly fragmentary leaf specimens and select a complete leaf from the Talbragar locality as the holotype.
Pinnate to bipinnate leaves assigned to Pachypteris crassa by Hill et al. (1966) from the Injune Creek Group have a slightly different preservation (ferruginous to siliceous coatings on impressions in tuffaceous siltstones) to those from Talbragar.However, the morphological characters of these specimens (Hill et al., 1966, p. J.8, pl. J IV, figs. 2-4) are indistinguishable from the examples of K. purlawaughensis from Talbragar.For similarites to other lobed spieces see "Remarks" section of K. tiruchirapalliense (above).
Emended diagnosis: Typically once-pinnate leaves generally with alternately arranged pinnae.Maximum length unknown, but >62 mm and maximum width ~30 mm.Pinnae lanceolate to falcate, or slender rhombic.Pinnule margins typically entire or slightly sinuous.Bases acute, tapering gradually.Pinnae sessile, acroscopically constricted and decurrent.Apices bluntly acute.Secondary veins sub-parallel, bifurcating typically once; emerging at low angles and recurving slightly near midvein, then remaining relatively straight at angles of <30 • .
Comparison and remarks: Specimens from Hope Bay, Antarctic Peninsula, originally attributed to Thinnfeldia constricta, were described extensively by Halle (1913), Gee (1987), and Rees and Cleal (2004).The species is represented by incomplete leaf impressions reaching >62 mm long by up to 30 mm wid.We provide illustrations to highlight the variation in leaf and pinna forms (Plate VI) and note circular, dark interveinal markings (Plate VI, 7) that are interpreted as impressions of resin bodies(for discussion see Section 4.3 Resin bodies).
The specimens described by Halle (1913) were later deemed to be conspecific with Indian specimens identified as K. indica by Rees and Cleal (2004) and Birkenmajer and Ociepa (2008).Although the Indian and some Antarctic leaves have similar gross leaf shapes, they differ slightly in pinna arrangement and venation pattern (for comparisons see "Remarks" section of Komlopteris indica, above).In addition, the significant paleogeographical and temporal (40-million-year) separation of the Late Jurassic Antarctic Peninsula and Early Cretaceous Indian forms means that they likely constitute distinct species.We retain them as separate taxa pending the availability of micromorphological details and reproductive organs of both species.We propose a new combination for the Antarctic material: Komlopteris constricta (Halle).Some specimens of Komlopteris constricta have markedly dimorphic pinnules on opposite sides of the leaf (Plate VI 1) and the Hope Bay population of leaves encompasses a morphological continuum between simple pinnae (traditionally assigned to 'T. constricta' or K. indica; Plate VI 2-9, 12-13) and highly dissected pinnae (traditionally assigned to Pachypteris dalmatica F. v. Kern.or Pachypteris hallei Frenguelli, 1943;Plate VI 10-11, 14-16)-see also the illustrations by Halle (1913, pl. 4, figs 23-35) and Gee (1989, pl. 4, figs 36, 37).Birkenmajer and Ociepa (2008) identified one such 'transitional' leaf from Hope Bay as K. indica (Feistmantel) Barback, emend. Rees and Cleal. This specimen (Birkenmayer and Ociepa 2008, fig. 27, right) has apical pinnae similar to K. indica but the proximal pinnae are dissected into inclined lobes with parallel secondary venation apparently lacking bifurcations.Given this variation in a single specimen, and the range of leaf forms represented in the Hope Bay population, we regard P. hallei as a junior synonym of K. constricta.
Some specimens attributed to Pachypteris hallei by Gee (1989) are leaves with deeply dissected pinnae with inclined pinnules.Similar specimens, illustrated by Fuenzalida et al. (1972) from Middle Jurassic strata of northeastern Snow Island as Pachypteris lanceolata and Scleropteris sp., have narrow pinnae with attenuated apices that appear to conform better to Pachypteris crassa (Halle) Townrow, 1965 than Komlopteris.
Distribution: Komlopteris constricta is known only from the Oxfordian strata of Hope Bay, Antarctic Peninsula.

(caption on next page)
M. Slodownik et al.Species: Komlopteris artabeae (Herbst et Gnaedinger 2002)  Comparison and remarks: Herbst and Gnaedinger (2002) reassigned specimens from the Jurassic Nestares Formation previously identified as a typical Triassic taxon, Kurtziana brandmayri (Arrondo and Petriella, 1980;Petriella and Arrondo, 1982;Artabe et al., 1991), to a new genus and species: Alicurana artabei.Alicurana is herein regarded to be a junior synonym of Komlopteris (for discussion see 'Remarks' section of Komlopteris).Herbst and Gnaedinger (2002) regarded the description provided by Artabe et al. (1991, pp. 366-372, for then assigned to Kurtziana brandmayri) as the diagnosis for the species.The key defining characters cited from Artabe et al. (1991) are as follows: monopinnate (imparipinnate) leaves, with robust rachis, abutting or overlapping, sessile, lanceolate pinnae inserted at 70-85 • , each having a contracted cordate to auriculate symmetrical base, entire margin, asymmetrically obtuse or rounded apex, and alethopteroid venation except in basal pinnae where there is a tendency towards odontopteroid venation.The leaves are amphistomatic with isodiametric, quadrangular, triangular or polygonal interveinal cells and proportionately longer rectangular, triangular or polygonal cells over the veins; stomata are imperfectly dicyclic with the pore sunken below rings of 6-7 subsidiary cells.
Komlopteris artabeae is clearly differentiated from all other species in the genus by its distinctive auriculate base, ovate pinnae, lack of marginal lobing and, apart from K. khatangiensis, by far the highest L:W ratio.Although the macromorphology is clearly distinct from other Komlopteris species, it conforms with the cuticular morphology of other species: The stomata share the well-developed polar extensions, sunken aperture, prominent stomatal ledges (apparently equivalent to the inner ring of subsidiaries reported by Herbst and Gnaedinger, 2002) and interveinal distribution with K. cenozoicus, and a weakly amphistomatic habit and parallel stomatal orientation with K. indica.
Comparison and remarks: Alicurana nestarensis was established by Herbst and Gnaedinger (2002) for Jurassic leaves with organic preservation that were otherwise similar to Triassic Kurtziana catcheutensis (Kurtz) Frenguelli, 1942, which so far lacks organic preservation.Herbst and Gnaedinger (2002) noted that the key features defining A. nestarensis are its small imparipinnate leaves with a thick, striate rachis bearing laterally attached oblong-linear, contiguous, pinnae with constricted bases, subparallel margins, obtusely rounded to truncate apices, alethopteroid venation with twice-divided secondary veins, and epidermis with isodiametric to rectangular cells and bi-and tetraperigenous stomata.We consider that this species conforms to the diagnostic characters of Komlopteris and here reassign the species to that genus.
Although Herbst and Gnaedinger (2002) included cuticular details in their diagnosis of this species citing details provided in an earlier study Petriella and Arrondo (1982), we note that it is unclear whether the cuticles illustrated for this taxon derive from Jurassic specimens (of Komlopteris nestarensis) or from Triassic leaves (of Kurtziana catcheutensis).Until cuticular details become available from unequivocal K. nestarensis specimens from the Alicura type locality we consider the cuticular micromorphology of this species as unresolved.We further note that the interpretation of Herbst and Gnaedinger (2002) of "obtusely rounded to truncate apices" may be a result of preservation bias as the holotype's pinna tips do not seem to be fully preserved (e.g., Arrondo and Petriella, 1980, pl. 3, fig. c).A specimen from the same locality published by Morel et al. (2013) appears generally similar to the holotype, but exhibits bluntly acute to rounded pinna tips.Komlopteris nestarensis can be distinguished from other Komlopteris species by a combination of characters, including the lack of lobing, lower L:W ratio of pinnae and generally wide basal and apical angles.Komlopteris nestarensis is strongly similar in gross morphology to Triassic Kurtziana cacheutensis leaves and a cuticular study of these species is required to determine if they are conspecific.Komlopteris nestarensis differs from the co-occuring K. artabeae in having primarily lanceolate pinnae with decurrent bases.
amarjolense Sharma et al. (1972) is represented by linear leaf fragments with semicircular to short oblong pinnae that are more consistent with the morphology of Rintoulia McLoughlin and Nagalingum (Mcloughlin et al., 2002).The report of Thinnfeldia sp. from the Lower Cretaceous Athgarh Formation of the Ghantikhal district, Katak, Odisha, India, by Adyalkar and Nageswara Rao (1960) cannot be verified in the absence of an illustrated specimen.A leaf portion illustrated by Sah (1965, pl. 1, fig. 3) as ?Thinnfeldia sp.from Khatangi Hill (upper Dubrajpur Formation: Upper Jurassic-Lower Cretaceous), Rajmahal Hills, has short rhombic pinnae with gently divergent secondary venation and might represent the basal portion of a Komlopteris indica leaf but more complete material is needed to verify this identification.Finally, a specimen (MPEF PB 6881) attributed to Pachypteris sp. from the Las Leoneras Formation (Sinemusian-Pliensbachian) at Cerro Bayo, Chubut province, Argentina (Escapa et al., 2022, pl. 2, fig. D), is strikingly similar in leaf and venation morphology to the weakly lobed species of Komlopteris recorded here from the Early Cretaceous of eastern Australia.Although the Argentinean specimen lacks cuticle, it may represent an early representative of the genus and further collections of this material are warranted to assess its affinities.

Stratigraphic and geographic range
Many early studies reported 'Thinnfeldia' to be abundant in Southern Hemisphere Triassic strata.However, these examples, have almost universally been reassigned to Dicroidium or Kurtzianatwo diagnostic Gondwanan seed fern taxa (Anderson andAnderson, 1983, 1989).Dicroidium (Umkomasiales), with forked leaves, is abundant across Gondwana throughout the Triassic, defining a distinctive middle-highlatitude deciduous vegetation type (Anderson and Anderson, 1983;McLoughlin, 2001).Kurtziana (variably attributed to Matatiellales, Umkomasiales, or Peltaspermales), with unforked leaves, represents a subsidiary component (typically <1% of specimens in any one assemblage) of Gondwanan Triassic floras (Anderson and Anderson, 2003;Holmes and Anderson, 2005).Kurtziana was apparently a casualty of the end-Triassic extinction.We conclude from their strong morphological similarities and temporal range, that some component of the Kurtziana group (confined to the Middle and Late Triassic), may have given rise to the lineage represented by Komlopteris that appeared in the latest Triassic (in low to middle paleolatitudes) and persisted through the Jurassic in both hemispheres, becoming confined to high southern latitudes of Gondwana in the Cretaceous and early Paleogene.
There is a general shift in distribution of southern representatives of the genus from western to southeastern Gondwana through the Mesozoic (Fig. I).The oldest examples of Komlopteris yet recorded are sparse Early Jurassic forms previously assigned to Dicroidium (although they lack the diagnostic forked rachis of that genus) or to Kurtziana and Alicurana (Artabe et al., 1991;Herbst and Gnaedinger, 2002;Sagastia et al., 2019).Middle to Late Jurassic records from Gondwana derive primarily from eastern Australia, where they represent prominent components of the floras of the lower Injune Creek Group (Callovian-Oxfordian: Wainman et al., 2015Wainman et al., , 2018)).Komlopteris remains prominent in the Late Jurassic Talbragar Fossil Fish Bed flora (Oxfordian-Tithonian: Walkom, 1921;Bean, 2006) of eastern Australia.Late Jurassic examples are also moderately common in the Hope Bay flora (upper Mount Flora Formation, Botany Bay Group) of the Antarctic Peninsula (Halle, 1913;Gee, 1989;Rees and Cleal, 2004;Birkenmajer and Ociepa, 2008) dated to 163.5 Ma (Oxfordian or younger: Scasso et al., 2022).By the Valanginian-Cenomanian, Komlopteris was confined to southeastern Gondwana (India, southwestern and southeastern Australia, and New Zealand), where it is always a minor component of fossil plant assemblages (Walkom, 1944;Douglas, 1969;Drinnan and Chambers, 1986;Maheshwari, 1986;Daniel, 1989;McLoughlin, 1996;McLoughlin et al., 2002).Thus far, there are no convincing records of Komlopteris from the relatively sparsely studied Jurassic-Cretaceous plant assemblages of Africa.
There are no convincing Upper Cretaceous records of Komlopteris from post-Cenomanian deposits of Gondwana, although major parts of that region are under-represented by fossiliferous continental strata of this age (Anderson and Anderson, 1983;Dettmann et al., 1992;McLoughlin et al., 2010).The youngest confirmed representative is K. cenozoicus, of Ypresian (Eocene) age, from western Tasmania.This also represents the last known occurrence of the group collectively known as 'Mesozoic seed ferns' globally (McLoughlin et al., 2008).The persistence of this lineage in austral temperate rainforest associations into the Cenozoic is matched by the apparent survival of Bennettitales into the Paleogene at two localities in southeastern Australia (McLoughlin et al., 2011).Bomfleur et al. (2018) argued that moist, temperate, high-latitude settings provided a refuge for these archaic lineages during the global expansion of angiosperms.A combination of abrupt cooling in southern Gondwana at the close of the Paleogene, combined with continued diversification of angiosperms, and aridification accompanying the northward movement of Australia and the rise of the Andes in Patagonia, may have all contributed to this lineage's ultimate demise in austral temperate forests.

Foliar morphology
Komlopteris is characterized by considerable inter-and intra-specific variation in leaf form, and even substantial heteromorphy in pinnae within individual leaves, which may represent developmental stages of the plant or a response to environmental factors, such as climate or illumination.Early Jurassic species (K.artabeae and K. nestarensis) are strictly pinnate and are not dissimilar to Triassic Kurtziana leaves.The largest pinnae of Middle to Late Jurassic K. purlawaughensis are markedly dissected and become fully bipinnate.However, both small leaves, and the apical portions of some large leaves retain a once-pinnate architecture with more-or-less entire-margined pinnae with decurrent bases.By contrast, the Late Jurassic K. antarcticus retains entiremargined pinnae throughout the leaf, although individual pinnules may vary markedly in shape from base to apex and even on either side of the rachis.At least one Early Cretaceous form (K. tiruchirapalliense) retains markedly dissected pinnae throughout the leaf.Other Early Cretaceous species (e.g., K. boolensis, K. victoriensis and K. khatangiensis) generally have mid-leaf pinnae that are lobed but not deeply so.The exception is K. indica, which has very slender pinnae with entire margins.The Eocene K. cenozoicus lacks lobing and has proportionally broader (ovate to broadly lanceolate) pinnae.Although a uniform trend is not evident throughout its stratigraphic range, representatives of Komlopteris appear to have adopted generally more dissected architectures in the Late Jurassic and returned to simpler pinnate forms with entire-margined pinnules towards the end of the generic range; no convincing latitudinal patterns could be identified.Variation of lobing of seed fern pinnae has been shown to reflect differences in leaf illumination, e.g., leaves with lobed pinnae representing shade leaves and entire pinne sun leaves, based on correlation with multiple cuticular characters (Barbacka and van Konijnenburg-van Cittert, 1998;Kustatscher et al., 2007).This raises the hypothesis that the variation of lobing in some Komlopteris species (e.g., K. boolensis, K. victoriensis) may also be a response to differential illumination within the canopy.This could be tested once specimens with cuticular preservation become available.The common preservation of Komlopteris as whole leaves with long petioles in densely matted layers (e.g., Zeba-Bano et al., 1979;this study), suggests that this lineage might have been deciduous.Most Gondwanan occurrences are in fern-rich assemblages that reflect humid conditions (Dettmann et al., 1992;McLoughlin et al., 2002) and from high paleolatitudes (>50 • S), where deciduousness may have been an advantage for surviving low light levels in winter.
Possible hair bases and papillae were reported to be present on K. artabeae (Artabe et al., 1991).However, they appear to be too sparse to significantly interfere with arthropod settling or movement over the leaf surface (cf.Pott et al., 2012) or to serve a function against other abiotic stressors, such as drought or salinity (see review by Kaur and Kariyat, 2020, pp. 21-22).
Although most Komlopteris leaves have pinnules with acute apices, they are never extended into spines or narrow projections (drip tips) that would specifically aid water shedding.The somewhat sunken stomata suggest some requirement for water loss regulation beyond simple closure of the guard cells, but strong stomatal protection can also be a feature of plants in everwet mire environments (Coulter, 1889;McLoughlin and Drinnan, 1996).
The cuticles of Komlopteris species are <8 μm thick and have minimal surface ornamentation apart from sparse short hairs in at least once species.Komlopteris artabeae has cuticle containing granular and laminar sublayers (Artabe et al., 1991).The cuticle of K. indica consists of a broad electron-lucent inner zone (up to 7 μm thick), a narrow electron-dense outer zone (680 nm thick), and remnants of epicuticular waxes (Bajpai and Maheshwari, 2000).The stomata are sunken below an epistomal chamber surrounded by a thickened rim, but there are no overhanging papillae that would suggest strong adaptation to water stress.Komlopteris indica is also known to host a range of epiphyllous and saprotrophic fungi (Bajpai and Maheshwari, 1987;Maheshwari and Bajpai, 1996;Bajpai, 1997) that suggest growth in humid environments.Overall, the leaf morphology suggests adaptation to humid conditions but with some need for water-loss regulation.

Reproductive structures
To date, no ovuliferous or polleniferous organs have been found attached to Komlopteris foliage.This remains one of the fundamental constraints to unequivocal assignment of these post-Triassic leaves to any established plant family.However, some dispersed cone-like and saccate organs known from the same strata are potential affiliates with this foliage.Among these, Stenorhachis scanicus (Nathorst) occurs in the same beds as the holotype of Komlopteris nordenskioeldii and is a lax cone-like structure with forked, cupulate appendages that, individually, are reminiscent of Umkomasia cupules (Nathorst, 1902).Rare pinnate or lax cone-like structures with unbranched cupulate appendages are also known from the type locality of Komlopteris purlawaughensis (White, 1981, fig. 250) but these have yet to be studied in detail.Townrow (1965) likened the pollen organ Stachyopitys cf.annularioides Shirley, described by Halle (1913) from the type formation of Komlopteris constricta (Mount Flora Formation, Hope Bay Antarctic Peninsula), to Pteruchus petasatus Townrow except that the former has a more circular receptacle (microsporophyll).Gee (1989) reassigned this microsporangiate organ to Kachchhia schopfii, but neither she, nor Halle (1913) could ascribe a definitive biological affinity to this reproductive organ.It remains possible that this taxon represents the pollen organ of Komlopteris constricta.Somewhat similar pollen bearing organs (pinnate microsporophylls with circular to oval receptacles bearing densely packed pendant pollen sacs on the abaxial surface) were found associated with Komlopteris leaves in Hettangian strata of Franconia, Germany, and originally assigned to Pteruchus septentrionalis (Kirchner and Müller, 1992), but later reassigned to Muelkirchium septentrionalis (Kirchner and Müller) Anderson (in Anderson et al., 2019).A similar dense cluster of pollen sacs (originally interpreted as an equisetalean leaf whorl) is copreserved with Komlopteris boolensis in Lower Cretaceous strata of southeastern Australia (McLoughlin et al., 2002, fig. 8A).The regular, though sparse, association of Pteruchus-like microsporangiate organs with Komlopteris foliage in both hemispheres reinforces the probable umkomasialean affinity of these leaves.

Higher affiliation
The architectural similarities between the foliage of Jurassic-Eocene Komlopteris species and Triassic leaves assigned to Dicroidium and Kurtziana are striking.Kurtziana, in particular, shares the unforked, generally pinnate leaf architecture, variable pinnule lobing, and odontopteroid venation pattern with Komlopteris.On this basis alone, we argue that Komlopteris and Kurtziana belong to the same plant lineage.Anderson and Anderson (2003) noted a weak co-preservational affiliation between Kurtziana foliage and the (?)cupulate, strobiloid fructification Matatiella in assemblages from the Upper Triassic Molteno Formation of South Africa.They used this perceived association to establish the order Matatiellales.However, the precise architecture of Matatiella fructifications (particularly the nature of the cupules and their insertion on the parent axis) remains ambiguous and this group might, alternatively, be placed in Umkomasiales.Discoveries of both reproductive organs and organically preserved leaves (especially of the Jurassic forms) from which cuticular data is retrievable will be necessary to resolve the higher affiliation of Komlopteris.

Insect damage
Subtle evidence of insect damage is represented on leaves of at least two Komlopteris species.The damage includes margin and hole feeding with prominent reaction rims (Plate IV 8-9;McLoughlin et al., 2015), and possible interveinal surface feeding (Plate IV 10), and oviposition scars (Plate VI,4,9).No Komlopteris leaves host intense arthropod damage, and it is unlikely that high levels of herbivory contributed significantly to the demise of this lineage.

Conclusions
Using mostly macromorphological criteria, supplemented by cuticular details for some forms, we distinguish 10 species of Komlopteris from Lower Jurassic to lower Eocene strata of Gondwana.The genus appears to represent a long-surviving lineage of Umkomasiales (or Matatiellales) that persisted beyond the end-Triassic extinction when most representatives of this order disappeared from the Southern Hemisphere.The foliage was resiniferous, hosts only modest insect damage, and occurs in matted leaf beds suggestive of a deciduous habit.Its micromorphological features and matted co-preservation with fern-rich leaf assemblages indicate adaptation to generally moist seasonal conditions.Adaptation to cool, humid, high-latitude forests may have buffered this lineage from abrupt intervals of global warming and facilitated its survival through the hyperthermal end-Triassic biotic crisis and the end-Cretaceous asteroid impact-induced extinction event.
Etymology: After the state of Victoria, Australia.
. Komlopteris indica is known only from Lower Cretaceous strata.