Ibericarpus E.M.FRIIS, P.R.CRANE, K.R.PEDERSEN, M.M.MENDES et J.KVAČEK, 2022

Friis, Else Marie, Crane, Peter R., Pedersen, Kaj Raunsgaard, Mendes, Mário Miguel & Kvaček, Jiří, 2022, The Early Cretaceous Mesofossil Flora Of Catefica, Portugal: Angiosperms, Fossil Imprint 78 (2), pp. 341-424 : 370-373

Genus Ibericarpus E.M.FRIIS, P.R.CRANE, K.R.PEDERSEN, M.M.MENDES et J.KVAČEK gen. nov.

Type. Ibericarpus cuneiformis E.M.FRIIS, P.R.CRANE, K.R.PEDERSEN, M.M.MENDES et J.KVAČEK gen. et sp. nov.

Plant Fossil Names Registry Number.

PFN0002790 (for new genus).

Etymology. From the Iberian Peninsula where the fossil was collected.

G e n e r i c diagnosis. Pistillate structure with numerous, densely spaced carpels borne in a spiral arrangement along a slender axis, with no remains of perianth parts or stamens. Carpels obconical to pyriform, sessile, uniovulate. Style lacking, stigmatic region slightly bulging. Fruit indehiscent. Epidermal cells of fruit with isodiametric facets. Ovule/seed obovate with micropyle pointing towards the base of the carpel. Embryo tiny. Seed coat unspecialized.

C o m m e n t s o n t h e g e n u s. There are no scars from bracts, perianth parts or stamens on the axis below the carpels, and there are no traces of a perianth or stamens associated with the individual carpels. The structure of the carpel is uncertain, but its shape and the lack of an obvious suture suggests that it is ascidiate.

Interpreting the floral structure is not straightforward. One possibility is that the carpel-bearing axis of Ibericarpus cuneiformis is a simple, unbranched inflorescence bearing numerous ebracteate pistillate flowers, each consisting of only a single carpel. Under this interpretation, Ibericarpus cuneiformis shows some similarity to floral structures of Chloranthaceae . Flowers of Chloranthaceae have simple, typically naked flowers, that are borne in elongated inflorescences and the carpels are ascidiate and uniovulate without a style. Among extant Chloranthaceae , Hedyosmum and Ascarina also have unisexual flowers. However, in extant Chloranthaceae the flowers are typically in the axil of a distinct bract and only the staminate flowers of Hedyosmum are ebracteate. Because no bracts are present associated with the individual carpels in Ibericarpus cuneiformis , we regard the inflorescence interpretation as unlikely. This conclusion is also supported by the unspecialized seed coat of Ibericarpus . In all chloranthoid seeds so far described from the Cretaceous, the seed coat is endotestal with crystalliferous endotestal cells.

An alternative interpretation of Ibericarpus is that the fruiting structure is derived from a pistillate, perhaps naked, flower with an apocarpous gynoecium of numerous free carpels. Among extant angiosperms, taxa with an apocarpous gynoecium of many carpels arranged spirally along a long, slender floral axis occur in Kadsura and Schisandra ( Schisandraceae , Austrobaileyales ), in Magnoliaceae (Magnoliales) , and also in Galbulimima F.M.BAILEY ( Himantandraceae , Magnoliales ).

Flowers of Magnoliaceae differ from those of Ibericarpus cuneiformis in being bisexual, and typically with well developed, often leathery, perianth parts that leave distinct scars after flowering. The carpels also have a distinct style, and each contains two or more ovules. Flowers of Galbulimima also differ from those of I. cuneiformis in being bisexual, but they are more like the fossils in being naked and in having uniovulate carpels that lack a style. Fruits of Galbulimima are drupes, while those of Ibericarpus are nuts or one-seeded berries. Carpels in both Magnoliaceae and Himantandraceae are plicate or intermediate plicateascidiate.

If the carpels of Ibericarpus cuneiformis are correctly interpreted as ascidiate then in this feature they are more similar to the carpels of Austrobaileyales . Flowers of Kadsura and Schisandra are similar to Ibericarpus cuneiformis in their unisexual organization as well as having carpels that lack a style, but flowers of both extant genera have a distinct perianth and also have one to several ovules per carpel.

Against this background, while we think that Ibericarpus cuneiformis is most likely an elongated receptacle bearing numerous fruitlets, and while a relationship to extant Austrobaileyales seems the most likely possibility, I. cuneiformis cannot be included securely in any extant angiosperm family or order.

Among the fossil floral structures that have a multicarpellate and apocarpous gynoecium, species of Atlantocarpus E.M.FRIIS, P.R.CRANE et K.R.PEDERSEN from the Early Cretaceous floras of Puddledock, Virginia, USA, and also Buarcos and Vale de Água, Portugal ( Friis et al. 2020a), are the most similar to Ibericarpus . Atlantocarpus has a very long receptacle and apparently ascidiate carpels that are uniovulate and lack a style. However, fossils of Atlantocarpus have distinct remains of floral organs below the carpels and the receptacle is obconical, rather than slender and stalk-like as in Ibericarpus .

Floral structures of Choffaticarpus compactus E.M.FRIIS, P.R.CRANE et K.R.PEDERSEN first described from the Torres Vedras mesofossil flora ( Friis et al. 2019a), and Anacostia ? sp. from the Puddledock flora of eastern North America ( Friis et al. 2020a), are also similar to Ibericarpus in having tightly packed carpels spirally arranged along an elongated receptacle. However, Anacostia ? sp. differs in having a distinct joint between pedicel and flower with remains of other floral parts below the carpels and Choffaticarpus compactus differs having strongly compressed carpels with a distinct ventral depression (see above).