CANNABACEAE / MORACEAE WOOD TYPE 2 FIG. 15A–I Description— Growth rings distinct (Fig. 15A, B), ring width 0.2–0.9 mm. Wood ring-porous or semi-ring porous. Wide vessels in earlywood solitary and in pairs, tangential diameters average 182 (SD 29) µm, range 123– 247 µm; narrow vessels solitary, in radial multiples, and occasionally in clusters (Fig. 15A, B). Perforation plates exclusively simple(Fig. 15C); intervessel pitting crowded alternate, oval to polygonal (Fig. 15D, E), 8–11 – 13 µm in horizontal diameter; vessel-ray parenchyma pits with reduced borders, oval in outline (Fig. 15F). Helical thickenings observed in some of the narrowest vessels (Fig. 15C). Vessel element lengths difficult to measure because of the abundant tyloses (Fig. 15G), 205–333 µm (n=3). Axial parenchyma hard to distinguish in cross sections, probably vasicentric, confluent, and marginal (initial) bands, commonly in strands of 2–4 (Fig. 15D). Rays commonly 4–5 (–6)-seriate, uniseriate rays rare (Fig. 15G); multiseriate rays heterocellular with procumbent body cells and usually 1–3 (-4) marginal rows of square to upright cells (Fig. 15H), weakly defined sheath cells in a few rays (Fig. 15G); average multiseriate ray height 295 (SD=99) µm, range 128–539 µm; ~4-5 rays per mm. Non-storied. Solitary prismatic crystals occasional in upright ray cells (Fig. 15I) Specimen— UF 278-84906, estimated maximum diameter 11 cm. Occurrence— Dietz Hill (UF 278). Comments— It is difficult to determine if this is a vine or a ring-porous wood (shrub, sapling, or tree branch) with very narrow growth rings. The cross-sectional areas of woody vines typically have a high proportion of vessels (e.g., Carlquist 1991, Baas et al. 2004, Angyalossy et al. 2015), as is seen in this fossil. When ring-porous arborescent plants have narrow growth rings, those rings will consist mostly of earlywood and have a high proportion of vessels; consequently, samples from such areas can resemble woody vines (see Fig. 8 in IAWA Committee 1989). Figure 15. Caption on pg. 32. Moreover, because of the narrow rings the porosity type is not obvious. Other difficulties with this sample include: 1) determining latewood vessel arrangement because of the narrow growth rings, and 2) differences between fibers and axial parenchyma in transverse section are not obvious, so axial parenchyma distribution is difficult to determine, but given the parenchyma strands visible in longitudinal sections, we suspect marginal (wide initial bands) as well as vasicentric and confluent are present. Comparisons with modern woods— We began by searching InsideWood’s modern wood database for this combination of features: 1p (growth rings distinct), vessels not exclusively solitary (9a), exclusively simple perforation plates (13p 14a), intervessel pits alternate, polygonal in outline and medium- to large-sized (22p, 23p, 24a, 25a), vessel-ray parenchyma pits with reduced borders (30a), helical thickenings in narrower vessel elements (39p), tangential vessel diameter not narrow (40a, 41a), tyloses common (56p), non-septate fibers with simple pits (61p, 66p), larger rays 4–10-seriate (98p), less than 1 mm high (102a), not of two distinct sizes (103a), heterocellular with procumbent body cells (104a, 105a, 109a), tile cells absent (111a), storied structure, radial canals, and oil/mucilage cells absent (118–121a, 124–126a, 130a). This search returned six matches: Rhus chinensis Mill. (1768) (Anacardiaceae), Celtis laevigata (Cannabaceae), Maclura pomifera (Raf.) C.K. Schneid. (1906), Morus alba, Mo. nigra L. (1753), Mo. rubra, and Mo. serrata Roxb. (Moracee). We think it unlikely that this fossil has affinities with Rhus because ray structure and appearance are quite different. Both Celtis and Maclura include species that are ring-porous or vines. Consequently, we suggest that this wood has affinities with the urticalean group, and most probably with Moraceae. According to ter Welle et al. (1986), the climber Maclura cochinchinensis (Lour.) Corner (1962) ranges from Sri Lanka and India to China and Japan, and throughout the Malesian Archipelago; it has variable anatomy and includes samples that are semi-ring- to ring-porous and that have helical thickenings. Comparison with fossil woods— When we searched the Fossil Hardwood database using equivalent features in the Fossil Hardwood menu for that database, the results only included descriptions with most diagnostic features coded as unknown. Allowing one mismatch, returned 17 matches, 12 of which lack information on vessel-ray parenchyma pitting. Among the results were two reports of Moroxylon xinhuanensis Yin, Liu, and Cheng (2013), a distinctly ring-porous wood from the Neogene of China (Yin et al. 2013, Cheng et al. 2018). It is possible that UF 278-84906 is a ring-porous wood with narrow rings, but the diagnosis of Moroxylon Selmeier (1993) includes “groups of small vessels, especially in the latewood, forming a pattern of tangential or discontinuous oblique bands” and in this wood with its narrow rings that is not visible. Although it is likely this wood is Moraceae, we think it better to refer to it as Cannabaceae / Moraceae Wood Type 2 because of the possibility of a relationship with Celtis.

Published as part of Wheeler, Elisabeth A., Manchester, Steven R. & Baas, Pieter, 2023, A late Eocene wood assemblage from the Crooked River Basin, Oregon, USA, pp. 1-55 in PaleoBios 40 (14) on pages 29-32, DOI: 10.5070/P9401462457, http://zenodo.org/record/10913330