AND TROCHANTHERA A LEPIDOTA GEN. NOV., FOSSIL ANGIOSPERM SP. INFLORESCENCE DOMINICAN AMBER IN George L Chambers Kenton 0. Poinar, Jr. Department ofZoology Department of Botany and Plant Pathology Oregon Oregon State University State University Oregon 9733 Oregon 9733 USA. Corvallis, U.S.A. Corvallis, 1, 1, ABSTRACT d by a capitate male infk Each composed 3-4 tacle. flower is of stamens, with dist dehiscence. The perianth represented by a short, appr< is of stem below the head and minutely puberulei is leafless Keywords: amber, Balanophoraceae, Dominican Republii INTRODUCTION Dominican amber is well known as a preservative agent of both animal and plant remains. Flowers may retain their structural details so clearly that they often can be assigned to extant genera (Poinar 1992) and & can be used in paleontological reconstructions (Poinar Poinar 1999). During an Dominican we investigation of floral remains in amber, discovered a complete, globose male inflorescence of uncertain affinity. Its morphology is compared with the families Balanophoraceae and Moraceae, with which many The shares essential characteristics. exceptional preservation allows it fossil's Due microscopic study of the flowers, which are here illustrated. to unique combination of features, the its new plant described below genus and is as a species. AND METHODS MATERIALS The fossil originated from mines in the northern mountain range (Cordillera Septentrional) of the Dominican Republic, between the cities of Puerto Plata and Santiago. Dating of Dominican amber controversial, with is & the youngest proposed age of 20-15 mya based on foraminifera (Iturralde-Vincent MacPhee 1996) and the 45-30 mya oldest as based on coccoliths (Cepek in Schlee 1999). Most of the amber secondarily deposited is Mamey in turbiditic sandstones of the Upper Eocene to Lower Miocene Group (Draper 1994). et al. mm mm The amber piece containing the fossil is rectangular in outline, measuring 15 in length, 10 in mm width and 8 in depth. The holotype deposited in the Poinar amber collection maintained Oregon is at State University. DESCRIPTION & Chambers Brown, G.O. Poinar, K.L. A.E. gen. nov. Type Species: Trochanthem lepidota G.O. Poinar, K.L. — unknown. Head com- Stem and minutely puberulent, length depressed-globose, Diagnosis. leafless total covered with persistent glabrous, flat-topped, closely connivent bracts in ± a hexagonal pattern. pletely Staminate flowers numerous, emerging over the entire head from between the edges of the bracts, each 3-4 consisting of a column of distally connivent filaments and anthers; anthers connivent, not fused, each with 2 unilocular thecae dehiscing extrorsely by longitudinal Perianth represented by a short, ovate, slits. unknown, from truncate, erose sheath enclosing base of emergent filament column. Pistillate flowers absent the described head or hidden beneath the bracts. & Trochanthera lepidota G.O. Poinar, K.L. Chambers A.E. Brown, sp. nov. (Figs. 1-6). Type: ] mm mm, mm; Description.— Staminate flower head: width 7.3 length 5.3 attached stem piece 2.6 long, mm 1.2 thick; peltate bracts of inflorescence 408(371-468) urn in diameter, anthers of emergent stamens 0-193 280(226-323) urn long, 113(96-145) urn wide; length of exposed filaments urn; perianth a short (up 5-12 60 pm) sheath enclosing base of filament column; pollen grains spherical, urn in diameter, out of to position for detailed view of exine morphology. — name Etymology. Genus name from Greek "trochos" ball, and "antheros" flowery. Species from Greek "lepidotos" scaly. We consider the distinctive unisexual inflorescence, consisting of a small, globose receptacle completely fossil's between which numerous simple staminate covered with flat-topped (presumably peltate) bracts, project — flowers, to have its closest analogs in two modern angiosperm groups family Balanophoraceae and family modern Moraceae, Dorstenieae (formerly Brosimeae). Both these groups have representatives in the tribe We Caribbean region. have not been able to identify a modern genus to which the fossil can confidently be The where assigned, although shares characteristics with both families. following discussion, Trochanthera it compared with each of these putative relatives, calls attention to the striking parallelisms in inflorescence is — morphology between two which otherwise markedly in habit small holoparasitic herbs the taxa, differ versus tropical forest trees. tall Balanophoraceae are a morphologically distinctive group of holoparasitic angiosperms occurring in New Harms tropical and subtropical regions of the Old and Worlds (Hooker 1856; Engler 1889; 1935, Hansen Nickrent 1997 onwards; Stevens 2001 onwards; Gonzales Nevling 1960; Kuijt 1969; 1976, 1980; from 2004). They are obligate subterranean parasites, usually forming a tuberous attachment to host roots which naked or bracteate inflorescence-bearing stems arise. By upward growth of the stem, the flowering portion is exposed at the soil surface or elevated on a fleshy, usually upright stalk, the plants at this stage often being described as "mushroom-like." Inflorescences are variously shaped, forming a loose or compact branched) raceme in some genera and a flat-topped or club-shaped to spherical head in others (Kuijt (tightly 1969, 125). The flowers are unisexual, the plants being dioecious or monoecious with the two sexes, in p. the latter case, either mixed or in separate parts of the inflorescence (or on separate heads from the same minute and extremely numerous, while the usually synangial flowers, in particular, are often tuber). Pistillate A 3-6-lobed anthers of staminate flowers take a great diversity of forms (Kuijt 1969, pp. 128, 130). peri- may anth most often present in staminate flowers and act as a protective covering over the anthers as the is A young pushes through kind protection such inflorescences, occurring, inflorescence the different for soil. New for example, in the World genera Helosis and Corynea, is a complete covering of peltate bracts, fitting tightly edge-to-edge in a geometric pattern and giving the capitulum a somewhat pineapple-like a m Poinar et Trochanthera lepidota gen. and sp. al., ETM| "^'MH P-* - 9 1 journal of the Botanical Research Institute of Texas 2(2) = tions of peltate bracts (B). Scale bar 130 urn. (Hooker 1856, pis. XIII, XV; Hansen 1980, fig. 11). The development and function of these bracts is well described by Hansen (1978). In the two genera mentioned above, the peltate bracts are deciduous following emergence of the capitulum, and the small staminate flowers then become evident as they emerge from a dense, mat-like covering of clavate hairs (Hooker 1856, pi. XIII). The tiny pistillate flowers remain embed- ded in these hairs with only the styles and stigmas exposed. among Hansen Trochanthera differs from the condition described for Helosis and Corynea, others (see 1978, 1980), in that a covering of tight-fitting bracts persists at anthesis, and the staminate flowers force their way out between the connivent edges of the bracts (Figs. 1-4, In the process, some flowers elon- 6). gate farther than others, and occasional anthers are broken and left behind, as it were. It seems remarkable some that the delicate tubular perianth (Figs. 3, 6) is also able to emerge, at least in flowers, in the process A of the staminal column forcing itself between the bracts. similar system for the emergence of staminate flowers from between protective bracts occurs in the monoecious species Balanophora wrightii, illustrated by Nickrent (1997 onwards), but in the male flowers are fewer on the head, they possess a broadly 3-lobed it some perianth, and the anthers form a synandrium with transverse dehiscence, as in other Balanophora and taxa. In his discussion of inflorescence structure in the related Asian Balanophora species B. harlandii B. wilderi, Hansen (1978) refers to the clavate bracts covering the pistillate portion of the heads as "spadicles," may although this usage be inappropriate (R.K. Eberwein pers. comm.). These intergrade with the broader We sub-peltate type in the staminate part of the heads. cannot be sure to what extent Hansen's term spadicles and we term even though 3-dimensional form cannot applies to Trochanthera, prefer the peltate bracts their be determined from the fossil. The stamens of Balanophoraceae are extremely diverse, especially the construction of the anthers. Anthers similar to those of Trochanthera are found in the South African genus Mystropetalon; that they is, are of an unspecialized, bithecal type, dehiscent extrorsely through two longitudinal Mystropetalon slits. happens also to be characterized by the putatively plesiomorphic trait of male and female flowers each with vestigial organs of the opposite sex (Nickrent et 2005; Kuijt 1969, 127), whereas pistillodes are not al. p. present in male flowers of Trochanthera. Compared with Helosis and Corynea, and with the Asian Balanophora New taxa mentioned above (Hansen 1972), Trochanthera anthers are notably larger than in the 2 World 's genera (see illustrations in Hooker 1856; Hansen 1980) and are extrorse rather than introrse (Figs. 1-4, 6). The anthers are not fused into synandrium that releases pollen through horseshoe-shaped or transverse a common may slits, pores, etc., as is in Balanophora (Hansen 1972). The filaments of the fossil be proximally fused, as in these other genera, but are only closely connivent distally (Figs. 2-4, 6). The short, tubular, sheathing, unlobed perianth of Trochanthera most resembles that of Corynea crassa On Hook. var. crassa (Hooker 1856, pi. XIII; Hansen 1980, fig. 11). the other hand, a prominently 3-lobed f. and type characteristic of staminate flowers in C. crassa var. sprucei also of Helosis, while in Mystropetalon is An and Balanophora species the perianth well developed and 3-6-lobed. unlobed, sheathing perianth is is present in staminate flowers of the Asian Rhopalocnemis (Harms 1935, fig. 163), but its anthers form a single multichambered synangium. The stalk or stem of Trochanthera is leafless, at least distally (Fig. This is comparable to various genera 1). of Balanophoraceae (Hooker 1856; Kuijt 1969; Hansen 1980; Nickrent 1997 onwards). More or less bracteate A remarkably similar form of staminate inflorescence to the above occurs in the Brosimum alliance of & Moraceae, tribe Dorstenieae (Berg 1989; Rohwer 1993; Datwyler Weiblen 2004; Zerega et al. 2005; for- merly Brosimeae of Engler 1889; Hutchinson 1967; Berg 1972). Congested, often globose flowering heads are frequent in Moraceae, as for example the members of tribes Castilleae (Olmedieae) and Dorstenieae may (Brosimeae) described and illustrated by Berg (1972). Heads be bisexual (one or more pistillate flow- mixed ers with the staminate ones) or unisexual (staminate or pistillate only, Berg 1989, fig. 11.3). These are borne on short, axillary peduncles, one to several per node (Woodson 1960; Berg 1972) and are in the same size range as Trochanthera. Because of the covering of peltate bracts in Trochanthera, between which the staminate flowers emerge, the most promising comparisons are with similarly bracteate Brosimum species. now Other genera having small, globose heads of male flowers, such as Coussapoa (Woodson 1960, in fam- & ily Cecropiaceae; Rohwer 1993; Datwyler Weiblen 2004), Helicostylis of tribe Dorstenieae (Olmedieae of & (Woodson Berg 1972), or Poulsenia 1960, assignable to tribe Castilleae per Datwyler Weiblen 2004) lack peltate bracts and have flowers with a 3-4-parted perianth and stamens either free and more numerous or 1-2 and often connate. The Trochanthera inflorescence appears completely staminate, whereas species of Brosimum often, but not always, have bisexual heads with a single, deeply embedded pistillate flower accompanying numerous may may staminate ones. However, the style of this female flower wither early, and the head then appear Woodson staminate. (1960, 130) quotes Standley as to the difficulty of finding "with ordinary herbarium p. We specimens. .even one pistillate flower." could not find an exposed style in the fossil. . 2-4 Flowers with stamens occur in Brosimum species of section Brosimopsis (genus Brosimopsis of Hutchinson 1967, p. 172), but a well developed, 4-lobed, pubescent perianth is present, and stamens are clearly separate with slender filaments (Berg 1972, 178). In section Piratinera (Berg 1972, 181) a perianth p. p. may is absent in the staminate flowers and be replaced by an enclosing bract (e.g., B. gaudichaudii, p. 185). However, there only stamen, which has a broad anther connective unlike Trochanthera. Other Brosimum is 1 species with a reduced or absent perianth and only stamen are found in section Brosimum (Berg 1972, 1 p. 169); they differ from Trochanthera in having either a peltate anther (B. alicastrum) or a tiny one only 0.2 mm in length and breadth (B. acutifolium). The peltate bracts of these and other Brosimum taxa differ from Trochanthera in being circular in outline, rather than closely connivent and more or less hexagonal (Woodson 1960, pp 131-133; Berg 1972, figs. 62, 66, 71, 73). In Brosimum subgenus Ferolia (Berg 1972, p. 188) the stamens, more than are divergent with slender filaments. The sheathing perianth, present, minutely if 1, if is ciliate apically. The peltate bracts are usually puberulent. The stamen, if 1, usually has the anther bent at 90°, with a thickened connective (all data from Berg 1972). In this comparison, therefore, the fossil cannot be accommodated in the genus Brosimum, and its description as a separate taxon is supported. Recognition of Twchanthera as a distinct genus will assist in future comparison of this unique fossil with extant members of Balanophoraceae and Moraceae, both of which are diverse in floral morphology and we modern genus inflorescence structure. In Balanophoraceae, prefer not to assign the plant to a variable may like Balanophora (Hansen 1972), since taxonomic circumscriptions change in the future, especially DNA A when methods are applied to the family. principal authority on the family, Bertel Hansen, was admit- tedly conservative in his taxonomic approach (Friis 2006), and he merged 9 later synonyms into the single & Weber genus Balanophora. Additionally, other generic unions have recently been proposed (Eberwein New when 2004), which strengthen Old and World taxonomic affinities in the family. Further, comparing between Twchanthera with representatives of Moraceae, again unclear that a close relationship exists it is it and modern Brosimum and particular genera, including allied taxa. angiosperm have advanced Molecular studies of the phylogenetic relationships of parasitic families rapidly in recent years (summarized in Barkman et 2007). Based on such work, the Balanophoraceae have al. & been firmly assigned to order Santalales (Nickrent Franchina 1990; Nickrent 1997 onwards; Stevens 2001 onwards), but their exact position was uncertain until the work of Nickrent et (2005). Based on 4 loci al. members from the plastid, nuclear, and mitochondrial genomes, this study clearly placed 3 selected of this and family within a clade containing 3 genera representative of Santalaceae, Olacaceae, Loranthaceae. Knowledge of the relationships of Moraceae with allied families, and of generic groups within the fam- & has been by recent molecular studies (Sytsma 2002; Datwyler Weiblen 2004; Zerega ily, clarified et al. et 2005). Changes in the familial and tribal assignments of certain genera mentioned above are based on al. Datwyler and Weiblen (2004) and Zerega et al. (2005). The latter paper involved a combined phylogenetic 26S and ndhF gent sequences and included proposals phylogeography and divergence times analysis of for mybp of the higher taxa. For Moraceae tribe Dorstenieae, a stem group age of 48 was estimated, while for Brosimum and 2 the crown group age was placed 28 mybp. Because the family place- closest relatives, at its may we ment of Twchanthera is uncertain, and our analysis not have exhausted all possibilities, can only point out that the fossil falls within an age of interest for presumed evolutionary diversification in both of abound and the families discussed above. Morphological parallelisms in the angiosperms, in this case, a more similar form of inflorescence in quite different families prevents us from proposing a exact assignment for this distinctive Tertiary fossil. ACKNOWLEDGMENTS New The authors thank the herbarium curators of the York Botanical Garden (NY) and Missouri Botanical We Garden (MO) for the loan of specimens of Balanophoraceae and Moraceae used in our study. are grateful We to P.K. Endress, R.K. Eberwein, and an anonymous reviewer for their comments and suggestions. also thank Strother providing us with samples of Brosimum. for J. REFERENCES DNA Barkman, T.J., J.R. McNeal, S-H. Lim, G. Coat, H.B. Croom, N.D. Young, and C.W. Depamphilis. 2007. Mitochondrial angiosperms and genomic chimerism suggests at least origins of parasitism in reveals in parasitic plants. 1 1 BMC Evol. Biol. 7:248. Berg, C.C. 1972. Olmedieae, Brosimeae (Moraceae). Neotropica 7:1-228. Fl. Berg, C.C. 989. Systematics and phylogeny of the Urticales. In: RR. Crane and S. Blackmore, eds. Evolution, sys- 1 tematics and history of the Hamamelidae. Vol. Clarendon Press, Oxford. Pp. 93-220. fossil 2. 1 Datwyler, and G.D. Weiblen. 2004. On the origin of the phylogenetic relationships of Moraceae from ndhF S.L. fig: .767-777. Bot. 91 J. r. 1173 nov. id sp. TA Donovan and Caribbean geology: an Jackson, eds. G, Mann, and 994. Hispaniola. Draper, Lewis. In: S. J.F. P. 1 129-150. introduction. The University of the West Indies Publishers' Association, Kingston, Jamaica. Pp. Eberwein, R.K. and A. Weber. 2004. Exorhopala ruficeps (Balanophoraceae): morphology and transfer to Helosis. Bot. 146:513-517. Linn. Soc. J. and Die Naturlichen Pflanzenfamilien Verlag Engler, A. 1889. Balanophoraceae. In: A. Engler K. Prantl, eds. III. 1. 243-263. Wilhelm Engelmann, Leipzig. Pp. Wilhelm and Die Naturlichen Pflanzenfamilien Verlag Engler, A. 894. Moraceae. In: A. Engler K. Prantl, eds. 1 1 . Engelmann, 66-98. Leipzig. Pp. Hansen 932-2005). Taxon 55:1055-1056. 2006. Bertel Friis, (1 1. and Henderson, D.W. Stevenson, Heald, eds. 2004. Balanophoraceae. N. Smith, Mori, A. S.V. Gonzalez, In: S.I. F. 41-43. Flowering plants of the Neotropics. Princeton University Press, Princeton, NJ. Pp. A monograph. Dansk The genus Balanophora and taxonomic Bot. Arkiv. 28: 1972. G. Forster. Hansen, J.R. B. Balanophoraceae. Malesiana 7:783-805. Hansen, 1976. B. Fl. Balanophoraceae. Neotropica 23:1-80. 1980. Hansen, B. Fl. Lophophytoideae and Scybalioideae (Balano- Balanophoroideae, Hansen, and Engell 978. Inflorescences in B. K. 1 77-1 phoraceae). BotTidsskr. 72:1 87. Harms, H. 1935. Balanophoraceae. In: A. Engler and H. Harms, eds. Die Naturlichen Pflanzenfamilien 16b. Verlag 296-339. Wilhelm Engelmann, Leipzig, pp. London On and of Balanophoreae. Trans. Linn. Soc. 22:1-68. Hooker, 1856. the structure affinities J.D. (Angiospermae). Clarendon Oxford. 1967.The genera of flowering plants Press, Hutchinson, J. Age and paleogeographic of Dominican amber. Science M.A. and Macphee. 1996. origin Iturralde-Vincent, R.D.E. 273:1850-1852. 1969. The biology of parasitic flowering plants. University of California Press, Berkeley. Kuijt, J. 47:303-308. 960. Flora of Panama: Balanophoraceae. Ann. Missouri Bot. Gard. Nevling, LI. 1 1997 onwards. The plant connection, <www.parasiticplants.siu.edu> parasitic Nickrent, D.I. and Molec. and CR. Franchina. 1990. Phylogenetic relationships of the Santalales relatives. J. Evol. Nickrent, D.I. 31:294-301. mushroom" ofthe"Maltese and F.E.Anderson. 2005. Discovery of the photosynthetic relatives Nickrent, J.R Der, D.I., BMC Cynomorium. 5:38 Evol. Biol. G.0. 992. amber. Stanford University Press, Stanford. Poinar, Life in Jr., 1 G.O. and 999. The amber forest. Princeton University Press. Princeton, NJ. Poinar, R. Poinar. Jr., 1 Rohwer, JG. 993. Moraceae. Kubitzki, J.G. Rohwer, and V. Bittrich, eds. The families and genera of vascular In: K. 1 438-453. plants. Springer-Verlag, Berlin. Pp. Schlee, D. 999. Das Bernstein-Kabinett. Stuttgarter Beitr. Naturk., Ser. C, 28. 1 more PR 2001 onwards. Angiosperm Phylogeny Website. Version June 2007 [and or less continuously 8, Stevens, http://www.mobot.org/MOBOT/research/APweb updated since]. SnsMA, Morawetz, M. Nepokroeff, Conti, M. Zjhra, J.C Hall, and M.W. Chase. 2002. Urticalean rosids: J.C. Pires, E. K.J., J. circumscription, rosid ancestry, and phylogenetics based on rbcL, trnL-F, and ndhF sequences. Amer. J. Bot. 89:1531-1546. 14-178. Woodson, 1960. Flora of Panama. Moraceae. Ann. Missouri Bot. Gard. 47:1 R.E. W.L Clement, Si. Datwyler, and G.D. Weiblen. 2005. Biogeography and divergence times in the mulberry Zerega, N.J.C., 37:402-416. family (Moraceae). Molec. Phylogen. Evol.