Amer. J. Bot. 7 l(2): 192-202. 1984. CUTICULAR ANATOMY OF ANGIOSPERM LEAVES FROM THE LOWER CRETACEOUS POTOMAC GROUP. I. ZONE I LEAVES1 Department of Paleobiology, Smithsonian Institution, Washington, D.C. 20560 ABSTRACT Angiosperm leaf cuticles from the oldest part of the Potomac Group reinforce previous paleobotanical evidence for a Cretaceous flowering plant diversification. Dated palynologically as Zone I of Brenner (Aptian?), these remains show a low structural diversity compared to later Potomac Group and modem angiosperms. All cuticle types conform to a single plan of stomatal construction that is unusual in its extraordinary plasticity: both the number of subsidiary cells and their arrangement vary greatly on a single epidermis, such that the stomatamight be classified as paracytic, anomocytic, laterocytic, and intermediate. Such stomatal diversity is uncommon in extant angiosperms but is known from a few Magnoliidae. Many species possess secretory cells comparable to the oil cells of modem Magnoliidae, and a few show the bases of probable uniseriate hairs. None of the cuticle types can be assigned to a single modem family, but several show similarities with Chloranthaceae and Illiciales. These results support the concept that subclass Magnoliidae includes some of the most primitive living angiosperms. WITHINTHE PAST 15 YEARS there has been a One largely untouched source of evidence is major re-evaluation of the Cretaceous flow- cuticular anatomy. While systematic studies of ering plant record and the role of fossils in both Mesozoic gymnosperms and Tertiary an- angiosperm phylogeny. Formerly, most bota- giosperm leaves have long relied on cuticle nists believed that fossils could provide little structure as an important complement to ex- direct evidence on angiosperm evolution, since ternal morphology (e.g., Harris, 1932, 1964; even the earliest remains were almost always Krausel and Weyland, 1950; Dilcher, 1974), related to modem families and genera (Axel- most studies of Early Cretaceous flowering plant rod, 1952, 1970). This view has been contra- leaves have neglected the cuticle and dealt only dicted by more recent studies of early angio- with features such as venation and shape. As sperm pollen and leaves, which show that most a consequence, I began a study of angiosperm of the older identifications are incorrect and leaf cuticles from the Lower Cretaceous Po- that Cretaceous flowering plants exhibit a pro- tomac Group of Virginia and Maryland, the gressive stratigraphic increase in both struc- most intensively studied sequence of early tural diversity and the percentage of forms re- flowering plant remains in North America latable to extant taxa (Doyle, 1969, 1973; (Brenner, 1963; Doyle, 1969, 1973; Wolfe et Muller, 1970, 198 1; Wolfe, Doyle and Page, al., 1975; Doyle and Hickey, 1976; Hickey and 1975; Doyle and Hickey, 1976; Hickey and Doyle, 1977). While at first cuticular remains Doyle, 1977). These results indicate that the were only known from the upper part of the Cretaceous was a period of major angiosperm Potomac Group (Subzone 11-B of Brenner, diversification and that paleobotanical studies probably middle to late Albian; Mersky, 1973; should continue to yield new evidence on the Upchurch, 1979), more recently a locality with group's early evolution. numerous organically preserved leaves was discovered from the lower part (Upchurch, 1978). This site, located on the James River Received for publication 8 February 1982; revision south of Richmond, Virginia, at Drewry's Bluff, accepted 24 August 1984. yields a pollen flora assignable to upper Zone I would like to thank James A. Doyle and Leo J. Hickey for helpful discussions and criticisms of the manuscript, I of Brenner (Aptian?), similar to one described and James U. McClammer for assistance in the field. Field by Doyle and Hickey (1976) from a compa- work was supported by grants from the Scott Turner award rable elevation at the south end of the exposure and Rackham School of Graduate Studies, University of (Doyle, pers. commun.). The anatomical fea- Michigan. This paper represents part of a doctoral disser- tures of these leaves and their possible evo- tation submitted to the Department of Botany, University of Michigan, Ann Arbor. lutionary implications are described below, a Current address: Department of Biology, Indiana Uni- systematic treatment involving both leaf ar- versity, Bloomington, IN 47405. chitecture and cuticular anatomy will be the February, 19 841 UPCHURCH -CRETACEOUS ANGIOSPERM LEAVES 193 subject of future publications. Comparisons will processing; the cuticles were then mounted in be made with contemporaneous dispersed cu- glycerine jelly on slides or in glycerine between ticle, later Potomac Group dicots, and extant two paraffin-sealed cover slips. Naturally ma- angiosperms. cerated cuticle, found at Dutch Gap, was mounted in glycerine on slides pending further MATERIALSAND METHODS- Angiosperm study. No dispersed cuticle was observed with leaves were collected from a 4-cm-thick bed SEM due to the small number of specimens of gray clay exposed at the northern end of available. Drewrys Bluff approximately 9 m above high tide level. Every fragment with angiospermous RESULTS-Over30 specimens belonging to venation was saved for laboratory analysis, and five distinct dicotyledonous leaf types are cuticular specimens were coated with acrylic known from the Drewrys Bluff locality. At least plastic to minimize fragmentation upon drying. 20 represent a new elongate leaf type that has Larger specimens were identified on the basis numerous glandular teeth, simple craspedod- of leaf architecture, using the classifications of romous secondary venation, and elongate ar- Fontaine (1 889), Ward (1 905), and Beny (1 9 11) eolation (D.B. Leaf Type # 1-Fig. 1, 2). This in addition to the informal morphological form does not closely resemble any previously groupings of Doyle and Hickey (1976) and described species of Early Cretaceous angio- Hickey and Doyle (1 977). Small fragments were sperm leaf. The others can be related to pre- identified on the basis of fine venational and viously described genera and species. Four cuticular similarities to larger specimens. Leaf specimens are assignable to Eucalyptophyllum architectural terminology follows Hickey oblongifolium Font. on the basis oftheir oblong (1973), while cuticular terminology follows shape, numerous irregularly spaced secondary Stace (1965), Dilcher (1974), and den Hartog veins, and prominent intramarginal vein (Fig. and Baas (1978). 4). Three leaf fragments show a random retic- The cuticles of the Drewrys Bluff leaves were ulate pattern of tertiary and higher order ve- studied with both light and scanning electron nation similar to that found in Ficophyllum microscopy (SEM). Light microscope (LM) Font. (cf. Ficophyllum-Fig. 3). Two speci- preparations were made by first removing the mens have the elliptical shape and marginal acrylic plastic with acetone, demineralizing serrations characteristic of the genus Celastro- overnight in HF, then macerating in Schulze's phyllum, but differ from all known Potomac solution (conc. HNO, and KClO, crystals) fol- Group species in having a low number of sec- lowed by 5% KOH. The freed cuticles were ondary vein pairs and in lacking a distinct pet- then stained in a 1% aqueous solution of Saf- iole (Celastrophyllum s~.-Fig. 5). Finally, one ranin 0 and mounted in glycerinejelly on slides. obovate fragment has the closely spaced sec- Preparations for SEM varied with the surface ondary veins and random reticulate pattern of studied. The outer cuticle surface was observed tertiary venation characteristic of the Celas- on leaf fragments which were cleaned in HF, trophyllum obovatum Font. complex from Bal- rinsed in 3 x distilled water, then glued to SEM timore, but differs by its much smaller size (cf. stubs. The inner cuticle surface was observed C. obovatum-Fig. 6). Certain characteristic on macerated cuticle which was rinsed in 3 x Zone I leaf types, such as pinnately lobate forms distilled water, then dried down to SEM stubs (Vitiphyllum Font.), elongate obovate leaves coated with Duco Cement in a chamber sat- with entire margins (Rogersia Font.), and re- urated with acetone vapors. All preparations niform leaves with basally congested second- were coated with gold for three minutes and ary veins (Proteaephyllum reniforme Font.), observed at 15 kv. are absent. Roughly a third of the leaf archi- Dispersed cuticle was analyzed to determine tectural variation from Zone I in the Potomac what proportion of Zone I cuticular diversity Group is present in this assemblage, since about was present in whole leaves. Samples of dis- 12 leaf types have been previously recognized persed cuticle were prepared from the Drewrys for this interval (Hickey and Doyle, 1977). Bluff angiosperm bed and a clay bed at Dutch Gap Canal of lower Zone I age that has yielded Upper cuticle-The upper cuticle ranges from rare angiosperm leaves (Hickey and Doyle, very thin in Celastrophyllum (Fig. 7) and cf. 1977). Plant fragments were obtained by mac- Ficophyllum (Fig. 8) to medium in D.B. Leaf erating the sediment in HF or Na,CO, and Type #1 (Fig. 10) and thick in Eucalyptophyl- sieving the slurry through fine-mesh screen. lum (Fig. 9). The upper epidermal cells in each Unoxidized remains, found at Drewrys Bluff, leaf type have a mixed shape and four to many were treated like materials from the whole anticlinal walls, which are predominantly leaves except that a centrifuge was used for straight in all groups except D.B. Leaf Type 194 AMERICAN JOURNAL OF BOTANY [Vol. 7 1 February, 19841 UPCHURCH -CRETACEOUS ANGIOSPERM LEAVES 195 #1, where they are curved. Stomata occur on Ficophyllum (Fig. 13). Complex striate sculp- the upper cuticles of Celastrophyllum (Fig. 7) ture is present in D.B. Leaf Type #1 and in and cf. Ficophyllum (Fig. 8). Their occurrence Eucalyptophyllum. In D.B. Leaf Type #1 there is correlated with a thin upper cuticle. are striations on the guard cells concentric to External sculpture on the upper cuticle is the stomatal pore, other striations that radiate smooth in every species except D.B.