BLUMEA 21 (1973) 153—178 Comparative leaf anatomy of Kokoona and Lophopetalum (Celastraceae) W.T. Jansen and P. Baas Rijksherbarium, Leiden Contents Summary 153 Introduction 153 Materials and methods 154 Results 156 Generic and specific descriptions 156 in Development of the stomatal complex Kokoona ovatolanceolata! 173 Discussions 173 Stomata 173 Vascular of and midrib system petiole 175 Variability of anatomical characters in certain species 175 Diagnostic and taxonomic value ofsome leaf anatomical characters 177 General conclusions 177 Acknowledgements 178 References 178 Summary is The leaf anatomy of Kokoona and Lophopetalum described in detail. Separation of the two genera as effectuated differences vascular of the distal end of the by Hou (1963) is supported by in anatomy petiole, is in Other which invariably more complex Lophopetalum than in Kokoona. differential characters (p. ooo) ofthe of As based on leaf itis define only apply to part species Lophopetalum. anatomy impossible to groups the of closely related species within the genera because of a lack of mutual correlation between different characters. in shows leaf anatomical Lophopetalum particular some striking anatomical features such as a very variable and complex arrangement of vascular tissue in petiole and midrib, complex types of cyclocytic and and Most of these characters in of anisocytic stomata, crystarque cells, epidermal papillae. occur part the The genus only. variabilitybelow the species level ofcharacters such as epidermal wall thickness, number of cells dimension of cell absence of in subsidiary per stoma, guard pairs, presence or stomata upper epider- Because mis etc. is reported for some species. of the great range of anatomical variation it is impossible to identify species, using leaf anatomical characters only. Introduction In his revision of the family Celastraceae for the Flora Malesiana Hou considered the Kokoona and Arn. the genera Thwaites Lophopetalum Wight ex as distinct, on grounds of differences in floral and seed characters (1963: 264). Herewith he refutedKurz's suggestion (1870: 73) to reduce Kokoona to Lophopetalum. A natural classification of the wood of This carried the first author under *) paper reports on the results ofa post graduate study out by the super- vision of the second author. 153 BLUMEA VOL. No. 154 XXI, I, 1973 Kokoona and Lophopetalum coincides with Hou's botanical classification (Balan Menon 1964: 20). A study of pollen morphology by Hou (1969) resulted in the discovery of for both Kokoona another character supporting the generic status and Lophopetalum. This of leaf undertakenin order to have a more study anatomy was complete survey of of whether the characters these genera and to find out there is leaf anatomical support of for their separation. Furthermore, the results such a study might be helpful in the identification is for these of sterile material which very difficult genera (Hou 1963: 260 and A further contribution the ofthe leaf in 264). to knowledge anatomy Celastraceae was regarded as another justification of this detailed study. and Some data on the leaf anatomy of some species of Kokoona Lophopetalum can be found Stenzel in publications by (1893: 50, 70), Metz (1903: 27—28), Loesener (1892: Solereder and 192—193), (1908: 87—96), Metcalfe and Chalk (1950: 389—391). In the present paper descriptions and tables of characters ofall species of Kokoona andLophopeta- known and will be lum, at present (Hou, 1962, Byrnes, 1971), given. In order to gain some impression of leaf anatomical variation within a species (be it either ecologically or genetically controlled) 11 specimens of Kokoona littoralis and selected for These known show Lophopetalum javanicum were study. species are to much Of studied. Ofall leaf morphological variation (Hou, 1963). K. sessilis 6 specimens were remaining species at least i, mostly 2 specimens were studied. MATERIALS AND METHODS All material taken from dried in the was specimens present Rijksherbarium, Leiden, for a of Kokoona which was except specimen coriacea, kindly provided by Mr. J. P. M. Brenan, Keeper ofthe Herbarium, Royal Botanic Gardens, Kew. Because of the possibility of misidentification of material the leaves taken from non-flowering were always speci- with flowers. taken from mens They were a branchlet just below the basis of a flowering shoot in order to be certain that fully developed leaves were studied. mounted Permanent preparations, in Euparal, were made of transverse microtome sections of the central part of the laminaand both distal and basal end of the petiole and of free hand sections of the leaf surface. Transverse sections were stained with a safranin- haematoxylin mixture. Hand sections were stained with safranin only. This procedure of was followed for one specimen each species. Semi-permanent shdes, mounted in made glycerine jelly, were ofall other specimens of which transverse sections were made. made Cuticular preparations of all specimens were using equal volumes of hydrogen- peroxyde 30% and glacial acetic acid. The cuticles were stained with Sudan IV and mounted in glycerine jelly. the width of the cell Measurements were carried out as follows: guard pairs at their from measured with widest part and the length pole to pole were a linear eye piece micrometer in all cuticular preparations. Lamina and cuticle thickness were measured in distinction between glycerine jelly preparations. In most specimens no sharp was present uncutinised cell wall, cuticular layer, and cuticle (cf. Stace, 1965). Thereforethe total outer wall thickness of the in the anticlinal periclinal epidermal cells halfway between walls was the cuticle. indices calculated measured instead of the thickness of Stomatal were according to the formula: number of stomata S.I. — x 100 number of epidermal cells + stomata + subsidiary cells This formula deviates from the one given by Salisbury (1927) in the presence of the number of subsidiary cells in the fraction's denominator. I have not found whether the W. T. JANSEN & P. BAAS: Leaf anatomy of Kokoona and Lophopetalum 155 of of vascular tissue in midrib and in Kokoona and Fig. 1. Types arrangement petiole Lophopetalum. and 6—8 Types 1 and 3 are simple, the other types are complex. Type 1 and 3—5 are ‘open’, types 2 are ‘closed’. cells should be counted or taken with the stomata in subsidiary apart together any publi- cation in which the stomatal index is used. It seems logical to include the number of cells total subsidiary in the sum of cell types used in the fraction's denominator.Probably Salisbury worked with plants with anomocytic stomata which would account for the omission of subsidiary cells. Average values of stomatal sizes are based on 25 measurements. Stomatal indices are based in the of the lamina between midrib and leaf on 10 counts part margin at a 156 BLUMEA VOL. XXI, No. I, 1973 of least from either of them. These carried distance at 3 mm counts were out in cuticu- of lar preparations using a grid eye piece micrometer at a magnification X 1000. For the description of stomatal types the terminology of Van Cotthem (1970) was The and for followed. terms complex anisocytic complex cyclocytic were used in addition of where subdivisions of sub- types occurring in some species Lophopetalum one or more cells evident. As in between aniso- sidiary are shown fig. 24, transitional types cyclocytic, cytic, and the complex types occur (arrows). In many specimens two or more of the types and in the leaf. of fig. 24 intermediates occur same of the of different vascular the of Because occurrence complex systems in petioles Lophopetalum, the types were classified as in fig. I. The classification is artificial and was made for convenience only. Collection numbers and abbreviations, used according to Hou (1969: 103) are recorded before each specific description. Numbers marked with* were used for both transverse cuticular sections and cuticular preparations. All unmarked numbers were used for preparations only. Abbreviations used in the text: 1 = long S.V. = in surface view w = wide T.S. = in transverse section RESULTS The leaf anatomical data have been recorded using both descriptions and tables. The the generic descriptions are as comprehensive as possible, but in specific descriptions only those characters not suitable for listing in a table are mentioned. The specific descriptions and with the are therefore unbalanced and should be used together with table I or II generic description. The complex vascular systems in Lophopetalum are not elaborately described, but only reference to figures is made. GENERIC AND SPECIFIC DESCRIPTIONS Kokoona & Thwaites (Table I; Plate I, 2 3, III, 2; Fig. 24) S.V.: Cuticle smooth to granular. Unspecialized epidermal cells polygonal, with straight curved anticlinal walls which from thin thick. Thin areas of to slightly range to very adaxial cuticle or pit-like structures absent. Minor veins mostly not prominent in epi- dermis, not to fairly prominent in abaxial epidermis. Stomataconfined to abaxialepidermis, for one of Kokoona or except specimen ovatolanceolata (see specific description), cyclocytic with anisocytic, 2—9 subsidiary cells, 13—38 /im 1., 10—34 fim w., length-width ratios 1. Outlines of cells difficultto observe in cuticular 0.9—1.5, mostly ca. subsidiary prep- because of the cuticle the with thick arations being very thin here except in specimens anticlinal epidermal cell walls where only those adjacent to the guard cells have in- conspicuous cuticular flanges. Outer and inner stomatal ledges not conspicuous. Polar Granular T-pieces absent. material sometimes present in the stomatal pore. Subsidiary index 6.2 cells partly or almost completely underlying the guard cells. Stomatal —12.4. Cork warts present in varying amounts. dorsiventral or to isobilateral, thick. Unspecialized T.S.: Lamina tending 140—470 fxm epidermal cells rectangular, tall to flattened, those of abaxial epidermis usually lower than those of adaxial epidermis. Anticlinalwalls often tapering towards the mesophyll, in some the middle walls bulging in (Plate I, i).