lAW A Bulletin n.s., Vol. 1 (1-2), 1980 43

CYSTOLITHS IN THE SECONDARY XYLEM OF SPARATTANTHELIUM (HERNANDIACEAE)

by

Ben J. H. ter Welle Institute of Systematic Botany, University of Utrecht, Heidelberglaan 2, Utrecht, The Netherlands

Abstract Cystoliths were observed in the secondary the leaves of Ficus. As cystoliths have also been xylem of Sparattanthelium (Hernandiaceae). described for the leaves of Sparattanthelium Their shape, size, distribution and chemical (Solereder, 1899) and have been accepted as a composition is described. The systematic value feature of systematic value (e.g. Kubitzki, of cystoliths in the Hernandiaceae as well as in 1969), a supplementary examination of the general is discussed. wood of other species was undertaken.

Introduction Methods Cystoliths are internal, stalked outgrowths Both stained and unstained sections were of the cell wall that project into the cell lumen. studied. A double staining of Astrablue and They consist of cellulose and are impregnated Safranin was applied to differentiate between with calcium carbonate, they are irregular in unlignified and lignified cell components. The shape and sometimes fill up the cell completely. usual light-microscope, polarized-light and SEM­ Cystoliths may occur in parenchymatous cells techniques were used. Finally, the mineral in various parts of the plant, including even composition of the cystoliths was determined xylem and phloem rays. They are most fre­ with an Energy Dispersive X-ray Analyzer. quently found in the epidermis, in hairs or in special large cells which are termed litho cysts Materials (Fahn, 1967). In spite of the presence of cal­ According to Kubitzki (1969), the genus cium carbonate, this type of cell inclusions was Sparattanthelium comprises 13 species, lianas not included in the comprehensive study on or shrubs, restricted to the neotropics. For this the occurrence and shape of crystals made by study wood samples of 5 species were available. Chattaway (1955, 1956). Others, like Metcalfe In addition samples of other Hernandiaceae & Chalk (1950), Esau (1963) and Fahn (1967) and of a few taxa of the Opiliaceae and of the also described cystoliths as a separate category, Acanthaceae were investigated. apart from crystals. The occurrence of cystoliths in various parts Material seen: of the plant, but especially in the leaves, is re­ Hernandiaceae - Gyrocarpus americanus stricted to a few families (Metcalfe & Chalk, Jacq. ssp. americanus, FHOw 70 I, Sri Lanka; 1950; Pireyre, 1961), particularly in Cannabi­ Pulle 3217, New Guinea; Hernandia didymantha naceae, Moraceae and Urticaceae. Donn., Uw 10383, P.H. Allen, Costa Rica, ex Cystoliths in the secondary xylem were up USw 30169; H. guianensis Aubl., Uw 380, till now only known in a single family, Opilia­ Stahel 380, Suriname; Uw 1509, Lanjouw & ceae (Record, 1925, 1927; Kanehira, 1921; Lindeman 1547, Suriname; H. peltata Meissn., Metcalfe & Chalk, 1950). Record described B.W. 7670, New Guinea; Illigera pentaphylla cystoliths as 'deposits of calcium carbonate in Welw., Uw 22123, Versteegh & den Outer 588, special structures in the rays of 7 genera, viz. Ivory Coast; Sparattanthelium aruakorum Agonandra, Cansjera, Champereia, Lepionurus, Tutin, Uw 12148, van Donselaar 3754, Suri­ Meliantha, Opilia and Rhopalopilia'. According name; S. borororum Mart., Uw 24192, R.C. to Metcalfe & Chalk (1950), Agonandra is the Gill 43, Equador, ex SJRw 36125; S. glabrum only genus which lacks the cystoliths. Rusby, Uw 24073, Schunke 2539, Peru, ex By chance I observed in a section of Sparat­ USw 41876; S. guianense Sandw. Uw 24087, tanthelium deposits which looked like a bunch Fanshawe, F.D. 3938, Guyana, ex FHOw of grapes. At first sight, these deposits appear­ 13266; S. wonotoboense Kosterm., Uw 24191, ed to be similar to the well-known cystoliths in A.C. Smith 3390, Guyana, ex SJRw 35917.

Downloaded from Brill.com09/26/2021 07:44:50AM via free access 44 IAWA Bulletin n.s., Vo1.1 (1-2),1980

Fig. 1-9. Light-microscopical appearance of cystoliths in the secondary xylem. -- I, 2 & 3: Spa­ rattanthelium aruakorum Tutin. (Uw 12148). Cross section, x 35, x 90 & x 220 resp. -- 4: Sparat­ tan thelium guianense Sandw. (Uw 24087). Tangential section, x 35. -- 5: Sparattanthelium wono­ toboense Kostenn. (Uw 24191). Radial section, x 35. -- 6 & 8: Sparattanthelium aruakorum Tutin. (Uw 12148). Tangential and radial section resp., x 220. --7: Trichantheragigantea (Humb. & Bonp1.) Steud. (Uw 5476). Radial section, x 220. -- 9: Champereia manillana Merrill. (Jacobs 7611). Radial section, x 220.

Downloaded from Brill.com09/26/2021 07:44:50AM via free access IAWA Bulletin n.s., Vol. I (1-2), 1980 45

Opiliaceae - Agonandra silvatica Ducke, Uw Chemical composition - A thorough exami­ 3402, Lindeman 4995, Suriname; Uw 6806, nation of the chemical composition of the cys­ Schulz 8329, Suriname; Champereia manillana toliths is beyond the scope of this study. How­ Merrill, Jacobs 7611, Philippines; Opilia cf ever, in order to establish the cystolith-nature amentacea Roxb., Jacobs 9600, Papua New of the observed inclusions, some information Guinea; O. celtidifolia (Guill. & Perr.) End!. ex on this subject is necessary. Therefore, sections Walp., Breteler 2261, Cameroun. were double-stained with Safranin and Astrablue Acanthaceae - Aphelandra tetragona (Vah!) according to the method recommended by Von C.G. Nees ab Esenb., Uw 10932, Florschiitz & Aufsess (1973) to differentiate between lignin Maas 2431, Suriname; Trichanthera gigantea and cellulose. In both transverse and longitudi­ (Humb. & Bonpl.) Steud., Uw 200 I, Lanjouw nal sections the cystoliths are stained bright & Lindeman 3433, Suriname; Uw 5476, Geys­ blue. Nevertheless the intensity of the blue col­ kes s.n., Suriname. our shows various grades. There is no trace of red stained material in the cystoliths which in­ Results dicates the absence of lignin. If pedicels do oc­ cur, they remain uncoloured. Sparattanthelium Subsequently SEMjEDXA techniques were Cystoliths were found in all species of Spa­ used to determine the mineral components of rattanthelium examined. Their shape, size and the cystoliths. The only element identified is distribution was variable, both within a sample calcium (Fig. 14, 15 & 16). In polarized light and between different samples. the cystoliths show little or no birefringence. Shape - Slightly oval and oblong in S. aruakorum, S. glabrum and S. guianense. In The SEM investigations gave some indication for the presence of a thin layer of unidentified S. borororum and S. wonotoboense only ob­ long cystoliths occurred. nature between the wall of the lithocyst and the cystoliths (Fig. 13). This layer stained blue Size - The oval cystoliths measure from c. with the Astrablue-Safranin staining. Although 45 /lm up to 60 /lm and the oblong ones from cystoliths usually are not classified as crystals, 100 x 40 to 160 x 60 /lm. In S. wonotoboense, this layer is comparable with integuments de­ a cystolith of 240 x 64/lm was observed. They scribed for crystals in the secondary xylem of usually fill up the cell completely. e.g. Flacourtiaceae (Miller, 1975). Structure - The surface is always more or The occurrence of the cystoliths in Sparat­ less granular, but without the sharp points tanthelium is restricted to a certain part of the characteristic for crystals like druses. Only oc­ stem. Shutts (1960) already noticed in this casionally the cystoliths show faint radiating genus the presence of two distinct types of ves­ lines (Fig. 10). On cross sections concentric sels in one sample, indicated by diameter and lines around the centre of the cystoliths can of­ element length. On a transverse section these ten be noticed (Fig. 2 & 3). two different types are very clear. It is remark­ Distribution - Average values for the num­ able that the cystoliths often occur at the place ber of cystoliths per square mm (cross section), where the diameter of the vessels changes ab­ based on 10 measurements, were computed. ruptly. The averages vary from 2 to 29, but in general According to Kubitzki (1969), cystoliths oc­ there are less than 10 per square mm (Fig. I). cur in the leaves of all species of Sparattanthe­ Lithocysts - The shape of these cells com­ lium and Gyrocarpus, together constituting the monly resembles that of the cystoliths, since the latter fill up the cell completely. The out­ tribe Gyrocarpoideae. However, contrary to line of the lithocysts in longitudinal sections is the results obtained with the secondary xylem rounded (oval or oblong, Fig. 4 & 5). and rath­ of Sparattanthelium, no cystoliths were observ­ er irregular in cross sections (Fig. 3). ed in the wood of Gyrocarpus. In view of the fact that occurrence of cystoliths is restricted Cystolith-containing cells in Sparattanthelium to a small number of families (Solereder, 1899, are always part of a parenchyma strand, but they are much larger than the other parenchy­ 1908; Pireyrc, 1961), it seemed worthwhile to examine the wood of some families from the ma cells. In general the cells show about the leaves of which cystoliths have been reported. same size and shape as oil or mucilage cells, well known from the wood of e.g. Lauraceae. Acanthaceae Pedicel - The presence of a pedicel is of es­ Cystoliths are abundant in the ray cells of sential importance for the development of cys­ two samples examined of Trichanthera gigantea. toliths (Fahn, 1967). However, in the material The ray cells are not enlarged and similar to the examined pedicels were rarely observed and other ray cells. Cystoliths occur in both square short. and upright cells. As a result their shape varies

Downloaded from Brill.com09/26/2021 07:44:50AM via free access 46 lAW A Bulletin n.s., Vol. 1 (1-2), 1980

Downloaded from Brill.com09/26/2021 07:44:50AM via free access IAWA Bulletin n.s., Vol. I (1-2),1980 47 from more or less square or round to very of­ fIrst stage of its formation, the body is incrust­ ten oblong, respectively c. 40 !.1m and from 120 ed or impregnated with amorphous calcium x 30 to 180 x 30 !.1m (Fig. 7). They never com­ carbonate (Frey-Wyssling, 1935; Esau, 1963; pletely fill the cell lumen. In polarized light the Pireyre, 1961). The few tests used in my inves­ cystoliths are birefringent or not. Pedicels are tigation to obtain information on the chemical rare or absent, but if present they are very faint. composition of the cystoliths in Sparattanthe­ The cystoliths are very fragile. A pretreatment lium are in agreement with these data from the of boiling in water completely destroyed most literature. All evidence clearly demonstrates of them. In the particular case of the Acantha­ that the structures observed in the secondary ceae, the following method turned out to be xylem of various Sparattanthelium-species are useful: use cold water when sectioning, remove real cystoliths. air by boiling in water for less than 20 seconds and mount in glycerin. Finally, the best results The occurrence of cystoliths in various parts were obtained using sections of 30-40 !.1m of the plants has always been considered as a thick. useful character in identification and it has of­ ten proved to be of systematic value (Solereder, Boraginaceae 1899 & 1908; Metcalfe & Chalk, 1950; Fahn, Seven species of Tournefortia and Cordia 1967; Pireyre, 1961). Their presence in the were examined. Cystoliths were not observed leaves of Gyrocarpus and Sparattanthelium is in the secondary xylem. of systematic value (Solereder, 1908; Kuhitzki, Discussion 1969), as it is one of the characters used to di­ The shape of the cystoliths has been describ­ vide the family of the Hernandiaceae into two ed as very variable (e.g. Sole reder, 1899, 1908), tribes. As already mentioned before, cystoliths but as a result of her investigations, Pireyre occur in the wood of all examined samples of (1961) limited the number of types to three: Sparattanthelium. Although size, shape and dis­ a. round cystoliths; tribution are variable, their occurrence is con­ b. oblong cystoliths the pedicel of which is at­ stant and therefore they constitute a character tached at one end of the long axis, or oblong of systematic value. The cystoliths were not cystoliths the pedicel of which is attached observed by Shutts (1960) who studied the sec­ perpendicular to the long axis; ondary xylem of the Hernandiaceae. In Sparat­ c. cystolith-like structures. tanthelium he noticed the occurrence of 'secre­ There is no doubt that the cystoliths observed tory cells, swollen, common in vertical paren­ in Sparattanthelium belong to type b. This con­ chyma'. Before sectioning he softened the clusion is strenghtened by the occurrence of a wood samples in a I : I solution ofliydrofluoric thin-walled layer between cystoliths and the acid and 50 % ethyl alcohol for three to six parenchyma cell wall. According to Pireyre weeks. As boiling in water sometimes destroys (1961) such an integument is usually connect­ the cystoliths, it seems plausible that his pre­ ed with cystoliths of type b. Pedicels are rare treatment completely dissolved the cystoliths. and usually weakly developed in both stained As Shutts (1960) described these 'swollen se­ and unstained sections of Sparattanthelium. cretory cells' not only for Sparattanthelium However, after a pretreatment with hydro­ but also for Hernandia, I studied the secondary chloric acid the calcium carbonate dissolves xylem of some samples of Hernandia myself. and in some cystoliths the pedicels become Neither cystoliths nor swollen parenchyma clearly visible. Cystoliths of type b often show cells were observed. This result is rather striking short or invisible pedicels (Pireyre, 1961). since these cells were not only recorded by In general, the pedicel and the cystolith Shutts but also by Garratt (1933) and Record body are composed of callose, cellulose and (1944) but not, however, by Brazier & Frank­ pectin (Fahn, 1967; Pireyre, 1961). After the lin (1961).

Fig. 10-16. SEM and Energy Dispersive X-ray Analyzer appearance of cystoliths in the secondary xylem of Sparattanthelium aruakorum Tutin. (Uw 12148). -- 10: Cross section through a cysto­ lith showing radiating lines, x 630. -- II: Radial section. At the top of the cystolith a short pedi­ cel can be seen; x 320. -- 12: Radial section showing the parenchymatous cell wall of the idio­ blast, x 320. -- 13: Radial section. A thin layer between the cystolith and the cell wall is indicated by the arrow. -- 14, IS & 16: Radial section. Two cystoliths and an empty idioblast. In Fig. 16 the X-ray element map shows the location of calcium in these cystoliths.

Downloaded from Brill.com09/26/2021 07:44:50AM via free access 48 IAWA Bulletin n.s., Vol. I (I~2), 1980

The positive results of my own observations sues. I. Trop. Woods 102: 55~74. on the occurrence of cystoliths in Trichanthera - 1956. Crystals in woody tissues. 2. Trop. (Acanthaceae) might be an indication of a Woods 104: 100-124. more widespread distribution of these deposits. Esau, K. 1963. Plant anatomy. Wiley & Sons, As most of the cystoliths in this genus are de­ New York. stroyed by boiling in water, a careful prepara­ Fahn, A. 1967. Plant anatomy. 2nd ed. Perga­ tion is essential. Finally, a more detailed exami­ mon Press, Oxford. nation of the crystals described by Chattaway Frey-Wyssling, A. 1935. Die Stoffausscheidung (1955) as 'druses, attached to the cell wall by a der hoheren Pflanzen. Springer, Berlin. peg' should be considered. These crystals occur Garratt, G.A. 1933. Bearing of wood anatomy in Celtis-species (Ulmaceae). This family to­ on the relationships of the Myristicaceae. gether with the Urticaceae. Moraceae and Can­ Trop. Woods 36: 20~44. nabinaceae constitutes the order Urticales. Es­ Kanehira, R. 1921. Anatomical characters and pecially in this order cystoliths are a common identification of Formosan woods. Bureau character for the leaves. of Productive Industries, Taihoku. Kubitzki, K. 1969. Monographie der Heman­ Acknowledgements diaceae. Bot. Jahrb. 89: 78~209. I wish to thank Dr. J. Burley (Commonwealth Metcalfe, C.R. & L. Chalk. 1950. Anatomy of Forestry Institute, Oxford), Dr. R.H. Eyde the dicotyledons. Clarendon Press, Oxford. (Smithsonian Institution, Washington) and Dr. Miller, R.B. 1975. Systematic anatomy of the R.C. Koeppen (U.S. Forest Products Laborato­ xylem and comments on the relationships ry, Madison) for the loan of wood samples. Dr. of Flacourtiaceae. J. Am. Arb. 56: 20~102. W. Berendsen and Mr. J. Pieters (Dept. of Mole­ Pireyre, N. 1961. Contribution a I'etude mor­ cular Cell Biology, Utrecht) are acknowledged phologique, histologique et physiologique for their competent assistance with the SEM des cystolithes. Rev. Cyt. et BioI. Veget. and Energy Dispersive X-ray Analyzer. The 23: 93~320. photographic plates were kindly prepared by Record, S.J. 1925. Cystoliths in wood. Trop. Mrs. A. Kuiper and T. Schipper. Last but not Woods 3: 1O~12. least thanks are due to Mr. L.Y. Westra for cor­ - 1927. Occurrence of calcium carbonate de­ recting the English text. posits in wood. Trop. Woods 12: 22-26. -- 1944. Keys to American woods: XV. Fibres References with conspicuous bordered pits; XVI. Aufsess, H. von. 1973. Mikroskopische Darstel­ Woods with oil (or similar) cells. Trop. lung des Verholzungsgrades durch Farbe­ Woods 80: 10-15. methoden. Holz Roh- u. Werkst. 31: 24~ Shutts, C.F. 1960. Wood anatomy of Heman­ 33. diaceae and Gyrocarpaceae. Trop. Woods Brazier, J.D. & G.L. Franklin. 1961. Identifica­ 113: 85~123. tion of hardwoods. For. Prod. Res. Bull. Solereder, H. 1899 & 1908. Systematische Ana­ no. 46. tomie der Dicotyledonen & Erganzungs­ Chattaway, M.M. 1955. Crystals in woody tis- band. F. Enke, Stuttgart.

Downloaded from Brill.com09/26/2021 07:44:50AM via free access