A Study of Stomatal Structure 10 Pandanaceae

Many species of Pandanaceae have unspecialized stomata. In species of Pandanus there is a range of stomatal structure involving increasing elaboration of papillae arising from subsidiary and neighbouring cells so that guard cells, which are otherwise constant in structure throughout the family, are increasingly provided with a canopy of papillae which, together with sinking of the stomatal apparatus, produces an outer stomatal chamber. In the most specialized stomata this chamber is itself virtually occluded by lobed papillae. Freycinetia may show significant differences from Pandanus. This linear trend of specialization can often be partly or wholly demonstrated on a single individual because stomata are always unspecialized on the first scalelike leaves of each renewal shoot, thereafter being increasingly specialized on subsequent leaves until the maximum potential elaboration characteristic of the species is realised on foliage leaves. This is briefly discussed in relation to taxonomy, physiology, and ecology. THIS ARTICLE is the partial result of a recent study of the systematic anatomy of the Pandanaceae in which limited material of all three of its genera, Freycinetia, Pandanus, and Sararanga, has been examined. Insufficient material has been studied to permit worthwhile comment about the diagnostic value within the family of microscopic structures, except that anatomy is relatively uniform throughout the Pandanaceae and emphasizes its naturalness. In the survey a considerable range in stomatal structure was found, especially in Pandanus, a range not fully described by previous observers like Kofler (1918), Solla (1884), or Solereder and Meyer (1933). Observations were extended to indicate the range of stomatal structure throughout a single plant, whereupon it was discovered that, in certain instances, the full range of stomatal variation within the genus could be demonstrated within a single individual. This can be explained when the growth habit of Pandanus is understood. It was felt that a general summary of stomatal structure in Pandanaceae would be a significant contribution to the literature on monocotyledonous anatomy. The present account deals 1 Fairchild Tropical Garden, Miami 56, Florida. Manuscript received September 27, 1963. 38 largely with Pandanus, notes on F.,-eycinetia and Sararanga being inc.luded when they show outstanding features. Material of some 50 species has been examined during the course of this study, which-in view of the richness of Pandanus, estimated at 500-600 species, and of Freycinetia, at 180-200 species (St. John, 1960) -indicates how provisional this study is. It is not, therefore, intended as an exhaustive survey of the subject, but gives a clear indication of the known range of stomatal structure in Pandanaceae and points the way to more intensive study. . MATERIAL AND METHODS The survey is largely based on herbarium material of Pandanus supplied from specimens in the Bernice P. Bishop Museum, Honolulu, Hawaii, by Dr. B. C. Stone, the material being selected to represent a wide range of subgenera. It was supplemented by material collected by myself and others in the wild, and by material in cultivation at the Fairchild Tropical Garden and the U. S. Department of Agriculture Experiment Station, Old Cutler Road, Miami, Florida. Slides and specimens in the collections of the Jodrell Laboratory, Royal Botanic Gardens, Kew, Stomatal Structure in Pandanaceae-TOMLINSON England, have also been examined. The cultivated material was in many instances unnamed and was used largely to confirm observations on named species and to provide large samples for studies in stomatal variability. Observation on unnamed species, with few exceptions, has not been tabulated in this account. A full list of named material and sources is given in the appendix. Only simple technical methods have been used. Dried material was restored to a consistency suitable for anatomical study by boiling for several minutes in water to which a little Clorox was added. This revived material was stored in formalin-acetic-alcohol and proved quite satisfactory for anatomical study. Some distortion which could not be eliminated by the reviving technique did not obscure anatomical details. This point is emphasized, because it means that stomatal studies in Pandanaceae can proceed satisfactorily using herbarium material. Fresh material has ..been examined either unfixed or fixed in 'formalin-acetic-alcohol. Preparations showing the surface layers in surface view were made by the simple scraping technique employed at the Jodrell Laboratory and described in detail by Metcalfe (1960: Ix). Such preparations were· either made permanent in Canada Balsam, after staining in safranin and Delafield's haematoxylin, or were examined as temporary mounts in chlor-zinc-iodide. All material to be sectioned was desilicified by soaking overnight in 50% hydrofluoric acid, followed by prolonged washing in running water. Freehand sections of the demineralized but unembedded material, in 70% alcohol, were cut at 10-20 fJ. in transverse and longitudinal sections on a Reichert sliding microtome. Most observations were made on sections in temporary glycerine mounts, either unstained or after treatment with a variety of temporary stains, such as phloroglucin and concentrated hydrochloric acid, Sudan IV, chlor-zinc-iodide, iodine-potassium iodide, and 70% sulfuric acid. A few permanent preparations of the freehand sections were made by a method described elsewhere (Tomlinson, 1961: