Aust. J. Plant Physiol., 1995, 22, 521-30 Photosynthetic Pathway-related Ultrastructure of C,, C, and C3-like C,-C, Intermediate Sedges (Cyperaceae), with Special Reference to Eleocharis Jeremy J. ~ r u h l *and ~ Sue perryA *~axonomyLaboratory, Research School of Biological Sciences, Australian National University, Canberra, ACT 2601, Australia. B~resentaddress and address for correspondence: Department of Botany, University of New England, Armidale, NSW 2351, Azistralia; email: jbruhl@metz.une.edu.au Abstract. The ultrastructure of photosynthetic organs (leaf blades and culms) was investigated in eight species from four genera of sedges: Fimbristylis (C, fimbristyloid anatomy), Pycreus (C4 chlorocyperoid anatomy), Rhynchospora (C4 rhynchosporoid anatomy)-all NADP-ME (malic enzyme) type, and uninvestigated C3, C, (eleocharoid anatomy, NAD-ME type) and C3-like C3-C4 intermediate species of Eleocharis. Ultrastructural characteristics previously reported for the former anatomical types are largely confirmed, though some evidence of poorly developed peripheral reticulum in C, rhynchosporoid sedges is presented. Sedges, regardless of anatomical and biochemical type, possess a suberised lamella in photosynthetic organs which is invariably present in and confined to the mestome sheath cell walls, though it is often incomplete in the radial walls. By contrast with other C, sedges, NAD-ME Eleocharis species and the C3-like C3-C, intermediate E. pusilla possess abundant mitochondria and chloroplasts with well-stacked grana in the photosynthetic carbon reduction (PCR) (Kranz)/bundle sheath cells. Peripheral reticulum is well developed in NAD-ME species in both PCR and photosynthetic carbon assimilation (PCA) (C4 mesophyll) chloroplasts, but differs from that seen in chlorocyperoid and fimbristyloid type sedges. The suberised lamella and starch grains (well preserved), and granal stacks (poorly preserved) are identifiable in dried herbarium material (Eleocharis). Prediction of C4 biochemical type of sedges should be possible by combining anatomical, ultrastructural and 613c value data. The significance of the ultrastructural similarities between the C, NAD-ME and C3-C, intermediate Eleocharis species is discussed. Introduction The fimbristyloid, chlorocyperoid, eleocharoid and rhynchosporoid C4 anatomical types discussed are described in terms of primary vascular bundles (Bruhl 1990, 1995; Bruhl et al. 1987; see Ueno and Koyama 1987; Ueno and Samejima 1989; and Ueno et al. 1989 for contrasting classification of some anatomical variants). (1)Fimbristyloid C4 anatomy comprises three bundle sheaths: the inner border parenchyma cells are large and chlorenchymatous, constituting the photosynthetic carbon reduction (PCR or 'Kranz') tissue, and interrupted laterally by the metaxylem vessel elements; the mestcme sheath of small, ach!orenchymatous, thickwalled cells; and a complete (unless interrupted by sclerenchyma) parenchymatous bundle sheath (PBS), which is usually smaller and less chloroplast laden than the surrounding primary carbon assimilation (PCA or C4 mesophyll) tissue (a PBS also surrounds the secondary bundles). (2) Chlorocyperoid C4 anatomy is essentially similar, but here the PBS is restricted to one or a few cells lateral to the metaxylem vessel elements or, in a few genera, is completely absent (and is always absent from the secondary bundles). (3)The border parenchyma cells also constitute the PCR tissue in eleocharoid C4 anatomy, but are usually not interrupted by the metaxylem vessel elements, and the PBS is absent. (4) The mestome sheath constitutes the PCR site in rhynchosporoid C4 species, and the PBS is present but irregularly incomplete. Ultrzistructural differences in the chloroplasts (presence or absence of peripheral reticulum and convolute or parallel arrangement of the thylakoid systems) of some C4 sedges possessing chlorocyperoid, fimbristyloid and rhynchosporoid anatomy (Bruhl et al. 1987), have been related to differences in the proximity of the PCR and PCA tissues in relation to metabolite transport (Ueno et al. 1988). 0310-7841/95/040521$05.00 J. J. Bruhl and S. Perry Apart from this, and in contrast with the existence of a substantial body of information on C4 anatomical variation in the Cyperaceae (e.g. see Lerman and Raynal 1972; Sharma and Mehra 1972; Raynal 1973; Takeda et al. 1985; Bruhl 1990; Bruhl et al. 1992), ultrastructure of the family has received relatively little attention (e.g. see Black and Mollenhauer 1971; Laetsch 1971; Carolin et al. 1977; Gilliland and Gordon-Gray 1978; Jones et al. 1981; EstelitaTeixeira and Handro 1987; Ueno and Samejima 1989; Estelita 1992, 1993). The assumption that all C4 sedges are NADP-ME (malic enzyme) has probably been a disincentive, but the desirability of extended comparative ultrastructure work became apparent with the discovery of the eleocharoid anatomical-type and its association with NAD-ME-type photosynthesis (Bruhl et al. 1987; cf. Ueno and Samejima 1989). For example, if the ultrastructural/biochemical correlations obtained in grasses apply to sedges (see e.g. Laetsch 1974; Hattersley 1987 and references therein), the NAD-ME species would be expected to have PCR cell chloroplasts with abundant grana and mitochondria, in contrast with the NADP-ME species which have agranal PCR cell chloroplasts and about a one-to-one ratio of mitochondria in the PCR and PCA cells. In fact, Ueno and Samejima (1989) found two C4, NAD-ME Eleocharis species to have PCR cell chloroplasts with abundant grana and mitochondria. Before the discovery of NAD-ME Eleocharis species with the PCR in the border parenchyma position (Bruhl et al. 1987), it had seemed that biochemical type could be predicted from anatomy alone (Ueno et al. 1986). Indeed, it seems that further C4 Eleocharis species with eleocharoid anatomy (i.e. with the PCR intervening between the metaxylem vessel elements and the mestome sheath, and without an obvious parenchymatous bundle sheath; Bruhl et al. 1987; see also Ueno and Samejima 1989, figs 1-2) will prove to be NAD-ME. However, the NAD-ME Eleocharis retrojlexa subsp. retrojlexa now seems to be variable for the anatomical trait (i.e. the metaxylem vessel elements often interrupt the border parenchyma cells), as is the recently discovered C4 E. retrojlexa subsp. subtilissima (Bruhl 1990), which casts some doubt on the prediction of their biochemical type from anatomy. Further, Ascolepis capensis (related to species with chlorocyperoid anatomy), and Bulbostylis paradoxa and Nemum equitans (related to species with fimbristyloid anatomy, and possessing a well defined PBS) also have PCR tissue in the border parenchyma which intervenes between the metaxylem vessel elements and the mestome sheath (Bruhl 1990; Bruhl et al. 1992). These three species also await biochemical typing. On the basis of physiological studies, Sonnenberg and Botha (1992) have suggested that the chlorocyperoid Cyperus fastigiatus may be NAD-ME or phosphoenolpyruvate carboxykinase (PCK). Therefore, these species cannot be confidently typed on the basis of anatomy alone. Ultrastructural/biochemical correlates may allow reliable biochemical typing of sedges on the basis of anatomical/ultrastructural criteria, possibly even from herbarium material (cf. Hattersley and Perry 1984). This study compares the ultrastructure of previously described variants in Eleocharis with that of the other known C4 and Cg anatomical types. Dried material of E. retrojlexa and E. subcancellata has been examined to assess the possibility of using ultrastructure to type C4 sedges from herbarium material, and to extend the comparison of the suberised lamella between fresh and dried material (Hattersley and Perry 1984). Materials and Methods Plants were grown from field-collected vegetative material or seed, under half shade in glasshouses maintained between 35°C (day maximum) and 15°C (night minimum), and regularly fertilised with Ruakura nutrient solution (Smith et al. 1983). All identities were checked using appropriate literature. Vouchers have been lodged at the Australian National Herbarium: Eleocharis acuta, JJB125; E. geniculata, JJB231; E. pusilla, JJB682; E. caespitosissima, JJB399; E. retroflexa subsp. retroflexa, C2967; Fimbristylis tetragona, JJB546; Pycreus polystachyos, JJB309; Rhynchospora rubra, JJB573. The mid-third of the youngest, fully-expanded leaves or culms were sliced into small segments under fixative. Culms of conventionallydried specimens of E. retroflexa and E. subcancellata were also prepared according to method (1) below. In order to improve the quality of fixation and infiltration, and to permit close comparison with the results of Ueno et al. (1988), three methods were applied to the fresh material: 2.5% glutaraldehyde in 50 m M cacodylate buffer, pH 7.0, for 2 h, washed twice in buffer over 1 h, post-fixed in 1%0 s 0 4 for 2 h at room temperature, washed in water, dehydrated in a graded ethanol series (15, 30, 50, 70, 90, 95 twice and 100% twice, for 15 min each), infiltrated for 24 h in LR White at room temperature, and embedded in fresh LR White for 12 h at 60°C under vacuum; 2.5% glutaraldehyde and 2% formaldehyde in 30 mM Pipes buffer, 16 h at 4"C, washed three times in buffer over 30 min, left in buffer for 12 h at 4"C, post-fixed in 1% 0 s 0 4 for 16 h at 4"C, washed in buffer three times over 1 h, dehydrated in an ethanol series, as above but 30 min per step, infiltrated in LR White with four changes over I week, and embedded in fresh LR White as above; and method similar to that of Ueno et al. (1988): 5% glutaraldehyde in 50 InM sodium phosphate buffer at 4OC, washed briefly in buffer three times, and overnight at 4"C, post-fixed in 2% 0 s 0 4 for 6 h at P C , rinsed in buffer and left overnight in buffer at 4"C, washed in distilled H,O, dehydrated and embedded as for (2). ~ransverse~sections were cut with a LKB 2128 ultramicrotome, double-stained with uranyl acetate and lead citrate, and examined in a Hitachi 600 electron microscope. The ultrastructural features of interest here were constant regardless of the fixation schedule employed, althcugh longer infiltratiun times resulted in fewer sectioning artifacts. Results and Discussion Table 1 summarises the ultrastructural features observed in a sample chosen to represent all the known biochemical and anatomical types in the Cyperaceae, paying particular attention to the variation in Eleocharis. The observations on Table 1. Ultrastructural characteristics of leaves and culms in the Cyperaceae sampled C3-C4 = C3-C4 intermediate; C = chlorocyperoid C4 anatomy; E = eleocharoid C4 anatomy; F = fimbristyloid C4 anatomy; R = rhynchosporoid C4 anatomy; NADP-ME, NADP-malic enzyme; NAD-ME, NAD-malic enzyme; MS = mestome sheath; PCR = primary carbon reduction ('Kranz'); grana: + = well developed, - = absent or at most stacks of three or fewer appressed thylakoids. Mitochondria1profiles are expressed on a per cell basis, BP = border parenchyma, PCA = primary carbon assimilation. Peripheral reticulum: + = poorly developed, ++ = well developed. n.a. = not applicable, 'blanks' = not scored. Photosynthetic pathway status, and C, anatomical- and biochemical-type data from Bruhl et al. (1987) - Photosynthetic pathway1 species c3 Eleocharis acuta Eleocharis geniculata c3-c4 Eleocharis pusilla c4 Eleocharis caespitosissima Eleocharis relrojlexa subsp. retrojlexu Fimbristylis tetragonu Pycreus polystachyos Rhynchospora rubra Photosynthetic C4 organ anatomical type C4 biochemical tYPe Location of suberised lamella Border parenchyma andlor PCR cells chloroplasts Position Location Mi tochondrial profiles BP :PCA ratio Peripheral reticulum Thylakoid systems Grana culm culm n.a. n.a. n.a. n.a. MS MS scattered scattered border parenchyma border parenchyma parallel parallel + + culm n.a. n.a. MS scattered border parenchyma parallel + >I + culm E NAD-ME MS scattered border parenchyma parallel + >1 ++ culm culm leaf leaf E F C R NAD-ME NADP-ME NADP-ME NADP-ME MS MS MS MS scattered centrifugal centrifugal centrifugal border parenchyma border parenchyma border parenchyma mestome sheath parallel convoluted convoluted parallel ++ ++ ++ + + >1 - -1 - -1 - -1 J. J. Bruhl and S. Perry Fig. 1. Transmission electron micrographs of fresh leaf blades and photosynthetic culms of sedges (cut transversely). (A) Eleocharis acuta (C3): only mestome sheath (MS) cell walls have a suberised lamella (SL and arrow heads). The suberised lamella (arrow heads) is discontinuous in the radial walls (RW). Inner tangential wall (ITW). The border parenchyma (BP) cells contain mitochondria (M) and one to few chloroplasts with grana (G), none are apparent in mestome sheath or parenchymatous bundle sheath (PBS) cells. Scale = 5 ym. (B) Pycreus polystachyos (CJ: the 'photosynthetic carbon reduction' (PCR) cell in the border parenchyma position, internal to the mestome sheath (MS, without chloroplasts), has three centrifugal chloroplasts. Note the electron-dense stroma (ST), convoluted thylakoid system, and peripheral reticulum (PR). Mitochondrion (M). The suberised lamella (arrow heads) occupies the mestome sheath and is discontinuous in the radial walls (RW). Scale = 2 ym. ( C ) Rhynchospora rubra (C4): mestome sheath cell (PCR site of this species) with chloroplasts showing electron-dense stroma (ST), parallel agranal thylakoid system, and abundant starch grains (SG). The arrows indicate the position of the peripheral reticulum, appearing as a less Photosynthetic Ultrastructure of Cyperaceae C3 (Fig. 1A) and C4 chlorocyperoid (Fig. 1B), fimbristyloid and rhynchosporoid (Fig. 1C and E) species largely confirm earlier observations on C3 and NADP-ME species by Laetsch (1971), Carolin et al. (1977), Gilliland and GordonGray (1978), Jones et al. (1981) and Ueno et al. (1988). Nevertheless, confidence in generalisations drawn from the available ultrastructural observations on Cyperaceae must be tempered with caution, given the small size of the sample. For rhynchosporoid anatomy this comprises only Rhynchospora rubra, which has been examined in three studies with more or less consistent results (Gilliland and Gordon-Gray 1978; Ueno and Samejima 1989; Fig. 1C and E). The fimbristyloid and chlorocyperoid types have been better surveyed: 3 general11 species (Carolin et al. 1977; Estelita-Teixeira and Handro 1987; Ueno and Samejima 1989; Table 1) and 6 general32 species (Black and Mollenhauer 1971; Laetsch 1971; Jones et al. 1981; EstelitaTeixeira and Handro 1987; Ueno and Samejima 1989; Estelita 1992, 1993; Fig. 1B) respectively, although no Rhynchospora species with chlorocyperoid anatomy has been investigated. The following ultrastructural characteristics of NADPME sedges with chlorocyperoid, fimbristyloid or rhynchosporoid anatomy are confirmed here: (1)the convoluted thylakoid membrane pattern (Fig. 1B) in PCR chloroplasts of fimbristyloid and chlorocyperoid sedges"; (2) the more usual parallel arrangement of the thylakoid system in the rhynchosporoid species (Fig. 1C); (3)the agranal or at best poorly stacked thylakoid membranes; and (4) the more or less 1:1 ratio of mitochondria in the photosynthetic carbon reduction (PCR or Kranz) cells to those in the primary carbon assimilation (PCA or C4 mesophyll) cells. NAD-ME species with eleocharoid anatomy (Fig. 2A-C) and the intermediate Eleocharis pusilla (Bruhl et al. 1987; Fig. 2 0 and F) exhibit a distinct suite of ultrastructural features (Table 1); i.e. well-stacked grana in the PCRhundle sheath chloroplasts and abundant mitochondria in the PCRhundle sheath cells. Fewer sections were examined of either of the C3 species than the other species sampled, but they also exhibit a few granal chloroplasts in the border parenchyma position, while numbers of mitochondria1 profiles for these cells (Fig. 1A) vary from equalling to slightly exceeding numbers for mesophyll cells. The stroma of the bundle sheath chloroplasts appears to be more electron-dense than that of the mesophyll cells in all the C4 species examined (e.g. Fig. 2C). Starch grains are larger and generally more abundant in the PCR chloroplasts of Rhynchospora rubra (Fig. 1C) and in the PCR (Fig. 2A, C and E) and PCA (Fig. 2 0 ) chloroplasts of the C4 Eleocharis species than in any of the other species examined. A suberised lamella has been found in walls of cells in the mestome sheath position in all sedges examined, regardless of anatomical or biochemical type (Figs 1A-E and 2A and C-F; Carolin et al. 1977; Ueno et al. 1988; Ueno and Samejima 1989). Ultrastructurally the suberised lamella seen in these sedges resembles that found in the Poaceae (cf. O'Brien and Carr 1970; Laetsch 1974; Hattersley and Browning 1981; cf. Gunning and Steer 1975; Ueno et al. 1988). It consists in transection of two parallel osmophilic bands separated by a lighter zone (Fig. 1D and E). Where the suberised lamella traverses plasmodesmata, it widens (Figs 1E and 2A and D) and consists of several alternating osmophilic and light bands. In leaves of C4 Poaceae, suberised lamellae vary in occurrence and position. For example, most NADP-ME grasses possess a complete suberised lamella in the outer tangential wall of the PCR tissue (which constitutes the mestome sheath) and only a patchy one aligning the inner tangential and radial walls. Most NAD-ME grasses possess a complete suberised lamella in the walls of the mestome sheath, but the PCRJPBS cell walls lack a suberised lamella. Most PCK species have a complete suberised lamella in the walls of the mestome sheath and the suberised lamella of the PCRPBS cell is complete in the outer tangential walls but * ~ oall t species have convoluted thylakoids and this has led Gilliland and Gordon-Gray (1978) to describe the thylakoid system in Kyllinga (chlorocyperoid anatomy) and Fimbristylis dichotoma (fimbristyloid anatomy) as parallel, while Carolin et al. (1977) described the thylakoid system in PCR chloroplasts of K. brevifolia as contorted, but still with large areas of parallel flattened lamellae. See also Estelita (1992, 1993). electron-dense zone. Mestome sheath (arrow heads). Radial wall (RW, cf. Fig. 1E). Scale = 1 pm. (D) Eleocharis retroflexa subsp. retroflexa (CJ: showing the junction of two mestome sheath (MS) cells. The suberised lamella in places (SL) has a tramline appearance, and is discontinuous in the radial wall (RW). The arrow heads indicate the extent of the suberised lamella, but note that the 'gap' is relatively electron-dense. Primary carbon reduction cell (PCR). Scale = 1 pm. (E) Rhynchospora rubra (C4): the radial wall with pit field and suberised lamella (arrow heads; Fig. 1C) at a higher magnification. The upper left suberised lamella (arrow) has a 'tramline' appearance. The upper suberised lamella is discontinuous, while the lower one is continuous and widened where it is traversed by plasmodesmata (PD, sectioned obliquely) in the pit field. Scale = 0.5 pm. J. J. Bruhl and S. Perry Fig. 2. Transmission electron micrographs of fresh (A-D and F) and dried (E) photosynthetic culms of Eleocharis (cut transversely). (A) E. caespitosissirna (CJ: part of a border parenchyma cell (the PCR site), internal to the mestome sheath (top left), with chloroplasts and mitochondria (M). Note the well-stacked grana, starch grains (SG), and well-developed peripheral reticulum (PR) with abundant electron transparent regions. Suberised lamella (arrow heads) in the mestome sheath wall is traversed by plasmodesmata (PD). Scale = 1 pm. (B) E. caespitosissirna (CJ: part of chloroplast of PCR cell showing well-stacked gram (G), peripheral reticulum (PR) and associated electrontransparent areas (asterisk). Scale = 0.2 pm. ( C )E. caespitosissirna (C4): photosynthetic carbon reduction cell (PCR) in the border parenchyma position with abundant chloroplasts (some with prominent starch grains) and mitochondria (M). A suberised lamella (arrow heads) occupies the walls of the mestome sheath (MS). Note the peripheral reticulum (PR) of chloroplasts in both the C4 mesophyll or 'primary carbon assimilation' (PCA) and PCR cells. Scale = 4 pm. Photosynthetic Ultrastructure of Cyperaceae patchy along the radial walls (Hattersley 1987; Prendergast et al. 1987). By contrast, in the Cyperaceae the suberised lamella always occupies the inner and the outer tangential walls of the mestome sheath (cf. Carolin et al. 1977; Ueno et al. 1988; Ueno and Samejima 1989). Although Ueno et al. (1988) observed the suberised lamella in the mestome sheath in R. rubra, they avoided naming these cells mestome sheath (see also Gilliland and Gordon-Gray 1978). The distribution of a suberised lamella in the radial walls of the mestome sheath of sedges, however, is somewhat variable. A continuous suberised lamella in radial walls, beautifully illustrated by Ueno et al. (1988 pp. 147-148), appears to be the exception rather than the rule. More often the suberised lamella is discontinuous in at least one of the radial walls of adjacent cells (Figs 1A-E and 20). The adjacent segments of the discontinuous lamellae sometimes end at the same distance from the opposite tangential walls (Fig. ID). In any case, the zone near the ends of the suberised lamella or between them is somewhat more electron-dense than much of the remainder of the secondary wall (Fig. ID-E), consistent with the presence of a casparian strip or of a similarly water-impermeable material (cf. Bocher and Oleson 1978; Hattersley 1987). Considering putative apoplastic transport between PCA and PCR tissue, the PCR cell chloroplasts are separated from PCA cells by a suberised lamella in both outer and inner tangential walls of the mestome sheath in species where the PCR tissue is in the border parenchyma position. By contrast, where the PCR tissue occupies the mestome sheath, the PCR cell chloroplasts are separated only by a suberised lamella in the outer tangential walls of the mestome sheath. Chloroplast position in C4 grasses may represent different compromises between the demands of maximising rates of PCR-PCA metabolic transport on the one hand, and of reducing rates of C 0 2 leakage on the other (Hattersley and Browning 1981). The correlation between biochemical type and chloroplast position in the sedges does not always correspond to the situation seen in grasses (e.g. NAD-ME grasses with centripetal chloroplasts and NAD-ME sedges with scattered chloroplasts; see Hattersley 1987). In C4 sedges, the presence of a suberised lamella irrespective of biochemical type could be seen as removing one of the constraints, as it relates to C 0 2 leakage, on chloroplast position within the PCR cells (cf. Ueno and Samejima 1989). Prominent peripheral reticulum, representing invagination of the inner of the two chloroplast envelope membranes (cf. Laetsch 1974; Gunning and Steer 1975), was identified by Carolin et al. (1977) in chloroplasts of chlorocyperoid and fimbristyloid-type species. Ueno et al. (1988) also generally found peripheral reticulum in these types, though they indicated that it was more abundant in PCR than PCA chloroplasts, and even tabulated its absence from the PCA chloroplasts of some species. In Rhynchospora rubra they noted it only in PCA chloroplasts. In the present study, prominent peripheral reticulum was confirmed not only in species with chlorocyperoid and fimbristyloid anatomy, but also in association with eleocharoid anatomy (Fig. 2A-C; cf. Ueno and Samejima 1989). In the former types the electron-transparent areas are often dumb-bell shaped (cf. Carolin et al. 1977; Jones et al. 1981) while in the C4 Eleocharis species they are more or less circular in profile (Fig. 2A and B). Our observations suggest that there is a poorly developed peripheral reticulum, or at least a peripheral reticulum-like zone in the PCR chloroplasts of R. rubra (Fig. 1C). The peripheral reticulum bounds or defines electron-transparent regions of the chloroplasts referred to by Carolin et al. (1977) and Ueno et al. (1988) as vesicles. However, their status as vesicles (rather than being some form of reticulate intermembrane region) needs confirming, e.g. via serial sections, to demonstrate that each transparent region is indeed a separate entity. The phenomenon of a peripheral reticulum and associated light regions may be an artifact of fixation (see Laetsch 1971, 1974): artifactual vesiculation has been described by Mersey and McCully (1978). Even so, it is a constant feature, notwithstanding different fixation schedules and preparation in different laboratories. Where well developed, it occurs in relatively closely related groups of species (Bruhl 1990, 1995), and evidently has some taxonomic value. Ueno et al. (1988) proposed that the development of the peripheral reticulum in species with chlorocyperoid and fimbristyloid anatomy is related to rapid metabolite transport, compensating for the higher resistance proposed ( D ) E. pusilla (C3-C4): border parenchyma cell (BP) showing mitochondria (M), and three scattered chloroplasts with well-stacked grana (G). Suberised lamella (arrow heads) in the mestome sheath (MS) walls is traversed by plasmodesmata (PD) and is discontinuous in the radial walls (RW). Inner tangential wall (ITW). Scale = 2 pm. ( E )E. retropexu subsp. retropexu (C4): 'photosynthetic carbon reduction' (PCR) cell in the border parenchyma position with disrupted chloroplasts. Note grana (arrows) and starch grains (unlabelled). The suberised lamella (SL) in the mestome sheath (MS) cell walls is intact. 'Primary carbon assimilation' (PCA) cell contents poorly preserved except for the starch grains (SG). Scale = 2 pm. ( F )E. pusilla (C3-C4): border parenchyma (BP) cells, with chloroplasts and abundant mitochondria (M), adjacent to the mestome sheath cell walls (top) with suberised lamella (arrow heads). Scale = 2 pm. 528 for C4 sedges where the PCR is separated from the PCA by the mestome sheath, as compared with species possessing rhynchosporoid anatomy where PCR and PCA tissues are adjacent. Peripheral reticulum in grasses seems never to have been critically examined in this context with respect to the C4 subtypes. The presence of well developed peripheral reticulum in the sedges with eleocharoid anatomy, where the border parenchyma cells constitute the PCR tissue, is consistent with the notion of its involvement in metabolite transport (Ueno et al. 1988). However, the different appearance of the peripheral reticulum in the eleocharoid species, together with their being biochemically distinct, suggests that it may have evolved independently here in connection with some as yet unknown function or a related but not identical one. The suberised lamella in E. retrojlexa and E. subcancellata remains intact after drying (Fig. 2E), as in grasses (Hattersley and Perry 1984). Not surprisingly, the chloroplasts are poorly preserved in such material: the peripheral reticulum and mitochondria are mostly not recognisable or are obviously altered, and the electron transparent regions (or vesicles) are greatly enlarged; the stroma and the thylakoid systems are disrupted, though grana are apparent; and starch grains are well preserved (Fig. 2 0 . The consistency of the ultrastructural features that were observed in both fresh and herbarium material of E. retrojlexa (which has also been biochemically typed) allows for some confidence in the typing of material from herbarium material. Thus, on the basis of abundant PCR chloroplasts with grana and many starch grains, and a C4 613c value and eleocharoid anatomy (Bruhl et al. 1987; Bruhl 1990), E. subcancellata is predicted to be C4 and NAD-ME. Prediction of C3-like C3-C intermediates in 3 ~ Eleocharis could be based on C3-like 61 C values combined with C4-like culm anatomy, poorly developed starch grains and less abundant bundle sheath chloroplasts. A striking feature of NAD-ME sedges and of the apparently C3-C4 intermediate, E. pusilla, is the abundance of mitochondria in the border parenchyma cells (Fig. 2A, C, D and F). These chlorenchymatous cells typically contain more than six mitochondria per profile and often many more (with some counts of 30 per profile). By contrast, there are usually fewer than six mitochondria per PCR cell profile in chlorocyperoid, fimbristyloid, and rhynchosporoid species. Both chloroplasts and mitochondria are less abundant in the border parenchyma cells of C3 sedges, but there is sufficient variability within and between the C3 species to suggest a need for more detailed examination, which might usefully start with C3 Eleocharis species. Size profiles of the border parenchyma mitochondria were not calculated but, in the NAD-ME and intermediate species, they appear to be at least as large as those of the mesophyll. The greater number of mitochondria in the PCR cells of the NAD-ME species is not surprising, given that J. J. Bruhl and S. Perry mitochondria are the site of decarboxylation in NAD-ME species (Hatch and Mau 1973; Hatch and Kagawa 1974a, 1974b; Kagawa and Hatch 1975). Abundant bundle sheath mitochondria are also a feature of C3-like C3-C4 intermediates (see e.g. Holaday et al. 1984; Hylton et al. 1988; Rawsthome et al. 1988). These mitochondria are involved in a glycine shuttle, where glycine from the mesophyll is transported to the bundle sheath cells for decarboxylation in the mitochondria (Monson et al. 1984; see also Edwards and Ku 1987). Hylton et al. (1988) demonstrated, via in situ immunogold labelling of leaves, that in C3-C4 intermediates of Flaveria and Moricandia glycine decarboxylase is present only in bundle sheath mitochondria. More detailed studies of E. pusilla are needed to establish whether glycine decarboxylase is restricted to its bundle sheath mitochondria (cf. Hylton et al. 1988; Rawsthome et al. 1988) together with more elaborate gas exchange experiments, designed to test whether the response of C 0 2 compensation point to changing O2 pressures fits the predictions of the model for C3-like C3-C4 intermediates (see Caemmerer 1989). The functional and evolutionary significance of the similarities between the NAD-ME and C3-like C3-C4 intermediate sedges are not known. Both have abundant mitochondria associated with scattered, granal chloroplasts located in the border parenchyma position. In both, the mitochondria probably serve as the site for decarboxylation (albeit perhaps for different metabolites). The sedges concerned are relatively closely related, all being members of Eleocharis, a helophytic to hydrophytic genus. The recent discovery of intraspecific variation of photosynthetic pathway in E. vivipara (Ueno et al. 1988; with C3 submerged forms and C4 NAD-ME terrestrial forms) , which may even be reversible, invites speculation as to the origin and stability of the C4-like features in E. pusilla, and should encourage investigation of the genetic and molecular basis of C4 expression in this genus. Ecological considerations suggest that the extreme and changeable microenvironments inhabited by its species may have been decisive in the appearance of these photosynthetic pathway variations (see also Bruhl 1994). Acknowledgments We are grateful to Paul Hattersley and Les Watson for discussions and critical comments. We thank Harold Brown, Beth Gibbs-Russell, and Toby Kellogg for critically reviewing an earlier version of the manuscript. Thanks also to Philip Sharpe and Mike Godwin for field assistance. JJB is also grateful to the Herbarium of the Northern Territory for support on field work; the CEO of the Australian Nature Conservation Agency for permission to collect in service areas; the Director of the Australian National Herbarium for use of herbarium facilities; and the Plant Culture services, Photosynthetic Ultrastructure of Cyperaceae Electron Microscope and Photography Units at RSBS. JJB acknowledges receipt of an Australian Postgraduate Research Award. References Black, C. C. J., and Mollenhauer, H. H. (1971). Structure and distribution of chloroplasts and other organelles in leaves with various rates of photosynthesis. Plant Physiology 47, 15-23. Bocher, T. W., and Oleson, P. (1978). Structural and ecophysiological pattern in the xerohalophytic C, grass Sporobolus rigens (Tr.) Desv. Biologiske Skrifter Kongelige Danske Videnskabemes Selskab 22, 1 4 8 . Bruhl, J. J. (1990). Taxonomic relationships and photosynthetic pathways in the Cyperaceae. PhD Thesis, Australian National University. Bruhl, J. J. (1994). Amphicarpy in the Cyperaceae, with novel variation in the wetland sedge Eleocharis caespitosissima Baker. Australian Journal of Botany 42, 441448. Bruhl, J. J. (1995). Sedge genera of the world: relationships and a new classification of the Cyperaceae. Australian Systematic Botany 8, 125-305. Bruhl, J. J., Stone, N. E., and Hattersley, P. W. (1987). C4 acid decarboxylation enzymes and anatomy in sedges (Cyperaceae): first record of NAD-malic enzyme species. Australian Journal of Plant Physiology 14, 7 19-728. Bruhl, J. J., Watson, L., and Dallwitz, M. J. (1992). Genera of Cyperaceae: interactive identification and information retrieval. Taxon 41, 225-234. Caemmerer, S. von (1989). A model of photosynthetic CO, assimilation and carbon-isotope discrimination in leaves of certain C3-C4 intermediates. Planta 178, 463-474. Carolin, R. C., Jacobs, S. W. L., and Vesk, M. (1977). The ultrastructure of Kranz cells in the family Cyperaceae. Botanical Gazette 138, 4 13-419. Edwards, G. E., and Ku, M. S. B. (1987). Biochemistry of C3-C, intermediates. In 'Photosynthesis'. The Biochemistry of Plants. Vol. 10. (Eds M. D. Hatch and N. K. Boardman.) pp. 275-325. (Academic Press: London.) Estelita, M. E. M. (1992). Origin and structure of the Kranz tissues in Cyperaceae. Bolm Botdnica 1 3 , 4 1 4 8 . Estelita, M. E. M. (1993). Remirea Aubl. (Cyperaceae), a new Kranz genus. Revista Brasileira de Botanica 16, 137-141. Estelita-Teixeira, M., and Handro, W. (1987). Kranz pattern in leaf, scape and bract of Cyperus and Fimbristylis species. Revista Brasileira de Botanica 10, 105-1 11. Gilliland, M. G., and Gordon-Gray, K. D. (1978). Kranz and non-Kranz cells in Cyperaceae. Proceedings of the Electron Microscopy Society of Southern Africa 8, 85-86. Gunning, B. E. S., and Steer, M. W. (1975). 'Plant Cell Biology: an Ultrastructural Approach.' (Edward Arnold: London.) Hatch, M. D., and Kagawa, T. (1974~).NAD malic enzyme in leaves with C, pathway photosynthesis and its role in C, acid decarboxylation. Archives of Biochemistry and Biophysics 160, 346-349. Hatch, M. D., and Kagawa, T. (1974b). Activity, location and role of NAD malic enzyme in leaves with C, pathway photosynthesis. Australian Journal of Plant Physiology 1, 357-369. 529 Hatch, M. D., and Mau, S. (1973). Association of NADP- and NAD-linked malic enzyme activities in Zea mays: relation to C, pathway photosynthesis. Archives of Biochemistry and Biophysics 179, 361-369. Hattersley, P. W. (1987). Variations in photosynthetic pathway. In 'Grass Systematics and Evolution'. (Eds T. R. Soderstrom, K. W. Hilu, C. S. Campbell and M. E. Barkworth.) pp. 49-64. (Smithsonian Institution Press: Washington DC.) Hattersley, P. W., and Browning, A. J. (1981). Occurrence of the suberized lamella in leaves of different photosynthetic types. I. In parenchymatous bundles sheaths and PCR (Kranz) sheaths. Protoplasma 109, 371-401. Hattersley, P. W., and Perry, S. (1984). Occurrence of the suberized lamella in leaves of grasses of different photosynthetic types. 11. In herbarium material. Australian Journal of Botany 32, 465473. Holaday, A. S., Shieh, Y., Lee, K. W., and Chollet, R. (1984). Anatomical ultrastructural and enzymic studies of leaves of Moricandia awensis, a C3-C, intermediate species. Biochimica et Biophysica Acta 637, 334-341. Hylton, C. M., Rawsthorne, S., Smith, A. M., Jones, D. A., and Woolhouse, H. W. (1988). Glycine decarboxylase is confined to the bundle-sheath cells of leaves of C3-C, intermediate species. Planta 175, 452459. Jones, M. B., Hannon, G. E., and Coffey, M. D. (1981). C, photosynthesis in Cyperus longus L., a species occurring in temperate climates. Plant, Cell and Environment 4, 161-168. Kagawa, T., and Hatch, M. D. (1975). Mitochondria as a site of C, acid decarboxylation in C,-pathway photosynthesis. Archives of Biochemistry and Biophysics 167, 687-696. Laetsch, W. M. (1971). Chloroplast structural relationships in leaves of C, plants. In 'Photosynthesis and Photorespiration'. (Eds M. D. Hatch, C. B. Osmond and R. 0 . Slatyer.) pp. 323-349. (Wiley-Interscience: New York.) Laetsch, W. M. (1974). The C, syndrome: a structural analysis. Annual Review of Plant Physiology 25, 27-52. Lerman, J. C., and Raynal, J. (1972). La teneur en isotopes stables du carbone chez les CypCracCes: sa valeur taxonomique. Compte Rendu de l'dcade'mie des Sciences, Paris 275, 1391-1394. Mersey, B., and McCully, M. E. (1978). Monitoring of the course of fixation of plant cells. Journal of Microscopy 114,49-76. Monson, R. K., Edwards, G. E., and Ku, M. S. B. (1984). C3-C, intermediate photosynthesis in plants. Bioscience 34, 563-574. O'Brien, T. P., and Carr, D. J. (1970). A suberized layer in the walls of the bundle sheath of grasses. Australian Journal of Biological Sciences 23, 275-287. Prendergast, H. D. V., Hattersley, P. W., and Stone, N. E. (1987). New structuravbiochemical associations in leaf blades of C, grasses. Australian Joumal of Plant Physiology 14, 403-420. Rawsthorne, S., Hylton, C. M., Smith, A. M., and Woolhouse, H. W. (1988). Photorespiratory metabolism and immunogold localization of photorespiratory enzymes in leaves of C3 and C3-C, intermediate species of Moricandia. Planta 173, 298-308. Raynal, J. (1973). Notes cypCrologiques. 19. Contribution 2i la classification de la sous-famille des Cyperoideae. Adansonia 2 13, 145-171. 530 Sharma, 0 . P., and Mehra, P. N. (1972). Systematic anatomy of Fimbristylis Vahl (Cyperaceae).Botanical Gazette 133, 87-95. Smith, G. S., Johnston, C. M., and Cornforth, I. S. (1983). Comparison of nutrient solutions for growth of plants in sand culture. New Phytologist 94, 537-548. Sonnenberg, B. J., and Botha, C. E. J. (1992). An investigation of leaf-blade anatomy and photosynthetic characteristics of four Cyperaceae species from the Albany and Bathurst districts in the eastern Cape. South African Journal of Botany 58,297-303. Takeda, T., Ueno, O., Samejima, M., and Ohtani, T. (1985). An investigation for the occurrence of C4 photosynthesis in the Cyperaceae from Australia. Botanical Magazine Tokyo 98, 393411. Ueno, O., and Koyama, T. (1987). Distribution and evolution of C4 syndrome in Rhynchospora (Rhynchosporeae-Cyperaceae). Botanical Magazine Tokyo 100, 63-85. Ueno, O., and Samejima, M. (1989). Structural features of NADmalic enzyme type C4 Eleocharis: an additional report of C, acid decarboxylation types of the Cyperaceae. Botanical Magazine Tokyo 102, 393402. J. J. Bruhl and S. Perry Ueno, O., Samejima, M., and Koyama, T. (1989). Distribution and evolution of C4 syndrome in Eleocharis, a sedge group inhabiting wet and aquatic environments, based on culm anatomy and carbon isotope ratios. Annals of Botany 64, 425-438. Ueno, O., Samejima, M., Muto, S., and Miyachi, S. (1988). Photosynthetic characteristics of an amphibious plant, Eleocharis vivipara: expression of C4 and Cg modes in contrasting environments. Proceedings of the National Academy of Sciences, USA 85,6733-6737. Ueno, O., Takeda, T., and Murata, T. (1986). C4 acid decarboxylating enzyme activities of C4 species possessing different Kranz anatomical types in the Cyperaceae. Photosynthetica 20, 111-116. Manuscript received 3 August 1994, accepted 28 September 1994