Journal oft/it Torrev Botanical Society 132(3). 25. pp. 377—383

Origin and nature of vessels in . 7. and

Edward L. Schneider”2 and Sherwin Cariquist Santa Barbara Botanic Garden, 1212 Mission Canyon Road, Santa Barbara, CA 93105

SCHNEIDER, E L. AND S. CARLQU,ST (Santa Barbara Botanic Garden. 1212 Mission Canyon Road. Santa Barbara, CA 93105. Origin and nature of vessels in monocotyledons. 7. Philydraceae and Haemodoraceae. J. Torrey Bot. Soc. 132: 377—383. 2005.—SEM studies of maccrated stems and roots showed that long scalariform perforation plates are present in roots and probably also stems of (Philydraceae). In Haemodoraceae. Clearly recognizable vessel elements are present in roots: perforation plates range from scalariform to simple. In stems of Haemodoraceae, however, less clearly recognizable vessel elements are present: the presence of thread like pit membrane remnants is one criterion that argues for vessel presence. Such tracheary elements can be considered transitional between tracheids and vessel elements. End walls of transitional tracheary elements may have narrower bars between perforations, less prominent borders, and pit membrane remnants rather than non porose pit membranes. However, only one or two of these three character expressions may be present, suggesting that clear definitions of vessel elements and tracheids may not be possible. Haemodoraceae show a marked difference, or disjunction. in morphology between vessel elements rn roots and the vessel elements transhional to tracheids of stems. The physiological and evolutionary correlations of this morphological disjunction may be related to seasonality of water availability. Key words: Haemodoraceae. Philydraceae. tracheids. vessel elements, xylem evolution

Philydraceae and Haemodoraceae have long families Acoraceae (Carlquist been considered closely related to each other, an and Schneider 1997) and Hanguanaceae idea confirmed in recent DNA-based phyloge (Schneider and Carlquist 2005). Such transition nies (Chase 2004, Davis et al. 2004). These two al tracheary elements are important in showing families are now considered 10 form, along with the phylogenetic origin of vessels in monocot , Hanguanaceae, and Pontederi yledons. Study of xylem by means of scanning aceae. an order, Haemodorales (APG 11 2003 electron microscopy (SEM) clarifies the nature Chase 2004, Davis et al. 2004). The present of transitional tracheary elements, and thus il composition of this order, as noted by Chase luniinates the precise structural and ultrastruc (2004), results from molecular studies; in earlier tural nature of Ihis transition. Studies by means phylogenies, these families were placed in sev of light microscopy can show only part of the eral orders. Nomenclature follows MacFarlane nature of transitional tracheary elements, as is et al. (1987), Simpson (1998), and Goldblatt and evident from Cheadle’s (t968) notation of Manning (2000). “questionable perforations” in stem tracheary Philydraceae and Haemodoraceae are of spe elements of Anigozanthos r,’a Labill. cial interest with respect to xylem histology be Our sampling of Haemodoraceae and Phily cause Cheadle (1968) claimed that they had draceae is not as large as that of Cheadle (1968). clearly defined vessels in roots, but no vessels Cheadle studied the tracheary elements of 22 (or questionable ones in Anigozanthos of the Haemodoraceae and three Philydraceae. Simp Haemodoraceae) and thus only tracheids in son and Dickison (1981) studied the tracheary stems. Such a mode of occurrence of tracheary elements of Lachnanthes caroliniana and report elements offers the possibility of finding trache vessels with simple perforation plates in roots, my elements transitional between tracheids and vessels with scalariform perforation plates in vessel elements (“vessel-tracheids” of Palm rhizomes. and tracheids in leaves. We believe 1954). Such elements have been reported in the that the nature of tracheary elements in these Haemodoraceae and Philydraceae can be estab lished basis Author for communication: E-mail: eschneider@ on the of the material we have ex sbbg.org amined. Hopefully, our results offer a basis for 2 The authors thank Virginia Hayes for material of further examination of transitional tracheary el thyrsifloro from the Lotusland Founda ements in a range of monocotyledons. The fam tion. We are grateful for textual assistance offered by Michael G. Simpson and an anonymous reviewer. ilies we have studied to date (Carlquist and Received for publication November 15. 2004, and Schneider 1997. 1998a, 1998b; Schneider and in revised form May 26. 2005. Carlquist 1997. 1998, 2005) have been selected

377 .3 ItS JOURNAL OF THE TORREY BOTANICAL SOCIETY [VOL. 132

because they represent Lineages in which Iran end wall facet shown (Fig. 3) appeals to us as a sitional tracheary elements seemed likely to oc probable perforation plate because of its wide cur. Families in which vessel elements are clear scataciform perforation plates. slender bars (of ly different from tracheids have not been includ ten forked), and absence of pit membranes. To ed in this series of papers. be sure, absence of pit membranes can occur because of handling of materials. Consequently, Materials and Methods. Roots and under in attempting to identify perforation plates in ground stems of living specimens of Philydrum stems, we sought features relating to pit mem lanuginosum Banks & Solander ex Gaertn., An brane remnants. In Fig. 4, a portion of a perfo igozanthos flavidus DC. ex Red., and Wachen ration plate shows threads and porose pit mem dorfia thyrsifiora Burin. were collected from brane fragments in what may be perfocations; under cultivation in Santa Barbara. Ma the presumptive lateral wall (Fig. 4. left) pos terials were fixed in 70% aqueous ethanol and sesses intact pit membranes. Pit membranes con then maccrated with Jeffrey’s Fluid (Johansen taining pores or holes of various sizes (Fig. 5) 1940). Macerations were spread onto aluminum were observed near the end of a presumptive stubs bearing electroconductive pads, dried in perforation plate. The delicate pit membranes il air, sputter coated with gold, and examined with lustrated in Fig. 5 are probably thinner than pit a Hitachi 52600>4 5PM using an accelerating membranes of typical lateral wall pits. and thus voltage of 25 kv. Care was taken to minimize can be interpreted as pits transitional to a per exposure of delicate pit membranes to the elec foration. tron beam, as such exposure at high power re Anigozanihos Ha vidus (Fig. 6—9). Vessel el sults in perceptible shrinkage and destruction of ements are identified pit membranes. easily in the root (Fig. 6— 8). Scalariform perforation plates may he seen Our criteria for recognizing perforation plates in narrower (likely early metaxylem) vessel el are as follows. Perforation plates are end walls ements. (Fig. 6—7). Scalariform perforation distinguishable from lateral walls by one or plates, whether longer (Fig. 6) or shorter (Fig. more of these features: absence of pit mem 7), appeared sunken (collapsed inward toward branes or presence of them as pit membrane the lumen) in our preparations, perhaps because remnants only; bars narrower between perfora the bars are so slender that they yield to the ten tions than corresponding secondary wall por sion involved in drying. The difference between tions of lateral wall pitting: and transitions be perforation plates and lateral wall pitting is read tween lateral wall pitting and perforation plates ily apparent in these elements (Fig. 6—7). Vessel visible as porose thin pit membranes: or pit bor elements with simple perforation plates (Fig. 8) ders narrower in perforation plates than on lat were somewhat more common eral wall pitting. than those with scalariform ones. Simple perforation plates are readily identifiable by their circular or oval out Results. Philydrurn laviuginosum (Fig. 1—5). lines, their smooth edges, and their obvious con Roots (Fig. 1—2) have scalariform perforation trast with lateral wall pitting (Fig. 8). plates with numerous bars in metaxylem vessel In stems, no clearly identifiable perforation elements (Fig. I). These were identified not only plates were observed on tracheary elements. On because of being on end walls of tracheary el end walls, a few adjacent pits were observed to ements, but also by virtue of having slender bars contain threadlike pit membranes (Fig. rather and no obvious pit membrane remnants in per 9) than laminar pit membranes. However, such pits forations.. A curious protoxylem tracheary ele are associated with thick intervening bars and rnent, representing a cell with transitions be have prominent borders (Fig. 9), features usually tween annular and helical thickenings and with associated with lateral wall pitting. conspicuous circuLar openings in primary wall areas, was observed (Fig. 2). We cannot present Wachendorfia thyrsifiora (Fig. /0—13). In this with certainly as a vessel element, if only the roots of W. thyrsiflora. perforation plates ap because so few protoxylem vessel elements have peared to be simple (Fig. 10—Il). However, been identified and illustrated in monocotyle close examination of some perforation plates re dons, and thus available comparisons are few. vealed probable bars at the end of the perfora In stems of P. lanugnurcum, a few tracheary tion plate near (he cell tip (Fig. 10). and thus element facets referable to the concept of per one might designate a few plates as scalariform. foretlon plates were observed (Pig, 3—5). The although not in the ordinary sense. Vessel dc. 2005] SCXNEIDER AND CARI.QUIST ORIGIN OF VESSELS IN MONOCOTYLEDONS 379 I ii I I :3. I - so,

1’ a

1 aa e -L

a 5 Fir,. 1—5, SEM photographs of tracheary elements in roots (Fig. 1—2) and stems (Fig. 3—5) of Philydrum lwlugtnUsum. Fig. I. Perforation plate of a melaxylem vessel. Fig. 2. Possible perforation plate of a proioxylem element. Fig. 3. Apparent perforation plate in a metaxylem element. Fig. 4. Portion of a facet of a tracheary element: threads and pit membrane portions suggesting the presence of perforations. Fig. 5. Thin pit membranes with holes pores and of various sizes, from an end-walt facet of a tracheary etement. Fig- 1.3, scale = 50 [am; Fig. 2, 4, scale 5 tim; Fig, 5, scale = 1 p.m. 1’

380 JOURNAL OF THE TORREY BOTANICAL SOCIETY [Vol. 132

—

— i.I —Jr.

Fio. 6—9. SEM phr.nographs of tracheary elements in roots (Fig. 6—8) and stems (Fig. 9). of Anigoanthos fluvidus. Fig. 6. L.ong sealariform perforation plate of an early metaxylem vessel. Fig. 7. Moderalely long scalariform perforation plate of an early metaxylern vessel. Fig. 8. Simple perforation plate of a late melaxylem vessel. Fig. 9. Pits with threadlike pit membrane remnants from a tracheary element Itracheary element oriented transversely in this image). Fig. 6, scale = 50 xrn; Fig. 7—8. scale = 20 1km; Fig. 9, scale = 5 ram. 200$) SCHNEIDER AND CARLQUIST: ORIGIN OF VESSELS IN MONOCOTYLEDONS 381

-% -

—cc

4’.’ I j 12 13

FIG. 0—13. SEM photographs of tracheary elements of roots (Fig- 10—Il) and stems (Fig. 12—13) of Wach endorfia rhyrsiflora. Fig. 10. Vessel element with scalariform lateral wall pitting: the perforation plate appears nearly simple, but has seseral bars at top. Fig. II. Vessel element with a simple perforalion plate and alternate lateral wall pitting. Fig. 12. Portion of a tracheary element facet with threadlike pit membrane remnants and vestigial borders on bars, Fig. 12. Tracheary element with non-porose pit membranes (left) and porose pit membranes (center, right). Fig. 10—Il, scale = 20 pm; Fig. 12—13, scale = 5 pm. .382 JOURNAL OF THE TORREY BOTANICAL SOCIETY [VOL. 32

ments have either scalariform IFig. 10) or alter deed have perforated pits.” We would tend to nate (FigS II) pitting. concur with that judgment, for the threadlike pit In stems. clearly defined vessel elements were membranes we located on end-wall facets in not present. However, some end walls appear stem tracheary elements of .4nigozanrhus flavi similar to perforation plates (Fig. 12). Thread this and contain like pit membranes and relatively slender bars threadlike pit membranes not unlike those i]lus suggest that seine elements have perforation trated for the perforation plates of dicotyledons plates, at least on the basis of these two criteria. with highly primitive wood (Carlquist 1992). Lateral walls of stem vessel elements (Fig. 13) SEM study shows the intermediacy of such tra contain both intact nonporose pit membranes cheary elements, which might be called transi (left) and porose pit membranes (center, right): tional elements or “vessel-tracheids” in the the latter may be transitional to the kind of facet sense of Fahn (1954). shown in Fig. (2). The stetn vessel elements of Haemodoraceae and Philydraceae are illustrative of the stages of Discussion and Conclusions. Cheadle (1968) origin of vessel elements from iracheids, as in found vessel elements with long scalarifonn per the monocotyledon family Hanguanaceae foration plates in roots of Philydrum lanuginos (Schneider and Carlquist, 2005). Those who urn, but did not find vessel elements in sterns of wish a simple definition of vessel elements, a that species. We confirm the observation of root precise line between the two types of tracheary vessels, but in addition, we located some vessels elements, are thus left with only at best an ar in stems of P. lanuginosum that are similar to bitrary demarcation. Narrowing of bars, lessen those of the root. These stem vessels may not be ing of borders, and increasing porousness of pit frequent, however, and the possibility exists that membranes in perforation plates (as compared tracheids as well as tracheidlike vessel elements to these features in lateral wall pitting in any are present in the stems of Philydrum. species with transitional tracheary elements) do In Haemodoraceae, Cheadle (1968) reported not occur in sharply demarcated stages. One of simple perforation plates in roots, except for An these can occur without similar expressions in igozanthos manglesii D. Don, Dilatris pillansii the other two features, as in Acoraceae and Ar W. F Barker, and Tribonanrhes australis EndI.; aceae (Carlquist and Schneider 1997, 1998b; in these species scalariform perforation plates Schneider and Carlquist 1998). were reported in roots. We can add Anigozan Philydraceae are notable for the retention of thos flavidus to this list. Cheadle (1968) illus notably primitive vessel elements in roots, ac trated a root perforation plate in Haemodorum cording the criteria of Cheadle (1942. 19441. ausrroqueenslandicum Domin in which there is who claimed that the most specialized vessel el a single large opening, with several bars distally. ements in any given species of monocotyledon We reported such a perforation in the root of are to be expected in the roots, with progres Wachendorfia rhyr.cifiora (Fig. 10). With the ex sively less specialized vessel elements in stems ception of the species named above in this par and leaves in that order. Our data support those agraph. all of the vessel elements thus far de concepts, and also the marked disjunction in scribed in roots of Haemodoraceae have simple Haemodoraceae between (mostly) specialized perforation plates. The situation with respect to vessel elements in roots combined with highly perforation plates on tracheary elements of primitive vessel elements—arguably tracheids— stems of Haemodoraceae is much less clear. in stems. As in that instance, the physiological Cheadle (1968) figured a long scalariform end advantage of such a disjunction may lie in the wail [or a tt-acheary element of Anigozanthos ru ephetneral functional life-span of roots that con fus, with stippling on the pits (perforations?) of duct rapidly (in the moist part of a highly sea the end wall to indicate his uncertainty. He also sonal environment) combined with perennial reports similar questionable vessel elements in stems that persist through dry seasons and con stems of Lachnanthe.c tincroria (F. Gmel.) duct more slowly tCarlquist )975). These ideas Sprague H Lachnanthes caroliniana (Lam.) have been confirmed by the work of Ewers et Dandy] and Xiphidium album Lam. [ Xiphi al. (1992). cjiu Pr? caeruieum Aubi.]. With regard to shoot tracheary elements of Haemodoraceae, Cheadle Literature Cited states that ‘end walls of tracheids in stems ap APG II. 2003. An update of the Angiosperm Phylog pear much like perforation plates and may in- eny Group cianificatlon for the orders and families 2OO5 SCHNEIDER AND CARLQUIST: ORIGIN OP VESSELS IN MONOCOTYLEDONS 383

of flowering plants: APG 11. Ba. J. Linn. Soc. 141: Root—stem junctions of a desert monocotyledon 399—436. and a dicotyledon: hydraulic consequences under CARLQUIST, S. 1975. Ecological strategies of xylem wet conditions and during drought. New Phytol. evolution. University of California Press, Berkeley. 121: 377—385. CA. EARN, A. l954. Metaxylem elements in some families CARLQUIST, S. 1992. membrane Pit remnants in per of the Monocotyledoneae. New Phytol. 53: 530— foration plates of primitive dicotyledons and their 540. significance. Am. J. Bot. 79: 660—672. GOLDBLATT. P. AND J. MANNtNG. 2000. Haemodora CARLQI.TIST. S. AND F. L. SCHNEIDER 1997. Origins arid ceae. Pages 92—93 in Cape Plants: a Conspectus of nature of vessels in monocotyledons. 1. Acorus. mt. the Cape Flora of Soulh Africa. Strelitzia 9. Na J. Sci. 158: 51—56. tional Botanical Institute of SouthAfrica, MBG CARLQUIST, S. AND E. L. SCHNEIDER. 1998a. Origins Press, Missouri Botanical Garden, St. Louis, MO. and nature of vessels in monocotyledons. I Low J0HAN5EN. D. 1940. Plant microtechnique. McGraw iuceae. with comments on rhizome anatomy. Blu Hill, New York, NY. rnea 43: 219—224. MACFARLANE, T. a, S. D. HOPPER, It W. PLRDIE. A. CARLQUISI, S. AND F. L. SCHNEIDER. 1998b. Origin and S. GEORGE, AND S. K. PATRICK. 1987. Haemodor nature of vessels inonocotyledons. in 5. Araceae aceae. Pages 55—148 in Flora of : Hyda subfamily Colocasioideae. Bot. J. LiIIn. Soc. 128: tellaceae to Liliaceae. Australian Government Pub 71—86. lishing Service. Canberra. CHASE, M. 2004. Monocot relationships: an overview. SCHNEIDER. B. I.. AND S. CARLQLIST. 1997. Origins and Am. 3. Dot. 9: 1645—1655. nature of vessels in monocotyledons. 2. Juncagi CHEADI.E, V. I. 1942. The occurrence and types of ves naceae and Scheuchzeriaceae. Nordic J. Bol. 17: sels in the various organs of the plant in the Mon 397-40 I. ocotyledoneae. Am. J. Bot. 29: 441—450. SCHNEtDER, E. L. AND S. CARLQUIST. 1998 Origin and CHEADLE. V. 1. 1944. SpecializatiDn of vessels within nature of vessels in monocotyledons. 4. Araceae the xylem of each organ in the Monocotyledoneae. subfamily Philodendroideae. I Torrey Bot. Soc. Am. J. Bat. 31: 81—92. 125: 253—260. CHEADLE, V. 1. 1968. Vessels in Haemodorales. Phy SCHNEIDER, F. L. AND S. CARLQUIST. 2005 Origin and tomorphology 18: 412—420. nature of vessels in monocotyledons. 6. ifanguana DAVIS, J., STEVENSON, 0. W. 6. PETERESEN, 0. SEBER0. (Hanguanaceae). Pacific Sci. 59: 393—398. L. M. CAMPHELL, J. V FRELDeNSTEIN, 0. H. GOLD SIMPSON, M. 6. 1998. Haemodoraceae. Pages 212—222 MAN, C R. HARDY. A. MICHELANOELI, F M. P. SIM in The families and genera of flowering plants- IV MONs. C. SPECHT. D. F VERGARA-SILVA, AND M. Flowering plants. Monocotyledons: Alistinatanae GALDOLPO. 2004. A phylogeny of the monocots, as and Commelinanae (except Gramineae) (K. Ku inferred from rbcL and atpA sequence variation, bitzki, ed). Springer Verlag, New York. NY. and a comparison of methods for calculating jack SIMPSON, M. G. AND W. C. DtciusoN. 1981. Compar nife and bootstrap values. Syst. Ba. 29: 467—5 10. ative anatomy of Lnchnanthes and Lophiola (Hae EWERS. W.. F 6. B. NORTH. AND P. S. NOHEL. t992. modoraceae). Flora 171: 95—113.