~
o
~
~
~$&he
c Limean
a ~ So&&
~ o (1998),
~ ~ 128:
l 71-86. With 33 figures
Article ID: bt980 181
Origin and nature of vessels in monocotyledons.
5. Araceae subfamily Colocasioideae
SHERWIN CARLQUIST FLS AND EDWARD L. SCHNEIDER
Santa Barbara Botanic Garden, 1212 Mission Canyon Road, Santa Barbara,
CA 93105, U.S.A.
Received JanuaZy 1998; accepted f o r publication April 1998
Tracheary elements from macerations of roots and stems of one species each of five genera
of Araceae subfamily Colocasioideae were studied by means of SEM (scanning electron
microscopy). All of the genera have vessel elements not merely in roots, as previously reported
for the family as a whole, but also in stems. The vessel elements of stems in all genera other
than Syngonium are less specialized than those of roots; stem vessel elements are tracheid-like
and have porose pit membrane remnants in perforations. The perforationswith pit membrane
remnants demonstrate probable early stages in evolution of vessels from tracheids in primary
xylem of monocotyledons. The vessel elements with such incipient perforation plates lack
differentiation in secondary wall thickenings between perforation plate and lateral wall, and
such vessel elements cannot be identified with any reliability by means of light microscopy.
The discrepancy in specialization between root and stem vessel elements in genera other
than Syngonium is ascribed to probable high conductive rates in roots where soil moisture
fluctuates markedly, in contrast with the storage nature of stems, in which selective value
for rapid conduction is less. Syngunium stem vessels are considered adapted for rapid conduction
because the stems in that genus are scandent. Correlation between vessel element morphology
and ecology and habit are supported. Although large porosities in vessel elements facilitate
conduction, smaller porosities may merely represent rudimentary pit membrane lysis.
0 1998 The Linnean Society of London
ADDITIONAL. KEY WORDS:-aroids - basal monocotyledon evolution
plant physiology - tracheary elements - vessel elements - xylem evolution.
-
conductive
CONTENTS
Introduction . . . .
Material and methods
Results . . . . .
Alocasia . . . .
Caladium . . .
Syngoniurn . . .
Colacasia . . .
Xanthosuma . . .
Conclusions . . . .
References . . . .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. .
. .
. .
. .
. .
. .
. .
. .
. . . . . . . .
. . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. .
. .
. .
. .
. .
. .
. .
. .
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
. .
.
.
.
.
.
.
. . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . .
0024-4074/98/090071+ 16 $30.00/0
71
72
73
74
74
76
80
80
80
83
86
0 1998 The Linnean Society of London
Cheadle (1942) and Il’agner (1977) summarized data on vessel occurrence in
monocotyledons. The) concluded that vessels originated in roots of monocotyledons,
and phylogenetically entered stems and leaves, in that order. Both of these authors
agree that specialization of vessels (as indicated by reduction of number of bars per
perforation plate) similarly progresses phylogenetically from roots to stems to leaves
in monocotyledons. Families judged to be primitive within monocotyledons by
Cheadle (1942) were those that have vessels in roots only (a few reduced aquatic
monocotyledons lack vessels entirely, presumably by secondary loss, and thus form
an exception to this gencralization). We are interested in the monocotyledon families
found by Cheadle (1942) and Ll’agner (1977) to have vessels in roots only, because
we thought earlier stages in evolution of vessels in primary xylem would likely be
revealed in these families. However, such hypothetical early stages in vessel evolution
are not visible with light microscopy. End walls of tracheary elements appear very
much like lateral walls in primary xylem as seen with light microscopy, whereas
SEhl can reveal presence of perforations, with or without porous or weblike pit
membrane remnants. \Ye found such elements in SEM studies of tracheary elements
of Xymphaeales (e.g. Schneider & Carlquist, 1995; Schneider & Carlquist, 1996).
M’e have now applied SEM methods to several families of monocotyledons thought
to have vessels in roots only: Acoraceae (Carlquist & Schneider, 1997),Juncaginaceae
and Scheuchzeriaceae (Schneider & Carlquist, 1997), Lowiaceae (Carlquist &
Schneider, in press) and Araceae subfamily Philodendroideae (Schneider & Carlquist,
in press). Our findings re\eal that all of these families have vessels in roots, but they
1iaF.e vessels in stems (rhizomes) as well. The vessels in rhizomes are less specialized
than those of roots, and have end walls similar to those of lateral walls. The end
~ a l l of
s tracheary elements in stems of most of the species studied have pit membrane
remnants that range from Nebs to near-intact pit membranes that contain porosities
of \.arious sizes. The occurrence of these pit membrane remnants is of significance
to us. because it is interpretable as the earliest stage in vessel evolution, in accordance
m-ith data from secondary xylem of woody dicotyledons (Carlquist, 1992)and primary
xylem of Nymphaeales (see abo\.e). These monocotyledons are giving us new insights
into the histological mode of origin of vessels. Vcssels were reported in stems of
aroids on1) in Pothos by Solereder & hleyer (1928), although those authors found
\msels in neither stems nor roots of Caladium. Hotta (1971) reported stem vessels
in the scandent aroids E)$mnnum, Rhaphidophora, Pothos, and Srindapsus, although
interestingly. Hotta (197 1) reported “vesselform tracheids” in stems of ten other
genera of aroids. Clear demonstration of pit membrane presence or absence in
tracheary elements is possible only with SEhI, so that Hotta’s ‘vesselform tracheid’
categor). sipifics tracheary elements with differentiated end walls, but he was unable
to demonstrate lack of pit membranes in the end walls of those species because he
used light microscopy.
The comparative nature of our studies permits us to see differences in vessel
specialization within families related to habit and ecology. Thus, the only genus of
Nymphaeaceae that is not entirely aquatic, Barclaya, has the most prominent
development of porosities in end walls of tracheary elements (Schneider & Carlquist,
1 995). In Araceae subfamily Philodendroideae, accelerated specialization of vessels
in roots apparently characterizes ilnthurium, a genus that ranges from terrestrial to
epiphytic. and Zantedexhza, a genus with storage corms that produce roots during
VESSELS IN ARACEAE SUBFAMILY COLOCASIOIDEAE
13
wet seasons. A relationship between greater degree of habitat seasonality and
accelerated specialization of vessels was hypothesized for monocotyledons (Carlquist,
1975). This hypothesis assumes that the shorter the season during which moisture
is available to roots, the greater the selective value of vessels that could conduct
water more rapidly (vessels with fewer bars per perforation plate, and, by extension,
with perforations unimpeded by pit membrane remnants). Araceae are of considerable importance in searching for this correlation. The family is a large one,
and occupies a relatively great range of ecological habitats, although these habitats
are less highly seasonal and extreme than those in which monocotyledons with more
specialized vessels (e.g. Poaceae) grow. The nature of correlation between vessel
distribution and morphology and ecology is best demonstrated within families such
as Araceae in which genera and species that are relatively closely related to each
other genetically have evolved into diverse habitats.
We are structuring our survey of Araceae on the basis of subfamilies. We are
using the subfamilies and tribes recognized by Behnke (1995). Behnke’s inframilial
classification is much like those of earlier authors (e.g. Dahlgren, Clifford & Yeo,
1985), but includes original data developed by Behnke (1995). Tribal classification
is highly tentative at present (see French, Chung & Hur, 1995) but the genera in
the present study are assigned by Behnke (1995) to the following tribes of subfamily
Colocasioideae: Caladium, Xanthosoma (Caladieae); Syngonium (Syngonieae); Alocasia,
Colocasia (Colocasieae). We have not included any genera Behnke assigned to the
tribes Steudnereae or Thomsonieae. The cladogram produced by French, Chung
& Hur (1995) on the basis of chloroplast DNA evidence shows genera typically
assigned to Colocasieae on clades separated from each other by other Araceae
(Areae and Arophyteae), and thus Colocasieae becomes a polyphyletic group.
Considering those results, citation of the number of genera and species in Colocasieae
would be premature.
MATERIAL AND METHODS
Roots and stem portions were preserved ih 50% aqueous ethanol. Macerations
of portions of roots and stems richest in vascular tissue were prepared with Jeffrey’s
Fluid (Johansen, 1940) and stored in 50% ethanol. Macerations were spread onto
aluminium stubs, dried, sputter coated, and examined with a Bausch and Lomb
Nanolab SEM.
The species studied and their sources were: Alocasia X amazonica Andri: (commercial
nursery, Santa Barbara); A. macrorhiza Schott (commercial nursery, Santa Barbara);
Caladium bkolor Vent. (commercial nursery, Santa Barbara); Colocasia exulenta (L.)
Schott (cultivated,University of California at Santa Barbara); Syngonium sp. (cultivated,
University of California at Santa Barbara); Xanthosoma sagttfolium (L.) Schott (cultivated, Hope Ranch Park, Santa Barbara). Roots and stems of one plant per species
were fixed, although several roots and usually one stem per species were studied.
Primary walls of primary xylem tracheary elements of Araceae are delicate, and
even with very careful maceration, some tearing occurs. These tears are clearly
different from porosities one sees in end walls of vessel elements (e.g. Figs 4, 5, 15,
30, 3 1) or kinds of webbing (Figs 32, 33). An end wall with small porosities but with
a prominent tear is shown in Figure 17, emphasizing the difference in outline
74
S. CARLQUIST AND E. L. SCHNEIDER
between natural lysis and rips due to handling. Any opening in a primary wall that
did not clearly appear natural by virtue of circular to oval outline and smooth edges
was not considered in interpreting our preparations. Porosities due to lysis are
apparent only in primary walls of tracheary elements, not in primary walls of other
cell types. Thin-walled parenchyma cells, particularly in starchy rhizomes, collapse
during macerations and may be deposited on the surface of tracheary elements
when preparations are made. Only observation of more numerous elements provides
the basis for reliable obsenations on walls of tracheary elements when such
parenchyma cell remnants are present. Although our macerations were air dried,
we believe that the porosities in pit membranes and the presences or absences of
pit membranes we report do not represent artefacts, or at most represent minimal
alteration of natural conditions. In our experience, as with studies of xylem in
dicotyledons (Meylan & Butterfield, 1978; Carlquist 1992), tears in pit membranes
can be induced by handling techniques, but the ovd to circular pores reported are
not likely induced by air drying or handling. The distribution of porosities within
tracheary elements is significant in this regard: pit membranes with porosities were
observed at the upper or lower ends of perforation plates in which the central
perforations lacked membranes entirely, and in no case did we observe porose pit
membranes in lateral walls of vessel elements. The distribution patterns of porose
pit membranes are consistent with those reported in our earlier studies of monocotyledons as well as in studies of dicotyledons (Meylan & Butterfield, 1978; Carlquist
1992). Larger porosities were visible with light microscopy in pit membranes of
perforation plates in some dicotyledons for which liquid-preserved material was
present, and in which the material was in liquid at all stages during processing.
Elizabeth Wheeler (pers. comm.) claims seeing porosities in pit membranes of
tracheary elements of Etz3 that have been air dried, and believes that these structures
may he related to this technique. The phenomenon we term ‘striations’ may
represent, at least to some degree, a product of air drying.
RESULTS
Alocasia (figs 1-15}
Portions of vessel elements of small-diameter roots are illustrated in Figures 1-5.
In these vessel elements, bands of secondary wall material are predominantly helical
or transitional to scdariform. Elements with annular bands of secondary wall material
ordinarily do not survive the maceration process, and one finds isolated rings of
secondary wall material in preparations such as that shown in Fi<pre20, upper left.
In Fi‘gure 1 , there is a short perforation plate (traversed by six bars) that is rather
similar in morphology to lateral walls, but the presence of primary walls on other
surfaces of the vessel element is readily evident. In Figure 2, a lanceolate perforation
plate (centre) is so similar to lateral walls that it could not have been detected by
means of light microscope. Figure 3 illustrates the lateral walls of a vessel element
in which axially-oriented striations are evident on the primary wall portions. The
element in Figure 3 is of interest in that it shows transitions between a helical and
a scdariform pattern: weak vertical bands of secondary wall material interconnect
the stronger horizontal bands (near left), whereas forkings of the horizontal bands
VESSELS IN ARACEAE SUBFAMILY COLOCASIOIDEAE
75
Figures 1 4 . Vessel element portions from small-diameter roots (1-5) and large-diameter roots (6) of
Alocasiu mwrhizu. Fig. 1. Tip of a vessel element, showing a perforation plate with six bars at the base
of an end wall facet. Fig. 2. Perforation plate portion, partially covered at tip by lateral walls of
adjacent vessel elements, left and right. Fig. 3. Lateral wall of vessel, showing axially-oriented striations
in primary wall and nature of secondary wall framework. Fig. 4. Perforation plate portion to show
absence of pit membranes (three perforations at left) and, elsewhere, perforation plates with pit
membranes that contain porosities of various sizes. Fig. 5. Pit membrane remnant in a perforation,
showing porosities with a wide range of sizes. Fig. 6. Transition between a perforation plate (above,
indicated by dark holes) and lateral wall (below, pit membranes intact) from vessel shown in Fig. 7
Scale bar = 10 pm; except in Figs 4 & 5 where scale bars = 5 pm.
7fi
S C.4RLQCIST XTD E. L. SCHNEIDER
(far right, bottom) also indicate deviation from a strictly helical pattern. Scalariform
patterns of secondary wall material occur in a few late metaxylem elements (Fig. 4).
Modification of primary wall remnants into perforations by means of development
of large porosities or even (left) complete dissolution of pit membranes is evident.
One pit membrane for a scalariform pit is shown in Figure 5; this membrane is
interesting in that oval (lower right) or circular (most of the remainder) porosities
are present, and these porosities vary from relatively large to a size barely resolvable
with our equipment.
Figures 6-14 show portions of vessel elements from metaxylem of a large-diameter
root (c. 4 mm in diameter). The perforations on the end walls of the vessels in
Figures 6-8 are only slightly wider than the spaces between secondary wall bands
on lateral walls, and thus such perforation plates might not be readily identified
with light microscopy, or at least, the limits of such perforation plates might be
difficult to define (e.g. Fig. 6, top, shows perforations and porosities whereas Fig. 6,
bottom, shows lateral intact pit membranes). However, with SEM, there is no
question about the limits of the perforation plates in these cells. The perforation
plate in Figure 9, with its wide perforations and the reticulate pattern of bands of
secondary wall material, is easily distinct. Even in the perforation plates that are
relatively different from adjacent areas of primary wall, transitions from the end of
the perforation plate into the lateral wall occur, and only SEM can delineate where
primary wall material is present and where perforations occur.
Vessel elements of large diameter from the wider-diameter roots of A. macrorhiza
are shown in Figures 10- 14. The vessel elements of Figures 10 and 1 1 are noteworthy
in that not only do the oblique terminal facets qualify as perforation plates, adjacent
lateral facets also lack pit membranes, although other lateral wall areas with pit
membranes are present in both of these vessel elements. Both of the perforation
plate facets of Figure 10 are transitional between helical and scalariform in pattern,
in contrast with the lateral walls (Fig. 10, bottom), which are scalariform. In Figure
1 1, extensive interconnections between the bars of the perforation plates are
evident. Similarly, interconnections that could be termed reticulate characterize the
perforation plates; some perforations are much larger than others. In both Figures
12 and 13, the end wall facet is clearly a perforation plate, and all other facets, in
which primary walls are evident, are lateral walls. Figure 14 illustrates a portion of
a perforation plate from a near-helical vessel element; delicate weblike remnants of
primary wall material may be seen.
Vessel elements are difficult to obsene in rhizomes of Alocasia because the vascular
bundles are sparse and are embedded in large quantities of parenchyma, from which
the tracheary elements do not separate readily. However, in A. arnazonica, we observed
a few vessel element end walls in which pit remnants containing porosities of various
sizes are present (Fig. 15).
Caladium (Figs I 6- I9)
Tracheary elements from roots of C. bicolor are shown in Figures 16 and 17. The
patterns of secondary walls are nearly helical (Fig. 16) to nearly scalariform (Fig.
17). The tracheary elements shown in Figure 16 illustrate difficulties encountered
in obsemation: at far left, a parenchyma cell covers part of the left tracheaiy element;
the tracheary element at right might be a tracheid, but since one cannot see all
VESSELS IN ARACEAE SUBFAMILY COLOCASIOIDEAE
77
Figures 7-1 1. End walls of vessel elements from large root of Alocm'a macrorhira. Fig. 7. Long perforation
plate; perforations at the bottom are transitional to lateral wall pits and contain porosities (see Fig. 6);
some pit membranes contain porosities. Fig. 8. Perforation plate with numerous slender bars. Fig. 9.
Perforation plate with few bars and several interconnectingstrands. Fig. 10. Wide vessel with perforation
plates on two end wall facets; bottom quarter of photograph = lateral wall. Fig. 1 1. Wide vessel with
perforation plate on end wall facet, but also perforations on wall at right, below; lateral wall with pit
membranes at left. Scale bars = 10 ,urn.
faces of this cell, one cannot be sure that perforation plates are absent. The porosities
in pit membrane remnants of the vessel element in Figure 17 are mostly small,
ranging downwards in size to the limit of resolution of our SEM equipment.
78
S. CAKLQUIST AR'D E. L. SCHNEIDER
Figures 12-1 5. Vessel portions Alocmiu. Figs 12-14. Vessels from large root of A. macrurhiza. Fig. 12.
Tip of wide vessel, with near-transverse perforation plate traversed with numerous slender anastomosing
bars. Fig. 13. Tip of wide vessel, with perforations ranging widely in size and shape. Fig. 14. Perforation
in which web-like pit membrane remnants are present. Fig. 15. Two perforations from vessel of stem
of rllocasia umaz-unicu; pit membrane remnants are porose. Scale bars = 10 pm; except in Fig. 15 where
scale bar = 5 pm.
In Figures 18 and 19, portions of two perforation plates from stems of C. bicolor
are illustrated. The two perforation plates contrast with each other. In Figure 18,
the perforations are merely large openings in primary wall, openings that cannot
be distinguished from the adjacent primary wall by means of light microscopy. In
VESSELS IN ARACEAE SUBFAMILY COLOCASIOIDEAE
79
Figures 16-19. Portions of tracheary elements of Caladium bkOl07. Figs 16 & 17. Tracheary elements
from root. Fig. 16. Two tracheary elements (the one at left partially covered by a collapsed parenchyma
cell) to show lateral wall pattern. Fig. 17. Portion of a perforation plate; porosities of various sizes are
present in the pit membranes (a large tear present in membrane, second perforation from bottom).
Figs 18 & 19. Perforation plates from vessels of stem. Fig. 18. Perforation plate in vessel with helical
bands; the perforations are oval holes dissolved from areas of primary wall; remnant primary wall
webbing at top. Fig. 19. Narrow perforations with pit membrane remnants that contain circular
porosities of various sizes. Scale bars = 10 pm; except in Fig. 17 where scale bar = 5 pm.
80
S. CARLQUIST AND E. L. SCHNEIDER
citing this fact, we wish to emphasize not the limitations of equipment but the
curious morphology of the perforation plate, a morphology that is not illustrated in
textbooks or reviews. In the two perforations at the top of Figure 18, fine remnant
webs of primary wall material persist. In the perforation plate portion illustrated in
Figure 19, we see notably narrow perforations from a perforation plate that had
both narrow and wide perforations. In the narrow perforations illustrated, we can
see small circular porosities in the remnant pit membranes.
Syngonium (Figs 20-23)
The root vessel element of Syngonium (Fig. 20) shows a long, wide perforation
plate; near the bottom of the portion shown, lateral walls covered with primary wall
are evident because the perforation plate has buckled somewhat and thus the side
M-alls are brought into view. The perforations of this vessel are devoid of pit
membrane remnants or nearly so. The helices of secondary wall are uniform and
do not change in thickness or in spacing on the perforation plate as compared to
lateral walls. O n some other root vessel elements, perforations bear pit membrane
remnants (Fig. 21), but the remnants are never common. In vessel elements of
rhizomes, pit membrane remnants are also not common. The perforation plate of
Ficgure 22 is devoid of pit membrane remnants. Because the vessel element of Figure
22 bears helical secondary wall bands, the only feature defining the perforation
plate is the absence of primary wall in the oval perforations. In Figure 23, pit
membranes remnants are porose and confined to lateral ends of otherwise clear
perforations.
Colocasia (Fks 24, 25)
In Colocusiu, there is a marked contrast between vessel elements of roots and those
of rhizomes. A helical vessel element of a root (Fig. 25) is long, with perforations
clear of pit membrane remnants. A helical vessel element from a rhizome (Fig. 24)
bears extensive porose pit membrane remnants; the pores in the area shown are
extremely small.
Xanthosoma (Figs 26-33)
The roots of X. sugittzjilium studied are relatively large (c. 4 mm) in diameter, and
thus the wide metaxylem vessel elements (Figs 26-29) are to be expected, based on
a similar correlation in Alocmiu. Each of the four vessel elements has a clearly defined
end wall perforation plate. However, the perforation plates differ in pattern. In
Figure 26, the secondary wall bands outlining perforations form a reticulate pattern,
verging on what would be called multiperforate in vessels of dicotyledons. In the
vessels of Figures 27 and 28, the bars of the perforation plates are mainly transverse,
hut are interconnected by diagonal and vertical strands of secondary wall material.
The lateral walls of the vessel in Figure 27 have a typical scalariform pattern, with
strands of secondary wall material either horizontal or vertical, thus delimiting pits
that are rectangular to oval. The lateral walls of the vessel in Figure 28, however,
VESSELS IN ARACEAE SUBFAMILY COLOCASIOIDEAE
81
Figures 20-25. Vessel portions from Syngonium sp. (2G23) and Colocasia esculenta (24, 25). Fig. 20. Long
perforation plate from root. Fig. 21. Perforation plate portion from root, showing pit membrane
remnants with weblike conformation. Fig. 22 Perforation plate portion from stem; perforations are
oval in openings in primary wall; helical bands of secondary wall thickening traverse the perforation
plate. Fig. 23. Perforation plate portion from stem, showing porose pit membrane remnants at ends
of perforations. Fig. 24. Perforations from vessel of stem; pit membranes have striations and very small
pores. Fig. 25. Perforation plate from root. Scale bars= 10pm; except in Figs 21, 23 where scale
bars = 5 pm.
S. C.\RLQCISI' AYI) E. L. SCHNEIDER
Figures 26-29. Perforation plates fiam nietaxylrm vrssels of wide root of .k'mzn,2t/iosomn . i a ~ t t ~Fig.
~ ~ ~ ~ i .
26. Trans\.erse perforation plate with multiperforate conformation. Fig. 27. Oblique perforation plate
Lvith interconnections among bars. Fig. 28. Oblique perforation plate with interconnections among
liars; interconnections behvcen horizontal secondary waU bands of lateral wall are oblique to vertical.
Fig. 29. Portion of long highly oblique perforation plate with rather thick bars. Scale bars = 10 pm.
feature mainly diagonal strands of secondary wall material interconnecting the
horizontal bands, and thus the pits are trapezoidal in shape. The perforation plate
in Fi'gurc 29 is an almost perfectly scalariform plate, with only a few interconnections
between the bars.
VESSELS IN ARACEAE SUBFAMILY COLOCASIOIDEAE
83
By contrast, the vessel elements of the rhizome of X.sagitt$lium are scalariform
and bear extensive pit membrane remnants (Figs 30-33). A portion of what we
believe should be termed a perforation plate in Figure 30 contrasts with the adjacent
lateral wall' (at right in Fig. 30) by showing thin weblike pit membranes with
rudimentary circular to oval small porosities (wider irregular openings are interpretated as artifacts due to handling). The perforation plate of Figure 30 is of
special significance because we see a vessel element minimally different from a
tracheid.
In Figure 31, the portion of a perforation plate shown illustrates a range of pit
membrane conditions, from smooth with porosities to strands separating large spaces.
The perforation plate portion of Figure 32 shows perforations variously covered
with pit membrane remnants, but the remnants are highly porose and weblike. In
Figure 33, perforations are mostly clear of pit membranes, but delicate webs and
strands of primary wall material remain.
CONCLUSIONS
The vessel elements of Araceae subfamily Colocasioideae are of critical importance
in understanding the origin of vessels in primary xylem of angiosperms, because
they show features and stages not evident in other groups. These features and stages
are revealed by SEM and have been described only in very partial'ways in light
microscopy studies. Traditionally, the evidence for existence of perforation plates in
primary xylem has been the difference between end walls and lateral walls in
thickness of bars of secondary wall material and spacing of the bars. Our studies
show that origin of vessels in primary xylem begins with thinning and lysis of the
pit membrane, resulting in porose or weblike pit membranes (such a beginning stage
is strikingly shown in Fig. 30). Other similar vessel elements show greater degrees
of lysis, leaving weblike or strandlike pit membrane remnants (e.g. Figs 31-33). Such
alterations in pit membranes are, in helical tracheary elements, often not accompanied
by differences between end walls and lateral walls, and thus perforation plates of
these elements would not be detected by light microscopy. The wider scalariform
tracheary elements of the species studied have end walls that are clearly end walls,
which would be revealed by either light microscopy or by means of SEM. However,
SEM does clarify limits of perforation plates even in the scalariform tracheary
elements, in which transitions between intact lateral wall pit membranes and
perforations with porose pit membrane remnants (e.g. Fig. 6) or no remnants occur.
The significance of these details of vessel element structure is not merely that SEM
is the instrument of choice for revealing them, but that in instances such as the ones
we have cited, we can see the transition between tracheids and incipient vessel
elements, how this transition occurs, and what stages in disappearance of pit
membranes follow. Araceae prove to be an ideal group for showing modes of origin
of vessel elements and opening stages in vessel specialization.
Both Cheadle (1942) and Wagner (1977) claimed vessels for roots only in the
Araceae they studied. To be sure, the number of species they sampled was small.
Solereder and Meyer (1928) had earlier reported vessels in roots of Pothos, and Hotta
(197 1) had reported vessels for stems of Epipremnum, Pothos, Rhaphidophora, and
Sandupsus. All of the Araceae we have studied, both in our earlier study (Schneider
S ChKLQLIST XYD E. L. SCHNEIDER
E'i<gurcs30--33. Portions of perforation plates from stem of Xanthosoma sugittlfolium. Fig. 30. Perforation
plate (left! with thin porose pit membrane remnants (tears in membranes are artifacts); lateral wall at
righr. Fig. 31. Perforations with pores of various sizes. Fig. 32. Perforations with webllke pit membrane
remnants. Fig. 33. Perforations pius piinlay wall of' lateral wall, left and right); perforations contain
delicatr threadlike pit membrane remnaiits. Scale bars = 5 pm.
and Carlquist, in press) and the present one, have vessels in stems (rhizomes) as well
as roots. The only exception to this was zantedeschia albo-maculata Baill., in which
\.essels were not obsemed with certainty in stems. Presence of vessels also characterized
stems as well as the roots of other monocotyledon families we have examined:
VESSELS IN ARACEAE SUBFAMILY COLOCASIOIDEAE
85
Acoraceae (Carlquist & Schneider, 1997); Juncaginaceae and Scheuchzeriaceae
(Schneider & Carlquist, 1997); and Lowiaceae (Carlquist & Schneider, in press). In
all of these families, both Cheadle (1942) and Wagner (1977) reported vessels in
roots only. Our demonstration of vessels in stems of all of these probably is based
on the fact that vessels in stems in these families are less specialized than those of
the roots, and the tracheidlike nature of the stem vessel elements in the monocotyledons we have studied thus far prevented identifjrlng them as vessel elements
when light microscopy was used. Our findings are not strongly dissimilar from those
of Cheadle (1942) and Wagner (1977), in that the concept they endorsed of vessels
originating in roots of monocotyledons and progressing upward in the plant appears
to be confirmed; however, what they interpreted as tracheids in stems prove to be
vessel elements with a low level of specialization.
The role of habit and ecology in degree of specialization of vessel elements is
clearly demonstrated by our studies. T o be sure, habit and ecology are not entirely
separable: for example, corm-like stems that withstand periods when soil dries (e.g.
Zantedeschia)represent a habit or growth form related to ecology. In Araceae subfamily
Colocasioideae, the genera Alocasia, Caladium, and Colocasia have this habit: stems
are corms that store both starch and water. In this growth form, roots form as soil
moisture becomes available, and if the moisture is seasonally limited, the roots must
absorb water with appropriate rapidity. In the three colocasioid genera studied with
this growth form, the vessels of roots show a higher degree of specialization (long
and/or wide perforation plates with no pit membrane remnants) for rapidity of
conduction. In the three genera, the stem vessel elements show a much lower degree
of specialization (pit membranes often intact, the membranes with pores of various
sizes). The difference between root and stem vessel elements in these genera is rather
marked, and relates to the habit, which in turn is based upon adaptation to a wet
season of finite duration. Xanthosoma sagittfolium superficially appears to have a
different habit, in that its stems are upright, rather than condensed, cormlike, and
subterranean. Nevertheless, the stems of X . sagittzjilium are very thick and succulent,
and as organs for storage of water and starch are functionally essentially like the
stems of Alocasia, Caladium, and Colocasia. Thus, the vessel element pattern in X.
sagittzjilium (wide and with clearly defined perforation plates in roots, narrower
with porose and weblike pit membrane remnants in perforations in stems) is
understandable, and is like the vessel pattern in the three genera listed.
Syngonium, by contrast, has relatively vessel elements of similar degrees of specialization (perforation plates scalariform, with small porose pit membrane remnants)
in stems and in roots. Syngonium has markedly elongate, scandent stems in which
the conductive characteristics would be expected to simulate those of scandent
dicotyledons or of monocotyledon roots in habitats that dry periodically. By definition,
the adventitous roots of a scandent monocotyledon occur in a habitat with short
periods of moisture availability.
Degrees of specialization of vessels in monocotyledons, and the organographic
distribution of vessels with different degrees of specialization throughout the plant
body in monocotyledons, were hypothesized to be correlated with ecology (Carlquist,
1975). Our studies confirm those hypotheses. The colocasioid and philodendroid
aroids, in addition, show that particular habits and habitats are correlated with
degrees of vessel specialization and their organographic distribution, and detailed
knowledge of both habit and of habitat of each species in monocotyledons will be
of value in analysis of xylem configurations.
86
S. CARLQ'UIST AWD E. L. SCHNEIDER
The occurrence of holes (which we term porosities) of various sizes in pit
membranes or pit membrane remnants invites the question of what function such
porosities might serve, assuming they are not artefacts. Stem perforation plates of
Araceae subfamily Colocasioideae show that such porosities vary in size from large
(representing a fraction of a perforation) to extremely small (barely resolvable with
our SEM equipment). The coexistence in a single perforation plate and even a
single perforation of such diverse sizes of pores suggests that the porosities represent
early stages and various degrees of lysis, leading to dissolution of pit membranes
that results in perforation plate formation. If these porosities served a particular
conductive function, one would expect them to range within a more limited range
of diameters (e.g. sieve plate pores or pores in the margos of gymnosperm tracheid
pit membranes).
REFERENCES
Behnke H-D. 1995. P-type sieve-element plastids and the systematics of the Arales (sensu Cronquist
1988)-with S-type plastids in Pistia. Plant Systematics and Evolution 195: 87-1 19.
Carlquist S. 1975. Ecological strategus of qlm evolution. Berkeley and Los Angeles: University of
California Press.
Carlquist S. 1992. Pit membrane remnants in perforation plates of primitive dicotyledon and their
significance. Ammkan Journal of B ~ t a g79: 660-672.
Carlquist S, Schneider EL. 1997. Origins and nature of vessels in monocotyledons. 1. Acorns.
Intmiaiional Journal of Plant Sciences 158: 5 1-56.
Carlquist S, Schneider EL. Origins and nature of vessels in monocotyledons. 3. Lowiaceae. Blumea,
in press.
Cheadle VI. 1942. The occurrence and types of vessels in the various organs in the Monocotyledoneae.
American Journal 0sBotany 29: 44 1 4 5 0 .
Dahlgren RMT, Clifford HT, Ye0 PF. 1985. Thefamilies ofthe monocogledons. Berlin: Springer Verlag.
French JC, Chung MG, Hur YK. 1995. Chloroplast DNA phylogeny of the Araceae. In: Rudall
PJ, Cribb PJ, Huniphries CJ,. eds. itilonoco@edon systematics and evolution Kew: Royal Botanic Gardens,
225-275.
Hotta M. 1971. Study of the family Araceae. General remarks. J a ~ a n e s e ~ o u m
OfBotany
l
20: 269-310.
Johansen DA. 1940. Plant microtechnique. New York McGraw Hill.
Meylan BA, Butterfield BG. 1976. The structure oj'.lrm zealand woods. D S I R Bulletin 222. Wellington:
New Zealand Department of Scientific and Industrial Research.
Schneider EL, Carlquist S. 1995. Vessels in the roots of Barclaya rotund$lia. Amm.canJouma1 OfBotaV
82: 1343 1349.
Schneider EL, Carlquist S. 1996. Vessels in Brasenia (Cabombaceae): new perspectives on vessel
origin in primary xylem of angiosperms. American 3ournal of Botany 83: 1236-1 240.
Schneider EL, Carlquist S. 1997. Origins and nature of vessels in monocotyledons. 2. Juncaginaceae
and Scheuchzeriaceae. Aordic Journal ofBotuny 17: 3 9 7 4 0 1.
Schneider EL, Carlquist S . In press. Origin and nature of vessels in monocotyledons. 4. Araceae
subfamily Philodendroideae. Jounlal of The Zrry Botanical Sock9
Solereder H, Meyer 8.1928. $shatische Anatomie der Monokopledonen. III. Princ;Pes, Synanthae,
.SpathiJorrw. Berlin: Gebriider Borntraeger.
Wagner P. 1977. Vessel types in the monocotyledons: a survey. Botanirk Notiseer 130: 383-402.