Research
Vascularization of the Selaginella rhizophore: anatomical
fingerprints of polar auxin transport with implications for the
deep fossil record
Kelly K. S. Matsunaga1, Nevin P. Cullen2 and Alexandru M. F. Tomescu3
1
Department of Earth and Environmental Sciences, University of Michigan, Ann Arbor, MI 48109, USA; 2Department of Biology, San Francisco State University, San Francisco, CA 94132,
USA; 3Department of Biological Sciences, Humboldt State University, Arcata, CA 95521, USA
Summary
Author for correspondence:
Kelly K. S. Matsunaga
Tel: +1 707 826 3229
Email: matsunagakelly@gmail.com
Received: 30 August 2016
Accepted: 16 January 2017
New Phytologist (2017)
doi: 10.1111/nph.14478
Key words: fossil, lycophyte, polar auxin
transport, rhizophore, Selaginella, tracheid.
The Selaginella rhizophore is a unique and enigmatic organ whose homology with roots,
shoots, or neither of the two remains unresolved. Nevertheless, rhizophore-like organs have
been documented in several fossil lycophytes. Here we test the homology of these organs
through comparisons with the architecture of rhizophore vascularization in Selaginella.
We document rhizophore vascularization in nine Selaginella species using cleared wholemounts and histological sectioning combined with three-dimensional reconstruction.
Three patterns of rhizophore vascularization are present in Selaginella and each is comparable to those observed in rhizophore-like organs of fossil lycophytes. More compellingly, we
found that all Selaginella species sampled exhibit tracheids that arc backward from the stem
and side branch into the rhizophore base. This tracheid curvature is consistent with acropetal
auxin transport previously documented in the rhizophore and is indicative of the redirection of
basipetal auxin from the shoot into the rhizophore during development.
The tracheid curvature observed in Selaginella rhizophores provides an anatomical fingerprint for the patterns of auxin flow that underpin rhizophore development. Similar tracheid
geometry may be present and should be searched for in fossils to address rhizophore homology and the conservation of auxin-related developmental mechanisms from early stages of
lycophyte evolution.
Introduction
Rhizophores, seen only in the lycophyte Selaginella, are downward growing axial organs located at branching points of the
shoot and which produce roots near their apex upon reaching the
soil (Jernstedt et al., 1994). Developmentally, rhizophores originate exogenously from meristems positioned in the branching
angle of the shoot, termed angle meristems. Two angle meristems
form at each branching point, a dorsal and a ventral one, either
or both of which can develop into a rhizophore. The homology
of the Selaginella rhizophore has puzzled plant morphologists for
well over a century, owing to a set of morphological and developmental characters that combines root and stem features. Stemlike features of the rhizophore include exogenous developmental
origination (Jernstedt et al., 1992), the expression of class 1
KNOX genes at the apex (like in the shoot apical meristem and
unlike in the root apex of Selaginella; Kawai et al., 2010), and the
capacity of rhizophores and angle meristems to develop as shoots
under specific conditions (Harvey-Gibson, 1902; Williams,
1937; Webster, 1969; Wochok & Sussex, 1976; Jernstedt et al.,
1994; Sanders & Langdale, 2013). However, like roots, the rhizophore does not typically produce leaves, it exhibits positive
gravitropism, and has been demonstrated to have acropetal polar
Ó 2017 The Authors
New Phytologist Ó 2017 New Phytologist Trust
auxin transport (PAT) (Wochok & Sussex, 1974), an auxin transport pattern which in euphyllophytes is seen only in roots. Various authors have argued for or against interpreting the
rhizophore as a stem homolog, a root, or as a novel sui generis
organ, but currently there is no scientific consensus on the nature
of the rhizophore (Jernstedt et al., 1994, and references therein).
Although no structures comparable to the selaginellalean rhizophore are known in extant plants, the deep fossil record of the
lycophyte clade includes several plants exhibiting two types of
structures that share some features with the rhizophore. These
include subaxillary tubercles and the branches derived from
them, which have been reported in several zosterophyll-grade
lycophytes of the Devonian: Crenaticaulis, Gosslingia,
Deheubarthia, Thrinkophyton and Anisophyton (Høeg, 1942;
Banks & Davis, 1969; Edwards, 1970, 1994; Edwards & Kenrick, 1986; Hass & Remy, 1986; Kenrick & Edwards, 1988;
Edwards et al., 1989). Such structures are regularly positioned in
the immediate vicinity of, and basal to, the angle formed at
branching points of axes, and for this reason they are also referred
to as angular organs (Hass & Remy, 1986) (Fig. 1a). While the
function and developmental fate of subaxillary tubercles is
unknown in many of these taxa, branches derived from them
may have functioned as rooting organs in some cases.
New Phytologist (2017) 1
www.newphytologist.com
New
Phytologist
2 Research
(a)
(b)
(c)
Fig. 1 Branching patterns associated with rhizophore-like structures
among lycophytes. (a) Angular organs of basal lycophytes. Diagrammatic
example (at left) shows the divergence of angular organs proximal to the
branching angle of axes. In fossils of Crenaticaulis verruculosus (at right)
the divergence point of the angular organ, which overlaps the main axis, is
indicated by the arrowhead (Banks & Davis, 1969, fig. 18; published with
permission from the Botanical Society of America). Bar, 5 mm. (b)
Divergence of the Selaginella uncinata rhizophore occurs laterally in the
immediate vicinity of the branching angle of the shoot. Bar, 2 mm. (c)
Rooting axes derived from K-branches arise from lateral branches of the
shoot (arrowhead), shown here in a drepanophycalean lycophyte from
Cottonwood Canyon, Wyoming, USA. Bar, 10 mm.
Additionally, there are the root-bearing axes of some Devonian
drepanophycalean lycophytes (Schweitzer & Giesen, 1980; Xu
et al., 2013; Matsunaga & Tomescu, 2016, 2017). These are produced by two closely spaced dichotomies of the shoot, in a manner that is referred to as H- or K-branching (Fig. 1c), and
resemble morphologically equivalent nonroot-bearing axes with
inferred downward growth seen in some Devonian zosterophylls:
Zosterophyllum, Sawdonia and Bathurstia (Lang, 1927; Walton,
1964; Rayner, 1983; Gensel et al., 2001; Hao et al., 2010).
Like the Selaginella rhizophore (Fig. 1b), both the subaxillary
axes and those produced by K-branching, documented in fossil
plants, (1) arise consistently in association with branching points
of the subtending axis; and (2) develop, in many cases, in a
New Phytologist (2017)
www.newphytologist.com
direction different from that of the regular axes or shoots of the
parent plant. Subaxillary branches have been found to extend in a
direction perpendicular to the plane defined by the forking of the
subtending axis (e.g. Banks & Davis, 1969; Kenrick, 1988) and
the axes arising from K-branching point away from the regular
axes or shoots of the parent plant, which sometimes means
downwards (Walton, 1964; Rayner, 1983; Gensel et al., 2001;
Hao et al., 2010; Matsunaga & Tomescu, 2016). In such cases,
the downward direction of growth is probably cued by a positive
gravitropic response (Walton, 1964; Matsunaga & Tomescu,
2016), like that of the Selaginella rhizophore. Furthermore,
like the rhizophore, some drepanophycalean axes arising from
K-branching bear roots (Schweitzer & Giesen, 1980; Xu et al.,
2013; Matsunaga & Tomescu, 2016).
The notable similarities between rhizophores and the organs
described above have prompted direct comparisons with
Selaginella (Banks & Davis, 1969; Rayner, 1983; Hass & Remy,
1986; Kenrick, 1988; Edwards et al., 1989). However, these similarities have not been seriously explored due to the great phylogenetic distance between Selaginella and the fossil lycophytes
characterized by subaxillary branching and K-branching (Kenrick
& Crane, 1997). Nevertheless, while Selaginella is highly derived
among lycophytes, the existence of shared morphological patterns
that span broad ranges of systematic diversity within the clade
indicates that fundamental body plan features can be conserved
over hundreds of million years of evolutionary history. For
instance, morphological identity and homology of the rooting system of extant Isoetes with that of Paleozoic rhizomorphic lycophytes (e.g. Paralycopodites and Sigillaria; Stewart, 1947; Karrfalt,
1980; Jennings et al., 1983; Rothwell & Erwin, 1985; Hetherington et al., 2016; Hetherington & Dolan, 2016) demonstrates conservation of developmental characters and body plan organization
over > 300 Myr. Or, to consider an even broader context, every
major lineage of lycophytes exhibits rooting systems derived from
or incorporating downward growing stems or undifferentiated
axes (Matsunaga & Tomescu, 2017). Together, these observations
imply that hypotheses of homology between the Selaginella rhizophore and the rhizophore-like organs of early lycophytes should
not be regarded as far-fetched, and that addressing them within a
comparative morphology framework is a worthy pursuit.
Branching patterns and vascular architecture have been documented in basal lycophytes with rhizophore-like organs (e.g.
Lang, 1927; Banks & Davis, 1969; Edwards, 1970; Kenrick,
1988; Matsunaga & Tomescu, 2016), but comparable studies
that can address the question of homology have not been conducted on Selaginella. Here we test the hypothesis of homology
between the selaginellalean rhizophore and rhizophore-like
organs of early lycophytes based on comparative anatomy. We
further explore whether anatomical signatures unique to rhizophore development, related to the hormone auxin and applicable to studying the plant fossil record, are present in Selaginella.
Materials and Methods
We surveyed the architecture of rhizophore vascularization in
nine Selaginella species kept in the Dennis K. Walker Glasshouse
Ó 2017 The Authors
New Phytologist Ó 2017 New Phytologist Trust
New
Phytologist
(Humboldt State University): S. involvens, S. wallacei,
S. kraussiana, S. pallescens, S. delicatula, S. doederleinii, S. wildenovii, and two additional species. While the latter two could not
be identified and are hereafter referred to as Selaginella sp. 1 and
Selaginella sp. 2, they are distinctly different from the other seven.
Information on the origin and identity of these two species was
lost from the glasshouse records, and Selaginella species (totaling
c. 700–800 world-wide; Zhou et al., 2016) are notoriously difficult to identify in the absence of information on their geographic
origin. The Selaginella species sampled here span phylogenetic
diversity within the family, covering six different clades as
resolved in the phylogeny of Weststrand & Korall (2016), and
include species with both dorsal and ventral rhizophores and a
range of stele types (Korall & Kenrick, 2002, 2004; Weststrand
& Korall, 2016).
Shoot segments centered around branching points were sampled from each species and prepared by clearing and by paraffin
embedding and serial sectioning on a rotary microtome. For
clearing we followed Ruzin’s (1999) sodium hydroxide–chloral
hydrate protocol, including full-strength commercial bleach
(sodium hypochlorite), followed by safranin staining, and dehydration in an ethanol series; slides were mounted in epoxy resin.
Paraffin embedding followed the standard protocol for plant tissue. Serial sections were cut at 10 lm thickness and stained using
Walker’s Sam quadruple stain protocol, which uses Weigert’s
iron hematoxylin (1% in 10% ethanol), followed by Bismark
Brown (1% in 50% ethanol), then phloxine (1% in 95%
ethanol) and 1 : 1 Fast Green – Orange G (saturated in clove oil).
Slides were mounted using Eukitt mounting medium (O.
Kindler GmbH, Freiburg, Germany).
Micrographs were taken using a Nikon Coolpix E8800 digital
camera mounted on a Nikon Eclipse E400 compound microscope (Nikon Inc., Melville, NY, USA) and an Olympus DP73
digital camera on an Olympus SZX16 microscope (Olympus,
Center Valley, PA, USA). High-magnification images of tracheids in Selaginella were digitally stacked along the z-axis to
provide a consistent focal plane. Focal stacking and figure construction were done in PHOTOSHOP CC (Adobe, San Jose, CA,
USA). Volume renderings based on serial sections were produced
using AMIRA 3D software (FEI, Hillsboro, OR, USA).
Results
Three patterns of rhizophore vascularization
In the nine Selaginella species we surveyed, the architecture of the
vascular supply of the rhizophore falls into three main types: (1)
the vascular bundle supplying the rhizophore (rhizophore trace)
diverges centrally, from the angle between stem stele and branch
stele (referred to as central divergence); (2) the rhizophore trace
diverges from the stele of the main stem, close to but above the
divergence of the branch stele (stem divergence); and (3) the rhizophore trace diverges from the stele of the side branch above the
divergence of the latter from the main stem stele (branch divergence). These patterns are independent of the external position of
rhizophores or angle meristems, which are always positioned in
Ó 2017 The Authors
New Phytologist Ó 2017 New Phytologist Trust
Research 3
the branching angle. Of the nine species, seven exhibit central
divergence, one species is characterized by stem divergence (S. sp.
2), and one by branch divergence (S. wallacei) (Table 1). While
the number of meristeles (one to three) forming the stele of the
main stem (Table 1) adds complexity to the architecture of vascularization in some cases, this does not interfere with the three
main types of rhizophore vascularization. We describe below the
architecture of rhizophore vascularization in four species that
illustrate the range of variation observed: Selaginella kraussiana
(two meristeles) and S. wildenovii (three meristeles) for central
divergence, S. sp. 2 (variable, one or two meristeles) for stem
divergence, and S. wallacei (one meristele) for branch divergence.
Selected species descriptions
S. kraussiana stems have two meristeles (Fig. 2a). At shoot
branching points, one of the meristeles of the main stem splits to
produce two vascular segments. One of the segments continues as
one of the main stem meristeles while the other segment fuses
with the second meristele producing a plexus of vascular tissue.
From this plexus three vascular segments diverge and continue
distal to the branching point: the second main stem meristele,
the stele of the side branch and the stele of the rhizophore. The
three diverge at the same level, with the rhizophore stele diverging from the angle between the other two and extending laterally
and perpendicular to the plane defined by the main stem and
lateral branch (Fig. 3a,b; Supporting Information Video S1).
S. wildenovii stems have three meristeles represented by plates
of vascular tissue flattened in the plane of shoot branching (defined by the main stem and lateral branch) (Fig. 2b). At shoot
branching points, each stem meristele diverges to give rise to one
vascular segment that becomes a meristele of the lateral branch.
At the same level, the three meristeles anastomose. The plexus of
vascular tissues connecting them forms a bridge. This bridge
extends vertically in the angle between the stem and lateral
branch, forming a ridge of vascular tissue developed in a direction
perpendicular to the plane of branching. The vascular supply of
the two rhizophores diverges from the two ends of this bridge,
laterally and perpendicular to the plane of branching (Fig. 3c,d;
Video S2). Overall, the rhizophore vascular strands thus diverge
from a central position in the angle between the steles of the main
stem and lateral branch.
Table 1 The position of rhizophore divergence in nine species of
Selaginella surveyed
Selaginella species
No. of meristeles
Position of rhizophore divergence
S. involvens
S. wallacei
S. kraussiana
S. pallescens
S. sp. 1
S. sp. 2
S. delicatula
S. doederleinii
S. wildenovii
1
1
2
2
2
1 or 2
3
3
3
Central
Lateral branch
Central
Central
Central
Main stem
Central
Central
Central
New Phytologist (2017)
www.newphytologist.com
New
Phytologist
4 Research
(a)
(b)
(c)
(d)
Fig. 2 Serial cross sections through branching points of four Selaginella species showing the vascular architecture of rhizophore vascularization. (a)
Selaginella kraussiana. The main stem has two meristeles (s) and the rhizophore stele (r) arises from the divergence point of the main stem and lateral
branch (b) steles. Bar, 500 lm. (b) Selaginella wildenovii has a complex stelar architecture consisting of three meristeles in both the main stem and the
lateral branch. Two rhizophores are produced, and diverge centrally from the angle between the steles of the main stem and lateral branch (see also
Fig. 3g,h). Bar, 500 lm. (c) Selaginella sp. 2 showing divergence of the rhizophore stele from the stele of the main stem, distal to the divergence of the
lateral branch. Bar, 500 lm. (d) Selaginella wallacei. The rhizophore stele diverges from the stele of the lateral branch. Bar, 500 lm.
In S. sp. 2 rhizophores are produced at branching points of
shoots that have either one meristele (Fig. 2c) or two meristeles.
In both cases, the vascular supply of the rhizophore arises from
the stele of the main stem, distal to the divergence of the lateral
branch. The rhizophore stele initially diverges from the stem stele
in the plane of shoot branching and in the same direction as the
stele of the branch (Fig. 2c); distally, the rhizophore stele curves
in a direction perpendicular to the plane of branching (Fig. 3e,f;
Video S3).
S. wallacei stems are characterized by a simple stele consisting
of a single vascular segment (Fig. 2d). In this species, the
New Phytologist (2017)
www.newphytologist.com
rhizophore stele diverges from the stele of the lateral branch, and
therefore distal to the divergence of the branch stele from the stele
of the main stem. Furthermore, the direction of rhizophore divergence is oblique to the plane of shoot branching (Fig. 2d, 3g,h;
Video S4).
Rhizophore divergence and tracheid geometry
In all the species surveyed, the shape, arrangement and orientation of tracheids at the point of rhizophore divergence exhibit a
consistent geometry. The xylem bundle of the rhizophore is
Ó 2017 The Authors
New Phytologist Ó 2017 New Phytologist Trust
New
Phytologist
Research 5
(a)
(c)
(e)
(g)
(b)
(d)
(f)
(h)
Fig. 3 Three-dimensional reconstructions of the vascular architecture at rhizophore divergence in Selaginella kraussiana (a, b), Selaginella wildenovii (c,
d), Selaginella sp. 2 (e, f) and Selaginella wallacei (g, h). (a, c, g) Views looking down at branching points from positions distal to them. (b, f, h) Views in
the horizontal plane from positions lateral to branching points. (d), (e) are oblique views from positions slightly above branching points. The stele of the
main stem (s) and lateral branch (b) are shown in red, while the rhizophore stele is shown in blue. Note the presence of more than one rhizophore trace in
S. wildenovii. Leaf traces are shown translucent (red) in (a, b, f–h) for clarity. Green structures on the periphery of the stems represent leaves, or parts of
leaves, present on the stem segments.
connected to the subtending shoot by tracheids that arc from the
stele of the main shoot and branch shoot back into the rhizophore vascular supply, toward the rhizophore apex (Fig. 4a–c,
g–k). This consistent pattern is very conspicuous starting with
early stages of rhizophore development and, when two rhizophores are associated with one branching point, is seen in the
vascular supply of both rhizophores. It is worth noting that in
some cases the high tracheid density at branching points, especially when associated with thick sclerified cortical tissues that
hindered clearing, obscured fine details of tracheid arrangement
and shape, making it difficult to obtain good photographs that
illustrate the curved tracheids clearly (Fig. 4l). Importantly, the
geometry of tracheids associated with rhizophore divergence is
clearly different from the geometry of tracheids documented
around leaf traces and shoot branching points that lack rhizophores. In these instances, tracheids consistently form a
basipetally convergent pattern with no tracheids that arc back
apically (Fig. 4d–f).
Discussion
The architecture of the rhizophore vascular supply is
variable
The rhizophore-like axes of Devonian lycophytes exhibit several
patterns of vascular architecture. These patterns provided the
starting point for our comparative approach to the homology of
the Selaginella rhizophore. In some plants that produce subaxillary branches, such as Gosslingia breconensis, the vascular supply
of the subaxillary branch diverges from the stele of the main axis
distal to the divergence of the branch stele (Edwards, 1970)
(Fig. 5a). In lycophytes with K-branching, the vascular strand of
the rhizophore-like organ diverges from the stele of the side
Ó 2017 The Authors
New Phytologist Ó 2017 New Phytologist Trust
branch (Matsunaga & Tomescu, 2016) (Fig. 5b). Finally, in
other lycophytes with subaxillary branches, such as Deheubarthia
splendens (Edwards et al., 1989), the vascular supply of the subaxillary branch seems to diverge centrally, from in between and at
the same level as the stele of the main axis and the stele of the side
branch (Fig. 5c).
The different patterns of vascular architecture at rhizophore
divergence, documented in the nine Selaginella species, cover all
three types of architecture seen in the vascular supply of Devonian rhizophore-like axes (Table 1). The pattern documented in
S. sp. 2 (stem divergence) corresponds to the architecture of
Gosslingia subaxillary branches; the pattern seen in S. wallacei
(branch divergence) fits the vascular architecture of K-branching;
the central divergence observed in the other seven Selaginella
species is comparable to the architecture of the subaxillary branch
vascular supply in Deheubarthia. This broad range of variation
documented within Selaginella, along with the taxonomic diversity of Devonian lycophytes that exhibit matching patterns, indicates that vascular architecture cannot directly provide useful
evidence relevant to discussions of homology between Devonian
rhizophore-like axes and the Selaginella rhizophore. However, the
variation seen in Selaginella indicates that a broad spectrum of
vascular patterns can be produced by the same set of developmental processes – in this case those associated with rhizophore development.
Tracheid geometry and patterns of auxin transport in the
development of the Selaginella rhizophore
PAT is the unidirectional flow of the hormone auxin and is mediated by PIN auxin efflux carriers positioned in the plasma membrane. The localization of PIN proteins at one end of the cell
polarizes and canalizes the flow of auxin away from a source of
New Phytologist (2017)
www.newphytologist.com
New
Phytologist
6 Research
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
(j)
(k)
(l)
Fig. 4 Tracheid curvature at shoot–rhizophore junctions in nine Selaginella species. Comparisons between branching points with rhizophores (a–c, g–l) and
those without rhizophores (d–f). Note the arcing of individual tracheids (arrows) from the stele of the main stem (s) and lateral branches (b) into the
rhizophore stele (r), and the absence of such tracheids in branching points without rhizophores. The thin vascular strand (l) in (e) represents a leaf trace and
not a rhizophore trace. Also note that curvature is more apparent in some species than in others, and cannot always be clearly shown (e.g. in l) owing to
tracheid density, the orientation of mounted specimens or the limitations on photographing a thick three-dimensional structure through multiple focal
planes. (a) Selaginella kraussiana; bar, 250 lm. (b) Selaginella sp. 2; bar, 50 lm. (c) Selaginella wildenovii; bar, 50 lm. (d) S. kraussiana branching point
without a rhizophore; bar, 100 lm. (e) S. sp. 2, no rhizophore; bar, 50 lm. (f) S. wildenovii, no rhizophore; bar, 50 lm. (g) Selaginella delicatula; bar,
100 lm. (h) Selaginella pallescens; bar, 50 lm. (i) Selaginella sp. 1; bar, 100 lm. (j) Selaginella involvens; bar, 50 lm. (k) Selaginella doederleinii; bar,
50 lm. (l) Selaginella wallacei; bar, 50 lm.
New Phytologist (2017)
www.newphytologist.com
Ó 2017 The Authors
New Phytologist Ó 2017 New Phytologist Trust
New
Phytologist
(a)
(b)
(c)
Fig. 5 Diagrammatic representations of vascular architecture in basal
lycophytes with rhizophore-like rooting structures: zosterophyll-grade
lycophytes – Gosslingia (a) and Deheubarthia (c) – with subaxillary
branches (angular organs) and basal lycophytes with K-branching (b).
Gosslingia exhibits stem divergence of the vascular supply to the
rhizophore-like axis, K-branching involves branch divergence and
Deheubarthia exhibits central divergence. If these rooting structures
exhibit acropetal polar auxin transport resulting from redirection of auxin
from the shoot system, then curved tracheids should be found in the
regions indicated by the blue boxes.
higher concentration, leading to elongation of those cells that
transport auxin and their differentiation as vascular tissues (Leyser & Day, 2003; Bennett et al., 2014). Polar gradients of auxin
within the plant body have been demonstrated, through numerous auxin application and inhibition experiments, as sufficient
for inducing vascular tissue differentiation in plants (e.g.
Wangermann, 1967; Sachs, 1969; Mattsson et al., 1999; Berleth
et al., 2000). Moreover, the auxin flux through tissues is responsible for the shape and orientation of xylem tracheary elements, as
documented in both natural and experimental conditions (Sachs,
1981; Sachs & Cohen, 1982; Lev-Yadun & Aloni, 1990). Specifically, developing tracheids elongate along a polar auxin gradient.
The direct relationship between PAT and vascular architecture
indicates that vascular tissues provide an anatomical record of
physiological processes, i.e. polar auxin transport during
development.
Ó 2017 The Authors
New Phytologist Ó 2017 New Phytologist Trust
Research 7
Patterns of PAT have been documented for both shoots and
rhizophores of Selaginella. In shoots, PAT is basipetal, consistent
with patterns of PAT documented in shoots of seed plants
(Wochok & Sussex, 1973; Sanders & Langdale, 2013). By contrast, PAT is acropetal in the rhizophore, as in the roots of seed
plants (Wochok & Sussex, 1974). These patterns of PAT are
consistent with the orientation of tracheids at rhizophore junctions, in which tracheids arc from the steles of the main stem and
side branch into the rhizophore vascular supply (Fig. 4).
Together, these suggest the following developmental scenario: at
the rhizophore–shoot junction, auxin is redirected from its
basipetal flow in the main stem and lateral branch to flow
acropetally into the rhizophore, and these changes in auxin flow
are recorded in the geometry of xylem tracheids (Fig. 6). This
pattern is seen in all species sampled, regardless of whether the
rhizophore develops from a dorsal or ventral angle meristem.
Although the specific polarity of auxin flow between rhizophores
and shoots (e.g. from shoot to rhizophore vs rhizophore to shoot)
cannot be directly inferred based on tracheid curvature alone, our
interpretation is the only one that is consistent with the patterns
of auxin transport documented in Selaginella.
There is a significant amount of experimental evidence suggesting that angle meristem specification and rhizophore development are at least partially under the control of, and can be
manipulated by, auxins (Williams, 1937; Webster, 1969;
Wochok & Sussex, 1973, 1974, 1976; Jernstedt et al., 1994; Sanders & Langdale, 2013) – higher concentrations of auxin are
associated with development of the angle meristems into rhizophores and maintenance of rhizophore identity, while lower
auxin concentrations or auxin inhibition are correlated with
development of shoots. We suspect that the developmental identity of angle meristems (shoot vs rhizophore) is specified, at least
in part, by the local distribution of auxin concentrations around
branching points. At least one study has shown differences in
auxin concentrations on dorsal vs ventral sides of Selaginella
shoots at branching points (Wochok & Sussex, 1973), where the
pattern of converging meristeles may generate an asymmetric distribution of auxin. It will be interesting to test, in the future, (1)
whether dorsi-ventral variation in auxin concentrations at shoot
branching points is correlated with dorsal and ventral rhizophore
position, and (2) if auxin maxima are formed at rhizophore
apices, like they are in roots (e.g. Petrasek & Friml, 2009; Robert
& Friml, 2009; Hayashi et al., 2014). Answers to these questions
would provide important insights into rhizophore development
and, more broadly, into developmental pathways common to all
rooting structures, regardless of homology.
An anatomical fingerprint for testing hypotheses on the
evolution of the lycophyte body plan
The arced tracheid geometry we documented in Selaginella at the
shoot–rhizophore junction is consistent with vascular patterns
associated with changes in the polarity of auxin transport in flowering plants (Sachs, 1981). More broadly, this provides another
example of tracheid geometry serving as an anatomical fingerprint
(e.g. Rothwell et al., 2014) for patterns of auxin flow during organ
New Phytologist (2017)
www.newphytologist.com
New
Phytologist
8 Research
Fig. 6 Patterns of polar auxin transport at the rhizophore–shoot junction.
Auxin is redirected from its basipetal flow in the main stem and lateral
branch to flow acropetally into the rhizophore. This redirection is reflected
in the geometry of tracheids at the base of the rhizophore: these tracheids
arch from the steles of the main stem and lateral branch into the
rhizophore stele. Red arrows indicate the flow of auxin and solid blue lines
represent differentiated vascular tissues.
exhibit positive gravitropism. Based on these we hypothesize that
acropetal PAT is independent of organ identity and is correlated
with positive gravitropic responses in organs with roles in absorption and anchoring. An interesting test of this hypothesis would
be to investigate the polarity of auxin flow in the positively gravitropic rhizomes seen in the monocot Cordyline, for which developmental plasticity of axillary buds borne on horizontal rhizomes
is at least in part controlled by auxin concentrations (Fisher,
1972), as well as in tubers of the lycophyte Phylloglossum – downward growing shoots modified for starch storage and perennation
(Bower, 1885; Eames, 1936).
Because auxin is an important regulator of plant development,
anatomical signatures of PAT in vascular tissues provide important means of studying developmental mechanisms in the fossil
record. If acropetal PAT is associated with positively gravitropic
organs independent of organ identity, then we can hypothesize
that the rhizophore-like organs of early lycophytes – subaxillary
branches with rooting function and rooting axes derived from Kbranches – must have had acropetal PAT regardless of their developmental homology as stems or above-ground axes. The tracheid
geometry of the shoot–rhizophore junction in Selaginella provides an excellent anatomical fingerprint that can be used to test
this hypothesis (Fig. 5). Xylem tracheary elements have among
the highest fossil preservation potential of all plant tissues and are
preserved even in the oldest known vascular plant fossils
(Edwards et al., 1992). Thus, arcing tracheids such as those seen
at the base of the Selaginella rhizophore could, and should, be
searched for in the vascular tissues of anatomically preserved fossils of basal lycophytes with rhizophore-like organs (Fig. 5). The
anatomy of subaxillary branch vascularization in Gosslingia
breconensis, as documented by Edwards (1970), is encouraging in
this respect and re-examination of those specimens could be a
first step in the hypothesis-testing process. Should such tracheids
be discovered, they would provide evidence for shared developmental mechanisms and patterns of PAT.
Conclusions
development. A well-documented example of this uses the ‘auxin
swirls’ seen in the wood of several plant lineages that evolved secondary growth independently (lignophytes (progymnosperms and
seed plants), lycophytes and sphenophytes; Lev-Yadun & Aloni,
1990; Rothwell & Lev-Yadun, 2005; Rothwell et al., 2008;
Decombeix et al., 2010) to infer a shared underlying mechanism
of polar auxin transport through the cambial layer (Rothwell &
Lev-Yadun, 2005; Sanders et al., 2011; Rothwell et al., 2014).
Sanders et al. (2011) used the same anatomical fingerprint,
auxin swirls, to demonstrate that rhizomorphs, the rooting organs
of extinct lepidodendralean lycophytes, had acropetal PAT. Aside
from lepidodendralean rhizomorphs, acropetal PAT has only
been demonstrated in the Selaginella rhizophore (Wochok & Sussex, 1974) and is well known in the roots of seed plants. Interestingly, whereas these three types of organs are not homologous
(the rhizomorph is best explained as a shoot-homolog modified
for rooting – Rothwell & Erwin (1985) – and the homology of
the rhizophore is unresolved), they all serve rooting functions and
New Phytologist (2017)
www.newphytologist.com
The rhizophore of Selaginella is a unique organ that has putative
counterparts in fossil basal lycophytes of the Devonian period. A
survey of the anatomy of rhizophore vascularization in nine
Selaginella species revealed that the rhizophore stele can diverge in
three positions in the branching angle of a shoot, depending on
the species: centrally, from the angle formed by the steles of the
main stem and lateral branch; from the stele of the main stem, distal to the divergence of the branch stele; or from the stele of the lateral branch. This broad range of anatomical variation within the
genus precludes the use of the architecture of rhizophore vascularization as a feature relevant to discussions of the homology of rhizophore-like axes identified in Devonian basal lycophytes, but
nevertheless indicates that the same set of developmental processes
can produce a wide range of vascular architectures.
Concurrently, this investigation revealed evidence for an
anatomical fingerprint of physiological processes involved in the
development of the rhizophore. This fingerprint consists of arced
tracheids that connect the stele of the rhizophore to the steles of
Ó 2017 The Authors
New Phytologist Ó 2017 New Phytologist Trust
New
Phytologist
the main stem and lateral branch. The geometry of these tracheids reflects redirection of the basipetal PAT of the stem and
branch into the acropetal auxin stream of the rhizophore. This
anatomical fingerprint has potential for applications in the study
of Devonian lycophytes with rhizophore-like organs, in which it
could be used to infer patterns of PAT. Documenting patterns of
PAT is relevant to characterizing developmental patterns and
mechanisms in these fossil plants. In turn such information is
crucial for addressing hypotheses of homology over broad taxonomic scales and the evolution of development in deep time.
Applying knowledge of plant development derived from studies of extant plants to the study of fossils is a necessary step in
advancing our understanding of the evolution of plant development. In studies of animals, integration of morphological and
anatomical data from the fossil record with developmental and
genetic data from extant organisms has played an important role
in the assembly of comprehensive evo-devo perspectives that span
phylogeny and geologic time, and wherein features of the fossils
are used to understand the evolution of developmental regulators
and vice versa (e.g. Shubin, 2008; Erwin & Valentine, 2013).
However, similar progress in plant studies lags behind. As we
advance our understanding of how plant physiology and developmental processes are reflected in anatomy, investigating these
aspects of plant biology in the deep fossil record will become
more feasible. This will provide an important means for harnessing the rich store of data inherent in the fossil record, and enable
us to both generate and test hypotheses on development and
homology in the evolution of plant body plans.
Acknowledgements
We thank Austin Browder for producing the three-dimensional
renderings of vascular architecture (Fig. 3 and Videos S1–S4) in
Amira, the handling editor Dr Elena Kramer, and three anonymous referees for helpful comments and suggestions.
Author contributions
K.K.S.M., N.P.C. and A.M.F.T. planned and designed the
research. N.P.C. and A.M.F.T. collected and analyzed the data.
K.K.S.M. and A.M.F.T. analyzed the data and wrote the
manuscript. K.K.S.M. rendered figures and illustrations.
References
Banks HP, Davis MR. 1969. Crenaticaulis, a new genus of Devonian plants allied
to Zosterophyllum, and its bearing on the classification of early land plants.
American Journal of Botany 56: 436–449.
Bennett T, Hines G, Leyser O. 2014. Canalization: what the flux? Trends in
Genetics 30: 41–48.
Berleth T, Mattsson J, Hardtke CS. 2000. Vascular continuity and auxin signals.
Trends in Plant Science 5: 387–393.
Bower FO. 1885. On the development and morphology of Phylloglossum
drummondii. Philosophical Transactions of the Royal Society of London 176:
665–678.
Decombeix A-L, Taylor EL, Taylor TN. 2010. Epicormic shoots in a
Permian gymnosperm from Antarctica. International Journal of Plant
Sciences 171: 772–782.
Ó 2017 The Authors
New Phytologist Ó 2017 New Phytologist Trust
Research 9
Eames AJ. 1936. Morphology of vascular plants. New York, NY, USA: McGrawHill.
Edwards D. 1970. Further observations on the Lower Devonian plant, Gosslingia
breconensis Heard. Philosophical Transactions of the Royal Society of London B:
Biological Sciences 258: 225–243.
Edwards D. 1994. Toward an understanding of pattern and process in the growth
of early vascular plants. In: Ingram DS, Hudson A, eds. Shape and form in
plants and fungi. London, UK: Academic Press, 39–59.
Edwards D, Davies KL, Axe L. 1992. A vascular conducting strand in the early
land plant Cooksonia. Nature 357: 683–685.
Edwards D, Kenrick P. 1986. A new zosterophyll from the Lower Devonian of
Wales. Botanical Journal of the Linnean Society 92: 269–283.
Edwards D, Kenrick P, Carluccio LM. 1989. A reconsideration of cf. Psilophyton
princeps (Croft & Lang, 1942), a zosterophyll widespread in the Lower Old
Red Sandstone of South Wales. Botanical Journal of the Linnean Society 100:
293–318.
Erwin DH, Valentine JW. 2013. The Cambrian Explosion: the construction of
animal biodiversity. Englewood, CO, USA: Roberts & Co.
Fisher JB. 1972. Control of shoot–rhizome dimorphism in the woody
monocotyledon, Cordyline (Agavaceae). American Journal of Botany 59: 1000–
1010.
Gensel PG, Kotyk ME, Basinger JF. 2001. Morphology of above- and
belowground structures in Early Devonian (Pragian–Emsian) plants. In: Gensel
PG, Edwards D, eds. Plants invade the land. New York, NY, USA: Columbia
University Press, 83–102.
Hao S, Xue J, Guo D, Wang D. 2010. Earliest rooting system and root: shoot
ratio from a new Zosterophyllum plant. New Phytologist 185: 217–225.
Harvey-Gibson RJ. 1902. Contributions toward a knowledge of the anatomy of
the genus Selaginella. Annals of Botany 16: 449–466.
Hass H, Remy W. 1986. Angular-organe – Ausdruck einer wenig beachteten
morphogenetischen Strategie devonischer Pflanzen. Argumenta Palaeobotanica
7: 155–177.
Hayashi K, Nakamura S, Fukunaga S, Nishimura T, Jenness MK, Murphy AS,
Motose H, Nozaki H, Furutani M, Aoyama T. 2014. Auxin transport sites are
visualized in planta using fluorescent auxin analogs. Proceedings of the National
Academy of Sciences, USA 111: 1–6.
Hetherington AJ, Berry CM, Dolan L. 2016. Networks of highly branched
stigmarian rootlets developed on the first giant trees. Proceedings of the National
Academy of Sciences, USA 113: 6695–6700.
Hetherington AJ, Dolan L. 2016. The evolution of lycopsid rooting structures:
conservatism and disparity. New Phytologist. doi: 10.1111/nph.14324.
Høeg OA. 1942. The Downtonian and Devonian flora of Spitsbergen. Norges
Svalbard- og Ishavs-Undersokelser Skrifter 83: 1–228.
Jennings JR, Karrfalt EE, Rothwell GW. 1983. Structure and affinities of
Protostigmaria eggertiana. American Journal of Botany 70: 963–974.
Jernstedt JA, Cutter EG, Gifford EM, Lu P. 1992. Angle meristem origin and
development in Selaginella martensii. Annals of Botany 69: 351–363.
Jernstedt JA, Cutter EG, Lu P. 1994. Independence of organogenesis and cell
pattern in developing angle shoots of Selaginella martensii. Annals of Botany 74:
343–355.
Karrfalt EE. 1980. A further comparison of Isoetes roots and stigmarian
appendages. Canadian Journal of Botany 58: 2318–2322.
Kawai JY, Tanabe Y, Soma S, Ito M. 2010. Class 1 KNOX gene expression
supports the Selaginella rhizophore concept. Journal of Plant Biology 53: 268–
274.
Kenrick P. 1988. Studies on Lower Devonian plants from South Wales. PhD thesis,
University of Wales, Cardiff, UK.
Kenrick P, Crane PR. 1997. The origin and early diversification of land plants.
Washington, DC, USA: Smithsonian Institution Press.
Kenrick P, Edwards D. 1988. A new zosterophyll from a recently discovered
exposure of the Lower Devonian Senni Beds in Dyfed, Wales. Botanical Journal
of the Linnean Society 98: 97–115.
Korall P, Kenrick P. 2002. Phylogenetic relationships in Selaginellaceae based on
rbcL sequences. American Journal of Botany 89: 506–517.
Korall P, Kenrick P. 2004. The phylogenetic history of Selaginellaceae based on
DNA sequences from the plastid and nucleus: extreme substitution rates and
rate heterogeneity. Molecular Phylogenetics and Evolution 31: 852–864.
New Phytologist (2017)
www.newphytologist.com
New
Phytologist
10 Research
Lang WH. 1927. Contributions to the study of the Old Red Sandstone flora of
Scotland. VI. On Zosterophyllum myretonianum, Penh., and some other plantremains from the Carmyllie Beds of the Lower Old Red Sandstone. VII. On a
specimen of Pseudosporochnus from the Stromness Beds. Transactions of the
Royal Society of Edinburgh 55: 443–455.
Lev-Yadun S, Aloni R. 1990. Vascular differentiation in branch junctions of
trees: circular patterns and functional significance. Trees 4: 49–54.
Leyser O, Day S. 2003. Mechanisms in plant development. Malden, MA, USA:
Blackwell Publishing.
Matsunaga KKS, Tomescu AMF. 2016. Root evolution at the base of the
lycophyte clade: insights from an Early Devonian lycophyte. Annals of Botany
117: 585–598.
Matsunaga KKS, Tomescu AMF. 2017. An organismal concept for Sengelia
radicans gen. et sp. nov. – morphology and natural history of an Early
Devonian lycophyte. Annals of Botany. doi: 10.1093/aob/mcw277.
Mattsson J, Sung ZR, Berleth T. 1999. Responses of plant vascular systems to
auxin transport inhibition. Development 126: 2979–2991.
Petrasek J, Friml J. 2009. Auxin transport routes in plant development.
Development 136: 2675–2688.
Rayner RJ. 1983. New observations on Sawdonia ornata from Scotland.
Transactions of the Royal Society of Edinburgh: Earth Sciences 74: 79–93.
Robert HS, Friml J. 2009. Auxin and other signals on the move in plants. Nature
Chemical Biology 5: 325–332.
Rothwell GW, Erwin DM. 1985. The rhizomorph apex of Paurodendron:
implications for homologies among the rooting organs of Lycopsida. American
Journal of Botany 72: 86–98.
Rothwell GW, Lev-Yadun S. 2005. Evidence of polar auxin flow in 375 millionyear-old fossil wood. American Journal of Botany 92: 903–906.
Rothwell GW, Sanders H, Wyatt SE, Lev-Yadun S. 2008. A fossil record for
growth regulation: the role of auxin in wood evolution. Annals of the Missouri
Botanical Garden 95: 121–134.
Rothwell GW, Wyatt SE, Tomescu AMF. 2014. Plant evolution at the interface
of paleontology and developmental biology: an organism-centered paradigm.
American Journal of Botany 101: 899–913.
Ruzin SE. 1999. Plant microtechinque and microscopy. Oxford, UK: Oxford
University Press.
Sachs T. 1969. Polarity and the induction of organized vascular tissues. Annals of
Botany 33: 263–275.
Sachs T. 1981. The control of the patterned differentiation of vascular tissues.
Advances in Botanical Research 9: 151–262.
Sachs T, Cohen D. 1982. Circular vessels and the control of vascular
differentiation in plants. Differentiation 21: 22–26.
Sanders H, Rothwell GW, Wyatt SE. 2011. Parallel evolution of auxin
regulation in rooting systems. Plant Systematics and Evolution 291:
221–225.
Sanders HL, Langdale JA. 2013. Conserved transport mechanisms but distinct
auxin responses govern shoot patterning in Selaginella kraussiana. New
Phytologist 198: 419–428.
Schweitzer H-J, Giesen P. 1980. Uber Taeniophyton inopinatum, Protolycopodites
devonicus und Cladoxylon scoparium aus dem Mitteldevon von Wuppertal.
Palaeontographica Abteilung B 173: 1–25.
Shubin N. 2008. Your inner fish: a journey into the 3.5-billion-year history of the
human body. New York, NY, USA: Pantheon Books.
New Phytologist (2017)
www.newphytologist.com
Stewart WN. 1947. A comparative study of stigmarian appendages and Isoetes
roots. American Journal of Botany 34: 315–324.
Walton J. 1964. On the morphology of Zosterophyllum and some other Early
Devonain plants. Phytomorphology 14: 155–160.
Wangermann E. 1967. The effect of the leaf on differentiation of primary xylem
in the internode of Coleus blumei Benth. New Phytologist 66: 747–754.
Webster TR. 1969. An investigation of angle-meristem development in excised
stem segments of Selaginella martensii. Canadian Journal of Botany 47:
717–722.
Weststrand S, Korall P. 2016. Phylogeny of Selaginellaceae: there is value in
morphology after all!. American Journal of Botany 103: 2136–2159.
Williams S. 1937. Correlation phenomena and hormones in Selaginella. Nature
139: 966.
Wochok ZS, Sussex IM. 1973. Morphogenesis in Selaginella. Auxin transport in
the stem. Plant Physiology 51: 646–650.
Wochok ZS, Sussex IM. 1974. Morphogenesis in Selaginella. II. Auxin transport
in the root (rhizophore). Plant Physiology 53: 738–741.
Wochok ZS, Sussex IM. 1976. Redetermination of cultured root tips to leafy
shoots in Selaginella willdenovii. Plant Science Letters 6: 185–192.
Xu HH, Feng J, Jiang Q, Wang Y. 2013. Report of Drepanophycus Goppert
(Lycopsida) from the Middle Devonian of Xinjiang, China. Journal of
Systematics and Evolution 51: 765–772.
Zhou X-M, Rothfels CJ, Zhang L, He Z-R, Le Pechon T, He H, Lu NT, Knapp
R, Lorence D, He X-J et al. 2016. A large-scale phylogeny of the lycophyte
genus Selaginella (Selaginellaceae: Lycopodiopsida) based on plastid and
nuclear loci. Cladistics 32: 360–389.
Supporting Information
Additional Supporting Information may be found online in the
Supporting Information tab for this article:
Video S1 Three-dimensional reconstruction of Selaginella kraussiana.
Video S2 Three-dimensional reconstruction of Selaginella wildenovii.
Video S3 Three-dimensional reconstruction of Selaginella sp. 2.
Video S4 Three-dimensional reconstruction of Selaginella wallacei.
Please note: Wiley Blackwell are not responsible for the content
or functionality of any Supporting Information supplied by the
authors. Any queries (other than missing material) should be
directed to the New Phytologist Central Office.
Ó 2017 The Authors
New Phytologist Ó 2017 New Phytologist Trust