The Evolution of Growth Forms with Expanded Root and Shoot Parenchymatous Storage Is

Correlated across the Eudicots

Author(s): David J. Hearn, Travis Poulsen, and Rachel Spicer
Source: International Journal of Plant Sciences, Vol. 174, No. 7 (September 2013), pp. 1049-
1061
Published by: The University of Chicago Press
Stable URL: http://www.jstor.org/stable/10.1086/671745 .
Accessed: 06/09/2013 15:04

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

The University of Chicago Press is collaborating with JSTOR to digitize, preserve and extend access to
International Journal of Plant Sciences.

http://www.jstor.org

This content downloaded from 129.72.2.27 on Fri, 6 Sep 2013 15:04:41 PM

All use subject to JSTOR Terms and Conditions
Int. J. Plant Sci. 174(7):1049–1061. 2013.
䉷 2013 by The University of Chicago. All rights reserved.
1058-5893/2013/17407-0005$15.00 DOI: 10.1086/671745

THE EVOLUTION OF GROWTH FORMS WITH EXPANDED ROOT AND SHOOT

PARENCHYMATOUS STORAGE IS CORRELATED ACROSS THE EUDICOTS
David J. Hearn,1,*,† Travis Poulsen,* and Rachel Spicer‡
*Department of Biological Sciences, Towson University, Towson, Maryland 21252, USA; †J. Craig Venter Institute, Rockville, Maryland 20850,
USA; and ‡Department of Botany, Connecticut College, New London, Connecticut 06320, USA

Premise of research. Plant succulence in leaves, stems, hypocotyls, and roots is heralded as an adaptation
to aridity. Subterranean succulence is an underappreciated dimension of succulent plant diversity. Remarkable
parallel evolution of proliferation of parenchyma and vascularization of stem and root parenchyma reported
in previous case studies led us to predict that the evolution of stem succulence is tightly coupled to the evolution
of subterranean succulence more broadly across the eudicots.
Methodology. We applied a taxonomic rank-based measurement of relatedness to examine the distances
between eudicot taxa with aboveground succulence (representing 120 separate origins of succulence) and their
closest relatives with subterranean succulence. We compared these distances to the distances between a random
sample of eudicot taxa that lack aboveground succulence (mesophytes) and their closest relatives with sub-
terranean succulence.
Pivotal results. Using a statistically conservative test, the ancestral distance test, we found that taxa with
aboveground succulence were significantly more closely related to taxa with subterranean succulence than are
mesophytes. Strikingly, all groups in which extensive aboveground succulence evolved are closely related to
taxa with subterranean succulence, with the exception of the Fouquieriaceae.
Conclusions. The phylogenetic coupling between aboveground and belowground succulence is a broad
and strong pattern across a huge clade of plants, the eudicots. Despite high relatedness among taxa with
aboveground and belowground succulence, the adaptive benefits of parenchymatous tissue are complex, and
they likely differ for aboveground and belowground plant organs. The apparent lability of evolutionary change
among succulent forms may reflect a common developmental program for parenchyma production whose
spatial expression in different plant organs is altered during evolutionary diversification. These results en-
courage an expanded view of the evolutionary convergence among succulent plant forms that includes both
aboveground and belowground components.

Keywords: ancestral distance test, cacti, Euphorbia, evolutionary convergence, parenchyma, stem succulence,
switch hypothesis, tuber.

Online enhancement: appendix table.

Introduction swollen stems in plants are poorly understood. Similarly, al-

though plants with swollen stems are often described as suc-
If charismatic megafauna are the large animal species with culent, the implied function in water storage has been dem-
the most popular appeal and conservation interest, then plants onstrated in only a few widely studied groups. Recent work
with large, swollen stems such as cacti, succulent Euphorbia emphasizing the broad phylogenetic diversity and multiple evo-
L., baobabs (Adansonia L.), and other “elephant trees” (e.g., lutionary origins of plants with swollen stems (fig. 1; Nyffeler
Pachycormus Coville and Bursera Jacq. ex L.) are surely the et al. 2008; Eggli and Nyffeler 2009; Ogburn and Edwards
charismatic megaflora of the plant kingdom. Interest in suc- 2010, 2012; Arakaki et al. 2011; Gardner et al. 2012; Horn
culent plants extends back to at least the 1600s, when suc- et al. 2012) highlights repeated evolutionary convergence,
culent groups were first characterized (Bauhin and Cherler while studies emphasizing ecological strategies (Chapotin et al.
1619; Rowley 1976), and nearly 300 periodicals worldwide 2006a, 2006c; Nyffeler et al. 2008; Eggli and Nyffeler 2009;
have been dedicated to their study and appreciation (Eggli Ogburn and Edwards 2010; M. Evans, X. Aubriot, D. J.
1985, 1987). Despite this long-standing interest, the devel- Hearn, S. Lavergne, C. Cruaud, P. Lowry, and T. Haevermans,
opmental and evolutionary processes that lead to distinctly unpublished manuscript) and evolutionary-developmental or-
igins (Olson and Carlquist 2001; Olson 2003; Hearn 2006,
1
Author for correspondence; e-mail: dhearn@towson.edu. 2009a, 2013; Olson and Rosell 2006) show great variation in
Manuscript received January 2013; revised manuscript received May 2013; elec- both the form and the function of aboveground storage tissue
tronically published August 7, 2013. in stems. This convergence to stem succulence through multiple

1049

This content downloaded from 129.72.2.27 on Fri, 6 Sep 2013 15:04:41 PM

All use subject to JSTOR Terms and Conditions
1050 INTERNATIONAL JOURNAL OF PLANT SCIENCES

Fig. 1 Ordinal phylogeny of core eudicots from Stevens (2001–) illustrating the phylogenetic diversity of stem succulents. Asterisks indicate
additional lineages with stem succulent taxa that are not illustrated. Subterranean storage occurs in all indicated orders; there are also lignotubers
in Myrtales. Within the majority of the indicated lineages, relatively few taxa are succulent. A, Tylecodon paniculatus (L. f.) Toelken (Crassulaceae).
Photo by Marco Schmidt. B, Cyphostemma juttae (Dinter and Gilg) Desc. (Vitaceae). Photo by M. Schmidt. C, Adenia pechuelii Harms
(Passifloraceae). Public domain. D, Dorstenia gigas Schweinf. ex Balf. f. (Moraceae). Photo by Stan Shebs. E, Dendrosicyos socotrana Balf. f.
(Cucurbitaceae). Photo by Edson Gentile. F, Pelargonium carnosum (L.) L’Hér. (Geraniaceae). Photo by Juergen Menzel. G, Adansonia digitata
L. (Malvaceae). Photo by Ferdinand Reus. H, Moringa ovalifolia Dinter and Berger (Moringaceae). Photo by Violet Gottrop. I, Echinopsis
mamillosa Gürke (Cactaceae). Photo by S. Shebs. J, Pachypodium lealii Welw. (Apocynaceae). Photo by Hans Hillewaert. K, Plectranthus ernstii
Codd (Lamiaceae). No copyright information. L, Ipomoea arborescens var. pachylutea Gentry (Convolvulaceae). Photo by Ruddy Benezet. M,
Brighamia insignis A. Gray (Campanulaceae). Photo by Daderot.

evolutionary and developmental paths provides a natural lab- 2009; Ogburn and Edwards 2010, 2012). Cacti and succulent
oratory to investigate the evolution of syndromes of traits un- Euphorbia are the most widely studied examples and may be
derlying plant adaptation and evolutionary diversification (Og- considered representative of a large number of succulent taxa
burn and Edwards 2010). that store water in close proximity to photosynthetic tissues
Although definitions of the term “succulent” vary, it is typ- in leaves, hypocotyls, or stems. Indeed, some authors restrict
ically used to describe plants or plant organs with a large the definition of succulent taxa to those in which storage tissues
volume of parenchyma and correspondingly high tissue water are adjacent to photosynthetic tissues (Ihlenfeldt 1985), and it
content, representing adaptation to aridity (Eggli and Nyffeler is in these taxa that a functional role in water storage has been

This content downloaded from 129.72.2.27 on Fri, 6 Sep 2013 15:04:41 PM

All use subject to JSTOR Terms and Conditions
 HEARN ET AL.—EVOLUTION OF GROWTH FORMS CORRELATED ACROSS EUDICOTS 1051

most convincingly demonstrated (Mauseth 1995, 2000; Nobel sue (Mauseth and Stone-Palmquist 2001; Stone-Palmquist and
2006). The fact that these well-studied succulents are adapted Mauseth 2002).
to arid environments—often using the crassulacean acid me- Even greater volumes of storage tissue are formed when
tabolism (CAM) photosynthetic pathway, in which a four-car- additional cell divisions occur within the stem and root pa-
bon acid is used for internal storage and release of CO2—has renchyma itself (Jones and Price 1996; Mauseth 2004; Hearn
led to the view that the defining feature of succulence is the 2009b). In parenchymatous woods such as those of Adenia
functional property of water storage or water content. (Janse van Vuuren 1970; Hearn 2009a, 2009b) and even in
Other plants variably described as succulent include the pa- the storage tissue in turnips (Namikawa 1932), parenchyma
chycauls (woody plants with greatly enlarged, trunk-like resulting from proliferation at these novel sites of meristematic
stems) and geophytes (plants with various forms of under- activity can account for a large proportion of the storage tissue.
ground perennating storage organs). Functional studies in In some cases, supplemental vascular strands differentiate in
these groups are rare, and it is simply their gross morphology or permeate storage tissue of both stems and roots (Mauseth
and prevalence in arid environments that has led to a succulent and Sajeva 1992; Mauseth 1993a, 2006b; Carlquist 2001;
designation. Rather than a demonstrated ability to store water, Hearn 2009b; Noh et al. 2010), resulting in additional pro-
what is shared among these taxa, along with the more classic duction of parenchyma, xylem, and phloem away from the
succulent cacti and Euphorbia, are greatly expanded regions main vascular cambium. Indeed, it has been argued that the
of parenchymatous tissue. development of successive (i.e., concentric) vascular cambia
While aboveground storage tissues have been studied in a allows for the production of wide bands of parenchyma with
number of taxa, subterranean storage has been largely ignored. extensive vascular connections, as seen in trunks of pachycaul
Some notable exceptions include Pate and Dixon (1981), Mau- trees such as Ipomoea arborescens (Humb. and Bonpl. ex
seth and Stone-Palmquist (2001), Olson and Carlquist (2001), Willd.) G. Don and storage roots of beets. That proliferation
Hearn (2006, 2009a), Olson and Rosell (2006), Chapotin et of parenchyma and associated vascularization are key pro-
al. (2006a), and M. Evans, X. Aubriot, D. J. Hearn, S. La- cesses (Hearn 2009b; Ogburn and Edwards 2013) in the evo-
vergne, C. Cruaud, P. Lowry, and T. Haevermans (unpublished lutionary expansion of storage tissue is also suggested by the
manuscript), who highlighted the morphological diversity and wide, vascularized cortices of many cacti, which contrast with
variation in the development and location of subterranean the narrow, nonvascularized cortices of Euphorbia species
(Mauseth 2004, 2006b).
storage organs. In addition to the fact that they are hidden
Given the diversity of developmental strategies that give rise
underground, subterranean tissues are particularly difficult to
to parenchymatous storage tissue, it is interesting to ask
study because storage organs, whether developing from roots,
whether there are evolutionary patterns that connect these var-
stems, or hypocotyls, must be extracted from the soil. This
ied forms and processes. For one, it is clear from a phylogenetic
typically requires destructive sampling and limits functional
depiction of the lineages in which enlarged storage organs oc-
studies, especially if the substrate is compacted or rocky. Even
cur that greatly expanded zones of parenchyma evolved mul-
fewer studies have focused directly on the diversity of subter-
tiple times. Individual case studies (e.g., cacti [Mauseth and
ranean storage organs (Pate and Dixon 1981; Hearn 2006)
Landrum 1997; Mauseth et al. 1998; Mauseth 1999, 2006a],
and the connection between aboveground and belowground
Euphorbia [Steinmann 2001; Horn et al. 2012], Portulacineae
storage (Olson and Carlquist 2001; Stone-Palmquist and Mau-
[Nyffeler et al. 2008], Adenia [Hearn 2006], Begonia L.
seth 2002; Hearn 2006). [Thomas et al. 2011], Cyphostemma (Planch.) Alston [Hearn
Within the various organs that may be enlarged for storage et al. 2012]) suggest that, even within “succulent groups,” taxa
(leaves, roots, stems, hypocotyls, even fruits), the storage tissue with greatly expanded storage tissue evolved from ancestors
itself may occur in the epidermis, cortex, mesophyll, xylem, with little to no storage tissue multiple times and that tuberous
phloem, or pith (Mauseth 1993b, 2004; Jones and Price 1996; roots were gained or lost repeatedly. There is also evidence of
Steinmann 2001). In the majority of cases, storage tissue is separate evolutionary origins of supplemental vascular bundles
composed of masses of parenchyma cells (with the unusual occurring within storage tissue (Mauseth 2004; Hearn 2009b).
exception of wide-band tracheids in most small, globular cacti; Finally, evolutionary transitions between parenchymatous and
Mauseth and Plemons-Rodriguez 1998), but whether the di- nonparenchymatous forms occur rapidly in case studies that
visions producing these cells occur in primary (i.e., shoot and inferred absolute divergence times (!1 million years in Hearn
root apical meristems) or secondary (i.e., lateral meristems, or 2006; Arakaki et al. 2011).
cambia) meristems is highly variable. Although it is difficult Despite the morphological diversity observed in storage tis-
to generalize, storage tissue adjacent to photosynthetic tissue sues, the above observations suggest that there may be com-
often occurs in the epidermis, mesophyll, and cortex and is mon mechanisms at the core of these evolutionary-develop-
thus usually derived from a primary meristem. In contrast, the mental processes. Hearn (2006) proposed the switch
expanded regions of parenchyma in pachycaulous stems and hypothesis to account for these patterns in the occurrence of
many underground organs (e.g., root and shoot tubers) are storage tissue in the genus Adenia (fig. 2; paralleling Mauseth
often derived from the vascular cambium, a secondary meri- and Sajeva’s [1992] model for the evolution of stem cortical
stem. Here, storage parenchyma is nearly always associated vascular bundles in cacti in the subfamily Cactoideae). Evo-
with the xylem, with either ray or axial parenchyma forming lutionary switches that alternate where a conserved develop-
expanded zones, as in baobabs (Adansonia; Chapotin 2006b) mental program is expressed in a plant may account for much
and Adenia Forssk. (Hearn 2009a), whereas in succulent roots of the contrasting variation in succulent life-forms. When ex-
of cacti, greatly expanded succulent phloem is the storage tis- pressed predominantly in roots, geophytes evolve (fig. 2B, 2C).

This content downloaded from 129.72.2.27 on Fri, 6 Sep 2013 15:04:41 PM

All use subject to JSTOR Terms and Conditions
1052 INTERNATIONAL JOURNAL OF PLANT SCIENCES

Fig. 2 Examples of succulent growth habits that highlight related taxa with highly contrasting growth forms. Lines at bases of stems represent
soil level. Examples from Adenia (Passifloraceae): A, A. wightiana (Wall. ex Wight and Arn.) Engl., a vine with little secondary growth in either
roots or shoots, in which stems remain intact throughout the year; C, A. digitata (Harv.) Engl., a tuberous vine with subterranean perennation
whose vines die to ground level each year during the dry season; E, A. spinosa Burtt Davy, a caudiciform with greatly expanded basal stem and
hypocotyls but minimal subterranean succulence; F, A. venenata Forssk., a pachycaul with photosynthetic stems and reduced belowground
succulence; H, A. fruticosa Burtt Davy, a pachycaul with both photosynthetic stems and expanded storage roots. From Euphorbiaceae: B, Jatropha
berlandieri Torr., a geophyte with a large subterranean tuber and annual stems; D, Euphorbia cylindrifolia var tuberifera Raud with succulent
leaves, narrow succulent stems, and a caudex that is partially aboveground. (Euphorbia cylindrifolia var cylindrifolia lacks a large storage caudex.
There are many examples of geophytes, caudiciforms, succulent shrubs, and pachycaul trees in the genus Euphorbia.) From Pachypodium
(Apocynaceae): G, P. brevicaule Baker, an extremely squat succulent with stems quickly tapering into the root system; I, P. namaquanum (Wyley
ex Harvey) Welw., a spiny tree succulent. (Mature roots are unknown to us, although seedlings have stems that quickly taper into fibrous roots.)
Illustrations by D. J. Hearn not to scale.

When expressed predominantly in stems, stem succulents ground storage, and we assess the relatedness between plants
evolve (fig. 2D–2I), and when downregulated, more typical with parenchymatous storage in aboveground stems and hy-
mesophytes evolve (fig. 2A). More recently, Hearn (2013) pocotyls and plants with subterranean parenchymatous stor-
found that storage tissue can consist of a set of traits that age. We predict that the evolution of aboveground succulence
evolve as a module. He suggested that evolutionary modularity is correlated with the evolution of subterranean storage across
of storage tissue dissociates the evolution of storage tissues the eudicots; we test this prediction using a statistically con-
from other functions, such as mechanical support and con- servative comparative test, the ancestral distance test; and we
duction, releasing evolutionary constraint and fostering evo- discuss the implications of our results with regard to the ecol-
lutionary lability in storage tissues. Ogburn and Edwards ogy, evolution of development, and conceptualization of
(2010) argue further that since water-storage tissues have succulence.
evolved flexibly in all plant organs, there are likely to be few
developmental constraints in the evolution of highly succulent Material and Methods
tissues.
Here we seek to expand the examination of parenchymatous In outline, we (1) assembled a taxonomic database of eudicot
storage tissues, broadly construed as “succulence,” by consid- species, (2) identified the subset of eudicot species that are
ering the occurrence of both aboveground and belowground traditionally considered stem succulent plants, (3) randomly
storage organs throughout the eudicots. We suggest that the sampled taxa from the eudicot database that lack aboveground
evolution of parenchymatous storage tissue in stems, hypo- succulence (mesophytes) or that have aboveground succulence,
cotyls, and roots is tightly coupled, despite previous notions and (4) measured the relatedness between these sampled taxa
of differing functions for storage tissues in roots (starch stor- and their closest relatives with belowground succulence. Spe-
age) and shoots (water storage). We use prior surveys of stem cial care was made to sample separate evolutionary origins of
succulent taxa together with expert knowledge from 22 bot- the traits to reduce phylogenetic pseudoreplication (Felsenstein
anists to characterize presence of subterranean and above- 1985). If the evolution of stem succulence is correlated with

This content downloaded from 129.72.2.27 on Fri, 6 Sep 2013 15:04:41 PM

All use subject to JSTOR Terms and Conditions
 HEARN ET AL.—EVOLUTION OF GROWTH FORMS CORRELATED ACROSS EUDICOTS 1053

the evolution of subterranean succulence, then the taxa with Hypothesis Testing
aboveground succulence will be more closely related to the
taxa with belowground succulence than the mesophytic taxa If the evolution of aboveground succulence in eudicots is
(Hearn and Huber 2006). tied to the evolution of belowground succulence, we predicted
that the taxa with aboveground succulence would be more
Database of Eudicot Taxonomy closely related to taxa with belowground succulence than ex-
pected by chance. To test this prediction, we applied the an-
Although succulence is found in early-diverging angio-
cestral distance test of Hearn and Huber (2006; fig. 3). We
sperms, monocots, and even in gymnosperms and cycads (Og-
considered two binary traits: (1) presence or absence of sub-
burn and Edwards 2010), we restrict attention to eudicots, a
terranean succulence and (2) presence or absence of above-
large clade with more than 70% of angiosperm alpha diversity.
ground succulence. If the evolution of belowground and above-
We constructed a MySQL relational database of eudicot taxa
ground succulence is correlated, Hearn and Huber (2006)
for the purpose of generating our comparative taxonomic sam-
showed that taxa with aboveground succulence (state D1 in
ples. The list of taxa we compiled came from three sources:
fig. 3) and taxa with subterranean succulence (state D2 in fig.
Tropicos.org, with the gracious assistance of J. Soloman (MO);
3) will be more closely related to each other than will taxa
the Web query resources at the International Plant Names In-
that lack aboveground succulence (mesophytes; fig. 3A) and
dex website (http://www.ipni.org/); and the complete list of
taxa with subterranean succulence.
genera and families from Mabberley’s (1996) opus, provided
The relatedness between taxa is measured as the distance
by R. Scotland, who used these data previously for analyses
(i.e., the ancestral distance) from a sampled taxon (either with
of taxonomic rank structuring (Scotland and Sanderson 2004).
aboveground succulence or, contrastingly, a mesophyte) to a
Taxonomic rank-level classifications were based on Mabberley
most recent ancestor (MRA) with at least one descendant with
(1996) and the Angiosperm Phylogeny Group (2009) III web-
subterranean succulence. Shorter distances mean higher relat-
site (Stevens 2001–).
edness. Here, specifically, we estimated the distances from 20
randomly sampled species with aboveground succulence to
Characterization of Succulent Taxa, Both
their MRA with at least one descendant with subterranean
Aboveground and Belowground
storage (fig. 3D). We also estimated distances from 20 ran-
Even though stem, hypocotyl, and root succulence are not domly sampled species that lack stem succulence to their MRA
binary traits (Ogburn and Edwards 2010, 2012), we compiled with at least one descendant with subterranean storage (fig.
a list of taxa considered to be stem succulent (here used in- 3B). Our MySQL taxonomic database allowed for selection
terchangeably with having “aboveground storage”) based on of N random eudicot species using the left and right indexing
previous encyclopedic characterizations of stem succulent life- scheme for hierarchical data of Celko (1999) as follows: se-
forms (Jacobsen 1978; Rowley 1980, 1987; Albers and Meve lect distinct taxon_id, taxon, rank from taxon where
2002; Eggli 2002, 2003). We considered taxa to have below- right_index!p (select right_index from taxon where taxon p
ground succulence (here used interchangeably with having ‘eudicots’) and left_index1p (select left_index from taxon
“subterranean storage”) if one or more of three conditions where taxonp‘eudicots’) and rank p ‘species’ order by rand()
were met: their roots were greatly expanded into tuberous limit N.
forms, the hypocotyl was succulent and predominantly sub- Our sampling procedure, illustrated in figure 3, assured that
terranean, and/or plants had underground tuberous stems. We only separate evolutionary origins of aboveground and below-
also included lignotubers and subterranean haustoria as sub- ground succulence were sampled so as to avoid phylogenetic
terranean succulence, even though these structures can be quite pseudoreplication (Felsenstein 1985). Independent origins of
woody and lack the extensive parenchyma found in more typ- succulence were assessed using a parsimony framework (Mad-
ical tuberous stems and roots. Stem succulent taxa whose dison and Maddison 2000). In particular, after sampling a
aboveground storage tissue tapered into a nonsucculent sub- species, we excluded all taxa from further sampling that de-
terranean root system were not considered to have subterra- scended from an MRA (e.g., descendants of the ancestral nodes
nean succulence. indicated by an asterisk in fig. 3) or that descended from a
Most species descriptions, monographs, and floras we con- deeper ancestor in which subterranean or aboveground suc-
sulted failed to provide descriptions of subterranean mor- culence originated (fig. 3E).
phology. So in addition to our personal knowledge and de- Ancestral distances can be estimated using one of many mea-
scriptions in monographs and other treatments (see table A1, surements of relatedness: time, branch lengths, or number of
available online), we based our assessment of the presence of speciation events from a sampled taxon with stem succulence
subterranean storage on the knowledge of taxonomic experts. to its MRA with at least one descendant with subterranean
We contacted plant systematics experts who studied relatives succulence (fig. 3, asterisks). Additionally, in plants, Hearn and
of sampled taxa to help ascertain subterranean morphology. Huber (2006) found that ancestral distances could be conser-
These experts were identified through lists of botanists at the vatively estimated using taxonomic ranks. The rank ancestral
Index Herbariorum (Thiers 2012) and the American Society distance was recorded as 0, 1, 1.5, 2, 3, or 4 when subterranean
for Plant Taxonomy website member’s directory. Each expert storage was first found in the species, genus, tribe/subfamily,
was asked whether a sampled species had subterranean stor- family, order, or a rank level above order, respectively. We
age, what its closest relative was with subterranean storage, alternated between sampling a stem succulent and a meso-
and what additional published resources, if any, described sub- phyte, removing taxon groups for each sampled taxon, as de-
terranean morphology. scribed above, to avoid phylogenetic pseudoreplication.

This content downloaded from 129.72.2.27 on Fri, 6 Sep 2013 15:04:41 PM

All use subject to JSTOR Terms and Conditions
1054 INTERNATIONAL JOURNAL OF PLANT SCIENCES

Fig. 3 Ancestral distance methodology. A, Branch of a potentially much larger phylogeny with two traits mapped (e.g., white p underground
succulence, dashed lines p aboveground succulence). Circles represent generic groupings, and the square represents a family grouping. B, A
mesophyte (thick black branch) has been randomly selected. The asterisk indicates its closest ancestor with at least one tuberous descendant
(indicated by an exclamation point), so its rank ancestral distance is 1 at the genus level. C, Taxa that descended from the ancestor (B, asterisk)
have been removed from further sampling. D, A species with aboveground succulence (thick white branch) has been randomly selected. It is
also tuberous, so its rank ancestral distance is 0. E, Descendants representing the ancestor (D, exclamation point) that gave rise to subterranean
succulence in D are excluded from further analysis to assure that separate evolutionary origins are sampled. This procedure would continue for
a larger sample of taxa with aboveground succulence and a larger sample of mesophytes, and the ancestral distances of the aboveground succulent
taxa would be compared to the ancestral distances of the mesophytes. The null hypothesis is no difference in ancestral distances, whereas if the
evolution of stem succulence is positively correlated with the evolution of tubers, then ancestral distances of stem succulent taxa would be shorter
than those of mesophytes.

We used rank ancestral distances rather than other metrics ability to detect differences in ancestral distances between
for three reasons. First, a fully resolved phylogeny of the eu- aboveground succulent and mesophytic taxa. Despite the as-
dicots is not currently available, so estimates of ancestral dis- sistance of taxonomic experts, we were unable to assess the
tances of succulent taxa using time, branch lengths, or number presence or absence of subterranean storage for all taxa in
of speciation events is not currently feasible. Second, using a some of the sampled groups. So we also included a “conser-
rank-based metric alleviates the onerous task of assessing the vative” assessment (table A1) of ancestral distances, in which
presence of underground storage in a large set of relatives of the ancestral distance of mesophytes was decreased to accom-
a sampled species, as only one exemplar with subterranean modate uncertainty in our assessment of the presence of un-
succulence need be found, if present, within a given taxonomic derground storage. Specifically, if we were unable to survey
group. Third, hypothesis tests based on rank ancestral dis- the subterranean morphology of all taxa of a particular tax-
tances result in statistically conservative tests, so false rejection onomic rank, we set the “conservative” ancestral distance to
of a null hypothesis occurs infrequently (Hearn and Huber the rank with uncharacterized taxa, even when it was highly
2006). Using the Mabberley plant data, Hearn and Huber unlikely for the group to have members with subterranean
(2006) examined rank-based assessments of ancestral distances succulence. For example (table A1), although it is unlikely that
in simulated data sets. The null hypothesis was rarely rejected any members of the large genus Tarenna Gaertn. have sub-
when using ranks, and when the null was rejected, the evo- terranean succulence, we conservatively considered this group
lutionary correlation between traits was high. to possess belowground succulence. These conservative as-
In all cases of uncertainty, we used data that decreased our sessments decreased the length of the mesophyte ancestral dis-

This content downloaded from 129.72.2.27 on Fri, 6 Sep 2013 15:04:41 PM

All use subject to JSTOR Terms and Conditions
 HEARN ET AL.—EVOLUTION OF GROWTH FORMS CORRELATED ACROSS EUDICOTS 1055

tances, making them more similar to those of the stem suc-

culents, so our ability to detect statistical differences between
stem succulents and mesophytes was decreased. Further, taxa
with lignotubers were most closely related to mesophytic taxa
(mimosoid legumes and Myrtales) but not to stem succulent
taxa, so inclusion of lignotubers decreased ancestral distances
for mesophytic taxa. Collectively, usage of rank ancestral dis-
tances, testing with conservative ancestral distances, and in-
clusion of morphologies such as lignotubers that do not fit our
definition of parenchymatous storage organs all reduced our
ability to reject the null hypothesis. We applied the Mann-
Whitney test, as implemented by the R ‘wilcox.test’ function
(R Development Core Team 2012), to test the null hypothesis
of no difference between ancestral distances of stem succulent
taxa and mesophytes.

Results

Database of Eudicot Taxonomy

The database we compiled of eudicot taxa is composed of
293,028 species names (including synonyms). Readers can
browse the taxonomic structure using the Taxon Tracker util-
Fig. 4 Box plot of rank ancestral distances for taxa that lack
ity at Diversity Mapper (http://loco.biosci.arizona.edu/ua
aboveground succulence (mesophytes) and taxa with aboveground suc-
_collections/diversity_mapper), which is currently hosted by culence. The complete list of sampled taxa and their ancestral distances
the M. Sanderson lab. is in table A1, available online. This figure presents the best estimate
of ancestral distances (table A1, “Rank distance”). As expected, stem
Characterization of Succulent Taxa, Both succulent taxa have shorter ancestral distances than mesophytes.
Aboveground and Belowground
succulent taxa in the Fouquieriaceae (Fouquieria purpusii
Table A1 presents the taxa that we sampled, the character
Brandegee, Fouquieria fasciculata [Willd. ex Roem. and
states, their rank ancestral distances, and, where applicable,
Schult.] Nash, Fouquieria columnaris [Kellogg] Kellogg ex
the taxonomic experts who aided in character state assessment.
Curran; Henrickson 1969), but we did not identify any Fou-
quieriaceae or taxa in the closely related Polemoniaceae (Porter
Hypothesis Testing and Johnson 2010) with extensive subterranean storage. How-
We expected the rank ancestral distances of the taxa with ever, juvenile plants of the succulent Fouquieria Kunth do have
aboveground succulence to be statistically shorter than the a greatly expanded hypocotyl compared with the stem, but
rank ancestral distances of the mesophytes. Using either the this hypocotyl is not subterranean.
original ancestral distances (table A1, “Rank distance”) or the
conservative ancestral distances (table A1, “Conservative Discussion
rank”), the ancestral distances of the stem succulent taxa were
significantly shorter than the ancestral distances of the mes- The extreme morphological similarity among more classi-
ophytic taxa (fig. 4; W p 373, P ! 5e ⫺ 5; conservative W p cally defined succulent lineages (e.g., cacti, succulent euphorbs,
325, P ! 0.001); across the eudicots, stem succulent taxa are and the carrion flower stapeliads) is the textbook example of
more closely related to taxa with subterranean storage than evolutionary convergence in plants (Futuyma 1997). The work
expected by chance. In each clade that we investigated in which we describe here suggests that evolutionary convergence of
aboveground succulence evolved, there were closely related stem succulence may be part of a much broader picture of
taxa with subterranean succulence (exemplars are in table 1), eudicot evolution, one in which the production of extensive
and most clades with aboveground succulent taxa included parenchyma in aboveground as well as belowground organs
geophytic taxa with reduced storage in aboveground stems but plays a pivotal role in evolutionary transitions among disparate
highly expanded subterranean storage (including, but not lim- growth habits, such as geophytes and stem succulents.
ited to, Adenia, Begonia, Ceropegia L., Cyphostemma, Dor- Previous accounts of convergence of succulent habits (Fu-
stenia L., Euphorbia, Fockea Endl. in Endlicher and Fenzl, tuyma 1997; Niklas 1997; Stearns and Hoekstra 2005) have
Ipomoea L., Jatropha L., Oxalis L., Pelargonium L’Hér. ex focused almost entirely on aboveground stem and leaf suc-
Aiton, Portulaca L., etc.) and vice versa, with stem succulent culence associated with photosynthetic tissues, where the mor-
taxa with reduced storage underground usually found in these phological, anatomical, and ecophysiological traits (such as
groups as well. CAM photosynthesis, pleated stems, stems with stomata, de-
There was only one exception to the rule that taxa with creased internode lengths, specialized apical meristems, re-
aboveground succulence are closely related to taxa with sub- duced/caducous leaves, and epicuticular waxes) are all viewed
terranean storage: the Fouquieriaceae. There are a few highly as adaptations to arid environments. In contrast, the adaptive

This content downloaded from 129.72.2.27 on Fri, 6 Sep 2013 15:04:41 PM

All use subject to JSTOR Terms and Conditions
 Table 1
Some Major Succulent Lineages
Liana
Family/genus Tuberous representatives Stem succulent representatives representatives
Adenia Forssk. A. pyromorpha (H. Perrier) W. J. de Wilde; A. A. venenata Forssk.; A. lapiazicola Bardot- A. gummifera
racemosa W. J. de Wilde; A. digitata (Harv.) Vaucoulon (Harv.)
Engl. Harms; A.
densiflora
(Baker) Harms
Adenium Roem. A. oleifolium Stapf; A. somalense Balf. f. A. socotranum Vierh.; A. arabicum Balf. f. None
and Schult.
Aizoaceae Mestoklema tuberosum (L.) N. E. Br. M. tuberosum; Trichodiadema bulbosum None
Schwantes
Asteraceae Senecio tropaeolifolius MacOwan ex F. Muell.; Notoniopsis (Senecio) fulgens B. Nord.; Mikania Willd.
Othonna bulbosa L. Othonna cacalioides L. f. spp. (vines)
Begonia L. B. X tuberhybrida Voss B. dregei Otto and A. Dietr. None
Brassica L. B. rapa L. subsp. rapa B. oleracea var. gongylodes L. None
Burseraceae Bursera fagaroides (Kunth) Engl. (juvenile) B. microphylla A. Gray; Commiphora dinteri None
Engl.; C. gracilifrondosa Dinter ex Van Der
Walt
Cactaceae Grusonia (Opuntia) pulchella (Engelm.) H. Grusonia (Opuntia) pulchella; many None (epiphytic
Rob.; many scramblers)
Campanulaceae Cyphia P. J. Bergius spp. Brighamia A. Gray spp. None (vines)
Caricaceae Jacaratia corumbensis Kuntze Carica L. spp. None
Ceropegieae Brachystelma R. Br. spp.; Ceropegia rendallii N. Stapelia (Stapelianthus) L. spp.; Ceropegia None
E. Br.; C. woodii Schltr. stapeliformis Haw.; C. cimiciodora Oberm.
Convolvulaceae Ipomoea batatas (L.) Lam. I. arborescens (Humb. and Bonpl. ex Willd.) G. Many
Don
Crassulaceae Sedum tuberosum Coss. and Letourn. S. oxypetalum Kunth; Tylecodon Toelken spp. None (few vines)
Cucurbitaceae Marah Kellogg spp.; Ibervillea Greene spp.; Dendrosicyos socotrana Balf. f.; Momordica Fevillea L. spp.;
Kedrostis Medik. spp. rostrata A. Zimm. many
Dorstenia L. D. barnimiana Schweinf.; D. benguellensis D. gigas Schweinf. ex Balf. f.; D. gypsophila None
Welw.; D. cuspidata Hochst. ex A. Rich. Lavranos; D. foetida Schweinf.
Euphorbia L. Euphorbia primulifolia Baker; E. trichadenia E. obesa Hook. f.; E. candelabrum Tremaut ex None
Pax; E. orbiculifolia S. Carter; E. platycephala Kotschy; E. cooperi N. E. Br. ex A. Berger; E.
Pax; E. cylindrifolia Rauh and Marn.-Lap. bupleurifolia Jacq.
subsp. tuberifera Rauh
Fabaceae Erythrina flabelliformis Kearney; Dolichos L. Delonix floribunda (Baill.) Capuron Mucuna Adans.
spp.; Neorautanenia ficifolia (Benth. ex Harv.) spp.; many
C. A. Sm.
Fockea Endl. in F. angustifolia K. Schum. F. multiflora K. Schum. F. multiflora
Endlicher and
Fenzl
Fouquieriaceae None Fouquieria columnaris (Kellogg) Kellogg ex None
Curran; F. purpusii Brandegee
Geraniaceae Pelargonium schizopetalum Sweet; many P. carnosum (L.) L’Hér.; Sarcocaulon Sweet spp. None (some
geophytes scramblers)
Impatiens L. I. tuberosa H. Perrier I. mirabilis Hook. f. None (some
scramblers)
Jatropha L. J. tuberosa Elliot; J. macrophylla Pax and K. J. podagrica Hook. None
Hoffm.
Lamiaceae Plectranthus crassus N. E. Br.; P. dolomiticus P. ernstii Codd None (few vines)
Codd; Aeollanthus buchnerianus Mart. ex
Spreng.
Malvaceae Adansonia digitata L. (juvenile); Ceiba Adansonia digitata; Pseudobombax ellipticum None
aesculifolia (Kunth) Britten and Baker f. (Kunth) Dugand
Moringaceae Moringa pygmaea Verdc.; M. borziana Mattei Moringa drouhardii Jum. None
Oxalis L. O. pachyrrhiza Wedd. O. gigantea Barnéoud None
Pachypodium Lindl. Pachypodium bispinosum A. DC.; P. P. lamerei Drake; P. densiflorum Baker None
succulentum (L. f.) Sweet
Pedaliaceae Pterodiscus angustifolius Engl.; Harpagophytum Sesamothamnus lugardii N. E. Br.; Uncarina None (some
DC. ex Meisn. spp. Stapf spp. scramblers)
Portulacaceous Talinum Adans. spp.; Portulaca argentinensis Portulacaria afra Jacq.; Portulaca molokiniensis None (vines,
cohort Speg.; P. macrorhiza R. Geesink; Pterocactus Hobdy; Carnegiea gigantea (Engelm.) Britton scramblers)
tuberosus (Pfeiff.) Britton and Rose and Rose
Pyrenacantha Hook. P. kaurabassana Baill. P. malvifolia Engl. Pyrenacantha
spp.

This content downloaded from 129.72.2.27 on Fri, 6 Sep 2013 15:04:41 PM

All use subject to JSTOR Terms and Conditions
 HEARN ET AL.—EVOLUTION OF GROWTH FORMS CORRELATED ACROSS EUDICOTS 1057

Table 1
(Continued )
Liana
Family/genus Tuberous representatives Stem succulent representatives representatives
Vitaceae Cyphostemma congestum (Baker) Desc. ex Wild Cyphostemma juttae (Dinter and Gilg) Desc.; Multiple
and R. B. Drumm.; Cissus trifoliata (L.) L. Cyphostemma currorii (Hook. f.) Desc.;
Cissus tiliacea Kunth; Cissus quadrangularis
L.
Note. Tuberous, stem succulent, and lianous representatives are included where applicable. All stem succulent lineages investigated, with
the exception of Fouquieriaceae, have tuberous representatives. Many succulent lineages are derived from lianous ancestors, but this pattern
is not universal.

significance of expanded regions of parenchyma in large studies found a strong association between increased aridity
woody stems that lack photosynthesis (e.g., in some pachy- and the evolution of subterranean succulence. Parenchyma tis-
cauls) and subterranean organs (e.g., in geophytes) is far less sue makes up the bulk of root and stem cortex, pith, and leaf
clear. Despite this, we show clear evidence that across the eu- mesophyll, and its function in plants is highly variable. Pa-
dicots, the occurrence of expanded storage tissue in above- renchyma containing chloroplasts functions in photosynthesis,
ground stems is associated with the occurrence of subterranean whereas nonchlorophyllous parenchyma functions in storage,
storage tissue among close relatives far more frequently than secretion, transport, and wound response (Evert 2006; Pruyn
would be expected by chance. This pattern holds not only for and Spicer 2012). Parenchyma cells in the cortex of strictly
the pachycauls but also for all stem succulent groups, including defined succulents are specialized to store water (and, in the
the classic stem succulents in Euphorbia, Cactaceae, and Apo- case of many alternative photosynthetic pathways, to store
cynaceae (table 1). The one exception was in the succulent four-carbon acids) in a greatly expanded central vacuole (Mau-
Fouquieriaceae, which does not appear to have any close rel- seth 1995, 2006b; Nobel 2006), whereas parenchyma cells in
atives with subterranean storage. tubers are typically believed to function in starch storage. High
What processes resulted in the high relatedness between starch concentrations can also be found in parenchyma of the
aboveground succulent taxa and belowground succulent taxa cortex, the xylem, the phloem, and, to a lesser extent, the pith
within the eudicots? The relative roles of adaptive processes of stems (Mauseth and Sajeva 1992; Hearn 2009a).
in concert with evolutionary-developmental mechanisms are A large percentage of swollen-stemmed plants included in
unclear. Two core hypotheses are discussed here, broken into this study are caudiciforms or pachycauls, which have either
ecological and evolutionary-developmental dimensions of suc- an expanded basal stem or an enlarged, monopodial trunk,
culence evolution. respectively. Often described as stem succulents, these plants
have expanded zones of axial and/or ray parenchyma within
Ecological and Functional Dimensions of Succulence the secondary xylem (i.e., xylem produced by the vascular
cambium, synonymous with “wood”). To ascribe a solely wa-
Could similar selection pressures foster the evolution of ex-
panded storage tissue in aboveground and belowground or- ter-storage function to these stems is likely an oversimplifi-
gans of close relatives that ancestrally occurred in arid habi- cation, as this putative role is rarely tested and the physiolog-
tats? If yes, the phylogenetic clustering of stem and root ical functions of xylem parenchyma are complex and poorly
succulence in specific clades could be attributed to ancestral understood. Work in baobabs (Adansonia) suggests that the
adaptation to aridity and subsequent adaptive specialization expanded storage tissue in stems and roots does not buffer
within a clade. However, an affirmative answer to this question diurnal use of water for photosynthesis but instead supports
is unlikely. Although aridity is often evoked to explain the rapid leaf growth prior to the onset of seasonal rains (Chapotin
evolution of succulence, most studies concern cacti (which rep- et al. 2006a, 2006c). Our own field observations suggest that
resent a small percentage of the total number of evolutionary storage tissues in tuberous and stem succulent Adenia and
origins of stem succulence), and cacti are most frequent in areas Cyphostemma may serve a similar function, as these plants
with seasonal and reliable precipitation without prolonged produce growth from fleshy rootstocks and flower immediately
freezing, not necessarily areas with extreme aridity (Nobel prior to commencement of the rainy season. Although leaf
1980; Gibson and Nobel 1986; Burgess and Shmida 1988; photosynthesis is typically cited as the main driver of photo-
Pierson and Turner 1998; Godı́nez-Álvarez et al. 2003). Bur- synthesis in pachycauls (Burgess and Shmida 1988), stem pho-
gess and Shmida (1988) highlighted that water storage as an tosynthesis is likely to be a major contributor to photosynthetic
adaptation to aridity is likely to be only part of the story. When output in at least a few species of pachycauls, such as Adenia,
the relationship between climate and the occurrence of above- for which leaves are rapidly caducous and/or small and stems
and belowground storage was examined explicitly in Malagasy retain a photosynthetic cortex with epidermal stomata (e.g.,
Euphorbia (M. Evans, X. Aubriot, D. J. Hearn, S. Lavergne, Adenia fasciculata, Adenia venenata) throughout life (Hearn
C. Cruaud, P. Lowry, and T. Haevermans, unpublished man- 2009a).
uscript) and Adenia (Hearn 2004, 2013), no statistically sig- Ray parenchyma in woody stems serves as a route of ex-
nificant links were found between increased aridity and the change between xylem and phloem (van Bel 1990), and there
evolution of aboveground stem storage. In contrast, these same is likely significant overlap between functions in carbohydrate

This content downloaded from 129.72.2.27 on Fri, 6 Sep 2013 15:04:41 PM

All use subject to JSTOR Terms and Conditions
1058 INTERNATIONAL JOURNAL OF PLANT SCIENCES

metabolism and water balance. Seasonal changes in starch lev- bility can facilitate adaptive response to divergent selection
els of parenchyma in secondary xylem suggest that these cells pressures for storage tissue under different ecological contexts
function in carbohydrate storage in both temperate and and functional constraints.
(sub)tropical species (Bullock 1992; Höll 2000; Barbaroux and
Breda 2002; Landhäusser and Leiffers 2003; Chapotin et al. Evolutionary-Developmental Dimensions of Succulence
2006b), where the timing of reserve depletion is often linked
to leaf phenology. However, diurnal changes in parenchyma Olson (2003) was among the first to link the evolution of
starch content suggest that carbohydrate reserves may play a succulence development among seemingly disparate growth
role in refilling air-filled vessels (e.g., embolism repair; Bucci habits: lianas and pachycauls. He suggested that many pachy-
et al. 2003; Salleo et al. 2004, 2009; Secchi and Zwieniecki cauls (in, e.g., Dendrosicyos socotrana, Cyphostemma, Ad-
2011), and in some species, seasonal fluctuations appear neg- enia, and Ipomoea) originated from a lianous ancestor, with
ligible (Bullock 1992; Mooney et al. 1992). The fact that car- the now self-supporting pachycaul descendants having lost ten-
bohydrates may signal water stress, are able to function as an drils or other mechanisms of climbing. Olson suggested that
osmoticum to direct water movement, and may be released the large volume of xylem parenchyma in lianoid stems—hy-
into the xylem transpiration stream suggests that water and pothesized to protect against damaging stem torsion in many
carbohydrate storage functions are interconnected. Moreover, lianas—is a preadaptation for increased stem storage tissue,
subterranean storage organs have been shown to supply car- such that parenchyma was co-opted to function in water stor-
bohydrates and, in some cases, proteins, mineral nutrients, and age in descendant pachycauls. Pachycauls in all of these groups
larger oligosaccharides to aboveground tissues (Pate and are also either tuberous themselves or closely related to tu-
Dixon 1981), but to our knowledge direct measurements of berous taxa. The idea that a storage function could be mod-
water storage and links to aboveground utilization are lacking, ularized and co-opted was expanded by Hearn (2006, 2013),
and their role in photosynthesis is unknown or absent. Thus, who suggested that evolutionary-developmental switches
storage organs cannot be viewed as either water-storage or where succulence develops can account for succulent life-form
starch-storage organs, as these functions appear to be cou- diversity in Adenia. The taxonomically broader pattern of as-
pled—rather, storage tissues serve the dual functions of car- sociation between stem and subterranean succulence that we
bohydrate and water usage to varying degrees. The close re- report can also be parsimoniously explained by this switch
latedness that we detected between stem succulent taxa and hypothesis.
storage-rooted taxa suggests, minimally, that evolutionary If a switch mechanism is involved in the patterning of suc-
transition between starch- and water-storage function is labile. culence evolution, then shared genetic-developmental path-
There are fundamental differences between the aboveground ways are expected in regions of both primary and secondary
and belowground environment that could contribute to different growth that produce succulent tissues. In an emerging evolu-
selection pressures for aboveground or belowground storage. tionary synthesis, Carroll (2008) and others (Prud’homme et
Storing materials aboveground has costs: aboveground storage al. 2007) argued that most morphological evolution is likely
organs can suffer from exposure to fire (Thomas 1991; Thomas to be the result of evolutionary shifts in the timing and posi-
and Goodson 1992), and they may experience greater temper- tioning of activities of core conserved genes. As our under-
ature extremes. The protection from weather exposure and fire standing of molecular genetic modules and networks improves,
afforded by underground organs may effectively select for sub- we are likely to see an increasing number of traits in which a
terranean storage, particularly in arid regions where water is at modest shift in the position or timing of expression leads to
a premium and growth (i.e., carbon assimilation) tends to be a dramatic change in anatomy or morphology. For example,
slow. On the other hand, storage organs with high water and/ the distinction between herbaceous and woody growth forms
or carbohydrate contents are susceptible to rot, especially in cool can be quite dramatic, but it has become increasingly clear
or, alternately, heat-stressed environments, so switching from that the secondary growth characteristic of woody plants
subterranean to aboveground storage may minimize this danger shows overlapping gene expression with the primary growth
by reducing the duration of exposure to moisture (Hearn 2004). characteristic of herbaceous plants (Groover 2005; Du and
This may also explain the association between increased aridity Groover 2010). Notably, the KANADI and Class III homeo-
and the evolution of tubers, as increased aridity may reduce the domain-leucine zipper (HD-ZIP) genes that are responsible for
moisture content of the substrate and thereby reduce the risk of establishing the polarity of organs originating at the shoot
rot. apical meristem (SAM) are also expressed in the vascular cam-
So, extensive parenchyma can perform many different over- bium (Schrader 2004). Recently, it was shown that overex-
lapping functions (water storage, carbohydrate storage, wound pressing a microRNA-resistant copy of the Class III HD-ZIP
response, salt tolerance; we did not consider halophytes in our gene popREVOLUTA led to an additional cambium forming
study, and succulence deriving from salt osmoregulation adds among parenchyma in the cortex (Robischon et al. 2011) and
an additional layer of complexity). The functional roles of that this cambium had reversed polarity such that xylem was
expanded parenchyma in both woody stems and subterranean produced toward the outside of the stem, with phloem to the
organs are complex and varied, and no one adaptive regime inside; thus, upregulation of a key gene can result in large-
is likely to account for their origins. The fact that we find scale changes to cambial patterning. KNOX genes involved in
expanded parenchyma repeatedly appearing in related taxa, specifying the stem cell niche in the SAM are also expressed
often in different organs, suggests not a common selective pres- in the cambial zone, including ARK1 (Groover et al. 2006)
sure but instead a flexible evolutionary-genetic mechanism that and ARK2 (Du et al. 2009), both of which affect the extent
switches positioning of parenchyma development. This flexi- of secondary growth when expression levels are perturbed. In

This content downloaded from 129.72.2.27 on Fri, 6 Sep 2013 15:04:41 PM

All use subject to JSTOR Terms and Conditions
 HEARN ET AL.—EVOLUTION OF GROWTH FORMS CORRELATED ACROSS EUDICOTS 1059

fact, the vascular cambium and concentric rings of secondary Conclusions

xylem and phloem that it produces are continuous with the
separate primary vascular bundles above, such that a true con- We have detected a previously undocumented but wide-
tinuum is formed, underscoring that the distinction between spread evolutionary pattern across the eudicots: the evolution
these two conditions is blurry at best (Spicer and Groover of stem succulence is correlated with the evolution of subter-
2010). Thus, even though the sites of development of paren- ranean succulence. The flexible evolutionary transition in the
chyma may vary from one evolutionary instance to another, positioning and amount of parenchyma development under-
the above molecular evidence suggests, more generally, that pins evolutionary changes among multiple growth forms. Thus
similar genes with differing levels of expression participate in far, the underground dimensions of this story in evolutionary
meristem function at different sites of development. convergence have been almost completely ignored by tradi-
To our knowledge, no studies have examined the molecular- tional discussions of succulence convergence (Futuyma 1997;
genetic dimensions of succulence evolution and diversification. Niklas 1997; Stearns and Hoekstra 2005), yet processes oc-
The switch hypothesis offers many testable predictions. In par- curring belowground are central to a more thorough under-
ticular, if the same genes are up- or downregulated when suc- standing of this quintessential example of evolutionary con-
culence evolved, one would expect similar patterns of differ- vergence. Adaptive benefits of parenchymatous tissue are
ential expression in the cambia of succulent taxa when complex and likely differ for above- and belowground plant
compared to closely related mesophytes. Our results further organs, despite the high relatedness among taxa with these
predict that evolutionary distance in gene space may not be features. Further, this apparent lability of evolutionary change
very large between disparate growth forms such as geophytes among succulent forms very likely reflects shared develop-
and pachycauls. These results provide exciting avenues of mo- mental pathways for parenchyma production that are switched
lecular-genetic research, which hint at deep evolutionary- between stem and root expression during evolutionary diver-
developmental linkages among succulent forms. sification. We hope these data and results will encourage re-
searchers of plant habit to reconsider the story of evolutionary
Last, case studies suggest that root succulence may precede
convergence in succulent plants by recognizing connections
the appearance of stem succulence. In available case studies,
between subterranean and aboveground dimensions of eudicot
root succulence evolved first in ancestors of many stem suc-
plant form. We highlighted gaps in our knowledge and indicate
culent taxa, and then subsequent temporal and spatial changes
several potential avenues of future research.
to succulence development resulted in the diversity of stem and
root succulent plant forms within these lineages (fig. 2). Ex-
amples of succulent groups that are likely to be ancestrally Acknowledgments
tuberous include Pereskia Mill. as a “basal” cactus (Edwards
and Donoghue 2006), Moringa Adans. (Olson and Carlquist We would like to thank Q. Cronk and A. Groover for or-
2001; Olson and Rosell 2006), Adenia (Hearn 2006, 2009a), ganizing the woody plants working group and the NSF NESCent
Fockea (Bruyns and Klak 2006), and Cyphostemma (Hearn for generously hosting our meeting. This work would not have
et al. 2012). The ancestral distance methodology used here been possible without the expertise of 22 botanists, including
cannot assess the evolutionary direction of transitions among L. A. Alice, F. Almeda, D. M. Bates, R. J. Bayer, C. Cupido, A.
stem, root, and hypocotyl storage, but if this pattern proves Freire-Fierro, R. L. Hartman, W. C. Holmes, K. Hummer, S.
to be more general than these few examples, then it suggests Kelso, J. Kirkbride, R. Levin, A. Martins, L. McDade, A. N.
that there are fewer constraints on subterranean succulence Meullner, J. Morawetz, J. F. B. Pastore, R. Roberts, J. Santos
evolution (i.e., that it is “easier” to evolve) and that further Silva, V. Steinmann, R. D. Stone, and K. Wurdack. Their interest
specialization of function and form is required to achieve stem and contributions are greatly appreciated. R. Scotland and J. C.
succulence and its correlates (e.g., specialized photosynthetic Solomon graciously provided extensive taxonomic data sets, and
pathways, additional mechanisms to reduce water loss). To M. Sanderson is kindly hosting the taxonomic database. Towson
further assess this hypothesis, more resolved phylogenies are University FDRC provided faculty development funds that made
required for multiple stem succulent groups, and, more im- this work feasible. Thank you to A. Estes, J. Mauseth, L.
portantly, the subterranean morphology needs more careful McDade, and two anonymous reviewers for reading previous
attention. versions of the manuscript and providing constructive feedback.

Literature Cited

Albers F, U Meve 2002 Illustrated handbook of succulent plants: As- porous sessile oak and diffuse-porous beech trees. Tree Physiol 22:
clepiadaceae. Springer, Berlin. 1201–1210.
Angiosperm Phylogeny Group 2009 An update of the Angiosperm Barneby RC 1991 Sensitivae censitae: a description of the genus Mi-
Phylogeny Group classification for the orders and families of flow- mosa L. (Mimosaceae) in the New World. Mem NY Bot Gard 65:
ering plants: APG III. Bot J Linn Soc 161:105–121. 1–835.
Arakaki M, P-A Christin, R Nyffeler, A Lendel, U Eggli, RM Ogburn, Bauhin J, JH Cherler 1619 Historiae plantarum generalis novae et
E Spriggs, MJ Moore, EJ Edwards 2011 Contemporaneous and absolutiss quinquaginta annis elaboratae iamprelo commissae prod-
recent radiations of the world’s major succulent plant lineages. Proc romus. Typographia Societatis Caldorianae, Yverdon, Switzerland.
Natl Acad Sci USA 108:8379–8384. Bruyns PV, C Klak 2006 A systematic study of the Old World genus
Barbaroux C, N Breda 2002 Contrasting distribution and seasonal Fockea (Apocynaceae-Asclepiadoideae). Ann Mo Bot Gard 93:535–
dynamics of carbohydrate reserves in stem wood of adult ring- 564.

This content downloaded from 129.72.2.27 on Fri, 6 Sep 2013 15:04:41 PM

All use subject to JSTOR Terms and Conditions
1060 INTERNATIONAL JOURNAL OF PLANT SCIENCES

Bucci SJ, FG Scholz, G Goldstein, FC Meinzer, LDL Sternberg Hearn D, M Huber 2006 The ancestral distance test: what relatedness
2003 Dynamic changes in hydraulic conductivity in petioles of two can reveal about correlated evolution in large lineages with missing
savanna tree species: factors and mechanisms contributing to the character data and incomplete phylogenies. Syst Biol 55:803–817.
refilling of embolized vessels. Plant Cell Environ 26:1633–1645. Hearn DJ 2004 Growth form evolution in Adenia (Passifloraceae)
Bullock SH 1992 Seasonal differences in nonstructural carbohydrates and a model of the evolution of succulence. PhD diss. University of
in two dioecious monsoon-climate trees. Biotropica 24:140–145. Arizona, Tucson.
Burgess TL, A Shmida 1988 Succulent growth-forms in arid environ- ——— 2006 Adenia (Passifloraceae) and its adaptive radiation: phy-
ments. Pages 383–395 in EE Whitehead, CF Hutchinson, BN Tim- logeny and growth form diversification. Syst Bot 31:805–821.
mermann, RG Verity, eds. Arid lands today and tomorrow. West- ——— 2009a Descriptive anatomy and evolutionary patterns of an-
view, Boulder, CO. atomical diversification in Adenia (Passifloraceae). Aliso 27:13–38.
Carlquist S 2001 Comparative wood anatomy: systematic, ecological ——— 2009b Developmental patterns in anatomy are shared among
and evolutionary aspects of dicotyledon wood. 2nd ed. Springer, separate evolutionary origins of stem succulent and storage root-
Berlin. bearing growth habits in Adenia (Passifloraceae). Am J Bot 96:1941–
Carroll SB 2008 Evo-devo and an expanding evolutionary synthesis: 1956.
a genetic theory of morphological evolution. Cell 134:25–36. ——— 2013 Dissection of evolutionary networks to assess their role
Celko J 1999 Joe Celko’s SQL for smarties: advanced SQL program- in the evolution of robustness, function, and diversification. Evo-
ming. 2nd ed. Academic Press, San Diego, CA. lution 67:2273–2283.
Chapotin SM, JH Razanameharizaka, NM Holbrook 2006a Baobab Hearn DJ, B Wolf, M McGinty, J Wen 2012 Phylogeography of Cy-
trees (Adansonia) in Madagascar use stored water to flush new phostemma (Vitaceae). Paper presented at Botany 2012: the next
leaves but not to support stomatal opening before the rainy season. generation, Columbus, OH, July 7–11.
New Phytol 169:549–559. Henrickson J 1969 Anatomy of periderm and cortex of Fouquieri-
——— 2006b A biomechanical perspective on the role of large stem aceae. Aliso 7:97–126.
volume and high water content in baobab trees (Adansonia spp.: Höll W 2000 Distribution, fluctuation and metabolism of food re-
Bombacaceae). Am J Bot 93:1251–1264. serves in the wood of trees. Pages 347–362 in R Savidge, J Barnett,
——— 2006c Water relations of baobab trees (Adansonia spp. L.) R Napier, eds. Cell and molecular biology of wood formation. BIOS
during the rainy season: does stem water buffer daily water deficits? Scientific, Oxford.
Plant Cell Environ 29:1021–1032. Horn JW, BW van Ee, JJ Morawetz, R Riina, VW Steinmann, PE Berry,
Du J, A Groover 2010 Transcriptional regulation of secondary KJ Wurdack 2012 Phylogenetics and the evolution of major struc-
growth and wood formation. J Integr Plant Biol 52:17–27. tural characters in the giant genus Euphorbia L. (Euphorbiaceae).
Du J, SD Mansfield, AT Groover 2009 The Populus homeobox gene Mol Phylogenet Evol 63:305–326.
ARBORKNOX2 regulates cell differentiation during secondary Ihlenfeldt H-D 1985 Lebensformen und Überlebensstrategien bei
growth. Plant J 60:1000–1014. Sukkulenten. Ber Dtsch Bot Ges 98:409–423.
Edwards EJ, MJ Donoghue 2006 Pereskia and the origin of the cactus Jacobsen H 1978 Handbook of succulent plants. Vols 1, 2. Blandford,
life-form. Am Nat 167:777–793. Poole.
Eggli U 1985 A bibliography of succulent plant periodicals. Bradleya Janse van Vuuren DR 1970 ‘n Morphologiese studie van die stingels
3:103–119. van die houtagtige verteenwoordigers van die genus Adenia Forsk.
——— 1987 More bibliographical data on succulent plant periodi- in Suid-Afrika. PhD diss. University of Pretoria, South Africa.
cals. Bradleya 5:101–102. Jones CS, RA Price 1996 Diversity and evolution of seedling Baupläne
——— 2002 Illustrated handbook of succulent plants: Dicotyledons. in Pelargonium (Geraniaceae). Aliso 14:281–295.
Springer, Berlin. Kubitzki K, ed 1990– The families and genera of vascular plants.
——— 2003 Illustrated handbook of succulent plants: Crassulaceae. Springer, Berlin.
Springer, Berlin. Lammers TG 2007 Campanulaceae. Pages 26–56 in JW Kadereit, C
Eggli U, R Nyffeler 2009 Living under temporarily arid conditions: Jeffrey, eds. The families and genera of vascular plants. Vol 8. Flow-
succulence as an adaptive strategy. Bradleya 27:13–36. ering plants: eudicots: Asterales. Springer, Berlin.
Evert RF 2006 Parenchyma and collenchyma. Pages 175–187 in Landhäusser SM, VJ Lieffers 2003 Seasonal changes in carbohydrate
Esau’s plant anatomy. 3rd ed. Wiley, Hoboken, NJ. reserves in mature northern Populus tremuloides clones. Trees 17:
Felsenstein J 1985 Phylogenies and the comparative method. Am Nat 471–476.
125:1–15. Mabberley D 1996 The plant-book: a portable dictionary of the
Freire-Fierro A 1992 Estudos taxonomicos dos subgeneros de Mon- higher plants. Press Syndicate of the University of Cambridge,
nina Ruiz & Pavón (Polygalaceae), baseados em species brasileiras Cambridge.
e equatorianas. MS diss. Universidade de São Paulo, Brasil. Maddison DR, WP Maddison 2000 MacClade version 4: analysis of
Futuyma DJ 1997 Evolutionary biology. 3rd ed. Sinauer, Sunderland, phylogeny and character evolution. Sinauer, Sunderland, MA.
MA. Marques MC 1989 Monnina Ruiz et Pavón (Polygalaceae) no Brasil.
Gardner AG, M Vaio, M Guerra, E Emshwiller 2012 Diversification Rodriguésia 67:3–33.
of the American bulb-bearing Oxalis (Oxalidaceae): dispersal to Mauseth JD 1993a Medullary bundles and the evolution of cacti. Am
North America and modifications of the tristylous breeding system. J Bot 80:928–932.
Am J Bot 99:152–164. ——— 1993b Water-storing and cavitation-preventing adaptations in
Gibson AC, PS Nobel 1986 The cactus primer. Harvard University wood of cacti. Ann Bot 72:81–89.
Press, Cambridge, MA. ——— 1995 Collapsible water-storage cells in cacti. Bull Torrey Bot
Godı́nez-Álvarez H, T Valverde, P Ortega-Baes 2003 Demographic Club 122:145–151.
trends in the Cactaceae. Bot Rev 69:173–203. ——— 1999 Anatomical adaptations to xeric conditions in Mai-
Groover AT 2005 What genes make a tree a tree? Trends Plant Sci huenia (Cactaceae), a relictual, leaf-bearing cactus. J Plant Res 112:
10:210–214. 307–315.
Groover AT, SD Mansfield, SP DiFazio, G Dupper, JR Fontana, R Millar, ——— 2000 Theoretical aspects of surface-to-volume ratios and
Y Wang 2006 The Populus homeobox gene ARBORKNOX1 reveals water-storage capacities of succulent shoots. Am J Bot 87:1107–1115.
overlapping mechanisms regulating the shoot apical meristem and the ——— 2004 The structure of photosynthetic succulent stems in
vascular cambium. Plant Mol Biol 61:917–932. plants other than cacti. Int J Plant Sci 165:1–9.

This content downloaded from 129.72.2.27 on Fri, 6 Sep 2013 15:04:41 PM

All use subject to JSTOR Terms and Conditions
 HEARN ET AL.—EVOLUTION OF GROWTH FORMS CORRELATED ACROSS EUDICOTS 1061

——— 2006a Blossfeldia lacks cortical bundles and persistent epi- regulatory evolution. Proc Natl Acad Sci USA 104(suppl):8605–
dermis: is it basal within Cactoideae? Bradleya 24:73–82. 8612.
——— 2006b Structure-function relationships in highly modified Pruyn ML, R Spicer 2012 Parenchyma. In Encyclopedia of life sci-
shoots of Cactaceae. Ann Bot 98:901–926. ence. Wiley, Chichester. http://www.els.net. http://dx.doi.org/10
Mauseth JD, JV Landrum 1997 Relictual vegetative anatomical char- .1002/9780470015902.a0002083.pub2.
acters in Cactaceae: the genus Pereskia. J Plant Res 110:55–64. R Development Core Team 2012 R: a language and environment for
Mauseth JD, BJ Plemons-Rodriguez 1998 Evolution of extreme xe- statistical computing. R Foundation for Statistical Computing, Vi-
romorphic characters in wood: a study of nine evolutionary lines in enna. http://www.R-project.org/.
Cactaceae. Am J Bot 85:209–218. Robischon M, JA Du, E Miura, A Groover 2011 The Populus class
Mauseth JD, M Sajeva 1992 Cortical bundles in the persistent, pho- III HD ZIP, popREVOLUTA, influences cambium initiation and pat-
tosynthetic stems of cacti. Ann Bot 70:317–324. terning of woody stems. Plant Physiol 155:1214–1225.
Mauseth JD, ME Stone-Palmquist 2001 Root wood differs strongly Rowley GD 1976 The rise and fall of the “Succulentae.” Cact Succ
from shoot wood within individual plants of many Cactaceae. Int J 48:184–189.
J Plant Sci 162:767–776. ——— 1980 Name that succulent: keys to the families and genera of
Mauseth JD, T Terrazas, S Loza-Cornejo 1998 Anatomy of relictual succulent plants in cultivation. S Thornes, Cheltenham.
members of subfamily Cactoideae, IOS Group 1a (Cactaceae). Brad- ——— 1987 Caudiciform and pachycaul succulents: pachycauls, bot-
leya 16:31–43. tle-, barrel- and elephant-trees and their kin: a collector’s miscellany.
Mooney HA, C Chu, SH Bullock, R Robichaux 1992 Carbohydrate, Strawberry, Mill Valley, CA.
water and nitrogen storage in vines of a tropical deciduous forest. Salleo S, MA Lo Gullo, P Trifilo, A Nardini 2004 New evidence for
Biotropica 24:134–139. a role of vessel-associated cells and phloem in the rapid xylem re-
Namikawa K, M Endo 1932 Anatomical analysis of root growth of filling of cavitated stems of Laurus nobilis L. Plant Cell Environ 27:
turnip cv. Sugukina (Brassica rapa L.). Agric Hort (Nogyo Oyobi 1065–1076.
Engei) 7:3–16. Salleo S, P Trifilo, S Esposito, A Nardini, MA Lo Gullo 2009 Starch-
Niklas KJ 1997 The evolutionary biology of plants. University of Chi- to-sugar conversion in wood parenchyma of field-growing Laurus
cago Press, Chicago. nobilis plants: a component of the signal pathway for embolism
Nobel PS 1980 Morphology, surface temperatures, and northern lim- repair? Funct Plant Biol 36:815–825.
its of columnar cacti in the Sonoran Desert. Ecology 61:1–7. Schrader J, J Nilsson, E Mellerowicz, A Berglund, P Nilsson, M Hertz-
——— 2006 Parenchyma-chlorenchyma water movement during berg, G Sandberg 2004 A high-resolution transcript profile across
drought for the hemiepiphytic cactus Hylocereus undatus. Ann Bot the wood-forming meristem of poplar identifies potential regulators
97:469–474. of cambial stem cell identity. Plant Cell 16:2278-2292.
Noh SA, HS Lee, EJ Huh, KH Paek, JS Shin, JM Bae 2010 SRD1 is Scotland RW, MJ Sanderson 2004 The significance of few versus
involved in the auxin-mediated initial thickening growth of storage many in the tree of life. Science 303:643.
root by enhancing proliferation of metaxylem and cambium cells in Secchi F, MA Zwieniecki 2011 Sensing embolism in xylem vessels:
sweetpotato (Ipomoea batatas). J Exp Bot 61:1337–1349. the role of sucrose as a trigger for refilling. Plant Cell Environ 34:
Nyffeler R, U Eggli, M Ogburn, E Edwards 2008 Variations on a 514–524.
theme: repeated evolution of succulent life forms in the Portulacineae Spicer R, A Groover 2010 Evolution of development of vascular cam-
(Caryophyllales). Haseltonia 14:26–36. bia and secondary growth. New Phytol 186:577–592.
Ogburn RM, EJ Edwards 2010 The ecological water-use strategies of Stearns SC, RF Hoekstra 2005 Evolution: an introduction. 2nd ed.
succulent plants. Adv Bot Res 55:179–225. Oxford University Press, Oxford.
——— 2012 Quantifying succulence: a rapid, physiologically mean- Steinmann VW 2001 The evolution of succulence in the new world
ingful metric of plant water storage. Plant Cell Environ 35:1533– species of Euphorbia (Euphorbiaceae). PhD diss. Claremont Grad-
1542. uate University, Claremont, CA.
——— 2013 Repeated origins of three-dimensional leaf venation re- Stevens PF 2001– Angiosperm Phylogeny website. Version 12, July
leases constraints on the evolution of succulence in plants. Curr Biol 2012. http://www.mobot.org/MOBOT/research/APweb/. Accessed
23:722–726. December 11, 2012.
Olson ME 2003 Stem and leaf anatomy of the arborescent Cucur- Stone RD 2006 Phylogeny of major lineages in Melastomataceae,
bitaceae Dendrosicyos socotrana with comments on the evolution subfamily Olisbeoideae: utility of nuclear glyceraldehyde 3-phos-
of pachycauls from lianas. Plant Syst Evol 239:199–214. phate dehydrogenase (GapC) gene sequences. Syst Bot 31:107–121.
Olson ME, S Carlquist 2001 Stem and root anatomical correlations Stone RD, K Andreasen 2010 The Afro-Madagascan genus Warne-
with life form diversity, ecology, and systematics in Moringa (Mor- ckea (Melastomataceae): molecular systematics and revised infra-
ingaceae). Bot J Linn Soc 135:315–348. generic classification. Taxon 59:83–92.
Olson ME, JA Rosell 2006 Using heterochrony to detect modularity Stone-Palmquist M, JD Mauseth 2002 The structure of enlarged stor-
in the evolution of stem diversity in the plant family Moringaceae. age roots in cacti. Int J Plant Sci 163:89–98.
Evolution 60:724–734. Thiers B 2012 Index Herbariorum: a global directory of public her-
Pate JS, KW Dixon 1981 Plants with fleshy underground storage or- baria and associated staff. New York Botanical Garden’s Virtual
gans: a Western Australian survey. Pages 181–215 in JS Pate, AJ Herbarium, http://sweetgum.nybg.org/ih/. Accessed December 15,
McComb, eds. The biology of Australian plants. University of West- 2012.
ern Australia Press, Nedlands. Thomas DC, M Hughes, T Phutthai, S Rajbhandary, R Rubite, WH
Pennington TD, BT Styles, DAH Taylor 1981 Meliaceae. Flora Neo- Ardi, JE Richardson 2011 A non-coding plastid DNA phylogeny
tropica Monograph 28. New York Botanical Garden Press, Bronx. of Asian Begonia (Begoniaceae): evidence for morphological ho-
Pierson EA, RM Turner 1998 An 85-year study of saguaro (Carnegiea moplasy and sectional polyphyly. Mol Phylogenet Evol 60:428–444.
gigantea) demography. Ecology 79:2676–2693. Thomas PA 1991 Response of succulents to fire: a review. Int J Wild-
Polhill RM 1976 Genisteae (Adans.) Benth. and related tribes. Bot land Fire 1:11–22.
Syst 1:143–368. Thomas PA, P Goodson 1992 Conservation of succulents in desert
Porter JM, LA Johnson 2010 A phylogenetic classification of Pole- grasslands managed by fire. Biol Conserv 60:91–100.
moniaceae. Aliso 19:55–91. van Bel AJE 1990 Xylem-phloem exchange via the rays: the under-
Prud’homme B, N Gompel, SB Carroll 2007 Emerging principles of valued route of transport. J Exp Bot 41:631–644.

This content downloaded from 129.72.2.27 on Fri, 6 Sep 2013 15:04:41 PM

All use subject to JSTOR Terms and Conditions