ArticlePDF Available

Disintegrating Portulacaceae: A new familial classification of the suborder Portulacineae (Caryophyllales) based on molecular and morphological data

Authors:

Abstract and Figures

Traditional classifications of the suborder Portulacineae recognize six families: Basellaceae, Cactaceae, Didiereaceae, Halophytaceae, Hectorellaceae, and Portulacaceae. However, phylogenetic analyses based on molecular sequence data indicate that the traditional family Portulacaceae is paraphyletic and consists of three distinct lineages that also include Cactaceae, Didiereaceae, and Hectorellaceae. We use sequence data from the chloroplast genes matK and ndhF representing 64 species of Portulacineae and outgroups to reconstruct their phylogenetic relationships with Bayesian and maximum parsimony inference methods. Evidence from these molecular phylogenetic analyses as well as from comparative morphological investigations allow us to propose a revised familial classification of the suborder Portulacineae. We recognize eight monophyletic families: Anacampserotaceae (Anacampseros, Grahamia, Talinopsis), Basellaceae, Cactaceae, Didiereaceae (incl. Calyptrotheca, Ceraria, Portulacaria), Halophytaceae, Montiaceae (incl. Hectorellaceae, Calandrinia, Cistanthe, Claytonia, Lewisia, Montia, Phemeranthus), Portulacaceae (Portulaca only), and Talinaceae (Amphipetalum, Talinella, Talinum). We provide a synopsis for this revised family classification with an identification key mainly based on habit and fruit characters, and family diagnoses with information on distribution, taxonomic diversity, and a brief discussion on phylogenetics and classification.
Content may be subject to copyright.
227
Nyffeler & Eggli •
New familial classification of Portulacineae
TAXON 59 (1) • February 2010: 227–240
INTRODUCTION
Molecular sequence techniques and rigorous analytical
methods have had a profound impact on the higher classi-
fication of flowering plants (e.g., Soltis & al., 2005; Judd &
al., 2007). Recently, a well-supported clade consisting of the
families Basellaceae, Cactaceae, Didiereaceae, Halophyta-
ceae, Hectorellaceae, and Portulacaceae was identified on the
basis of several molecular systematic studies (Hershkovitz &
Zimmer, 1997; Applequist & Wallace, 2001; Cuénoud & al.,
2002; Hilu & al., 2003; Müller & Borsch, 2005; Applequist &
al., 2006; Nyffeler, 2007). This monophyletic group largely
corresponds to the suborder Portulacineae as originally pro-
posed by Engler (1898) and recently recognized by Thorne
(1976, 2000 [as Cactineae]) and Takhtajan (1997). Further-
more, these molecular studies indicate that the traditional fam-
ily Portulacaceae consists of three distinct lineages of which
one includes Cactaceae (e.g., Hershkovitz & Zimmer, 1997;
Nyffeler, 2007), one includes Didiereaceae (Applequist & Wal-
lace, 2001, 2003), and one includes Hectorellaceae (Applequist
& al., 2006; Wagstaff & Hennion, 2007). Basellaceae and the
monotypic Halophytaceae represent two additional major lin-
eages of suborder Portulacineae (Cuénoud & al., 2002; Müller
& Borsch, 2005). The circumscription of Portulacineae previ-
ously received support from studies in palynology (Nowicke,
1996) and wood anatomy (Carlquist, 1997).
The family Portulacaceae was first established by Adanson
(1763) and was later taken up by de Jussieu (1789), to whom
the name must be ascribed under ICBN Art. 13.1(1) (McNeill
& al., 2006). Adanson included a total of 34 genera belong-
ing to more than 20 currently recognized families, including
Portulacaceae, Aizoaceae, and Cactaceae, and the unrelated
Begoniaceae, Cuscutaceae, Saxifragaceae, Theophrastaceae,
and Turneraceae. De Jussieu (1789) provided a considerably
narrower circumscription of the family Portulacaceae (as
‘Portulaceae’) and placed it between Cactaceae (as ‘Cacti’)
and Aizoaceae (as ‘Ficoideae’). Floral characters such as petal
number and arrangement, ovary position, and fruit morphol-
ogy were used by him to differentiate Portulacaceae from their
allies. He included the genera Portulaca L., Talin u m Adans.,
Montia L., and Claytonia L. along wit h some genera cu rrently
placed in Aizoaceae, Caryophyllaceae, Gisekiaceae, and Mol-
luginaceae and the unrelated families Plantaginaceae sensu
lato (s.l.), Tamaricaceae, and Turneraceae.
For more than two hundred years the circumscription of
Portulacaceae was debated, modified, and adjusted up to the
most recent treatment by Carolin (1993). Portulacaceae, as
traditionally (i.e., in the sense of Carolin, 1993) understood,
comprise 30 genera and about 450 species, mainly character-
iz ed by the pre sence of two se paloids, often five fast ly withe r-
ing petaloids, and capsular fruits consisting of usually three
fused carpels. This suite of characters is primarily responsible
for the traditional circumscription of Portulacaceae remaining
unchallenged for so long (e.g., Cronquist, 1981; Car olin, 1993;
Takhtajan, 1997, 2009). The claim that Portulacaceae are “a
very natural and easily recognized family” (Brummitt, 2002:
Disintegrating Portulacaceae: A new familial classification of the
suborder Portulacineae (Caryophyllales) based on molecular and
morphological data
Reto Nyffeler1 & Urs Eggli2
1 Institut für Systematische Botanik, Universität Zürich, Zollikerstrasse 107, 8008 Zürich, Switzerland
2 Sukkulenten-Sammlung Zürich, Grün Stadt Zürich, Mythenquai 88, 8002 Zürich, Switzerland
Author for correspondence: Reto Nyffeler, reto.nyffeler@systbot.uzh.ch
Abstract
Traditional classifications of the suborder Port ulacineae recognize six families: Basellaceae, Cactaceae, Didiereaceae,
Halophytaceae, Hectorellaceae, and Portulacaceae. However, phylogenetic analyses based on molecular sequence data indicate
that the traditional family Portulacaceae is paraphyletic and consists of three distinct lineages that also include Cactaceae,
Didiereaceae, and Hectorellaceae. We use sequence data from the chloroplast genes matK and ndhF representing 64 species
of Portulacineae and outgroups to reconstruct their phylogenetic relationships with Bayesian and maximum parsimony infer-
ence methods. Evidence from these molecular phylogenetic analyses as well as f rom comparative morphological investigations
allow us to propose a revised familial classification of the suborder Portulacineae. We recognize eight monophyletic families:
Anacampserotaceae (Anacampseros, Grahamia, Talinop sis), Basellaceae, Cactaceae, Didiereaceae (incl. Calyptrotheca, Ceraria,
Portulacaria), Halophytaceae, Montiaceae (incl. Hectorellaceae, Calandrinia, Cistanthe, Claytonia, Lewisia, Montia, Phem-
eranthus), Portulacaceae (Portulaca only), and Talinaceae (Amphipetalum, Talinella, Talinu m). We provide a synopsis for this
revised family classification with an identification key mainly based on habit and fruit characters, and family diagnoses with
information on distribution, taxonomic diversity, and a brief discussion on phylogenetics and classification.
Keywords
Anacampserotaceae; description; familial classification; molecular phylogenetics; Montiaceae; Portulacaceae;
Tal inaceae
228
TAXON 59 (1) • February 2010: 227–240Nyffeler & Eggli •
New familial classification of Portulacineae
36) reflects the selective characterization of this taxon mainly
by sustaining an established tradition. Basellaceae (Sperling
& Bittrich, 1993) and the traditional Didiereaceae (Kubitzki,
1993) also have flowers with two sepaloids very similar to
those of Portulacaceae, but have been maintained as distinct
families mainly based on their twining or cactus-like xero-
phytic habit. Furthermore, a crit ical comparative evaluation of
all relevant taxa of Portulacineae for different morphological
and anatomical characters was hampered by the very distinc-
tive nature of the family Cactaceae. For decades, this family
was retained in a separate monofamilial order allied with Pa-
rietales (e.g., Engler, 1925) or Cucurbitales (e.g., Hutchinson,
1973), respectively, despite solid contrasting arguments (e.g.,
de Candolle, 1828; Schumann, 1899; Chorinsky, 1931). Only
recently, the family Cactaceae was considered for comparative
investigations of the closely related families now included in
Portulacineae (e.g., Hershkovitz, 1993).
Starting very early, two different concepts of the family
Portulacaceae were proposed: (1) a narrower concept favored
by de Candolle (1828) and Bartling (1830) that includes gen-
era of current Portulacaceae and Basellaceae, and (2) a much
broader concept proposed by Fenzl (1836, 1839), and supported
by Endlicher (1840) and Baillon (1886a), that in addition in-
cludes also some genera of current Aizoaceae, Caryophylla-
ceae, and Molluginaceae. Later, the treatments by Pax (1889)
and Pax & Hoffmann (1934), that adhered to the narrower
concept as outlined by Bentham (1862a,b), brought along a
consensus to recognize Portulacaceae with its narrower cir-
cumscription. Franz (1908), based on comparative morpho-
logical and anatomical investigations of various characters
(i.e., inflorescences, flowers, fruits, and pollen as well as stem
vasculature and stomata), argued to retain the genera of Ba-
sellaceae in Portulacaceae. In contrast, Pax (1889) and Pax
& Hoffmann (1934) excluded them and placed them in their
own family. Furthermore, while Pax (1889) placed Hectorella
Hook. f. in Portulacaceae, Pax & Hoffmann (1934) included it
in Caryophyllaceae along with the closely related genus Lyallia
Hook. f. During the past few decades, separate families (e.g.,
Halophytaceae, Hectorellaceae) were recognized for taxa with
ambiguous or unresolved relationships (e.g., Dahlgren, 1980;
Kubitzki & al., 1993; Takhtajan, 1997; Thorne, 2000). Only
Cronquist (1981) and Cronquist & Thorne (1994) opted to re-
tain Hectorella and Lyallia in Portulacaceae and Halophytum
Speg. in Chenopodiaceae. On the other hand, recent molecular
phylogenetic studies provided ample evidence that the genera
Dendroportulaca Eggli and Pleuropetalum Hook. f., which
were originally included in Portulacaceae (Hooker, 1846;
Eggli, 2002), are representatives of Amaranthaceae (Appleq-
uist & Pratt, 2005; Müller & Borsch, 2005). Furthermore, simi-
lar studies confirm that Taline lla Baill. unambiguously belongs
in Portulacaceae (e.g., Hershkovitz & Zimmer, 1997; Appleq-
uist & Wallace, 2001; Nyffeler, 2007), despite its unique fruit
type (mucilaginous berry rather than dry capsule). On the other
hand, Calyptrotheca Gilg, first described as a member of Cap-
paraceae, was subsequently included in Portulacaceae and is
currently included in an expanded concept of Didiereaceae
(Applequist & Wallace, 2001, 2003).
The first infrafamilial classification of traditional Portu-
lacaceae was proposed by Franz (1908). He recognized two
subfamilies, one of which was divided into two tribes and
four subtribes using morphological features (i.e., pollen mor-
phology, form of the ovary base, number of carpels, number
of ovules, micropyle orientation, and number of sepaloid or-
gans). Pax & Hoffmann (1934) only slightly modified this
classification by recognizing the tribe Baselleae of subfamily
Montioideae as a separate family, and by treating the genera
Ceraria Pearson & Stephens, Portulacaria Jacq., and Philip-
piamra Kuntze as “intermediate between Portulacaceae and
Basellaceae” (Pax & Hoffmann, 1934: 244). More recently,
infrafamilial classifications by McNeill (1974), Carolin (1987,
1993), and Nyananyo (1990) recognized either seven, five or
four, and eight tribes, respectively (for details see table 1 in
Applequist & Wallace, 2001). The number of recognized gen-
era ranges from 17 (Nyananyo, 1990) or 18 (McNeill, 1974) up
to 28 or 29 (Carolin, 1987, 1993). These observations clearly
indicate the widely differing views on the relationships among
the taxa of Portulacaceae and their circumscriptions.
Evidence is growing that the traditional, vaguely character-
ized family Portulacaceae re present s an assembly of different
evolutionary lineages (e.g., Bittrich, 1993a; Hershkovitz &
Zimmer, 1997; Applequist & Wallace, 2001, Cuénoud & al.,
2002; Nyffeler, 2007), and this is the main rationale for the
present study. Available data from molecular phylogenetic
investigations and from comparative investigations of morpho-
logical characters allow us to propose a revised classification
of the suborder Portulacineae. We suggest a recircumscription
of the family Portulacaceae and the recognition of three ad-
ditional families that are either new or have not been recently
used. The objectives of the present study are (1) to present a
molecular phylogenetic analysis of Portulacineae based on the
cpDNA markers matK and ndhF from a r epresent ative sample
of individuals, (2) to transform the inferred relationships into
a hierarchical classification of monophyletic families, and (3)
to provide a concise taxonomic treatment of all families rec-
ognized in Portulacineae. We chose to use a ‘supermatrix’
approach (de Queiroz & Gatesy, 2006) by relying primarily on
the ndhF data to provide information on the core relationships
and by further resolving relationships towards the tips with a
more densely sampled matK dataset.
MATERIALS AND METHODS
Taxon sampling and markers. —
In our analysis of
suborder Portulacineae we included 59 representatives of
the traditional families Basellaceae, Cactaceae, Didierea-
ceae, Halophytaceae, Hectorellaceae, and Portulacaceae.
These ingroup taxa were carefully selected from published
sequences in order to fully represent the taxonomic diversity
of the study group. Furthermore, we included five samples
from the families Aizoaceae, Molluginaceae, Nyctaginaceae,
and Phytolaccaceae as outgroups. Overall, 53 sequences of
matK and 37 sequences of ndhF from previously published
studies (Olmstead & al., 2000; Applequist & Wallace, 2001;
229
Nyffeler & Eggli •
New familial classification of Portulacineae
TAXON 59 (1) • February 2010: 227–240
Cuénoud & al., 2002; Nyffeler, 2002, 2007; Edwards & al.,
2005; Müller & Borsch, 2005; Applequist & al., 2006) were
obtained from GenBank (http://www.ncbi.nlm.nih.gov/) and
seven matK sequences were newly generated. Four species of
Talinum (incl. Phemeranthus Raf.) were newly added in order
to address the recent observation that this genus, as tradition-
ally circumscribed, might be polyphyletic (Hershkovitz &
Zimmer, 1997; Applequist & Wallace, 2001; Ferguson, 2001).
Two species of Portulaca were also added to the present study
to increase the taxon sampling of this species-rich lineage.
Finally, the sampling of Cactaceae subfamily Opuntioideae
was expanded by sequencing Maihueniopsis subterranea (R.E.
Fr.) E.F. Anderson.
DNA
extraction,
PCR
amplification and sequencing.
Total DNA was extracted from silica gel–dried stem or
leaf material using the DNeasy Plant Mini Kit (Qiagen Corp.).
External primers trnK-3914F and trnK-2R were used for ampli-
fication of trnK/matK (Johnson & Soltis, 1994). The 20-μl PCR
reactions contained 10 mM Tris-HCl, 50 mM KCl, 2.5 mM
MgCl2, 0.2 mM of each dNTP, 0.4 mM of each primer, and
0.8 units of AmpliTaq polymerase (Perkin-Elmer Applied Bio-
systems). The PCR temperature profile for the reactions was
95°C for 5 min, then 35 cycles of 94°C for 30 s, 48°C for 60 s,
72°C for 90 s, followed by a final extension of 72°C for 5 min.
Double-stranded PCR products were cleaned using QIAquick
columns (Qiagen Corp.) and directly sequenced for the matK
gene using the internal primers trnK-23F, trnK-41R, trnK-44F,
trnK-52F, and trnK-71R (Nyffeler, 2002). The products were
cleaned with Microspin G-50 (Amersham Pharmacia Biotech)
using multiscreen plates to remove excess Big Dye Termina-
tor before loading on the automated sequencer ABI PRISM
3100 Genetic Analyzer (Perkin-Elmer Applied Biosystems).
All seven newly created matK sequences were checked and
assembled using the software Sequencher v.4.2 (Gene Codes
Corporation, 2000) and are available from GenBank (Acces-
sion numbers EU834746–EU834752; see Appendix 1).
Phylogenetic analyses. —
For 29 out of the 64 repre-
sentatives only one of the two cpDNA markers were available
(see Appendix 1). Combining sequences of several different
molecular markers into a combined matrix, in particular if
they are derived from the same genome, is widely used today
(e.g., de Queiroz & Gatesy, 2006). Furthermore, the issue of
missing data is no longer regarded as a major problem when
dealing with incompletely coded taxa (Kearny, 2002) if “there
are sufficient characters in one broadly sampled dataset to al-
low the position of these taxa to be resolved” (Wiens, 2003:
536). Alignment of the two partitions was done by eye and
these were combined into a single matrix. Only very few in-
formative indels were located, and therefore not further con-
sidered. We conducted maximum parsimony (MP) and Bayes-
ian inference (BI) analyses using PAUP* v.4.0b10 (Swofford,
2002) and MrBayes v.3.1.1 (Huelsenbeck & Ronquist, 2001).
The parsimony analysis was based on 100,000 heuristic search
replicates using random taxon addition and TBR branch swap-
ping with MULTREES on. Bootstrap support values were cal-
culated using 10,000 replicates, each with simple taxon addition
and MAXTREES set to 1000. We also conducted individual
analyses of the matK and ndhF datasets comprising 60 and 37
samples, respectively. Then, we compared the topologies of the
two majority-rule consensus trees from the bootstrap analyses
for possible incongruence and sampling effects.
The Bayesian inference analysis was performed on the
combined dataset, allowing model parameters to be indepen-
dently estimated for the two partitions. We used MrModeltest
v.2.2 (Nylander, 2004) to identify the best available model of
molecular evolution where the Akaike information criterion
favored the GTR + G + I model for both partitions. We con-
ducted three independent Markov chain Monte Carlo runs,
each consisting of four linked chains with standard settings
that were run for five million generations (sampling every 100
generations). The burn-in was set to 500,000 generations, after
comparing plots of log-likelihood values against generation
time across independent runs. Furthermore, topologies derived
from the three majority rule consensus trees and clade pos-
terior probabilities were compared to check for good mixing
during the BI analysis. The 135,003 post-burn-in trees of the
three independent runs were then combined into a single ma-
jority rule consensus phylogram using the SUMT command,
which also estimated the branch lengths. The aligned data
matrix and the majority rule consensus tree of the Bayesian
inference analysis are available at TreeBase (http://www.tree
base.org/; study accession number = S2283, matrix accession
number = S2283).
RESULTS
Dataset. —
The aligned matrix of the combined matK and
ndhF dataset comprises 3706 characters, of which 2306 are
constant, 632 are variable but parsimony uninformative, and
768 are parsimony informative. Overall, the average percent-
age of missing characters for the combined dataset is 33.4%
(nucleotides in gaps of the aligned matrix not considered).
In the matK data partition (1563 characters aligned) 16.4%
of nucleotides are missing, while in the ndhF partition (2143
characters aligned) 45.9% of nucleotides are missing.
Parsimony analysis. —
The MP analysis of the com-
bined dataset yielded 98,559 most parsimonious trees (length
= 2852; consistency index, CI = 0.656; retention index, RI =
0.704; RI excluding uninformative characters = 0.539). The
strict consensus tree, of which the topology largely corre-
sponds to the majority rule consensus tree of the BI analysis
(Fig. 1), resolves 45 clades, but does not provide unambiguous
information for a sister-group relationship between ANAC
(Anacampseros L., Grahamia Hook., Talino p sis A. Gray) and
CACT (Cactaceae), nor for the monophyly of MONT (Phemer-
anthus as part of a clade including Cistanthe Spach, Lewisia
Pursh, and Montia). Bootstrap support values are listed below
the branches onto the majority-rule consensus tree derived
from the BI analysis (Fig. 1). The bootstrap majority-rule
consensus trees of the individual analyses identified almost all
highly supported clades that are also present in the combined
analysis, and did not yield any incongruence in the topologies
and underlying datasets.
230
TAXON 59 (1) • February 2010: 227–240Nyffeler & Eggli •
New familial classification of Portulacineae
Fig. 1.
Majority rule consensus of 135,003 trees derived from a Bayesian inference (BI) analysis of the suborder Portulacineae. Dashed branches
collapse in the strict consensus cladogram of the maximum parsimony (MP) analysis. Numbers above branches denote posterior probabilities
of the BI analysis, numbers below branches indicate bootstrap values of the MP analysis. Eight distinct major clades are marked with character
codes; we propose that they are recognized as the families Anacampserotaceae (ANAC), Basellaceae (BASE), Cactaceae (CACT), Didierea-
ceae (DIDI), Halophytaceae (HALO), Montiaceae (MONT), Portulacaceae (PORT), and Talinaceae (TALI). An asterisk marks new species
combinations (see Appendix 2).
Delosperma cooperi
Mirabilis jalapa
Phytolacca americana/dioica
Limeum africanum
Mollugo verticillata
Phemeranthus teretifolius
Phemeranthus multiflorus
Phemeranthus punae*
Cistanthe grandiflora
Calyptridium umbellatum
Parakeelya volubilis
Hectorella caespitosa
Lewisia cantelovii/pygmaea
Montia parvifolia
Halophytum ameghinoi
Anredera cord ifolia
Basella alba
Ullucus tuberosus
Ceraria fruticulosa
Portulacaria afra
Calyptrotheca somalensis
Alluaudia ascendens/humbertii
Decarya madagascariensis
Didierea trollii
Talinum caffrum
Talinum lineare
Talinum polygaloides
Talinum paniculatum
Talinella pachypoda
Talinum portulacifolium
Talinum triangulare
Portulaca bicolor
Portulaca fluvialis
Portulaca oleracea
Portulaca eruca
Portulaca grandiflora
Talinopsis frutescens
Grahamia bracteata
Anacampseros vulcanensis
Anacampseros kurtzii
Anacampseros coahuilensis*
Anacampseros australiana
Anacampseros papyracea
Anacampseros albissima
Anacampseros recurvata
Anacampseros telephiastrum
Anacampseros karasmontana
Anacampseros retusa
Anacampseros subnuda
Pereskia guamacho
Pereskia zinniiflora
Pereskia aculeata
Pereskia stenantha
Maihuenia patagonica
Opuntia quimilo
Pereskiopsis diguetii
Austrocylindropuntia vestita
Maihueniopsis subterranea
Blossfeldia liliputana
Echinocactus platyacanthus
Calymmanthium substerile
Copiapoa bridgesii
Rhipsalis floccosa
Stetsonia coryne
CACT
ANAC
PORT
TALI
DIDI
BASE
MONT
HALO
1.00
100
1.00
82
0.90
-
1.00
100
1.00
90
1.00
98
1.00
54
1.00
100
0.92
-
1.00
91
1.00
55
0.83
-
1.00
99 1.00
98
1.00
98 1.00
100 0.98
64 0.97
87
1.00
95
0.99
90
1.00
96
1.00
100 1.00
90
0.98
60 1.00
97
1.00
100 0.92
75 1.00
100
0.93
73
0.94
-
1.00
100 1.00
100 0.80
78 1.00
94 1.00
100
1.00
96
1.00
100 1.00
85
1.00
100 1.00
100
1.00
99
0.77
69
0.85
-
1.00
100 1.00
100
0.95
82 1.00
90 1.00
88 0.82
68
231
Nyffeler & Eggli •
New familial classification of Portulacineae
TAXON 59 (1) • February 2010: 227–240
Bayesian analysis. —
The three independent BI runs
yielded similar topologies with minimal variation in clade pos-
terior probabilities of the post–burn-in majority-rule consensus
trees. The three individual tree pools were combined into one
majority-rule consensus tree (Fig. 1). Posterior probability
values for the different clades are given above the branches
(Fig. 1). This consensus tree does not support a sister-group re-
lationship of Ba sellacea e with Halophy taceae. In contrast, such
a close relationship was indicated by the strict consensus of
the MP analysis, though with less than 50% bootstrap support.
DISCUSSION
Comparison with previous molecular phylogenetic
studies. —
The present study is based on a combined analy-
sis of the two molecular markers matK and ndhF from the
chloroplast genome, which have been used most often so far
to infer relationships within and among families of Portula-
cineae (Applequist & Wallace, 2001; Cunéoud & al., 2002;
Applequist & al., 2006; Nyffeler, 2007). For the first time,
this study provides a concise but well-balanced sampling of
all eight major lineages of this suborder to infer their interrela-
tionships and to recircumscribe all the families included. The
resulting topology (Fig. 1) is congruent, for clades that receive
reasonable statistical support, with previous studies relying on
the same molecular markers (e.g., Applequist & Wallace, 2001;
Applequist & al., 2006; Nyffeler, 2007). A major challenge
remains to identify possible sister-group relationships among
the eight major lineages identified here. So far we encounter
the strongest support for a clade consisting of ANAC, CACT,
PORT, and TALI (Fig. 1; ACPT clade of Nyffeler, 2007).
Character interpretation. —
Flor al envelope chara cters,
in combination with aspects of the habit, have been used in
the past for the circumscription of the family Portulacaceae
(e.g., Carolin, 1993). In particular, the presence of two, of-
ten unequal sepaloid floral parts is commonly listed as a key
character for Portulacaceae (e.g., Geesink & al., 1981; Cullen,
1997), even though it is not consistent nor unique. Many spe-
cies of Lewisia have five to nine sepaloid elements, while the
members of the traditionally segregated families Basellaceae
and Didiereaceae also have two sepaloids. Furthermore, this
character is inconsistently identified and termed either as ‘se-
pals’ (e.g., Geesink, 1969; Cronquist, 1981; Carolin, 1993),
‘sepaloid bracts’ (Legrand, 1949), ‘pseudosepals’ (Legrand,
1953), or ‘bracteoles’ (Zomlefer, 1995; Judd & al., 2007). Pax
& Hoffmann (1934) were the first to recognize that these ‘se-
pals’ correspond to involucral bracts, and that the ‘petals’ are
elements of a perigone. Already Payer (1857; cited in Friedrich,
1956) pointed out that Portulacaceae do not have ‘true’ petals.
This inter pret ation is now widely r ecogniz ed (see also E ckardt,
1976). Erbar & Leins (2006), in a study of Didiereaceae, sug-
gest that this condition applies to all higher Caryophyllales.
Traditional Portulacaceae and related families show a
wealth of different fruit types. Capsular fruits are promi-
nent, either opening at the base or near the top by valves (e.g.,
Anacampseros, Calandrinia Kunth, Calyptrotheca, Talinum,
etc.) or by a circumscissile lid (i.e., Amphipetalum Bacigalupo,
Lewisia, Lewisiopsis Govaerts, Portulaca). More rare are vari-
ous forms of indehiscent nutlets (e.g., Halophytum, Hectorella,
Lyallia, Philippiamra), which are in some cases enclosed by
different parts of the floral envelope (e.g., Ceraria, Portula-
caria, Basellaceae), and berries (Talinella). Furthermore, in
Halophytum the one-seeded nutlets are deeply sunken into the
inflorescence axis and form a slightly woody infructescence
(Bittrich, 1993b). Most taxa have two to three carpels, though,
one carpel is typical for Ceraria and Portulacaria, five to
eight and up to a dozen for Cactaceae, Lewisia, and Portulaca.
Berry-like (baccate) fruits are considered diagnostic for
Talinell a, and are unique in the suborder Portulacineae. Ap-
plequist (2005: 50) reports, however, that thin-walled dry cap-
sules apically dehiscing by shallow valves have been found in
a single specimen, and argues that capsular fruits might be
more common in the genus than reported so far. Moreover,
young ovaries have in the past been described as having two or
more distinct locules (Baillon, 1886b), which would be unique
within the suborder. Anatomical studies are needed to cor-
roborate or falsify this undocumented observation.
Seed characters have not been explored for many repre-
sentatives of Por tulacineae. However, various observations on
testa architecture, embryo shape, and amounts of perisperm
indicate a potential as informative characters for larger taxa.
Furthermore, various types of fleshy or spongy appendages
(e.g., aril, strophiola) are reported and should be investigated in
more detail. The embryo is only slightly curved in Anacamp-
seroteae, but reported to be st rongly curve d or an nular i n most
other families of Portulacineae. Perisperm is generally copi-
ous, but scanty in Basellaceae (Sperling & Bittrich, 1993).
Conspicuous axillary outgrowths are characteristic for
Cactaceae, Didiereoideae (= Didiereaceae sensu stricto [s.str.];
Applequist & Wallace, 2003), Anacamperos, and Portulaca.
In the case of Cactaceae, the axillary outgrowth is termed an
areole, and it conforms to a contracted short shoot (brachy-
blast) of which the leaves have been transformed into spines,
and which usually also produces abundant trichomes. The
architecture of the axillary outgrowths of Didiereoideae
is similar. Their brachyblasts differ from cactus areoles in
having a small and definite number of spines, and no as-
sociated trichomes (Rauh, 1956). The axillary outgrowths of
Anacampseros and Portulaca have been regarded as modi-
fied stipules (de Candolle, 1827: 186; Pax, 1889; Schönland,
1903). Chorinsky (1931) has shown, however, that such a
derivation is unlikely, and since stipules are absent from all
Portulacineae (Geesink, 1969), this interpretation can be dis-
missed. Chorinsky (1931) found that the axillary outgrowths
of Anacampseros and the hairs and bristles of cacti are basi-
cally similar in nature, while Gerbaulet (1992: 489) stresses
that the outgrowths of Anacampseros and Portulaca are
homologous, but without reaching a conclusion about their
possible anatomical derivation. Geesink (1969), based on the
data presented by Chorinsky (1931), concluded that the axil-
lary hairs and bristles of Portulaca are clearly derived from
the axillary meristem of the foliage leaves. Recently, Ogburn
(2007) identified proleptic leaves in the axils of Talinum s.str.
232
TAXON 59 (1) • February 2010: 227–240Nyffeler & Eggli •
New familial classification of Portulacineae
and noted that they are products of the axillary buds. It is
therefore tempting to suggest that the axillary outgrowths
of Anacampseros, Portulaca, and probably also Talinum, are
the only remnants of a highly condensed axillary short shoot,
and are thus homologous to the areoles of cacti, and hence,
represent a potential synapomorphy for a large subclade of
Portulacineae that also receives high statistical support from
molecular phylogenetic analyses. Similar areolar structures
are also present in Didiereaceae, indicating a possible close
relationship with the ACPT clade (Nyffeler, 2007).
Revised family classification. —
The present phylo-
genetic analysis of suborder Portulacineae identifies eight
major lineages that are with the exception of the two clades
DIDI and MONT (Fig. 1), well supported. However, only the
clades BASE, CACT, and HALO correspond to traditionally
circumscribed families, viz., Basellaceae, Cactaceae, and
Halophytaceae, respectively. Members of Portulacaceae in
a traditional sense are found to be part of the clades ANAC,
DIDI (incl. Didiereaceae), MONT (incl. Hectorellaceae),
PORT, and TALI. These findings, in particular in combina-
tion with mor phological con siderations (see below), lead us to
propose the following families in the suborder Portulacineae:
Anacampserotaceae, fam. nov. (ANAC; Fig. 1), Basellaceae
(BASE), Cactaceae (CACT), Didiereaceae (DIDI), Halophy-
taceae (HALO), Montiaceae (MONT), Portulacaceae (PORT),
and Talinaceae (TALI).
Nomenclatural note on the name Portulacineae. —
The suborder name Portulacineae was published by Engler
(1898) in the second edition of his Syllabus der Pflanzenfami-
lien. It is predated by the suborder name Cactineae, published
by Bessey (1895) in the eighth volume of Johnson’s Universal
Cyclopedia. Since priority does not apply at ranks above fam-
ily based on ICBN Art. 11.10 (McNeill & al., 2006), we disre-
gard Recommendation 16B of the ICBN (McNeill & al., 2006),
and continue to use the more familiar name Portulacineae (i.e.,
Nowicke, 1996; Carlquist, 1997; Nyffeler, 2007; Nyffeler &
al., 2008; Ogburn & Edwards, 2009), which also more aptly
circumscribes the taxon that includes all major lineages previ-
ously referred to traditional Portulacaceae.
SYNOPSIS OF FAMILIES OF PORTULACINEAE
Here, we provide a synopsis of the eight monophyletic
families to be recognized in the suborder Portulacineae. The
artificial key to the families is largely based on fruit and habit
characteristics. Further comparative studies that include all
relevant taxa will certainly make additional distinctive char-
acteristics available.
Key to the families of Portulacineae
1. Fruits dry capsules, utricles, or nutlets . . . . . . . . . . . . . . 2
1. Fruits fleshy berries (rarely dry and irregularly dehiscent
at maturity). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2. Fruits aggregated into a dry infructescence; flowers uni-
sexual, wind-pollinated . . . . . . . . . . . . . . Halophytaceae
2. Fruits discrete, not aggregated into an infructescence;
flowers animal-pollinated . . . . . . . . . . . . . . . . . . . . . . . .3
3. Fruits indehiscent (i.e., utricles or nutlets). . . . . . . . . . . .4
3. Fruits dehiscent (i.e., capsules) . . . . . . . . . . . . . . . . . . . . 6
4. Shrubs to trees, usually stem-succulent, often with
spines . . . . . . . . . . . . . . . . . . . . . . . . . . Didiereaceae p.p.
4. Plants herbaceous to suffruticose, spineless . . . . . . . . . .5
5. Stems well-developed, usually herbaceous and semi-suc-
culent, trailing to scandent and vine-like; inflorescences
dichasia, spikes, racemes, or panicles; fruits enveloped in
dry to fleshy perianth remains. . . . . . . . . . . . Basellaceae
5. Stems contracted or well-developed, hardly succulent but
stiff; inflorescences cymose, usually condensed; fruits
enveloped in dry perianth remains . . . . Montiaceae p.p.
6. Fruit dehiscence circumscissile, top portion shed intact as
a lid (operculum) . . . . . . . . . . . . . . . Portulacaceae s.str.
6. Fruit dehiscence variable but valvate, without true oper-
culum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .7
7. Exocarp and endocarp not separating . . . . . . . . . . . . . . .8
7. Exocarp and endocarp separating . . . . . . . . . . . . . . . . . 10
8. Fruit dehiscence valvate starting at the top . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Montiaceae p.p.
8. Fruit dehiscence circumscissile at the base and splitting
upwards into valves in the upper part . . . . . . . . . . . . . . . 9
9. Sparsely branched shrubs . . Didiereaceae (Calyptrotheca)
9. Herbs with a basal sessile rosette of succulent leaves. . . .
. . . . . . . . . . . . . . . . . . . . . . . . . . . .Montiaceae (Lewisia)
10. Fruit dehiscence basal; seeds usually black and glossy, with
a strophiola; embryo curved . . . . . Talinace ae (Talinum)
10. Fruit dehiscence apical; seeds usually pale, without strophi-
ola; embryo rather straight. . . . . . . Anacampserotaceae
11. Sarmentose lianoid shrubs; spines absent; f lowers small,
with 2–5 petaloids . . . . . . . . . . . . .Talinaceae (Talinella)
11. Usually spiny stem-succulents with mostly very reduced
leaves; flowers usually showy, with 5 to many petaloids
in a graded series . . . . . . . . . . . . . . . . . . . . . . . Cactaceae
Anacampserotaceae Eggli & Nyffeler, fam. nov. – Type:
Anacampseros L., Opera Var.: 232. 1758.
Herbae vel suffrutices succulentae perennes axillis folio-
rum pilis axillaribus vel squamis albis foliis obtegentibus in-
structis; flores tricarpellati; fructi capsulae elaboratae partes
exocarpi et endocarpi separandae, partes exocarpi caducis;
seminis pelliculis siccis pallidis instructae.
Small shrubs to thick-stemmed perennial herbs, muci-
laginous (except Grahamia), sometimes with a basal f leshy
caudex or tuberous main root; leaves spiral, succulent to very
succulent, terete to globose, rarely flattened, glabrous or to-
mentose; axils with hairs, bristles, or a pergamentaceous scale
(Anacampseros sect. Avonia (Fenzl) Gerbaulet); inflorescence
lateral or terminal few-flowered thyrsoids, sometimes with
contracted internodes, sometimes with scorpioid partial inflo-
rescences; flowers small to medium-sized, bisexual, usually
showy; sepaloids 2, fleshy, persistent and becoming dry in
fruit; petaloids 5; stamens 5–25; ovary superior, of 3 united
carpels; calyptra formed by the perianth remains and stamens
persistent at fruiting stage (Grahamia, Talinops is) or deciduous
233
Nyffeler & Eggli •
New familial classification of Portulacineae
TAXON 59 (1) • February 2010: 227–240
as an e ntity (Anacampseros); fruits loculicidally dehiscent cap-
sules with the caducous exocarp separating from the endocarp
(exce pt Grahamia, Gerbaulet, 1992: 506; Hershkovitz, 1993),
and the endocarp valves forming a small basket; seeds usu-
ally somewhat angular and voluminous, usually pale-colored
to white, without strophiola or elaiosome, testa two-layered,
the outer testa layer usually partially or almost completely
separating from the inner layer of the seed; embryo parallel to
the perisperm and rather straight (Franz, 1908; Kowal, 1961).
Distribution. – Southern and eastern Africa, Australia,
Argentina, Bolivia, Mexico, United States.
Genera and number of species. – Anacampseros L. (ca.
34 species; incl. Avonia (Fenzl) G.D. Rowley, incl. Talinaria
Brandegee, incl. Xenia Gerbaulet); Grahamia Hook. (1 spe-
cies); Talinopsis A. Gray (1 species).
Important taxonomic literature. – Gerbaulet (1992 – mono-
graph Anacampseros), Rowley (1994, 1995 – illustrated syn-
opses).
Discussion. – Anacampserotaceae are easily recognized
by the combination of elaborate fruits and pale-colored seeds
with the outer testa layer becoming separate from the inner
layer. The more derived representatives are found in Austra-
lia and arid southern and eastern Africa, while the species
of the basal grade occur in North and South America. The
migration between the continents was accompanied by the
evolution of specialized diminutive leaf-succulent life forms
(Nyffeler, 2007).
We propose, on the basis of our molecular phylogenetic
analyses and previous morphological investigations (Nyffeler,
2007), that the genus Anacampseros is recircumscribed to in-
clude all dwarf herbaceous species with rosulate leaf arrange-
ment, including Talinaria c oahuile n sis (S. Watson) P. Wilson
and Xenia vulcanensis (Añón) G erbau let. We retain the mono-
typic genera Grahamia and Talin opsis for the two species
from North and South America that form woody subshrubs
with distinct internodes (Nyffeler, 2007). The new combina-
tion required for the former Talin a r ia species is provided in
Appendix 2.
Basellaceae Raf., Fl. Tellur. 3: 44. 1837 (nom. cons.) – Type:
Basella L.
Vines or trailing herbs, usually glabrous, slightly to dis-
tinctly fleshy, mucilaginous, sometimes with tuberous roots;
leaves alternate to subopposite at the stem base; inf lorescences
axillary or terminal spikes, racemes, panicles, or dichasia;
flowers bisexual (functionally unisexual in Anredera vesicaria
(L am.) C.F. Gae rt.), r ather small and in conspic uous, somet imes
cleist ogamous; sepaloids t wo, f ree or par tly united , sometimes
hardly different from the petaloids; petaloids (4–)5(–13), con-
nate only at the ba se to more tha n half their leng th, somet imes
becoming black in fruit; stamens (4–)5(–9), basally connate
and adnate to the petaloids; ovary superior, consisting of three
united carpels, with a single basal ovule; fruits thin-walled
nutlets surrounded by the dry or fleshy perianth remains.
Distribution. – Tropics and subtropics of the New World,
few species in Africa and Madagascar, one species pantropical
due to cultivation.
Genera and number of species. – Anredera Juss. (ca. 12
species; incl. Boussingaultia Kunth); Basella L. (5 species);
Tour n onia Moq. (1 species); Ullucus Caldas (1 species).
Important taxonomic literature. – Eriksson (2007 – synop-
sis), Sperling (1987 – family monograph), Sperling & Bittrich
(1993 – synopsis).
Discussion. – Basellaceae are well-characterized by the
combination of a subsucculent to herbaceous, trailing to scan-
dent, vine-like growth form, often spicate inflorescences with
small, pale-colored flowers, and nutlets or drupes with one
seed enclosed by the perianth remains.
Cactaceae Juss., G en. Pl.: 310. 1789 (nom. cons.) – Type: Cac-
tus L. (nom. rejic. Mammillaria Haw., nom. cons.).
Perennial trees to shrubs, or dwarfs, usually stem succulent
and mucilaginous; roots fibrous, rarely tuberous; foliage leaves
usually absent (but present as minute microscopical vestiges;
Mauseth, 2007) or present, if present either flat and weakly to
distinctly fleshy (Pereskia), or terete, then either persistent for
a vegetation period (Maihuenia) or early caducous and only
present on young growth (Opuntioideae); axils developed into
a spiniferous (rarely spineless) areole, usually with some wool
or felt; flowers solitary (rarely several together or in succession)
from the areoles, small to very large, usually showy, usually
bisexual, usually actinomorphic, short- to long-lived, consist-
ing of a pericarpel with some to many spiniferous or spine-
less areoles, a perianth tube of varying length and with few
to many spiniferous or spineless areoles, and a usually graded
series of perianth elements varying from scales to sepaloids to
petaloids; stamens usually numerous, inserted in one or two
distinct series, or over the length of the perianth tube; ovary
inferior (superior or semi-inferior in Pereskia), included in the
pericarpel, composed of up to ten and more carpels, unilocular
with numerous ovules; fruit normally a fleshy to juicy berry,
rarely spontaneously dehiscing capsules (e.g., Copiapoa) or
slowly weathering over time (e.g., Tephrocactus); seeds vari-
able, often with diagnostic color and testa cells, sometimes with
a conspicuous spongy hilum-micropyle region (Cactoideae) or
completely enveloped into a hard bony aril (Opuntioideae); em-
bryo strongly curved around the perisperm to almost straight.
Distribution. – North and South America (southern Canada
to South-Central Argentina and South Chile) and 1 species
(Rhipsalis baccifera (J.S. Muell.) Stearn) also in Africa, Mada-
gascar, Sri Lanka and various islands of the Indian Ocean.
Number of genera and species. – 126 genera and ca. 1900
species (Anderson, 2001, 2005); 124 genera and 1438 species
(Hunt, 2006). Refer to these two sources for recent, slightly
contrasting lexicographic treatments of the family.
Important taxonomic literature. – Anderson (2001 – lexi-
con; 2005 – updated lexicon), Barthlott & Hunt (1993 – synop-
sis; 2000 – seed atlas), Butterworth & al. (2002 – phylogeny),
Buxbaum (1950 – morphology), Endler & Buxbaum (1973
– classification), Hunt (1967 – classification; 2006 – lexicon),
Leuenberger (1976 – palynology), Mauseth (2006 – morphol-
ogy, anatomy), Stuppy (2002 – synopsis Opuntioideae).
Discussion. – All members of Cactaceae are immediately
recognized on account of the usually spine-bearing areoles, the
234
TAXON 59 (1) • February 2010: 227–240Nyffeler & Eggli •
New familial classification of Portulacineae
wide-spread stem-succulence in conjunction with the lack of
foliage leaves, and the flower morphology involving a pericarp
formed by stem tissue with areoles, and a graded series of
perianth elements. Cacti are a very prominent group of stem
succulents, and it is often thought that this applies to the whole
family. However, succulence is only vaguely present in the
cladistically most basal and paraphyletic genus Pereskia Mill.
(Edwards & al., 2005), and has been lost again to a large de-
gree in highly specialized epiphytes such as Rhipsalis Gaertn.,
Disocactus Lindl. p.p. or Epiphyllum Haw. p.p.
A concise overview of the suprageneric classification of
Cactaceae is found in Anderson (2001, 2005). The traditional
division of the family into the three subfamilies Pereskioi-
deae, Opuntioideae and Cactoideae dates back to Schumann
(1897–1898). Maihuenioideae were recently erected for the
single genus Maihuenia Phil. (Fearn, 1996).
Didiereaceae Radlk. in Engler & Prantl, Nat. Pflanzenfam.
3(5): 462. 1896 – Type: Didierea Baill.
Slightly stem-succulent trees or shrubs, sometimes with
spines (D idier eoid eae), medu lla and cor tex with muc ilage ducts
and older stems with conspicuous tannin deposits (Didiere-
oideae and Portulacarioideae, unknown for Calyptrotheca);
leaves deciduous, leathery to succulent, flat to terete; axil-
lary buds of primary leaves developing into short spur-shoots
(Calyptrotheca), or as contracted short-shoots producing only
leaves or spines and leaves; inflorescences panicles to cymes,
often fasciculate, often many-flowered, or flowers in small
groups; flowers regular, unisexual (but rudiments of the oppo-
site sex present) on dioecious or gynodioecious (some Ceraria;
Swanepoel, 2007) plants, or bisexual (Calyptrotheca, Portula-
caria), minute to small (or large and showy in Calyptrotheca
and Alluaudiopsis marnieriana Rauh); sepaloids 2, persistent
and dry at fruiting time (except Calyptrotheca); petaloids 4
or 5; stamens (4–)5–12 (up to 60 in Calyptrotheca); ovary su-
perior, formed by (2–)3(–4) united carpels, ovule 1 or up to 6
(Calyptrotheca), basal; fruits 1-seeded indehiscent nutlets en-
closed by dry bracts (Didiereoideae), indehiscent dry or slightly
fleshy nutlets with membranous wings (Portulacarioideae), or
basally circumscissile, 6-valved, 1- (rarely 2-) seeded capsules
(Calyptrotheca); seeds with a small funicular strophiole or an
aril; embryo strongly curved around the perisperm.
Distribution. – Southern and eastern Africa, Madagascar.
Genera and number of species. – Alluaudia (Drake) Drake
(6 species), Alluaudiopsis Humbert & Choux (2 species), Ca-
lyptrotheca Gilg (2 species), Ceraria Pearson & Stephens (4–5
species), Decarya Choux (1 species), Didierea Baill. (2 spe-
cies), Portulacaria Jacq. (2 species).
Important taxonomic literature. – Applequist & Wallace
(2000 – molecular phylogeny; 2003 – expansion of family, inf-
rafamilial classification), Kubitzki (1993 – synopsis), Nowicke
(1996 – palynology), Rauh (1956 – morphology, anatomy; 1961
– growth form; 1963 – monograph), Rauh & Reznik (1961 –
chemistry), Rauh & Schölch (1965 – lower morphology, em-
bryology), Rowley (1992 – illustrated synopsis).
Discussion. – The systematic position of the Didiereaceae
was enigmatic for a long time, and it was variously associated
with Euphorbiaceae or placed in Sapi ndale s (e.g., Hutchinson,
1969). Its placement in core Caryophyllales was first suggested
by Radlkofer (1896), and later confirmed by Rauh & Reznik
(1961) based on the presence of betalains. Rauh (1961) and
Rauh & Reznik (1961) stress the morphological and anatomi-
cal similarities (i.e., long and short shoot organization, short
shoots as areoles, presence of oxalate druses and conspicuous
mucilage idioblasts in the primary cortex) with Cactaceae.
Hence, some authors (e.g., Rowley, 1992) refer to them as “cacti
of the Old World”. Palynologically, the family is readily recog-
nizable due to the 5–7-zonocolpate pollen with a finely spinate
aperture, which is unique for the whole order (Nowicke, 1996).
As traditionally circumscribed, the family consisted only
of the four Madagascan genera now included in subfamily
Didiereoideae. Molecular studies by Applequist & Wallace
(2001, 2003) and Nyffeler (2007) have shown that the two
genera Ceraria and Portulacaria, traditionally placed in Por-
tulacaceae s.l., are closely related to Didiereaceae s.str. and
should be placed here. In addition, the genus Calyptrotheca,
also formerly included in Portulacaceae, is part of this major
lineage of Portulacineae too. These additions, and in particular
Calyptrotheca, make Didiereaceae a rather heterogeneous as-
semblage as to gross vegetative morphology and floral char-
acters. The presence of tannin deposits (no reports available
for Calyptrotheca), otherwise only known for Talinaceae, is
potentially diagnostic for this family.
Halophytaceae A. Soriano in Bol. Soc. Argent. Bot. 23: 161.
1984 – Type: Halophytum Speg.
Annual, glabrous, leaf-succulent monoecious herbs; leaves
sessile, alternate, subterete with flattened upper face, without
axillary elements, but occasionally fascicled on short shoots;
flowers unisexual, small, usually with 2 (male flowers) or
2–4 (female flowers) bracts or bracteoles, female flowers 4–5
together in the axils of upper leaves, male flowers numerous,
densely aggregated in a condensed spike-like inflorescence
from the axils of the upper leaves; sepaloids none; petaloids
none in female flowers, 4 in male f lowers, membranous, whit-
ish; stamens 4; ovary superior, 3-carpellate, unilocular, with 1
ovule; fruit a thin-walled, indehiscent, 1-seeded nutlet partly
embedded into the axial tissue of the inflorescence, which as a
whole becomes hard and forms a fusiform syncarp consisting
of several nutlets; embryo annular.
Distribution. – Argentina.
Genera and number of species. – Halophytum Speg. (1 spe-
cies only: H. ameghinoi Speg.).
Important taxonomic literature. – Bittrich (1993b – syn-
opsis).
Discussion. – The phylogenetic relationships of this mono-
typic family remained unresolved for long, and it was either
associated with Aizoaceae (esp. Tet rago nia L.) or Chenopo-
diaceae. Ehrendorfer (1976: 102) placed it in Portulacaceae as
a “more isolated derivative”, and Bittrich (1993b) associated
the taxon with the ‘portulacoid’ group of families. In the pres-
ent study Halophytum is found to be part of a polytomy that
also includes Basel lacea e, Didierea ceae a nd the ACPT clade.
Cuboidal pollen and f lowers, as well as fruits, embedded into
235
Nyffeler & Eggli •
New familial classification of Portulacineae
TAXON 59 (1) • February 2010: 227–240
the floral axis are also found in Basellaceae (esp. Basella
excavata Scott-Eliot). Further investigations are needed to
clarify whether, indeed, Halophytum might share closer re-
lationships with Basellaceae. Anemophily is unique in the
whole suborder.
Montiaceae Raf. in Ann.n. Sci. Phys. 5: 349. 1820 – Type:
Montia L. (incl. Hectorellaceae Philipson & Skipw. in
Trans. Roy. Soc. New Zealand, Bot. 1: 31. 1961).
Perennial to annual herbs, frequently stemless, rarely
subshrubs, very rarely semiaquatic (Montia spp.), sometimes
with thickened roots and/or stems; leaves spiral, often in ro-
settes, commonly succulent, sometimes with clasping base,
usually glabrous, leaf axils naked; inflorescences terminal or
lateral, usually cymose, often scorpioid, or flower solitary and
axillary, flowers sessile to pedicellate, bisexual (bisexual or
unisexual in Hectorella), actinomorphic; sepaloids 2 or more
(up to 9 in Lewisia), often persistent and dry at fruiting stage;
petaloids 4 or 5 or up to 19 (Lewisia), usually free, sometimes
basally connate; stamens as many as petaloids, or numerous
(to 100); ovary superior, unilocular, consisting of 2–8 united
carpels; fruits 2 to 3-valved capsules with usually persistent
valves (deciduous in Phemeranthus), or basally circumscissile
(Lewisia, Lewisiopsis), or 1-seeded utricles (indehiscent or
tardily dehiscent in Lenzia, irregularly dehiscent or indehis-
cent in Philippiamra), or 1 to 2-seeded indehiscent capsules
disintegrating with time (Hectorella and Lyallia), sometimes
with a deciduous calyptra formed by the dry perianth remains
and stamens; seeds often minutely papillate, with or without
a strophiole or elaiosome, rarely with a thin-textured fleshy
or chartaceous aril (‘pellicle, Phemeranthus); embryo curved
around the perisperm.
Distribution. – North and South America (predominantly
western parts), northern Asia to northern Europe (circumbo-
real), Australia, New Zealand.
Genera and number of species. – Calandrinia Kunth
(14 species; incl. Monocosmia Fenzl); Calyptridium Torr. &
A. Gray (14 species; incl. Spraguea Torr.); Cistanthe Spach
(20 species); Claytonia L. (27 species; incl. Limnia Haw.);
Hectorella Hook. f. (1 species); Lenzia Phil. (1 species); Lewi-
sia L. Pursh (16 species; incl. Erocallis Rydb., Oreobroma
Howell); Lewisiopsis Govaerts (1 species); Lyallia Hook. f.
(1 species); Montia L. (12 species; incl. Claytoniella Jurtzev,
Crunocallis Rydb., Limnalsine Rydb., Maxia O. Nilsson, Mona
O. Nilsson, Montiastrum Rydb., Naiocrene Rydb., Neopaxia O.
Nilsson); Montiopsis Ku ntze (40 sp ecies; i ncl. Calandriniopsis
E. Franz); Parakeelya Hershk. (40 species); Phemeranthus Raf.
(ca. 30 spe cie s); Philippiamra Kuntze (8 species; incl. Diazia
Phil., Silvaea Phil.); Schreiteria Carolin (1 species).
Important taxonomic literature. – Appleqist & al. (2006 –
relationships of Hectorellaceae), Carolin (1993 – synopsis Por-
tulacaceae), Davidson (2000 – monograph Lewisia), Heenan
(1999 – monograph Montia p.p. [Neopaxia]), Hershkovitz
(1991 – phylogeny Portulacaceae), Hershkovitz (1993 – phy-
logeny Portulacaceae), Hershkovitz (2006 – phylogeny Por-
tulacaceae), Hershkovitz & Hogan (2003 – flora monograph
Lewisia), Kiger (2003 – flora monograph Phemeranthus),
Mathew (1989 – monograph Lewisia), Miller & Chambers
(2006 – monograph Claytonia), Philipson (1993 – Hectorel-
laceae synopsis), Wagstaff & Hennion (2007 – relationships
of Hectorellaceae).
Discussion. – The genera included here (except for Hec-
torella and Lyallia, formerly Hectorellaceae), were previously
placed in Portulacaceae (Carolin, 1993). The genera now
placed in the family Montiaceae were previously dispersed
among different tribes of Portulacaceae s.l. (see McNeill,
1974). However, the present circumscription was suggested
by Hershkovitz (1993, 2006) and Hershkovitz & Zimmer
(2000), as their informal “Western American Portulacaceae”
subg roup, on the basis of veget ative mor pholog y as well as m o-
lecular phylogeny analysis. The reestablishment of the genus
Phemeranthus exemplif ies very wel l the prev ious lack of solid
knowledge to resolve issues in Portulacaceae classification:
molecular phylogenetic investigations provide clear evidence
that the traditional concept of Talinum is poly phyletic a nd con-
sists of two distinct lineages; one (i.e., Phemeranthus) takes up
a cladistically basal position in Montiaceae and the other (i.e.,
Talinum s.str.) forms the sister-group to a clade consisting of
the families Anacampserotaceae, Cactaceae, and Portulaca-
ceae s.str. (Hershkovitz & Zimmer, 1997; Applequist & Wal-
lace, 2001). Differences in the morphology between the two
clades (i.e., Phemeranthus and Talinum s.str.) were discussed
by Ferguson (2001) and the necessary new combinations for
the Phemeranthus species of North America have been pub-
lished in the recent past (Hershkovitz & Zimmer, 1997; Kiger,
2001; Ocampo, 2002, 2003). However, so far no combination
for Phemeranthus is available for the disjunct species Tal in um
punae (R.E. Fr.) Carolin from higher altitudes in northern Ar-
gentina. Our phylogenetic analysis (Fig. 1) supports its close
relationships with North American Phemeranthus species,
and the necessary new combination is provided in Appendix
2. Morphologically, P. punae gr oups wit h othe r Phemeranthus
species on account of the terete leaves and the scapose, richly
branched inflorescence.
The family Montiaceae is by far the most diverse group
within Portulacineae. A predominance of herbaceous plants
with weakly expressed succulence is notable, and some gen-
era (e.g., Montiopsis, Lenzia, Hectorella, Lyallia) can hardly
be termed succulent at all. The placement of the monotypic
Schreiteria in Montiaceae is preliminary. This enigmatic ge-
nus has not been found again in the past 80 years or so, and
was not available for our analysis.
Portulacaceae Juss., Gen. Pl.: 312. 1789 – Type: Portulaca L.
Perennial to annual, usually succulent and mucilaginous
herbs with fibrous to tuberous roots, sometimes minute and
ephemeral, rarely somewhat suffrutescent (Portulaca suf-
frutescens Engelm.); stems herbaceous to slightly succulent,
rarely somewhat woody, or strongly succulent with flaking
bark (P. molokiniensis R.W. Hobdy); leaves alternate or rarely
opposite, flat to terete, succulent, sessile, glabrous or rarely
tomentose, axils appearing naked or commonly with few
to numerous short to long hairs or scales (P. somalica N.E.
Br., P. wightiana Wall.); inflorescence terminal, basically
236
TAXON 59 (1) • February 2010: 227–240Nyffeler & Eggli •
New familial classification of Portulacineae
cymose but much congested and head-like; flowers sessile
to pedicellate; sepaloids 2; petaloids (4–)5(–8), very shortly
connate, very delicate, usually showy in bright colors; sta-
mens usually numerous, or as few as 4; ovary semi-inferior,
composed of (4–)5–8 carpels, unilocular, ovules numerous;
fruits circumscissile capsules (pyxidia) with few to numer-
ous seeds, capsule lid (operculum) falling off intact together
with the dry perianth remains, stamens and style as a cap-like
structure (calyptra); seeds yellow, brown to black or grey,
often with iridescent gloss, testa cells usually forming an
intricate stellate pattern, sometimes with tubercles or short
to long projecting spines; hilum with a small to large spongy
aril; embryo curved.
Distribution. – Worldwide in the tropics and subtropics,
very rare in temperate climates.
Genera and number of species. – Portulaca L. (116 spe-
cies; incl. Lamia Endl., Lemia Vand., Merida Neck., Sedopsis
Exell & Mendonça).
Important taxonomic literature. – Carolin (1993 – synop-
sis), Geesink (1969 – monograph Indo-Pacific and Australia),
Gilbert & Phillips (2000 – monograph Africa and Arabia),
Legrand (1962 – monograph New World), Phillips (2002 –
monograph East Africa).
Discussion. – According to our results, Portulacaceae
has to be restricted to the single genus Portulaca, which is in
sharp contrast to the traditional circumscription of the fam-
ily (e.g., Carolin, 1993; Eggli & Ford-Werntz in Eggli, 2002).
The isolated position of Portulaca within the Portulacaceae
s.l. was already recognized by Pax & Hoffmann (1934), who
assigned it to the monogeneric subtribe Portulacinae, as well
as by McNeill (1974), who recognized the monogeneric tribe
Portulaceae. The contracted, head-like inf lorescences and
the operculate capsules (pyxidia) are absolutely diagnostic.
Such capsules are not known for other Portulacineae, but are
found elsewhere in Caryophyllales (e.g., Aizoaceae subfam.
Sesuvioideae, several genera of Amaranthaceae [Townsend,
1993]). Portulacaceae are also anatomically unique, as the
leaves show K ranz anatomy a ssociated with C 4 photos ynt hesis
(Nyananyo, 1988).
The genus Portulaca is usually divided into two subgenera:
P. subg. Portulaca (leaves alternate or rarely opposite, axil-
lary hairs present or seemingly absent, inflorescence capi-
tate or flowers solitary; distribution world-wide) and P. subg.
Portulacella (F. Muell.) Legrand (leaves opposite, axillary
hairs absent, inflorescence somewhat lax cymes; Australia,
Africa). Currently, molecular phylogenetic studies are in prog-
ress to evaluat e this classification based on morphologica l data
(Ocampo, pers. comm.).
Talinaceae Doweld, Tent. Syst. Pl. Vasc. (Tracheophyta): 42
[xlii]. 2001 – Type: Ta lin um Adanson.
Dwarf shrubs with often tuberous roots or rootstock; leaves
alternate, flat and slightly succulent, mucilaginous, entire,
glabrous or tomentose, axils appearing naked but usually
with a rudimentary axillary short shoot; inf lorescence ter-
minal and basically paniculate, or flowers solitary from leaf
axils; flowers small to medium-sized and showy, bisexual,
actinomor phic; sepaloids 2, deciduou s or persistent at f ruiting
time; petaloids usually 5, sometimes 2–4 and not clearly sepa-
rated from the sepaloids (Taline lla, Amphipetalum); stamens
15–35; ovary superior, unilocular, composed of 3(–5) carpels;
fruits many-seeded loculicidal capsules, or mucilaginous ber-
ries (Talinell a), capsules covered by the dry remains of peri-
anth, stamens and style which are shed in their entirety as a
calyptra, capsules dehiscent from the tip and/or base and the
valves deciduous, or the caducous exocarp separating from
the persistent endocarp; seeds usually black and glossy, with
a strophiola; embryo curved.
Distribution. – America, Africa, Madagascar, Tal inum
paniculatum and T. triang ulare pantropical weeds.
Genera and number of species. – Amphipetalum Baciga-
lupo (1 species), Talin ella (12 species; incl. Sabouraea Lean-
dri), Talin u m (ca. 15 species; excl. Phemeranthus Raf. which
is now included in Montiaceae).
Important taxonomic literature. – Applequist (2005 –
monograph Talin ella), Eggli (1997 – monograph Taline lla),
Tölken (1969 – monograph Talinum South Africa).
Discussion. – The small family Talinaceae is rather het-
erogeneous. The genus Talinella (endemic to Madagascar)
is easily recognized on morphological grounds (sarmentose
lianoid shrubs, inconspicuous and often numerous flowers
in congested inflorescences, berry-like fruits unique for the
whole suborder). Molecular phylogenetic studies repeatedly
placed it in a clade together with Talinum s.str. (Hershkov-
itz & Zimmer, 1997; Applequist & Wallace, 2001; Nyffeler,
2007), which is certainly unexpected in view of the previous
uncertainties regarding its relationships (Pax & Hoffmann,
1934; Nyananyo, 1986, Carolin, 1993). Interestingly enough,
Talinell a was closely associated with Didiereaceae s.l. by
Hershkovitz (1993: 349). Pending further research towards a
complete species phylogeny of Talinum s.l., we refrain from
formally transferring Taline lla species to Talinum here.
The axils of Talinum are usually described as naked, but
Ogburn (2007) found scale-like and often paired prophylls.
Talinum thus appears to possess axillary contracted short-
shoots that are most likely homologous to the ‘areoles’ of the
Cactaceae and Didiereaceae, and the axillary hairs and scales
of Anacampserotaceae and Portulacaceae s.str.
The monotypic genus Amphipetalum from Paraguay and
Bolivia was not available for study. It is placed here based on
general habit and inflorescence morphology.
Excluded genera.
Dendroportulaca and Pleuropeta-
lum are excluded from Portulacaceae s.l. These genera clearly
belong to the family Amaranthaceae, as confirmed by recent
investigations (Applequist & Pratt, 2005; Müller & Borsch,
2005).
Relationships of Portulacineae. —
The closest rela-
tionships of Portulacaceae s.l. has been debated in the past.
Fenzl (1836) claimed that Aizoaceae (tribes Ficoideae and
Mesembryanthemeae) and Caryophyllaceae (tribes Alsineae
and Paronychieae) might represent the most closely related
families. This view was shared by Pax (1889), who pointed out
the half-inferior ovary of Portulaca and the elevated number
237
Nyffeler & Eggli •
New familial classification of Portulacineae
TAXON 59 (1) • February 2010: 227–240
of sepals and petals of Lewisia as chara cters that are ind icative
of a close relationship to the family Aizoaceae. Other authors
stressed the similarity between some members of Portulaca-
ceae (i.e., Anacampseros and Portulaca) and Cactaceae on
the basis of the presence of multiseriate hair-like structures
(Chorinsky, 1931; see discussion above). Finally, Hutchinson
(1969) suggested a close affinity to Primulaceae.
The most broadly sampled molecular phylogeney of Caryo-
phyllales identifies various members of Molluginaceae (e.g.,
Adenogramma Rchb., Glinus L., and Suessenguthiella
Fried-
rich
) as the closest relatives of a generally well-supported Por-
tulacineae (Cuénoud & al., 2002). However, sequence data
available so far does not provide adequate information to settle
this question.
CONCLUDING REMARKS AND OUTLOOK
Resolving the family classification of traditional Portulaca-
ceae in the light of new findings from phylogenetic analyses
has been made public as a test case for the sensibility of tradi-
tional taxonomic practice to contrast the current obsession to
create monophyletic taxa (Brummitt, 2002, 2006). Molecular
phylogenetic analyses as well as insights derived from com-
parative morphological data of fruit characters clearly indi-
cate that the traditional classification of Portulacineae into six
families (i.e., Basellaceae, Cactaceae, Didiereaceae, Halophy-
taceae, Hectorellaceae, and Portulacaceae; Kubitzki & al.,
1993) does not reflect the phylogenetic relationships among
the members of this suborder, and makes inconsistent use of
morphological characteristics to circumscribe these taxa. Hid-
ing behind established tradition of Portulacaceae classification
with the notion that “there are no characters by which anyone
has ever thought to divide it [Portulacaceae] into two families”
(Brummitt, 2002: 36) is not helpful to plant systematics in the
long term. Discordance between molecular phylogenies and
traditional classification practice should rather be seen as a
challenge to further investigate morphological characteristics
potentially useful for circumscribing and identifying taxa de-
rived from inferred phylogenetic relationships.
We argue that we are in a much better position to work
towards overcoming this obsolete traditional family classifi-
cation and to replace it with one that better reflects phyloge-
netic hypotheses as well as provides well-supported overall
taxon circumscriptions. In line with arguments by Albach
(2008) we maintain that taxonomic stability will ultimately
only be reached by making taxon delimitation congruent
with well supported monophyletic groups of extant species.
For the suborder Portulacineae we suggest that eight families
(i.e., Anacampserotaceae, Basellaceae, Cactaceae, Didierea-
ceae, Halophytaceae, Montiaceae, Portulacaceae, Talinaceae)
should be recognized as outlined in our synoptical treatment.
This revised classification forms the framework for further
phylogenetic analyses on the basis of molecular markers from
the nuclear genome as well as detailed comparative struc-
tural investigations (Nyffeler & al., 2008; Ogburn & Edwards,
2009).
ACKNOWLEDGEMENTS
We thank the Sukkulenten-Sammlung of Grün Stadt Zürich for
supplying plant material. Three anonymous reviewers provided ad-
vice and helped to improve the current publication. The G. und A.
Claraz-Schenkung, University of Zürich, provided financial support
for this project.
LITERATURE CITED
Adanson, M. 17 63. Familles des plantes, vol. 2. Paris: Vincent.
Albach, D.C. 2008. Further arguments for the rejection of paraphyletic
taxa: Ve ron ica subgen. Pseudolysimachium (Plantaginaceae).
Taxo n 57: 1–6.
Anderson, E.F. 2001. The cactus family. Portland: Timber Press.
Anderson, E.F. 2005. Das grosse Kakteen-Lexikon. Stuttgart: Ulmer.
[Corrected and supplemented German translation of Anderson,
2001.]
Applequist, W.L. 2005. A revision of the Malagasy endemic Talinel la
(Portulacaceae). Adansonia, ser. 3, 27: 47–80.
Applequist, W.L. & Pratt, D.B. 2005. The Malagasy endemic Den-
droportulaca (Portulacaceae) is referable to Deeringia (Ama-
ranthaceae): Molecular and morphological evidence. Taxon 54:
681–687.
Applequist, W.L., Wagner, W.L., Zimmer, E.A. & Nepokroeff,
M. 2006. Molecular evidence resolving the systematic position of
Hectorella (Portulacaceae). Syst. Bot. 31: 310 319.
Applequist, W.L. & Wallace, R.S. 2000. Phylogeny of the Madagas-
can endemic family Didiereaceae. Pl. Syst. Evol. 221: 157–166.
Applequist, W.L. & Wallace, R.S. 2001. Phylogeny of the portulaca-
ceous cohort based on ndhF sequence data. Syst. Bot. 26: 40 6– 419.
Applequist, W.L. & Wallace, R.S. 2003. Expanded circumscription
of Didiereaceae and its division into three subfamilies. Adansonia
ser. 3, 25: 13–16.
Baillon, H. 1886a . Histoires des Plantes: Monographie des Aristo-
lochia cées, Cactacées, Mésembryanthémacées et Portulacacées.
Paris: Librairie Hachette.
Baillon, H. 1886b. Un nouveau type réduit de Portulacacées. Bull.
Mens. Soc. Linn. Paris 1: 569.
Barthlott, W. & Hunt, D.R. 1993. Cactaceae. Pp. 161–197 in: Ku-
bitzki, K., Rohwer, J.G. & Bittrich, V. (eds.), The families and gen-
era of vascular plants, vol. 2, Flowering plants: Dicotyledons;
Magnoliid, hamamelid and caryophyllid families. Berlin: Springer.
Barthlott, W. & Hunt, D.R. 2000. Seed-diversity in the Cactaceae
subfamily Cactoideae. Succ. Pl. Res. 5: 1–176.
Bartling, F.G. 1830. Ordines naturales plantarum eorumque cha-
racteres et affinitates adjecta generum enumeratione. Göttingen:
Dietrich.
Bentham, G. 1862a. Notes on the Caryophylleae, Portulaceae, and
some allied 27 orders. J. Proc. Linn. Soc., Bot. 6: 55–77.
Bentham, G. 1862b. Ordo XXIII. Portulaceae. Pp. 155 –159 in:
Bentham, G. & Hooker, J.D. (eds.), Genera plantarum, vol. 1.
London: L. Reeve.
Bessey, C. E. 1895. Vegetable kingdom. Pp. 453–464 in: Adams, C.K.
(ed .): Johnson’s Universal Cyclopedia, vol. 8. New York: D. Apple-
ton & A.J. Johnson.
Bittrich, V. 1993a. Introduction to Centrospermae. Pp. 13–19 in:
Kubitzki, K., Rohwer, J.G. & Bittrich, V. (eds.), The families and
genera of vasc ular plan ts, vol. 2, Flowering plants: Dicotyledons;
Magnoliid, hamamelid and caryophyllid families. Berlin: Springer.
Bittrich, V. 1993b. Halophytaceae. Pp. 320–321 in: Kubitzki, K., Ro-
hwer, J.G. & Bittrich, V. (eds.), The families and genera of vas-
cular plants, vol. 2, Flowering plants: Dicotyledons; Magnoliid,
hamamelid and caryophyllid families. Berlin: Springer.
238
TAXON 59 (1) • February 2010: 227–240Nyffeler & Eggli •
New familial classification of Portulacineae
Brummitt, R.K. 2002. How to chop up a tree. Ta xon 51: 31–41.
Brummitt, R.K. 2006. Am I a bony fish? Ta xon 55: 268–269.
Butterworth, C.A., Cota-Sanchez, J.H. & Wallace, R.S. 2002. Molec-
ular syst ematics of tribe Ca cteae (Cactaceae: Cactoideae): A phylog-
eny based on rpl16 intron sequence variation. Syst. Bot. 27: 257–270.
Buxbaum, F. 1950. Morphology of cacti. Pasadena: Abbey Garden Press.
Candolle, A.P. de. 1827. Revue de la famille des Portulacacées. Mém.
Soc. Hist. Nat. Paris 4: 174–194.
Candolle, A.P. de. 1828. Revue de la famille des Cactées. Mém. Mus.
Hist. Nat. 17: 1–119.
Carlquist, S. 1997. Wood anatomy of Portu lacaceae and Hectorellaceae:
Ecological, habital, and systematic implications. Aliso 16: 137–153.
Carolin, R.C. 1987. A review of the family Portulacaceae. Austral. J.
Bot. 35: 383–412.
Carolin, R.C. 1993. Portulacaceae. Pp. 544–555 in: Kubitzki, K.,
Rohwer, J.G. & Bittrich, V. (eds.), The families and genera of vas-
cular plants, vol. 2, Flowering plants: Dicotyledons; Magnoliid,
hamamelid and caryophyllid families. Berlin: Springer.
Chorinsky, F. 1931. Vergleichend-anatomische Untersuchungen der
Haargebilde bei Portulacaceen und Cactaceen. Österr. Bot. Z.
308–327.
Cronquist, A. 1981. An in teg rated sys tem of cl ass ifi cat ion of flow eri ng
plants. New York: Columbia Univ. Press.
Cronquist, A. & Thorne, T.F. 1994. Nomenclatural and taxonomic
history. Pp. 87–21, in: Behnke, H.D. & Mabry, T.J. (eds.), Caryo-
phyllales: Evolution and systematics. Berlin: Springer.
Cuénoud, P., Savolainen, V., Chatrou, L.W., Powell, M., Grayer,
R.J. & Chase, M.W. 2002. Molecular phylogenetics of Car yophyl-
lales based on nuclear 18S rDNA and plastid rbcL, atpB, and matK
DNA sequences. Amer. J. Bot. 89: 132–144.
Cullen, J. 1997. The identification of flowering plant families, 4th ed.
Cambridge: Cambridge Univ. Press.
Dahlgren, R.M.T. 1980. A revised system of classification of an-
giosperms. Bot. J. Linn. Soc. 80: 91–124.
Davidson, B.L. 2000. Lewisias. Portland: Timber Press.
Eckardt, T. 1976. Classical morphological features of centrospermous
families. Pl. Syst. Evol. 126: 5–25.
Edwards, E.J., Nyffeler, R. & Donoghue, M.J. 2005. Basal cactus
phylogeny: Implications of Pereskia (Cactaceae) paraphyly for
the transition to the cactus life form. Amer. J. Bot. 92: 1177–1188.
Eggli, U. 1997. A synopsis of the woody Portulacaceae in Madagascar.
Bull. Mus. Natl. Hist. Nat., B, Adansonia 19: 45–59.
Eggli, U. (ed.). 2002. Illustrated handbook of succulent plants. Dicoty-
ledons. Berlin: Springer.
Ehrendorfer, F. 1976. Closing remarks: Systematics and evolution of
centrospermous families. Pl. Syst. Evol. 126: 99–105.
Endler, J. & Buxbaum, F. 1973. Die Pflanzenfamilie der Kakteen:
Ein systematischer Wegweiser für Liebhaber und Erwerbszüchter
mit einer kompletten Liste der Gattungssynonyme, 3rd ed. Minden:
Philler.
Endlicher, S. 1840. Ordo CCVI. Portulaceae. Pp. 946 –955 in: Genera
plantarum secundum ordines naturales disposita. Wien: Beck.
Engler, A. 1898. Syllabus der Pf lanzenfamilien: Eine Uebersicht über
das gesamte Pflanzensystem mit Berücksichtigung der Medicinal-
und Nutzpflanzen zum Gebrauch bei Vorlesungen und Studien
über spezielle und medicinisch-pharmaceutische Botanik, 2nd ed.
Berlin: Gebrüder Borntraeger.
Engler, A. 1925. Reihe Opuntiales: Historische Entwicklung der An-
sichten über die systematische Stellung der Reihe. Pp. 592–594 in:
Engler, A. & Prantl, K. (eds.), Die natürlichen Pflanzenfamilien,
2. Aufl., vol. 21. Leipzig: Engelmann.
Erbar, C. & Leins, P. 2006. Floral ontogeny and systematic position
of the Didiereaceae. Pl. Syst. Evol. 261: 265–285.
Eriksson, R. 2007. A synopsis of Basellaceae. Kew Bull. 26: 297–320.
Fear n, P.J. 1996. A review of the origins of the Cactus Family and
the search for a system of classification. […]. Wimborne, U.K.:
Published by the author.
Fenzl, E. 1836. Monographie der Mollugineen und der Steudelieen,
zweier Unterabtheilungen der Familie Portulaceen. Ann. Wiener
Mus. Naturgesch. 1: 337–384.
Fenzl, E. 1839. Monographie der Mollugineen. Zweiter Artikel. Ann.
Wiener Mus. Naturgesch. 2: 244–310.
Ferguson, D.J. 2001. Phemeranthus and Talinum (Portulacaceae) in
New Mexico. New Mexico Botanist 20: 1–7.
Franz, E. 1908. Beiträge zur Kenntnis der Portulacaceen und Ba-
sellaceen. Bot. Jahrb. Syst. 42 (Beibl. 97): 1–46.
Friedrich, H.C. 1956. Studien über die nat ürliche Verwandtschaft der
Plumbaginales und Centrospermae. Phyton (Horn) 6: 220–263.
Geesink, R. 1969. An account of the genus Portulaca in Indo-Australia
and the Pacific (Portulacaceae). Blumea 17: 275–301.
Geesink, R., Leeuwenberg, A.J.M., Ridsdale, C.E. & Veldkamp,
J.F. 1981. Thonner’s analytical key to the families of flowering
plants. The Hague: Martinus Nijhoff Publ.
Gene Codes Corporation. 2000. Sequencher, version 4.2. Ann Arbor:
Gene Codes Corporation. http://www.genecodes.com/.
Gerbaulet, M. 1992. Die Gattung Anacampseros L. (Portulacaceae).
I. Untersuchungen zur Systematik. Bot. Jahrb. Syst. 113: 477–564.
Gilbert, M.G. & Phillips, S.M. 2000. A review of the opposite-leaved
species of Portulaca in Africa and Arabia. Kew Bull. 55: 769–8 02.
Heenan, P.B. 1999. A taxonomic revision of Neopaxia O. Nilss. (Por-
tulacaceae) in New Zealand. New Zealand J. Bot. 37: 213–234.
Hershkovitz, M.A. 1991. Phylogenetic assessment and revised cir-
cumscription of Cistanthe Spach (Portulacaceae). Ann. Missouri
Bot. Gard. 78: 1009–1021.
Hershkovitz, M.A. 1993. Revised circumscriptions and subgeneric
taxonomies of Calandrinia and Montiopsis (Portulacaceae) with
notes on phylogeny of the Portulacaceous alliance. Ann. Missouri
Bot. Gard. 80: 333–365.
Hershkovitz, M.A. 2006. Ribosomal and chloroplast DNA evidence
for diversification of Western American Portulacaceae in the An-
dean region. Gayana, Bot. 63: 13–74.
Hershkovitz, M.A. & Hogan, S.B. 2003. Lewisia. Pp. 476–485 in:
FNA Editorial Committee (ed.), Flora of North America north of
Mexico, vol. 4. New York: Oxford Univ. Press.
Hershkovitz, M.A. & Zimmer, E.A. 1997. On the evolutionar y origins
of the cacti. Ta x on 46: 217–242.
Hershkovitz, M.A. & Zimmer, E.A. 2000. Ribosomal DNA evidence
and disjunctions of Wester n American Portulacaceae. Molec. Phy-
log. Evol. 15: 419–439.
Hilu, K.W., Borsch, T., Müller, K., Soltis, D.E., Soltis, P.S.,
Savolainen, V., Chase, M.W., Powell, M.P., Alice, L.A., Evans,
R., Sauquet, H., Neinhuis, C., Slotta, T.A.B., Rohwer, J.G.,
Campbell, C.S. & Chatrou, L.W. 2003. Angiosperm phylogeny
based on matK sequence information. Amer. J. Bot. 90: 1758–1776.
Hooker, J.D. 1846. Description of Pleuropetalum, a new genus of
Portulaceae, from the Galapagos Islands. London J. Bot. 5: 108–109
Huelsenbeck, J.P. & Ronquist, F. 2001. MRBAYES: Bayesian infer-
ence of phylogenetic trees. Bioinformatics 17: 754 –755.
Hunt, D.R. 1967. The genera of the Cactaceae. Pp. 427–467 in: Hutch-
inson, J. (ed.), The genera of flowering plants, vol. 2. Oxford: Ox-
ford Univ. Press.
Hunt, D.R. (ed.). 2006. The new cactus lexicon, 2 vols. Milborne Port:
DH Books.
Hutchinson, J. 1969. Evolution and phylogeny of flowering plants:
Dicotyledons; Facts and theory. London: Academic Press.
Hutchinson, J. 1973. The families of flowering plants, 3rd. ed. Oxford:
Clarendon Press.
Johnson, L.A. & Soltis, D.E. 1994. matK DNA-sequences and phy-
logenetic reconstruction in Saxifragaceae s.str. Syst. Bot. 19:
143–156.
Judd, W.S., Campbell, C.S., Kellogg, E.A., Stevens, P.F. & Dono-
ghue, M.J. 2007. Plant systematics: A phylogenetic approach, 2nd.
ed. Sunderland, Massachusetts: Sinauer.
Jussieu, A.L. de. 178 9. Genera plantarum. Paris: Herrisat & Barrois.
239
Nyffeler & Eggli •
New familial classification of Portulacineae
TAXON 59 (1) • February 2010: 227–240
Kearny, M. 2002. Fragmentary taxa, missing data, and ambiguity:
Mistaken assumptions and conclusions. Syst. Biol. 51: 369–381.
Kiger, R.W. 2001. New combinations in Phemeranthus Rafinesque
(Portulacaceae). Novon 11: 319–321.
Kiger, R.W. 2003. Phemeranthus. Pp. 488–495 in: Flora of North
America Editorial Committee (ed.), Flora of North America north
of Mexico, vol. 4. New York: Oxford Univ. Press.
Kowal, K. 1961. Studia nad morphologia i anatomia nasion Portu-
lacaceae Reichb. Monogr. Bot. 12: 1–48.
Kubitzki, K. 1993. Didiereaceae. Pp. 292–295 in: Kubitzki, K., Ro-
hwer, J.G. & Bittrich, V. (eds.), The families and genera of vas-
cular plants, vol. 2, Flowering plants: Dicotyledons; Magnoliid,
hamamelid and caryophyllid families. Springer Verlag, Berlin.
Kubitzki, K., Rohwer, J.G. & Bittrich, V. (eds.). 1993. The families
and genera of vascular plants, vol. 2, Flowering plants: Dicoty-
ledons; Magnoliid, hamamelid and caryophyllid families. Berlin:
Springer.
Legrand, C.D. 1949. Las especies del género Portulaca e n la Ar gen ti na .
Lilloa 17: 311–376.
Legrand, C.D. 1953. Demarcación del género Portulaca. Comun. Bot.
Mus. Hist. Nat. Montevideo 3: 1–16.
Legrand, C.D. 1962. Las especies americanas de Portulaca. Anales
Mus. Hist. Nat. Montevideo 7: 1–147.
Leuenberger, B.E. 1976. Die Pollenmorphologie der Cactaceae und
ihre Bedeutung für die Systematik. Dissertationes Botanicae 31.
Vad u z: C rame r.
Mathew, B. 1989. The genus Lewisia. Kew: Royal Botanic Gardens.
Mauseth, J.D. 2006. Structure–function relationships in highly modi-
fied shoots of Cactaceae. Ann. Bot. (Oxford) 98: 901–926.
Mauseth, J.D. 2007. Tiny but complex foliage leaves occur in many
“leafless” cacti (Cactaceae). Int. J. Pl. Sci. 168: 845 853.
McNeill, J. 1974. Synopsis of a revised classification of the Portu-
lacaceae. Taxo n 23: 725–728.
McNeill, J., Barrie, F.R., Burdet, H.M., Demoulin, V., Hawksworth,
D.L., Marhold, K., Nicolson, D.H., Prado, J., Silva, P. C., Skog,
J.E., Turland, N.J. & Wiersema, J.H. 2006. International code
of botanical nomenclature (Vienna Code). Regnum Vegetabile
146. R uggell: G ant ner.
Miller, J.M. & Chambers, K.L. 2006. Systematics of Claytonia (Por-
tulacaceae). Syst. Bot. Monogr. 78: 1–236.
Müller, K. & Borsch, T. 2005. Phylogenetics of Amaranthaceae based
on matK/trnK sequence data: Evidence from parsimony, likelihood,
and Bayesian analyses. Ann. Missouri Bot. Gard. 92: 66–102.
Nowicke, J.W. 1996. Pollen morphology, exine structure and the rela-
tionships of Basellaceae and Didiereaceae to Portulacaceae. Syst.
Bot. 21: 187–208.
Nyananyo, B.L. 1986. The taxonomic position of Talinel la (Portu-
lacaceae). Feddes Repert. 97: 771–773.
Nyananyo, B.L. 1988. Leaf anatomical studies in the Portulacaceae
(Centrospermae), with regards to photosynthetic pathways. Folia
Geobot. Phytotax. 23: 99–101.
Nyananyo, B.L. 1990. Tribal and generic relationship in the Portu-
lacaceae (Centrospermae). Feddes Repert. 101: 237–241.
Nyffeler, R. 2002. Phylogenetic relationships in the cactus family
(Cactaceae) based on evidence from trnK/matK and trnL-trnF
sequences. Amer. J. Bot. 89: 312–326.
Nyffeler, R. 2007. The closest relatives of cacti: Insights from phylo-
genetic analyses of chloroplast and mitochondrial sequences with
special emphasis on relationships in the tribe Anacampseroteae.
Amer. J. Bot. 94: 89–101.
Nyffeler, R., Eggli, U., Ogburn, M. & Edwards, E. 2008. Variations
on a theme: Repeated evolution of succulent life forms in the Por-
tulacineae (Caryophyllales). Haseltonia 14: 26–36.
Nylander, J.A.A. 2004. MrModeltest, version 2.2. Evolutionary Biol-
ogy Centre, Uppsala University: Program distributed by the author.
http://www.abc.se/~nylander/.
Ocampo, G. 2002. Transferencia de tres especies mexicanas de
Talinu m Adans. a Phemeranthus Raf. (Portulacaceae). Acta Bot.
Mex. 59: 75–80.
Ocampo, G. 2003. Una combinación nueva en Phemeranthus (Portu-
lacaceae). Acta Bot. Mex. 63: 55–57.
Ogburn, R.M. 2007. Anatomical variation in Cactaceae sensu lato.
Master Thesis, University of Missouri, St. Louis.
Ogburn, R.M. & Edwards, E.J. 2009. Anatomical variation in Cac-
taceae and relatives: Trait lability and evolutionary innovation.
Amer. J. Bot. 96: 391–408.
Olmstead, R.G., K im, K.J., Jansen, R.K. & Wagstaff, S.J. 2000. The
phylogeny of the Asteridae sensu lato based on chloroplast ndhF
gene sequences. Molec. Phylog. Evol. 16: 96–112.
Pax, F. 1889. Portulacaceae. Pp. 51–60 in: Engler, A. & Prantl, K. (eds.),
Die natürlichen Pflanzenfamilien, vol. 3(1b). Leipzig: Engelmann.
Pax, F. & Hoffmann, K. 1934. Portulacaceae. Pp. 234–262 in: Engler,
A. & Harms, H. (eds.), Die natürlichen Pf lanzenfamilien, ed. 2,
vol. 16c. Leipzig: Engelmann.
Payer, J.B. 1857. Trai dorganogénie comparée de la fleur. Paris:
Librairie de Victor Masson.
Philipson, W.R. 1993. Hectorellaceae. Pp. 331–334 in: Kubitzki, K.,
Rohwer, J.G. & Bittrich, V. (eds.), The families and genera of vas-
cular plants, vol. 2, Flowering Plants: Dicotyledons; Magnoliid,
hamamelid and caryophyllid families. Berlin: Springer.
Phillips, S.M. 2002. Portulacaceae. Pp. 1–40 in: Beentje, H.J. (ed.),
Flora of tropical East Africa. Rotterdam: Balkema.
Queiroz, A. de & Gatesy, J. 2006. The supermatrix approach to sys-
tematics. Trends Ecol. Evol. 22: 34–41.
Radlkofer, L. 1896. Sapindaceae (Nachträge). Pp. 460–462 in: Engler,
A. & Prantl, K. (eds.), Die natürlichen Pflanzenfamilien, vol. 3,
part 5. Leipzig: Engelmann.
Rauh, W. 1956. Morphologische, entwicklungsgeschichtliche, his-
togenetische und anatomische Untersuchungen an den Sprossen
der Didiereaceen. Akad. Wiss. Mainz, Abh. Math.-Naturwiss. Kl.
1956: 343–444.
Rauh, W. 1961. Weitere Untersuchungen an Didiereaceen. 1. Teil:
Beitrag z ur Kenntnis der Wuchsformen der Didiereaceen, unter be-
sonderer Berücksichtigung neuer Arten. Sitzungsber. Heidelberger
Akad. Wiss., Math.-Naturwiss. Kl. 1960/61: 185–300.
Rauh, W. 1963. Didieréacées–121e famille. In: Humbert, H. (ed.),
Flore de Madagascar et des Comores. Paris: Muséum National
d’Histoire Naturelle.
Rauh, W. & Reznik, H. 1961. Zur Frage der systematischen Stellung
der Didiereaceen. Bot. Jahrb. Syst. 81: 94–105.
Rauh, W. & Schölch, H.F. 1965. Weitere Untersuchungen an Didie-
reaceen. 2. Teil. Infloreszenz-, blütenmorphologische und em-
bryologische Untersuchungen mit Ausblick auf die systematische
Stellung der Didiereaceen. Sitzungsber. Heidelberger Akad. Wiss.,
Math.-Naturwiss. Kl. 1965: 221–434.
Rowley, G.D. 1992. Didiereaceae. ‘Cacti of the Old World’. Richmond,
U.K.: British Cactus and Succulent Society.
Rowley, G.D. 1994. Anacampseros and allied genera: A reassessment.
Bradleya 12: 105–112.
Rowley, G.D. 1995. Anacampseros, Avonia, Grahamia: A grower’s
guide. Hull Road: British Cactus and Succulent Society.
Schönland, S. 1903. Morphological and biological observations on
the genus Anacampseros L. (Rulingia Ehrh.). S. African Rep. Sci.
1: 295–301.
Schumann, K. 1897–1898. Gesamtbeschreibung der Kakteen (Mono-
graphia Cactacearum). Neudamm: Neumann.
Schumann, K. 1899. Die Verbreitung der Cactaceae im Verhältnis zu
ihrer systematischen Gliederung. Abh. Königl. Akad. Wiss. Berlin
2: 1–114.
Soltis, D.E., Solt is, P.S., Endress, P.K. & Chase, M.W. 2005. Phylogeny
and evolution of angiosperms. Sunderland, Massachuset ts: Sinauer.
Sperling, C.R. 1987. Systematics of the Basellaceae. Ph.D. thesis,
Harvard University, Department of Organismic and Evolutionary
Botany, Cambridge.
240
TAXON 59 (1) • February 2010: 227–240Nyffeler & Eggli •
New familial classification of Portulacineae
Sperling, C.R. & Bittrich, V. 1993. Basellaceae. Pp. 143–146 in:
Kubitzki, K., Rohwer, J.G. & Bittrich, V. (eds.), The families and
genera of vasc ular plan ts, vol. 2, Flowering plants: Dicotyledons;
Magnoliid, hamamelid and caryophyllid families. Berlin: Springer.
Stuppy, W. 2002. Seed characters and the generic classification of the
Opuntioideae (Cactaceae). Succ. Pl. Res. 6: 25–58.
Swanepoel, W. 2007. Ceraria kaokoensis, a new species from Namibia,
with notes on gynodioecy in the genus. Bothalia 37: 202–206.
Swofford, D.L. 2000. PAU P*: Ph yloge net ic ana lysis using pa rsimo ny
(*and other methods), version 4.0b10. CD-ROM. Sunderland, Mas-
sachusetts: Sinauer.
Tak htajan, A. 1997. Diversity and classification of f lowering plants.
New York: Columbia Univ. Press.
Tak htajan, A. 2009. Flowering plants, 2nd ed. Dordrecht: Springer
Science + Business Media.
Thorne, R.F. 1976. A phylogenetic classification of the Angiospermae.
Evol. Biol. 9: 35–106.
Thorne, R.F. 2000. The classif ication and geography of the flowering
plants: Dicotyledons of the class Angiospermae (subclasses Mag-
noliidae, Ranunculidae, Caryophyllidae, Dilleniidae, Rosidae,
Asteridae, and Lamiidae). Bot. Rev. 66: 441—647.
Tölken, H.R. 1969. The genus Talinum (Portulacaceae) in southern
Africa. Bothalia 10: 19–28.
Townsend, C.C. 1993. Amaranthaceae. Pp. 70–91 in: Kubitzki, K.,
Rohwer, J.G. & Bittrich, V. (eds.), The families and genera of vas-
cular plants, vol. 2, Flowering plants: Dicotyledons; Magnoliid,
hamamelid and caryophyllid families. Berlin: Springer.
Wagstaff, S.J. & Hennion, F. 2007. Evolution and biogeography of
Lyallia and Hectorella (Portulacaceae), geographically isolated
sisters from the Southern Hemisphere. Antarc. Sci. 19: 417–426.
Wiens, J.J. 2003. Missing data, incomplete taxa, and phylogenetic
accuracy. Syst. Biol. 52: 528–538.
Zomlefer, W.B. 1995. Guide to f lowering plant families. Chapel Hill:
Univ. of North Carolina.
Appendix 1.
Accessions of Portulacineae and outgroups used for the molecular phylogenetic analysis. A dash indicates that the sequence was not avail-
able. Voucher specimen information is given for previously unpublished sequences. For taxon names with an asterisk new combinations are provided in
Appendix 2 of this publication.
Tax on, voucher, GenBank accession number: matK (original publication), ndhF (original publication).
PORTULACINEAE: Alluaudia ascendens (Drake) Drake, –, AY042541 (Cuénoud & al., 2002); Alluaudia humbertii Choux, AF194832 (Applequist
& Wallace, 2001), –; Anacampseros albissima Marloth, –, DQ855856 (Nyffeler, 2007; as Avonia albissima); Anacampseros australiana J.M. Black, –,
DQ855855 (Nyffeler, 2007; as Grahamia australiana); *An ac am ps er os coahuilensi s (S. Watson) Eggli & Nyffeler, –, DQ855854 (Nyffeler, 2007; as Graha-
mia coahuilensis); Anacampseros kurtzii Bacigalupo, –, DQ855853 (Nyffeler, 2007; as Grahamia kurtzii); Anacampseros
ka
rasmontana Dinter, DQ855872
(Nyffeler, 2007), DQ855859 (Nyffeler, 2007); Anacampseros papyracea Fenzl, –, DQ855857 (Nyffeler, 2007; as Avonia papyracea); Anacampseros re-
curvata Schönland, –, DQ855858 (Nyffeler, 2007; as Avonia recurvata); Anacampseros retusa Poelln., DQ855873 (Nyffeler, 2007), DQ855860 ( Nyffeler,
2007); Anacampseros subnuda Poelln., DQ855874 (Nyffeler, 2007), DQ855861 (Nyffeler, 2007); Anacampseros telephiastrum DC., DQ855875 (Nyffeler,
2007), DQ855862 (Nyffeler, 2007); Anacampseros vulcanensis Añón, –, DQ855852 (Nyffeler, 2007; as Grahamia vulcanensis); Anredera cordifolia (Ten.)
Steenis, –, AY042547 (Cuénoud & al., 2002); Austrocylindropuntia vestita (Salm-Dyck) Backeb., DQ855878 (Nyffeler, 2007), AY015278 (Nyffeler, 2002);
Basella alba L., AF194834 (Applequist & Wallace, 2001), AY042553 (Cuénoud & al., 2002); Blossfeldia liliputana Werdermann, –, AY015284 (Nyffeler,
2002); Calymmanthium substerile F. Ritter, –, AY015291 (Nyffeler, 2002); Calyptridium umbellatum (Torr.) Greene, AF194840 (Applequist & Wallace,
2001), –; Calyptrotheca somalensis Gilg, AF194839 (Applequist & Wallace, 2001), AY042563 (Cuénoud & al., 2002); Ceraria fruticulosa H. Pearson &
Stephens, AF194841 (Applequist & Wallace, 2001), AY875371 (Edwards & al., 2005); Cistanthe grandiflora (Lindl.) Hershk., AF194842 (Applequist &
Wallace, 2001), AY042568 (Cuénoud & al., 2002); Copiapoa bridgesii (Pfeiff.) Backeb., DQ855879 (Nyffeler, 2007), AY015293 (Nyffeler, 2002); Decarya
madagascariensis Choux, AF194844 (Applequist & Wallace, 2001), AY042574 (Cuénoud & al., 2002); Didierea trollii Capuron & Rauh, AF194845 (Ap-
ple quist & Wal lace, 2001), AY0 42576 (C uénoud & a l., 200 2); Echinocactus platyacanthus Link & Otto, –, AY015287 (Nyffeler, 2002); Grahamia bracteata
Gi ll.; A F19484 6 (Ap pleq uis t & Wa lla ce, 2 001), AY015273 (Ny ffe ler, 200 2); Halophytum amegh inoi Speg., –, AY514852 (Müller & Borsch, 20 05); Hectorella
caespitosa Hook. f.; DQ093963 (Applequist & al., 2006); DQ267197 (Applequist & al., 2006); Lewisia cantelovii J.T. Howell, –, AY042607 (Cuénoud &
al., 2002); Lewisia pygmaea (A. Gray) B.L. Rob., AF194847 (Applequist & Wallace, 2001), —; Maihuenia patagonica (Phil.) Britton & Rose; DQ855877
(Nyffeler, 2007), AY015281 (Nyffeler, 2002); Maihueniopsis subterranea (R.E. Fr.) E.F. Anderson, –, EU834746 (this study; Bolivia: Potosí, Rausch s.n.;
ZSS 28414); Montia parvifolia (DC.) Greene, AF194851 (Applequist & Wallace, 2001), AY042616 (Cuénoud & al., 2002); Opuntia quimilo K. Schum., –,
AY015279 (Ny ffele r, 20 02); Parakeelya volubilis (Benth.) Hershk., AF194838 (Applequist & Wallace, 2001; as Calandrinia volubilis), –; Pereskia aculeata
Mill., DQ855876 (Nyffeler, 2007), DQ855863 (Nyffeler, 2007); Pereskia guamacho F.A.C. Weber, –, AY015275 (Nyffeler, 2002); Pereskia stenantha F.
Ritter, –, AY015276 (Nyffeler, 2002); Pereskia zinniif lora DC., –, AY015277 (Nyffeler, 2002); Pereskiopsis diguetii (F.A.C. Weber) Britton & Rose, –,
AY015280 ( Nyf feler, 2 002); Phemeranthus multiflorus (Rose & Standley) Ocampo, –, EU834747 (this study; Mexico: Queretaro, Ocampo & Morales 1484;
ZSS 27389); *Phemeranthus punae (R.E. Fr.) Eggli & Nyffeler, –, EU834748 (this study; Argentina: Salta, Leuenberger & Eggli 4867a; ZSS 23769); Phe-
meranthus teretifolius Raf., –, EU834749 (this study; ex cult. Huntington Botanical Garden; HNT); Portulaca cf. bicolor F. Muell., DQ855870 (Nyffeler,
2007), DQ855848 (Nyffeler, 2007); Portulaca eruca Hauman, –, DQ855849 (Nyffeler, 2007); Portulaca fluvialis D. Legrand, –, EU834750 (this study;
Urug uay: Río Negro, Nyffeler & Eggli 1652; ZSS 26796); Portulaca grandif lora L., AF194853 (Applequist & Wallace, 2001), EU834751 (th is study; Urug uay:
Paysandú, Nyffeler & Eggli 1673; ZSS 26698); Portulaca oleracea L., AY194867 (Applequist & Wallace, 2001), DQ855850 (Nyffeler, 2007); Portulacaria
afra Jacq., AF194857 (Applequist & Wallace, 2001), AY042637 (Cuénoud & al., 2002); Rhipsalis floccosa Pfeiff., –, AY015342 (Nyffeler, 2002); Stetsonia
coryne (Salm-Dyck) Britton & Rose, –, AY015320 (Nyffeler, 2002); Talinella pachypoda Eggli, DQ855868 (Nyffeler, 2007), DQ855846 (Nyffeler, 2007);
Tal ino psi s frute scens A. Gray, DQ855871 (Nyffeler, 2007), DQ855851 (Nyffeler, 2007); Talinum caffrum (Thunb.) Eckl. & Zeyh., AY194859 (Applequist
& Wallace, 2001), AY042662 (Cuénoud & al., 2002); Talinum lineare Kunth, –, EU834752 (this study; Mexico: Michoacán, Ocampo & Morales 1460; ZSS
27415); Talinum paniculatum (Jacq.) Gaertn., DQ855866 (Nyffeler, 2007), AY015274 (Nyffeler, 2002); Talinum polyg alo ide s Arn., DQ855867 (Nyffeler,
2007), DQ855845 (Nyffeler, 2007); Talinum portulacifolium (Forssk.) Schweinf., DQ855869 (Nyffeler, 2007), DQ855847 (Nyffeler, 2007); Ta lin um tri-
angulare (Jacq.) Willd., DQ855865 (Nyffeler, 2007), DQ855844 (Nyffeler, 2007); Ullucus tuberosus Caldas, AF194865 (Applequist & Wallace, 2001), –.
OUTGROUP SPECIES: Delosperma cooperi L. Bolus, DQ855864 (Nyffeler, 2007), DQ855843 (Nyffeler, 2007); Limeum africanum L., –, AY042608
(Cuénoud & al., 2002); Mirabilis jalapa L., AF194826 (Applequist & Wallace, 2001), AY042614 (Cuénoud & al., 2002); Mollugo verticillata L., AF194827
(Applequist & Wallace, 2001), DQ267195 (Applequist & al., 2006); Phytolacca americana L., AF130229 (Olmstead & al., 2000), –; Phytolacca dioica L.,
–, AY042631 (Cuénoud & al., 2002).
Appendix 2.
New combinations.
Anacampseros coahuilensis (S. Watson) Eggli & Nyffeler, comb. nov. Basionym: Talinum coahuilense S. Watson in Proc. Amer. Acad. Arts 26: 132. 1891.
Phemeranthus punae (R.E. Fr.) Eggli & Nyffeler, comb. nov. Basionym: Calandrinia punae R.E . Fr. in Nova Act a Reg iae Soc. S ci. U psa l., s er. 4 , 1: 149. 1905.
... One of the most enigmatic groups is the genus Copiapoa, comprising about 32 species endemic to coastal Atacama Desert ( Larridon et al. 2015). This genus has been consid- ered "orphan," because its tribal and phylogenetic placement remains uncertain within Cactoideae ( Nyffeler and Eggli 2010). This mono- phyletic group shows very low genetic variation both with plastid and nuclear markers ( Larridon et al. 2015). ...
... Tribe Notocacteae, composed mainly of solitary or shrubby cacti, were shown to be paraphyletic in molecular phylogenies (Nyffeler and Eggli 2010). However, there is a core Notocacteae clade that has been consistently recovered as monophyletic with good support val- ues ( Figure 2; Nyffeler 2002; Hernández-Hernández et al. 2011). ...
... In the western Andes, the highest diversity of Notocacteae is southward displaced (22-34°S) ( Guerrero et al. 2011b), since the hyper arid Atacama Desert severely constrains the colonization of major groups of plants and animals ( Guerrero et al. 2013). Phylogenies with limited sampling suggest Parodia is mono- phyletic while Eriosyce s.l. is not (Nyffeler 2002;Nyffeler and Eggli 2010;Guerrero et al. 2011a;Hernández-Hernández et al. 2011). ...
Article
Full-text available
Members of the cactus family are keystone species of arid and semiarid biomes in the Americas, as they provide shelter and resources to support other members of ecosystems. Extraordinary examples are the several species of flies of the genus Drosophila that lay eggs and feed in their rotting stems, which provide a model system for studying evolutionary processes. Although there is significant progress in understanding the evolution of Drosophila species, there are gaps in our knowledge about the cactus lineages hosting them. Here we review the current knowledge about the evolution of Cactaceae, focusing on phylogenetic relationships and trends revealed by the study of DNA sequence data. During the last several decades, the availability of molecular phylogenies has considerably increased our understanding of the relationships, biogeography, and evolution of traits in the family. Remarkably, although succulent cacti have very low growth rates and long generation times, they underwent some of the fastest diversifications observed in the plant kingdom, possibly fostered by strong ecological interactions. We have a better understanding of the reproductive biology, population structure and speciation mechanisms in different clades. The recent publication of complete genomes for some species has revealed the importance of phenomena such as incomplete lineage sorting. Hybridization and polyploidization are common in the family, and have been studied using a variety of phylogenetic methods. We discuss potential future avenues for research in Cactaceae, emphasizing the need of a concerted effort among scientists in the Americas, together with the analyses of data from novel sequencing techniques.
... Portulacaceae currently includes only the genus Portulaca L. (Nyffeler and Eggli 2010) with ca. 100 species (Legrand 1962;Eggli and Ford-Werntz 2002) distributed worldwide, with centers of diversity in South America and Africa (Ocampo and Columbus 2012). Phylogenetic analyses recovered two major lineages this genus, one clade containing species with opposite leaves (OL clade) and the other comprising species with alternate leaves taxa (AL clade) (Ocampo andColumbus 2010, 2012). ...
... (María and Slanis 2010), segregated from Portulacaceae s.l. (Nyffeler and Eggli 2010). ...
Article
Full-text available
Portulaca minensis is a poorly known species of Portulacaceae, considered exclusive to the campo rupestre vegetation in Minas Gerais, Brazil. It shares morphological features with members of the Cryptopetala clade, composed of P. mucronata, P. cryptopetala, and P. hirsutissima. The purpose of this paper is to expand the knowledge about the taxonomy, distribution, and micromorphology of P. minensis, comparing it to its morphologically related species occurring in the Espinhaço Range, along Eastern Brazil. We analyzed specimens of this group in herbaria and in the field and gathered data on micromorphology of seeds surface and pollen grains using scanning electron microscopy. In addition, we also analyzed leaf anatomy using light microscopy for the first time to this species. Portulaca minensis can be distinguished from its allies within the P. hirsutissima complex by glabrous leaves and sepaloid (vs. pilose) and from P. mucronata, it differs by reddish sepaloid structures (vs. greenish). As this species showed leaf anatomy with intermediate metabolism C 3-C 4 , we consider it as a probable member of the Cryptopetala clade. The pantocolpate pollen grains were similar to most Portulaca species (except the P. hirsutissima complex), but the slightly convex ornate seeds with par-dome type projections were considered diagnostic to P. minensis. We also indicate new localities of occurrence to P. minensis in the state of Bahia, expanding the species distribution toward the Northeastern region. This species is classified as endangered according to the IUCN criteria (EN B1 + b(i, ii, iii), with EOO = 3,557,978 km 2 and AOO = 28,000 km 2 .
... The Portulacineae comprise ~ 2200 species in nine families and 160 genera (Nyffeler and Eggli 2010a; Angiosperm Phylogeny Group 2016) that are inferred to have diverged from its sister group, the Molluginaceae, ~54 Mya ( Arakaki et al. 2011). Species in this clade are incredibly morphologically diverse, ranging from annual herbs to long-lived stem succulents to trees, and include several iconic plants such as the cacti and Malagasy Didiereaceae ( fig. ...
Article
Several plant lineages have evolved adaptations that allow survival in extreme and harsh environments including many families within the plant clade Portulacineae (Caryophyllales) such as the Cactaceae, Didiereaceae, and Montiaceae. Here, using newly generated transcriptomic data, we reconstructed the phylogeny of Portulacineae and examined potential correlates between molecular evolution and adaptation to harsh environments. Our phylogenetic results were largely congruent with previous analyses, but we identified several early diverging nodes characterized by extensive gene tree conflict. For particularly contentious nodes, we present detailed information about the phylogenetic signal for alternative relationships. We also analyzed the frequency of gene duplications, confirmed previously identified whole genome duplications (WGD), and proposed a previously unidentified WGD event within the Didiereaceae. We found that the WGD events were typically associated with shifts in climatic niche but did not find a direct association with WGDs and diversification rate shifts. Diversification shifts occurred within the Portulacaceae, Cactaceae, and Anacampserotaceae, and while these did not experience WGDs, the Cactaceae experienced extensive gene duplications. We examined gene family expansion and molecular evolutionary patterns with a focus on genes associated with environmental stress responses and found evidence for significant gene family expansion in genes with stress adaptation and clades found in extreme environments. These results provide important directions for further and deeper examination of the potential links between molecular evolutionary patterns and adaptation to harsh environments.
... Formerly, Montiaceae were included in the traditional but paraphyletic family Portulacaceae (Nyffeler & Eggli 2010). Its conceptual origins trace to Hershkovitz (1993, who referred to a group he called Western American Portulacaceae. ...