TAXON 59 (3) • June 2010: 905–910 Douglas & Spellenberg • Tribal classification of Nyctaginaceae

A new tribal classification of Nyctaginaceae

Norman Douglas1 & Richard Spellenberg2

1 Department of Plant Biology, North Carolina State University, Campus Box 7612, Raleigh, North Carolina 27695-7612 U.S.A.
2 Department of Biology, MSC 3AF, New Mexico State University, Las Cruces, New Mexico 88003-8001, U.S.A.
Author for correspondence: Norman Douglas, [email protected]

Abstract Recent phylogenetic work shows that existing tribal concepts within Nyctaginaceae are incompatible with the
principle of recognizing monophyletic taxa. We review the history of supergeneric classification in Nyctaginaceae, clarify
issues pertaining to priority of certain generic names, and discuss the application of the conserved family name to the tribe
Nyctagineae. Pisoniella and Phaeoptilum are moved from tribe Nyctagineae to Pisonieae and Bougainvilleeae, respectively,
while tribe Abronieae, containing Abronia and Tripterocalyx, is submerged into Nyctagineae. Two distinctive genera, Caribea
and Colignonia, are assigned to their own tribes, recognizing both their uniqueness and the uncertainty of their phylogenetic
relationships. Finally, subtribes are not recognized in tribe Nyctagineae. Updated descriptions are provided for each tribe and
one new tribe, Caribeaeae Douglas & Spellenberg, is recognized. Plant habit and general geographic distribution seem to be
at least as pertinent as the often-homoplasious morphological details which had been emphasized in previous classifications.

Keywords conserved names; nomenclature; Nyctaginaceae; phylogenetics; supergeneric taxonomy

Introduction known forest species from Central America, possesses winged

fruits quite unlike those of other trees in the family. Its fruits
Diversity, distribution, and morphology. — The family are more similar in general appearance to those of Phaeoptilum
Nyctaginaceae comprises 300–400 species of trees, shrubs, and the xerophytic herbs found in North America, Tripteroca-
and herbs classified in approximately 30 genera (Mabberley, lyx and some Acleisanthes. Okenia, with one or two species,
1987; Bittrich & Kühn, 1993). Species in Nyctaginaceae are has unique geocarpous fruits.
found in all warmer areas of the world, but the bulk of the Nyctaginaceae have provided considerable fodder at the
diversity at the generic and specific levels occurs in two re- suprageneric and infrageneric levels for several generations
gions: the Neotropics, and arid western North America. In the of taxonomists. At the species level, perceived intergradation
Neotropics, the majority of species are shrubs or small trees across morphospecies, in combination with a proliferation of
found in the three large genera Neea, Guapira, and Pisonia. In names, has led to much taxonomic confusion, e.g., Mirabilis,
addition, the familiar genus Bougainvillea, known primarily (Reed, 1969; Spellenberg, 2003a), and spicate species of Boer-
for the horticulturally important B. glabra, B. spectabilis and havia (Kearney & Peebles, 1964; Reed, 1970; Spellenberg,
hybrids, is endemic to South America, and is especially diverse 2002). Some of the differences in species concepts result from
in Bolivia. Species of Colignonia are restricted to the Andean actual incomplete differentiation of populations, perhaps in
region. Diversity at the generic level is greater in the arid re- response to recently changing environments as in geologically
gions of North America where nearly half of the recognized diverse western North America (Boerhavia, Abronia, Mirabi-
genera in the family are present, the most diverse being the lis), and in response to changing taxonomic fashion with regard
herbaceous and suffrutescent Abronia, Boerhavia, Mirabilis, to species concepts (i.e., “lumping” and “splitting”) over the
and Acleisanthes. Commicarpus, with roughly 30 species in past century (Weakley, 2005). In the arborescent tropical gen-
Africa and western Asia, is also found in North and South era, comparatively poor preservation in herbarium specimens
America with five endemic species. of many characters, combined with incomplete collections of
The number of species in genera of Nyctaginaceae follows these often-dioecious plants, has meant that species in the gen-
a classic “hollow curve” pattern of diversity (Willis, 1922). The era Neea, Guapira, and Pisonia remain poorly understood.
few large genera differ in geographic distribution and are mor- In contrast, differences in classification at the generic level
phologically variable. There are many genera of low diversity, and above more likely result from the paucity of characters
14 of those being monotypic. While in some cases monotypic provided by the Nyctaginaceae flower with its simple and often
genera, or genera with very few species, represent minor off- reduced perianth, and the simple, single-carpel, uni-ovulate
shoots of dubious distinction (for instance, three genera for ovary. Given the apparent ease with which the perianth (and
four species in tribe Boldoeae), it is also apparent that many thus the anthocarp) is modified in response to pollination and
small genera are well differentiated from other members of dispersal pressures, characters of these organs are perhaps less
the family. For example, the monotypic genus Phaeoptilum is conserved than are characters of the gynoecium wall in other
a unique xeromorphic shrub with winged fruits. This endemic families. We draw this inference from the fact that the acces-
of southwestern Africa is the only genus not occurring in the sory fruit characters (e.g., wings, ribs, glands) previously used
Americas. Likewise, the monotypic genus Grajalesia, a poorly in classifications display substantial homoplasy (Douglas &

905
 19968175, 2010, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/tax.593018 by Universidad De Chile, Wiley Online Library on [25/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Douglas & Spellenberg • Tribal classification of Nyctaginaceae TAXON 59 (3) • June 2010: 905–910

Manos, 2007). Thus, while ovary and fruit characteristics carry were to be placed into a single genus, as Standley preferred.
morphological information that is phylogenetically important In his last treatment of the family, Heimerl (1934) maintained
in many families (e.g., Lawrence, 1951; Stebbins, 1974), artifi- the three genera; naturally, he maintained Boldoeae. However,
cial groupings have resulted when similar confidence has been many treatments (e.g., Spellenberg, 2001) have followed Stan-
extended to the characteristics of the anthocarps in Nyctagina- dley in treating Boldoa as a synonym of Salpianthus.
ceae. Such artificial grouping was clearly demonstrated at the Another issue that has never been satisfactorily addressed
generic level, when three previously recognized genera, Aclei- concerns the proper name for the tribe containing the type of
santhes s.str. (long perianth, vespertine, ribbed anthocarp), the family name Nyctaginaceae. Nyctago, the generic name
Ammocodon (= Selinocarpus) chenopodoides (short perianth, upon which the family name is based (Art. 10.6 of the ICBN,
matutinal, winged anthocarp), and the remaining species of McNeill & al., 2006) is a superfluous and illegitimate name
Selinocarpus (long perianth, vespertine, winged anthocarp) to be typified by the type of Mirabilis. Nowicke (1970) noted
(see Fowler & Turner, 1977) each were found to be non-mono- that Article 19 of the Code required that Heimerl’s (1934)
phyletic separately, but instead to comprise multiple lineages tribe Mirabileae, subtribe Boerhaaviinae be changed to tribe
embedded within a single clade (Levin, 2000). They are now Nyctagineae, subtribe Nyctagininae, establishing these names
all considered to belong within Acleisanthes (Levin, 2002). by reference to Heimerl (1934). However, she referred to Mi-
Similarly, changing concepts of the boundaries between groups rabileae in two subsequent publications on pollen morphology
at the generic level have resulted in very different taxonomies, (Nowicke, 1975; Nowicke & Skvarla, 1979). Bogle (1974) dis-
e.g., Mirabilis (Standley, 1918; Spellenberg, 2003a), Boerhavia cussed a conflict that existed in the 1966 and 1972 Codes that
(Fosberg, 1978; Spellenberg, 2003b). pertained to families with conserved names, noting that both
Supergeneric classifications. — The family was first rec- Heimerl’s and Nowicke’s names could be considered correct,
ognized by Jussieu (1789). Early treatments of the family by depending on whether one emphasized the requirement that
Meisner (1841), Choisy (1849) and Bentham & Hooker (1880) subfamilial taxa conform to the “correct” name of an included
established the outlines of a tribal classification by establishing genus, or the requirement that the names of such groups be
Mirabileae Meisner, Pisonieae Meisner, Bougainvilleae Choisy, based on the same stem as the next higher taxon. Article 19.3
and Leucastereae Benth & Hook. Heimerl (1889) redrew the was revised in the Sydney Code (Voss & al., 1983) so that such
three tribes recognized by Bentham & Hooker. Standley (1918) a taxon was to be based on the “type of the adopted, legitimate
recognized Heimerl’s subtribes of tribe Mirabileae as tribes name of the family to which it is assigned, but without citation
Abronieae, Bougainvilleae and Colignonieae. His publication of of an author’s name”. This wording is maintained in the Vienna
Abronieae and Colignonieae at tribal rank is valid by virtue of Code Article 19.4 except that the proscription of author citations
the inclusion of a dichotomous key to the tribes. Heimerl (1934) in tribal and subtribal names disappeared with the Tokyo Code
updated the tribal classification of Nyctaginaceae, recognizing (Greuter & al., 1988). In this case, Nyctago nom. illeg. is the
five tribes. The tribe Pisonieae included genera that represented type genus. Article 19.5 further clarifies this issue by stating
the majority of the tropical arborescent taxa, except the three that subfamilial names based on illegitimate generic names
genera in Leucastereae. The largest of Heimerl’s tribes, Mira- are legitimate if they are also the base of a conserved family
bileae, was further divided into four subtribes, one of which, name, meaning that the proper name of the tribe that includes
Boerhaaviinae, contained most of the herbaceous and suffrutes- Mirabilis is Nyctagineae. Bittrich & Kühn (1993), like Nowicke
cent desert taxa. Bittrich & Kühn (1993) updated the treatment (1970), recognized the bulk of the genera in Heimerl’s (1934)
of the family and made several adjustments to Heimerl’s 1934 tribe Mirabileae as tribe Nyctagineae, citing no author as per
treatment, including segregating from subtribe Boerhaaviinae the final clause in Art. 19.3 then in effect. In fact, the first use of
a new subtribe, Nyctagininae, which represented genera with this tribal name was by Horaninov (1847: 105–107), whose tribe
substantial bracts forming an involucre (Fig. 1). Nyctagineae was equivalent to the modern Nyctaginaceae and
The proper name for the tribe Boldoeae (Heimerl, 1889) was one of four tribes in a broadly conceived “Allioniaceae”,
has been confused in the literature for over 90 years. The tribe along with Plumbagineae, Staticeae, and Plantagineae. If a
contains three genera that are similar to a degree that Standley subtribe including Mirabilis were recognized (which we will
included in Salpianthus Humb. & Bonpl. (1807) the genera not do) its proper name would be Nyctagininae Nowicke.
Boldoa (commonly cited as Cav. ex Lag. 1816), and Crypto-
carpus Kunth (1817) (Standley, 1918, 1931). When he combined
these genera, he renamed the tribe Boldoeae as Salpiantheae. Taxonomic implications of
However, he apparently erred in determining the priority to be molecular studies
with Salpianthus. The original publication of the name Boldoa
is as follows: Cavanilles, Hortus regius matritensis (1803: 8–9, The first molecular phylogenetic study of Nyctaginaceae
tab. 7). This was later cited in Lagasca, Nova genera et species (Levin, 2000) explicitly focused on some members of tribe
(1816: 10) (Heimerl, 1889). Both works include Latin descrip- Nyctagineae. Due to sampling limitations and poor backbone
tions of Boldoa and Boldoa purpurascens (likely explaining resolution, the only result pertinent to higher-level classification
Standley’s error) but the former also has an illustration of the was the suggestion that Boerhavia and Allionia were relatively
species and clearly represents valid publication. Thus Boldoa closely related, calling into question the separation of subtribes
has priority over Salpianthus if the four species in the group Boerhaviinae and Nyctagininae. A recent phylogenetic study

906
 19968175, 2010, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/tax.593018 by Universidad De Chile, Wiley Online Library on [25/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TAXON 59 (3) • June 2010: 905–910 Douglas & Spellenberg • Tribal classification of Nyctaginaceae

Bentham &
Hooker, 1880 Heimerl, 1934 Bittrich & Kühn, 1993 Douglas & Spellenberg
Mirabileae-Boerhaavieae Mirabileae-Boerhaaviinae Nyctagineae-Boerhaviinae Nyctagineae Boerhavia
Mirabileae-Boerhaavieae Mirabileae-Boerhaaviinae Nyctagineae-Boerhaviinae Nyctagineae Okenia
Mirabileae-Boerhaavieae Mirabileae-Boerhaaviinae Nyctagineae-Nyctagininae Nyctagineae Nyctaginia
Mirabileae-Boerhaavieae Mirabileae-Boerhaaviinae Nyctagineae-Boerhaviinae Nyctagineae Anulocaulis
Mirabileae-Boerhaavieae Mirabileae-Boerhaaviinae Nyctagineae-Boerhaviinae Nyctagineae Cyphomeris
Mirabileae-Boerhaavieae Mirabileae-Boerhaaviinae Nyctagineae-Nyctagininae Nyctagineae Allionia
Mirabileae-Boerhaavieae Mirabileae-Boerhaaviinae Nyctagineae-Boerhaviinae Nyctagineae Commicarpus
Mirabileae-Boerhaavieae Mirabileae-Boerhaaviinae Nyctagineae-Nyctagininae Nyctagineae Mirabilis
Mirabileae-Abronieae Mirabileae-Abroniinae Abronieae Nyctagineae Tripterocalyx
Mirabileae-Abronieae Mirabileae-Abroniinae Abronieae Nyctagineae Abronia
Mirabileae-Boerhaavieae Mirabileae-Boerhaaviinae Nyctagineae-Boerhaviinae Nyctagineae Acleisanthes
Pisonieae Pisonieae Pisonieae Pisonieae Pisonia
Pisonieae Pisonieae Pisonieae Pisonieae Neea
Pisonieae Pisonieae Pisonieae Pisonieae Guapira
Pisonieae Pisonieae Nyctagineae-Colignoniinae Pisonieae Pisoniella
Mirabileae-Bougainvillaeeae Mirabileae-Bougainvilleinae Bougainvilleeae Bougainvilleeae Bougainvillea
Bougainvilleeae Bougainvilleeae Belemia
Mirabileae-Phaeoptilinae Nyctagineae-Phaeoptilinae Bougainvilleeae Phaeoptilum
Mirabileae-Boldoeae Colignonieae Nyctagineae-Colignoniinae Colignonieae Colignonia
Mirabileae-Boldoeae Boldoeae Boldoeae Boldoeae Salpianthus
Leucastereae Leucastereae Leucastereae Leucastereae Leucaster
Leucastereae Leucastereae Leucastereae Ramisia
Leucastereae Leucastereae Leucastereae Leucastereae Andradea
Mirabileae-Boldoeae Leucastereae Leucastereae Leucastereae Reichenbachia

Nyctagineae-Boerhaviinae Caribeeae Caribea

Mirabileae-Boldoeae Boldoeae Boldoeae Boldoeae Boldoa
Leucastereae Boldoeae Boldoeae Boldoeae Cryptocarpus
Pisonieae Pisonieae Grajalesia
Pisonieae Pisonieae Pisonieae Pisonieae Cephalotomandra
Pisonieae Pisonieae Neeopsis
Mirabileae-Boerhaaviinae Nyctagineae-Nyctagininae Nyctagineae Cuscatlania

Fig. 1. Incongruence of historically influential classification schemes for Nyctaginaceae with respect to the phylogenetic hypothesis of Douglas
& Manos (2007) and the new classification proposed here. All branches in the phylogenetic tree were supported at least >70% in parsimony
or likelihood bootstrap or >95 bayesian posterior probability; otherwise they were collapsed. Hyphenated names correspond to tribe-subtribe;
names in bold represent changes from Bittrich & Kühn (1993); genera unconnected to the tree were not included in the 2007 analysis.

of Nyctaginaceae that included nearly all currently accepted distinction between the subtribes Boerhaviinae and Nyctag-
genera (Douglas & Manos, 2007) found that all previous clas- ininae is artificial due to the highly homoplasious distribution
sifications of Nyctaginaceae at the tribal level are incongruent of involucral bracts in the North American xerophytic clade
with the evolutionary relationships among the genera dem- (Douglas & Manos, 2007). Finally, Caribea was included in
onstrated by molecular evidence (Fig. 1). With respect to the subtribe Boerhaviinae, but a preliminary result based on a frag-
most recent (Bittrich & Kühn, 1993), the tribe Nyctagineae is mentary ndhF sequence indicated that this poorly known taxon
not monophyletic due the basal position of the genus Acleisan- is probably more closely related to Bougainvillea or Pisoniella
thes (tribe Nyctagineae subtribe Boerhaviinae, now including than to any members of the Nyctagineae. Now that the relation-
Selinocarpus and Ammocodon; Levin, 2002) with respect to a ships among the genera of the family are better understood, a
clade that contains Abronia and Tripterocalyx (tribe Abronieae) formal reclassification of the family into monophyletic tribes
and the remaining members of tribe Nyctagineae subtribes is warranted. We propose the following classification, which
Nyctagininae and Boerhaviinae. The inclusion in tribe Nyctag- accomplishes the goal of recognizing monophyletic groups, is
ineae of the distantly related Phaeoptilum and Colignonia (as conservative with respect to previous classifications, and which
the monogeneric subtribes Phaeoptilinae and Colignoniinae, accommodates remaining phylogenetic uncertainty with the
respectively) is incompatible with a monophyletic tribe. The recognition of two monogeneric tribes:

907
 19968175, 2010, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/tax.593018 by Universidad De Chile, Wiley Online Library on [25/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Douglas & Spellenberg • Tribal classification of Nyctaginaceae TAXON 59 (3) • June 2010: 905–910

(1) The tribe Colignonieae Standl. (1918) is recognized, of these genera. It should be noted that the winged fruits of
containing the genus Colignonia. The ambiguity in the exact Grajalesia, an arborescent species from Mexico and Cen-
placement of this distinctive genus in the molecular phylog- tral America, bear at least superficial similarity to those of
eny of the family (Douglas & Manos, 2007) means that our Phaeo­ptilum, but at present we lack any convincing evidence
recognition of the tribe Colignonieae is likely to remain stable to suggest its placement elsewhere, or its removal from, the
even if phylogenetic resolution is improved. Furthermore, the Pisonieae. Based on general morphological similarity, the other
genus Pisoniella will be recognized in Pisonieae, as it had been unsampled genera, Neeopsis, Cephalotomandra, and Cuscat-
in every classification until Bittrich & Kühn (1993) moved it lania, seem likely to remain within the tribes where they are
into tribe Nyctagineae, subtribe Colignoniinae. The molecular currently placed.
results place Pisoniella sister, with high support, to the other Descriptions of the tribes follow below. Characters of the
genera in Pisonieae. tribes are gleaned from the literature, and where possible from
(2) Similarly, the genus Caribea is removed from tribe specimens when such were available. Genera included, their
Nyctagineae and placed in a distinct tribe, Caribeaeae. This size, and general geographic distribution are noted.
distinctive Cuban endemic is known only from the type col-
lection (Alain & Lopez P. 7013, Cuba, Oriente, 1959. Holotype Tribe 1. Leucastereae Benth. & Hook. f., Gen. Pl. 3: 3. 1880
at LS, isotype at NY!). The present phylogenetic uncertainty (‘Leucasterae’).
(Douglas & Manos, 2007) refutes the present classification Trees or scandent shrubs, unarmed. Leaves alternate,
without clearly suggesting a justifiable alternative placement. petiolate, elliptic to lanceolate, with scurfy stellate hairs or
This, in combination with the unique morphology of the genus scales; base symmetric to slightly asymmetric; margin entire.
(Alain, 1960), leads us to conclude that tribal status will be Inflorescences of axillary paniculate cymes or racemes; bracts
stable in the face of new evidence if and when this genus is minute, triangular, at base of pedicel or absent. Flowers perfect,
rediscovered and can be studied in more detail. rotate; perianth contracted in the middle or tubular throughout,
(3) The monotypic genus Phaeoptilum is transferred to the tomentose or not, accrescent, 3–5-lobed, the lobes spreading
Bougainvilleeae. This transfer, in combination with proposed or reflexed. Stamens 2 or 3 (12–20), connate at base, included.
changes 1 and 2, will remove from tribe Nyctagineae those Pollen 3-colpate, 20–35 μm, exine reticulate or spinulose.
genera that are demonstrably not closely related to the clade, Style linear or thickened, or essentially absent; stigma lateral,
which comprises the bulk of tribe Nyctagineae sensu Bittrich crested, or sulcate. Anthocarp with 12 ribs or anthocarp usually
& Kühn (1993). We note that Heimerl (1901) mentioned several absent. Embryo hooked. Andradea (1 sp., E Brazil); Ramisia
similarities between Phaeoptilum spinosum and Bougainvillea (1 sp., SE Brazil); Reichenbachia (2 spp., trop. South Amer.);
patagonica (= B. spinosa), especially in a particular short hair Leucaster (1 sp., SE Brazil).
type, and leaf form and arrangement, and for a time placed
the two genera in the same subtribe (Heimerl, 1897). Like Tribe 2. Boldoeae Heimerl in Engler & Prantl, Nat. Pflanzen-
Leucastereae, this tribe is essentially native to the southern fam. 3(1b): 21, 31. 1889.
hemisphere. Plants suffrutescent or herbaceous, often subshrubs, un-
(4) The tribe Abronieae Standl. (1918) is no longer rec- armed. Roots unknown. Leaves alternate, petiolate, thickish,
ognized. The two genera within it, Abronia and Triperoca- elliptic to rhomboid, tomentose or not, viscid, hairs straight or
lyx, are now in the tribe Nyctagineae. These two genera are hooked; bases symmetric or nearly so; margin entire. Inflo-
clearly related based on morphological similarity and molecular rescences congested axillary or terminal paniculate cymes of
evidence, and they are distinctive within the family based on glomerules or racemes, bostryches (Bittrich & Kühn, 1993),
characters of pollen morphology, stigma shape, inflorescence bracts absent. Flowers perfect, rotate, 3–5-lobed, 2–5 mm,
architecture, fruit morphology, and embryo shape. This con- tubular to campanulate, not contracted above the ovary, pu-
centration of unique apomorphic characters makes this clade bescent. Stamens 3–5, exserted, filaments free. Pollen tricol-
a poor fit with any other tribe, which is why it has often been pate, 20–46 μm, exine spinulose. Style short, linear-filiform,
recognized as distinct. The molecular evidence clearly indi- stigma delicate, fimbriate. Anthocarp globose or subglobose,
cates that it is derived from within the North American xero- coriaceous. Embryo curved. Boldoa (2–3 spp., Mexico); Cryp-
phytic clade. tocarpus (1 sp., S Mexico, Centr. Amer., NW South Amer.);
(5) We do not recognize any subtribes within tribe Nyc- Salpianthus (1 sp., Mexico, Cent. Amer., N South Amer.).
tagineae. The close relationship of Nyctaginia to Anulocaulis
(previously in separate subtribes) and low support values for the Tribe 3. Colignonieae Standl. in Britton, N. Amer. Fl. 21:
relationships between Allionia, Commicarpus, and Mirablis, 195. 1918.
preclude any justifiable grouping of these genera into subtribes. Lianas or scandent shrubs, unarmed. Roots tuberous or
Nevertheless, the Nyctagineae now represents a coherent group fibrous, also adventitious. Leaves opposite or whorled, short-
of mostly herbaceous genera based the North American xero- petiolate, deltoid, ovate, or elliptic, glabrous, puberulous, or
phytic clade. rufo-pilose, trichomes 3–4-celled; base truncate; margin en-
Four additional recognized genera were not sampled for tire. Inflorescence a cymose, umbel-like condensed dichasium;
the phylogenetic study; without evidence to justify alternative bracts white, green, or reddish, broadly lanceolate to obovate.
assignments, we propose no changes in the tribal assignment Flowers perfect, rotate; perianth lobes 3(–4)–5, campanulate or

908
 19968175, 2010, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/tax.593018 by Universidad De Chile, Wiley Online Library on [25/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
TAXON 59 (3) • June 2010: 905–910 Douglas & Spellenberg • Tribal classification of Nyctaginaceae

spreading, connate only at base. Stamens 5, exserted, episepal- margins entire, crenate, undulate or sinuate, glandular pubes-
ous, basally connate; filaments flattened, nectariferous; anthers cent to glabrous. Inflorescences terminal or axillary spikes,
subglobose. Pollen 12-pantoporate, 17–30 μm, exine spinulose. cymes, umbels or flowers solitary. Involucres of 3–20 connate
Ovary stalked, style clavate, stigma penicillate. Anthocarp or free bracts, or 1–2 often minute, persistent or cauducous
present, winged or angled. Embryo curved. Colignonia (6 spp., bracts subtending individual flowers or terminal cymes; bracts
Andean South Amer.). oblong, linear, acuminate, or lanceolate, green or scarious.
Flowers actinomorphic or slightly to strongly zygomorphic,
Tribe 4. Bougainvilleeae Choisy in Candolle, Prodr. 13: 427, campanulate to tubular to salverform, constricted above the
436. 1849. (‘Bougainvilleae’) ovary (4–)5-lobed (in some cleistogamous flowers often pres-
Trees, or shrubs, sometimes scandent, perennial, occa- ent, these quite reduced). Stamens (1–)2–5(–18), included or
sionally with spines. Leaves alternate, opposite, or fasciculate exserted, united at base and sometimes inserted on perianth
on short branches, petiolate or nearly sessile, ovate, orbicular, tube. Pollen 100–200 μm, pantoporate with spinulose exine,
to linear-lanceolate, succulent to membranous, glabrous or or tricolpate with reticulate exine. Style included or exserted,
pubescent; base symmetric or nearly so, margin entire. In- filiform, stigma linear, capitate, or peltate. Anthocarp globose,
florescence cymose or racemose or flowers solitary, bracts turbinate, clavate, obpyramidal to fusiform, radially symmetric
absent or 3, often showy, the flowers often borne on costae (gibbous in Cyphomeris, cymbiform with 2 rows of teeth in
of bracts. Flowers perfect or imperfect (plants then possibly Allionia), coriaceous (spongy and geocarpous in Okenia), 3–10
polygamous, Stannard, 1988), rotate, tubular or salverform. costate, often with membranaceous wings or viscid glands;
Stamens 5–12, often connate at base, included or didymous sulci smooth, pappilose, or rugose. Embryo hooked. Abronia
with longer stamens exserted. Pollen tricolpate with reticulate (ca. 20 spp., W and C North Amer.); Acleisanthes (17 spp., SW
exine or pantocolpate. Style short, filiform or stout, stigma lin- and SC North Amer., with 1 sp. in NE Africa); Allionia (1–2
ear to penicilliate or multifid. Anthocarp fusiform and 5-ribbed spp., Americas); Anulocaulis (5 spp., SC and SW North Amer.);
or with 4 translucent, scarious wings. Embryo curved. Belemia Boerhavia (ca. 40 spp., worldwide in tropical and subtropi-
(1 sp., E Brazil); Bougainvillea (14–18 spp., Centr. & trop. South cal regions, annuals especially diverse in SW North Amer.);
Amer.); Phaeoptilum (1 sp., SW Africa). Commicarpus (30–35 spp., nearly worldwide in tropical and
subtropical regions, especially diverse in Africa and W Asia);
Tribe 5. Pisonieae Meisn., Pl. Vasc. Gen. 10: Tab. Diagn. 318, Cyphomeris (2 spp., SC North Amer.); Mirabilis (ca. 60 spp.,
Comm. 230. 1841. Americas and 1 sp. in SC Asia); Nyctaginia (1 sp., SC North
Trees, shrubs, or scandent shrubs, unarmed or with paired Amer.); Okenia (1–2 spp., Mexico, Centr. America); Triptero-
axillary spines. Leaves alternate, opposite, whorled, or verticil- calyx (4 spp., SW North Amer.).
late, sometimes anisophyllous, petiolate, lanceolate, elliptic to
(ob)ovate, membranous to coriaceous or fleshy, glabrous to Tribe 7. Caribeeae Douglas & Spellenb., tr. nov. – Type: Ca-
glandular puberulent; base symmetric; margins entire. Inflo- ribea Alain in Candollea 17: 113. 1960.
rescences axillary and terminal paniculate cymes, corymbs, or Fruticulus perennis valde abbreviatus, depressus; caules
glomerules; bracts 2–3 beneath each flower, persistent or cau- numerosi, in base lignescentes, dense foliosi, striate, glandu-
ducous. Flowers perfect or imperfect, or polygamous, campan- losi. Folia opposita. Flores 3–5-bracteolati, solitarrii, termi-
ulate, urceolate, rotate-salverform, or tubular, the limb 5-lobed. nales. Perianthium infundibulare, tubo supra ovarium con-
Stamens (2–)5–10(–many), exserted or included, connate at stricto, limbo 5-lobato. Stamina 2, filamentis capillaribus ad
base, adnate to base of pistil in perfect flowers, filaments un- basim perianthii adnati. Ovarium globoso-oblongum, stylus
equal, Pollen generally tricolpate-spinulose. Style exserted, filiformis, exsertus, stigmate capitato. Anthocarpium globoso-
stigma penicillate or pappilose. Anthocarps oblong, clavate, or oblongum, laeve (adapted from Alain, 1960).
ellipsoid, 5-ribbed, coriaceous and glandular-sticky, or globose, Tufted, compact mat-forming, taprooted perennial. Leaves
fleshy and glabrous. Embryo straight. Cephalotomandra (1–3 petiolate, opposite, forming a stipuliform sheath at base, oblan-
spp., Colombia); (Grajalesia (1 sp., Mexico); Guapira (ca. 70 ceolate, fleshy or succulent, glabrous, margin entire. Inflores-
spp., trop. Amer.); Neea (ca. 80 spp., trop. Amer.); Neeopsis (1 cence terminal, flowers solitary, subtended by an involucre of
sp., Guatemala); Pisonia (ca. 40 spp., pantropical, but especially 3–5 free narrow bracts. Perianth constricted above the ovary,
diverse in the Americas and SE Asia); Pisoniella (1 sp., Mexico, distil portion nearly urceolate, with 5 shallow lobes. Stamens
Centr. & S South Amer.). 2, exserted, filiments adenate to perianth base. Style filiform,
exserted, stigma capitate. Anthocarp subglobose, smooth. Em-
Tribe 6. Nyctagineae Horan., Char. Ess. Fam.: 106. 1847. bryo unknown. Caribea (1 sp., Cuba).
Woody or suffrutescent subshrubs, or annual or perennial
herbs, sometimes scandent, unarmed, in some with bands of
viscid exudate on internodes. Roots slender and fleshy to tu- Acknowledgements
berous or spongy-woody taproots, rarely rhizomatous. Leaves
opposite, frequently anisophyllous, sessile or petiolate, mem- The authors wish to thank Alexander Krings and John Strother
branous to fleshy, linear, cordate, ovate, or orbicular, glabrous for helpful discussion during the preparation of this manuscript.
or pubescent, often glandular; base symmetric to asymmetric;

909
 19968175, 2010, 3, Downloaded from https://onlinelibrary.wiley.com/doi/10.1002/tax.593018 by Universidad De Chile, Wiley Online Library on [25/11/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Douglas & Spellenberg • Tribal classification of Nyctaginaceae TAXON 59 (3) • June 2010: 905–910

Literature cited McNeill, J., Barrie, F.R., Burdet, H.M., Demoulin, V., Hawksworth,
D.L., Marhold, K., Nicolson, D.H., Prado, J., Silva, P.C., Skog,
Alain. 1960. Novedades en la flora Cubana, XIII. Candollea 17: 113–121. J.E., Wiersema, J.H. & Turland, N.J. (eds.). 2006. International
Bentham, G. & Hooker, J. 1880. Genera plantarum ad exemplaria code of botanical nomenclature (Vienna Code): Adopted by the
imprimis in herbariis kewensibus servata definita. London: Reeve. Seventeenth International Botanical Congress Vienna, Austria,
Bittrich, V. & Kühn, U. 1993. Nyctaginaceae. Pp. 473–486 in: Ku- July 2005. Regnum Vegetabile 146. Ruggell: Gantner.
bitzki, K., Rohwer, J.G. & Bittrich, V. (eds.), The families and Meisner, K.F. (ed.). 1841. Plantarum vascularium genera secundum
genera of flowering plants, vol. 2. Berlin: Springer. ordines: Tabulis diagnosticis expositae. Leipzig: Libraria Weid-
Bogle, A.L. 1974. The genera of Nyctaginaceae in the southeastern mannia.
United States. J. Arnold Arbor. 55: 1–37. Nowicke, J.W. 1970. Pollen morphology in the Nyctaginaceae, part 1:
Cavanilles Palop, A.J. 1803. Hortus regius matritensis. Madrid: Real Nyctagineae, Mirabileae. Grana 10: 79–88.
Jardín Botánico. Nowicke, J.W. 1975. Pollen morphology in the order Centrospermae.
Choisy, J.D. 1849. Nyctaginaceae. Pp. 425–458 in: de Candolle, A., Grana 15: 51–78.
(ed.), Prodromus systematis naturalis regni vegetabilis, vol. 13. Nowicke, J.W. & Skvarla, J.J. 1979. Pollen morphology: The potential
Paris: Masson. influence in higher order systematics. Ann. Missouri Bot. Gard.
Douglas, N.A. & Manos, P.S. 2007. Molecular phylogeny of Nyctagi- 66: 633–700.
naceae: Taxonomy, biogeography, and characters associated with Reed, C.F. 1969. Nyctaginaceae. Pp. 151–220 in: Lundell, C.L. (ed.),
a radiation of xerophytic genera in North America. Amer. J. Bot. Flora of Texas, vol. 2. Renner, Texas: Texas Research Foundation.
94: 856–872. Reed, C.F. 1970. Nyctaginaceae. Pp. 574–599 in: Correll, D.S. & John-
Fosberg, F.R. 1978. Studies in the genus Boerhavia Nyctaginaceae, ston, M.C. (eds.), Manual of the vascular plants of Texas, vol. 6.
parts 1–5. Smithsonian Contr. Bot. 39: 1–20. Renner, Texas: Texas Research Foundation.
Fowler, B.A. & Turner, B.L. 1977. Taxonomy of Selinocarpus and Spellenberg, R. 2001. Nyctaginaceae. Fasc. 93 of: Rzedowski Rot-
Ammocodon (Nyctaginaceae). Phytologia 37: 177–208. ter, J. & de Rzedowski, G.C. (eds.), Flora del Bajío y de regiones
Greuter, W., Burdet, H.M., Chaloner, W.G., Demoulin, V., Nicolson, adyacentes. Mexico, Patzcuaro: Instituto de Ecología.
D.H., Silva, P.C. (eds.). 1988. International code of botanical Spellenberg, R. 2002. Boerhavia coulteri var. palmeri, a new varietal
nomenclature: Adopted by the Fourteenth International Botani- combination for Boerhavia (Nyctaginaceae) of southwestern North
cal Congress, Berlin July-August 1987. Regnum Vegetabile 118. America. SIDA 20: 151–155.
Königstein: Koeltz Scientific Books. Spellenberg, R. 2003a. Mirabilis. Pp. 40–56 in: Flora of North America
Heimerl, A. 1889. Nyctaginaceae. Pp. 14–32 in: Engler, A. & Prantl, K. Editorial Committee (ed.), Flora of North America north of Mexico,
(eds.), Die natürlichen Pflanzenfamilien, vol. 1b. Leipzig: Engel­ vol. 4. New York: Oxford Univ. Press.
mann. Spellenberg, R. 2003b. Nyctaginaceae. Pp. 14–74 in: Flora of North
Heimerl, A. 1897. Beiträge zur Systematik der Nyctaginaceen. XXIII. America Editorial Committee (ed.), Flora of North America north
Jahresber. K.K. Staats-Oberrealschule im XV. Bezirke Wien (Fünf­ of Mexico, vol. 4. New York: Oxford Univ. Press.
haus) 13: 23. Standley, P.C. 1918. Allioniaceae. Pp. 171–254 in: Britton, N.L. (ed.),
Heimerl, A. 1901. Monographie der Nyctaginaceen. I. Denkschr. Kai- North American flora, vol. 21. New York: New York Botanical
serl. Akad. Wiss., Wien, Math.-Naturwiss. Kl. 70: 97–141. Garden.
Heimerl, A. 1934. Nyctaginaceae. Pp. 86–134 in: Engler, A. & Prantl, Standley, P.C. 1931. The Nyctaginaceae of northwestern South Amer-
K. (eds.), Die natürlichen Pflanzenfamilien, 2nd ed., vol. 16c. ica. Publ. Field Mus. Nat. Hist., Bot. Ser. 11: 71–126.
Leipzig: Engelmann. Stannard, B.L. 1988. Nyctaginaceae. Pp. 16–37 in: Launert, E. (ed.),
Horaninov, P.F. 1847. Characteres essentiales familiarum ac tribuum Flora zambesiaca, vol. 9, part 1. Kew: Royal Botanic Gardens.
regni vegetabilis et amphorganici. Petropoli: Wienhöberianis. Stebbins, G.L. 1974. Evolutionary processes and the origin of higher
Jussieu, A.L. 1789. Genera plantarum. Paris: Herrisant. categories. Pp. 12–28 in: Flowering plants: Evolution above the
Kearney, T.H. & Peebles, R.H. 1964. Nyctaginaceae. Pp. 270–279 species level. Cambridge: Harvard Univ. Press.
in: Kearney, T.H. & Peebles, R.H. (eds.), Arizona flora, 2nd ed. Voss, E.G., Burdet, H.M., Chaloner, W.G., Demoulin, V., Hiepko, P.,
Berkeley: Univ. of California Press. McNeill, J., Meikle, R.D., Nicolson, D.H., Rollins, R.C., Silva,
Lagasca y Segura, M. 1816. Genera et species plantarum. Madrid: P.C. & Greuter, W. (eds.). 1983. International Code of Botanical
Matriti ex Typographia Regia. Nomenclature: Adopted by the Thirteenth International Botanical
Lawrence, G.H.M. 1951. Taxonomy of vascular plants. New York: Congress, Sydney, August 1981. Regnum Vegetabile 111. Utrecht:
MacMillan. Bohn, Scheltema & Holkema.
Levin, R.A. 2000. Phylogenetic relationships within Nyctaginaceae Weakley, A.S. 2005. Change over time in our understanding of the
tribe Nyctagineae: Evidence from nuclear and chloroplast ge- flora of the southeastern United States: Implications for plant sys-
nomes. Syst. Bot. 25: 738–750. tematics, bioinformatics, and conservation. Dissertation, Duke
Levin, R.A. 2002. Taxonomic status of Acleisanthes, Selinocarpus, and University, Durham, North Carolina.
Ammocodon (Nyctaginaceae). Novon 12: 58–63. Willis, J.C. 1922. Age and area: A study in geographical distribution
Mabberley, D.J. 1987. The plant book. Cambridge: Cambridge Univ. and origin of species. Cambridge: Cambridge Univ. Press.
Press.

910