Phylogenetic Relationships in Euphorbieae (Euphorbiaceae) Based on ITS and ndhF Sequence Data

38 Pages·2002·27.8 MB·

by  Victor W Steinmann

89
Volume Annals
Number
the
of
2002
Missouri
Botanical
Garden
PHYLOCENETIC
and
Steuimann^-^
Victor W.
Mark
Porter-
RELATIONSHIPS
IN ,/.
EUPHORBIEAE
BASED
(EUPHORBIACEAE)
AND
ON
ndhF
ITS
SEQUENCE
DATA'
Abstkact
Th(^ nK)n()|)livtv and pin lot^cnctic relati()nslii|ts iif the Irihe Kii|)h<n hicac (Kui»hnrhiaeeae) w<'re evaluated uslii^ se[)-
inDNA
spacer
arale vvei^liled niaximuin parsirnnriy analyses of nueh-olide sctiuein-es of the inlernal IranserihiMi (1 TS)
re<^iori arul cpDN A c(tditig re-^ioti iidhV. Tlie sliid\ IncluchMl 22."^ inn;n)U[) species n^presenliiig nearly all of llie previously
subgenera, and sections within the Euplunbieae and oulj^roup taxa from \\\v tribe lli[)[>otnaneae
rect>^nized jicneia.
1
(Knpliorbiac(^ae). 1'^ ^^^fhV aiial}ses suppitrl the innnophvly of Kuphorbieae in ad<litit)n to the nioriuplivly
I^'>ll' tl><' ^>'i*'
I
major
Within FAiphorbiinae. tlure an- four
of three subtribes. Anthosteininae. Neogiiillainniniinae. anil Kupliorbiiiiai:.
its
chuh^s. onlv one of which corresponds with a |)rfviously recon;iii/<'d laxori: the three remaining clades are con^lornerales
of various subgenera and sectioris. The niajorily of the subtiihe is comjtosed of a paraplnletic FAiplwilnd. All other
Synadcniiim and Endndi'inum
currently recognized in the subtril)e are nested within F.iiphorlu'd. In addition.
«i<'neia
are nested within Monadenium. Within Hiipltnrhia, the niajnritv of the eurrentl\ n-co^ni/ed snb^<niera are eitlu^r par-
aphvletic or [xtlvphvletic. Bio*;eo^raphical patterns exatnined in li^lit of the molecular evidence suggest that the tribe
Kuphorbieae arose Africa. possii>iy Ijefore the !>reaku|) of (ittudwanaland. at whiih time the major lineages of subtril)e
in
We are extremely gratehil to all those who have assist^^l with this research. In particular we thank Susan (barter,
'
Oady
MePherson. and Webster reviewing
Columbus, FdizalxMh Victoria llnllowell. (,ordnn for critically
Travis Kriar.
J.
the manuscri]tt. For permission to examine tluir c(j|lectiotr^ and to sam[)lc herbarium uialerial, we thank the curators
and MUZ. CAS. FNCB. K CB. K. MO, NY. USA. SCO, Sl\ and TEX. The following institntions ami
I)A\.
staffs at
Chuck Hanson
Dylan llannon; of Arid
inilividuals extended assistance by permitting access to their living collections:
bands (weenhouses; Mark Mayfield; Meetia Singh; S<'ymonr bindf^n; John Trager and Jot- ChMuents of the llnntington
Herman Cordon McTherson and Sebastian
Snkkulenten-Sammlung and Schwartz.
Botanical Carden; I'rs Eggli of /ihich:
collected material .W'of^iiillaandnid itcopalnt and Kupliorlmi grtfiKiitui, respecti\ ely. Fieldwork, herbarium
Teillier
(A'
and laboiatory use and materials were made possible by the giMu-rous support of the Andrew W. M<'llon Fonn-
visits,
Hardman
Howard and Phoebe Brown Scholarship, and the
American of Taxonomists, the
ihition, the Socii'tv IMarit
C
and
Native FMant Besearch Award. I.anren Baz jirovided ITS secjucnees for Cluimui'syce degcneri, C, hyper'uljoiia,
We and graduate
Kenneth Wnrdack gave useful information about outgroup selection. are Indebted to ihe staff
prnslnita.
Hancho Santa Ana Botanic Carden enconragement and support, especiall} Eric Uoalson for his
students for
at tlte
and
(n)ntinuous hclji ad\ice.
sleinmarm®
01711, U.S.A.
M^ancho Santa Ana Botanic Carden, Colh^g*' Ave., Claremont. California
ir>(K)
i\.
inecdibajio.edu.mx.
61600 Michoacan,
Hairrent address: Institnto dv Fcologia, A.C.. Centro Begional del WhjUk A.B. ;U;6, Batzcuaro,
M<'xico.
Gahd. 453-190. 2002.
Ann. MissoLiii lior. H9:
454
Annals
the
of
Garden
Missouri Botanical
were
FiijilKirhiinae alreaily |»resetil. It is ar^iunl tlial [\\v best .solution for Euphorbia tlabsifuation is a )roa(i (ircuin-
1 I
scriptioii oi Kupliorhia lliat contains all of the ahont 2000 species ol the subtiil)e Eupli(nLiinat\
Kvy
words:
classifunliori. Knp}\nr})i(i, Kupliorhii^ir. ITS, udhV.
was
It only a year after lanitaeiis (17r>;?) first eir- periaiilli; in NeDguillauiiiiniinae (»nly llie pistillate
genus Euphorbia
riittiscriJKMl llie L, (Eiipliorhi- flowers possess a perianlli; and in Eu()horbiinae
aceae) to ineliide all of the iben-knovvn inenih<Ms neither the pistillate nor the starninate flowers pos-
of the triJH^ lM4)liorl)iea(Mhat other botanists began sess a W(^ll-developed periantlu although a rudi-
genus
to divid(» the into several smaller gt^nera nientary ealyx-Jike strnelure present below
the
is
The
(Miller, 1754; Trew, 1751). controversy has pistillate flowers in a few specit^s.
continued since, and almost 250 )ears later the is- Aiithnsteniinae consisi of two genera of tropical
sue as to whetluM- Euphorbia shiuild be recognized forest [revs: Anfhosfci7Ui A. Juss. disjunct
(1 s|>[). in
In its initial broad sense or be separated into many west tropical Africa and Madagascar) Dirhos-
atid
Wcu
smaller genera tcmma
still lias not resolved. Eiiphor- Pierre sp. west This
in troj)ical Africa).
(1
bia ami Kuphorbieae are generally considered tax- sid>tribe Is considered the leasl speciali/<'(l because
nnd
oiioiincniiy (lifficiill. a cdnsidtMal)!*' (]<'gn'r of of the perianth on both the starninate and
pistillate
uncertainly has always existed about the relation- flowers. Further, each individual iiuolucral bra<i
ships ol the groups within them. This probh^m is closely envelops a clust(T of many braclcoles and
ditc in great part to extreme moiphological di\('r- starninate fIow(^rs. The cvathia art^ bist^xiial or sta-
rmmber
sity, a large i)f species, and a subcosmo- minale. In Anthostema {\\v cyathia are arrang<'d in
politan distribution. Ftnv workers have been able condensed cymes. The
to axillary involucre
r»iade
Is
galn a com[»lcte understanding of the tribe through- up of four united bracts and zvgomorphie
slightly
is
out its immenst^ rang(\ and there has never been a because is spread In an ojumi half circle. There
it
universally accc[)t(nl classificalitm. This lack of are five large glands along involucral bract mar-
tiie
consistency has hindered and to some extent dis- gins and betwet^n the clusters of stamlnate
flow<'rs.
comaged
researih within the genus Euphorbia as The 3-locidar pistlllale flower not contaitUMJ
is in
well as the Kuphorbieae as a whole. Thus, a clear the involucre but instead lies at the base of the
imderstanding
of relatioii>hI|)s within the tribe is of involucre's open side. Rastnl on open involucral
its
great imi>ortanct* in ordtM* to provide subsiMiuent morphology and the possible position of the
]at<'ral
workers with a |)hylogenetic franunvork oti whiih |)istillate flower, this gemis probabi) most closely
t(»
bast* their studies. resembles
the ancestral inflorescence moq)hology
Kuphorbieae
Th<* tribe characterized by
is its of the Eui)li(»rbicac. In />^/r//as7cmmf/ the cvathia are
sjnapomorphic
pseudanthial inflorescence (termed arranged in loose, open, axillary or terminal cymes,
cyathinm) composed
a of a gland-bearing involucre In contrast to Anthostema, the four monochasla-
of.^cxeral united bracts and their associated flowers <*ontainIn": itnolucral bracts are
imite<l in a
rinir
and Kach
bracleoh^s. bract subt<^nds monochasial and
a the Involucre com[)lelely and
is i-loseil acti-
staminate infloresceiu-e, and these nionochasia sur- nomorphic. The four involucral glands are con-
The
roinnl a single pistillate flower. indi\idual flow- taiiUMl within this structure and attaclicd the
to in-
ers in Kuphorbieae are highly reduced atul repre- ner walls of the bracts and the base of ihe
seiitctl b\ a single stamt^n oxary. with or without gynophore.
<»r In the center of the inflor(\se(mce
is ei-
The
a p<'rianlh. slaminate flowers terminate slend(M- ther a minute pistillode or a 4-locular pistillate
and
l>edictds, the [)islillate flow<'rs can be long-|)ed- flower.
icellale or sid>sessile. This structure com- Kike Anthoslemina(\
is hi<rlilv the sublribe Neoguillau-
and
plex, there art* still doubts as its exact natuit* mlniinac also contains two genera: Calyropeplus
tt)
(5
and
evolution
(Gilbert. 1994). spp. in Australia) and Ncoguillauminia Croizal
(I
According
to the most recent synopsis Ku- Nt*w Cah-donia).
(»f tht^ sj). In In contrast to AnlhosttMiii-
phoibiaceae
(VJcbsler, 1991), the Euphi^rbieae con- nae, ihe starninate flowtM's lack a perianth, llow-
tain II gentM-a placed into subtribes: Anthos-
thret^ ever, a perianth pn^sent on
Is the pistillate flower,
It^minae (Haill.) G. K. Webster, Neoguillamniniinae and this combination of features defines the sub-
and Kuphorbiinae. The presence
Cr(»izat, or ab- tribe. In addition, the bracts of lh<M'n\olucre do not
sence of a ptMlanlh (piesnmabK a calw) on the tightU envelop the stamiriaie mouochasia. althoudi
flowers distitigiiishes iht^se taxa. In Anthosteminae the latter are enclosed within large braclcoles. Ca-
both Ihe slamiuale atul pistillate flowers possess a iycojtrjdus are xerophytie shrubs with small, oppo-
Volume Number Steinmann & 455
4 Porter
89,
Euphorbieae
2002 Phylogenetic Relationships
in
site leaves and cyathia arranged in axillary clusters. niosl fasrinalinii features of the sul)trilH^ is its "-reat
The involucre cupular and made up diversity of growth forms, l/dv^c. forest tices, shrubs,
is (hstiiictly is
and
of four bracts. The glands are located Ix^lvveen the p(*renuial herbs, geophytes, aiuuials, a great
lobes and allaclied to the wall of the cupular in- (Hversity of succulents are all well represented. Tn
vonicre There a central pistiUate flower sur- comparison to most other genera of Euphorbiinae,
hi is
rounded by four staminale monochasia that are op- le cyathial morphology of Kuphor^DiUa is nMhaintivvef^d y
tl
posite the iuv(tlucral lobes. The sole representative unspecializeil. The cyathia are actinomorphic and
of Neoguillanuiinia a mesic forest tree or shrub generally possess one to five separate glands situ-
is
arranged leaves. The cyathia are aled on the rim of the involuc-re, and this i)lesiom-
witli large. s|)irally
gemis.
long-pedunculate and arranged few-eyathiate ax- orphic feature uniti's the
in
iMary or subter'minal groups. The involucn' Is (H)m- With about .*5()() species, Cluimaesyce is llu^ larg-
from occurs wide-
posed of four six bracts that extend into large, est segregate g(^n us /^///>//o/-/>m. It
to
New
petaloid a[)pendages. There are eight to twelve ly, but most sp(^cies are coufuuMl to the World,
glands arrang(Hl in pairs between the f(»ur to six Distinguished on the basis of vegetative morphol-
many
staminate monochasia and attached base of ogy, cyathia are nearly identical to those of
to th(^ its
Euphorbia Agalonid House.
and sometimes also the base of the spt^cies of subg. (Haf.)
the involuci(^ to
(Ihiunaesyce characterized by many unusual syn-
yno[)hore. Th<'re a single central pistidate flow- is
tr is
Neogiiillduminld notewordiy bet:ause the in- a|)omort)liies: apiial abortion of the main shoot and
er. is
volucres are gcn(^rally composed of five bracts, and subsecjuent sympodial growth: interp<'tit)lar stip-
and
5-merous involucre also characterizes Euphorbi- ules; opposite, frequently asynmietrical leaves;
a
Numer-
(Koutnik, 1984, 19H7).
inae, disciiss(Ml below. Calyropeplus, as well as 4/?- C, photosyntliesis
ihostema and Dichostemma, j)ossess 1-merous in- ous Fnphorbieae specialists ((\g., Carler. 19<U]a,
subgenus
1W21); Gilbert, 1987) retain as a of
volucres.
it
W 2000 and subcosmopolitan
about species a Eiipliorhia.
ith
Euphorhia
distributi(»n. die largest and most complex subtribe riie other five segregate genera ol are
Euphorbiinae. characterized by the lack of also easily identifiable, but their differences, as
is It is
Three
and How- mentioned above, involve involucral features.
a |)erianlh on l>oth the staminate pistillate
zygomorphic
although rudimentan calvx-like structure getuMa possess cyathia: in /\'^/i/a/i//n^.'^
ers, a is
present below the pistillate flowers of a few species. (15 spp., primarily Mexico) there are tw(» to six
made up glands enclosed within an adaxial, spurlike exlen-
In addition, the involucre of five nnitetl
is
Culmnthus Cuba
bracts, not four as generally are found in the other sion of the involucre; in (-*? spp..
subtribes of Euphorbieae, and the bracteolcs diat and llisj)aniola) there are two glands united into a
surround the staminate m(mochasia are generally shieldlike structure on the outside of the Involucre;
reduced. The glands are mostly locaI(Ml along the and in Monadeniiim (ca. 70 spp., Africa) the glands
rim of a cupular involucre. Again following tlie cir- are united into a single horseslu)e-sliat)ed structure,
cutnscriplion of Webster (1994), the subtrilx^ con- The two nMuaiuing segregate genera possess acti-
m
Synadcnium
seven genera: Chanuiesyce Gray, Cubdnlhus nomorphic cyathia: (20 sj>p., Africa)
tains
Emlddcnium Leach, Euphorbid. the cyathia possess five united glands that form a
(I5oiss.) iVlillsp.,
Mofiadeniiun Pax, Pedilanlhus Necker ex Poil., and complete ring around the top, and in Endddrniuru
cyathium
Syfiddenium 8oiss. PoinscUia Graham and Elaeo' Angola) the closed rim of the is
s[).,
(1
Stapf an' also sometimes recognized as (lis- not a gland but instead au apparently eglandular
;^Aori/{/
from Euphorbid, but their status not as extension of the involucral wall with a ring of nec-
lincl is
widely accepted, and most authors treat diese as tar-bearing depressions on the inside of the invo-
Euphorbid. Extreme (^mphasis lucre.
infrageneric laxa of
placed on variation in the configuration of the
is
cyathium, and with the exception of Chaiudesyce, HlsroevY OK THK El lMK)i:iUEAE
TA\()N()\n(: Titnn-:
genera segiegated from Euphorbid are distin-
the
laxonomic works treating Euphorbieae are
'Yhv
guished on the basis of involucral features.
Euphorbiinae are dominated by Euphorbia, !un»ierous, and ordy a brief overview, focusing o!i
80% and those with broad and significant implications to
which accounts for ca. of the species oc-
modern Euphorbieae taxonomy, wdl be provided
curs throughout the geographic range of the sub-
common A more account found Stein-
The genus best known for the \\cn\ detailed is in
tribe. is
maun
Christmas poinsettia {Euphorbid pulcherrimd \^'illd. (2001).
prominent laxonomic treatments
ex Klotzsch) and popular with horticulturists ])e- ()n<' of the first
is
One was Klotzsch and (larcke
Euphori)icae of
cause of the ])revalence of succulents. of the of the that
s
456
Annals
the
of
Garden
Missouri Botanical
(IHriO, 1860). Altliouyli llu'se aulliors segrrgiUrd lluU of Pax and HofTiiiaiin (1931) Engler's Die
in
both Antliostema and
Pcdildntluis separate
into natiirUclien Pflanzenfamilien. 1liey ess(Mi(ialIy
f<»l-
and
trihes, llin An(lH»s!etn<\u' Pedila!illiea<% ifiey lowed Rentluuns elassifieation wiin only minor
ill I)
eoneede
did their elose relationship with Eu[)hor- niodilicatiuns aeeoninuMlate newly deserlhed
to
hieae. \\\v\v K.iiphorhieae contained 408 species, taxa. Again, a single the Euphorbiea(% was
tribe,
all nu^nibeM-s of prese^il-day Kuphorbiinae. Fuphor- recognized without any further divisions, Anlhostc-
hia was reeognl/ed in a restricted sens*% with only ma, DiclioslrmnKK and Calycoprphis were
Ireat-
all
27 The
speei(^s. remainder were placed 17 seg- ed Euphorbia
in as distinct. contiiuK^d recog-
to l>e
regale genera, organized in two sublribes: Aniso- nlztMJ in the broad sense, although the mirnber of
])hyllae Klotzsch Garcke and Tithymalae
S: segn^gate geniTa had increased. In adtlition to
t!ie
Klotzseh Garcke.
recognized and
earlier Pedilanlluis S)fia(lrn!uni,
<Ji
and monograph
'Ilu* firsl last of the (Millr*^ Ku- tlu^se authors segn^gated the gcMK^'a Monadcuium,
phorbi(\i(^ and the basic framework for the classi- Stcnadcniuni Pax, Elaeophorbia, and Diploiy-
fication that is still in use today, was providtMl by athiiun Schmidt. '\\w VmWv two an*
II. s(*gr(^gat(*s
1862
Hoissii^r In (supplenu^nt 1866) for de Can- now treated \s\\\\m Euphorbia (Webst*M\ 1991), and
dolle's Prodronius. Boissit^r followed Klotzsch and Stenadcnium synou>m
is currently treat<'d as a of
Careke
by recognizing Anlhost<Miieae as distinct, Monadrnium 1959, Cubanthus was
(Bally, 1901).
l)Ut differed from them by subnuMging Pedilantheae relegated back to a section oi Pedilotuhus.
740
into Fuphorbii^a*'. Euph<»rl)ieae coulaine<l spe- Tn 1937, Croizat descril>e(l the genus Ncoguil-
cies. Also in contrast to Klotzsch and (Garcke, Bois- launiinia and the subtribe Neoguillauminiinae
to
sier recogin'zed Euphorbia in a broad sense, and aeeommodatt^ The was
type, cicopatra,
fust
it. A^.
his Kui)h<»rbieae contained only three genera, all descri])ed by Haillon (1861) as a species of Eu-
eurrentl) [)laced in Eu[)horbiinae: Pedilanthus phorhia and placed monolypie Dr-
(1.5 in his E, sect.
Synadenlum
spp.), and Euphorbia radenia
(2 s[)p.), (72;? 8aill. Hoissier (18()2) treated this species
spp.). 11ie taxa oi Euphorbia wore positioned into within E. seel. Tithvnialu.s.
27 stHlions and two "series": A[»pendiculatae Shortly
after Croizat's contribution, heeler
\\
(conospondlng
Boiss. Klotzsch and Garcke
to s (1913) |)ub!isluMl a broad classification of the entire
ubtribt^ Aniso[)hyllae) and Exappendiculatae Kupliorbleae. This was major conspectus
the fust
Boiss. (eorn^sponding Klotzsch and Garckes
to to advocate the use of subgtMuis as the primary di-
subtribe TilhymalatO- <^V^/;co/vc/>//i,s and A^eog-zaV/au- vision i){ Eupliorlna, and is this rank that pre-
it
nnuia were
treatiMl in Euphorbia^ and Cuhanfhus dominates current Euphorbiinae
In classilicalion.
was W
(T(M'led as a section of Pcdtlanlhus. heeler recogmzItziMmjI a ssinniigiilee rKauinphnoorrbieae but
Benlham
(1878) expoiindtMl greatly Fu- abandoncd The
iui tht^ (^roizat's subtribes. gt^iera that he
phorbiac(Vie as a prelude to his treatment of the included were the same as tliose of Pax and
ll(»ff-
family Centra (Bentham. He maun
in Pla/itaruni 1880). (1931) except Diploryalhium was
that rele-
agreed dial Euphorbia should be r<M'ognized in the gattxl to synonymy within Euphorbia, and Croizat
broad sense and
fi)llo\V(H] Boissiers tnnitment widi newly describtul Neoguillauminia was recognized,
oidy minor modificalltms. Andiostemae was com- Euphorbia
eonsisl(Hl of eight subgiMiera, and
in
bined with Euphorbieae.
Calycopcplus, first de- general, these c"orres|)onded to tiie siotions n'cog-
srrlbed by Planchon 1861
in but subse(|uenlly nized by Pax and Iloffmami
(1931).
synonym
lreat<"d as a of Eitphorbia by Bolssier Dressier (19S7) pr<»\ided a mi)nograph of Pcdi-
was
(1862), resnrr(^ct(M] at the rank of genus. Ben- laulhus h\ which he convinclndy demonslrated thai
tluims change concerned
grt^atesl ihe rank of Bois- the sp<'clt\s o{ Cubanllins do not belong within Pcd-
sierV s<'<-tions of Eujdiarbia, and he proposed a sys- ilaulhus and should
treated separately. Dressier
l)e
t(Mn containing only six sections, under whicli also not<nvordi\ because he ad\ocaled Eu-
tlie is that
maj<iril) of Boissier's sections were loduced sub- phorbia shoidd be
h» recogniz<Ml in the narrow sense
sections. OlherwIs<\ Betilliam did modify and Old
little to restricted to Worl.i su<-culents.
grouping
Boissier's of species.
In the years after Diessl<M's treatm<Mit, \\ebster
Around
turn of the century, Pax made
llu^ (189'la) (l<- important contributions Euphorbieae
to tax-
Monadenium
scribed and Pierre (1896) described onomy. He suggested Chamae-
the recognition of
Duhoslcnima, second known genus mod-
\\\v. in the sycr as a genus and the same time strongly sug-
at
ern sublribi- AnthostemIna<\ two decades
N(^arly gest(ul that Elaeophorbia should be treated
williin
Cubauthus
later, IVlills[)augh (19i;5) first recognIz<*d Euphorbia (Webster, 1967). Eight years Web-
later,
as a dislinci genus.
circumscribed
sler (1975)
present.
lln' trilu^ to its
The
major
nt^xl treatment of Euphorbieae was generally accepKnl Me
lh<^ configuralion.
creal(Ml the
Volume Number 4 Steinmann & Porter 457
89,
Euphorbieae
2002 Phylogenetic Relationships
in
Aiilliosleminae acrcornmodale Anthoste- entire Eupliorhiaceae also supports iheir alfmity
sul>tril)e to
nia and Dichostcmma and resurrected N(M)j;uillau- (Kenneth Wurda<*k, pers. comin.). An alt<'nipl was
made
niiniinae include Calycopeplus and Neo^uillaU' to include as broad a sanipl<^ as possi[)le from
to
Four 4
suhtribe Euphorbiiiiae, Euphorl)it^a<\ S{)ecies y\nthost<'minae,
ruinia. In his third lie t)f
and 215
recognized seven genera: Euphorbia, Chanuicsyce, species of Neoguillauminiinae, spiH'ies of
Cuhdiithus, Etuliidenium, Mouddeninm, Pedilan- Fuphorbiinae sequenctnl. These included all
and Synddcnium. genera (except (Admnthus) recognized by Websti^r
//h/.s',
Endadanum,
The
(filbert (1987) and CarttM" (1985, 1988a) liave (1994). sole representative of
made the most recent mothfications in Knjihorbi- eight species of Pedilauthus^ one species of .Sjm/-
and
inae classification. Their changes primarily in\olve dcnium, eight species of Cluinidcsycc, five s|)e-
Mouddcnium were One hundred
African nKMubeis and the elevation of various pre- eies of inchuied.
viously recognized sections of Euphorhia to sub- ninety-two species ol Euphorbia were represented,
Bougiurn
generic rank. including all sections (exct*pt sects. IJoiss.
In summary, the time of this writing tnost au- and Cduhinthium Boiss.) treated l)y Boissi<»r (1862)
at
thors agree that Kuphorbieae is a single Iribt* with and most other imjiortant taxorKjmic grou[)s recog-
three sublrilx^s: Anthosteminae, the N(M)guillau- nized by subsecjuent work(Ms (e.g., Webstt^r, 1967;
tlu^
niiniinae, and the Euphorbiinae. Within Antlios- (iilbcrt, 1987; Carter, 1985, 1988a).
216
leminae there are two getiera, Anthostema and Di- he ITS analysis included species (4 out-
I
Anthosleminae, 2 Neoguillauminiinae,
ihostcmma. Within Neoguillauminiinae are groups, 3
thtM'e
also two genera, Calycopeplus and Neoguillaumi- and 207 Eu[)horbiinae). The sample for the ndhV
The was smaller and included 114 species
nid. subtribc Kui>horbiina(M'ontains ab(»ut sev- analysis (4
en genera. Those that are univtusally acee|>ted are oulgroups, 4 Anthostenunae, 3 Neoguillauminiinae,
Euphorbia, Etidddenium, Monadenium, Sy/iaden- and 103 Kupliorl)iinae). Following a ])reliminary
ium, Pedilaulhus. and Cubanlhus. Genera are anal\sis of the ITS sequence* data, a subsist of the
tliat
less frequentlv accepted are Chaniaescye. Elaeo- sampled taxa ri^Mesenliug the major clades and
was sequenced
and Other genera wcII-sup{)orted lineages for the
phorbia, Poinsettia. segregate*
have not even gairuxl minor acceptance. Within &/- adhV analysis. In addition, 15 species were includ-
ndhV
were
phorbia, the connnon [)rimary division emj>loyed is ed in the analysis itiat not includ(*d in
the rank of subgemis, and there are 9 to 11 gen- the TfS analysis because of problems ol>laining
recognized AgalonuL Cbarttarsyce ''clean" ITS seejuences for these taxa (s(*e A])pen-
crally <^\\\)gcm^ny.
Kaf. (when not treated as a genus), Poinscltia {Cra- di\
I),
DNA
ham) House (when Esula genomic was isolated from either
not treated as a geiuis). Total
PtMs., Eremophyton (Boiss.) F. C. Wheeler, Eu- fresh, silica gel-dried, or ht^rbarium material using
CTAB &
phorbid. Lanuilhis (Raf.) M. G. Gilbert, TirucalU a modified method (Doyle Doyh*, 1987).
Two
(H(»iss.) S. Cart(^r, Trichadenia (Pax) S. Ciuter; Rhi- g(Miic Regions were em|>loyed in the |»huvylioge
cpDNA
zautbium (Boiss.) L. C. Wheeler, and Lyriopsis nclic rtn'onstructions: the cofhng region
urDNA
(Boiss.) L. C. Wheeler. However, a modern, ghtbal. ndhV and tlie internal transcribed spacer
coherent classification of genus lacking, and region (ITS). ITS amplification using the polymer-
tlu' is
some proposed sections oi Euphorbia have not been ase chain reaction (PCH) followed tfie proee(hnes
cuncndy Baldwin and Baldwin
adecjuately cU:commodated within a rec- described by (1992) et al.
ndhV
ognized subgenus E. sects. Arlhrothamnus (1995). Amplificalion of the region generally
(e.g.,
& Garcke) Drnisophorbia (Lean- followed protocols described by Olmstead and
(Klotzsch Boiss.,
llie
Kim The 5'
and and Jansen
Croizat. and Deuteroadii Croizat). Detailed ac- Sw(mm(^ (1994) (1995).
(hi)
was excluded due
counts tluM'urrently recognized infrageneric taxa (|uarter of the ndliV n^gion to
<jf
50 bp
of Euphorbia and their taxonomic histories aie [)re- |)r<)blcnis amplifying it. Also, a [)rlmer ca.
AIT AIT
3' end '^21 (5'-TCA
sent(^d in the discussion section. internal to the lOKi'^
riA rCA A-3') was d(\signed ben-ause many
(J(/r
taxa would not am])lify using primer "21 lOK.'' Four
M
AM)
ATKHI ALS Mi'/mOlJS
were designed
additional prirn(Ts si)ecifically for
TTC AAT ATC TYT ATG
227 species were sam[)led (Appendix study: (5'-TTA
In total, 1), this (1)
GGG CCG CAT AKA GAT
including 4 outgroup and 223 ingroup species. Out- TAA-3'), (5'-TAA
(2)
GAA GAA C^T
were chosen from Hippomancae be- A'lT TAA-3'), (5'-TA(i 'ITC
gr<»ups tribe
(3)
KAA KGG
ATF
TYA ATC
and (5'-lTG
be AA-3'),
cause traditionally cot»sidered to closely
is (4)
it
AAT TCC
and
related to Fuphorbieae (Webster, 1994), a TA-3').
tlu^
The PCI{ products were ehxtrophoresed using a
broad molecular phylogen<4ic reconstruction the
of
1 )
458
Annals
the
of
Garden
Missouri Botanical
ITS and
Tiililc 1. Cliaiiiclcrlslics ol llic porlioii of the iidliV regions iiicliiilcd this stud).
iti
TTS ITSl 8 irs2
5. mill
J !•
Raw 59 660 210-267 161-167 202-24
(hp)
lcii};tli 16 5(
1 )(
I i
Aiifincd iciij^th 7:59 69 267 5M
;5();i
1 ]
Varialdc (pro|)ortion) 527 272 35(0.21) 220 710
sit<'s (0.71) (().<X)) (0.»2) (0.47)
Parsininnv-infonnatiNc
sites
(projwirtion) 470 (0.61) 244 (O.ai) 22 201 519
(0. (0.76) (il'M])
1 .{)
CC
mean
content, (mn-r) O.SH (0.5(MK7()) O.SH 40-0.7 0.56(0.51-0.58) 0.61 (0.18-0.76) (0.:i(MU3)
0.;^2
(0. 1)
l.S%
aj^rose gel in a TIIK (pH 8.3) huflei; a TI/TV step matrix, eitiployed rnaxinunTi
l).^^\ in fiirthcr
staino(l with cthidiuni l)r()ini(l(', and then cleaned jxifsinioiiy analyses. This weighting srlienie was
using
thi' 1M^(/ piecij>itati()n protocol (Nifkirnl. eniploye<l in (u<ltT to rTuulrl niotv clostdy niax-
th<"
Cycle-s(MjiMMicing adhered mamifac-
199()). to tlie iniiini |)arsinu)iiy analyses lo the givtMi data set. In-
PRISM®
turcr's spec ifn:at ion using (he Dye- dels vscre coded as nnssing data. Mtdtiph* most par-
Deoxy®
Ternnnat<»r Kit (Perkin-FJmer, Inc.). Cy- simonious were combined
trees
in a strict
ele-s(n]uencing was owe
d by ethanol
foil consensus Tree robnslness was
tree. estitnated us-
|>nririealioii. and seciuencing ustnl an Applied Rio-
ing 10,000 "last addition" henrisiic l)oolslrap
re[)-
DNA
M'^\
syst*Mns IVlod(d Anl(Mnat<'d Sequencing
licates.
em.
Syst
DNA
were
Sc(|ueiK"es ass(»tnbhM! from autotnated
RKsui;t\s
sequence ehromalograms using Sequencher 3.0
(Gene Codes
Corpoiation, ITS seipiences
Inc.).
SeqiK^nce \ariation for b<»th ITS and n(UiV
is
were itiilially aligtted with (llnstalW 1.4 (Thomn- t itc
v.
i * i t-
sliown Ml lali )li e i. lor IS setjnences, pairwise ilev-
T,,,,., I
son et al,., 1994), usi. ng a ga|» cust:gap extensi, on
"p"
els of div(Tgenee (uncorreet(Ml values) for the
cost rati(» of 10:.5, l(»llovved by visual mcxbhcalions;
entire Eu[)horbieae ranged from 34.4% (between
UiUiV se(juenees were aHgned Because
visually. of
Anthostcma
sp. nov. and Euphorbia iusuhiud)
to
number
divergence and
(he large of taxa
liigli in-
1.1% (between and
E. alia £. sjHilhulala), F(»r sub-
eluded in lh(* study, abgnnienl was problematic h)r
tribe Anlliost(^minae, levels of diverg(^nce
variinl
certain highly variable regions of ITS sequences.
9.4%
from (between DuJioslcnuun and
glaiicescens
How(*viw\ the difficulties niostlv oetanred ali^niim
Afilhos/cnai madagascaricnsc) 2.2% (between
(o A.
the major lineages of Fuphorbieae each
to other,
madagascariense and
sp. nov.); for ihe two in-
.1.
and alignment williin major lineages was less prob-
...
AT nv eluded species of subtiibe Neoguillanminiinae
r T^ ^
1lemati•e. Ahgrmu'nt, I ,1 IIS and ;/ * .
ol l>oth I ndliv nitro-
(Neogudlauminia cleopalra and Calycopeplus
cas-
„
(1l.u. c-cd1 g..aps i:.nt. o s(»me sequences, and1 *tiliese si•.tes J t t
. . .
uarinoidcs), the level of (liverg<Miee was 7.6%;
for
were
inchi<led the analyses. Missing data were
in
subtribe Euphorbiinae levels of divergence varied
coded mark
with a question in matrix. vMign-
llic
30.9% pamh-
from (between and
E. truholnrfja E.
ment maliiees have been
de])osited the librarv
at
E
aha
(between and spathu-
l-^^/^ E,
Raneho Ana »**
of Santa Botanic Cardt^n and snbmit-
i^^''"*'^^'-^'^*^0
As ndhV
lata). expected, levels of divergence for
TrcidJASE
led to (hi lp://lu^rl>aria. harvard.edu/
much
secpiences were lower than levels observed
Ireebase.index.html).
The ndhV '" ^^^ se(]uences. For the entire Fuplunbieae,
aligned ffS and sequt^nce matrices
ranged from Atuhostema
*'^^^^ .7%r (belw(MMi sp.
wen* analyzed separately using t.Obla 1 1
RAL4'^'^ for
oaxacana) 0.1% (between bdubala
M;ieiiilosh® (Swofford, 2000) on a Macintosh® '^**^'' ^^'^'' ^- to E,
(;3,
For
Due maximum ^- exstipidala). subtriln^ Ant!iost<Mninae,
to the larg*' data sets, parsimony '*»**'
2.6%
using heuristic searches (Aectran, 10 random ad- levels of diverginice vari(Ml from (belweeu/ln-
TBR
dition eyeh\s, branch swapjiing, steepest de- thostetna sp. nov. and DIchostrmma glaiiresrcns) to
0.4%
scent option not in eflect) was employi^d. Maximum (between Anthostcma sp. nov. and niada-
/I.
likelihooH estimates of transit iori/transv<M-sion (Tl/ gascarirfuse): for sublribe Neoguillauminiinae di-
2%
TV) biases were measured for both the rfS and verg(^ne(^ levels varied from (between Calyco-
fidhh data sets individuall) as implemented in pi'jdiis collinits and C. pamifalius) to 1.4% (between
PAl P* 4.0b4a under the IKY inofhd nucleotide Neoguillauminia cleopatra and Calycopeplus paa-
<if
I
and
subslilulion using the ecpial-vveightcd parsi- cifolius): for sublribe Euphorbiinae h^v(4s of
div(M'-
niony trees. These estimates wrro us(mI to produce genee varied from 9.6% (between aphylla and
E.
2
1
Volume Number Steinmann & 459
4 Porter
89,
Euphorbieae
2002 Phylogenetic Relationships
in
udhV
'ral)le 2. Statistics fruni ut-i^litcti inaxinuini parsimony analyses of ITS an(] dalascts anil statistics for un-
uci^litcd characlers of same Fur each analysis, nuniher and length of mo^t parsimonious trees, eonsist(Muy
llic to[)ol()<:;y.
index (CI), retention index (RI), and resealed consistency index (RC) are reported.
Nutnher
KC
Anal vsis of Irees Tree l(^n^ll» Ct Rl
0J219
2160 0.1813 0.1:^09
rrS (weit^hted) 883J5.2
60 (Ami 0.1802 0,7126 0.1281
ITS 2
(unutM^i^hted)
1
19012 18ia.2 0.522 0.7926 0.1158
(weighted)
/idliV
90 1985 0.5141 0.7992 0.1519
fiilhV (un\v<M*j;li(ed)
1
1
E, (xi.xacana) lo 0.1% (between F. hilohaUi and K. conducted for tiic 99 species in connnon between
ndhV The
ibe ITS and data sets. results aic not
exstipithiia).
'IVansition/transv torsion biases wore calculated lo presenlod here because results from the indepen-
udhV
1.8 for tbe XV^ data set and 0.85 for die tidhV dent ITS and anidyses agree stron<i;ly witli
b(^
data set. These biases were used in the weighted each oth(M-, and [he combitu'd analysis does lutt pio-
maximutn [)arsimony analyses. vide novel insights. For the instances in which tree
The ITS anal) sis resulted in the recoveiy of 21()0 structure n^sidting from the ITS ami udliV analyses
most parsimonious trees of 6887 steps (equal differed, th(^ combined analysis generally provided
ndhV
same
weighted). I1ic ndliV analysis n'sidted in die the structure as die analysis,
r(*-
1985
eovery of 19,012 most parsimonious trees of
steps (e(|ual weighted). Statistics (or both analyses Discission
The
are detailetl in Table 2. strict cons(Misus of the
am)
rnir> of ihk udliv irs kk(;ions r\
2160 trees obtained in the ITS analysis is de{)ictcd I
ACKAK AND COMPMilSON OK
I'HOnin A rilK
Kl
The
and
in Figures 1, 5, 7, 9, 10, 12, 13. slricl
niM.OGENIKS
UKSLlLriNi;
ndhV
19012
consensus of trees obtained the
tiie iti
and
analysis depicted in Figures 2. 6, 8. 11, 14.
is being one of the largest angios})erm fam-
l)es{)ite
W^ith the exception of Figures 3 and phylograms
4, and an important componenl. espc-
floristie
ilies
Stcinmami
are presented (2001).
in have
ciallv in tropical regions, the p]u|)horbiaccac
For the ITS analysis, the l)00tstrap 509r majority
been comprehensive
no! previously the subject of a
on consensus
ruh' tree (values dc[)icted strict tree)
niolecular systcmaties investigation. If ihe results
consensus
is siructurallv identical to die strict in- obtained here are any indication, both ndhV and
Many
sofar as the majority rule tree n^solv(ML of
is
the ITS data a|)pear useful inferring relationships
iti
the terminal clades are supported with high boot-
within the family. In general, the two reconstruc-
strap (bs) percentage values. However, there is no tions of the Fuithorbieae are similar, and thus cor-
majority of the basal internal nodes.
sup|)ort for The
tlic roborate evidence of th<Mr phylogenetic signal.
excepti(»ns are the ancestral nodes that de-
N(»tal>le same
overall structure of bodi analyses the (Figs.
is
fine the tribe Euphorl)ieae and its subtribes. In the and when differences do these gcner-
occiu".
1, 2),
analysis, the bootstrap 507^ majority rule tree
//r///F
^^^ involve clades that lack bootstrap sup])ort in
(values depicted on strict consensus tree in Fig, 2) There few
analyses. are ver)
^^^^ ^j. j^^^j], ,^f ^j^^
consensus
structurally identical to tbe strict with
is instances which well-supported clades an^ po-
in
one significant difference. In die bootstrap analysis,
^jtioned differently in the separate analyses, e.g.,
A
elade not placed as the sister taxon of clade &
is arrangement Euphorbia meemie
of
lihitt.
,|^^
B but instead located at the earliest diverging
is McCann, Thorm.,
E, nhdclkuri Balf.f., E. drupifeni
gnmp C and and
clade a that contains clades D.
in and Pax
E. polssonii (Figs. 10, 11).
77%.
supported by bootstrap value This
this a of
is
ma-
incongruency suggests that tbe t(>pology of the
AND
MUNOIMI^IA KlTMIOUtUKAK
ITS
Ol- riil-:
Euphorbiinae shouhl
jor clades within the subtribe
AND WITHIN
suirrmitK
siiBTRiT^Fs t>Atr\tMivi.Y
The
be viewed with caution. maj<»rily of clades
EUPHOUIiltNAE
consensus sup-
found in the ndliF strict tree are
Euphorbieae form motiophy-
ported in the bootstrap analysis, and in contrast to Iti ])oth analyses. a
ITS (see Fig. manv of internal- basal nodes letic assembhige with respect to the four outgroups.
tht^
1),
also possess significant bootstrap supj)ort. However, there is lutt high bootstrap support foi- this
The imusual synapo-
worth noting that a combined analysis was grouping in cither analysis.
is
It
460
Annals
the
of
Garden
Missouri Botanical
D
clade
(Widespread)
m
C
clade
82
(Widespread)
B
clade
>
(Primarily Laurasia)
tn
95
92
A
clade
(Africa and Madagascar)
2
O
Calycopeplus casuannoidcs
100 (Australia)
Z^
—
>
L-
CD
67 rr
i,
Ncoi^iiillaiim'mid vlvopatra
>
(New
Caledonia)
2
Anlhostemu mada^asamensc >
(Madauascar)
00
1
Authostema
sp. nov.
100
(Madagascar) 9
2
>
Dkhostemniii ^lauccscens
West
(Tropical
Africa)
Sapium
schiferum
100
OmaUmlhus C
popidifolius
O
Sehiistiama rornuta C
98
C/3
"
SlWWngla spinulosa
consensus
Kignr<» I. StricI of 2U)() In-cs hascd on a wcit^liU'tl niaxinuim parsiinuny analysis of ITS region.
tlic
50%
Koolslnip values giraU-r than are shown ahovo the liranchos. Tlic hihes and sul>lrih(^s of the classirication of
Wfl»sliT (1991) arc indlcalod. Irnhvidual sirirl consensuses for clades A-1) are found Figures and
in S. 7, 9, 10. 12, 1.'?.
morpliic iiidoresceru'r slrLiclure of the tribe there- Webster s overall elassifiealion of the tribe Eiiplior-
and
fore sLi|»porls this relalionstiip, it is Inghly bieae (1075) is eonsisleiit with tlu^ TJiolieeeeiiUir re-
probable' llial dn^ tribe is i?ideed m()no[)hy]<'tie. All sults.
W
three sul)lril)es of Kuphorbieae also form nionophy- ithin Ku[)lu)rbiiiiae, Ijoth aruilyses di-nionslrate
b^lic groups (T'igs. 1, 2). and these elad(^s possess that Cliinnarsyre, Pedilunf}}us, Monddcnium, Syn-
fiigh bo(»(s(ra|» support, l(>09r for all three in tlie adrnium, and Endadcnium are all nested williin a
ndhV
analysis, Kurtlu'r, In bodi analyses sublribe paraphyletic EupfiorbUi (see ela(h\s C and Kigs.
T).
Aiithosleiiiinae is the (^jrliesl diverging siiblribe in 9-1 I). Also, b<»tli Symtdcnium and Endiidrniuni are
Euphorbieae and
sublril)es Neoguillauniiriiinae and ne^sted within a paraphyletic MDnadenitun,
Euj)horbiinae are sister each Anfhosfema
to othiM".
rnn\
and Calycoprplus are stnmgly supported be iii:i--Vn(>.\Siurs A.Nl) MAJ(iK I.INI: \(;ks TIIK
\\
to
00%
niioiiiJUiNAK
monophyh^lie and
(bs ndh\\ i:i
](){y/( in res[)e<-
and presumably
tively), so too the monotypie
ar(^ l?oth analyses support diat four major
th<'rc ar<'
Diihostrmma and
ISvoiiuillanminia. Therefore,
lineag<*s within the sjublribe KuplKnbiinae, here
&
Volume Number 4 Steinmann Porter 461
89,
Euphorbieae
2002 Phylogenetic Relationships
in
74 D
clade
91
(Widespread)
11
C
75
C
clade
100 (Widespread)
86
B
clade
>
(Primarily Laurasia)
100
A
clade
100
1
and Madagascar)
(Africa
^^^^^^^
r
o
Culycopeplus collinus
o
90
(Australia)
r— dd
Calycopeplus paucifolius
o
100
(Australia)
deopatra
Ncoi^uilhiuminia
62
7.
(New > >
Caledonia)
Afffhostenui nHuhii^ascariense
78
(IVlada>:ascar) >
Antlioslcnia nov.
100 sp.
(Madagascar)
100
Anthostciua sene^aloisc
West
(Tropical Africa) >
Dichi)stc}}un(i i^laitcescens
West
(Tropical Africa)
Sapium
schifenuu
KM)
O
0}nahinthits popuUfolius H
O
O
cornuta
Sehdstiiuiia
99
spinulosci
Stilliiii^id
Fi^uiv 2. Stricl const'ii^us of 19,012 liocs based on a v\(m^IiI('(1 nm\iinuin f)arsimuny analysis of !lu' lulhV n^gion.
Bootslrap greater than 50%; are shown above brandies. The tribes and snhtribes of the ekissifieatinn of
\alut's liie
A-D
\^ebsler (1994) are indicated. Individual siriet eonsensnses for clades are found in Figures 6. 8, and 11.
I I .
designattMl (dade-s A-D. With (he ext;e|)li(m of chuh- Diic to the lack of a coniprehi-nsive^ olassifica-
ao
many Euphorbia have
B composed htMhaceous sections not l)een
(primarily of the temi)(n-at<' lion, o(
group of Euphorbia suhg. Esula). nunc of these lin- conunodatcd within a currently nn-ognized subg*^-
some have suggested
eages closely correspond any previonsl\ recog- mis. Also, sections that l)een
to
subgenus demonstrated
nizcd taxon. Instead, they are conglomt^ratcs of var- lo heh)ng to a particular an^
ious sections and sid)genera. A o f tliese nujjor [o im\ l>e ch>sely related. Therefore, the following
lineages possess signihcani bootstrap support in discussion aboul ihc major gnnips of FAiphorbiinat*
llie
A
analysis 74—100%), while only clade organized by laxa of various ranks.
ftdhY^ (l)s is is
ITS analysis (bs 92%). Tliere
supi)orted in still
th(^
lamaesyce
Clu unaesvce. Fight species un CI
some question, however, as to the exact relation- fr<
is
ndhV
ship among these Hneages. Although both the ITS were included. In both the ITS and analyses
and julhF consensus trees show that (dades (Figs. 12, 14), these sfiecies form a well-sup|)orted,
strict
91%
and
A and and logether diese monophyletic gronj) (bs 99%;, respectivi'-
are sister to eacdi otlier
1?
are sister lo clades C and D, there is no ])ootstrap ly) in clade I). Species previously thought lo inter-
with Chamacsyce (see WebsU^r, 1967), such
support to this grouping in eitlier analysis. g\\\{\c
462
Annals
of the
Garden
Missouri Botanical
H
D
^'^"'
' - '] clade
gucrichkma
E.
M
E. primuUfi C
''"""]
clade
E. tan^uahuete
E. (richotonui
B
I
clade
E. calyptrata
E. acalyphoid A
clade
J
£. antso
Calycopeplus casuarinoides
CT
Neoguillauminia
cleop<itra
Anihostema tnadagascariense
C
Anlhosttnui nov.
sp.
Dichosteftuna glaucesce/is
Sapium
sehiferum
—
Offuiianthits populifoiius
Seixistiama cornuta
{
Sliliingia spinulosa
10 changes
one
Ki^iiri' 3. Pli\l<)«];rarri of of llie eqiuilly iiiosi |unsiiti()iiioiis Irces rrsiilling from ihr ITS analysis, showing hraiuli
lengtlis williin Fuplittrhicac. Onlv tlu- sfXM-ies with {\w least and most clian^i^s arc shown for each of the major clatlos
A-O =
orsuI)lriln' Kiifihorhiiiuu'. Irnil\i(hial plnlo^rams (if cladcs arc ijiven in Steinniann (20(1 K. Eunhorhin.
1).
as Kuj>h()rl)i(i inuocua K. C. Wheeler and E. peper- Eiidadcnium iiesled inside Moiiiliold^ieallvn
it. ll»e
owioidcs Boiss.. share
a|>iK»ar to iheir vegetative genera are utiiled in the possession of fused invo-
diuMo
siinihnily eonv<Mg<'nre, as they do not group lueral glands. They belong to a weil-suppi»r(ecl
Chamaesycc
elose lo the chu]c (Figs. 9, 12, 13). clade C 78% ITS, KKK/r
(l>s in in ndliF) that also
Sitnilailv. E, rhnroh'pis Fiseh. C, A. Mey., sug-
contains Euphorbia suhg. Euphorbia and
(?v Licanthis.
Prokhanof
gesltMJ hy {V)V)) possibly
to l)e inter-
mediate between
(^hainarsycc aiul species of Pedihuilhns. y\s their unusual zygornorphie ev-
(»!!i<'r
Euphorbi<K is not support(Hl h) be closely related athia suggest, the eight s{)ecies o{ Pedilanlhus
in-
(Figs. 12, II). At huisl based on the molecular eluded heic form 100%
re- a \v(dl-support(Ml (bs
in
these
suits, putalivt* relalives an^ clearly (uMside both analyses), nionophyh^lic group
(Figs.
9, 11).
Chamarsyce, and
the charact<M\s present<'d in \hc Bast;d primarily on involucral appen<lages and
its
introduction can distinguish all true sptM-ies of predominantly Mexican distribution. Dressier
Chamarsyre. As
previously suggestcnl by Dressier (1957) and Webster (1967) suggested IWilan-
that
and Webster Chamarsyre
(1957) (1907), derived
is ihus aiose froin Euphorbia snbg. AgaJoma, with
E.
from Ag(doma
within snbg.
E. (Figs. 12,
14).
potential
/' clost'st
Sanipli ig of Chauuicsyce was not hioad
i-iiougli living intermcdiatt;. Howrvcr, this analysis docs not
lo d»-t.Tmin<- the motK.pliyly of the siil.seclions pro-
any helween
.su|)[)()rt ielatiuu.,liip these two groups.
j)ose(l l)y Roissier (IH()2). Il()\vever, it is noteworthy Tn fcicl, tlu'v do iiol cvfii Ix'Iong lo the same major
acnla and
that C. C. nnp^usla, both Chaiiiacsyre
iti lad
/
suhsecl. Aciitar Roiss.. lie sister lo the remainder
other menibtM-s of subgetuis 11-13)
/l^'-aA/m^/ (Figs.
of the speeies sampled IVom
ClidiiKicsycc suhseets.
belonging Beyond member
to elade D. being a of
Cymnadcmac
Chvlonvac CJumuwsy-
Boiss., Hoiss..
clade^Crthe
exact position of
Av/yA,/,//m,s- not
is
llyprnrijohar Chamarsyce
re, an.l Ro.ss.
snl.seet.
,(,,„„^,y ,„pp,,rt^H. In the ITS analysis
(Figs.
9,
10),
\rutar the only grou,. u.tlun Chamacsycv not
,s
^
ohj
,;,., ,;,,,.,. j^ ,.],j,, \Xorld taxa that
ii,^. ,,f
j)ossessu»g the deri\ed C^ photosyiUluMic pathway
includes (he Synadt^niuni alliance and subg. Eu-
E.
.
. /
olhtMvvise <'haraclens(ic of (WVbster
n^ °<j;(*rnis i^i e^ti ti
.,,_^ ...... t \ y i ph/ orh/ia andi uuanlh.i/ s. now<'ver, \.ihvve i• s no ibimt-
,
1975; Maydeld, 190|).
al.,
support ndhV
straj) for the relationship. Th(^ anal-
Synadenium/Mona<lenium/faidatleniuni. Al- y^i^ (Fig. I) sugg(\sls a r(dati<»nshi}) with £". data
1
lougnh t[lUutMMr composition differs between the ITS in a weakly su[)porteH rlad(* (bs 727f) of various
ll
and
/;f///F analyses, a total of six speeies from these Neotropical speeies. At least I)ased on biogeogra-
g<Miera, here relerriMl to as the Synadcniuni alii- [)hy, this association is more plausible. Involucral
ance, wrvr included. They form a well-suf)portcd, apj^eridages have arisen on various occasions
witli-
monophylelic group 100%
(bs in both analyses, see in Fni>horbiinae, and their Pcdilaruhns
i)r«*senee in
MS
ndhV
Huth and
I'igs. 10, 11). also suggest that api)aren(ly represents an independent
dci'I\alion of
Monadrtiium
paraj)hylt*tic with Syruidenium and
is this feature.