Table Of ContentTWO NEW
GENERA
OF
Hans Georg
Jens Rohwer,^ Richter^
G,
and Henk van der
NEOTROPICAL LAURACEAE Werff^
AND REMARKS ON
CRITICAL
THE GENERIC DELIMITATION
I
Abstract
Two genera, Chlorocardium and Paraia, are described, illustrated, and discussed. Chlorocardium includes two
Ocotea and known from Guyana and Surinam (C and Ecuador (C venenosum).
species previously placed in rodiei),
is
Paraia includes one species, P, bracteata, from the States of Para and Amazonas, Brazil. Problems with generic
placement of several other species are discussed. The usefulness of two-celled vs. four-celled stamens as a character
to define genera discussed, and examples of polyphyletic genera based solely on the two-celled condition are given.
is
&
Chlorocardium
The genera of the Lauraceae, with very few Rohwer, Richter van der
Chlorocardium
exceptions, are circumscribed by a combination of Werff, gen. nov. HOLOTYPE:
&
characters rather than by single exclusive char- rodiei (Schomb.) Rohwer, Richter van der
However, Kostermans (1957) noted that Werff.
acters.
more
"Apparently the combination of characters
is
Lauracearum
Ceteris generibus differt foliis oppositis,
and almost combinations are
or
less indefinite, all
plerumque
floribus exinvolucratis tetrameris vel irregu-
represented." Although statement must not be 3-5
this laribus, staminibus 12-20, verticillis dispositis, subae-
omnibus
taken because mathematical qualibus, quadrilocellatis latrorso-extrorsis, ses-
literally, in strictly
biglandulosis, ovario pubescente.
terms millions of combinations would be possible ^^'^^^^^^ ^'^^'
with the given characters, does emphasize the
it
variation within the family. Naturally,
reticulate
Chlorocardium Rohwer,
rodiei (Schomb.)
The more
not combinations are equally frequent.
all
&
Richter van der Werff, comb. nov.
frequent combinations correspond to the larger
Nectandra Schomb.
BASIONYM: rodiei in
genera, while rare combinations have been treated
Hook., London Bot. 616. 1844. type:
3:
J.
two different ways. They either have been placed
in
20-50
Guyana. mi. inland, along the rivers
own
genus Anaueria, Dicypel-
in a of their (e.g.,
Essequibo, Cuyuni, Demerara, Pomeroon,
Phyiloslemonodaphne, Vrhanodendron)^
or
liuni,
Schomhurgk 1004
K;
Berbice, (holotype,
iso-
they have been included as aberrant species one
in
1703
types, G, L; probable isotypes with no.
we
paper
of the larger genera. In the present de-
GZU).
B,
in
two new genera based on rare combinations
scribe
we some
of characters, and discuss of the discor- Nectandra leucantha var. rodiaei (Schomb.) Griseb.,
Fl.
W.
282. 1860.
dant elements in larger genera, as well as several Brit, I.
Ocotea rodiei (Schomb.) Mez, Jahrb. Konigl. Bot. Gart.
cases of tisfactory generic delimitation.
236. 1889.
Berlin 5:
The of our new genera has the most unusual
first
flower structure. Ironically, based on a species Chlorocardium rodiei a species of consider-
it is is
among wood
that normally would be considered to be able economic importance. Its (''Green-
the better known Lauraceae. heart,'' hence the name Chlorocardium) among
is
We NY
thank the curators of F, INPA, and for the loan of specimens and/or for providing duplicates. L. E.
'
Skog (US) provided additional flowers of Chlorocardium rodiei^ Kallunki (NY) provided the wood sample of Paraia
J.
hracteata, and John Myers made the habit drawing of Paraia bracteata.
D-2000 Hamburg Germany.
- Institut fur Allgemeine Botanik, Ohnhorststrasse 18, 52, Federal Republic of
^ Institut Holzbiologie and Holzschutz, Bundesforschungsanstalt Forst- und Holzwirtschaft, Leuschnerstrasse
fiir fiir
D-2050 Hanihurg Germany.
91, 80, Federal Republic of
Box 63166-0299,
^ Missouri Botanical Garden, P.O. 299, St. Louis, Missouri U.S.A.
388-400.
Ann, Missouri Bot. Card. 78: 1991.
Volume Number Rohwer 389
2
78. et
al.
New
Genera Lauraceae
1991 of Neotropical
D
mm
1
> i
—
Figure Chlorocardiam rodiei. A. Floral diagram of a regular flower with four whorls of stamens with four
1.
— —
^B.
stamens each. Flower, nearly tepals removed. C. Stamen of the whorl, seen from outside. D. Stamen
all first
—
Stamen
of the whorl, seen from side (note placement on the inside of the floral tube). E. of the third whorl,
first
— —
seen from outside. F. Stamen of the fourth whorl, seen from outside. G. Ovary. Voucher: Pipoly 7530.
the most important commercial timbers in the fam- the flower as trimerous, with nine fertile stamens,
We
Of
Although the species has been known for more could not find this condition. the 14 flowers
ily.
than 200 years, poorly represented in the that we examined only one appeared to be trim-
it is still
major herbaria, and very few flowering collections erous, but had 12 fertile stamens. Eight flowers
it
we
are known. This scarcity of flowers permitted a were clearly tetramerous, of these found:
Schom-
misinterpretation perpetuated until today.
with 4 tepals and 4 whorls of 4 stamens each
1
burgk mentioned in the original description that
2 with 8 tepals and 3 whorls of 4 stamens each
more
most flowers appeared to be abnormal, with
3 with 8 tepals and 4 whorls of 4 stamens each
number
than the usual of parts. Similar statements
2 with 8 tepals and 5 whorls of 4 stamens each.
can be found subsequent accounts on the
in
all
species. Until today everyone who dissected a flow- The remaining five flowers were irregular, usu-
number
er (including one of the authors) believed that he ally with less than the full of stamens in
One
when had
or she had picked a monstrosity the flower the inner whorls. of the irregular flowers
turned out to be tetramerous or irregular, yet no only six tepals, and another had eight tepals but
one wanted to dissect a second flower because there four whorls of five stamens each. Tepals with one
was so material. or two apparently fully developed pollen sacs were
little
Asitturnsout now, with more material available, found in both a regular and an irregular flower.
the only regular flowers are indeed tetramerous The stamens (Fig. IC-F) are almost equal, broadly
(Fig. lA), and irregular flowers are common in all sessile, and more or less tongue-shaped, each with
&
specimens. Bentley Trimen (1880) described two rather inconspicuous glands at the base and
390 Annals of the
Garden
Missouri Botanical
with four locules, of which the upper ones are leaves no doubt that the two species are extremely
lalrorse and the lower ones latrorse-extrorse. To- close. Chlorocardium venenosum is known only
ward the center of the flower the stamens become from the border region of Colombia and Ecuador,
we have Chlorocardium
smaller, and the inner whorls are inserted while seen material of
slightly
at successively lower levels within the floral tube. roc^«ei only from Guyana. It also occurs in Surhiam,
much
This configuration reminiscent of the closely but reports of a wider distribution (as in Mell
is
&
related Monimiaceae, but without parallel in the Brush, 1913) appear to have no foundation.
Lauraceae. may be note The structure of the secondary xylem of the two
interesting to that also
It
the native name Guyana shared with a Monimi- species described here under Chlorocardium
in is
is
The
aceae. reported to be ''Bebearu," ''Bibiru,'' highly unusual in two aspects: interconnecting
It is
—
or undoubtedly the pairs between vessels as well as between vessels
''Sipeira,'* *'Sipiri," ''Sipiru'' pit
same and parenchyma and parenchyma)
root from which Aublet (1775) derived the cells (rays axial
name (3-4
Siparuna. are extremely small diam.), a condition
//ni
Schomburgk reported that times of food short- absent not only from Nectandra and Ocotea, but
in
age the seeds of the Greenheart tree were used by from Lauraceae with the exception of the small
all
Neocinnamomum. More
way genus impor-
the natives in a similar to cassava, they Asiatic of
i.e.,
were grated, and the starch was repeatedly washed tantly, parenchyma cells are connected to adjacent
compound
to deprive of the bitter alkaloids. vessels through unilaterally pitting with
it
more
There some uncertainty concerning the spell- large, simple pits subtending two or of the
is
ing of the specific epithet. The species named small bordered pits in the vessel elements. These
is
Navy who parenchyma
for ''Dr.'' Rodie, a Royal physician first large pits in the cells are often tra-
described the medicinal uses of the plant. In the versed by branched or simple filiform protrusions
original publication the epithet spelled 'VoJ/o^uV of cell wall material. In addition, vessel elements
is
with the oe printed as one character. Sandwith in both species bear at the ends a partial or com-
(1939) suggested that was indistinguishable plete ring of large, windowlike pits that give the
this
from ae, and that the correct spelling should be impression of a crown or corona, hence the term
name Kostermans
'Vo(//aci," because the Rodie would be latin- ''coronated vessel elements'" used by
phenomenon
ized as "Ro(haeus." However, from the following et al. (1969) to describe this quite
text of description obvious that oe and ae unique in xylem anatomy.
tlie is
it
The
are indeed diiTerent characters. Since the spelling other aspect considered here the complete
is
with oe lacks any foundation, we suggest that [in lack of secretory cells in the secondary xylem.
What
accordance with Art. 73.10. of the Coc/e (Greuter, might have been a chance observation for
1988)] should be corrected as has Chlorocardium venenosum described from only
to 'Vori/et,''
it
been done by Mcz (1889) and most subsequent one specimen so far (Kostermans et 1969)
al.,
becomes The
authors. significant for C. rodiei: countless
The secondary xylem
second species of the genus C/i/ororart^/um descriptions dealing with the of
poorly known: taxon have not once revealed the presence of
this
is
Thus
secretory absence, although gen-
their
cells.
„, erally not diagnostic in ,Lau1raceae, mieht justly be
,, .,. o o
- venenosum rr>-mk-
i^hlorocardiuiTi (Kosterm. I ^
cv /. j
. . 1
„ ^T uscd lu cou uiictiou With otHcr charactcrs to seri-
&o
, . Koh1 wcr, rK>ich1 tcr van d1er Weriirr, combi.
ley) -^
ously question the traditional taxonomic affinities
&
nov. BASIONYM: Ocotea venenosa Kosterm.
two
of these species.
Mus. Harvard
Pinkley, Bot. Leaf?. Univ. 22:
These findings are corroborated by the equally
241. 1969. TYPE: Colombia. Putumayo: Rio
unusual structure of the secondary phloem of C.
77°05'W,
Giiamues, Santa Rosa, 1,060
alt. ft.,
rodiei (for detailed description see Richter, 1981),
00°19'N, 26 Nov. 1966, Pinkley 555
(ho-
which characterized by special features different
is
GH; ECON,
lotype, isotypes, S, U).
from the prevailing structural patterns in Laura-
In the original description of this species the ceae, among them the complete lack of secretory
buds
authors state that ''Most of the flower avail- cells (sic).
able for dissection were abnormal as a result of Thus, the wood and bark structure supports a
fungal attack.'* In fact, they are so few and im- rather isolated position for Chlorocardium, but
mature on the type specimen that does not look unlike the flower structure does not suggest an
it it
worthwhile to dissect any of them. The general affinity with the Monimiaceae.
aspect of the plant agrees very well with Cliloro- Fruits of Chlorocardium rodiei have been de-
cm cm
cardium rodiei^ and the wood structure described scribed as large (6-7 long, 5 diam.) and
——
Volume Number 2 Rohwer 391
78, et
al.
New
1991 Genera Neotropical Lauraceae
of
mm
1
mm
1
G
I
— — —
new
Figure Undescribed genus. A, Flower. B. Flower, tepals removed. C. Receptacle with ovary, outer
2.
—
^D.
stamens removed. Stamen of the whorl, seen from outside. E. Stamen of the whorl, seen from inside.
first first
—
F. Stamen of the first whorl, seen from side (curved inward). G. Stamen of the second whorl, seen from outside.
— —
H. Stamen of the second whorl, seen from inside. I. Stamen of the second whorl, seen from side. J. Ovary, its
position the receptacle schematically shown. Voucher: Vds(^uez 6020.
in
With
ellipsoid to ovoid-globose, seated in a shallow, sin- shallow cupule. the exception of the recently
cm
gle-rnargined cupule, 2-2.5 diam. (Mez, 1889; described Ocotea megacarpa^ of the two
fruit size
A
Kostermans, 1936). recently collected specimen Chlorocardium species far exceeds that of any
MO)
{Pipoly 8616^ shows a not very deep, Ocotea or Nectandra species.
len-
cm
ticellate cupule^ ca. 1.5 diam. and a clearly
we
Recently found another species with opposite
cm
and
exserted, ovoid ca. 1.5 long. Fruits
fruit,
and
leaves variable staminal configuration that
in
cupule are attached to the twigs. The tree was
still
some Chlorocardium
characters resembles but also
and
flowering at the time of collection, possible
it is
shows marked differences Like other
(Fig. in
2). all
were immature.
that the fruits Fruits of C. twne-
Lauraceae Chlorocardium)
neotropical (except the
nosum
were described as depressed-globose, 4.5
and
flowers are trimerous, invariably with six tepals
cm cm
high and 5.5 diam., almost completely
more The
not than three whorls of stamens.
fertile
A
enclosed by the cupule. recent collection with
many
which Lauraceae
fourth whorl, stami-
in
is
fallen fruits agrees very well with this description;
seemed
nodial, to be represented only in the frag-
cm cm
the are up 6 high and 7 diam.,
fruits to
ments of one broken flower. In 25 reasonably intact
the cupule largely encloses the leaving an
fruit,
we
examined found
flowers the following configu-
cm
orifice of about 2.5 diam. Fruits are used by
rations:
Kofan make arrow The
the Indians to poison. very
large cupule of C. venenosum readily separates 10 flowers with whorl and 4-locular, whorl
I II
much
from
this species C. rodiei with its smaller, III 2-locular
392 Annals the
of
Garden
Missouri Botanical
6-11
8 flowers with whorl and 4-locular, whorl utroque costae latere e nervo medio sub angulo
II
I
40-80° gemmae
prodientibus. Inflorescentiae e terminali
absent
III
squamarum quam
prodientes, anguste thyrsoides,
axillis
4 flowers with three whorls 4-locular
all
cm
multo breviores (3-11 longae, pedunculis 1.2-
folia
2 flowers irregular cm
8 dense ferrugineo-tomentosae. Flores breviter
longis),
ramu-
pedicellati vel subsessiles in apice inflorescentiae
±
lorum glomerati, bracteis conspicuis subtenti, plerum-
stamens upper
In the 2-locular the pair of
it is que tubo receptaculi conspicuo. Perianthium flore plena
mm
locules that lost. The lower pair, like in Chlo- subglobosum, haud patens, 2-3.5 diametro. Tepala
is
suborbiculares, basi crassa, margine hyalino, externe basi
rocardium, latrorse-extrorse stamens; the
in
all
is
media tomento
et linea ferrugineo, interne epapillosa gla-
upper pair can be latrorse or lat -introrse. Like mm
braque. Stamina novem, ca. 0.8-1.5 longa, epapil-
Pleurothyriam the open toward the abaxial
in flaps
i^sa' quadrilocellata vel serie prima raro bilocellata, sex
side of the stamens. The outer stamens (especially exteriora eglandulosa filamentis gracilibus locellis subap-
whorl 2D-F) are bent inward, overarching icalibus arcuatim vel linea unica juxtapositis, tria interiora
Fig.
I,
111
basi biglandulosa filamentis crassiusculis locellis subapicali-
tbp innpr ^!;^mpn^ All bsitd;iiimpindci;o liivp di^tin^t hurrooadrilj
lilt; iiiiier i^iciiiidib* /\ii iicivt> iiioiiiici^ o
Ji ateralii-ibus, htami*nodii-a parva, subulata, dense tomentosa.
The
filaments, and there no trace ofglands. large
is
Receptaculi tubus profundus, saltem longitudine stami-
ovary is inserted in a deep, cup-shaped receptacle. num, angustus, intus dense tomentosus. Pistillum glabrum,
mm
All parts of the flower, including the ovary and the elongatum, ca. 2-2.4 longum, stylo ovarium ae-
q^^^te vel eo paulo longiore. Cupula obconica vel sub-
inside of the receptacle, are covered with papillae.
mm mm
12-15 6-8
hemisphaerica, diametro, profunda,
r
rr.1 . • • • • • •
Ihis combIination ch1 aracters unique withI in
ot
is ^ uiupli-ic•imargm• at.a, margme ext*eri• ore ilongi• ore * periant*unn
et
the family, and we have no doubt that our material
lobis incrassatis manifeste sexlobato. Bacca ellipsoidea vel
mm mm
new However, 14-24 10-12
represents a genus. as our obser- plerumque ovoidea, longa
ca. et
diametro.
vations are based on only one collection Vds-
(/?.
quez 6020), we postpone the formal description m
Shrub 8
or small tree, usually to ca. (only
tall
more becomes
until material available. m
2220
Silva from a tree 20 Twigs round,
tall).
The new we
second genus that describe in^this
0.4-
^^^ youngest with a dense cover of long (ca.
paper unusual staminal configu- mm
is far less in its ±
long), straight to curled, erect, relatively
1
ration. Yet combination of characters does not
its
brown becoming
thick, reddish hairs, quickly
felt-
known
any
of the genera.
fit
worn Terminal buds covered by
like, easily off.
mm)
(4-7 somewhat
relatively large scales, spread-
&
Paraia Rohwer, Richter van der Werff, gen.
^^^^ ^-^^ indument
-^^^ ^-^^^ ^f as the twigs
^j^^
nov. HOLOTYPE: Paraia bracteata Rohwer,
on outside (though not always dense at the mar-
&
Richter van der WerfF.
glabrous on Leaves drying with a
gins), inside.
characteristic grayish green to olive-brown color,
Lauracearum
Ceteris generibus apice
differt foliis in
ramulorum subverticillatim congestis, floribus subglobosis alternate but with few exceptions clustered at the
bracteis conspicuis plerumque numerosis subtentis, tepalis of the branchlets (sometimes a second, older
tip
margine
hyalinis interne
epap^^^^^^^^ 14.35 ^^ 5.12.5 cm
^^^^^j ^^.j, ^^.^^^^^
,
novem, anthens quadnlocellatis loceihs arcuatim juxta-
o r\
-i
^'^^^ ^^^ 2r».3-4(/-4.5) times longer t1han wide, ob1 -
positis, ovario bene evoluto in tubo profundo, cupula du-
lanceolate, obovate or oblong, widest the middle
margine in
plicimarginata exteriore longiore et perianthii
segmentis auctis sexlobato. or mostly slightly above, toward the apex acute to
acumen,
obtuse, the tip itself with a distinct base
&
Paraia bracteata Rohwer, van ±
Richter der
mostly long, cuneate or sometimes acute (at the
Mun.
Werff, sp, nov. TYPE: Brazil, Para: Orixi- very base occasionally abruptly obtuse), lateral veins
mina, Rio Trombetas, right bank, Porto Trom- 6-11 on each side of the midrib, diverging at 40-
km
betas, road to Minera^ao Rio Norte, 60, 80°, midrib and secondaries above ± im-
slightly
beyond Bauxite mine, 29 Aug. 1980
pressed but the vein convex, below prominent,
(fl, fr), itself
Cidetal MO; HBG,
7S90(holotype,
isotypes, above
tertiary reticulation alrtiost invisible to im-
INPA 96055
NY).
not seen, Figures 3, 4. pressed and slightly convex, below slightly raised
Indument
to prominent. of leaves consisting of the
m
Frutex vel arbor ad 8 (20 m?) alta. Hamuli teretes,
same on even
kind of hairs as the twigs, in the
gemmaeque
hornotini dense ferrugineo tomentosi. Folia
513 mm
youngest leaves above sparse, sometimes dense on
petiolis crassiusculis longis canaliculatis, fere
ramulorum
onuiia apice congesta, oblanceolata vel obova- midrib, below moderately sparse and highly un-
14-35 cm 5-12.5 cm
ta vel oblonga, adulta longa, lata, ^•^^1. 11
distributed, often dense on midrib, or some-
apice ex acuto vel obtuso manifeste acuminata, basi cu-
t™^^ subglabrous fr rom t1he 1beg•mnm• g; mature 1leaves
neata vel rarius acuta (basi ipso interdum obtusa), in sicco
± entirely glabrous or below often with rudiments of
olivacea, adulta supra opaca vel nitidula glabra, subtus
peropaca subglabra, margine piano, nervis secundariis the indument* Inflorescences springing from the
Volume Number 2 Rohwer et 393
78, al.
New
Genera Lauraceae
Neotropical
1991 of
E
£
—
— —
^B.
Figure Paraia bracteata. A. Habit. Fruit. C. Cupule, cut open, showing the double margin. D.
3.
Detail of leaf base. Vouchers: A, D, Cid 1890; B, C, A. S. Silva et al. 72.
cm more numerous
scales of the terminal bud, 2-8(-l long with bracts, which are often than the
1)
a peduncle of L5-5.5(-8) cm, narrowly thyrsoid, flowers but (at least partly) fall off during anthesis.
cm 2-3.5 mmdiam.,
the branches not more than long Perianth subglobose, ca. opening
lateral (at
1
anthesis; often enlarged afterward). Flowers ap- only slightly during anthesis. Tepals 6, suborbicu-
parently bisexual (but see below), sessile or with a lar, fleshy at the base, hyalinous toward the mar-
short pedicel (to mm), clustered at the tip of the gins, only on the outside at the base and in the
1
inflorescence branches, subtended by relatively large middle with reddish hairs, otherwise glabrous, with-
394 Annals of the
Garden
Missouri Botanical
mm
2
y
mm
1
F
—
Figure Paraia hracteata. A. Flower, globular form.^B. Flower, tepals removed. C. Stamen of the
4. first
—
whorl (of A), seen from inside.— D. Stamen of the first whorl, seen from side. E. Stamen of the second whorl, seen
—
—
from inside. F. Stamen of the third whorl, seen from outside. G. Receptacle with ovary, staminode and two inner
—
—
stamens, cut open, tepals and outer stamens removed. H. Flower, elongate form. Stamen of the whorl (of
I. first
—
H), 2-locular form, with a rudimentary third locule, seen from inside. J. Stamen of the second whorl, seen from
—
inside. K. Stamen of the second whorl, seen from side. Vouchers: A-G, Cid 1890; H-K, Ducke s.n.
out any papillae. Stamens 9, 4-thecous or in the pals (which, however, seem to break off easily).
mm
1424
first whorl rarely 2-thecous, without papillae, the Berry ellipsoid to ovoid, long, 10 12
mm mm
0.8-
outer six 1.4 long, their slender filaments diam., often with truncate
tip.
mm
0.50.8
long, at least at the base and on
Additional specujiens examined. BRAZIL. AMAZONAS:
some
adaxial side with hairs, the anth wiaer
id
Manaus-Itacoatiara Highway, Colonia Santo Antonio, 27
m
than long, thecae subapical, introrse, arranged ^
q^^ ^g^^^ ^//^^ 330 {UBG. NY); Rio
(f^j^ i f^^^/j^^^
an arc or (especially in the first whorl) almost in a Cuieiras Manaus km 6 Apr. 1974 Campbell
trail, 1, (fl),
stamens 0.8-1.6 Pra^ct- 27907 (HBG, MO, NY); Manaus, Estrada
horizontal line, the inner three 11 e/ a/, in
mm
long, their filaments thick, only slightly nar- ^^ ^^""'^^ (^^^^ ^^ M^^^)' ^^^- ^V^ ^'^^' ^' ^"""^^'''^ ''''
^
n. no mm (HBG, INPA 46534), near IgarapeMindu, 12 July 1929
me
rower an . U.4-U.O
th tner, lone, at least n
tl ^n\ o c
' (n), uurk/ e s.n. (/LtJt)\, ali ong roadi *to Aaliei• xo, iiz Aaug.-li hep.
.
laterally hairy, each with a pair of conspicuous
1936 Krukoff7996 (F, NY); Rio Cuieiras, 2 kin
(fr),
glands at base, the thecae subapical in two below mouth of Rio Brancinho, Sep. 1973 France
its 1 1 (fl),
ai 77749 (HBG, NY); Manaus, Igarape do Rinda, 19
same
lateral pairs, almost at the level (the adaxial et
=
3466
ones .Hghtly higher than the abaxial ones). Stami-
..^^^^^^^^
P„^i,^?^/,i^^' .f,«/^ifv^^^f km
INPA
(HBG, 10016, NY); ManauH-Caracarai'
road,
mm
0.30.6
nodia three, long, subulate, very hairy. & L
3317
20 1961 Rodrigues Coelho
19, Sep.
(fr),
mm
Receptacular tube 1.2 2 deep, narrow, dense- (NY), para: Belem, Agronomico do Norte, 3
Instituto
mm
ly hairy inside. Pistil slender, ca. 2-3.4 long, Aug. 1962 (fl juv.), Allen 79 (NY); Mun. Orixiniina, Rio
Troinheta., bank, Lago Moura, 22 Aug. 1980
style as long as the ovary or slightly longer. Cuuule left at (fl),
„ ^ Cid ai 1825 (HGB, MO); Lago Cunari, Planalto de
-^ .\ o mm et
, , . , - , - , r
broadly obcomcal subhemisphencal, 1 - 1
to 5
1 2 SPVEA,
1 gantarem, area of forest inventory by IAN, and
mm
-8
double-nmmed,
diam., 6 deep, distmctly
pAO, 15 Apr. 1955 FrSes 31668 {NY); Mocambo,
(fl),
EMBRAPA km
the outer margin longer, crowned by enlarged te- Forest Reserve, ca. 10 from Belem, 13
Volume Number Rohwer 395
2
78, et
al.
New
1991 Genera of Neotropical Lauraceae
Nov. 1984 (st), Gentry 48956 (MO), 10 Oct. 1966 (fr), posits and silica absent. Oil cells common, associ-
&
Pires N. Silva 10245 (HBG), 28 Sep. 1967
T. (fr ^^^j ^j^^ ^xial and ray parenchyma; very large
&
10947 1967 /
imm.), Pires N. Silva (NY), 2 Sep.
j&^N T T. moAQi^^w ^^,^,^ d,i. amcters up *to -7t0a txm andj axi•ali extensi• ons
Ana
(^.\ PPi;r.e.sc N. S<i^ilhv.an 10949 (NY), 3-^ Aug. 1iQ9?7"3^ (^\
(fr), T. (fl), mm
Pires 13219 (HBG, NY); Mun. Almerim, Estrada do up to 0.35 (associated with marginal ray cells)
mm
Pedr5o, km 40, 12 Nov. 1978 Santos 284 and 0.70 (associated with parenchyma strands).
(fr), (F);
AMZA
km
Serra dos Carajas, 13 from headquarters, road
18 1977 72 (HBG,
to sawmill, Oct. (fr), A, S. Silva et al.
km BARK
DESCRIPTION
MO); Santar^m, Estrada do Palhao, 35, near Igarape
&
Curupira camp, 24 Aug. 1969 M, Silva Souza
(fr),
Secondary phloem ground composed
tissue of
2402 (HBG, MO, Monte
NY); Rio Dourado, Planalto
Jarl,
B., between Pilao and Repartimento, 28 Oct. 1968 conducting (sieve and companion) cells and phloem
(fr),
T
Silva 1328 (HBG, NY); road from Bandeira parenchyma. numerous
A^. Jari, Interspersed are oval or
km
to Pil5o 39 18 June 1969 imm.), Silva
fr A^. T.
(fl, diamond-shaped, extended
rarely tangentially scler-
2220 (HBG, NY).
n
r
ei- di i- silandis consi• st• mg oi verti• cally eli ongate, coli um-
nar sclereids. Secondary phloem fibers are absent.
WOOD
DESCRIPTION
Toward the periphery the secondary phloem
is
The wood and bark descriptions are based on delimited by a continuous ring of small, irregularly
The specimen comes from shaped and
the type collection only. largely isodiametric sclereids enclosing
m
a small tree ca. 7 The wood sample has an few and widely spaced groups of primary phloem
tall.
average diameter of 7 cm, covered by a grayish fibers. The sclereid ring followed by cortical
is
mm
Wood
brown bark ca. 3 thick. and bark are parenchyma gradually phasing into several rows
slightly aromatic. of rectangular, radially oriented phelloderm cells
gum
(replete with dark colored deposits or scleri-
Wood
General. uniformly yellowish brown;
and a narrow (2-3 rows of similar but empty
fied)
heartwood probably dark brown as indicated by
terminal phellem. Just inside the sclereid ring
cells)
the presence of darker colored patches around a
few
isolated radially flattened sclereids are present.
damaged
zone by insect larvae; evenly textured,
Phloem rays resemble wood rays dimensions and
in
dense and heavy (specific gravity ca. 0.85/cm^).
composition. Their radial course through the sec-
Wood
Anatomy, diffuse porous throughout, ondary phloem ragged due to the spatial rear-
is
pores evenly distributed, thin-walled and angular rangement of other constituents following the early
2-5
in outline; solitary and radial multiples of in collapse of conducting tissues. Ray cells become
100-150
when
variable proportions; largest pores fim sclerified traversing sclereid islands. Oil cells
15-22/mm^.
diam., frequency Vessel perforations are scattered throughout the secondary phloem and
parenchyma.
exclusively simple; intervascular pits invariably al- the cortical Crystalline inclusions are
ternate, mostly circular in outline and not crowded, deposited in ray and axial parenchyma in consid-
9-1
pit apertures included; pits 1 ixm diam. Tyloses erablc quantities, usually as minute acicular crys-
common, no
thin-walled to sporadically sclerotic; tals.
gum medium The
deposits observed. Fibers libriform, of Note. sclereid island constituents (colum-
more
wall thickness; regularly septate; generally ca. 15 nar sclereids) are or less rectangular in cross
jum diam., rarely up to 20 ixm; pits very small and section, with the larger side extending in radial
The narrow have
Inconspicuous, mostly confined to radial walls; direction. tangential faces pointed
marked The
growth increments distinct microscopically, ends, thus resembling fibers in tangential view.
by 1-2-seriate tangential bands of wide-bodied polylamellate structure of the cell wall, the distinct
fi-
bers (up to 30 fim diam.); no imperforate tracheary pit canals, and the type of lignification exposed by
Parenchyma
elements observed. basically paratra- differential staining, however, betray their probably
cheal sparse to irregularly vasicentric; rather in- parenchyma-derived sclerenchymatic nature.
conspicuous unless interspersed with large Parala bracteata was recognized as a new
idio- first
blasts (oil cells) as part of the vasicentric sheath; species by Beulah Coe-Teixeira. She annotated sev-
4-6
vertical strands composed of individual eral specimens as Nectandra paraensis Coe-Tei-
cells;
pits to vessels enlarged, oval to gashlike and ex- xeira, but she never published the name. While
Wood
tended horizontally. rays predominantly working toward a revision of the genus Nectandra^
bi-
seriate, heterogeneous with 1-2, rarely up to 4, Rohwer found that this taxon has nothing in
J.
common
marginal rows of upright and/or square height with Nectandra except the technical
cells;
mm;
0.5-0.8
of largest rays vessel ray pits en- character of "four locules of the anther arranged
brown an There no Nectan-
larged, oval to irregular or gashlike; reddish in arc." are other species of
gum
deposits sometimes present. Crystalline de- dra with the leaves distinctly crowded at the tip
396 Annals of the
Garden
Missouri Botanical
2-thecous Paraia tendency toward
of the branchlets, with hyalinous tepals entirely in indicates a
were reduced
lacking papillae, or with clearly double-rimmed cu- reduction. If the th entirely in
pules. Also, the general aspect of the plant, the outer whorls, the resulting staminodia would
its
Licara
flowers or flower parts, do not suggest Nectandra. be quite similar to those of several species.
and widened apex,
Sterile material often annotated as Aniba, which with a slender filament a flat,
is
more
be quite similar. Vegetatively and even in Thus, the possibility that Paraia represents a
its
inflorescences Paraia bracteata strikingly resem- primitive offshoot of the line leading to Licaria
Rhodostemonodaphne miranda (Sandw.) cannot be ruled out.
bles
When
Rohwer, but this similarity does not extend to the looked at individually, the flowers of
Paraia bracteata appear be invariably bisexual.
inner organization of the flower. After discussing to
when
new we However, flowers from several different spec-
the possible relationships of this species,
concluded that we would have to describe as a imens are dissected, seems that there are two
it
it
new genus. For every known genus we could find types of flowers, diff"ering in proportions. In the
three important reasons why should not type (Fig. 4A-G), the flowers are nearly glob-
at least first
it
from
the receptacular tube not being apparent
be included there. ular,
Judged by rather commonplace lauraceous the outside. The length of the stamens is approx-
its
wood Paraia bracteata cannot be imately equivalent to the depth of the receptacle,
structure alone,
4H-L)
safely diff'erentiated from but a few neotropical In the second type (Fig. the stamens, al-
all
L Mezilaurus, Clinostemon, Car- though apparently fertile, are noticeably smaller.
e.g.,
,
The
yodaphnopsis. Nearly and negative receptacular tube deeper, almost twice as
positive is
all
from
affirmative characters circumscribing the prevail- long as the stamens, visible also the outside.
The more same
ing structural patterns of Lauraceae are present. ovary, however, is or less the size
In combination with the very specific bark struc- in both types. It is possible that the diff"erent pro-
ture, however, most neotropical genera of this fam- portions indicate an incipient separation of the sex-
can be excluded as possible matches. Very good es. It may, on the other hand, be just an expression
ily
Such
structural agreement can be found only in a small of geographic diff'erentiation. a differentiation
group of taxa comprising most species of Aniba obvious in leaf shape: specimens from the sur-
is
much
Manaus
and Licaria, and a few species currently ascribed roundings of have relatively nar-
When
to Ocotea, taking into account the less di- rower leaves than those from the state of Para,
wood
agnostic but nevertheless helpful general
haracters such as color and density, the scope
can be narrowed down further to the few species DISCUSSION OF GENERIC DELIMITATION IN
constituting subg. Canella of Licaria (Kurz, 1982; THE Lauraceae
and two
Richter, 1985), strikingly similar species
OCOTEA RUBRA
Aniba caneliUa (HBK) Mez, A. ferrea Ku-
of {A.
A
Even without supporting evidence from ex- highly unusual combination of characters
bitzki). is
known
omorphic characters, an attribution to the genus also found in a species as Ocotea rubra
more
Nectandra, as proposed by Coe-Teixeira, Mez. This species has or less obovate, gla-
initially
can definitely be ruled out. brous leaves, crowded at the tip of the branches.
Following the traditional "number's game" of It therefore resembles Mezilaurus rather than
The
would seem Paraia Ocotea vegetative characters. flowers,
generic delimitation, that in
its
it
known
was not at close to Licaria, which has only however, are unlike any other species of
all
three stamens (whorl each with two Lauraceae. The tepals are shorter than the floral
fertile III),
Only double-rimmed cupule would tube and usually more or unequal, the outer
thecae. the less
fit
genus than any However, when ones smaller than the inner ones. The stamens are
that better other.
considering additional characters (which have tongue-shaped and extremely papillose. Their four
floral
received attention up to now), a connection thecae are introrse and arranged in an arc on the
little
between Paraia and Licaria does not seem that stamens of the and second whorl, but latrorse
first
common
far In Licaria quite that the and in two pairs above each other in the third
off. it is
tepals are short and scarcely spreading at anthesis, whorl. The staminodia of whorl four are almost as
and that only the style protrudes beyond the nearly large as the inner stamens, but unlike very large
columnar staminodia other genera they are not glandular,
globular perianth. Also the thickish, sta- in
The
mens of the third whorl with their subapical thecae ovary has a very slender style, as, e.g., in
The com-
are somewhat similar to Licaria, The fact that the Licaria or Mezilaurus. fruit is almost
most
stamens of the whorl are sometimes only pletely surrounded by a deep cupule during
first
Volume Number 2 Rohwer 397
78. et
al.
New
1991 Genera Neotropical Lauraceae
of
of its development, but about half exserted at ma- der Werff, 1987, 1988). In nearly all of its char-
wood
turity. acters, including the structure, A. costari-
There no other species of Ocolea with unequal censis (Mez) Kosterm. member
clearly a of the
is
is
tepals, and most of the other characters of O. ruftra Ocotea insularis group in the sense of Rohwer
are very rare. Anatomical data (Richter, 1981) (1986). so close to Ocotea skutchii C. K.
It is
make
further underscore the isolated position of this spe- Allen as to very difficult to separate the
it
cies. Obviously, there no reason to retain in two in fruit. Aiouea lundelliana C. K. Allen, a
is
it
member
Ocotea other than reluctance to further increase of the same group, even more obviously
is
number
the of monotypic genera. In several char- intermediate between the generic concepts in that
acters (unequal tepals, broadly sessile stamens, red- its staminal configuration is variable within the
dish brown wood with very large cross-field and same inflorescence van der Werff, 1984; Roh-
(cf.
intervascular O. rubra resembles the Asian wer, 1986). In other Central American species of
pits),
genus Nothaphoebe, but the fruit very different. Aiouea the ties to a particular species or species
is
Nothaphoebe does not have a well-developed cu- group of Ocotea are not as clear, yet there no
is
somewhat
pule, but only a enlarged pedicel with doubt that they are closer to that genus than to
the tepals persistent under the the South American species of Aiouea,
fruit.
In South America, Aiouea very
similar to
is
species that used to be ascribed to the genus Phoe-
TWO-LOCULAR FOUR-LOCULAR ANTHERS
VS.
i Ai gam, . mai* n ^^rr • • i number
the difference the of
be. in
is
All of the taxa discussed above go beyond the thecae, but has long been known that in several
it
current delimitation of the genera simply because of the ''P/^oe&e" species only the outer anthers
they represent rare combinations of characters. are 4-thecous, whereas the inner ones are 2-the-
They
are therefore systematically isolated, but do cous. '^Phoebe^"* placed in quotation marks here
is
not jeopardize the delimitation as such. There because in the Neotropics anything but a good
is, it is
however, also an increasing number of cases that genus. The type of Phoebe from Asia, and
is in all
lead to serious doubts about the value of the current of the palaeotropical species the subtended
fruit
is
system. Traditionally, one of the most important by erect, enlarged, lignified tepals on a scarcely
characters used for the delimitation of genera thickened pedicel, clasping the base of the
is fruit.
the number of thecae per anther, two vs. four. Such a structure does not occur any of the
in
Although this character indeed constant for the neotropical species. In some of them there are
is
overwhelming
majority of species within well-de- persistent tepals, but they are not enlarged, and
fined natural groups, indiscriminate application either erect on the margin of a well-developed
its
may
lead to artificial groupings. cupule or spreading from a swollen pedicel or a
Several neotropical species have the staminal small, disklike cupule. Others have a normal, more
configuration of one genus, yet they can be rec- or less bowl-shaped cupule, as in most species of
Many
ognized as belonging to another genus based on Ocotea. of the neotropical species currently
all
The many
other characters. current system inconsis- ascribed to Phoebe^ in particular Central
is
Some
tent in treating these species. are placed with American and Andean species, probably belong to
Cmnawiomum,
their allies despite a staminal configuration that as suggested by Kostermans (1961).
makes impossible to key them out to the correct However, transferring of them indiscriminately
it all
Cinnamomum
genus, whereas others are placed within the genus to seems rather precipitant. Some
may
to which they would key out, regardless of their are clearly Ocotea, several go to Aiouea after
obvious affinity to species of a different genus. a revision of the generic delimitation, and oth-
still
may
Thus, Persea cuneataMeissn.j P, bilocularis Kopp^ ers represent a genus of their own.
and Caryodaphnopsis A
inaequalis (A. C. Smith) third example of a probably polyphyletic genus
&
New
van der Werff Richter, e.g., have 2-thecous in the World Endlicheria. defined by
is It is
anthers, although their respective genera are de- the combination of two characters, which indicate
common
fined by 4-thecous anthers, and Beilschmiedia sul- a level of advancement rather than a
&
common
cata (Ruiz Pavon) Kosterm. is the only species origin: (1) anthers 2-thecous, and (2) flow-
Some
with 4-thecous anthers in this otherwise 2-thecous ers unisexual. species of Endlicheria are
genus
extremely Rhodoste-
close to certain species of
Examples of species placed according to their monodaphne. Endlicheria metallica Kosterm. and
staminal configuration despite obvious to spe- R. grandis (Mez) Rohwer can only be separated
ties
cies of a different genus are Central American by counting the number of thecae per anther.
all
It
Aiouea (compare van
species currently placed in questionable whether they should even be treated
is