The taxonomy and morphology of Macranthera flammea (Orobanchaceae)

16 Pages·2002·8.6 MB·English

by  J D Alford

THE TAXONOMY AND MORPHOLOGY
OF
MACRANTHERA FLAMMEA
(OROBANCHACEAE)
C
and Anderson
D.AIford Loran
Jill
Department of Biological Science
RESUMEN
INTRODUCTION
Macranthera monotypic; the single species M.Jlammea endemic to the
is is
commonly
southeastern coastal plain of the United States and referred to as
is
"flame flower" or "hummingbird flower." listed as endangered by the state
It is
of Florida (Coile 2000). The genus was formerly placed in the Scrophulariaceae,
shown
but recent studies have that family to be a heterogenous group (Olmstead
Orobanchaceae.
et 2001); the hemiparasitic taxa are better placed in the
al.
map
Here, an updated species description and range are provided. More
detail on the history and pollination biology documented by Alford (2000).
life is
many
The names
that have been proposed for this species indicate the diffi-
culty that early botanists had in understanding the taxon and generic place-
its
ment (Table William Bartram (1791) described M.Jlammea as Gerardia
first
1).
flammea during his travels through the southeastern states during 1773-1777.
In 1834, after examining an unlabeled specimen in the herbarium of the
Academy
of Natural Sciences, Philadelphia (PH), Nuttall published the genus
flammea
Russelia
(Bartr.) Raf.
Toxopus gymnanthes
Raf.
7(Bartr)Pennell
Conradia as a dedication to the late Solomon W. Conrad, a Professor of Botany
name
at the University of Pennsylvania. Nuttall proposed the Conradia
fuchsioides, because "the corolla and long exerted stamens put one in mind of
some
gigantic Fuchsia" (Nuttall 1834).
Bentham name Macr
(1835) applied the anthera fuchsioides (Nutt.) Benth.,
because Conradia had been previously applied genus Gesneriaceae by
to a of
Martins. The name Macranthera means "long anther" Bentham and LeConte
were both originally cited as authors the genus (Hooker although
for 1835),
Bentham
(1835) felt that Torrey alone should receive credit for the generic name.
Torrey (1837) described a second species, M. kcontei (sometimes spelled
from specimen He
lecontii), a collected in Georgia in 1831 by LeConte. believed
was
distinct because the segments of the calyx were entire, linear-lanceolate,
it
and approximately
one-third the length of the corolla. In M.fuchsioides the seg-
ments and
of the calyx are dentately lobed often as long as the corolla tube [on
name
specimen
LeConte's (NY), Torrey applied the Conradia kcontei along
Torr.,
with a note stating the segments of the calyx are The specimen
entire]. itself is
damaged, but an attached
illustration depicts the entire sepals. Torrey actually
C
published the species (1837) not as kcontei, but as M. kcontei, since Bentham
had determined Conradia was
(1835) that already in
use.
Pennell (1935) questioned the authenticity of the collection of M.
site
which was Altamaha
purportedly along
h'contei, the River in Liberty County,
Georgia, because that was removed from other known occurrences
site far of
the species. However, the species has been observed along the Altamaha River
as recently as 1997 (M. Hopkins, comm.).
pers.
Rafmesque proposed many new names Macra
(1837) for nthera (see Table
1).
Macranthera flammea Raf. was proposed on the basis of Bartram's original
However, Rafinesque thought the genus name Macranthera was inappro-
and
priate because he did not beUeve the anthers to be of considerable length,
name meaning He
he suggested the Toxopus, "incurved peduncles." substituted
Toxopus gymnanthes Raf. for M. hcontei Torr. and Toxopus calycinus Raf. for M.
He meaning
fuchsioides (Nutt.) Benth. further proposed the genus Tomilix "cut
calyx" for M. fuchsioides, suggesting that this species be treated as either a sub-
new and
genus or genus because of the condition of the calyx the presence of
what he felt to be a bifid style. He proposed the binomial Tomilix hracteata
name The new
Raf. as an additional for M. fuchsioides (Nutt.) Benth. genus
Flamaria, as in Flamaria coccinea Raf., was also proposed
new
Bertoloni (1853) described a genus and species, Dasystoma tuhulosa,
Chapman
after examination of collections in GH, NY, and PH. (1860) treated M.
and Kuntze genus
lecontei as a variety of M. fuchsioides, (1891) resurrected the
Conradia on the basis of the type specimen that Torrey had originally identi-
fied as C. lecontei.
combined two Macranthera
Pennell (1913) the species of into one, writing:
may
"this difference [m the calyx] due wholly to age of the flower, as be readily
is
seen in the field; young flowers are as in lecontei, while with age the sepals con-
tinue growing and become lobed while the corolla shrivels and shortens in situ,
so bringing about \ht fuchsioides state." Pennell suggested that, because Bartram
name must become Macranthera
originally described the species, the accepted
named
flammea (Bartr.) Pennell. He argued that Bartram his species for its
which was
Gerardia
"flame-coloured" flowers a fitting description. rejected
is
name synonym
as a generic because a for Agalinis.
is
it
many
Later, were concerned about the validity of Bartram's binomials
(Fernald 1944; Rickett 1944; Merri 1945). Rickett (1944) argued that Bartram's
1 1
binomials are inappropriate because they appear in a publication that not
is
always consistent with the Linnean system of binomial nomenclature. Merrill
(1945) defended Bartram's binomials, writing: "the proportion of binomials to
polynomials, 358 to too great in favor of the former to justify the elimina-
2, is
Wilbur argued Bartram binomial
tion of Bartram's work." (1971) that, a ac-
if is
companied by an adequate description, the publication of the binomial should
be considered
valid.
Additional controversy surrounds Bartram's description of his collection
He specimen Macranthera with
probably confused the locality of his of
site.
that of the mint Calamintha coccinea (Nutt.) Benth. While traveling along the
Tensaw Baldwin Alabama), Bartram
River near "Taensa" (near or in Co, (1791)
recounts the following observations: "These stony gravelly heights produce a
variety of herbaceous plants, but one in particular shall mention on account
I
of singular beauty. believe a species of Gerardia (GerardiaJlammeaX
its I it is it
grows stem from branching
erect, a single a root, three or four feet in height,
very regularly from about one half length upwards, forming a cone or pyra-
its
mid, profusely garnished with large tubular labiated scarlet or flame coloured
which
flowers, give the plant a very splendid appearance even at a great dis-
tance." This description closely describes, for the most part, the form and ap-
pearance of M.Jlammea, but the locale and scarlet flower color are more appro-
priate for the mint Calamintha coccinea (Merrill 1945). Unlabeled specimens
Mu-
of both plants were found in Bartam's collection the Natural History
at
seum London
in (Pennell 1935). Bartram's description applicable in part to
is
both plants (Table but Pennell (1935) believed Bartram's description to be
2),
most M.Jlammea.
applicable
to
SPECIES DESCRIPTION
Macrantheraflammea generally an herbaceous biennial of the Orobanchaceae
is
on
Rosette leaves juvenile plants are oblanceolate to spatulate. Leaves are gla-
brous with entire or pinnately lobed margins, depending on developmental
number and 3-46
stage. Rosettes are variable in leaf leaf size, having leaves per
cm cm
rosette with lengths of 3-33 and widths of 0.8-3.6 (Fig.
1).
may
more
Plants be appropriately classified as monocarpic perennials be-
m
may
remain more
cause they the rosette phase than one year Rosettes
for
that have attained a critical size are triggered to reproductive states by an in-
day
crease in length. Before a rosette bolts into a flowering adult, begins to
it
produce pinnately lobed
leaves.
dm
Reproductive plants of M.Jlammea are mostly 15-30 with stems
tall,
that are obtusely four-angled and essentially nonbranchmg below the
candelabralike spreading inflorescence. Cauline leaves are opposite and
sessile
cm cm
have winged Lower and 2-6
or short, petioles. leaves are 8.5-16 long wide,
and
lanceolate to ovate, pinnately lobed. Leaves are progressively reduced up-
ward along the stem; the uppermost are 4-7 cm long and 1-2.5 cm wide, slightly
and
lobed, toothed or entire. Blades are glabrous, or with minutely ciliate mar-
and
gins. Rosettee lower leaves are usually shed by the time of anthesis.
branches Flowers
(Fig. 2). are axillary to a reduced, usually entire, foliar bract
m
cm
Godfrey and Wooten 1-2 becom-
(see illustration 1981). Pedicels are long,
The
ing ref lexed in fruit. calyx fused with five short lobes that are equal and
is
mm
cm
and The
linear 7.5-15 long. corolla is bright orange, fleshy, 2-2.5 long,
short pubescent without, and glabrous or slightly pubescent the base within.
at
mm
A
chamber
whitish nectar at the corolla base 4 long.
is
The corolla turns brown to black with age. The tube cylindrical and much
is
mm)
longer (20-23 than the two-lipped lobes (3-4.5 mm), with an adaxial
lip
gravelly heights
iy
spreading
that two-lobed and erect and a three-lobed, spreading abaxial The four
lip.
IS
contiguous stamens are equal in length and exerted. The filaments are orange
mm.
and become about twice as long as the corolla tube, up to 46 The ovary
is
mm
ovoid and has a slender style that to 46 and a linear-clavate stigma. Ma-
is
mm mm
ture bicarpellate capsules are 9.5-13 long, 7.5-10.5 wide, loculicidal,
mm
and The brownish
slightly pubescent. seeds are black, 2.5-3 long, half as
wide, irregular in shape (angular-lunate to triangular) with two or three fluted
membranous wings. Specimens M.flammea become black upon drying (due
of
compounds
to presence of orobanchin or iridoid as in related parasitic taxa).
The by examination (Appen-
species description verified of collections
is
dix from major and regional herbaria: Angus K. Gholson (personal collection
1)
AKG; become
designated will eventually part of FLAS), University of Florida
it
(FLAS), Florida State University (FSU), Gray Herbarium (GH), University of
HATT
Southern Mississippi (no official abbreviation, designated here for con-
venience), Jacksonville State University Louisiana State University (LSU),
(JSU),
New
Missouri Botanical Garden (MO), York Botanical Garden (NY), Old Do-
Academy
minion University (OD), of Natural Sciences of Philadelphia (PH),
Smithsonian Institution (US), University of Georgia (GA), University of Missis-
sippi (MISS), University of South Alabama (USAM), Valdosta State University
(VSC), and Vanderbilt University (VDB).
The supplemented by measurements taken from
species description also
is
GWD
Thomas Greenwood
two
several plants at field sites: Co., Georgia, at or
{Alford 1077 at FSU) and Liberty Co., Florida, in the Apalachicola National For-
ANF
est or (Alford 1064 at FSU).
Range and Habitat.— The range of M.flammea (Fig. 3) extends along the
lower coastal plain from Georgia to southeast Louisiana. This species requires
mesic habitats of bogs and thickets and ecotonal, occurring in the sandy acid
is
pine latwoods along the margins of seepage slopes and shrub-tree bogs
soils of
f
or bays (Small 1933). can occasionally be found in shallow water of cypress-
It
&
gum Wooten
ponds or depressions (Godfrey 1981). does not tolerate stand-
It
(Determann
1997).
et
al.
flammea
Macranthera
Am
Woody mmea
plants associated with M.fla include: Acer rubruniL.,
Ckthra
arbutifolia (L.) Ell., alnijolia var. tomentosa (Lam.) Michx., Clifto
monophylla
(Lain.) Britt. ex Sarg., Cyrilla racemiflora L, Fraxinus carolinU
Chapm.,
Mill., Ilex coriacea (Pursh) Ilex glabra (L.) Gray, Itea virginica
Liriodendron tulipijera Lyonia lucida (Lam.) K. Koch, Magnolia virginu
L.,
Myrica
Nyssa
L., cerifera L., syhatica var. biflora (Walt.) Sarg., Persea palus^
(Raf.) Sarg., Pinus elliottii En^ lus palustris Mill, Pinus serotina Michx.,
Rhododendron viscosumd.)! urnum nudum Herbaceous
associated
L.
taxa include: Andropogon Aristida Michx., Arnoglossum
vir L, stricta
Muhl,
sulcatum
(Fern.) H. Robins., naria gigantea (Walt.) Coreopsis
Muhl,
gladiata Walt., Elephantopus Gray, Erianthus giganteus (Walt.)
W
Hypericum
integrijolium 'ianthus angustifolius
L.,
Hypericum Hypericum
Lam., galioides Lam., tetrapetalum Lam.,
Hyptisala (Raf.) Shinners,Juncus valtaus Coville, ludwigia hirtella Raf., Os-
2
^amomeaL., Panicum scoparium Lam., and Pteridium aquilinum
Kuhn.
(L.)
Common Name.— commonly
This species referred to as flame flower or
is
hummingbird one was Spanish
flower. In at least instance, referred to as prin-
it
name
cess (Pickens 1955). Flame flower an appropriate for M.flammea.
is It is
representative of the specific epithet because of the flame-like appearance of
the orange-yellow corolla (and the candelabra branching habit lends a flame
Hummingbird
shape to the inflorescence). flower less desirable as could be
is is
any number
applied to of species that have red or orange tubular corollas.
t
—
F. 3. Documented range of M.C
MORPHOU
E\AMirN\IIO\C)l VARIABILITY
:)GICAL
lax onon base pla ntclesciipr ions on the morphology of repeating modular
iistb
akhough numbei
uni tso[ constiuction ic lca\cs tloweis etc because, the rof
can\aiyg] with and environmer
J size ot the \ hole plain teatly age ital
s
m
d.no and form modular
coil tern size of
1Q77) DiKeK llLLslKtUL cnspccicsc lepend on the morphological form
(Ik.
Ol cptat ing units Vj nation in l hataaeisnnay exist across the range of a s pe-
1
mmate
ndetei as defined by Harper (1977).
and and an
lowers leaves, the position of
1
By measuring changes
individual flo^ the iniloiescer ce letlects age.
its
how
mined
infloralchaia. detei each individual flower
affected by architecture of an individual plant and
is
i
in stating that the differences (in calyx
length and lobmg) between two
the species recognized that tnne (M.
at
m
fuchsioides and M. hcontei) were
fact attributable to age.
We
examined herbarium specimens (Appendix determine
to the extent
1)
morphological
of variability of characters across the species' range. Characters
considered were calyx length, corolla tube length, corolla lobe length, corolla
width, pedicel length, bract length, fruit length, and fruit width. Because was
it
know
not always possible to which part of the plant was represented by the
herbarium we
specimen, standardized measurements by choosing buds
floral
development
at the stage of just prior to the separation of the corolla lobes.
Lengths and width of mature fruits and, when available, the length and width
of the lower and uppermost (just below the inflorescence) stem leaves were
examine
To the extent of variability within a population and also within a
we randomly
from
single plant, selected five plants the established study
site at
GWD
Thomas
in Georgia, that represented the range environmental
Co., of
conditions for the Three flower buds (upper, median, and lower) were
site.
marked on each of three racemes of the inflorescence (terminal, upper
axillary,
and
lower measured buds
axillary), for a total of nine per plant Charac-
(Fig. 4).
previously mentioned were then measured on marked buds
ters every four days
We
for a period of one month. measured fruit and foliage on 12 plants randomly
from
selected the population.
Means and
standard were measured
deviations calculated charac-
for all
from herbarium specimens Means and
ters (Table standard deviations of flo-
3).
GWD
from
ral characters the sample population were calculated from charac-
ters measured on 28 August 1997 in order to determine the amount of variability
for characters within the plant architecture at a given time (Table
3).
The
standard deviations of measured characters (Table herbarium
3) for
specimens although components
reveal that the floral tissue (corolla tube) of
m
show
the flower little variability variability higher the chlorophyll con-
is
GWD
components and The
taining (calyx, pedicel, bract). population exhib-
more
and
ited variability overall in corolla tube calyx lengths than the her-
barium specimens because smaller and buds were
larger included.
GWD
Plant 59 of the sample population was selected growth
to illustrate
changes over time in corolla tube length (Fig. 4) and calyx length (Fig. 5) for
each bud position of the three racemes examined. For each raceme, the corolla
On
tube expanded rapidly (Fig. the terminal and lower axillary racemes,
4).
mm
expanded 22-24
the tubes to before anthesis. Those of the upper axillary
mm
mm.
raceme reached only 20-22 Corolla tube width was between 4 and
5
flowers examined.
for all
The calyces of the lower and median buds on the terminal raceme
(Fig.
5,
had
a) nearly reached their limits of expansion the measuring, but
at initial
from
orolla lengths different
p
plateau. Calyces buds on the upper and lower axillary racemes (Fig. b-c)
oi
5,
The
demonstrated rapid expansion before leveling calyces of the up-
also off.
per buds were much shorter than those of the median and lower buds for all
growth upper buds racemes
Figure 6 illustrates calyx for the of the three
three of the plants comprising the sample population. shows that the up-
for It
and
per axillary raceme had shorter calyces than did the lower axillary termi-
nal racemes for plants of the sample population.
Pedicels exhibited the same growth pattern as calyces. Bracts generally did
not expand further once anthesis had occurred for an individual bud. Bracts