Content uploaded by Pilar Catalan
Author content
All content in this area was uploaded by Pilar Catalan on Feb 10, 2014
Content may be subject to copyright.
A NEW CIRCUMSCRIPTION
OF FESTUCA TRICHOPHYLLA (GAUDIN)
K.
RICHTER (GRAMINEAE)
by
A.-K.
K. A.
AL-BERMANI
*, P.
CATALÁN
••
& C. A.
STACE•
Resumen
AL-BERMANI,
A.-K.
K. A., P.
CATALÁN
& C. A.
STACE
(1992).
Una
nueva circunscripción
de Festuca trichophylla (Gaudin)
K.
Richter (Gramineae). Anales Jará.
Bot.
Madrid 50(2):
209-220
(en
inglés).
Se revisan
la
identidad
y la
posición taxonómica
de
diez táxones que,
en un
tiempo
u
otro,
han sido considerados próximos
a
Festuca
trichophylla (Drucros
ex
Gaudin)
K.
Richter.
Se
propone
un
nuevo concepto,
el
del
grupo
de
F. trichophylla,
que se
compone
de
cuatro espe-
cies:
F.
nevadensis (Hackel)
K.
Richter,
F.
rothmaleri (Litard.) Markgr.-Dann.,
F.
tricho-
phylla
y
F. paucispicula Fuente García
&
Sánchez-Mata.
La
propia F. trichophylla incluye,
en nuestra propuesta, cuatro especies distintas
de las
reconocidas
por
MARKGRAF-DANNEN-
BERG
(1980)
más una
subespecie situada
por
la
autora bajo
F.
rubra
L.;
nosotros distingui-
mos,
dentro
de
la
especie, tre& Subespecies: subsp. trichophylla, subsp. scabrescens (Hackel
ex Trabut) Catalán
&
Stace, comb. nov.,
y
subsp. asperifolia (St.-Yves) Al-Bermani, comb.
nov.
Se
aporta
un
breve comentario sobre
la
sinonimia
y la
distribución geográfica
de los
seis
táxones del grupo.
F.
rubra var. gaetula Maire
ex
St.-Yves
se
transfiere
a
F. nevadensis como
var. gaetula (Maire
ex
St.-Yves) Al-Bermani
&
Stace, comb.
nov.
Palabras clave: Festuca, grupo
de
F. trichophylla, Gramineae, taxonomía.
Abstract
AL-BERMANI,
A.-K. K. A., P.
CATALÁN
& C. A.
STACE
(1992). A new circumscription of
Festuca trichophylla (Gaudin)
K.
Richter (Gramineae). Anales Jará.
Bot.
Madrid 50(2):
209-220.
The identity
and
taxonomic disposition
of
ten taxa that have
at one
time
or
another been con-
sidered
as
cióse relatives
of
Festuca trichophylla (Ducros
ex
Gaudin)
K.
Richter
are
revie-
wed.
A
new
concept,
the
F. trichophylla group,
is
presented
in
which
are
placed four species:
F. nevadensis (Hackel)
K.
Richter,
F.
rothmaleri (Litard.) Markgr.-Dann.,
F.
trichophylla
and F. paucispicula Fuente García
&
Sánchez-Mata.
F.
trichophylla itself includes four sepá-
rate species recognized
by
MARKGRAF-DANNENBERG
(1980)
and
a
subspecies that
she
placed
under F. rubra
L.; we
divide
it
into three subspecies: subsp. trichophylla, subsp. scabrescens
(Hackel
ex
Trabut) Catalán
&
Stace, comb.
nov., and
subsp. asperifolia (St.-Yves) Al-Ber-
mani, comb. nov.
A
brief synonymy
and
geographical range
of
the six taxa
are
given. F. rubra
var.
gaetula
Maire
ex
St.-Yves
is
transferred
to
F.
nevadensis
as
var.
gaetula
(Maire
ex
St.-Yves)
Al-Bermani
&
Stace, comb.
nov.
Key words: Festuca,
F.
trichophylla group, Gramineae, taxonomy.
* Department
of
Botany, University
of
Leicester. Leicester
LEÍ 7RH
(England).
** Present address: Departamento
de
Agricultura
y
Economía Agraria, Universidad
de
Zaragoza. Miguel
Ser-
vet, 177. 50013 Zaragoza (España).
INTRODUCTION
The Festuca rubra
L.
aggregate
Twenty-one species
(nos.
56-89, excl.
57
and
62) of the 170
species recognized
in
Flora Europaea
by
MARKGRAF-DANNEN-
BERG (1980) would fall into 'Festuca
ru-
bra
L.
sensu amplissimo1
as
defined
by
HAC-
KEL
(1882:
128).
Some
of
these
21
species
(e.g.
F.
heterophylla Lam.)
are
very distinct
210ANALES JARDÍN BOTÁNICO DE MADRID, 50(2) 1992
from all the rest, while others form recogni-
zable groupings within which specífic limits
are more difficult to describe. Two oí these
may be known as the F. rubra and the F. tri-
chophylla (Ducros ex Gaudin) K. Richter
groups. Together, these two groups have
been called the F. rubra aggregate, and
correspond with species 65-78 of MARK-
GRAF-DANNENBERG'S (1980) treatment.
The
Festuca
trichophylla (Ducros ex Gaudin)
K. Richter group
A future paper (AL-BERMANI & STACE,
inprep.) will presenta revisión of the F. ru-
bra aggregate and define precisely its two
groups and give full descriptions of its segre-
gates.
The F. trichophylla group is characte-
rized by a range of vegetative morphologi-
cal and anatomical characters, which taken
together set it apart from the F. rubra
group. All the characters, however, are
variable, and taken in isolation none of
them can be regarded as absolutely diagnos-
tic.
The F trichophylla group may be distin-
guished from the F. rubra group as follows:
1.
The leaf-blades are usually scabrid. The
scabridity is caused by antrorsely di-
rected prickle-hairs on the abaxial leaf
surface. In the F. rubra group such
prickle-hairs are absent or few and con-
fined to the apical región of the
leaf.
The leaf-sheaths are rnostly fused almost
to the mouth, as in the F. rubra group,
but usually a few of the tillers have free,
overlapping leaf-sheaths (as in the F. ovi-
na aggregate).
The abaxial leaf-blade surface usually
possesses rows of small protuberances
c. 12-15 am across where the files of long-
cells meet (fig. 1). We have never seen
these in the F. rubra group.
The silica-cells are usually mostly kid-
ney-shaped or rounded and each is cha-
racteristically infolded by the adjaccnt
cork-celh whereas they are usually rec-
tangular or square and not infolded in
the F. rubra group (fig. 2).
The longitudinal walls of the long-cells
on the abaxial epidermis are usually
thicker and more strongly sinuate than in
the F. rubra group (fig. 2).
As seen in leaf-blade transverse section,
the abaxial strands of sclerenchyma are
markedly unequal in size: either the
median strand alone, or sometimes the
median as well as the two marginal
strands, are conspicuously larger than
the others (fig. 3).
Fig.
1
.- — •-Leaf-sections: A, F. trichophylla subsp. trichophylla, toshow abaxial protuberances (arrowed); B, F. rubra
subsp. rubra.
A.-K. K. A. AL-BERMAN1 & AL.: FESTUCA TRICHOPHYLLA211
Fig.
2.
— Abaxíal epidermis to show silica-cells (SC) and cork-cells (CC): A, F. trichophylla subsp. trichophylla;
B,
F. arenaria.
Fig.
3.
— Leaf-sections: A, F. ¡richophylla subsp. Irichophylla; B, F. nevadensis var. gaetula; C, F. nevadensis var.
nevadensis; D, F. rubra subsp. rubra; E, F. rubra subsp. juncea; F, F. arenaria.
212ANALES JARDÍN BOTÁNICO DE MADRID, 50(2) 1992
Sometimes
the
submarginal
and
marginal
strands merge
to
form
one
larger strand
at
each margin,
but
otherwise
the
strands
are
distinct.
In the F.
rubra group, where
the
strands
are
discrete
all are
approximately
of
the same size,
but in
other members
of the
group
the
strands merge
to
varying degrees
(fig.
3), in
extreme cases
(F.
arenaria
Os-
beck) forming
a
continuous
are.
Our concept
of the F.
trichophylla group
has gradually developed during taxonomic
and biosystematic studies
of the F.
rubra
aggregate;
the
numbers
of
taxa that
we
ascribed
to it
gradually grew
as
further bino-
mials were investigated. Concurrently
the
number
of
species that
we
recognized
de-
creased
as
more material
was
examined.
At
present
we
recognize four species:
F. ne-
vadensis (Hackel)
K.
Richter,
F.
rothmaleri
(Litard.) Markgr.-Dann.,
F.
trichophylla
and
F.
paucispicula Fuente García
& Sán-
chez-Mata.
All
the
taxa that
we
place under
F. tri-
chophylla sensu stricto
are, as far as we
know, hexaploid
(2n =
42),
but we
have
in-
sufficient living material
for a
thorough
cy-
tological survey. Segregates that have been
studied
in
this respect
are F.
trichophylla,
F. cyrnea (St.-Yves
&
Litard.) Markgr.-
Dann.
and F.
iberica (Hackel)
K.
Richter.
On
the
other hand,
F.
rothmaleri
is
octo-
ploid
(2n = 56) and F.
nevadensis
is
deca-
ploid
(2n = 70)
(table
1), and the
stomatal
length (mean
55 um) of the
type specimen
of
F.
paucispicula suggests that this might
also
be at a
higher level than hexaploid.
The
purpose
of
this paper
is to
review
the F. tri-
chophylla group
and to
present evidence
for
a
new,
wider circumscription
of F.
tricho-
phylla sensu stricto that includes four
se-
párate species recognized
by MARKGRAF-
DANNENBERG
(1980)
(one of
them placed
by
her
outside
the F.
rubra aggregate)
and a
subspecies that
she
placed under
F.
rubra.
MATERIALS
We have examined herbarium material
from
BC, BM, G,
GDAC, JACA,
K,
LISI,
LTR,
MA, MAF, P and W,
including
the
type specimens
of
eight
of the ten
taxa
dis-
cussed,
and
living material
of F.
tricho-
phylla,
F.
cyrnea
and F.
iberica,
as
well
as of
F. nevadensis
and F.
rothmaleri.
TABLE
1
ORIGIN
OF
MATERIAL PROVIDING
NEW
CHROMOSOME COUNTS
IN THE
GENUS FESTUCA
Festuca cyrnea (St.-Yves
&
Litard.) Markgr.-Dann.
1.
St.
Zacharie, Var(France)
2« = 42
2.
Allauch, Bouches-du-Rhóne (France)
2n = 42
F. iberica (Hackel)
K.
Richter
1.
Sierra Nevada, Granada (Spain)
2n = 42
2.
Prado
de las
Pozas, Sierra
de
Gredos, Ávila (Spain)
2n = 42
3.
Morcuera
to
Rascafría, Sierra
de
Guadarrama, Madrid (Spain)
2n = 42
F. nevadensis (Hackel)
K.
Richter
1.
By
road from Granada
to
Veleta, near Fuente Bajada,
1900 m,
Granada (Spain)
. 2/» = 70
F. rothmaleri (Litard.) Markgr.-Dann.
1.
Puerto
de la
Peña Negra, Ávila (Spain)
2n = 56
F. rubra
L.
subsp. juncea (Hackel)
K.
Richter
1.
Great Orme Copper Mine, Caernarvonshire (Wales)
2n = 42
2.
Tarbat Ness, East Ross (Scotland)
2n = 42
3.
Near Liége (Belgium)
2n = 56
F. trichophylla (Ducros
ex
Gaudin)
K.
Richter subsp. trichophylla
1.
Montagne
de
Lachens, 1700
m, Var
(France)
2n = 42
A.-K.
K. A.
AL-BERMANI
& AL.:
FESTUCA TRICHOPHYLLA213
METHODS
All
the
established morphological
and ana-
tomical characters
of
valué
in
Festuca taxo-
nomy have been utiüzed. Spikelet length,
as
is normal,
is
taken
to the tip of the
fourth
lemma only, excluding
the awn. Leaf-
sections
and
epidermal preparations were
taken from
the
middle
of a
tiller
leaf. Sec-
tions were made with
a
freezing microtome
after
the
leaf
had
been boiled
in
water.
Epi-
dermises were obtained
by the
lactic acid/
scraping method. Chromosome counts
were made
on
root-tips
of
mature plants
that were pre-treated
in
saturated aqueous
hexachloro-cyclohexane (gammexane)
and
stained with acetic-orcein.
REVIEW OFTAXA
IN THE
F. TRICHOPHYLLA GROUP
The distinct
and
well-known species
F. ne-
vadensis
and F.
rothmaleri
(see
FUENTE
GAR-
CÍA
&
SÁNCHEZ-MATA,
1987) are not co-
vered here, except
in the key, but ten
other
taxa that fall into
the F.
trichophylla group
or
have been implicated
in its
taxonomy
are dis-
cussed below.
1.
F.
rubra trichophylla Ducros
ex Gau-
din,
Fl.
Helv.
1: 288
(1828), variously inter-
preted
as
variety
or
subspecies,
was des-
cribed from
wet
places
in the
Jura
and
Swiss
Alps.
The
leaves were said
to be
very
narrow
and to
resemble those
of F.
ovina,
from which
it was
distinguished
by the pre-
sence
of
rhizomes.
It was
raised
to
specific
status
by
RICHTER (1890:
100). F.
tricho-
phylla
is a
well-known species from
the
mountains
of
southern Europe
and,
although
we have
not
been able
to
trace type mate-
rial, we have examined many accurately
determined specimens.
We
have living
ma-
terial from
Var,
France, which
is
hexaploid
(2n
= 42)
(table
1).
KERGUÉLEN
&
PLONKA
(1989:
283)
stated that
the
type
(?
lectotype)
from Mont Reculet,
Ain
(Jura)
had
been
seen
by
them
in LAU.
Gaudin
did not
mention leaf scabridity,
from which
one
could assume that
it was
absent
or
slight.
In our
experience this
taxon
is
rarely strongly scabrid — usually
the scabridity
is
weak and/or confined
to the
apical part
of the leaf.
Material that
we
have
seen almost always agrees with Gaudin's
description
in its
very narrow leaf-blades
(usually 5S0.6
mm
diameter)
and
long,
often numerous rhizomes.
2.
F.
rubra
var.
iberica Hackel, Monogr.
Festuc.
Eur.: 136
(1882)
was
described from
Spain under
F.
rubra subsp. violacea
(Schleicher
ex
Gaudin) Hackel, most
va-
riants
of
which were differentiated from
ssp.
eu-rubra
in
having
the
ovary hairy
at
the apex.
The
ovary
of var.
iberica,
ho-
wever,
was
rightly described
as
glabrous,
and this variety
was
separated
in
Hackel's
key from most varieties
of
subsp. eu-rubra
by
the
densely tufted, scarcely rhizomatous
habit
and
setaceous leaves.
It was
raised
to
specific rank
by
RICHTER (1890:
99) and
kept within
the F.
violacea Schleicher
ex
Gaudin aggregate
by
MARKGRAF-DANNEN-
BERG
(1980). Hackel described
the
leaves
as
scabrid
and as
having three (central
and two
marginal) sclerenchyma strands conspicuous-
ly larger than
the
others. Material
we
have
seen confirms
all the
above characters.
F. iberica occurs
in the
mountains
of
Spain
and
North África.
We
have living
material from central
and
south-eastern
Spain; this
is
hexaploid
(2M
=
42)
(table
1).
Within
his var.
iberica,
HACKEL
(1882:
136-137) distinguished
two
subvarieties;
typica
and
subscabra Hackel,
the
latter
dif-
fering
in its
leaf-blades being scabrid only
near
the
apex,
its
smaller spikelets
and its
smooth
(not
scabrid) lemmas;
it was
said
to
be confined
to the
Pyrenees, where subvar.
typica
was
unknown. KERGUÉLEN
&
PLON-
KA
(1989:
193)
also recognized only subvar.
subscabra from
the
Pyrenees; they claimed
that
it
lacks rhizomes.
Var. iberica
was
lectotypified
by
FUENTE
GARCÍA
& al.
(1988:
515):
"Sierra Nevada,
in valle
sup.
fluv. Jenil, 29.6.1876"
(W no.
5182).
Subvar. subscabra
has
been lectoty-
pified
by
FUENTE GARCIA
&
SÁNCHEZ-
MATA
(1986a:
170):
"Gavarnie, Htes. Pyré-
nées,
juillet
1877,
Bordére"
(W no.
5183).
We have seen both lectotypes.
F. iberica
in
some ways comes
at the
opposite
end of the
spectrum from
F. tri-
214ANALES JARDÍN BOTÁNICO
DE
MADRID,
50(2) 1992
chophylla sensu stricto,
as it is
strongly
sca-
brid
and has a
densely tufted growth-habit,
making
it a
characteristic species
of the
southern
and
central Spanish mountains.
It
shares with
F.
trichophylla, however,
the
very narrow leaves (usually
=S
0.55 mm) and
it sometimes produces
a few
long rhizomes
that eventually form further dense tufts
(field note
by
P.C.). Subvar. subscabra
has
longer, much less scabrid leaves
and is pos-
sibly less densely tufted,
but we
have
not
seen
it in the
field;
it
comes closer
to F. tri-
chophylla.
3.
F.
scabrescens (Hackel
ex
Trabut)
Batt.
&
Trabut
was
described
by
Trabut,
Fl.
Algérie
Mon.: 215
(1895)
and
BATTANDIER
& TRABUT (1904) from
1900 m on
Djebel
Mzi,
Sud
Oranais, Algeria.
We
have been
unable
to
trace
any
type
or
authentic mate-
rial,
but
have examined
two
specimens
in P
collected
by R.
Maire
in the
Grand
and
Moyen Atlas
in 1921 and 1923
respectively.
The basionym (TRABUT,
1895) was
written
"F.
[RUBRA] SCABRESCENS Hackel,
in
litt",
and
reference
to the
Introduction
(p.ix)
and to the
footnote
on p. 4 of the
Dicotylédones volume
of
Flore
de
VAlgérie
(1888) suggests that
the
rank intended
was
race
or
subspecies. St.-Yves (1922:
17) re-
ferred
to it as a
variety
and
considered that
in
its
acuminate young leaves, pubescent
leaf-sheaths, presence
of
well-developed
bulliform cells
and
scabrid leaf-blades
it
shows affinity with
F.
nevadensis
(a
member
of
the F.
trichophylla group
in our
sense).
ST.-YVES
(1913:
125)
earlier stated that
he
had received material sent
by
Trabut from
Djebel
Mzi,
some
of
which possessed
the
diagnostic characters
"peu
devéloppés".
He
also (ST.-YVES,
1922: 17) saw the
same
Maire material from Grand Atlas that
we
haveseen.
This taxon
is
cióse
to F.
iberica
and we
here unite them;
at
subspecific level
we
believe scabrescens
is the
earliest epithet
to
have been used.
4.
F.
rubra subvar. asperifolia St.-Yves,
Annuaire Conserv. Jard.
Bot.
Genéve
17:
125 (1913)
was
described from
the
Alpes
Maritimes, France under subsp. eu-rubra
var. genuina.
It was
said
to
differ from other
subvarieties
of var.
genuina
in its
glaucous,
scabrid leaves,
and
from
var.
trichophylla
in
its fíat culm-leaves. Other characters stressed
were
the
presence
of
rhizomes,
the
well
developed leaf-blade sclerenchyma strands,
the relatively thick (0.7-0.8
mm
diameter)
leaf-blades,
and the
long
(8-10 mm)
spike-
lets.
ST.-YVES (1913:125) remarked that
his
subvar. asperifolia
is
"tres distinct" from
F. (rubra) scabrescens,
the
latter having
lar-
ger, auriculate ligules
and
well developed
bulliform cells, although these characters
were
not
always well developed.
This taxon
was
raised
to
subspecific level
by
MARKGRAF-DANNENBERG
(1976: 143).
We have seen many specimens
in G, and
of
the
syntypes
we
select
the
following
as
lectotype: "Alpes Maritimes,
St.
Etienne
de Tinée,
á
Cascai, points frais, silice,
1500
m,
18.VIL 1908, St.-Yves".
Our exa-
minations confirm
the
characters given
by
St.-Yves. This taxon shares
the
rhizoma-
tous habit with
F.
trichophylla
and the leaf-
blade scabridity with
F.
iberica,
but it has
much thicker leaf-blades than either. Other
characters placing
it in the F.
trichophylla
group rather than
the F.
rubra group
are the
leaf-blade sclerenchyma pattern
and the
presence
of
abaxial protuberances, charac-
teristic silica-bodies
and
some overlapping
leaf-sheaths.
ST.-YVES
(1913:
125)
noted specimens
from
the
Alpes Maritimes that were inter-
mediate between subsp. asperifolia
and
subsp. juncea (Hackel)
K.
Richter,
and
others intermediate between subsp. aspe-
rifolia
and F.
trichophylla. According
to
MARKGRAF-DANNENBERG
(1980) subsp.
as-
perifolia
is
endemic
to
mountains
of
south-
ern Europe,
but
KERGUÉLEN
&
PLONKA
(1989) also gave localities north
to
Switzer-
land
and
Bretagne. KERGUÉLEN
&
PLONKA
(1989) suggested that decaploid
(2« = 70)
plants from
the
Pyrénées Atlantiques
"rap-
prochent
un peu" F.
rubra subsp. asperifo-
lia
but
suggested that they might
be F.
rubra
subsp. megastachys Gaudin
or a new
taxon
(see under
F.
nevadensis
var.
gaetula
for
further discussion).
5.
F.
rubra
var.
yvesiana, Litardiére
&
Maire,
Mém. Soc. Sci. Nat.
Maroc
4: 25
A.-K. K. A. AL-BERMANI & AL.: FESTUCA TRICHOPHYLLA215
(1924) was described from the Grand and
Moyen Atlas under subsp. eu-rubra. The
leaves were said to be scabrid, setaceous
and in exposed habitats with very strong
marginal and central strands of sclerenchy-
ma; the plants were described as denseíy
tufted with few short rhizomes. It was raised
to specific rank by PATZKE (1964: 195) as
F. pseudotrichophylla; this was accepted
and placed next to F. trichophylla by MARK-
GRAF-DANNENBERG (1980), who gave the
distribution as endemic to mountains of
France, Spain and Portugal. LITARDIÉRE
(1945:
134) gave records from Var and Bas-
ses-Pyrénées, France, and said that forms
transitional to F. trichophylla occur in
Spain. KERGUÉLEN & PLONKA (1989: 149),
however, implied that the Var records were
errors for F. cyrnea, and those from the
Pyrénées for F. trichophylla. According to
MARKGRAF-DANNENBERG
(1980), F. pseu-
dotrichophylla (and F. cyrnea) differ from
F. trichophylla in their more densely tufted
habit with shorter rhizomes and more per-
sistent leaf-sheaths. We have seen a spe-
cimen in P that we select as lectotype: "Ma-
roc,
Grand Atlas, Ourika, pentes N.N.W.
du Djebel Tachdirt, porphyre, 3200 m,
25.VII.1922, R. de Litardiére". This speci-
men has poorly developed sclerenchyma
strands, but the small abaxial protube-
rances are evident.
This taxon is cióse to F. iberica.
6. F. rubra var. cyrnea St.-Yves & Li-
tard., Bull. Soc. Bot. France 71: 122 (1924)
was described from Corsica (ST.-YVES,
1924),
under subsp. eu-rubra, and was rai-
sed to specific rank by MARKGRAF-DAN-
NENBERG (1978: 237). The leaves were said
to be ca-pillary to subsetaceous, glaucous
and very smooth, and transitional plants to
F. rubra subvar. glaucescens Wahlenb, and
to F. trichophylla were noted. (The precise
identity of subvar. glaucescens is unknown,
but is probably referable to F. rubra.)
F. cyrnea differs from F. pseudotricho-
phylla in its longer rhizomes and smoother
leaf-blades. KERGUÉLEN & PLONKA (1989:
149),
however, said that the latter character is
very variable in F. cyrnea, which may have
leaves "plus ou moins scabre". We agree.
MARKGRAF-DANNENBERG
(1980) stated
that F. cyrnea was endemic to Italy and
Corsica,
but
MARKGRAF-DANNENBERG
(1982:
491) added Sardinia, KERGUÉLEN
(1983:
10) added Bouches-du-Rhóne, and
KERGUÉLEN & PLONKA (1989: 149) added
south-eastern France (Var and Alpes Mari-
times).
We have living material from Var
and Bouches-du-Rhóne; both are hexa-
ploid (2n = 42) (table 1). We have also seen
the lectotype (G), designated by KERGUÉ-
LEN
(1983). When grown side-by-side in the
garden with F. trichophylla no differences
were detected, and the rhizomatous habit of
both was conspicuous.
7.
F. rubra subsp. font-queri Litardiére,
Candollea 10: 133 (1945) was described
from Tarragona, Spain as having a pu-
bescent ovary apex, long rhizomes, and
capillary, smooth leaf-blades. It was said to
be "valde distincta" but with affinity "satis
remota" to F. pyrenaica Reuter,and F. vio-
lacea Schleicher ex Gaudin. Its most dis-
tinctive character is the very depauperate
panicle with only 2-4 spikelets, and it was
raised to specific rank as F. paucispicula
by FUENTE GARCÍA & SÁNCHEZ-MATA
(1986b: 443), who designated the lectotype
in BC. They provided drawings of leaf-sec-
tions which showed a strongly pronounced
median sclerenchyma strand. and stated
that the leaf-blades were slightly scabrid.
The only material or record known to us
are the syntypes, which we have seen. Our
observations confirm those of FUENTE GAR-
CÍA
&
SÁNCHEZ-MATA
(1986b). In addition
we can confirm that the ovary is sparsely
pubescent, a character unknown in either
the F. trichophylla or F. rubra groups (ex-
cept for F. cretacea T. Popov & Proskorj.).
Despite this, its scabrid leaves with abaxial
protuberances and trichophylla-like scle-
renchyma pattern and silica-bodies suggest
to us that it is a member of the F. tricho-
phylla group, but probably a distinct spe-
cies.
The collection of more material is
highly desirable.
8. F. duriotagana Franco & Rocha
Afonso, Bol. Soc. Brot., ser. 2, 54: 91
(1980) was described from the valleys of the
Douro and Tagus in Portugal. It was said to
216ANALES JARDÍN BOTÁNICO
DE
MADRID,
50(2) 1992
have been included
in F.
rubra
by
previous
Portuguese authors,
but to
have leaves
±
sacabrid near
the
apex
and to be
closer
to
F. trichophylla, from which
it
differs
in its
thicker stems
and
longer upper glume.
The
rhizomes were described
as
long.
The only material
we
have seen
is the
type
in
LISI.
The
leaf-blade sclerenchyma
pattern,
the
lack
of
abaxial protuberances,
and
the
very weakly developed scabridity
place this taxon closer
to F.
rubra than
to
F.
trichophylla.
In
silica-cell shape
and
long-cell wall sinuation
it is
more
or
less
in-
termediate between
the F.
rubra
and F. tri-
chophylla groups.
The
leaf-sheaths.
ho-
wever. present
a
mixture
of
open
and
closed,
and henee suggest
a
closer affinity with
F. trichophylla. Assignation
of
this taxon
to
one group
or the
other must await
the
acqui-
sition
of
further material.
9.
F.
rubra subvar. juncea
was
describ-
ed
by
Hackel, Monogr. Festuc.
Eur.: 139
(1882) under subsp. eu-rubra
var.
genuina
from
a
range
of
localities
on
lake
and
river
banks
in
central Europe
and
Sweden.
It was
raised
to
subspecific level
by
RICHTER
(1890:
99) and
accepted
at
that rank
by
MARKGRAF-DANNENBERG
(1980) and
KER-
GUÉLEN
&
PLONKA
(1989). Hackel des-
cribed
it as
strongly creeping, with thick,
/««CMS-like leaf-blades with strong scleren-
chyma strands. Since then
it has
been
re-
corded more widely
in
central Europe,
extending
to the
British Isles (MARKGRAF-
DANNENBERG,
1952).
FUENTE GARCÍA
&
SÁNCHEZ-MATA (1989:
248)
found that
it
also oceurs
in the
Pyrenees (Gerona
to
Huesca),
and
they compared
the
Spanish
material with
the
syntypes
in W, but
with-
out making
a
lectotypif¡catión.
We
have
also examined
the
syntypes
and
select
the
lectotype
as
follows: "Bóhmen. Prag.
Im
Bergschutte
am
Fusse
des
Kuchelbader
Berger.
15.
VI. 1879".
As noted previously, ST.-YVES
(1913:
125)
recorded specimens intermediate between
F. rubra subsp. asperifolia
and
subsp. juncea.
However,
in all
characters (except perhaps
leaf sclerenchyma,
q.v.)
subsp. juncea
comes much nearer
F.
rubra than
F. tri-
chophylla,
and we are
convinced that
it
belongs within
the
former species, whereas
subsp. asperifolia falls much closer
to F. tri-
chophylla.
In
fact
A.-K.
A.-B.
and C. A. S.
consider subsp. juncea indistinguishable
from
F.
rubra subsp. pruinosa (Hackel)
Piper,
for
which
it is an
earlier ñame.
Its
pattern
of
leaf-sclerenchyma
(fig. 3E)
often
resembles that
of
F.
trichophylla rather than
that
of
F. rubra, with
the
median
and
margi-
nal strands conspicuously larger than
the
others,
as
shown
in the
drawings
of
Hackel
on
the
lectotype sheet
and
FUENTE GARCÍA
& SÁNCHEZ-MATA (1989:
250), and
less
so
in those
of
KERGUÉLEN
&
PLONKA (1989:
267).
We
have taken leaf-sections from
many specimens,
and
find that
the
drawings
of Fuente García
&
Sánchez-Mata
and
Hackel represent
the
extreme
of the
range,
the other extreme
of
which
is
typical
of the
F. rubra group.
F. rubra subsp. juncea
is
hexaploid
(2«
= 42) or
octoploid
(2n = 56)
(table
1).
10.
F.
rubra
var.
gaetula Maire
ex St.-
Yves,
Candollea
1: 17
(1922)
was
described
from 2000
m in
Morocco under subsp. neva-
densis Hackel
(=F.
nevadensis). This
was
raised
to
specific level
by
KERGUÉLEN
(1979:
545).
ST.-YVES (1922:
20)
considered that
the
taxa 'nevadensis', 'gaetula', 'scabrescens',
'cyrnea'
and
'rubra' formed
a
'chaine'
con-
necting
F.
rubra subsp. rubra
and
subsp.
nevadensis, 'gaetula' differing from typical
'nevadensis' mainly
in its
more angular
leaf-
section
and
smaller, more equal-sized scle-
renchyma strands.
LITARDIÉRE
(1945:
137)
recognized intermediates
in all
three combi-
nations between 'rubra', 'gaetula',
and
typi-
cal 'nevadensis',
and
recorded
F.
nevaden-
sis from France (Pyrenees, Cevennes
and
Alpes).
Later (LITARDIÉRE
1947: 121) he
identified plants from
the
región between
the Cevennes
and
Alpes (Dróme, Vauclu-
se)
as
intermediate between
F.
rubra subsp.
eu-rubra
and F.
rubra subsp. nevadensis
var. gaetula. Subsequently
(e.g.
CLAUS-
TRES,
1960;
KERGUÉLEN,
1979, 1983), all
French
and
Pyrenean material referred
to
F. nevadensis
was
considered
to
represent
var. gaetula.
The
combination F. nevadensis
var. gaetula, used
by
FUENTE GARCÍA
&
A.-K. K. A. AL-BERMANI & AL.: FESTUCA TRICHOPHYLLA217
SÁNCHEZ-MATA (1987:371), appears not to
have been made validly before now. The
latter two authors considered that F. ne-
vadensis (including var. gaetula) does not
occur north of south-eastern Spain.
FUENTE GARCÍA & SÁNCHEZ-MATA
(1989) reviewed the Pyrenean records of
var. gaetula, and concluded that they re-
present F. rubra subsp. juncea. According
to KERGUÉLEN & PLONKA (1989: 297) some
of the French Pyrenean material that has
been placed under F. nevadensis is referable
to F. rubra subsp. asperifolia (q.v.), and
some to subsp. juncea. KERGUÉLEN
(1983:
13) stated that plants named F. nevadensis
from the Alps and Cevennes need checking;
we have not seen specimens ñor a subse-
quent reference to them, and similary do
not know the source of MARKGRAF-DAN-
NENBERGS (1980) record of the species from
Corsica.
One collection from the Pyrénées Atlan-
tiques is, like F. nevadensis, decaploid
(2/Í = 70) (KERGUÉLEN & PLONKA, 1988:
226).
It was originally collected as var.
gaetula
(KERGUÉLEN
1975: 178) but in our
opinión it is F. rubra subsp. megastachys
(=F. diffusa Dumort. = F. heteromalla
Pourret), a possibility also mentioned by
KERGUÉLEN & PLONKA (1989: 297). In cul-
tivation it has fíat leaves and a diffuse inflo-
rescence.
An isotype of F. rubra var. gaetula in P
has been examined by us. It bears a Uni-
versité d'Alger label in Maire's writing
agreeing in all repects with the locality given
in the protologue: "Djebel Beni Smir,
foréts de Quercus, gres, 2000 m, R. Maire,
2.
VI. 1918". The specimen also agrees clo-
sely with the type description. In our opi-
nión it clearly belongs to F. nevadensis; it
has some open leaf-sheaths, slightly scabrid
leaves with very hairy adaxial ribs and aba-
xial protuberances, and rounded silica-
bodies. The stomatal length (50-70 um,
mean 61.5 u.m) is typical of a decaploid. It
differs from typical F. nevadensis in the cha-
racters emphasized by St.-Yves, and in the
adaxial sclerenchyma strands mostly being
absent or rudimentary. We do not, ho-
wever, consider that these characters merit
the description by St.-Yves of the taxon as
"exactement intermediaire" between F. ru-
bra and F. nevadensis. It would be more
accurate to describe it as intermediate
between F. nevadensis and F. trichophylla
subsp. trichophylla.
We have seen specimens in herb. St.-Yves
(G) collected by Sennen in Pyrénées Orien-
tales,
France (Angoustrine and Odeillo)
that were determined by St.-Yves as F. ru-
bra subsp. nevadensis var. gaetula. We fully
agree with these determinations, and con-
sider that this taxon should be regarded as a
variety of F. nevadensis. We have not seen
specimens from the Spanish Pyrenees, but
surely they exist.
DISCUSSION
As with many groups, taxa in the past
have often been based solely on a few her-
barium specimens that show distinctive cha-
racters, and when a much greater range of
specimens is examined and material is
grown in the botanic garden the distinctions
tend to break down. It has become very
noticeable to us that characters such as scle-
renchyma development, leaf rigidity and
scabridity, leaf-sheath persistence, rhizome
length and panicle size and shape are great-
ly affected by cultural conditions.
Other characters that have been stressed
by some authors in this group, such as
leaf-
blade or sheath pubescence and various spi-
kelet characters, are in our experience not
valuable in taxon recognition as they vary in
a parallel fashion across many taxa. For
example, populations frequently contain
both glabrous and pubescent individuáis;
hybridization experiments have shown
these characters to be genetically fixed, but
they are of no taxonomic valué.
The anatomical characters used by pre-
vious authors to sepárate taxa within this
group have not been found by us to be consis-
tently useful. Bulliform cells are. for exam-
ple,
well developed in at least some indivi-
duáis of all taxa, despite the considerable re-
liance on their presence or absence placed by
ST.-YVES (1913 and later) and MARKGRAF-
DANNENBERG (1980, key), who used the
character also within the F. rubra group.
218ANALES JARDÍN BOTÁNICO
DE
MADRID,
50(2) 1992
The key in Flora Europaea
(MARKGRAF-
DANNENBERG, 1980)
was not
actually cons-
tructed
by
Markgraf-Dannenberg
but was
compiled from
the
descriptions supplied
by
her;
it is
highly unsatisfactory
¡n
many
res-
pects.
The
three characters (sheath decay;
vein number; presence
of
intravaginal
shoots) used
in the key
(couplet 58)
to
sepá-
rate
F.
iberica from
the
rest
of the
taxa
in
this group
are
totally unreliable.
From
our
studies involving many charac-
ters
of a
large number
of
specimens,
we
conclude that
F.
nevadensis,
F.
rothmaleri,
F. paucispicula
and F.
trichophylla
are
four
related species together forming what
we
cali
the F.
trichophylla group. Other taxa
that
we
include within
F.
trichophylla itself
are
F.
iberica,
F.
scabrescens,
F.
pseudotri-
chophylla,
F.
cyrnea
and F.
rubra subsp.
as-
perifolia. These were obviously considered
closely related by
MARKGRAF-DANNENBERG
(1980) also, except
for F.
rubra subsp. aspe-
rifolia, which
was
placed
in the F.
rubra
group,
and F.
iberica, which was placed
out-
side
the
whole
F.
rubra aggregate.
Al-
though
the
latter
two
taxa, especially F.
ibe-
rica,
are
quite distinct
in
their typical
or
extreme state, they
are
linked
to F.
tricho-
phylla
by
intermediates.
It is
unfortunate
that under
our new
classification
the
ñame
iberica
has to be
replaced
by
scabrescens.
Within
F.
trichophylla there
are
three
nodes
of
variation that
we
consider worth
recognizing
as
subspecies. They
are
less
distinct than
the
four species that
we re-
cognize,
but
almost
all
specimens
can be
placed within
one or the
other,
and
there
is
some geographical distinction betwecn
them.
KEY TO TAXA
IN
THE
F.
TRICHOPHYLLA GROUP
1.
Tiller leaf-blades with marginal sclerenchyma
strands distinctly larger than those
of
main
lateral veins
2
1'.
Tiller leaf-blades with marginal sclerenchyma
strands smaller than
to
slightly larger than
those
of
main lateral veins
4
2.
Adaxial ribs
of
leaf-blades sparsely pubes-
cent; panicle rather diffuse
... F.
rothmaleri
2\
Adaxial ribs
of
leaf-blades densely pubes-
cent; panicle rather contracted
3
3.
Tiller leaf-blades with well-developed scle-
renchyma strands
in
adaxial ridges;
leaf-
blades
not or
scarcely angular
in
section
. . .
F. nevadensis
var.
nevadensis
3'.
Tiller leaf-blades without
or
with very sparse
sclerenchyma
in
adaxial ridges; leaf-blades
distinctly angular
in
section
F. nevadensis
var.
gaetula
4.
Panicle with
2-4
spikelets; ovary sparsely
pubescent
F.
paucispicula
4".
Panicle with many spikelets; ovary glabrous
(F. trichophylla)
5
5.
Leaf-diameter mostly 0.65-1.3
mm .... 6
5".
Leaf-diameter mostly 0,3-0.6
mm 7
6. Leaf-blades strongly scabrid throughout
abaxial surface, with scabrid
or
sparsely
pubescent adaxial ridges; stomatal length
(37.5-)41.5-46.5(-50)um
F. trichophylla subsp. asperifolia
6'. Leaf-blades weakly scabrid
or
scabrid only
towards apex
on
abaxial surface, with densely
pubescent adaxial ridges; stomatal length
50(-61.5-)75
um ... F.
nevadensis var. gaetula
7.
Plant distinctly rhizomatous,
not
forming
large tufts; leaves slightly scabrid
or
scabrid
only towards apex
F. trichophylla subsp. trichophylla
7".
Plant
not or
sparsely rhizomatous, forming
large tufts; leaves usually moderately
to
strongly scabrid along whole length
F. trichophylla subsp. scabrescens
TAXONOMIC TREATMENT
OFTHE
F.
TRICHOPHYLLA GROUP
1.
F.
nevadensis (Hackel)
K.
Richter,
Pl.
Eur.
1: 101
(1890)
Basionym:
F.
rubra subsp. nevadensis
Hackel, Monogr. Festuc.
Eur. 146
(1882).
Lectotypified by
FUENTE GARCIA
&
SÁNCHEZ-MATA
(1986a).
a.
var.
nevadensis
(fig. 3B)
Synonym:
F.
rubra [subsp. nevaden-
sis]
var.
hackelii Litard.
&
Maire
ex
Litard., Arch.
Bot. 1
(Bull. Mens.
4):
56 (April 1927),
nom. nov. pro
"subsp. nevadensis Hack.
s.
str.".
South-eastern Spain; north-eastern
Morocco.
b.
var.
gaetula (Maire
ex
St.-Yves)
Al-Bermani
&
Stace, comb. nov.
(fig.
3C)
Basionym:
F.
rubra [subsp. nevaden-
A.-K.
K. A.
AL-BERMANI
&
AL.: FESTUCA TRICHOPHYLLA219
sis]
var.
gaetula Maire
ex
St.-Yves,
Candolleal: 17(1922).
Other synonym: F. gaetula (Maire
ex
St.-Yves) Claustres ex Kerguélen,
Fl.
Descr. Illustr. France, Suppl.
5: 545
(1979).
North-eastern Morocco, Spain, French
Pyrenees.
2.
F.
rothmaleri (Litard.) Markgr.-Dann.,
Bot.
J.
Linn, Soc. 76: 325 (1978)
Basionym:
F.
rubra [subsp. eu-rubra]
var. rothmaleri Litard., Cavanillesia
8:
57 (1936).
Central and north-western Spain; north-
ern Portugal; ? Corsica.
Lectotypified
by
FUENTE GARCÍA
&
SÁNCHEZ-MATA
(1986b).
3.
F.
trichophylla (Ducros
ex
Gaudin)
K. Richter, Pl. Europ.
1:100(1890)
Basionym: F. rubra trichophylla Ducros
ex Gaudin, Fl. Helv. 1: 288 (1828).
Other synonym:
F.
rubra
var.
tricho-
phylla (Ducros
ex
Gaudin) Hackel,
Monogr. Festuc. Eur. 142 (1882).
a. subsp. trichophylla (fig. 3A)
Synonyms:
F.
rubra [subsp. violacea
var.
iberica]
súbvar. subscabra Hackel,
Monogr. Festuc.
Eur. 137
(1882);
F. iberica subsp. subscabra (Hackel)
K. Richter, Pl. Eur. 99 (1890); F.
ru-
bra [subsp. eu-rubra] var. cyrnea St.-
Yves
&
Litard., Bull.
Soc. Bot.
France 71:122 (1924); F. cyrnea (St.-
Yves
&
Litard.) Markgr.-Dann.,
Bot.
J.
Linn. Soc. 76: 327 (1978).
Central
and
south-central Europe,
from Spain
to
Romanía;
?
North
África.
b.
subsp. scabrescens (Hackel
ex Tra-
but) Catalán & Stace, comb. nov.
Basionym:
F.
rubra [subsp.] scabres-
cens Hackel
ex
Trabut,
Fl.
Algérie
Mon. 215 (1895).
Other synonyms:
F.
rubra [subsp.
violacea]
var.
iberica Hackel (incl.
subvar. typica), Monogr. Festuc.
Eur. 136 (1882);
F.
iberica (Hackel)
K. Richter, Pl. Eur. 99 (1890); F. sca-
brescens (Hackel
ex
Trabut) Batt.
&
Trabut,
Fl.
Algérie Tunisie
384
(1904);
F.
rubra [subsp. eu-rubra]
var. yvesiana Litard.
&
Maire, Mém.
Soc.
Sci. Nat.
Maroc
4: 25
(1924);
F. pseudotrichophylla Patzke,
De-
cheniana 117:195 (1964).
Spain and North África; ? France and
Portugal.
c. subsp. asperifolia (St.-Yves) Al-Ber-
mani, comb. nov.
Basionym:
F.
rubra [subsp. eu-rubra
var. genuina] subvar. asperifolia
St.-
Yves,
Annuaire Conserv. Jard.
Bot.
Genéve 17:125 (1913).
Synonym:
F.
rubra subsp. asperifolia
(St.-Yves) Markgr.-Dann.,
Veróff.
Geobot. Inst. Rübel Zürich 56:
143
(1976).
Southern Europe; ? exact range.
4.
F.
paucispicula Fuente García
& Sán-
chez-Mata, Candollea 41: 443 (1986)
Nom. sust.:
F.
rubra subsp. font-queri
Litard., Candollea 10:133 (1945).
Tarragona (Spain).
ACKNOWLEDGEMENTS
We are grateful
to
the curators
of
the herbaria
Usted
for
lending material,
to
Dr.
M.
Kerguélen
for helpful correspondence
and for
providing
viable material
of
F. cyrnea,
and to Dr. A. T.
Romero and Miss E. Ortúñez for donating viable
material
of
F. iberica and F. rothmaleri.
BlBLlOGRAPHIC REFERENCES
BATTANDIER,
J. A. & L. C.
TRABUT
(1904). Flore Ana-
lytique
el
Synoptique
de
l'Algérie
el de la
Tunisie.
Alger.
CLAUSTRES,
G.
(1960). Festuca
des
Pyrénées,
1. La
repartición géographyque
des
taxa dans l'ensemble
de
la
chaine. Bull.
Soc.
Hist.
Nat.
Toulouse
95:
111-123.
FUENTb
GARClA,
V. Dt LA & D.
SÁNCHEZ-MATA
(1986a). Tipificación
de
algunos táxones ibéricos del
género Festuca L. (Gramineae) descritos por E. Hac-
kel.
Candollea
41:
163-171.
FutNTfc
GARCÍA.
V. DE LA & D.
SÁNCHEZ-MATA
(1986b). Datos taxonómico!» sobre
el
género Festuca
L. (Gramineae)
en la
Península Ibérica. Candollea
41:441-448.
FUENTE
GARCÍA,
V. i)fc LA & D.
SÁNCHEZ-MATA
(1987).
Datos sobre Festuca rothmaleri (Litard.)
Markgr.-Dannenb.
y F.
nevadensis (Hackel)
K.
Richter (Gramineae). Anules
Jard.
Bot. Madrid 43:
361-373.
220ANALES JARDÍN BOTÁNICO
DE
MADRID,
50(2) 1992
FUENTE GARCÍA.
V. DE LA & D.
SA\CHEZ-MATA
(1989).
Sobre Festuca rubra
L.
subsp. juncea
(Hac-
kel)
K.
Richter
en la
Península Ibérica. Collect.
Bot.
(Barcelona)
17:
247-253.
FUENTE GARCÍA,
V. DE LA. D.
SANCHEZ-MATA
& J. C.
MORENO
SAIZ(1988).
Sobre
el
género Festuca
L.
(Gramineae). Tipificaciones
en el
herbario original
de
E.
Hackel. Candollea 43:513-520.
HACKEL,
E.
(1882). Monographia Festucarum Euro-
paearum. Kassel
&
Berlín.
KEROUÉLEN,
M.
(1975).
Les
Gramineae (Poaceae)
de la
flore francaise. Essai
de
mise
au
point taxonomique
et nomenclaturale. Lejeunia,
n. ser., 75.
KERGUÉLEN,
M.
(1979).Flore Descriptive
et
Illustrée
de
la France. Cinquiéme Supplement: Graminées.
Paris.
KERGUÉLEN,
M.
(1983).
Les
Graminées
de
France
au travers
de
'Flora Europaea'
et de la
"Flore'
du
C.N.R.S. Lejeunia,
n. ser., 110.
KERGUÉLEN,
M. & F.
PLONKA
(1988). Festuca des
Pyrenées.
Du
nouveau
sur
quelques taxons. Monogr.
Inst.
Pir.
Ecología [Homenaje
a
Pedro Montserrat]
4:
225-229.
KERGUÉLEN,
M. & F.
PLONKA
(1989). Les Festuca de la
Flore
de
France (Corsé Comprise). Bull.
Soc. Bot.
Centre-Ouest,
n. ser.,
num. spéc.
10.
LITARDIERE,
R. DE
(1945). Contribution
á
l'étude
du
genre Festuca. Candollea 10:103-146.
LITARDIERE,
R.
DE (1947). Festuca nouveaux
ou
rares
de France
et
d'Espagne, principalement
des
Pyre-
nées.
Bull.
Soc.
Hist.
Nat.
Toulouse&2: 110-122.
MARKGRAF-DANNENBERG.
I.
(1952). Studien
an
Iris-
chen Festuca-Rassen.
Veróff,
Geobot. Inst. Rübel
Zürich 25: 114-142.
MARKGRAF-DANNENBERG.
I.
(1976).
Die
Gattung
Fes-
tuca
in
Griechenland.
Veroff.
Geobot. Inst. Rubel
Zürich
56:
92-182.
MARKGRAF-DANNENBERG.
I.
(1978).
New
taxa
and
ñames
in
European Festuca (Gramineae).
Bot. J.
Linn.
Soc. 76:
322-328.
MARKGRAF-DANNENBERG.
I.
(1980). Festuca. //i.T.
G.
Tutin
& al.
(eds.). Flora Europaea
5:
125-153.
Cam-
bridge.
MARKGRAF-DANNENBERG.
I.
(1982). Festuca.
In:
S. Pignatti (ed.). Flora
a"Italia
3:
478-501.
Bologna.
PATZKE.
E.
(1964). Gliederung
der
Festuca rubra
L. in
Deutschland. Dechemana
117:
191-196.
RICHTER.
K.
(1890). Plantae Europeae
1.
Leipzig.
ST.-YVES,
A.
(1913).
Les
Festuca
de la
section Eu-Fes-
tuca
et
leurs variations dans
les
Alpes Maritimes.
AnnuaireConserv. Jará.
Bot.
Genéve
17:
1-218.
ST.-YVES,
A.
(1922).
Les
Festuca (subg. Eu-Festuca)
de
l'Afrique
du
Nord
et des
lies Atlantiques. Candollea
1:1-63.
ST.-YVES,
A.
(1924). Festucarum varietates novae
(subg. Eu-Festuca). Bull.
Soc. Bot.
FrancelX:
28-43.
119-135.
TRABUT.
L. C.
(1985).
In: J. A.
Battandier
& L. C. Tra-
but. Flore
de
l'Algérie.
2.
Monocotylédones. Paris
&
Alger.
Aceptado para publicación: 18-VI1I-1992
