4
C. den Hartog and J. Kuo
usually not developing; endosperm absent. Tannin
cells absent.
The family consists of three genera, viz. Zostera,
Heterozostera, and Phyllospadix.
There is no doubt about the monophyletic status of
the family Zosteraceae. This has already been concluded by Tomlinson (1982) on merely anatomical
and morphological grounds. A further confirmation
comes from molecular phylogenetic studies of the
families of the subclass Alismatidae, using chloroplast rbcL (Les et al., 1997; Procaccini et al., 1999a).
Les et al. (1997) demonstrated that the Zosteraceae
are more closely related to the Potamogetonaceae
and the Zannichelliaceae than with the other seagrass groups. These studies, however, did not support
the recognition of Heterozostera tasmanica as representing a distinct genus, but accepted it as a distinct
species within the subgenus Zosterella (Les et al.,
1997, 2002). Based on the matK gene sequence data,
Tanaka et al. (2003) also show a similar result. On the
other hand, Kato et al. (2003) proposed to divide the
Zosteraceae into three genera: Phyllospadix, Zostera
and Nanozostera, the genus Nanozostera containing
two subgenera, Zosterella and Heterozostera. In this
case, by priority Heterozostera, which was established more than 30 years earlier, should be used as
the generic name instead of Nanozostera. There is
nothing known about the possible ancestors of the
family. Originally it was thought that Archeozostera
(Koriba and Miki, 1931, 1960) from the Cretaceous
of Japan was a protozosterid (den Hartog, 1970),
but Kuo et al. (1989) have shown convincingly that
Archeozostera is not a seagrass at all, and possibly
not even a plant.
Key to the Genera
1a. Rhizome monopodial, herbaceous, with two
vascular bundles in the cortical layer; a short
lateral branch at each node. . . . . . . . . . . . . . . . . 2
1b. Rhizome sympodial, but sometimes monopodial, often ligneous, with 4–10 vascular bundles
in the cortical layer. Stems erect, not branched,
with distichous leaves . . . . . . . . . Heterozostera
2a. Monoecious. Spadix always enclosed within the
spathal sheath. Retinacula, if present, membranous, not nerved, smaller than the ovaries or
stamens. Peduncle of the spathe partly coalescent with the axis from which it springs. Fruit
ovoid or ellipsoid. Rhizome with elongate internodes, and with two, or more, long, thin
roots at each node. Vegetative parts not sclerenchymatic. Leaf-blades thin, translucent; margin entire or rarely slightly denticulate. Sheaths
deciduous; sometimes some scaly basal parts
remaining. . . . . . . . . . . . . . . . . . . . . . . . . . Zostera
2b. Dioecious (rarely monoecious). Spadix, when
mature, projecting out of the spathal sheath.
Retinacula coriaceous, one-nerved, larger than
the ovaries or stamens. Peduncle free. Fruit
crescent-shaped. Rhizome with very short, thick
internodes, and with two or more thick roots
at each node. Vegetative parts usually strongly
sclerenchymatic. Leaf-blades coriaceous, with
‘fin cells’ along the margin. Basal parts of the
sheaths decaying with age into bundles of very
fine, woolly fibres . . . . . . . . . . . . . Phyllospadix
Zostera L. Sp. Pl. ed. 1 (1753) 986.
Type species: Zostera marina L.
The genus consists of two well-distinguished subgenera, Zostera and Zosterella.
In subgen. Zostera the rhizome has the fibre bundles in the outermost part of the cortex, and there
are always two groups of roots at each node; the
leaf-sheaths are tubular and rupturing with age; the
generative shoots are terminal and retinacula are absent. In the old literature this subgenus is referred to
as subgen. Alega; according to the rules of botanical
nomenclature the correct name is subgen. Zostera,
as it contains the type of the genus. In subgen.
Zosterella (Ascherson) Ostenfeld (type species Z.
nana Mertens ex Roth = Zostera noltii Hornemann)
the fibre bundles occur in the innermost layers of
the outer cortex of the rhizome; the leaf-sheaths
are open with two membranous flaps; the generative
shoots develop lateral and in the inflorescences the
retinacula are always present. Recently, Tomlinson
and Posluszny (2001) upgraded the latter subgenus
to a genus in its own right, Nanozostera, mainly because they considered the differences between this
taxon and the subgenus Zostera of the same order as
its differences with Heterozostera. Meanwhile Kuo
(2005) has discovered that the genus Heterozostera
is also not homogeneous from a morphological point
of view.
The subgenus Zostera is widely distributed in the
northern temperate coastal waters of the Atlantic and
the Pacific. Up to now four species have been accepted (den Hartog, 1970), but a further analysis
may show that this number is too low. At present
C. den Hartog and J. Kuo
usually not developing; endosperm absent. Tannin
cells absent.
The family consists of three genera, viz. Zostera,
Heterozostera, and Phyllospadix.
There is no doubt about the monophyletic status of
the family Zosteraceae. This has already been concluded by Tomlinson (1982) on merely anatomical
and morphological grounds. A further confirmation
comes from molecular phylogenetic studies of the
families of the subclass Alismatidae, using chloroplast rbcL (Les et al., 1997; Procaccini et al., 1999a).
Les et al. (1997) demonstrated that the Zosteraceae
are more closely related to the Potamogetonaceae
and the Zannichelliaceae than with the other seagrass groups. These studies, however, did not support
the recognition of Heterozostera tasmanica as representing a distinct genus, but accepted it as a distinct
species within the subgenus Zosterella (Les et al.,
1997, 2002). Based on the matK gene sequence data,
Tanaka et al. (2003) also show a similar result. On the
other hand, Kato et al. (2003) proposed to divide the
Zosteraceae into three genera: Phyllospadix, Zostera
and Nanozostera, the genus Nanozostera containing
two subgenera, Zosterella and Heterozostera. In this
case, by priority Heterozostera, which was established more than 30 years earlier, should be used as
the generic name instead of Nanozostera. There is
nothing known about the possible ancestors of the
family. Originally it was thought that Archeozostera
(Koriba and Miki, 1931, 1960) from the Cretaceous
of Japan was a protozosterid (den Hartog, 1970),
but Kuo et al. (1989) have shown convincingly that
Archeozostera is not a seagrass at all, and possibly
not even a plant.
Key to the Genera
1a. Rhizome monopodial, herbaceous, with two
vascular bundles in the cortical layer; a short
lateral branch at each node. . . . . . . . . . . . . . . . . 2
1b. Rhizome sympodial, but sometimes monopodial, often ligneous, with 4–10 vascular bundles
in the cortical layer. Stems erect, not branched,
with distichous leaves . . . . . . . . . Heterozostera
2a. Monoecious. Spadix always enclosed within the
spathal sheath. Retinacula, if present, membranous, not nerved, smaller than the ovaries or
stamens. Peduncle of the spathe partly coalescent with the axis from which it springs. Fruit
ovoid or ellipsoid. Rhizome with elongate internodes, and with two, or more, long, thin
roots at each node. Vegetative parts not sclerenchymatic. Leaf-blades thin, translucent; margin entire or rarely slightly denticulate. Sheaths
deciduous; sometimes some scaly basal parts
remaining. . . . . . . . . . . . . . . . . . . . . . . . . . Zostera
2b. Dioecious (rarely monoecious). Spadix, when
mature, projecting out of the spathal sheath.
Retinacula coriaceous, one-nerved, larger than
the ovaries or stamens. Peduncle free. Fruit
crescent-shaped. Rhizome with very short, thick
internodes, and with two or more thick roots
at each node. Vegetative parts usually strongly
sclerenchymatic. Leaf-blades coriaceous, with
‘fin cells’ along the margin. Basal parts of the
sheaths decaying with age into bundles of very
fine, woolly fibres . . . . . . . . . . . . . Phyllospadix
Zostera L. Sp. Pl. ed. 1 (1753) 986.
Type species: Zostera marina L.
The genus consists of two well-distinguished subgenera, Zostera and Zosterella.
In subgen. Zostera the rhizome has the fibre bundles in the outermost part of the cortex, and there
are always two groups of roots at each node; the
leaf-sheaths are tubular and rupturing with age; the
generative shoots are terminal and retinacula are absent. In the old literature this subgenus is referred to
as subgen. Alega; according to the rules of botanical
nomenclature the correct name is subgen. Zostera,
as it contains the type of the genus. In subgen.
Zosterella (Ascherson) Ostenfeld (type species Z.
nana Mertens ex Roth = Zostera noltii Hornemann)
the fibre bundles occur in the innermost layers of
the outer cortex of the rhizome; the leaf-sheaths
are open with two membranous flaps; the generative
shoots develop lateral and in the inflorescences the
retinacula are always present. Recently, Tomlinson
and Posluszny (2001) upgraded the latter subgenus
to a genus in its own right, Nanozostera, mainly because they considered the differences between this
taxon and the subgenus Zostera of the same order as
its differences with Heterozostera. Meanwhile Kuo
(2005) has discovered that the genus Heterozostera
is also not homogeneous from a morphological point
of view.
The subgenus Zostera is widely distributed in the
northern temperate coastal waters of the Atlantic and
the Pacific. Up to now four species have been accepted (den Hartog, 1970), but a further analysis
may show that this number is too low. At present