Geoscience Reference
In-Depth Information
present in Warsteinia paprothii Rowe). The taxon-
omy of those early seed representatives is based
mainly on variation in the integument, including
various degrees of fusion of the integumentary
lobes and/or characteristics of the cupule. The
apical modification of the nucellus is remarkably
uniform (see Fig. 3). The nucellar apex is modified
into a pollen chamber (Gordon 1941) closed by a
pollen chamber floor and extended by a cylindrical
structure (Hilton & Edwards 1996) referred to as a
salpinx (Gordon 1941); the pollen chamber contains
a central dome of parenchymatous cells. This com-
bination of traits characterizes the hydrasperman
reproductive syndrome (Rothwell 1986; Rothwell
& Scheckler 1988).
The Moresnetia-type includes Archaeosperma
arnoldii (Pettitt & Beck 1968), Elkinsia polymorpha
(Rothwell et al. 1989 emend. Serbet & Rothwell
1992), Glamorgania gayerii (Hilton 2006), Kerrya
mattenii (Rothwell &Wight 1989), Lenlogia krysto-
fovichii (Petrosyan in Lepekhina et al. 1962 emend.
Krassilov & Zakharova 1995), Moresnetia zalesskyi
(Stockmans 1948 emend. Fairon-Demaret &
Scheckler 1987) and Xenotheca devonica (Arber
& Goode 1915 emend. Hilton & Edwards 1999). It
is characterized by a four-unit cupule, formed by
two successive cruciate dichotomous divisions
(Prestianni 2005) (Fig. 3a). Cupules define a well-
circumscribed space where up to four seeds (excep-
tionally six, Matten et al. 1980) are found. The
hydrasperman nucellus is surrounded by several
variably fused integumentary lobes. Which exact
part of this plant is dispersed remains unknown.
The fossil record includes Moresnetia-type parts
ranging from large branching systems to isolated
cupules. The branching systems are probably
not the propagules (dispersed part of the plant).
Rather, they represent whole plants that have been
buried before dispersal of their seeds (Fairon-
Demaret & Scheckler 1987). Indeed, a close
observation of the cupules shows that many of
them lack seeds (Fairon-Demaret & Scheckler
1987). Additionally, the collected branching
systems are of variable size and range from 1 time
dichotomizing axe to up to 15 times dichotomizing
axis, highlighting the random break of these axes.
These observations suggest that seeds are presum-
ably the propagules of the Moresnetia-type plants.
No dispersed seeds of this type have yet been recog-
nized in the fossil record, however, but this may
reflect sedimentation and/or collecting bias. The
cupules and integuments of all the plants of the
Moresnetia-type do not show clear dispersion
traits. A notable exception is the seed from Port
Alleghany Archaeosperma arnoldii, where the
lower part of the integument is covered by spines
or hairs which have been interpreted as being
related to dispersal (Pettitt & Beck 1968).
As with most of the seeds of the Moresnetia-
type, no clear modification for dispersal is observed
in Aglosperma-type seeds, comprising Aglosperma
quadrapartita (Hilton & Edwards 1996), Aglos-
perma avonensis (Hilton 1998b), Pseudosporogo-
nites hallei (Stockmans 1948) and Xenotheca
bertrandii (Stockmans 1948) (Fig. 3d). In this
type, the integument is formed by three to four flat
lobes fused up to their lower third. The cupule is a
tiny structure that inconspicuously covers the base
of the integument. Similarly, many extant seeds
(or fruit) do not present any evident morphological
traits (Ridley 1930).
Dispersal of these may be performed by rain or
wind in the case of sufficiently small seeds. Zooch-
ory (dispersal by animals) cannot be dismissed.
During the Late Devonian, arthropods are the only
putative dispersal candidates as vertebrates were
confined to aquatic environments. Labandeira
(2006) has divided the evolution of plant-arthropod
associations into four phases based on functional
feeding groups. The Famennian conforms to the
end of the first phase (420-360 Ma). This phase
is characterized by small herbivorous myriapods
and apterygotes. Since these two groups do not
have a strong dispersal potential and no adaptation
to arthropod attraction is observed on seeds, we con-
sider zoochory unlikely in early seed plant dispersal.
Other seed types suggest adaptations for wind
dispersal. Seeds of the Dorinnotheca-type are
represented by a single species, Dorinnotheca stree-
lii (Fairon-Demaret 1996b). They show an ovular
morphology similar to that of the Aglosperma-type
with three to four lobes (Fig. 3b). However, the
cupule is much larger and is composed of eight
proximally fused parts forming a cup, with the
distal segments divided into at least 40 free
endings. In this, the cupule is interpreted to be part
of the propagule as all dispersed seeds were still
closely attached to it. This dispersal apparatus can
be compared to the Corolline or Sepaline fruits
described by Ridley (1930), which are related to
wind dispersal.
Similarly, seeds of the Condrusia-type -
Condrusia rumex (Stockmans 1948), C. minor
(Stockmans 1948) and C. brevis (Petrosyan in
Lepekhina et al. 1962) - are enclosed in a well-
developed wing which is composed of two flat
cupule segments adpressed against each other
(Fig. 3e). This wing is interpreted as a dispersal
apparatus. Finally, the Warsteinia-type, represented
by Warsteinia paprothii (Rowe 1997), has been
found devoid of cupules and hence is considered
as acupulate (Fig. 3c). The integument is made up
of four winged lobes adnate to the nucellus. The
presence of integumentary wings is strongly sugges-
tive of anemochory, even although hydrochory
cannot be totally dismissed.
Search WWH ::




Custom Search