18
C. den Hartog and J. Kuo
and may be classified into a genus of its own.
Beautifully preserved fossils of C. serrulata have
been collected from the Miocene of Sulawesi (formerly Celebes) as C. micheloti (Laurent and Laurent,
1926).
From these few confirmed fossils it can be concluded that seagrasses already developed at an early
stage of the evolution of the angiosperms. Taking
into consideration that at present three exclusive
seagrass families can be recognized, and a fourth
family with two seagrass subfamilies, the possibility may not be excluded that the evolutionary transition from terrestrial plants to fully submersed marine
plants may have taken place just as many times. Les
et al. (1997) concluded after a phylogenetic analysis that among the seagrasses three lineages can be
recognized giving evidence that these lineages independently entered the marine environment. These
lineages are (i) the Zosteraceae, (ii) the Cymodoceaceae complex consisting of the Posidoniaceae,
the Cymodoceaceae and the Ruppiaceae, and (iii)
the Hydrocharitaceae.
The Ruppiaceae and the Zannichelliaceae occur
in waters with a very diverse chemical composition, such as brackish and continental salt waters,
but also in hard fresh waters. Ruppia species sometimes occur in the marine environment, particularly in the intertidal zone of muddy or sandy flats
where seagrasses are absent; after the introduction
of Zostera japonica in North America Ruppia decreased markedly. In temperate Australia Lepilaena
(Zannichelliaceae) occurs with Ruppia in brackish
waters. Both genera have developed one species that
is fully restricted to very sheltered marine conditions, and these two species generally form together
a community. These species are probably the most
recent seagrasses in the evolutionary history.
It is apparent from the very limited fossil record
that the distribution of the seagrasses must have
been quite different from the present situation, as
is demonstrated by the fossil occurrence of Cymodocea and a Thalassodendron-like species in North
America, genera presently absent from America.
Another fact is that fossil material of Zosteraceae
has not been found in Cretaceous or Tertiary deposits, but only in Quaternary layers; the presence
of Z . noltii in the Caspian and Aral Seas shows,
however, that on the ground of the history of these
seas Zostera must have been in existence already
in the Miocene. Larkum and den Hartog (1989)
have attempted to use all kinds of geological data
to work out the history of the Australian seagrass
flora. The great handicap for this approach remains
the lack of fossil material. Of the 12 presently living seagrass genera some fossil remains of only four
are known, and this is insufficient to elaborate in
a reliable way the areas of origin of the various
families. One would expect that thick seagrass deposits, as presently known from Posidonia species
and Thalassodendron ciliatum, must have been
formed also in the past. Therefore, it is recommended that seagrass researchers should more cooperate with paleontologists when marine geological
deposits are explored.
V. Conclusion and Outlook
We acknowledge the important contribution made
by molecular technology [rbcL, matK (plastid DNA)
and ITS (nuclear DNA) gene sequences] in our understanding of phylogenetic relationships of seagrasses, particularly at the higher levels for which
this approach has been developed (the reader is referred to Waycott et al., Chapter 2, for an in-depth review of this topic). However, this powerful tool may
not always be suitable for defining the species and
cannot replace the morphological characters generally used for species identification. To consider taxa
which cannot be separated by the present molecular techniques as ‘phenotypic plasticity’ of a wider
molecularly defined ‘biological species’ is not really
a solution of the problem; new advanced techniques
may show that they are indeed different. There is a
great need to correlate the molecular data to morphological and physiological data.
In this connection we refer to the works of
McMillan, who studied isozymes (McMillan, 1980,
1982, 1991; McMillan and Williams, 1980) and sulfated flavonoids (McMillan et al., 1980; McMillan,
1983, 1986) in many seagrass species, but he generally did not give morphological descriptions of
the material used, with the exception of the taxa
within the genus Halophila. In the latter case the
isozyme and flavone patterns seem characteristic at
the species level, but unfortunately the number of
observations is low. More research is necessary, as
McMillan and Williams (1980) clearly state.
Admittedly the number of useful morphological characters that can be used to identify the seagrass species is very limited due to the relatively
simple morphological and anatomical features in
C. den Hartog and J. Kuo
and may be classified into a genus of its own.
Beautifully preserved fossils of C. serrulata have
been collected from the Miocene of Sulawesi (formerly Celebes) as C. micheloti (Laurent and Laurent,
1926).
From these few confirmed fossils it can be concluded that seagrasses already developed at an early
stage of the evolution of the angiosperms. Taking
into consideration that at present three exclusive
seagrass families can be recognized, and a fourth
family with two seagrass subfamilies, the possibility may not be excluded that the evolutionary transition from terrestrial plants to fully submersed marine
plants may have taken place just as many times. Les
et al. (1997) concluded after a phylogenetic analysis that among the seagrasses three lineages can be
recognized giving evidence that these lineages independently entered the marine environment. These
lineages are (i) the Zosteraceae, (ii) the Cymodoceaceae complex consisting of the Posidoniaceae,
the Cymodoceaceae and the Ruppiaceae, and (iii)
the Hydrocharitaceae.
The Ruppiaceae and the Zannichelliaceae occur
in waters with a very diverse chemical composition, such as brackish and continental salt waters,
but also in hard fresh waters. Ruppia species sometimes occur in the marine environment, particularly in the intertidal zone of muddy or sandy flats
where seagrasses are absent; after the introduction
of Zostera japonica in North America Ruppia decreased markedly. In temperate Australia Lepilaena
(Zannichelliaceae) occurs with Ruppia in brackish
waters. Both genera have developed one species that
is fully restricted to very sheltered marine conditions, and these two species generally form together
a community. These species are probably the most
recent seagrasses in the evolutionary history.
It is apparent from the very limited fossil record
that the distribution of the seagrasses must have
been quite different from the present situation, as
is demonstrated by the fossil occurrence of Cymodocea and a Thalassodendron-like species in North
America, genera presently absent from America.
Another fact is that fossil material of Zosteraceae
has not been found in Cretaceous or Tertiary deposits, but only in Quaternary layers; the presence
of Z . noltii in the Caspian and Aral Seas shows,
however, that on the ground of the history of these
seas Zostera must have been in existence already
in the Miocene. Larkum and den Hartog (1989)
have attempted to use all kinds of geological data
to work out the history of the Australian seagrass
flora. The great handicap for this approach remains
the lack of fossil material. Of the 12 presently living seagrass genera some fossil remains of only four
are known, and this is insufficient to elaborate in
a reliable way the areas of origin of the various
families. One would expect that thick seagrass deposits, as presently known from Posidonia species
and Thalassodendron ciliatum, must have been
formed also in the past. Therefore, it is recommended that seagrass researchers should more cooperate with paleontologists when marine geological
deposits are explored.
V. Conclusion and Outlook
We acknowledge the important contribution made
by molecular technology [rbcL, matK (plastid DNA)
and ITS (nuclear DNA) gene sequences] in our understanding of phylogenetic relationships of seagrasses, particularly at the higher levels for which
this approach has been developed (the reader is referred to Waycott et al., Chapter 2, for an in-depth review of this topic). However, this powerful tool may
not always be suitable for defining the species and
cannot replace the morphological characters generally used for species identification. To consider taxa
which cannot be separated by the present molecular techniques as ‘phenotypic plasticity’ of a wider
molecularly defined ‘biological species’ is not really
a solution of the problem; new advanced techniques
may show that they are indeed different. There is a
great need to correlate the molecular data to morphological and physiological data.
In this connection we refer to the works of
McMillan, who studied isozymes (McMillan, 1980,
1982, 1991; McMillan and Williams, 1980) and sulfated flavonoids (McMillan et al., 1980; McMillan,
1983, 1986) in many seagrass species, but he generally did not give morphological descriptions of
the material used, with the exception of the taxa
within the genus Halophila. In the latter case the
isozyme and flavone patterns seem characteristic at
the species level, but unfortunately the number of
observations is low. More research is necessary, as
McMillan and Williams (1980) clearly state.
Admittedly the number of useful morphological characters that can be used to identify the seagrass species is very limited due to the relatively
simple morphological and anatomical features in