126
R. J. Orth, M. C. Harwell and G. J. Inglis
Most programs have involved the transplantation of
adult ramets; use of seeds has been rare (Fonseca
et al., 1998). We believe that seeds should be incorporated into future restoration programs for a number of reasons. Many species produce regular crops
of large numbers of seeds that can be easily harvested
and stored. However, there is great variation among
species and meadows in the regularity with which
seeds are produced. Some of the larger tropical seagrasses appear to produce seeds infrequently (see
Duarte et al., 1997). Flowering in Posidonia oceanica, the dominant seagrass species in the Mediterranean, is apparently very rare in most locations
(Balestri and Cinelli, 2003). Because of this variation, identification of suitable source populations
and their reproductive phenology is a necessary first
step. Harvesting seeds has less impact on a donor
bed than harvesting adult plants and may have no
significant impact on long-term survival of the donor
bed. However, studies on the impacts of long-term
removal of seeds from an area are non-existant and
if restoration programs begin to target areas for seed
removal, it will be important to monitor not only the
donor beds but adjacent areas that may be recipient
sites for seeds that were exported from surrounding areas. Meadows created from transplanted ramets tend to be genetically more homogenous, grow
slower, produce fewer seeds and have poorer rates
of germination than more genetically diverse stands
(Williams, 2001). Although natural seed mortality
is generally high (Harrison, 1993; Orth et al., 2003),
the potential exists for increasing survivorship with
innovative techniques (e.g. Harwell and Orth, 1999).
Seeds have the potential for accelerating or enhancing restoration by establishing new patches, adding
to slowly recovering areas, or increasing species and
genetic diversity of meadows. Using seeds, possibly
in conjunction with adult plants, may prove to be
more cost and labor effective than using adult plants
in some seagrass species.
VII. Conclusions and Emerging Paradigms
Seed ecology represents an exciting opportunity for
seagrass research. Many of the intriguing issues discussed in terrestrial environments are likely to be
applicable for seagrass species. Seeds are likely to
be significantly more important in the dynamics of
bed development than previously believed. However,
generalizations may fail in extreme ranges of populations where flowering may be limited and asexual
propagation considered more important in meadow
maintenance and spread (Billingham et al., 2003).
New paradigms will emerge as the field evolves and
as questions are answered about the role seeds play
for different species in determining dispersal, genetic diversity and evolution (Reusch, 2002).
A. Dispersal
Dispersal, in particular, has been the subject of a
number of recent in-depth review articles in the
terrestrial literature, pointing to challenging areas
of research such as managing pristine, degraded
or restored areas as sources of diaspores, recognition of sites as sources or sinks of propagules, etc.
(Clark et al., 1998; Howe and Miriti, 2000; Nathan
and Muller-Landau, 2000; Wang and Smith, 2002;
Higgins et al., 2003; Levin et al., 2003). Emerging technologies are allowing us to address previously intractable questions about dispersal and connectivity within populations (e.g. molecular genetic
techniques, stable isotopes, radioactive labeling,
fluorescent microspheres) (Wang and Smith, 2002).
The use of ecological simulation models to examine
seagrass reproduction (Bearlin et al., 1999; Harwell,
2000) and geographic information systems (GIS) to
examine patch dynamics (e.g. Wilcox et al., 2000;
Duarte et al., Chapter 11) can be valuable tools
for bridging the gap between theoretical and actual
population and metapopulation dynamics. Longdistance dispersal has now been demonstrated in at
least four species, and it is possible that this phenomenon will be found in most, if not all, seagrass
genera although the mechanisms may be dramatically different. We predict that future studies will
demonstrate that seeds on the sediment surface do
not move far from where they settle except as a result
of extreme catastrophic events.
B. Dispersal Processes
Understanding the different aspects of dispersal
(Phase I and II, Chambers and MacMahon (1994)
will be very important in understanding actual dispersal distances and the speciation of seagrasses as
populations which may not be as isolated in geologic
history as once thought. Research should focus on
the relative proportion of seeds that are both retained
R. J. Orth, M. C. Harwell and G. J. Inglis
Most programs have involved the transplantation of
adult ramets; use of seeds has been rare (Fonseca
et al., 1998). We believe that seeds should be incorporated into future restoration programs for a number of reasons. Many species produce regular crops
of large numbers of seeds that can be easily harvested
and stored. However, there is great variation among
species and meadows in the regularity with which
seeds are produced. Some of the larger tropical seagrasses appear to produce seeds infrequently (see
Duarte et al., 1997). Flowering in Posidonia oceanica, the dominant seagrass species in the Mediterranean, is apparently very rare in most locations
(Balestri and Cinelli, 2003). Because of this variation, identification of suitable source populations
and their reproductive phenology is a necessary first
step. Harvesting seeds has less impact on a donor
bed than harvesting adult plants and may have no
significant impact on long-term survival of the donor
bed. However, studies on the impacts of long-term
removal of seeds from an area are non-existant and
if restoration programs begin to target areas for seed
removal, it will be important to monitor not only the
donor beds but adjacent areas that may be recipient
sites for seeds that were exported from surrounding areas. Meadows created from transplanted ramets tend to be genetically more homogenous, grow
slower, produce fewer seeds and have poorer rates
of germination than more genetically diverse stands
(Williams, 2001). Although natural seed mortality
is generally high (Harrison, 1993; Orth et al., 2003),
the potential exists for increasing survivorship with
innovative techniques (e.g. Harwell and Orth, 1999).
Seeds have the potential for accelerating or enhancing restoration by establishing new patches, adding
to slowly recovering areas, or increasing species and
genetic diversity of meadows. Using seeds, possibly
in conjunction with adult plants, may prove to be
more cost and labor effective than using adult plants
in some seagrass species.
VII. Conclusions and Emerging Paradigms
Seed ecology represents an exciting opportunity for
seagrass research. Many of the intriguing issues discussed in terrestrial environments are likely to be
applicable for seagrass species. Seeds are likely to
be significantly more important in the dynamics of
bed development than previously believed. However,
generalizations may fail in extreme ranges of populations where flowering may be limited and asexual
propagation considered more important in meadow
maintenance and spread (Billingham et al., 2003).
New paradigms will emerge as the field evolves and
as questions are answered about the role seeds play
for different species in determining dispersal, genetic diversity and evolution (Reusch, 2002).
A. Dispersal
Dispersal, in particular, has been the subject of a
number of recent in-depth review articles in the
terrestrial literature, pointing to challenging areas
of research such as managing pristine, degraded
or restored areas as sources of diaspores, recognition of sites as sources or sinks of propagules, etc.
(Clark et al., 1998; Howe and Miriti, 2000; Nathan
and Muller-Landau, 2000; Wang and Smith, 2002;
Higgins et al., 2003; Levin et al., 2003). Emerging technologies are allowing us to address previously intractable questions about dispersal and connectivity within populations (e.g. molecular genetic
techniques, stable isotopes, radioactive labeling,
fluorescent microspheres) (Wang and Smith, 2002).
The use of ecological simulation models to examine
seagrass reproduction (Bearlin et al., 1999; Harwell,
2000) and geographic information systems (GIS) to
examine patch dynamics (e.g. Wilcox et al., 2000;
Duarte et al., Chapter 11) can be valuable tools
for bridging the gap between theoretical and actual
population and metapopulation dynamics. Longdistance dispersal has now been demonstrated in at
least four species, and it is possible that this phenomenon will be found in most, if not all, seagrass
genera although the mechanisms may be dramatically different. We predict that future studies will
demonstrate that seeds on the sediment surface do
not move far from where they settle except as a result
of extreme catastrophic events.
B. Dispersal Processes
Understanding the different aspects of dispersal
(Phase I and II, Chambers and MacMahon (1994)
will be very important in understanding actual dispersal distances and the speciation of seagrasses as
populations which may not be as isolated in geologic
history as once thought. Research should focus on
the relative proportion of seeds that are both retained