Chapter 5 Seagrass Seeds and Dispersal Strategies
113
1. Species with seeds that have a fleshy or membranous seed coat and no distinct dormancy period
(e.g. Enhalus, Posidonia spp., Thalassia spp.),
2. species with seeds with a hard seed coat and
distinct dormancy period (e.g. Zostera spp.,
Halophila spp.), and
3. species with seeds that do not have a seed coat,
no distinct dormancy period, and where seedlings
develop for some time period on the parent plant
(e.g. Amphibolis spp., Thalassodendron spp.).
Seagrasses have evolved a variety of reproductive
strategies (dioecy, monoecy, and hermaphroditism)
and functional morphologies to insure successful
pollination in an aquatic environment (Cox, 1988;
Ackerman, 1995, 1997a,b; 2000; Ackerman, Chapter 4, this volume). Bottlenecks for successful pollination are expected to be more prominent in dioecious species (over 60% of seagrass species are
dioecious) where rarity of either sex could lead
to the reduction or absence of pollen or flowers,
or in species where pollen dispersal occurs over
short distances (Cox, 1988; Ruckelhaus, 1996), although recent research challenges the latter (Ackerman, 2002). Seed production would obviously be
problematic where meadows are entirely dominated
by one sex, or in beds with asynchronous flowering
(Caye and Meinesz, 1985; Buia and Mazzella, 1991;
Terrados, 1993). However, in the only study to examine the influence of unequal sex ratios on seed production, Williams (1995) showed that, despite the
rarity of male flowers in a sub-tidal population of
Phyllospadix torreyi, almost all ovules had been fertilized, most likely facilitated by the strong mixing
characteristics of the surrounding environment.
Seagrass seed size varies by two orders of magnitude among species, ranging from 0.2 mm in
Halophila to 20 mm in Posidonia. Most seagrass
genera produce only one seed per fruit, but some,
such as Halophila, set up to 60 seeds per fruit, although there is some variation within particular genera for both seed size and the number of seeds per
fruit (Table 1). Moles and Westoby (2003) showed
a ten-fold reduction in mean seed mass of terrestrial plants for every 23˚ increase in latitude (they
did show a wide range of seed mass within a latitude), possibly due to habitat type, plant growth
form, seed disperser assemblage, length of the growing period, or net primary productivity. Comparable
information for seagrass species is sparse (WyllieEcheverria et al., 2003).
Seagrass seed production (densities m
−2 ) vary by
more than two orders of magnitude among the various genera (Table 1) with the largest number produced by those species that have relatively hard seed
coats and some period of dormancy. In terrestrial
plants, significant variation in seed production occurs in space and time (e.g. masting), so that seed
production in some places and years contribute to
more long-term population success than others. Similarly, large interannual (Orth and Moore, 1986; Inglis and Lincoln Smith, 1998) and spatial (Inglis
and Lincoln Smith, 1998; Inglis, 2000b) fluctuations
have been recorded in the reproductive output of
seagrasses, but the relationship between seed production and seedling recruitment is unclear.
III. Seed Banks, Dormancy,
and Seed Germination
Published estimates of sediment seed bank densities in seagrass communities are generally comparable to those of most plant community types
(Fenner, 1995; Orth et al., 2000). Like seed banks
in terrestrial systems, they are highly variable both
in space and time and may bear little or no relationship to the spatial distribution of plants that produce
the seeds. Harrison (1993) documented the spatial
and temporal dynamics of a Zostera marina seed
bank in The Netherlands, showing high variability
in a single population. A significant portion of the
seeds in the seed bank was lost to natural mortality
and a small percentage of the seeds contributed to
a persistent seed bank. Inglis (2000b) found significant heterogeneity in Halodule uninervis seed banks
at multiple spatial scales and reported “no obvious
spatial relationships between densities of flowering
shoots, vegetative shoots, developing ovules, and
seed bank densities that might indicate consistent
patches of seed production and storage.” Harwell
and Orth (2002b) also found extreme variability in
Z. marina seed banks similar to those of Inglis
(2000b). Future studies of seagrass seed banks must
recognize the dynamic nature of a seed bank and that
the vagaries of currents, storm events, waves, and
bottom structure probably determine more about the
location and density of seed deposition than do floral
characteristics, i.e. flowering shoot density (Inglis,
2000b; Inglis and Waycott, 2001).
Seed banks are often classified as transient
(turnover in less than one year) or persistent (remaining viable longer than one year) according to how
113
1. Species with seeds that have a fleshy or membranous seed coat and no distinct dormancy period
(e.g. Enhalus, Posidonia spp., Thalassia spp.),
2. species with seeds with a hard seed coat and
distinct dormancy period (e.g. Zostera spp.,
Halophila spp.), and
3. species with seeds that do not have a seed coat,
no distinct dormancy period, and where seedlings
develop for some time period on the parent plant
(e.g. Amphibolis spp., Thalassodendron spp.).
Seagrasses have evolved a variety of reproductive
strategies (dioecy, monoecy, and hermaphroditism)
and functional morphologies to insure successful
pollination in an aquatic environment (Cox, 1988;
Ackerman, 1995, 1997a,b; 2000; Ackerman, Chapter 4, this volume). Bottlenecks for successful pollination are expected to be more prominent in dioecious species (over 60% of seagrass species are
dioecious) where rarity of either sex could lead
to the reduction or absence of pollen or flowers,
or in species where pollen dispersal occurs over
short distances (Cox, 1988; Ruckelhaus, 1996), although recent research challenges the latter (Ackerman, 2002). Seed production would obviously be
problematic where meadows are entirely dominated
by one sex, or in beds with asynchronous flowering
(Caye and Meinesz, 1985; Buia and Mazzella, 1991;
Terrados, 1993). However, in the only study to examine the influence of unequal sex ratios on seed production, Williams (1995) showed that, despite the
rarity of male flowers in a sub-tidal population of
Phyllospadix torreyi, almost all ovules had been fertilized, most likely facilitated by the strong mixing
characteristics of the surrounding environment.
Seagrass seed size varies by two orders of magnitude among species, ranging from 0.2 mm in
Halophila to 20 mm in Posidonia. Most seagrass
genera produce only one seed per fruit, but some,
such as Halophila, set up to 60 seeds per fruit, although there is some variation within particular genera for both seed size and the number of seeds per
fruit (Table 1). Moles and Westoby (2003) showed
a ten-fold reduction in mean seed mass of terrestrial plants for every 23˚ increase in latitude (they
did show a wide range of seed mass within a latitude), possibly due to habitat type, plant growth
form, seed disperser assemblage, length of the growing period, or net primary productivity. Comparable
information for seagrass species is sparse (WyllieEcheverria et al., 2003).
Seagrass seed production (densities m
−2 ) vary by
more than two orders of magnitude among the various genera (Table 1) with the largest number produced by those species that have relatively hard seed
coats and some period of dormancy. In terrestrial
plants, significant variation in seed production occurs in space and time (e.g. masting), so that seed
production in some places and years contribute to
more long-term population success than others. Similarly, large interannual (Orth and Moore, 1986; Inglis and Lincoln Smith, 1998) and spatial (Inglis
and Lincoln Smith, 1998; Inglis, 2000b) fluctuations
have been recorded in the reproductive output of
seagrasses, but the relationship between seed production and seedling recruitment is unclear.
III. Seed Banks, Dormancy,
and Seed Germination
Published estimates of sediment seed bank densities in seagrass communities are generally comparable to those of most plant community types
(Fenner, 1995; Orth et al., 2000). Like seed banks
in terrestrial systems, they are highly variable both
in space and time and may bear little or no relationship to the spatial distribution of plants that produce
the seeds. Harrison (1993) documented the spatial
and temporal dynamics of a Zostera marina seed
bank in The Netherlands, showing high variability
in a single population. A significant portion of the
seeds in the seed bank was lost to natural mortality
and a small percentage of the seeds contributed to
a persistent seed bank. Inglis (2000b) found significant heterogeneity in Halodule uninervis seed banks
at multiple spatial scales and reported “no obvious
spatial relationships between densities of flowering
shoots, vegetative shoots, developing ovules, and
seed bank densities that might indicate consistent
patches of seed production and storage.” Harwell
and Orth (2002b) also found extreme variability in
Z. marina seed banks similar to those of Inglis
(2000b). Future studies of seagrass seed banks must
recognize the dynamic nature of a seed bank and that
the vagaries of currents, storm events, waves, and
bottom structure probably determine more about the
location and density of seed deposition than do floral
characteristics, i.e. flowering shoot density (Inglis,
2000b; Inglis and Waycott, 2001).
Seed banks are often classified as transient
(turnover in less than one year) or persistent (remaining viable longer than one year) according to how