42
Michelle Waycott, Gabriele Procaccini, Donald H. Les and Thorsten B. H. Reusch
incidence of dioecy may also reflect the sexual conditions of monocotyledonous plants ancestral to seagrasses (Waycott and Les, 1996; Les et al., 1997).
Long-term demographic censuses are often unavailable for many seagrass populations. Unravelling the spatial pattern of ramets and genets in a
seagrass meadow uncovers parts of its local demographic history. Hence, as a potential alternative to census data, important demographic information may be inferred from a genetic analysis of
the spatial arrangement of clones. For example, the
detection of identical multi-locus genotypes among
distant vegetation patches may indicate the fragmentation of previously continuous meadow. Moreover,
minimal estimates of meadow age can be inferred
from rates of lateral spread of clones in conjunction
with the aerial extension of clones (Reusch et al.,
1999d). In several northern Baltic Sea populations
of Zostera marina, vegetation patches are composed
of shoots with one identical genotype (TBH Reusch
and C. Bostr¨ om, unpublished data), suggesting that
the initiation of patches often starts with a single
seedling.
Within-population genetic structure not only comprises the ramet–genet dichotomy but also involves
the dispersal capability of pollination and sexual
progeny. Given that seeds and pollen are thought
to typically travel only a few meters (Ruckelshaus,
1996) (but cf. Harwell and Orth, 2002), germinating
seedlings will often be related genetically to nearby
plants. Thus, additional population structure can be
expected at the scale of meters. Spatial autocorrelation techniques are a powerful tool to detect such
structure (Heywood, 1991). In autocorrelation analyses, the genetic similarity among pairs of ramets
is analysed as a function of their pairwise distance.
If autocorrelation values are positive, genetic relatedness is higher as expected under random mating,
and vice versa. As clonal plants, seagrasses require
special attention because the inclusion of members
of the same clone into an autocorrelation analysis inflates small-scale kinship structures (Reusch et al.,
1999b). In a recent study using a high-resolution spatial sampling grid (H¨ ammerli and Reusch, 2003a),
a significant positive spatial autocorrelation in kinship coefficients was found at distances between 2
and 5 m even when counting each clone only once.
Significant positive spatial autocorrelation was also
found between 1 and 11 m in Cymodocea nodosa and
between 2 and 7 m in Zostera noltii (Ruggiero and
Procaccini, unpublished). Thus, for any focal shoot,
the landscape of neighbouring plants is complex and
consists of several levels of genetic sub-structuring.
Clone affiliation, clone spatial arrangement and local kinship patterns interact to influence mating patterns, clonal competition and reproductive output in
apparently homogeneous seagrass meadows.
C. Mating Systems
Plant mating systems are one of the most significant factors in the transmission of genetic diversity
between generations (Clegg, 1980). The measurement of plant mating systems is complicated since an
overwhelming majority of angiosperm species possess both male and female flowers on the same plant
(Richards, 1997). These bisexual plants are therefore capable of self-pollination either from the same
flower (autogamy) or separate flowers on the same
plant (geitonogamy). Most seagrass species are dioecious (individuals of separate sex) (Waycott and Les,
1996; Les et al., 1997) and as such cannot self pollinate. However, even dioecious plants can be subject to inbreeding through the mating of siblings or
parent-offspring. In this context, two components of
seagrass mating systems are particularly significant,
first the movement of pollen between flowers within a
meadow and second the genetic structuring of plants
within populations as described in the proceeding
section.
Sexual reproduction in the marine environment
is limited by the ability of pollen to reach stigmas
(Pettitt et al., 1981; Cox, 1988; Ackerman, 1995;
Verduin, 1996; Reusch, 2003; Ackerman, Chapter 4). Seagrass pollination occurs by hydrophily,
that is, water mediated, abiotic pollination (Fægri
and van der Pijl, 1979). Our understanding of pollination in seagrasses is limited by the interaction of
the timing of pollen release/capture and the hydrodynamic environment in which these events occur. A
correlation between hydrodynamic environment and
the efficiency of pollen movement between flowers,
and thus outcrossing, has been proposed by Waycott
and Sampson (1997) for Posidonia australis. However, the spatial arrangement of flowers in a seagrass
meadow is critical to the outcome of any mating
event as the greater the area a flowering genet covers
the greater the chance there will be self-pollination
as found by Reusch (2001a) where outcrossing
rate was positively correlated with clonal diversity
(Fig. 8). For example in populations of Northern European Zostera marina, pollen limitation has been
Michelle Waycott, Gabriele Procaccini, Donald H. Les and Thorsten B. H. Reusch
incidence of dioecy may also reflect the sexual conditions of monocotyledonous plants ancestral to seagrasses (Waycott and Les, 1996; Les et al., 1997).
Long-term demographic censuses are often unavailable for many seagrass populations. Unravelling the spatial pattern of ramets and genets in a
seagrass meadow uncovers parts of its local demographic history. Hence, as a potential alternative to census data, important demographic information may be inferred from a genetic analysis of
the spatial arrangement of clones. For example, the
detection of identical multi-locus genotypes among
distant vegetation patches may indicate the fragmentation of previously continuous meadow. Moreover,
minimal estimates of meadow age can be inferred
from rates of lateral spread of clones in conjunction
with the aerial extension of clones (Reusch et al.,
1999d). In several northern Baltic Sea populations
of Zostera marina, vegetation patches are composed
of shoots with one identical genotype (TBH Reusch
and C. Bostr¨ om, unpublished data), suggesting that
the initiation of patches often starts with a single
seedling.
Within-population genetic structure not only comprises the ramet–genet dichotomy but also involves
the dispersal capability of pollination and sexual
progeny. Given that seeds and pollen are thought
to typically travel only a few meters (Ruckelshaus,
1996) (but cf. Harwell and Orth, 2002), germinating
seedlings will often be related genetically to nearby
plants. Thus, additional population structure can be
expected at the scale of meters. Spatial autocorrelation techniques are a powerful tool to detect such
structure (Heywood, 1991). In autocorrelation analyses, the genetic similarity among pairs of ramets
is analysed as a function of their pairwise distance.
If autocorrelation values are positive, genetic relatedness is higher as expected under random mating,
and vice versa. As clonal plants, seagrasses require
special attention because the inclusion of members
of the same clone into an autocorrelation analysis inflates small-scale kinship structures (Reusch et al.,
1999b). In a recent study using a high-resolution spatial sampling grid (H¨ ammerli and Reusch, 2003a),
a significant positive spatial autocorrelation in kinship coefficients was found at distances between 2
and 5 m even when counting each clone only once.
Significant positive spatial autocorrelation was also
found between 1 and 11 m in Cymodocea nodosa and
between 2 and 7 m in Zostera noltii (Ruggiero and
Procaccini, unpublished). Thus, for any focal shoot,
the landscape of neighbouring plants is complex and
consists of several levels of genetic sub-structuring.
Clone affiliation, clone spatial arrangement and local kinship patterns interact to influence mating patterns, clonal competition and reproductive output in
apparently homogeneous seagrass meadows.
C. Mating Systems
Plant mating systems are one of the most significant factors in the transmission of genetic diversity
between generations (Clegg, 1980). The measurement of plant mating systems is complicated since an
overwhelming majority of angiosperm species possess both male and female flowers on the same plant
(Richards, 1997). These bisexual plants are therefore capable of self-pollination either from the same
flower (autogamy) or separate flowers on the same
plant (geitonogamy). Most seagrass species are dioecious (individuals of separate sex) (Waycott and Les,
1996; Les et al., 1997) and as such cannot self pollinate. However, even dioecious plants can be subject to inbreeding through the mating of siblings or
parent-offspring. In this context, two components of
seagrass mating systems are particularly significant,
first the movement of pollen between flowers within a
meadow and second the genetic structuring of plants
within populations as described in the proceeding
section.
Sexual reproduction in the marine environment
is limited by the ability of pollen to reach stigmas
(Pettitt et al., 1981; Cox, 1988; Ackerman, 1995;
Verduin, 1996; Reusch, 2003; Ackerman, Chapter 4). Seagrass pollination occurs by hydrophily,
that is, water mediated, abiotic pollination (Fægri
and van der Pijl, 1979). Our understanding of pollination in seagrasses is limited by the interaction of
the timing of pollen release/capture and the hydrodynamic environment in which these events occur. A
correlation between hydrodynamic environment and
the efficiency of pollen movement between flowers,
and thus outcrossing, has been proposed by Waycott
and Sampson (1997) for Posidonia australis. However, the spatial arrangement of flowers in a seagrass
meadow is critical to the outcome of any mating
event as the greater the area a flowering genet covers
the greater the chance there will be self-pollination
as found by Reusch (2001a) where outcrossing
rate was positively correlated with clonal diversity
(Fig. 8). For example in populations of Northern European Zostera marina, pollen limitation has been