422
B. I. van Tussenbroek, J. A. Vonk, J. Stapel et al.
Possible explanations for the existence of this disjunctive population were that floating fruits from
existing populations reached the island (improbable,
as this would take ∼2 months), or ‘hitch-hiking’ of
fruits or seeds in ballast water of ships or transported
coastal sediments (Rollon et al., 2001).
VII. Population Dynamics
A. Demography
To date, demographic analyses of the genets (an individual originated from a seed, sensu Harper, 1977) of
Thalassia have not been attempted because following a genet in the field is virtually impossible. Tracing the shoots of an individual in the field is difficult
because their interconnecting horizontal rhizomes
occur belowground and are entangled. Also, the connections between the shoots break with time, either
through mechanical damage or decay of the oldest
end of the horizontal rhizomes, thus a genet consists
of many separate rhizomes sections (i.e. clones), and
it is impossible to tell one genet apart from another.
In contrast, the shoots arising from the horizontal
rhizomes, which can considered being the ramets
(sensu Harper, 1977), are easily identified, and techniques of age determination of the shoots (Patriquin,
1973; Duarte et al., 1994) make it possible to study
their demography.
The longevity of the shoots can be determined
by ‘translating’ their size (expressed as the number
of scars of abscised leaves) into age using information based on the number of leaves produced yearly.
For example, a shoot with 30 leaf scars plus standing
leaves, which produces ten leaves per year, has an age
of 3 years. There exist various techniques for the determination of annual leaf production, or its derivative, annual mean Plastochron Interval, which was
defined by Patriquin (1973) as the difference in age
between two successive leaves, measured when they
have reached the same stage of development. The
techniques applied to Thalassia spp. are: (1) Leaf
marking, which determines the rate of leaf initiation
per foliar shoot (Durako, 1994; van Tussenbroek,
1994a; Kaldy et al., 1999; Peterson and Fourqurean,
2001), (2) differences in internode’s length of the vertical rhizomes (Gallegos et al., 1993; Marb` a et al.,
1994), (3) number of leaves formed between successive flower scars (Gallegos et al., 1992; van Tussenbroek, 1994a), and (4) annual cohort analysis based
on size frequency histograms of the vertical rhizomes (Durako, 1994; Duarte et al., 1994). All of
these techniques, except for leaf marking, utilize
cyclical variations caused by seasonal differences
in vertical rhizome morphology to indicate the passage of a year, and the number of leaf scars between these seasonal imprints is considered to be
the number of leaves produced in that year. The important underlying assumption of all techniques of
age-determination is that the number of leaves produced annually does not differ significantly between
years, but only a few authors have tested this, and results vary from place to place. T. testudinum showed
large inter-annual variability in annual leaf formation rates in the Laguna Madre, Texas (Kaldy et al.,
1999), whereas this species showed similar annual
rates of leaf production over a period of 13 years
in a tropical Mexican reef lagoon (van Tussenbroek,
2002), and neither Durako (1994) nor Peterson and
Fourqurean (2001) found significant differences in
annual mean leaf emergence rates between two consecutive sampling years at a large number of sites in
Florida Bay.
The ideal study of shoot demography would be to
mark a large number of shoots and follow their development through time, allowing for production of
a dynamic life-table, but this requires excavation of
the rhizomes, which is virtually impossible without
considerable damage. Therefore, all demographic
studies have involved examination of the shoot population sampled at one particular time, resulting in
a static age-distribution. Applications of such static
age-distributions have been many-fold; and amongst
others they have been used to retrospectively assess
flowering events and the effects of a hurricane on
T. testudinum in the Mexican Caribbean (Gallegos
et al., 1992; Marb` a et al., 1994; van Tussenbroek,
1994a,b). The principal application of the static agedistributions has been for the derivation of population growth rate and related parameters (such as
mortality, recruitment, shoot turnover rates, and half
life), according to the model proposed by Duarte
et al. (1994), which has been used for populations of both T. testudinum (Gallegos et al., 1993;
Durako, 1994; Peterson and Fourqurean, 2001) and
T. hemprichii (Vermaat et al., 1995; Rollon et al.,
2001). There are questions, however, over whether
these static models can be applied to determine population growth (see Durako and Duarte, 1997; Jensen
et al., 1997; Kaldy et al., 1999; van Tussenbroek,
2000), or whether the prerequisites of constant
B. I. van Tussenbroek, J. A. Vonk, J. Stapel et al.
Possible explanations for the existence of this disjunctive population were that floating fruits from
existing populations reached the island (improbable,
as this would take ∼2 months), or ‘hitch-hiking’ of
fruits or seeds in ballast water of ships or transported
coastal sediments (Rollon et al., 2001).
VII. Population Dynamics
A. Demography
To date, demographic analyses of the genets (an individual originated from a seed, sensu Harper, 1977) of
Thalassia have not been attempted because following a genet in the field is virtually impossible. Tracing the shoots of an individual in the field is difficult
because their interconnecting horizontal rhizomes
occur belowground and are entangled. Also, the connections between the shoots break with time, either
through mechanical damage or decay of the oldest
end of the horizontal rhizomes, thus a genet consists
of many separate rhizomes sections (i.e. clones), and
it is impossible to tell one genet apart from another.
In contrast, the shoots arising from the horizontal
rhizomes, which can considered being the ramets
(sensu Harper, 1977), are easily identified, and techniques of age determination of the shoots (Patriquin,
1973; Duarte et al., 1994) make it possible to study
their demography.
The longevity of the shoots can be determined
by ‘translating’ their size (expressed as the number
of scars of abscised leaves) into age using information based on the number of leaves produced yearly.
For example, a shoot with 30 leaf scars plus standing
leaves, which produces ten leaves per year, has an age
of 3 years. There exist various techniques for the determination of annual leaf production, or its derivative, annual mean Plastochron Interval, which was
defined by Patriquin (1973) as the difference in age
between two successive leaves, measured when they
have reached the same stage of development. The
techniques applied to Thalassia spp. are: (1) Leaf
marking, which determines the rate of leaf initiation
per foliar shoot (Durako, 1994; van Tussenbroek,
1994a; Kaldy et al., 1999; Peterson and Fourqurean,
2001), (2) differences in internode’s length of the vertical rhizomes (Gallegos et al., 1993; Marb` a et al.,
1994), (3) number of leaves formed between successive flower scars (Gallegos et al., 1992; van Tussenbroek, 1994a), and (4) annual cohort analysis based
on size frequency histograms of the vertical rhizomes (Durako, 1994; Duarte et al., 1994). All of
these techniques, except for leaf marking, utilize
cyclical variations caused by seasonal differences
in vertical rhizome morphology to indicate the passage of a year, and the number of leaf scars between these seasonal imprints is considered to be
the number of leaves produced in that year. The important underlying assumption of all techniques of
age-determination is that the number of leaves produced annually does not differ significantly between
years, but only a few authors have tested this, and results vary from place to place. T. testudinum showed
large inter-annual variability in annual leaf formation rates in the Laguna Madre, Texas (Kaldy et al.,
1999), whereas this species showed similar annual
rates of leaf production over a period of 13 years
in a tropical Mexican reef lagoon (van Tussenbroek,
2002), and neither Durako (1994) nor Peterson and
Fourqurean (2001) found significant differences in
annual mean leaf emergence rates between two consecutive sampling years at a large number of sites in
Florida Bay.
The ideal study of shoot demography would be to
mark a large number of shoots and follow their development through time, allowing for production of
a dynamic life-table, but this requires excavation of
the rhizomes, which is virtually impossible without
considerable damage. Therefore, all demographic
studies have involved examination of the shoot population sampled at one particular time, resulting in
a static age-distribution. Applications of such static
age-distributions have been many-fold; and amongst
others they have been used to retrospectively assess
flowering events and the effects of a hurricane on
T. testudinum in the Mexican Caribbean (Gallegos
et al., 1992; Marb` a et al., 1994; van Tussenbroek,
1994a,b). The principal application of the static agedistributions has been for the derivation of population growth rate and related parameters (such as
mortality, recruitment, shoot turnover rates, and half
life), according to the model proposed by Duarte
et al. (1994), which has been used for populations of both T. testudinum (Gallegos et al., 1993;
Durako, 1994; Peterson and Fourqurean, 2001) and
T. hemprichii (Vermaat et al., 1995; Rollon et al.,
2001). There are questions, however, over whether
these static models can be applied to determine population growth (see Durako and Duarte, 1997; Jensen
et al., 1997; Kaldy et al., 1999; van Tussenbroek,
2000), or whether the prerequisites of constant