168
M. A. Mateo, J. Cebri ´
an, K. Dunton, and T. Mutchler
and R = 0.37, p = 0.05 for phosphorus content;
Fig. 4A and B, respectively) are possibly suggestive, but they do not allow for any robust conclusion. Clearly, more work is needed to understand
the existing controversial observations of the association between seagrass leaf nutrient content and
the intensity of herbivory. Given the high spatial and
temporal variability that herbivory may show in nature (see above), it seems possible that the different
spatial and temporal scales covered by these studies
confounds any firm conclusion.
An important point to be made is that, of the
carbon consumed by herbivores, only a very small
fraction will be assimilated and effectively incorporated in the secondary production compartment.
For example, it has been shown that the carbon assimilation efficiency of the main grazer of Neptune
grass in the Mediterranean, the fish Sarpa salpa,
is as low as 0.2% (Velimirov, 1984). In fact, it is
not uncommon to observe entire, green fragments of
Neptune grass leaves (ca. 0.8 cm
2 ) being defecated
to the bed sediments from schools of this fish (personal observation). For P. australis on the East coast
of Australia, the value was 3% for the main grazing fish, Monacanthus chinensis (Conacher et al.,
1979). In tropical seagrass stands, grazing by macroinvertebrates can be substantial. Gammarids of the
genus Ampithoe in Fiji grazed half the leaf carbon
production of Syringodium isoetifolium, but assimilated only 10% of it (Mukai and Iijima, 1995). More
than half of the carbon grazed was respired and the
rest excreted and defecated. Thus, most grazing rates
given in the literature may be large overestimates of
the actual carbon flux from producers to consumers.
A detailed knowledge of assimilation rates for the
various herbivores should be acquired and used to
correct the fluxes accordingly.
B. Export
The export of materials from seagrass beds has
many important implications for surrounding communities and ecosystems (Romero et al., Chapter
9; Kenworthy et al., Chapter 25; Bell et al., Chapter 26). Since most exported detritus is decomposed
in downstream systems (Mann, 1988), the quantity
of detritus exported sets the limits to the levels of
secondary production that the bed can support beyond its boundaries (i.e. allochthonous secondary
production; see Chapters 25 and 26). Since export
represents a nutrient loss for the bed, these losses
must be compensated by exogenous nutrient inputs
(Duarte and Cebri´ an, 1996; Mateo and Romero,
1997; Romero et al., Chapter 9).
Despite the importance of export in seagrassdominated coastal ecosystems, few reports exist on
its impact on bed economy. This oversight may be
due to three inherent technical difficulties.
First, and most importantly, seagrass beds are
often open systems and have widespread exchange with offshore waters, driven by the interaction of several forces, including wind, tides,
and waves. This renders measurement of detrital export difficult. Most measurements of detrital export are limited to specialized systems
connected to open waters through narrow outlets (e.g. coastal lagoons).
Second, the boundaries of seagrass beds, which
define the location at which export measurements are taken, are sometimes difficult to
define with certainty, making measurements
somewhat arbitrary.
Third, detrital traps used to derive direct estimates
of export are difficult to deploy.
These methodological limitations have discouraged researchers from quantifying detrital export
from seagrass beds and have resulted in a scarcity of
publications on the issue. These problems emphasize
the importance of developing alternative methods.
Romero et al. (1992) proposed an indirect method:
in the hypothetical absence of export, litter stocks
in the bed depend on inputs from leaf fall (the main
source of variation being depth) and outputs due to
remineralization. Leaf fall rates can be estimated as
the difference between primary production (using
the method described in Zieman, 1974) and biomass
increase. Decay rates can be approximated using the
classical in situ litter bag incubations. This can be
expressed mathematically in order to predict litter
accumulation in a given moment and a given place
in the bed.
L
i = F i e
(−kt/2)
+ L i−t e
(−kt)
(3)
where L
i is the predicted standing litter at time i, F i is
the weight of leaf fallen between times i and i − 1, t
is the time interval between consecutive samplings, k
is the decay rate for this period and area (e.g. depth),
and L i−t is the standing litter observed in situ at
time i − t (i.e. before the initiation of the period).
Knowing the standing litter stock at the end of the
M. A. Mateo, J. Cebri ´
an, K. Dunton, and T. Mutchler
and R = 0.37, p = 0.05 for phosphorus content;
Fig. 4A and B, respectively) are possibly suggestive, but they do not allow for any robust conclusion. Clearly, more work is needed to understand
the existing controversial observations of the association between seagrass leaf nutrient content and
the intensity of herbivory. Given the high spatial and
temporal variability that herbivory may show in nature (see above), it seems possible that the different
spatial and temporal scales covered by these studies
confounds any firm conclusion.
An important point to be made is that, of the
carbon consumed by herbivores, only a very small
fraction will be assimilated and effectively incorporated in the secondary production compartment.
For example, it has been shown that the carbon assimilation efficiency of the main grazer of Neptune
grass in the Mediterranean, the fish Sarpa salpa,
is as low as 0.2% (Velimirov, 1984). In fact, it is
not uncommon to observe entire, green fragments of
Neptune grass leaves (ca. 0.8 cm
2 ) being defecated
to the bed sediments from schools of this fish (personal observation). For P. australis on the East coast
of Australia, the value was 3% for the main grazing fish, Monacanthus chinensis (Conacher et al.,
1979). In tropical seagrass stands, grazing by macroinvertebrates can be substantial. Gammarids of the
genus Ampithoe in Fiji grazed half the leaf carbon
production of Syringodium isoetifolium, but assimilated only 10% of it (Mukai and Iijima, 1995). More
than half of the carbon grazed was respired and the
rest excreted and defecated. Thus, most grazing rates
given in the literature may be large overestimates of
the actual carbon flux from producers to consumers.
A detailed knowledge of assimilation rates for the
various herbivores should be acquired and used to
correct the fluxes accordingly.
B. Export
The export of materials from seagrass beds has
many important implications for surrounding communities and ecosystems (Romero et al., Chapter
9; Kenworthy et al., Chapter 25; Bell et al., Chapter 26). Since most exported detritus is decomposed
in downstream systems (Mann, 1988), the quantity
of detritus exported sets the limits to the levels of
secondary production that the bed can support beyond its boundaries (i.e. allochthonous secondary
production; see Chapters 25 and 26). Since export
represents a nutrient loss for the bed, these losses
must be compensated by exogenous nutrient inputs
(Duarte and Cebri´ an, 1996; Mateo and Romero,
1997; Romero et al., Chapter 9).
Despite the importance of export in seagrassdominated coastal ecosystems, few reports exist on
its impact on bed economy. This oversight may be
due to three inherent technical difficulties.
First, and most importantly, seagrass beds are
often open systems and have widespread exchange with offshore waters, driven by the interaction of several forces, including wind, tides,
and waves. This renders measurement of detrital export difficult. Most measurements of detrital export are limited to specialized systems
connected to open waters through narrow outlets (e.g. coastal lagoons).
Second, the boundaries of seagrass beds, which
define the location at which export measurements are taken, are sometimes difficult to
define with certainty, making measurements
somewhat arbitrary.
Third, detrital traps used to derive direct estimates
of export are difficult to deploy.
These methodological limitations have discouraged researchers from quantifying detrital export
from seagrass beds and have resulted in a scarcity of
publications on the issue. These problems emphasize
the importance of developing alternative methods.
Romero et al. (1992) proposed an indirect method:
in the hypothetical absence of export, litter stocks
in the bed depend on inputs from leaf fall (the main
source of variation being depth) and outputs due to
remineralization. Leaf fall rates can be estimated as
the difference between primary production (using
the method described in Zieman, 1974) and biomass
increase. Decay rates can be approximated using the
classical in situ litter bag incubations. This can be
expressed mathematically in order to predict litter
accumulation in a given moment and a given place
in the bed.
L
i = F i e
(−kt/2)
+ L i−t e
(−kt)
(3)
where L
i is the predicted standing litter at time i, F i is
the weight of leaf fallen between times i and i − 1, t
is the time interval between consecutive samplings, k
is the decay rate for this period and area (e.g. depth),
and L i−t is the standing litter observed in situ at
time i − t (i.e. before the initiation of the period).
Knowing the standing litter stock at the end of the