Chapter 7 Carbon Flux in Seagrasses
165
production and import and the loss processes of
decomposition, export, and refractory accumulation.
The following sections contain a selective discussion
of the variability, controls, and consequences of the
routes listed in Eqs. (1) and (2). The two main objectives of this discussion are to (i) summarize some
important patterns in our current knowledge and (ii)
identify gaps that merit further research.
A. Herbivory
1. The Extent of Herbivory in Seagrass Beds
Herbivory may vary widely both within and among
seagrass species. Some species, such as the mediterranean Neptune grass (Posidonia oceanica), tend
to have modest levels of herbivory (i.e. <10% of
aboveground production consumed by herbivores;
Pergent et al., 1994; Cebri´ an et al., 1996a), although occasionally this species has been shown to
support substantial herbivory (Kirkman and Young,
1981; Shepherd, 1987). Other species, such as turtle grass (T. testudinum), often support larger levels
of consumption, although this may have been more
important in pre-columban (ca. 1500 AD) times
when abundances of turtles, manatees, dugongs, and
other large grazers were high, before the current
centuries of heavy human exploitation. Indeed, intense grazing on turtle grass does not seem uncommon (Valentine and Heck, 1991; Heck and Valentine, 1995; Valentine et al., 2000; Valentine and
Duffy, Chapter 20). However, even for species that
often show large losses to herbivores, the spatial
and temporal variability in herbivory within the
species can be substantial. For instance, Cebri´ an
and Duarte (1998) compared four turtle grass populations in the Caribbean and found that the percentage of primary production consumed by herbivores ranged between <1 and 30% among the
populations. Many other examples of large withinspecies variability in herbivory exist (Cebri´ an et al.,
1996b; Valentine and Heck, 1999; Valentine et al.,
2000).
So what generalities can be made about the
magnitude of herbivory for seagrasses? Figure 3A
and B includes an extensive compilation of published values of herbivory made for this chapter
(data set and further explanations are available at
‘http://ecosystemslab.disl.org’ under ‘data sets’). It
can be seen that the percentage of above-ground production removed by herbivores is skewed to the right,
with most populations losing <10% of the aboveground production to herbivores. While herbivory is
an inherently variable process, the number of studies
reporting modest levels of herbivory largely exceeds
those reporting intense herbivory (but see Valentine
and Duffy, Chapter 20).
An important difference emerges depending on
whether herbivory is regarded as absolute consumption or as the percentage of seagrass production consumed. While the percentage formulation tends to be
modest, because of the high levels of above-ground
production that seagrasses often reach, the values of
absolute consumption tend to be higher and similar
to those observed for many other aquatic and terrestrial producers (Cebri´ an and Duarte, 1994; Cebri´ an
et al., 1998; Cebri´ an, 1999, 2002).
Thus, in general, seagrasses transfer a significant
quantity of biomass to herbivores despite appearing to have only a modest percentage of production
removed. This dichotomy suggests two important
corollaries. First, it appears that herbivores generally
have a limited role in constraining seagrass biomass
since they often remove <10% of the plant production. Second, and despite the seemingly modest role
of herbivores, seagrasses seem to have the capacity
to fuel significant levels of herbivore production in
comparison to the levels supported by other aquatic
and terrestrial producers. The large variability in herbivory found within and among seagrass species indicates that numerous mechanisms can influence this
process. One such mechanism is herbivore abundance. The influence of herbivory abundance has
been particularly well demonstrated for sea urchins.
Sea urchin densities are often strongly regulated by
the intensity of predation on young recruits (Sala,
1997; Sala et al., 1998). Under relaxed predation, sea
urchins may become abundant and inflict substantial damage in nearby seagrass beds (Keller, 1983;
Larkum and West, 1990; Klumpp et al., 1993; Greenway, 1995; Rose et al., 1999). The increase in sea
urchin density may be enormous, leading to population outbreaks and the subsequent decimation of
large seagrass areas (Camp et al., 1973; Larkum and
West, 1990; Macia and Lirman, 1999; Rose et al.,
1999).
Another well-known source of variability is the
‘cultivation’ feeding pattern that some vertebrate
herbivores, such as green turtles (Bjorndal, 1980;
Zieman et al., 1984) and dugongs (De Iongh et al.,
1995; Preen, 1995) exhibit. These herbivores feed
165
production and import and the loss processes of
decomposition, export, and refractory accumulation.
The following sections contain a selective discussion
of the variability, controls, and consequences of the
routes listed in Eqs. (1) and (2). The two main objectives of this discussion are to (i) summarize some
important patterns in our current knowledge and (ii)
identify gaps that merit further research.
A. Herbivory
1. The Extent of Herbivory in Seagrass Beds
Herbivory may vary widely both within and among
seagrass species. Some species, such as the mediterranean Neptune grass (Posidonia oceanica), tend
to have modest levels of herbivory (i.e. <10% of
aboveground production consumed by herbivores;
Pergent et al., 1994; Cebri´ an et al., 1996a), although occasionally this species has been shown to
support substantial herbivory (Kirkman and Young,
1981; Shepherd, 1987). Other species, such as turtle grass (T. testudinum), often support larger levels
of consumption, although this may have been more
important in pre-columban (ca. 1500 AD) times
when abundances of turtles, manatees, dugongs, and
other large grazers were high, before the current
centuries of heavy human exploitation. Indeed, intense grazing on turtle grass does not seem uncommon (Valentine and Heck, 1991; Heck and Valentine, 1995; Valentine et al., 2000; Valentine and
Duffy, Chapter 20). However, even for species that
often show large losses to herbivores, the spatial
and temporal variability in herbivory within the
species can be substantial. For instance, Cebri´ an
and Duarte (1998) compared four turtle grass populations in the Caribbean and found that the percentage of primary production consumed by herbivores ranged between <1 and 30% among the
populations. Many other examples of large withinspecies variability in herbivory exist (Cebri´ an et al.,
1996b; Valentine and Heck, 1999; Valentine et al.,
2000).
So what generalities can be made about the
magnitude of herbivory for seagrasses? Figure 3A
and B includes an extensive compilation of published values of herbivory made for this chapter
(data set and further explanations are available at
‘http://ecosystemslab.disl.org’ under ‘data sets’). It
can be seen that the percentage of above-ground production removed by herbivores is skewed to the right,
with most populations losing <10% of the aboveground production to herbivores. While herbivory is
an inherently variable process, the number of studies
reporting modest levels of herbivory largely exceeds
those reporting intense herbivory (but see Valentine
and Duffy, Chapter 20).
An important difference emerges depending on
whether herbivory is regarded as absolute consumption or as the percentage of seagrass production consumed. While the percentage formulation tends to be
modest, because of the high levels of above-ground
production that seagrasses often reach, the values of
absolute consumption tend to be higher and similar
to those observed for many other aquatic and terrestrial producers (Cebri´ an and Duarte, 1994; Cebri´ an
et al., 1998; Cebri´ an, 1999, 2002).
Thus, in general, seagrasses transfer a significant
quantity of biomass to herbivores despite appearing to have only a modest percentage of production
removed. This dichotomy suggests two important
corollaries. First, it appears that herbivores generally
have a limited role in constraining seagrass biomass
since they often remove <10% of the plant production. Second, and despite the seemingly modest role
of herbivores, seagrasses seem to have the capacity
to fuel significant levels of herbivore production in
comparison to the levels supported by other aquatic
and terrestrial producers. The large variability in herbivory found within and among seagrass species indicates that numerous mechanisms can influence this
process. One such mechanism is herbivore abundance. The influence of herbivory abundance has
been particularly well demonstrated for sea urchins.
Sea urchin densities are often strongly regulated by
the intensity of predation on young recruits (Sala,
1997; Sala et al., 1998). Under relaxed predation, sea
urchins may become abundant and inflict substantial damage in nearby seagrass beds (Keller, 1983;
Larkum and West, 1990; Klumpp et al., 1993; Greenway, 1995; Rose et al., 1999). The increase in sea
urchin density may be enormous, leading to population outbreaks and the subsequent decimation of
large seagrass areas (Camp et al., 1973; Larkum and
West, 1990; Macia and Lirman, 1999; Rose et al.,
1999).
Another well-known source of variability is the
‘cultivation’ feeding pattern that some vertebrate
herbivores, such as green turtles (Bjorndal, 1980;
Zieman et al., 1984) and dugongs (De Iongh et al.,
1995; Preen, 1995) exhibit. These herbivores feed