Chapter 20 Grazing in Seagrass Ecosystems
487
and Gammarus mucronatus (Duffy and Harvilicz
2001). The two amphipod taxa each essentially eliminated periphyton from eelgrass leaves and reached
nearly identical final population biomasses, supporting the general conclusion (Edgar 1993) that a common resource ceiling limited their population sizes
and production. Nevertheless, subtle differences in
their feeding preferences produced marked differences in the total biomass and composition of the
epiphyte community that developed under their influence (Duffy and Harvilicz 2001). In fact, total
epiphyte mass in Gammarus treatments actually exceeded the fouling mass in grazer-free controls due
to Gammarus facilitation of the filamentous epiphyte Polysiphonia sp., which grew luxuriously only
where Gammarus was present. A second, similar example comes from a seaweed-dominated system; in
this case macroalgal community composition, and
particularly biomass of the dominant seaweed Sargassum filipendula, differed dramatically between
treatments with versus without the herbivorous amphipod Ampithoe longimana, despite there being
no difference in total epifaunal abundance (nor,
presumably, production) among treatments (Duffy
and Hay 2000). Ampithoe longimana was the only
mesograzer that fed significantly on macroalgae
in that system. Finally, the frequent destruction of
seagrasses by ampithoids and idoteid isopods in
predator-free mesocosms (Kirkman 1978; Williams
and Ruckelshaus 1993; Short et al. 1995; Duffy et al.
2001, 2003), and similar defoliation of macroalgal
beds during occasional amphipod or isopod outbreaks in nature (Kangas et al. 1982; Haahtela 1984;
Tegner and Dayton 1987) suggests that some factor
other than food availability normally keeps populations of these species in check.
In summary, even if total secondary production
is limited by resource availability, as appears true
(Edgar 1993; Edgar and Klumpp et al. 2003), predation may remain critical in preventing the explosion of particular keystone or dominant species that
could severely damage macrophyte populations and
the communities that depend on them. That is, while
metabolic rate-based estimates are important indicators of energy flow within ecosystems, they may tell
us little about the community-level impacts of grazers (Paine 1980). Realistic experimental tests of the
importance of predation in influencing assemblage
structure of mesograzers, and their cascading impacts on primary producers in seagrass beds, remain
an important challenge for future research.
B. Eutrophication, Overfishing, and the
Decline of Estuarine Vegetation
The question of what regulates mesograzer population density and community structure is of far
more than academic interest. If mesograzers are as
important to facilitating seagrass dominance as existing evidence suggests they are, then any factors—
natural or anthropogenic—that affect mesograzer
populations may indirectly affect the health of
seagrasses. A case in point involves eutrophication, which is a serious threat to seagrass ecosystems worldwide (Short and Wylie-Echeverria 1996;
Duarte 2002; Kenworthy et al., Chapter 25). Increased nutrient loading to coastal ecosystems generally favors fast-growing algae over seagrasses,
with phytoplankton dominating in deeper basins
with low water turnover rate, and benthic algae dominating in shallower basins and where flushing is
stronger (Valiela et al. 1997). Thus, dominance by
seagrasses is potentially threatened by eutrophication. However, experiments in greenhouses (Neckles
et al. 1993), outdoor mesocosms (Lin et al. 1996),
and the field (Williams and Ruckelshaus 1993; Heck
et al. 2000) have demonstrated that nutrient loading
by itself often has little effect on the balance between
seagrasses and algae, as long as natural grazer populations are present to crop the algae. Models also predict that natural levels of grazing can often eliminate
detrimental effects of nutrient loading on seagrasses
(Wetzel and Neckles 1986). These considerations
suggest that negative impacts of eutrophication on
seagrass ecosystems are more complex than commonly envisaged. Specifically, they suggest that eutrophied systems, which are often impacted by multiple stressors, may degrade as much as a consequence
of altered food webs as of nutrient loading by itself
(Fig. 6).
A trophic cascade is usually understood to refer to the penetration of predator influence downward through one or more intermediate levels in a
food chain to influence plant biomass. Classic examples include the cascading effects of piscivorous
fish, through zooplanktivorous fishes and zooplankton, to phytoplankton biomass in northern temperate
lakes (Carpenter et al. 1985), and the penetration of
killer whale influence, through sea otters and their
sea urchin prey, to regulate the abundance and distribution of kelp beds in the northeastern Pacific
(Estes and Duggins 1995; Estes et al. 1998). As
mentioned above, many seagrass beds, particularly
487
and Gammarus mucronatus (Duffy and Harvilicz
2001). The two amphipod taxa each essentially eliminated periphyton from eelgrass leaves and reached
nearly identical final population biomasses, supporting the general conclusion (Edgar 1993) that a common resource ceiling limited their population sizes
and production. Nevertheless, subtle differences in
their feeding preferences produced marked differences in the total biomass and composition of the
epiphyte community that developed under their influence (Duffy and Harvilicz 2001). In fact, total
epiphyte mass in Gammarus treatments actually exceeded the fouling mass in grazer-free controls due
to Gammarus facilitation of the filamentous epiphyte Polysiphonia sp., which grew luxuriously only
where Gammarus was present. A second, similar example comes from a seaweed-dominated system; in
this case macroalgal community composition, and
particularly biomass of the dominant seaweed Sargassum filipendula, differed dramatically between
treatments with versus without the herbivorous amphipod Ampithoe longimana, despite there being
no difference in total epifaunal abundance (nor,
presumably, production) among treatments (Duffy
and Hay 2000). Ampithoe longimana was the only
mesograzer that fed significantly on macroalgae
in that system. Finally, the frequent destruction of
seagrasses by ampithoids and idoteid isopods in
predator-free mesocosms (Kirkman 1978; Williams
and Ruckelshaus 1993; Short et al. 1995; Duffy et al.
2001, 2003), and similar defoliation of macroalgal
beds during occasional amphipod or isopod outbreaks in nature (Kangas et al. 1982; Haahtela 1984;
Tegner and Dayton 1987) suggests that some factor
other than food availability normally keeps populations of these species in check.
In summary, even if total secondary production
is limited by resource availability, as appears true
(Edgar 1993; Edgar and Klumpp et al. 2003), predation may remain critical in preventing the explosion of particular keystone or dominant species that
could severely damage macrophyte populations and
the communities that depend on them. That is, while
metabolic rate-based estimates are important indicators of energy flow within ecosystems, they may tell
us little about the community-level impacts of grazers (Paine 1980). Realistic experimental tests of the
importance of predation in influencing assemblage
structure of mesograzers, and their cascading impacts on primary producers in seagrass beds, remain
an important challenge for future research.
B. Eutrophication, Overfishing, and the
Decline of Estuarine Vegetation
The question of what regulates mesograzer population density and community structure is of far
more than academic interest. If mesograzers are as
important to facilitating seagrass dominance as existing evidence suggests they are, then any factors—
natural or anthropogenic—that affect mesograzer
populations may indirectly affect the health of
seagrasses. A case in point involves eutrophication, which is a serious threat to seagrass ecosystems worldwide (Short and Wylie-Echeverria 1996;
Duarte 2002; Kenworthy et al., Chapter 25). Increased nutrient loading to coastal ecosystems generally favors fast-growing algae over seagrasses,
with phytoplankton dominating in deeper basins
with low water turnover rate, and benthic algae dominating in shallower basins and where flushing is
stronger (Valiela et al. 1997). Thus, dominance by
seagrasses is potentially threatened by eutrophication. However, experiments in greenhouses (Neckles
et al. 1993), outdoor mesocosms (Lin et al. 1996),
and the field (Williams and Ruckelshaus 1993; Heck
et al. 2000) have demonstrated that nutrient loading
by itself often has little effect on the balance between
seagrasses and algae, as long as natural grazer populations are present to crop the algae. Models also predict that natural levels of grazing can often eliminate
detrimental effects of nutrient loading on seagrasses
(Wetzel and Neckles 1986). These considerations
suggest that negative impacts of eutrophication on
seagrass ecosystems are more complex than commonly envisaged. Specifically, they suggest that eutrophied systems, which are often impacted by multiple stressors, may degrade as much as a consequence
of altered food webs as of nutrient loading by itself
(Fig. 6).
A trophic cascade is usually understood to refer to the penetration of predator influence downward through one or more intermediate levels in a
food chain to influence plant biomass. Classic examples include the cascading effects of piscivorous
fish, through zooplanktivorous fishes and zooplankton, to phytoplankton biomass in northern temperate
lakes (Carpenter et al. 1985), and the penetration of
killer whale influence, through sea otters and their
sea urchin prey, to regulate the abundance and distribution of kelp beds in the northeastern Pacific
(Estes and Duggins 1995; Estes et al. 1998). As
mentioned above, many seagrass beds, particularly