(PDF) Role of Larger Herbivores in Seagrass Communities

Estuaries Vol. 7, No. 4A,

p.

351376 December 1964

Role of Larger Herbivores

Seagrass Communities1

in

GORDON W. THAYER

Beaufort Laboratory

Southeast Fisheries Center

National Marine Fisheries Service, NOAA

Beaufort, North Carolina 28516

KAREN A. BJORNDAL

Department of Zoology

University of Florida

Gainesville, Florida 3261 I

JOHN C. OGDEN

West Indies Laboratory

Teague Bay

Christiansted, St. Croix

U.S. Virgin Islands 00820

SUSAN L. WILLIAMS

Marine Sciences Research Center

State University of New York

Stony Brook, New York 11794

JOSEPH C. ZIEMAN

Department of Environmental Science

University of Virginia

Charlottesville, Virginia 22093

ABSTRACT: The nutritional ecology

of macroherbivores in seagrass meadows and the roles of

grazing by urchins, fisbes and green turtles in tropical systems and waterfowl in temperate systems

are discussed in this review. Only a few species of animals graze on living seagrasses, and apparently

only a small portion of the energy and nutrients in seagrasses is usually channeled through these

herbivores. The general paucity of direct seagrass grazers may be a function of several factors in

the composition of seagrasses, including availability of nitrogen compounds, presence of relatively

high amounts of structural cell walls, and presence of toxic or inhibitory substances. The macroherb-

ivores, however, can have a profound effect on the seagrass plants, on other grazers and fauna

associated with the meadow, and on chemical and decompositional processes occurring within the

meadow. Grazing can alter the nutrient content and digestibility of the plant, as well as its pro-

ductivity. Removal of leaf material can influence interrelations among permanent and transient

faunal residents. Grazing also interrupts the detritus cycle. Possible consequences of this disruption,

either through acceleration or through decreased source input, and the enhancement of intersystem

coupling by increased export and offsite fecal production, are discussed. The extent and magnitude

of these effects and their ecological significance in the overall functioning of seagrass meadows only

can be speculated, and probably are not uniform or of similar importance in both tropical and

temperate seagrass systems. However, areas grazed by large herbivores provide natural experiments

in which to test hypotheses on

many functional relations in seagrass meadows.

’

Contribution No. 108, West Indies Laboratory,

Marine Sciences Research Center, State University of

Fairleigh Dickinson University; Contribution No. 392,

New York.

0 1964

Estuarine Research Federation

351

0160-6347/64/04A0351-376$01.50/O

352

G.

W. Thayer et al.

Introduction

Worldwide, seagrasses constitute one of

the most conspicuous and common coastal

ecosystem types. Submerged seagrass

meadows frequently contribute large por-

tions of the primary productivity of coastal

ecosystems. The production of these sea-

grass meadows provides food, shelter, and

nursery areas for many marine animals (see

reviews by Kikuchi 1980; McRoy and Helf-

ferich 1980; Ogden 1980; Zieman 1982;

Thayer et al., in press), including all but the

most pelagic of U.S. commercial fishery

species, 10 endangered species, and some

organisms inhabiting the ocean’s abyss

(Wolff 1980).

Organic matter derived from seagrass

production is transferred to secondary con-

sumers through three pathways: herbivores

that consume living plant matter; detriti-

vores that exploit dead material as partic-

ulate organic matter; and microorganisms

that take up seagrass-derived organic mat-

ter. Direct grazing on seagrasses has been

considered relatively unimportant, and de-

tritivores have emerged as the major benefi-

ciaries of energy fixed by seagrasses (Kik-

uchi et al. 1973). The basis of this conclusion

is primarily the low numbers of large sea-

grass herbivores; sea urchins, fish, water-

fowl, green turtles and sirenians are the ma-

jor seagrass macroherbivores. The general

paucity of species that are direct grazers may

be a function of several factors in the com-

position of seagrasses, including availability

of nitrogen compounds, presence of rela-

tively high amounts of structural cell wall

compounds, and the presence of toxic or

inhibitory chemicals. However, direct uti-

lization may be substantial in certain lo-

cations. In tropical regions, particularly the

Caribbean, herbivores comprise a greater

proportion of the fauna that depend on sea-

grasses for food than in temperate regions

(Ogden 1976; Ogden and Lobe1 1978;

McRoy and Helfferich 8980; McRoy and

Lloyd 1981; Zieman 1982).

A substantial information base exists on

herbivore-plant interactions in terrestrial

and algal-based marine systems (e.g., Matt-

son and Addy 19 7 5; Ogden and Lobe1 19 7 8;

McNaughton 1979; Hay 1981; Gaines and

Lubchenco

1982;

Vadas et al.

1982; Hay et

al. 1983). Herbivores have been shown to

alter plant productivity, distribution, com-

munity structure, and nutrient relations and

tissue nutrient content. Theories based on

algal systems may not apply to seagrasses,

however, since seagrass communities may

respond differently to grazing pressure.

Many of the larger seagrass herbivores (e.g.,

green turtles, sirenians, and many waterfowl

species) have been reduced in number due

to heavy exploitation and to reduction of

available foraging areas. However, even at

present day population levels, grazing by

these and other macroherbivores can have

a profound effect on seagrass communities,

although at a more reduced scale than prob-

ably occurred historically.

This paper is subdivided into four mini-

reviews that address the role of herbivores

in seagrass ecosystems: nutritional ecology

of large herbivores (KAB) and conse-

quences of grazing by fishes and urchins

(JCZ), green turtles (JCO and SLW), and

temperate waterfowl (GWT). Portions of

original reviews provided were extracted and

incorporated into either or both Seagrasses

As A Food Source and Concluding Com-

ments.

SEAGRASSES ASAFOOD SOURCE

In coastal marine ecology, one elusive

question continues to be why there are such

relatively few grazers on the abundant and

extensive seagrass meadows, especially on

a global basis. These meadows would seem

to be a vastly underutilized resource. A

common speculation was that the blades

represented an undesirable or at least a low

yield food source; however, recent literature

on the composition of seagrasses does not

support this speculation (see Zieman 1982;

Thayer et al., in press). .Although a critical

comparison of values among different stud-

ies frequently is not possible due to the wide

variety of non-comparable techniques used,

some useful generalizations have emerged.

The value of seagrasses as a food source is

a function of their availability (i.e., distri-

bution, abundance, morphology, and pro-

duction), as well as their chemical compo-

sition, but we will discuss only the latter.

Seagrasses have relatively high ash (min-

eral) content that varies with plant part, age

of the plant, and season (Zieman 1982;

Thayer et al., in press). Denton et al. (1980)

found very high levels of some minerals in

the tissues of the dugong, Dugong

dugon.

Some of these high concentrations were cor-

related with concentrations in their seagrass

diet. Dietary imbalances of minerals may

be significant for seagrass herbivores, but

neither the influence of mineral content or

concentration of individual components on

seagrass consumers and their feeding ecol-

ogy is well understood. However, several

species of macroherbivores recrop areas,

apparently feeding preferentially on young

leaves that, among other things, are lower

in ash content than older leaves.

Fiber (cell wall constituents) makes up a

fairly high percentage of the organic matter

in seagrasses (Bjomdal 1980; Vicente et al.

1980). Lignin increases with the age of

Tha-

lassia testudinum

blades, with a corre-

sponding drop in digestible cell contents

(Bjomdall980) that may be similar in other

seagrasses. This increase is important to

those herbivores that digest cell walls, be-

cause lignin has repeatedly been identified

as the major chemical component control-

ling the digestibility of cell walls (Van Soest

1982). The majority of seagrass consumers

apparently lack enzymes to digest structural

carbohydrates (Lawrence 1975) and, with

the exception of turtles and possibly man-

atees, do not have a gut flora capable of such

digestion. Thus, most macroconsumers of

seagrasses depend on the cell content of sea-

grasses and the attached epiphytes for food,

and must have a mechanism for the efficient

maceration of the material.

The nutritional value of seagrass leaves,

when expressed as nitrogen or calories on

an ash-free dry weight basis, falls within the

range of terrestrial vegetation and macro-

algae (e.g., Birch 1975; Bjomdal 1980;

Dawes and Lawrence 1980; Hocking et al.

198 1; Augier et al. 1982) and varies with

plant part, age of the plant, and season. The

nitrogen content of seagrasses, expressed as

protein, varies from 3 to 30%, but the ma-

jority of values are in the range of 10 to 16%

protein (see Zieman 1982; Thayer et al., in

press). By comparison, Vicente et al. (1980)

found that 10 species of common tropical

forage grasses contained 5.1 to 9.1% pro-

tein. Other studies have shown that, when

Role of Seagrass Macroherbivores

353

using similar techniques, the protein con-

tent of seagrasses is greater than that of algae

from the same locations. Off the central west

coast of Florida, Dawes et al. (1979) re-

ported a range for

Thalassia

from 3 to 12%

protein while the range for 14 species of co-

occurring macroalgae varied from 2 to 5

percent. Similarly, Lowe and Lawrence

(1976) found that in the Tampa Bay area

Thalassia

averaged 9% protein while 5

species of algae fed upon by sea urchins were

1.6 to 4.3% protein.

The nutrient content of a seagrass leaf is

a function of the age of the leaf and the

degree of epiphytism. One is tempted to call

this the nutritional value, but that would be

misleading as seagrass herbivores have de-

veloped a variety of grazing strategies to

obtain maximum nutritional value from

seagrasses. The protein (or nitrogen) content

is highest in fresh, green leaves and green

leaf bases. It declines with age, with senes-

cent leaf tips commonly having only half

the nitrogen content (Fig. 1) of the leaf bases

(Zieman et al. 1984). As these leaves grow,

the outer portions become coated with epi-

phytes, and the C/N ratio rises (Fig. 1) from

initial values of about 16 (range 13.8-l 9.8)

for the leaf bases, to 23 (21.1-25.2) for the

epiphytized tips (including the epiphytes),

to 34 (27.6-41.5) for senescent unepiphy-

tized tips (Zieman et al. 1984). A similar

trend occurs in

Zostera

leaves. In Nova Sco-

tia, Harrison and Mann (1975) found nitro-

gen content (dry weight percent) to be 2.9

in green leaves, 2.0 in senescent leaves and

1.3 in dead leaves, while in Beaufort, North

Carolina, Thayer et al. (1977) found epi-

phytized green leaves containing 1.9% ni-

trogen to drop to 1.3% as the leaves reached

senescence.

Phenols, sulphated phenols and sulphat-

ed flavones are present in many seagrass

species (Zapata and McMillan 1979;

McMillan et al. 1980). These secondary

substances may be toxic, unpalatable or may

bind proteins and carbohydrates making

them unavailable to both herbivores and

detritivores. Phenols and flavonoids inhibit

herbivory (Harbome 1979; Swain 1979),

and water soluble extracts (believed to con-

tain phenolic acids) from blades of

Zostera

marina

inhibited amphipod grazers, epi-

phytic algae and microorganisms (Harrison

C/N

I I

I

Epiphytes

r

354

G. W. Thayer et al.

Leaf

r

- -

%N

r

0

;:

3

0 10 20 30

Dark Green to Brown;

Senescent; Commonly

Heavily Epiphytized

Older

Decreased Epiphytism

Young, Green,

Rapidly Growing,

Few Epiphytes

%N

C/N

Epiphytes

Fig. 1. General relationships among nitrogen. carbon/nitrogen ratios, and degree of epiphytism along a

vertical plane for a seagrass blade.

1982).

The concentration of phenolic com-

pounds varies with age of a seagrass leaf,

being most concentrated in subapical re-

gions of the young leaf and decreasing sub-

stantially in older leaves (Harrison, in press).

The chemical forms of these secondary sub-

stances and their effect on seagrass herbi-

vores and detritivores need further study.

Nutrition of Seagrass Herbivores

The nutrition of seagrass herbivores-

their feeding behavior, rate of feeding and

digestive efficiency-determines their role

in the cycling of energy and nutrients in the

seagrass ecosystem. DiRerent feeding be-

haviors can alter the effect a herbivore has

on a seagrass bed, and seagrass herbivores

vary greatly in the amount of food they con-

sume and the degree to which they digest

seagrasses. This review will discuss feeding

behavior, feeding rates and digestive effi-

ciencies of the large herbivores that graze

on live seagrass blades: sea urchins, fish,

green turtles, sirenians, and birds.

Sea Urchins

Many sea urchins that feed in seagrass

beds consume mainly detritus, but some do

ingest live blades. Sea urchins chew their

food, thus making the easily digested cell

contents available to their digestive en-

zymes; feeding habits, gut pH and digestive

enzymes of herbivorous sea urchins have

been thoroughly reviewed by Lawrence

(1975, 1982). Although sea urchins have

large numbers of bacteria in their gut (2.6 x

10” per ml, Lasker and Giese 1954; 2 x

lo8 to 6 x lo9 per ml, Fong and Mann

1980), active cellulases generally are lacking

(Lawrence 1975; Prim and Lawrence 1975).

However, Fong and Mann (1980) did dem-

onstrate cellulose breakdown and end-prod-

uct utilization in

Strongylocentrotus droe-

bachiensis

on pure, carbon- 14 labelled,

powdered cellulose suspended in agar blocks.

Elyakova (1972) found low levels of cellu-

lase and chitinase in hepatopancreas tissue

from S.

intermedius

and

Echinarachnius

parma.

Passage rates measured in sea urchins

feeding

ad libitum

range from 8 to 12 h in

Diadema antillarum

(Lewis

1964), 14

to

24

h in

D. antillarum

(Hawkins 198 l), and 24

h in

Lytechinus variegatus

(Vadas et al.

1982). These rapid rates suggest that, at least

when sea urchins are eating seagrasses

ad

libitum,

gut bacteria do not have sufficient

time to attach to and digest cell walls to any

extent. For example, Lasker and Giese

(1954) found that gut bacteria from S.

pur-

Role of Seagrass Macroherbivores

355

puratus

required a week

in vitro

to break

down a small piece of the alga

Iridophycus

jlaccidum.

In another study (Prim and Law-

rence 1975), physical breakdown of plants

due to the activity of bacteria

in vitro

from

the gut of

L. variegatus

was first noted at 7

days for

Eucheuma nudum, 14

days for Ulva

lactuca,

and 44 days for both

Halodule

wrightii

and

Thalassia testudinum.

When

bacteria from

Arbacia punctulata

were

tested, signs of degradation of

E. nudum

were noted at 30 days, and of U.

lactuca

at

44

days; no breakdown was observed in

either

H. wrightii

or

T. testudinum

during

the 44-day experiment. The authors con-

cluded that bacteria contribute little to

digestion in sea urchins when food is abun-

dant. Lawrence (1976) found no evidence

of cell wall degredation in photomicro-

graphs of

T. testudinum

fragments in feces

from four species of sea urchins. However,

intestinal microflora can play an important

role in sea urchin nutrition by fixing nitro-

gen (Johnson and Mann 1982).

The digestive efficiencies (i.e., the net

amount of dietary constituent removed dur-

ing passage through the digestive tract, ex-

pressed as a percent of the amount ingested)

of sea urchins (Table 1) vary considerably

among species and among studies. The di-

gestibility of dry matter (DM) of

T. testu-

dinum

in

L. variegatus

varied from less than

19 to 65%. [The lower end of the range is

less than 19% because 19% is given for or-

ganic matter (OM) digestibility in Table 1.

DM digestibility is less than OM digest-

ibility because OM is more digestible than

ash and OM + ash = DM.]

The feeding rates of sea urchins (Table 2)

are difficult to compare because of the dif-

ferent units used. Where possible, the rates

were converted to g DM per urchin per day;

it was not possible, with the available data,

to convert to the more useful measures of

g DM or g OM per g live weight per day.

Again, considerable variation is seen among

species and studies.

Fish

Herbivorous fish apparently are depen-

dent upon disruption of plant cell walls,

either through chemical or mechanical

means, to gain nutrients from their diet. Fish

that rely on chemical lysis have a low stom-

ach pH that lyses some algae, generally those

with thin cell walls (Moriarty 1973; Lobe1

1980, 1981; Edwards and Horn 1982).

Montgomery and Gerking (1980), however,

have questioned the extent of acid hydroly-

sis of algae polysaccharides in fish stom-

achs. The cell walls of seagrasses appear

immune to disruption by acidic gastric se-

cretions (Bjomdal, in press). Therefore,

fish that utilize a significant portion of the

nutrients in green seagrass blades would

seem to be limited to those species that can

mechanically break cell walls, either with a

pharyngeal mill (Scaridae and Hemiram-

phidae, Randall 1967) or a muscular, tritu-

rating stomach (some Acanthuridae, Ogden

and Lobe1 1978).

Several studies on herbivorous fish have

found no cellulases or enzymes capable of

breaking down complex algal structural

polysaccharides (I&poor et al. 1975; Chiu

and Benitez 198 1). Cellulases have been re-

ported from some fish (Stickney and Shum-

way 1974; Lewis and Peters 198 1; Wein-

stein et al. 1982). Their presence does not

always correlate with diet (Stickney and

Shumway 1974) but, in the case of Atlantic

menhaden (Lewis and Peters 198 1) and pin-

fish (Weinstein et al. 1982) both species are

known to consume plant material. Also, the

rapid passage rate of food through the ali-

mentary canal of tropical herbivorous fish

would seem to preclude a significant nutri-

tional benefit to the fish from cellulolytic

enzymes. The Cortez damselfish

(Eupo-

macentrus rectzjraenum)

has passage rates

from 3.2 to 4.2 h, and food passes through

the gut of the giant blue damselfish

(Mi-

crospathodon dorsalis)

in 8.6 h (Montgom-

ery and Gerking 1980). Parrotfish fill and

empty their digestive tracts several times

each day (Ogden 1980).

Fishes that feed on seagrasses are iden-

tified in several reviews (Randall 196 5,1967;

Cat-r and Adams 1973; Adams 1976; McRoy

and Helfferich 1980). The bucktooth par-

rotfish,

Sparisoma radians,

feeds almost en-

tirely on seagrasses, particularly

Thalassia

testudinum

with epiphytes (Ogden 1980).

However, as can be seen in Table 1, only a

small part of the

Thalassia

blade (7% of the

organic matter) is utilized (Lobe1 and Ogden

198 1). Other seagrasses and

Thalassia

with

epiphytes are utilized to a much greater ex-

356

G.

W. Thayer et al.

Table 1. Digestive efficiencies (DE) of herbivores feeding on seagrass blades. Superscripts indicate the method

used: 1 = total collection trial, 2 = ash used as undigestible marker, 3 = lignin used as undigestible marker.

DM = dry matter, OM = organic matter, N = nitrogen, NDF = neutral detergent fiber (-cellulose, hemicellulose,

and lignin), ADF = acid detergent fiber (-cellulose and lignin), P = phosphorus. Ranges in values for green

turtles represent means from different size classes.

Diet

NUtIi.%t

DE (O/o)

Reference

Sea

Urchins

Diadema

antillarum

Thalassia

testudinum

Thalassia

testudinum

Thalassia

testudinum

Thalassia

testudinum

Syringodium

jiliforme

Phyllospadix

iwatensis

OM

Lipids

Proteins

Soluble

Carbohy-

drates

OM

Lipids

Proteins

Soluble

Carbohy-

drates

OM

Lipids

Proteins

Soluble

Carbohy-

drates

DM

OM

Lipids

Proteins

Carbohy-

drates

OM

Lipids

Proteins

Carbohy-

drates

DM

N

P. scouleri

OM

Thalassia

testudinum

DM

OM

Lipids

Proteins

Soluble

Carbohy-

drates

T. testudinum

with epiphytes

T. testudinum

w/o

epiphytes

Syringodium

jihiforme

Halodule wrightii

OM

T. testudinum

OM

Energy

28 rf: 2

51 Ik 9

76 t 5

73 t 2

Lawrence 19762

Lawrence 1976

37 *

18

46 + 23

85 + 6

71 + 6

Lawrence 1976

Echinometra

lucunter

E. viridis 35 + 17

58 k 20

86 + 3

80 + 4

Lawrence 1976

Lytechinus

variegatus

54-57

65

19 zk I

67 k 6

47 k 2

13 2 9

Moore and McPherson 1965’

Greenway 1 9762

Lowe and Lawrence 1 9762

Lowe and Lawrence 1976

6?9

57 k 7

-9 ?I 14

4*

10

Lowe and Lawrence 1976

Strongylo-

centrotus

intermedius

S. purpuratus

Tripneustes

ventricosus

32

Fuji 1967 (in Lawrence 1975)’

67

Fuji 1967 (in Lawrence 1975)

52k 17

Leighton 1968 (in Lawrence 1975)’

Moore and McPherson 1965’

Lawrence 1 9762

Lawrence 1976

52-56

30 k 5

71 k 14

78 + 4

15 t 3

Fish

Sparisoma

radians

54

7

Lobe1 and Ogden 1981

I

Lobe1 and Ogden 198 1

54

57

Reptiles

Chelonia mydas 45-61

Bjomdal 19803

65-71

Bjomdal 1979b

34-62 Bjomdal 1980

64-69

Bjomdal 1979b

Role of Seagrass Macroherbivores

357

Table 1. Continued.

Diet

Nutrient

DE (%)

Reference

Birds

Branta bernicIa

Sirenia

Dugong dugon

Zostera noltii

HaIophila sovaIis

Carbon

Cellulose

Hemicellu-

lose

Hemicellu-

lose

Total N

Organic N

Energy

OM

NDF

ADF

N

P

63-75

Thayer et al. 1982”

85-89 Bjomdal 1980

77-94

Bjomdal 1979b

53-75 Bjomdal 1980

78-94

Bjomdal 1979b

15-54 Bjomdal 1980

25-44 Thayer et al. 1982

67 Chartnan 1 9792

85 Murray et al. 1 9773

84 Murray et al. 1977

82 Murray et al. 1977

70 Murray et al. 1977

63 Murray et al. 1977

tent (Table 1). No explanation for the low

digestion of clean

Thalassia

blades was of-

fered by the authors. Survivorship of S.

ra-

dians

on pure diets of clean

Thalassia

blades

was lower than survivorship on mixed diets

or on pure diets of

Syringodium, Halodule

or

Thalassia

with microalgal epiphytes, but

was higher than survivorship on pure diets

of the macroalgae

Dictyota, Caulerpa, Hal-

imeda

or

Penicillus.

Two estimates of feeding rates for S.

ra-

dians

were calculated from data presented

by Lobe1 and Ogden (198 1). Both are based

on mixed diets composed mainly of sea-

grasses (Table 2).

Green Turtles

The only reptile that consumes seagrasses

to any extent is the green turtle,

Chelonia

mydas,

which feeds on the ‘blades of many

species of seagrasses as well as algae (Mor-

timer 1976, 198 1, 1982). In the Miskito

Cays, Nicaragua, the major :feeding area for

green turtles in the Caribbean,

Thalassia

testudinum

makes up 87% by dry weight of

the diet; other seagrasses

(Syringodiumfili-

forme

and

Halodule wright@

account for

another 5% (Mortimer 1976). Mortimer

(1976) found no

T. testudinum

rhizomes in

the stomachs of 202 green turtles. Only 3.6%

of the dry weight of S.

filiforme

in the stom-

achs and 0.9% ofthe dry weight of

H. wrightii

was rhizomes. Rhizomes of

T. testudinum

are ingested by green turtles in the southern

Bahamas, but apparently only when they

are uncovered by the burrowing shrimp

Upogebia

sp. (Bjorndal 1979a).

Green turtles graze selectively on

T. tes-

tidinum

by biting at the base of tall leaves,

and allowing the upper, older portions to

float away (Bjorndal 1980); the base has a

higher nitrogen and lower C/N ratio than

the upper portion (Fig. 1). They also main-

tain grazing plots, repeatedly cropping the

young regrowth in grazed areas (Bjorndal

1980; Ogden et al. 1983). These feeding be-

haviors result in a diet that is higher in pro-

tein and lower in lignin (Bjorndal 1980).

Zieman et al. (1984) have suggested that

these grazing patterns result from green tur-

tles avoiding epiphytes high in calcium car-

bonate. Although green turtles may select a

diet low in calcium, the two hypotheses pro-

posed by Zieman et al. (1984) to explain the

negative effects of ingesting calcium car-

bonate on green turtles-significant de-

crease in the digestion of seagrass blades and

a sudden influx of calcium ions into the

bloodstream-are possible but unlikely. Lit-

tle digestion of seagrass blades occurs until‘

the food bolus reaches the caecum (Bjorndal

1979b, in press). Therefore, if stomach pH

were raised by calcium carbonate, it would

have little effect on digestibility of seagrass

blades. Also, absorption of calcium ions in

the proximal small intestine is facilitated by

low gastric pH. If gastric pH is buffered by

carbonate, calcium ions would precipitate

and be unavailable for absorption.

Digestive efficiencies (Table 1) and feed-

358

G.

W. Thayer et al.

Table 2. Feeding rates of herbivores consuming seagrasses. Superscripts indicate units measured: 1 = g wet wt

per animal per day, 2 = g dry wt per animal per day, 3 = mg wet wt per g fresh wt per day, 4 = g dry wt per kg

fresh wt per day. 2 was estimated from 1 by assuming 20% dry matter for seagrasses.

Diet

Feeding Rate

Reference

Sea Urchins

Lytechinus

variegatis

Tripneustes

ventricosus

Strongylocentrotus

intermedius

Fish

Sparisoma

radians

Reptiles

Chelonia mydas

Birds

Branta bernicla

Anas acuta

A. penelope

A. platyrhynchos

Cygnus olor

Sirenia

Dugong dugon

Thalassia

testudinum

Syringodium

jiliforme

Thalassia

testudinum

Phyhospadix

iwatensis

Mixed-mainly

seagrasses

Thalassia

testudinum

Zostera noltii

Z. noltii

Z. marina

Seagrasses

l-6’ = 0.2-1.22

Moore and McPherson 1965

0.14-0.182

Greenway 1976

0.62 Vadas et al. 1982

0.062 Lowe 1975

0.072 Lowe 1975

3.1-3.3’ = 0.62-0.66*

Moore and McPherson 1965

20.3’ Fuji 1967 (in Lawrence 1975)

232-236’ ^I 46-472

3.0-3.74

lOO-2002

1002 or 714

672 or 844

542 or 904

84* or 764

57Y or 774

5,6002

Lobe1 and Ogden 198

1

Bjomdal 1980

Charman 1979

Jacobs et al. 198 1

Mathiasson 1973

Heinsohn et al. 1977

ing rates (Table 2) of four size classes of

green turtles have been measured using lig-

nin as an undigestible marker, by collecting

feces for a 24 h period from free-swimming

turtles that were feeding

ad libitum

on

T.

testudinum

(Bjorndal 1980). Concentra-

tions of the end products of the fermenta-

tion-volatile fatty acids (VFA) and lac-

tate- were measured along the gut of green

turtles at time of death, and rates of VFA

production were measured

in vitro

(Table

3). The gases produced

in vitro

during a 5

h incubation of caecum contents under N2

were collected and analyzed. Composition

of the gas phase was: CH4, 0.03 5%; HZ, 2.1%;

CO*, 1.9% (Bjorndal 1979b). Gut contents

from these turtles also were analyzed for

carbon and organic nitrogen (Thayer et al.

1982) and for organic matter, cellulose,

hemicellulose, lignin and energy content

(Bjomdal 1979b). Digestive efficiencies were

calculated using a lignin ratio (Table 1).

Birds

In temperate waters, wildfowl are major

seagrass herbivores. Birds that feed on sea-

grasses, like sea urchins and fish, are ap-

parently dependent on physical maceration

of the blades in their gizzards to expose cell

contents to digestion. Most authors assume

that no cellulose is digested (Ranwell and

Downing 1959; Prins et al. 1980) based on

three lines of evidence: Mattocks (197 1)

found no cellulose digestion in the goose

Anser

anser,

microscopic inspection of sea-

grass fragments from the feces of brant show

intact cell walls, with cell contents evenly

distributed within the cells (Ranwell and

Downing 1959); and passage rates are very

rapid.

Values for feeding rates of

B. bernicla,

Anas acuta, A. Penelope, A. platyrhynchos,

and

Cygnus olor

(Mathiasson 1973) are

similar when adjusted for body mass (Table

2). Digestive efficiencies of wildfowl feeding

on seagrasses have not been well studied.

Data from Charman (1979) are rough es-

timates (Table 1).

A grazing pattern similar to that de-

scribed for the green turtle has been ob-

served in

B. bernicla

grazing on

Plantago

maritima

(Prins et al. 1980). Brant regraze

the same plots at four-day intervals, re-

moving about 30% of each rosette’s leaf ma-

terial during each grazing pass. Clipping

trials showed that this harvesting regime

yielded the highest regrowth of new tissue.

Sirenia

Sirenia- manatees and dugongs- are the

only marine mammals that feed on sea-

grasses (Bertram and Bertram 1968). The

Caribbean manatee,

Trichechus manatus,

moves between fresh and salt water (Camp-

bell and Irvine 1977), wlnile the dugong, Du-

gong dugon,

is strictly marine (Heinsohn

1972).

In marine waters, manatees feed primar-

ily on seagrasses and may prefer

Syringo-

diumjXiforme

to

Thalassia testudinum;

there

is no apparent selection for young versus

older blades or for blades with or without

epiphytes (Hartman 1979). Hartman (1979)

Role of Seagrass Macroherbivores

359

Table 3. Concentrations of lactate and volatile fatty acids (VFA) at time of death (TJ in two green turtles

collected from Miskito Cays that had fed on

Thalassia testudinum

and one dugong that fed on

Halophila ovalis,

and rates of production of lactate and VFA in caecal fluid of the two green turtles. Green turtle data are from

Bjomdal(1979b) and dugong data are from Murray et al. (1977).

Green Turtles

Concentrations at T,

Esophagus

Stomach

Small intestine

Caecum

Anterior colon

Mid-colon

Rectum

Rate of production

Caecum

Dugong

Concentrations at T,

Stomach

Small intestine

Caecum

Large intestine

Lactate

mM per Liter

0.4

0.7

0.9

2.8

2.0

0.6

0.5

mM/l per hr

Total VFA mM per Liter

30.6

7.6

57.8

156.1

191.4

207.0

62.3

12.0

mM/l per hr

16

18

183

236

Acetate

Propionate

Molar %

Butyrate

Molar %

75.2

22.0 2.8

96.9

3.1 0

92.5

2.0

5.5

92.7

1.7 5.6

82.9

2.1

15.0

78.4

7.5 14.1

70.4

11.2 18.4

75.1

1.4 23.5

82

6 12

84

6

10

57

17

25

50

17 32

reported that manatees feed only on the

leaves of seagrasses, but Zieman (1982) de-

scribes manatees digging up rhizomes and

consuming entire plants. Feeding scars ap-

proximately 30 by 60 cm were left from

which nearly all of the rhizomes had been

removed. Zieman (1982) suggested that this

behavior would limit manatees to feeding

in areas with an unconsolidated sediment.

There are no published studies on nutri-

tion of manatees, other than accounts of

their feeding habits (Husar 1977), and the

only estimates of feeding rates are based on

animals held in captivity and fed primarily

lettuce. Based on these studies, manatees

consume from 20% to 25% of their body

weight each day (Hartman 1979; Zieman

1982).

Manatees are thought to have a hindgut

microbial fermentation similar to that in the

dugong (Murray et al. 1977). This is a rea-

sonable assumption based on the anatomy

and histology of the digestive tract which

has a midgut caecum (Reynolds 1980). The

fact that manatees constantly relieve flatus

(Hartman 1979) also suggests the presence

of a microbial fermentation.

The nutrition of dugongs has been more

thoroughly studied than that of manatees.

360

G. W. Thayer et al.

The diet of the dugong is composed almost

exclusively of seagrasses (Jarman 1966;

Heinsohn and Birch 1972; Lipkin 1975;

Johnstone and Hudson 198 l), although du-

gongs will feed on algae if the availability

of seagrasses is limited (Heinsohn and Spain

1974; Lipkin 1975). The more tender sea-

grass species are usually selected; tougher

ones, such as

Enhalus acoroides,

are avoid-

ed (Heinsohn and Birch 1972; Husar 1975;

Hudson 198 1). Dugongs usually ingest en-

tire plants, but when feeding on tall forms,

such as

Amphibolus antarctica,

the leaves

are stripped from the stem (Anderson 198 1).

Dugongs leave a characteristic feeding trail

that is a series of bare areas 19-25 cm wide

and l-5 m long. They dig about 3-5 cm

into the substrate and remove up to 86% of

the seagrass biomass; the average amount

removed is 63% (Heinsohn et al. 1977).

These grazing scars can be very extensive,

and Ogden and Rainey (pers. comm.) have

observed several hectares of densely over-

lapping scars in an intertidal seagrass flat at

Thursday Island, Australia. Based on the

intake of two captive dugongs that were

hand-fed seagrasses, it has been estimated

that an adult dugong will consume about

5.6 kg of seagrasses (dry weight) per day

(Heinsohn et al. 1977).

The dugong has a midgut caecum, and its

large intestine is twice as long as its small

intestine (Marsh et al. 1978). The caecum

and hindgut harbor an active microbial fer-

mentation (Murray et al. 1977). Samples

removed from the digestive tract of an adult

female dugong that had been feeding on

Halophila ovalis

were analyzed for nutrient

content and volatile fatty acid concentra-

tions (Table 3). The high VFA levels indi-

cate a rapid rate of fermentation, and the

percentages of organic matter, nitrogen and

fiber digested (Table 1) are high (Murray et

al. 1977).

Status of Research on Herbivore Nutrition

Seagrass herbivores can be divided into

two groups, based on how they extract nu-

trients from their diet. One group (sea ur-

chins, fish and wildfowl) lyses cell walls, di-

gests cell contents and passes the cell walls

with little or no degradation. Green turtles,

manatees and dugongs, on the other hand,

rely on a microbial fermentation in their

hindgut to degrade the cell walls and release

the cell contents. Cell walls are transformed

by the microflora into a source of energy for

the host in the form of VFA and lactate and

both cell walls and cell contents are digest-

ed.

Unfortunately, the few quantitative stud-

ies on the nutrition of seagrass herbivores

limit our ability to evaluate how feeding

behavior, feeding rates and digestive effi-

ciencies differ between these groups. Ac-

curate measurements of intake and digest-

ibilities are difficult to achieve; the wide

variation in the data presented in Tables 1

and 2 reflects this difficulty. Some of this

variation is due to normal individual vari-

ation, which is exacerbated by the small

sample sizes in many of the studies, and

some is due to differences in experimental

design and techniques.

Intake and digestive efficiency are influ-

enced by the physiological state of the an-

imal, and by diet quality, including the pres-

ence of anti-herbivore compounds. In

addition, digestive efficiencies vary due to

feeding rate and the physical state of the diet

presented; increased feeding rates generally

result in decreased digestibilities, and ground

diets may be more digestible than un-

ground. The nutrient quality of the diet is

very important in determining both intake

and digestive efficiencies, and these data

need to be presented with results of intake

and digestibility studies for comparison

purposes.

Feeding rates are presented in a variety

of units. Generally, the most useful measure

is g of dry matter ingested per g or kg of

herbivore (live weight) per day, but fre-

quently weights of study animals are not

reported (Table 2). Similarly, because dif-

ferent nutrients are studied in digestibility

trials, results cannot be compared. Results

often are given for only one nutrient, usually

dry matter, organic matter or energy. Often

the author can greatly increase the value of

the study by making a few simple calcula-

tions and expressing the digestibility values

for at least one more variable.

Two techniques are generally used to

measure digestibility: trials in which undi-

gestible markers are used, and total collec-

tion trials in which all food consumed and

all feces produced are weighed and samples

Role of Seagrass Macroherbivores

361

of each are taken for analyses. These differ-

ent approaches present no problems for

comparisons if appropriate markers are

used.

Study of comparative nutrition of sea-

grass herbivores has been hampered by the

paucity of quantitative studies and by vari-

ations in techniques used and measure-

ments recorded. Future research should be

directed toward studying more species on a

wider range of diets, including both sea-

grasses and algae. Also, emphasis should be

given to studies of possible anti-herbivore

compounds in seagrasses. Careful measure-

ments of intake and digestibility, expressed

in units that will allow comparison, will in-

crease understanding of large herbivores in

the seagrass ecosystem and will enable

quantification of their influence on energy

and nutrient flow patterns.

Grazing Patterns of Urchins and Fishes

in Tropical Seagrass Meadows

Consumers derive energy and nutrients

from seagrasses and associated epiphytes

from three distinct pathways: 1) direct her-

bivory, 2) detrital food webs and 3) ex-

ported material that is consumed in other

systems either as macrolplant material or as

detritus (Zieman 1982). While the detrital

food web still appears to be the largest path-

way, in many regions both the direct utili-

zation pathway and the export of material

may be quantitatively large and even be-

come locally dominant. Although grazing is

present at low levels in many regions, the

diversity of grazers and intensity of grazing

are highest in the seagrass beds of the Ca-

ribbean and southern Florida (Ogden 1980).

The most abundant and common grazers

are small fish and sea urchins, with grazing

by gastropods, green turtles, and manatees

existing in varying amounts through the en-

tire region.

Major Grazers

The number of species that derive their

primary nutrition from direct herbivory on

seagrasses is numerically small, although the

number of individuals can be abundant in

seagrass beds (McRoy and Helfferich 1980).

The major invertebrate herbivores on sea-

Role of Larger Herbivores in Seagrass Communities

Abstract and Figures