Nutrition and grazing behavior of the green turtle, Chelonia mydas, a seagrass herbivore


Material Information

Nutrition and grazing behavior of the green turtle, Chelonia mydas, a seagrass herbivore
Physical Description:
ix, 73 leaves : ill. ; 28 cm.
Bjorndal, Karen Anne, 1951-
Publication Date:


Subjects / Keywords:
Green turtle   ( lcsh )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )


Thesis--University of Florida.
Includes bibliographical references (leaves 66-72).
Statement of Responsibility:
by Karen Anne Bjorndal.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000089669
notis - AAK5055
oclc - 05739437
lcc - QL666.C5 B5
System ID:

Full Text








I would like to thank the members of my committee: Dr. Archie

Carr, my chairman, for his continuing inspiration and encouragement;

Dr. John Kaufmann for his careful editing; Dr. John Moore for making his

laboratory facilities available to me with cheerful generosity and for

introducing me to the study of nutrition; and Dr. Hugh Popenoe for his

help with nitrogen determinations. I would also like to thank Dr. Frank

Nordlie for his time and equipment; John Funk for teaching me the NDF,

ADF and lignin analysis techniques with much-appreciated patience; Dr.

Paul Smith and Dr. David Boone for their advice and assistance in analyz-

ing fermentation end-products; Donna Gillis for her wise counsel and

skillful typing; and Esta Belcher for preparing the figures. I am

grateful to Alan Bolten, Jeanne Mortimer, Ken Prestwich and Doug Simmons

who have provided advice and encouragement throughout the study.

The funding for this study was provided by the Inagua Project of

the Caribbean Conservation Corporation. I am indebted to Morton Bahamas

Limited, particularly Mr. Charles Bremer, and to the Bahamas National

Trust, especially Mr. Michael Lightsbourn, for their invaluable logistic

support. I acknowledge the support of NSF IDOE Living Resources Program

grant to the Seagrass Ecosystem Study Group for the work done aboard

the Alpha Helix.

Finally, I would like to express my thanks to Sammy and Jimmy

Nixon, Bahamas National Trust wardens on Inagua, without whose help

this study would not have been possible. They not only provided muscle

for turtle catching and fence building, but also shared their extensive

knowledge of the natural history of Inagua with me. I am grateful to

them for their companionship, for their assistance, and for introducing

me to boxfish roasted whole in its shell.









Feeding Behavior . .
Digestibility and Intake .
Changes along the Digestive Tract.


Feeding Behavior .. .
Nutrient Composition and Digestibi
Changes along the Digestive Tract.
Seagrass Herbivores .
Algae as an Alternative Diet .




i i














Table Page

1 Percentagesof Thalassia blades in green turtle
feces. . . ... ..... 17

2 Percentages of Thalassia rhizomes in green turtle
feces. . . ... ..... 18

3 Percentages of sponges in green turtle feces .. 19

4 Components of Thalassia blades from grazed patches 21

5 Components of Thalassia blades from ungrazed patches 22

6 Organic matter content of Thalassia blade feces from
green turtles. . . ... ... 23

7 Energy content of Thalassia blade feces from green
turtles. . . 26

8 Neutral detergent fiber content of Thalassia blade
feces from green turtles . .... .27

9 Acid detergent fiber contents of Thalassia blade feces
from green turtles . .... .28

10 Sulfuric acid lignin content of Thalassia blade feces
from green turtles . . 29

11 Nitrogen content of Thalassia blade feces from green
turtles and the apparent digestibilities of nitrogen 30

12 Quantities of Thalassia blade feces produced daily by
green turtles. . . .... 31

13 Quantities of Thalassia blades consumed daily by
green turtles . .. .. 32

14 Means and results of Duncan's multiple range tests
for green turtle diet variables and digestibility
coefficients . . 38

15 Changes in nutrient concentrations along the
digestive tracts of two green turtles and one
dugong . . 41



16 Concentrations of VFA's and lactate along the
digestive tracts of a green turtle and one
dugong . . . 42

17 Time-zero rates of production of VFA's and lactate
in green turtle cecum fluid. ............... 40



1 Apparent digestibilities of organic matter .

2 Apparent digestibilities of energy .

3 Apparent digestibilities of neutral detergent fiber

4 Apparent digestibilities of acid detergent fiber..

5 Apparent digestibilities of cellulose .

6 Apparent digestibilities of hemicellulose .. ...

7 Maximum and minimum water temperatures in the
study area . . .

8 Consumption rates of green turtles .


. 33

S 34




. 36

* 39


Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy



Karen Anne Bjorndal

June 1979

Chairman: Archie Carr
Major Department: Zoology

The apparent digestibility coefficients of the components of

Thalassia testudinum leaves were determined for four size classes (8,

30, 48 and 66 kg) of green turtles, Chelonia mydas. The lignin ratio

technique was used. The apparent digestibility coefficients for organic

matter ranged from 32.6% to 73.9%; for energy from 21.5% to 70.7%; for

neutral detergent fiber from 62.3% to 80.4%; for acid detergent fiber

from 64.9% to 85.2%; for cellulose from 71.5% to 93.7%; for hemicellulose

from 40.3% to 90.8%; and for protein from 14.4% to 56.6%. Digestive

efficiency varied with water temperature and turtle size. Total 24-hour

fecal collections were taken and, again with a lignin ratio, the rates

of consumption for the four size classes were calculated. A carrying

capacity of 138 adult green turtles per hectare of Thalassia testudinum

was derived.

The concentrations of volatile fatty acids (VFA) along the diges-

tive tracts of two green turtles were measured, and the rate of VFA

production in the cecum was determined in vitro. It is estimated that


14.2% of a green turtle's daily energy budget could be supplied by the

production of VFA's in the cecum. The cecum fermentation evolves

hydrogen gas, not methane, which is the usual gas given off by gut

fermentations of other herbivores.

Green turtles are selective grazers. In the study area, they

maintained grazing plots of young Thalassia leaves by consistent re-

cropping, and thus consumed a forage higher in protein and lower in

lignin than the ungrazed, older Thalassia leaves. The selectivity of

green turtles for either a seagrass or an algal diet may reflect the

specificity of their intestinal microflora.


Large areas of the shallow, sheltered marine regions of the world

are covered with seagrass beds. In the Caribbean, most seagrass beds are

made up of combinations of Thalassia testudinum (turtle grass), Syringo-

dium filiforme (manatee grass) and Halodule wrightii (shoal grass), which

vary from monospecific stands to any combination of the three. Thalassia

testudinuw is the dominant species; H. wrightii accounts for the smallest

proportion of the biomass. A fourth species of seagrass, Halophila

baillonis, normally grows in monospecific stands (Hartog, 1970).

The seagrass ecosystem is one of the most productive in terms of

biomass production (Westlake, 1963; Odum, 1971; Fenchel, 1977). Of that

production, much of the greater part enters the detrital food chain

(Fenchel, 1970 and 1977). Ogden (in press) estimates that in the

Caribbean, which is apparently unique for its high number of seagrass

herbivores (Kikuchi and Peres, 1977), about 90% of Thalassia blades

enter the detrital system. Outside the Caribbean more than 90% of

Thalassia leaf production is cycled through the detrital food chain.

Ogden (in press) presents a review of seagrass herbivores in the

Caribbean. Of the invertebrate seagrass consumers, sea urchins are the

dominant species. The five common species of regular sea urchins

found in the Caribbean are Diadema antillarum, Echinometra lucunter,

Eucidaris tribuloides, Lytechinus variegatus and Tripneustes ventricosus;

all feed on seagrasses (Lawrence, 1975). The queen conch, Strombus


gigas, and the asteroid Oreaster both feed on seagrasses and their


All but one of the classes of vertebrates--the amphibia--contain

at least one seagrass consumer. Today, the fishes are the dominant

vertebrate seagrass consumers. The major fish groups that feed on

Thalassia blades include the parrotfishes (Scaridae), the surgeonfishes

(Acanthuridae), the porgies (Sparidae) and the halfbeaks (Hemiramphidae).

Carr and Adams (1973) examined the gut contents of 21 species of fish

inhabiting seagrass flats near Crystal River, Florida. They found that

only one species, the halfbeak Hyporhamphus unifasciatus, fed on sea-

grasses to a significant degree. Two other species, Diplodus holbrooki

and Lagodon rhomboides, fed on algae epiphytes. Randall (1967) found

that the diet of over 30 species of fish in the Caribbean consisted at

least partly of seagrasses.

The brant, Branta bernicla, is probably the bird best known for

feeding on seagrasses. Prior to the 1930's, the diet of the brant was

almost exclusively eelgrass, Zostera marina (Cottam et at., 1944).

Following the wasting disease that destroyed most of the eelgrass flats

along the Atlantic coasts of North America and Europe, the brant popu-

lation plummeted. The remaining birds fed on sea lettuce, UZva spp.

A recent study of brant feeding habits in New Jersey (Penkala, 1975)

showed that Ulva Zactuca is still the major food item, and Zostera is

second in importance. Several species of sea ducks and geese feed on

Zostera to varying degrees. These include the Canadian goose, Branta

canadensis, the emperor goose, Anser canagicus, the king eider,

Somateria spectabilis, the greater scaup, Aythya maria, the lesser

scaup, Aytha affinis, and the mute swan, Cygnus otor (Johnsgard, 1975).

McMahan (1970) found that the seagrass Halodule wrightii made up 84% by

volume of the winter diet of redheads, Aythya americana, 88% of the diet

of the pintail, Anas acuta, and 22% of the diet of lesser scaup, Aythya


The mammalian seagrass consumers, other than man (Felger and Moser,

1973), are limited to the order Sirenia--the dugong, Dugong dugon, and

two of the three species of manatee, Trichechus senagalensis and

T. manatus. The diet of the dugong consists almost entirely of sea-

grasses, particularly Cymodocea and HaZodule In northern Australia

(Heinsohn and Birch, 1972; Heinsohn et al., 1977) and Halodule, Halophila

and Cymodocea in the Red Sea (Lipkin, 1975). Manatees eat seagrasses,

but they also consume large quantities of marine algae and freshwater

plants (Bertram and Bertram, 1968). Even before the manatee populations

were drastically reduced by man, they probably did not graze seagrasses

to a significant degree.

The only reptilian seagrass consumer is the green turtle, Chelonia

mydas, which, prior to man's over-exploitation of green turtle popu-

lations, was certainly the major seagrass consumer in tropical and sub-

tropical waters. The correlation between green turtle distribution and

seagrass distribution has been noted (Parsons, 1962). However, green

turtles feed on algae in some areas where seagrasses are lacking such as

the coast of Brazil (Ferreira, 1968), in the Galapagos Islands

(Pritchard, 1971), and in the Gulf of California (Felger and Moser,


Mortimer (1976) has recently completed a thorough review of the

literature on green turtle feeding habits. She examined 202 stomachs

and found that in the Miskito Cays, Nicaragua, the major feeding grounds

of the green turtle in the Caribbean, green turtles consume Thalassia

testudinwn, Syringodium filiforme and Halodule wrightii. On a dry

weight basis, she found that Thalassia made up 87.34% of the combined

stomach contents; Syringodiwn accounted for 4.70%; and Halodule, 0.40%.

Seagrasses, therefore, constituted 92.44% of the dry weight of 202

stomach contents.

The degree to which these seagrass herbivores digest their food has

not been well studied. A series of papers dealing with the digestive

enzymes, gut microbes, feeding rates and digestive efficiencyof sea urchins

that feed on seagrasses has been published (Prim and Lawrence, 1975;

Greenway, 1976; Fuji, 1962; Lowe, 1974; Moore et at., 1963a, 1963b; Moore

and McPherson, 1965). These were reviewed by Lawrence (1975).

Digestion in vertebrate seagrass consumers has been almost totally

neglected. Fish studies have been limited to gut content and selectiv-

ity studies. Nothing is known of the digestive capabilities of fishes.

The digestive systems ofavian seagrass consumers have not been studied.

The extent of Zostera breakdown in the brant is not known. The diges-

tive systems of manatees and dugongs are better known, although much work

remains to be done. The anatomy of the digestive system of the dugong

has been described by several authors (Marsh et at., 1977; Kenchington,

1972; Osman-Hill, 1945). Fiber digestion and digestive efficiency in

dugongs was.studied by Murray et at. (1977). The digestive efficiency

of captive manatees on diets of lettuce and water hyacinths, Eichhonia

crassipes, has been measured (Lomolino, 1977). It has been estimated

that an adult manatee may consume 20 kg wet weight of freshwater vege-

tation per day (National Academy of Sciences, 1976).

Other than stomach content analyses and brief anatomical descrlp-

tions, there have been no studies of the digestive system of the green

turtle. In fact, the digestive systems of few herbivorous reptiles have

received much attention. Aspects of the gut anatomy of herbivorous

lizards are discussed by Iverson (in press), Beattie (1926), and Lonnberg

(1902). Harlow et al. (1976) determined the digestive efficiency of

Dipsosaurus dorsalis feeding on rabbit pellets and the effect of tempera-

ture on the digestive efficiencies. Throckmorton (1973) calculated the

digestive efficiency of Ctenosaura pectinata when fed on a very low-fiber

diet of sweet potatoes. Nagy and Shoemaker (1975) calculated the

assimilation efficiency for the chuckwalla, Sauromalus obesus, on a

natural diet. Unfortunately, the only attempt to test for cellulose

digestion in an herbivorous reptile gave uninterpretable results (Nagy,

1977). General discussions of the evolution of herbivory in lizards,

the necessary anatomical adaptations, and the selective pressures for

and against herbivory are given by Pough (1973), Sokol (1967), Ostrom

(1963), and Szarski (1962). Wilson and Lee (1974) estimated the energy

expenditures of the herbivorous lizard, Egernia cunninghami.

Obviously, the green turtle's importance in the seagrass ecosystem

had decreased as its populations have declined. It is well known that,

prior to the advent of European man, the green turtle populations were

orders of magnitude greater than they are today. Many of the beaches

described in the logs of 16th and 17th century ship captains as

important turtling beaches are no longer visited by turtles. Turtles

no longer nest on Grand Cayman, formerly the site of a hugh green

turtle rookery about which Long wrote in 1774: ". insomuch that it

is affirmed, that vessels, which have lost their latitude in hazy weather,

have steered entirely by the noise which these creatures make in swimming,

to attain the Caymana isles" (Lewis, 1940; p. 57). Similarly, Little

Cayman once was the site of a major green turtle rookery which no longer

exists. In July 1630, Pieter Adriaensz Ita saw a beach on the northwest

coast of Little Cayman ". where from May to October great numbers of

edible turtles come to lay their eggs in the sand in a single night

one or two thousand can be taken .. ." (Parsons, 1962; p. 28). Cuba,

also, had impressive numbers of turtles. The Gulf of Batabano on the

south coast of Cuba apparently still has a small turtle fishery, but in

1494 Bernaldez, Columbus' scribe, described encountering a group of

turtles so large "that it seemed as if the ships would run aground on

them and their shells actually clattered" along the hulls (Morison,

1942; p. 150).

Considering only the green turtle population of the greater

Caribbean area and the rookeries that at one time contributed to it, we

know that most of these rookeries are now gone, or nearly so: Bermuda,

mainland Florida, Dry Tortugas, Grand Cayman, Little Cayman and Alta

Vela (Parsons, 1962). The remaining important rookeries--Tortuguero,

Aves Island and Surinam--have evidently been greatly reduced. This

accounting does not consider the numerous, scattered nesting beaches

that were extirpated before they were recorded.

Because we know so little of their feeding ecology today, we

cannot even attempt to extrapolate back in time and estimate the prim-

itive impact of the green turtle on the seagrass flats. Most studies

of marine turtle biology have centered upon the adult females at the

nesting beaches, where they gather in relatively dense aggregations.

Because the turtles are thinly dispersed on their feeding grounds, little

is known of their ecology or behavior in that habitat.

I have addressed my study to gaps in the three areas discussed

above. I worked with the digestive physiology and feeding ecology of

the green turtle in order to 1) learn something about the digestive

system of a vertebrate seagrass consumer and the green turtle's role in

the seagrass ecosystem, 2) further the study of herbivory in reptiles,

and 3) gather new information on the ecology of green turtles away

from their nesting beaches.


Most of the data reported here are the result of one and a half

years of field work (April 1975 to August 1976) at Union Creek, a tidal

bay on the north shore of Great Inagua, the southernmost island in the

Bahamas. Shallow, sheltered tidal bays, called "creeks" by Bahamians,

are favored feeding places for the indigenous green turtle populations

throughout the Bahamas.

In 1964, the Caribbean Conservation Corporation, with the help of

the Bahamas National Trust, walled off Union Creek, thus impounding

several square miles of turtle grass (Thalassia testudinwn) flats. Stone

walls, 1.5 m wide, were erected at both openings to the bay--one approxi-

mately 120 m long and the second about 300 m--to keep the turtles in the

bay. After several years, culverts were put into the walls to increase

tidal flow, and a fence was put in on the inner side of the wall.

Some local green turtles were no doubt already in the bay at the

time the walls were finished. During the next few years their numbers

were increased by shipments of hatchling green turtles flown in from

Tortuguero, Costa Rica. These hatchlings were raised in pens of small-

mesh wire within the Creek, fed on chopped fish and conch, and released

into the Creek to feed on their own when they were about 15 cm long.

Over the years, the population there has been slowly increased, as small

local turtles are put in the Creek and lab-reared Tortuguero green

turtles are released there when they outgrow the facilities in Archie

Carr's lab at the University of Florida. Also, storms have broken

down the walls several times, allowing turtles to move into and out of

the Creek at will. Thus, today we do not know how many turtles are in

Union Creek, or what their origin has been. The number is probably at

least 100, but, as the area is large, and the turtles tend to hide from

motor boats, population estimates are difficult.

Most of my data were obtained from forage samples and 24 hour fecal

collections taken concurrently. During the first five months of the

study, various techniques of measuring total consumption and/or fecal

output were tested. Initially, six bottomless, 2 mby 3 m cages were

built with cyclone fencing and erected over thick seagrass patches. A

green turtle, 25 to 35 kg in weight, was put in each. I had hoped to

be able to determine the average daily food consumption of a green turtle

by estimating the standing crop of turtle grass in a cage before putting

the turtle in, and after a set period of occupancy. The turtles, however,

would not feed while so incarcerated; they divided their time between

sitting on the bottom and swimming into the fencing. They had to be

released, and the experiment discontinued because they cut their faces

badly and showed no signs of habituation.

A new approach was made by fencing off a 3 hectare area of the

Creek. This area contained lush Thalassia flats, interspersed with

algae: Penicillus capitatus, Batophora oerstedi, Acetabularia crenulata,

Udotea sp., Avrainvillea nigricans, CauZerpa spp., Halimeda sp. and the

very common red alga Spyridia filamentosa. Twelve turtles--three each

of the four size classes of 8 kg, 30 kg, 48 kg and 66 kg--were caught in

the Creek and put in the study area. Since the turtle supply was limited,

turtles were considered to be in a given size class if their weight

varied from the given value by less than 5%.

After several attempts, I developed a bag that could be fastened

around a turtle's cloaca to collect feces. These "diapers" were plastic

bags covered with a double layer of fine-woven cotton cloth. The mouths

of the bags were made to fit snugly around the cloaca and tail but were

not water tight. Because green turtles excrete a liquid urine composed

mainly of urea (Bjorndal, in press), urinary nitrogen was not collected

with the feces to any great extent. To avoid having to suture into

the turtle's skin each time a bag was attached, non-corrosive nickel-

cadmium wire loops were inserted around the cloaca. The bags could then

be sewn to the rings instead of directly into the skin. These rings

occasionally had to be replaced due to tissue necrosis, but this system

greatly reduced tissue necrosis and the risk of introducing an infection.

Twice a month, we caught the turtles in the study area, fastened

on bags, and released them back into the study area. After 24 hours, we

recaptured them, removed the bags, and returned the turtles to the study

area. Taking each bag separately, I sorted the feces into three

categories: Thalassia rhizomes, the sponge Chondrilla nucula, and the

remainder which was composed of Thalassia blades and metabolic excretions.

Hereafter, this last category will be referred to as Thalassia blade

feces. On the few occasions on which algae were encountered, these were

separated. The wet weights of each feces component were recorded for

each turtle, and the feces were dried to constant weight at 800C.

During the drying, any urinary nitrogen contamination would have been

driven off as ammonia. The dry weights were recorded, and the samples

placed in plastic bags in desiccators until analyses could be run upon

my return to the University of Florida.

On the days that fecal collections were made, I also collected

grab samples of four forage categories: Thalassia blades from ungrazed

areas, Thalassia blades from grazed areas, Thalassia rhizomes, and

sponges. These samples were treated in the same manner as the fecal


Samples were collected twice a month during the 12 month period

from September 1975 to August 1976. Each month's samples of each

feces component for each size class were combined. For example, all

of the rhizome feces from 30 kg turtles collected in March were combined

to provide a large enough sample on which to run the following series of

analyses. During the year I collected samples, I returned to the

University of Florida every three months to determine the energy content

of the food and fecal samples in a Parr bomb calorimeter (Parr, 1960).

The acid detergent fiber (ADF), neutral detergent fiber (NDF), sulfuric

acid lignin and total nitrogen determinations were not carried out until

I returned to the University at the end of my field work. Ash-free

NDF is composed of the cellulose, hemicellulose and lignin fractions

of the forage; ash-free ADF is made up of cellulose and lignin. Thus,

hemicellulose content can be calculated by subtracting ash-free ADF

from ash-free NDF, and cellulose can be calculated by subtracting

lignin from ash-free ADF.

For organic matter determinations, samples were ashed at 600C

for three hours. Standard procedures were followed for the ADF and

sulfuric acid lignin analyses (Van Soest, 1963) except that an acid-washed

asbestos mat was laid down over the scintered glass disc in the

crucibles. This mat was necessary to prevent the fecal samples from

clogging the pores in the disc. NDF analyses were carried out accord-

ing to standard procedure (Van Soest and Wine, 1967; Robertson and

Van Soest, 1977) except that the residue was collected on glass wool

in Gooch crucibles and rinsed with acetone. Total nitrogen was determined

with a micro-Kjeldahl technique. Following digestion and distillation,

the sample was titrated against HCI using a two part methyl red to one

part methylene blue indicator solution.

The apparent digestibility coefficient (ADC) of a nutrient is the

percentage of ingested nutrient that is apparently digested in the animal's

gut. That is:

ADC =Nutrient ingested nutrient egested x 100
Nutrient ingested

Theword "apparent" is used because feces contain microbes, gut secretions

and sloughed cells from the gut lining that might lead to slight mis-

calculations of the actual digestive coefficient of the dietary


The lignin ratio is a commonly used technique to calculate the

total amount consumed or defecated by an animal when either one of the

two is already known, and to determine ADC's of various dietary nutrients.

The lignin ratio is based on the assumption that lignin is not digested,

and therefore can be used as an internal marker with which to measure the

quantity of nutrient digested. Using the lignin ratio, the above equa-

tion becomes:
% Indicator % Nutrient
in diet in feces
ADC = 100 inx x 100
% Indicator % Nutrient
in feces in diet

Van Dyne (1968) has reviewed the studies in which lignin ratios have been


In order to test the assumption that green turtles do not digest

lignin, I tried to run a series of trials with green turtles in cement

tanks located on Great Inagua where total intake as well as total output

would be measured. Only one small (4.5 kg) turtle belatedly began to

feed. For four days its food intake was measured, and the end of that period

was marked with a dye pellet fed to the turtle. As the turtle had not

fed for over a week, and had not defecated for several days, feces were

collected from the time they first appeared until the dye passed out.

Later analyses and calculations showed a 99.18% recovery of lignin.

I did not realize that the turtles were only feeding on young

Thalassia blades from grazed patches, and therefore did not collect

those leaves until December. As there was no seasonal change in lignin

content, the values for the nine months were averaged and a mean value of

4.55 was used as the percent lignin indicator in the diet in the above


Behavioral observations were made on the turtles both in the study

area and in the larger area of the Creek. I spent many hours underwater

watching the turtles both with and without their diapers. There was no

apparent behavioral change when turtles had collection bags. Also,

there was no visible leakage of feces from the bags.

The Inagua work was supplemented by a later follow-up study on

board the R/V AZpha Helix. In November, 1977, I spent two weeks in the

Miskito Cays, Nicaragua', on a Seagrass Ecosystem Study Group cruise.

Miskito Bank is the major feeding ground of the green turtle in the

Caribbean. While there, I obtained the complete digestive tracts of

two green turtles (50 kg and 82 kg) that had been caught the night

before by Nicaraguan turtlers and later consumed by them. Seven sections

along the gut were sampled: the esophagus, stomach, small intestine,

cecumm," anterior colon, mid-colon and rectum. The cecumm" in the green

turtle, though distinct from the rest of the colon, is a section of the

continuous gut tube rather than a blind out-pocketing of the gut.

Although it is thus not a true cecum, it functions like one. It lies

just posterior to the ileo-colic valve, and is wider and has a much

higher fluid content than the rest of the gut. In starved adults, the

entire gut is empty except for the cecumm," which always contains a

dark green fluid with small pieces of vegetation. In pre- and post-

hatching green turtles, the cecumm" is apparent as an enlarged section

of the intestine which contains a green paste, while the rest of the

tract is empty.

After the pH of each section was measured, two sets of samples

were taken from each section. One set was dried at 80C to constant

weight, and stored in plastic bags until my return to the University,

where ADF, NDF, lignin, organic matter and energy content determinations

could be made. The second set were liquid samples obtained by squeezing

gut contents through two layers of cheese cloth, and preserved with

NaOH for later volatile fatty acid (VFA), ethanol and lactate analyses.

The ADF, NDF, lignin, organic matter and energy content determina-

tions were carried out as described above. The liquid samples were

analyzed for VFA's and ethanol on a Packard 800 gas chromatograph.

Lactate levels were determined using the lactate test kit prepared by

Mannheim Boehringer.

When the above two sets of samples had been collected, I strained

the remaining contents of the cecum from each turtle through two layers

of cheesecloth. Five 100 ml samples were put into 250 ml bottles. In

order to determine the rates of VFA, ethanol and lactate production,

four of these samples were gassed with CO2, sealed and incubated at 300C,

the average body temperature of green turtles (Hirth, 1962; Mrosovsky and

Pritchard, 1971). One bottle was removed from the incubator each hour

for four hours, and a sample of the liquid was preserved in NaOH and

analyzed as above for VFA's, ethanol and lactate.

In order to determine what gases were evolved during fermentation

and their relative amounts, the fifth sample bottle was gassed with N2

and incubated at 30C for 5 hours. At the end of the incubation period,

the bottle contents were shaken to equilibrate the dissolved and free

gas phases. The gas was transferred to a tube, sealed, and later

analyzed for CO2, H2 and CHq on a Packard 800 gas chromatograph.

The fermentation in the first turtle was killed, probably from

mishandling, so no data on VFA, ethanol and lactate production rates or

gas evolution were obtained. The data here presented on those parameters

are from the second turtle only.

Replicates of organic matter, energy content, nitrogen, VFA and

lactate determinations were acceptable with 1% error. Replicates of

ADF, NDF and lignin were accepted within a range of 2%.


Feeding Behavior

The turtles in the study area consumed three food items: Thalassia

blades, Thalassia rhizomes, and the chicken liver sponge, Chondrilla

nucula. In Tables 1 to 3, the percent dry weight of each of these

components in the feces of each turtle size class, for each month, is

shown. These values underestimate the percentage of Thalassia blades

in the diet in two ways. First, the sponges usually have pieces of

shell or Halimeda ossicles incorporated in their tissue. These fragments

greatly increase their dry weight and make it seem as ff more sponge

tissue has been ingested than actually has. Secondly, the Thalassia

blades are apparently digested to a greater extent than either the

rhizome or the sponges. This greater breakdown of Thalassia blades

decreases its percentage in the feces from its percentage of food


Mortimer (1976) found no Thalassia rhizomes in the green turtle

stomachs she examined from Nicaragua. I found small pieces of rhizomes

not only in the feces of green turtles in the study area, but also in the

guts of green turtles caught around Inagua and butchered in Matthewtown.

I never saw any evidence of turtles either uprooting entire plants of

Thalassia, which have a very dense, interwoven mat of rhizomes, or

digging up rhizomes. The burrowing shrimp, Upogebia sp., is found in

great numbers around Inagua, as evidenced by their mounds. As these



,-o oo

O CO CM -9
C N cn oo Ui

) O O -
0u O o
C3 00

* 4- < o N m 0
c < 0

0 ~

- 0C O

u0 m O-

OO >
V) )4-1 r2 m 0o m0

L .
E X m o 0c 0'

0 IO
- L* C) CL LA
-aC L
1o Co 0)

0) C 4-J

.1 400 U\ r CM )9
ILO0 0 N 0 C
m LA 0 L 'D 0 '4 -

O a- oo0 -t
)0)> 0 4 0

4J 0 0 44
CO- )
a O) n n
LU 00 N- 0 '.0
0 0U

a) IO0 U
-) U 1u
-a 4) (u o

C0)Oi- i C L o

0 m 0 -m C)
u N- 0"

0) t L L* A

r.l U '.0 Li N- nA

0) N-4- 3o a
0r -- Or o Co

-0 0 ) ) 0 00 %
i- a) r~ r> o v
a) w r. en en e

C 0 -
4J C

r,. )

-0 0)
-: 15 -
4o-1 L 4- C
(M.0 4- *-
E0 4-
()0< (C
CL 4-1 O
>a0 <( -
L- fi E O.
S.- X
EJ 0

C 4- I 0
o LIea
d) >-

O (O

03 :3 f
N 4c1 01

O 3 *-

m Od
' E- >O .
So 0 r

S- 1

4 01 '4-

a'4- in C
mU 0 0
. .-- O

J 4-' t 0


^ N ig
L. '" _^

N 4 M

'0 -: t
N N 0o

o r- r

Lf %\ p
M cr\

o N

- \o

0r1 in

N c
'0 \

0 00 .0
en _r \%o

CD 0

(D E 4- -
4-' C
o- lo c.

ro E 0

C I- -

W 0
Sn) L

r- > X
; 4 0 0
L 4-J
3 C
41 m w
4-Jo 0" a

i n n 0)

C3 U -
0 *-

4- W (4-

0 "a 0

0 0
m -C )

c V c
S0 -

L- 411
in C-a
.c / r- Q.


-0 04
U 0 U

a) uir^ a
--c i n-


4 N

3 -- -

0 O~n i

0' ('4

(' ('

N 01
- 0

o --

- (C

( '


- \

0o c


0 C -

o CO

o L

C) *U

O cO
e> -* o

mounds are built and eroded, Thalassia rhizomes are exposed. Turtles

eat the rhizomes so exposed in the study area. I believe this is the

only common access green turtles have to Thalassia rhizomes, and that it

may explain the geographic variation in rhizome consumption.

Because the study area was separated from the rest of the Creek by

a wide, shallow strip which the turtles rarely crossed, the study area

had not been grazed by turtles.for some time. When the experimental

animals were added to the area there was a great increase in the amount

of Thalassia blades washing up on the shore on one side of the study

area. Examination of these blades showed that they had been bitten off

across the bottom, and the original rounded tip, or dead upper portions

were still present. Obviously, the turtles were grazing by biting the

lower parts of the leaves and allowing the uppermost parts to float

away. After a few weeks, the volume of the blades washed ashore had

decreased to approximately the pre-turtle level.

After the turtles had been in the study area for three months, I

realized that the area of grazed Thalassia was no longer increasing.

The grazed areas of short Thalassia blades were being recropped, while

adjacent stands of tall (average 20 cm) blades were untouched. Often

there were sharp boundaries between the grazed and ungrazed areas.

These boundaries remained essentially unchanged for the year of my

study. In Tables 4 and 5, the nutritional values of the grazed and

ungrazed Thalassia patches are shown. By re-cropping grazed areas,

the turtles maintain a source of forage of higher nutritional value.

The existence of such plots needs to be.corroborated in wild popu-


0 0



-3? I

* 0



> 0



*- In


0- -
L c f


m L

L. cI -


C1 C

mu- L

U 4-

m -- IU

C O*-

C- O




-- U >-

4uJ L

- E C
O Lu.

r- N-

o 04

Lo -" 4

a i O
-O 01 -~ r

* I
00 -:


(U O

U 0 I

1 >.
N L. 0
C 0
o L

4.-- ci


- 0-

m c
- E --

mI m

*- 0

cL I

S- uC

C -

) 0 E 0

O -O

E Oc- en

cr% \ c


Li. Li en
o 0
z < -J

NU -


- -T
M m


\o cr\

0e1 -

cr- o

in 0

LA 0








c 41)
C 4-1

i E



% 0- r- rN- r.


>c 0 0 U' -3

% 0

>C m m :3


L )

O) E
00 1 N '. e

LO >
lD z I- r-.. rN-

w 4- 0. LA co -3

0< %0 r" 1

U 0
3 U 7
> c
0 o 0
U 0 U 0 Ln ren -3

4 -0 LA
N m *r) C Cr C9

In' E to .0c *

L(- U E
0 N- N- N- N-
C) rf

00 N- N- N-T cn

n r-n M

The green turtles in Union Creek begin to move out of the sleeping

areas shortly after dawn. These movements are sporadic on the part of

individual turtles, with frequent stops either at the surface or on the

bottom. Usually, feeding does not begin until two hours after dawn.

Throughout the day it is intermittent, but there are peaks in feeding

between 0800 hours and 1000 and between 1400 and 1700. Most feeding

occurs during daylight, but I observed turtles occasionally feeding on

bright, moonlit nights. I attached chem-lights to three turtles and

released them on moonless nights. Once they had settled down, they

moved only to come to the surface to breathe and return to the bare

rock bottom where most of the turtles in the study area usually slept.

By watching for and listening to blowing turtles at night, I could

determine whether they were staying in one area or moving about. There

is apparently very little activity at night except for the occasional

coming up for air.

Both in the study area and in the rest of Union Creek, turtles

rested in specific areas both during the day and at night. These areas

were the deepest places in the Creek; about 7 m deep, with bare sand

and rocky bottom. The turtles spent a good part of each day at these

resting places, especially during the mid-day "break" from 1000 to 1400.

Within the resting area, individual turtles did not have specific resting

spots. At no time during my hours of observation did I see any

aggressive behavior between turtles. In fact, there was very little

interaction of any kind, and no indication that any kind of a

hierarchy existed.

Digestibility and Intake

The results of the analyses of acid detergent fiber (ADF), neutral

detergent fiber (NDF), lignin, energy, total nitrogen and organic matter

of Thalassia blades from both grazed and ungrazed plots and of Thalassia

blade feces from the four size classes of turtles are given in Tables 4

to 11. There are several ways of manipulating these data, and since

methods vary from author to author, I am including the data in this form

to facilitate comparisons between my data and future studies by other

workers. The apparent digestibility coefficients (ADC) of organic

matter, energy, NDF, ADF, cellulose and hemicellulose are given in

Figures 1 to 6. The ADC's were calculated using a lignin ratio as

explained earlier.

The lignin ratio technique can also be used to estimate the amount

a turtle consumes each day if the amount it defecates each day is known.

If you assume that a turtle's daily fecal output is the amount of feces

produced from one day's consumption, a reasonable assumption for a

regularly feeding herbivore, then the quantity of Thalassia blade dry

matter consumed in one day is calculated from the equation (using

percent dry matter values):

amount consumed (% lignin in feces)(amount of feces per day)
per day % lignin in food

The average amounts of Thalassia blades defecated in one 24-hour period

for each size group for each month are shown in Table 12. Only samples

from bags that showed no evidence of tearing loose were used in calcu-

lating these mean values. The average amounts of Thalassia blades

consumed daily were calculated using the above equation and are given

in Table 13.

o o r i
tn D C;

x I s M C4 C-


crN -M C14
0 *
~ } -N O

E -

0 --
m0 C, A



N- 0 N

L 4-

E u) cN c
3l L

C 0
0 N CM N Cn

_2 1 N LIn
00 mU N 0 N N

-4) -.
3 U, -I it 2

3 M -

cn r C M m

C L 00 OT

a) 1 4 m r N
U* o0 N- 0 -

4-1) 3 N N N-
O- < cMJ f f f
E0 4N

U 3

1c C o c* r 0

L 0

C -

N- 4-
a, -0 S -

eU Cu N N Lr 0

0 -

- r- -3 -

0 a,01 Q M Cr\ r

4-- 0 < 4 \
2* 0 N a)
4- T M 1 M

( t3 0 o -


a). t C N O -
4) co em

LLLu ( N M r ( L T
L- EI c: *n *

4- 0 a,
M 4- -0
d) E

0 M0 3-

I.- 1* ( 0 (M
4C m 4) m .*

0 CM c C


a m c 3 0

Lf l f r


I -


-- -in O O
S, .
L m

c r^- Do o o

0 E



U 0 a

l ,N c

00 N N
I- 4.

u -
L. E

41 4- A M
0 C

-o O0 0 '0 0
u Z M NM

.) 0

4J .- N

Lo in 0 r. cc cc
1" -
0) 0 '0 N N0

4- N C CM

Li -- Fri 0

a -f o D c
U) N N N




O 0 N

C (N c (N

a -10 P- n

C % T

ScC <

m c-
L- '0;

a- )
u 00 N D .4C


4L .
C 00 LA 0 CO
f0a n 0 LA


0 U (- -

0 .C

)iu m Tr\

O -

- 0 0 c)%
cre m -I mr
-)3 EO 00 f (r>

L *

C ap
C- n *I '.0 LA

a) CO 0 (N (N

I.-3 y m
.0 ~~ U) CO 0 LA

The nitrogen content on a percent organic matter basis
for Thalassia blade feces collected from green turtles
in January and August, 1976, and the apparent
digestibility coefficients.


% OM








% OM




4.4 56.6 5.8




Table 11.


0 0 0
AA 'N-

mf fA C

c 0 C4 r -T

A- Pn

'r- o
C *


4) 1fc
EO. U ) N 0. 0 r-.

"4-) 3 r0- 0 0
4) -

r uc "N o

un r-

u,- >- ,

'4- z AM N- N-

-t L


N) -0 a)

- L - 3 .

*0>- 04 -

eo o o

c3 L.
E~~ oM (14 -T

4)4 )LI cr4 0r -

u0 C) -1
S41 L4

- 0u -T oM
(U 0 -

rJ N- > Co

4 4) QN CM OD u 0

o u0 \ 3 \o CM

O s


-: o- 0
N -4

- A 0


o Ln -


Q )

o- -

c u 4-1
o .
0 -c

m i
-o -

0 E

O m

o 4.' ,-0

.1 E
4-0 X
*- 0

0 0 m

O L 0
*- a) X

TrcO *-
3r --
CT 0) <
V)1 T
C= L
TO tfl
*M 'C
v)t 0
.C 3 -
c- l

- 0 -
CM -

-^ 4 0~ LA

C CM) co M
c" N N

4- n in
Ln LA (r

M r-. M 4

r^ \o co

a0. 0 00 '0
P> 43- '.0

- a,
-J N


0 NC

1- 60




20 1 I 1 1 1

Figure 1.

..- S .


SA--.- 8 kg e--- 48kg
A---A 30kg o ..... 66kg



Apparent digestibilities of organic matter in four size
classes of green turtle feeding on Thalassia testudinwn

A.---- 8 kg *--* 48kg
A----A 30kg o-.---.. 66kg o

,. /........o...... > "
S" .0 "" /

N /
A... "-A' /

.- '

/ \\ .



\ .-


Id -


0 60-




9 20-



Apparent digestibilities of energy in four size
classes of green turtles feeding on Thalassia
testudinuw blades.

O I 1 I


Figure 2.



A---A- 8 kg *-48kg
A----A 30kg o----.o 66kg



Apparent digestibilities of neutral detergent fiber
in four size classes of green turtles feeding on
Thalassia testudinum blades.

6---6- 8 kg -- 48kg
A----A 30kg o ..... -o 66kg

60 1I-


Figure 4. Apparent digestibilities of acid detergent fiber in
four size classes of green turtles feeding on
Thalassia testudinum blades.



a 80


Figure 3.





Apparent digestibilities
classes of green turtles
testudinum blades.

of cellulose in four size
feeding on Thalassia

A--.-6 8 kg ---* 48 kg
A---- 30kg o .-oo 66kg


Apparent digestibilities of hemicellulose in four
size classes of green turtles feeding on Thalassia
testudinum blades.


Figure 5.




Figure 6.

Several general patterns can be distinguished from the data

presented in Tables I to 13, Figures 1 to 6, and the accompanying results

of ANOVA and Duncan multiple range tests shown in Table 14. First, as

would be expected, each size class consumes and defecates quantities that

are different statistically at the 0.01 level. The 8 kg turtles are

consuming significantly more sponges and Thalassia rhizomes and digesting

a significantly smaller percentage of the nutrients, except cellulose,

than the larger turtles. The 30 kg size class is intermediate--they

consume the same proportions of Thalassia blades, rhizomes and sponges

as the larger turtles; digest the same percentage of cellulose and

protein: but digest less hemicellulose, energy and organic matter.

There is no significant difference between the 48 kg and the 66 kg classes

for any of the parameters. Apparently the digestive system of 48 kg

turtles has attained the adult functional level, as there is no increase

in digestive efficiences between the 48 and 66 kg size classes.

Water temperature also affects digestive efficiency as would be

expected for a poikilotherm which at best maintains a temperature 2 to

3C above ambient (Hirth, 1962; Mrosovsky and Pritchard, 1971). As can

be seen from Figures 1 to 6, there is a greater variation in digestive

efficiences during the cooler months of October to March (see Figure 7).

To test whether smaller turtles were affected by temperature variation to

a greater extent than the larger turtles, the mean square errors of the

regression of each size class's performance over time were determined for

each of the digestibility coefficients, food consumption, feces production

and percentage of leaves, rhizomes and sponges in the feces. Comparing

the mean square errors of the different size classes within each category


> c


-1 0 c
(n Ca ) .-
0 w E E
c 0
4-O C


iO n 41
4-, .C-"

00 -
aN.- -
.r tn 4)
Io >


"0 0

m 0- C
wJ C

C0 0

y- O n

4- 4

vC 4J 4-
ta 4-

E E- In -
In 4- M u
O U 4- 4-
U00 I) M)
0n) o)
u- MD N 4-
O 4- .- 0
c In 41- 0

- 0 0 Si
10 0.4- C
> u0
-0 0 0-
C: C -C I-

Em m-'
E In 4- In -
4-1 0 41-'
In c 0.3
L- W L- I
m0 U U U -
D 4) In C
>4- 4- L. M
4- 00
0 0 C
S0 = 3 3
U 1 "










- 0
3 /

nO u
wU 3
U -0



U "

.0 -0 .0

o in

0 0 -3

0 e n mo

o -3

0 aM


ED .0 0
U*% C)

0 0

EU -0 0
N- ~ ~

-1 r-.
-I '0 F



.0 U

N 00
0 in

U -0

N- -
- (M
r- OC

0 00 N0
m -T %


o----0 MINIMUM


w "o--o-- ,,o ,o.
Q- 25- / 'O--0--0o--O' '0

\ ,o--o--o--o--o--o
20 -

Figure 7. Maximum and minimum water temperatures in the study
area recorded on each collection date for the
preceding time interval.

revealed no correlation between body size and effect of temperature at

the 0.01 level. This lack of correlation is also expected, as there is

not a significant increase in body temperature with size.

Changes along the Digestive Tract

The samples taken from along the digestive tracts of two turtles

caught on Miskito Bank, Nicaragua, were analyzed for ADF, NDF, lignin,

energy, organic matter, volatile fatty acids (acetate, propionate and

butyrate), lactate and ethanol. The results from these determinations

are given in Tables 15 and 16. No ethanol was found in the samples, so

it is not shown. The small intestine of turtle 1 was empty, so no

determinations were possible.

The gas sample following the five hour incubation of cecal contents

at 30C under nitrogen contained CHq = 0.035%, H2 = 2.1%, CO2 = 1.9%.

The amounts of VFA's and lactate produced, and their rates of production

during incubation at 30C under CO2 are shown in Table 17.

Table 17. Time-zero rates of production of total
VFA's (sum of acetate, propionate and
butyrate) and lactate in green turtle
cecum fluid.

Lactate Total VFA Acetate Propionate Butyrate
mM/l'hr mM/1-hr molar % molar % molar %

0.46 11.97 75.10 1.42 23.48

The lignin ratio was again used to calculate changes in concen-

trations of nutrients along the digestive tracts of the turtles. Changes

in cellulose, hemicellulose, energy and organic matter are shown in

Table 15.

Percentages of organic matter (OM), energy, cellulose (CEL), and
hemicellulose (HC) remaining in each region of the digestive
tract, calculated by a lignin ratio. The dugong digestibility
data are recalculated from Murray et at. (1977); the dugong pH
values are from Kenchington (1972). Total ADC is the apparent
digestibility coefficient along the entire gut. See text for
cellulose and hemicellulose calculations.

OM Energy CEL HC
pH (%) (%) (%) (%)

Green Turtle #1
(50 kg)
Small intestine
Anterior Colon
Mid Colon

Green Turtle #2
(82 kg)
Small intestine
Anterior Colon
Mid Colon

Ant. Sm. Intes.
Post. Sm. Intes.
Anterior Colon
Mid Colon
Posterior Colon

Total ADC:

Total ADC:


Total ADC:

Table 15.









9 %







Table 16. Concentrations of VFA's and lactate along the
digestive tract at time of death. Dugong data
are from Murray et at. (1977).

Lactate Total VFA Acetate Propionate Butyrate
mM/1 mM/1 molar % molar % molar %
Green Turtle

Esophagus 0.40 30.57 75.17 21.98 2.85

Stomach 0.69 7.64 96.86 3.14 0

Small Intestine 0.93 57.78 92.46 1.99 5.52

Cecum 2.79 156.13 92.69 1.74 5.57

Anterior Colon 2.75 191.36 82.89 2.07 14.95

Mid Colon 2.00 206.96 78.41 7.54 14.05

Rectum 0.60 62.28 70.46 11.17 18.37


Stomach 16 82 6 12

Small Intestine 18 84 6 10

Cecum 183 57 17 25

Large Intestine 236 50 17 32


Feeding Behavior

Green turtles display a certain degree of selectivity in their

feeding, as demonstrated by the different proportions of diet components

consistently found in different size classes when they feed over the same

area. Selective feeding is also shown by the manner in which turtles

graze the plots of young Thalassia blades that they maintain. Tables 4

and 5 show that lignin levels in ungrazed stands are 100% greater than

in leaves from grazed stands. Also, nitrogen (or, by multiplying by the

6.25 conversion factor, protein) is increased by 11% in grazed stands in

January and 6% in August. By re-grazing plots of Thalassia, green turtles

select a food which is higher in protein and lower in lignin. Lower

lignin levels are associated with higher digestibility in vegetation,

because lignin forms complexes with cellulose and hemicellulose, blocking

these structural carbohydrates from digestive enzymes. Thus, the

digestive efficiencies of green turtles are enhanced by their specialized

feeding behavior. Other herbivore species have been shown to select

plants or plant parts with low lignin levels (Field, 1976; Moss, 1977;

Moss and Miller, 1976; McNaughton, 1976).

Thalassia is well known for carrying heavy epiphytic loads of

both algae and invertebrates. The question has been raised as to what

importance these epiphytes have in the nutrition of seagrass grazers

(Mortimer, 1976; Ogden, in press). Some fish feed solely on the epiphytes

without ingesting any seagrasses (Carr and Adams, 1973), while others

ingest the seagrass but apparently only utilize the epiphytes (Ogden,

in press). Mortimer (1976) found very few epiphytic algae in the stomach

contents of green turtles, and therefore reasoned that epiphytes are

not important to green turtle nutrition. This conclusion is supported

by my observations that green turtles graze on young leaves which, since

epiphyte loads increase with the length of time a leaf surface is

exposed, have few epiphytes.

Nutrient Composition and Digestibility


Both the data from Inagua and the data from the Miskito Cays are

calculated using a lignin ratio. The use of lignin as an indigestible

tracer was necessary, as I was unable to measure directly the amount of

food consumed. Van Dyne (1968) reviewed the various tracer techniques.

Of the many techniques available, lignin is the one most widely used.

Conover (1966) used ash as a tracer in his work with zooplankton. He

used the formula:

U = (I- E)F x 100

where U is the percent utilization, F is the percent organic matter of

the food, and E is the percent organic matter of the feces. He found

that the formula results agreed well with the direct measurement of

utilizations from total fecal collections. Other workers have used

Conover's formula (Altig and McDearman, 1975; Johnson and Maxell,

1966; studies cited in Lawrence, 1975), but many have not tested the

accuracy of the technique for their species before using it. When the

data for percent organic matter of esophagus contents (food) and

rectum contents (feces) from the two Miskito Cays turtles were used

in Conover's equations, U's of -146% and -151% were obtained. That

is, organic matter was not absorbed but created along the turtles'

digestive tracts according to the ash tracer method. Obviously Conover's

equation does not apply to sea turtles and it should only be used with

zooplankton until its accuracy has been verified for other species.


Approximately 60% of the organic matter of Thalassia blades is

composed of NDF. Cellulose is the major component of the NDF (averaging

45.3%), and hemicellulose makes up a much smaller proportion (averaging

9.0%). Lignin accounts for an average of 4.6% (see Table 4).

Cellulose is digested with equal efficiency in all of the turtle

size classes. In Figure 5, it is clear that during the warmer months,

cellulose digestion is consistently 90% for all size classes. During

the cooler months, there is more variation, with values ranging between

72% and 91%. The fact that there is no significant difference among

the size classes in the percentage of cellulose digested implies that

once the cellulolytic gut microflora is established, the size of the

turtle or length of the gut are not significant in the digestion of


The digestibility coefficients of hemicellulose presented in

Figure 6 vary greatly, probably because the experimental errors in the

calculations of hemicellulose ADC's are high due to the small amount

of hemicellulose in grazed Thalassia blades.

Dugong dugon is a non-ruminant, herbivorous, marine mammal that

also feeds on seagrasses and has an active post-gastric fermentation

(Murray et al., 1977). These similarities provide the basis for a useful

comparison of digestive efficiencies. By comparing the above figures

for the green turtle with the value of 90% given in Table 5 for the

dugong, it is clear that even an 8 kg green turtle with an average

body temperature of 30C (Hirth, 1962; Mrosovsky and Pritchard, 1971)

digests cellulose as efficiently as a 250 kg mammal with a temperature

regulated in the upper 30's (Blair Irvine, pers. comm. for manatees--

dugong data not available).

Comparing green turtles and ruminants is more difficult because of

the differences in their food. Since variations in diet (e.g., lignin and

fiber content) can cause the apparent digestibility coefficients of

individual animals to change (Hungate, 1966), comparisons between

animals on different diets must be made with caution. On a diet of

orchardgrass, Daetytis glomerata, which has a cellulose and lignin

content similar to that of T. testudinwn, sheep have an apparent

digestibility coefficient of 67.4% for cellulose and 76.1% for hemi-

cellulose (Keys et al., 1969). These data, when compared with.the

green turtle data in Table 15, show that the green turtle is at least

as efficient as a ruminant with respect to fiber digestion.

Unfortunately, there are no fiber digestibility coefficients

available for other reptiles. There is, however, a study that reports

high cellulase activity, and assumes cellulose digestion, in the colon

of the herbivorous lizard, Sauromatus obesus (Nagy, 1977). The results

reported are difficult to evaluate for the following reasons: 1) the

cellulase activity was determined from in vitro experiments in which

natural conditions were not simulated. 2) If cellulose were digested

in the lizard, it would be reflected in a decrease of organic matter,

regardless whether the organic acid end products were assimilated, because

the gaseous end products (H2, CO2 and CH4) would be lost. However, the

author reports there was no decrease in organic matter in the colon and

rectum. 3) Manganese was used as a marker for organic matter changes

along the lizard's gut. In mammals, manganese is known to be repeatedly

absorbed and excreted back into the gut (Cotzias, 1962). Manganese,

therefore, may be an inappropriate marker in lizards. The only conclusions

that can be drawn concerning S. obesus are that there seems to be a

potential, as determined by in vitro study, for cellulase activity in

the colon, which was apparently not realized by these lizards at the

time of sampling. Obviously, more work needs to be done with S. obesus

to explore these implications.


The apparent digestibility coefficients of energy are given in

Figure 2. As previously mentioned, the ADC's of the 8 kg and 30 kg

turtles are significantly different from each other and from the 48 and

66 kg size classes which are not statistically different. The two

larger size classes have an average energy ADC of 60%. This value is

rather high for an herbivorous reptile, particularly one with a 30C

body temperature. Harlow et at. (1976) showed a clear relationship

between temperature and energy digestive efficiency in the desert iguana,

Dipsosaurus dorsalis, when fed rabbit pellets. At 330C, the iguanas had

an ADC of 54.3%. At 280C their digestive system failed, and they died

as a result of massive distension of the esophagus and stomach. The

group of iguanas cycled between 410C and 28C, the regime most similar

to natural conditions, had an ADC of 57%. Nagy and Shoemaker (1975)

give an energy ADC of 56% for free-ranging Sauromalus obesus, an

herbivorous lizard with a preferred body temperature of 37C. They

also give a value of 57% energy ADC for D. dorsalis, the same as Harlow

et al. (1976). The higher energy ADC's of the green turtle are probably

due to the fermentation of fiber in their gut.


Thalassia blades have a relatively high nitrogen content (Table 4).

The organic matter of young Thalassia blades contains 3.6% nitrogen, or

22.5% protein. As a percentage of dry matter, these figures are 2.7% and

16.7%, respectively. The apparent digestibilities of protein in the green

turtle are low (Table 11). The ADC's increase with the size of the

turtle, but only the 8 kg size class is significantly different at the

0.01 level.

The nitrogen ADC's of the green turtle are lower than the 70% ADC

given by Murray et al. (1977) for the dugong, determined from Kjeldahl

analyses, and the 70% ADC of Sauromalus obesus, an herbivorous lizard

(Nagy and Shoemaker, 1975). No explanation for the low protein ADC's

can be given at this time. The possibility of contamination from

urinary nitrogen exists, but seems unlikely.

Changes along the Digestive Tract

The data for the two turtles taken during the R/V Alpha Helix

cruise to the Miskito Cays, which are presented in Table 15, are similar

except for the lower apparent digestibility coefficients of turtle 2,

and the differences between the turtles in the nutrient values for the

ceca. Individual variation accounts for some of the difference between

the ADC's. Also, these data are not based on one food bolus passing

through the gut, but are calculated, using a lignin ratio, from successive

samples along the gut. Although both digestive tracts contained leaves

of Thalassia throughout their entire lengths, the variation between ADC

values could result in part from differences in the lignin content since

lignin increases with age in Thalassia leaves. As has been pointed out,

turtles select young leaves, and therefore low lignin food, so that the

entire range of lignin values for Thalassia would not be found in a

turtle's gut.

The difference between the cecum values results from differences

in the degree of digestion. The contents of the cecum of turtle 1 had

apparently been there some time, as the cecum fluid and contents were

well-mixed and resembled the colon contents. In contrast, turtle 2 had

one food bolus in the small intestine, just anterior to the ileo-colic

valve. In this turtle the contents of the cecum were not mixed with the

cecal fluid, and resembled the contents of the small intestine rather

than the more-digested contents of the colon. The cecum values for

turtle 1 represent food that has already undergone cecum digestion,

while the values for cecum nutrients in turtle 2 represent pre-cecum

digestion (as does the small intestine value).

To determine the percentage of total digestion that occurs posterior

to the ileo-colic valve, the rectal nutrient value is subtracted from the

cecal nutrient and divided by the total ADC. The data for turtle 2 show

that 83% of digested organic matter, 82% of the digested cellulose and

58% of the digested hemicellulose are digested in the cecum and large

intestine. The data for turtle 1 are inappropriate for these calculations

because of the advanced state of cecal digestion, which would result in

an underestimate of percent digestion in the cecum and beyond. The

morphological characters of the cecum, and these results, demonstrate

that the cecum is the initial site of cellulolytic activity.

The amounts of VFA's and lactate found along the gut at the time

of death are shown in Table 16. Ethanol was not found and so is not

included in the table. As can be seen, VFA and lactate levels are low

in the esophagus, stomach and small intestine, showing a low level of

carbohydrate fermentation. In the cecum and colon the levels rise

significantly, indicating a great increase in fermentation in that region.

The sharp drop in VFA and lactate concentrations from the mid-colon to the

rectum indicates that most of the VFA and lactate end products are absorbed

in the cecum and large intestine. Acetate is by far the major VFA

component, followed by butyrate and propionate. Isobutyrate was present

only in trace amounts and is not included here.

No other reptile is known to possess an active gut fermentation,

and there are no data on reptilian VFA production with which to compare

these data. However, the quantities of VFA's, particularly acetate, in

the cecum and colon of the green turtle are greater than those found in

the hindgut of the horse (Church, 1971) or the rumen of sheep and cattle

(Moir, 1968). The data in Table 16 for the dugong, Dugong dugon, from

Murray et al. (1977), correspond closely to the values for the green

turtle in Table 16. The dugong is a large marine mammal which, like the

green turtle, feeds on seagrasses.

The most common relative molar concentration of VFA's found in

microbial fermentation in the guts of vertebrates is acetate > pro-

pionate > butyrate (Hungate, 1966). The pattern of proportions in the

green turtle (acetate > butyrate > propionate) is, however, shared with

the dugong (Murray et al., 1977), the quokka, Setonix brachyurus, a small

marsupial (Moir, 1968), and the porcupine, Eretizon dorsatumw (Johnson and

McBee, 1967). The low proportion of propionate in the green turtle may

reflect a substrate low in soluble carbohydrates, as rumen microbes

produce high levels of propionate when the diet is high in soluble

carbohydrates (Hungate, 1966). The relatively high butyrate levels may

be a result of end product inhibition of the butyrate-to-acetate break-

down caused by the very high acetate concentration.

The ratio of % acetate:% butyrate for rates of production in the

cecum is much lower than the same ratio for initial concentrations in the.

cecum: 3.20 and 16.64, respectively. This difference suggests that either

butyrate is absorbed more rapidly than acetate from the cecum, or that a

significant quantity of acetate is converted to butyrate in vitro. If

the former explanation is correct, the rates of absorption from the cecum

are butyrate > propionate > acetate. This sequence is the same as that

for rock ptarmigan, Lagopus mutus (Gasaway, 1976a), willow ptarmigan,

L. lagopus (Gasaway, 1976b), sheep and cattle (Church, 1971).

The increase in the proportions of propionate and butyrate relative

to acetate in successive sections of the colon probably indicates

differential absorption rates of acetate > butyrate > propionate in that

region. This pattern is different from that suggested for the cecum. The

changes in VFA proportions in the lower colon might also result from a

change in the VFA's produced in this region. But, as the change is an

increase in propionate, and as the contents of the lower colon are high

in fiber and low in soluble carbohydrates, the former explanation of

differential absorption is more plausible than the latter explanation of

differential production.

McBee and West (1969) found ethanol to be one of the major fermen-

tation end products in most of the samples from the willow ptarmigan,

Lagopus lagopus, and they found low levels of lactate in several of the

ptarmigan. No ethanol was found in any of the samples from the green

turtles. Lactate was found in low levels and is certainly not a signifi-

cant energy source.

Table 17 shows the time-zero rates of production of VFA's and

lactate in one liter of cecum fluid per hour. The percent contribution

of acetate, butyrate and propionate to the total VFA production are also

given. The amount of energy a green turtle gains from the VFA's and

lactate formed in the cecum can be estimated from the data for turtle 2.

An 82 kg green turtle defecates approximately 74 g of organic matter

each day (see Figure 8). Using the 65% organic matter ADC for turtle 2,

an ingestion rate of 211 g of organic matter per day is obtained. From

Table 4, the mean energy content of one gram of Thalassia blade organic

matter is 18.9 kJ; the turtle thus ingests approximately 4000 kJ each

day. With turtle 2's energy ADC of 65%, the energy absorbed each day is

estimated to be 2600 kJ. The energy derived from VFA production can be

calculated from the energy values of acetate (899 J/mM), propionate

(1534 J/mM) and butyrate (2270 J/mM). Turtle 2's cecum contained I liter

of fluid. If we assume that the rate of fermentation in the cecum is





< ,o'

40- /


OO ,
0 1 I I



0 20 40 60 80 100 120 140

Figure 8. Consumption rates of green turtles feeding on
Thalassia testudinwn blades. Data for the
126 kg turtles are from Bjorndal (unpubl.).

constant throughout the day, we arrive at a figure of 370 kJ/day of VFA

being produced in the cecum. If most of the organic acid end products

are absorbed, the cecum provides 14.2% of the green turtle's daily

energy budget.

The gut fermentation in the green turtle provides more energy than

the value calculated for the cecum. The colon is very long in green

turtles, and there is active fermentation along much of its length, as

evidenced by both the increase in VFA concentration past the cecum and

the rapid swelling of the gut due to accumulated gases once the intestine

is removed from the turtle. Therefore, the cecum value is only a portion

of the total contribution of the gut fermentation to the energy balance

of the green turtle.

The gas sample following the 5 hour incubation of cecum fluid con-

tained the following: CH4 = 0.035%, H2 = 2.1%, CO2 = 1.9%. The gases

evolved during cecum fermentation are not in the proportions normally

found in gut fermentation in herbivores (Hungate, 1966). The CH4 value

is quite low and the H2 value is high. The low pH of the cecum (Table

15) blocks the methanogenic pathway, resulting in the accumulation of

free H2 (P. E. Smith, pers. comm.). Farther along the colon, as the pH

rises, the fermentation may become methanogenic.

The proportions of gases produced are similar to those reported for

the quokka, Setonix brachyurus, a macropodid marsupial (Moir, 1968).

Moir found that the gas produced soon after fermentation began had a

much higher hydrogen content and lower methane content than the rumen

gas of cattle. Gasaway (1976c) found that methane production in rock

ptarmigan was not an accurate estimate of cecum fermentation, as it is

in cattle, and suggested that hydrogen might also be given off, although

he had not tested for the presence of that gas.

Seagrass Herbivores

As previously mentioned, over most of its range the green turtle

feeds primarily on seagrasses and, prior to man's destruction of green

turtle populations, it played a major role in nutrient and energy cycling

in the seagrass ecosystem. Little is known concerning the Inter-relation-

ship of green turtles and seagrasses; the primary aim of this study was

to increase our knowledge of this association.

It is difficult to estimate the effect the green turtle had on the

seagrass system in~their former, high numbers. One obvious effect was

that a greater proportion of leaf biomass was channeled into the herbivore

food chain, away from the detrital chain. As discussed in greater detail

elsewhere (Thayer et at., in prep.), grazing by green turtles and their

depositing of feces with a higher nitrogen to carbon ratio than the

Thalassia blades which they consume, act to shorten the energy and

nutrient cycling time in seagrass beds.

The green turtle also serves as an energy and nutrient link between

the nutrient-rich seagrass system and other, poorer, communities. This

nutrient transport is associated with predation, daily movements and

reproductive migrations. Now that large turtles are regularly killed by

man, their biomass is removed from the seagrass system. Prior to man,

the major predators on turtles inhabiting seagrass beds were sharks,

which may or may not remain over the grass flats. Turtles do not spend

all of their time over grass flats, but often move into barren rocky areas

or reefs to rest. Therefore, not all of the rich turtle feces are

deposited in the seagrass beds. The reproductive migrations that adult

males and females make to their usually distant breeding grounds may

have represented a significant transport of energy and nutrients when

green turtle populations were high. As green turtles require high wave

energy beaches for nesting, and Thalassia grows in low wave energy areas,

the breeding and feeding grounds of green turtles are necessarily separate.

Prior to the arrival of European man, the majority of adult turtles

probably died at, or on their way to, the nesting beach, as this is the

period of greatest physiological stress. The nutrients in the many eggs

laid by each female turtle remain largely within the nesting beach and

near-offshore region due to unsuccessful clutches and predation on both

eggs and hatchlings. It is estimated that 1% of hatchlings reach adult-

hood; only a slightly higher percentage arrives at the seagrass feeding

grounds, as much the greater part of predation occurs while the turtles

are in their planktonic, "lost year," developmental stages.

The carrying capacity of Thalassia for the green turtle can be

estimated by combining standing crop and turnover rates for ThaZassia

from the literature and figures for grazing rates presented here. An

average standing crop of 250 g dry weight per square meter has been

measured at both Jamaica (Greenway, 1976) and Venezuela (Gessner, 1971).

Iverson and Bittaker (1978) reported an average of 10 turnovers per year

for tropical ThaZassia flats. Combining these figures, an average of

2.5 kg dry weight of Thalassia blades are produced per square meter

each year. If this value is used directly with the average consumption

rate per year for an adult green turtle, the resulting carrying capacity

of one hectare of Thalassia is 250 adult green turtles.

A better estimate can be made if the effect of grazing on blade

production is included. Cropping stimulates blade growth, but continual

recropping leads to a decrease in growth rate, presumably as rhizome

stores decrease (Greenway, 1974). Working with Thalassia in Jamaica,

Greenway found a blade biomass turnover rate of 8.8 per year in ungrazed

stands (Greenway, 1976) and 5.5 turnovers per year in stands cropped to

2.5 cm above the leaf base (Greenway, 1974), the average cropping height

of green turtles. Substituting the 5.5 turnover rate for the value used

in the preceding calculations, the carrying capacity of one hectare of

Thalassia drops to 138 adult green turtles, or 1 turtle per 72 square

meters, still a very impressive figure. This estimate of the carrying

capacity would be improved if we had an estimate of the age distribution

of green turtles and, using the feeding rates from Figure 8, calculated

a carrying capacity based on all size classes.

If the Seagrass Ecosystem Study Group is successful in its attempt

to chart and quantify the extent of seagrass beds in the Caribbean, the

carrying capacity of the Caribbean for the green turtle, and thus the

maximum potential population for the green turtle under natural con-

ditions, can be estimated.

In comparing the digestive efficiency of seagrass herbivores, the

green turtle and dugong (and, probably, the manatee) are comparable in

terms of fiber digestion, as already indicated. The invertebrate sea-

grass consumers apparently lack the digestive enzymes necessary to break

down the structural carbohydrates of seagrasses (Lawrence, 1975).

Although the digestive system of the brant, Branta bernicZa, has not

been studied, cellulose is not digested to any significant degree by the

goose Anser anser (Mattocks, 1971). Digestive studies are also lacking

for fish that graze on ThaZassia. Ogden (in press) states that parrot-

fish void their gut contents several times each day, indicating a low

digestive efficiency and lack of an active cellulolytic gut microflora.

For organic matter ADC's, the only other parameter for which there

is information with which to compare the green turtle data, Lowe (1974)

found an ADC of 19 + 7% for the urchin, Lytechinus variegatus, while

Moore and McPherson (1965) report 52-57% ADC's for both L. variegatus

and Tripneustes esculentus. The urchins fed on Thalassia in the above

studies. Fuji (1962) measured an organic matter ADC of 32% for the sea

urchin Strongylocentrotus intermedius feeding on the seagrass

PhyZZospadix awatensis. Murray et at. (1977) calculated an organic

matter ADC of 84% for the dugong which they dissected. The green turtle

organic matter ADC's (Figure 1) for the 48 kg and 66 kg size classes

average 66% while the two AZpha Helix turtles gave values of 65% and 77%

(Table 15). To judge from data from one animal, the dugong appears to

have a greater organic matter digestibility than the green turtle and

the sea urchins. Depending on the species of urchin and on the study,

green turtles range between being much more efficient (19% vs. 66%) or

slightly more efficient (57% vs. 66%) than sea urchins.

Algae as an Alternative Diet

All of the above seagrass consumers are capable of feeding and

existing on algae. In urchins and herbivorous marine fish, differential

feeding on seagrasses and algae seems to be directly related to avail-

ability (Ogden, 1976). In brant the original change from feeding mainly

on seagrasses to feeding on algae was caused by the disappearance of

Zostera marina in the 1930's (Cottam et al., 1944). Now that Zostera

has returned to much of its former range, feeding choice in brant also

seems to be regulated by relative abundances of Zostera and Ulva lactuca

(Penkala, 1975). Heinsohn and Birch (1972) in a detailed study of

dugong feeding habits concluded that algae, although present in the

feeding area, are only taken rarely and apparently unintentionally.

However, Spain and Heinsohn (1973) found that following a hurricane that

disrupted the dugong's seagrass feeding habitat, there was a significant

increase in the quantities of algae ingested. Lipkin (1975), working in

the Red Sea, also found that dugongs fed almost entirely on seagrasses,

although, again, algae were available. In the six animals he examined,

only one had ingested more than a trace of algae. This specimen had

approximately 1% by volume of three Caulerpa species in its large

intestine and trace amounts in its cecum and small intestine. Green

turtles consume algae in significant quantities only when seagrasses

are not available at levels necessary to sustain grazing. Mortimer

(1976) found only 2.54% (by dry weight) of algae in the 202 green turtle

stomachs she examined from the Miskito Cays, Nicaragua. In my study

area, the percentage of algae in the standing crop biomass was much

greater than the percentage of algae in the turtles' diet.

A pattern emerges, then, of green turtles and dugongs specializing

on either algae or seagrasses, even in the presence of the alternate

food source, while brant, sea urchins and fish select food largely on

the .basis of availability. That is, those herbivores that have an

active fermentive gut microflora are specialists in their feeding

habits and those without are generalists. I believe this is due to the

selectivity forced on them by their gut microbes.

Gut microflora are dynamic systems, capable of changing and adjust-

ing to different diets. Rumen microflora have been shown to vary not

only in relative population proportions but also in numbers of species

with moderate changes in diet (Hungate, 1966; Church, 1971). For

example, a deer can browse on many species and, because the structural

carbohydrate (cellulose) is the same for all species, only minor

variations in rumen microflora would result. Cellulose, the major

structural carbohydrate in seagrasses, is present only in very small

amounts in algae (Percival, 1964). Most algae contain complex structural

carbohydrates such as glucan, mannan, xylan, agar, carrageenan, alginic

acid and uronic acid in place of cellulose (Chapman and Chapman, 1973).

It follows that if algae-eating green turtles have an active gut

fermentation, the cecal microflora of a green turtle that feeds on

algae would be significantly different from one that feeds on sea-

grasses. In dugongs and green turtles, a change from a diet of seagrasses

to algae would require radical changes in gut microflora, from cellulase-

secreting microbes to microbes that secrete enzymes capable of breaking

down the structural carbohydrates in algae. I believe it is this

specificity of cecal microflora that causes the green turtle and dugong

to specialize. For an animal to change from one diet to another, or to

regularly consume both seagrasses and algae, would result in lowered

efficiency. I am not suggesting that a green turtle with a gut micro-

flora adapted to seagrasses would starve on a diet of algae, only that

such a turtle would have a lower digestive efficiency and would absorb

only the cell contents of cells broken by mastication until its gut

microflora adjusted to the diet change.

There is further evidence in support of the above theory. The few

times that my study animals ingested a small amount of algae (Sargassum

or Batophora), it passed through their intestine apparently unaltered, in

striking contrast to the unrecognizable mush of Thalassia blades. Lipkin

(1975) observed the same phenomenon in the dugong that had ingested

Caulerpa. He stated that the "algal material was not digested at

all even in the large intestine Even the cytoplasm squeezed out of

the CauZerpa fragments coming from the large intestine looked unaffected by

digestive fluids when examined with a microscope" (p. 94). He had earlier

described the very well-digested appearance of seagrasses in this region

of the gut. In addition, Felger and Moser (1973) reported that the Seri

Indians recognize two kinds of green turtles--"sweet" turtles that feed

on eelgrass on the east coast of Tiburon Island in the Gulf of California

and "stinking" turtles that feed on algae on the west coast of Tiburon

Island. The distance separating these two groups is within swimming

distance for green turtles. For the diet to be consistent enough to

flavor their flesh, the turtles must be maintaining relatively long-term

diet specificity; this specificity may be controlling their local dis-

tribution. Mortimer (unpublished) has reported a similar disparity

between algae-consuming turtles near Set Net Point, Nicaragua, which are

considered to have an inferior flavor by Miskito housewives, and ThaZassia-

feeding turtles from Miskito Bank.

There is another interesting aspect of algae consumption in green

turtles. Within the last few years there has been a series of papers on

the effect of herbivory on algae species diversification and relative

abundance (reviewed by Lubchenco, 1978). As mentioned earlier, in some

areas green turtles feed entirely on algae. Also, green turtles,

particularly young individuals possibly in the midst of their develop-

mental migrations, are often seen in the Caribbean, swimming and feeding

over reefs (personal observations and Archie Carr, pers. comm.). While

in this habitat, they can be seen feeding on algae. Certainly, if green

turtles ever approached the carrying capacity of Thalassia as Randall

(1965) has suggested, algae were consumed more frequently than today.

Frequently, however, in discussions of possible algae-herbivore co-evolu-

tion the once dominant influence of the green turtle is ignored (Ogden,

1976; Ogden and Lobel, 1978). Ogden (1976) stated that species in the

genera Cauterpa, Dictyota, Laurencia, Penicillus and Turbinaria remain

lush in the presence of herbivores (fish and urchins) and that some of

these are also rejected by parrotfish and urchins in preference tests.

From these data, he suggested that these algal species have evolved

chemical or structural anti-herbivore defenses. These genera, however,

are repeatedly included in lists of stomach contents of green turtles

(Balazs, 1977; Ferreira, 1968; Frazier, 1971; Hughes, 1974; Mortimer,

1976; Pritchard, 1971), sometimes as the only species present in the

stomach. Greenway (1976) found that the urchin Lytechinus variegatus

consumed Thalassia at a rate of 1 g dry weight per urchin per week. A

small (30 kg) green turtle, a size commonly seen in reef system

habitats (usually gorgonian patches between reef sections), consumes

an average of 574 g dry weight per turtle per week (from Figure 8).

That Is, one turtle consumes roughly as much as 574 L. variegatus

consume. Certainly, sea urchins are important grazers in the Caribbean,

as are herbivorous fishes (for which, unfortunately, there are no feeding

rate estimates), but, when dealing with evolutionary questions, green

turtles must be considered. Certainly, during the time algae and

herbivores were "co-evolving" in the Caribbean, the influence of green

turtle grazing was strong.

When the present diversity and relative frequencies of algae

species are considered, the influence of the green turtle is minor in

most habitats. However, when species diversity and relative abundances

are considered as an end product of plant-herbivore interactions over

evolutionary time, the answers are not as simple. The complex of

interactions must be evaluated as they were, not as they are--particularly

when the controlling factors in the past and present differ as markedly

as do the grazing patterns of sea urchins and green turtles. Tropical

seagrass and reef ecologists, then, are in a difficult position when

they attempt to evaluate their present ecosystem in terms of evolutionary

time, because the major, herbivore is gone. It is as if the Serengeti were

cleared of its ungulates, leaving the grasslands ecologists to study

energy and nutrient cycling through crickets and hares. Undoubtedly

crickets and hares are important, but who would suggest that they have

provided the major selective pressures in forming the Serengeti?


1. The green turtle is a selective grazer, recropping plots of Thalassia

testudinum blades and thus maintaining a low lignin, high protein

food source.

2. Low water temperatures decrease digestive efficiency; all size classes

of turtles were affected to the same degree.

3. Cellulose is digested efficiently by the colon microflora in the

green turtle. There was no significant difference in percent

cellulose digested among the size classes.

4. The highly variable apparent digestibilities of hemicellulose were

possibly an artifact of the low hemicellulose concentrations in


5. The energy and organic matter digestibilities are quite similar.

The 48 kg and 66 kg size classes digest significantly more than the

30 kg group, which in turn digests significantly more than the 8 kg


6. Only the 8 kg size.class has significantly different protein apparent

digestibilities. Because of gut secretions, the nitrogen data give

a poor indication of what becomes of dietary protein.

7. In the cecum, VFA's are produced in the following proportions:

acetate > butyrate > propionate. The probable rates of absorption

are: butyrate > propionate > acetate.

8. The VFA's produced in the cecum provide approximately 14.2% of a

green turtle's daily energy budget. This percentage would be

higher if the VFA's produced in the colon were included.


9. Hydrogen is the major gas evolved by the cecum fermentation.

10. The consumption rates of ThaZassia blades were estimated for the

8, 30, 48 and 66 kg green turtle size classes to be 8.8, 29.9,

64.6 and 79.6 kg dry weight per year, respectively.

11. The carrying capacity of Thalassia testudinwn for the green turtle

was calculated to be 138 adult green turtles per hectare.

12. The gut microflora of the green turtle may be responsible for its

feeding selectively on either algae or seagrasses.

13. Prior to man's destruction of green turtle populations, the green

turtle played a major role in nutrient and energy cycling in the

seagrass ecosystem. The past importance of the green turtle is

often forgotten in discussions of seagrass ecology.


Altig, R. and W. McDearman. 1975. Percent assimilation and clearance
times of five anuran tadpoles. Herpetologica 31: 67-69.

Balazs, G. H. 1977. Ecological aspects of green turtles at Necker
Island. A report based on an expedition to the Hawaiian Islands
National Wildlife Refuge, Northwestern Hawaiian Archipelago.
Mimeo 27 pp.

Beattie, J. 1926. The ileo-caecal region in the reptilia--1. The
ileo-caecal region of Tupinambis teguexin. Proc. Zool. Soc. Lond.
28: 931-939.

Bertram, G. C. L. and C. K. R. Bertram. 1968. Bionomics of dugongs and
manatees. Nature 218: 423-426.

Bjorndal, K. A. In press. Urine concentrations of ammonia, urea and
uric acid in the green turtle, Chelonia mydas. Comp. Biochem.

Carr, W. E. S. and C. A. Adams. 1973. Food habits of juvenile marine
fishes occupying seagrass beds in the estuarine zone near Crystal
River, Florida. Trans. Am. Fish. Soc. 102(3): 511-540.

Chapman, V. J. and D. J. Chapman. 1973. The Algae. Second ed.
Macmillan Press, London.

Church, D. C. 1971. Digestive Physiology and Nutrition of Ruminants.
0. S. U. Book Stores, Corvalis, Oregon

Conover, R. J. 1966. Assimilation of organic matter by zooplankton.
Limnol. Oceanog. 11: 338-345.

Cottam, C., J. Lynch and A. I. Nelson. 1944. Food habits and management
of the sea brant. J. Wildl. Manag. 8: 36-46.

Cotzias, G. C. 1962. Manganese. In C. L. Comar and F. Bronner (eds.)
Mineral Metabolism Vol. 2, Part B. Academic Press, New York.

Felger, R. and M. B. Moser. 1973. Eelgrass (Zostera marina) in the
Gulf of California: Discovery of its nutritional value by the
Seri Indians. Science 181: 355-356.

Fenchel, T. 1970. Studies on the decomposition of organic detritus
derived from the turtle grass, Thalassia testudinum. Limnol.
Oceanogr. 15(1): 14-20.

Fenchel, T. 1977. Aspects of the decomposition of seagrasses. In:
C. P. McRoy and C. Helfferich (eds.) Seagrass Ecosystems: A
Scientific Perspective. Marcel Dekker, Inc., New York.

Ferreira, M. M. 1968. Sobre a alimentacao da aruana, Chelonia mydas,
ao long da costa do estado do Ceara. Arq. Est. Biol. Mar. Univ.
Fed. Ceara 8(1): 83-86.

Field, C. R. 1976. Palatability factors and nutrient values of the food
of buffaloes (Syncerus caffer) in Uganda. E. Afr. Wildl. J. 14:

Frazier, J. 1971. Observations on sea turtles at Aldabro Atoll. Phil.
Trans. Roy. Soc. Lond. B 260: 373-410.

Fuji, A. 1962. Studies on the biology of the urchin. V. Food con-
sumption of Strongylocentrotus intermedius. Jap. J. Ecol. 12(5):

Gasaway, W. C. 1976a. Seasonal variation in diet, volatile fatty acid
production and size of the cecum of rock ptarmigan. Comp. Biochem.
Physiol. 53A: 109-114.

Gasaway, W. C. 1976b. Volatile fatty acids and metabolizable energy
derived from cecal fermentation in the willow ptarmigan. Comp.
Biochem. Physiol. 53A: 115-121.

Gasaway, W. C. 1976c. Methane production in rock ptarmigan (Lagopus
mutus). Comp. Biochem. Physiol. 54A: 183-185.

Gessner, F. 1971. The water economy of the seagrass Thalassia testudinum.
Mar. Biol. 10: 258-260.

Greenway, M. 1974. The effects of cropping on the growth of Thalassia
testudinum in Jamaica. Aquaculture 4: 199-206.

Greenway, M. 1976. The grazing of Thalassia testudinum in Kingston
Harbour, Jamaica. Aquatic Bot. 2: 117-126.

Harlow, H. J., S. S. Hillman and M. Hoffman. 1976. The effect of
temperature on digestive efficiency in the herbivorous lizard,
Dipsosaurus dorsalis. J. Comp. Physiol. 11l: 1-6.

Hartog, C. den. 1970. The Seagrasses of the World. North-Holland
Publ. Co., Amsterdam.

Heinsohn, G. E. and W. R. Birch. 1972. Foods and feeding habits of the
dugong, Dugong dugon, in Northern Queensland, Australia. Mammalia
36: 414-422.

Heinsohn, G. E., J. Wake, H. Marsh and A. V. Spain. 1977. The dugong
(Dugong dugon) in the seagrass system. Aquaculture 12: 235-248.

Hirth, H. F. 1962. Cloacal temperatures of the green and hawksbill
sea turtles. Copeia 1962: 647-648.

Hughes, G. R. 1974. The sea turtles of South-East Africa. Vol. 2. The
biology of the Tongaland loggerhead turtles Caretta caretta with
comments on the leatherback turtle DermocheZys coriacea and the
green turtle Chelonia mydas in the study region. S. African Assoc.
for Mar. Biol. Res. Durban, South Africa.

Hungate, R. E. 1966. The Rumen and its Microbes. Academic Press,
New York.

Iverson, J. B. In press. Colic modifications in iguanine lizards.
J. Morph.

Iverson, R. L. and H. F. Bittaker. 1978. Patterns of tropical seagrass
productivity. Presented at the Conference on Ecological Processes
in Coastal and Marine Systems. April 13-15, Tallahassee, Fl.

Johnsgard, P. A. 1975. Waterfowl of North America. Indiana Univ.
Press, Bloomington.

Johnson, D. R. and M. H. Maxell. 1966. Energy dynamics of Colorado pikas.
Ecology 47: 1059-1061.

Johnson, J. L. and R. H. McBee. 1967. The porcupine cecal fermentation.
J. Nutr. 91(4): 540-546.

Kenchington, R. A. 1972. Observations of the digestive system of the
dugong, Dugong dugon. J. Mammal. 53: 884-887.

Keys, Jr., K. E., P. J. Van Soest and E. P. Young. 1969. Comparative
study of the digestibility of forage cellulose and hemicellulose
in ruminants and nonruminants. J. Anim. Sci. 29: 11-15.

Kikuchi, T. and J. M. Peres. 1977. Consumer ecology of seagrass beds.
In: C. P. McRoy and C. Helfferich (eds.) Seagrass Ecosystems: A
Scientific Perspective. Marcel Dekker, Inc., New York.

Lawrence, J. M. 1975. On the relationships between marine plants and
sea urchins, p. 213-286. In: H. Barnes (ed.) Oceanogr. Mar. Biol.
Ann. Rev. 13. Geo. Allen and Unwin Ltd., London.

Lewis, C. B. 1940. The Cayman Islands and marine turtle. Bull. of the
Inst. of Jamaica Sci. Series, No. 2: 56-65.

Lipkin, Y. 1975. Food of the Red Sea Dugong from Sinai. Israel J.
of Zool. 24: 81-98.

Lomolino, M. V. 1977. The ecological role of the Florida manatee
(Trichechus manatus latirostris) in water hyacinth-dominated
ecosystems. Master's Thesis, Univ. of Fla.

Lonnberg, E. 1902. On some points of relation between the morphological
structure of the intestine and the diet of reptiles. Bihang. Till K.
Svenska Vet.-Akad. Handlingar. Band 28(IV) No. 8: 3-53.

Lowe, E. F. 1974. Absorption efficiencies, feeding rates and food
preferences of Lytechinus variegatus for various marine plants.
Master's Thesis, Univ. South Fla., Tampa.

Lubchenco, J. 1978. Plant species diversity in a marine intertidal
community: Importance of herbivore food preference and algal
competitive abilities. Am. Nat. 112(983): 23-39.

Marsh, H., G. E. Heinsohn and A. V. Spain. 1977. The stomach and duodenal
diverticula of the dugong, Dugong dugon. In: R. J. Harrison (ed.)
Functional Anatomy of Marine Mammals. Vol. 3. Academic Press,

Mattocks, J. G. 1971. Goose feeding and cellulose digestion. Wildfowl
22: 102-113.

McBee, R. H. and G; C. West. 1969. Cecal fermentation in the willow
ptarmigan. Condor 71: 54-58.

McMahan, C. A. 1970. Food habits of ducks wintering on Laguna Madre,
Texas. J. Wildl. Manag. 34(4): 946-949.

McNaughton, S. J. 1976. Serengeti migratory wildebeest: Facilitation of
energy flow by grazing. Science 191: 92-94.

Moir, R. J. 1968. Ruminant digestion and evolution. In C. F. Code
(ed.) Handbook of Physiology. Sec. 6, Vol. V. American Physiological
Society, Washington, D.C.

Moore, H. B., T. Jutare, J. C. Bauer and J. A. Jones, 1963a. The
biology of Lytechinus variegatus. Bull. Mar. Sci. Gulf Carib.
13: 23-53.

Moore, H. B., T. Jutare, J. A. Jones, B. F. McPherson and C. F. E. Roper.
1963b. A contribution to the biology of Tripneustes esculentus.
Bull. Mar. Sci. Gulf Carib. 13: 267-281.

Moore, H. B. and B. F. McPherson. 1965. A contribution to the study
of the productivity of the urchins Tripneustes esculentus
Lytechinus variegatus. Bull. Mar. Sci. Gulf Carib. 15(4): 855-871.

Morison, S. E. 1942. Adniral of the Ocean Sea, a Life of Christopher
Columbus. Little, Brown and Co., Boston.

Mortimer, J. A. 1976. Observations on the feeding ecology of the green
turtle, Chelonia mudas, in the Western Caribbean. Master's Thesis,
Univ. of Fla.

Moss, R. 1977. The digestion of heather by red grouse during the spring.
Condor 79: 471-477.

Moss, R. and G. R. Miller. 1976. Production, dieback and grazing of
heather (Calluna vulgaris) in relation to numbers of red grouse
(Lagopus 1. scoticus) and mountain hares (Lepus timidus) in North-
east Scotland. J. Appl. Ecol. 13: 369-377.

Mrosovsky, N. and P. C. H. Pritchard. 1971. Body temperatures of
Dermochelys coriacea and other sea turtles. Copeia 1971: 624-631.

Murray, R. M., H. Marsh, G. E. Heinsohn and A. V. Spain. 1977. The
role of the midgut caecum and large intestine in the digestion of
sea grasses by the dugong. Comp. Biochem. Physiol. 56A: 7-10.

Nagy, K. A. 1977. Cellulose digestion and nutrient assimilation in
Sauromalus obesus, a plant-eating lizard. Copeia 1977: 355-362.

Nagy, K. A. and V. H. Shoemaker. 1975. Energy and nitrogen budgets of
the free-living desert lizard Sauromalus obesus. Physiol. Zool.
48(3): 252-262.

National Academy of Sciences. 1976. Making Aquatic Weeds Useful: Some
Perspectives for Developing Countries. National Academy of
Sciences, Washington, D.C.

Odum, E. P. 1971. Fundamentals of Ecology. 3rd ed. W. B. Saunders,
Philadelphia. 574 pp.

Ogden, J. C. 1976. Some aspects of herbivore-plant relationships on
Caribbean reefs and seagrass beds. Aquatic Bot. 2: 103-116.

Ogden, J. C. In press. Faunal relationships in Caribbean seagrass beds.
In: R. C. Phillips (ed.) Seagrass FLosystems. Marcel Dekker, New York.

Ogden, J. C. and P. S. Lobel. 1978. The role of herbivorous fishes and
urchins in coral reef communities. Env. Biol. Fish. 3(1): 49-63.

Osman-Hill, W. C. 1945. Notes on the dissection of two dugongs. J.
Mammal. 26: 153-175.

Ostrom, J. H. 1963. Further comments on herbivorous lizards.- Evolution
17: 368-369.

Parr Instrument Co. 1960. Oxygen bomb calorimetry and combustion methods.
Technical Manual No. 130: 1-56.

Parsons, J. J. 1962. The Green Turtle and Man. Univ. of Fla. Press,
Gainesville, Fla.

Penkala, J. M. 1975. Winter food habits and body weight of Atlantic
brant. Mimeo 9 pp.

Percival, E. 1964. Algal polysaccharides and their biological relation-
ships, p. 18-35. In: D. de Virville and J. Feldman (eds.) Proc. IV
International Seaweed Symposium. Pergamon Press, London.

Pough, F. H. 1973. Lizard energetic and diet. Ecology 54(4): 837-844.

Prim, P. and J. M. Lawrence. 1975. Utilization of marine plants and
their constituents by bacteria isolated from the gut of echinoids.
Mar. Biol. 33: 167-173.

Pritchard, P. C. H. 1971. Galapagos sea turtles--preliminary findings.
J. Herp. 5: 1-9.

Randall, J. E. 1965. Grazing effect on sea grasses by herbivorous reef
fishes in the West Indies. Ecology 46(3): 255-260.

Randall, J. E. 1967. Food habits of reef fishes of the West Indies.
Stud. Trop. Oceanogr. No. 5: 665-847.

Robertson, J. B. and P. J. Van Soest. 1977. Dietary fiber estimation
in concentration feedstuffs. Presented at 69th meeting of the
American Society of Animal Science, Univ. of Wisc., Madison.
July 23-27, 1977.

Sokol, 0. M. 1967. Herbivory in lizards. Evolution 21: 192-194.

Spain, A. V.and G. E. Heinsohn. 1973. Cyclone associated feeding changes
in the dugong. Mammalia 37: 678-680.

Szarski, H. 1962. Some remarks on herbivorous lizards. Evolution 16: 529.

Thayer, G. W., K. A. Bjorndal and M. W. LaCroix. In prep. Aspects of
the nutrition and role of the green turtle, Chelonia mydas, in the
ecology of seagrass beds.

Throckmorton, G. 1973. Digestive efficiency in the herbivorous lizard
Ctenosaura pectinata. Copeia 1973(3): 431-435.

Van Dyne, G. M. 1968. Measuring quantity and quality of the diet of
large herbivores. In: F. B. Golley and H. K. Buechner (eds.)
A Practical Guide to the Study of the Productivity of Large
Herbivores. Blackwell Scientific Publications, Oxford.

Van Soest, P. J. 1963. Use of detergents in the analysis of fibrous
feeds 11. A rapid method for determination of fiber and lignin.
J. Assoc. Off. Agr. Chem. 46: 829.


Van Soest, P. J. and R. H. Wine. 1967. Use of detergents in the analysis
of fibrous feeds IV. Determination of plant cell-wall constituents.
J. Assoc. Off. Agr. Chem. 50: 53.

Westlake, D. F. 1963. Comparisons of plant productivity. Biol. Rev.
38: 385-425.

Wilson, K. J. and A. K. Lee. 1974. Energy expenditure of a large
herbivorous lizard. Copeia 1974: 338-348.


Karen Anne Bjorndal was born 22 February 1951 in Berkeley,

California. She attended Miramonte High School in Orinda, California,

until June, 1968. She was graduated with honors from Occidental College,

Los Angeles, California,with a Bachelorsdegree in Biology in June, 1972.

In September, 1973, she entered the Department of Zoology of the

University of Florida where she completed the requirements for the

Doctor of Philosophy major in zoology.

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

Archie Carr, Chairman
Graduate Research Professor of Zoology

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

Jhh H. Kaufmann
P fessor of Zoology

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

Sohn E. Moore
Professor of Animal Sciences

I certify that I have read this study and that in my opinion it
conforms to acceptable standards of scholarly presentation and is fully
adequate, in scope and quality, as a dissertation for the degree of
Doctor of Philosophy.

Hu'gh L. P/penoe
Profess of Geography

This dissertation was submitted to the Graduate Faculty of the Department
of Zoology in the College of Liberal Arts and Sciences and to the Graduate
Council, and was accepted as partial fulfillment of the requirements of
the degree of Doctor of Philosophy.

June, 1979

Dean, Graduate School


3 1262 08285 400 01l
3 1262 08285 400 0