Nutritional ecology and digestive physiology of the hoatzin, Opisthocomus hoazin, a folivorous bird with foregut fermentation

MISSING IMAGE

Material Information

Title:
Nutritional ecology and digestive physiology of the hoatzin, Opisthocomus hoazin, a folivorous bird with foregut fermentation
Physical Description:
x, 109 leaves : ill. ; 28 cm.
Language:
English
Creator:
Grajal, Alejandro, 1957-
Publication Date:

Subjects

Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1991.
Bibliography:
Includes bibliographical references (leaves 94-106).
Statement of Responsibility:
by Alejandro Grajal.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001693363
notis - AJA5442
oclc - 25222069
System ID:
AA00003722:00001

Full Text










NUTRITIONAL ECOLOGY AND DIGESTIVE PHYSIOLOGY OF THE HOATZIN,
OPISTHOCOMUS HOAZIN. A FOLIVOROUS BIRD WITH FOREGUT FERMENTATION














BY


ALEJANDRO GRAJAL


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY



UNIVERSITY OF FLORIDA


1991


























Copyright 1991

by

Alejandro Grajal






























to my parents, Alejandro and Carmina.













ACKNOWLEDGEMENTS


I would like to express my gratitude to Stuart D. Strahl and Rodrigo Parra

for their initial guidance with the project. Special thanks go to Ornella Parra

and the staff of the Laboratory of the Animal Production Institute, Universidad

Central de Venezuela at Maracay for their enthusiasm, patience, and sample

analyses. Thanks go to Helena Puche, Gerry Casaday, Gustavo Hernandez,

Daniel Carrillo, Gregg Andraso, Jaime Aranguren, Robert Eddington, Ann C.

Wilkie, and Gilman S. Marshall for their assistance.


Thanks go to the Otero family for their hospitality at Maracay, to Tomas

Blohm for his hospitality and logistic support at Fundo Pecuario Masagural and

to Antonio Branger for providing housing and logistic support at Hato Pifiero.

Thanks go to Fabian Michelangeli, Maria G. Dominguez and Marie C. Landais of

the CBB-IVIC, Venezuela, for their field logistic support at Hato Pifiero, to

Alexis Arends, Angela Martin6, and Clara Alarc6n for their hospitality at Coro.

Hoatzins were collected under permission of the Venezuelan Environment

Ministry, M.A.R.N.R. Thanks go to the people of the Department of Zoology,

University of Florida, for their support and to Richard A. Kiltie for the use of

computer video equipment.


This study was funded by grants of The Nixon Griffis Fund of the New York

Zoological Society, The Chicago Zoological Society, The Alexander Wetmore

Memorial Fund of the American Ornithologists Union (AOU), a Sigma-Xi Grant-

in-aid, a scholarship by the Organization of American States (OAS), the









Venezuelan National Council for Science and Technology CONICIT and a

research assistantship of the Department of Zoology, University of Florida.


Karen A. Bjorndal, Douglas J. Levey, Brian K. McNab, Richard A. Kiltie, John

F. Anderson and Kent H. Redford served on the advisory committee and

provided helpful comments and criticism. Finally, special thanks go to my

wife, Helena, for her patience, friendship and support at all times.












TABLE OF CONTENTS



ACKNOWLEDGEMENTS ................................................ i v

TABLE OF CONTENTS ................................................. v i

KEY TO ABBREVIATIONS ............................................. viii

ABSTRACT ......................................................... ix

CHAPTERS

1. GENERAL INTRODUCTION ........................................ 1


2. STRUCTURE AND FUNCTION OF THE DIGESTIVE TRACT OF THE HOATZIN. 7

Introduction ............................................. 7
Materials and Methods .................................... 9
Results .................... ............................. 11
Digestive tract morphology ........................... 11
Gut contents ........................................ 14
Particle dynamics ................................... 14
D discussion ............... ............................... 15
Digestive morphology ................................ 15
Gut contents ......................................... 16
Particle dynamics .................................... 17
Conclusions ............................................. 18

3. DIGESTION EFFICIENCY OF THE HOATZIN .......................... 27

Introduction ............................................. 27
M materials and Methods ..................................... 29
Animal husbandry .................................... 29
Diet composition ...................................... 30
Intake and digestibility ................................ 31
Results ....................... ......... .................. 33
Intake and digestibility ............................... 33
D discussion ................ ............................... 35









4. RETENTION TIMES AND PARTICLE PASSAGE RATES OF DIGESTA
MARKERS IN THE HOATZIN GUT ................................. 44

Introduction ............................................. 44
Materials and Methods .................................... 46
Results ........ ...... ..... ...... ... ............ ....... 49
Fecal excretion rates ................................. 49
M arker recovery ..................................... 49
Mean retention times ................................. 49
Transit times .................. ......... ............. 50
Discussion .......................... ..................... 51

5. FERMENTATION RATE IN THE CROP AND ESOPHAGUS OF THE HOATZIN.. 57

Introduction ............................................. 57
Materials and Methods .................................... 59
Laboratory study: Fermentation in captive hoatzins .. .59
Field study: Fermentation in wild hoatzins .............. 62
Comparison of hoatzin and cow in vitro digestibilities .. 63
Comparison of the miniature and standard in vitro
techniques ............... ............... .......... 64
Gas-liquid chromatography .......................... 65
Statistical analyses .................................... 65
Results ................ ................................ 66
VFA production rate in captive and wild hoatzins ..... 66
In itro fiber fermentation and comparisons with cow
ruminal fermentation ................................. 67
Comparison of the miniature and standard in vitro ....... 67
D discussion ................. ............................. 68
VFA production rate in captive and wild hoatzins ......... 68
In vitro fiber fermentation ........................... 69
Miniature and standard in vitro techniques ............. 70
Conclusions ............... ............................. 70

6. RATE OF METABOLISM IN THE HOATZIN .......................... 77

Introduction ............................................ 77
Materials and Methods .................................... 79
Results ................................................... 81
Discussion ................. .............................. 81

7. GENERAL DISCUSSION ....................................... 86

Evolution of Foregut Fermentation .......................... 86
Herbivory in Birds .......................................... 88
Suggestions for Further Studies ............................ 92
Coda ................. ................................. 93

LIST OF REFERENCES ........ ............................... ...... 94

BIOGRAPHICAL SKETCH ................... ....................... 107















KEY TO SYMBOLS OR ABBREVIATIONS


ADF Acid Detergent Fiber
ANOVA Analysis of Variance
BRM Basal Rate of Metabolism
Cm Thermal Conductance
Cr chromium
CS Concentrate Selectors
CW Cell Wall
d.f. degrees of freedom
DM Dry Matter
EDTA Chromium Ethylene-Diamine Tetra Acetic acid
g grams
gDM grams Dry Matter
GR Grass and Roughage eaters
h hour
IF Intermediate Feeders
IVOMD In Vitro Organic Matter Digestibility
KJ Kilojoule
kg kilogram
1 liter
MEC Metabolizable Energy Coefficient
mg milligram
min minute
ml milliliter
mm2 Square millimeter
mmol millimol
MPS Mean Particle Size
NDF Neutral Detergent Fiber
CM Organic Matter
pers. obs. personal observation
pers. comm. personal communication
Ta Ambient Temperature
Tb Body Temperature
TT Transit Time
VFA Volatile Fatty Acids
V02 Rate of Oxygen Consumption
Yb ytterbium
OC degrees centigrade


viii















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


NUTRITIONAL ECOLOGY AND DIGESTIVE PHYSIOLOGY OF THE HOATZIN,
OPISTHOCOMUS HOAZIN. A FOLIVOROUS BIRD WITH FOREGUT FERMENTATION

By

Alejandro Grajal

August, 1991



Chairman: Karen A. Bjorndal
Major Department: Zoology


The hoatzin is the only known obligate folivorous bird with a well-

developed foregut fermentation system. Most fermentation takes place at the

crop and caudal esophagus, where pH and volatile fatty acid (VFA) levels are

similar to those of foregut fermenting mammals. Contents from fermentation

organs represent 77% of the total gut capacity and 10% of the adult body mass

(average 650g). Large particles are retained longer than small particles at the

anterior fermentation sites. Food particle size is reduced by microbial

fermentation and grinding by the keratinous interior lining of the muscular

crop. Dry matter digestibilities by captive hoatzins were high (70-80%). Fiber

digestibilities were higher than values previously reported for other avian

herbivores (35-71% neutral detergent fiber, 47-63% cellulose). Passage rates

of liquid and solid digesta were measured by giving a single pulse dose of a

liquid marker (Cr-EDTA) and three solid markers ytterbiumm mordanted to









fiber and cuts of plastic tape of 1 and 4mm2). Mean retention times in the

hoatzin are among the longest ever recorded for a bird (18h for liquid, 24h for

ytterbium, 34h for lmm2 and 44h for 4mm2 markers). Production rates of VFA

are the highest recorded for a bird, and provide energy for more than 60% of

the hoatzin's basal rate of metabolism. In vitro fiber digestibility was similar

between hoatzin and cow inoculum. The basal rate of metabolism in hoatzins is

low (70% of the expected value for endotherms and 43% of the expected for

nonpasserine birds). Body temperature is maintained at 38.50C at

environmental temperatures between 120C and 360C.

Some of the extreme adaptations in the hoatzin are more similar to those

of mammals with foregut fermentation than to those of a bird. Microbial

fermentation breaks down fiber and makes a significant contribution to the

energy balance of hoatzins. Additionally, long retention times and selective

particle retention enhance digestion of cell wall and cell contents. Other

nutritional benefits, such as detoxification of plant secondary compounds and

microbial synthesis of essential amino acids, may have been important in the

evolution of foregut fermentation in this unique bird.















CHAPTER 1
GENERAL INTRODUCTION



Foregut fermentation as a method to digest plant fiber has been

reported for mammals such as ruminants (Parra 1978, Van Soest 1982), colobid

monkeys (Bauchop and Martucci 1968), sloths (Bauchop 1978, Montgomery and

Sunquist 1978) and macropod marsupials (Dellow et al. 1983, Hume 1982, Hume

and Dellow 1980, Moir et al. 1956). Hoatzins are a unique among birds because

they are the only known bird with a foregut fermentation system.

Hoatzins are one of the few avian obligate folivores (leaf-eaters)

(Morton 1978). Less than 3% of extant bird species feed extensively on green

leaves or buds (Morton 1978). Although some birds can digest plant fiber, it is

generally of little nutritional value. The reasons for the rarity of avian

folivores are that leaves are difficult to digest, bulky, and usually have

defensive chemical compounds. These dietary characteristics can impose

serious constraints on the high energy demands of powered flight, by

increasing the weight to be carried and decreasing the rate of energy

extraction from leaves. Consequently, most birds that eat large amounts of

plant material maximize the rate of energy or nutrient uptake and minimize

the weight of the digesta by extracting the readily digestible cell contents and

quickly excreting the undigested bulk of the fiber or cell wall. Examples are

geese and ducks (Buchsbaum et al. 1986, Burton et al. 1979, Dawson et al. 1989,

Kingsford 1989, Marriot and Forbes 1970, Muztar et al. 1977), some birds within

the Galliformes (Inman 1973), Takahe (Notornis mantelli), and kakapo or owl

parrot (Strigops habroptilus) (Morton 1978). Some ratites, such as the ostrich









(Struthio camelus) (Mackie 1987, Withers 1983), and the emu (Dromaius

novaholliandeae) (Herd and Dawson 1984) can digest plant fiber. Several

species of the family Tetraonidae can ferment fiber in enlarged paired caeca

(Gasaway 1976a, Gasaway et al. 1976, Hill et al. 1968, Inman 1973, Moss 1973,

Moss 1977, Moss and Parkinson 1972, Moss and Trenholm 1987, Pulliainen et al.

1968, Suomalainen and Arhimo 1945).

Hoatzins are obligate folivores (Grajal et al. 1989). In the Venezuelan

Llanos, 78.3% of all observed foraging time (>17,000 minutes) was spent on new

leaves and shoots, 8.5% on mature leaves, 6.1% on flowers and 7.2% on fruits.

Green leaves comprise 86.7% of the hoatzin diet (as % observed foraging time)

(Grajal et al. 1989, Strahl and Parra 1985). Among all species included in the

diet, samples of portions of the plant that are eaten are significantly higher in

water content, nitrogen, and hemicellulose, and lower in total cell wall,

cellulose, and lignin than non-eaten portions of the same plants (Grajal et al.

1989). Although hoatzins forage on the leaves of more than 45 species of

plants, only a few plants make up the bulk of the diet. For example, two

species, Zanthoxylum culantrillo and Acacia articulata, comprise nearly 50% of

the observed overall diet in the Central Venezuelan Llanos (Strahl and Parra

1985). In contrast, the remainder of the diet is made up of over 40 plant

species, most of which account individually for less than 1% of the observed

diet. These include several legumes (Fabaceae) and many additional species in

over 20 other families.

Foregut fermentation in the hoatzin is achieved by some unique

morphological and physiological adaptations. Their digestive system is

unusual both in structure and function among birds. In fact, the hoatzin is

the only flying vertebrate with foregut fermentation. The voluminous crop

and caudal esophagus have become functional fermentation chambers,










analogous to those of mammalian foregut fermenters. The crop and esophagus

are situated in front of a greatly reduced sternal carina that leaves little area

for flight muscle attachment. Consequently, hoatzins are not powerful fliers,

preferring to hop from branch to branch (Grajal et al. 1989).

Foregut fermentation in hoatzins is a theoretical anomaly. Present

models predict a limit of 6-10 kg of body mass, below which foregut

fermentation cannot fulfill the nutritional needs of an endotherm (Demment

and Van Soest 1985, Parra 1978). At this mass, the predicted total metabolic

requirements surpass the rate of energy made available from plant fiber

fermentation. These models suggest that herbivores below this threshold

should be omnivores with hindgut fermentation. Thus, the small mass of the

hoatzin (650g) is an order of magnitude lower than the predicted minimum

body mass for a foregut fermenter.

Some of the advantages of foregut fermentation in mammals include an

extensive use of plant fiber and an efficient use of microbial byproducts and

nitrogen. Microbial byproducts such as volatile fatty acids (VFA) are directly

absorbed at the fermentation sites and readily used by the host as an energy

source (Blaxter 1962). Microbes also produce essential nutrients such as amino

acids and vitamins that are absorbed in the lower gut (Hungate 1966).

Furthermore, foregut fermentation can be an important way to synthesize

bacterial protein from non-protein nitrogen (e.g., ammonia and urea) (Nolan

and Leng 1972, Prins 1977, Van Soest 1982). Finally, foregut fermentation can

be an effective method to detoxify some plant secondary compounds before

they reach the absorptive tissues of the lower gut (Freeland and Janzen 1974,

Mackie 1987).

Foregut fermentation also has some important drawbacks. For example,

foregut fermentation can decrease the digestive efficiency of a small










herbivore, because gut flora establishes one or more trophic levels between

the host and the readily digestive fractions in the diet. These additional

trophic levels can significantly decrease the rate of energy extraction by the

host. Fermentation products from microbial metabolism, mainly heat, C02 and

methane, can cause a substantial loss of nutrients and energy for the host.

Thus, some of these nutritional costs can be especially detrimental in a small

endotherm with high energy and nutrient turnover rates (Demment and Van

Soest 1985, Parra 1978).

The study of the digestive system of the hoatzin can provide numerous

insights into the evolution of foregut fermentation in vertebrates and about

the evolutionary constraints of herbivory in birds. Therefore, the objectives

of this study were to

1) Analyze the structure and function of the gastrointestinal tract of

hoatzins,

2) Characterize the dynamics of their digestive process

3) Determine their digestive efficiency

4) Analyze the energetic balance of the hoatzin as a folivorous bird, and

5) Explore the possible selective forces of this unique digestive system

in a bird.

This study has been divided in individual chapters that explore these

objectives. A final general discussion provides some evolutionary and

ecological implications of foregut fermentation in hoatzins, digestive

strategies of avian herbivores and the general implications of the evolution of

foregut fermentation in vertebrates.

Hoatzins live in gallery forests, forest swamps and oxbow lakes of the

Orinoco and Amazon drainages, ranging locally from the Guianas and

Venezuela throughout Amazonian Brazil, Colombia, Ecuador, Peru and Bolivia.










Although their range is quite extensive, hoatzins have strict habitat

requirements, with narrow local distributions.

The peculiarities of the hoatzin's anatomy were the subject of many

descriptive studies in an attempt to establish the evolutionary affinities

between the hoatzin and other birds. As a consequence, the taxonomic

position of the hoatzin has been one of the most debated topics of avian

systematics (Banzhaf 1929, Beebe 1909, B6ker 1929, Brigham 1919, Cherrie 1909,

Gadow 1891, Garrod 1879, Huxley 1898, L'Hermenier 1837, Parker 1891, Perrin

1877, Pycraft 1895, Verheyden 1956). The presence of functional wing claws

in hoatzin chicks, the reduced sternal carina and poor flying abilities of the

hoatzin were regarded as the primitive characteristics of a "missing link"

between the first fossil birds such as Archaeopteryx and modem birds

(Brigham 1919, Huxley 1898, Parker 1891, Young 1888). Later, it was classified

within the Galliformes, or chicken-like birds (Huxley 1898). Modem

electrophoretic (Sibley and Ahlquist 1973), morphological (de Queiroz and Good

1988, Verheyden 1956) and DNA-DNA hybridization (Sibley et al. 1988) studies

have consistently classified the hoatzin as the only member of the family

Opisthocomidae, included in the order Cuculiformes and closely related to anis

(Crothophaga) and guira cuckoos (Guira).

Surprisingly, the gastrointestinal tract and nutritional ecology received

little attention until recently (Grajal et al. 1989). Some early authors attempted

to relate the large gut capacity to a folivorous diet (Beebe 1909, Boker 1929,

Gadow 1891, L'Hermenier 1837). Also, some authors described the smell of the

gut contents as that of fresh cow manure (Beebe 1909, Goeldi 1886, Young

1888). None of these authors, however, suggested foregut fermentation as the

primary function of the large gut capacity in the hoatzin.










Young hoatzins are fed regurgitated leaves. Their growth is

comparatively slow; they require up to 70-80 days after hatching to fly. During

this long growth period, the young can be vulnerable to predators. Thus,

young hoatzins display some unique predator-escape mechanisms. Wing claws

are actively used to climb branches and vines. Young birds can dive into

water and can readily swim underwater in case of imminent danger after the

first few days from hatching (Strahl 1988).

During the breeding season (May-Oct. in Venezuela), hoatzins live in

social groups that consist of a breeding pair and sometimes up to four helpers

at the nest. The social units defend small (200-1000 m2) multi-use territories.

The nests usually are on vegetation overhanging watercourses. The non-

breeding season coincides with the dry season in Venezuela, and during this

period hoatzins become gregarious, when flocks of up to 200 individuals move

to permanent water bodies with green vegetation (Ramo and Busto 1984, Strahl

1988, Strahl and Schmitz 1990).















CHAPTER 2
STRUCTURE AND FUNCTION OF THE DIGESTIVE TRACT OF THE HOATZIN



Introduction

The hoatzin, Opisthocomus hoazin. is a Neotropical folivorous bird that

inhabits oxbow lakes, flooded forests and swamps of the Guianas, Orinoco and

Amazon basins. Up to 87% of its diet consists of leaves (Grajal et al. 1989).

Obligate folivory is unusual in birds because leaves are bulky, have low

nutritional value and can have noxious chemicals. These properties can be in

direct conflict with the flying ability and energy demands typical of most

birds. Much organic matter of plant tissues is structural carbohydrate in cell

walls. Cellulose is one of the main components of cell walls and is the most

common organic compound in nature. No vertebrate produces the enzymes

necessary to digest cellulose. Therefore, many herbivores have enlarged

chambers in their gut where anaerobic microbes secrete these enzymes and

digest cellulose.

The hoatzin is the only known bird with a well-developed foregut

fermentation system (Grajal et al. 1989). The voluminous crop and caudal

esophagus have become functional fermentation chambers, analogous to those

of mammalian foregut fermenters. The crop and esophagus are situated in

front of a greatly reduced sternal carina, leaving little area for flight muscle

attachment (Fig. 2.2). Indeed, hoatzins are not powerful fliers, preferring to

hop from branch to branch. The peculiarities of the hoatzin's anatomy were

the subject of many early descriptive studies in an attempt to establish the

evolutionary affinities between the hoatzin and other birds (Brigham 1919,









Goeldi 1886, Huxley 1898, L'Hermenier 1837, Parker 1891, Perrin 1877, Pycraft

1895, Verheyden 1956). The presence of functional wing claws in hoatzin

chicks, the reduced sternal carina and poor flying abilities of the hoatzin were

regarded as the primitive characteristics of a "missing link" between the first

fossil birds such as Archaeopteryx and modern birds (Brigham 1919, Parker

1891). Present systematic studies place the hoatzin within the Cuculiformes

(de Queiroz and Good 1988, Sibley and Ahlquist 1973, Sibley et al. 1988). Some of

the early authors made an attempt to relate the large gut capacity to a

folivorous diet in the hoatzin (Boker 1929, Gadow 1891, L'Hermenier 1837).

Moreover, some authors described the smell of the gut contents as that of fresh

cow manure (Goeldi 1886, Young 1888). None of these authors suggested

foregut fermentation as the primary function of the large gut capacity in the

hoatzin.

Foregut fermentation in a 680 g flying endotherm is unexpected on

theoretical grounds. In most vertebrate herbivores, gut capacity scales

directly with body mass, while metabolism scales with body mass at a power of

0.75 (Demment and Van Soest 1983, Parra 1978). Accordingly, an endotherm

below 3-5 kg should not be able to support its normal metabolic requirements

on foregut fermentation alone. Moreover, large fermentation chambers place

an additional constraint on flying ability, because power requirements scale

directly with body mass (Pennycuick 1969). Foregut fermentation is also

unexpected in birds, because they do not have the dental adaptations to reduce

food particle size as do mammals. While birds can grind their food in the

muscular stomach or gizzard before it reaches the main digestion sites of the

hindgut, a bird with foregut fermentation requires significant particle size

reduction before or during fermentation to increase plant matter digestibility.

In fact, particle size is an important factor affecting plant matter digestibility










(Bjorndal et al. 1990), as demonstrated by the independent evolutionary

origins of rumination in the typical ruminants (Tragulids and Pecorans),

camels (Tylopodidae) and kangaroos (Macropodidae) (Hume and Dellow 1980,

Hume and Warner 1980, Langer 1974, 1980, 1984). Finally, selective particle

retention is another important gut function that enhances the nutritional use

of plant matter by foregut fermenters (Warner 1981b).

This study describes the gross anatomy and function of the

gastrointestinal tract of the hoatzin, and then compares the hoatzin's

gastrointestinal tract to other herbivorous birds and foregut fermenting

mammals. I measured the gut capacity of hoatzins and explored relevant

functions, such as particle dynamics and the nutritional and physical

characteristics of gut contents. If foregut fermentation is nutritionally

important for hoatzins, then it can be expected that gut capacity would be

similar to that of mammalian foregut fermenters. Additionally, the hoatzin's

digestive tract should be able to reduce particle size and show selective particle

retention to optimize the nutritional use of plant cell wall and cell contents.

The understanding of the anatomy and function of the hoatzin digestive tract

can provide insights into the evolutionary limits of foregut fermentation in

vertebrates and in birds in particular.



Materials and Methods

Birds were captured at several sites in the Llanos of Venezuela (see

Table 2.1). The total body mass of each bird was recorded immediately after

capture with a portable spring scale (+1 g). Then the gastrointestinal tract

was removed and weighed. The gut was divided with string knots into anterior

esophagus, crop, posterior esophagus, proventriculus, gizzard, small intestine,

caeca, and large intestine. The wet mass of the contents of each section was









determined by subtraction of the mass of each section with and without its

contents. The pH of the contents from each segment was measured in situ with

a portable pH-meter, usually within 20 min of the bird's death. Samples from

each segment were fixed with concentrated sulfuric acid and frozen in dry ice

for later measurement of volatile fatty acid (VFA) concentration. Other fresh

samples were weighed and dried at 100*C to constant mass for determination of

dry matter. Samples from some segments were fixed in buffered formalin for

particle size analysis and the remaining contents were frozen and later dried

at 600C to constant mass for nutritional analysis. Tissue samples from the gut

were fixed in 10% buffered formalin for histological analysis.

Gut contents were analyzed for dry matter, cell wall, nitrogen, and ash.

Fiber content was determined following the neutral detergent (NDF) method of

Goering and Van Soest (1970). Nitrogen content was determined by the

Kjeldahl method. The concentration of VFA was determined using gas

chromatography (Wilkie et al. 1986). Mean particle size in some gut sections

was measured using a computerized particle analysis video system with a

camera mounted on a microscope. The small sample sizes at specific hindgut

sites did not allow an accurate measurement of particle size, so the contents of

all hindgut sites were pooled.

Particle retention at various portions of the gut was measured on a

captive adult hoatzin. The bird was previously acclimated to a maintenance

diet for more than 60 days (Grajal et al. 1989). The maintenance diet consisted

of romaine lettuce, soybean protein powder, ground alfalfa pellets and fresh

young shoots of Enterolobium cyclocarpum, Pithecellobium saman, Guazuma

ulmifolia and Phthirusa af. orinocensis. The hoatzin was force-fed a gel

capsule with plastic markers of three sizes (10, 4 and 1 mm2) in a single pulse

dose (Warner 1981b). The inert plastic markers were pieces of brightly-










colored commercial flagging tape. This material has the advantage that its

specific gravity is almost one (1.01), so it resembles the specific gravity of wet

food particles in the fermentation chambers (Warner 1981b). The captive

hoatzin was housed in an individual custom-made metabolic cage with

removable floor trays and given food ad libitum. All feces were collected after

the pulse dose, and all markers present in the feces were counted. After 24 h

of administration of the single pulse dose, the bird was killed and the plastic

particles at each gut portion were counted. Acclimating hoatzins to captivity

is an expensive and time-consuming effort, so this experiment was not

repeated with more than one bird.

The characteristics of the gut contents and mean particle size were

compared at different sites of the gut using two-tailed statistical tests with an

alpha level of 0.05. Individual birds were considered experimental units for

the tests. Standard deviations are shown in parentheses.

All sacrificed birds were used for other complementary experiments on

in itro fermentation rates, microbial population studies, and general

histology (Grajal et al. 1989, also see Chapter 5). Additionally, complete

skeletons of these birds were prepared for museum collections and deposited at

the MARNR Museum at Maracay, Aragua state, Venezuela and the Florida

Museum of Natural History at Gainesville, Florida, U.S.A.



Results


Digestive tract morphology

Mean body mass of 24 adult hoatzins was 687.3 g (.77.1, Table 2.1).

Although males were on average heavier than females (730.7 vs. 705.9 g,

respectively,) the difference between sexes was not statistically significant.










Similarly, no significant differences in body mass were found between

capture sites or times of capture. The fresh contents the large crop and

posterior esophagus averaged a mass equivalent to 9% of total body mass,

roughly equivalent to 77% of the mass of the total digestive tract contents (Fig.

2.1, Table 2.2).

The mouth region has been partially described by early authors

(Banzhaf 1929, B6ker 1929). The general structure of the bill was more

Galliform than cuckoo-like, which may explain the classification of the

hoatzin as a Galliform for many years (Banzhaf 1929, Huxley 1898). The bill

had sharp edges that help in cutting leaves. The lanceolate tongue had sharp

caudally directed papillae, like backward-pointing spines, which probably

assist in swallowing large pieces of leaves. A pair of large sublingual

mandibular salivary glands (glandula mandibularis external) (sensu McLelland

1979) were evident. Although I did not measure the composition of the saliva

from these glands, it was quite thick and sticky. Other salivary glands in the

corner of the mouth (glandla an.guli .ri.a) and in the cheeks were relatively

smaller (F. Michelangeli, pers. comm.). The upper esophagus was quite

smooth, soft and elastic, with almost no muscle. Near the entrance of the crop,

the upper esophagus started to show some inner longitudinal ridges and thick

muscle tissue, resembling the upper crop.

The crop was a large muscular organ folded into two chambers and

wrapped by mesenteries. The two crop chambers were connected through a

constricted zone with circular muscles that resembled the pillars found in

ruminant stomachs. The crop extended ventrally and was harbored in a

concave depression of the sternum keel (Fig. 2.2). The muscle wall of the crop

was thick, with several circular muscle layers. The interior lining was

covered by a hard epithelium and showed parallel longitudinal ridges and










folds. The ridges were generally higher (up to 4 mm) on the ventral side of

the crop, and shorter and stouter on the dorsal side of the crop. The terminal

portion of the second crop chamber had the shortest ridges. The crop ended in

a narrow pillar zone connecting to the posterior esophagus. The crop contents

were a heterogeneous green mixture of fully recognizable leaves, partially

broken leaves and unrecognizable plant material.

The posterior esophagus was also heavily muscular and quite rigid. Its

hard inner lining also showed longitudinal ridges, but these ridges were

shorter and less uniform. The posterior esophagus consisted of a series of

small sacculated chambers. Most of these chambers were separated by pillars

and constriction zones, sometimes completely circular or otherwise

resembling semilunar folds. Most of these muscular folds and constrictions

were longitudinally connected, resembling short haustrations. The contents

in the posterior esophagus seemed to be drier than those in the crop and were

less diverse in size. No complete leaves were recognized in the posterior

esophagus, except some small leaves, such as Acacia spp. (approx. 4 x 2 mm).

The glandular stomach or proventriculus was small, barely wider, and

less muscular than the connecting posterior esophagus. An abrupt change in

pH (Table 2.2) suggested that the proventriculus is the secretary region of

gastric acids. The gizzard was also small but muscular, with a hardened

keratinous inner lining. Two transversal muscle types were found in the

gizzard, but none was thicker than the muscles of the crop. No grit was

present in any hoatzin gizzard, as expected, considering that the birds rarely

go to the forest floor. The contents of the gizzard were thoroughly ground,

and only a few leaf veins and petioles could be identified.

The small intestine was uniform in diameter. The soft and elastic

intestinal walls were only covered with thin muscle layers. The small










intestine was never completely full, and the contents were generally

distributed in lumps. The contents in this region were not green as in the rest

of the anterior gut, but orange-brown. Almost no recognizable particles could

be found. The plant matter of the small intestine was mixed with a thick,

sticky mucous substance. The paired caeca were relatively small for a

herbivorous bird (Gasaway et al. 1975, Inman 1973, McLelland 1979, Ziswiler

and Farner 1979) and lined with thin muscle. The caeca were partially full

with an homogeneous dark green-brown material with the consistency of

thick pudding. The large intestine was short and not clearly differentiated

from the small intestine. No obvious morphological differentiation between

the large intestine and the cloaca was evident (Fig. 2.1). In two individuals,

white mucous streaks were found at the end of the large intestine. Whether

these streaks were thick mucous aggregations or refluxed uric acid was not

determined.



Gut contents

The dry matter (%DM) of the crop contents was significantly lower than

the average %DM of the young tender leaves that constitute the typical hoatzin

diet (Grajal et al. 1989) (Mann-Whitney U, P = 0.006, n = 5). The %DM of the

contents of the posterior esophagus were significantly higher than those of

the crop (Mann-Whitney U, P = 0.016, n = 5) but similar to those of the

proventriculus and the gizzard. The hindgut had the lowest %DM contents

(Table 2.2).

Nutritional characteristics of gut contents changed along the gut (Table

2.2). Cell wall levels were significantly higher in the esophagus than in the

crop (Mann-Whitney U, P = 0.009, n = 5). Cell wall levels were significantly

different among all three measured gut sites (Kruskal-Wallis one way ANOVA,










P = 0.002, n = 5; Fisher PLSD post-hoc test). The hindgut had the lowest cell wall

levels. Nitrogen and organic matter levels were significantly higher in the

esophagus than in the crop and much lower in the hindgut (Mann-Witney U,

both P = 0.009, n = 5).



Particle dynamics

Mean particle size was smaller in the caudal esophagus than in the crop

(Table 2.2), although the difference was barely significant (Mann-Whitney U,

P = 0.047, n = 5). Mean particle size was significantly smaller (and less

variable) at the hindgut than at either foregut site (Mann-Whitney U, P =

0.009, n = 5). Mean particle size was significantly different at all three gut

sites (Kruskal-Wallis one way ANOVA P = 0.004, n = 5; Fisher PLSD post-hoc test).

The experiments on particle retention showed that the larger the particle, the

longer it remains in the anterior fermentation organs (Table 3). After 24

hours, 92.5% of the large (10 mm2) plastic markers remained at the crop and

esophagus, none was found in the hindgut, and only a few (3.7%) were

excreted. Interestingly, all the excreted 10 mm2 plastic markers were tightly

folded in half. A higher proportion of the 4 mm2 plastic markers were

excreted in the 24 h period and none of these markers was folded. The small 1

mm2 markers were present almost everywhere in the gut. No plastic markers

were found in the caeca.










Discussion


Digestive morphology

In the hoatzin, obligate folivory has produced remarkable anatomical

specializations. The crop and esophagus are the primary organ for digestion

and fermentation. As a consequence, the morphology of the gut is more

similar to that of small mammals with foregut fermentation (Hofmann 1989)

than to any known herbivorous bird (Fig. 2.1). Indeed, the crop and the

esophagus are the functional equivalent of multi-chambered fermentation

organs. The relative capacities at these sites are among the largest

fermentation capacities of any bird (Dawson et al. 1989, Herd and Dawson

1984), and roughly equivalent to the relative capacity of mammals with

foregut fermentation (Demment and Van Soest 1983, Demment and Van Soest

1985, Parra 1978) (see Fig. 2.3). Similarly, the pH and VFA levels are within the

range of mammals with foregut fermentation (Grajal et al. 1989). Since VFA

can be actively absorbed at the fermentation sites, the inner folds of the crop

and esophagus increase area for VFA absorption and probably help in the

selective passage of particles. The dark red color of the crop muscles suggests

a high blood supply that probably enhances oxygen supply and absorption of

VFA (F. Michelangeli, pers. comm.).

The crop and posterior esophagus probably are important sites for

selective retention of the solid over the liquid fraction. The thick muscle

tissues at the crop and esophagus probably squeeze the digesta, resulting in a

gradual increase in the %DM from the crop to the esophagus. The low %DM of

the crop contents, relative to the average hoatzin diet, suggests that saliva

secretions into the first portion of the fermentation chambers are significant.









The abrupt decrease in %DM contents between the gizzard and the small

intestine suggests an increased absorption of water and digestible nutrients in

solution.



Gut contents

The large volume, pH, and VFA concentrations in the crop and posterior

esophagus demonstrate that these are the main fermentation sites where most

cell walls are broken down and microbially digested. Usually, as the cell walls

are broken by physical abrasion and microbial fermentation, digestible cell

contents disappear rapidly. This study, however, could not discern whether

the cell contents are more heavily used by foregut microbes or moved on to

the lower gut to be absorbed by the host. In addition, gastric digestion of

hemicellulose can be important in the overall disappearance of the cell wall

fraction (Dawson et al. 1989, Keys et al. 1969, Parra 1978). Finally, the pH and

VFA levels in the paired caeca demonstrate additional fermentation in the

hindgut. Caecal fermentation is probably important in water and nitrogen

recycling and microbial production of essential vitamins (Mead 1989,

Remington 1989). Higher microbial density can explain the significantly

higher levels of nitrogen and organic matter in the esophagus.



Particle dynamics

The constrictions and sacculations of the crop and posterior esophagus

are presumably important adaptations for selective particle retention. A large

proportion of the large and medium plastic markers remained in the crop and

esophagus after 24 hours. The observation that almost all the excreted 10 mm2

plastic markers were folded supports the idea that there is a minimum size

threshold for escape to the lower gut. The 4 mm2 plastic markers behaved









similarly, but they were not folded. None of the markers entered the caeca,

suggesting that caecal filling can be highly selective (Bjomhag 1989). I

suppose that even the 1 mm2 plastic markers were too large to enter the caeca,

which were filled with an homogeneous thin paste.

The relatively long retention time of large plastic markers in the

foregut was probably artificially high, since the markers could not be broken

into smaller particles or attacked by microbes. Normally, large food particles

are broken by a combination of physical abrasion and microbial fracture of

the cell walls. Evidently, these plastic markers were inert to these digestive

processes. The markers were appropriate to measure selective passage for two

reasons. First, the behavior of these markers closely resembled that of food

particles, since plastic tape has specific gravity similar to normal food

particles. Second, the standardized particle sizes allowed a quantitative count

of particles at the gut sites.

The crop and esophagus are also important sites for reduction and

homogenization of particle size. This is probably achieved by the combined

action of muscular pressure, abrasion by the hardened lining of the crop and

intense microbial attack on the cell walls. The result is a functional

equivalent to the re-mastication that gives ruminants their name, but with the

added advantage that fermentation and trituration occur at the same site.

Particle size reduction is an important factor in overall plant material

digestion. Indeed, particle size reduction in toothless vertebrates is a crucial

component of cell wall and cell contents digestion, because smaller plant

particles can be more easily attacked by fermenting bacteria (Bjorndal et al.

1990). Further panicle reduction probably takes place in the mid-gut, where

the combined effect of gastric digestion in the proventriculus and physical

grinding in the gizzard result in significantly smaller particles at the hindgut.









Although the small sample sizes did not allow measurement of particle size in

the caeca, the appearance of their contents suggests that the caeca are sites

for selective entrance of fluid and small particles.



Conclusions

The hoatzin's strategy to deal with a leafy diet is unique, leading to some

extreme morphological, physiological and behavioral adaptations (Strahl

1988). This bird is the only known non-mammalian vertebrate with a foregut

fermentation digestive system. These results demonstrate that the hoatzin

crop and posterior esophagus are the primary site for digestion of its leafy

diet. The anatomy and function of the hoatzin gut are unique for birds.

Indeed, they are more similar to those of mammals with foregut fermentation,

with the difference that the hoatzin is almost an order of magnitude smaller

than the smallest mammals with well-developed foregut fermentation. This is

probably achieved by a unique set of morphological adaptations in the hoatzin

gut. In the hoatzin, food is effectively broken down into smaller particles at

the fermentation chambers, increasing digestive efficiency. The selective

retention of solid food particles at the foregut sites has not been reported for

birds (Warner 1981b). Indeed, most other birds eating a bulky diet are able to

either regurgitate or pass refractory solids faster than the more digestible

liquids (Bjornhag 1989, Duke and Rhoades 1977, Levey 1976, Warner 1981b).

The hoatzin digestive strategy, however, seems to use of both cell contents and

cell wall as nutritional sources.

The relative capacity of the hoatzin's fermentation structures is similar

to the capacity of mammals in which foregut fermentation supplies a

significant amount of the metabolic requirements. The levels of VFA at the

crop and esophagus are similar to those of foregut fermenting mammals,









suggesting that microbial fermentation at the foregut sites is important for

the overall metabolism of hoatzins. The contents of the anterior crop chamber

are dry compared to the dry matter of contents in ruminants or other foregut

fermenting mammals (Parra 1978, Van Soest 1982). Although the salivary

glands were not large, the saliva was thick and probably contained

mucoproteins and buffering salts. It is not clear how hoatzins regulate pH

levels at the foregut fermentation sites. The ridges at the interior lining of the

crop increase the absorption area, diminishing the acidifying effect of VFA

accumulation in the fermentation organs. High microbial populations at the

posterior esophagus may explain the increase in organic matter and nitrogen

concentrations from the crop to the posterior esophagus.

The presence of a well-developed foregut fermentation system in the

hoatzin provides new insights into the morphological and functional

constraints of foregut fermentation in vertebrates. Gut capacity, particle

reduction, and selective retention are important characteristics for an

efficient use of plant leaves as a food source. Indeed, relative gut capacity,

particle reduction and dynamics, pH and VFA levels in the hoatzin are quite

similar to mammals with foregut fermentation systems. These similarities

across taxonomic classes suggest similar functional constraints and selective

pressures on the evolution of foregut fermentation.
















































10 cm


Figure 1) The digestive tract of the hoatzin. Its unique form and function is more
similar to that of mammals with foregut fermentation than to any known bird.



















































Figure 1) Schematic representation of the anterior gut of an adult hoatzin seen from
the left, showing the crop (a), caudal esophagus (b), proventriculus (c), and gizzard
(d). The anterior sternum is much reduced to room the voluminous fermentation
chambers, with a drastic reduction of the area for flight muscle attachment to the
sternal carina (e). A "resting" pad (f) at the end of the sternum is used while
perching with a full crop.


E F













100






O, 1
10





E .4: hoatzin
C. 1
U)





.01
.1 1 10 100 1000
.01 ----I---I---i---


Body mass (kg)




Figure 2.3) Relationship between body mass (kg) and fermentation contents
(kg) of wild ruminants from Demment and Van Soest (1983). The line
represents the regression log y = -1.02 + 0.998 log x (R2 = 0.95). The
fermentative capacity of the crop and esophagus of the hoatzin falls within
the 95% confidence limits of the regression line.










Table 2.1) Mass (in g), sex and capture site of hoatzins used in this study.
Capture sites correspond to the following geographic coordinates: Masaguaral
(670 35' W, 8* 34' N), Guarico River (670 28' W, 80 33' N), Suapure (66* 20' W, 6 0
08'), Pifiero (680 04' W, 80 82' N). Not all birds were sexed. Suapure data from
unpublished observations by Rodrigo Parra.

Mass (g) Date Site Sex

765 12-2-84 Suapure male
785 12-2-84 Suapure male
681 12-2-84 Suapure female
653 12-2-84 Suapure female
650 25-5-88 Masaguaral
600 27-5-88 Masaguaral
520 27-5-88 Masaguaral
740 13-6-88 Masaguaral
660 28-6-88 Masaguaral
450 28-6-88 Masaguaral
730 29-6-89 Guarico River female
695 15-7-89 Guarico River female
685 22-7-89 Guarico River male
700 20-7-89 Guarico River female
720 01-8-89 Guirico River male
740 07-8-89 Guarico River female
740 18-9-89 Guarico River female
680 18-9-89 Guarico River male
740 11-7-90 Pifiero male
640 11-7-90 Pifiero female
760 11-7-90 Pifiero female










Table 2.2) Characteristics of the gut contents of hoatzins. Sample sizes were n =
5 for all parameters except for relative capacity, which is presented as
percentage of body mass (mean body mass for this sample was 712 g _56.6, n =
8). Mean values of organic matter, nitrogen and cell wall are presented on a
dry matter basis. Large intestine values for mean particle size, organic matter,
nitrogen and cell wall represent the pooled contents of the caeca, large
intestine and lower small intestine. Volatile fatty acids (VFA) are presented in
mmol/1 of contents. Standard deviations are in parentheses.


Posterior Small Large
Crop Esoph. Provent. Gizzard Intest. Caeca Intest.


Length (cm)


25 15 3 3 63 3 15


Relative capacity
(% of body mass)
%DM


Mean particle size
(microns)
%Organic matter


%Nitrogen


%Cell wall (NDF)


VFA (mmol/l)


%Acetic


1.4 0.1 0.2


(1.2)
22.9


(0.3)
28.3


(0.0)
27.7


0.2 0.6


(3.0) (2.6) (9.9) (6.0) (3.6) (1.4) (2.2)


467.2


279.6


(158.4) (122.7)


92.4


93.4


(0.2) (0.2)
4.4 4.7
(0.1) (0.1)
51.0 59.3
(2.3) (2.4)
6.4 6.6 2.1
(0.4) (0.3) (0.3)
114.5 170.3
(62.3) (121.0)


68.1


69.8


(5.8) (3.6)


%Propionic


%Butyric


%Isobutyric


13.2
(4.8)
8.3


13.9
(1.3)
7.7


(2.3) (3.1)


10.4
(1 6i


8.6
(1 R)


(0.0)
30.8


(0.3)
20.3


(0.1)
19.3


(0.2)
19.9


138.6
(5.1)
91.0
(0.8)
4.1
(0.1)
37.1
(3.0)


(0.1)
94.7
(42.1)
77.4
(0.6)
13.3
(0.6)


13.6
(9 .5













Table 3) Percentage of plastic markers found at gut sites after 24 hours from a
single pulse dose. The total number of markers given were 27 large (10 mm2),
38 medium (4 mm2), and 39 small (1 mm2).


10 mm2


Plastic marker type
4 mm2


1 mm2


Gut site
Crop 48.1 23.7 25.6
Posterior Esophagus 44.4 31.6 15.4
Proventriculus 3.7 2.6 5.1
Gizzard 0.0 5.3 2.6
Small Intestine 0.0 0.0 2.6
Caeca 0.0 0.0 0.0
Large Intestine 0.0 0.0 7.7
Excreted 3.7 36.8 41.0















CHAPTER 3
DIGESTIVE EFFICIENCY OF THE HOATZIN



Introduction

Plant leaves are difficult to digest, bulky and usually have defensive

chemical compounds. The digestion of leaves generally results in a low rate of

energy extraction that can conflict with the high energy demands of flight

and endothermy. Consequently, few birds rely on plant fiber digestion for

their nutritional needs. About 3% of extant bird species feed extensively on

green leaves or buds (Morton 1978). The main reason for the rarity of

herbivory in birds seems to be related to the conflict between eating a bulky

diet of low nutritional value and the energy demands of flight and endothermy

(Morton 1978). Although some birds can digest plant fiber, it is generally of

little nutritional value. Therefore, most birds that eat significant amounts of

plant material only extract the readily digestible cell contents, quickly

excreting the bulk of the cell wall or fiber. Examples are herbivorous

Anseriformes (geese and ducks) (Buchsbaum et al. 1986, Dawson et al. 1989,

Marriot and Forbes 1970), some Galliformes (Inman 1973), takahe, and kakapo

(Morton 1978). Large ratites may digest significant amounts of fiber (Herd and

Dawson 1984, Mackie 1987, Withers 1983). Within the Galliformes, species of

the family Tetraonidae (grouse and ptarmigan) can derive significant

nutritional benefit from the fermentation of fiber in enlarged paired caeca

(Gasaway 1976, Gasaway et al. 1976, Hill et al. 1968).

Most birds that digest significant amounts of fiber have fermentative

chambers in the posterior part of the gut (hindgut) (e.g. grouse and










ostriches). Others, such as geese, ducks and emus have no specialized gut

fermentative chambers (Buchsbaum et al. 1986, Dawson et al. 1989, Herd and

Dawson 1984, McLelland 1979, Ziswiler and Farner 1979). Foregut fermentation

is essentially restricted to mammals such as ruminants, colobid monkeys,

kangaroos and tree sloths. The hoatzin, Opisthocomus hoazin, is unique among

birds. It is one of the few known obligate avian folivores and the only known

bird with a well-developed foregut fermentation system (Grajal et al. 1989).

Although specialization to a folivorous diet was reported by early studies

(Beebe 1909, Grimmer 1962), the hoatzin's nutritional ecology received little

attention until recently (Grajal et al. 1989). In the hoatzin, the crop and caudal

esophagus are enlarged (see Chapter 2), with a relative gut capacity similar to

the fermentative structures of mammalian herbivores (Demment and Van

Soest 1983, Parra 1978). Furthermore, the pH and concentrations of

fermentation by-products such as volatile fatty acids (VFA) in the anterior

part of the gastrointestinal tract are comparable to known foregut fermenters

(Grajal et al. 1989). Therefore, hoatzins are the only known flying vertebrate

with a well-developed foregut fermentation system. This digestive system is

unique both in structure and function among birds.

Foregut fermentation in hoatzins is unexpected. Current allometric

models of foregut fermentation predict a lower limit of = 8 kg body mass for

endotherms with foregut fermentation (Demment and Van Soest 1983,

Demment and Van Soest 1985, Parra 1978). The rate of energy available from

the fermentation of plant fiber does not fulfill the predicted total metabolic

requirements for an endotherm below this critical mass. Hoatzins, however,

are an order of magnitude lower than the predicted minimum body mass for a

mammal with foregut fermentation (650 g).










To evaluate the function of this unique digestive strategy, it is

necessary to estimate the digestive efficiency of the hoatzin. Additionally, a

comparison of hoatzin digestive efficiency with other mammalian and avian

herbivores can provide new insights into the evolution of foregut

fermentation.

Hoatzin digestive efficiency was studied using balance trials under

captive conditions with three experimental diets of various fiber levels.

Previous attempts to keep these birds in captivity failed, probably due to

nutritional imbalances (Grimmer 1962) and wide fluctuations in ambient

temperature (Webb 1965). After extensive field work on the dietary and

thermoregulatory constraints of wild populations (Grajal et al. 1989), I was

able to keep hoatzins in captivity in 1986.



Materials and Methods



Animal husbandry

Hoatzins were captured along the Guarico River (670 28' W, 80 33' N), an

affluent of the Orinoco River in central Venezuela. Two birds were used for

the balance trials in 1986, two in 1988, and five in 1989. The birds were kept in

outdoor aviaries at Fundo Pecuario Masaguaral, a private ranch in the central

llanos of Venezuela. The birds were acclimated to captivity by a slow and

progressive change from their natural diet to experimental diets. Hoatzins are

extremely neophobic towards unknown foods, and acclimation required

dedication and persistence. After an acclimation period of more than 60 days,

the birds were moved to the Animal Production Institute of the Universidad

Central de Venezuela (UCV) campus at Maracay. The birds were kept indoors in

1 x 1 x 2 m custom-made metabolic cages for a 20 day adjustment period before










the start of the experiments. The cages had removable floor trays for

quantitative recovery of feces. Food during the adjustment period was offered

ad libitum twice daily, in the morning and in late afternoon.



Diet composition

The experimental diets were a "salad" of romaine lettuce, sprinkled with a

powdered mix of varying proportions of ground alfalfa hay pellets, ground

Timothy grass hay, ground roasted soybeans, and a multi-vitamin and mineral

complement. Three diets were offered (Table 3.1). All three diets were

different not only in their fiber and protein content, but also in the quality of

the fiber, with different lignin/cellulose and hemicellulose/cellulose ratios.

Diet A was an acclimation diet with high protein and low fiber levels that

maintained the animals in stable condition and helped to overcome the initial

stress of confinement. Diet A, however, had a relatively poor quality fiber

fraction, since most of the fiber came from soybean hulls and alfalfa hay.

Consequently, Diet A had a very low (negligible) hemicellulose content. This

resulted in low overall fiber digestibilities (see results below). The

digestibility of this diet was studied in 1986. Diet B was used in 1988, and was

designed to resemble the high nitrogen and low fiber portions of the hoatzin's

natural diet. Consequently, Diet B had levels of hemicellulose and lignin

similar to these portions of the hoatzin's natural diet. Diet C resembled the

high fiber portions of the hoatzin's natural diet, and was used in 1989. The

higher fiber content of diet C was achieved by increasing the grass hay

contribution to the diet. This increased the hemicellulose content and

decreased the lignification ratio. None of the three diets was identical to the

estimated natural diet. Diet C, however, was the most similar in nitrogen level,

overall fiber content and fiber composition.










Intake and digestibility

Digestibility experiments lasted 7 days, with a previous acclimation

period of 10 days to the experimental diet. During the experiments, feces were

collected daily on pre-weighed aluminum foil on the floor trays. Offered and

rejected food were measured twice daily in separate food trays. Food and feces

were dried in a forced air oven at 650C for 48 h and cumulatively stored for

later analysis. Sub-samples were dried at 1050C for absolute dry matter (DM)

content. All samples were ground and chemically analyzed following the

detergent methods of Goering and Van Soest (Goering and Van Soest 1970).

Neutral detergent fiber (NDF) consisted of cellulose, lignin, and hemicellulose.

Acid detergent fiber (ADF) consisted of cellulose and lignin. Lignin was

determined by treating the ADF fraction with concentrated sulfuric acid.

Hemicellulose was the difference between ADF and NDF. Cellulose was

calculated by subtracting lignin and ash from the ADF fraction. Cell contents

were estimated by subtracting NDF from the original sample dry weight.

Organic matter was estimated as the difference between DM and ash after

incineration at 500C. Nitrogen was determined by the Kjeldahl method. Uric

acid from whole feces was determined by elimination using the method of

Tepstra and deHart (1973).

Since the experimental diets were composed of two substrates with

different nutrient compositions and physical properties (fresh lettuce and a

dry powder mix), substrate selection by the hoatzins was unavoidable. A

separate experiment was performed to estimate the relative intake of each

substrate by three hoatzins. The hoatzins were kept in the same metabolic

cages and fed Diet C twice daily for 5 days immediately after the digestibility

trials. All three mixed diets had similar physical properties, so Diet C was used

as a representative. Lettuce and powder mix were separately weighed before










they were combined and offered. The rejected fraction was removed twice

daily. All rejected lettuce was physically separated from the rejected powdered

mix by washing under running water, and collecting the powder on a 0.3 mm2

wire mesh sieve. The separated rejected lettuce and powder mix were dried in

a forced air oven at 650C for 48h and relative rejection of each substrate was

calculated on a dry matter basis. The relative rejection percentages were used

to estimate the average composition of the intake in the three diets (Table 3.1).

Average intake of lettuce and powdered mix was found to be significantly

different from the average lettuce and powdered mix ratio of the offered diet

(see Results section).

Since hoatzins were actively selecting substrates of the offered diet, dry

matter intake (DM intake) was estimated as [(gDM offered lettuce + gDM offered

powdered mix) gDM total rejected]. This formula was also used to calculate

intake for each nutritional fraction, such as nitrogen or NDF. Apparent

digestibility for each nutritional fraction was calculated as [(DM intake of the

nutritional fraction DM fecal excretion of the nutritional fraction) / (DM

intake of the nutritional fraction)] x 100. Organic matter digestibility of the

experimental diets by live hoatzins was compared to an in vitro organic matter

digestibility (IVOMD) using cow ruminal inoculum (Alexander and McGowan

1966). The IVOMD provided a comparison of the organic matter digestibility of

the experimental diets. Gross energy from lettuce, powdered mix, rejected

food, and feces was measured with a Parr Bomb calorimeter. Metabolizable

Energy Coefficients (MEC) were calculated as [(Gross energy of diet x DM

intake Gross energy of fecal excretion x DM fecal excretion) / (Gross energy

of diet x DM intake)] x 100.

All statistical tests were two-tailed with an alpha level of 0.05.

Individual birds were considered as the experimental units for each test.










Standard deviations are shown in parentheses. Digestibilities (expressed as

percentages) were transformed to their square root and compared using

unpaired t-tests, unless otherwise specified.



Results

The body mass of captured birds ranged from 550 to 690 g, with a mean

of 616.5 g (.85.1) at the time of the experiment. All birds maintained body mass

on the three experimental diets. Hoatzins at the end of the experiments were

between 99% and 103% of their original body mass (mean 100.9% + 1.8).



Intake and digestibility

Lettuce was actively selected over the powdered mix by captive hoatzins.

On average, hoatzins rejected 66% (8) of the offered powdered mix and 40%

(18) of the lettuce. The average intakes of dry matter lettuce and dry matter

powdered mix were 58.9% and 41.1% (17.6) from the total dry matter intake,

respectively. The relative DM intakes of lettuce and powdered mix were

significantly different (Wilcoxon matched pairs signed rank test, P = 0.005, n =

3). Total DM intakes (gDM/day) or mass-specific intakes (gDM/kg body mass

day) were not significantly different among diets (Table 3.2) (unpaired t-

tests). Similarly, gross energy intakes (KJ/day or KJ/kg body mass day) were

not significantly different (unpaired t-tests).

Apparent digestibilities and metabolizable energy coefficients (MEC)

are shown in Table 3.3. Dry matter digestibilities and cell contents

digestibilities were significantly different among diets B and C (P = 0.02 and P =

0.001, respectively). Organic matter and nitrogen digestibilities were not

significantly different among diets (unpaired t-tests). In vitro organic matter

digestibilities of the experimental diets were similar to the in vivo average










organic matter digestibility for the three diets (Table 3.3). Diet A had lower

MEC than Diet B and Diet C (unpaired t-tests, P = 0.03).

Neutral detergent fiber (NDF) digestibilities were lower for Diet A and

Diet B than those for Diet C (unpaired t-tests, P = 0.003 and P = 0.006,

respectively). Diets A and B did not show significantly different NDF

digestibilities. In general, the higher the NDF content of the diet, the higher

the NDF digestibility. Since cellulose digestibilities were only different among

diets B and C (P = 0.04), and lignin digestibilities were not significantly

different, most of the difference in NDF digestibilities was the result of

differential hemicellulose digestibilities. Indeed, hemicellulose digestibilities

were significantly different among all diets, with the highest digestibility for

Diet C, and very low ("negative") hemicellulose digestibility for Diet A. This

"negative" hemicellulose digestibility was an artifact, because Diet A had low

hemicellulose levels, and the analytical methods failed to detect this fraction

in such small amounts.

The extraction of uric acid from the feces demonstrated its importance

in the calculation of nitrogen digestibilities. Average nitrogen digestibility

without uric acid extraction was low (56.2% 11.4), while uric acid extraction

showed a more realistic average nitrogen digestibility (78.3% 9.3) (Table 3.3).

Although nitrogen digestibilities were not much different among diets, the

differences between nitrogen digestibilities with and without uric acid

extraction were highly significant (paired t-test, P = 0.002).










Discussion

Comparisons of the results on hoatzin intake and digestibilities with

other studies of herbivorous birds or mammals are difficult to interpret

because of major differences in experimental methods, diets, digestive

strategies and body mass. Furthermore, most studies on ruminants are based

on relatively large, domesticated grazers, such as sheep and cows.

Hoatzins seem to have mass-specific intake levels within the range of

the emu, a large ratite herbivorous bird (Herd and Dawson 1984) and within

the levels reported for grouse (Gasaway 1976, Gasaway et al. 1976, Inman 1973).

Hoatzin intakes are lower than intake levels in geese (Marriot and Forbes

1970), probably because geese make little nutritional use of fiber. However,

hoatzins have intakes levels lower than ruminant intakes under similar diets

(Van Soest 1982).

Mass-specific DM intake levels in hoatzins were constant and

independent of diet type. This relatively constant intake in the hoatzin under

three different diet compositions is in contrast with the variable intakes of

almost all herbivorous birds under different dietary fiber levels (Gasaway

1976, Herd and Dawson 1984, Hill et al. 1968, Miller 1984, Moss and Parkinson

1972, Moss and Trenholm 1987). This may suggest that hoatzin intake is

regulated by gut fill rather than by cell wall level, at least for the range of cell

wall levels offered. The range of cell wall levels in the experimental diets was

within the range of similar experiments on herbivorous birds (Gasaway 1976,

Herd and Dawson 1984, Hill et al. 1968, Miller 1984, Moss and Parkinson 1972,

Moss and Trenholm 1987), but relatively limited compared to similar

experiments on ruminants (Van Soest 1982). The cell wall range of the

experimental diets, however, reflected natural variation in cell wall levels of

the natural diet of hoatzins (Strahl and Parra 1985).










Energy intakes were similar among diets. This is explained by the

similarities in DM intake and energy content of the experimental diets. The

resulting MEC's were higher than any values of MEC reported for other

herbivorous birds under diets with similar energy content (Karasov 1990).

Apparent digestibilities of many dietary fractions were high (Table 3.3).

Digestibilities of cell contents, organic matter, and DM by hoatzins were as

high as those reported for larger mammalian herbivores (Parra 1978, Van

Soest 1982). Dry matter (DM) digestibilities in the hoatzin were higher than

DM digestibilities by other herbivorous birds (Dawson and Herd 1983, Dawson

et al. 1989, Gasaway 1976, Gasaway et al. 1976, Herd and Dawson 1984, Inman

1973, Moss and Trenholm 1987). High overall DM digestibilities also indicate

that hoatzins are able to digest both the cell contents and cell walls of plant

material to a greater extent than most avian herbivores. Furthermore, the

high DM digestibilities may explain the relatively low DM intake of hoatzins

when compared to other herbivorous birds and most mammals.

Total nitrogen digestibilities were not different among diets. The

relatively low values of nitrogen digestibility are always a methodological

artifact of balance trials on vertebrates that excrete feces and urinary

products through a common cloaca (e.g., birds and reptiles). Uric acid and

other urinary nitrogenous compounds are present in bird feces, inflating the

amount of total nitrogen excretion and reducing the apparent digestibility of

the nitrogenous fraction. The extraction of uric acid from feces showed

significantly higher nitrogen digestibilities, well within the range of other

herbivorous birds (Buchsbaum et al. 1986, Marriot and Forbes 1970).

Organic matter digestibilities were within the range reported for

ruminants (Hoppe 1977a, Parra 1978, Van Soest 1982). The similarities between

organic matter digestibility by live hoatzins and in itro organic matter










digestibility (IVOMD) by cow ruminal inoculum shows that the hoatzin

fermentation system is comparable to that of the ruminant in extracting and

digesting the organic component of the diets. High organic matter

digestibility is the result of the breakdown of the cell wall structural fiber,

which in turn makes the cell contents available for digestion.

Digestibilities of all fiber fractions by hoatzins were equal to or higher

than fiber digestibilities recorded for herbivorous birds (Buchsbaum et al.

1986, Dawson et al. 1989, Gasaway 1976, Gasaway et al. 1976, Herd and Dawson

1984, Inman 1973, Marriot and Forbes 1970, Moss and Trenholm 1987). Some

studies (Dawson et al. 1989, Herd and Dawson 1984) show higher NDF

digestibilities than hoatzins on Diet A, but lower than NDF digestibilities of

Diets B or C. When compared to herbivorous mammals, these hoatzin fiber

digestibilities are more similar to reported fiber digestibilities for ruminants

and higher than for some non-ruminant mammalian herbivores under

similar diets (Demment and Van Soest 1985, Hume and Dellow 1980, Parra 1978,

Van Soest 1982).

Digestibilities of the NDF fraction were positively correlated to diet fiber

content. The results for NDF digestibility of Diet A appear to be relatively low

compared to digestibilities of the other two experimental diets or digestibilities

recorded in ruminants (Parra 1978, Van Soest 1982). The low fiber digestibility

of Diet A can be attributed to the poor fiber quality of this diet. A large portion

of the NDF fraction was composed of relatively indigestible fiber, mainly

soybean hulls and the lignified portions of alfalfa. Both soybean hulls and the

lignified alfalfa are mainly composed of refractory fiber compounds, such as

cutin and lignin, which are essentially indigestible (Van Soest 1969). In

contrast, Diet C showed high NDF digestibilities.









Since cellulose, ADF or lignin digestibilities were not different among

diets, the differences in NDF digestibilities can be explained by the differential

digestibility of the hemicellulose fraction of each diet. Diet C had more

hemicellulose than the other two diets, and consequently hemicellulose

digestibilities of Diet C were the highest of the three. In fact, hemicellulose

digestibilities by hoatzins are among the highest recorded for avian

herbivores (Buchsbaum et al. 1986, Dawson et al. 1989, Herd and Dawson 1984).

Hemicellulose can be microbially digested to the same extent as cellulose (Keys

et al. 1969). In addition, hemicellulose may be hydrolyzed by gastric enzymes

under low pH levels (Dawson et al. 1989, Herd and Dawson 1984, Keys et al. 1969,

Parra 1978). The combination of microbial and gastric digestion may explain

the high hemicellulose digestibility, and therefore its contribution to the

higher NDF digestibility of Diet C.

High cellulose digestibility in hoatzins is not predicted by current

models of cellulose digestibility as a function of body mass (Demment and Van

Soest 1983, Parra 1978). Small herbivores have limited capacity to digest fiber

because of the size limitations in gut capacity and high energy turnover of a

small endotherm. Consequently, the relationship of body mass to cellulose

digestion in mammalian herbivores has been described as y = 14.5 + 5.4x,

where y is cellulose digestion and x is body mass (in kg 0.25) (Van Soest et al.

1983). This model predicts a cellulose digestibility of 19.4% for an herbivore of

a mass of 0.65 kg, as the hoatzin. Although the hoatzin is not a grazer and

assuming that the equation is valid for a folivorous bird, the empirical

measurements of cellulose digestibility for the three diets were approximately

60%, well above the predicted values of the equation. Therefore, cellulose

digestion in hoatzins is higher than predicted by body mass or by size

limitations in gut capacity (Demment and Van Soest 1983, Parra 1978). Other









factors, including the unique gut structure and function (see Chapters 2 and

4), and the low rate of metabolism (see Chapter 6), contribute to the relatively

high cellulolytic activity for a small herbivore such as the hoatzin.

Additionally, the digest are retained for long periods of time in the anterior

chambers of the gastrointestinal tract, particularly in the crop and caudal

esophagus (Grajal et al. 1989). Thus providing enough time for thorough

microbial fermentation. Finally, food particles are reduced in size in the crop

and esophagus, possibly by a combination of microbial attack and physical

grinding by the crop's internal epithelium (Grajal et al. 1989, also see Chapter

2). Particle size reduction increases the surface area for microbial and

enzymatic attack of the digesta.

The existence of foregut fermentation in an animal of this mass is

intriguing. Foregut fermentation can decrease the digestive efficiency of a

small selective browser, because the high quality proteins and carbohydrates

of plant cell contents can be microbially fermented before the host absorbs

these components in the intestine (Demment and Van Soest 1983, Demment and

Van Soest 1985, Parra 1978). Additionally, fermentation products from

microbial metabolism, mainly CO2 and methane, can be a substantial energy

loss for a host with high mass-specific energy requirements.

Most other avian herbivores on similar plant diets increase their rate of

energy and nutrient extraction by trading fiber digestion for higher intake

rates. In the hoatzin, foregut fermentation results in efficient digestion of

both plant cell contents and cell walls. The question is then why hoatzins do

not have higher intake rates (with the concomitant reduced fiber use) or why

do they bother to digest fiber if other avian herbivores on similar diets do not.

One reason may be the detoxification of plant secondary compounds. Foregut

fermentation is advantageous as a detoxification mechanism because most









secondary compounds are readily degraded by gut bacteria (Barry and Blaney

1987, Freeland and Janzen 1974). Moreover, the internal lining of

fermentative structures in ruminants is impermeable to most types of

secondary compounds, allowing detoxification before the food reaches the

absorptive tissues of the gastrointestinal tract (Freeland and Janzen 1974).

Hoatzins in the Venezuelan llanos feed on plants that seem to have a wide

array of secondary compounds (pers. obs.). Unfortunately, almost nothing is

known about the secondary compound biochemistry of most tropical plants

found in the hoatzin's diet. Another possible explanation may be the microbial

synthesis of essential amino acids and vitamins that provide the host with

balanced nutrition from an otherwise unbalanced diet (Purser 1970, Van Soest

1982).












Table 3.1) Chemical composition of the three experimental diets and the
estimated composition of the natural diet of hoatzins at the study area. Values
presented as percentage of dry matter (% DM) except for energy content
(KJ/gDM). Natural diet values from (Strahl and Parra 1985).


Experimental diet


composition
Diet B


Diet C


Natural diet
composition


Dry matter
(DM 65C)

Organic matter
(DM 100*C ash)

Energy
(KJ/gDM)

Nitrogen


Cell wall
(NDF)

Cell contents
(100-NDF)

Acid detergent fiber
(ADF)

Cellulose


Hemicellulose
(NDF-ADF)

Lignin


Lignin/cellulose
ratio

Hemicellulose/cellulose
ratio


Diet A


41.4


84.9


17.8


4.1


29.2


70.8


22.0


14.5


7.2


6.6


0.46


0.50


42.2


85.7


18.3


4.6


35.7


64.3


21.3


13.2


14.4


7.4


0.56


1.09


57.0


87.7


17.9


2.9


39.0


61.0


18.7


13.9


20.3


4.6


0.33


1.46


30.4


70.1


2.9


47.8


52.2


34.3


17.4


13.5


10.5


0.60


0.78















Table 3.2: Mean intake, fecal excretion rates and average body
hoatzins for the three experimental diets. Standard deviations
parentheses.


mass of captive
are in


Diet A Diet B Diet C

Dry matter intake 43.6 40.8 35.2
(gDM/day) (0.6) (5.9) (5.4)

Dry matter intake 64.9 66.4 60.6
(gDM/kg body mass day) (2.6) (0.4) (13.9)

Gross energy intake 694.0 738.2 616.2
(KJ/day) (10.9) (106.2) (97.5)

Gross energy intake 1032.6 1200.9 1059.4
(KJ/kg body mass day) (43.4) (6.8) (240.5)

DM fecal excretion 13.7 8.5 9.71
(gDM/day) (2.3) (0.1) (1.3)

Average body mass 672.5 615.0 594.6
(g) (17.7) (91.9) (100.6)

number of birds 2 2 5












Table 3.3) Apparent digestibilities, metabolizable energy coefficients (MEC)
and in vitro organic matter digestibilities (IVOMD) of the three experimental
diets. Standard deviations in parentheses. Significant differences between
Diet A and C are marked by a (X), differences between Diet A and Diet B diets
are marked by a (Y) and differences between Diet B and Diet C diets are marked
by a (Z). (Unpaired t-tests, P < 0.05).

Digestibilities (%)


lipt A


Diet R


nDit C


Dry matter
(DM 650C)


Organic matter
(DM 100C ash)

Nitrogen
(without uric acid extraction)

Nitrogen
(with uric acid extraction)

Cell wall
(NDF)

Cell contents
(100-NDF)


Acid detergent fiber
(ADF)

Cellulose


Hemicellulose
(NDF-ADF)


61
(14)


negative


Lignin


MEC


In vitro organic
matter digestibility


72
(5)

9
(0.3)

12
(0.1)

35
(7)

77
(3)


74
(3)

9
(1)

13
(1)

71
(4)

74
(2)

58
(6)

63
(8)


(X)(Z)


(Z)


79
(3)

80
(3)

10
(1)

13
(1)

41
(14)

87
(2)

31
(5)

47
(2)

68
(4)

67
(11)

80
(3)

80


(X)(Y)(Z)


39
(11)


(X)(Z)


if- cc..















CHAPTER 4
RETENTION TIMES AND PARTICLE PASSAGE RATES OF DIGESTA MARKERS IN THE
HOATZIN GUT



Introduction

The passage rates of birds are generally short, with a few common

trends (Karasov 1990, Warner 1981b). In particular, herbivorous birds have

slower passage rates than other birds of similar size (Warner 1981b). Within

this group, birds with active caecal fermentation (e.g., grouse and ptarmigan)

have longer retention times than herbivorous birds with little or no

fermentation (e.g., geese, emu) (Karasov 1990, Warner 1981b).

Only a few studies have analyzed selective particle retention in birds;

Some birds pass refractory solids faster than the more digestible liquids

(Bjornhag and Sperber 1977, Warner 1981b). Furthermore, herbivorous birds

with caecal fermentation seem to retain the liquid phase almost twice as long

in the caeca than the solid phase (Gasaway et al. 1975). Foregut fermenting

mammals generally show the opposite trend: solids are retained longer than

liquids (Warner 1981a, Warner 1981b). The particular gastrointestinal

morphology of foregut fermenting mammals results in selective particle

retention, with smaller particles and liquids passing faster along the gut than

larger particles. This is an important trait for foregut fermenting mammals

because larger fiber particles are retained in the fermentation chambers for

further digestion, while the digestible liquids and small solid particles are

passed to the lower gut where assimilation takes place. This selective particle










retention seems to occur at the foregut chambers and not at other gut sites

(Grovum and Williams 1973, Warner 1981a).

The hoatzin is the only known obligate avian folivore with a well-

developed foregut fermentation system (Grajal et al. 1989). Most of the

fermentation takes place in the anterior portion of the gut (i.e., crop and

caudal esophagus). The capacity of these foregut sections is approximately

10% of the adult hoatzin's body mass (Grajal et al. 1989, see Chapter 2). The

morphology of these organs is unique. The crop is divided by a fold into two

connected chambers, while the caudal esophagus is heavily sacculated, with

multiple semilunar folds and constrictions. Particle size is significantly

reduced at these fermentation sites (Grajal et al. 1989, see Chapter 2),

suggesting a combined abrasive action by the internal lining of the muscular

crop and intense microbial attack on the fiber components of the diet. The

proventriculus and gizzard are much reduced in size. Some additional

fermentation takes place in the small paired caeca.

In the hoatzin, digesta dynamics are probably more similar to the trends

in foregut fermenting mammals than trends in other herbivorous birds.

Differential passage of solid and liquid digesta or small and large particles are

important attributes of foregut fermentation digestive systems, because these

traits can increase nutrient and energy extraction from a herbivorous diet.

Therefore, this study examined differential particle passage rates and

retention times of digesta in the hoatzin.










Materials and Methods

The experiments were performed with two captive hoatzins in 1988 and

five captive hoatzins in 1989. All birds had ad libitum access to a diet

consisting of romaine lettuce with a powdered mix of ground alfalfa hay

pellets, ground Timothy grass hay and ground roasted soybeans (see Chapter

3). Additionally, fresh young shoots of plants in their natural diet (e.g.,

Enterolobium cyclocarpum. Pithecellobium saman, Guazuma ulmifolia and

Phthirusa cf. venezuelensis) were offered ad libitum twice daily. The birds

were housed individually indoors in adjacent 1 x 1 x 2 m cages with removable

floor trays for quantitative collection of feces. All trials started in the

morning (730-1000 h) just before the routine morning feeding.

Hoatzins were force-fed a gel capsule with markers as a single pulse

dose (Warner 1981b). Two markers, Cr-EDTA and ytterbium mordanted on

fiber, were used to measure differential passage rates of liquid and fiber

phases of the digesta, respectively. Plastic markers were used to measure

differential particle size passage rates along the gut. Plastic markers consisted

of two sizes (1mm2 and 4mm2) of squares of brightly-colored pink and orange

commercial flagging tape. The specific gravity of this flagging tape was 1.01,

similar to the specific gravity of wet fiber plant fractions (Warner 1981b).

The liquid phase was marked with a Cr-EDTA (chromium ethylene-diamine

tetra-acetic acid) complex prepared following the procedure of Binnerts et al.

(1968). The Cr concentration in fecal excretion was measured using atomic

absorption spectrometry (Williams et al. 1962). Fiber from a mature grass hay

was marked with ytterbium (Yb) oxide, using the procedure of Ellis et al.

(1982). Before marking, the fiber was sieved to lmm2 particles and purified

using neutral detergent fiber extraction (Goering and Van Soest 1970). After

the NDF extraction, the fiber was dried over absorbent filter paper at ambient









temperature for 24 h and then stored in a desiccator until needed for marking.

The ytterbium concentration was measured by atomic absorption spectrometry

using a nitrous oxide acetylene flame.

The doses for the trials were estimated at approximately 16 mg Cr, 15 mg

Yb (mordanted to 2 g fiber), 500 1mm2 plastic markers and 200 4mm2 plastic

markers per bird. The exact dose of Cr and Yb for each animal was not known

because the sizes of the gel capsules were slightly different, and some material

remained in the bill and mouth of the birds. Therefore, recovery of these

markers was calculated based on estimated doses. Feces were quantitatively

collected on thick black plastic film liners on the cage's floor tray. The black

plastic liners on which feces were collected provided a color contrast that

enhanced the recovery of the tiny 1mm2 plastic markers. Feces were collected

at regular intervals of 2-3 h the first day, 4-5 h the second day, 8 h the third

day and 10 h the fourth day. For each batch of feces, the total number by size-

class of plastic markers were counted and removed. All feces from each

collection were removed from the plastic liners, dried in a forced-air oven at

65 C to constant mass and kept individually in plastic bags until analyzed.

Feces from each sample were ashed at 600 C. Some samples were too

small (e.g., from the early feces), and were pooled for a larger sample

corresponding to the later sample time. Both Cr and Yb from the ash were

simultaneously extracted based on the procedure to extract Cr by Christian and

Coup (1954) and modified for simultaneous extraction of Cr and Yb by Siddons

et al. (1985). The Yb extraction was similar to that of Siddons et al. (1985), but

instead of centrifugation after ash extraction, the extract was filtered through

WhatmanTm N 41 filter paper and stored until analyzed. Standards for both Cr

and Yb were prepared using feces from previous experiments that contained

no Cr or Yb (Christian and Coup 1954, Siddons et al. 1985).









Transit time was the time of first appearance of a marker in the feces.

Mean retention times were calculated as MRT = Ymiti/mmi, where mi is the

amount of marker excreted per unit dry matter feces at the ith defecation at

time ti after dosing (Blaxter et al. 1956). This method makes no assumptions

about the frequency distribution of dye excretion (Warner 1981b), an

important advantage because fecal excretion curves are generally not

uniform in shape. Although a high proportion of marker recovery is

desirable, this formula is advantageous because it does not depend on the total

recovery of ingested marker, but rather on the amount excreted.

Trials in 1988 measured retention times using plastic markers. In 1989,

one trial consisted of a pulse dose of both plastic markers and Cr-EDTA to five

birds (18 Sept.) and another trial consisted of Cr-EDTA and Yb markers (2 Oct.).

All other markers were given once to each bird.

All statistical analyses were two-tailed with an alpha level of 0.05. Each

individual combination of bird-marker was considered as the experimental

unit. Mean retention time and transit time were transformed to the inverse of

the square root to reduce heteroscedasticity (heterogeneous variances) in the

ANOVA tests. Similarly, percentage recovery was transformed to the square

root (Sokal and Rohlf 1981). Post hoc multiple comparisons were performed

using the Games-Howell test (Games and Howell 1976). This test is a

conservative and robust procedure under unequal sample sizes, heterogeneous

variances and violations of normality (Jaccard et al. 1984). Standard deviations

are shown in parentheses.










Results



Fecal excretion rates

Fecal excretion rates of hoatzins were relatively constant. No daily

fluctuations were evident at the scale used. Nevertheless, handling of birds

during the administration of the marker pulse lowered the fecal excretion rate

for the first 4-6 h of the trials. This initial depression of the fecal excretion

rate was present in most birds, but its length was different for each individual

(Fig. 4.1). Solid markers first appeared well after the fecal excretion rates

stabilized. The effect of initial low fecal excretion rates on liquid transit time

remains unknown.



Marker recovery

The percentage recovery for all four markers was high (Table 4.1). No

difference was found in mean percentage recovery among markers (One

factor ANOVA, F = 0.187, d.f. = 3, P = 0.91). The Cr-EDTA concentration was below

the sensitivity of the atomic absorption spectrometer in some samples, so a

higher dose of this marker or larger intervals between feces collections would

have been desirable. These results, however, did not affect the outcome of the

experiment.



Mean retention times

One bird in 1989 (Y-chick) had longer retention times for all markers

(One factor ANOVA, F = 2.2, d.f. = 5, P = 0.12, Games-Howell test). This bird was a

growing fledging that developed flight feathers and increased body mass

during the experiments. This was the only fledging bird in the group, so it

was not included in further calculations.









Marker concentration curves were clearly skewed. Mean retention

times for each marker are shown in Table 4.1. No difference in mean

retention time was found among individual hoatzins, but the difference was

highly significant among markers (Two factor ANOVA, F = 2.22, d.f. = 4, P = 0.27

for birds and F = 14.07, d.f. = 3, P = 0.03 for markers). The interactions between

bird and marker were not significant (F = 0.6, d.f. = 7, P = 0.74). The liquid

marker (Cr-EDTA) had a shorter retention time than all solid markers (Games-

Howell test). Mean retention time was shortest for the liquid marker, followed

by Yb and 1mm2 plastic particles. The longest retention time was recorded for

4mm2 plastic particles; more than twice the liquid retention time (Table 4.1).

Small particles passed faster than large particles. The larger variation of

mean retention times for the plastic markers obscured some of these

differences. For example, no significant differences could be detected between

1mm2 and 4mm2 plastic markers. The effect of particle size was evident when

significant differences in mean retention times were found between Yb and

4mm2 markers (Games-Howell test). Interestingly, there was no difference

between Yb and 1mm2, as expected because the fiber marked with Yb was

sieved to 1mm2 particles.



Transit times

Transit times were fast for Cr-EDTA (Table 4.1). This marker appeared in

the first collection after the single pulse dose, so the smallest detectable transit

time was on average 2.57h (- 0.54). Transit times were similar for Yb and

1mm2 plastic particles. The longest (and most variable) transit time was

recorded for 4mm2 plastic particles (Table 4.1). Overall, transit times were

significantly among markers (One factor ANOVA, F = 32.93, d.f. = 3, P = 0.0001),










but multiple comparisons (Games-Howell test) show that the only significant

difference was between Cr-EDTA and the particulate markers.



Discussion

All markers seemed satisfactory to measure mean retention times as well

as selective particle size retention. The marker recovery rate was relatively

high (Sklan et al. 1975), and the markers did not seem to affect the birds. The

fact that mean retention times of 1mm2 plastic and 1mm2 Yb particles were not

different, demonstrates that small plastic markers can provide a quick method

to estimate passage rates. Some of the drawbacks of plastic markers include

the intensive labor required to obtain a satisfactory recovery. Additionally,

the variability of mean retention times measured with plastic markers seems

to be inherently higher than that measured with chemical markers. This

variability is probably related to the behavior of the plastic markers in the

gut. Although food particles are chemically and physically attacked by

microbes, enzymes, and the grinding action at the crop, esophagus and

gizzard, plastic markers remain completely inert and do not change in

composition or size.

Mean retention times in hoatzins are among the longest ever recorded

for a herbivorous bird and are similar to those of ruminants and arboreal

folivores (Karasov 1990, Warner 1981b). The short transit time of the Cr-EDTA

suggests a fast movement of liquids through the hoatzin's gut. Fast liquid

transit times have been reported for other herbivorous birds (Bjornhag 1989,

Clemens et al. 1975a, Duke 1988) and herbivorous mammals (Clemens et al.

1975b, Warner 1981b). The effect of the initial low defecation rates produced

by the handling of birds upon liquid transit time could not be discerned with










the methodology used. Since hoatzins rarely drink, it was not possible to

administer the liquid marker without handling the animals.

Liquids passed faster than solid particles. Selective retention of solid

particles against the liquid phase has been reported in mammals with foregut

fermentation (Grovum and Williams 1973, Hume and Dellow 1980, Warner

1981a, Warner 1981b). Within this group, macropodid marsupials can separate

the two phases to a higher degree than can ruminants (Hume and Dellow 1980,

Warner 1981a). The separation of liquid and solid markers in the hoatzin does

not reach the extent seen in macropod mammals; it is more similar to marker

separations seen in ruminants. Faster passage of the liquid fraction can be

important for a small vertebrate with foregut fermentation, such as the

hoatzin. This pattern allows the passage of the more digestible substrates to

the lower gut where enzymatic digestion and absorption take place.

Meanwhile, increased residence of larger particles in the foregut allows more

time for microbial attack of the cell wall and enough time for microbial

population turnover. One of the possible drawbacks of a fast liquid passage is

that fermentative microbes in solution would be rapidly washed away and

digested in the lower gut. This may result in a detrimental reduction of the

time allowed for microbial population turnover. Microscopic observations of

the crop contents of hoatzins indicate that most bacteria are firmly attached to

the cell walls (F. Michelangeli, pers comm.). As a result, most bacteria seem to

avoid being washed down the gut with the liquid phase.

In hoatzins, larger particles are retained longer than smaller ones.

Although this is common in foregut fermenting mammals (e.g., Blaxter et al.

1956, Warner 1981b), it is rarely seen in birds. For example, some herbivorous

birds show no selective particle retention (Herd and Dawson 1984), while most

herbivorous and frugivorous birds show the opposite trend larger particles









are excreted significantly faster than the smaller particles and the liquid

fraction (BjOnmhag and Sperber 1977, Gasaway et al. 1975, Levey and Grajal

1991, Warner 1981b). This pattern is especially prevalent in birds with caecal

fermentation (e.g., family Tetraonidae), in which the caeca seem to selectively

retain small particles and liquids and reject the undigestible larger food

particles (Bjmrnhag 1989, Bjmrnhag and Sperber 1977). The differential

particle selectivity probably increases the rate of energy intake in birds with

caecal fermentation (Gasaway et al. 1975, Remington 1989). Although a similar

mechanism may occur in the caeca of the hoatzin, its effect on particle

retention through the gut is probably negligible.

Long retention times in the hoatzin increase the nutritional use of both

cell contents and cell walls. In contrast, most herbivorous birds use the

readily digestible nutrients of plant cell contents at the expense of more

thorough nutritional use of cell walls. This is important, because microbial

fermentation of readily digestible cell contents can insert an additional

trophic level between the food and the host. Thus, microbial metabolic losses

(e.g., methane, CO2 and heat) can decrease the overall energy available to a

herbivorous bird. These losses can be significant for a small vertebrate. In

the hoatzin, these microbial metabolic losses may be offset by other

nutritional benefits such as an increased nutritional use of cell walls, VFA

(volatile fatty acid) production and detoxification of secondary compounds.

Another benefit of differential passage rates is that the more refractory

cell wall fraction remains longer in the crop and esophagus, where microbial

fermentation takes place. Indeed, hoatzins digest fiber components to an

extent rarely seen in birds. On an experimental diet with 39% neutral

detergent fiber (NDF) and 3% nitrogen, hoatzins digested 63% cellulose, 78%

hemicellulose and 71% NDF (see Chapter 3). Moreover, the high VFA










production rate at the crop and esophagus provides energy for about 60% of

the basal rate of metabolism of adult hoatzins (see Chapters 5 and 6). These

high fiber digestibilities probably result from the combined effect of long

retention times, intense microbial fermentation, and selection of a highly

fermentable diet. Indeed, hoatzins in their natural habitat select plant parts

that are low in cell wall, lignin and high in protein and water content (Grajal

et al. 1989). The selectivity of high quality plant parts is possible because

hoatzins fly and therefore can track resources that are patchy in space and

time.

These patterns of differential passage rates optimize the energy and

nutrient extraction from the hoatzin's leafy diet (Grajal et al. 1989, also see

Chapter 3). Foregut fermentation is possible because hoatzins select particular

plant parts that are low in fiber and high in protein, maintain a high rate of

microbial fermentation, and have very long retention times of the digesta.

Foregut fermentation and the long digesta retention times in the hoatzin

result in a unique evolutionary adaptation that provides an efficient use of a

herbivorous diet by a flying bird.









0.4




S0.3 -

o0






LL.
0.1
l 0.1*J---------
0 20 40 60 80 100

hours




Figure 4.1) Average change in fecal excretion rate (in grams dry
matter/hour) over time. Fecal excretion rates were low at the start and then
remained relatively constant throughout the experiments. Bars represent
standard deviations (n = 4).














Table 4.1) Average mean retention time (MRT), transit time (TT), percent
recovery and sample size for liquid and solid markers. Liquid marker was Cr-
EDTA (chromium ethylene-diamine tetra acetic acid). Solid markers were
ytterbium (Yb) mordanted to lmm2 particles of hay fiber and two sizes (1mm2
and 4mm2) of cuts of commercial plastic flagging tape. All markers were
orally given as a single pulse dose in a gel capsule. Standard deviations are
shown in parentheses. Significant differences (P < 0.05) among markers are
denoted by an asterisc.


Liquid
Cr-EDTA


Solid
Yb 1mm2


4mm2


MRT (hours) 17.9 24.4 33.3 44.4 (*)
(3.4) (2.3) (16.8) (15.4)

TT (hours) 2.6 8.3 7.5 10.7 (*)
(0.5) (2.9) (1.0) (4.6)

Recovery (%) 66.7 73.0 70.15 67.3
(17.6) (15.9) (8.4) (4.3)

(n) 7 4 4 3















CHAPTER 5
FERMENTATION RATE IN THE CROP AND ESOPHAGUS OF THE HOATZIN



Introduction

The hoatzin is the only known bird with an active foregut fermentation

digestive system and the only instance of such a digestive system outside the

mammals (Grajal et al. 1989). This distant relative of the cuckoos (Sibley and

Ahlquist 1973, Sibley et al. 1988) inhabits gallery forests, forest swamps and

oxbow lakes of the Orinoco and Amazon drainages (Strahl 1988). Unusual

characteristics such as functional wing claws in the first and second digits of

the wings of young hoatzins were first seen as evidence of a "missing link"

between the ancient Archaeopteryx and modem birds (Banzhaf 1929, Garrod

1879, Huxley 1898, Parker 1891).

The nutritional ecology and digestive physiology received little

attention until recently (Grajal et al. 1989, Strahl 1988, Strahl and Schmitz

1990). Hoatzins are one of the few avian obligate folivores: up to 87% of their

diet is composed of green leaves of plants (Grajal et al. 1989). Nutritional

analyses have shown that preferred plant parts (shoots, buds and new leaves)

are lower in fiber content and higher in nitrogen and water than non-

preferred parts (Grajal et al. 1989, Strahl 1985).

The structure and function of the gastrointestinal tract of the hoatzin is

unique. Although early descriptions included the disproportion in the sizes of

the crop and the proventriculus (or gastric stomach), no fermentation activity

was suggested (Boker 1929, Gadow 1891). In fact, the greatly enlarged crop and

caudal esophagus have a relative gut capacity similar to the fermentative










structures of mammalian herbivores (Demment and Van Soest 1985, Parra

1978).

In herbivores with gut fermentation chambers, volatile fatty acids

(VFA) represent the most important microbial by-products in terms of energy

benefits to the host. In the hoatzin, VFA concentrations and pH levels in the

anterior part of the gastrointestinal tract are comparable to mammals with

foregut fermentation (Grajal et al. 1989). Additional fermentation takes place

in the paired caeca of the lower gut. Microbial densities in the crop and caudal

esophagus are the same order of magnitude as in ruminants and other

mammals with foregut fermentation (Grajal et al. 1989). At least three species

of protozoans have been found at the foregut sites (F. Michelangeli, pers.

comm.). Studies with captive hoatzins showed that fiber digestibilities were

among the highest ever recorded for herbivorous birds under equivalent diets

(Grajal et al. 1989, see also Chapter 3).

Foregut fermentation in hoatzins is a theoretical anomaly. Present

models predict a limit of 6-10 kg of body mass below which foregut

fermentation cannot fulfill the energetic needs of an endotherm (Demment

and Van Soest 1985, Parra 1978). At this mass, the predicted total metabolic

requirements surpass the rate of energy available from plant fiber

fermentation. The small mass of the hoatzin (650 g) is an order of magnitude

lower than the predicted minimum body mass for a foregut fermenter. A study

of the foregut fermentation system of hoatzins can enhance the

understanding of the physiological limits of foregut fermentation.

This study examines the contribution of foregut microbial fermentation

to the metabolism of hoatzins. The basal rate of metabolism in hoatzins is

relatively low for a bird of its size (see Chapter 6) -about 68% of the expected

value from allometric models (Kleiber 1961, McNab 1988). A measurement of










the rate of VFA production can estimate the contribution of fermentation to

the overall metabolic expenditure of live hoatzins. One experimental approach

to determine this contribution is to measure in vitro fermentation activity.

Standard techniques have been developed to study domestic herbivores (i.e.,

cows and sheep) (Alexander and McGowan 1966, Tilley and Terry 1963). These

techniques, however, can not be used in small (< 3kg) herbivores, because the

small sample size is a limitation. Therefore, a "miniature" in vitro technique

was designed to culture fermenting microbes from a small sample (<150g fresh

mass compared to 2,500 g for the standard technique).

The objectives of this study were a) to determine the rate and extent of

microbial fermentation in the hoatzin using in vitro techniques, b) to

determine the energy contribution of microbial fermentation to the

metabolism of the live hoatzin, c) to compare the fiber fermentation

capabilities of microbial inocula from hoatzin crop and cow rumen incubated

under the same in vitro conditions and d) to determine the replicability and

reliability of the "miniature" in vitro technique.



Materials and Methods



Laboratory study: Fermentation in captive hoatzins

Hoatzins were captured along the Guairico River (670 28' W, 80 33' N), a

northern affluent of the Orinoco River in central Venezuela. The birds were

kept in outdoor aviaries at Fundo Pecuario Masaguaral, a private ranch and

biological station in the central llanos of Venezuela. The birds were

acclimated to captivity by a slow and progressive change from their natural

diet to an artificial diet. The latter was a "salad" of romaine lettuce, sprinkled

with a powdered mix of ground alfalfa hay pellets, ground Timothy grass hay,










ground roasted soybeans and a vitamin-mineral supplement. The nutritional

composition of this diet was (on a dry matter basis): 84.9% organic matter, 4.1%

nitrogen, 29.2% neutral detergent fiber (NDF), 22% acid detergent fiber (ADF),

14.5% cellulose, 7.2% hemicellulose, 6.6% lignin.

After more than 60 days of acclimation, the birds were moved to large

outdoor aviaries at the Animal Production Institute of the Universidad Central

de Venezuela campus at Maracay. The acclimated hoatzins were used for other

studies on passage rates and in vivo digestibilities (see Chapter 3). One bird

was used 10 December 1986, and two birds were used 18 July 1988. The birds

were killed and the gut rapidly removed, with the foregut sections (crop and

caudal esophagus) separated by string knots. These fermenting sections were

weighed, and the contents were rapidly passed to a previously weighed baby

food blender vase with a continuous flow of pre-heated CO2 The container

was kept closed in a water bath at 390C. A 1:5 (w/w) dilution was made by

adding the necessary volume of pre-heated buffered artificial saliva

(McDougall 1948) to the vase. A 2.5% of a solution of 2.64% ammonium sulfate

solution was added, as a source of nitrogen for fermenting bacteria.

The blender was turned on at high speed for 5 seconds and then stopped

for 10 seconds. This sequence was repeated three times to allow the separation

of fermenting bacteria from the substrate. This solution was then filtered in a

special anaerobic filter assemblage (Fig. 5.1) through a double layer of

cheesecloth. This filter was designed to maintain anaerobic and isothermic

conditions and to achieve a high filtering efficiency from a small sample.

After filtration, the inoculum was gently shaken every other minute and kept

under a continuous flow of pre-heated CO2 in a water bath at 390C during

inoculation.










The fermenting substrate for the laboratory in vitro experiments was a

finely (1mm) ground sample of alfalfa pellets (2.9% nitrogen and 36.1% NDF).

Approximately 100 mg of the sample were added to 20 ml Hungate tubes with

screw-on caps with internal rubber stoppers. To avoid caking of the dry

substrate during inoculation, two drops of artificial saliva were used to moisten

the alfalfa substrate 30 min before inoculation. The Hungate tubes were

individually gassed with C02 before inoculation and then 10 ml of the

inoculum were added to each tube with a repeating inoculation syringe. The

mixture was gently stirred and thoroughly flushed with a flow of C02 and then

tightly capped. Blank samples consisted of 10 ml of inoculum in tubes without

substrate. The tubes were gently shaken by hand 4 times every 24 h, and each

time the screw-on caps were slightly unscrewed to release the gas buildup.

At each time period of 0, 1, 2, 2.5, 4, 5, 6, 12, 18, 24, 48 and 72 h after

inoculation, two tubes and a blank were taken from the incubation and the

fermentation stopped. Fermentation in tubes corresponding to 0, 1, 2, 2.5, 4

and 5 h was stopped with approximately 1 ml of concentrated sulfuric acid.

These tubes were used to estimate the production rate of VFA by extrapolation

from a linear regression model (zero-time method) (Carroll and Hungate 1954).

Daily energy available from VFA (in KJ/day) was calculated using the model

by Prins et al. (1984) as



Daily energy available = W x FC x DM x Y x Ex (1/0.6) x 24 x 10-4



where W is the body mass in kg, FC is the mass of fermentation contents as

percentage of body mass in kg, DM is the dry matter content of the

fermentation organ contents, Y is the invitro fermentation rate in mmol

VFA/gDM h, E is the energy equivalent of the mix of acetic, propionic, butyric










and other VFA found in the individual birds, using the energy equivalents of

VFA given by (Blaxter 1962) and 0.6 is the average utilization efficiency of the

metabolizable energy from VFA, assuming 85% efficiency for maintenance

and 35% efficiency for other "production" activities, such as mating, stress or

social interactions (Blaxter 1962).

The fermentation in the other tubes was stopped with 2 ml of toluene.

These tubes were analyzed for in vitro cell wall digestibility (Van Soest 1982).

Fiber content (as neutral detergent fiber) of the alfalfa substrate and amount

of fiber digested by crop microbes was measured using standard detergent

fiber analysis (Goering and Van Soest 1970). Fiber digestibility was estimated

as (dry matter mass of fiber in the tube before fermentation dry matter mass

of fiber in the tube after fermentation) x 100/dry matter mass of fiber in the

tube before fermentation.



Field study: Fermentation in wild hoatzins

This study was designed to measure the VFA production rate of wild

hoatzins eating their natural diet. Between 11-12 July 1990, three adult

hoatzins were shot at Hato Piflero (680 04' W, 80 56' N), a cattle ranch and

biological station on the Cojedes River, another northern affluent of the

Orinoco River. The dead birds were immediately taken to a close-by field

laboratory and the foregut contents removed. The time between death and the

start of inoculation was no more than 35 min. The same buffering saliva as in

the laboratory study was used for a 1:5 (w/w) dilution. Most of the separation

and filtration procedures were similar to the laboratory study, but instead of a

blender, the container with the gut contents and the saliva was vigorously

shaken by hand for 20-40 seconds to separate the attached bacteria from the

surface of the cell walls.










The fermenting substrate for the field in vitro experiments was a finely

(1 mm) ground mix of leaves representing the natural diet of hoatzins at the

study site. The substrate mix was composed of 12.8% Guazuma ulmifolia leaves,

7.2% f.. ulmifolia buds, 6.8% Phthirusa cf. orinocensis. 27.8 % Enterolobium

cyclocarpum, and 45.3% Lonchocarpus cruciarubierae. The nutritional

composition of this combination was 92.9% organic matter, 2.6% nitrogen,

39.3% NDF, 24.5% ADF, 13.7% cellulose, 15.3 hemicellulose, 10.8% lignin. This

diet was taxonomically similar to the estimated average natural diet

composition eaten by wild hoatzins (Strahl and Parra 1985). Approximately

100 mg of the sample was added to 20 ml Hungate tubes with screw-on caps

with internal rubber stoppers. Small (5ml) syringes were punched on the

rubber caps every hour to alleviate the gas pressure buildup inside the tubes

once fermentation started. Inoculation proceeded as in the laboratory study.

At 0, 1, 2, 2.5, 4 and 5 h after inoculation, two tubes and a blank were taken

from the incubation and the fermentation stopped with approximately 1 ml of

a 0.5 M solution of sulfuric acid.



Comparison of hoatzin and cow in vitro digestibilities

This study compared the fiber fermenting capabilities of hoatzin crop

contents and cow rumen contents. The hoatzin crop contents were treated in

the same way as in the laboratory study. The ruminal contents were extracted

from the ventral part of the rumen of a fistulated Holstein cow eating mature

grass hay. The ruminal contents were kept under isothermic and anaerobic

conditions until filtering. After removing part of the ruminal liquid through

a double layer of cheese cloth, the contents were treated in the same way as

with the hoatzin crop contents. A dried, ground sample of alfalfa hay (3.1%

nitrogen and 39.5% NDF) was used as the fermenting substrate. Again, 100 mg










of the sample were added to 20 ml Hungate tubes with screw-on caps with

internal rubber stoppers. At 2, 6, 12, 24, 48 and 72 h after inoculation, two

tubes and a blank were taken from the incubation and the fermentation

stopped by adding 2 ml of toluene and stored in a refrigerator at -40C until

fiber analyses were done. The rates of in vitro fiber digestibility were

estimated as the slope of regression curves of the natural logarithm (In)

transformation of apparent digestibilities on hours of fermentation.



Comparison of the miniature and standard in vitro techniques

This study was designed to compare the level of fiber digestibility of the

miniature in vitro technique to that of the standard technique. Inoculum

from a fistulated cow eating grass hay was used to compare both in vitro

techniques. The substrate used was a ground mature grass hay (61.65% NDF).

Both experiments consisted of 10 tubes (replicates), and were run

simultaneously. For the miniature technique, 100 mg of the sample were added

to the same Hungate tubes as in the previous experiments. The standard in

vitro technique was a version of the Tilley and Terry method (Tilley and Terry

1963), but without the acid pepsin digestion stage. For the standard technique,

500 mg of the sample were added to 100 ml glass centrifuge tubes, capped with

Bunsen gas release valves. The buffering solution was the same as in the

other experiments. In both in vitro techniques, the dilution was 1:5 (w/w) of

fermentation contents to buffering saliva solution. Both in vitro

fermentations were stopped after 24 h by adding 3 ml of Toluene. Then the

tubes were stored in a refrigerator at -40C until NDF analyses were done.










Gas-liquid chromatography

Volatile fatty acid (VFA) concentrations were measured using gas-liquid

chromatography. Samples were acidified in the field with 2 drops of

concentrated sulfuric acid, frozen in solid C02, and later stored in a freezer at

-10C until analyzed. For the chromatographic analysis, the samples were

thawed and centrifuged at 10,000 r.p.m. for four minutes. A volume of 1.35 ml

of the supernatant was acidified with 0.15 ml of 20% phosphoric acid to a

concentration of 2% (V/V). The sample was injected into a glass column of

10% SP-1000 on 100/120 Chromosorb W/AW (Supelco Inc.). The column was

maintained at 140*C with nitrogen as the carrier gas at 40 ml/min. The

injector was set at 160*C and the flame ionization detector was set at 2000C.



Statistical analyses

All statistical tests were two-tailed with an alpha level of 0.05.

Individual in vitro tubes were considered as the experimental units for the in

vitro regressions. Individual birds were considered as the experimental unit

for comparisons of digestibilities or VFA production rates. Digestibilities

(expressed as percentages) were transformed to their square root and

compared using unpaired t-tests, unless otherwise specified. Standard

deviations are shown in parentheses.














VFA production rate in captive and wild hoatzins

The concentration of VFA increased linearly with incubation time both

in the captive and in the wild hoatzin in vitro fermentations (Fig. 5.2). These

results suggest there was no significant interference by the accumulated end-

products from fermentation during the 5-6 h of the trial. The VFA production

rates were estimated from the slopes of the regression lines of VFA

concentration over time. The overall in vitrl VFA production rate was

significantly higher in wild hoatzins than in captive hoatzins (t-test for the

comparison of two regression coefficients, t = 3.437, d.f. = 47, P = 0.001) (Zar

1984). Additionally, the proportions of individual VFA were different in each

experiment, reflecting differences in the fermentative capabilities of captive

and wild hoatzins, probably due to the different diets of captive and wild

hoatzins, and different substrates in the tubes of each experiment (Blaxter

1962) (Table 5.1). For example, the acetic:propionic ratio was higher in captive

hoatzins than in wild hoatzins (Table 5.1). In fact, production rates of all

individual VFA were higher in the fermentation from wild hoatzins than from

captive hoatzins. The production rate from captive hoatzins was, on average,

53 mmol/kgDM h or 21 KJ/day while for wild hoatzins the average production

rate was 136 mmol/kgDM h or 102 KJ/day. These energy contributions

represent about 14% of the basal rate of metabolism of captive hoatzins and

62% of the basal rate of metabolism of wild hoatzins.









In vitro fiber fermentation and comparison with cow ruminal fermentation

Fiber fermentations by hoatzin and cow inoculum are represented in

Fig. 5.3. The resulting fermentation regressions for the first 24 h of

incubation were: y = 3.841 + 0.017x for hoatzin inoculum (R2 = 0.81, slope

standard error = 0.003) and y = 3.475 + 0.022x for cow inoculum (R2 = 0.63, slope

standard error = 0.007). The fermentation rates (regression coefficients) were

not significantly different, but the intercept was significantly higher for

hoatzin than for cow inocula (t-test for the comparison of intercepts of two

regression lines, t = 2732.4 d.f. = 13, P < 0.001) (Zar 1984). In yjvit fiber

digestibilities stabilized after 48 h of incubation, both for cow rumen contents

and hoatzin crop contents. Total fiber digestibilities after 48 h were not

significantly different between hoatzin and cow inoculum (unpaired t-test).



Comparison of the miniature and standard in vitro techniques

In vitro fiber digestibility measured by the miniature technique was

higher than that measured by the standard technique (unpaired t-test, n = 10,

p = 0.038). Average fiber digestibility was 37.7% (+3.92) for the small tubes and

34.6% (1.97) for the large tubes. Although the average digestibilities were

not very different, the small variability among the large tubes made the

difference significant.













VFA production rate in captive and wild hoatzins

The VFA production rate of captive hoatzins was lower than the range of

VFA production rates reported from mammals with foregut fermentation

(Table 5.2). In wild hoatzins, however, the VFA production rates are similar to

low rates measured for domestic ruminants, but below the rates of ruminants

that feed selectively on young plant parts (concentrate selectors sensu

Hofmann 1989) (Table 5.2). Therefore, VFA production rate in the hoatzin

provides a significant proportion of its energy requirements. In fact, VFA

production rate in hoatzins is the highest recorded for a bird. For example,

VFA production rate in captive hoatzins is higher than the average production

rate for willow ptarmigan, while VFA production rate in wild hoatzins is almost

twice of the maximum rate measured for willow ptarmigan (McBee and West

1969) (Table 5.2).

These fermentation rates are probably the result of the fermentation of

both cell walls (fiber) and cell contents. In the hoatzin, VFA production rate

does not meet all its energy requirements, but the contribution of VFA

production to the hoatzin's metabolism is quite significant. The energy

contribution of fermentation to the metabolism of hoatzins is much larger

than in any other known bird with fiber fermentation (Annison et al. 1968,

Clemens et al. 1975, McBee and West 1969), and similar to that of other foregut

fermenting mammals (Dreschen-Kaden 1977, Hoppe et al. 1983, Parra 1978).

The difference between in vitro fermentation in captive and wild

hoatzins is related to the differences in their diets and probably reflect

different microbial communities. Similar differences have been reported in

other comparisons between captive and wild herbivores (Foley et al. 1989,










Hoppe 1977, Hoppe et al. 1977, Hume 1977). Moreover, different fermentation

rates and proportions of individual VFA generally reflect the different dietary

compositions (Blaxter 1962). For example, diets with a less lignified cell wall

are generally fermented at a faster rate than highly lignified mature plant

material (Smith et al. 1972, Smith et al. 1971). As a consequence, the

proportions of individual VFA are also different. The ratio of acetic to

propionic acid in the fermentation from captive hoatzins was about 5.47, while

this ratio was 1.99 in wild hoatzins. The higher proportion of propionic acid in

the fermentation from wild hoatzins is an important contribution to the

energy balance of wild hoatzins, because propionic acid is one of the main

precursors of glucose (Blaxter 1962, Miller 1979). The composition and high

production rate of VFA in wild hoatzins reflect not only a more fermentable

diet but possibly a better developed microbial community.



In vitro fiber fermentation

The fiber fermenting capabilities of the hoatzin microbial community

are remarkable, especially for a small herbivore. Fiber fermentation from

hoatzin fermentation contents was not different during the first 48 h than

from cow fermentation contents. The extrapolation from the regression

models, however, indicates that fiber fermentation from hoatzin fermentation

contents started at a higher level. These differences probably reflect the

effect of the different diets upon the composition of the microbial community

of the two species. The offered substrate was more similar to the diet of captive

hoatzins than to the diet of the fistulated cow. Therefore, the hoatzin inoculum

may have adapted to a familiar substrate more rapidly and fermented it more

rapidly than the cow inoculum did. After 48 h, both fiber digestibilities










become nearly asymptotic, probably due to substrate disappearance and

accumulation of unusable end-products.



Miniature and standard in vitro techniques

Although the difference in digestibilities measured in small and large

tubes was not great, it was significant. The miniature in itro technique

overestimated fiber digestibility when compared to the standard technique.

The difference between small and large tubes is not easy to explain. One

methodological difference is that the large tubes had Bunsen gas release

valves, while the small tubes were sealed and the gas buildup pressure was

released only three times during the 24 h fermentation. As a consequence, it

is possible that the accumulation of gas pressure in the miniature technique

affected in vitro fiber fermentation. The use of small (5ml) syringes punched

in the rubber caps can alleviate the gas pressure buildup inside the small

tubes.

This study demonstrated that the miniature in ito technique can be

performed with acceptable accuracy, permitting comparisons using the same

miniature technique without a loss of precision or repeatability. Comparisons

with the standard technique, however, should take into consideration that the

miniature technique seems to marginally overestimate fiber digestibilities.



Conclusions

Microbial foregut fermentation in the hoatzin provides a unique

nutritional use of cell wall and cell contents for a bird. Cell wall fermentation

has two main advantages: first it makes the highly digestible cell contents

available to the host and the microbial community. Second, cell wall microbial

fermentation produces VFA that in turn make a significant contribution to the










energy requirements of the hoatzin. Moreover, in vitro estimates of the

contribution of fermentation to the metabolism of the whole animal usually

underestimate this contribution (Blaxter 1962). In fact, the energy

contribution of fermentation in the live hoatzin is probably higher, because

additional fermentation takes place in the paired caeca (Grajal et al. 1989) and

optimal conditions for fermentation are better maintained in the live animal.

Fermentation in the hoatzin is unlike that of any other herbivorous

bird, and more similar to that of selective browsing ruminants. As a

consequence, the fermentation rate from hoatzin crop microbial communities

is higher than in any other bird and more similar to that of ruminants with

low fermentation rates. Fermentation in hoatzins optimizes the nutritional use

of both the fiber and cell content fractions of their leafy diet. Additionally,

VFA production rate can sustain a large portion of the energy requirements of

the hoatzin. Finally, foregut microbial fermentation in the hoatzin may

provide other nutritional advantages, such as microbial production of

vitamins and amino acids and the detoxification of plant secondary compounds.

These latter aspects need further study.








Inoculating syringe


Figure 5.1) Diagram of in vitro filter system used to achieve a high filtering
efficiency while keeping anaerobic and isothermic conditions. When the
filter clogged, the lid was used as a piston to press the cheesecloth with the
contents and increase the amount of filtrate.


CC02
-=-





73



80

S 70 o Wild hoatzins
Captive hoatzins
7 60-
E

c 50 o
0
S40-

C 30
S 20

5.L 20

10
0 1 2 3 4 5 6
Incubation time (hours)





Figure 5.2) Change in VFA concentration (in mmol/l of fermentation
contents) with time from the crop and caudal esophagus inoculum of captive
and wild hoatzins.









90

80

70

a 60


~0 l Hoatzin inoculum

0o 40 /- Cow inoculum

30 ,
0 20 40 60 80
Incubation time (hours)



Figure 5.3) Cell wall fermentation capabilities of cow ruminal microbial
inoculum and hoatzin crop microbial inoculum using a miniature in vitro
technique. Curves represent logarithmic regression models: y = 9.2 + 38.7 log x
(R2 = 0.92) for cow inoculum and y = 33.7 + 25.5 log x (R2 = 0.91) for hoatzin
inoculum.














Table 5.1) Total VFA production rates (in mmol/kg dry matter of fermentation
contents h) and proportional contribution of each VFA for wild and captive
hoatzins. The proportional contribution of individual VFA production rates to
the basal rate of metabolism (% BRM) was calculated using the model by Prins
et al. (1984) with energy equivalents of individual VFA from Blaxter (1962).
Fermentation contents refer to contents of the crop and caudal esophagus.


CAPTIVE HOATZINS
1 2


Body mass (kg)
Ferm. contents (kg)
VFA production rate
% Acetic
% Propionic
% Isobutyric
% Butyric
% Isovaleric
% Valeric


0.64 0.65
0.034 0.040


79.5 80.1
18.2 12.2


TOTAL (mmol/kgDM h) 55.41


5.7
1.0
0.3
50.49


13.23 14.65


WILD HOATZINS
1 2 3
0.75 0.64 0.76
0.052 0.045 0.066


61.5 54.7 36.5
22.2 30.1 17.9
0.6 0.6 23.4
11.0 11.2 6.2


139.12 152.29 116.14
54.23 61.46 69.70


% of BRM















Table 5.2) Comparative table showing fermentation rates (as maximum
reported VFA production rate in mmol/kgDM of fermentation contents h) of
selected mammals with foregut and hindgut fermentation, herbivorous birds,
and hoatzins in this study. Feeding strategies represent categories defined by
(Hofmann 1989), as CS = concentrate selector, IF = intermediate (mixed) feeder
and GR = grass and roughage eater.


DIGESTIVE
STRATEGY


VFA MAIN
PRODUCTION FEEDING FERMENTATION
DA'I CTD ATEtGV ITTTL


REFERENCF


Mammals with foreaut fermentation


Suni 629 CS
Kirk's Dikdik 542 CS
Colobid monkey 475 CS
Thompson's gazelle 420 IF
Grant's gazelle 356 IF
Greater Kudu 175 CS
Quokka 135 IF
Zebu cattle 126 GR
Mammals with hindgut fermentation


Rabbit 205
Howler monkey 250
Birds
Willow Ptarmigan 74
Captive hoatzin 53
Wilrd hnat7in I 'f


IF
CS


rumen (Hungate et al. 1959)
rumen (Hoppe et al. 1983)
forestomach (Bauchop and Martucci
rumen (Hoppe et al. 1977)
rumen (Hoppe et al. 1977)
rumen (Giesecke and Gylswyk
forestomach (Moir et al. 1956)
rumen (Hungate et al. 1959)


caecum
caecum


caeca
crop
crnn


1968)



1975)


(Hoover and Clarke 1972)
(Milton and McBee 1983)


(McBee and West 1969)
this study
this .tlniv















CHAPTER 6
RATE OF METABOLISM IN THE HOATZIN



Introduction

Basal rates of metabolism in birds are correlated to body mass and food

habits (McNab 1988). Among all possible food habits, folivory is rare in birds

(Morton 1978). This rarity seems to result from a conflict between the

processing and digestion of a bulky diet and the energy requirements for

flight (Sibly 1981). Fermentation of leaves of plants requires large

fermentation chambers in the gastrointestinal tract where anaerobic

microbes break down cell walls. This study examines the rate of metabolism of

the hoatzin, Opisthocomus hoazin, a unique folivorous bird with a well-

developed foregut fermentation system.

The hoatzin is one of the most folivorous of all birds: up to 85% of its

natural diet consists of plant leaves (Grajal et al. 1989). It is the only known

vertebrate with a foregut fermentation digestive system outside the mammals.

Moreover, it is the smallest vertebrate with such a digestive system. Folivory

in the hoatzin has resulted in dramatic morphological, physiological and

behavioral adaptations (Grajal et al. 1989, Strahl 1988). The sternal carina is

reduced to accommodate the voluminous crop and caudal esophagus, where

fermentation occurs. As a result, there is little area for flight muscle

attachment. Indeed, hoatzins are not powerful fliers, preferring to hop from

branch to branch. Other life history characteristics, such as functional wing

claws in young hoatzins, might be related to the energy constraints of the

hoatzin's folivorous habits (Grajal et al. 1989).










In mammals, arboreal folivorous food habits are accompanied by low

basal rates of metabolism (McNab 1978). It is not clear, however, how

folivorous food habits affect basal rates of metabolism in birds. This is partly

caused by the lack of a universally acceptable standard basal rate of

metabolism for birds. For example, birds of the order Passeriformes seem to

have significantly higher basal rates of metabolism than other (non-

passerine) birds (Dawson and Hudson 1970, Lasiewski and Dawson 1967).

Furthermore, Aschoff and Pohl (1970) found that birds have higher basal rates

of metabolism during the active phase of the daily cycle. These data sets

reflect potential biases because most measured birds are from temperate

habitats and come from a narrow taxonomic spectrum when compared to world

bird diversity. Moreover, a confounding factor in these data sets is that most

measured Passeriformes are of small body mass (<300g), while measured non-

passerines are of medium or large body mass (McNab 1988, Prinzinger and

Hlnssler 1980). Indeed, when small non-passerines are included in the

allometric models, no appreciable differences in basal rates of metabolism can

be found between Passeriformes and non-passerines (Prinzinger and Hinssler

1980).

As a result, comparisons of the basal rate of metabolism of folivorous

birds are difficult. For example, folivorous Grouse and Ptarmigan (family

Tetraonidae) have higher rates of metabolism than mammals of similar size

(Kendeigh et al. 1977). On the other hand, partially folivorous tropical

mousebirds (Colius) have relatively low rates of metabolism when compared to

other birds (Bartholomew and Trost 1970, Prinzinger et al. 1981). This study

explores the possible relationship between rate of metabolism and folivorous

food habits in birds in general and in hoatzins in particular. If this

relationship is similar between mammals and birds, then it would be expected










that an avian arboreal folivore, such as the hoatzin, would have the relatively

low rate of metabolism found in arboreal folivorous mammals.



Materials and Methods

Three adult hoatzins were captured at the GuArico River, in the North-

Central Llanos of Venezuela (670 35'N, 8 34' W). The birds were of unknown

age and sex, since no external sexual dimorphism is present in this species.

The birds were progressively acclimated from their natural diet to an artificial

diet composed of romaine lettuce and a mix of alfalfa pellets, soybean protein

concentrate and a vitamin supplement (Grajal et al. 1989). After 35 days, the

birds were acclimated to captivity and maintained a stable body mass. During

the study, 2-9 October 1989, the hoatzins were housed in 1 x 1 x 0.5 m wire cages

in a temperature-controlled room (280C +-30C) and controlled light cycle (12:12

hours). In the wild, hoatzins are most active late in the morning and late in

the afternoon. Outside these activity peaks, wild hoatzins spend most of their

time resting (pers. obs.). Consequently, the hoatzins were fed twice daily

during their normal activity peaks, while measurements of their metabolism

were performed at other times. No attempts were made to starve the hoatzins,

since their gut has to be close to full capacity at all times for a regular and

substantial rate of fermentation.

Rate of oxygen consumption (VO2) was measured using a controlled

temperature, negative pressure, open flow system with an Applied

Electrochemistry S-3A Oxygen Analyzer. Water vapor and carbon dioxide were

removed with drierite and ascarite before entering the respirometry chamber.

Two temperature-controlled chambers of 36 and 42 1 were used for 32

measurements. Each hoatzin was measured in both chambers for the same

ambient temperature. Flow rates were adjusted to the volume of each chamber









and ranged from 90 to 200 ml/min. Measurements lasted 2-5 hours, and were

terminated when a low and constant V02 was obtained. Body temperatures (Tb)

were measured before and after the V02 determination, using an electronic

telethermometer with a thermocouple probe inserted 3-5 cm in the cloaca for

15-30 seconds. Hoatzin Tb and body mass were measured immediately before

and after each VO2 measurement.

Thermal conductance was calculated as the mean conductance from

individual conductances Cm = V02 / (Tb Ta) at each measurement, in which

Cm is the thermal conductance (McNab 1980) and Ta is the ambient

temperature (air temperature inside the chamber). To compare the hoatzin

basal rate of metabolism with expected values of mass-specific basal rate of

metabolism for endotherms, the regression model from Kleiber (1961) was

used: V02 / M = 3.4 M 0.25 cm3 02 g-1 h-1. Additionally, the basal rate of

metabolism in hoatzins was compared to the mass-specific non-passerine

model of Prinzinger and Hanssler (1980): V02 / M = 6.8 M -0.28 cm3 02 g-1 h-

1. Minimal resting mass-specific conductance was calculated from Aschoff

(1981) Cm / M = 0.95 M -0.53 cm3 02 g-1 h-1 C -1

Average environmental temperature in the natural habitat of the

hoatzin in the Llanos of Venezuela is 27C with a year minimum of 190C and a

maximum of 430C (Troth 1979). Therefore, the temperatures at which we tested

the oxygen consumption of hoatzins are within the range of temperatures

experienced by hoatzins in their natural habitat.










Results

The relationships of body temperature and rate of metabolism with

ambient temperature are shown in Fig. 6.1. No differences in rate of

metabolism between day and night measurements were found. The three

hoatzins maintained a constant body temperature of 38.50C (s.d. 1.20C, n = 32)

at Ta between 120C and 360C. At ambient temperatures over 36.50C, the hoatzins

were restless and became dangerously hyperthermic (Fig. 6.1).

The rate of metabolism for hoatzins within the thermoneutral zone was

0.48 cm3 02 g-1 h-1, about 69.8% of the expected value for an endotherm with a

body mass of 598 g (Kleiber 1961) and 43% of the expected basal rate of

metabolism for a non-passerine bird (Prinzinger and Hiinssler 1980). The

thermoneutral zone was between 26.5 and 36.50C. Thermal conductance below

the thermoneutral zone was 0.039 cm3 02 g-1 h-1 oC-1 (s.d. = 0.006, n = 18),

which represents 105% of the expected value for birds (Aschoff 1981).

Conductance decreased to 89% (0.033 cm3 02 g-1 h-1 oC-1) at Ta below 180C.



Discussion

Hoatzins maintain homeothermy over a wide range of environmental

temperatures. In the hoatzin, constant temperature is probably important for

the optimization of microbial and enzymatic activity (Grajal et al. 1989).

Furthermore, in wild hoatzins, constant body temperature is probably

maintained using special behavioral patterns. Overall energy expenditure is

probably reduced by long periods of inactivity and thermoregulatory

behaviors, such as selection of shady or cool microhabitats at high Ta

Moreover, hoatzins adopt a specialized sunbasking posture during early

morning or after heavy rains, opening the wings, ruffling the rump feathers,

and exposing the dark skin underneath (pers. obs., Strahl 1985). Hoatzins










change conductance at low Ta, suggesting they are able to mix chemical and

physical thermoregulation (McNab 1980). Lower conductance at Ta below 18C

may be another technique for energy conservation as a result of decreased

peripheral circulation and changes in feather position (McNab 1989).

The low basal rate of metabolism of the hoatzin agrees with the general

pattern of folivorous arboreal mammals (McNab 1978, McNab 1983) and is

much lower than rates of metabolism for non-passerine birds (Prinzinger and

Hinssler, 1980). The small number of studies on the rate of metabolism of

folivorous birds, however, prevent broad generalizations (McNab 1988). Other

highly folivorous birds, such as Grouse and Ptarmigan (family Tetraonidae)

have comparatively high rates of metabolism (Kendeigh et al. 1977). These

results are difficult to compare with the rate of metabolism in hoatzins due to

several factors: Hoatzins are exclusively tropical, eat mostly young leaves and

shoots of angiosperm plants and fermentation occurs in the foregut. In

contrast, Grouse occur only in temperate boreal forests and tundra habitats,

eat a variety of plants, including lichens, conifer needles and seeds, and their

fermentation site is located in the hindgut (Davis 1987, Gasaway 1976b,

Gasaway 1976c, Martin and Martin 1984). Additionally, other folivorous birds

do not have significant fiber fermentation capabilities and still can maintain

normal to high levels of metabolism (Crawford and Schmidt-Nielsen 1967,

Kendeigh et al. 1977). These contrasting trends probably reflect the diversity

of digestive strategies of herbivorous birds (see Chapter 7).

In the hoatzin, a unique combination of selective pressures probably

resulted in a well-developed foregut fermentation digestive system and

accompanying low basal rates of metabolism. These selective pressures reflect

the energy conflicts between homeothermy and flight ability on one side, and

leaf fiber and secondary compound concentration on the other. Leaves are










bulky, and their fermentation releases assimilable energy at a slow rate, when

compared to other food items. Most folivorous birds avoid these energy

constraints by an increase in processing rate. For example, folivorous birds

with little or no fermentation process their bulky diet at high rates

(Buchsbaum et al. 1986, Dawson and Herd 1983, Dawson et al. 1989, Mackie 1987,

Mattocks 1971). Faster processing rates, however, result in decreased digestive

efficiency (Van Soest 1982). In avian herbivores, fast food processing

decreases per unit dry matter digestibility but can increase the total rate of

energy intake. In fact, simple gastric digestion of cell contents and sometimes

of the hemicellulose fraction of the cell walls seem to supply most of the

energy needs of some of these birds (Buchsbaum et al. 1986, Dawson et al. 1989,

Herd and Dawson 1984, Mackie 1987).

Evidently, the digestive patterns of these avian folivores are not always

possible. Most dicotiledoneous plants have lower levels of hemicellulose than

grasses (Agricultural Research Council 1980, Van Soest 1969) and usually

contain a large variety of secondary compounds (Freeland and Janzen 1974).

Fermentation in the hoatzin has resulted in significantly slower passage rate

and comparatively lower intake rate than nonfermenting folivorous birds

(Grajal et al. 1989, also see Chapters 3 and 4). This digestive pattern provides

several advantages: a) enhanced fiber digestion (Van Soest 1982), b) reduced

dependence on continuous foraging to meet energy requirements (with more

possibilities for food selectivity) and c) detoxification of plant defensive

chemical compounds (Freeland and Janzen 1974). On the other hand, effective

fermentation of leaves requires gut fermentation chambers of large capacity

and delayed passage rates (Parra 1978). In the hoatzin, the enlarged capacity

of the crop and esophagus has caused extensive anatomical and behavioral









modifications that reduce flight performance (Boker 1929, Gadow 1891, Grajal

et al. 1989, L'Hermenier 1837).

Foregut fermentation reduces the hoatzin's dependence on continuous

foraging and increases the time for microbial attack of fiber and secondary

compounds. Microbial metabolic losses of fermentation can reduce the rate of

energy available for assimilation to the host. Therefore, the low basal rate of

metabolism in the hoatzin is probably the result of an evolutionary tradeoff

between the benefits of enhanced fiber digestion, greater selectivity,

detoxification of secondary compounds, and the costs of a low rate of energy

availability and reduced flight ability. As a consequence, the low basal rate of

metabolism is probably one of the most important physiological adaptations

for the evolution of a foregut fermenting system in a flying bird.





85







10 15 20 25 30 35 40
,I ,, U 43 O
41 o
8 6 oo, o -3 c
C I-
1.2 37




o 0.8
E 0.6
0.6
0 o

> 0.4

0.2
S10 15 20 25 30 35 40
cr Environmental Temperature ( 0 C)



Figure 6.1). Mass-specific rates of metabolism and body temperatures of three
adult hoatzins (each individual denoted by a different symbol) as a function of
ambient temperature.















CHAPTER 6
GENERAL DISCUSSION



Evolution of Foregut Fermentation

The hoatzin's way to deal with a leafy diet is unique, leading to some

extreme morphological, physiological and behavioral adaptations. The hoatzin

crop and posterior esophagus are the primary site for digestion of its leafy

diet. This foregut fermentation system is unique among birds, and it is more

similar to foregut fermentation systems in mammals. The hoatzin, however, is

almost an order of magnitude smaller than the smallest mammal with a well

developed foregut fermentation.

The evolution of foregut fermentation has been interpreted as a

digestive strategy that takes advantage of diets low in nitrogen and high in

fiber (Demment and Van Soest 1985, Hume and Warner 1980, Janis 1976, Parra

1978). Indeed, the rapid radiation of foregut fermenting Artiodactyls and

Macropods probably occurred with a simultaneous expansion of grasslands

during the Miocene and Pliocene (Janis 1976). As a consequence, most

evolutionary explanations of the presence of foregut fermentation have

emphasized the advantages of cell wall digestion as an important selective

force in the evolution of foregut fermentation systems (Janis 1976). Although

some of the most advanced foregut fermenters, such as ruminants, do indeed

take advantage of a highly fibrous diet, foregut fermentation has evolved

independently in other taxa that are not grassland dwellers, but tropical forest

inhabitants. This group includes tree-kangaroos (Hume 1978, Hume 1982),

tree-sloths (Bauchop 1978, Montgomery and Sunquist 1978), colobid monkeys










(Bauchop 1978, Bauchop and Martucci 1968, Ohwaki et al. 1974), tragulids

(primitive Artiodactyls) (Langer 1974) and the hoatzin (this study, Grajal et al.

1989). Hume and Warner (1980) proposed that the presence of foregut

fermentation in these forest animals is probably not related to the nutritional

use of grasses, but to the use of tropical forest plants. Tropical forest plants

are usually available year-round, but generally have high levels of secondary

compounds (McKey et al. 1981, McLeod 1974, Moreno-Black and Bent 1982,

Robbins et al. 1987). Foregut fermentation can be an adaptive foraging

strategy in these habitats, because foregut microbes can detoxify secondary

compounds before they reach the lower gut where absoption takes place

(Barry and Blaney 1987, Freeland and Janzen 1974, Mackie 1987). In addition,

foregut fermentation can enhance the quality of nitrogen levels in the diet

and allow the use of plant fiber as a nutrient source.

Therefore, the presence of foregut fermentation in another small,

arboreal no-mammalian vertebrate, seems to indicate that foregut

fermentation has evolved several times not only as a response to the use of

tropical forest plants as a resource and not to the nutritional use of grasses.

Other lines of evidence suggest that indeed most foregut fermentation systems

have evolved from ancestral forest forms that have later radiated into

grasslands. Indeed, Tragulids are tropical forest Artiodactyls, and considered

"primitive" ruminants because they have retained ancestral characteristics of

the original ruminants (Langer 1974).










Herbivory in Birds

Given the advantages of foregut fermentation, it is not clear why

hoatzins are the only birds with this digestive system. Indeed, foregut

fermentation may not be advantageous for birds. Microbial fermentation of

readily digestible cell contents inserts an additional trophic level between the

food and the host, increasing microbial metabolic losses (e.g., methane, C02

and heat) and decreasing the overall energy available to the host. Given the

high energy cost of flight and endothermy, such metabolic losses may not be

acceptable for most birds (Morton 1978).

These and other costs may explain why only 3% of the extant species of

birds consume plant leaves as a significant proportion of their diet (Morton

1978). Most herbivorous birds increase the digestion of cell contents at the

expense of a reduced nutritional use of cell walls. Although a more thorough

investigation of digestive patterns of herbivorous birds is needed, some

general trends may explain how herbivorous birds deal with different kinds of

leafy diets and habitat constraints. Some of these major categories include:

1) Caecal fermenters: This group includes some well-studied taxa (e.g.,

ptarmigan and grouse of the family Tetraonidae) and others not so well studied

(e.g., screamers of the family Anhimidae, and large ratites such as ostriches).

Ptarmigan and grouse eat some of the most refractory diets for a bird,

including pine needles, twigs, and catkins (Davis 1987, Gasaway 1976b, Hill et

al. 1968, Leopold 1953, Moss 1974, Moss 1977, Moss and Parkinson 1972,

Pendergrast and Boag 1971, Ponce 1985, Ponce 1987, Pulliainen et al. 1968,

Pulliainen and Tunkkari 1983). The digestive pattern of these advanced

herbivores can be summarized as an optimization of the nutritional use of cell

contents until food arrives in the enlarged caeca. The caeca fill selectively

with highly fermentable smaller particles and liquid, while most of the










largely undigested cell wall is excreted (Duke 1989, McLelland 1979, Ziswiler

and Farner 1979). Although significant amounts of fiber are digested by

caecal fermenters (Gasaway 1976a, Inman 1973, Moss 1973, Moss and Trenholm

1987, Pulliainen et al. 1968, Suomalainen and Arhimo 1945), cell wall digestion

may not be the digestive goal (Bjmrnhag 1989, Remington 1989). Caecal

fermentation provides important benefits in addition to some fiber

fermentation. For example, caecal fermentation enhances the use of urinary

nitrogen for microbial growth (Skadhauge, 1976) and provides the energy

benefits of microbial VFA production (Bjornhag 1989, Gasaway 1976b, Gasaway

1976c, Gasaway et al. 1976, Mackie 1987, Remington 1989, Skadhauge 1976,

Withers 1983). Another advantage of caecal fermentation is that it allows

wider dietary niches. For example, when high quality food sources as seeds

and fruits are seasonally available, caecal fermenters can reduce the ingestion

of leaves and take advantage of these higher quality food items (Davis 1987,

Ponce 1985, Ponce 1987). A similar dietary plasticity is found in most other

avian herbivores (probably except hoatzins), but in caecal fermenters it is

combined with an efficient fermentation system.

2) Non-fermenting grazers and aquatic plant eaters: This group

includes many members of the family Anatidae (ducks and geese) and some

coots and gallinules of the family Rallidae, including some flightless species

(Buchsbaum et al. 1986, Dawson et al. 1989, Kingsford 1989, Mulholland and

Percival 1982, Reid 1974). Most of these birds have little or no fermentation

(Clemens et al. 1975), with fast passage rates and high food intakes (Bj6mhag

and Sperber 1977, Buchsbaum et al. 1986, Burton et al. 1979, Dawson et al. 1989,

Ebbinge et al. 1975, Halse 1984, Marriot and Forbes 1970, Miller 1984, Muztar et

al. 1977). Most eat grass or aquatic plants with low nitrogen and high fiber

levels (Ebbinge et al. 1975, Esler 1989, Hardin et al. 1984, Kingsford 1989,










Montalbano et al. 1979, Mulholland and Percival 1982, Owen 1975, Reid 1974).

The rate of nutrient and energy uptake is optimized by increasing food intake

at the expense of a thorough digestion of both cell contents and cell walls.

Some fiber is digested, mainly by acid degradation of hemicellulose at the

stomach and some microbial fermentation (Buchsbaum et al. 1986, Dawson et

al. 1989, Marriot and Forbes 1970).

3) Frugivores-folivores: This is a heterogeneous group with varying

proportions of leaves in the diet. In South America, Passeriformes of the

family Phytotomidae and Saltatoridae ingest large amounts of leaves during

parts of the year (pers. obs.), but not detailed studies are available yet.

Similarly, birds of the family Colidae are partially folivorous (Bartholomew

and Trost 1970, Prinzinger et al. 1981). The New Zealand fruit pigeon subsists

for more than eight months of the year exclusively on leaves of a few plants

in riparian forests of temperate New Zealand (Clout et al. 1986). Finally, some

Galliformes include large proportions of leaves in their diets, including

members of the families Phasianidae (Young et al. 1991) and Cracidae (largely

unstudied). It is unclear why these frugivores ingest leaves or what is the

proportional composition of their diet. Leaves may supplement the low

nitrogen or mineral deficiencies of a frugivorous diet (Morton 1978).

4) Fiber manipulators: This is the niche of another oddity among avian

herbivores. The kakapo or owl parrot, Strigops habroptilus. uses its dexterous

bill and tongue to literally "chew" tussocks, grass blades and rhizomes,

squeezing the cell contents and leaving dried clumps of squeezed plant leaves

(Stivens 1964). Whether fermentation takes place in the lower gut or in the

enlarged crop remains unclear (B6ker 1929). Sadly, this species is one of the

rarest birds in the world, and its survival as a species depends on active

management by humans (Merton 1977). Therefore, research on this bizarre