Citation
Green pigmentation in neotropical frogs

Material Information

Title:
Green pigmentation in neotropical frogs
Creator:
Jones, Duvall Albert, 1933-
Publication Date:
Language:
English
Physical Description:
vii, 154 leaves. : illus. ; 28 cm.

Subjects

Subjects / Keywords:
Eggs ( jstor )
Frogs ( jstor )
Hemoglobins ( jstor )
Liver ( jstor )
Oxygen ( jstor )
Pigmentation ( jstor )
Pigments ( jstor )
Species ( jstor )
Tadpoles ( jstor )
Water temperature ( jstor )
Animals -- Color ( lcsh )
Dissertations, Academic -- Zoology -- UF
Frogs ( lcsh )
Zoology thesis Ph. D
City of Gainesville ( local )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida, 1967.
Bibliography:
Bibliography: leaves 143-153.
General Note:
Manuscript copy.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
This item is presumed in the public domain according to the terms of the Retrospective Dissertation Scanning (RDS) policy, which may be viewed at http://ufdc.ufl.edu/AA00007596/00001. The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator(ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
022373061 ( ALEPH )
13692006 ( OCLC )

Downloads

This item has the following downloads:


Full Text















GREEN PIGMENTATION IN NEOTROPICAL FROGS












By
DUVALL ALBERT JONES


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











UNIVERSITY OF FLORIDA
December, 1967

















ACKNOWLEDGMENTS


I wish to thank Dr. Coleman J. Goin for suggesting the challenging topic of this paper and for his continued interest and support during the various phases of its study. Dr. Goin has aided in the identification of frogs used in this study and has extended many other courtesies which are deeply appreciated. I am grateful to Dr. Mildred Griffith, Dr. Frank J. S. Maturo and Dr. James Nation for their aid during this study and for suggestions which improved the dissertation. To Olive B. Goin, I owe thanks for many kind acts and stimulating discussions.

Several individuals have made laboratory and field facilities available to me. Among these were: Dr. James Nation, Dr. Ward Noyes, Dr. Albert Laessle, Dr. Frank Nordlie, Dr. Paul Elliott, and Dr. James Gregg, all of the University of Florida; Dr. Harold Heatwole, University of Puerto Rico (El Yunke Biological Station); Mr. Bernard Lewis and Dr. Thomas Farr, Institute of Jamaica; Dr. Ivan Goodbody, University of West Indies, Kingston, Jamaica; Dr. Jocelyn Crane, Simla Biological Station, Trinidad; and Mr. Walter Polder, Paramaribo, Surinam.

To the following persons, I am grateful for the opportunity to study the specimens of their respective institutions: Dr. Ernest Williams, Museum of Comparative Zoology, Harvard University; Mr. C. M. Bogert and Dr. Richard Zweifel, American Museum of Natural History; ii











Dr. James Bohlke, Philadelphia Academy of Natural Sciences; Dr. Doris Cochran and Dr. James Peters, United States National Museum; Mr. Neil Richmond and Dr. Clarence McCoy, Carnegie Museum; Dr. Charles F. Walker, University of Michigan Museum of Zoology; Dr. Robert Inger, Chicago Natural History Museum; Dr. Walter Auffenberg, Florida State Museum; Mr. Bernard Lewis and Dr. Thomas Farr, Institute of Jamaica.

I am grateful to Dr. Thomas Goreau, University of the West Indies, Dr. Margaret Stewart, State University of New York at Albany, and others for enlightening discussions. Bonneye Greene, Thomas Quarles, Robert MacFarlane and John Anderson kindly rendered technical assistance. Forest service personnel in Jamaica and Surinam made it possible for me to visit several remote localities. I appreciate the aid rendered by members of the U. S. Consular Service. Grants for travel expenses were received from the University of Florida Graduate School, the Sigma Xi- RESA Fund and the National Science Foundation (GB-3644, to Dr. Coleman J. Goin).

The services of Rhea Warren, Stephen Bass, Murray de la Fuente and others who aided in the collection of specimens are appreciated greatly. Janet Bungay and Geraldine Lennon aided in the preparation of the manuscript. My wife, Dorothy, assisted in the histological portion of the study, in preparation of the manuscript, and in giving moral support.
















TABLE OF CONTENTS


Page

ACKNOWLEDGMENTS........................................... ii

LIST OF TABLES.............................................. vi

LIST OF FIGURES............................................ vii

INTRODUCTION.............................................. 1

LITERATURE REVIEW.......................................... 3

Naturally Occurring Green Pigments.................... 3
Characteristics of Biliverdin......................... 6
Sources of Biliverdin................................. 9
Causes of Chlorosis................................... 12
Respiratory Studies................................... 13
Relation of Respiration to External
Environment......................................... 22
Structure and Function of Frog Liver.................. 23

SOURCES AND CARE OF TADPOLES AND FROGS..................... 24

1. CAUSATIVE PIGMENT AND INCIDENCE OF CHLOROSIS........... 27

Pigment Identification................................ 27

11. CONSIDERATION OF ECOLOGICAL FACTORS ASSOCIATED
WITH CHLOROSIS........................................ 35

Methods of Study...................................... 35
Observation of Frog Habitats and Behavior............. 36
Resume of Ecology and Breeding Habits................. 72

IlIl. PHYSIOLOGICAL FACTORS RELATED TO CHLOROSIS........... 74

Attempts to Reproduce Green Pigmentation
In Vivo............................................. 75
Physiological Characteristics of Chlorotic
and Non-Chlorotic Frogs............................. 81
Disease as a Cause of Chlorosis in Frogs.............. 91

iv













DISCUSSION ................................................ 93

High Rate of Red Cell Hemolysis...................... 93
Defective Transport of Bile Pigment into
the Liver Cell..................................... 99
Defective Bile Pigment Conjugation................... 100
Disturbed Bile Pigment Excretion..................... 100

CONCLUSIONS............................................... 104

APPENDICES................................................ 107

LITERATURE CITED.......................................... 143

BIOGRAPHICAL SKETCH....................................... 154


v

















LIST OF TABLES


Table Page

1 Occurrence of Green Color in Tissues and
Fluids of Adult Anurans................................... 108

2 Occurrence of Green Pigment Among Metamorphosing Neotropical Anuran Tadpoles....................... 113

3 Locality Data of Tadpole Habitats......................... 114

4 Water Characteristics of Tadpole Habitats................. 118

5 Iron and Phosphate Concentrations (Parts
Per Million) in Tadpole Habitats.......................... 123

6, Blood Characteristics..................................... 124

7 Average Dimensions of Frog Red Blood Cells................ 127

8 Respiratory Rates of Tadpoles in Warburg
Respirometers at Different Temperatures................... 128

9 Tadpole Respiratory Rates at 250 C.
(Measured by Jones Method)................................ 129

10 Froglet Respiratory Rates Measured In Warburg Respirometers..................................... 131


vi

















LIST OF FIGURES


Figure Page

I Absorption Spectrum of Biliverdin in
5 Per Cent Hydrochloric Acid Methanol
So lution.................................................. 132

2 Absorption Spectrum of Liver Extract
From Hyla septentrionalis in Aqueous
Solution.................................................. 132

3 Absorption Spectrum of Lymph of
Osteocephalus taurinus.................................... 133

4 Absorption Spectrum of Coelomic Fluid
(bile) From Osteocephalus taurinus........................ 133

5 Absorption Spectra of Bile Solutions of
Siren lacertina........................................... 134

6 Absorption Spectrum of Pigment Extracted
From Hyla marianae in 70 Per Cent Ethanol................. 134

7 Respiratory Rates of Hyla brunnea Tadpoles
at Different Temperatures................................. 135

8 Respiratory Rates of Tadpoles at 250 C.................... 136

9 Normal Red Blood Cells of Bufo terrestris
Four Hours After Collection............................... 137

10 Red Blood Cells of the Same Individual as Above, After Four Hours at 450 C.......................... 137

11 Red Blood Cells of Hyla sguirella After Four Hours at 450 C....................................... 138

12 Untreated Blood Cells of Hyla leucophyllata................ 138

13 Stained Smear of Untreated Blood of Bufo typhonius................................................. 139












Figure Page

14 Stained Smear of Untreated Blood of
Phrynohyas venulosa..................................... 139

15 A Stained Smear of Blood from Hyla
punctata, a Species With Green Tissues.................. 140

16 Stained Smear of Untreated Phyllomedusa
bicolor Blood........................................... 140

17 Stained Section of Apparently Normal
Liver of Hyla rubra..................................... 141

18 Stained Section of Hyla punctata Liver.................. 141

19 Stained Section of Liver From Adult Male
Pseudis paradoxus....................................... 142

20 Stained Section of Liver From Adult Male
Hyla maxima............................................. 142















INTRODUCTION


A wide variety of colors is demonstrated by anurans, particularly those of tropical areas. Frogs and toads which spend most of their time on tree trunks, among the dead leaves of the forest floor,

about rocks, or underground, are usually brown, gray or black due to the presence of melanin in melanophores. Red or yellow colors are usually due to lipid pigments in the lipophores, and white results from reflection of visible light by guanine crystals of guanophores. Green hues are often seen among anurans, especially in the skin of those closely associated with green vegetation.

Studies have shown that the green skin color of many frogs in life is the result of combined effects of three types of chromatophores located in the upper portion of the dermis. Lipophores containing yellow pigment are found just under the basement membrane of the epidermis, in close association with the guanophores immediately below them. Melanophores are found below the guanophores, but have processes extending into the guanophore layer. Visible light is diffracted as it

impinges upon the guanine crystals of the guanophores; the blue and green portions of the spectrum tend to be reflected, while light of longer wave lengths is absorbed by the melanophores. As the light of shorter wave lengths passes back through the lipophores, the blue portion is absorbed and only green is reflected from the skin. Modification of lipophores following fixation of the frog causes the skin to appear









2

blue. Thus, the green color of frog skin of this type results from the physical arrangement of three pigments, none of which is green (Schmidt, 1919, 1920, 1921; Kawaguti, Kamishima and Sato, 1965; Elias, 1943).

Though most frogs with green skin appear to fit the pattern described, some tree frogs do not. Notable are some of the small hylids such as Hyla wilderi. This species has few melanophores and appears to be a diffuse green color throughout the body. When preserved in alcohol, a green pigment is extracted and the frog becomes whitish in color. In this species, as well as in others, there is no blue color which supplants the green upon preservation. Indications are that structural effects are not involved and that this type of green color in frogs is due solely to a green pigment or pigments. Reports of internal green pigment substantiate this idea.

This paper is concerned with three facets of the biology of

green neotropical frogs: (1) identification of the green pigment or pigments; (2) ecology; and (3) physiology of green frogs, with emphasis upon possible factors influencing pigment accumulation. Since these three areas require different methods of study and yield various types of information, they will be presented in separate sections.















LITERATURE REVIEW

Naturally Occurring Green Pigments


Porphyrin compounds and their derivatives constitute the

vast majority of green pigments among living organisms. Among the best known of these are the chlorophylls, green hemoglobins and green bile pigments. Others include the chlorocruorin blood pigments of polychaetes; turacoverdin from the feathers of plantain-eating touracos (Fox, 1953); myeloperoxidase, the green enzyme of white blood cells and certain tumors (Agner, 1941); and verdohemochromes, which are

intermediate compounds that appear during the degradation of heme compounds to bile pigments in vitro (Lemberg and Legge, 1949).

Some tunicates have a pale green pigment which contains

vanadium. The structure of this pigment is not well known, but it may be similar to some of the bile pigments. Green pigments of crustacean hypodermis and eggs result from conjugation of a reddish carotene pigment with proteins (Fox, 1953).

Green pigments of frogs

In 1850, Kunde wrote the first paper concerning a green pigment in frogs (see Lemberg and Legge,1949, p. 506). He observed biliverdin in frogs' blood serum after liver extirpation. Lemberg (1935), Nisimaru (1931), Cabello (1943), Rodriguez Garay, Noir and Royer (1965), and Barrio (1965a) indicated that amphibian bile contains biliverdin as

3









4

its bile pigment. Lester and Schmid (1961) found biliverdin in some samples of frog bile, but believed bilirubin to be the main bile pigment of anurans. A claim was made by von Recklinghausen (1883) that formation of biliverdin takes place in sterile frog blood. Rich (1925) indicated that this had not been confirmed. Cabello (1943) caused the

formation of green serum in Bufo arenarum by administration of phenylhydrazine and ligation of the bile duct. Rodriguez Garay et al. (1965) found biliverdin glucuronide in the bile of Bufo arenarum.

The first report of externally visible green pigment in

frogs appears to be the one by Peters (1873), who indicated that the skeleton of Pseudis minuta was green. Camerano (1879) believed the green bones of Pseudis paradoxusto be due to the presence of ferrous phosphate. The recorder of the Zoological Record (Boulenger, 1880) took exception to Peters' paper by stating that Pseudis minuta does not have green bones. However, Boulenger (1883) took note of the green eggs of pseudids. Fernandez and Fernandez (1921), Miranda-Ribeiro (1926), Parker (1935), and Gallardo (1961) also noted green eggs or tissues in this family.

In 1910, Podiapolsky described a "chlorophyll" pigment from Hyla arborea and Rana esculenta, two European species. It may be noted here that reports by Schmidt (1919-1921) upon these two species, and by

Kawaguti, Kamishima and Sato (1965) upon Hyla arborea, do not support this. More recently, Dunn (1926), A. Lutz (1924, 1938), B. Lutz

(1948, 1954), Cochran (1955) Lynn (1958), and Bokermann (1964) have noted green coloration of epithelia, muscles and bones in certain










5

hylid species. Dunn (1931), B. Lutz (1947), and Savage (1967) noted the presence of green bones in certain species of centrolenids.

Barrio (1965a), working with frogs from Argentina, found

that five of twelve hylid species and all three of the pseudid species which he studied had green tissues. Individuals of twenty-two species of four other families lacked the green pigment in their tissues.

The species that had green tissues were similar in that the muscular and subcutaneous tissue, lymph, walls of the digestive tract, eggs of mature females, and especially the bones were strongly pigmented. He

attributed this situation to high concentrations of biliverdin in the blood. The highest concentration, 11.9 mg/100ml, was found in the serum of Hyla punctata. In addition, he pointed out that biliverdin is decomposed by formalin. This point clarified an earlier controversy (Boulenger, 1880) and also indicated a reason for so little attention being given to this phenomenon.

Another paper by Barrio (1965b) brought out the fact that

populations of Hyla pulchella differ greatly in their serum concentrations of biliverdin; two of the five subspecies lack the pigment altogether.

Green tissues of other animals

Green tissues are not limited to frogs. Peters (1873) noted that the green bones of Pseudis minuta were similar in color to those of the marine garfish Belone. The relatively recent papers of Wagenaar (1939), Fontaine (1941a, 1941b), Willstaedt (1941), Caglar (1945), and Fox (1953) concerning green bones of teleost fish pointed toward









6


biliverdin or other hematin derivatives as the green pigment. Lonnberg (1934) noted green color in tissues of the African lungfish, Protopterus annectens. Lemberg and Legge (1949, p. 506, pp. 569-570) listed other vertebrate structures and invertebrates which contain biliverdin. Identification of the green pigment

Barrio (1965a) has amply demonstrated by chemical and

spectrophotometric means that biliverdin is primarily responsible for the green color of tissues in the frogs which he studied. Through

electrophoretic fractionation of green serum, he found that most of the biliverdin was combined with globulin proteins and a lesser amount was associated with albumin. Lester and Schmid (1961) have shown that adult anurans have an enzyme for conjugating bile pigments and Rodriguez Garay, Noir and Royer (1965) have found small quantities of water soluble biliverdin glucuronide in Bufo arenarum. Barrio (1965a) considered the biliverdin to be unconjugated, since it migrated at the same rate as free biliverdin on paper chromatographs, was completely soluble in chloroform, and insoluble in water.


Characteristics of Biliverdin

Chemical properties and reactions

Biliverdin (C33H3406N4) is composed of four pyrrole groups linked together by three methyne bridges. Its structural formula may be drawn in the form of a chain, but is more precisely represented by the incomplete ring form. There is an obvious similarity of biliverdin to two of its precursors, protoporphyrin IXa and heme of hemoglobin. It differs from them in that its ring structure is incomplete; also,










7


it does not contain iron as does heme. Mesobiliverdin is formed from biliverdin by reduction of the vinyl side chains to ethyl groups.

Chemical reactions of biliverdin include the Gmelin reaction. Addition of fuming nitric acid to tetrapyrrole bile pigments converts all initially to biliverdins. These are oxidized further to yield a succession of pigments from blue-green through violet, red and yellow to colorless compounds.

As already indicated, reduction of the vinyl side chains of biliverdin results in mesobiliverdin. Reduction of the middle methyne bridge yields bilirubin, a common bile pigment of mammals and other vertebrates. Bilirubin may be reduced further to other bile pigments.

Biliverdin is destroyed by heating with concentrated sulfuric acid, but mesobiliverdin is not. Stable hydrochlorides or hydrobromides result from treatment with the appropriate acid. Complex salts may be formed with iron, zinc and copper. Biliverdin also undergoes esterification. Unlike bilirubin, biliverdin does not couple with diazotized sulfanilic acid to give the diazo reaction (Lemberg and Legge, 1949).

Physical properties

Biliverdin is moderately soluble in ether, in which it has a greenish-blue color. It may be extracted from ether by I percent hydrochloric acid, in which it has a blue-green color. These deep colors result from the conjugation of double bonds throughout the length of the molecule. Biliverdin is also soluble in methanol (Lemberg and Legge, 1949) and chloroform, but is insoluble in water (Barrio, 1965a).











8

The absorption curve of biliverdin has been shown (Figure 1). Mesobiliverdin has a similar curve, but its absorption maxima lie about 10 millimicrons further from the infrared than do those of biliverdin.

Neither of these compounds has an absorption peak between 400-440 millimicrons (Soret band), which is characteristic of compounds with the closed porphyrin ring (Holden and Lemberg, 1939).

Biological properties

The biological properties of biliverdin are related to its physical and chemical properties. Of considerable importance is its solubility in different solvents. The fact that Barrio (1965a) did not find biliverdin in nervous tissue, ocular fluids, or urine may be directly related to the insolubility of unconjugated biliverdin in water.

Insolubility of biliverdin in water presents a problem in

regard to its transport within the body and its excretion. In mammals, as well as some birds and reptiles and Bufo arenarum, this problem is not so important because the liver conjugates variable amounts of the bile pigments (either bilirubin or biliverdin) with glucuronic acid to form water-soluble glucuronides (Rodriguez Garay et al., 1965). These investigators also found a sodium salt of biliverdin, free biliverdin, and a substance which had chromatographic characteristics of a complex of bile acid and sodium biliverdinate in Bufo arenarum. The electrophoretic studies of Barrio (1965a) indicate that in the circulatory system biliverdin is associated primarily with globulin proteins, but some is conjugated with albumin. Non-esterified fatty acids appear to










9

displace bile pigment from serum proteins in new-born infants (Melichar, Polacek and Novak, 1962). Assuming that this also occurs in frogs, one would expect that when biliverdin exceeds the carrying capacity of the serum proteins, it will be deposited in the non-fluid tissues, staining them with its color. Evidence of the staining by biliverdin is available in many frogs, particularly where the gall bladder lies adjacent to the stomach wall.


Sources of Biliverdin

Hemoglobin and other hemoproteins such as myoglobin and

oxidizing enzymes constitute the main sources of bile pigments. Although most studies of bile pigment formation are concerned with the origin of bilirubin in mammals, we may consider the origins of biliverdin to be the same, since it is generally considered to be a precursor of bilirubin.

In a recent paper, Israels et al. (1966) proposed a scheme in which there are four components of human bilirubin. Each of these components is characterized by a peak concentration of radioactive bilirubin at a specific interval after the labeling of bilirubin precursors. The major bilirubin component is formed from hemoglobin of old red blood cells and makes its appearance about 100 days after the radioactive isotope is administered. The lesser components appear more rapidly; collectively, they are referred to as early bilirubin. One of

these reaches a peak three to five days after ingestion of the radioactively labeled compound and is believed to originate from a heme loss











10


during erythropoiesis. The third component is formed during the first twenty-four hours, probably beginning during the second hour after administration. The fourth fraction appears most rapidly--within a few minutes of intake. The latter fraction was detected in rat liver homogenate after five minutes' incubation with C14 labeled deltaaminolevulinic acid (a precursor of protoporphyrin). Presence of such a component was suspected when the radioactive bilirubin recovered during the first two hours was found to account for one-third of the twenty-four-hour total. A block of the bile duct or similar obstruction

is shown to increase the latter bilirubin fraction. Perfusion experiments and other data show that the latter two fractions are of hepatic rather than erythropoietic origin. The main non-erythropoietic component of bilirubin, which is relatively slow in being formed, 'is probably related to the turnover of the major hemeproteins," such as liver catalase. The smaller component arises rapidly from a pool of heme-protein, heme, or heme precursors which has a high turnover rate. Barrio (1965a) suggested that high concentrations of biliverdin in frogs may be due to increased formation of early biliverdin. Mechanisms of bile pigment formation

Earlier hypotheses. -- Much study has gone into the chemistry of bile pigments and their formation from hemoglobin. Most of these studies have dealt with non-enzymatic reactions of bile pigments and related compounds carried out in vitro. Early studies of hemoglobin degradation employed harsh techniques and detected such substances as hematin and hematoporphyrin. With such supposed intermediate compounds













as these, and the knowledge that globin and iron are separated from the tetrapyrrole, it was generally assumed that hemoglobin was degraded as follows: hemoglobin, hematin, hematoporphyrin, bilirubin, with biliverdin a secondary oxidation product of bilirubin (Lemberg and Legge, 1949). Later, Lemberg and others (see Lemberg and Legge, 1949; Foulkes, Lemberg and Purdom, 1951) succeeded in finding a sequence of relatively mild reactions which degraded hemoglobin to biliverdin. Although these reactions were not well understood and yielded only 15 per cent of the biliverdin expected, workers in the field generally accepted Lemberg's series of reactions as a hypothetical model of hemoglobin degradation in cells.

Enzymatic degradation of hemoglobin. -- Recent studies of

Nakajima et al. (1963) have demonstrated an enzymatic pathway which is capable of degrading hemoglobin to biliverdin. These workers have characterized an enzyme which oxidizes the alpha-methyne bridge of the heme group to form a possible precursor of biliverdin. Most of this work was done with the pyridine hemichrome rather than hemoglobins, but their study of substrates is of considerable interest. The enzyme did

not act upon alkaline hematin or protoporphyrin IX, and only weakly upon hemoglobin (contains ferrous ion) and hemiglobin (contains ferric ion). However, reaction of the enzyme with a complex substrate of hemoglobin and haptoglobin (a plasma protein which combines with extracellular hemoglobin), produced 49 per cent of the theoretical yield of biliverdin. With the hemoglobin-haptoglobin complex as a substrate, there was a lag period of five minutes before any noticeable reaction took place. There










12

was no lag period when a substrate of hemiglobin-haptoglobin was used. In addition, the enzymatic activity was about 50 per cent higher with hemiglobin-haptoglobin or carboxyhemoglobin-haptoglobin than with hemoglobin-haptoglobin as a substrate. Thus, it would appear that hemiglobin-haptoglobin is the primary substrate of the enzyme and that the lag in degradation of the hemoglobin-haptoglobin represents the period during which hemoglobin is being changed into hemiglobin. The enzyme acts only in the presence of oxygen and is considered an oxidase, rather than a peroxidase. It is known as heme alpha-methenyl oxygenase and is found largely in the liver and kidney, being nearly absent from the spleen and bone marrow. The product of this enzymatic reaction is then acted upon by a second enzyme, heme alpha-methenyl formylase, to yield biliverdin, iron and formaldehyde.

An interesting observation which may be inserted at this point is that biliverdin conversion to bi lirubin has been shown to be under enzymatic control in the laboratory rat (Lester et al. 1966).


Causes of Chlorosis

Barrio (1965a) noted that the three families of Neotropical

frogs which include chlorotic species--Hylidae, Centrolenidae and Pseudidae-also have in common an intercalary cartilage between the ultimate and penultimate phalanges of the digits. He suggested that these two common factors indicate a close relationship among these families. However, the intercalary bone of pseudids is believed to represent an adaptation for swimming, similar to the situation in aquatic mammals (Goin and Goin, 1962a). Most recent writers have considered the presence










13

of intercalary cartilages in the several families of tree frogs as an example of parallelism (Goin, 1961; Griffiths, 1963; Lynch and Freeman, 1966). Goin and Goin (1962a, p. 231) stated, "The extra joint thus provided allows the last phalanx, with its adhesive disc, to be placed flat against the surface regardless of the position of the foot--an obvious advantage to the climbing form."

No unusual conditions such as high rates of hemolysis were noted by Barrio (1965a). He suggested that the chlorotic conditions which he witnessed were due to formation of early bile pigment. Hemolytic conditions are not recorded for Amphibia, but might be expected during metamorphosis when tadpole hemoglobin is replaced by frog hemoglobin (McCutcheon, 1936). Varela and Sellares (1938) noted

a rapid decrease in the red cell count of Bufo arenarum at the end of the breeding period, a change which indicates rapid hemolysis.

It appears that no satisfactory explanation of chlorosis has

been published. However, presence of biliverdin in the chlorotic species studied by Barrio (1965a) implicates hemoproteins, particularly hemoglobin. Through hemoglobin one might expect involvement of various structures and functions related to gaseous exchange.


Respiratory Studies

Embryonic and larval respiratory structures

Considerable research has been done on respiration as one of the most important processes of living organisms. Much of it has been directed toward micro-organisms, the mammalian species, and animals










14

possessing unusual respiratory structures (Negus, 1965). The several types and combinations of respiratory organs shown by species of Amphibia have been the interest of many investigators of respiration. Most reports have been concerned with the structure of respiratory

organs and surfaces, rates of oxygen uptake, and the role of the blood and circulatory system in gaseous transport. Some work has been done on larval amphibian respiration, especially on those species of salamanders and frogs commonly used in embryological and physiological research, despite their small size. There are few studies, however,

of adaptations in relation to the ecological situations of larval or adult amphibians (Foxon, 1964).

Structure and function of respiratory and associated circulatory organs of amphibians in general have been reviewed by Noble (1931) and more recently by Goin and Goin (1962a) and Foxon (1964). The present survey will be concerned specifically with the respiration of amphibian tadpoles and similar aquatic forms, as well as with hemoglobin and red blood cells of amphibians generally.

As the amphibian embryo develops, gaseous exchange first occurs through the egg surface, later through the external body surfaces (including gills) and finally through lungs, if present (Foxon, 1964). Among the species of anurans, the respiratory surface decreases proportionately as the diameter of the egg increases, since the volume of the egg increases as a cubic function of the radius, while the area

increases as a squared function. No examplesof enlarged surface area due to folding or other changes in the egg membrane are known (Noble, 1931).













However, there are considerable differences in size of amphibian eggs. Panton (1952) noted that aquatic eggs of Hyla brunnea were only one-half

millimeter in diameter, whereas those of another frog (probably Eleutherodactylus jamaicensis) that did not develop in water, were several millimeters in diameter.

The sites of egg deposition and tadpole development appear related to respiratory processes. Dickerson (1906) believed that the jelly membranes of amphibian eggs retain heat longer than the surrounding water. Experiments by Savage (1950), using the eggs of Rana temporaria, showed that the mean temperature difference between eggs with jelly envelopes and plain water was 0.630 C. Moore (1940) suggested that the compact jelly masses (up to 10 centimeters in diameter) of Rana sylvatica slow the diffusion of oxygen to the embryos, and that this becomes critical as the metabolic rate increases in response to a temperature of about 250 c. However, Savage (1950) pointed out that water moves freely between individual jelly capsules of Rana temporaria, thus requiring diffusion of oxygen through only a few millimeters rather than several centimeters of jelly. Algae associated with jelly membranes of Rana sylvatica (Dickerson, 1906), Rana aurora (Wright and Wright, 1949) and Rana temporaria (Savage, 1961) may affect the respiration of these frogs' eggs.

Goin and Goin (1962b) noted that amphibian eggs may be laid

singly or in clusters and may be attached to submerged objects, floating or settled in water, carried by parents, or be laid on land or vegetation.












There are few indications of respiratory rates of embryos being correlated with these types or sites of egg deposition. Panton's observation

(1952) that a set of large eggs, probably those of Eleutherodactylus jamaicensis, failed to develop when placed in water may indicate that embryos which develop terrestrially receive insufficient oxygen when subjected to the reduced oxygen tension of water. Noble (1931) suggested that large eggs undergoing rapid development require more oxygen than they can obtain in water when surrounded by the egg capsule.

As the embryos develop and hatch, other respiratory structures are formed and behavioral patterns related to respiration arise. Organs of external respiration in tadpoles are the skin, the internal surface of the operculum, external and internal gills, the vascularized surface

of the food filtering apparatus, and the lungs.

The most evident respiratory structures to appear in about the time of hatching are the external gills. These may be small as in Hyla vasta or enlarged as in Hyla rosenbergi tadpoles, although both species dwell in tropical streams (Noble, 1931). Dunn (1926) noted the reduction of gill size in Jamaican hylids, which breed in water which collects in bromeliads. Likewise, tadpoles of the genus Hoplophryne, which live in water that collects in banana leaves, also have reduced gills (Noble, 1929). Terrestrial embryos of the genus Eleutherodactylus may or may not have external gills, their highly vascular tails serving as important respiratory organs (Lynn, 1961). Among the more unusual are the gills of some Gastrotheca which are large and bell-shaped (Noble, 1931).










17

Studies of Babak (1907a, 1907b) upon Rana temporaria and

Rana esculenta, of Drastich (1925) upon Salamandra maculosa, of Bond (1960) on Salamandra maculosa, Ambystoma opacum and Ambystoma jeffersonianum, of Conant (1958) concerning adult Necturus maculosus and Freeman (1963) on adult Pseudobranchus striatus, all showed an inverse relationship between external gill size and oxygen concentration. Most indicated that external gills increased in size when animals were put into water of lower oxygen concentration. In addition, Conant, Bond, and Freeman all noted that gill movements and expansion, as well as blood supply were correlated with oxygen tension.

Pulmonary respiration is important to the majority of adult amphibians and is found in a number of larvae as well. The lungs of stream dwelling species are reduced in size (Noble, 1931). Savage

(1961) has described the well developed lungs of Rana temporaria larvae and still larger ones from a Papuan tadpole; both of these species were taken from standing pools or puddles of water. In the Papuan species, it appeared that tail muscles pressed the lung against the notochord to bring about ventilation.

Savage stated further that:

The respiratory systems in tadpoles are connected with the
ecology in the following way. If a tadpole lives in an environment rich in food, as many temporary or polluted ponds are, it does not need to pump water to get its good, and so does not need large gill filters. To use the oxygen under these
conditions, however, it must have large gills, and needs
lungs to tide it over emergencies. If, however, it lives
in the oligotrophic type of pond, with plentiful and almost constant supplies of oxygen but with a low concentration of
food, it needs to pump much water, and so must have large
gill filters. With these, gills might not be necessary,
because of the large surface of the filters (or rather, in










.0


view of Strawinski'-s work, more probably, the associated large operculum). In ordinary ponds, intermediate
as habitats, the arrangements might be expected to be
intermediate also.

There is a great deal of conjecture in all this, but
the microhylids seem to provide examples. Some have such
large gill filaments that they trail in the opercular
cavity in a way quite unlike those of Rana, others have no
gills but have enormous gill filters, for example Glyphoglossus molossus. Some, such as Hypopachus aquae, are
intermediate and live in ordinary ponds. Rana temporaria
also have a moderate development of gills, filters and lungs,
and lives in ordinary ponds.(P. 56)

Among amphibians, we may turn to the work of Strawinski (1956) for an estimation of the relative importance of tadpole respiratory surfaces. In studying the density of capillary networks of Rana

esculenta, he reported that most gaseous exchange of early stages took place through the skin, with the internal surface of the operculum also important. It was his belief that the external and internal gills, and vascularized filtering apparatus were of little importance. As

the lungs developed, their portion of the total respiratory capillaries increased rapidly, reaching 65 per cent at metamorphosis, the same proportion as in the adult. Vascularization of the skin also increased until metamorphosis, when these capillaries accounted for 34 per cent of the total.

Blood transport of oxygen

Among vertebrates, blood contained within a closed circulatory system is largely responsible for the transport of gases in the body. Hemoglobin pigments, located in red blood cells and responsible for their color, are the main carriers of oxygen in vertebrates. Exceptions










19

to this are the Antarctic ice fishes, Chaenichthyidae (Ruud, 1959). Blood of the several species of this family is nearly colorless and contains only white cells (1 per cent of blood volume). The cold water in which these fish live has a high oxygen content and the low temperature tends to lower the respiratory rates, so that oxygen diffusion through the naked skin sufficiently supplements that dissolved in the blood to maintain their low rate of metabolism. Small eel larvae also lack hemoglobin until they reach the elver stage (Andrew, 1965) and occasional specimens of the frog, Xenopus laevis, have been found without hemoglobin (de Graaf, 1957; Ewer, 1959).

Among the indicators of oxygen-carrying capacity of blood

are measurements of red cell size, erythrocyte counts and determinations of hemoglobin content. Amphibian erythrocyte sizes and counts have been listed by Prosser et al. (1950), Vernberg (1955), Freeman (1963), and Hartman and Lessler (1964). Of all vertebrates, salamanders have the largest red cells, ranging from 30-62 microns in longest dimension. According to these sources, erythrocytes of frogs, which are also oval in shape, range from 18-23 microns in length. The number of red blood cells usually varies inversely to cell size. Counts of salamander cells vary from 28,000 to 197,000 per microliter; those of anurans, from 380,000 to 670,000 per microliter. While one would expect larger cells

and higher cell counts to represent greater oxygen-carrying capacity, a trurer measure lies in the direct determination of hemoglobin. Kirkberger (1953), Stuart (1951), and Goin and Jackson (1965) have published values










2C

for hemoglobin concentrations of amphibian blood. These range from

8.7 to 11.5 grams of hemoglobin per 100 milliliters of blood.

In regard to shape, it should be noted that the oval disc

form of amphibian erythrocytes presents a greater surface area per unit of volume than would a sphere of like volume, thus presenting a greater surface for gaseous diffusion (Andrew, 1965) Greatest efficiency of oxygen transport among amphibians appears in the genus Batrachoseps, which may have up to 90 per cent erythroplastids or enucleate erythrocytes (Emmel, 1924). Lack of a nucleus makes the cells flatter and thus increases cell count per unit volume, as well as reducing the oxygen uptake of the red cel l by the amount consumed by the nucleus (Foxon, 1964).

Transfer of oxygen to respiring tissue

Once oxyhemoglobin is carried to the tissues from the respiratory organs, oxygen dissociates from it and diffuses into the cells. Krogh (1941) has shown that the rate of oxygen diffusion in tissues is variable, but less than half that of water. Less is known of conditions which promote dissociation of oxyhemoglobin of the type found in tadpoles. Since the great affinity which tadpole hemoglobin has for oxygen is not modified by changing pH, dissociation would appear more difficult than it is for oxyhemoglobin of adult erythrocytes (Manwell, 1966). Oxygen utilization and metabolic rate

Boell (1948) has reviewed earlier work which related to oxygen consumption of early amphibian embryos. Much of this was concerned with respiration of various regions of the gastrula stage. The work of










21

Brachet (1935) upon eggs of Rana fusca (= Rana temporaria) showed no significant difference between respiratory rates of fertilized and unfertilized eggs.

Among the first studies of oxygen uptake in anuran larvae

were those of Helff (1926) and Etkin (1934), working upon Rana pipiens and Rana catesbeiana respectively. Atlas (1938) studied Rana pipiens

and Rana sylvatica from fertilization to tadpole states and found that respiratory rates increase from fertilization to overgrowth of the operculum. This was in agreement with the studies of Godlewski (1900) and Bialaszewicz and Bledowski (1915) on Rana temporaria. Parnas and

Krasinska (1921) found the rate of oxygen consumption rose sharply at the onset of gastrulation, neurulation, and external gill formation, with no noticeable changes between. Atlas (1938) suggested that the latter conclusions were based upon insufficient data which wereadversely affected by temperature fluctuation. It is interesting to note that the respiratory rates obtained by Bialaszewicz and Bledowski for Rana temooraria are similar to Atlas' data for the closely related Rana sylvatica, but somewhat different for Rana pipiens or Rana catesbejana. Atlas concluded that the increased rate of respiration per embryo was

a function of cell number rather than of intensity of respiration of individual cells. The basis for this statement is obscure, since no cell counts were reported, nor was the subject discussed in the main

text of the paper.










22


Relation of Respiration to External Environment

Although a number of studies have dealt with respiration in

regard to external factors such as temperature and partial pressure of oxygen, little has been done to relate respiratory form and function of a tadpole to its specific environment. The most thorough study along these lines was done by Savage (1961), who compared several species of tadpoles in relation to feeding and respiration in different environments. His conclusions were quoted previously.

Laessle (1961) noted that the tadpoles of the common Jamaican tree frog, Hyla brunnea, lived in stagnant water of very low oxygen content, which collects in the central reservoirs and leaf bases of bromeliads. He suggested that the agitation of the long tail increased oxygen diffusion into the water and that the large surface area of the tail provided additional respiratory surface. The tadpoles of the other three species of Jamaican tree frogs also develop in the oxygen-poor water which accumulates in bromeliads (Dunn, 1926). Tree frogs, closely

related to the Jamaican species, on the island of Hispaniola, are known to breed in streams and torrents (Noble, 1927) ; the Cuban tree frog, Hyla septentrionalis, is known to develop in brackish water, as well

as in cisterns and pools (Grant, 1940).

It may be noted here that Powers et al. (1932), working upon fish respiration in relation to environment, found that ''The number of red blood corpuscles is increased with a decrease in the oxygen and by

an increase in the carbon dioxide tension of the water and vice versa."










23


Structure and Function of Froa Liver

Studies of liver structure in frogs have been few; most were considered by Elias and Bengelsdorf (1952). These authors found that the walls separating neighboring sinusoids are predominantly two cells thick in frogs, but only one cell thick in mammals. Similarly, relatively little experimental work has been done on the excretion of bile pigment by frog liver. Nisimaru (1931) carried out perfusion experiments on the liver of Rana catesbeiana. He found that the rate of bile pigment excretion changed when blood pressure in the liver circulation was altered.

Papers concerning frog liver structure or function do not aid in the explanation of chlorosis in these forms.














SOURCES AND CARE OF TADPOLES AND FROGS


The live frogs and tadpoles used for most of this study were collected from their normal habitats during field trips to the West Indies and South America. The species included: Hyla septentrionalis collected from Grand Cayman and Southern Florida; Hyla brunnea, Hyla lichenata, Hyla marianae, and Hyla wilder from Jamaica; Hyla dominicensis, Hyla vasta and Hyla heilprini from Haiti; Leptodactylus

albilabrus of Puerto Rico; and Hyla maxima and Pseudis paradoxus from Surinam. Incidental observations were made of Phyllomedusa tadpoles on Trinidad and several other hylid species in Surinam. Collections were made at the following places during the months indicated: Jamaica -June-July, 1965; August, 1965; October, 1965; May, 1966 and August, 1966; Haiti -- October, 1965; May, 1966 and August, 1966; Grand Cayman -July, 1965 and October, 1965; Puerto Rico -- May, 1966; Southern Florida -- July, 1966; Trinidad -- July, 1966; Surinam -- July, 1966.

Tadpoles of the species listed above were collected with a tea strainer or small dip net. They were transferred to plastic bags which were partially filled with water from the immediate environment of the tadpole. When suitable tap water was available, it was used to replace water brought from the field. Thus the frog larvae were transferred to water which was apparently free of noxious substances, detritus, and other organisms. Water was poured off and replaced with fresh tap water as often as required, usually at one- to three-day intervals. Tadpoles did not show marked reactions to addition of tap 24










25

water from localities in South America, the West Indies, or Southern Florida. Care was taken to keep the temperature of the fresh tap water about the same as that which had been removed.

Upon returning to the laboratories in Gainesville, Florida, the larvae were transferred to culture dishes of appropriate size. Since Gainesville tap water was found to be lethal to both Hyla brunnea and Hyla septentrionalis, spring water or Holtfreter's solution was ordinarily substituted. During May, 1966, Holtfreter's solution proved unsatisfactory and its use was discontinued. At one point, the spring water appeared harmful and filtered pond water was used in its place.

By avoiding direct sunlight wherever possible and using other precautions, tadpoles were maintained in the field with few losses due to excessive temperature. At the University of Florida, the culture

dishes containing the tadpoles were moved from one laboratory to another in an attempt to maintain a moderate temperature; however, groups of tadpoles were exposed to water temperatures between known extremes of 160 C. and 280 C.

Thus, with reasonable care, most of the tadpoles were collected and transferred to Gainesville with little difficulty. A notable exception was the Hyla heilprini tadpoles, which did not survive more than three days after capture. Of the other species, most demonstrated an ability to live for two weeks or more without being fed. However, in the laboratory and whenever possible in the field, the tadpoles were fed Gerber's strained foods at regular intervals. Hyla septentrionalis, Hyla dominicensis, Hyla vasta, and Leptodactylus albilabrus were fed a










26


mixture of peas, carrots, and spinach. Hyla brunnea, Hyla lichenata, and Hyla maranae were fed strained egg yolk. Tadpoles were maintained satisfactorily for two months or more on these diets. Of those on the above diets, individuals of all species but Hyla septentrionalis reached metamorphosis under laboratory conditions. Following metamorphosis, the young frogs were kept in covered jars, and were fed termites and other insects.

Adult frogs were collected by hand, using a flashlight at night to locate them at their calling locations, or by searching for them in their resting sites during daylight hours. They were transported in jars or in plastic or cloth bags. In the laboratory, they were kept in jars or terraria. Crickets or other insects were used to feed the adult frogs at approximately weekly intervals.

In addition to live specimens, hundreds of preserved frogs

and tadpoles were surveyed for unusual characteristics. Most of these specimens were of West Indian hylids and were obtained from the museums mentioned under Acknowledgments.















I. CAUSATIVE PIGMENT AND INCIDENCE OF CHLOROSIS Pigment Identification


During the present study, filtered tissue extracts and body fluids were tested spectrophotometrically and chemically. Figures 2-6 show that the spectral characteristics of several of these green fluids are very similar to those of biliverdin (Figure 1). Addition of fuming nitric acid to the fluids was followed by the succession of colors known as the Gmelin reaction, generally considered to indicate the presence of bi le pigments. These two tests, along with characteristics of color and solubility, appear to eliminate from consideration all known green pigments except the two naturally occurring green bile pigments, biliverdin and mesobiliverdin. Since green bone is rapidly bleached by concentrated sulfuric acid, mesobiliverdin does not appear to be present, since it is stable in this substance. Thus, it was independently concluded during the present study that biliverdin is the pigment primarily responsible for the green color of these frog tissues.

Survey of green pigmentation among Neotropical anurans

During the course of several field trips to the West Indies and one to Surinam, it was possible to study a variety of tropical anurans under natural conditions. Many of these did not exhibit any green color, while others showed different levels of green pigmentation in the tissues and body fluids. Observations made during the present 27













study and by others are summarized in Tables I and 2. Points to be noted from Table 1 include the fact that the color of the bile was usually green. In many individuals, the bile was so concentrated as to appear blue when seen through the gall bladder wall. When soft tissues were green, it was usually a rather general phenomenon and often was accompanied by a relatively high concentration of plasma pigment, a condition which might be termed chlorosis. Sex and its relationship to chlorosis in frogs

Due to the relatively small number of individuals collected from each species, it is difficult to make a generalization concerning the relationship between chlorosis and sex. Chlorotic and non-chlorotic individuals of both sexes were among specimens of Hyla septentrionalis from Grand Cayman. Table 2 shows several species in which all tadpoles observed had green pigment at the time of metamorphosis or before. The

general occurrence of bi liverdin in these species indicates that the sex of the individual is not important to the development of chlorosis at this stage. The presence of green pigmentation in calling males of several species and in eggs of certain Hyperolius, Acalychnis, Centrolenella, Pseudis and Hyla indicates that sex is not a very important factor in the development of chlorosis. Age distribution of chlorotic anurans

Among tadpoles, no green pigmentation was noted in young tadpoles of Hyla brunnea, Hyla boesemani, Hyla dominicensis, Hyla leucophyllata, Hyla maxima, Hyla rubra, Hyla septentrionalis, Hyla vasta, or Hyla wilderi.










25

When green pigmentation was present in tadpoles, usually it made its appearance at the onset of metamorphosis and reached a peak upon the completion of metamorphosis. All of the species of Hyla listed in Table 2 are from the West Indies and they present similar appearances at the completion of metamorphosis, except for Hyla marianae. Most of their skeletons were green, particularly the limb bones and vertebrae; the green pigmentation was not so pronounced where the marrow maintained its hemopoietic function; melanin pigmentation had developed, especially on the dorsal side; pigmentation of the ventral side was not well developed in Hyla brunnea, Hyla lichenata, or Hyla marianae, all of Jamaica (stomach contents could be seen through the skin); a white substance, presumably guanine, had developed in the skin of the other species, which are usually found in more exposed habitats than the Jamaican species; green pigmentation of soft parts was particularly intense in the region of the throat and pectoral girdle; the small size of most of these species of tadpoles made it difficult to determine the color of the plasma. In Surinam, the tadpoles of Pseudis paradoxus were found to be darkly pigmented, externally by melanin and internally by biliverdin and other pigments. The plasma of this species is green prior to metamorphosis. The coiled intestines of these gigantic tadpoles are packed with green plant materials and fill most of the body cavity.

During the present study, a young Rana heckscheri collected

near Gainesville, Florida, at the completion of metamorphosis was found to have gray-green bone marrow. This appeared to be a stage which represented degenerating red marrow of the tadpole. This color appeared













to be restricted to the marrow, since neither the bone nor the plasma demonstrated green pigment. Although the origin of this pigment probably was the same as that of the Neotropical forms, the color was almost obscured by the bone.

Green pigmentation of the adults is summarized in Table 1.

It is worthwhile to compare the amount of green pigment in adults. with that of tadpoles of the same species, where possible. Tadpoles of Pseudis paradoxus, Hyla dominicensis, Hyla lichenata, Hyla septentrionalis and Hyla vasta generally had much more extensive chlorosis of soft tissues than did the adults of these species. Two adult specimens of Hyla heilprini showed chlorosis as extensive as that of metamorphosing tadpoles, but melanin pigments of the dorsum were better developed in the adult than in the tadpoles. While scores of tadpoles of Hyla brunnea invariably had green pigmentation, dozens of adults showed no biliverdin in tissues, including bone. Green pigment is absent from the tissues of Hyla marianae, which are orange in color.

As indicated previously, there was individual variation of green pigmentation within populations of Hyla septentrionalis. Small species such as Hyla wilderi have bones of a more intense green color than do larger species such as Hyla maxima or Pseudis paradoxus. Calcification of bones tends to obscure the green pigment, as in the larger species, but the color of bi liverdin is quite apparent in poorly calcified bones as in the distal limb structure of Hyla wiIderi.










31


Seasonal changes in chlorosis

It should be noted that most of the specimens for the present study were collected from June through August. This period constitutes the main breeding period for most of the species studied, a fact which should be kept in mind when considering the physiology of these organisms.

It is interesting to note that Hy la geographica had green bones, but no green pigment in the plasma. On the other hand, Hyla misera had light green plasma, but white bones. These two examples indicate that high concentrations of green pigment in the plasma are temporary in some species. Different shades of green in plasma and bones of other species tend to support this idea. Concentric layers of different shades were seen in Barrio's (1965a) photograph of bone from Lysapsus mantidactylus, and in a femur of Osteocephalus taurinus. Like growth rings of a tree, these suggest a seasonal change in conditions during development of the tissue.

Phyloqenetic distribution of chlorotic frogs

An understanding of the phylogenetic relationships of chlorotic frogs to those which lack tissue biliverdin should be of value. However, Tables I and 2 indicate no clear boundaries along phylogenetic lines. As previously mentioned, adult individuals of Hyla septentrionalis and Hyla dominicensis may fall into either category. In Hyla brunnea, the pigment is always present in young frogs, but has never been observed in the adults. Hyla pulchella has some populations which are chlorotic and others which are not. Similarly, it can be seen that neither the genus Hyla nor the family Hylidae shows uniformity in this characteristic.











32

The degree of pigmentation is no more helpful than its presence or absence. Of the species collected in Surinam, Hyla punctata, Hyla crepitans, Sphaenorhynchus aurantiacus and Phrynohyas venulosa had the highest concentration of green pigment of the frogs collected, but they certainly do not constitute a homogeneous group of species. Of the eight species of Hyla that were studied in life in the West Indies,

seven had green pigmentation at metamorphosis or later, but Hyla marianae lacks green pigmentation in these stages, being orange instead.

There is at least one relationship between chlorosis and frog phylogeny. Thus far, all chlorotic frogs have been members of only four families. Most of the green species are members of the family Hylidae, while the others are included in the Centrolenidae, Pseudidae, and Rhacophoridae. The chlorotic hylids, centrolenids and rhacophorids are all tree frogs, while pseudids are aquatic for extensive periods. All four of these families have in common an intercalary cartilage between the ultimate and penultimate phalanges of the digits.

A close relationship between the four families is unlikely

even though they have the intercalary cartilage, green pigmentation and tropical distribution in common. Hyperolius, the African genus of rhacophorid frogs which has green eggs, has a firmisternal pectoral girdle and diplasiocoelous vertebrae which distinguishes it from chlorotic frogs of the other three families, which are procoelous, have an arciferal pectoral girdle and are Neotropical. There appears to be considerable agreement that the presence of intercalary cartilages in










33


these families is the result of parallel evolution. All things considered, there appears to be no clear relationship between phylogeny and chlorosis among frogs.

Geographic distribution

Frogs with green tissues or eggs appear to be tropical or

sub-tropical in their distribution. Below is a list of green pigmented

-'rogs according to the countries or regions in which they are found. Sources of information are given in Table 1.

Central Africa: Hyperolius sp.

Jamaica: Hyla brunnea, Hyla lichenata, Hyla wilder.

Hispaniola: Hyla dominicensis, Hyla heilprini, Hyla vasta.

Cayman Islands (presumably also Cuba and Bahama Islands): Hyla septentrionalis.

Mexico and Central America: Agalychnis dacnicolor, Centrolenella albomaculata, Centrolenella granulosa, Centrolenella ilex, Centrolenella spinosa, Centrolenella prosoblepon, Centrolenella pulveratum.

Northern South America, including Brazil: Lysapsus limellum laevis, Pseudis paradoxus, Anotheca coronata, Hyla albofrenata, Hyla albomarginata, Hyla boesemani, Hyla calcarata, Hyla crepitans, Hyla cuspidata, Hyla geooraphica. Hyla langsdorffi, Hyla maxima, Hyla misery, Hyla punctata, Osteocephalus taurinus, Phrynohyas venulosa, Sphaenorhynchus, Trachycephalus niqromaculata, Centrolenella vanzolinii.

Argentina: Lysapsus limellum limellum, Lysapsus mantidactylus, Pseudis paradoxus platensis, Hyla berthae, Hyla nasica, Hyla pulchella,










34

Hyla phrvnoderma, Hyla ounctata rubrolineata, Hyla raniceps, Hyla siemersi, Hyla squalirostris, Hyla trachytorax, Phrynohyas venulosa.

Judging from this list, it appears that nearly all green

pigmented frogs are to be found in the Neotropical Zone. It should be noted that, except for the works of Barrio (1965a, 1965b) and the present writer, the references to green tissues or eggs of frogs are few, and these are often obscure. Although a number of individuals have volunteered personal observations concerning coloration, there appears to be a general reluctance to publish such observations. It seems quite possible, if not likely, that additional records will be forthcoming from Africa or other parts of the Old World tropics.















11. CONSIDERATION OF ECOLOGICAL FACTORS ASSOCIATED WITH CHLOROSIS


In attempting to determine the cause of a condition such as chlorosis, ecological factors should be considered. It has already been noted that the majority of chlorotic anurans are tree frogs which are restricted to tropical or sub-tropical regions. However, not all tropical tree frogs have green tissues; conversely, the chlorotic pseudids are not tree frogs, but are aquatic. Since the tropical climate and general habitat (aquatic, terrestrial or arboreal) do not completely account for either presence or absence of chlorosis, one should consider the importance of specific habitats, as well as behavioral adaptations to such environments. This section includes a summary of information concerning specific habitats and the frogs associated with them. Greater emphasis was placed upon ecological study of breeding sites because of their accessibility and the appearance of chlorosis during the larval stages. Special attention was given to those factors which might be related to hemoglobin or red cell formation and function, including oxygen tension, temperature and iron concentration. Additional factors were surveyed in order to find conditions which were markedly different from those ordinarily encountered by anurans.


Methods of Study

Concentration of dissolved oxygen in the habitats of tadpoles was measured with a Precision Scientific Portable Oxygen Analyzer, which

was calibrated in air.


35











36

Temperature readings of tadpole habitats were taken with the thermistor component of the oxygen analyzer wherever possible; otherwise, they were made with standard mercury thermometers, graduated from -100 C. to 1000 C.

Ferrous and total iron, and phosphate concentrations of water were measured by colorimetric methods using a Hach portable colorimeter and the appropriate techniques (Hach Chemical Company, no date). Salinity, alkalinity and hardness were measured by titrametric methods,

using the appropriate kits and techniques developed by the La Motte Chemical Company. Water pH was measured to the nearest half unit by means of pHydrion paper.


Observations of Frog Habitats and Behavior

During this study, frogs were collected in a variety of

habitats from tropical rain forest to open, grassy areas to residential districts. Although eggs and tadpoles were found in or near the environments of the adults of their respective species, the larval environs were more uniform in appearance. Most of the tadpoles were found in standing water, but several species of tadpoles were found in flowing streams. While the temperature and dissolved substances of these habitats were found to vary considerably, one may satisfactorily divide tadpole habitats into flowing stream and standing water types. Within the latter type, special consideration will be given to the bromeliad microhabitat of Jamaican tree frogs. Ecological data collected during this study are presented in Tables 3-5.













The bromeliad microhabitat

Members of the pineapple family, Bromeliaceae, ordinarily do not constitute the dominant plants of a habitat, although they may be an important part of the flora. On the West Indian island of Jamaica, the bromeliads have undergone considerable adaptive radiation (Dr. Richard Proctor, personal communication). They may be large or small, epiphytic or terrestrial, and are found in shaded and open areas. This appears to be a very fortunate circumstance since the water which is caught in the leaf bases and central reservoirs of bromeliads is the most reliable supply for small animals, including the four species of tree frogs on Jamaica.

The importance of the close relationship between the Jamaican tree frogs and the bromeliads should not be underestimated. Perkins (1948) noted that the water level in the bromeliads ("wild pines") is maintained by dew which condenses and runs down into the reservoir in the center of the plant; very little direct sunlight tends to reduce evaporation from the wild pines. She states further (p. 87), "In view of the many creatures that depend on the wild-pine for moisture it would seem that these plants hold an important place in the economy of the countryside, for surely our wildlife would be largely depleted during a severe drought, were it not for these hidden stores of water." My observations during the drought which continued into July, 1965, substantiate this position. Except for an occasional pool in stream beds, or the largest rivers which continued to flow, there was no surface











38

water other than that in the bromeliads. Steepness of the hills and porosity of the limestone substrate are partly responsible for rapid run-off of water.

Observations of the adults of the three smaller Jamaican hylid

species indicate that they prefer to be covered by water, at least in a lighted area. Under these circumstances, they may remain completely immersed for minutes at a time, and then slowly rise until only the external nares and eyes protrude above the surface. It is interesting to note that the four Jamaican species differ from their relatives of Cuba and Hispaniola in having a more truncate snout with the external nares at the most anterodorsal point (Dunn, 1926). This may be interpreted as an adaptation to living in the reservoirs of bromeliads.

Laessle (1961) studied the ecology of Jamaican bromeliads, in which he found the following ranges for the water which they contained: dissolved oxygen, 0.0 8.0 ppm; dissolved carbon dioxide,

4.0 67.0 ppm; pH, 4.0 7.0; temperature, 17.5" 30.00. He estimated the maximal quantity of water in the reservoir of a large bromeliad at

200 milliliters.

Hyla brunnea. -- The brown tree frog of Jamaica is the most widely distributed species of Hyla on the island. It is absent from the Blue Mountains above 1600 meters elevation, and from arid areas such as Kingston and the Helishire Hills along the south coast (Lynn, 1940). This frog is most likely to be found in localities where the large tank bromeliads, including species of Hohenbergia and Aechmea, are readily available as breeding sites and resting places. While this frog













may be found in bromeliads which grow at ground level on shaded, limestone hillsides or in the epiphytes high above ground in well developed forests, they are frequently found in towns or among the trees which line the roads in agricultural areas. Like the other tree frogs of Jamaica, Hyla brunnea appears to be largely dependent upon the bromeliads

for the water which is retained at the leaf bases, but several Jamaicans have told me that the brown tree frog is found among banana leaves and stalks.

Breeding habits of Hyla brunnea are unusual in some respects, particularly as they relate to adaptations to life in the bromeliads. The breeding season begins during May, as indicated by a record cited by Dunn (1926). Panton (1952) has stated that the strongest choruses of the tree-toad are heard at the end of May or beginning of June, but that the time fluctuates because of weather conditions. During the latter part of May, 1966, and early June, 1965, I noted very few tadpoles, but found eggs more common than later in the year. Tadpoles were not uncommon during October, 1965. The eggs of Hyla brunnea are deposited in the central reservoir of bromeliads in most cases, but eggs or tadpoles may be found at the bases of outer leaves; on one occasion, eggs were found in water that had collected on a hollow branch.

Development of the tadpole stages has been described by Schreckenberg (1956), who studied the embryonic development of the thyroid gland in Hyla brunnea. The mode of development differs little from that seen in Rana and hylids of the United States. The main differences appear to be the result of adaptations to the bromeliad habitat.












The jelly mass from which the tadpoles hatch remains in the bromeliad reservoir longer than do the tadpoles themselves. Panton (1952) has suggested that the presence of the jelly reduces evaporation and moderates temperature changes. The low pH of the water probably prevents the rapid decay of the jelly mass or infertile eggs. In his micro-limnological study of Jamaican bromeliads, Laessle (1961) found the following ranges of readings in five bromeliads which contained eggs or tadpoles of Hyla brunnea: dissolved oxygen, 0.03 2.3 parts per million; dissolved carbon dioxide, 23.0 41.0 ppm; pH, 4.0 4.5; temperature, 23.0 25.00 C. Additional measurements made during the present study increased these ranges to: dissolved oxygen, 0.03 2.7 ppm; pH, 4.0 6.5; temperature, 23.0 28.00 C.

During the present study, the contents of a number of bromeliad reservoirs were poured into waterproof containers. When jelly masses without eggs were present, they were not firm and had a tendency to separate into capsules 1-2 centimeters in diameter. These probably represent capsules which contained four to six eggs each, as mentioned by Dunn (1926). Within the bromeliad, the mixture of water and jelly has the consistency of glycerine, as mentioned by previous writers. While I have noted tadpoles, especially small ones, moving about near the surface on several occasions, I have also noted larger ones moving vertically within the reservoir. The larger individuals tend to remain under leaf fragments when the reservoir is exposed to the sun. They come to the surface at intervals to take air and then return to lower depths. This respiratory behavior apparently was not observed previously, since













Dunn (1926) was unable to reconcile the reduced gill structure with the low oxygen content of the environment. Lungs are seen on either side of the vertebral column, and appear like small bubbles. Lungs are present in very small tadpoles as well as older ones. They appear to be used as accessory respiratory structures when branchial and cutaneous respiration are insufficient. In this respect, the tadpoles would be similar to the Australian lungfish, which utilizes pulmonary respiration largely at night when it becomes more active (Grigg, 1965). While the

long, narrow tails of Hyla brunnea tadpoles probably are important to cutaneous respiration as Laessle (1961) indicated, it seems likely that their primary purpose is to propel the organisms through their viscous environment.

The diet of Hyla brunnea tadpoles consists largely of frog eggs, as noted by Dunn (1926) and Laessle (1961). Most often the eggs are probably those of its own species, since it is by far the most common tree frog of Jamaica. In addition, it is the only hylid species known from the eastern third of the island so that any hylid eggs found in Hyla brunnea tadpoles there, would have to be of the same species (Laessle, 1961).

Several structural characters of the tadpoles can be correlated with their unusual diet. The digestive tract is expanded into a sac in which eggs may be found more than a week after the last feeding. It is relatively straight, not coiled as in vegetarian tadpoles. Since the gills are not important in collection of food, their structure is










42

simplified. Finally, the mouth has only a single row of teeth about

it (Dunn, 1926), since these are not needed for scraping food from surfaces.

The length of time required for development from fertilization to metamorphosis is not known since neither Lynn (1940) nor I was able to follow a single group through the entire period. From Lynn's work, nearly a week passes between fertilization and hatching. After hatching, at least a month probably passes before metamorphosis is completed. While tadpoles have been kept in captivity for more than six weeks, it is likely that metamorphosis could be completed approximately six weeks after fertilization under optimal conditions.

At the time that the forelegs penetrate the opercular fold,

the green pigmentation was always quite obvious. Jarring of the bromeliad or container in which such tadpoles were present caused them to climb upward very rapidly. They have no difficulty in climbing out of a bromeliad reservoir, a water glass, or a plastic bag. Once they leave the water, the tail is resorbed in twenty-four to thirty-six hours. This is in agreement with the finding of Schreckenberg (1956) that there is intense thyroid secretory activity and sudden release of colloid from the thyroid gland at the stage of tail resorption. The dark green pigmentation of

soft tissues in the gular region remains for two weeks or more after metamorphosis and then gradually fades away in the living animal. Green pigmentation of bones remained as long as the young frogs lived, or about six weeks after metamorphosis, in the laboratory. In nature, no immature










43

frogs were found in bromeliads, ant or termite nests, or elsewhere, so that the post-metamorphic development and ecology of the young are unknown.

Hyla lichenata. -- This giant tree frog, which attains a

length of 117 millimeters, is restricted to the central and western hills of Jamaica, generally above 300 meters in elevation. Since this frog is rarely seen and less often collected, knowledge of its habits and distribution is based largely upon hearing its distinctive snoring call (see Lynn, 1940).

In regard to the habits of this species, Panton (1952) first found one in a bromeliad on a dead candlewood tree in woodland, but all subsequent specimens were taken from hollow trees. Dunn (1926) collected three of these frogs from small hollow trees with openings 4 to 12 feet above the ground, but noted that they ordinarily call from greater heights. He stated further that the ''bony head is obviously of use in plugging the hole after the frog is inside.'' Lynn and Dent (1943) traced the unmistakable call of this species to a clump of bamboos near Chapelton. A female specimen of Hyla lichenata was collected by Dr. Thomas Farr of the

Institute of Jamaica, near the entrace to St. Clair Cave, St. Catherine Parish on June 19, 1965. During several days in captivity at the Institute of Jamaica, this individual deposited a number of eggs. It showed no interest in a cockroach which was offered as food. This individual produced a skin secretion which became gum-like on the hands of those who held it. Gosse (1851) recorded an instance where the skin secretion of this species caused severe irritation to the human eye.













During my visits to Jamaica I heard Hyla lichenata as far

east as localities near Moneague, Ewarton and Lluidas Vale. These calls could be heard for nearly half a mile and invariably came from wooded slopes. This species did not venture out into flat, open valleys, as did Hyla brunnea. In attempting to collect this species while it called at night, I found the vocal individuals to be in trees, with one exception. North of Mandeville, it seemed that one frog was calling from underground. This observation, along with the presence of Dr. Farr's specimen near a cave entrance, leads one to suspect that these large frogs may take refuge in cave entrances and crevices in limestone, both of which are present in quantity in the range of Hyla lichenata.

Dunn (1926) found four tadpoles of Hyla lichenata in a bromeliad 20 feet up in a small tree, in August, 1925. I collected two large

tadpoles of this species from a large bromeliad situated on the ground under the shade of small trees. These were taken in Rose Valley on October 25, 1965. These large specimens could be distinguished from tadpoles of Hyla brunnea by their large size, relatively shorter and more muscular tail, and slightly different mouth structure. They were only slightly darker than Hyla brunnea tadpoles, not black as were those collected by Dunn. Like other Jamaican hylid tadpoles, they feed upon frog eggs. One of my two specimens underwent metamorphosis and died about six days later. At the time of its death, it weighed 660 milligrams and had a snout-vent length of 18.7 millimeters. This individual developed green pigmentation, similar to that of Hyla brunnea, in its bones and other tissues.









1 45

Hyla wilderi. -- This small, green species is said to be quite common in bromeliads about Mandeville, Jamaica, "and appears to have a fairly wide range in the central part of the island above 1,000 feet" (Lynn, 1940). Dr. Albert Laessle (personal communication) found this species common on Juan de Bolas mountain, from which there is a large series of specimens in the collection of the Museum of the Institute of Jamaica. I was unable to collect a specimen there although I heard a call which may have been of this species. The call of this species is a faint clicking sound, which becomes louder as the frog continues to call. However, in localities near Moneague and Fishbrook, I traced similar calls to frogs which appeared to be of the genus Eleutherodactylus. Unfortunately, both of these frogs escaped. During August, 1965, I collected three adults from the Cockpit Country four miles north of Quickstep, and a tadpole near Moneague. Another individual was taken from Crown Lands in the Cockpit Country during May, 1966.

During the present study, adults of this species were found in large as well as small bromeliads. My experience was similar to Dunn's (1926) in that this species appeared most often in open woods, usually with a southerly exposure. While Dunn (1926) collected one Hyla wilder tadpole for each seven of Hyla brunnea, only one Hyla wilderi was collected during the present study, along with more than 500 Hyla brunnea tadpoles. The tadpole was collected from a large bromeliad in one of several trees in a pasture. It was not darker than the Hyla brunnea tadpoles collected at the same time, but did appear to have a greater density of melanophores. The hind limbs on this specimen did not have










46

green pigment. Tadpoles of Hyla wilderi have been collected in March, April I and August (Lynn, 1940).

The tadpole of Hyla wilderi has poorly developed gills,

similar to those of Hyla brunnea, and like other Jamaican tadpoles, its body is depressed, and has no fin.

Dunn (1926) found that the specimens which Barbour (1910) considered pale green young of Hyla brunnea were actually Hyla wilder. In fact, Dunn believed that the non-ossified head and green bones of adult Hyla wilderi represented a neotenic condition because of its resemblance to the young of Hyla brunnea and related forms. He suggested that Hyla wilder, as well as Hyla lichenata and Hyla marianae arose from a

frog of the Hyla brunnea type through sympatric speciation, because of differential growth rates.

Hyla marianae. -- Of the Jamaican hylids, least is known of Hy]a marianae. Dunn (1926) originally considered these yellowish-green or greenish-brown frogs to be the young of Hyla brunnea, but then recognized them to be of a different species, which he described. The range of this species appears to be the most restricted of the Jamaican hylids. Most of the specimens have been collected in or near the Cockpit Country of west-central Jamaica; two specimens came from Hollymount, Mt. Diabolo, further to the east (Lynn, 1940; Goin and Cooper, 1950). This area has a combination of high elevation (above 400 meters), greater rainfall and less disturbance of habitat than surrounding areas. Limestone cliffs covered by lianas and hillsides covered with jagged, honeycomb limestone usually form the steep sides of the ''cockpits." Vegetation on these










47

hillsides varies from sparse, scrubby growth, to saplings which shade an understory of terrestrial bromeliads, to large trees. The areas of sparse vegetation usually are higher on the hills and have a more southerly exposure than do the areas of denser vegetation. Dunn took his frogs from "wild pines' in rather thick woods. I collected one male, two females and four metamorphosing tadpoles of this species from small, compact bromeliads along the border between a sunny clearing and a thick growth of saplings on the slope above. Hyla wilder and Hyla brunnea adults were collected from the same site.

Three of the four Hyla marianae tadpoles were in the same small bromeliad, but between different sets of leaves. All of them were engorged with eggs. The tadpole mouthparts had been lost and the mouths of these tadpoles were quite wide. Although the adults of this species (28-38 mm) were slightly more than half the length of Hyla brunnea (56-57 mm) adults, the tadpoles of Hyla marianae appeared to be about the same size as Hyla lichenata tadpoles at metamorphosis. Like the tadpoles of Hyla brunnea, those of Hyla marianae were quite active and tended to move upward at metamorphosis. The color of these tadpoles was brown. The adults which I collected could change from orange to brown and back. The bones of adults and tadpoles were orange with no trace of green in the soft tissues.

Pond habitats

During the course of this study, a relatively large number of frog species wae found breeding in bodies of standing water. In addition to the bromeliad habitat already considered, such environments included:










48

ponds, pools, puddles and wheel ruts filled with water; roadside ditches and drainage canals; flooded, grassy meadows; and cattail marshes. These waters varied widely in size, amount of exposure to sun and wind, and in water characteristics (Tables 4 and 5). Temperatures at mid-day ranged to 400 C. in shallow open waters, but general ly were less than 350 C. when exposed to direct sunlight for only a few hours; where direct sunlight did not reach the pools, temperatures remained under 300 C.

Oxygen concentrations varied from less than one part per million to ten parts per million and appeared to be affected by air currents and photosynthesis as well as temperature.

The pH of standing surface waters in the West Indies appeared to average somewhat higher than the pH of water in bromeliads in the region. This may be due to the fact that the surface strata of most of the West Indies consist of limestone. At a locality near Juan de Bolas in Jamaica (Table 4, Habitat 7), the pH of water in three bromeliads without an obvious layer of dust on the leaves ranged from 5.5 to 6.5; three bromeliads near the road had a heavy covering of dust on the leaves and the pH of the water in the reservoirs was 7. Presumably the dust from the limestone gravel of the road reduced the acidity of water in bromeliads when it was washed into the central reservoirs by rain or dew. Other water characteristics do not show definite patterns (Tables 4 and 5).

It is difficult to make a distinction between exposed and shaded pool habitats. There may be greater difficulty in assigning a particular species to one or the other. Hyla dominicensis and Hyla









49

septentrionalis are quite opportunistic in regard to selection of breeding areas. However, in Surinam, several species appeared to be associated almost exclusively with forest habitats. Included in this group are Hyla ceoqraphica, Hyla calcarata, Hyla lanciformis, and Osteocephalus taurinus, all of which are medium-sized and brown dorsally, and the large species, Phyllomedusa bicolor, which is green above.

Hyla geographica. -- A silent male of Hyla aeocraphica was collected from vegetation on a rocky ledge of Princess Irene Falls, Brownsberg, Surinam. Wooded slopes were present on both sides of the falls and the stream which flowed over it. Tadpoles were present at the base of the falls, but these were believed to be larvae of Hyla maxima, which was calling at the locality. This specimen of Hyla geographica had light green bones but no other green organs.

Hyla calcarata. -- The single female was collected from a bridge over a small stream which flowed into a swamp. This individual had green

bones, but no other green pigmentation.

Hyla lanciformis. -- Most of the specimens of Hyla lanciformis were collected in wooded areas. One was collected while calling near the edge of the forest in the vicinity of temporary pools. No eggs or tadpoles were found. Goin and Layne (1958) stated that its note was regularly heard in open lands such as wet meadows near Leticia, Colombia.

They noted that the specimens collected were perched on bushes or other plants one or two feet above ground. Four specimens from Surinam lacked green pigmentation in bone or other tissues.










50

Osteocephalus taurinus. -- A female with green bones was taken from the trunk of a tree on the upper coastal plain in Surinam. The tree was beside a road through a well-developed forest; a stream and standing puddle were nearby.

Bokermann (1965) has recently given the first description of

the breeding habits of Osteocephalus taurinus. Rainfall in the vicinity of Marmelo, in forests of western Brazil, is strongly seasonal, but reaches 2250 millimeters annually. Osteocephalus taurinus choruses were heard at the time of the first rain in November. Two of the breeding pools used by the species were in the forest; in one of these ponds, there was breeding activity at 1:00 P.M. The surfaces of these ponds were covered by eggs of this species which were 5 millimeters in diameter, including the jelly envelopes. A temperature reading from one of them

indicated the water temperature to be 270 C., while the air temperature was 340 C.

Of particular interest was a third breeding pond found by

Bokermann. This pool was in an open area and contained the dead crown of a tree which made difficult the capture of frogs beneath the branches. However, as the collectors approached the pool, the frogs followed their habit of rapidly climbing upward to escape capture until they reached the ends of the tree's bare branches. After eight minutes in this exposed position, the first frog fell into the pool, motionless but not dead. Although its skin had dried in several places, it recovered. Air temperature in the sun was 460 C. and the water temperature was 320 C. at several points of the pond. Two days later, evaporation had greatly










51

reduced the size of this pond. Water temperatures rose to 380 C. and the embryos were dead.

Phyliomedusa bicolor. -- This is a large arboreal species which appears to breed near stagnant water which may be in open areas with some exposure to the sun. One adult male was collected from the surface of a main road on the upper coastal plain of Surinam. Another was heard calling from a site high up in a tree in the forest near Powakka. The male which was collected had no green pigmentation in the tissues. As

with other members of this genus, this species moves very slowly, never hopping, but walking on all fours. The whitish color of muscular tissue in this individual may be an indication of poor vascularization or a relative lack of myoglobin.

Phyllomedusa eggs and tadpoles believed to be those of

Phyllomedusa bicolor were observed at the New York Zoological Society's biological station at Simla, Trinidad. The large, yellow eggs were attached to the upper side of a large shaded leaf which overhung a pool. Water lilies and water hyacinths covered part of the surface of this small,

ornamental pool, which harbored the Phyllomedusa tadpoles of various sizes. The tadpoles were slaty-blue in color, much the same color as the adult dorsum in preservative. They fed upon aquatic vegetation and were observed coming to the surface. Lutz (1954) stated that the larvae of Brazilian species of Phyllomedusa use their rudimentary lungs as hydrostatic organs. Movement of these tadpoles is accomplished with a minimum of effort. The main propulsive structure is the narrow, upturned tip of the tail which vibrates constantly. This causes a slow, steady










52

movement of the larva; occasionally, the tadpole moves more rapidly by using the entire tail to propel itself. Both the tadpole and the adult of this species give the impression of being quite slow. The assumption might be made that this indicates a low rate of metabolism, but this was not measured in frog or larva. As in the adult, no green

pigmentation were found in the tadpole.

A relatively large number of frog species was found in exposed

habitats where there was standing water. Neither tadpoles nor adult frogs were seen often in such habitats during daylight hours. Tadpoles generally avoided bright sunlight by swimming under aquatic vegetation.

Elachistocleis ovale. -- In Surinam, this small microhylid was found calling from shallow roadside ditches containing standing water. The ditches had a moderate growth of grass and were generally exposed to the sun. This species was found in Habitat 25 (Tables 3 and 4). Neither eggs nor tadpoles were found. The three adults examined did not show any indication of green tissues (Table 1).

Pseudis paradoxus. -- This highly aquatic medium-sized frog and its exceptionally large tadpoles were found in roadside ditches and drainage canals in Surinam. Water in these ditches usually exceeded 50 centimeters in depth, was semi-permanent and stagnant. Vegetation in this habitat was varied and served as food for Pseudis tadpoles. While the eggs of this species were not seen in Surinam, Gans (1956) found that this species lays its eggs in a frothy mass which floats on ponds in Trinidad. The black tadpoles grow to more than 15 centimeters in total length, a third of which is body length. Body length of the largest










53

tadpoles approximates that of the adults. Mr. Walter Polder (personal communication) observed that Pseudis tadpoles grow to this large size during the course of a single rainy period of four to six months. Thus, the growth rate of this species is phenomenally high. In comparison

with other species, the adults of which are the size of adult Pseudis, metamorphosis of this tadpole is greatly delayed. The slow, deliberate movements of this tadpole reflect its low metabolic rate (Table 9). Although typical habitats (such as Habitats 23 and 24 of Tables 3 and 4) of this species were visited during the day, neither tadpoles nor frogs were seen. Both forms appeared to be more active at night, when they could be seen floating nearly vertically in the water. Its dark pigmentation, low metabolic rate, and perhaps decreased diurnal activity are

probably important adaptations of this tadpole to its exposed and warm environment. The adult can swim and hop quite rapidly. Thousands of

individuals were seen hopping across the road on a rainy night toward the end of the wet season (late July).

Both the tadpoles and the adults of this species have green pigmentation in the tissues, but it is particularly pronounced in the tadpoles. In the larger tadpoles, green pigment permeates the tissues so that they appear almost as dark internally as externally. Development of green pigmentation in this species does not appear to be related to metamorphosis, since it is well developed before that time. Green pigmentation in the adults appears limited to the well-calcified bones. Their other internal tissues are usually light in color and the blood is bright red. Thus, there is a marked difference between adult and tadpole with regard to green pigmentation.












Smaller hylas. -- jr Surinam, several species of small or

medium-sized hylids were found calling or breeding in open, grassy ponds or flooded fields with a dense growth of weeds. Both of these situations

appeared to be temporary due to increased rainfall at that season (July).
Two very small species, Hyla misera and Hyla minuta, as well as Hyla boesemani, Hyla egleri, Hyla leucophyliata, Hyla rubra, Hyla punctata, and the medium-sized Hvla crepitans, were included in this group. Goin and Layne (1958) recorded Hyla rubra, Hyla misera, Hyla punctata, Hyla leucophyllata, and Hyla lanc-Formis from similar habitats near Leticia in southern Colombia.

While several of these species were found in the same pond in Surinam, they often had different habits. Hyla misera usually called

from a perch on stalks of grass which were surrounded by water; Hyla crepitans was found floating in the water near clumps of grass; and Hyla boesemani called from the grasses at the edge of the same pond. Within this group, Hyla rubra was most likely to be found in the vicinity of Paramaribo, becoming less frequent at higher elevations to the south.

Hyla punctata was found perched on weeds in a flooded field and Hyla leucophyllata called from bushes along a drainage canal.

Of the eight species collected in Surinam in this habitat, Hyla egleri (one individual), Hyla leucophyllata (three), Hyla minuta (two), and Hyla rubra (four) showed no green pigmentation. Hyla boesemani (one), Hyla crepitans (two), and Hyla punctata (two) had green bones and the latter two species had green plasma as well. The single specimen of Hyla misery had light green plasma.











55

Phrynohyas venu1osa. -- This is a medium-slzed species which has a well developed pattern of melanin pigmentation which largely

obscures the extensive internal green pigmentation of bones, soft tissues, and plasma. The poisonous secretion from the skin of this species is well known; breaks in the skin make one particularly susceptible to inflammation caused by contact with this secretion (Goin and Layne, l958). Phrynohyas venulosa was heard or collected from swamps, shallow roadside ditches near forests or thickets, and from residential areas in Surinam. Tadpoles of this species were collected from standing water in roadside ditches near wooded areas.

Zweifel (1964) described the life history of Phrynohyas

venulosa from Panama. The eggs were laid on the surface film of marshy ponds. When brought into the laboratory, they hatched in less than a day and metamorphosis took place after 37 days. The tadpole was of the pond type, with globose body, lateral eyes, and well-developed tail fin. The early larvae have well-developed external gills, which spread at the surface as the tadpole hangs vertically, and conspicuous lungs are present at a later larval stage. These two characters, along with the flotation of eggs, were considered to be adaptations to low oxygen concentrations of the environment. The temperature range in the natural habitat was 25 330 C.

Phyllomedusa hypochondrialis. -- This medium-sized frog is

bright green on its dorsal surface in life. In preservation, this green color fades to slaty-blue. There are no indications of a green pigment in the skin or in other tissues of this species. Phyllomedusa










56

hypochondrialis was found on leaves of bushes and small trees at the edge of wooded areas. The large, yellow eggs of Phyllomedusa hypochondrialis are probably attached to leaves above standing water, into which the larvae drop after hatching. Movements of these pond-type

tadpoles are rather slow.

Sphaenorhynchus aurantiacus. -- This small green hylid was collected from two similar localities in Surinam. In both habitats, the frogs called from cattails in standing water, 50 centimeters or more in depth. In a locality near Domberg, the frogs were found at an intersection of two drainage canals in which there was considerable aquatic vegetation. Along these canals were some shrubs and small trees. The second locality, south of Paramaribo, was a small cattail marsh, perhaps

30 meters in diameter and surrounded by forest except where it emptied into a roadside canal. Goin (1957) noted that all of the known species

of this genus select areas with still waters as their breeding habitats, and that they call from the water, or from floating or emergent vegetation. Lutz (1954) stated that members of this genus lay eggs on leaves, but their development seems to be unknown.

Goin (1957) specifically noted that certain members of this genus have green bones (Table 1). From the description of color in

other structures, it appears that green pigmentation of tissues is a general and striking characteristic of the genus Sphaenorhynchus.

Hyla dominicensis. -- Of the hylid species on Hispaniola, Hyla dominicensis is certainly the most common and has the widest distribution. It is very similar to Hyla brunnea, the most common and widespread of










57

Jamaican tree frogs, and to Hyla septentrionalis, the only tree frog of Cuba and surrounding islands. In the Dominican Republic, Mertens (1939) found Hyla dominicensis in all but the most arid habitats, from near sea level to more than 1000 meters in elevation. Barbour (1914) noted that the Museum of Comparative Zoology has many specimens from all

parts of the island and Cochran (1941) noted that it is common in collections.

During October, 1965, I collected an adult and tadpoles of

this species from Port-au-Prince and heard adults calling from trees in Petionville (400 meters elevation) and Cap Haitien. Adults were heard in May and August of 1966, most often during or after rain. Tadpoles were also collected during these visits. The first large chorus of this species which Mertens (1939) heard was on the evening of February 21 after the first hard rain. He noted breeding activity until September 29 and concluded that there is no definite breeding period. While Lynn (1958) collected adults and tadpoles of Hyla dominicensis in Haiti during midApril, 1953, he did not hear them calling. He also noted that this frog produces a skin secretion which is irritating to cuts and scratches, and that the bones are green.

Mertens (1939) took note of the very rapid development of Hyla dominicensis tadpoles. He collected a number of these from a street puddle filled with rust-colored water on March 3. He stated that the larger tadpoles had a covering of grass-green algae on the back and only small hind limb buds, but developed into frogs in 10 days. Similarly, larvae that he collected from a cistern on February 21 had only rudimentary









58

hind limbs, but were in metamorphosis on March 4. Mertens found Hyla dominicensis in bromeliads in the coniferous forests, but did not state that they bred there. Noble (1923) indicated that he found tadpoles of this species at 2500 meters elevation in wheel ruts. Lynn (1958) found tadpoles of this species in a small pool at the head of a stream.

I found this species only in stagnant water: in a shaded pool in a dry stream bed, in a flooded wiregrass pasture exposed to the sun, in wheel ruts of a muddy road which were also exposed to the sun, and in concrete-lined pits which had been used for tanning hides. In the latter two habitats, the water contained so much solid material and algae that the tadpoles could not be seen except when they came to the surface to take air. The constant agitation of the water by the dozens or hundreds of tadpoles appeared to be largely responsible for the turbidity.

When the flooded pasture was visited in May, a large cluster of tadpoles was noted in one flooded corner. After tadpoles from the cluster came to the surface, they ordinarily returned to the group. This behavior continued until my presence caused the group to disperse. It is of particular interest since the temperature of the water was above 320 C. and the oxygen concentration averaged 0.7 parts per million. Brattstrom (1962) has shown that such aggregations of dark tadpoles absorb more radiant heat than do isolated individuals. He found water temperatures near such groups to be higher than water temperatures at a distance. He reasoned that the resultant higher body temperature caused an increase in metabolic rate, which decreased the time required for development.









59

After the group of tadpoles dispersed, the behavior of individuals was noted. Most appeared to be resting near the bottom of the flooded area while others appeared to be feeding along the bottom. At intervals, they came rapidly to the surface to gulp air and then returned to the bottom to rest or continue feeding. Four individuals which were observed for five or more intervals had average interval lengths of 17, 22, 27, and 31 seconds. It appeared that the smaller and more active ones surfaced more often. Several hours after some of these tadpoles had been collected, their tails appeared quite red. When this habitat was visited again in August, 1966, a flocculent precipitate covered most of the surface and no tadpoles were present.

Young larvae of Hyla domfnicensis are black, but they become

brown or gray and develop green pigmentation in bone and soft tissues during metamorphosis. The black pigmentation of the young may be an adaptation to an environment exposed to the sun, since this character is seen in Hyla vasta and Hyla septentrionalis in similarly exposed environments, but not in the Jamaican hylid tadpoles which develop in shaded habitats. Gills are well developed in this species of tadpole, but they appear to lack lungs until metamorphosis. Therefore, the air which they take at the surface must be kept in the mouth where gaseous diffusion takes place, and then the bubble is released at or near the surface prior to taking in a fresh gulp of air. The complex gill structure probably is important in filtering food, primarily algae, from the water. Like most vegetarian tadpoles, these have a long, coiled intestine. While the tail of this species is not particularly muscular,









60

it is similar to that of stream tadpoles in lacking the fin extension on the back.

Hyla septentrionalis. -- Of the native West Indian hylids,

Hyla septentrionalis is the most widely distributed. This medium-sized tree frog is found throughout Cuba, where it is most abundant in the banana groves of the lowlands (Barbour and Ramsden, 1919). In addition it is common and perhaps has been introduced into the Cayman Islands, the Bahama Islands, the Florida Keys and southern Florida (Barbour, 1937).

Barbour (1931, 1937) believed that this species was introduced into Key West on freight cars from Cuba. There are indications of subsequent introductions and range extensions in south Florida (King and Krakauer, 1966). Grant (1940) summarized references which indicated that this species was accidentally introduced into the Cayman Islands, but did not maintain itself on Cayman Brac. However, William Greenhood (personal communication) noted this species on Cayman Brac during the summer of 1965. Although there has been some question about its distribution on

Grand Cayman, I have found it there at two localities--near North Side and just east of Boddentown--as well as in the Georgetown area. Specimens have been seen from all of the major islands of the Bahamas as far southeast as Acklin Island. A specimen of Hyla septentrionalis in the Institute of Jamaica was collected during the spring of 1965 at Highgate, St. Mary's Parish, Jamaica. This specimen resulted from the introduction of tadpoles from Hialeah, Florida, during the spring of 1961 (Edwin Todd, personal communication); there is no evidence that

this species has bred in Jamaica.











Thus, the Cuban tree frog has become established in a number of localities through the activity of man, and its original range outside of Cuba would be difficult to determine. Part of the adaptability of this species may be related to its accommodation to human habitations; it is also found in mesophytic situations, but not in pineland or prairie habitats (Duellman and Schwartz, 1958; Barbour and Ramsden, 1919; Grant, 1940; Stejneger, 1905; Peterson, Garrell and Lantz, 1952; Neill, 1958).

During the present study, habitats of the Cuban tree frog

were studied on Grand Cayman, on New Providence Island in the Bahamas, in the vicinity of Miami, Florida, and near Highgate, St. Mary's Parish, Jamaica. Cuba was not visited due to political restrictions. The two days' stay on New Providence, during May, 1966, was unproductive due to drought and a widespread lack of surface water.

In Miami, Florida, tadpoles of Hyla septentrionalis were found

at a tropical fish hatchery and at the Serpentarium, a commercial enterprise concerned largely with snakes. Large larvae undergoing metamorphosis were noted in concrete tanks in both places. In addition, thousands of small tadpoles were found in the long, shallow, ornamental pool in front of the Serpentarium (Table 3, Habitat 21). Since this pool had been drained and cleaned just one week before, the tadpoles had hatched and grown to 5 millimeters in body length in less than a week. Although the water temperature of this pool was 33.90 C., the oxygen concentration was 9.2 parts per million. The pool had been drained to remove large amounts of algae, but the water appeared clear at the time of measurement; also,









62

the sky was overcast during much of the day so that it appeared unlikely that the high oxygen content was due to photosynthesis. At the time of measurement, a strong breeze was blowing and may have been responsible for the presence of so much oxygen in the shallow pool. Although the tadpoles were moderately active at the high temperature, they rose to the surface infrequently, and then as often to feed from the surface film as to take air.

The majority of my observations on Hyla septentrionalis were

made during two visits to Grand Cayman--July 5-8, 1965 and October 28-30, 1965. At the time of the July visit, most of the island had been without rainfall for more than a month and was quite dry. Adult Hyla septentrionalis were found only in the vicinity of Georgetown; they became quite active following a shower on the evening of July 7. Tadpoles were not found in any of the cow wells inspected, but a number were collected from a small cistern which was well shaded. The cistern was located behind an abandoned church 6.4 kilometers southeast of Georgetown and was about 50 meters from the shoreline.

During the October visit, tadpoles were still present in this locality. Considerable rain had fallen during the two weeks prior to my visit so that a number of temporary pools of rain water had collected. Tadpoles were collected from several small woodland pools near South Sound, from a small shaded pond and from an open pond near North Side. The woodland pond and pools were less than 20 centimeters deep; the open pond was about 50 centimeters in depth.










63

The tadpoles and adults of Hyla septentrionalis are very

similar to those of Hyla dominicensis; in fact, the two forms have been considered races of the same species (Barbour, 1937). The tadpoles of Hyla septentrionalis are black when young and become brown toward metamorphosis. They are primarily vegetarian, but on one occasion I observed a larger tadpole rapidly ingest a smaller one. The gills are well developed, but lungs are not. Body shape is rounded, and, like that' of other hylid tadpoles in the West Indies, does not bear an extension of the tail fin. The rate of development is very rapid as previously mentioned. Grant

(1940) stated that C. Bernard Lewis found "that eggs laid in a tiny puddle in coral or tree trunk during the night will be active tadpoles by mid-day." English (1912) found that these frogs breed any time of the year after a good rain, with eggs developing to frogs within a few weeks.

He also indicated that the tadpole and frog are enemies of mosquitos which are abundant on Grand Cayman. This is probably true, since the smaller bodies of water in which these tadpoles are found are ordinarily clear and free of other obvious forms of animal life, such as mosquito larvae.

Hyla septentrionalis tadpoles have been found in brackish

pools on the Florida Keys (Neill, 1958). Larvae of this species often have a clear envelope of fluid surrounding the body which is contained within an outer layer of skin. This would seem to be an adaptation which slows diffusion of water or salts between the tadpole body and its environment. It should also reduce heat exchange. It is possible that these tadpoles are capable of slight osmoregulation, since an individual lived nearly four days in sea water diluted by distilled water to 10 parts per thousand salinity. At 15 and 20 parts per thousand salinity, the










64

tadpoles died within three hours, but at 5 parts per thousand salinity, a tadpole lived for ten days before it was returned to fresh water.

Since an isotonic solution for these tadpoles is presumably 7 parts per thousand, the first tadpole was able to maintain itself in a hypertonic solution for near four days.

A number of preserved young Hyla septentrionalis from Grand Cayman showed green bone pigmentation, as did a number of the larger specimens. Of more than 260 live adults collected from Miami, Florida, during June, 1966, none showed green pigmentation in bones; however, Clarence McCoy (personal communication) collected specimens with green bones from Homestead, Florida, during August, 1967. I was unable to determine more about the development of green pigmentation in this species because it did not undergo metamorphosis in the laboratory. I found the skin secretion of this species to be extremely irritating if rubbed into the eyes; it is also difficult to wash from the eyes.

Stream habitats

Flowing streams constitute the primary habitat of several species of tadpoles in the West Indies and South America. Temperature and oxygen concentration within these streams apparently have a narrow range of daily fluctuations. Streams inhabited by tadpoles were found to have temperatures between 200 C. and 250 C. in most cases. Since these streams were in hilly or mountainous areas, it seemed that the higher altitude was related to the relatively low water temperatures. Quite regularly, direct sunlight was blocked by clouds during the afternoon. These factors,










65

along with the shade provided by forest vegetation, probably account for stream temperatures remaining below air temperatures, which are also relatively low. Stream turbulence appears to keep oxygen concentration between 7 and 10 parts per million, but this is decreased where there is little turbulence and the stream moves more slowly (Table 4, Habitat 29). With a pH range of 6.0 8.6, the streams were similar to other surface waters, but were less acidic than water which collected in bromeliads. Dissolved solids demonstrated no definite trends.

Hyla maxima. -- This large, brownish species was found calling at Princess Irene Falls, Brownsberg, Surinam. Individuals called from tree sites 2 meters or more above ground. At least six individuals called from trees around the perimeter of the falls for three successive nights, even though there was no rain; none were heard elsewhere on the small mountain. Tadpoles in the pools at the base of the falls were believed to be of this species, as were eggs and approximately 50 small tadpoles found in a shallow rock basin (3 x 25 x 40 centimeters). The basin was in a rock ledge approximately 2 meters to one side of the main falls and 60 centimeters above the base of the falls. It received a steady supply of water from spray and droplets which fell into it. Since tall trees nearby surrounded the base of te falls, the basin remained in the shade except for about an hour at mid-day. Shortly after sunrise (7:35 A.M.), temperature of the basin water was 23.80 C. Near the end of the period of insolation (12:30 P.M.), the water temperature in the basin reached 29.90 C., while the air temperature was 28.10 C. Two hours later, the water temperature in the basin had dropped to 25.60 C.









66

Thus, water temperatures in the basin ranged from 23.80 C. to 29.90 C., while the temperature ofthe water in a pool at the base of the falls ranged from 23.40 C. to 25.50 C. Iron concentration in the basin was about three times as great as in the pool (Table 5). This may be important to young tadpoles, since it was noted that young Hyla brunnea tadpoles are more sensitive to iron deficiency than are the older ones. Concentrations of phosphate and pH were simi lar for both the basin and the stream pool. Oxygen concentration in the pool ranged from 8.3 to 8.9

parts per million; during the same period, the basin contained 2.0 to

3.1 parts per million of oxygen. Concentration of carbon dioxide in the basin shortly after it had been in the sun was 24.0 parts per million; at the same time, the concentration of carbon dioxide in the pool was 10.0 parts per million.

The eggs in the basin were of medium size--2 to 3 millimeters in diameter, not including the jelly--and rested on the detritus at the bottom of the basin. It was not known whether the eggs originally floated on the water surface. The yolk of the eggs was white rather than yellow and the animal hemisphere was darkly pigmented. Tadpoles in the basin were observed feeding on the eggs, which filled their alimentary tracts. They also came to the surface to take air, but do not appear to have lungs. When the water temperature was highest, these tadpoles were observed coming to the surface at intervals of 10 to 25 seconds.

None of the tadpoles observed showed green pigmentation. However, none had well-developed hind legs. The length of time required










67

for developnmnt was not determined. Three adult males were collected from this locality and all had green bones, soft tissues and plasma (Table 1).

Dowling (1960) found the tadpoles of this species in the Arima River of Trinidad. She found the jet black tadpoles--about 2 to 5 centimeters in length--swimming in schools. These dense aggregations of tadpoles, as many as 177 in a group, did not keep to sun or shade. Neither did the individuals in masses appear to feed; tadpoles observed feeding were scattered in shallow water.

Hyla vasta. -- Noble (1923) found this giant tree frog in the

northern coastal mountains of Hispaniola and also at 2500 meters elevation in the central range which runs east to west. Mertens (1939) also found it in forested areas in the central range, but did not consider it exclusively montane, since he collected the species at 220 meters in one locality. It has been found by the present writer and others near Furcy in southeastern Haiti at 1350 meters elevation. In this unforested

and agricultural area the adults take cover in crevices and under roots and sod which overhang the stream bank. I also heard a small chorus in a ravine near the Riviere Froid at an elevation of 200 meters.

The life history of the giant Hispaniolan tree frog was first recorded by Noble (1923). In the rain-soaked northern mountains of the Dominican Republic, he found Hyla vasta along the streams which flow through the rain forest. He stated: ''After the sun has set, the giant tree frog, Hyla vasta, leaves his hiding place among the tree tops and descends to some rocky ravine. There flattened out on a mossy boulder









68

in midstream, he rests for hours, seemingly enjoying the cool mists which arise from the torrents. So closely does the frog resemble the

moss and lichen of his surroundings that he would rarely be observed were it not for his big shiny eyes, which are conspicuous even when closed, the lower eyelid being translucent"(p. 56).

While Noble (1923) found that the skin secretion of this species caused inflammation and irritation to his hands, I did not find this to be true of the frogs collected from Furcy. Neither the boys who caught them, nor 1, after handling them, experienced any discomfort.

Noble (1927) described the early development and habits of

the young as follows: "Hyla vasta laid its eggs in little basins in the gravel and stones on the edge of the pools in the mountain torrent

(one observation). Six days after hatching, the larvae made their way out of one of the basins over wet stones into the torrent pool. As they grew older they developed better stream lines than the tadpoles of Hyla dominicensis. They were equipped with more rows of teeth. The mouth was larger and better adapted to holding on to rocks in the stream. Its tail was thicker and more muscular than that of the stagnant-pool

tadpole"( p. 108).

He also reported that the eggs of Hyla vasta are pigmented and are stuck to rocks at the bottom of the basin, and that they give rise to dark tadpoles with small external gills, which do not rise to the surface. The temperature of the stream he recorded as 21.7 25.60 c., averaging 23.70 C.









69

My observations near Furcy were similar, except that I was

unable to find a basin in which the young were developing. Half-grown to metamorphosing larvae were found at the bottoms of pools. The mottled

gray pattern of the larger tadpoles makes them rather obvious against the silt on the pool bottoms. Occasionally these larvae will move under stones in the stream bed or attach themselves to rocks by means of their large sucking mouths. The sucking mouth is important for stabilizing the tadpole in the stream current and also for scraping plant food from rock surfaces. When disturbed, they may allow themselves to be swept downstream by the current.

Like Hyla dominicensis tadpoles, the larvae of Hyla vasta have well-developed gills. In the stream habitat, these gills seem to serve them well, since they were never observed taking air from the surface. In the laboratory, six members of this species were kept in a graduated cylinder which contained water with 0.5 parts per million of oxygen. At first, the tadpoles rested at the bottom when not surfacing for air, but after several hours they remained near the surface and were inactive except for taking air.

Green pigmentation appears in the large tadpoles at about the beginning of metamorphosis. The bones of metamorphosed young are green and this color is present in some of the large adult individuals. Slightly older frogs found among rocks along the stream also had green bones.

Hyla heilprini. -- This medium-sized species of Hispaniola

was first seen by Noble (1923) as a brightly colored male which gave its shriek while sitting on a rock in the middle of a torrent in the northern








70
mountains. Although he did not capture that individual, he obtained tadpoles nearby. In cold streams above 2500 meters in the central

regions he heard again the loud call of Hyla heilprini above the roar of the cascades. While the average temperature in the first stream was 24.70 C., it was 12.60 C. in the second locality at the higher elevation. Noble found the tadpole of Hyla heilprini to be more streamlined than the previous two species and considered this an adaptation to living in torrents, rather than in slower moving streams or puddles. Since these larvae were obviously adapted to an environment with a high oxygen concentration, Noble had doubts about being able to maintain them in captivity, but he was successful in raising them through metamorphosis.

From Noble's (1923) observations in the Dominican Republic,

specimens from the Museum of Comparative Zoology from several localities along the southern peninsula of Haiti, and personal observations, it appears that Hyla heilprini is widespread on Hispaniola. I believe that I heard the species just west of Port-au-Prince in a ravine near the Riviere Froid, less than 200 meters above sea level. It would seem that wooded ravines are more important to this species than are high elevations. Cascading streams probably are necessary for its reproduction, as both Noble (1923) and I found its tadpoles only near waterfalls.

Green pigmentation was present in all of the dozen or more tadpoles which I collected. These were well-developed tadpoles, with either two or four legs, which were olive green with black spots dorsally, blue-green and white ventrally, and the distal portions of the limbs were orange in color. Since only the distal portions of the limbs were orange,










71

one wonders if this was the result of extreme effects of temperature or oxygen upon a pigment which was more generally distributed through the body. An adult of this species collected from a banana stalk was bright grass-green on the dorsal surface with white and blue ventrally and blue and yellow on the flanks. Later the back became quite dark. This specimen had green bones and green plasma. Noble (1923) described the dorsal color of a living specimen as "golden green."

Hyla pulchrilineata. -- Except for its light-colored longitudinal stripes, lack of an interocular bar, and larger size, this small greEntree frog of Hispaniola is quite like Hyla wilderi of Jamaica. It has been collected from widely scattered points on the island and Cochran (1941) noted that it seems to be common in certain localities. From the records of Noble (1923), Cochran (1941), and Lynn (1958), it seems likely that this species has a distribution similar to those of Hyla vasta and Hyla heilprini: in forests at higher altitudes and in wooded ravines at lower altitudes. This species appears to be the least common of the four hylids of Hispaniola; I did not encounter adults or tadpoles during my visits to Haiti. Noble (1923) and Lynn (1958) found adults calling from leaves in the vicinity of streams. Lynn (1958) described its call as a faint clicking like a small telegraph instrument, similar to that of Hyla wilder.

Noble (1927) stated that Hyla pulchrilineata young develop in

water as do the other hylas on the island, but apparently he has given no further details. There are no known specimens of Hyla pulchrilineata tadpoles in the major museums of eastern North America (including the









72

American Museum of Natural History where Noble deposited most of his specimens). Dr. Lynn and Dr. Albert Schwartz (personal communications), both of whom have collected in this area, are unaware of any tadpoles of this species which have been collected. It seems best to consider Noble's statement as a supposition on his part.

The color of the bones of this species has not been recorded.

Bufo marinus. -- This very large toad was found breeding in

several habitats including temporary pools in Surinam and near drainage canals and tropical fish hatcheries in Miami, Florida. It has been introduced into Jamaica, where it appears to be the only anuran species which breeds in ponds or slow-moving portions of streams. Neither young, which are vegetarian, nor adults are known to have green tissues under normal circumstances.

Leptodactylus albilabrus. -- Tadpoles of this Puerto Rican

species were collected from a roadside stream on El Yunke, the highest mountain on the island. Water flowed rapidly in the stream, thus accounting for its high oxygen concentration (Table 4). Both large and small tadpoles of this species came to the surface, evidently to take air, but their lungs do not appear to be developed at these stages. This species feeds on vegetation and does not develop green pigmentation as it undergoes metamorphosis.


Resume of Ecology and Breeding Habits

Considering the great variety of aquatic habitats available in the Neotropical Region, it seems that anurans have taken advantage of most of them. The species with green pigmentation are well










73

distributed throughout the Neotropical Region, but appear to be restricted

to certain habitats or families of frogs. As Barrio (1965a) noted, green pigmentation is found in only three families of South American frogs. Hylids and centrolenids are primarily arboreal following

metamorphosis, the period of life when they appear most likely to have green pigmentation. Pseudids are generally considered to be highly aquatic as adults. Although I found relatively little green pigment in some adult Pseudis paradoxus paradoxus in Surinam, Barrio (1965a) found high concentrations of biliverdin in Pseudis paradoxus platensis, Lysapsus limellus and Lysapsus mantidactylus adults. While green pigmentation appears in tree frog tadpoles with metamorphosis, it preceded metamorphosis by an extensive period in Pseudis paradoxus paradoxus. Thus, the pseudids differ from hylids and centrolenids in both the type of habitat and in the time of onset of pigment development. It seems logical to assume that the hyperbiliverdinemia of pseudids has a cause which is different from that of the two tree frog families. In addition, the point should be made that terrestrial and fossorial anuranstoads largely--have not been reported to have biliverdin in their tissues.

Since distributional and ecological factors have not yielded a satisfactory explanation for the accumulation of green pigment, it appears logical to turn to physiological studies.















Ill. PHYSIOLOGICAL FACTORS RELATED TO CHLOROSIS


Previous studies have shown that the immediate cause of

chlorosis is the production of biliverdin in quantities greater than those which are excreted. It has not yet been determined if chlorbsis in frogs is due to a high rate of biliverdin formation, impairment of bile pigment excretion, or both.

Since hemoglobin of the blood is ordinarily the main source

of biliverdin, it is logical that one should study this protein and the red blood cells in which it is found. If chlorosis is due primarily to a high rate of biliverdin formation, one might expect to find higher concentrations of hemoglobin or higher rates of hemolysis in chlorotic frogs than in others. In addition to chlorosis itself, high rates of hemolysis might be indicated by low red cell counts, or by blood smears which show immature erythrocytes or cell fragments. Should hemolysis or other process be shown to be the cause of increased biliverdin, then the cause of such a process should be determined. In the case of rates of hemolysis, it is known that they may be correlated with temperature, concentrations of chemical agents, or differences in red cell structure.

Impairment of bile pigment excretion would involve the liver and its associated ducts. Blockage of extra-hepatic ducts is readily determined by the absence of biliverdin from feces. Blockage or restriction of bile flow within the liver may be difficult to detect,

74










75

but histological study ordinarily shows the degeneration of liver which is associated with prolonged cholestasis.

Several factors indicate that the study of tadpole respiration may shed light upon chlorosis. First of all, the present study has

shown that green pigmentation first appears in the tadpole stage of several species (Table 2). Secondly, the low oxygen tensions of some

tadpole environments suggest that the larval stages have special respiratory adaptations or that tissue damage may result from oxygen deficient cy. The degradation of respiratory pigments--apparently the immediate cause of chlorosis--suggests a disruption or decrease of oxygen transport within the organism. Since the liver is quite sensitive to oxygen deficiency, one wonders if the preceding factors are concerned with possible liver malfunctions. Finally one should consider the high environmental temperatures to which these organisms are exposed, and the greater stress which this factor places upon respiratory and hepatic systems.

In order to determine the importance of biliverdin formation to chlorosis, several attempts were made to induce green pigmentation in vivo, using non-chlorotic species. This was followed by a broader

physiological study of Neotropical frog species--both chlorotic and nonchlorotic--to clarify the importance of biliverdin formation, liver

function, and other factors which might pertain to the green pigmentation.


Attempts to Reproduce Green Pigmentation In Vivo

During the early part of the study, attempts were made to duplicate the high concentration of tissue bi liverdin, particularly as it occurred in bone. These experiments were carried out upon species which









76

were readily available in Florida, none of which had green bones. Four methods were utilized in attempts to increase the concentrations of biliverdin in the experimental animals. Phenylhydrazine hydrochloride was used to destroy the red blood cells in the frog, thus causing its hemoglobin to be degraded into biliverdin more rapidly. The second method was the injection of additional hemoglobin, presumably of mammalian origin, again to increase the rate of biliverdin formation. Third was the direct injection of biliverdin into the animal. The fourth method was to induce an iron-deficiency anemia, such as that known to cause chlorosis in man.

Phenylhydrazine hydrochloride administration

Since biliverdin formation was induced in Bufo arenarum by

means of phenylhydrazine injection (Cabello, 1943), this method was used

in an attempt to develop green bones in species which were known not to have green bones.

Two adult specimens of Bufo terrestris, along with ten Hyla

cinerea, were given intraperitoneal injections of phenylhydrazine hydrochloride in water in October 28, 1964. The dose was at the rate of

1 microgram of phenylhydrazine hydrochloride per gram of body weight of the frog. On November 19, 1964, each of these frogs was given a second dose of 2 micrograms of phenyihydrazine hydrochloride per gram of body weight. During the period of the experiment these frogs were fed twice a week. The animals were observed daily until December 9.

Injection of phenylhydrazine hydrochloride caused destruction of red blood cells and formation of biliverdin, but like the other










77

methods.used, it did not induce detectable hyperbiliverdinemia. Severe anemia developed in specimens of Hyla cinerea as a result of injection of this drug. Two smaller individuals died several days after the second injection. Post-mortem examination showed that very few red cells were present in the blood vessels or elsewhere. There was slight indication of hemopoietic activity in the long bones; the spleen was pale and showed no increase in size. The kidneys were slight yellow in color, rather than the usual red. The frogs appeared emaciated. The liver was normal in the larger frog examined; the smaller frog had an enlarged gall bladder filled with green fluid and its liver was smaller than expected. None of the frogs and toads used in this experiment showed any evidence of green pigmentation in bone, blood or other tissues outside of the digestive tract.

Administration of hemoglobin solution

A second attempt to increase the amount of biliverdin in

living anurans involved the intraperitoneal injection of a 1 per cent hemoglobin solution. Specimens of Hyla septentrionalis (eleven individuals), Rana pipiens (four individuals), Rana catesbeiana (two individuals), Hyla gratiosa (one individual), Bufo terrestris (two individuals), and Hyla cinerea (seven individual) were injected with either 1 per cent

hemoglobin in phosphate buffer, or with phosphate buffer (pH 7.4) alone, or were given no treatment. Fluids voided by these animals were taken from the jars in which they were kept and analyzed with the Beckman DU spectrophotometer.










78

Intraperitoneal injection of hemoglobin solution resulted in no clear indication of green tissues in any of the species. Several individuals voided substances which had absorption peaks that corresponded to biliverdin or hemoglobin. There appeared to be no addition of green pigment to the tissues. Judging from the color of the fluids voided by these animals, bile pigments were formedand eliminated within twenty-four hours after injection of hemoglobin. However, it should be pointed out that green pigment was not excreted by all of the frogs which were injected with hemoglobin. In addition, one untreated specimen and one injected with only phosphate buffer demonstrated green pigment in voided fluids. This was also true of other untreated frogs in the laboratory.

Infection of biliverdin

Still a third method was used--the direct injection of

biliverdin in sesame oil or phosphate buffer (pH 7.4). On one occasion, an attempt was made to mobilize bone calcium prior to injection of biliverdin. Several specimens of Hyla cinerea were injected with 0.5 (.005 cc) units of Lilly's Parathyroid Extract per gram of body weight. The same dosage was repeated twelve hours later. After another six hours, an intraperitoneal injection of biliverdin in sesame oil was administered. Finally, they were given two drops of cod liver oil after an hour. The latter step was intended to stabilize calcium in bone because

of its vitamin D content (Cantarow and Schepartz, 1957).

Biliverdin solutions injected into the peritoneal cavity

caused no apparent change in the color or other visual characteristics










79

of the tissues. This treatment was discontinued after a relatively short period of time because it did not appear practical to continue. Attempts to decalcify bone or otherwise change its pigmentation in vivo were unsuccessful.

Induction of iron deficiency anemia in tadpoles

Iron deficiency is the most common cause of anemia in man and may become pronounced during pregnancy or during periods of growth. In one of its more serious forms (chlorosis) iron deficiency anemia causes the skin to become green. During the nineteenth century, this was found

regularly among girls and young women, particularly those wi th menstrual disorders.

Little is known about iron metabolism of frogs. Iron is

absorbed in the duodenal region of the intestine and incorporated into eggs and hemoglobin (Brown, 1964). However, it seemed possible that a lack of iron could be a cause of green pigmentation in frogs, as it is in man. The main difficulty in running an experiment to test this idea

in frogs or most tadpoles is that their intake of iron cannot be regulated easily. Live insects constitute the normal diet of frogs, and most tadpoles feed upon plant life and other organisms. Tadpoles of Hyla brunnea are different in this respect, since they feed upon frog eggs, usually those of their own species. These tadpoles develop normally when they are fed Gerber's strained egg yolk, and kept in spring water. The strained egg yolks were known to contain 3.07 milligrams of iron per hundred grams (Gerber Products Company, 1965). It was thought that

the high concentration of phosphates in egg yolk (290 milligrams of










80

phosphorus per hundred grams) would decrease iron absorption, due to the poor absorption of iron complexed with phosphates (Cantarow and Schepartz, 1957). The animals were kept in glass-distilled water,

rather than spring water, in order to reduce the iron concentration further.

Two matched groups of twelve tadpoles were placed in small

glass bowls which had been acid-cleaned, rinsed and fi lied with glassdistilled water. During the course of the experiment, the water was changed twice a week and fresh egg yolk was added immediately thereafter. An excess of food was always available to both groups. The only difference in treatment between the two groups was that ferrous sulfate (reagent grade) solution was added to the water of the first group. Individuals of the first group appeared to develop normally, as did the four larger individuals of the second group. Of the eight smaller tadpoles of the second group, all became pale yellow in color and five were dead after two weeks. At this time, ferrous sulfate was added to the water of the second group. After three days, the second group of tadpoles became the same light brown color as those of the first group. The darker pigmentation appeared to be the result of an increase in hemoglobin. These observations were taken to indicate that limitation of iron intake of rapidly growing tadpoles resulted in an anemia, which was relieved by addition of more iron. It is also possible that trace elements such as copper or cobalt may have been responsible for hemoglobin formation in both groups. In either case, the anemia did not cause development of green pigment in the skin of any of the tadpoles.










81

Physiological Characteristics of Chlorotic and Non-Chlorotic Frogs

This portion of the study was directed toward finding physiological differences between chlorotic and non-chlorotic frogs. Frog blood, tadpole respiratory rates, and histological sections of liver

were the main subjects of study.

Methods of study

Hemoglobin determinations. -- The amounts of hemoglobin present in anuran blood was determined by the acid-hematin test (Cohen and Smith, 1919). A Heilige hemoglobinometer was used for this purpose, since much of the work was done under field conditions.

Blood was collected from a large blood vessel or the heart

into heparinized capillary tubes. From these, 20 microliter samples of blood were transferred to a graduated test tube which contained 2 milliliter of 1 per cent hydrochloric acid solution. After mixing the contents of the tube, they were allowed to stand for ten minutes in order to insure-complete hemolysis of the red blood cells. Then distilled

water was added until the color of the solution matched the color standard of the hemoglobinometer. The calibration mark at the meniscus of the fluid indicated the hemoglobin concentration of the blood in grams

per hundred milliliters.

Red blood cell counts. -- Blood samples obtained as for hemoglobin determinations were di luted and counted by standard techniques using a Spencer Bright Line hemocytometer. Four counts were made for each individual and the average was used as the red blood cell count.









82

Red blood cell measurements. -- An optical micrometer was

used to measure length and width of twenty-five red blood cells from each individual. Measurements were made from smears prepared and stained in the field. Average cell length and width were determined and converted to microns. These blood smears were used to study abnormal cell types in the blood.

Effects of temperature. -- In order to determine the effect of temperature upon frog erythrocytes, blood from specimens of Hyla squirrella, Bufo terrestris and Scaphiopus holbrooki was collected in heparinized capillary tubes. The tubes were plugged with clay at both ends and kept in a water bath at 450 C. for varying periods of time. After this treatment, smears of the blood were made, stained and studied. Measurement of tadpole respiratory rates

Measurement of respiratory rates of tadpoles was carried out by two different methods. During the earlier part of the work, oxygen uptake was determined by use of Warburg constant volume respirometers as described by Umbreit, Burriss and Stauffer (1964).

The second method was improvised in order to determine the rate of oxygen utilization by tadpoles shortly after capture, under either controlled or field conditions. Necessary apparatus for field measurements were Erlenmeyer flasks with rubber stoppers, a Precision Scientific portable oxygen analyzer, and a wristwatch. The simple technique consisted of the following steps:

1. Clean 125 ml Erlenmeyer flasks were calibrated to determine
their actual volume when stoppered.

2. The appropriate number of clean Erlenmeyer flasks were filled
with clean tap water of the desired temperature.









83


3. The probe of the oxygen analyzer was calibrated in air.
After dipping the tip of the probe into clean water,
it was waved in the air while two ammeter readings were
taken. The ammeter of the oxygen analyzer measures electrical current between silver and lead electrodes located in
the tip of probe; this current is directly proportional to
the oxygen which diffuses through the covering membrane and into the space between the electrodes. Air temperature was determined at the same time by use of a thermistor attached
to the probe. These data were used to calculate the sensitivity of the probe under prevailing conditions.

4. The probe of the oxygen analyzer (which had been taped to
fit snugly into the neck of the Erlenmeyer flask) was
lowered carefully into the neck of the flask until the probe
formed a watertight seal with the neck of the flask.
After this was accomplished, the tip of the probe projected
slightly below the neck of the flask. When done properly,
no bubbles were trapped in the flask.

5. While holding the flask and probe together, they were
rotated vertically through a 90 degree arc, so that their
axis was in a horizontal plane. From this position, the
flask was shaken from side to side in order to cause water
to flow past the tip of the probe. This procedure was
necessary to improve the accuracy of the method. In the laboratory, a magnetic stirrer was used to cause a flow
across the tip of the probe.

6. Once the indicator of the ammeter became steady, a reading
was made. The ammeter was then turned off for a short time
and then turned on again. While the shaking continued a
second reading was made and the average of the two readings was recorded. Probe sensitivity, ammeter reading and water
temperature was used to calculate the initial oxygen concentration.

7. Once the data necessary to determine the initial oxygen
concentration were obtained, a single tadpole was introduced into the flask, and the flask was stoppered. If an air bubble was trapped within the flask, sufficient water
from the original source was added so that the flask was
completely filled with water when stoppered. This was
necessary to prevent intake of air by the tadpoles and also
to prevent immeasurable exchange of gases between air and
water. Note was made of the time that the flask was
stoppered.










84

8. After an appropriate length of time, the flask was opened
and the final oxygen concentration determined in the same
manner as the initial oxygen concentration (steps 3 through
6).

9. The tadpoles were fixed in formalin and each individual was
labeled so that their weights could be measured upon return to the laboratory. Since formalin preserved tadpoles were found to lose nearly 10 per cent of their original weight, the weight of each preserved tadpole was multipled by the
conversion factor 1.094 to obtain a more accurate estimate
of the original wet weight.

10. Since the difference between initial and final oxygen concentrations was given in mg/l, it was necessary to make
additional calculations to obtain the respiratory rates in
terms of p1 02/g/hr. This was achieved by use of the
following equation:

Respiratory rate (p1 02/g/hr)
Change in 02 concentration (mg/I). Water volume (jul)
1.429 (conversion to ml of 02). Tadpole weight (g). Time (hrs)

Although the error incurred in this method approached 3 per cent or even 5 per cent under some circumstances, the necessary equipment was not nearly so cumbersome to transport as manometers and pipettes; cleaning was not a problem, nor was the equipment as likely to be damaged in transit. In addition, if one desired to maintain the animals at a specific temperature during an experiment, a wash basin or large pan could hold a dozen or more flasks at one time and the temperature of the water in the bath could be maintained within 10 C. of the desired temperature by

addition of ice or warmer water, as required; hence, no special bath, refrigerator or heater was required. In addition to technical advantages of the method, it allowed measurement of oxygen uptake under actual field conditions, although this opportunity was not realized. Since it was possible to make measurements shortly after the tadpoles were collected,










85

there was little time for them to become acclimated to different environmental conditions. With modifications in the size of the flask and diameter of probe, larger organisms could be studied.

Aside from the lower accuracy of the measurement of oxygen change, disadvantages in relation to manometric techniques included: slightly more variation in temperature due to lack of automatic controls

and constant stirring; the practical impossibility of beginning and ending an entire series simultaneously, or of taking readings at regular intervals during the experiment.

This technique was developed in the laboratory at Gainesville, where it was used to measure oxygen uptake of Hyla dominicensis, Hyla brunnea, Hyla vasta, and Leptodactylus albilabrus maintained at 250 C. in a water bath. Respiratory rates of Hyla maxima and Pseudis paradoxus were studied at 250 C. in Surinam; similarly oxygen uptake of Hyla vasta, Hyla dominicensis and Hyla heilprini were determined at 250 C. in Haiti.

Liver tissue sections. -- Blocks of frog liver tissue were

fixed in Bouin's fluid and then changed to 50 per cent ethanol for storage and transport back to the laboratory. These tissues were embedded, sectioned and stained with hematoxylin and eosin by standard techniques. Results

Hemoglobin concentrations and red cell counts. -- The small

green-boned species, Hyla punctata and Sphaenorhynchus aurantiacus, had the lowest concentrations of hemoglobin of the species studied. Another small form with green tissues, Hyla wilderi, along with Sphaenorhynchus aurantiacus, had the lowest red blood cell counts. Hyla punctata had the









86

lowest concentration of hemoglobin per cell, along with an individual of Phrynohyas venulosa. Yet species with maximal values in these categories also have green tissues. These include Hyla maxima and Phrynohyas venulosa. However, the species which has the largest amount of hemoglobin per cell, Phyllomedusa hypochondriasis, does not have green tissues. This is evidently a reflection of the large size of the red blood cells in this species (Table 7). There appears to be no correlation between the blood characteristics studied and the presence of green pigmentation.

It should be noted that heparinized blood of Pseudis paradoxus tadpoles and Sphaenorhynchus aurantiacus males completely hemolyzed after being kept in their own blood plasma at room temperature overnight. This was not noted in other species with green tissues or in those without green tissues.

Red cell size. -- From Table 7, it can be seen that cell lengths range from 12.8 to 21.6 microns. Cell widths vary from 9.9 to 14.5 microns. To obtain an approximation of erythrocyte volume for each species, the average cell length was multiplied by the average width. The largest cells are those of Phyllomedusa hypochondriasis (20.6 x 14.5 microns) and Sphaenorhynchus aurantiacus (21.6 x 13.9 microns). Smallest cells were those of Pseudis paradoxus (13.2 x 12.0 microns), Phrynohyas venulosa (12.8 x 10.8 microns) and Osteocephalus taurinus (13.1 x 10.8 microns). The most variable cel ls were those of Hyla punctata. All of these species except Phyllomedusa hypochondrialis are chlorotic.

Examination of blood smears. -- Blood smears stained with

Wright's stain revealed certain characteristics which may or may not be










37

related to green pigmentation (Figures 12-16). A list of these and some others, according to the organisms in which they were observed, follows:

Pseudidae
Pseudis paradoxus -- adult male Cells of this individual tended to form cytoplasmic blebs; tadpole blood hemolyzed
upon standing overnight.

Bufonidae
Bufo typhonius -- adult an erythroplastid was observed
in the blood of this individual.

Atelopodidae
Atelopus sp. -- male erythrocytes showed coagulated
cytoplasm.

Hylidae
Hyla boesemani -- adult cells have coagulated cytoplasm,
as occurs after slight heating.

Hyla crepitans -- adult most cells hemolyzed.

Hyla egleri -- male cells not unusual.

Hyla geographica -- adult male nothing unusual.

Hyla lanciformis -- adult most of the cells were hemolyzed.

Hyla leucophyllata -- about half of erythrocytes have
coagulated cytoplasm.

Hyla minuta -- adult most of the cells have coagulated
cytoplasm as occurs on heating.

Hyla misera -- some cells have odd shapes.

Hyla maxima -- adult many lymphocytes present.

Hyla punctata -- adult cell membranes are irregular; green
cytoplasmic and extracellular granules present on stained
slides. Erythrocytes appear immature.

Hyla rubra -- adult nothing unusual.

Osteocephalus taurinus -- adult many erythrocytes with
vacuolar nuclei.










88

Phrvnohvas venulosa -- red cell cytoplasm coagulated.

Phy llomedusa bicolor -- appears to have many leukocytes.

Phyllomedusa hypochondrialis -- erythrocytes are basophilic.

Sphaenorhynchus aurantiacus -- cytoplasm of erythrocytes appears vacuolated, with greenish refractile material in
vacuoles (stained slide). Blood of three males hemolyzed
overnight.

Leptodactylidae

Leptodactylus pentadactylus -- some red cells show coagulated
cytoplasm.

Effect of temperature upon frog red cells. -- Slight changes in a few cells were noted after ten minutes at 450 C. After four hours at 450 C., cytoplasm had coagulated in nearly all of the red cells of Hyla squirrella and Bufo terrestris (Figures 9-11); the red cells of Scaphiopus holbrooki were abnormally shaped after this time, but the cytoplasm had not coagulated.

Respiratory rates of tadpoles. -- The respiratory rates of tadpoles and froglets of green boned species (Tables 8-10) are similar to those previously recorded for amphibian larvae (Hopkins and Handford, 1943). The basis of recording respiratory rate in the study was the uptake of microliters of oxygen per gram of wet weight of tadpole per hour. In order to convert this to an approximate dry weight basis, the rate of

oxygen uptake should be multiplied by eight. Size range of the tadpoles in the present study is much greater than in previous ones, but it can be seen that respiratory rate is a function of tadpole size and that the species is not very important in this respect (Figure 8). As is to be expected, respiratory rates increase with increasing temperature (Figure 7).








00


It should be noted that the respiratory rate of Hyla brunnea tends to drop sharply between 35 and 400 C. (Table 8, Group A), whereas the respiratory rate of Hyla septentrionalis does not (Table 8, Group C). When the Hyla septentrionalis tadpoles were removed from the flasks, they appeared to be in good condition (heart rate was about three beats per second). On the other hand, these Hyla brunnea tadpoles were limp after two hours at 400 C., and one did not recover. This is an indication that

the critical thermal maximum of Hyla brunnea larvae is lower than that of the Hyla septentrionalis tadpoles. This might be expected when one considers the higher temperatures to which Hyla septentrionalis tadpoles are often exposed (Table 4) These higher temperatures are largely due to the lower elevation and greater insolation of Hyla septentrionalis habitats. Hyla brunnea is more often found in shaded areas and at greater elevations than are available to Hyla septentrionalis over most of its range.

Along these lines, it may. be noted that the froglet of Hyla

lichenata appeared in poor condition after being kept at 350 C. for less than two hours (Table 10). It would appear that Hyla lichenata is more sensitive to this temperature than is Hyla brunnea at the same stage. This may account for the fact that while Hyla brunnea is found throughout the range of Hyla lichenata, the former species extends its range to lower elevations and more exposed habitats than the latter.

Tadpoles of Hyla brunnea kept in the laboratory were exposed

to low water temperatures during the early morning of November 30, 1965. This caused some mortality and the surviving tadpoles appeared pale,










90

similar to those which had iron deficiency. Very little movement of the tadpoles was evident when the water temperature in their containers was 16.80 C. Lynn (1940) suggested that the lack of large bromeliads above 1600 meters in the Blue Mountains may be the reason for Hyla brunnea not being found at such altitudes. Low temperatures at these altitudes may also be a factor which prevents this species from moving to higher elevations.

Liver histology. -- Study of liver sections of ten species of

frogs from Surinam is of interest (Figures 17-20). No dead or completely degenerated tissue was seen in these slides but there were several indications of conditions or activities which may cause changes in bile excretion. There did not appear to be any obstructions to the bile duct or gall bladder in the majority of individuals examined, but those with green tissues and plasma usually had very concentrated bile in the gall bladder, judging from its dark blue color. A resume of liver histological characteristics follows: Pseudidae
Pseudis paradoxus -- adult male liver apparently normal
except for very heavy pigmentation (the liver appeared black
before dissection);
-- adult female as in male, with possible cell fragments in blood of liver;
-- tadpole liver structure apparently normal; much less pigmentation than in adults; cell fragments
in blood of the liver; many immature erythrocytes present in
circulating blood.

Atelopodidae
Atelopus sp. -- male some liver cells appear enlarged and
have rarified cytoplasm; some pyknotic nuclei present; otherwise apparently normal.

Hylidae
Hyla geographica -- adult male veins of liver partly occluded; considerable pigment present in liver cells.









91

Hyla leucophyllata -- adult female liver apparently
normal; contains many erythrocytes.

Hyla maxima -- adult male fat deposition taking place in
one area; erythrocytes appear to be undergoing disintegration
in liver; relatively little pigment present.

Hyla minuta -- adult male some irregularly shaped nuclei;
considerable pigment present.

Hyla punctata -- adult ma le many nuclei present in liver tissue; these are quite variable in size, shape and staining characteristics; some may be pyknotic; very little
pigment is present in the liver; many unusual cells are
present in blood, perhaps immature erythrocytes.

Osteocephalus taurinus -- adult female liver pigmentation
less than normal; odd-shaped nuclei present in liver.


Disease as a Cause of Chlorosis in Frogs

Early in this study, it was assumed that disease was not principally responsible for the green pigmentation of frogs. Since several writers had accepted green bones or other green tissues as traits which were characteristic of certain species, it was not logical to believe that a bacterial or viral disease would affect all the members of a species or population in a nearly uniform manner. Neither was it logical that a genetic disease should be passed from generation to generation with such regularity. Thus, it was tentatively assumed that the normal metabolism of chlorotic frogs was responsible for their green pigmentation.

On the other hand, disease cannot be ruled out as a possible cause of biliverdin accumulation. Individual adult specimens of Hyla septentrionalis and Hyla dominicensis may have dark green, light green or white bones. Here is an example of what one might expect if some of the frogs were diseased, but such individual variation may also have other causes.









92

A better example of the effect of disease might be demonstrated by a specimen of Osteocephalus taurinus which arrived in the laboratory shortly after its death. The individual had been kept in another laboratory for several months, but did not eat during the latter part of its stay there. Dissection showed the internal organs to be darker than in most frogs. Blood vessels were a bluish-purple rather than red in color. An olive-green fluid was found in the coelomic cavity, while the dorsal lymph sac was nearly filled with a clear, viscous, blue-green fluid. The gall bladder was filled and appeared dark blue externally, but contained a clear fluid and a solid brown substance, perhaps bile salts. The dark color of the internal organs was due primarily to the presence of biliverdin, which appeared to be in most of the tissues, including bone. A cross-section of femur showed that the green pigment was found throughout the bone, but that alternating concentric lamellae of light and dark green were present. Microscopic examination showed that the liver cells had completely degenerated. A histochemical test for liver glycogen was made by means of the Periodic Acid-Schiff reaction. Aggregations of dark pigment were the only structures which were PASpositive. These were believed to be a lipofuscin pigment from the degeneration of red blood cells. Unlike melanin, lipofuscin is PAS-positive (Dubin, 1958). These liver sections appeared simi lar to photographs of liver in acute viral hepatitis published by Popper and Schaffner (1957, p. 437). While the actual cause* cannot be pinpointed, an obvious pathological condition of the liver prevented the normal excretion of bile pigment and evidently caused it to pass into the lymph.




Full Text
2
blue. Thus, the green color of frog skin of this type results from
the physical arrangement of three pigments, none of which is green
(Schmidt, 1919, 1920, 1921; Kawaguti, Kamishima and Sato, 1965; Elias,
1943) .
Though most frogs with green skin appear to fit the pattern
described, some tree frogs do not. Notable are some of the small
hylids such as Hyla wilderi. This species has few melanophores and
appears to be a diffuse green color throughout the body. When preserved
in alcohol, a green pigment is extracted and the frog becomes whitish
in color. In this species, as well as in others, there is no blue color
which supplants the green upon preservation. Indications are that
structural effects are not involved and that this type of green color
in frogs is due solely to a green pigment or pigments. Reports of
internal green pigment substantiate this idea.
This paper is concerned with three facets of the biology of
green neotropical frogs: (l) identification of the green pigment or
pigments; (2) ecology; and (3) physiology of green frogs, with emphasis
upon possible factors influencing pigment accumulation. Since these
three areas require different methods of study and yield various types
of information, they will be presented in separate sections.


Elias, H. and H. Bengelsdorf. 1952. The structure of the liver of
vertebrates. Acta Anatmica J_4: 297337-
Emmel, V. E. 1924. Studies on the non-nucleated elements of the blood:
II. The occurrence and genesis of non-nucleated erythrocytes
or erythroplastids in vertebrates other than mammals. American
Journal of Anatomy 33* 347-406.
English, T. M. 1912. Some notes on the natural history of Grand
Cayman. Handbook of Jamaica for 1912: 598-600. (From Grant,
1940)
Etkin, W. 1934. The phenomena of anuran metamorphosis II. Physiological
Zoology J\ 129-148.
Ewer, D. W. 1959- A toad (Xenopus laevis) without haemoglobin. Nature
183: 271.
Fernandez, K. and M. Fernandez. 1921. Sobre la biologa y reproduccin
de Batracios argentinos. Anales Sociedad Cientfica Argentina
91 (From Barrio, 1965a)
Fontaine, M. 1941a. A propos du pigment vert de l'orphie (Belone belone
L.). Bulletin Institu Oceanographi que Monaco 8: 793*
. 1941^.. Recherche sur quelques pigments serique et dermique de
poissons marins (Labrdes et Cyclopterides). Bulletin Institu
Oceanographi que Monaco 38: 792.
Foulkes, E. C., R. Lemberg and P. Purdom. 1951. Verdohaem and Vverdoglobi
Proceedings of Royal Society of London. Series B. 138: 386-402.
Fox, D. L. 1953. Animal Biochromes and Structural Colors. Cambridge:
University Press.
Foxon, G. E. H. 1964. Blood and Respiration. In: J. A. Moore, editor,
Physiology of the Amphibia. New York: Academic Press, pp. 151~
209.
Freeman, J. R. 1963. Studies on respiratory mechanisms of the salamander
Pseudobranchus striatus. Dissertation, Gainesville, Florida:
University of Florida.
Gallardo, J. M. 1961. On the species of Pseudidae (Amphibia, Anura).
Bulletin of Museum of Comparative Zoology 125: 111-134.
Gans, C. 1956. Frogs and paradoxes. In Animaland 23.: 2-4.
Gerber Products Company. 1965. Nutritive values of Gerber baby foods.
Fremont, Michigan: Gerber Products Company.


99
hypothesis for lack of green pigment in this species is that the
liver of this relatively large tadpole is better developed at metamor
phosis, and therefore can cope with the heavy load thrust upon it at
metamorphosis, including an increase of biliverdin.
Experiments reported on earlier in this paper indicate that
hemolysis alone is not likely to result in high concentrations of bili
verdin in the tissues. One cubic centimeter of a 1 per cent hemoglobin
solution injected into a 1 gram specimen of Hyla cinerea produced no
noticeable effect. This amount of hemoglobin (0.01 gram) would result
in the production of 0.46 milligrams of biliverdin in a rather short
period (hours or a few days). Apparently, it is excreted without diffi
culty and without green pigmentation developing in the tissues. If we
estimate that 5 per cent of such a frog consists of blood, that one-fifth
of that consists of red blood cells, and that hemoglobin constitutes
one-half of the red cell weight, we would calculate that it contains
0.005 grams of hemoglobin. Comparison of these two sets of calculations
would show that the frog can degrade an amount of hemoglobin equivalent
to twice that contained within its body, in a few days, without accumulat
ing biliverdin in the tissues.
Bile pigment formation in frogs has not been studied in regard
to the source of biliverdin. it is presumed that hemoglobin is the pri
mary source of bile pigment, and that other hemoproteins yield a minor
fraction.
Defective Transport of Bile Pigment into the Liver Cell
The importance of this factor in the accumulation of biliverdin
in frogs is not known.


95
several hours. In addition, Bokermann's (1965) observation that Osteo-
cephalus taurinus was subjected to air temperature of 46 C. shows that
the effect of heat is a likely cause of protein coagulation and hemo
lysis in Neotropical anurans. Heating of human blood to 40 C. for
extended periods decreases solubility of hemoglobin (Goldberg, 1958);
heating to 50 C. for ten minutes or more causes hemolysis (Kimber and
Lander, 1964). Cloudsley-Thompson (1965) found that heat death of
tropical lizards was due to physiological oxygen deficiency and another
factor which operated simultaneously, perhaps protein coagulation. Thus,
it appears that the oxygen transport system reaches the limit of its
capacity at about the same temperature which causes the degeneration of
the system because of coagulation of hemoglobin and hemolysis.
The blood cells of Hyla punctata are quite variable in size
and shape, and have basophilic cytoplasm. In fact, these cells appear
to be immature and closely resemble those of erythropoietic tissue. It
seems that these immature cells have been released into the circulating
blood to compensate for cell loss, probably through hemolysis. Erythro
cytes of Phyllomedusa hypochondria 1 is also appear basophilic, but are
relatively uniform in size and shape. It is possible that these cells
are immature, but it seems more likely that the basophilia is tied in
with the presumably low metabolic rate of these slow moving creatures.
Additional evidence of hemolysis is suggested by the presence
of cell fragments in the liver circulation of Pseud is paradoxus.
Since most chlorotic species do not have coagulated cytoplasm
in the erythrocytes present, and since most of the species with coagulated


22
Relation of Respiration to External Environment
Although a number of studies have dealt with respiration in
regard to external factors such as temperature and partial pressure of
oxygen, little has been done to relate respiratory form and function of
a tadpole to its specific environment. The most thorough study along
these lines was done by Savage (1961), who compared several species of
tadpoles in relation to feeding and respiration in different environ
ments. His conclusions were quoted previously.
Laessle (1961) noted that the tadpoles of the common Jamaican
tree frog, Hyla brunnea, lived in stagnant water of very low oxygen
content, which collects in the central reservoirs and leaf bases of
bromeliads. He suggested that the agitation of the long tail increased
oxygen diffusion into the water and that the large surface area of the
tail provided additional respiratory surface. The tadpoles of the other
three species of Jamaican tree frogs also develop in the oxygen-poor
water which accumulates in bromeliads (Dunn, 1926). Tree frogs, closely
related to the Jamaican species, on the island of Hispaniola, are known
to breed in streams and torrents (Noble, 1927); the Cuban tree frog,
Hyla septentriona1 i s, is known to develop in brackish water, as well
as in cisterns and pools (Grant, 1940).
It may be noted here that Powers et aj_. (1932), working upon
fish respiration in relation to environment, found that "The number of
red blood corpuscles is increased with a decrease in the oxygen and by
an increase in the carbon dioxide tension of the water and vice versa."


19
to this are the Antarctic ice fishes, Chaenichthyidae (Ruud, 1959).
Blood of the several species of this family is nearly colorless and
contains only white cells (1 per cent of blood volume). The cold water
in which these fish live has a high oxygen content and the low tempera
ture tends to lower the respiratory rates, so that oxygen diffusion
through the naked skin sufficiently supplements that dissolved in the
blood to maintain their low rate of metabolism. Small eel larvae
also lack hemoglobin until they reach the elver stage (Andrew, 1965)
and occasional specimens of the frog, Xenopus laevis, have been found
without hemoglobin (de Graaf, 1957; Ewer, 1959)-
Among the indicators of oxygen-carrying capacity of blood
are measurements of red cell size, erythrocyte counts and determinations
of hemoglobin content. Amphibian erythrocyte sizes and counts have been
listed by Prosser et a]_. (1950), Vernberg (1955) > Freeman (1963), and
Hartman and Lessler (1964). Of all vertebrates, salamanders have the
largest red cells, ranging from 30-62 microns in longest dimension.
According to these sources, erythrocytes of frogs, which are also oval
in shape, range from 18-23 microns in length. The number of red blood cells
usually varies inversely to cell size. Counts of salamander cells
vary from 28,000 to 197,000 per microliter; those of anurans, from
380,000 to 670,000 per microliter. While one would expect larger cells
and higher cell counts to represent greater oxygen-carrying capacity, a
trurer measure lies in the direct determination of hemoglobin. Kirkberger
(1953), Stuart (1951), and Goin and Jackson (1965) have published values


28
study and by others are summarized in Tables 1 and 2. Points to be
noted from Table 1 include the fact that the color of the bile was
usually green. in many individuals, the bile was so concentrated as
to appear blue when seen through the gall bladder wall. When soft
tissues were green, it was usually a rather general phenomenon and often
was accompanied by a relatively high concentration of plasma pigment,
a condition which might be termed chlorosis.
Sex and its relationship to chlorosis in frogs
Due to the relatively small number of individuals collected
from each species, it is difficult to make a generalization concerning
the relationship between chlorosis and sex. Chlorotic and non-chlorotic
individuals of both sexes were among specimens of Hyla septentriona 1 is
from Grand Cayman. Table 2 shows several species in which all tadpoles
observed had green pigment at the time of metamorphosis or before. The
general occurrence of biliverdin in these species indicatesthat the
sex of the individual is not important to the development of chlorosis
at this stage. The presence of green pigmentation in calling males of
several species and in eggs of certain Hypero1ius. Aaalychnis,
Centro lene 11a, Pseud? s and Hyla indicates that sex is not a very import
ant factor in the development of chlorosis.
Age distribution of chlorotic anurans
Among tadpoles, no green pigmentation was noted in young
tadpoles of Hy 1 a brunnea h'y la boeseman i Hy la domin ?cens? s Hy la
leucop'ny 1 lata Hyla maxi ma Hyla rubra, Hyla septentriona 1 i s Hyla vasta,


55
Phrvnohyas venulosa. --This is a medium-sized species which
has a well developed pattern of melanin pigmentation which largely
obscures the extensive internal green pigmentation of bones, soft tissues
and plasma. The poisonous secretion from the skin of this species is
well known; breaks in the skin make one particularly susceptible to
inflammation caused by contact with this secretion (Goin and Layne, 1958)
Phrvnohyas venulosa was heard or collected from swamps, shallow roadside
ditches near forests or thickets, and from residential areas in Surinam.
Tadpoles of this species were collected from standing water in roadside
ditches near wooded areas.
Zweifel (1964) described the life history of Phrynohyas
venulosa from Panama. The eggs were laid on the surface film of marshy
ponds. When brought into the laboratory, they hatched in less than a
day and metamorphosis took place after 37 days. The tadpole was of the
pond type, with globose body, lateral eyes, and well-developed tail fin.
The early larvae have we 11-developed external gills, which spread at the
surface as the tadpole hangs vertically, and conspicuous lungs are
present at a later larval stage. These two characters, along with the
flotation of eggs, were considered to be adaptations to low oxygen con
centrations of the environment. The temperature range in the natural
habitat was 25 33 C.
Phy1lomedusa hypochondri a iis. -- This medium-sized frog is
bright green on its dorsal surface in life. In preservation, this green
color fades to slaty-blue. There are no indications of a green pigment
in the skin or in other tissues of this species. Phy1lomedusa


33
these families is the result of parallel evolution. All things con
sidered, there appears to be no clear relationship between phylogeny
and chlorosis among frogs.
Geographic distribution
Frogs with green tissues or eggs appear to be tropical or
sub-tropical in their distribution. Below is a list of green pigmented
crogs according to the countries or regions in which they are found.
Sources of information are given in Table 1.
Central Africa: Hypero 1ius sp.
Jama ica: Hyla brunnea, Hyla lichenata, Hyla wi Ideri.
Hispaniola: Hyla dom?nicensis, Hyla heiIprini, Hyla vasta.
Cayman Islands (presumably also Cuba and Bahama Islands):
Hyla septentr?ona1is.
Mexico end Central America: Agalychnis dacnicolor, Centro le
ne 1 la a 1bomaculata, Centro lene 1 la granulosa, Centrolenella ?lex,
Centrolene1 la spinosa, Centrolene1 la prosoblepon, Centro lene 1 la
pul vera turn.
Northern South America, including Brazil: Lvsapsus 1 ime 1 1 urn
laevis, Pseudis paradoxus, Anotheca coronata, Hyla albofrenata, Hyla
a 1bomargina ta, Hyla boesemani Hyla calca rata, Hy la crepitans, Hyla
cuspidata. Hyla geographica, Hyla langsdorffi, Hyla maxima, Hyla mi sera,
Hyla punctata, Osteocepha1 us taurinus, Phrynohyas venulosa, Sphaenor-
hynchus Trachycepha1 us nigromaculata, Centro lene 1 la vanzolinli.
Argentina: Lysapsus 1 i me 11 urn limellum, Lysapsus mantidacty1 us,
Pseudis paradoxus platensis, Hyla berthae, Hyla nasica, Hyla puIche1 la,


49
septentrona1is are quite opportunistic in regard to selection of breed
ing areas. However, in Surinam, several species appeared to be associated
almost exclusively with forest habitats. Included in this group are
Hyla qeograohica, Hyla calcarata, Hyla lanciformis, and Osteoceoha1 us
taunjTus, all of which are medium-sized and brown dorsal ly, and the
large species, Phyllomedusa bicolor, which is green above.
Hyla qeoqraphica. -- A si lent male of Hyla aeocraphica was
collected from vegetation on a rocky ledge of Princess Irene Falls,
Brownsberg, Surinam. Wooded slopes were present on both sides of the
falls and the stream which flowed over it. Tadpoles were present at the
base of the falls, but these were believed to be larvae of Hyla maxima,
which was calling at the locality. This specimen of Hyla qeoqraphica
had light green bones but no other green organs.
Hyla calca rata. -- The single female was collected from a bridge
over a small stream which flowed into a swamp. This individual had green
bones, but no other green pigmentation.
Hyla lanciformis. -- Most of the specimens of Hyla lanciformis
were collected in wooded areas. One was collected while calling near
the edge of the forest in the vicinity of temporary pools. No eggs or
tadpoles were found. Goin and Layne (1958) stated that its note was
regularly heard in open lands such as wet meadows near Leticia, Colombia.
They noted that the specimens collected were perched on bushes or other
plants one or two feet above ground. Four specimens from Surinam lacked
green pigmentation in bone or other tissues.


92
A better example of the effect of disease might be demonstrated
by a specimen of Osteocepha1 us taurinus which arrived in the laboratory
shortly after its death. The individual had been kept in another
laboratory for several months, but did not eat during the latter part
of its stay there. Dissection showed the internal organs to be darker
than in most frogs. Blood vessels were a bluish-purple rather than red
in color. An olive-green fluid was found in the coelomic cavity, while
the dorsal lymph sac was nearly filled with a clear, viscous, blue-green
fluid. The gall bladder was filled and appeared dark blue externally,
but contained a clear fluid and a solid brown substance, perhaps bile
salts. The dark color of the internal organs was due primarily to the
presence of biliverdin, which appeared to be in most of the tissues, in
cluding bone. A cross-section of femur showed that the green pigment was
found throughout the bone, but that alternating concentric lamellae of
light and dark green were present. Microscopic examination showed that
the liver cells had completely degenerated. A histochemica1 test for
liver glycogen was made by means of the Periodic Acid-Schiff reaction.
Aggregations of dark pigment were the only structures which were PAS-
positive. These were believed to be a lipofuscin pigment from the degen
eration of red blood cells. Unlike melanin, lipofuscin is PAS-positive
(Dubin, 1958). These liver sections appeared similar to photographs of
liver in acute viral hepatitis published by Popper and Schaffner (1957
p. 437)- While the actual cause cannot be pinpointed, an obvious patho
logical condition of the liver prevented the normal excretion of bile
pigment and evidently caused it to pass into the lymph.


65
along with the shade provided by forest vegetation, probably account
for stream temperatures remaining below air temperatures, which are
also relatively low. Stream turbulence appears to keep oxygen concen
tration between 7 and 10 parts per million, but this is decreased where
there is little turbulence and the stream moves more slowly (Table 4,
Habitat 29). With a pH range of 6.0 8.6, the streams were similar to
other surface waters, but were less acidic than water which collected
in bromeliads. Dissolved solids demonstrated no definite trends.
Hyla maxima. -- This large, brownish species was found calling
at Princess Irene Falls, Brownsberg, Surinam. Individuals called from
tree sites 2 meters or more above ground. At least six individuals
called from trees around the perimeter of the falls for three successive
nights, even though there was no rain; none were heard elsewhere on the
small mountain. Tadpoles in the pools at the base of the falls were
believed to be of this species, as were eggs and approximately 50 small
tadpoles found in a shallow rock basin (3 x 25 x 40 centimeters). The
basin was in a rock ledge approximately 2 meters to one side of the main
falls and 60 centimeters above the base of the falls. It received a
steady supply of water from spray and droplets which fell into it.
Since tall trees nearby surrounded the base of the falls, the basin
remained in the shade except for about an hour at mid-day. Shortly
after sunrise (7:35 A.M.), temperature of the basin water was 23.8 C.
Near the end of the period of insolation (12:30 P.M.), the water temper
ature in the basin reached 29-9 C., while the air temperature was 28.1 C
Two hours later, the water temperature in the basin had dropped to 25.6 C


I
Figure Page
14 Stained Smear of Untreated Blood of
Phrynohyas venulosa 139
15 A Stained Smear of Blood from Hv la
punctata, a Species With Green Tissues 140
16 Stained Smear of Untreated Phy1lomedusa
bicolor Blood 140
17 Stained Section of Apparently Normal
Liver of Hyla rubra 141
18 Stained Section of Hyla punctata Liver 141
19 Stained Section of Liver From Adult Male
Pseud i s paradoxus 142
20 Stained Section of Liver From Adult Male
Hyla maxima 142
vi i i


96
cytoplasm are not chlorotic, these two characters appear to be mutually
exclusive to a large extent. Thus the possibility arises that the
erythrocytes of the chlorotic species are more susceptible to lysis (and
perhaps coagulation as well) than are the red cells of non-chlorotic
frogs. One may hypothesize that high temperatures cause cytoplasmic coagu
lation in both types of frogs, but that hemolysis, leading to biliverdin
formulation, occurs only in the chlorotic species.
Among frogs, two other types of increased hemolysis are likely.
Varela and Sellares (1938) found that Bufo arenarum in Brazil has a
peak number of red blood cells (one million per cubic millimeter) in
July and August with a rapid decrease in blood count in October, at the end
of the breeding season. During the present work, frogs studied in
Surinam in July (at the end of the breeding season) sometimes demonstrated
green plasma but did not have green bones. The assumption was made that
the high concentration of biliverdin was of a transitory nature and did
not last long enough to cause staining of relatively permanent tissues
such as bone. Similarly, the presence of green bones in animals without
green plasma gives additional evidence of seasonal peaks of biliverdin
concentration. Bone cross-sections showing alternate regions of dark and
light green color in Lysapsus mantidacty1 us (Barrio, 1965a) and in
Osteocephalus taurinus also indicate periodic hemolysis. One may suggest
that hemolysis at the end of the breeding season is due to hormonal changes
or to effects of higher temperatures encountered during the breeding
season, or both.


43
frogs were found in bromeliads, ant or termite nests, or elsewhere,
so that the post-metamorphic development and ecology of the young
are unknown.
Hyla lichenata. This giant tree frog, which attains a
length of 117 millimeters, is restricted to the central and western hills
of Jamaica, generally above 300 meters in elevation. Since this frog is
rarely seen and less often collected, knowledge of its habits and distri
bution is based largely upon hearing its distinctive snoring call (see
Lynn, 1940).
In regard to the habits of this species, Panton (1952) first
found one in a bromeliad on a dead candlewood tree in woodland, but all
subsequent specimens were taken from hollow trees. Dunn (1926) collected
three of these frogs from small hollow trees with openings 4 to 12 feet
above the ground, but noted that they ordinarily call from greater heights.
He stated further that the "bony head is obviously of use in plugging the
hole after the frog is inside." Lynn and Dent (1943) traced the unmis
takable call of this species to a clump of bamboos near Chapel ton. A
female specimen of Hyla 1 ichenata was collected by Dr. Thomas Farr of the
Institute of Jamaica, near the entrace to St. Clair Cave, St. Catherine
Parish on June 19, 1965- During several days in captivity at the Institute
of Jamaica, this individual deposited a number of eggs. It showed no
interest in a cockroach which was offered as food. This individual pro
duced a skin secretion which became gum-like on the hands of those who
held it. Gosse (1851) recorded an instance where the skin secretion of this
species caused severe irritation to the human eye.


39
may be found in bromeliads which grow at ground level on shaded, lime
stone hillsides or in the epiphytes high above ground in well developed
forests, they are frequently found in towns or among the trees which
line the roads in agricultural areas. Like the other tree frogs of
Jamaica, Hyla brunnea appears to be largely dependent upon the bromeliads
for the water which is retained at the leaf bases, but several Jamaicans
have told me that the brown tree frog is found among banana leaves and
sta1ks.
Breeding habits of Hyla brunnea are unusual in some respects,
particularly as they relate to adaptations to life in the bromeliads.
The breeding season begins during May, as indicated by a record cited
by Dunn (1926). Panton (1952) has stated that the strongest choruses
of the tree-toad are heard at the end of May or beginning of June, but
that the time fluctuates because of weather conditions. During the
latter part of May, 1966, and early June, 1965, I noted very few tadpoles,
but found eggs more common than later in the year. Tadpoles were not
uncommon during October, 1965- The eggs of Hyla brunnea are deposited
in the central reservoir of bromeliads in most cases, but eggs or
tadpoles may be found at the bases of outer leaves; on one occasion,
eggs were found in water that had collected on a hollow branch.
Development of the tadpole stages has been described by
Schreckenberg (1956), who studied the embryonic development of the
thyroid gland in Hyla brunnea. The mode of development differs little
from that seen in Rana and hylids of the United States. The main differ
ences appear to be the result of adaptations to the bromeliad habitat.


1000
700
500
400
OXYGEN 300
UPTAKE
(MICRO- 200
LITERS
PER GRAM
PER HOUR))00
70
50
40
30
20
p Pseudis paradoxus
10 I i L_l U I 1 I I l I I lI I I 1 > iI
0.01 0.02 0.03 0.05 0.07 0.1 0.2 0.3 --5 0.7 1.0 2.0 3-0 5-07-0 10.0
WEIGHT (GRAMS)
Figure 8. Respiratory rates of tadpoles at 25 C. (From Table 9)
O'


141
Figure 17. Stained section of apparently normal liver of Hyla rubra.
(400X)
Figure 18. Stained section of Hyla punctata liver. Note the numerous
nuclei, some of which appear pycnotic. (400X)


69
My observations near Furcy were similar, except that I was
unable to find a basin in which the young were developing. Half-grown
to metamorphosing larvae were found at the bottoms of pools. The mottled
gray pattern of the larger tadpoles makes them rather obvious against
the silt on the pool bottoms. Occasionally these larvae will move under
stones in the stream bed or attach themselves to rocks by means of their
large sucking mouths. The sucking mouth is important for stabilizing
the tadpole in the stream current and also for scraping plant food from
rock surfaces. When disturbed, they may allow themselves to be swept
downstream by the current.
Like Hyla dominicensis tadpoles, the larvae of Hyla vasta have
we 11-developed gills. In the stream habitat, these gills seem to serve
them well, since they were never observed taking air from the surface.
In the laboratory, six members of this species were kept in a graduated
cylinder which contained water with 0.5 parts per million of oxygen. At
first, the tadpoles rested at the bottom when not surfacing for air, but
after several hours they remained near the surface and were inactive
except for taking air.
Green pigmentation appears in the large tadpoles at about the
beginning of metamorphosis. The bones of metamorphosed young are green
and this color is present in some of the large adult individuals. Slightly
older frogs found among rocks along the stream also had green bones.
Hy la he?1prin ?. -- This medium-sized species of Hispaniola
was first seen by Noble (1923) as a brightly colored male which gave its
shriek while sitting on a rock in the middle of a torrent in the northern


86
lowest concentration of hemoglobin per cell, along with an individual
of Phrynohyas venulosa. Yet species with maximal values in these cate
gories also have green tissues. These include Hyla maxima and Phrynohyas
venulosa. However, the species which has the largest amount of hemoglobin
per cell, Phyllomedusa hypochondria 1is, does not have green tissues. This
is evidently a reflection of the large size of the red blood cells in
this species (Table 7)- There appears to be no correlation between the
blood characteristics studied and the presence of green pigmentation.
it should be noted that heparinized blood of Pseudis paradoxus
tadpoles and Sphaenorhynchus aurantiacus males completely hemolyzed
after being kept in their own blood plasma at room temperature overnight.
This was not noted in other species with green tissues or in those without
green tissues.
Red cell size. -- From Table 7 it can be seen that cell lengths
range from 12.8 to 21.6 microns. Cell widths vary from 9-9 to 14.5
microns. To obtain an approximation of erythrocyte volume for each species,
the average cell length was multiplied by the average width. The largest
cells are those of Phyllomedusa hypochondria 1 is (20.6 x 14.5 microns)
and Sphaenorhynchus aurantiacus (21.6 x 13-9 microns). Smallest cells
were those of Pseudis paradoxus (13-2 x 12.0 microns), Phrynohyas venulosa
(12.8 x 10.8 microns) and Osteocepha1 us taurinus (13-1 x 10.8 microns).
The most variable cells were those of Hyla punctata. All of these species
except Phyllomedusa hypochondria 1is are chlorotic.
Examination of blood smears. -- Blood smears stained with
Wright's stain revealed certain characteristics which may or may not be


85
there was little time for them to become acclimated to different environ
mental conditions. With modifications in the size of the flask and
diameter of probe, larger organisms could be studied.
Aside from the lower accuracy of the measurement of oxygen
change, disadvantages in relation to manometric techniques included:
slightly more variation in temperature due to lack of automatic controls
and constant stirring; the practical impossibility of beginning and end
ing an entire series simultaneously, or of taking readings at regular
intervals during the experiment.
This technique was developed in the laboratory at Gainesville,
where it was used to measure oxygen uptake of Hy1 a domin ? censis, Hyla
brunnea, Hyla vasta, and Leptodacty 1 us albilabrus maintained at 25 C.
in a water bath. Respiratory rates of Hyla maxima and Pseud is paradoxus
were studied at 25 C. in Surinam; similarly oxygen uptake of Hyla vasta,
Hyla dominicensis and Hyla hei1prin? were determined at 25 C. in Haiti.
Liver tissue sections. -- Blocks of frog liver tissue were
fixed in Bouin's fluid and then changed to 50 per cent ethanol for stor
age and transport back to the laboratory. These tissues were embedded,
sectioned and stained with hematoxylin and eosin by standard techniques.
Resu1ts
Hemoglobin concentrations and red cell counts. -- The sma11
green-boned species, Hyla punctata and Sphaenorhynchus aurantiacus, had
the lowest concentrations of hemoglobin of the species studied. Another
small form with green tissues, Hy la wi lderi, along with Sphaenorhynchus
aurantiacus, had the lowest red blood cell counts. Hyla punctata had the


17
Studies of Babak (1907a, 1907b) upon Rana temporaria and
Rana esculenta, of Drastich (1925) upon Salamandra maculosa, of Bond
(I960) on Salamandra maculosa, Ambystoma opacum and Ambystoma jefferson-
ianum, of Conant (1958) concerning adult Necturus maculosus and Freeman
(1963) on adult Pseudobranchus striatus, all showed an inverse relation
ship between external gill size and oxygen concentration. Most indicated
that external gills increased in size when animals were put into water
of lower oxygen concentration. In addition, Conant, Bond, and Freeman
all noted that gill movements and expansion, as well as blood supply
were correlated with oxygen tension.
Pulmonary respiration is important to the majority of adult
amphibians and is found in a number of larvae as well. The lungs of
stream dwelling species are reduced in size (Noble, 1931). Savage
(1961) has described the well developed lungs of Rana temporaria larvae
and still larger ones from a Papuan tadpole; both of these species
were taken from standing pools or puddles of water. In the Papuan
species, it appeared that tail muscles pressed the lung against the
notochord to bring about ventilation.
Savage stated further that:
The respiratory systems in tadpoles are connected with the
ecology in the following way. If a tadpole lives in an environ
ment rich in food, as many temporary or polluted ponds are,
it does not need to pump water to get its good, and so does
not need large gill filters. To use the oxygen under these
conditions, however, it must have large gills, and needs
lungs to tide it over emergencies. If, however, it lives
in the oligotrophic type of pond, with plentiful and almost
constant supplies of oxygen but with a low concentration of
food, it needs to pump much water, and so must have large
gill filters. With these, gills might not be necessary,
because of the large surface of the filters (or rather, in


34
Hyla phrynoderma, Hyla punctata rubro!?neata, Hyla raniceps, Hyla
slemersi, Hyla squa1 irostris, Hyla trachytorax, Phrynohyas venuiosa.
Judging from this list, it appears that nearly all green
pigmented frogs are to be found in the Neotropical Zone. It should be
noted that, except for the works of Barrio (1965a, 1965^) and the present
writer, the references to green tissues or eggs of frogs are few, and
these are often obscure. Although a number of individuals have volun
teered personal observations concerning coloration, there appears to
be a general reluctance to publish such observations. It seems quite
possible, if not likely, that additional records will be forthcoming
from Africa or other parts of the Old World tropics.


This dissertation was prepared under the direction of the
chairman of the candidate's supervisory committee and has been
approved by a 11 members of that committee. It was submitted to the
Dean of the College of Arts and Sciences and to the Graduate Council,
and was approved as partial fulfillment of the requirements for the
degree of Doctor of Philosophy.
December, 1967
Dean, Graduate School


137
Figure 9. Normal red blood cells of Bufo terrestris four hours
after collection. (400X)
Figure 10. Red blood cells of the same individual as above, after
four hours at 45 C. Note coagulated cytoplasm. (400X)


New York: McGraw-
Noble, G. K. 1931- The Biology of the Amphibia.
Hill Book Company, Inc.
Noir, B. A., E. Rodriguez Garay, and M. Royer. 1965. Separation and
properties of conjugated biliverdin. Biochimica et Biophysica
Acta J_00: 403-410.
Panton, E. S. 1952. Our ground and tree frogs--g1impses into their
life and habits. Natural History Notes of the Natural History
Society of Jamaica (53): 87-92. Reprinted from the Daily
Gleaner, Kingston, Jamaica, December, 1922.
Parker, H. W. 1935. The frogs, lizards and snakes of British Guiana.
Proceedings of the Zoological Society of London 1935: 505-530.
Pamas, J. K., and S. Krasinska. 1921. Uber den Stoffwechsel der
Amphibien-larven. Biochemische Zeitschrift 116: 108-137-
(From Atlas, 1938)
Perkins, L. 1948. Life in a wild pine. Natural History Notes of the
Natural History Society of Jamaica (31): 86-90.
Peters, W. 1873. Ueber die von Dr. J. J. v. Tschudi Beschriebenen
Batrachier aus Peru. Monatsberichte Akademie Berlin, pp. 622-699-
(From Zoological Record, 1873, p. 95)
Peterson, H. W., R. Garrell and J. P. Lantz. 1952. The mating period
of the giant tree frog Hy la domi n i cens i s. Herpe to logi ca 8_: 63.
Podiapolsky, P. P. 1910. (UeberChlorophy11 bei Froschen). Biologi-
cheski i Zhurnal, Moskva 5-9-
Popper, H. and F. Schaffner. 1957- Liver: Structure and Function. New
York; Blakiston.
Powers, E. B. et a_L 1932. The relation of respiration of fishes to
environment. Ecological Monographs 2_: 385-473
Prosser, C. L. e_t aj_. 1950. Comparative Physiology. Philadelphia:
W. B. Saunders Company.
von Recklinghausen, F. 1883 Handbuch der allgemeine Pathologie der
Kreislaufs und die Ernahrunq. Stuttgart: enke. (From Lemberg
and Legge,1949)
Reeder, W. G. 1964. The digestive system. in: J. A. Moore, editor,
Physiology of the Amphibia. New York: Academic Press.
Rich, A. R. 1925- The formation of bile pigment. Physiological Reviews
182-224.


APPENDICES
Tables 1 10
Figures 1 20


148
Hach Chemical Company. No date. Methods Manual for Mach Direct Read
ing Colorimeter, fifth edition. Ames, Iowa: Hach Chemical
Company.
Hartman, F. A. and M. A. Lessler. 1964. Erythrocytic measurements in
fishes, amphibia and reptiles. Biological Bulletin 126:
83-88.
Helff, 0. M. 1926. Studies on amphibian metamorphosis II. The oxygen
consumption of tadpoles undergoing precocious metamorphosis
following treatment with thyroid and di-iodotyrosine. Journal
of Experimental Zoology 45.: 69_93 -
Herner, A. E. and E. Frieden. 1961. Biochemical changes during anuran
metamorphosis. VIII. Changes in the nature of the red cell
proteins. Archives of Biochemistry and Biophysics 5.: 2535-
Holden, H. F. and R. Lemberg. 1939- The UV absorption spectra of bile
pigment iron compounds and of some bile pigments. Australian
Journal of Experimental Biology and Medical Science 17: 133-143-
Hopkins, H. S. and S. W. Handford. 1943. Respiratory metabolism during
development in two species of Amb1ystoma. Journal of Experi
mental Zoology 93.: 403-414.
Israels, L. G., M. Levitt, W. Novak, and A. Zipursky. 1966. The early
bilirubin. Medicine 4: 517-521.
Kawaguti, S., Y. Kamishima, and K. Sato. 1965. Electron microscopic
study on the green skin of the tree frog. Biological Journal,
Okayama University JJ_: 97-109-
Kimber, R. J. and H. Lander. 1964. The effect of heat on human red
cell morphology, fragility and subsequent survival in vivo.
Journal of Laboratory and Clinical Medicine 64: 922-933-
King, W. and T. Krakauer. 1966. The exotic herpetofauna of southeast
Florida. Quarterly Journal of the Florida Academy of Sciences
2£: 144-154.
KIrkberger, C. 1953- Temperaturadaptation der Sauerstoffbindung des
Blutes von Rana esculenta L. Zeitschrift fur Vergleichende
Physiologie 35: 153-158.
Krogh, A. 1941. The Comparative Physiology of Respiratory Mechanisms.
Philadelphia: University of Pennsylvania Press.
1850. De hepatis ranarum extirpatione. Dissertation, Berlin.
(From Lemberg and Legge, 1949)
Kunde.


84
8. After an appropriate length of time, the flask was opened
and the final oxygen concentration determined in the same
manner as the initial oxygen concentration (steps 3 through
6)
9- The tadpoles were fixed in formalin and each individual was
labeled so that their weights could be measured upon return
to the laboratory. Since formalin preserved tadpoles were
found to lose nearly 10 per cent of their original weight,
the weight of each preserved tadpole was multipled by the
conversion factor 1.094 to obtain a more accurate estimate
of the original wet weight.
10. Since the difference between initial and final oxygen con
centrations was given in mg/1, it was necessary to make
additional calculations to obtain the respiratory rates in
terms of pi 02/g/hr. This was achieved by use of the
following equation:
Respiratory rate (pi 02/g/hr)
_ Change in O7 concentration (mq/l). Water volume (pi)
1.429 (conversion to ml of O2). Tadpole weight(g). Time (hrs)
Although the error incurred in this method approached 3 per cent
or even 5 per cent under some circumstances, the necessary equipment was
not nearly so cumbersome to transport as manometers and pipettes; clean
ing was not a problem, nor was the equipment as likely to be damaged in
transit. In addition, if one desired to maintain the animals at a specific
temperature during an experiment, a wash basin or large pan could hold a
dozen or more flasks at one time and the temperature of the water In the
bath could be maintained within Io C. of the desired temperature by
addition of ice or warmer water, as required; hence, no special bath,
refrigerator or heater was required. In addition to technical advantages
of the method, it allowed measurement of oxygen uptake under actual field
conditions, although this opportunity was not realized. Since it was
possible to make measurements shortly after the tadpoles were collected,


TABLE 4 (Continued)
LOCATION, HABITAT AND DATE
Bromeliad of Puerto Rico habitat of
no known tadpoles
8. El Yunke Biological Station (6 V)
Ponds and standing waters of Haiti -
habitats of Hyla dominicensis
9. 8 km W of Port-au-Prince (12 V)
10. 32 km W of Port-au-Prince (4 VIII)
11. 4.8 km W of Port-au-Prince (10 V)
(12 V)
Ponds and standing waters of Grand
Cayman, B.W.I. potential habitats of
Hyla septentriona 1 ?s
12. 3-2 km N of Georgetown (29 X)
13. Hell (29 X)
14. South Sound (29 X)
16. 4.8 km E of Georgetown (29 X)
18. Boddentown to East End (29 X)
TEMPERATURE
OXYGEN(ppm)
HARDNESS(ppm)
21.0
3.1
-
5-2
32.0
0.7
480
7-5
39-2-40.2(2)
1.4-8.0(2)
220(1)
7-5
25.1
2.2
200
-
24.6
1.7
200
7
32
-
64
7-0
33.5
-
60
7-0
28-33(2)
-
84
7-5
32.5
-
50
7-0
28-29(3)
-
60-90
7-0


82
Red blood cell measurements. -- An optical micrometer was
used to measure length and width of twenty-five red blood cells from each
individual. Measurements were made from smears prepared and stained
in the field. Average cell length and width were determined and con
verted to microns. These blood smears were used to study abnormal cell
types in the blood.
Effects of temperature. In order to determine the effect of
temperature upon frog erythrocytes, blood from specimens of Hyla
squirrella, Bufo terrestris and Scaphiopus holbrooki was collected in
heparinized capillary tubes. The tubes were plugged with clay at both
ends and kept in a water bath at 45 C. for varying periods of time. After
this treatment, smears of the blood were made, stained and studied.
Measurement of tadpole respiratory rates
Measurement of respiratory rates of tadpoles was carried out
by two different methods. During the earlier part of the work, oxygen
uptake was determined by use of Warburg constant volume respirometers
as described by Umbreit, Burriss and Stauffer (1964).
The second method was improvised in order to determine the rate
of oxygen utilization by tadpoles shortly after capture, under either
controlled or field conditions. Necessary apparatus for field measure
ments were Erlenmeyer flasks with rubber stoppers, a Precision Scientific
portable oxygen ana lyzer, a-nd a wristwatch. The simple technique con
sisted of the following steps:
1. Clean 125 ml Erlenmeyer flasks were calibrated to determine
their actual volume when stoppered.
2. The appropriate number of clean Erlenmeyer flasks were filled
with clean tap water of the desired temperature.


12
was no lag period when a substrate of hemiglobin-haptoglobin was used.
In addition, the enzymatic activity was about 50 per cent higher with
hemiglobin-haptog1 obin or carboxyhemoglobin-haptoglobin than with
hemoglobin-haptoglobin as a substrate. Thus, it would appear that
hemiglobin-haptoglobin is the primary substrate of the enzyme and that
the lag in degradation of the hemoglobin-haptoglobin represents the
period during which hemoglobin is being changed into hemiglobin. The
enzyme acts only in the presence of oxygen and is considered an oxidase,
rather than a peroxidase. It is known as heme a 1pha-metheny1 oxygenase
and is found largely in the liver and kidney, being nearly absent from
the spleen and bone marrow. The product of this enzymatic reaction
is then acted upon by a second enzyme, heme a 1pha-metheny1 formylase,
to yield biliverdin, iron and formaldehyde.
An interesting observation which may be inserted at this point
is that biliverdin conversion to bilirubin has been shown to be under
enzymatic control in the laboratory rat (Lester e_t ajk 1966).
Causes of Chlorosis
Barrio (1965a) noted that the three families of Neotropical
frogs which include chlorotic species--Hy1idae, Centrolenidae and Pseudidae
also have in common an intercalary cartilage between the ultimate and
penultimate phalanges of the digits. He suggested that these two
common factors indicate a close relationship among these families.
However, the intercalary bone of pseudids is believed to represent an
adaptation for swimming, similar to the situation in aquatic mammals
(Goin and Goin, 1962a). Most recent writers have considered the presence


78
Intraperitonea1 injection of hemoglobin solution resulted in
no clear indication of green tissues in any of the species. Several
individuals voided substances which had absorption peaks that corres
ponded to biliverdin or hemoglobin. There appeared to be no addition of
green pigment to the tissues. Judging from the color of the fluids
voided by these animals, bile pigments were formedand eliminated within
twenty-four hours after injection of hemoglobin. However, it should
be pointed out that green pigment was not excreted by a 11 of the frogs
which were injected with hemoglobin. In addition, one untreated specimen
and one injected with only phosphate buffer demonstrated green pigment
in voided fluids. This was also true of other untreated frogs in the
laboratory.
Injection of biliverdin
Still a third method was usedthe direct injection of
biliverdin in sesame oil or phosphate buffer (pH 7-^0 On one occasion,
an attempt was made to mobilize bone calcium prior to injection of
biliverdin. Several specimens of Hyla cinerea were injected with 0.5
(.005 cc) units of Lilly's Parathyroid Extract per gram of body weight.
The same dosage was repeated twelve hours later. After another six
hours, an intraperitonea1 injection of biliverdin in sesame oil was
administered. Finally, they were given two drops of cod liver oil after
an hour. The latter step was intended to stabilize calcium in bone because
of its vitamin D content (Cantarow and Schepartz, 1957).
Biliverdin solutions injected into the peritoneal cavity
caused no apparent change in the color or other visual characteristics


83
3. The probe of the oxygen analyzer was calibrated in air.
After dipping the tip of the probe into clean water,
it was waved in the air while two ammeter readings were
taken. The ammeter of the oxygen analyzer measures electri
cal current between silver and lead electrodes located in
the tip of probe; this current is directly proportional to
the oxygen which diffuses through the covering membrane and
into the space between the electrodes. Air temperature was
determined at the same time by use of a thermistor attached
to the probe. These data were used to calculate the sen
sitivity of the probe under prevailing conditions.
4. The probe of the oxygen analyzer (which had been taped to
fit snugly into the neck of the Erlenmeyer flask) was
lowered carefully into the neck of the flask until the probe
formed a watertight seal with the neck of the flask.
After this was accomplished, the tip of the probe projected
slightly below the neck of the flask. When done properly,
no bubbles were trapped in the flask.
5. While holding the flask and probe together, they were
rotated vertically through a 30 degree arc, so that their
axis was in a horizontal plane. From this position, the
flask was shaken from side to side in order to cause water
to flow past the tip of the probe. This procedure was
necessary to improve the accuracy of the method. In the
laboratory, a magnetic stirrer was used to cause a flow
across the tip of the probe.
6. Once the indicator of the ammeter became steady, a reading
was made. The ammeter was then turned off for a short time
and then turned on again. While the shaking continued a
second reading was made and the average of the two readings
was recorded. Probe sensitivity, ammeter reading and water
temperature was used to calculate the initial oxygen con
centrad on.
7. Once the data necessary to determine the initial oxygen
concentration were obtained, a single tadpole was intro
duced into the flask, and the flask was stoppered. If an
air bubble was trapped within the flask, sufficient water
from the original source was added so that the flask was
completely filled with water when stoppered. This was
necessary to prevent intake of air by the tadpoles and also
to prevent immeasurable exchange of gases between air and
water. Note was made of the time that the flask was
stoppered.


98
toxic on the basis of degree of hemolysis produced. Reeder (1964)
has recently summarized the meager knowledge concerning amphibian bile
salts, none of which refers to species known to have biliverdin in
tissues.
Among West Indian hylids, only Hyla marianae has orange bones
during metamorphosis. The reasons for this species being different from
the others are unknown, but speculation is possible. The extract of Hyla
marianae skin contains at least one light orange pigment. Although the
pigment was not identified, it shares spectral characteristics (Figure 6)
with formyl derivatives of folic acid (Stokstad and Koch, 1967)- Coenzymes
containing folic acid are important in transfer of certain one-carbon
groups, oxidation-reduction reactions, synthesis of purines and some
pyrimidines, and in amino acid metabolism, including the transformation
of phenylalanine to tyrosine. Thus, there seems to be the possibility that
this species has a rich supply of folic acid or similar substance which
could allow the liver cells to divide rapidly (because of readily avail
able purines and pyrimidines for DNA synthesis) in order to excrete an
increased amount of biliverdin; produce more melanin (because of avail
ability of tyrosine) than pale, green forms such as Hyla wilderi; and
have sufficient folic acid left that it could be used to eliminate formyl
groups (perhaps from enzymatic biliverdin formation) by way of the skin.
An excess of this vitamin might account for the relatively large size of
these tadpoles at metamorphosis, as well as their aggressiveness. The
source of folic acid is not readily apparent, but might be the leaves
of the small bromeliads in which this species was found. An alternate


102
cause their extreme green pigmentation. It should be noted that
these tadpoles were in the process of metamorphosis and may have had
an excessively high respiratory rate for this reason.
The very low metabolic rate of Pseudis tadpoles may be an
adaptation to low oxygen concentrations within their environment or be
correlated with their large size, but they still develop green tissues.
Desiccation could well be a factor contributing to reducing
the flow of bile through the liver. With less internal water for physio
logical purposes, secretion and excretion of bile may be impaired, so
that bile pigments would occur in the blood. Since the tree frogs, in
their arboreal habitat, appear to be more susceptible to chlorosis than
do the ground frogs, one might suppose that their more exposed position
is the reason. Desiccation is also indicated by the concentrated con-
dition of the bile.
Obstruction of the extra-hepatic bile ducts by gall stones,
parasites or tumors was not observed in the frogs studied. One specimen
did have an inflamed duodenum which may have reduced bile flow into that
portion of the intestine. In most specimens, the gall bladder was filled
with green bile and the feces were dark, indicating that bile was passing
out of the liver, and into the intestine, respectively.
Advantages of the accumulation of green pigment in skin and other
organs seem to lie largely in the realm of protective coloration. In
a number of species, the green color is rarely visible externally, but in
some of the smaller speciesHy la wilderi, and species of Centro lene 1 la
and Sphaenorhynchusthe green pigment is quite apparent and blends well
with the green foliage on which these forms are found. This green


63
The tadpoles and adults of Hyla septentriona1is are very
similar to those of Hyla dom?nicens?s; in fact, the two forms have been
considered races of the same species (Barbour, 1937) The tadpoles of
Hyla septentriona1is are black when young and become brown toward meta
morphosis. They are primarily vegetarian, but on one occasion I observed
a larger tadpole rapidly ingest a smaller one. The gills are well developed,
but lungs are not. Body shape is rounded, and, like that' of other hylid
tadpoles in the West Indies, does not bear an extension of the tail fin.
The rate of development is very rapid as previously mentioned. Grant
(1940) stated that C. Bernard Lewis found "that eggs laid in a tiny
puddle in coral or tree trunk during the night will be active tadpoles
by mid-day." English (1912) found that these frogs breed any time of the
year after a good rain, with eggs developing to frogs within a few weeks.
He also indicated that the tadpole and frog are enemies of mosquitos which
are abundant on Grand Cayman. This is probably true, since the smaller
bodies of water in which these tadpoles are found are ordinarily clear
and free of other obvious forms of animal life, such as mosquito larvae.
Hyla septentriona1is tadpoles have been found in brackish
pools on the Florida Keys (Neill, 1958). Larvae of this species often
have a clear envelope of fluid surrounding the body which is contained
within an outer layer of skin. This would seem to be an adaptation
which slows diffusion of water or salts between the tadpole body and its
environment. It should also reduce heat exchange. It is possible that
these tadpoles are capable of slight osmoregulation, since an individual
lived nearly four days in sea water diluted by distilled water to 10 parts
per thousand salinity. At 15 and 20 parts per thousand salinity, the


54
Sma1ler hvlas. -- ¡P Surinam, several species of small or
medium-sized hylids were found calling or breeding in open, grassy ponds
or flooded fields with a dense growth of weeds. Both of these situations
appeared to be temporary due to increased rainfall at that season
(July).
Two very small species, Hy la mi sera and Hyla nvnuta, as well
as Hyla boesemani, Hyla eg leri Hyla leucophy1 lata, Hyla rubra, Hyla
punctata, and the medium-sized Hvla crepitans, were included in this
group. Goin and Layne (1958) recorded Hyla rubra, Hyla mi sera, Hyla
punctata, Hyla leucophy1 lata, and Hyla lanciformis from similar habitats
near Leticia in southern Colombia.
While several of these species were found in the same pond in
Surinam, they often had different habits. Hyla misera usually called
from a perch on stalks of grass which were surrounded by water; Hyla
crepitans was found floating in the water near clumps of grass; and
Hyla boesemani called from the grasses at the edge of the same pond.
Within this group, Hyla rubra was most likely to be found in the vicinity
of Paramaribo, becoming less frequent at higher elevations to the south.
Hyla punctata was found perched on weeds in a flooded field and Hyla
leucophy1 lata called from bushes along a drainage canal.
Of the eight ...species collected in Surinam in this habitat,
Hyla egleri (one individual), Hy la leucophy1 lata (three), Hyla minuta
(two), and Hyla rubra (four) showed no green pigmentation. Hyla
boesemani (one), Hyla crepitans (two), and Hyla punctata (two) had green
bones and the latter two species had green plasma as well. The single
specimen of Hyla mi sera had light green plasma.


75
but histological study ordinarily shows the degeneration of liver which
is associated with prolonged cholestasis.
Several factors indicate that the study of tadpole respiration
may shed light upon chlorosis. First of all, the present study has
shown that green pigmentation first appears in the tadpole stage of
several species (Table 2). Secondly, the low oxygen tensions of some
tadpole environments suggest that the larval stages have special respir
atory adaptations or that tissue damage may result from oxygen deficiaicy.
The degradation of respiratory pigmentsapparently the immediate cause
of chlorosissuggests a disruption or decrease of oxygen transport
within the organism. Since the liver is quite sensitive to oxygen
deficiency, one wonders if the preceding factors are concerned with
possible liver malfunctions. Finally one should consider the high environ
mental temperatures to which these organisms are exposed, and the greater
stress which this factor places upon respiratory and hepatic systems.
In order to determine the importance of biliverdin formation
to chlorosis, several attempts were made to induce green pigmentation
in vivo, using non-chlorotic species. This was followed by a broader
physiological study of Neotropical frog speciesboth chlorotic and non-
chloroticto clarify the importance of biliverdin formation, liver
function, and other factors which might pertain to the green pigmentation.
Attempts to Reproduce Green Pigmentation In Vivo
During the early part of the study, attempts were made to dupli
cate the high concentration of tissue biliverdin, particularly as it
occurred in bone. These experiments were carried out upon species which


LIST OF FIGURES
Figure Page
1 Absorption Spectrum of Biliverdin in
5 Per Cent Hydrochloric Acid Methanol
Solution 132
2 Absorption Spectrum of Liver Extract
From Hyla septentriona1is in Aqueous
Solution 132
3 Absorption Spectrum of Lymph of
Osteocepha 1 us taurinus 133
4 Absorption Spectrum of Coelomic Fluid
(bi le) From Osteocepha 1 us taurinus 133
5 Absorption Spectra of Bile Solutions of
Siren lacertina 134
6 Absorption Spectrum of Pigment Extracted
From Hyla marianae in 70 Per Cent Ethanol 134
7 Respiratory Rates of Hvla brunnea Tadpoles
at Different Temperatures 135
8 Respiratory Rates of Tadpoles at 25 C 136
9 Normal Red Blood Cells of Bufo terrestrls
Four Hours After Collection 137
10 Red Blood Cells of the Same Individual as
Above, After Four Hours at 45 C 137
11 Red Blood Cells of Hvla squire 1 la After
Four Hours at 45 C 138
12 Untreated Blood Cells of Hvla leucophvl lata.. 13 8
13 Stained Smear of Untreated Blood of Bufo
tvphoni us 139
vi i


ACKNOWLEDGMENTS
I wish to thank Dr. Coleman J. Goin for suggesting the
challenging topic of this paper and for his continued interest and
support during the various phases of its study. Dr. Goin has aided in
the identification of frogs used in this study and has extended many
other courtesies which are deeply appreciated. I am grateful to Dr.
Mildred Griffith, Dr. Frank J. S. Maturo and Dr. James Nation for their
aid during this study and for suggestions which improved the dissertation.
To Olive B. Goin, I owe thanks for many kind acts and stimulating
discussions.
Several individuals have made laboratory and field facilities
available to me. Among these were: Dr. James Nation, Dr. Ward Noyes,
Dr. Albert Laessle, Dr. Frank Nordlie, Dr. Paul Elliott, and Dr. James
Gregg, all of the University of Florida; Dr. Harold Heatwole, University
of Puerto Rico (El Yunke Biological Station); Mr. Bernard Lewis and
Dr. Thomas Farr, Institute of Jamaica; Dr. Ivan Goodbody, University of
West Indies, Kingston, Jamaica; Dr. Jocelyn Crane, Simla Biological
Station, Trinidad; and Mr. Walter Polder, Paramaribo, Surinam.
To the following persons, I am grateful for the opportunity
to study the specimens of their respective institutions: Dr. Ernest
Williams, Museum of Comparative Zoology, Harvard University; Mr. C. M.
Bogert and Dr. Richard Zweifel, American Museum of Natural History;


111. PHYSIOLOGICAL FACTORS RELATED TO CHLOROSIS
Previous studies have shown that the immediate cause of
chlorosis is the production of biliverdin in quantities greater than
those which are excreted. It has not yet been determined if chlorosis
in frogs is due to a high rate of biliverdin formation, impairment of
bile pigment excretion, or both.
Since hemoglobin of the blood is ordinarily the main source
of biliverdin, it is logical that one should study this protein and the
red blood cells in which it is found. If chlorosis is due primarily to
a high rate of biliverdin formation, one might expect to find higher
concentrations of hemoglobin or higher rates of hemolysis in chlorotic
frogs than in others. In addition to chlorosis itself, high rates of
hemolysis might be indicated by low red cell counts, or by blood smears
which show immature erythrocytes or cell fragments. Should hemolysis
or other process be shown to be the cause of increased biliverdin, then
the cause of such a process should be determined. In the case of rates
of hemolysis, it is known that they may be correlated with temperature,
concentrations of chemical agents, or differences in red cell structure.
Impairment of bile pigment excretion would involve the liver
and its associated ducts. Blockage of extra-hepatic ducts is readily
determined by the absence of biliverdin from feces. Blockage or
restriction of bile flow within the liver may be difficult to detect,
74


97
Another type of large-scale hemolysis found among anurans
occurs at metamorphosis. McCutcheon (1936) has shown that larval
hemoglobin of Rana catesbeiana is different from that of the adult.
Herner and Frieden (1961) have shown that the hemoglobins of larval
Xenopus laevls, Rana heckscheri and Rana catesbeiana are replaced by
adult hemoglobins during the course of metamorphosis. It seems likely
that this is the explanation for the appearance of biliverdin during
metamorphosis of most West Indian hylids. A similar change from fetal
to adult hemoglobin in man is partly responsible for neonatal jaundice.
Phytohemagglutinins may be used to hemolyze erythrocytes and
induce division of leukocytes in frog blood. However, the effects of
these water-soluble plant derivatives in vivo are poorly known (Boyd, 1963).
Deficiency of vitamin E may result in changes to the cell
membrane of red blood cells, causing them to hemolyze more readily.
This may be a factor in frogs which appear to have more fragile
erythrocytes.
A possibility which has been considered unlikely is that hemo
lysis may be a result of the effect of ultraviolet light upon hemoglobin
(Lemberg and Legge, 19^+9) In view of the fact that blue light may be
used to reduce circulating bilirubin in infants (Broughton ejt a_L 1965),
it is conceivable that the high incidences of ultraviolet light in tropi
cal areas may have an effect on circulating red cells.
Increase in circulating bile salts as a result of defective
liver function may be a cause of hemolysis (Grodins, Berman and Ivy, 1941).
These authors found the apocholate and deoxycholate salts to be the most


77
methods.used, it did not induce detectable hyperbi1iverdinemia. Severe
anemia developed in specimens of Hyla cinerea as a result of injection
of this drug. Two smaller individuals died several days after the
second injection. Post-mortem examination showed that very few red
cells were present in the blood vessels or elsewhere. There was slight
indication of hemopoietic activity in the long bones; the spleen was
pale and showed no increase in size. The kidneys were slight yellow in
color, rather than the usual red. The frogs appeared emaciated. The
liver was normal in the larger frog examined; the smaller frog had an
enlarged gall bladder filled with green fluid and its liver was smaller
than expected. None of the frogs and toads used in this experiment showed
any evidence of green pigmentation in bone, blood or other tissues outside
of the digestive tract.
Administration of hemoglobin solution
A second attempt to increase the amount of biliverdn in
living anurans involved the intraperitonea1 injection of a 1 per cent
hemoglobin solution. Specimens of Hyla septentriona1?s (eleven individuals),
Rana pipiens (four individuals), Rana catesbeiana (two individuals),
Hyla gratiosa (one individual), Bufo terrestris (two individuals), and
Hyla cinerea (seven individual^ were injected with either 1 per cent
hemoglobin in phosphate buffer, or with phosphate buffer (pH 7-^) alone,
or were given no treatment. Fluids voided by these animals were taken
from the jars in which they were kept and analyzed with the Beckman DU
spectrophotometer.


SOURCES AND CARE OF TADPOLES AND FROGS
The live frogs and tadpoles used for most of this study were
collected from their normal habitats during field trips to the West
Indies and South America. The species included: Hyla septentriona1is
collected from Grand Cayman and Southern Florida; Hyla brunnea, Hyla
1ichenata, Hyla marianae, and Hyla wi1deri from Jamaica; Hyla domini-
censis, Hyla vasta and Hyla heilprini from Haiti ; Leptodacty1 us
a 1bilabrus of Puerto Rico; and Hyla maxima and Pseudis paradoxus from
Surinam. Incidental observations were made of Phy1lomedusa tadpoles on
Trinidad and several other hy1id species in Surinam. Collections were
made at the following places during the months indicated: Jamaica
June-July, 1965; August, 1965; October, 1965; May, 1966 and August, 1966;
Haiti -- October, 1965; May, 1966 and August, 1966; Grand Cayman --
July, 1965 and October, 1965; Puerto Rico -- May, 1966; Southern
Florida July, 1966; Trinidad -- July, 1966; Surinam -- July, 1966.
Tadpoles of the species listed above were collected with a
tea strainer or small dip net. They were transferred to plastic bags
which were partially filled with water from the immediate environment
of the tadpole. When suitable tap water was available, it was used to
replace water brought from the field. Thus the frog larvae were trans
ferred to water which was apparently free of noxious substances,
detritus, and other organisms. Water was poured off and replaced with
fresh tap water as often as required, usually at one- to three-day
intervals. Tadpoles did not show marked reactions to addition of tap
24


8
The absorption curve of bil verdn has been shown (Figure 1).
Mesobi1iverdin has a similar curve, but its absorption maxima lie about
10 millimicrons further from the infrared than do those of biliverdin.
Neither of these compounds has an absorption peak between 400-440
millimicrons (Soret band), which is characteristic of compounds with
the closed porphyrin ring (Holden and Lemberg, 1939)-
Biological properties
The biological properties of biliverdin are related to its
physical and chemical properties. Of considerable importance is its
solubility in different solvents. The fact that Barrio (1965a) did not
find biliverdin in nervous tissue, ocular fluids, or urine may be
directly related to the insolubility of unconjugated biliverdin in
water.
Insolubility of biliverdin in water presents a problem in
regard to its transport within the body and its excretion. In mammals,
as well as some birds and reptiles and Bufo arenarum, this problem is
not so important because the liver conjugates variable amounts of the
bile pigments (either bilirubin or biliverdin) with glucuronic acid to
form water-soluble glucuronides (Rodriguez Garay et_ a_l_. 1965). These
investigators also found a sodium salt of biliverdin, free biliverdin,
and a substance which had chromatographic characteristics of a complex
of bile acid and sodiurn bi1 iverdi na te in Bufo arenarum. The electro
phoretic studies of Barrio (1965a) indicate that in the circulatory
system biliverdin is associated primarily with globulin proteins, but
some is conjugated with albumin. Non-esterified fatty acids appear to


10
during erythropoies is. The third component is formed during the
first twenty-four hours, probably beginning during the second hour
after administration. The fourth fraction appears most rapidlywithin
a few minutes of intake. The latter fraction was detected in rat
liver homogenate after five minutes' incubation with labeled delta-
aminolevulinic acid (a precursor of protoporphyrin). Presence of such
a component was suspected when the radioactive bilirubin recovered
during the first two hours was found to account for one-third of the
twenty-four-hour total. A block of the bile duct or similar obstruction
is shown to increase the latter bilirubin fraction. Perfusion experi
ments and other data show that the latter two fractions are of hepatic
rather than erythropoietic origin. The main non-erythropoietic component
of bilirubin, which is relatively slow in being formed, "is probably
related to the turnover of the major hemeproteinssuch as liver
catalase. The smaller component arises rapidly from a pool of heme-protein,
heme, or heme precursors which has a high turnover rate. Barrio (1965a)
suggested that high concentrations of biliverdin in frogs may be due to
increased formation of early biliverdin.
Mechanisms of bile pigment formation
Earlier hypotheses. -- Much study has gone into the chemistry
of bile pigments and their formation from hemoglobin. Most of these
studies have dealt with non-enzymatic reactions of bile pigments and
related compounds carried out in vi tro. Early studies of hemoglobin
degradation employed harsh techniques and detected such substances as
hematin and hematoporphyrin. With such supposed intermediate compounds


106
14. In general, it seems that the green pigmentation is due
to increased hemolysis coupled with a decreased ability of the liver
cells to excrete this bile pigment, so that it accumulates in plasma,
then stains the proteins of soft tissues and finally those of bone.
High environmental temperatures and hormonal changes are considered the
most likely causes of hemolysis, and perhaps impairment of liver function
as well.


59
After the group of tadpoles dispersed, the behavior of indi
viduals was noted. Most appeared to be resting near the bottom of the
flooded area while others appeared to be feeding along the bottom. At
intervals, they came rapidly to the surface to gulp air and then returned
to the bottom to rest or continue feeding. Four individuals which were
observed for five or more intervals had average interval lengths of 17,
22, 27, and 31 seconds. It appeared that the smaller and more active
ones surfaced more often. Several hours after some of these tadpoles
had been collected, their tails appeared quite red. When this habitat
was visited again in August, 1966, a flocculent precipitate covered
most of the surface and no tadpoles were present.
Young larvae of Hyla dominicensis are black, but they become
brown or gray and develop green pigmentation in bone and soft tissues
during metamorphosis. The black pigmentation of the young may be an
adaptation to an environment exposed to the sun, since this character
is seen in Hyla vasta and Hyla septentriona1?s in similarly exposed
environments, but not in the Jamaican hylid tadpoles which develop in
shaded habitats. Gills are well developed in this species of tadpole,
but they appear to lack lungs until metamorphosis. Therefore, the air
which they take at the surface must be kept in the mouth where gaseous
diffusion takes place, and then the bubble is released at or near the
surface prior to taking in a fresh gulp of air. The complex gill
structure probably is important in filtering food, primarily algae,
from the water. Like most vegetarian tadpoles, these have a long, coiled
intestine. While the tail of this species is not particularly muscular,


52
movement of the larva; occasionally, the tadpole moves more rapidly
by using the entire tail to propel itself. Both the tadpole and the
adult of this species give the impression of being quite slow. The
assumption might be made that this indicates a low rate of metabolism,
but this was not measured in frog or larva. As in the adult, no green
pigmentation were found in the tadpole.
A relatively large number of frog species was found in exposed
habitats where there was standing water. Neither tadpoles nor adult frogs
were seen often in such habitats during daylight hours. Tadpoles gener
ally avoided bright sunlight by swimming under aquatic vegetation.
Elachistocleis ovale. -- In Surinam, this small microhylid was
found calling from shallow roadside ditches containing standing water.
The ditches had a moderate growth of grass and were generally exposed to
the sun. This species was found in Habitat 25 (Tables 3 and 4). Neither
eggs nor tadpoles were found. The three adults examined did not show
any indication of green tissues (Table 1).
Pseudis paradoxus. -- This highly aquatic medium-sized frog and
its exceptionally large tadpoles were found in roadside ditches and
drainage canals in Surinam. Water in these ditches usually exceeded
50 centimeters in depth, was semi-permanent and stagnant. Vegetation
in this habitat was varied and served as food for Pseudis tadpoles.
While the eggs of this species were not seen In Surinam, Gans (1956) found
that this species lays its eggs in a frothy mass which floats on ponds
in Trinidad. The black tadpoles grow to more than 15 centimeters in
total length, a third of which is body length. Body length of the largest


7
it does not contain iron as does heme. Mesobi1 iverdin is formed from
biliverdin by reduction of the vinyl side chains to ethyl groups.
Chemical reactions of biliverdin include the Gmelin reaction.
Addition of fuming nitric acid to tetrapyrrole bile pigments converts
all initially to biliverdins. These are oxidized further to yield a
succession of pigments from blue-green through violet, red and yellow
to colorless compounds.
As already indicated, reduction of the vinyl side chains of
biliverdin results in mesobi1 iverdin. Reduction of the middle methyne
bridge yields bilirubin, a common bile pigment of mammals and other
vertebrates. Bilirubin may be reduced further to other bile pigments.
Biliverdin is destroyed by heating with concentrated sulfuric
acid, but mesobi1 iverdin is not. Stable hydrochlorides or hydrobromides
result from treatment with the appropriate acid. Complex salts may be
formed with iron, zinc and copper. Biliverdin also undergoes esteri
fication. Unlike bilirubin, biliverdin does not couple with diazotized
sulfanilic acid to give the diazo reaction (Lemberg and Legge, 1949).
Physical properties
Biliverdin is moderately soluble in ether, in which it has a
greenish-blue color. It may be extracted from ether by 1 percent
hydrochloric acid, in which it has a blue-green color. These deep
colors result from the conjugation of double bonds throughout the length
of the molecule. Biliverdin is also soluble in methanol (Lemberg and
Legge, 19^+9) and chloroform, but is insoluble in water (Barrio, 1965a).


II. CONSIDERATION OF ECOLOGICAL FACTORS ASSOCIATED
WITH CHLOROSIS
In attempting to determine the cause of a condition such as
chlorosis, ecological factors should be considered. It has already
been noted that the majority of chlorotic anurans are tree frogs
which are restricted to tropical or sub-tropical regions. However,
not all tropical tree frogs have green tissues; conversely, the chlorotic
pseudids are not tree frogs, but are aquatic. Since the tropical climate
and general habitat (aquatic, terrestrial or arboreal) do not completely
account for either presence or absence of chlorosis, one should con
sider the importance of specific habitats, as well as behavioral adapt
ations to such environments. This section includes a summary of
information concerning specific habitats and the frogs associated with
them. Greater emphasis was placed upon ecological study of breeding
sites because of their accessibility and the appearance of chlorosis
during the larval stages. Special attention was given to those factors
which might be related to hemoglobin or red cell formation and function,
including oxygen tension, temperature and iron concentration. Additional
factors were surveyed in order to find conditions which were markedly
different from those ordinarily encountered by anurans.
Methods of Study
Concentration of dissolved oxygen in the habitats of tadpoles
was measured with a Precision Scientific Portable Oxygen Analyzer, which
was calibrated in air.
35


14
possessing unusual respiratory structures (Negus, 1965) The several
types and combinations of respiratory organs shown by species of
Amphibia have been the interest of many investigators of respiration.
Most reports have been concerned with the structure of respiratory
organs and surfaces, rates of oxygen uptake, and the role of the blood
and circulatory system in gaseous transport. Some work has been done
on larval amphibian respiration, especially on those species of sala
manders and frogs commonly used in embryologica1 and physiological
research, despite their small size. There are few studies, however,
of adaptations in relation to the ecological situations of larval or
adult amphibians (Foxon, 1964).
Structure and function of respiratory and associated circul
atory organs of amphibians in general have been reviewed by Noble (1931)
and more recently by Goin and Goin (1962a) and Foxon (1964). The present
survey will be concerned specifically with the respiration of amphibian
tadpoles and similar aquatic forms, as well as with hemoglobin and
red blood cells of amphibians generally.
As the amphibian embryo develops, gaseous exchange first occurs
through the egg surface, later through the external body surfaces
(including gills), and finally through lungs, if present (Foxon, 1964).
Among the species of anurans, the respiratory surface decreases pro
portionately as the diameter of the egg increases, since the volume of
the egg increases as a cubic function of the radius, while the area
increases as a squared function. No examplesof enlarged surface area
due to folding or other changes in the egg membrane are known (Noble, 1931).


123
TABLE 5. IRON AND PHOSPHATE CONCENTRATIONS (PARTS PER MILLION) IN
TADPOLE HABITATS
HABITAT DATA AND LOCATION
IRON
PHOSPHATE
SPECIES OF TADPOLE
PRESENT
Bromeliads of Jamaica
1. Rose Valley
0.00
0.00
Hyla brunnea
Trace
0.40
None
0.05
2.2
None
0.02
0.30
None
3. 6 km S of Mandevi1le
0. 16
0. 10
Hyla brunnea
0.25
0.21
Hyla brunnea
0.73
3-4
None
0.30
2.7
None
Standing waters
10. Wheel ruts 32 km W of
1.13
1.35
Hyla dominicensis
Port-au-Prince, Haiti
1.80
4.80
Hyla dominicensis
24. Domberg, Surinam
7-5
0. 18
Pseud is paradoxus
25. 28 km S of Paramaribo
0.62
0.70
Phrynohyas venulosa
Surinam
Stream habi tats
28. Princess Irene Falls
0.06
0.02
Hyla maxima (larqe)
Basin near the Fa 11s
0.18
0.03
Hyla maxima (sma11)
30. Furcy, Haiti
0.
, 10-0.26
0.35
Hyla heilprini, Hyl
32. Troja, Jamaica
0.
,02-0.40
0.39-0.41
Bufo marinus
33. 4.8 km W of Bog Walk
Jamaica
0.30
0.25
None
vasta


125
TABLE 6 (Continued)
I
FAMILY AND SPECIES
HEMOGLOBIN
(Gram Per Cent)
RED CELL
COUNT
HEMOGLOBIN
PER CELL
(Picograms)
Hvla leucophy1 lata
7.0
460,000
152
Hyla marianae
543,000
*Hyla maxima
14.5
967,500
150
14.5
1,395,000
104
7.5
512,500
146
*Hvla punctata
3.0
615,000
49
Hyla rubra
13-5
965,000
140
7-5
675,000
111
10.0
790,000
127
*Hyla vasta
11.0
742,500
148
*Hyla wiIderi
212,000
*Osteocepha1 us taurinus
6.5
660,000
98
'"Phrynohyas venulosa
9-5
1,212,500
78
4.5
950,000
47
11.5
1,157,500
99
16.0
1,378,000
116
Phyllomedusa bicolor
5.5
322,500
171
Phyllomedusa hypochondri a 1is 6.0
255,000
235
6.5
287,500
226
5-5
360,000
153
7.0
478,000
146
''Sphaenorhynchus aurantiacus 7.5
585,000
128
2.0
217,000
92


V
139
Figure 13. Stained smear of untreated blood of Bufo typhonius. Note
coagulated cytoplasm. Green tissues are unknown in this
species. (400X)
Figure 14. Stained smear of untreated blood of Phrynohyas venulosa.
Note the clumped cytoplasm. (400X)


51
reduced the size of this pond. Water temperatures rose to 38 C. and
the embryos were dead.
Phyllomedusa bicolor. -- This is a large arboreal species which
appears to breed near stagnant water which may be in open areas with
some exposure to the sun. One adult male was collected from the surface
of a main road on the upper coastal plain of Surinam. Another was heard
calling from a site high up in a tree in the forest near Powakka. The
male which was collected had no green pigmentation in the tissues. As
with other members of this genus, this species moves very slowly, never
hopping, but walking on all fours. The whitish color of muscular tissue
in this individual may be an indication of poor vascularization or a
relative lack of myoglobin.
Phy1lomedusa eggs and tadpoles believed to be those of
Phyllomedusa bicolor were observed at the New York Zoological Society's
biological station at Simla, Trinidad. The large, yellow eggs were
attached to the upper side of a large shaded leaf which overhung a pool.
Water lilies and water hyacinths covered part of the surface of this small,
ornamental pool, which harbored the Phy1lomedusa tadpoles of various
sizes. The tadpoles were slaty-blue in color, much the same color as the
adult dorsum in preservative. They fed upon aquatic vegetation and were
observed coming to the surface. Lutz (195*0 stated that the larvae of
Brazilian species of Phy1lomedusa use their rudimentary lungs as hydro
static organs. Movement of these tadpoles is accomplished with a
minimum of effort. The main propulsive structure is the narrow, upturned
tip of the tail which vibrates constantly. This causes a slow, steady


5
hylid species. Dunn (1931), B. Lutz (1947), and Savage (19&7) noted
the presence of green bones in certain species of centrolenids.
Barrio (1965a), working with frogs from Argentina, found
that five of twelve hylid species and all three of the pseudid species
which he studied had green tissues. Individuals of twenty-two species
of four other families lacked the green pigment in their tissues.
The species that had green tissues were similar in that the muscular
and subcutaneous tissue, lymph, walls of the digestive tract, eggs of
mature females, and especially the bones were strongly pigmented. He
attributed this situation to high concentrations of biliverdin in the
blood. The highest concentration, 11.9 mg/100ml, was found in the serum
of Hyla punctata. In addition, he pointed out that biliverdin Is
decomposed by formalin. This point clarified an earlier controversy
(Boulenger, 1880) and also indicated a reason for so little attention
being given to this phenomenon.
Another paper by Barrio (I965b) brought out the fact that
populations of Hyla pul che 1 la differ greatly in their serum concentra
tions of biliverdin; two of the five subspecies lack the pigment
a 1 together.
Green tissues of other animals
Green tissues are not limited to frogs. Peters (1873) noted
that the green bones of Pseudis minuta were similar in color to those
of the marine garfish Be 1 one. The relatively recent papers of Wagenaar
(1939), Fontaine (1941a, 194lb), Willstaedt (1941), Caglar (1945), and
Fox (1953) concerning green bones of teleost fish pointed toward


126
TABLE 6 (Continued)
FAMILY AND SPECIES
HEMOGLOBIN
RED CELL
HEMOGLOBIN
(Gram Per Cent)
COUNT
PER CELL
(Pi cograms)
Leptodacty1idae
Leptodactv1 us
pentadacty1
I us 7-0
467,500
150
Leptodacty1 us
sp.
11.5
1, 100,000
105
Leptodacty1 us
sp.
6.5
850,000
76
Chlorotic species


30
to be restricted to the marrow, since neither the bone nor the plasma
demonstrated green pigment. Although the origin of this pigment
probably was the same as that of the Neotropical forms, the color was
almost obscured by the bone.
Green pigmentation of the adults is summarized in Table 1.
It is worthwhile to compare the amount of green pigment in adults, with
that of tadpoles of the same species, where possible. Tadpoles of
Pseudis paradoxus, Hyla dominicensis, Hyla lichenata, Hyla septentriona1is
and Hyla vasta generally had much more extensive chlorosis of soft
tissues than did the adults of these species. Two adult specimens of
Hyla he?Iprini showed chlorosis as extensive as that of metamorphosing
tadpoles, but melanin pigments of the dorsum were better developed in
the adult than in the tadpoles. While scores of tadpoles of Hyla brunnea
invariably had green pigmentation, dozens of adults showed no bil ¡verdn
in tissues, including bone. Green pigment is absent from the tissues
of Hyla marianae, which are orange in color.
As indicated previously, there was individual variation of
green pigmentation within populations of Hyla septentr?ona1is. Small
species such as Hyla wiIderi have bones of a more intense green color
than do larger species such as Hyla maxima or Pseudis paradoxus. Calcifi
cation of bones tends to obscure the green pigment, as in the larger
species, but the color of biliverdin is quite apparent in poorly calcified
bones as in the distal limb structure of Hyla wilderi.


72
American Museum of Natural History where Noble deposited most of his
specimens). Dr. Lynn and Dr. Albert Schwartz (personal communications),
both of whom have collected in this area, are unaware of any tadpoles
of this species which have been collected. It seems best to consider
Noble's statement as a supposition on his part.
The color of the bones of this species has not been recorded.
Bufo mar inus. This very large toad was found breeding in
several habitats including temporary pools in Surinam and near drainage
canals and tropical fish hatcheries in Miami, Florida. it has been
introduced into Jamaica, where it appears to be the only anuran species
which breeds in ponds or slow-moving portions of streams. Neither young,
which are vegetarian, nor adults are known to have green tissues under
normal circumstances.
Leptodacty1 us albilabrus. -- Tadpoles of this Puerto Rican
species were collected from a roadside stream on El Yunke, the highest
mountain on the island. Water flowed rapidly in the stream, thus account
ing for its high oxygen concentration (Table 4). Both large and small
tadpoles of this species came to the surface, evidently to take air,
but their lungs do not appear to be developed at these stages. This
species feeds on vegetation and does not develop green pigmentation as
it undergoes metamorphosis.
Resume of Ecology and Breeding Habits
Considering the great variety of aquatic habitats available
in the Neotropical Region, it seems that anurans have taken advantage
of most of them. The species with green pigmentation are well


144
Barrio, A. 1965b. Las subespecies de Hyla pulchel la. Physis 2:
1 15-128.
Bialaszewicz, K. and R. Bledowski. 1915- The influence of fertili
zation on the respiration of eggs. Proceedings of Scientific
Society of Warsaw 8^: 429-473-
Boell, E. J. 1948. Biochemical differentiation during amphibian
development. Annals of the New York Academy of Sciences 49:
773-800.
Bokermann, W. C. A. 1964. Dos nuevas especies de Hyla de Minas Gerais y
notas sobre Hyla alvarengai Bok. Neotropica J_0: 67 (From
Barrio, 1965a)
. 1965- Field observations on the hylid frog Osteocepha1 us
taur ?nus Fitz. Herpetologi ca 20_: 252-255-
Bond, A. N. I960. An analysis of the response of salamander gills to
changes in the oxygen concentration of the medium. Develop
mental Biology 2_: 1-20.
Boulenger, G. A. 1880. (Note). Zoological Record 1880: 10-11.
. I883. Notes on little known species of frogs. Magazine of
Natural History 8^: 1-32.
Boyd, W. C. 1963. The lectins: their present status. Vox Sanguinis
N. S. 8: 1-32.
Brachet, J. 1935- Etude du metabolisme de l'oeuf de Grenoui 1 le (Rana
fusca) au cours du developpement. Archives de Biologie 46:
1-24.
Brattstrom, B. H. 1962. Thermal control of aggregation behavior in
tadpoles. Herpeto logi ca J_8: 38-46.
Broughton, P. M. G., E. J. R. Rossiter, C. B. M. Warren, G. Goulis, and
P. S. Lord. 1965- Effect of blue light on hyperbilirubinemia.
Archives of Diseases of Childhood 40: 666-671 -
Brown, G. W., Jr. 1964. The metabolism of Amphibia. In: Physiology of
the Amphibia, J. A. Moore, ed. New York: Academic Press,
pp. 1-98.
Cabello, Ruz, J. 1943- Bi 1 iverdinemia del sapo. Revista Sociedad
Argentina Biolgica 81.
Caglar, M. 1945- Biliverdin as a pigment in a fish. Nature 155: 67O.


Respiratory rates of Hvia brunnea tadpoles at different temperatures (Table 8)
Figure .


DISCUSSION
93
High Rate of Red Cell Hemolysis 93
Defective Transport of Bile Pigment into
the Li ver Cell
Defective Bile Pigment Conjugation 100
Disturbed Bile Pigment Excretion 100
CONCLUSIONS 104
APPENDICES 107
LITERATURE CITED 143
BIOGRAPHICAL SKETCH 154
v


TABLE 3 (Continued)
LOCATION AND HABITAT DATA
Streams of Haiti
30. Furcy (habitat of Hyla vasta
and Hyla he i Iprini)
31. Riviere Froid, 4.8 km W of Port-au-Prince
Streams of Jamaica habitat of Bufo marinus
32. Vicinity of Troja (two streams)
33- 4.8 km W of Bog Walk (irrigation ditch)
34. Vicinity of Kellitts (two streams)
Streams of Puerto Rico habitat of
Leptodacty1 us albilabrus
35 El Yunke roadside stream
DATE
ELEVATION
(Meters)
AIR
TEMPERATURE
10-12 V 1966
1350
20.5-29.5
2 VI11 1966
1350
25.4-26.0
10 v 1966
150
26.7
9 VI11 1966
450
28.7-30.0
9 VIII 1966
350
28.2
17 v 1966
700
24.0-25-3
6 V 1966
750
22.2


LITERATURE REVIEW
Naturally Occurring Green Pigments
Porphyrin compounds and their derivatives constitute the
vast majority of green pigments among living organisms. Among the best
known of these are the chlorophylls, green hemoglobins and green bile
pigments. Others include the chlorocruorin blood pigments of poly-
chaetes; turacoverdin from the feathers of plantain-eating touracos
(Fox, 1953); myeloperoxidase, the green enzyme of white blood cells
and certain tumors (Agner, 1941); and verdohemochromes, which are
intermediate compounds that appear during the degradation of heme com
pounds to bile pigments in vi tro (Lemberg and Legge, 19^+9)
Some tunicates have a pale green pigment which contains
vanadium. The structure of this pigment is not well known, but it may
be similar to some of the bile pigments. Green pigments of crustacean
hypodermis and eggs result from conjugation of a reddish carotene pig
ment with proteins (Fox, 1953)
Green pigments of frogs
In I85O, Kunde wrote the first paper concerning a green
pigment in frogs (see Lemberg and Legge,1949, p. 506). He observed
biliverdin in frogs' blood serum after liver extirpation. Lemberg (1935),
Nisimaru (1931), Cabello (1943), Rodriguez Garay, Noi r and Royer (1965),
and Barrio (1965a) indicated that amphibian bile contains biliverdin as
3


41
Dunn (1926) was unable to reconcile the reduced gill structure with
the low oxygen content of the environment. Lungs are seen on either
side of the vertebral column, and appear like small bubbles. Lungs
are present in very small tadpoles as well as older ones. They appear
to be used as accessory respiratory structures when branchial and cutaneous
respiration are insufficient. In this respect, the tadpoles would be
similar to the Austra1ian lungfish, which utilizes pulmonary respiration
largely at night when it becomes more active (Grigg, 1965)- While the
long, narrow tails of Hyla brunnea tadpoles probably are important
to cutaneous respiration as Laessle (1961) indicated, it seems likely
that their primary purpose is to propel the organisms through their
viscous environment.
The diet of Hyla brunnea tadpoles consists largely of frog
eggs, as noted by Dunn (1926) and Laessle (I96I). Most often the eggs
are probably those of its own species, since it is by far the most
common tree frog of Jamaica. in addition, it is the only hylid species
known from the eastern third of the island so that any hylid eggs found
in Hyla brunnea tadpoles there, would have to be of the same species
(Laessle, 1961).
Several structural characters of the tadpoles can be correlated
with their unusual diet. The digestive tract is expanded into a sac
in which eggs may be found more than a week after the last feeding. it
is relatively straight, not coiled as in vegetarian tadpoles. Since the
gills are not important in collection of food, their structure is


44
During my visits to Jamaica I heard Hy 1 a 1 ichenata as far
east as localities near Moneague, Ewarton and Lluidas Vale. These calls
could be heard for nearly half a mile and invariably came from wooded
slopes. This species did not venture out into flat, open valleys, as
did Hyla brunnea. In attempting to collect this species while it called
at night, 1 found the vocal individuals to be in trees, with one exception.
North of Mandeville, it seemed that one frog was cal ling from underground.
This observation, along with the presence of Dr. Farr's specimen near
a cave entrance, leads one to suspect that these large frogs may take
refuge in cave entrances and crevices in limestone, both of which are
present in quantity in the range of Hyla lichenata.
Dunn (1926) found four tadpoles of Hvla lichenata in a bromeliad
20 feet up in a small tree, in August, 1925- 1 collected two large
tadpoles of this species from a large bromeliad situated on the ground
under the shade of small trees. These were taken in Rose Valley on
October 25, 1965 These large specimens could be distinguished from
tadpoles of Hyla brunnea by their large size, relatively shorter and
more muscular tail, and slightly different mouth structure. They were
only slightly darker than Hyla brunnea tadpoles, not black as were those
collected by Dunn. Like other Jamaican hylid tadpoles, they feed upon
frog eggs. One of my two specimens underwent metamorphosis and died about
six days later. At the time of its death, it weighed 660 milligrams and
had a snout-vent length of 18.7 millimeters. This individual developed
green pigmentation, similar to that of Hyla brunnea, in its bones and
other tissues.


133
Optica1
Density
Figure 3
Optica 1
Density
Figure 4. Absorption spectrum of coelomic fluid (bile) from
Osteocephalus taurinus.


88
Phrvnohvas venulosa -- red cell cytoplasm coagulated.
Phyllomedusa bicolor -- appears to have many leukocytes.
Phyllomedusa hypochondr1 a 1 is -- erythrocytes are basophilic.
Sphaenorhynchus aurantiacus -- cytoplasm of erythrocytes
appears vacuolated, with greenish refractile material in
vacuoles (stained slide). Blood of three males hemolyzed
overnight.
Leptodacty1idae
Leptodacty1 us pentadacty1 us some red cel Is show coagulated
cytoplasm.
Effect of temperature upon frog red cells. -- Slight changes in
a few cells were noted after ten minutes at 45 C. After four hours at
45 C., cytoplasm had coagulated in nearly all of the red cells of Hyla
squirre 1 la and Bufo terrestris (Figures 9-l0 ; the red cel Is of Scaphiopus
holbrooki were abnormally shaped after this time, but the cytoplasm had
not coagulated.
Respiratory rates of tadpoles. -- The respiratory rates of tad
poles and froglets of green boned species (Tables 8-10) are similar to
those previously recorded for amphibian larvae (Hopkins and Handford,
1943). The basis of recording respiratory rate in the study was the uptake
of microliters of oxygen per gram of wet weight of tadpole per hour. In
order to convert this to an approximate dry weight basis, the rate of
oxygen uptake should be multiplied by eight. Size range of the tadpoles
in the present study is much greater than in previous ones, but it can be
seen that respiratory rate is a function of tadpole size and that the
species is not very important in this respect (Figure 8). As is to be
expected, respiratory rates increase with increasing temperature (Figure 7)-


Figure 11. Red blood cells of Hvla squirelia after four hours at 45 C.
Note coagulated cytoplasm and tendency of cells to stick
to each other. (400X)
Figure 12. Untreated blood cells of Hyla leucophy1 lata. Note numerous
erythrocytes with coagulated cytoplasm. (400X)


105
8. In different taxa, the presence of green pigment may vary
at familial, generic, specific, subspecific or individual levels.
9. Frogs and tadpoles with green pigmentation have been
collected from habitats which differ widely in concentrations of oxygen,
dissolved salts, iron and phosphate, in addition to temperature, insol
ation, altitude and presence or absence of water currents. These
organisms also differ widely in their food, feeding habits and methods
of gaseous exchange.
10. The actual cause of the presence of high concentrations
of biliverdin is not known. However, it appears that the source of the
pigment is hemoglobin from red blood cells which are hemolyzed.
11. It seems likely that a combination of high temperatures
and fragile erythrocyte membranes may cause hemolysis in some breeding
adults. Hormonal changes at the end of the breeding season and during
metamorphosis may also account for some types of hemolysis.
12. While hemolysis apparently is the source of pigment, it
is unlikely that this alone could cause the green coloration of tissues.
Since hemolysis probably occurs in these species during metamorphosis and
at the end of the breeding season, both of which are times when the
liver has great stress from other activities, it appears that the liver
cells are unable to handle the increased bile pigment at this time.
Since the biliverdin appears to be relatively non-toxic, its elimination
presumably has a low priority.
13. Some species evidently have taken evolutionary advantage
of the green bile pigment accumulation by using it as protective coloration.


140
Figure 15. A stained smear of blood from Hyla punctata, a species with
green tissues. Note that the majority of cells are immature,
as indicated by their irregular shaped and large nuclei. (400X)
Figure 16. Stained smear of untreated Phyllomedusa bicolor blood.
Cells were basophilic and perhaps immature in this frog
which lacked green pigment. (400X)


101
which showed development of fatty areas, which can be caused by extended
use of steroid hormones or deficient diet.
Destruction of liver cells by chemicals such as carbon tetra
chloride or situations accompanying acute virus hepatitis may result in
intrahepatic obstruction (Williams, 1965)- Tannic acid is a chemical
which causes damage to the liver in mammals (Arhelger, Broom and Boler,
1965) and may be a cause of liver injury to tadpoles which live in waters
with significant concentration of this chemical.
Allen, Carstens and Olson (1967) studied veno-occ1 us ive disease
in monkeys. A single dose of monocrota1ine causes necrosis of liver cells
and partial blockage of the veins within the liver. This drug is believed
to be the ingredient of bush teas which cause veno-occ1 usive disease in
Jamaica. Venous occlusion was observed in Hyla geoqraphica in Surinam.
It may have been induced by a plant extract.
Another factor which has an effect upon bile flow is the con
centration of environmental oxygen. Anesthetized dogs which breathed air
with 15 per cent oxygen for thirty to forty-five minutes had the bile
flow reduced by 16 to 50 per cent. There was also a reduction of urine
formation (Schnedorf and Orr, 19^1). The tadpoles studied lived under
conditions of low oxygen tension, but all appeared to utilize atmospheric
oxygen when placed under stress. However, the conditions of low oxygen
tension and high temperature to which Hyla dominicensis tadpoles were
subjected, would appear to push the limit in this direction. Also, a high
respiratory rate, such as appears to be present in tadpoles of Hyla hei lprini
may result in a relative oxygen lack that would reduce bile flow and thus


70
mountains. Although he did not capture that individual, he obtained
tadpoles nearby. In cold streams above 2500 meters in the central
regions he heard again the loud call of Hyla hei1prini above the roar
of the cascades. While the average temperature in the first stream was
24.7 C., it was 12.6 C. in the second locality at the higher elevation.
Noble found the tadpole of Hyla heilprini to be more streamlined than
the previous two species and considered this an adaptation to living in
torrents, rather than in slower moving streams or puddles. Since these
larvae were obviously adapted to an environment with a high oxygen con
centration, Noble had doubts about being able to maintain them in captivi.ty,
but he was successful in raising them through metamorphosis.
From Noble's (1923) observations in the Dominican Republic,
specimens from the Museum of Comparative Zoology from several localities
along the southern peninsula of Haiti, and personal observations, it
appears that Hyla heilprini is widespread on Hispaniola. I believe that
1 heard the species just west of Port-au-Prince in a ravine near the
Riviere Froid, less than 200 meters above sea level. It would seem that
wooded ravines are more important to this species than are high elevations.
Cascading streams probably are necessary for its reproduction, as both
Noble (1923) and I found its tadpoles only near waterfalls.
Green pigmentation was present in all of the dozen or more
tadpoles which I collected. These were we 11-developed tadpoles, with
either two or four legs, which were olive green with black spots dorsally,
blue-green and white ventrally, and the distal portions of the limbs were
orange in color. Since only the distal portions of the limbs were orange,


INTRODUCTION
A wide variety of colors is demonstrated by anurans, particu
larly those of tropical areas. Frogs and toads which spend most of
their time on tree trunks, among the dead leaves of the forest floor,
about rocks, or underground, are usually brown, gray or black due to
the presence of melanin in me 1anophores. Red or yellow colors are usu
ally due to lipid pigments in the lipophores, and white results from
reflection of visible light by guanine crystals of guanophores. Green
hues are often seen among anurans, especially in the skin of those close
ly associated with green vegetation.
Studies have shown that the green skin color of many frogs in
life is the result of combined effects of three types of chromatophores
located in the upper portion of the dermis. Lipophores containing
yellow pigment are found just under the basement membrane of the epider
mis, in close association with the guanophores immediately below them.
Melanophores are found below the guanophores, but have processes extend
ing into the guanophore layer. Visible light is diffracted as it
impinges upon the guanine crystals of the guanophores; the blue and
green portions of the spectrum tend to be reflected, while light of
longer wave lengths is absorbed by the melanophores. As the light of
shorter wave lengths passes back through the lipophores, the blue portion
is absorbed and only green is reflected from the skin. Modification of
lipophores following fixation of the frog causes the skin to appear
1


20
for hemoglobin concentrations of amphibian blood. These range from
8.7 to 11.5 grams of hemoglobin per 100 milliliters of blood.
In regard to shape, it should be noted that the oval disc
form of amphibian erythrocytes presents a greater surface area per unit
of volume than would a sphere of like volume, thus presenting a greater
surface for gaseous diffusion (Andrew, 1965). Greatest efficiency of
oxygen transport among amphibians appears in the genus Batrachoseps,
which may have up to 90 per cent erythroplastids or enucleate
erythrocytes (Emmel, 1924). Lack of a nucleus makes the cells flatter
and thus increases cell count per unit volume, as well as reducing the
oxygen uptake of the red cel 1 by the amount consumed by the nucleus
(Foxon, 1964).
Transfer of oxygen to respiring tissue
Once oxyhemoglobin is carried to the tissues from the respira
tory organs, oxygen dissociates from it and diffuses into the cells.
Krogh (1941) has shown that the rate of oxygen diffusion in tissues is
variable, but less than half that of water. Less is known of conditions
which promote dissociation of oxyhemoglobin of the type found in tadpoles.
Since the great affinity which tadpole hemoglobin has for oxygen is not
modified by changing pH, dissociation would appear more difficult than
it is for oxyhemoglobin of adult erythrocytes (Manwell, 1966).
Oxygen utilization and metabolic rate
Boell (1948) has reviewed earlier work which related to oxygen
consumption of early amphibian embryos. Much of this was concerned with
respiration of various regions of the gastrula stage. The work of


TABLE 3 (Continued)
LOCATION AND HABITAT DATA
Open pool of Florida, U.S.A. habitat of
Hyla septentrionai?s
21. SW Miami (pool of Serpentarium)
Pool of Trinidad habitat of Phy1lomedusa
bicolor tadpoles
22. Simla Biological Station
Ponds and standing waters of Surinam
23. Albina Road at Commewijne River (road
side ditch)
24. Domberg (roadside ditch)
25. 28 km S of Paramaribo (ditch)
26. Overtoom Road (pond and pool)
27. Onverwacht (flood meadow)
Streams of Surinam
28. Brownsberg Princess Irene Falls
(habitat of Hyla maxima)
29. Base of Brownsberg Brown's Kreek
DATE
ELEVATION AIR
(Meters) TEMPERATURE
25 VI 1966
10 31.2-32.2
29 VI I 1966
30.2
13
VI 1
1966
10
29.
0
17
VI 1
1966
5
32.
,8
17
VI 1
1966
60
27
5
25
VI 1
1966
60
29.
.0-29.3
25
VI 1
1966
60
27
.0-30.2
20
VI 1
1966
350
24.7-25.1
21
VI 1
1966
350
24.5-25.0
21
VI 1
1966
200
27.8


6
billverdIn or other hematin derivatives as the green pigment. Lonnberg
(1934) noted green color in tissues of the African lungfish, Protopterus
annectens. Lemberg and Legge (1949, P- 506, pp. 569570) listed other
vertebrate structures and invertebrates which contain biliverdin.
Identification of the green pigment
Barrio (1965a) has amply demonstrated by chemical and
spectrophotometric means that biliverdin is primarily responsible for
the green color of tissues in the frogs which he studied. Through
electrophoretic fractionation of green serum, he found that most of the
biliverdin was combined with globulin proteins and a lesser amount was
associated with albumin. Lester and Schmid (1961) have shown that
adult anurans have an enzyme for conjugating bile pigments and Rodriguez
Garay, Noir and Royer (1965) have found small quantities of water soluble
biliverdin glucuronide in Bufo arenarum. Barrio (1965a) considered
the biliverdin to be unconjugated, since it migrated at the same rate
as free biliverdin on paper chromatographs, was completely soluble in
chloroform, and insoluble in water.
Characteristics of Biliverdin
Chemical properties and reactions
Biliverdin (C33H34O6N4) is composed of four pyrrole groups
linked together by three methyne bridges. Its structural formula may
be drawn in the form of a chain, but is more precisely represented by
the incomplete ring form. There is an obvious similarity of biliverdin
to two of its precursors, protoporphyrin IXa and heme of hemoglobin.
It differs from them in that its ring structure is incomplete; also,


61
Thus, the Cuban tree frog has become established in a number
of localities through the activity of man, and its original range
outside of Cuba would be difficult to determine. Part of the adapt
ability of this species may be related to its accommodation to human
habitations; it is also found in mesophytic situations, but not in pine-
land or prairie habitats (Duellman and Schwartz, 1958; Barbour and Ramsden,
1919; Grant, 1940; Stejneger, 1905; Peterson, Garrell and Lantz, 1952;
Neill, 1958).
During the present study, habitats of the Cuban tree frog
were studied on Grand Cayman, on New Providence Island in the Bahamas,
in the vicinity of Miami, Florida, and near Highgate, St. Mary's Parish,
Jamaica. Cuba was not visited due to political restrictions. The two
days' stay on New Providence, during May, 1966, was unproductive due to
drought and a widespread lack of surface water.
In Miami, Florida, tadpoles of Hyla septentriona1?s were found
at a tropical fish hatchery and at the Serpentariurn, a commercial enter
prise concerned largely with snakes. Large larvae undergoing metamorphosis
were noted in concrete tanks in both places. In addition, thousands of
small tadpoles were found in the long, shallow, ornamental pool in front
of the Serpentarium (Table 3, Habitat 21). Since this pool had been
drained and cleaned just one week before, the tadpoles had hatched and
grown to 5 millimeters in body length in less than a week. Although the
water temperature of this pool was 33-9 C., the oxygen concentration was
9-2 parts per million. The pool had been drained to remove large amounts
of algae, but the water appeared clear at the time of measurement; also,


21
Brachet (1935) upon eggs of Rana fusca (= Rana temporaria) showed no
significant difference between respiratory rates of fertilized and
unferti1ized eggs.
Among the first studies of oxygen uptake in anuran larvae
were those of Helff (1926) and Etkin (1934), working upon Rana pipiens
and Rana catesbeiana respectively. Atlas (1938) studied Rana pipiens
and Rana sylvatica from fertilization to tadpole states and found that
respiratory rates increase from fertilization to overgrowth of the
operculum. This was in agreement with the studies of Godlewski (1900)
and Bialaszewicz and Bledowski (1915) on Rana temporaria. Pamas and
Krasinska (1921) found the rate of oxygen consumption rose sharply at
the onset of gastrulation, neurulation, and external gill formation,
with no noticeable changes between. Atlas (1938) suggested that the
latter conclusions were based upon insufficient data which wereadversely
affected by temperature fluctuation. it is interesting to note that
the respiratory rates obtained by Bialaszewicz and Bledowski for Rana
temporaria are similar to Atlas' data for the closely related Rana
sy1 vatica, but somewhat different for Rana pipiens or Rana catesbeiana.
Atlas concluded that the increased rate of respiration per embryo was
a function of cell number rather than of intensity of respiration of
individual cells. The basis for this statement is obscure, since no
cell counts were reported, nor was the subject discussed In the main
text of the paper.


53
tadpoles approximates that of the adults. Mr. Walter Polder (personal
communication) observed that Pseudis tadpoles grow to this large size
during the course of a single rainy period of four to six months. Thus,
the growth rate of this species is phenomenally high. in comparison
with other species, the adults of which are the size of adult Pseud ? s,
metamorphosis of this tadpole is greatly delayed. The slow, deliberate
movements of this tadpole reflect its low metabolic rate (Table 9)
Although typical habitats (such as Habitats 23 and 24 of Tables 3 and 4)
of this species were visited during the day, neither tadpoles nor frogs
were seen. Both forms appeared to be more active at night, when they
could be seen floating nearly vertically in the water. Its dark pigment
ation, low metabolic rate, and perhaps decreased diurnal activity are
probably important adaptations of this tadpole to its exposed and warm
environment. The adult can swim and hop quite rapidly. Thousands of
individuals were seen hopping across the road on a rainy night toward
the end of the wet season (late July).
Both the tadpoles and the adults of this species have green
pigmentation in the tissues, but it is particularly pronounced in the
tadpoles. In the larger tadpoles, green pigment permeates the tissues
so that they appear almost as dark internally as externally. Development
of green pigmentation in this species does not appear to be related to
metamorphosis, since it is well developed before that time. Green
pigmentation in the adults appears limited to the we 11-caleified bones.
Their other internal tissues are usually light in color and the blood
is bright red. Thus, there is a marked difference between adult and tad
pole with regard to green pigmentation.


TABLE 4 (Continued)
LOCATION, HABITAT ANO DATE
Actual habitats of Hyla septentriona 1 ?s
15. South Sound (29 X)
17. 8 km E of Georgetown (29 X)
19- North Side (30 X)
20. North Side (30 X)
Open pool of Florida, U.S.A. habitat
of Hyla septentriona 1 is
21. SW Miami (25 VI)
Pool of Trinidad habitat of
Phyllomedusa bicolor
22. Simla Biological Station
Ponds and standing waters of Surinam
23. Albina Road at Commewijne River
(13 VII)
24. Oomberg (1J VII)
25. 28 km S of Paramaribo (17 VII) -
habitat of Phrynohyas venulosa
26.Overtoom Road (25 VII)
TEMPERATURE
OXYGEN(ppm)
HARDNESS(ppm)
E
29.0-34.0(4)
-
106-134
7-5
25
-
60
7-5
27
-
64
7.0
26
-
70
7-0
32.9-33.9(2)
8.1-9-2(2)
-
-
28.2
10. 1
-
-
27.8
3-4
-
-
33.0
-
-
-
25.0-29.4(2)
0.7
-
-
29.9-30.8(2)
6.5-7-1(2)
-
-


149
Laessle, A. M. I96I. A mi ero-limnologica1 study of Jamaican bromeliads.
Ecology 4_: 499-517-
Lemberg, R. 1935- Transformation of haemins into bile pigments. Bio
chemical Journal 29: 1322-1336.
Lemberg, R. and J. W. Legge. 19^9- Hematin Compounds and Bile Pigments.
New York: Interscience Publishers.
Lester, R. and R. Schmid. 1961. Bile pigment excretion in Amphibia.
Nature 190: 452.
Lester, R. et_ aj_. 1966. Biosynthesis of tritiated bilirubin and studies
of its excretion in the rat. Journal of Laboratory and Clinical
Medicine 61000-1012.
Lonnberg, E. 193^+ Den grona fargen hos skelettet av lungfisken,
Protopterus annectens. Fauna och Flora 29: 278-279-
Lutz, A. 1924. Sur les rainettes des environs de Rio de Janeiro.
Comptes Rendus Biologique Paris 90: 241.
Lutz, A. and B. Lutz. 1938. Two new Hylidae: H_. a 1 bosignata n. sp. and
H_. pickeli n. sp. Annaes Academia Bras i 1 ¡ero Sci enti fi cas J_0: 185
Lutz, B. 1947. Trends towards non-aquatic and direct development in
frogs. Copeia 1947: 242-252.
. 1948. Anfibios Anuros da Colecao Adolpho Lutz. II. Especies
verdes do genero Hyla do Leste-Meridional do Brasil. Memorias
do Instituto Oswaldo Cruz 52_: 155-238.
. 195^. Anfibios Anuros do Distrito Federal. Memorias do Instituto
Oswaldo Cruz 2: 155-238.
Lynch, J. D. and H. L. Freeman. 1966. Systematic status of a South
American frog, Allophryne ruthveni Gaige. University of Kansas
Publications, Museum of Natural History J493-502.
Lynn, W. G. 19^0. The Herpetology of Jamaica I. The Amphibians.
Bulletin of Institute of Jamaica, Science Series (1): 1-60.
. 1958. Some amphibians from Haiti and a new subspecies of Eleuthero
dacty 1 us schmidti Herpetologi ca J_4: 153-157-
. 1961. Types of amphibian metamorphosis. American Zoologist Jk
151-161.
Lynn, W. G. and J. N. Dent. 1943- Notes on Jamaican amphibians. Copeia
1943: 234-242.


57
Jamaican tree frogs, and to Hyia septentriona1is, the only tree frog of
Cuba and surrounding islands. in the Dominican Republic, Mertens
(1939) found Hyla dominicensis in all but the most arid habitats, from
near sea level to more than 1000 meters in elevation. Barbour (191
noted that the Museum of Comparative Zoology has many specimens from all
parts of the island and Cochran (19^1) noted that it is common in
co1lections.
During October, 1965, I collected an adult and tadpoles of
this species from Port-au-Prince and heard adults calling from trees in
Petionville (400 meters elevation) and Cap Haitien. Adults were heard
in May and August of 1966, most often during or after rain. Tadpoles
were also collected during these visits. The first large chorus of this
species which Mertens (1939) heard was on the evening of February 21 after
the first hard rain. He noted breeding activity until September 29 and
concluded that there is no definite breeding period. While Lynn (1958)
collected adults and tadpoles of Hyla dominicensis in Haiti during mid-
April, 1953j he did not hear them calling. He also noted that this frog
produces a skin secretion which is irritating to cuts and scratches, and
that the bones are green.
Mertens (1939) took note of the very rapid development of Hyla
dom?n?censis tadpoles. He collected a number of these from a street
puddle filled with rust-colored water on March 3- He stated that the
larger tadpoles had a covering of grass-green algae on the back and only
small hind limb buds, but developed into frogs in 10 days. Similarly,
larvae that he collected from a cistern on February 21 had only rudimentary


TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS 11
LIST OF TABLES v!
LIST OF FIGURES vi i
INTRODUCTION 1
LITERATURE REVIEW 3
Naturally Occurring Green Pigments 3
Characteristics of Biliverdin 6
Sources of Biliverdin 9
Causes of Chlorosis 12
Respiratory Studies 13
Relation of Respiration to External
Environment 22
Structure and Function of Frog Liver 23
SOURCES AND CARE OF TADPOLES AND FROGS 24
I. CAUSATIVE PIGMENT AND INCIDENCE OF CHLOROSIS 27
Pigment Identification 27
II. CONSIDERATION OF ECOLOGICAL FACTORS ASSOCIATED
WITH CHLOROSIS 35
Methods of Study 35
Observation of Frog Habitats and Behavior 36
Resume of Ecology and Breeding Habits. 72
III. PHYSIOLOGICAL FACTORS RELATED TO CHLOROSIS 74
Attempts to Reproduce Green Pigmentation
In Vivo 75
Physiological Characteristics of Chlorotic
and Non-Ch loroti c Frogs 8l
Disease as a Cause of Chlorosis in Frogs 91
i v


TABLE 4 (Continued)
LOCATION, HABITAT AND DATE
27. Onverwacht habitat of Hy1 a crepitans ,
Hyla punctata, Hyla minuta and Hyla
boesemani (25 VI I-5PM)
(25 VII-7:40PM)
Streams of Surinam
28. Princess Irene Falls habitat of
Hyla maxima (20-21 VII)
Rock basin (21 VII)
29. Brown's Kreek habitat of fish
(21 VII)
Streams of Haiti
30. Furcy habitat of Hyla vasta and
Hyla heiIprini (10-12 V)
(2 VIII)
31. Riviere Froid (10 V)
Streams of Jamaica
32. Troja habitat of Bufo marinus
(9 VIII)
TEMPERATURE
OXYGEN(ppm)
HARDNESS(ppm)
£H
30.9
3.4
-
-
27-5
0.6
-
-
23.4-25.5(6)
7.8-8.6(6)
-
-
23.8-29.9
2.0-3.1
-
-
25-2
4.9
-
-
19-5-24.5(14)
7.4-9-5(14)
40-90(3)
6-7(2)
20.1-25.2(5)
7-3-9.9(5)
60-90(5)
7
25-8
7-9
130
27.0-27.1(2)
7-9-8.0(2)


GREEN PIGMENTATION
IN NEOTROPICAL FROGS
By
DUVALL ALBERT JONES
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
December, 1967

ACKNOWLEDGMENTS
I wish to thank Dr. Coleman J. Goin for suggesting the
challenging topic of this paper and for his continued interest and
support during the various phases of its study. Dr. Goin has aided in
the identification of frogs used in this study and has extended many
other courtesies which are deeply appreciated. I am grateful to Dr.
Mildred Griffith, Dr. Frank J. S. Maturo and Dr. James Nation for their
aid during this study and for suggestions which improved the dissertation.
To Olive B. Goin, I owe thanks for many kind acts and stimulating
discussions.
Several individuals have made laboratory and field facilities
available to me. Among these were: Dr. James Nation, Dr. Ward Noyes,
Dr. Albert Laessle, Dr. Frank Nordlie, Dr. Paul Elliott, and Dr. James
Gregg, all of the University of Florida; Dr. Harold Heatwole, University
of Puerto Rico (El Yunke Biological Station); Mr. Bernard Lewis and
Dr. Thomas Farr, Institute of Jamaica; Dr. Ivan Goodbody, University of
West Indies, Kingston, Jamaica; Dr. Jocelyn Crane, Simla Biological
Station, Trinidad; and Mr. Walter Polder, Paramaribo, Surinam.
To the following persons, I am grateful for the opportunity
to study the specimens of their respective institutions: Dr. Ernest
Williams, Museum of Comparative Zoology, Harvard University; Mr. C. M.
Bogert and Dr. Richard Zweifel, American Museum of Natural History;

Dr. James Bohlke, Philadelphia Academy of Natural Sciences; Dr. Doris
Cochran and Dr. James Peters, United States National Museum; Mr. Neil
Richmond and Dr. Clarence McCoy, Carnegie Museum; Dr. Charles F. Walker,
University of Michigan Museum of Zoology; Dr. Robert Inger, Chicago
Natural History Museum; Dr. Walter Auffenberg, Florida State Museum;
Mr. Bernard Lewis and Dr. Thomas Farr, Institute of Jamaica.
1 am grateful to Dr. Thomas Goreau, University of the West Indies,
Dr. Margaret Stewart, State University of New York at Albany, and others
for enlightening discussions. Bonneye Greene, Thomas Quarles, Robert
MacFarlane and John Anderson kindly rendered technical assistance.
Forest service personnel in Jamaica and Surinam made it possible for
me to visit several remote localities. I appreciate the aid rendered
by members of the U. S. Consular Service. Grants for travel expenses
were received from the University of Florida Graduate School, the Sigma
Xi- RESA Fund and the National Science Foundation (GB-3644, to Dr.
Coleman J. Goin).
The services of Rhea Warren, Stephen Bass, Murray de la Fuente
and others who aided in the collection of specimens are appreciated
greatly. Janet Bungay and Geraldine Lennon aided in the preparation of
the manuscript. My wife, Dorothy, assisted in the histological portion
of the study, in preparation of the manuscript, and in giving moral
support.
i i i

TABLE OF CONTENTS
Page
ACKNOWLEDGMENTS 11
LIST OF TABLES v!
LIST OF FIGURES vi i
INTRODUCTION 1
LITERATURE REVIEW 3
Naturally Occurring Green Pigments 3
Characteristics of Biliverdin 6
Sources of Biliverdin 9
Causes of Chlorosis 12
Respiratory Studies 13
Relation of Respiration to External
Environment 22
Structure and Function of Frog Liver 23
SOURCES AND CARE OF TADPOLES AND FROGS 24
I. CAUSATIVE PIGMENT AND INCIDENCE OF CHLOROSIS 27
Pigment Identification 27
II. CONSIDERATION OF ECOLOGICAL FACTORS ASSOCIATED
WITH CHLOROSIS 35
Methods of Study 35
Observation of Frog Habitats and Behavior 36
Resume of Ecology and Breeding Habits. 72
III. PHYSIOLOGICAL FACTORS RELATED TO CHLOROSIS 74
Attempts to Reproduce Green Pigmentation
In Vivo 75
Physiological Characteristics of Chlorotic
and Non-Ch loroti c Frogs 8l
Disease as a Cause of Chlorosis in Frogs 91
i v

DISCUSSION
93
High Rate of Red Cell Hemolysis 93
Defective Transport of Bile Pigment into
the Li ver Cell
Defective Bile Pigment Conjugation 100
Disturbed Bile Pigment Excretion 100
CONCLUSIONS 104
APPENDICES 107
LITERATURE CITED 143
BIOGRAPHICAL SKETCH 154
v

LIST OF TABLES
Table Page
1 Occurrence of Green Color in Tissues and
Fluids of Adult Anurans 108
2 Occurrence of Green Pigment Among Metamor
phosing Neotropical Anuran Tadpoles 113
3 Locality Data of Tadpole Habitats 114
4 Water Characteristics of Tadpole Habitats 118
5 Iron and Phosphate Concentrations (Parts
Per Million) in Tadpole Habitats 123
6. Blood Characteristics 124
7 Average Dimensions of Frog Red Blood Cells 127
8 Respiratory Rates of Tadpoles in Warburg
Respirometers at Different Temperatures 128
9 Tadpole Respiratory Rates at 25 C.
(Measured by Jones Method) 129
10Frog let Respiratory Rates Measured In
Warburg Respi rorneters 131

LIST OF FIGURES
Figure Page
1 Absorption Spectrum of Biliverdin in
5 Per Cent Hydrochloric Acid Methanol
Solution 132
2 Absorption Spectrum of Liver Extract
From Hyla septentriona1is in Aqueous
Solution 132
3 Absorption Spectrum of Lymph of
Osteocepha 1 us taurinus 133
4 Absorption Spectrum of Coelomic Fluid
(bi le) From Osteocepha 1 us taurinus 133
5 Absorption Spectra of Bile Solutions of
Siren lacertina 134
6 Absorption Spectrum of Pigment Extracted
From Hyla marianae in 70 Per Cent Ethanol 134
7 Respiratory Rates of Hvla brunnea Tadpoles
at Different Temperatures 135
8 Respiratory Rates of Tadpoles at 25 C 136
9 Normal Red Blood Cells of Bufo terrestrls
Four Hours After Collection 137
10 Red Blood Cells of the Same Individual as
Above, After Four Hours at 45 C 137
11 Red Blood Cells of Hvla squire 1 la After
Four Hours at 45 C 138
12 Untreated Blood Cells of Hvla leucophvl lata.. 13 8
13 Stained Smear of Untreated Blood of Bufo
tvphoni us 139
vi i

I
Figure Page
14 Stained Smear of Untreated Blood of
Phrynohyas venulosa 139
15 A Stained Smear of Blood from Hv la
punctata, a Species With Green Tissues 140
16 Stained Smear of Untreated Phy1lomedusa
bicolor Blood 140
17 Stained Section of Apparently Normal
Liver of Hyla rubra 141
18 Stained Section of Hyla punctata Liver 141
19 Stained Section of Liver From Adult Male
Pseud i s paradoxus 142
20 Stained Section of Liver From Adult Male
Hyla maxima 142
vi i i

INTRODUCTION
A wide variety of colors is demonstrated by anurans, particu
larly those of tropical areas. Frogs and toads which spend most of
their time on tree trunks, among the dead leaves of the forest floor,
about rocks, or underground, are usually brown, gray or black due to
the presence of melanin in me 1anophores. Red or yellow colors are usu
ally due to lipid pigments in the lipophores, and white results from
reflection of visible light by guanine crystals of guanophores. Green
hues are often seen among anurans, especially in the skin of those close
ly associated with green vegetation.
Studies have shown that the green skin color of many frogs in
life is the result of combined effects of three types of chromatophores
located in the upper portion of the dermis. Lipophores containing
yellow pigment are found just under the basement membrane of the epider
mis, in close association with the guanophores immediately below them.
Melanophores are found below the guanophores, but have processes extend
ing into the guanophore layer. Visible light is diffracted as it
impinges upon the guanine crystals of the guanophores; the blue and
green portions of the spectrum tend to be reflected, while light of
longer wave lengths is absorbed by the melanophores. As the light of
shorter wave lengths passes back through the lipophores, the blue portion
is absorbed and only green is reflected from the skin. Modification of
lipophores following fixation of the frog causes the skin to appear
1

2
blue. Thus, the green color of frog skin of this type results from
the physical arrangement of three pigments, none of which is green
(Schmidt, 1919, 1920, 1921; Kawaguti, Kamishima and Sato, 1965; Elias,
1943) .
Though most frogs with green skin appear to fit the pattern
described, some tree frogs do not. Notable are some of the small
hylids such as Hyla wilderi. This species has few melanophores and
appears to be a diffuse green color throughout the body. When preserved
in alcohol, a green pigment is extracted and the frog becomes whitish
in color. In this species, as well as in others, there is no blue color
which supplants the green upon preservation. Indications are that
structural effects are not involved and that this type of green color
in frogs is due solely to a green pigment or pigments. Reports of
internal green pigment substantiate this idea.
This paper is concerned with three facets of the biology of
green neotropical frogs: (l) identification of the green pigment or
pigments; (2) ecology; and (3) physiology of green frogs, with emphasis
upon possible factors influencing pigment accumulation. Since these
three areas require different methods of study and yield various types
of information, they will be presented in separate sections.

LITERATURE REVIEW
Naturally Occurring Green Pigments
Porphyrin compounds and their derivatives constitute the
vast majority of green pigments among living organisms. Among the best
known of these are the chlorophylls, green hemoglobins and green bile
pigments. Others include the chlorocruorin blood pigments of poly-
chaetes; turacoverdin from the feathers of plantain-eating touracos
(Fox, 1953); myeloperoxidase, the green enzyme of white blood cells
and certain tumors (Agner, 1941); and verdohemochromes, which are
intermediate compounds that appear during the degradation of heme com
pounds to bile pigments in vi tro (Lemberg and Legge, 19^+9)
Some tunicates have a pale green pigment which contains
vanadium. The structure of this pigment is not well known, but it may
be similar to some of the bile pigments. Green pigments of crustacean
hypodermis and eggs result from conjugation of a reddish carotene pig
ment with proteins (Fox, 1953)
Green pigments of frogs
In I85O, Kunde wrote the first paper concerning a green
pigment in frogs (see Lemberg and Legge,1949, p. 506). He observed
biliverdin in frogs' blood serum after liver extirpation. Lemberg (1935),
Nisimaru (1931), Cabello (1943), Rodriguez Garay, Noi r and Royer (1965),
and Barrio (1965a) indicated that amphibian bile contains biliverdin as
3

4
its bile pigment. Lester and Schmid (1961) found biliverdin in some
samples of frog bile, but believed bilirubin to be the main bile pig
ment of anurans. A claim was made by von Recklinghausen (1883) that
formation of biliverdin takes place in sterile frog blood. Rich (1925)
indicated that this had not been confirmed. Cabello (1943) caused the
formation of green serum in Bufo arenarum by administration of phenyl-
hydrazine and ligation of the bile duct. Rodriguez Garay et_ aj_. (1965)
found biliverdin glucuronide in the bile of Bufo arenarum.
The first report of externally visible green pigment in
frogs appears to be the one by Peters (1873), who indicated that the
skeleton of Pseudis mlnuta was green. Camerano (1879) believed the
green bones of Pseudis paradoxus to be due to the presence of ferrous
phosphate. The recorder of the Zoological Record (Boulenger, 1880)
took exception to Peters' paper by stating that Pseudis minuta does
not have green bones. However, Boulenger (1883) took note of the green
eggs of pseudids. Fernandez and Fernandez (1921), Miranda-Ribeiro (1926),
Parker (1935), and Gallardo (1961) also noted green eggs or tissues in
this family.
In 1910, Podiapolsky described a "chlorophyll" pigment from
Hyla arbrea and Rana esculenta, two European species. It may be noted
here that reports by Schmidt (1919-1921) upon these two species, and by
Kawaguti, Kamishima and Sato (1965) upon Hyla arbrea, do not support
this. More recently, Dunn (1926), A. Lutz (1924, 1938), B. Lutz
(1948, 1954), Cochran (1955), Lynn (1958), and Bokermann (1964) have
noted green coloration of epithelia, muscles and bones in certain

5
hylid species. Dunn (1931), B. Lutz (1947), and Savage (19&7) noted
the presence of green bones in certain species of centrolenids.
Barrio (1965a), working with frogs from Argentina, found
that five of twelve hylid species and all three of the pseudid species
which he studied had green tissues. Individuals of twenty-two species
of four other families lacked the green pigment in their tissues.
The species that had green tissues were similar in that the muscular
and subcutaneous tissue, lymph, walls of the digestive tract, eggs of
mature females, and especially the bones were strongly pigmented. He
attributed this situation to high concentrations of biliverdin in the
blood. The highest concentration, 11.9 mg/100ml, was found in the serum
of Hyla punctata. In addition, he pointed out that biliverdin Is
decomposed by formalin. This point clarified an earlier controversy
(Boulenger, 1880) and also indicated a reason for so little attention
being given to this phenomenon.
Another paper by Barrio (I965b) brought out the fact that
populations of Hyla pul che 1 la differ greatly in their serum concentra
tions of biliverdin; two of the five subspecies lack the pigment
a 1 together.
Green tissues of other animals
Green tissues are not limited to frogs. Peters (1873) noted
that the green bones of Pseudis minuta were similar in color to those
of the marine garfish Be 1 one. The relatively recent papers of Wagenaar
(1939), Fontaine (1941a, 194lb), Willstaedt (1941), Caglar (1945), and
Fox (1953) concerning green bones of teleost fish pointed toward

6
billverdIn or other hematin derivatives as the green pigment. Lonnberg
(1934) noted green color in tissues of the African lungfish, Protopterus
annectens. Lemberg and Legge (1949, P- 506, pp. 569570) listed other
vertebrate structures and invertebrates which contain biliverdin.
Identification of the green pigment
Barrio (1965a) has amply demonstrated by chemical and
spectrophotometric means that biliverdin is primarily responsible for
the green color of tissues in the frogs which he studied. Through
electrophoretic fractionation of green serum, he found that most of the
biliverdin was combined with globulin proteins and a lesser amount was
associated with albumin. Lester and Schmid (1961) have shown that
adult anurans have an enzyme for conjugating bile pigments and Rodriguez
Garay, Noir and Royer (1965) have found small quantities of water soluble
biliverdin glucuronide in Bufo arenarum. Barrio (1965a) considered
the biliverdin to be unconjugated, since it migrated at the same rate
as free biliverdin on paper chromatographs, was completely soluble in
chloroform, and insoluble in water.
Characteristics of Biliverdin
Chemical properties and reactions
Biliverdin (C33H34O6N4) is composed of four pyrrole groups
linked together by three methyne bridges. Its structural formula may
be drawn in the form of a chain, but is more precisely represented by
the incomplete ring form. There is an obvious similarity of biliverdin
to two of its precursors, protoporphyrin IXa and heme of hemoglobin.
It differs from them in that its ring structure is incomplete; also,

7
it does not contain iron as does heme. Mesobi1 iverdin is formed from
biliverdin by reduction of the vinyl side chains to ethyl groups.
Chemical reactions of biliverdin include the Gmelin reaction.
Addition of fuming nitric acid to tetrapyrrole bile pigments converts
all initially to biliverdins. These are oxidized further to yield a
succession of pigments from blue-green through violet, red and yellow
to colorless compounds.
As already indicated, reduction of the vinyl side chains of
biliverdin results in mesobi1 iverdin. Reduction of the middle methyne
bridge yields bilirubin, a common bile pigment of mammals and other
vertebrates. Bilirubin may be reduced further to other bile pigments.
Biliverdin is destroyed by heating with concentrated sulfuric
acid, but mesobi1 iverdin is not. Stable hydrochlorides or hydrobromides
result from treatment with the appropriate acid. Complex salts may be
formed with iron, zinc and copper. Biliverdin also undergoes esteri
fication. Unlike bilirubin, biliverdin does not couple with diazotized
sulfanilic acid to give the diazo reaction (Lemberg and Legge, 1949).
Physical properties
Biliverdin is moderately soluble in ether, in which it has a
greenish-blue color. It may be extracted from ether by 1 percent
hydrochloric acid, in which it has a blue-green color. These deep
colors result from the conjugation of double bonds throughout the length
of the molecule. Biliverdin is also soluble in methanol (Lemberg and
Legge, 19^+9) and chloroform, but is insoluble in water (Barrio, 1965a).

8
The absorption curve of bil verdn has been shown (Figure 1).
Mesobi1iverdin has a similar curve, but its absorption maxima lie about
10 millimicrons further from the infrared than do those of biliverdin.
Neither of these compounds has an absorption peak between 400-440
millimicrons (Soret band), which is characteristic of compounds with
the closed porphyrin ring (Holden and Lemberg, 1939)-
Biological properties
The biological properties of biliverdin are related to its
physical and chemical properties. Of considerable importance is its
solubility in different solvents. The fact that Barrio (1965a) did not
find biliverdin in nervous tissue, ocular fluids, or urine may be
directly related to the insolubility of unconjugated biliverdin in
water.
Insolubility of biliverdin in water presents a problem in
regard to its transport within the body and its excretion. In mammals,
as well as some birds and reptiles and Bufo arenarum, this problem is
not so important because the liver conjugates variable amounts of the
bile pigments (either bilirubin or biliverdin) with glucuronic acid to
form water-soluble glucuronides (Rodriguez Garay et_ a_l_. 1965). These
investigators also found a sodium salt of biliverdin, free biliverdin,
and a substance which had chromatographic characteristics of a complex
of bile acid and sodiurn bi1 iverdi na te in Bufo arenarum. The electro
phoretic studies of Barrio (1965a) indicate that in the circulatory
system biliverdin is associated primarily with globulin proteins, but
some is conjugated with albumin. Non-esterified fatty acids appear to

9
displace bile pigment from serum proteins in new-born infants (Melichar,
Polacek and Novak, 1962). Assuming that this also occurs in frogs, one
would expect that when biliverdin exceeds the carrying capacity of
the serum proteins, it wi11 be deposited in the non-fluid tissues,
staining them with its color. Evidence of the staining by biliverdin
is available in many frogs, particularly where the gall bladder lies
adjacent to the stomach wall.
Sources of Biliverdin
Hemoglobin and other hemoproteins such as myoglobin and
oxidizing enzymes constitute the main sources of bile pigments. Although
most studies of bile pigment formation are concerned with the origin
of bilirubin in mammals, we may consider the origins of biliverdin to
be the same, since it is generally considered to be a precursor of
bi 1irubin.
In a recent paper, Israels e_t a_L (1966) proposed a scheme in
which there are four components of human bilirubin. Each of these
components is characterized by a peak concentration of radioactive
bilirubin at a specific interval after the labeling of bilirubin pre
cursors. The major bilirubin component is formed from hemoglobin of
old red blood cells and makes its appearance about 100 days after the
radioactive isotope is administered. The lesser components appear more
rapidly; collectively, they are referred to as early bilirubin. One of
these reaches a peak three to five days after ingestion of the radio-
actively labeled compound and is believed to originate from a heme loss

10
during erythropoies is. The third component is formed during the
first twenty-four hours, probably beginning during the second hour
after administration. The fourth fraction appears most rapidlywithin
a few minutes of intake. The latter fraction was detected in rat
liver homogenate after five minutes' incubation with labeled delta-
aminolevulinic acid (a precursor of protoporphyrin). Presence of such
a component was suspected when the radioactive bilirubin recovered
during the first two hours was found to account for one-third of the
twenty-four-hour total. A block of the bile duct or similar obstruction
is shown to increase the latter bilirubin fraction. Perfusion experi
ments and other data show that the latter two fractions are of hepatic
rather than erythropoietic origin. The main non-erythropoietic component
of bilirubin, which is relatively slow in being formed, "is probably
related to the turnover of the major hemeproteinssuch as liver
catalase. The smaller component arises rapidly from a pool of heme-protein,
heme, or heme precursors which has a high turnover rate. Barrio (1965a)
suggested that high concentrations of biliverdin in frogs may be due to
increased formation of early biliverdin.
Mechanisms of bile pigment formation
Earlier hypotheses. -- Much study has gone into the chemistry
of bile pigments and their formation from hemoglobin. Most of these
studies have dealt with non-enzymatic reactions of bile pigments and
related compounds carried out in vi tro. Early studies of hemoglobin
degradation employed harsh techniques and detected such substances as
hematin and hematoporphyrin. With such supposed intermediate compounds

as these, and the knowledge that globin and iron are separated from
the tetrapyrrole, it was generally assumed that hemoglobin was degraded
as follows: hemoglobin, hematin, hematoporphyrin, bilirubin, with bili-
verdin a secondary oxidation product of bilirubin (Lemberg and Legge,
1949). Later, Lemberg and others (see Lemberg and Legge, 1949; Foulkes,
Lemberg and Purdom, 1951) succeeded in finding a sequence of relatively
mild reactions which degraded hemoglobin to biliverdin. Although
these reactions were not well understood and yielded only 15 per cent
of the biliverdin expected, workers in the field generally accepted
Lemberg's series of reactions as a hypothetical model of hemoglobin
degradation in cells.
Enzymatic degradation of hemoglobin. Recent studies of
Nakajima e_t aj_. (1963) have demonstrated an enzymatic pathway which is
capable of degrading hemoglobin to biliverdin. These workers have
characterized an enzyme which oxidizes the a lpha-methyne bridge of
the heme group to form a possible precursor of biliverdin. Most of this
work was done with the pyridine hemichrome rather than hemoglobins, but
their study of substrates is of considerable interest. The enzyme did
not act upon alkaline hematin or protoporphyrin IX, and only weakly
upon hemoglobin (contains ferrous ion) and hemiglobin (contains ferric
ion). However, reaction of the enzyme with a complex substrate of hemo
globin and haptoglobin (a plasma protein which combines with extracellular
hemoglobin), produced 49 per cent of the theoretical yield of biliverdin.
With the hemoglobin-haptoglobin complex as a substrate, there was a lag
period of five minutes before any noticeable reaction took place. There

12
was no lag period when a substrate of hemiglobin-haptoglobin was used.
In addition, the enzymatic activity was about 50 per cent higher with
hemiglobin-haptog1 obin or carboxyhemoglobin-haptoglobin than with
hemoglobin-haptoglobin as a substrate. Thus, it would appear that
hemiglobin-haptoglobin is the primary substrate of the enzyme and that
the lag in degradation of the hemoglobin-haptoglobin represents the
period during which hemoglobin is being changed into hemiglobin. The
enzyme acts only in the presence of oxygen and is considered an oxidase,
rather than a peroxidase. It is known as heme a 1pha-metheny1 oxygenase
and is found largely in the liver and kidney, being nearly absent from
the spleen and bone marrow. The product of this enzymatic reaction
is then acted upon by a second enzyme, heme a 1pha-metheny1 formylase,
to yield biliverdin, iron and formaldehyde.
An interesting observation which may be inserted at this point
is that biliverdin conversion to bilirubin has been shown to be under
enzymatic control in the laboratory rat (Lester e_t ajk 1966).
Causes of Chlorosis
Barrio (1965a) noted that the three families of Neotropical
frogs which include chlorotic species--Hy1idae, Centrolenidae and Pseudidae
also have in common an intercalary cartilage between the ultimate and
penultimate phalanges of the digits. He suggested that these two
common factors indicate a close relationship among these families.
However, the intercalary bone of pseudids is believed to represent an
adaptation for swimming, similar to the situation in aquatic mammals
(Goin and Goin, 1962a). Most recent writers have considered the presence

13
of intercalary cartilages in the several families of tree frogs as
an example of parallelism (Goin, 1961; Griffiths, 1963; Lynch and
Freeman, 1966). Goin and Goin (1962a, p. 231) stated, "The extra
joint thus provided allows the last phalanx, with its adhesive disc,
to be placed flat against the surface regardless of the position of
the foot--an obvious advantage to the climbing form."
No unusual conditions such as high rates of hemolysis were
noted by Barrio (1965a)- He suggested that the chlorotic conditions
which he witnessed were due to formation of early bile pigment.
Hemolytic conditions are not recorded for Amphibia, but might be
expected during metamorphosis when tadpole hemoglobin is replaced by
frog hemoglobin (McCutcheon, 1936). Varela and Sellares (1938) noted
a rapid decrease in the red cell count of Bufo arena rum at the end of
the breeding period, a change which indicates rapid hemolysis.
It appears that no satisfactory explanation of chlorosis has
been published. However, presence of biliverdin in the chlorotic species
studied by Barrio (1965a) implicates hemoproteins, particularly hemo
globin. Through hemoglobin one might expect involvement of various
structures and functions related to gaseous exchange.
Respiratory Studies
Embryonic and larval respiratory structures
Considerable research has been done on respiration as one of
the most important processes of living organisms. Much of it has been
directed toward micro-organisms, the mammalian species, and animals

14
possessing unusual respiratory structures (Negus, 1965) The several
types and combinations of respiratory organs shown by species of
Amphibia have been the interest of many investigators of respiration.
Most reports have been concerned with the structure of respiratory
organs and surfaces, rates of oxygen uptake, and the role of the blood
and circulatory system in gaseous transport. Some work has been done
on larval amphibian respiration, especially on those species of sala
manders and frogs commonly used in embryologica1 and physiological
research, despite their small size. There are few studies, however,
of adaptations in relation to the ecological situations of larval or
adult amphibians (Foxon, 1964).
Structure and function of respiratory and associated circul
atory organs of amphibians in general have been reviewed by Noble (1931)
and more recently by Goin and Goin (1962a) and Foxon (1964). The present
survey will be concerned specifically with the respiration of amphibian
tadpoles and similar aquatic forms, as well as with hemoglobin and
red blood cells of amphibians generally.
As the amphibian embryo develops, gaseous exchange first occurs
through the egg surface, later through the external body surfaces
(including gills), and finally through lungs, if present (Foxon, 1964).
Among the species of anurans, the respiratory surface decreases pro
portionately as the diameter of the egg increases, since the volume of
the egg increases as a cubic function of the radius, while the area
increases as a squared function. No examplesof enlarged surface area
due to folding or other changes in the egg membrane are known (Noble, 1931).

15
However, there are considerable differences in size of amphibian eggs.
Panton (1952) noted that aquatic eggs of Hyla brunnea were only one-half
millimeter in diameter, whereas those of another frog (probably
Eleutherodacty1 us jamaicensis) that did not develop in water, were
several millimeters in diameter.
The sites of egg deposition and tadpole development appear
related to respiratory processes. Dickerson (1906) believed that the
jelly membranes of amphibian eggs retain heat longer than the surround
ing water. Experiments by Savage (1950), using the eggs of Rana
temporaria, showed that the mean temperature difference between eggs
with jelly envelopes and plain water was 0.63 C. Moore (19^0) suggested
that the compact jelly masses (up to 10 centimeters in diameter) of
Rana syIva tica slow the diffusion of oxygen to the embryos, and that
this becomes critical as the metabolic rate increases in response to a
temperature of about 25 C. However, Savage (1950) pointed out that
water moves freely between individual jelly capsules of Rana temporaria,
thus requiring diffusion of oxygen through only a few millimeters rather
than several centimeters of jelly. Algae associated with jelly membranes
of Rana svIva tica (Dickerson, 1906), Rana aurora (Wright and Wright, 19^9)
and Rana temporaria (Savage, 1961) may affect the respiration of these
frogs' eggs.
Goin and Goin (1962^) noted that amphibian eggs may be laid
singly or in clusters and may be attached to submerged objects, floating
or settled in water, carried by parents, or be laid on land or vegetation.

16
There are few indications of respiratory rates of embryos being corre
lated with these types or sites of egg deposition. Panton's observation
(1952) that a set of large eggs, probably those of E1 eutherodacty1 us
i amaicens ?s, fai led to develop when placed in water may indicate that
embryos which develop terrestrially receive insufficient oxygen when
subjected to the reduced oxygen tension of water. Noble (1931) suggested
that large eggs undergoing rapid development require more oxygen than
they can obtain in water when surrounded by the egg capsule.
As the embryos develop and hatch, other respiratory structures
are formed and behavioral patterns related to respiration arise. Organs
of external respiration in tadpoles are the skin, the internal surface
of the operculum, external and internal gills, the vascularized surface
of the food filtering apparatus, and the lungs.
The most evident respiratory structures to appear in about
the time of hatching are the external gills. These may be small as
in Hyla vasta or enlarged as in Hyla rosenbergi tadpoles, although both
species dwell in tropical streams (Noble, 1931). Dunn (1926) noted
the reduction of gill size in Jamaican hylids, which breed in water
which collects in bromeliads. Likewise, tadpoles of the genus Hoplophryne.
which live in water that collects in banana leaves, also have reduced
gills (Noble, 1929). Terrestrial embryos of the genus Eleutherodacty1 us
may or may not have external gills, their highly vascular tails serving
as important respiratory organs (Lynn, 1961). Among the more unusual
are the gills of some Gastrotheca which are large and bell-shaped (Noble,
1931) .

17
Studies of Babak (1907a, 1907b) upon Rana temporaria and
Rana esculenta, of Drastich (1925) upon Salamandra maculosa, of Bond
(I960) on Salamandra maculosa, Ambystoma opacum and Ambystoma jefferson-
ianum, of Conant (1958) concerning adult Necturus maculosus and Freeman
(1963) on adult Pseudobranchus striatus, all showed an inverse relation
ship between external gill size and oxygen concentration. Most indicated
that external gills increased in size when animals were put into water
of lower oxygen concentration. In addition, Conant, Bond, and Freeman
all noted that gill movements and expansion, as well as blood supply
were correlated with oxygen tension.
Pulmonary respiration is important to the majority of adult
amphibians and is found in a number of larvae as well. The lungs of
stream dwelling species are reduced in size (Noble, 1931). Savage
(1961) has described the well developed lungs of Rana temporaria larvae
and still larger ones from a Papuan tadpole; both of these species
were taken from standing pools or puddles of water. In the Papuan
species, it appeared that tail muscles pressed the lung against the
notochord to bring about ventilation.
Savage stated further that:
The respiratory systems in tadpoles are connected with the
ecology in the following way. If a tadpole lives in an environ
ment rich in food, as many temporary or polluted ponds are,
it does not need to pump water to get its good, and so does
not need large gill filters. To use the oxygen under these
conditions, however, it must have large gills, and needs
lungs to tide it over emergencies. If, however, it lives
in the oligotrophic type of pond, with plentiful and almost
constant supplies of oxygen but with a low concentration of
food, it needs to pump much water, and so must have large
gill filters. With these, gills might not be necessary,
because of the large surface of the filters (or rather, in

18
view of Strawinski's work, more probably, the assoc
iated large operculum). in ordinary ponds, intermediate
as habitats, the arrangements might be expected to be
intermediate also.
There is a great deal of conjecture in all this, but
the microhylids seem to provide examples. Some have such
large gill filaments that they trail in the opercular
cavity in a way quite unlike those of Rana, others have no
gills but have enormous gill fi Iters, for example Glypho-
glossus molossus. Some, such as Hypopachus aquae, are
intermediate and live in ordinary ponds. Rana temporaria
also have a moderate development of gills, filters and lungs,
and lives in ordinary ponds.(P- 56)
Among amphibians, we may turn to the work of Strawinski (1956)
for an estimation of the relative importance of tadpole respiratory
surfaces. In studying the density of capillary networks of Rana
esculenta, he reported that most gaseous exchange of early stages
took place through the skin, with the internal surface of the operculum
also important. It was his belief that the external and internal gills,
and vascularized filtering apparatus were of little importance. As
the lungs developed, their portion of the total respiratory capillaries
increased rapidly, reaching 65 per cent at metamorphosis, the same
proportion as in the adult. Vascularization of the skin also increased
until metamorphosis, when these capillaries accounted for 3^ per cent
of the tota 1.
Blood transport of oxygen
Among vertebrates, blood contained within a closed circulatory
system is largely responsible for the transport of gases in the body.
Hemoglobin pigments, located in red blood cells and responsible for
their color, are the main carriers of oxygen in vertebrates. Exceptions

19
to this are the Antarctic ice fishes, Chaenichthyidae (Ruud, 1959).
Blood of the several species of this family is nearly colorless and
contains only white cells (1 per cent of blood volume). The cold water
in which these fish live has a high oxygen content and the low tempera
ture tends to lower the respiratory rates, so that oxygen diffusion
through the naked skin sufficiently supplements that dissolved in the
blood to maintain their low rate of metabolism. Small eel larvae
also lack hemoglobin until they reach the elver stage (Andrew, 1965)
and occasional specimens of the frog, Xenopus laevis, have been found
without hemoglobin (de Graaf, 1957; Ewer, 1959)-
Among the indicators of oxygen-carrying capacity of blood
are measurements of red cell size, erythrocyte counts and determinations
of hemoglobin content. Amphibian erythrocyte sizes and counts have been
listed by Prosser et a]_. (1950), Vernberg (1955) > Freeman (1963), and
Hartman and Lessler (1964). Of all vertebrates, salamanders have the
largest red cells, ranging from 30-62 microns in longest dimension.
According to these sources, erythrocytes of frogs, which are also oval
in shape, range from 18-23 microns in length. The number of red blood cells
usually varies inversely to cell size. Counts of salamander cells
vary from 28,000 to 197,000 per microliter; those of anurans, from
380,000 to 670,000 per microliter. While one would expect larger cells
and higher cell counts to represent greater oxygen-carrying capacity, a
trurer measure lies in the direct determination of hemoglobin. Kirkberger
(1953), Stuart (1951), and Goin and Jackson (1965) have published values

20
for hemoglobin concentrations of amphibian blood. These range from
8.7 to 11.5 grams of hemoglobin per 100 milliliters of blood.
In regard to shape, it should be noted that the oval disc
form of amphibian erythrocytes presents a greater surface area per unit
of volume than would a sphere of like volume, thus presenting a greater
surface for gaseous diffusion (Andrew, 1965). Greatest efficiency of
oxygen transport among amphibians appears in the genus Batrachoseps,
which may have up to 90 per cent erythroplastids or enucleate
erythrocytes (Emmel, 1924). Lack of a nucleus makes the cells flatter
and thus increases cell count per unit volume, as well as reducing the
oxygen uptake of the red cel 1 by the amount consumed by the nucleus
(Foxon, 1964).
Transfer of oxygen to respiring tissue
Once oxyhemoglobin is carried to the tissues from the respira
tory organs, oxygen dissociates from it and diffuses into the cells.
Krogh (1941) has shown that the rate of oxygen diffusion in tissues is
variable, but less than half that of water. Less is known of conditions
which promote dissociation of oxyhemoglobin of the type found in tadpoles.
Since the great affinity which tadpole hemoglobin has for oxygen is not
modified by changing pH, dissociation would appear more difficult than
it is for oxyhemoglobin of adult erythrocytes (Manwell, 1966).
Oxygen utilization and metabolic rate
Boell (1948) has reviewed earlier work which related to oxygen
consumption of early amphibian embryos. Much of this was concerned with
respiration of various regions of the gastrula stage. The work of

21
Brachet (1935) upon eggs of Rana fusca (= Rana temporaria) showed no
significant difference between respiratory rates of fertilized and
unferti1ized eggs.
Among the first studies of oxygen uptake in anuran larvae
were those of Helff (1926) and Etkin (1934), working upon Rana pipiens
and Rana catesbeiana respectively. Atlas (1938) studied Rana pipiens
and Rana sylvatica from fertilization to tadpole states and found that
respiratory rates increase from fertilization to overgrowth of the
operculum. This was in agreement with the studies of Godlewski (1900)
and Bialaszewicz and Bledowski (1915) on Rana temporaria. Pamas and
Krasinska (1921) found the rate of oxygen consumption rose sharply at
the onset of gastrulation, neurulation, and external gill formation,
with no noticeable changes between. Atlas (1938) suggested that the
latter conclusions were based upon insufficient data which wereadversely
affected by temperature fluctuation. it is interesting to note that
the respiratory rates obtained by Bialaszewicz and Bledowski for Rana
temporaria are similar to Atlas' data for the closely related Rana
sy1 vatica, but somewhat different for Rana pipiens or Rana catesbeiana.
Atlas concluded that the increased rate of respiration per embryo was
a function of cell number rather than of intensity of respiration of
individual cells. The basis for this statement is obscure, since no
cell counts were reported, nor was the subject discussed In the main
text of the paper.

22
Relation of Respiration to External Environment
Although a number of studies have dealt with respiration in
regard to external factors such as temperature and partial pressure of
oxygen, little has been done to relate respiratory form and function of
a tadpole to its specific environment. The most thorough study along
these lines was done by Savage (1961), who compared several species of
tadpoles in relation to feeding and respiration in different environ
ments. His conclusions were quoted previously.
Laessle (1961) noted that the tadpoles of the common Jamaican
tree frog, Hyla brunnea, lived in stagnant water of very low oxygen
content, which collects in the central reservoirs and leaf bases of
bromeliads. He suggested that the agitation of the long tail increased
oxygen diffusion into the water and that the large surface area of the
tail provided additional respiratory surface. The tadpoles of the other
three species of Jamaican tree frogs also develop in the oxygen-poor
water which accumulates in bromeliads (Dunn, 1926). Tree frogs, closely
related to the Jamaican species, on the island of Hispaniola, are known
to breed in streams and torrents (Noble, 1927); the Cuban tree frog,
Hyla septentriona1 i s, is known to develop in brackish water, as well
as in cisterns and pools (Grant, 1940).
It may be noted here that Powers et aj_. (1932), working upon
fish respiration in relation to environment, found that "The number of
red blood corpuscles is increased with a decrease in the oxygen and by
an increase in the carbon dioxide tension of the water and vice versa."

23
Structure and Function of Froa Liver
Studies of liver structure in frogs have been few; most were
considered by Elias and Bengelsdorf (1952). These authors found that
the walls separating neighboring sinusoids are predominantly two cells
thick in frogs, but only one cell thick in mammals. Similarly, relatively
little experimental work has been done on the excretion of bile pigment
by frog liver. Nisimaru (1930 carried out perfusion experiments on the
liver of Rana catesbeiana. He found that the rate of bile pigment
excretion changed when blood pressure in the liver circulation was
a 1 tered.
Papers concerning frog liver structure or function do not
aid in the explanation of chlorosis in these forms.

SOURCES AND CARE OF TADPOLES AND FROGS
The live frogs and tadpoles used for most of this study were
collected from their normal habitats during field trips to the West
Indies and South America. The species included: Hyla septentriona1is
collected from Grand Cayman and Southern Florida; Hyla brunnea, Hyla
1ichenata, Hyla marianae, and Hyla wi1deri from Jamaica; Hyla domini-
censis, Hyla vasta and Hyla heilprini from Haiti ; Leptodacty1 us
a 1bilabrus of Puerto Rico; and Hyla maxima and Pseudis paradoxus from
Surinam. Incidental observations were made of Phy1lomedusa tadpoles on
Trinidad and several other hy1id species in Surinam. Collections were
made at the following places during the months indicated: Jamaica
June-July, 1965; August, 1965; October, 1965; May, 1966 and August, 1966;
Haiti -- October, 1965; May, 1966 and August, 1966; Grand Cayman --
July, 1965 and October, 1965; Puerto Rico -- May, 1966; Southern
Florida July, 1966; Trinidad -- July, 1966; Surinam -- July, 1966.
Tadpoles of the species listed above were collected with a
tea strainer or small dip net. They were transferred to plastic bags
which were partially filled with water from the immediate environment
of the tadpole. When suitable tap water was available, it was used to
replace water brought from the field. Thus the frog larvae were trans
ferred to water which was apparently free of noxious substances,
detritus, and other organisms. Water was poured off and replaced with
fresh tap water as often as required, usually at one- to three-day
intervals. Tadpoles did not show marked reactions to addition of tap
24

25
water from localities in South America, the West Indies, or Southern
Florida. Care was taken to keep the temperature of the fresh tap
water about the same as that which had been removed.
Upon returning to the laboratories in Gainesville, Florida,
the larvae were transferred to culture dishes of appropriate size.
Since Gainesville tap water was found to be lethal to both Hyla brunnea
and Hyla septentr1ona1is, spring water or Holtfreter's solution was
ordinarily substituted. During May, 1966, Holtfreter's solution proved
unsatisfactory and its use was discontinued. At one point, the spring
water appeared harmful and filtered pond water was used in its place.
By avoiding direct sunlight wherever possible and using other
precautions, tadpoles were maintained in the field with few losses
due to excessive temperature. At the University of Florida, the culture
dishes containing the tadpoles were moved from one laboratory to another
in an attempt to maintain a moderate temperature; however, groups of
tadpoles were exposed to water temperatures between known extremes of
16 C. and 28 C.
Thus, with reasonable care, most of the tadpoles were collected
and transferred to Gainesville with little difficulty. A notable except
ion was the Hyla heiIprin? tadpoles, which did not survive more than
three days after capture. Of the other species, most demonstrated an
ability to live for two weeks or more without being fed. However, in
the laboratory and whenever possible in the field, the tadpoles were
fed Gerber's strained foods at regular intervals. Hyla septentriona1is,
Hyla dom?nicensis, Hyla vasta, and Leptodacty1 us albilabrus were fed a

26
mixture of peas, carrots, and spinach. Hyla brunnea, Hyla iichenata,
and Hyla marianae were fed strained egg yolk. Tadpoles were maintained
satisfactorily for two months or more on these diets. Of those on the
above diets, individuals of all species but Hyla septentriona1is reached
metamorphosis under laboratory conditions. Following metamorphosis,
the young frogs were kept in covered jars, and were fed termites and
other insects.
Adult frogs were collected by hand, using a flashlight at
night to locate them at their calling locations, or by searching for
them in their resting sites during daylight hours. They were transported
in jars or in plastic or cloth bags. In the laboratory, they were kept
in jars or terraria. Crickets or other insects were used to feed the
adult frogs at approximately weekly intervals.
In addition to live specimens, hundreds of preserved frogs
and tadpoles were surveyed for unusual characteristics. Most of these
specimens were of West Indian hylids and were obtained from the museums
mentioned under Acknowledgments.

I. CAUSATIVE PIGMENT AND INCIDENCE OF CHLOROSIS
Pigment Identification
During the present study, filtered tissue extracts and body
fluids were tested spectrophotometrica1ly and chemically. Figures 2-6
show that the spectral characteristics of several of these green fluids
are very similar to those of biliverdin (Figure 1). Addition of fuming
nitric acid to the fluids was followed by the succession of colors
known as the Gme1in reaction, generally considered to indicate the
presence of bile pigments. These two tests, along with characteristics
of color and solubility, appear to eliminate from consideration all
known green pigments except the two naturally occurring green bile
pigments, biliverdin and mesobi1iverdin. Since green bone is rapidly
bleached by concentrated sulfuric acid, mesobi1iverdin does not appear
to be present, since it is stable in this substance. Thus, it was
independently concluded during the present study that biliverdin is
the pigment primarily responsible for the green color of these frog
tissues.
Survey of green pigmentation among Neotropical anurans
During the course of several field trips to the West Indies
and one to Surinam, it was possible to study a variety of tropical
anurans under natural conditions. Many of these did not exhibit any
green color, while others showed different levels of green pigmentation
in the tissues and body fluids. Observations made during the present
27

28
study and by others are summarized in Tables 1 and 2. Points to be
noted from Table 1 include the fact that the color of the bile was
usually green. in many individuals, the bile was so concentrated as
to appear blue when seen through the gall bladder wall. When soft
tissues were green, it was usually a rather general phenomenon and often
was accompanied by a relatively high concentration of plasma pigment,
a condition which might be termed chlorosis.
Sex and its relationship to chlorosis in frogs
Due to the relatively small number of individuals collected
from each species, it is difficult to make a generalization concerning
the relationship between chlorosis and sex. Chlorotic and non-chlorotic
individuals of both sexes were among specimens of Hyla septentriona 1 is
from Grand Cayman. Table 2 shows several species in which all tadpoles
observed had green pigment at the time of metamorphosis or before. The
general occurrence of biliverdin in these species indicatesthat the
sex of the individual is not important to the development of chlorosis
at this stage. The presence of green pigmentation in calling males of
several species and in eggs of certain Hypero1ius. Aaalychnis,
Centro lene 11a, Pseud? s and Hyla indicates that sex is not a very import
ant factor in the development of chlorosis.
Age distribution of chlorotic anurans
Among tadpoles, no green pigmentation was noted in young
tadpoles of Hy 1 a brunnea h'y la boeseman i Hy la domin ?cens? s Hy la
leucop'ny 1 lata Hyla maxi ma Hyla rubra, Hyla septentriona 1 i s Hyla vasta,

29
When green pigmentation was present in tadpoles, usually
it made its appearance at the onset of metamorphosis and reached a
peak upon the completion of metamorphosis. All of the species of Hyla
listed in Table 2 are from the West Indies and they present similar
appearances at the completion of metamorphosis, except for Hyla marianae.
Most of their skeletons were green, particularly the limb bones and
vertebrae; the green pigmentation was not so pronounced where the marro;
maintained its hemopoietic function; melanin pigmentation had developed,
especially on the dorsal side; pigmentation of the ventral side was not
well developed in Hyla brunnea, Hyla 1 ichenata, or Hyla marianae, all
of Jamaica (stomach contents could be seen through the skin); a white
substance, presumably guanine, had developed in the skin of the other
species, which are usually found in more exposed habitats than the
Jamaican species; green pigmentation of soft parts was particularly
intense in the region of the throat and pectoral girdle; the small size
of most of these species of tadpoles made it difficult to determine
the color of the plasma. In Surinam, the tadpoles of Pseudis paradoxus
were found to be darkly pigmented, externally by melanin and internally
by biliverdin and other pigments. The plasma of this species is green
prior to metamorphosis. The coiled intestines of these gigantic tadpoles
are packed with green plant materials and fill most of the body cavity.
During the present study, a young Rana heckscheri collected
near Gainesville, Florida, at the completion of metamorphosis was found
to have gray-green bone marrow. This appeared to be a stage which
represented degenerating red marrow of the tadpole. This color appeared

30
to be restricted to the marrow, since neither the bone nor the plasma
demonstrated green pigment. Although the origin of this pigment
probably was the same as that of the Neotropical forms, the color was
almost obscured by the bone.
Green pigmentation of the adults is summarized in Table 1.
It is worthwhile to compare the amount of green pigment in adults, with
that of tadpoles of the same species, where possible. Tadpoles of
Pseudis paradoxus, Hyla dominicensis, Hyla lichenata, Hyla septentriona1is
and Hyla vasta generally had much more extensive chlorosis of soft
tissues than did the adults of these species. Two adult specimens of
Hyla he?Iprini showed chlorosis as extensive as that of metamorphosing
tadpoles, but melanin pigments of the dorsum were better developed in
the adult than in the tadpoles. While scores of tadpoles of Hyla brunnea
invariably had green pigmentation, dozens of adults showed no bil ¡verdn
in tissues, including bone. Green pigment is absent from the tissues
of Hyla marianae, which are orange in color.
As indicated previously, there was individual variation of
green pigmentation within populations of Hyla septentr?ona1is. Small
species such as Hyla wiIderi have bones of a more intense green color
than do larger species such as Hyla maxima or Pseudis paradoxus. Calcifi
cation of bones tends to obscure the green pigment, as in the larger
species, but the color of biliverdin is quite apparent in poorly calcified
bones as in the distal limb structure of Hyla wilderi.

31
Seasonal changes in chlorosis
It should be noted that most of the specimens for the present
study were collected from June through August. This period constitutes
the main breeding period for most of the species studied, a fact which
should be kept in mind when considering the physiology of these organisms.
It is interesting to note that Hy la qeoqraph? ca had green bones,
but no green pigment in the plasma. On the other hand, Hy 1a mi sera had
light green plasma, but white bones. These two examples indicate that
high concentrations of green pigment in the plasma are temporary in some
species. Different shades of green in plasma and bones of other species
tend to support this idea. Concentric layers of different shades were
seen in Barrio's (1965a) photograph of bone from Lysapsus mantidactylus,
and in a femur of Osteocepha1 us taurinus. Like growth rings of a tree,
these suggest a seasonal change in conditions during development of
the tissue.
Phylogenetic distribution of chlorotic frogs
An understanding of the phylogenetic relationships of chlorotic
frogs to those which lack tissue biliverdin should be of value. However,
Tables 1 and 2 indicate no clear boundaries along phylogenetic lines.
As previously mentioned, adult individuals of Hyla septentriona11s and
Hyla dominicensis may fall into either category. In Hyla brunnea,
the pigment is always present in young frogs, but has never been observed
in the adults. Hyla pul che 1 la has some populations which are chlorotic
and others which are not. Similarly, it can be seen that neither the
genus Hyla nor the family Hylidae shows uniformity in this characteristic.

32
The degree of pigmentation is no more helpful than its presence or
absence. Of the species collected in Surinam, Hyla punctata, Hyla
crepitans, Sphaenorhynchus aurantiacus and Phrynohyas venulosa had the
highest concentration of green pigment of the frogs collected, but they
certainly do not constitute a homogeneous group of species. Of the
eight species of Hyla that were studied in life in the West Indies,
seven had green pigmentation at metamorphosis or later, but Hyla
marianae lacks green pigmentation in these stages, being orange instead.
There is at least one relationship between chlorosis and frog
phylogeny. Thus far, all chlorotic frogs have been members of only four
families. Most of the green species are members of the family Hylidae,
while the others are included in the Centrolenidae, Pseudidae, and
Pvhacophor i dae. The chlorotic hylids, centrolenids and rhacophorids
are all tree frogs, while pseudids are aquatic for extensive periods.
All four of these families have in common an intercalary cartilage
between the ultimate and penultimate phalanges of the digits.
A close relationship between the four families is unlikely
even though they have the intercalary cartilage, green pigmentation and
tropical distribution in common. Hyperolius, the African genus of
rhacophorid frogs which has green eggs, has a firmisternal pectoral
girdle and diplasiocoelous vertebrae which distinguishes it from chlorotic
frogs of the other three families, which are procoelous, have an arci-
feral pectoral girdle and are Neotropical. There appears to be con
siderable agreement that the presence of intercalary cartilages in

33
these families is the result of parallel evolution. All things con
sidered, there appears to be no clear relationship between phylogeny
and chlorosis among frogs.
Geographic distribution
Frogs with green tissues or eggs appear to be tropical or
sub-tropical in their distribution. Below is a list of green pigmented
crogs according to the countries or regions in which they are found.
Sources of information are given in Table 1.
Central Africa: Hypero 1ius sp.
Jama ica: Hyla brunnea, Hyla lichenata, Hyla wi Ideri.
Hispaniola: Hyla dom?nicensis, Hyla heiIprini, Hyla vasta.
Cayman Islands (presumably also Cuba and Bahama Islands):
Hyla septentr?ona1is.
Mexico end Central America: Agalychnis dacnicolor, Centro le
ne 1 la a 1bomaculata, Centro lene 1 la granulosa, Centrolenella ?lex,
Centrolene1 la spinosa, Centrolene1 la prosoblepon, Centro lene 1 la
pul vera turn.
Northern South America, including Brazil: Lvsapsus 1 ime 1 1 urn
laevis, Pseudis paradoxus, Anotheca coronata, Hyla albofrenata, Hyla
a 1bomargina ta, Hyla boesemani Hyla calca rata, Hy la crepitans, Hyla
cuspidata. Hyla geographica, Hyla langsdorffi, Hyla maxima, Hyla mi sera,
Hyla punctata, Osteocepha1 us taurinus, Phrynohyas venulosa, Sphaenor-
hynchus Trachycepha1 us nigromaculata, Centro lene 1 la vanzolinli.
Argentina: Lysapsus 1 i me 11 urn limellum, Lysapsus mantidacty1 us,
Pseudis paradoxus platensis, Hyla berthae, Hyla nasica, Hyla puIche1 la,

34
Hyla phrynoderma, Hyla punctata rubro!?neata, Hyla raniceps, Hyla
slemersi, Hyla squa1 irostris, Hyla trachytorax, Phrynohyas venuiosa.
Judging from this list, it appears that nearly all green
pigmented frogs are to be found in the Neotropical Zone. It should be
noted that, except for the works of Barrio (1965a, 1965^) and the present
writer, the references to green tissues or eggs of frogs are few, and
these are often obscure. Although a number of individuals have volun
teered personal observations concerning coloration, there appears to
be a general reluctance to publish such observations. It seems quite
possible, if not likely, that additional records will be forthcoming
from Africa or other parts of the Old World tropics.

II. CONSIDERATION OF ECOLOGICAL FACTORS ASSOCIATED
WITH CHLOROSIS
In attempting to determine the cause of a condition such as
chlorosis, ecological factors should be considered. It has already
been noted that the majority of chlorotic anurans are tree frogs
which are restricted to tropical or sub-tropical regions. However,
not all tropical tree frogs have green tissues; conversely, the chlorotic
pseudids are not tree frogs, but are aquatic. Since the tropical climate
and general habitat (aquatic, terrestrial or arboreal) do not completely
account for either presence or absence of chlorosis, one should con
sider the importance of specific habitats, as well as behavioral adapt
ations to such environments. This section includes a summary of
information concerning specific habitats and the frogs associated with
them. Greater emphasis was placed upon ecological study of breeding
sites because of their accessibility and the appearance of chlorosis
during the larval stages. Special attention was given to those factors
which might be related to hemoglobin or red cell formation and function,
including oxygen tension, temperature and iron concentration. Additional
factors were surveyed in order to find conditions which were markedly
different from those ordinarily encountered by anurans.
Methods of Study
Concentration of dissolved oxygen in the habitats of tadpoles
was measured with a Precision Scientific Portable Oxygen Analyzer, which
was calibrated in air.
35

36
Temperature readings of tadpole habitats were taken with the
thermistor component of the oxygen analyzer wherever possible; other
wise, they were made with standard mercury thermometers, graduated
from -10 C. to 100 C.
Ferrous and total iron, and phosphate concentrations of water
were measured by colorimetric methods using a Hach portable colorimeter
and the appropriate techniques (Hach Chemical Company, no date).
Salinity, alkalinity and hardness were measured by titrametric methods,
using the appropriate kits and techniques developed by the La Motte
Chemical Company. Water pH was measured to the nearest half unit by
means of pHydrion paper.
Observations of Frog Habitats and Behavior
During this study, frogs were collected in a variety of
habitats from tropical rain forest to open, grassy areas to residential
districts. Although eggs and tadpoles were found in or near the environ
ments of the adults of their respective species, the larval environs
were more uniform in appearance. Most of the tadpoles were found in
standing water, but several species of tadpoles were found in flowing
streams. While the temperature and dissolved substances of these
habitats were found to vary considerably, one may satisfactorily divide
tadpole habitats into flowing stream and standing water types. Within
the latter type, special consideration will be given to the bromeliad
microhabitat of Jamaican tree frogs. Ecological data collected during
this study are presented in Tables 3~5-

37
The brome liad microhabitat
Members of the pineapple family, Brome 1iaceae, ordinarily
do not constitute the dominant plants of a habitat, although they may
be an important part of the flora. On the West Indian island of
Jamaica, the bromeliads have undergone considerable adaptive radiation
(Dr. Richard Proctor, personal communication). They may be large or
small, epiphytic or terrestrial, and are found in shaded and open areas.
This appears to be a very fortunate circumstance since the water which
is caught in the leaf bases and central reservoirs of bromeliads is
the most reliable supply for small animals, including the four species
of tree frogs on Jamaica.
The importance of the close relationship between the Jamaican
tree frogs and the bromeliads should not be underestimated. Perkins
(1348) noted that the water level in the bromeliads ("wild pines")
is maintained by dew which condenses and runs down into the reservoir
in the center of the plant; very little direct sunlight tends to reduce
evaporation from the wild pines. She states further (p. 87), "In view
of the many creatures that depend on the wild-pine for moisture it would
seem that these plants hold an important place in the economy of the
countryside, for surely our wildlife would be largely depleted during
a severe drought, were it not for these hidden stores of water." My
observations during the drought which continued into July, 1965, sub
stantiate this position. Except for an occasional pool in stream beds,
or the largest rivers which continued to flow, there was no surface

38
water other than that in the bromeliads. Steepness of the hills and
porosity of the limestone substrate are partly responsible for rapid
run-off of water.
Observations of the adults of the three smaller Jamaican hylid
species indicate that they prefer to be covered by water, at least in a
lighted area. Under these circumstances, they may remain completely
immersed for minutes at a time, and then slowly rise until only the
external nares and eyes protrude above the surface. It is interesting
to note that the four Jamaican species differ from their relatives of
Cuba and Hispaniola in having a more truncate snout with the external
nares at the most anterodorsal point (Dunn, 1926). This may be inter
preted as an adaptation to living in the reservoirs of bromeliads.
Laessle (1961) studied the ecology of Jamaican bromeliads,
in which he found the following ranges for the water which they con
tained: dissolved oxygen, 0.0 8.0 ppm; dissolved carbon dioxide,
4.0 67-0 ppm; pH, 4.0 7-0; temperature, 17-5 30.0. He estimated
the maximal quantity of water in the reservoir of a large bromeliad at
200 mi 11 i 1iters.
Hyla brunnea. -- The brown tree frog of Jamaica is the most
widely distributed species of Hy1 a on the island. It is absent from
the Blue Mountains above 1600 meters elevation, and from arid areas
such as Kingston and the Hellshire Hills along the south coast (Lynn,
1940). This frog is most likely to be found in localities where the
large tank bromeliads, including species of Hohenberqia and Aechmea, are
readily available as breeding sites and resting places. While this frog

39
may be found in bromeliads which grow at ground level on shaded, lime
stone hillsides or in the epiphytes high above ground in well developed
forests, they are frequently found in towns or among the trees which
line the roads in agricultural areas. Like the other tree frogs of
Jamaica, Hyla brunnea appears to be largely dependent upon the bromeliads
for the water which is retained at the leaf bases, but several Jamaicans
have told me that the brown tree frog is found among banana leaves and
sta1ks.
Breeding habits of Hyla brunnea are unusual in some respects,
particularly as they relate to adaptations to life in the bromeliads.
The breeding season begins during May, as indicated by a record cited
by Dunn (1926). Panton (1952) has stated that the strongest choruses
of the tree-toad are heard at the end of May or beginning of June, but
that the time fluctuates because of weather conditions. During the
latter part of May, 1966, and early June, 1965, I noted very few tadpoles,
but found eggs more common than later in the year. Tadpoles were not
uncommon during October, 1965- The eggs of Hyla brunnea are deposited
in the central reservoir of bromeliads in most cases, but eggs or
tadpoles may be found at the bases of outer leaves; on one occasion,
eggs were found in water that had collected on a hollow branch.
Development of the tadpole stages has been described by
Schreckenberg (1956), who studied the embryonic development of the
thyroid gland in Hyla brunnea. The mode of development differs little
from that seen in Rana and hylids of the United States. The main differ
ences appear to be the result of adaptations to the bromeliad habitat.

40
The jelly mass from which the tadpoles hatch remains in the
bromeliad reservoir longer than do the tadpoles themselves. Panton
(1952) has suggested that the presence of the jelly reduces evaporation
and moderates temperature changes. The low pH of the water probably
prevents the rapid decay of the jelly mass or infertile eggs. In his
micro-limnological study of Jamaican bromeliads, Laessle (1961) found
the following ranges of readings in five bromeliads which contained
eggs or tadpoles of Hyla brunnea: dissolved oxygen, 0.03 2.3 parts
per million; dissolved carbon dioxide, 23.0 41.0 ppm; pH, 4.0 4.5;
temperature, 23.0 25.0 C. Additional measurements made during the
present study increased these ranges to: dissolved oxygen, 0.03 2.7 ppm
pH, 4.0 6.5; temperature, 23.0 28.0 C.
During the present study, the contents of a number of bromeliad
reservoirs were poured into waterproof containers. When jelly masses
without eggs were present, they were not firm and had a tendency to
separate into capsules 1-2 centimeters in diameter. These probably repre
sent capsules which contained four to six eggs each, as mentioned by
Dunn (1926). Within the bromeliad, the mixture of water and jelly has
the consistency of glycerine, as mentioned by previous writers. While
I have noted tadpoles, especially small ones, moving about near the
surface on several occasions, I have also noted larger ones moving verti
cally within the reservoir. The larger individuals tend to remain under
leaf fragments when the reservoir is exposed to the sun. They come to
the surface at intervals to take air and then return to lower depths.
This respiratory behavior apparently was not observed previously, since

41
Dunn (1926) was unable to reconcile the reduced gill structure with
the low oxygen content of the environment. Lungs are seen on either
side of the vertebral column, and appear like small bubbles. Lungs
are present in very small tadpoles as well as older ones. They appear
to be used as accessory respiratory structures when branchial and cutaneous
respiration are insufficient. In this respect, the tadpoles would be
similar to the Austra1ian lungfish, which utilizes pulmonary respiration
largely at night when it becomes more active (Grigg, 1965)- While the
long, narrow tails of Hyla brunnea tadpoles probably are important
to cutaneous respiration as Laessle (1961) indicated, it seems likely
that their primary purpose is to propel the organisms through their
viscous environment.
The diet of Hyla brunnea tadpoles consists largely of frog
eggs, as noted by Dunn (1926) and Laessle (I96I). Most often the eggs
are probably those of its own species, since it is by far the most
common tree frog of Jamaica. in addition, it is the only hylid species
known from the eastern third of the island so that any hylid eggs found
in Hyla brunnea tadpoles there, would have to be of the same species
(Laessle, 1961).
Several structural characters of the tadpoles can be correlated
with their unusual diet. The digestive tract is expanded into a sac
in which eggs may be found more than a week after the last feeding. it
is relatively straight, not coiled as in vegetarian tadpoles. Since the
gills are not important in collection of food, their structure is

42
simplified. Finally, the mouth has only a single row of teeth about
it (Dunn, 1926), since these are not needed for scraping food from
surfaces.
The length of time required for development from fertilization
to metamorphosis is not known since neither Lynn (1940) nor I was able to
follow a single group through the entire period. From Lynn's work, nearly
a week passes between fertilization and hatching. After hatching, at
least a month probably passes before metamorphosis is completed. While
tadpoles have been kept in captivity for more than six weeks, it is likely
that metamorphosis could be completed approximately six weeks after ferti
lization under optimal conditions.
At the time that the forelegs penetrate the opercular fold,
the green pigmentation was always quite obvious. Jarring of the bromeliad
or container in which such tadpoles were present caused them to climb up
ward very rapidly. They have no difficulty in climbing out of a bromeliad
reservoir, a water glass, or a plastic bag. Once they leave the water,
the tail is resorbed in twenty-four to thirty-six hours. This is in
agreement with the finding of Schreckenberg (1956) that there is intense
thyroid secretory activity and sudden release of colloid from the thyroid
gland at the stage of tail resorption. The dark green pigmentation of
soft tissues in the guiar region remains for two weeks or more after
metamorphosis and then gradually fades away in the living animal. Green
pigmentation of bones remained as long as the young frogs lived, or about
six weeks after metamorphosis, in the laboratory. In nature, no immature

43
frogs were found in bromeliads, ant or termite nests, or elsewhere,
so that the post-metamorphic development and ecology of the young
are unknown.
Hyla lichenata. This giant tree frog, which attains a
length of 117 millimeters, is restricted to the central and western hills
of Jamaica, generally above 300 meters in elevation. Since this frog is
rarely seen and less often collected, knowledge of its habits and distri
bution is based largely upon hearing its distinctive snoring call (see
Lynn, 1940).
In regard to the habits of this species, Panton (1952) first
found one in a bromeliad on a dead candlewood tree in woodland, but all
subsequent specimens were taken from hollow trees. Dunn (1926) collected
three of these frogs from small hollow trees with openings 4 to 12 feet
above the ground, but noted that they ordinarily call from greater heights.
He stated further that the "bony head is obviously of use in plugging the
hole after the frog is inside." Lynn and Dent (1943) traced the unmis
takable call of this species to a clump of bamboos near Chapel ton. A
female specimen of Hyla 1 ichenata was collected by Dr. Thomas Farr of the
Institute of Jamaica, near the entrace to St. Clair Cave, St. Catherine
Parish on June 19, 1965- During several days in captivity at the Institute
of Jamaica, this individual deposited a number of eggs. It showed no
interest in a cockroach which was offered as food. This individual pro
duced a skin secretion which became gum-like on the hands of those who
held it. Gosse (1851) recorded an instance where the skin secretion of this
species caused severe irritation to the human eye.

44
During my visits to Jamaica I heard Hy 1 a 1 ichenata as far
east as localities near Moneague, Ewarton and Lluidas Vale. These calls
could be heard for nearly half a mile and invariably came from wooded
slopes. This species did not venture out into flat, open valleys, as
did Hyla brunnea. In attempting to collect this species while it called
at night, 1 found the vocal individuals to be in trees, with one exception.
North of Mandeville, it seemed that one frog was cal ling from underground.
This observation, along with the presence of Dr. Farr's specimen near
a cave entrance, leads one to suspect that these large frogs may take
refuge in cave entrances and crevices in limestone, both of which are
present in quantity in the range of Hyla lichenata.
Dunn (1926) found four tadpoles of Hvla lichenata in a bromeliad
20 feet up in a small tree, in August, 1925- 1 collected two large
tadpoles of this species from a large bromeliad situated on the ground
under the shade of small trees. These were taken in Rose Valley on
October 25, 1965 These large specimens could be distinguished from
tadpoles of Hyla brunnea by their large size, relatively shorter and
more muscular tail, and slightly different mouth structure. They were
only slightly darker than Hyla brunnea tadpoles, not black as were those
collected by Dunn. Like other Jamaican hylid tadpoles, they feed upon
frog eggs. One of my two specimens underwent metamorphosis and died about
six days later. At the time of its death, it weighed 660 milligrams and
had a snout-vent length of 18.7 millimeters. This individual developed
green pigmentation, similar to that of Hyla brunnea, in its bones and
other tissues.

45
Hvla wIder i. -- This small, green species is said to be quite
common in bromeliads about Mandeville, Jamaica, "and appears to have
a fairly wide range in the central part of the island above 1,000 feet"
(Lynn, 1940). Dr. Albert Laessle (personal communication) found this
species common on Juan de Bolas mountain, from which there is a large
series of specimens in the collection of the Museum of the Institute of
Jamaica. I was unable to collect a specimen there although 1 heard a
call which may have been of this species. The call of this species is
a faint clicking sound, which becomes louder as the frog continues to
call. However, in localities near Moneague and Fishbrook, I traced similar
calls to frogs which appeared to be of the genus Eleutherodacty1 us.
Unfortunately, both of these frogs escaped. During August, 1965, I
collected three adults from the Cockpit Country four miles north of
Quickstep, and a tadpole near Moneague. Another individual was taken
from Crown Lands in the Cockpit Country during May, 1966.
During the present study, adults of this species were found in
large as well as small bromeliads. My experience was similar to Dunn's
(1926) in that this species appeared most often in open woods, usually
with a southerly exposure. While Dunn (1926) collected one Hyla wiIderi
tadpole for each seven of Hyla brunnea, only one Hyla w? Ideri was collect
ed during the present study, along with more than 500 Hyla brunnea
tadpoles. The tadpole was collected from a large bromeliad in one of
several trees in a pasture. It was not darker than the Hyla brunnea
tadpoles collected at the same time, but did appear to have a greater
density of melanophores. The hind limbs on this specimen did not have

46
green pigment. Tadpoles of Kyla wi1deri have been collected in March,
Apr!1 and August (Lynn, 1940).
The tadpole of Hyla wilderi has poorly developed gills,
similar to those of Hy1 a brunnea, and like other Jamaican tadpoles, its
body is depressed, and has no fin.
Dunn (1926) found that the specimens which Barbour (1910) con
sidered pale green young of Hyla brunnea were actually Hy1 a w1 lderi.
In fact, Dunn believed that the non-ossified head and green bones of
adult Hyla wilderi represented a neotenic condition because of its resemb
lance to the young of Hyla brunnea and related forms. He suggested that
Hyla wilderi, as well as Hyla lichenata and Hyla marianae arose from a
frog of the Hyla brunnea type through sympatric speciation, because of
differential growth rates.
Hyla marianae. -- Of the Jamaican hylids, least is known of
Hyla marianae. Dunn (1926) originally considered these yellowish-green
or greenish-brown frogs to be the young of Hyla brunnea, but then recognized
them to be of a different species, which he described. The range of this
species appears to be the most restricted of the Jamaican hylids. Most
of the specimens have been collected in or near the Cockpit Country of
west-central Jamaica; two specimens came from Hollymount, Mt. Diabolo,
further to the east (Lynn, 1940; Goin and Cooper, 1950). This area has
a combination of high elevation (above 400 meters), greater rainfall and
less disturbance of habitat than surrounding areas. Limestone cliffs
covered by lianas and hillsides covered with jagged, honeycomb limestone
usually form the steep sides of the "cockpits." Vegetation on these

47
hillsides varies from sparse, scrubby growth, to saplings which shade
an understory of terrestrial bromeliads, to large trees. The areas of
sparse vegetation usually are higher on the hills and have a more
southerly exposure than do the areas of denser vegetation. Dunn took
his frogs from "wild pines" in rather thick woods. 1 collected one male,
two females and four metamorphosing tadpoles of this species from small,
compact bromeliads along the border between a sunny clearing and a thick
growth of saplings on the slope above. Hyla wiIder? and Hyla brunnea
adults were collected from the same site.
Three of the four Hyla marianae tadpoles were in the same small
bromeliad, but between different sets of leaves. All of them were
engorged with eggs. The tadpole mouthparts had been lost and the mouths
of these tadpoles were quite wide. Although the adults of this species
(28-38 mm) were slightly more than half the length of Hyla brunnea
(56-57 mm) adults, the tadpoles of Hyla marianae appeared to be about
the same size as Hyla lichenata tadpoles at metamorphosis. Like the
tadpoles of Hyla brunnea, those of Hyla marianae were quite active and
tended to move upward at metamorphosis. The color of these tadpoles was
brown. The adults which 1 collected could change from orange to brown
and back. The bones of adults and tadpoles were orange with no trace
of green in the soft tissues.
Pond habitats
During the course of this study, a relatively large number of
frog species waie found breeding in bodies of standing water. In addition
to the bromeliad habitat already considered, such environments included:

48
ponds, pools, puddles and wheel ruts filled with water; roadside ditches
and drainage canals; flooded, grassy meadows; and cattail marshes.
These waters varied widely in size, amount of exposure to sun and wind,
and in water characteristics (Tables 4 and 5)- Temperatures at mid-day
ranged to 40 C. in shallow open waters, but generally were less than
35 C. when exposed to direct sunlight for only a few hours; where direct
sunlight did not reach the pools, temperatures remained under 30 C.
Oxygen concentrations varied from less than one part per million to ten
parts per million and appeared to be affected by air currents and
photosynthesis as well as temperature.
The pH of standing surface waters in the West Indies appeared
to average somewhat higher than the pH of water in bromeliads in the
region. This may be due to the fact that the surface strata of most of
the West Indies consist of limestone. At a locality near Juan de Bolas
in Jamaica (Table 4, Habitat 7), the pH of water in three bromeliads
without an obvious layer of dust on the leaves ranged from 5-5 to 6.5;
three bromeliads near the road had a heavy covering of dust on the
leaves and the pH of the water in the reservoirs was 7- Presumably the
dust from the limestone gravel of the road reduced the acidity of water
in bromeliads when it was washed into the central reservoirs by rain
or dew. Other water characteristics do not show definite patterns
(Tables 4 and 5)
It is difficult to make a distinction between exposed and
shaded pool habitats. There may be greater difficulty in assigning a
particular species to one or the other. Hyla dominicensis and Hyla

49
septentrona1is are quite opportunistic in regard to selection of breed
ing areas. However, in Surinam, several species appeared to be associated
almost exclusively with forest habitats. Included in this group are
Hyla qeograohica, Hyla calcarata, Hyla lanciformis, and Osteoceoha1 us
taunjTus, all of which are medium-sized and brown dorsal ly, and the
large species, Phyllomedusa bicolor, which is green above.
Hyla qeoqraphica. -- A si lent male of Hyla aeocraphica was
collected from vegetation on a rocky ledge of Princess Irene Falls,
Brownsberg, Surinam. Wooded slopes were present on both sides of the
falls and the stream which flowed over it. Tadpoles were present at the
base of the falls, but these were believed to be larvae of Hyla maxima,
which was calling at the locality. This specimen of Hyla qeoqraphica
had light green bones but no other green organs.
Hyla calca rata. -- The single female was collected from a bridge
over a small stream which flowed into a swamp. This individual had green
bones, but no other green pigmentation.
Hyla lanciformis. -- Most of the specimens of Hyla lanciformis
were collected in wooded areas. One was collected while calling near
the edge of the forest in the vicinity of temporary pools. No eggs or
tadpoles were found. Goin and Layne (1958) stated that its note was
regularly heard in open lands such as wet meadows near Leticia, Colombia.
They noted that the specimens collected were perched on bushes or other
plants one or two feet above ground. Four specimens from Surinam lacked
green pigmentation in bone or other tissues.

50
Osteocephalus taurinus. A female with green bones was taken
from the trunk of a tree on the upper coastal plain in Surinam. The tree
was beside a road through a we 11-developed forest; a stream and standing
puddle were nearby.
Bokermann (1965) has recently given the first description of
the breeding habits of Osteocepha1 us taurinus. Rainfall in the vicinity
of Marmelo, in forests of western Brazil, is strongly seasonal, but
reaches 2250 millimeters annually. Osteocepha1 us taurinus choruses were
heard at the time of the first rain in November. Two of the breeding
pools used by the species were in the forest; in one of these ponds, there
was breeding activity at 1:00 P.M. The surfaces of these ponds were
covered by eggs of this species which were 5 millimeters in diameter,
including the jelly envelopes. A temperature reading from one of them
indicated the water temperature to be 27 C., while the air temperature
was 34 C.
Of particular interest was a third breeding pond found by
Bokermann. This pool was in an open area and contained the dead crown
of a tree which made difficult the capture of frogs beneath the branches.
However, as the collectors approached the pool, the frogs followed their
habit of rapidly climbing upward to escape capture until they reached
the ends of the tree's bare branches. After eight minutes in this
exposed position, the first frog fell into the pool, motionless but not
dead. Although its skin had dried in several places, it recovered. Air
temperature in the sun was 46 C. and the water temperature was 32 C. at
several points of the pond. Two days later, evaporation had greatly

51
reduced the size of this pond. Water temperatures rose to 38 C. and
the embryos were dead.
Phyllomedusa bicolor. -- This is a large arboreal species which
appears to breed near stagnant water which may be in open areas with
some exposure to the sun. One adult male was collected from the surface
of a main road on the upper coastal plain of Surinam. Another was heard
calling from a site high up in a tree in the forest near Powakka. The
male which was collected had no green pigmentation in the tissues. As
with other members of this genus, this species moves very slowly, never
hopping, but walking on all fours. The whitish color of muscular tissue
in this individual may be an indication of poor vascularization or a
relative lack of myoglobin.
Phy1lomedusa eggs and tadpoles believed to be those of
Phyllomedusa bicolor were observed at the New York Zoological Society's
biological station at Simla, Trinidad. The large, yellow eggs were
attached to the upper side of a large shaded leaf which overhung a pool.
Water lilies and water hyacinths covered part of the surface of this small,
ornamental pool, which harbored the Phy1lomedusa tadpoles of various
sizes. The tadpoles were slaty-blue in color, much the same color as the
adult dorsum in preservative. They fed upon aquatic vegetation and were
observed coming to the surface. Lutz (195*0 stated that the larvae of
Brazilian species of Phy1lomedusa use their rudimentary lungs as hydro
static organs. Movement of these tadpoles is accomplished with a
minimum of effort. The main propulsive structure is the narrow, upturned
tip of the tail which vibrates constantly. This causes a slow, steady

52
movement of the larva; occasionally, the tadpole moves more rapidly
by using the entire tail to propel itself. Both the tadpole and the
adult of this species give the impression of being quite slow. The
assumption might be made that this indicates a low rate of metabolism,
but this was not measured in frog or larva. As in the adult, no green
pigmentation were found in the tadpole.
A relatively large number of frog species was found in exposed
habitats where there was standing water. Neither tadpoles nor adult frogs
were seen often in such habitats during daylight hours. Tadpoles gener
ally avoided bright sunlight by swimming under aquatic vegetation.
Elachistocleis ovale. -- In Surinam, this small microhylid was
found calling from shallow roadside ditches containing standing water.
The ditches had a moderate growth of grass and were generally exposed to
the sun. This species was found in Habitat 25 (Tables 3 and 4). Neither
eggs nor tadpoles were found. The three adults examined did not show
any indication of green tissues (Table 1).
Pseudis paradoxus. -- This highly aquatic medium-sized frog and
its exceptionally large tadpoles were found in roadside ditches and
drainage canals in Surinam. Water in these ditches usually exceeded
50 centimeters in depth, was semi-permanent and stagnant. Vegetation
in this habitat was varied and served as food for Pseudis tadpoles.
While the eggs of this species were not seen In Surinam, Gans (1956) found
that this species lays its eggs in a frothy mass which floats on ponds
in Trinidad. The black tadpoles grow to more than 15 centimeters in
total length, a third of which is body length. Body length of the largest

53
tadpoles approximates that of the adults. Mr. Walter Polder (personal
communication) observed that Pseudis tadpoles grow to this large size
during the course of a single rainy period of four to six months. Thus,
the growth rate of this species is phenomenally high. in comparison
with other species, the adults of which are the size of adult Pseud ? s,
metamorphosis of this tadpole is greatly delayed. The slow, deliberate
movements of this tadpole reflect its low metabolic rate (Table 9)
Although typical habitats (such as Habitats 23 and 24 of Tables 3 and 4)
of this species were visited during the day, neither tadpoles nor frogs
were seen. Both forms appeared to be more active at night, when they
could be seen floating nearly vertically in the water. Its dark pigment
ation, low metabolic rate, and perhaps decreased diurnal activity are
probably important adaptations of this tadpole to its exposed and warm
environment. The adult can swim and hop quite rapidly. Thousands of
individuals were seen hopping across the road on a rainy night toward
the end of the wet season (late July).
Both the tadpoles and the adults of this species have green
pigmentation in the tissues, but it is particularly pronounced in the
tadpoles. In the larger tadpoles, green pigment permeates the tissues
so that they appear almost as dark internally as externally. Development
of green pigmentation in this species does not appear to be related to
metamorphosis, since it is well developed before that time. Green
pigmentation in the adults appears limited to the we 11-caleified bones.
Their other internal tissues are usually light in color and the blood
is bright red. Thus, there is a marked difference between adult and tad
pole with regard to green pigmentation.

54
Sma1ler hvlas. -- ¡P Surinam, several species of small or
medium-sized hylids were found calling or breeding in open, grassy ponds
or flooded fields with a dense growth of weeds. Both of these situations
appeared to be temporary due to increased rainfall at that season
(July).
Two very small species, Hy la mi sera and Hyla nvnuta, as well
as Hyla boesemani, Hyla eg leri Hyla leucophy1 lata, Hyla rubra, Hyla
punctata, and the medium-sized Hvla crepitans, were included in this
group. Goin and Layne (1958) recorded Hyla rubra, Hyla mi sera, Hyla
punctata, Hyla leucophy1 lata, and Hyla lanciformis from similar habitats
near Leticia in southern Colombia.
While several of these species were found in the same pond in
Surinam, they often had different habits. Hyla misera usually called
from a perch on stalks of grass which were surrounded by water; Hyla
crepitans was found floating in the water near clumps of grass; and
Hyla boesemani called from the grasses at the edge of the same pond.
Within this group, Hyla rubra was most likely to be found in the vicinity
of Paramaribo, becoming less frequent at higher elevations to the south.
Hyla punctata was found perched on weeds in a flooded field and Hyla
leucophy1 lata called from bushes along a drainage canal.
Of the eight ...species collected in Surinam in this habitat,
Hyla egleri (one individual), Hy la leucophy1 lata (three), Hyla minuta
(two), and Hyla rubra (four) showed no green pigmentation. Hyla
boesemani (one), Hyla crepitans (two), and Hyla punctata (two) had green
bones and the latter two species had green plasma as well. The single
specimen of Hyla mi sera had light green plasma.

55
Phrvnohyas venulosa. --This is a medium-sized species which
has a well developed pattern of melanin pigmentation which largely
obscures the extensive internal green pigmentation of bones, soft tissues
and plasma. The poisonous secretion from the skin of this species is
well known; breaks in the skin make one particularly susceptible to
inflammation caused by contact with this secretion (Goin and Layne, 1958)
Phrvnohyas venulosa was heard or collected from swamps, shallow roadside
ditches near forests or thickets, and from residential areas in Surinam.
Tadpoles of this species were collected from standing water in roadside
ditches near wooded areas.
Zweifel (1964) described the life history of Phrynohyas
venulosa from Panama. The eggs were laid on the surface film of marshy
ponds. When brought into the laboratory, they hatched in less than a
day and metamorphosis took place after 37 days. The tadpole was of the
pond type, with globose body, lateral eyes, and well-developed tail fin.
The early larvae have we 11-developed external gills, which spread at the
surface as the tadpole hangs vertically, and conspicuous lungs are
present at a later larval stage. These two characters, along with the
flotation of eggs, were considered to be adaptations to low oxygen con
centrations of the environment. The temperature range in the natural
habitat was 25 33 C.
Phy1lomedusa hypochondri a iis. -- This medium-sized frog is
bright green on its dorsal surface in life. In preservation, this green
color fades to slaty-blue. There are no indications of a green pigment
in the skin or in other tissues of this species. Phy1lomedusa

hypochondria 1is was found on leaves of bushes and small trees at the
edge of wooded areas. The large, yellow eggs of Phy1lomedusa hypo
chondria 1 is are probably attached to leaves above standing water, into
which the larvae drop after hatching. Movements of these pond-type
tadpoles are rather slow.
Sphaenorhynchus aurantiacus. -- This small green hylid was
collected from two similar localities in Surinam. In both habitats,
the frogs called from cattails in standing water, 50 centimeters or more
in depth. In a locality near Domberg, the frogs were found at an inter
section of two drainage canals in which there was considerable aquatic
vegetation. Along these canals were some shrubs and small trees. The
second locality, south of Paramaribo, was a small cattail marsh, perhaps
30 meters in diameter and surrounded by forest except where it emptied
into a roadside canal. Goin (1957) noted that all of the known species
of this genus select areas with still waters as their breeding habitats,
and that they call from the water, or from floating or emergent vegetation.
Lutz (195^) stated that members of this genus lay eggs on leaves, but their
development seems to be unknown.
Goin (1957) specifically noted that certain members of this
genus have green bones (Table 1). From the description of color in
other structures, it appears that green pigmentation of tissues is a
general and striking characteristic of the genus Sphaenorhynchus.
Hy la dominicensis. Of the hylid species on Hispaniola, Hyla
dominicensis is certainly the most common and has the widest distribution.
It is very similar to Hyla brunnea, the most common and widespread of

57
Jamaican tree frogs, and to Hyia septentriona1is, the only tree frog of
Cuba and surrounding islands. in the Dominican Republic, Mertens
(1939) found Hyla dominicensis in all but the most arid habitats, from
near sea level to more than 1000 meters in elevation. Barbour (191
noted that the Museum of Comparative Zoology has many specimens from all
parts of the island and Cochran (19^1) noted that it is common in
co1lections.
During October, 1965, I collected an adult and tadpoles of
this species from Port-au-Prince and heard adults calling from trees in
Petionville (400 meters elevation) and Cap Haitien. Adults were heard
in May and August of 1966, most often during or after rain. Tadpoles
were also collected during these visits. The first large chorus of this
species which Mertens (1939) heard was on the evening of February 21 after
the first hard rain. He noted breeding activity until September 29 and
concluded that there is no definite breeding period. While Lynn (1958)
collected adults and tadpoles of Hyla dominicensis in Haiti during mid-
April, 1953j he did not hear them calling. He also noted that this frog
produces a skin secretion which is irritating to cuts and scratches, and
that the bones are green.
Mertens (1939) took note of the very rapid development of Hyla
dom?n?censis tadpoles. He collected a number of these from a street
puddle filled with rust-colored water on March 3- He stated that the
larger tadpoles had a covering of grass-green algae on the back and only
small hind limb buds, but developed into frogs in 10 days. Similarly,
larvae that he collected from a cistern on February 21 had only rudimentary

58
hind limbs, but were in metamorphosis on March 4. Mertens found Hyla
domin?censis in bromeliads in the coniferous forests, but did not
state that they bred there. Noble (1923) indicated that he found tad
poles of this species at 2500 meters elevation in wheel ruts. Lynn
(1958) found tadpoles of this species in a small pool at the head of a
stream.
I found this species only in stagnant water: in a shaded pool
in a dry stream bed, in a flooded wiregrass pasture exposed to the sun,
in wheel ruts of a muddy road which were also exposed to the sun, and
in concrete-1ined pits which had been used for tanning hides. In the
latter two habitats, the water contained so much solid material and algae
that the tadpoles could not be seen except when they came to the surface
to take air. The constant agitation of the water by the dozens or
hundreds of tadpoles appeared to be largely responsible for the turbidity.
When the flooded pasture was visited in May, a large cluster
of tadpoles was noted in one flooded corner. After tadpoles from the
cluster came to the surface, they ordinarily returned to the group.
This behavior continued until my presence caused the group to disperse.
It is of particular interest since the temperature of the water was
above 32 C. and the oxygen concentration averaged 0.7 parts per million.
Brattstrom (1962) has shown that such aggregations of dark tadpoles
absorb more radiant heat than do isolated individuals. He found water
temperatures near such groups to be higher than water temperatures at
a distance. He reasoned that the resultant higher body temperature
caused an increase in metabolic rate, which decreased the time required
for development.

59
After the group of tadpoles dispersed, the behavior of indi
viduals was noted. Most appeared to be resting near the bottom of the
flooded area while others appeared to be feeding along the bottom. At
intervals, they came rapidly to the surface to gulp air and then returned
to the bottom to rest or continue feeding. Four individuals which were
observed for five or more intervals had average interval lengths of 17,
22, 27, and 31 seconds. It appeared that the smaller and more active
ones surfaced more often. Several hours after some of these tadpoles
had been collected, their tails appeared quite red. When this habitat
was visited again in August, 1966, a flocculent precipitate covered
most of the surface and no tadpoles were present.
Young larvae of Hyla dominicensis are black, but they become
brown or gray and develop green pigmentation in bone and soft tissues
during metamorphosis. The black pigmentation of the young may be an
adaptation to an environment exposed to the sun, since this character
is seen in Hyla vasta and Hyla septentriona1?s in similarly exposed
environments, but not in the Jamaican hylid tadpoles which develop in
shaded habitats. Gills are well developed in this species of tadpole,
but they appear to lack lungs until metamorphosis. Therefore, the air
which they take at the surface must be kept in the mouth where gaseous
diffusion takes place, and then the bubble is released at or near the
surface prior to taking in a fresh gulp of air. The complex gill
structure probably is important in filtering food, primarily algae,
from the water. Like most vegetarian tadpoles, these have a long, coiled
intestine. While the tail of this species is not particularly muscular,

60
it is similar to that of stream tadpoles in lacking the fin extension
on the back.
Hyla septentriona1?s. Of the native West Indian hylids,
Hyla septentriona1is is the most widely distributed. This medium-sized
tree frog is found throughout Cuba, where it is most abundant in the banana
groves of the lowlands (Barbour and Ramsden, 1919). In addition it is
common and perhaps has been introduced into the Cayman Islands, the
Bahama Islands, the Florida Keys and southern Florida (Barbour, 1937).
Barbour (1931, 1937) believed that this species was introduced into
Key West on freight cars from Cuba. There are indications of subsequent
introductions and range extensions in south Florida (King and Krakauer,
1966). Grant (19^+0) summarized references which indicated that this
species was accidentally introduced into the Cayman Islands, but did
not maintain itself on Cayman Brae. However, William Greenhood (personal
communication) noted this species on Cayman Brae during the summer of
1965. Although there has been some question about its distribution on
Grand Cayman, I have found it there at two localitiesnear North Side
and just east of Boddentown--as well as in the Georgetown area.
Specimens have been seen from all of the major islands of the Bahamas
as far southeast as Acklin Island. A specimen of Hyla septentriona1?s
in the Institute of Jamaica was collected during the spring of 1965
at Highgate, St. Mary's Parish, Jamaica. This specimen resulted from
the introduction of tadpoles from Hialeah, Florida, during the spring
of 1961 (Edwin Todd, personal communication); there is no evidence that
this species has bred in Jamaica.

61
Thus, the Cuban tree frog has become established in a number
of localities through the activity of man, and its original range
outside of Cuba would be difficult to determine. Part of the adapt
ability of this species may be related to its accommodation to human
habitations; it is also found in mesophytic situations, but not in pine-
land or prairie habitats (Duellman and Schwartz, 1958; Barbour and Ramsden,
1919; Grant, 1940; Stejneger, 1905; Peterson, Garrell and Lantz, 1952;
Neill, 1958).
During the present study, habitats of the Cuban tree frog
were studied on Grand Cayman, on New Providence Island in the Bahamas,
in the vicinity of Miami, Florida, and near Highgate, St. Mary's Parish,
Jamaica. Cuba was not visited due to political restrictions. The two
days' stay on New Providence, during May, 1966, was unproductive due to
drought and a widespread lack of surface water.
In Miami, Florida, tadpoles of Hyla septentriona1?s were found
at a tropical fish hatchery and at the Serpentariurn, a commercial enter
prise concerned largely with snakes. Large larvae undergoing metamorphosis
were noted in concrete tanks in both places. In addition, thousands of
small tadpoles were found in the long, shallow, ornamental pool in front
of the Serpentarium (Table 3, Habitat 21). Since this pool had been
drained and cleaned just one week before, the tadpoles had hatched and
grown to 5 millimeters in body length in less than a week. Although the
water temperature of this pool was 33-9 C., the oxygen concentration was
9-2 parts per million. The pool had been drained to remove large amounts
of algae, but the water appeared clear at the time of measurement; also,

62
the sky was overcast during much of the day so that it appeared unlikely
that the high oxygen content was due to photosynthesis. At the time
of measurement, a strong breeze was blowing and may have been respon
sible for the presence of so much oxygen in the shallow pool. Although
the tadpoles were moderately active at the high temperature, they rose
to the surface infrequently, and then as often to feed from the surface
film as to take air.
The majority of my observations on Hyla septentriona1?s were
made during two visits to Grand CaymanJuly 5-8, 1965 and October 28-30,
1965. At the time of the July visit, most of the island had been without
rainfall for more than a month and was quite dry. Adult Hyla septentrio
nal i s were found only in the vicinity of Georgetown; they became quite
active following a shower on the evening of July 7- Tadpoles were not
found in any of the cow wells inspected, but a number were collected
from a small cistern which was well shaded. The cistern was located
behind an abandoned church 6.4 kilometers southeast of Georgetown and was
about 50 meters from the shoreline.
During the October visit, tadpoles were still present in this
locality. Considerable rain had fallen during the two weeks prior to
my visit so that a number of temporary pools of rain water had collected.
Tadpoles were collected from several small woodland pools near South Sound,
from a small shaded pond and from an open pond near North Side. The
woodland pond and pools were less than 20 centimeters deep; the open pond
was about 50 centimeters in depth.

63
The tadpoles and adults of Hyla septentriona1is are very
similar to those of Hyla dom?nicens?s; in fact, the two forms have been
considered races of the same species (Barbour, 1937) The tadpoles of
Hyla septentriona1is are black when young and become brown toward meta
morphosis. They are primarily vegetarian, but on one occasion I observed
a larger tadpole rapidly ingest a smaller one. The gills are well developed,
but lungs are not. Body shape is rounded, and, like that' of other hylid
tadpoles in the West Indies, does not bear an extension of the tail fin.
The rate of development is very rapid as previously mentioned. Grant
(1940) stated that C. Bernard Lewis found "that eggs laid in a tiny
puddle in coral or tree trunk during the night will be active tadpoles
by mid-day." English (1912) found that these frogs breed any time of the
year after a good rain, with eggs developing to frogs within a few weeks.
He also indicated that the tadpole and frog are enemies of mosquitos which
are abundant on Grand Cayman. This is probably true, since the smaller
bodies of water in which these tadpoles are found are ordinarily clear
and free of other obvious forms of animal life, such as mosquito larvae.
Hyla septentriona1is tadpoles have been found in brackish
pools on the Florida Keys (Neill, 1958). Larvae of this species often
have a clear envelope of fluid surrounding the body which is contained
within an outer layer of skin. This would seem to be an adaptation
which slows diffusion of water or salts between the tadpole body and its
environment. It should also reduce heat exchange. It is possible that
these tadpoles are capable of slight osmoregulation, since an individual
lived nearly four days in sea water diluted by distilled water to 10 parts
per thousand salinity. At 15 and 20 parts per thousand salinity, the

64
tadpoles died within three hours, but at 5 parts per thousand salinity,
a tadpole lived for ten days before it was returned to fresh water.
Since an isotonic solution for these tadpoles is presumably 7 parts
per thousand, the first tadpole was able to maintain itself in a hyper
tonic solution for near four days.
A number of preserved young Hyla septentriona1?s from Grand
Cayman showed green bone pigmentation, as did a number of the larger
specimens. Of more than 260 live adults collected from Miami, Florida,
during June, 1966, none showed green pigmentation in bones; however,
Clarence McCoy (personal communication) collected specimens with green
bones from Homestead, Florida, during August, 1967- I was unable to
determine more about the development of green pigmentation in this species
because it did not undergo metamorphosis in the laboratory. 1 found
the skin secretion of this species to be extremely irritating if rubbed
into the eyes; it is also difficult to wash from the eyes.
Stream habitats
Flowing streams constitute the primary habitat of several species
of tadpoles in the West Indies and South America. Temperature and
oxygen concentration within these streams apparently have a narrow range
of daily fluctuations. Streams inhabited by tadpoles were found to have
temperatures between 20 C. and 25 C. in most cases. Since these streams
were in hilly or mountainous areas, it seemed that the higher altitude
was related to the relatively low water temperatures. Quite regularly,
direct sunlight was blocked by clouds during the afternoon. These factors,

65
along with the shade provided by forest vegetation, probably account
for stream temperatures remaining below air temperatures, which are
also relatively low. Stream turbulence appears to keep oxygen concen
tration between 7 and 10 parts per million, but this is decreased where
there is little turbulence and the stream moves more slowly (Table 4,
Habitat 29). With a pH range of 6.0 8.6, the streams were similar to
other surface waters, but were less acidic than water which collected
in bromeliads. Dissolved solids demonstrated no definite trends.
Hyla maxima. -- This large, brownish species was found calling
at Princess Irene Falls, Brownsberg, Surinam. Individuals called from
tree sites 2 meters or more above ground. At least six individuals
called from trees around the perimeter of the falls for three successive
nights, even though there was no rain; none were heard elsewhere on the
small mountain. Tadpoles in the pools at the base of the falls were
believed to be of this species, as were eggs and approximately 50 small
tadpoles found in a shallow rock basin (3 x 25 x 40 centimeters). The
basin was in a rock ledge approximately 2 meters to one side of the main
falls and 60 centimeters above the base of the falls. It received a
steady supply of water from spray and droplets which fell into it.
Since tall trees nearby surrounded the base of the falls, the basin
remained in the shade except for about an hour at mid-day. Shortly
after sunrise (7:35 A.M.), temperature of the basin water was 23.8 C.
Near the end of the period of insolation (12:30 P.M.), the water temper
ature in the basin reached 29-9 C., while the air temperature was 28.1 C
Two hours later, the water temperature in the basin had dropped to 25.6 C

66
Thus, water temperatures in the basin ranged from 23.8 C. to 29-9 C.,
while the temperature ofthe water in a pool at the base of the falls
ranged from 23.4 C. to 25-5 C. Iron concentration in the basin was
about three times as great as in the pool (Table 5)- This may be import
ant to young tadpoles, since it was noted that young Hyla brunnea tadpoles
are more sensitive to iron deficiency than are the older ones. Concen
trations of phosphate and pH were similar for both the basin and the
stream pool. Oxygen concentration in the pool ranged from 8.3 to 8.9
parts per million; during the same period, the basin contained 2.0 to
3.1 parts per million of oxygen. Concentration of carbon dioxide in
the basin shortly after it had been in the sun was 24.0 parts per million;
at the same time, the concentration of carbon dioxide in the pool was
10.0 parts per million.
The eggs in the basin were of medium size--2 to 3 millimeters
in diameter, not including the jelly--and rested on the detritus at
the bottom of the basin. It was not known whether the eggs originally
floated on the water surface. The yolk of the eggs was white rather
than yellow and the animal hemisphere was darkly pigmented. Tadpoles
in the basin were observed feeding on the eggs, which filled their ali
mentary tracts. They also came to the surface to take air, but do not
appear to have lungs. When the water temperature was highest, these
tadpoles were observed coming to the surface at intervals of 10 to 25
seconds.
None of the tadpoles observed showed green pigmentation. How
ever, none had we 11-developed hind legs. The length of time required

67
for development was not determined. Three adult males were collected
from this locality and all had green bones, soft tissues and plasma
(Table 1) .
Dowling (i960) found the tadpoles of this species in the Arima
River of Trinidad. She found the jet black tadpolesabout 2 to 5
centimeters in length--swimming in schools. These dense aggregations
of tadpoles, as many as 177 in a group, did not keep to sun or shade.
Neither did the individuals in masses appear to feed; tadpoles observed
feeding were scattered in shallow water.
Hyla vasta. -- Noble (1923) found this giant tree frog in the
northern coastal mountains of Hispaniola and also at 2500 meters elevation
in the central range which runs east to west. Mertens (1939) also
found it in forested areas in the central range, but did not consider
it exclusively montane, since he collected the species at 220 meters in
one locality. It has been found by the present writer and others near
Furcy in southeastern Haiti at 1350 meters elevation. In this unforested
and agricultural area the adults take cover in crevices and under roots
and sod which overhang the stream bank. I also heard a small chorus in
a ravine near the Riviere Froid at an elevation of 200 meters.
The life history of the giant Hispaniolan tree frog was first
recorded by Noble (1923). In the rain-soaked northern mountains of the
Dominican Republic, he found Hyla vasta along the streams which flow
through the rain forest. He stated: "After the sun has set, the giant
tree frog, Hyla vasta, leaves his hiding place among the tree tops and
descends to some rocky ravine. There flattened out on a mossy boulder

68
in midstream, he rests for hours, seemingly enjoying the cool mists
which arise from the torrents. So closely does the frog resemble the
moss and lichen of his surroundings that he would rarely be observed
were it not for his big shiny eyes, which are conspicuous even when
closed, the lower eyelid being trans1ucent"(p. 56).
While Noble (1923) found that the skin secretion of this species
caused inflammation and irritation to his hands, I did not find this
to be true of the frogs collected from Furcy. Neither the boys who
caught them, nor I, after handling them, experienced any discomfort.
Noble (1927) described the early development and habits of
the young as follows: "Hyla vasta laid its eggs in little basins in the
gravel and stones on the edge of the pools in the mountain torrent
(one observation). Six days after hatching, the larvae made their way
out of one of the basins over wet stones into the torrent pool. As
they grew older they developed better stream lines than the tadpoles of
Hyla dominicensis. They were equipped with more rows of teeth. The
mouth was larger and better adapted to holding on to rocks in the stream.
Its tail was thicker and more muscular than that of the stagnant-pool
tadpole"( p. 108) .
He also reported that the eggs of Hyla vasta are pigmented and
are stuck to rocks at the bottom of the basin, and that they give rise
to dark tadpoles with small external gills, which do not rise to the
surface. The temperature of the stream he recorded as 21.7 25.6 C.,
averaging 23-7 C.

69
My observations near Furcy were similar, except that I was
unable to find a basin in which the young were developing. Half-grown
to metamorphosing larvae were found at the bottoms of pools. The mottled
gray pattern of the larger tadpoles makes them rather obvious against
the silt on the pool bottoms. Occasionally these larvae will move under
stones in the stream bed or attach themselves to rocks by means of their
large sucking mouths. The sucking mouth is important for stabilizing
the tadpole in the stream current and also for scraping plant food from
rock surfaces. When disturbed, they may allow themselves to be swept
downstream by the current.
Like Hyla dominicensis tadpoles, the larvae of Hyla vasta have
we 11-developed gills. In the stream habitat, these gills seem to serve
them well, since they were never observed taking air from the surface.
In the laboratory, six members of this species were kept in a graduated
cylinder which contained water with 0.5 parts per million of oxygen. At
first, the tadpoles rested at the bottom when not surfacing for air, but
after several hours they remained near the surface and were inactive
except for taking air.
Green pigmentation appears in the large tadpoles at about the
beginning of metamorphosis. The bones of metamorphosed young are green
and this color is present in some of the large adult individuals. Slightly
older frogs found among rocks along the stream also had green bones.
Hy la he?1prin ?. -- This medium-sized species of Hispaniola
was first seen by Noble (1923) as a brightly colored male which gave its
shriek while sitting on a rock in the middle of a torrent in the northern

70
mountains. Although he did not capture that individual, he obtained
tadpoles nearby. In cold streams above 2500 meters in the central
regions he heard again the loud call of Hyla hei1prini above the roar
of the cascades. While the average temperature in the first stream was
24.7 C., it was 12.6 C. in the second locality at the higher elevation.
Noble found the tadpole of Hyla heilprini to be more streamlined than
the previous two species and considered this an adaptation to living in
torrents, rather than in slower moving streams or puddles. Since these
larvae were obviously adapted to an environment with a high oxygen con
centration, Noble had doubts about being able to maintain them in captivi.ty,
but he was successful in raising them through metamorphosis.
From Noble's (1923) observations in the Dominican Republic,
specimens from the Museum of Comparative Zoology from several localities
along the southern peninsula of Haiti, and personal observations, it
appears that Hyla heilprini is widespread on Hispaniola. I believe that
1 heard the species just west of Port-au-Prince in a ravine near the
Riviere Froid, less than 200 meters above sea level. It would seem that
wooded ravines are more important to this species than are high elevations.
Cascading streams probably are necessary for its reproduction, as both
Noble (1923) and I found its tadpoles only near waterfalls.
Green pigmentation was present in all of the dozen or more
tadpoles which I collected. These were we 11-developed tadpoles, with
either two or four legs, which were olive green with black spots dorsally,
blue-green and white ventrally, and the distal portions of the limbs were
orange in color. Since only the distal portions of the limbs were orange,

71
one wonders if this was the result of extreme effects of temperature
or oxygen upon a pigment which was more generally distributed through
the body. An adult of this species collected from a banana stalk was
bright grass-green on the dorsal surface with white and blue ventrally
and blue and yellow on the flanks. Later the back became quite dark.
This specimen had green bones and green plasma. Noble (1923) described
the dorsal color of a living specimen as "golden green."
Hyla pulehr? 1 ?neata. -- Except for its light-colored longi
tudinal stripes, lack of an interocular bar, and larger size, this small
green tree frog of Hispaniola is quite like Hyla wilderi of Jamaica.
It has been collected from widely scattered points on the island and
Cochran (1941) noted that it seems to be common in certain localities.
From the records of Noble (1923), Cochran (1941), and Lynn (1958), it
seems likely that this species has a distribution similar to those of
Hyla vasta and Hyla he?Iprini: in forests at higher altitudes and in
wooded ravines at lower altitudes. This species appears to be the least
common of the four hylids of Hispaniola; I did not encounter adults or
tadpoles during my visits to Haiti. Noble (1923) and Lynn (1958) found
adults calling from leaves in the vicinity of streams. Lynn (1958)
described its call as a faint clicking like a small telegraph instrument,
similar to that of Hyla wiIderi.
Noble (1927) stated that Hyla pulehr i 1?neata young develop in
water as do the other hylas on the island, but apparently he has given no
further details. There are no known specimens of Hyla pulehr i 1 ineata
tadpoles in the major museums of eastern North America (including the

72
American Museum of Natural History where Noble deposited most of his
specimens). Dr. Lynn and Dr. Albert Schwartz (personal communications),
both of whom have collected in this area, are unaware of any tadpoles
of this species which have been collected. It seems best to consider
Noble's statement as a supposition on his part.
The color of the bones of this species has not been recorded.
Bufo mar inus. This very large toad was found breeding in
several habitats including temporary pools in Surinam and near drainage
canals and tropical fish hatcheries in Miami, Florida. it has been
introduced into Jamaica, where it appears to be the only anuran species
which breeds in ponds or slow-moving portions of streams. Neither young,
which are vegetarian, nor adults are known to have green tissues under
normal circumstances.
Leptodacty1 us albilabrus. -- Tadpoles of this Puerto Rican
species were collected from a roadside stream on El Yunke, the highest
mountain on the island. Water flowed rapidly in the stream, thus account
ing for its high oxygen concentration (Table 4). Both large and small
tadpoles of this species came to the surface, evidently to take air,
but their lungs do not appear to be developed at these stages. This
species feeds on vegetation and does not develop green pigmentation as
it undergoes metamorphosis.
Resume of Ecology and Breeding Habits
Considering the great variety of aquatic habitats available
in the Neotropical Region, it seems that anurans have taken advantage
of most of them. The species with green pigmentation are well

73
distributed throughout the Neotropical Region, but appear to be restricted
to certain habitats or families of frogs. As Barrio (1965a) noted,
green pigmentation is found in only three families of South American
frogs. Hylids and centrolenids are primarily arboreal following
metamorphosis, the period of life when they appear most likely to have
green pigmentation. Pseudids are generally considered to be highly
aquatic as adults. Although 1 found relatively little green pigment
in some adult Pseudis paradoxus paradoxus in Surinam, Barrio (1965a)
found high concentrations of biliverdin in Pseudis paradoxus platens is,
Lysapsus lime 11 us and Lysapsus mantidacty1 us adu1ts. While green
pigmentation appears in tree frog tadpoles with metamorphosis, it
preceded metamorphosis by an extensive period in Pseudis paradoxus
paradoxus. Thus, the pseudids differ from hylids and centrolenids in
both the type of habitat and in the time of onset of pigment development.
It seems logical to assume that the hyperbi1 iverdinemia of pseudids has
a cause which is different from that of the two tree frog families. In
addition, the point should be made that terrestrial and fossorial anurans
toads largely--have not been reported to have biliverdin in their tissues.
Since distributional and ecological factors have not yielded
a satisfactory explanation for the accumulation of green pigment, it
appears logical to turn to physiological studies.

111. PHYSIOLOGICAL FACTORS RELATED TO CHLOROSIS
Previous studies have shown that the immediate cause of
chlorosis is the production of biliverdin in quantities greater than
those which are excreted. It has not yet been determined if chlorosis
in frogs is due to a high rate of biliverdin formation, impairment of
bile pigment excretion, or both.
Since hemoglobin of the blood is ordinarily the main source
of biliverdin, it is logical that one should study this protein and the
red blood cells in which it is found. If chlorosis is due primarily to
a high rate of biliverdin formation, one might expect to find higher
concentrations of hemoglobin or higher rates of hemolysis in chlorotic
frogs than in others. In addition to chlorosis itself, high rates of
hemolysis might be indicated by low red cell counts, or by blood smears
which show immature erythrocytes or cell fragments. Should hemolysis
or other process be shown to be the cause of increased biliverdin, then
the cause of such a process should be determined. In the case of rates
of hemolysis, it is known that they may be correlated with temperature,
concentrations of chemical agents, or differences in red cell structure.
Impairment of bile pigment excretion would involve the liver
and its associated ducts. Blockage of extra-hepatic ducts is readily
determined by the absence of biliverdin from feces. Blockage or
restriction of bile flow within the liver may be difficult to detect,
74

75
but histological study ordinarily shows the degeneration of liver which
is associated with prolonged cholestasis.
Several factors indicate that the study of tadpole respiration
may shed light upon chlorosis. First of all, the present study has
shown that green pigmentation first appears in the tadpole stage of
several species (Table 2). Secondly, the low oxygen tensions of some
tadpole environments suggest that the larval stages have special respir
atory adaptations or that tissue damage may result from oxygen deficiaicy.
The degradation of respiratory pigmentsapparently the immediate cause
of chlorosissuggests a disruption or decrease of oxygen transport
within the organism. Since the liver is quite sensitive to oxygen
deficiency, one wonders if the preceding factors are concerned with
possible liver malfunctions. Finally one should consider the high environ
mental temperatures to which these organisms are exposed, and the greater
stress which this factor places upon respiratory and hepatic systems.
In order to determine the importance of biliverdin formation
to chlorosis, several attempts were made to induce green pigmentation
in vivo, using non-chlorotic species. This was followed by a broader
physiological study of Neotropical frog speciesboth chlorotic and non-
chloroticto clarify the importance of biliverdin formation, liver
function, and other factors which might pertain to the green pigmentation.
Attempts to Reproduce Green Pigmentation In Vivo
During the early part of the study, attempts were made to dupli
cate the high concentration of tissue biliverdin, particularly as it
occurred in bone. These experiments were carried out upon species which

76
were readily available in Florida, none of which had green bones. Four
methods were utilized in attempts to increase the concentrations of
biliverdin in the experimental animals. Phenylhydrazine hydrochloride
was used to destroy the red blood cells in the frog, thus causing its
hemoglobin to be degraded into biliverdin more rapidly. The second
method was the injection of additional hemoglobin, presumably of mammalian
origin, again to increase the rate of biliverdin formation. Third was
the direct injection of biliverdin into the animal. The fourth method
was to induce an iron-deficiency anemia, such as that known to cause
chlorosis in man.
Pheny 1 hydrazine hydrochloride administration
Since biliverdin formation was induced in Bufo arena rum by
means of phenylhydrazine injection (Cabello, 1943), this method was used
in an attempt to develop green bones in species which were known not
to have green bones.
Two adult specimens of Bufo terrestris, along with ten Hyla
cinerea were given intraperitonea1 injections of phenylhydrazine hydro
chloride in water in October 28, 1964. The dose was at the rate of
1 microgram of phenylhydrazine hydrochloride per gram of body weight of
the frog. On November 19, 1964, each of these frogs was given a second
dose of 2 micrograms of phenylhydrazine hydrochloride per gram of body
weight. During the period of the experiment these frogs were fed twice
a week. The animals were observed daily until December 9-
Injection of phenylhydrazine hydrochloride caused destruction
of red blood cells and formation of biliverdin, but like the other

77
methods.used, it did not induce detectable hyperbi1iverdinemia. Severe
anemia developed in specimens of Hyla cinerea as a result of injection
of this drug. Two smaller individuals died several days after the
second injection. Post-mortem examination showed that very few red
cells were present in the blood vessels or elsewhere. There was slight
indication of hemopoietic activity in the long bones; the spleen was
pale and showed no increase in size. The kidneys were slight yellow in
color, rather than the usual red. The frogs appeared emaciated. The
liver was normal in the larger frog examined; the smaller frog had an
enlarged gall bladder filled with green fluid and its liver was smaller
than expected. None of the frogs and toads used in this experiment showed
any evidence of green pigmentation in bone, blood or other tissues outside
of the digestive tract.
Administration of hemoglobin solution
A second attempt to increase the amount of biliverdn in
living anurans involved the intraperitonea1 injection of a 1 per cent
hemoglobin solution. Specimens of Hyla septentriona1?s (eleven individuals),
Rana pipiens (four individuals), Rana catesbeiana (two individuals),
Hyla gratiosa (one individual), Bufo terrestris (two individuals), and
Hyla cinerea (seven individual^ were injected with either 1 per cent
hemoglobin in phosphate buffer, or with phosphate buffer (pH 7-^) alone,
or were given no treatment. Fluids voided by these animals were taken
from the jars in which they were kept and analyzed with the Beckman DU
spectrophotometer.

78
Intraperitonea1 injection of hemoglobin solution resulted in
no clear indication of green tissues in any of the species. Several
individuals voided substances which had absorption peaks that corres
ponded to biliverdin or hemoglobin. There appeared to be no addition of
green pigment to the tissues. Judging from the color of the fluids
voided by these animals, bile pigments were formedand eliminated within
twenty-four hours after injection of hemoglobin. However, it should
be pointed out that green pigment was not excreted by a 11 of the frogs
which were injected with hemoglobin. In addition, one untreated specimen
and one injected with only phosphate buffer demonstrated green pigment
in voided fluids. This was also true of other untreated frogs in the
laboratory.
Injection of biliverdin
Still a third method was usedthe direct injection of
biliverdin in sesame oil or phosphate buffer (pH 7-^0 On one occasion,
an attempt was made to mobilize bone calcium prior to injection of
biliverdin. Several specimens of Hyla cinerea were injected with 0.5
(.005 cc) units of Lilly's Parathyroid Extract per gram of body weight.
The same dosage was repeated twelve hours later. After another six
hours, an intraperitonea1 injection of biliverdin in sesame oil was
administered. Finally, they were given two drops of cod liver oil after
an hour. The latter step was intended to stabilize calcium in bone because
of its vitamin D content (Cantarow and Schepartz, 1957).
Biliverdin solutions injected into the peritoneal cavity
caused no apparent change in the color or other visual characteristics

79
of the tissues. This treatment was discontinued after a relatively
short period of time because it di d not appear practical to continue.
Attempts to decalcify bone or otherwise change its pigmentation in vivo
were unsuccessful.
Induction of iron deficiency anemia in tadpoles
Iron deficiency is the most common cause of anemia in man and
may become pronounced during pregnancy or during periods of growth. In
one of its more serious forms (chlorosis) iron deficiency anemia causes
the skin to become green. During the nineteenth century, this was found
regularly among girls and young women, particularly those with menstrual
disorders.
Little is known about iron metabolism of frogs. Iron is
absorbed in the duodenal region of the intestine and incorporated into
eggs and hemoglobin (Brown, 1964). However, it seemed possible that a
lack of iron could be a cause of green pigmentation In frogs, as it is
in man. The main difficulty in running an experiment to test this idea
in frogs or most tadpoles is that their intake of iron cannot be regulated
easily. Live insects constitute the normal diet of frogs, and most tad
poles feed upon plant life and other organisms. Tadpoles of Hyla
brunnea are different in this respect, since they feed upon frog
eggs, usually those of their own species. These tadpoles develop normally
when they are fed Gerber's strained egg yolk, and kept in spring water.
The strained egg yolks were known to contain 3-07 milligrams of iron
per hundred grams (Gerber Products Company, 1965). It was thought that
the high concentration of phosphates in egg yolk (290 milligrams of

80
phosphorus per hundred grams) would decrease iron absorption, due to
the poor absorption of iron complexed with phosphates (Cantarow and
Schepartz, 1957)- The animals were kept in glass-distilled water,
rather than spring water, in order to reduce the iron concentration
further.
Two matched groups of twelve tadpoles were placed in small
glass bowls which had been acid-cleaned, rinsed and filled with glass-
distilled water. During the course of the experiment, the water was
changed twice a week and fresh egg yolk was added immediately thereafter.
An excess of food was always available to both groups. The only differ
ence in treatment between the two groups was that ferrous sulfate
(reagent grade) solution was added to the water of the first group.
Individuals of the first group appeared to develop normally, as did the
four larger individuals of the second group. Of the eight smaller tad
poles of the second group, all became pale yellow in color and five were
dead after two weeks. At this time, ferrous sulfate was added to the
water of the second group. After three days, the second group of tad
poles became the same light brown color as those of the first group.
The darker pigmentation appeared to be the result of an increase in hemo
globin. These observations were taken to indicate that limitation of
iron intake of rapidly growing tadpoles resulted in an anemia, which was
relieved by addition of more iron. It is also possible that trace
elements such as copper or cobalt may have been responsible for hemoglobin
formation in both groups. In either case, the anemia did not cause develop
ment of green pigment in the skin of any of the tadpoles.

81
Physiological Characteristics of Chlorotic and
Non-Chlorotic Frogs
This portion of the study was directed toward finding physio
logical differences between chlorotic and non-chlorotic frogs. Frog
blood, tadpole respiratory rates, and histological sections of liver
were the main subjects of study.
Methods of study
Hemoglobin determinations. -- The amounts of hemoglobin pre
sent in anuran blood was determined by the acid-hematin test (Cohen and
Smith, 1919). A Heilige hemoglobinometer was used for this purpose,
since much of the work was done under field conditions.
Blood was collected from a large blood vessel or the heart
into heparinized capillary tubes. From these, 20 microliter samples of
blood were transferred to a graduated test tube which contained j milli
liter of 1 per cent hydrochloric acid solution. After mixing the
contents of the tube, they were allowed to stand for ten minutes in
order to insure complete hemolysis of the red blood cells. Then distilled
water was added until the color of the solution matched the color stand
ard of the hemoglobinometer. The calibration mark at the meniscus of
the fluid indicated the hemoglobin concentration of the blood in grams
per hundred milliliters.
Red blood cell counts. Blood samples obtained as for hemo
globin determinations were diluted and counted by standard techniques
using a Spencer Bright Line hemocytometer. Four counts were made for
each individual and the average was used as the red blood cell count.

82
Red blood cell measurements. -- An optical micrometer was
used to measure length and width of twenty-five red blood cells from each
individual. Measurements were made from smears prepared and stained
in the field. Average cell length and width were determined and con
verted to microns. These blood smears were used to study abnormal cell
types in the blood.
Effects of temperature. In order to determine the effect of
temperature upon frog erythrocytes, blood from specimens of Hyla
squirrella, Bufo terrestris and Scaphiopus holbrooki was collected in
heparinized capillary tubes. The tubes were plugged with clay at both
ends and kept in a water bath at 45 C. for varying periods of time. After
this treatment, smears of the blood were made, stained and studied.
Measurement of tadpole respiratory rates
Measurement of respiratory rates of tadpoles was carried out
by two different methods. During the earlier part of the work, oxygen
uptake was determined by use of Warburg constant volume respirometers
as described by Umbreit, Burriss and Stauffer (1964).
The second method was improvised in order to determine the rate
of oxygen utilization by tadpoles shortly after capture, under either
controlled or field conditions. Necessary apparatus for field measure
ments were Erlenmeyer flasks with rubber stoppers, a Precision Scientific
portable oxygen ana lyzer, a-nd a wristwatch. The simple technique con
sisted of the following steps:
1. Clean 125 ml Erlenmeyer flasks were calibrated to determine
their actual volume when stoppered.
2. The appropriate number of clean Erlenmeyer flasks were filled
with clean tap water of the desired temperature.

83
3. The probe of the oxygen analyzer was calibrated in air.
After dipping the tip of the probe into clean water,
it was waved in the air while two ammeter readings were
taken. The ammeter of the oxygen analyzer measures electri
cal current between silver and lead electrodes located in
the tip of probe; this current is directly proportional to
the oxygen which diffuses through the covering membrane and
into the space between the electrodes. Air temperature was
determined at the same time by use of a thermistor attached
to the probe. These data were used to calculate the sen
sitivity of the probe under prevailing conditions.
4. The probe of the oxygen analyzer (which had been taped to
fit snugly into the neck of the Erlenmeyer flask) was
lowered carefully into the neck of the flask until the probe
formed a watertight seal with the neck of the flask.
After this was accomplished, the tip of the probe projected
slightly below the neck of the flask. When done properly,
no bubbles were trapped in the flask.
5. While holding the flask and probe together, they were
rotated vertically through a 30 degree arc, so that their
axis was in a horizontal plane. From this position, the
flask was shaken from side to side in order to cause water
to flow past the tip of the probe. This procedure was
necessary to improve the accuracy of the method. In the
laboratory, a magnetic stirrer was used to cause a flow
across the tip of the probe.
6. Once the indicator of the ammeter became steady, a reading
was made. The ammeter was then turned off for a short time
and then turned on again. While the shaking continued a
second reading was made and the average of the two readings
was recorded. Probe sensitivity, ammeter reading and water
temperature was used to calculate the initial oxygen con
centrad on.
7. Once the data necessary to determine the initial oxygen
concentration were obtained, a single tadpole was intro
duced into the flask, and the flask was stoppered. If an
air bubble was trapped within the flask, sufficient water
from the original source was added so that the flask was
completely filled with water when stoppered. This was
necessary to prevent intake of air by the tadpoles and also
to prevent immeasurable exchange of gases between air and
water. Note was made of the time that the flask was
stoppered.

84
8. After an appropriate length of time, the flask was opened
and the final oxygen concentration determined in the same
manner as the initial oxygen concentration (steps 3 through
6)
9- The tadpoles were fixed in formalin and each individual was
labeled so that their weights could be measured upon return
to the laboratory. Since formalin preserved tadpoles were
found to lose nearly 10 per cent of their original weight,
the weight of each preserved tadpole was multipled by the
conversion factor 1.094 to obtain a more accurate estimate
of the original wet weight.
10. Since the difference between initial and final oxygen con
centrations was given in mg/1, it was necessary to make
additional calculations to obtain the respiratory rates in
terms of pi 02/g/hr. This was achieved by use of the
following equation:
Respiratory rate (pi 02/g/hr)
_ Change in O7 concentration (mq/l). Water volume (pi)
1.429 (conversion to ml of O2). Tadpole weight(g). Time (hrs)
Although the error incurred in this method approached 3 per cent
or even 5 per cent under some circumstances, the necessary equipment was
not nearly so cumbersome to transport as manometers and pipettes; clean
ing was not a problem, nor was the equipment as likely to be damaged in
transit. In addition, if one desired to maintain the animals at a specific
temperature during an experiment, a wash basin or large pan could hold a
dozen or more flasks at one time and the temperature of the water In the
bath could be maintained within Io C. of the desired temperature by
addition of ice or warmer water, as required; hence, no special bath,
refrigerator or heater was required. In addition to technical advantages
of the method, it allowed measurement of oxygen uptake under actual field
conditions, although this opportunity was not realized. Since it was
possible to make measurements shortly after the tadpoles were collected,

85
there was little time for them to become acclimated to different environ
mental conditions. With modifications in the size of the flask and
diameter of probe, larger organisms could be studied.
Aside from the lower accuracy of the measurement of oxygen
change, disadvantages in relation to manometric techniques included:
slightly more variation in temperature due to lack of automatic controls
and constant stirring; the practical impossibility of beginning and end
ing an entire series simultaneously, or of taking readings at regular
intervals during the experiment.
This technique was developed in the laboratory at Gainesville,
where it was used to measure oxygen uptake of Hy1 a domin ? censis, Hyla
brunnea, Hyla vasta, and Leptodacty 1 us albilabrus maintained at 25 C.
in a water bath. Respiratory rates of Hyla maxima and Pseud is paradoxus
were studied at 25 C. in Surinam; similarly oxygen uptake of Hyla vasta,
Hyla dominicensis and Hyla hei1prin? were determined at 25 C. in Haiti.
Liver tissue sections. -- Blocks of frog liver tissue were
fixed in Bouin's fluid and then changed to 50 per cent ethanol for stor
age and transport back to the laboratory. These tissues were embedded,
sectioned and stained with hematoxylin and eosin by standard techniques.
Resu1ts
Hemoglobin concentrations and red cell counts. -- The sma11
green-boned species, Hyla punctata and Sphaenorhynchus aurantiacus, had
the lowest concentrations of hemoglobin of the species studied. Another
small form with green tissues, Hy la wi lderi, along with Sphaenorhynchus
aurantiacus, had the lowest red blood cell counts. Hyla punctata had the

86
lowest concentration of hemoglobin per cell, along with an individual
of Phrynohyas venulosa. Yet species with maximal values in these cate
gories also have green tissues. These include Hyla maxima and Phrynohyas
venulosa. However, the species which has the largest amount of hemoglobin
per cell, Phyllomedusa hypochondria 1is, does not have green tissues. This
is evidently a reflection of the large size of the red blood cells in
this species (Table 7)- There appears to be no correlation between the
blood characteristics studied and the presence of green pigmentation.
it should be noted that heparinized blood of Pseudis paradoxus
tadpoles and Sphaenorhynchus aurantiacus males completely hemolyzed
after being kept in their own blood plasma at room temperature overnight.
This was not noted in other species with green tissues or in those without
green tissues.
Red cell size. -- From Table 7 it can be seen that cell lengths
range from 12.8 to 21.6 microns. Cell widths vary from 9-9 to 14.5
microns. To obtain an approximation of erythrocyte volume for each species,
the average cell length was multiplied by the average width. The largest
cells are those of Phyllomedusa hypochondria 1 is (20.6 x 14.5 microns)
and Sphaenorhynchus aurantiacus (21.6 x 13-9 microns). Smallest cells
were those of Pseudis paradoxus (13-2 x 12.0 microns), Phrynohyas venulosa
(12.8 x 10.8 microns) and Osteocepha1 us taurinus (13-1 x 10.8 microns).
The most variable cells were those of Hyla punctata. All of these species
except Phyllomedusa hypochondria 1is are chlorotic.
Examination of blood smears. -- Blood smears stained with
Wright's stain revealed certain characteristics which may or may not be

87
related to green pigmentation (Figures 12-16). A list of these and
some others, according to the organisms in which they were observed,
fo 11ows:
Pseudidae
Pseudis paradoxus -- adult male Cells of this individual
tended to form cytoplasmic blebs; tadpole blood hemolyzed
upon standing overnight.
Bufnidae
Bufo typhonius -- adult an erythroplastid was observed
in the blood of this individual.
Atelopodidae
Atelopus sp. -- male erythrocytes showed coagulated
cytop 1 asm.
Hy 1 idae
Hyla boeseman? adult cells have coagulated cytoplasm,
as occurs after slight heating.
Hyla crepitan s -- adult most cells hemolyzed.
Hyla egleri -- male cells not unusual.
Hyla qeographica -- adult male nothing unusual.
Hy la lane? formis adult most of the cells were hemolyzed.
Hy la leucophy1 lata -- about half of erythrocytes have
coagulated cytoplasm.
Hyla minuta adult most of the cells have coagulated
cytoplasm as occurs on heating.
Hyla mi sera -- some cells have odd shapes.
Hyla maxima -- adult many lymphocytes present.
Hyla punctata -- adult cell membranes are irregular; green
cytoplasmic and extracellular granules present on stained
slides. Erythrocytes appear immature.
Hyla rubra adult nothing unusual.
Osteocepha1 us taurinus -- adult many erythrocytes with
vacuo la r nuc lei.

88
Phrvnohvas venulosa -- red cell cytoplasm coagulated.
Phyllomedusa bicolor -- appears to have many leukocytes.
Phyllomedusa hypochondr1 a 1 is -- erythrocytes are basophilic.
Sphaenorhynchus aurantiacus -- cytoplasm of erythrocytes
appears vacuolated, with greenish refractile material in
vacuoles (stained slide). Blood of three males hemolyzed
overnight.
Leptodacty1idae
Leptodacty1 us pentadacty1 us some red cel Is show coagulated
cytoplasm.
Effect of temperature upon frog red cells. -- Slight changes in
a few cells were noted after ten minutes at 45 C. After four hours at
45 C., cytoplasm had coagulated in nearly all of the red cells of Hyla
squirre 1 la and Bufo terrestris (Figures 9-l0 ; the red cel Is of Scaphiopus
holbrooki were abnormally shaped after this time, but the cytoplasm had
not coagulated.
Respiratory rates of tadpoles. -- The respiratory rates of tad
poles and froglets of green boned species (Tables 8-10) are similar to
those previously recorded for amphibian larvae (Hopkins and Handford,
1943). The basis of recording respiratory rate in the study was the uptake
of microliters of oxygen per gram of wet weight of tadpole per hour. In
order to convert this to an approximate dry weight basis, the rate of
oxygen uptake should be multiplied by eight. Size range of the tadpoles
in the present study is much greater than in previous ones, but it can be
seen that respiratory rate is a function of tadpole size and that the
species is not very important in this respect (Figure 8). As is to be
expected, respiratory rates increase with increasing temperature (Figure 7)-

39
It should be noted that the respiratory rate of Hyla brunnea
tends to drop sharply between 35 and 40 C. (Table 8, Group A), whereas
the respiratory rate of Hyla septentriona1?s does not (Table 8, Group C).
When the Hyla septentriona1is tadpoles were removed from the flasks, they
appeared to be in good condition (heart rate was about three beats per
second). On the other hand, these Hyla brunnea tadpoles were limp after
two hours at 40 C., and one did not recover. This is an indication that
the critical thermal maximum of Hyla brunnea larvae is lower than that
of the Hyla septentriona1is tadpoles. This might be expected when one
considers the higher temperatures to which Hyla septentriona 1 is tadpoles
are often exposed (Table 4). These higher temperatures are largely due
to the lower elevation and greater insolation of Hyla septentr?ona1?s
habitats. Hyla brunnea is more often found in shaded areas and at
greater elevations than are available to Hyla septentriona1is over most
of its range.
Along these lines, it may. be noted that the frog let of Hyla
lichenata appeared in poor condition after being kept at 35 C. for less
than two hours (Table 10). It would appear that Hyla lichenata is more
sensitive to this temperature than is Hyla brunnea at the same stage.
This may account for the fact that while Hyla brunnea is found throughout
the range of Hyla lichenata, the former species extends its range to lower
elevations and more exposed habitats than the latter.
Tadpoles of Hyla brunnea kept in the laboratory were exposed
to low water temperatures during the early morning of November 30, 1965.
This caused some mortality and the surviving tadpoles appeared pale,

90
similar to those which had iron deficiency. Very little movement of
the tadpoles was evident when the water temperature in their containers
was 16.8 C. Lynn (19^0) suggested that the lack of large bromeliads
above 1600 meters in the Blue Mountains may be the reason for Hyla brunnea
not being found at such altitudes. Low temperatures at these altitudes may
also be a factor which prevents this species from moving to higher
elevations.
Liver histology. -- Study of liver sections of ten species of
frogs from Surinam is of interest (Figures 17_20). No dead or completely
degenerated tissue was seen in these slides but there were several indi
cations of conditions or activities which may cause changes in bile
excretion. There did not appear to be any obstructions to the bile
duct or gall bladder in the majority of individuals examined, but those
with green tissues and plasma usually had very concentrated bile in the
gall bladder, judging from its dark blue color. A resume of liver histo
logical characteristics follows:
Pseudidae
Pseudis paradoxus -- adult male liver apparently normal
except for very heavy pigmentation (the liver appeared black
before dissection);
-- adult female as in male, with possible
cell fragments in blood of liver;
-- tadpole liver structure apparently nor
mal; much less pigmentation than in adults; cell fragments
in blood of the liver; many immature erythrocytes present in
circu lating blood.
Atelopodidae
Atelopus sp. -- male some liver cells appear enlarged and
have rarified cytoplasm; some pyknotic nuclei present; other
wise apparently normal.
Hy 1 idae
Hyla geographica -- adult male veins of liver partly
occluded; considerable pigment present in liver cells.

91
Hvla leucophv1 lata adult female liver apparently
normal; contains many erythrocytes.
Hvla maxima -- adult male fat deposition taking place in
one area; erythrocytes appear to be undergoing disintegration
in liver; relatively little pigment present.
Hyla minuta -- adult male some irregularly shaped nuclei;
considerable pigment present.
Hyla punctata adult male many nuclei present in liver
tissue; these are quite variable in size, shape and stain
ing characteristics; some may be pyknotic; very little
pigment is present in the liver; many unusual cells are
present in blood, perhaps immature erythrocytes.
Osteocepha1 us taurinus -- adult female liver pigmentation
less than normal; odd-shaped nuclei present in liver.
Disease as a Cause of Chlorosis in Frogs
Early in this study, it was assumed that disease was not prin
cipally responsible for the green pigmentation of frogs. Since several
writers had accepted green bones or other green tissues as traits which
were characteristic of certain species, it was not logical to believe that
a bacterial or viral disease would affect all the members of a species or
population in a nearly uniform manner. Neither was it logical that a
genetic disease should be passed from generation to generation with such
regularity. Thus, it was tentatively assumed that the normal metabolism
of chlorotic frogs was responsible for their green pigmentation.
On the other hand, disease cannot be ruled out as a possible
cause of biliverdin accumulation. Individual adult specimens of Hyla
septentriona1? s and Hyla dom? n ? censls may have dark green, light green
or white bones. Here is an example of what one might expect if some of
the frogs were diseased, but such individual variation may also have other
causes.

92
A better example of the effect of disease might be demonstrated
by a specimen of Osteocepha1 us taurinus which arrived in the laboratory
shortly after its death. The individual had been kept in another
laboratory for several months, but did not eat during the latter part
of its stay there. Dissection showed the internal organs to be darker
than in most frogs. Blood vessels were a bluish-purple rather than red
in color. An olive-green fluid was found in the coelomic cavity, while
the dorsal lymph sac was nearly filled with a clear, viscous, blue-green
fluid. The gall bladder was filled and appeared dark blue externally,
but contained a clear fluid and a solid brown substance, perhaps bile
salts. The dark color of the internal organs was due primarily to the
presence of biliverdin, which appeared to be in most of the tissues, in
cluding bone. A cross-section of femur showed that the green pigment was
found throughout the bone, but that alternating concentric lamellae of
light and dark green were present. Microscopic examination showed that
the liver cells had completely degenerated. A histochemica1 test for
liver glycogen was made by means of the Periodic Acid-Schiff reaction.
Aggregations of dark pigment were the only structures which were PAS-
positive. These were believed to be a lipofuscin pigment from the degen
eration of red blood cells. Unlike melanin, lipofuscin is PAS-positive
(Dubin, 1958). These liver sections appeared similar to photographs of
liver in acute viral hepatitis published by Popper and Schaffner (1957
p. 437)- While the actual cause cannot be pinpointed, an obvious patho
logical condition of the liver prevented the normal excretion of bile
pigment and evidently caused it to pass into the lymph.

DISCUSSION
Earlier in this paper, evidence was presented to show that
chlorosis is the result of the accumulation of biliverdin in the tissues.
Therefore, this condition corresponds to human jaundice, which is due to
accumulation of bilirubin. Presumably, the obscure causes of chlorosis
would parallel those of jaundice, which were recently outlined by
Williams (1965).
There appear to be four basic causes of jaundice. Abnormally
high rates of hemolysis cause an increase of bile pigment which may
exceed the liver's capacity of excretion. Defective absorption or con
jugation of bile pigment by the hepatic cells, or obstruction of the
various biliary ducts can cause jaundice. These basic causes of bile
pigment accumulation will be considered in relation to chlorosis of frogs.
High Rate of Red Cell Hemolysis
High rates of hemolysis and bile pigment formation might be
expected in frogs which have a high rate of red cell replacement because
of high erythrocyte counts or short red cell life spans. From Table 6,
one notes that there are chlorotic species with high (Phrynohyas venulosa).
low (Hy la wiIderi), and intermediate red cell counts. Species which lack
the green pigmentation demonstrate a similar range of red cell counts.
Little information regarding frog erythrocyte survival times is available,
but my unpublished data on this subject suggest that there is no significant
93

94
difference between red cell survival times of chlorotic and achlorotic
frogs.
Blood smears from adults of Hvla crepitans, Hyla lanciformis,
Sphaenorhvnchus aurantiacus and a tadpole of Pseud?s paradoxus have
demonstrated that the majority of erythrocytes of these Individuals
were hemolyzed within a few hours after the samples were taken. The
changes in cell structure or shape which preceded hemolysis are not known.
Likewise the actual cause of hemolysis is unknown. Since unconjugated
bilirubin is known to cause hemolysi s i n vi tro (Cheung e_t a_l_. 1966),
it seems likely that plasma bilverdin induced hemolysis in the species
listed above, with the exception of Hyla lanciformis. This species is
not chlorotic. Conversely, the red blood cells of several other species
did not hemolyze in their green plasma. From these observations it would
appear that biliverdin is not necessarily responsible for hemolysis,
or that it is effective only under certain conditions or in certain species.
Coagulation of red cell cytoplasm was another notable feature
of some blood smears. The species of Atelopus, Hyla and Leptodacty1 us which
demonstrated this character have little or no tendency toward chlorosis.
Within this group, the small species of Hyla, such as Hyla boesemani ,
Hyla leucophyllata. and Hyla minuta, as well as the chlorotic Sphaenorhynchus
aurantiacus are found in relatively exposed breeding habitats. Since it
has been shown that exposure to a temperature of 45 C. causes coagulation
of frog erythrocyte cytoplasm in vitro, it appears that high environmental
temperatures are responsible for this condition under natural conditions.
Tadpoles of Hyla dominicensis have been observed at 40 C. in nature for

95
several hours. In addition, Bokermann's (1965) observation that Osteo-
cephalus taurinus was subjected to air temperature of 46 C. shows that
the effect of heat is a likely cause of protein coagulation and hemo
lysis in Neotropical anurans. Heating of human blood to 40 C. for
extended periods decreases solubility of hemoglobin (Goldberg, 1958);
heating to 50 C. for ten minutes or more causes hemolysis (Kimber and
Lander, 1964). Cloudsley-Thompson (1965) found that heat death of
tropical lizards was due to physiological oxygen deficiency and another
factor which operated simultaneously, perhaps protein coagulation. Thus,
it appears that the oxygen transport system reaches the limit of its
capacity at about the same temperature which causes the degeneration of
the system because of coagulation of hemoglobin and hemolysis.
The blood cells of Hyla punctata are quite variable in size
and shape, and have basophilic cytoplasm. In fact, these cells appear
to be immature and closely resemble those of erythropoietic tissue. It
seems that these immature cells have been released into the circulating
blood to compensate for cell loss, probably through hemolysis. Erythro
cytes of Phyllomedusa hypochondria 1 is also appear basophilic, but are
relatively uniform in size and shape. It is possible that these cells
are immature, but it seems more likely that the basophilia is tied in
with the presumably low metabolic rate of these slow moving creatures.
Additional evidence of hemolysis is suggested by the presence
of cell fragments in the liver circulation of Pseud is paradoxus.
Since most chlorotic species do not have coagulated cytoplasm
in the erythrocytes present, and since most of the species with coagulated

96
cytoplasm are not chlorotic, these two characters appear to be mutually
exclusive to a large extent. Thus the possibility arises that the
erythrocytes of the chlorotic species are more susceptible to lysis (and
perhaps coagulation as well) than are the red cells of non-chlorotic
frogs. One may hypothesize that high temperatures cause cytoplasmic coagu
lation in both types of frogs, but that hemolysis, leading to biliverdin
formulation, occurs only in the chlorotic species.
Among frogs, two other types of increased hemolysis are likely.
Varela and Sellares (1938) found that Bufo arenarum in Brazil has a
peak number of red blood cells (one million per cubic millimeter) in
July and August with a rapid decrease in blood count in October, at the end
of the breeding season. During the present work, frogs studied in
Surinam in July (at the end of the breeding season) sometimes demonstrated
green plasma but did not have green bones. The assumption was made that
the high concentration of biliverdin was of a transitory nature and did
not last long enough to cause staining of relatively permanent tissues
such as bone. Similarly, the presence of green bones in animals without
green plasma gives additional evidence of seasonal peaks of biliverdin
concentration. Bone cross-sections showing alternate regions of dark and
light green color in Lysapsus mantidacty1 us (Barrio, 1965a) and in
Osteocephalus taurinus also indicate periodic hemolysis. One may suggest
that hemolysis at the end of the breeding season is due to hormonal changes
or to effects of higher temperatures encountered during the breeding
season, or both.

97
Another type of large-scale hemolysis found among anurans
occurs at metamorphosis. McCutcheon (1936) has shown that larval
hemoglobin of Rana catesbeiana is different from that of the adult.
Herner and Frieden (1961) have shown that the hemoglobins of larval
Xenopus laevls, Rana heckscheri and Rana catesbeiana are replaced by
adult hemoglobins during the course of metamorphosis. It seems likely
that this is the explanation for the appearance of biliverdin during
metamorphosis of most West Indian hylids. A similar change from fetal
to adult hemoglobin in man is partly responsible for neonatal jaundice.
Phytohemagglutinins may be used to hemolyze erythrocytes and
induce division of leukocytes in frog blood. However, the effects of
these water-soluble plant derivatives in vivo are poorly known (Boyd, 1963).
Deficiency of vitamin E may result in changes to the cell
membrane of red blood cells, causing them to hemolyze more readily.
This may be a factor in frogs which appear to have more fragile
erythrocytes.
A possibility which has been considered unlikely is that hemo
lysis may be a result of the effect of ultraviolet light upon hemoglobin
(Lemberg and Legge, 19^+9) In view of the fact that blue light may be
used to reduce circulating bilirubin in infants (Broughton ejt a_L 1965),
it is conceivable that the high incidences of ultraviolet light in tropi
cal areas may have an effect on circulating red cells.
Increase in circulating bile salts as a result of defective
liver function may be a cause of hemolysis (Grodins, Berman and Ivy, 1941).
These authors found the apocholate and deoxycholate salts to be the most

98
toxic on the basis of degree of hemolysis produced. Reeder (1964)
has recently summarized the meager knowledge concerning amphibian bile
salts, none of which refers to species known to have biliverdin in
tissues.
Among West Indian hylids, only Hyla marianae has orange bones
during metamorphosis. The reasons for this species being different from
the others are unknown, but speculation is possible. The extract of Hyla
marianae skin contains at least one light orange pigment. Although the
pigment was not identified, it shares spectral characteristics (Figure 6)
with formyl derivatives of folic acid (Stokstad and Koch, 1967)- Coenzymes
containing folic acid are important in transfer of certain one-carbon
groups, oxidation-reduction reactions, synthesis of purines and some
pyrimidines, and in amino acid metabolism, including the transformation
of phenylalanine to tyrosine. Thus, there seems to be the possibility that
this species has a rich supply of folic acid or similar substance which
could allow the liver cells to divide rapidly (because of readily avail
able purines and pyrimidines for DNA synthesis) in order to excrete an
increased amount of biliverdin; produce more melanin (because of avail
ability of tyrosine) than pale, green forms such as Hyla wilderi; and
have sufficient folic acid left that it could be used to eliminate formyl
groups (perhaps from enzymatic biliverdin formation) by way of the skin.
An excess of this vitamin might account for the relatively large size of
these tadpoles at metamorphosis, as well as their aggressiveness. The
source of folic acid is not readily apparent, but might be the leaves
of the small bromeliads in which this species was found. An alternate

99
hypothesis for lack of green pigment in this species is that the
liver of this relatively large tadpole is better developed at metamor
phosis, and therefore can cope with the heavy load thrust upon it at
metamorphosis, including an increase of biliverdin.
Experiments reported on earlier in this paper indicate that
hemolysis alone is not likely to result in high concentrations of bili
verdin in the tissues. One cubic centimeter of a 1 per cent hemoglobin
solution injected into a 1 gram specimen of Hyla cinerea produced no
noticeable effect. This amount of hemoglobin (0.01 gram) would result
in the production of 0.46 milligrams of biliverdin in a rather short
period (hours or a few days). Apparently, it is excreted without diffi
culty and without green pigmentation developing in the tissues. If we
estimate that 5 per cent of such a frog consists of blood, that one-fifth
of that consists of red blood cells, and that hemoglobin constitutes
one-half of the red cell weight, we would calculate that it contains
0.005 grams of hemoglobin. Comparison of these two sets of calculations
would show that the frog can degrade an amount of hemoglobin equivalent
to twice that contained within its body, in a few days, without accumulat
ing biliverdin in the tissues.
Bile pigment formation in frogs has not been studied in regard
to the source of biliverdin. it is presumed that hemoglobin is the pri
mary source of bile pigment, and that other hemoproteins yield a minor
fraction.
Defective Transport of Bile Pigment into the Liver Cell
The importance of this factor in the accumulation of biliverdin
in frogs is not known.

100
Defective Bile Pigment Conjugation
Although it is possible for small amounts of anuran biliverdin
to be enzymatically conjugated as a glucuronide (Noir, Rodriguez Garay
and Royer, 1965), Barrio (1965a) found that the bili verdn of frogs
reacts as if unconjugated. Thus, a deficiency or inhibition of this
conjugating enzyme system probably would not be important in the green
pigmented frogs.
Disturbed Bile Pigment Excretion
Excretion of bile pigment may be blocked either within the
liver or in the ducts leading away from the liver.
Intrahepatic cholestasis may be a constitutional disease which
is often familial. Dubin-Johnson syndrome (Dubin, 1958) is a chronic,
intermittent, benign form of jaundice in which the liver is often greenish
black and microscopically shows much pigment which is probably lipofuscin.
In these characteristics it resembles the situation seen in Pseud? s
paradoxus, which has a black liver in the adult.
Drugs can cause an intrahepatic blockage of bile due to their
effects upon hepatic cells. Important among these are certain steroid
hormones and chlorpromazine, as well as others (Sherlock, 1964). It is
possible that frog liver is affected by hormones since Bachmann, Goin and
Goin (1966) have shown that there is an increase in polyploidy of frog
liver cells with the onset of the breeding season. Steroid hormones are
known to increase polyploidy in other forms. Thus, it seems possible that
tropical frogs may have an extended breeding season which affects the
liver adversely. Evidence of this was seen in the liver of Hyla maxima

101
which showed development of fatty areas, which can be caused by extended
use of steroid hormones or deficient diet.
Destruction of liver cells by chemicals such as carbon tetra
chloride or situations accompanying acute virus hepatitis may result in
intrahepatic obstruction (Williams, 1965)- Tannic acid is a chemical
which causes damage to the liver in mammals (Arhelger, Broom and Boler,
1965) and may be a cause of liver injury to tadpoles which live in waters
with significant concentration of this chemical.
Allen, Carstens and Olson (1967) studied veno-occ1 us ive disease
in monkeys. A single dose of monocrota1ine causes necrosis of liver cells
and partial blockage of the veins within the liver. This drug is believed
to be the ingredient of bush teas which cause veno-occ1 usive disease in
Jamaica. Venous occlusion was observed in Hyla geoqraphica in Surinam.
It may have been induced by a plant extract.
Another factor which has an effect upon bile flow is the con
centration of environmental oxygen. Anesthetized dogs which breathed air
with 15 per cent oxygen for thirty to forty-five minutes had the bile
flow reduced by 16 to 50 per cent. There was also a reduction of urine
formation (Schnedorf and Orr, 19^1). The tadpoles studied lived under
conditions of low oxygen tension, but all appeared to utilize atmospheric
oxygen when placed under stress. However, the conditions of low oxygen
tension and high temperature to which Hyla dominicensis tadpoles were
subjected, would appear to push the limit in this direction. Also, a high
respiratory rate, such as appears to be present in tadpoles of Hyla hei lprini
may result in a relative oxygen lack that would reduce bile flow and thus

102
cause their extreme green pigmentation. It should be noted that
these tadpoles were in the process of metamorphosis and may have had
an excessively high respiratory rate for this reason.
The very low metabolic rate of Pseudis tadpoles may be an
adaptation to low oxygen concentrations within their environment or be
correlated with their large size, but they still develop green tissues.
Desiccation could well be a factor contributing to reducing
the flow of bile through the liver. With less internal water for physio
logical purposes, secretion and excretion of bile may be impaired, so
that bile pigments would occur in the blood. Since the tree frogs, in
their arboreal habitat, appear to be more susceptible to chlorosis than
do the ground frogs, one might suppose that their more exposed position
is the reason. Desiccation is also indicated by the concentrated con-
dition of the bile.
Obstruction of the extra-hepatic bile ducts by gall stones,
parasites or tumors was not observed in the frogs studied. One specimen
did have an inflamed duodenum which may have reduced bile flow into that
portion of the intestine. In most specimens, the gall bladder was filled
with green bile and the feces were dark, indicating that bile was passing
out of the liver, and into the intestine, respectively.
Advantages of the accumulation of green pigment in skin and other
organs seem to lie largely in the realm of protective coloration. In
a number of species, the green color is rarely visible externally, but in
some of the smaller speciesHy la wilderi, and species of Centro lene 1 la
and Sphaenorhynchusthe green pigment is quite apparent and blends well
with the green foliage on which these forms are found. This green

103
pigmentation of the skin appears to have advantages over the green color
produced by the physical arrangement of different types of chromato-
phores. It appears to be long lasting and apparently does not change
according to light, temperature, or hormonal conditions. As long as the
frog remains associated with green vegetation, this should not be parti
cularly disadvantageous. Perhaps the efficiency of energy utilization in
the pigmented forms provides a greater advantage. Since the green pig
mentation results from diffusion of an apparently harmless waste product
into skin and other organs, little or no metabolic energy is required
for the coloring process. In addition, chromatophores may be reduced
without selective disadvantages, thus decreasing the nutrients and energy
needed to maintain chromatophores.
Disadvantages where chlorosis is the normal condition are not
readily apparent, but Lutz and Lutz (1938) stated that Sphaenorhynchus
orophilus loses its green color when not in good health, a situation
opposite to that which would be expected if the green indicated a diseased
condition.

CONCLUSIONS
1. The green coloration of bones, soft tissues, and plasma
of Neotropical anurans has been shown to be due to the presence of a
green pigment.
2. Identification of the pigment as biliverdin has been
confirmed.
3. The high concentrations of biliverdin have been found only
in frogs which inhabit the Neotropical Region. It appears that this
phenomenon is restricted to tropical tree frogs and to the aquatic frogs
of the family Pseudidae.
4. For the first time, chlorosis has been found in tadpoles
of several species.
5. Incidence of the pigmentation does not appear to be
restricted to either sex or to any particular stage of the development,
although it does not appear until metamorphosis in the tree frogs.
6. There are indications that the pigment concentration is
higher during metamorphosis and at the end of the breeding season, than
at other times.
7- While it appears that the green pigmentation is restricted
to three families of tree frogs and the Pseudidae, this is not considered
to be an indication of relationship between the Pseudidae and the tree
frogs.
104

105
8. In different taxa, the presence of green pigment may vary
at familial, generic, specific, subspecific or individual levels.
9. Frogs and tadpoles with green pigmentation have been
collected from habitats which differ widely in concentrations of oxygen,
dissolved salts, iron and phosphate, in addition to temperature, insol
ation, altitude and presence or absence of water currents. These
organisms also differ widely in their food, feeding habits and methods
of gaseous exchange.
10. The actual cause of the presence of high concentrations
of biliverdin is not known. However, it appears that the source of the
pigment is hemoglobin from red blood cells which are hemolyzed.
11. It seems likely that a combination of high temperatures
and fragile erythrocyte membranes may cause hemolysis in some breeding
adults. Hormonal changes at the end of the breeding season and during
metamorphosis may also account for some types of hemolysis.
12. While hemolysis apparently is the source of pigment, it
is unlikely that this alone could cause the green coloration of tissues.
Since hemolysis probably occurs in these species during metamorphosis and
at the end of the breeding season, both of which are times when the
liver has great stress from other activities, it appears that the liver
cells are unable to handle the increased bile pigment at this time.
Since the biliverdin appears to be relatively non-toxic, its elimination
presumably has a low priority.
13. Some species evidently have taken evolutionary advantage
of the green bile pigment accumulation by using it as protective coloration.

106
14. In general, it seems that the green pigmentation is due
to increased hemolysis coupled with a decreased ability of the liver
cells to excrete this bile pigment, so that it accumulates in plasma,
then stains the proteins of soft tissues and finally those of bone.
High environmental temperatures and hormonal changes are considered the
most likely causes of hemolysis, and perhaps impairment of liver function
as well.

APPENDICES
Tables 1 10
Figures 1 20

108
TABLE 1. OCCURRENCE OF
GREEN
COLOR IN TISSUES AND
ANURANS
FLUIDS
OF ADULT
P Green color present
A Green
color absent
FAMILY AND SPECIES
BONE
SOFT
PLASMA
Bl LE
SOURCE
TISSUES
Rhinophrynldae
Rhinophrynus dorsalls(l)
A
A
A
-
Ranidae
Phv1lobates sp.(1)
A
A
A
RhacophorIdae
Hyperolius sp.
Microhy1idae
-
p (eggs)
-
-
M. Stewart
(Pers. Comm.)
Elachistocleis ovale(3)
A
A
A
P
Pseudidae
Lysapsus limellum laevis
P
-
_
-
Parker, 1935
Lysapsus limellum limellum -
P
Barrio, 1965a
Lysapsus mantidacty1 us
P
-
P
-
Barrio, 1965a
Pseudis minuta
P
-
-
-
Peters, 1873
Pseudis paradoxus
paradoxus
P
p
A
P
Pseudis paradoxus
pa radoxus(3)
-
A
A
P
Pseudis paradoxus
pla tens i s
P
-
P
-
Barrio, 1965a
Camerano, 1879
Bufnidae
Bufo typhonius(3)
A
A
A
P

109
TABLE 1 (Continued)
FAMI LY
AND SPECIES
BONE
SOFT
PLASMA
BILE
SOURCE
TISSUES
Atelopodidae
Atelopus sp.
A
A
A
P
Hy 1 idae
Anotheca coronata
P
-
-
-
C. J. Goin
Hy la
a Ibofrenata
P
-
-
-
(Pers. Comm.)
Cochran, 1955
Hy la
albomarqinata
P
-
-
-
Cochran, 1955
Hy la
berthae
-
-
A
-
Barrio, 1965a
Hy la
brunnea
A
A
A
P
Hy la
boesemani (1)
P
P
-
P
Hy la
ca lea rata (1)
P
P
P
P
Hv la
crepitans(2)
P
P
P
P
Hy la
cuspidata
P
-
-
-
B. Lutz, 1954
Hy la
dominicensis
P or A
_
_
_
Lynn, 1958
Hy la
eq leri(2)
A
A
A
P
Hy la
qeoqraphica(1)
P
A
A
P
Hv la
hei1prini (1)
P
P
P
P
Hy la
land formi s (4)
A
A
A
P
Hyla
1 anqsdorffi
P
P
-
-
B. Lutz, 1954
Hv la
leucophy1 la ta (1)
A
A
A
P
Hy 1 a
1ichena ta (1)
P
Hv la
mar ianae (7)
A
A
A
-
Hy 1 a
maxima(3)
P
P
P
P
Hy la
mi sera (1)
A
A
P
P

TABLE 1 (Continued)
FAMILY AND SPECIES
BONE
SOFT
PLASMA
B! LE
SOURCE
TISSUES
Hyla nasica
-
-
P
-
Barrio,
1965a
Hyla pulchella andina
-
-
P
-
Barri0,
1965a
Hyla pulchella riojana
P
Barrio,
1965a
Hyla pulchella cordobae
.
_
P
Barri0,
1965a
Hyla pulchella pulchella
_
A
Barrio,
1965a
Hyla pulchella prasina
.
.
A
Barrio,
1965a
Hyla phrynoderma
-
-
A
-
Barrio,
1965a
Hyla punctata(2)
P
P
P
P
Hyla punctata
P
-
P
-
Barrio,
1965a
Hyla raniceps
-
-
A
-
Barrio,
1965a
Hyla rubra(4)
A
A
A
P
Hyla septentriona1 is P
or A
A
P or A
P
Hyla siemersi
-
-
P
-
Barrio,
1965a
Hyla squa1irostris
-
-
A
-
Barri0,
1965a
Hyla trachytorax
-
-
A
-
Barrio,
1965a
Hyla vasta P
or A
-
-
P
Hyla wavrin!
P
-
-
-
Rivero,
1961
Hyla wilderi(4)
P
P
-
-
Osteocepha1 us taurinus(2)
P
P or A
P or A
P
Phyllomedusa bicolor(l)
A
A
A
Phyllomedusa helenae
-
green eggs
-
-
Starrett, I960
Phy1lomedusa
hypochondria 1 is (5)
A
A
A
P

TABLE 1 (Continued)
FAMILY AND SPEC IES
PLASMA
BONE SOFT
TISSUES
Phyllomedusa sauvagii
Phrynohyas venulosa P
Phrynohyas venulosa(4) P P
Sphaenorhynchus
aurantiacus (5) P P
Sphaenorhynchus dorisae P
Sphaenorhynchus habrus P
Sphaenorhynchus orophilus P
A
P
P
P
T rachycepha1 us
nigromaculatus P
Leptodacty 1 idae
Eleutherodacty1 us
(several species) A
Leptodacty1 us sp.(2) A
Leptodacty1 us
pentadacty 1 us(1) A
Centrolenidae
A A
A A
Centro lene 1 la albomaculata P
Centrolenel la
fleischmanni
A
-
Centro lene 1 la
fleischmanni
green eggs
-
Centrolenella qranulosa
P
Centrolenella ilex
P
-
Centro 1 ene 1 la
prosoblepon
P P
-
BILE SOURCE
Barrio, 1965a
Barrio, 1965a
Coin, 1957
Coin, 1957
Coin, 1957
Cochran, 1955
P
P
Savage, 1967
Dunn, 1931
Starrett, I960
Savage, 1967
Savage, 1967
Dunn, 1931

TABLE 1 (Continued)
FAMILY AND SPECIES BONE SOFT PLASMA
TISSUES
Centrolenella pulveratum P P
Centrolene1 la reticulata green eggs
Centrolene1ia spinosa P -
112
BILE SOURCE
Dunn, 1931
Starrett, i960
Savage, 1967
Barrio, 1965a
Centrolene1 la vanzolinii P

113
TABLE 2. OCCURRENCE OF GREEN PIGMENT AMONG METAMORPHOSING NEOTROPICAL
ANURAN
TADPOLES
SPECIES
BONE
SOFT TISSUES
PLASMA
BILE
PSEUD 1 DAE
Pseud is paradoxus
P
P
P
P
HYLIDAE
Hyla brunnea
P
P
-
P
Hyla dominicensis
P
P
-
P
Hy la hei1prini
P
P
P
P
Hyla lichenata
P
P
-
-
Hyla marianae
A
A
-
-
Hyla septentriona1is
P
P
-
-
Hyla vasta
P
P
-
-

TABLE 3- LOCALITY DATA OF TADPOLE HABITATS
LOCATION AND HABITAT DATA
Bromeliads of Jamaica habitat of Jamaican
tree frogs and tadpoles
I Rose Va 1 ley
2. 10 km NW of Troy
3. 6 km S of Mandevi1 le
4. Point Hill
5. Cavalier District, St. Andrew
6. 1.6 km W of Wakefield
7. 6.4 km NW of Juan de Bolas
Bromeliad of Puerto Rico habitat of no
known tadpoles
8. El Yunke Biological Station
Ponds and standing waters of Haiti
Hyla dominicensis
habitats of
DATE
ELEVATION AIR
(Meters) TEMPERATURE
25 X 1965
490
-
14 V 1966
490
28.5-33.5
10 VIII 1966
490
26.4-31.2
16 V 1966
750
26.0-27.5
8 VI11 1966
550
24.7-28.2
17 V 1966
600
25.0
27 X 1965
600
26.7
27 x 1965
350
28
27 x 1965
600
24
6 V 1966
800

TABLE 3 (Continued)
LOCATION AND HABITAT DATA
9. 8 km W of Port-au-Prince (flooded
wi regrass pasture)
10. 32 km W of Port-au-Prince (wheel ruts)
11. 4.8 km W of Port-au-Prince
(shaded pool in stream bed)
t
Ponds and standing waters of Grand Cayman,
B.W.I. habitats of Hyla septentriona1is
12. 3-2 km N of Georgetown (pond)
13- Hell (roadside ditch)
14. South Sound (two mangrove swamps)
15- South Sound (four woodland ponds)
16. 4.8 km E of Georgetown (cow well)
17- 8 km E of Georgetown (cistern)
18. Boddentown to East End (three ponds)
19- North Side (woodland pond)
20. North Side (large open pond)
DATE
ELEVATION
(Meters)
AIR
TEMPERATURE
12 V 1966
1
30.2
4 VI11 1966
10
33-2-35.2
10 V 1966
200
26.8
12 V 1966
200
27-2
29
X
1965
4
31.7
29
X
1965
4
30.5
29
X
1965
2
29-5-30.5
29
X
1965
2
30.5-32.2
29
X
1965
2
29.5
29
X
1965
2
28.3
29
X
1965
3
27.2-28.9
30
X
1965
3
27.2
30
X
1965
3
27.7

TABLE 3 (Continued)
LOCATION AND HABITAT DATA
Open pool of Florida, U.S.A. habitat of
Hyla septentrionai?s
21. SW Miami (pool of Serpentarium)
Pool of Trinidad habitat of Phy1lomedusa
bicolor tadpoles
22. Simla Biological Station
Ponds and standing waters of Surinam
23. Albina Road at Commewijne River (road
side ditch)
24. Domberg (roadside ditch)
25. 28 km S of Paramaribo (ditch)
26. Overtoom Road (pond and pool)
27. Onverwacht (flood meadow)
Streams of Surinam
28. Brownsberg Princess Irene Falls
(habitat of Hyla maxima)
29. Base of Brownsberg Brown's Kreek
DATE
ELEVATION AIR
(Meters) TEMPERATURE
25 VI 1966
10 31.2-32.2
29 VI I 1966
30.2
13
VI 1
1966
10
29.
0
17
VI 1
1966
5
32.
,8
17
VI 1
1966
60
27
5
25
VI 1
1966
60
29.
.0-29.3
25
VI 1
1966
60
27
.0-30.2
20
VI 1
1966
350
24.7-25.1
21
VI 1
1966
350
24.5-25.0
21
VI 1
1966
200
27.8

TABLE 3 (Continued)
LOCATION AND HABITAT DATA
Streams of Haiti
30. Furcy (habitat of Hyla vasta
and Hyla he i Iprini)
31. Riviere Froid, 4.8 km W of Port-au-Prince
Streams of Jamaica habitat of Bufo marinus
32. Vicinity of Troja (two streams)
33- 4.8 km W of Bog Walk (irrigation ditch)
34. Vicinity of Kellitts (two streams)
Streams of Puerto Rico habitat of
Leptodacty1 us albilabrus
35 El Yunke roadside stream
DATE
ELEVATION
(Meters)
AIR
TEMPERATURE
10-12 V 1966
1350
20.5-29.5
2 VI11 1966
1350
25.4-26.0
10 v 1966
150
26.7
9 VI11 1966
450
28.7-30.0
9 VIII 1966
350
28.2
17 v 1966
700
24.0-25-3
6 V 1966
750
22.2

TABLE 4.
WATER CHARACTERISTICS OF TADPOLE HABITATS
LOCATION. HABITAT AND DATE
TEMPERATURE
OXYGEN(ppm)
HARDNESS(ppm)
pH
Bromeliads of Jamaica habitat of
Jamaican tree frogs and tadpoles
1. Rose Valley (25 X, in sun)
24.0-28.0(6)
-
-
5-7
(25 X, in shade)
23.0-26.0(17)
-
-
5.5-7
(14 V, in sun)
25-5-28.5(4)
1.4-2.4(4)
20
5
(14 V, in shade)
25.5-25.7(3)
1 .2-2.4(3)
40
-
(10 VI11 in sun)
27.2
2.7
-
5-5
(10 VIII, in shade)
24.8-25.3(3)
3.0-3.2(3)
-
6.5-7.0
2. 10 km NW of Troy (16 V)
24.0-26.1(3)
1.6-1.9(3)
50
5-5
3. 6 km S of Mandevilie (8 VI1 I)
22.9-23-7(4)
2.3-6.6(4)
-
5.6
4. Point Hi 11 (17 V)
22.5
1. 1
-
5.5
5- Cavalier District (27 X)
25.0-27.0(8)
-
8-40
5-6
6. 1.6 km W of Wakefield (27 X)
25.0-25.5(3)
-
10
5-5-5
7. 6.4 km NW of Juan de Bolas (27 X)
(without dust on leaves)
24.0
-
16-30(3)
5-5-6.5
(with dust on leaves)
24.5
-
70-90(3)
7.0
\

TABLE 4 (Continued)
LOCATION, HABITAT AND DATE
Bromeliad of Puerto Rico habitat of
no known tadpoles
8. El Yunke Biological Station (6 V)
Ponds and standing waters of Haiti -
habitats of Hyla dominicensis
9. 8 km W of Port-au-Prince (12 V)
10. 32 km W of Port-au-Prince (4 VIII)
11. 4.8 km W of Port-au-Prince (10 V)
(12 V)
Ponds and standing waters of Grand
Cayman, B.W.I. potential habitats of
Hyla septentriona 1 ?s
12. 3-2 km N of Georgetown (29 X)
13. Hell (29 X)
14. South Sound (29 X)
16. 4.8 km E of Georgetown (29 X)
18. Boddentown to East End (29 X)
TEMPERATURE
OXYGEN(ppm)
HARDNESS(ppm)
21.0
3.1
-
5-2
32.0
0.7
480
7-5
39-2-40.2(2)
1.4-8.0(2)
220(1)
7-5
25.1
2.2
200
-
24.6
1.7
200
7
32
-
64
7-0
33.5
-
60
7-0
28-33(2)
-
84
7-5
32.5
-
50
7-0
28-29(3)
-
60-90
7-0

TABLE 4 (Continued)
LOCATION, HABITAT ANO DATE
Actual habitats of Hyla septentriona 1 ?s
15. South Sound (29 X)
17. 8 km E of Georgetown (29 X)
19- North Side (30 X)
20. North Side (30 X)
Open pool of Florida, U.S.A. habitat
of Hyla septentriona 1 is
21. SW Miami (25 VI)
Pool of Trinidad habitat of
Phyllomedusa bicolor
22. Simla Biological Station
Ponds and standing waters of Surinam
23. Albina Road at Commewijne River
(13 VII)
24. Oomberg (1J VII)
25. 28 km S of Paramaribo (17 VII) -
habitat of Phrynohyas venulosa
26.Overtoom Road (25 VII)
TEMPERATURE
OXYGEN(ppm)
HARDNESS(ppm)
E
29.0-34.0(4)
-
106-134
7-5
25
-
60
7-5
27
-
64
7.0
26
-
70
7-0
32.9-33.9(2)
8.1-9-2(2)
-
-
28.2
10. 1
-
-
27.8
3-4
-
-
33.0
-
-
-
25.0-29.4(2)
0.7
-
-
29.9-30.8(2)
6.5-7-1(2)
-
-

TABLE 4 (Continued)
LOCATION, HABITAT AND DATE
27. Onverwacht habitat of Hy1 a crepitans ,
Hyla punctata, Hyla minuta and Hyla
boesemani (25 VI I-5PM)
(25 VII-7:40PM)
Streams of Surinam
28. Princess Irene Falls habitat of
Hyla maxima (20-21 VII)
Rock basin (21 VII)
29. Brown's Kreek habitat of fish
(21 VII)
Streams of Haiti
30. Furcy habitat of Hyla vasta and
Hyla heiIprini (10-12 V)
(2 VIII)
31. Riviere Froid (10 V)
Streams of Jamaica
32. Troja habitat of Bufo marinus
(9 VIII)
TEMPERATURE
OXYGEN(ppm)
HARDNESS(ppm)
£H
30.9
3.4
-
-
27-5
0.6
-
-
23.4-25.5(6)
7.8-8.6(6)
-
-
23.8-29.9
2.0-3.1
-
-
25-2
4.9
-
-
19-5-24.5(14)
7.4-9-5(14)
40-90(3)
6-7(2)
20.1-25.2(5)
7-3-9.9(5)
60-90(5)
7
25-8
7-9
130
27.0-27.1(2)
7-9-8.0(2)

TABLE 4 (Continued)
LOCATION, HABITAT AND DATE
33- 4.8 km W of Bog Walk (9 VIII)
34. Kellitts (17V)
Streams of Puerto Rico
35- El Yunke habitat of Leptodacty1 us
a 1bilabrus (6 V)
TEMPERATURE
OXYGEN(ppm)
HARDNESS(ppm)
J2
24.5
8.6
-
7
22.0-22.5(2)
8.2-8.4(2)
80-130(2)
6.5-7-0
21.0
3.1
20
6

123
TABLE 5. IRON AND PHOSPHATE CONCENTRATIONS (PARTS PER MILLION) IN
TADPOLE HABITATS
HABITAT DATA AND LOCATION
IRON
PHOSPHATE
SPECIES OF TADPOLE
PRESENT
Bromeliads of Jamaica
1. Rose Valley
0.00
0.00
Hyla brunnea
Trace
0.40
None
0.05
2.2
None
0.02
0.30
None
3. 6 km S of Mandevi1le
0. 16
0. 10
Hyla brunnea
0.25
0.21
Hyla brunnea
0.73
3-4
None
0.30
2.7
None
Standing waters
10. Wheel ruts 32 km W of
1.13
1.35
Hyla dominicensis
Port-au-Prince, Haiti
1.80
4.80
Hyla dominicensis
24. Domberg, Surinam
7-5
0. 18
Pseud is paradoxus
25. 28 km S of Paramaribo
0.62
0.70
Phrynohyas venulosa
Surinam
Stream habi tats
28. Princess Irene Falls
0.06
0.02
Hyla maxima (larqe)
Basin near the Fa 11s
0.18
0.03
Hyla maxima (sma11)
30. Furcy, Haiti
0.
, 10-0.26
0.35
Hyla heilprini, Hyl
32. Troja, Jamaica
0.
,02-0.40
0.39-0.41
Bufo marinus
33. 4.8 km W of Bog Walk
Jamaica
0.30
0.25
None
vasta

124
TABLE 6. BLOOD CHARACTERISTICS
FAMILY AND SPECIES
HEMOGLOBIN
RED CELL
HEM0GL0BI!
(Gram Per Cent)
COUNT
PER CELL
(P i cograms'
Microhy1idae
Elachistocleis ovale
8.0
597,500
134
13.0
805,000
161
9-0
725,000
124
Pseudidae
Pseudis paradoxus
7-0
887,500
79
10.0
782,500
128
9.0
767,500
117
8.0
833,000
96
Buronidae
Bufo typhonius
7.5
505,000
149
6.0
460,000
130
Hy1idae
'*H v 1 a calca rata
7.0
517,500
135
*Hyla crepitans
8.0
962,500
83
*Hyla qeoqraphica
8.0
682,500
117
*Hyla heilprini
11.4
400,000
285
Hyla lanciformis
5.5
320,000
172
6.5
345,000
188
7.5
477,500
157

125
TABLE 6 (Continued)
I
FAMILY AND SPECIES
HEMOGLOBIN
(Gram Per Cent)
RED CELL
COUNT
HEMOGLOBIN
PER CELL
(Picograms)
Hvla leucophy1 lata
7.0
460,000
152
Hyla marianae
543,000
*Hyla maxima
14.5
967,500
150
14.5
1,395,000
104
7.5
512,500
146
*Hvla punctata
3.0
615,000
49
Hyla rubra
13-5
965,000
140
7-5
675,000
111
10.0
790,000
127
*Hyla vasta
11.0
742,500
148
*Hyla wiIderi
212,000
*Osteocepha1 us taurinus
6.5
660,000
98
'"Phrynohyas venulosa
9-5
1,212,500
78
4.5
950,000
47
11.5
1,157,500
99
16.0
1,378,000
116
Phyllomedusa bicolor
5.5
322,500
171
Phyllomedusa hypochondri a 1is 6.0
255,000
235
6.5
287,500
226
5-5
360,000
153
7.0
478,000
146
''Sphaenorhynchus aurantiacus 7.5
585,000
128
2.0
217,000
92

126
TABLE 6 (Continued)
FAMILY AND SPECIES
HEMOGLOBIN
RED CELL
HEMOGLOBIN
(Gram Per Cent)
COUNT
PER CELL
(Pi cograms)
Leptodacty1idae
Leptodactv1 us
pentadacty1
I us 7-0
467,500
150
Leptodacty1 us
sp.
11.5
1, 100,000
105
Leptodacty1 us
sp.
6.5
850,000
76
Chlorotic species

127
TABLE 7. AVERAGE DIMENSIONS OF FROG RED BLOOD CELLS
FAMILY AND SPECIES AVERAGE CELL AVERAGE CELL
LENGTH +
2 S.E.
WIDTH +
2 S.E
Pseudidae
'Pseudis paradoxus
13.2 +
0.73
12.0 +
0.30
Bufnidae
Bufo typhonius
16.3 +
0.92
12.9 +
0.39
Hy1idae
*Hyla boesemani
16.3 +
0.63
11.9 +
0.38
Hyla eqleri
17.4 +
1.48
12.6 +
0.62
Hyla qeoqraphica
15 8 £
0.30
9.9
1.74
Hyla leucophy1 lata
16.7 +
0.61
13.4 +
0.44
Hy 1 a maxima
18.3 +
0.48
11.0 +
0.33
Hyla minuta
15.1 +
0.40
11.7 +
0.43
*Hyla misera
14.0 +
0.54
12.4 +
0.48
*Hyla punctata
17.9 +
1.05
12.8 +
2.00
Hyla rubra
13.9
0.43
11.4 +
0.24
'Osteocepha lus taurinus
13.1
0.36
10.8 +
0.40
Phyllomedusa bicolor
17.0 +
0.73
13.7 +
0.66
Phyllomedusa hypochondria 1is
20.6 +
0.89
14.5 +
1.29
*Phrynohyas venulosa
12.8 +
1.11
10.8 +
0.27
'Sphaenorhynchus aurantiacus
21.6 +
0.47
13.2 +
0.39
Leptodacty1idae
Leptodactv1 us pentadactv1 us
17.2 +
0.42
12.4 +
0.33
Figures represent average for twenty-five cells from a single
of each species.
PRODUCT OF CELL
LENGTH x WIDTH
158.4
210.3
194.0
219.2
156.4
223.8
201.3
176.7
173.6
229.1
158.5
141.5
232.9
298.7
138.2
285.1
213.3
i ndi vidua1
* Chlorotic species

128
TABLE 8. RESPIRATORY RATES OF TADPOLES IN WARBURG RESPIROMETERS
AT DIFFERENT TEMPERATURES
SPECIES AND GROUP WEIGHT(mg) OXYGEN UPTAKE (ul 02/g/hr)
15
20
O
LA
CM
O
O
0
40
Group A
Hvla brunnea
170
236
303
364
403
275
264
399
483
319
280
306
428
547
646
286
253
00
CA
460
390
Group B
Hvla brunnea
152
41
108
133
252
113
147
186
260
123
152
225
Group C
Hvla septentriona1is
101
232
365
409
421
152
239
352
536
574
21 1
273
425
506
497
219
260
317
395
371
Group D
Hvla septentriona1is
179
108
131
273
190
82
88
138
201
113
132
217
191
106
117
257

129
TABLE 9. TADPOLE RESPIRATORY RATES AT 25 C. (MEASURED BY JONES METHOD)
SPECIES WEIGHT(mg) OXYGEN UPTAKE (ul Oy/g/hr)
Hyla dominicens!s
315
262
Hyla dominicensis
481
241
Hyla dominicensis
398
187
Hyla dominicensis
248
216
Hyla dominicensis
226
206
Hyla dominicensis
597
252
Hyla hei1prini
1,203
255
Hyla heilprini
1,444
265
Hyla hei1prini
1,420
277
Hyla vasta
476
203
Hyla vasta
O
CTn
248
Hyla vasta
370
329
Hyla vasta
83
663
Hyla vasta
84
669
Hyla vasta
170
560
Hyla maxima
69
739
Hyla maxima
70
714
Hyla maxima
78
756
Hyla maxima
84
678
Pseud is paradoxus
3,696
53
Pseud is paradoxus
5,948
31
Pseud is paradoxus
4,577
44

130
TABLE 9 (Continued)
SPECIES
WEIGHT(mq)
OXYGEN UPTAKE (ul 0?/q/hr)
Pseudis paradoxus
3,013
35
Pseudis paradoxus
6,782
41
Pseudis paradoxus
5,806
36

131
TABLE 10.
FROGLET RESPI
RATORY RATES MEASURED
IN WARBURG RESPiROMETERS
SPECIES
WEIGHT(mq)
OXYGEN CONSUMED
(25 C.)
(ul 0?/q/hr)
OXYGEN CONSUMED
(35 C.)
(ul 0?/q/hr)
Hyla brunnea
1 12
167
482
120
101
548
128
279
61 1
135
134
395
215
152
406
231
211
471
237
159
459
286
119
331
Four days beyond
metamorphos Is
Hyla brunnea
225
549
Hyla lichenata
282
487

132
Opt ica1
Dens ity
Figure 1.
Optica1
Density
Figure 2.
Absorption spectrum of blllverdln In 5 per cent hydro
chloric acid methanol solution.
Absorption spectrum of liver extract from Hvla
septentriona11s in aqueous solution.

133
Optica1
Density
Figure 3
Optica 1
Density
Figure 4. Absorption spectrum of coelomic fluid (bile) from
Osteocephalus taurinus.

134
Figure 5. Absorption spectra of bile solutions of Siren lacertina.
200 300 400 500 600
WAVE LENGTH IN MILLIMICRONS
Figure 6. Absorption spectrum of pigment extracted from Hyla
marianae in 70 per cent ethanol.

Respiratory rates of Hvia brunnea tadpoles at different temperatures (Table 8)
Figure .

1000
700
500
400
OXYGEN 300
UPTAKE
(MICRO- 200
LITERS
PER GRAM
PER HOUR))00
70
50
40
30
20
p Pseudis paradoxus
10 I i L_l U I 1 I I l I I lI I I 1 > iI
0.01 0.02 0.03 0.05 0.07 0.1 0.2 0.3 --5 0.7 1.0 2.0 3-0 5-07-0 10.0
WEIGHT (GRAMS)
Figure 8. Respiratory rates of tadpoles at 25 C. (From Table 9)
O'

137
Figure 9. Normal red blood cells of Bufo terrestris four hours
after collection. (400X)
Figure 10. Red blood cells of the same individual as above, after
four hours at 45 C. Note coagulated cytoplasm. (400X)

Figure 11. Red blood cells of Hvla squirelia after four hours at 45 C.
Note coagulated cytoplasm and tendency of cells to stick
to each other. (400X)
Figure 12. Untreated blood cells of Hyla leucophy1 lata. Note numerous
erythrocytes with coagulated cytoplasm. (400X)

V
139
Figure 13. Stained smear of untreated blood of Bufo typhonius. Note
coagulated cytoplasm. Green tissues are unknown in this
species. (400X)
Figure 14. Stained smear of untreated blood of Phrynohyas venulosa.
Note the clumped cytoplasm. (400X)

140
Figure 15. A stained smear of blood from Hyla punctata, a species with
green tissues. Note that the majority of cells are immature,
as indicated by their irregular shaped and large nuclei. (400X)
Figure 16. Stained smear of untreated Phyllomedusa bicolor blood.
Cells were basophilic and perhaps immature in this frog
which lacked green pigment. (400X)

141
Figure 17. Stained section of apparently normal liver of Hyla rubra.
(400X)
Figure 18. Stained section of Hyla punctata liver. Note the numerous
nuclei, some of which appear pycnotic. (400X)

142
Figure 19. Stained section of liver from adult male Pseud is paradoxus.
Note the large amount of pigment present. (200X)
Figure 20. Stained section of liver from adult male Hyla maxima.
Note degeneration of cytoplasm toward the lower right.
(400X)

LITERATURE CITED
Agner, K. 1941. Verdoperoxidase. A ferment Isolated from leucocytes.
Acta Physiologica Scandinavica 2_:Supplement 8.
Allen, J. R., L. A. Carstens, and B. E. Olson. 1967. Veno-occlus ive
disease in Macaca sped osa monkeys. American Journal of
Pathology 0: 653 667-
Andrew, W. 1965- Comparative Hematology. New York: Grue and Stratton.
Arhelger, R. B., J. S. Broom, and R. K. Boler. 1965- U1trastructura1
hepatic alterations following tannic acid administration to
rabbits. American Journal of Pathology 46: 409434.
Atlas, M. 1938. The rate of oxygen consumption of frogs during
embryonic development and growth. Physiological Zoology.11:
278-291
Babak, E. 1907a. Uber die funktionelle Anpassung der ausseren Kiemen
bei Sauerstoffmange 1. Zentralblatt fur Physiologie 2_1_: 97-
Babak, E. 1907^- Untersuchungen uber die Warmelahmung und die Wirkung
des Sauerstoffmange1s bei Rana fusca und esculenta. Zentralblatt
fur Physiologie 2 1: 6.
Bachmann, K., 0. B. Goin and C. J. Goin. 1966. Hy 1 id frogs: polyploid
classes of DNA in liver nuclei. Science 154: 650-651.
Barbour, T. 1910. Notes on the herpetology of Jamaica. Bulletin of
Museum of Comparative Zoology 2_: 273-301.
. 1914. A contribution to the zoogeography of the West Indies.
Memoirs Museum of Comparative Zoology 44: 209-259-
- 1931* Another introduced frog In North America. Copeia 1931: 140.
. 1937. Third List of Antillean Amphibians and Reptiles. Bulletin
of Museum of Comparative Zoology 82_: 77-166.
Barbour, T. and C. T. Ransden. 1919- The herpetology of Cuba. Memoirs
Museum of Comparative Zoology 471-213.
Barrio, A. 1965a. Cloricia fisiolgica en batracios anuros. Physis 25:
137-142.
143

144
Barrio, A. 1965b. Las subespecies de Hyla pulchel la. Physis 2:
1 15-128.
Bialaszewicz, K. and R. Bledowski. 1915- The influence of fertili
zation on the respiration of eggs. Proceedings of Scientific
Society of Warsaw 8^: 429-473-
Boell, E. J. 1948. Biochemical differentiation during amphibian
development. Annals of the New York Academy of Sciences 49:
773-800.
Bokermann, W. C. A. 1964. Dos nuevas especies de Hyla de Minas Gerais y
notas sobre Hyla alvarengai Bok. Neotropica J_0: 67 (From
Barrio, 1965a)
. 1965- Field observations on the hylid frog Osteocepha1 us
taur ?nus Fitz. Herpetologi ca 20_: 252-255-
Bond, A. N. I960. An analysis of the response of salamander gills to
changes in the oxygen concentration of the medium. Develop
mental Biology 2_: 1-20.
Boulenger, G. A. 1880. (Note). Zoological Record 1880: 10-11.
. I883. Notes on little known species of frogs. Magazine of
Natural History 8^: 1-32.
Boyd, W. C. 1963. The lectins: their present status. Vox Sanguinis
N. S. 8: 1-32.
Brachet, J. 1935- Etude du metabolisme de l'oeuf de Grenoui 1 le (Rana
fusca) au cours du developpement. Archives de Biologie 46:
1-24.
Brattstrom, B. H. 1962. Thermal control of aggregation behavior in
tadpoles. Herpeto logi ca J_8: 38-46.
Broughton, P. M. G., E. J. R. Rossiter, C. B. M. Warren, G. Goulis, and
P. S. Lord. 1965- Effect of blue light on hyperbilirubinemia.
Archives of Diseases of Childhood 40: 666-671 -
Brown, G. W., Jr. 1964. The metabolism of Amphibia. In: Physiology of
the Amphibia, J. A. Moore, ed. New York: Academic Press,
pp. 1-98.
Cabello, Ruz, J. 1943- Bi 1 iverdinemia del sapo. Revista Sociedad
Argentina Biolgica 81.
Caglar, M. 1945- Biliverdin as a pigment in a fish. Nature 155: 67O.

145
Carnerario, L. 1879. Colorazione naturale del le ossa di una specie di
anfibio anuro. Atti Accademia Torino _1: 789-794.
Cantarow, A. and B. Schepartz. 1957- Biochemistry, second edition,
Philadelphia: W. B. Saunders Company, p. 601.
Cheung, W. H. e_t aj_. 1966. The effect of bilirubin on the mammalian
erythrocyte. Transfusion 6^: 475-486.
Cloudsley-Thompson, J. L. 1965. Rhythmic activity, temperature-
tolerance, water-relations and mechanisms of heat death in
a tropical skink and gecko. Journal of Zoology 146: 55-69-
Cochran, D. M. 1941. The Herpetology of Hispaniola. U. S. National
Museum Bulletin 177-
. 1955- Frogs of Southeastern Brazil. U. S. National Museum
Bulletin 2 06.
Cohen, B. and A. H. Smith. 1919- The colorimetric determination of
hemoglobin. A practical procedure. Journal of Biological
Chemistry 489-496.
Conant, R. 1958. A Field Guide to Reptiles and Amphibians. Boston:
Houghton Mifflin Company, p. 199-
Dickerson, M. C. 1906. The Frog Book. New York: Doubleday, Page
and Company.
Dowling, M. I960. Interlude at Simla. Animal Kingdom 6: 137-139-
Drastich, L. 1927- Uber das Leben der Salamanderlarven bei hohem und
niedrigem Sauerstoffpartialdruck. Zeitschrift fur Vergleichende
Physiologie 2\ 632-657- (From Foxon, 1964)
Dubin, I. N. 1958. Chronic idiopathic jaundice. American Journal of
Medicine 24: 268-292.
Duellman, W. E. and A. Schwartz. 1958. Amphibians and reptiles of
southern Florida. Bulletin, Florida State Museum : 181-324.
Dunn, E. R. 1926. The frogs of Jamaica. Proceedings of Boston Society
of Natural History 8: 111 -130.
1931- The amphibians of Barro Colorado Island. Occasional Papers
of Boston Society of Natural History 403-421.
Elias, H. 1943. Cause for blue, green and red color in Anura. Anatomical
Record 8]_ (Supplement): 440.

Elias, H. and H. Bengelsdorf. 1952. The structure of the liver of
vertebrates. Acta Anatmica J_4: 297337-
Emmel, V. E. 1924. Studies on the non-nucleated elements of the blood:
II. The occurrence and genesis of non-nucleated erythrocytes
or erythroplastids in vertebrates other than mammals. American
Journal of Anatomy 33* 347-406.
English, T. M. 1912. Some notes on the natural history of Grand
Cayman. Handbook of Jamaica for 1912: 598-600. (From Grant,
1940)
Etkin, W. 1934. The phenomena of anuran metamorphosis II. Physiological
Zoology J\ 129-148.
Ewer, D. W. 1959- A toad (Xenopus laevis) without haemoglobin. Nature
183: 271.
Fernandez, K. and M. Fernandez. 1921. Sobre la biologa y reproduccin
de Batracios argentinos. Anales Sociedad Cientfica Argentina
91 (From Barrio, 1965a)
Fontaine, M. 1941a. A propos du pigment vert de l'orphie (Belone belone
L.). Bulletin Institu Oceanographi que Monaco 8: 793*
. 1941^.. Recherche sur quelques pigments serique et dermique de
poissons marins (Labrdes et Cyclopterides). Bulletin Institu
Oceanographi que Monaco 38: 792.
Foulkes, E. C., R. Lemberg and P. Purdom. 1951. Verdohaem and Vverdoglobi
Proceedings of Royal Society of London. Series B. 138: 386-402.
Fox, D. L. 1953. Animal Biochromes and Structural Colors. Cambridge:
University Press.
Foxon, G. E. H. 1964. Blood and Respiration. In: J. A. Moore, editor,
Physiology of the Amphibia. New York: Academic Press, pp. 151~
209.
Freeman, J. R. 1963. Studies on respiratory mechanisms of the salamander
Pseudobranchus striatus. Dissertation, Gainesville, Florida:
University of Florida.
Gallardo, J. M. 1961. On the species of Pseudidae (Amphibia, Anura).
Bulletin of Museum of Comparative Zoology 125: 111-134.
Gans, C. 1956. Frogs and paradoxes. In Animaland 23.: 2-4.
Gerber Products Company. 1965. Nutritive values of Gerber baby foods.
Fremont, Michigan: Gerber Products Company.

147
Godlewski, E. 1900. Ueber die Einwirkung des Sauerstoffs auf
Entwicklung und uber den Gaswechsel in den ersten Entwicklung-
stadien von Rana temporaria. Bulletin Internationale de .1 '
Academie des Sciences, Cracovie. pp. 232-255-
Goin, C. J. 1957- Status of the frog genus Sphoenohyla with a synopsis
of the species. Caldasia 8: 11-31.
. 1961. Synopsis of the genera of hylid frogs. Annals of
Carnegie Museum 6: 5-18.
Goin, C. J. and B. W. Cooper. 1950. Notes on a collection of amphibians
from Jamaica. Occasional Papers of the Museum of the Institute
of Jamaica No. 4: 1-9.
Goin, C. J. and 0. B. Goin. 1962a. Introduction to Herpetology. San
Francisco: W. H. Freeman and Company.
and 1962'-. Amphibian eggs and the amphibian environment.
Evolution 16: 364-371-
Goin, C. J. and C. G. Jackson. 1965- Hemoglobin values of some amphibians
and reptiles from Florida. Herpetologica 2J_: 145-146.
Goin, C. J. and J. N. Layne. 1958. Notes on a collection of frogs from
Leticia, Colombia. Publication of Research Division Ross Allen
Reptile Institute J_: 97-104.
Goldberg, C. A. J. 1958. The ferrohemoglobin solubility test. Clinical
Chemistry 4: 146-149.
Gosse, P..H. 1851. A Naturalist's Sojourn in Jamaica. London.
de Graaf, A. R. 1957- A note on the oxygen requirements of Xenopus
laevis. Journal of Experimental Biology 4: 173-176.
Grant, C. 1940. The Herpetology of the Cayman Islands. Bulletin of
the Institute of Jamaica, Science Series. (2): 1-56.
Griffiths, I. 1963. The phylogeny of the Salientia. Biological Reviews
8: 241-292.
Grigg-:, G. C. 1965. Studies on the Queensland lungfish, Neoceratodus
forsteri (Krefft). III. Aerial respiration in relation to
habits. Australian Journal of Zoology J_3_: 413-42!.
Grodins, F. S., A. L. Berman and A. C. Ivy. 1941. Observations on the
toxicities and choleretic activities of certain bile salts.
Journal of Laboratory and Clinical Medicine 27: 181 -186.

148
Hach Chemical Company. No date. Methods Manual for Mach Direct Read
ing Colorimeter, fifth edition. Ames, Iowa: Hach Chemical
Company.
Hartman, F. A. and M. A. Lessler. 1964. Erythrocytic measurements in
fishes, amphibia and reptiles. Biological Bulletin 126:
83-88.
Helff, 0. M. 1926. Studies on amphibian metamorphosis II. The oxygen
consumption of tadpoles undergoing precocious metamorphosis
following treatment with thyroid and di-iodotyrosine. Journal
of Experimental Zoology 45.: 69_93 -
Herner, A. E. and E. Frieden. 1961. Biochemical changes during anuran
metamorphosis. VIII. Changes in the nature of the red cell
proteins. Archives of Biochemistry and Biophysics 5.: 2535-
Holden, H. F. and R. Lemberg. 1939- The UV absorption spectra of bile
pigment iron compounds and of some bile pigments. Australian
Journal of Experimental Biology and Medical Science 17: 133-143-
Hopkins, H. S. and S. W. Handford. 1943. Respiratory metabolism during
development in two species of Amb1ystoma. Journal of Experi
mental Zoology 93.: 403-414.
Israels, L. G., M. Levitt, W. Novak, and A. Zipursky. 1966. The early
bilirubin. Medicine 4: 517-521.
Kawaguti, S., Y. Kamishima, and K. Sato. 1965. Electron microscopic
study on the green skin of the tree frog. Biological Journal,
Okayama University JJ_: 97-109-
Kimber, R. J. and H. Lander. 1964. The effect of heat on human red
cell morphology, fragility and subsequent survival in vivo.
Journal of Laboratory and Clinical Medicine 64: 922-933-
King, W. and T. Krakauer. 1966. The exotic herpetofauna of southeast
Florida. Quarterly Journal of the Florida Academy of Sciences
2£: 144-154.
KIrkberger, C. 1953- Temperaturadaptation der Sauerstoffbindung des
Blutes von Rana esculenta L. Zeitschrift fur Vergleichende
Physiologie 35: 153-158.
Krogh, A. 1941. The Comparative Physiology of Respiratory Mechanisms.
Philadelphia: University of Pennsylvania Press.
1850. De hepatis ranarum extirpatione. Dissertation, Berlin.
(From Lemberg and Legge, 1949)
Kunde.

149
Laessle, A. M. I96I. A mi ero-limnologica1 study of Jamaican bromeliads.
Ecology 4_: 499-517-
Lemberg, R. 1935- Transformation of haemins into bile pigments. Bio
chemical Journal 29: 1322-1336.
Lemberg, R. and J. W. Legge. 19^9- Hematin Compounds and Bile Pigments.
New York: Interscience Publishers.
Lester, R. and R. Schmid. 1961. Bile pigment excretion in Amphibia.
Nature 190: 452.
Lester, R. et_ aj_. 1966. Biosynthesis of tritiated bilirubin and studies
of its excretion in the rat. Journal of Laboratory and Clinical
Medicine 61000-1012.
Lonnberg, E. 193^+ Den grona fargen hos skelettet av lungfisken,
Protopterus annectens. Fauna och Flora 29: 278-279-
Lutz, A. 1924. Sur les rainettes des environs de Rio de Janeiro.
Comptes Rendus Biologique Paris 90: 241.
Lutz, A. and B. Lutz. 1938. Two new Hylidae: H_. a 1 bosignata n. sp. and
H_. pickeli n. sp. Annaes Academia Bras i 1 ¡ero Sci enti fi cas J_0: 185
Lutz, B. 1947. Trends towards non-aquatic and direct development in
frogs. Copeia 1947: 242-252.
. 1948. Anfibios Anuros da Colecao Adolpho Lutz. II. Especies
verdes do genero Hyla do Leste-Meridional do Brasil. Memorias
do Instituto Oswaldo Cruz 52_: 155-238.
. 195^. Anfibios Anuros do Distrito Federal. Memorias do Instituto
Oswaldo Cruz 2: 155-238.
Lynch, J. D. and H. L. Freeman. 1966. Systematic status of a South
American frog, Allophryne ruthveni Gaige. University of Kansas
Publications, Museum of Natural History J493-502.
Lynn, W. G. 19^0. The Herpetology of Jamaica I. The Amphibians.
Bulletin of Institute of Jamaica, Science Series (1): 1-60.
. 1958. Some amphibians from Haiti and a new subspecies of Eleuthero
dacty 1 us schmidti Herpetologi ca J_4: 153-157-
. 1961. Types of amphibian metamorphosis. American Zoologist Jk
151-161.
Lynn, W. G. and J. N. Dent. 1943- Notes on Jamaican amphibians. Copeia
1943: 234-242.

150
Manwell, C. 1966. Metamorphosis and gene action I. Electrophoresis
of dehydrogenases, esterases, phosphatases, hemoglobins and
other soluble proteins of tadpole and adult bullfrogs. Compar
ative Biochemistry and Physiology 17: 805-823.
McCutcheon, F. H. 1936. Hemoglobin function during the life history of
the bullfrog. Journal of Cellular and Comparative Physiology
8: 63-81.
Me 1 ichar, V. K. Polacek arid M. Novak. 1962. The relationship between
bilirubin concentration and the level of non-esterified fatty
acids in the blood of new-born infants. Biologia Neonatorum
4: 94-101.
Mertens, R. 1939- Herpetologische Ergebnisse einer Reise nach der Insel
Hispaniola, Westindien. Abhandlungen senckenburg natur
Gesellschaft 449: 1-84.
Miranda-Ribeiro, A. 1926. Notas para servirem ao estudo dos gymnobat-
raquios (Anura) brasileiros. Archivios Museo Nacional do
Rio de Janeiro 2_7: 7- (From Barrio, 1965a)
Moore, J. A. 1940. Adaptive differences in the egg membranes of frogs.
American Natura 1ist 4: 89-93-
Nakajima, H., T. Takemura, 0. Nakajima, and K. Yamaoka. 1963- Studies
on heme aIpha-metheny1 oxygenase I. The enzymatic conversion
of pyridine-hemichromogen and hemoglobin-haptoglobin into a
possible precursor of biliverdin. Journal of Biological
Chemistry 28: 3784-3796.
Negus, V. 1965- The Biology of Respiration. Baltimore: Williams and
Wi1kins Company.
Neill, W. T. 1958. The occurrence of amphibians and reptiles in salt
water areas, and a bibliography. Bulletin of Marine Science of
the Gulf and Caribbean 8_: 1 -97-
Nisimaru, Y. 1931. Bile pigment formation in the liver from hemoglobin.
American Journal of Physiology 97: 654-657-
Noble, G. K. 1923. In pursuit of the giant tree frog. Journal of the
American Museum of Natural History 2_: 104-116.
. 1927. The value of life history data in the study of evolution
of the Amphibia. Annals of the New York Academy of Science 30:
31-128.
. 1929- The adaptive modifications of the arboreal tadpoles
Hoplophryne and the torrent tadpoles of Staurois. Bulletin
of the American Museum of Natural History 8: 291-334.

New York: McGraw-
Noble, G. K. 1931- The Biology of the Amphibia.
Hill Book Company, Inc.
Noir, B. A., E. Rodriguez Garay, and M. Royer. 1965. Separation and
properties of conjugated biliverdin. Biochimica et Biophysica
Acta J_00: 403-410.
Panton, E. S. 1952. Our ground and tree frogs--g1impses into their
life and habits. Natural History Notes of the Natural History
Society of Jamaica (53): 87-92. Reprinted from the Daily
Gleaner, Kingston, Jamaica, December, 1922.
Parker, H. W. 1935. The frogs, lizards and snakes of British Guiana.
Proceedings of the Zoological Society of London 1935: 505-530.
Pamas, J. K., and S. Krasinska. 1921. Uber den Stoffwechsel der
Amphibien-larven. Biochemische Zeitschrift 116: 108-137-
(From Atlas, 1938)
Perkins, L. 1948. Life in a wild pine. Natural History Notes of the
Natural History Society of Jamaica (31): 86-90.
Peters, W. 1873. Ueber die von Dr. J. J. v. Tschudi Beschriebenen
Batrachier aus Peru. Monatsberichte Akademie Berlin, pp. 622-699-
(From Zoological Record, 1873, p. 95)
Peterson, H. W., R. Garrell and J. P. Lantz. 1952. The mating period
of the giant tree frog Hy la domi n i cens i s. Herpe to logi ca 8_: 63.
Podiapolsky, P. P. 1910. (UeberChlorophy11 bei Froschen). Biologi-
cheski i Zhurnal, Moskva 5-9-
Popper, H. and F. Schaffner. 1957- Liver: Structure and Function. New
York; Blakiston.
Powers, E. B. et a_L 1932. The relation of respiration of fishes to
environment. Ecological Monographs 2_: 385-473
Prosser, C. L. e_t aj_. 1950. Comparative Physiology. Philadelphia:
W. B. Saunders Company.
von Recklinghausen, F. 1883 Handbuch der allgemeine Pathologie der
Kreislaufs und die Ernahrunq. Stuttgart: enke. (From Lemberg
and Legge,1949)
Reeder, W. G. 1964. The digestive system. in: J. A. Moore, editor,
Physiology of the Amphibia. New York: Academic Press.
Rich, A. R. 1925- The formation of bile pigment. Physiological Reviews
182-224.

152
Rivero, J. A. 1961 Salientia of Venezuela. Bulletin of the Museum
of Comparative Zoology 126: 1-207.
Rodriguez Garay, E., B. Noir, and M. Royer. 1965- Biliverdin pigments in
green biles. Biochimica et Biophysica Acta 1001 411-417-
Ruud, R. T. 1959. Vertebrates without blood pigment: A study of the
fish family Chaenichthyidae. Proceedings of the XVth Inter
national Congress of Zoology, Section 6: 526-528.
Savage, J. M. 1967. A new tree frog (Centrolenidae) from Costa Rica.
Copeia 1967: 325-331.
Savage, R. M. 1950. A thermal function of the envelope of the egg of
the common frog, Rana temporaria temporaria Linn., with
observations on the structure of the egg clusters. British
Journal of Herpetology 3.: 57-
. 1961. The Ecology and Life History of the Common Frog (Rana
temporaria temporaria). London: Pitman and Sons, Ltd.
Schmidt, W. J. 1919. Vollzieht sich Ballung und Expansion des Pigmentes
in Melanophoren von Rana nach Art amoboider Bewegungen oder
durch intraze11u la re Kornchenstroming? Biologische Zentralblat
39: 140-194.
. 1920. Uber das Verhalten der verschiedenartigen Chromatophoren
beim Farbenwechse1 des Laubfrosches. Archiv fur mikroscopische
Anatomie3.: 414-455-
. 1921. Uber die Xantholeukosomen von Rana esculenta. Jena
Zeitschrift 2 (N- s- 50): 219-228.
Schnedorf, J. G. and T. G. Orr. 1941. The effect of anoxemia and oxygen
therapy upon the flow of bile and urine in the newbutalized
dog. 11. Its possible relationship to the hepatorenal syndrome.
American Journal of Digestive Diseases _8: 356-358.
Schreckenberg, M. G. 1956. The embryonic development of the thyroid
gland in the frog, Hyla brunnea. Growth 20_: 295-313-
Sherlock, S. 1964. Jaundice due to drugs. Proceedings of the Royal
Society of Medicine 881.
Starrett, P. 1960. Descriptions of tadpoles of Middle American frogs.
Miscellaneous Publications of the Museum of Zoology, University
of Michigan (1 10): 5~37-
Stejneger, L. 1905- Batrachians and land reptiles of the Bahama Islands.
In: G. B.Shattuck, editor, The Bahama Islands. New York:
MacMillan Company.

153
Stokstad, E. L. R. and J. Koch. 1967- Folic acid metabolism. Physio
logical Reviews 47.: 83-116.
Strawinski, S. 1956. Vascularization of respiratory surfaces in
ontogeny of the edible frog, Rana esculenta L. Zoolgica
Poloniae 327365-
Stuart, L. C. 1951. The distributional implications of temperature
tolerances and hemoglobin values in the toads Bufo mar inus
(Linnaeus) and Bufo bocourti Brocchi. Copeia 1951: 220-229.
Taylor, E. H. 1942. Tadpoles of Mexican Anura. University of Kansas
Science Bulletin 28.: 3 7_55
Umbreit, W. W., R.H. Burrissand J. F. Stauffer. 1964. Manometric
Techniques. Minneapolis: Burgess Publishing Company.
Varela, M. E. and Sellares, M. E. 1938. Variations annuel les du sang
du crepaud Bufo arenarum Hensel. Comptes Rendus Societe de
Biologie J2£: 1248-1249-
Vernberg, F. J. 1955. Hematological studies on salamanders in relation
to their ecology. Herpetologica _M_: 129-133-
Wagenaar, M. 1939- Inquiry into the identity of the green colour from
the spine of the sea pike. Archives neerlandaises de Zoologie
4: 103-105.
Williams, A. W. 1965- Causes and laboratory diagnosis of jaundice.
New Zealand Medical Journal 64: 486-491.
Willstaedt, H. 1941. Zur Kenntnis der grunen Farbstoffe von Seefischen.
Enzimo logia S_: 260.
Wright, A. H. and A. A. Wright. 1949- Handbook of Frogs and Toads of the
United States and Canada, third edition. Ithaca, New York:
Comstock Publishing Company, Inc.
Zweifel, R. G. 1964. Life history of Phrynohyas venulosa (Salientia:
Hylidae) in Panama. Copeia 1964: 201-208.

BIOGRAPHICAL SKETCH
Duvall Albert Jones was born October 17, 1933, near Hurlock,
Maryland. In June, 1951, he was graduated from Kenwood High School,
Essex, Maryland. He received the degree of Bachelor of Arts from
Western Maryland College in May, 1955- Following two years of military
service, he enrolled in the Graduate School of the University of
Maryland, from which he received the degree of Master of Science.
Mr. Jones taught biology courses at Madison College, Harrisburg,
Virginia, for two years. He accepted a position at Ferrum Junior
College, Ferrum, Virginia, in September, 1962. He took a leave-of-
absence from that position to enter the Graduate School of the University
of Florida in September, 1963- Later, he accepted positions at West
Liberty State College and Carnegie-Me1 Ion University while continuing
his graduate studies.
Duvall A. Jones is married to the former Dorothy Ann Paul.
He is a member of the American Association for the Advancement of
Science, the American Institute of Biological Sciences, the American
Society of Zoologists, the American Society of Ichthyologists and
Herpetologists, the Herpetologists' League and the Association of
Southeastern Biologists.
154

This dissertation was prepared under the direction of the
chairman of the candidate's supervisory committee and has been
approved by a 11 members of that committee. It was submitted to the
Dean of the College of Arts and Sciences and to the Graduate Council,
and was approved as partial fulfillment of the requirements for the
degree of Doctor of Philosophy.
December, 1967
Dean, Graduate School



BIOGRAPHICAL SKETCH
Duvall Albert Jones was born October 17, 1933, near Hurlock,
Maryland. In June, 1951, he was graduated from Kenwood High School,
Essex, Maryland. He received the degree of Bachelor of Arts from
Western Maryland College in May, 1955- Following two years of military
service, he enrolled in the Graduate School of the University of
Maryland, from which he received the degree of Master of Science.
Mr. Jones taught biology courses at Madison College, Harrisburg,
Virginia, for two years. He accepted a position at Ferrum Junior
College, Ferrum, Virginia, in September, 1962. He took a leave-of-
absence from that position to enter the Graduate School of the University
of Florida in September, 1963- Later, he accepted positions at West
Liberty State College and Carnegie-Me1 Ion University while continuing
his graduate studies.
Duvall A. Jones is married to the former Dorothy Ann Paul.
He is a member of the American Association for the Advancement of
Science, the American Institute of Biological Sciences, the American
Society of Zoologists, the American Society of Ichthyologists and
Herpetologists, the Herpetologists' League and the Association of
Southeastern Biologists.
154


64
tadpoles died within three hours, but at 5 parts per thousand salinity,
a tadpole lived for ten days before it was returned to fresh water.
Since an isotonic solution for these tadpoles is presumably 7 parts
per thousand, the first tadpole was able to maintain itself in a hyper
tonic solution for near four days.
A number of preserved young Hyla septentriona1?s from Grand
Cayman showed green bone pigmentation, as did a number of the larger
specimens. Of more than 260 live adults collected from Miami, Florida,
during June, 1966, none showed green pigmentation in bones; however,
Clarence McCoy (personal communication) collected specimens with green
bones from Homestead, Florida, during August, 1967- I was unable to
determine more about the development of green pigmentation in this species
because it did not undergo metamorphosis in the laboratory. 1 found
the skin secretion of this species to be extremely irritating if rubbed
into the eyes; it is also difficult to wash from the eyes.
Stream habitats
Flowing streams constitute the primary habitat of several species
of tadpoles in the West Indies and South America. Temperature and
oxygen concentration within these streams apparently have a narrow range
of daily fluctuations. Streams inhabited by tadpoles were found to have
temperatures between 20 C. and 25 C. in most cases. Since these streams
were in hilly or mountainous areas, it seemed that the higher altitude
was related to the relatively low water temperatures. Quite regularly,
direct sunlight was blocked by clouds during the afternoon. These factors,


38
water other than that in the bromeliads. Steepness of the hills and
porosity of the limestone substrate are partly responsible for rapid
run-off of water.
Observations of the adults of the three smaller Jamaican hylid
species indicate that they prefer to be covered by water, at least in a
lighted area. Under these circumstances, they may remain completely
immersed for minutes at a time, and then slowly rise until only the
external nares and eyes protrude above the surface. It is interesting
to note that the four Jamaican species differ from their relatives of
Cuba and Hispaniola in having a more truncate snout with the external
nares at the most anterodorsal point (Dunn, 1926). This may be inter
preted as an adaptation to living in the reservoirs of bromeliads.
Laessle (1961) studied the ecology of Jamaican bromeliads,
in which he found the following ranges for the water which they con
tained: dissolved oxygen, 0.0 8.0 ppm; dissolved carbon dioxide,
4.0 67-0 ppm; pH, 4.0 7-0; temperature, 17-5 30.0. He estimated
the maximal quantity of water in the reservoir of a large bromeliad at
200 mi 11 i 1iters.
Hyla brunnea. -- The brown tree frog of Jamaica is the most
widely distributed species of Hy1 a on the island. It is absent from
the Blue Mountains above 1600 meters elevation, and from arid areas
such as Kingston and the Hellshire Hills along the south coast (Lynn,
1940). This frog is most likely to be found in localities where the
large tank bromeliads, including species of Hohenberqia and Aechmea, are
readily available as breeding sites and resting places. While this frog


31
Seasonal changes in chlorosis
It should be noted that most of the specimens for the present
study were collected from June through August. This period constitutes
the main breeding period for most of the species studied, a fact which
should be kept in mind when considering the physiology of these organisms.
It is interesting to note that Hy la qeoqraph? ca had green bones,
but no green pigment in the plasma. On the other hand, Hy 1a mi sera had
light green plasma, but white bones. These two examples indicate that
high concentrations of green pigment in the plasma are temporary in some
species. Different shades of green in plasma and bones of other species
tend to support this idea. Concentric layers of different shades were
seen in Barrio's (1965a) photograph of bone from Lysapsus mantidactylus,
and in a femur of Osteocepha1 us taurinus. Like growth rings of a tree,
these suggest a seasonal change in conditions during development of
the tissue.
Phylogenetic distribution of chlorotic frogs
An understanding of the phylogenetic relationships of chlorotic
frogs to those which lack tissue biliverdin should be of value. However,
Tables 1 and 2 indicate no clear boundaries along phylogenetic lines.
As previously mentioned, adult individuals of Hyla septentriona11s and
Hyla dominicensis may fall into either category. In Hyla brunnea,
the pigment is always present in young frogs, but has never been observed
in the adults. Hyla pul che 1 la has some populations which are chlorotic
and others which are not. Similarly, it can be seen that neither the
genus Hyla nor the family Hylidae shows uniformity in this characteristic.


94
difference between red cell survival times of chlorotic and achlorotic
frogs.
Blood smears from adults of Hvla crepitans, Hyla lanciformis,
Sphaenorhvnchus aurantiacus and a tadpole of Pseud?s paradoxus have
demonstrated that the majority of erythrocytes of these Individuals
were hemolyzed within a few hours after the samples were taken. The
changes in cell structure or shape which preceded hemolysis are not known.
Likewise the actual cause of hemolysis is unknown. Since unconjugated
bilirubin is known to cause hemolysi s i n vi tro (Cheung e_t a_l_. 1966),
it seems likely that plasma bilverdin induced hemolysis in the species
listed above, with the exception of Hyla lanciformis. This species is
not chlorotic. Conversely, the red blood cells of several other species
did not hemolyze in their green plasma. From these observations it would
appear that biliverdin is not necessarily responsible for hemolysis,
or that it is effective only under certain conditions or in certain species.
Coagulation of red cell cytoplasm was another notable feature
of some blood smears. The species of Atelopus, Hyla and Leptodacty1 us which
demonstrated this character have little or no tendency toward chlorosis.
Within this group, the small species of Hyla, such as Hyla boesemani ,
Hyla leucophyllata. and Hyla minuta, as well as the chlorotic Sphaenorhynchus
aurantiacus are found in relatively exposed breeding habitats. Since it
has been shown that exposure to a temperature of 45 C. causes coagulation
of frog erythrocyte cytoplasm in vitro, it appears that high environmental
temperatures are responsible for this condition under natural conditions.
Tadpoles of Hyla dominicensis have been observed at 40 C. in nature for


108
TABLE 1. OCCURRENCE OF
GREEN
COLOR IN TISSUES AND
ANURANS
FLUIDS
OF ADULT
P Green color present
A Green
color absent
FAMILY AND SPECIES
BONE
SOFT
PLASMA
Bl LE
SOURCE
TISSUES
Rhinophrynldae
Rhinophrynus dorsalls(l)
A
A
A
-
Ranidae
Phv1lobates sp.(1)
A
A
A
RhacophorIdae
Hyperolius sp.
Microhy1idae
-
p (eggs)
-
-
M. Stewart
(Pers. Comm.)
Elachistocleis ovale(3)
A
A
A
P
Pseudidae
Lysapsus limellum laevis
P
-
_
-
Parker, 1935
Lysapsus limellum limellum -
P
Barrio, 1965a
Lysapsus mantidacty1 us
P
-
P
-
Barrio, 1965a
Pseudis minuta
P
-
-
-
Peters, 1873
Pseudis paradoxus
paradoxus
P
p
A
P
Pseudis paradoxus
pa radoxus(3)
-
A
A
P
Pseudis paradoxus
pla tens i s
P
-
P
-
Barrio, 1965a
Camerano, 1879
Bufnidae
Bufo typhonius(3)
A
A
A
P


9
displace bile pigment from serum proteins in new-born infants (Melichar,
Polacek and Novak, 1962). Assuming that this also occurs in frogs, one
would expect that when biliverdin exceeds the carrying capacity of
the serum proteins, it wi11 be deposited in the non-fluid tissues,
staining them with its color. Evidence of the staining by biliverdin
is available in many frogs, particularly where the gall bladder lies
adjacent to the stomach wall.
Sources of Biliverdin
Hemoglobin and other hemoproteins such as myoglobin and
oxidizing enzymes constitute the main sources of bile pigments. Although
most studies of bile pigment formation are concerned with the origin
of bilirubin in mammals, we may consider the origins of biliverdin to
be the same, since it is generally considered to be a precursor of
bi 1irubin.
In a recent paper, Israels e_t a_L (1966) proposed a scheme in
which there are four components of human bilirubin. Each of these
components is characterized by a peak concentration of radioactive
bilirubin at a specific interval after the labeling of bilirubin pre
cursors. The major bilirubin component is formed from hemoglobin of
old red blood cells and makes its appearance about 100 days after the
radioactive isotope is administered. The lesser components appear more
rapidly; collectively, they are referred to as early bilirubin. One of
these reaches a peak three to five days after ingestion of the radio-
actively labeled compound and is believed to originate from a heme loss


60
it is similar to that of stream tadpoles in lacking the fin extension
on the back.
Hyla septentriona1?s. Of the native West Indian hylids,
Hyla septentriona1is is the most widely distributed. This medium-sized
tree frog is found throughout Cuba, where it is most abundant in the banana
groves of the lowlands (Barbour and Ramsden, 1919). In addition it is
common and perhaps has been introduced into the Cayman Islands, the
Bahama Islands, the Florida Keys and southern Florida (Barbour, 1937).
Barbour (1931, 1937) believed that this species was introduced into
Key West on freight cars from Cuba. There are indications of subsequent
introductions and range extensions in south Florida (King and Krakauer,
1966). Grant (19^+0) summarized references which indicated that this
species was accidentally introduced into the Cayman Islands, but did
not maintain itself on Cayman Brae. However, William Greenhood (personal
communication) noted this species on Cayman Brae during the summer of
1965. Although there has been some question about its distribution on
Grand Cayman, I have found it there at two localitiesnear North Side
and just east of Boddentown--as well as in the Georgetown area.
Specimens have been seen from all of the major islands of the Bahamas
as far southeast as Acklin Island. A specimen of Hyla septentriona1?s
in the Institute of Jamaica was collected during the spring of 1965
at Highgate, St. Mary's Parish, Jamaica. This specimen resulted from
the introduction of tadpoles from Hialeah, Florida, during the spring
of 1961 (Edwin Todd, personal communication); there is no evidence that
this species has bred in Jamaica.


37
The brome liad microhabitat
Members of the pineapple family, Brome 1iaceae, ordinarily
do not constitute the dominant plants of a habitat, although they may
be an important part of the flora. On the West Indian island of
Jamaica, the bromeliads have undergone considerable adaptive radiation
(Dr. Richard Proctor, personal communication). They may be large or
small, epiphytic or terrestrial, and are found in shaded and open areas.
This appears to be a very fortunate circumstance since the water which
is caught in the leaf bases and central reservoirs of bromeliads is
the most reliable supply for small animals, including the four species
of tree frogs on Jamaica.
The importance of the close relationship between the Jamaican
tree frogs and the bromeliads should not be underestimated. Perkins
(1348) noted that the water level in the bromeliads ("wild pines")
is maintained by dew which condenses and runs down into the reservoir
in the center of the plant; very little direct sunlight tends to reduce
evaporation from the wild pines. She states further (p. 87), "In view
of the many creatures that depend on the wild-pine for moisture it would
seem that these plants hold an important place in the economy of the
countryside, for surely our wildlife would be largely depleted during
a severe drought, were it not for these hidden stores of water." My
observations during the drought which continued into July, 1965, sub
stantiate this position. Except for an occasional pool in stream beds,
or the largest rivers which continued to flow, there was no surface


4
its bile pigment. Lester and Schmid (1961) found biliverdin in some
samples of frog bile, but believed bilirubin to be the main bile pig
ment of anurans. A claim was made by von Recklinghausen (1883) that
formation of biliverdin takes place in sterile frog blood. Rich (1925)
indicated that this had not been confirmed. Cabello (1943) caused the
formation of green serum in Bufo arenarum by administration of phenyl-
hydrazine and ligation of the bile duct. Rodriguez Garay et_ aj_. (1965)
found biliverdin glucuronide in the bile of Bufo arenarum.
The first report of externally visible green pigment in
frogs appears to be the one by Peters (1873), who indicated that the
skeleton of Pseudis mlnuta was green. Camerano (1879) believed the
green bones of Pseudis paradoxus to be due to the presence of ferrous
phosphate. The recorder of the Zoological Record (Boulenger, 1880)
took exception to Peters' paper by stating that Pseudis minuta does
not have green bones. However, Boulenger (1883) took note of the green
eggs of pseudids. Fernandez and Fernandez (1921), Miranda-Ribeiro (1926),
Parker (1935), and Gallardo (1961) also noted green eggs or tissues in
this family.
In 1910, Podiapolsky described a "chlorophyll" pigment from
Hyla arbrea and Rana esculenta, two European species. It may be noted
here that reports by Schmidt (1919-1921) upon these two species, and by
Kawaguti, Kamishima and Sato (1965) upon Hyla arbrea, do not support
this. More recently, Dunn (1926), A. Lutz (1924, 1938), B. Lutz
(1948, 1954), Cochran (1955), Lynn (1958), and Bokermann (1964) have
noted green coloration of epithelia, muscles and bones in certain


145
Carnerario, L. 1879. Colorazione naturale del le ossa di una specie di
anfibio anuro. Atti Accademia Torino _1: 789-794.
Cantarow, A. and B. Schepartz. 1957- Biochemistry, second edition,
Philadelphia: W. B. Saunders Company, p. 601.
Cheung, W. H. e_t aj_. 1966. The effect of bilirubin on the mammalian
erythrocyte. Transfusion 6^: 475-486.
Cloudsley-Thompson, J. L. 1965. Rhythmic activity, temperature-
tolerance, water-relations and mechanisms of heat death in
a tropical skink and gecko. Journal of Zoology 146: 55-69-
Cochran, D. M. 1941. The Herpetology of Hispaniola. U. S. National
Museum Bulletin 177-
. 1955- Frogs of Southeastern Brazil. U. S. National Museum
Bulletin 2 06.
Cohen, B. and A. H. Smith. 1919- The colorimetric determination of
hemoglobin. A practical procedure. Journal of Biological
Chemistry 489-496.
Conant, R. 1958. A Field Guide to Reptiles and Amphibians. Boston:
Houghton Mifflin Company, p. 199-
Dickerson, M. C. 1906. The Frog Book. New York: Doubleday, Page
and Company.
Dowling, M. I960. Interlude at Simla. Animal Kingdom 6: 137-139-
Drastich, L. 1927- Uber das Leben der Salamanderlarven bei hohem und
niedrigem Sauerstoffpartialdruck. Zeitschrift fur Vergleichende
Physiologie 2\ 632-657- (From Foxon, 1964)
Dubin, I. N. 1958. Chronic idiopathic jaundice. American Journal of
Medicine 24: 268-292.
Duellman, W. E. and A. Schwartz. 1958. Amphibians and reptiles of
southern Florida. Bulletin, Florida State Museum : 181-324.
Dunn, E. R. 1926. The frogs of Jamaica. Proceedings of Boston Society
of Natural History 8: 111 -130.
1931- The amphibians of Barro Colorado Island. Occasional Papers
of Boston Society of Natural History 403-421.
Elias, H. 1943. Cause for blue, green and red color in Anura. Anatomical
Record 8]_ (Supplement): 440.


as these, and the knowledge that globin and iron are separated from
the tetrapyrrole, it was generally assumed that hemoglobin was degraded
as follows: hemoglobin, hematin, hematoporphyrin, bilirubin, with bili-
verdin a secondary oxidation product of bilirubin (Lemberg and Legge,
1949). Later, Lemberg and others (see Lemberg and Legge, 1949; Foulkes,
Lemberg and Purdom, 1951) succeeded in finding a sequence of relatively
mild reactions which degraded hemoglobin to biliverdin. Although
these reactions were not well understood and yielded only 15 per cent
of the biliverdin expected, workers in the field generally accepted
Lemberg's series of reactions as a hypothetical model of hemoglobin
degradation in cells.
Enzymatic degradation of hemoglobin. Recent studies of
Nakajima e_t aj_. (1963) have demonstrated an enzymatic pathway which is
capable of degrading hemoglobin to biliverdin. These workers have
characterized an enzyme which oxidizes the a lpha-methyne bridge of
the heme group to form a possible precursor of biliverdin. Most of this
work was done with the pyridine hemichrome rather than hemoglobins, but
their study of substrates is of considerable interest. The enzyme did
not act upon alkaline hematin or protoporphyrin IX, and only weakly
upon hemoglobin (contains ferrous ion) and hemiglobin (contains ferric
ion). However, reaction of the enzyme with a complex substrate of hemo
globin and haptoglobin (a plasma protein which combines with extracellular
hemoglobin), produced 49 per cent of the theoretical yield of biliverdin.
With the hemoglobin-haptoglobin complex as a substrate, there was a lag
period of five minutes before any noticeable reaction took place. There


50
Osteocephalus taurinus. A female with green bones was taken
from the trunk of a tree on the upper coastal plain in Surinam. The tree
was beside a road through a we 11-developed forest; a stream and standing
puddle were nearby.
Bokermann (1965) has recently given the first description of
the breeding habits of Osteocepha1 us taurinus. Rainfall in the vicinity
of Marmelo, in forests of western Brazil, is strongly seasonal, but
reaches 2250 millimeters annually. Osteocepha1 us taurinus choruses were
heard at the time of the first rain in November. Two of the breeding
pools used by the species were in the forest; in one of these ponds, there
was breeding activity at 1:00 P.M. The surfaces of these ponds were
covered by eggs of this species which were 5 millimeters in diameter,
including the jelly envelopes. A temperature reading from one of them
indicated the water temperature to be 27 C., while the air temperature
was 34 C.
Of particular interest was a third breeding pond found by
Bokermann. This pool was in an open area and contained the dead crown
of a tree which made difficult the capture of frogs beneath the branches.
However, as the collectors approached the pool, the frogs followed their
habit of rapidly climbing upward to escape capture until they reached
the ends of the tree's bare branches. After eight minutes in this
exposed position, the first frog fell into the pool, motionless but not
dead. Although its skin had dried in several places, it recovered. Air
temperature in the sun was 46 C. and the water temperature was 32 C. at
several points of the pond. Two days later, evaporation had greatly


I. CAUSATIVE PIGMENT AND INCIDENCE OF CHLOROSIS
Pigment Identification
During the present study, filtered tissue extracts and body
fluids were tested spectrophotometrica1ly and chemically. Figures 2-6
show that the spectral characteristics of several of these green fluids
are very similar to those of biliverdin (Figure 1). Addition of fuming
nitric acid to the fluids was followed by the succession of colors
known as the Gme1in reaction, generally considered to indicate the
presence of bile pigments. These two tests, along with characteristics
of color and solubility, appear to eliminate from consideration all
known green pigments except the two naturally occurring green bile
pigments, biliverdin and mesobi1iverdin. Since green bone is rapidly
bleached by concentrated sulfuric acid, mesobi1iverdin does not appear
to be present, since it is stable in this substance. Thus, it was
independently concluded during the present study that biliverdin is
the pigment primarily responsible for the green color of these frog
tissues.
Survey of green pigmentation among Neotropical anurans
During the course of several field trips to the West Indies
and one to Surinam, it was possible to study a variety of tropical
anurans under natural conditions. Many of these did not exhibit any
green color, while others showed different levels of green pigmentation
in the tissues and body fluids. Observations made during the present
27


23
Structure and Function of Froa Liver
Studies of liver structure in frogs have been few; most were
considered by Elias and Bengelsdorf (1952). These authors found that
the walls separating neighboring sinusoids are predominantly two cells
thick in frogs, but only one cell thick in mammals. Similarly, relatively
little experimental work has been done on the excretion of bile pigment
by frog liver. Nisimaru (1930 carried out perfusion experiments on the
liver of Rana catesbeiana. He found that the rate of bile pigment
excretion changed when blood pressure in the liver circulation was
a 1 tered.
Papers concerning frog liver structure or function do not
aid in the explanation of chlorosis in these forms.


87
related to green pigmentation (Figures 12-16). A list of these and
some others, according to the organisms in which they were observed,
fo 11ows:
Pseudidae
Pseudis paradoxus -- adult male Cells of this individual
tended to form cytoplasmic blebs; tadpole blood hemolyzed
upon standing overnight.
Bufnidae
Bufo typhonius -- adult an erythroplastid was observed
in the blood of this individual.
Atelopodidae
Atelopus sp. -- male erythrocytes showed coagulated
cytop 1 asm.
Hy 1 idae
Hyla boeseman? adult cells have coagulated cytoplasm,
as occurs after slight heating.
Hyla crepitan s -- adult most cells hemolyzed.
Hyla egleri -- male cells not unusual.
Hyla qeographica -- adult male nothing unusual.
Hy la lane? formis adult most of the cells were hemolyzed.
Hy la leucophy1 lata -- about half of erythrocytes have
coagulated cytoplasm.
Hyla minuta adult most of the cells have coagulated
cytoplasm as occurs on heating.
Hyla mi sera -- some cells have odd shapes.
Hyla maxima -- adult many lymphocytes present.
Hyla punctata -- adult cell membranes are irregular; green
cytoplasmic and extracellular granules present on stained
slides. Erythrocytes appear immature.
Hyla rubra adult nothing unusual.
Osteocepha1 us taurinus -- adult many erythrocytes with
vacuo la r nuc lei.


81
Physiological Characteristics of Chlorotic and
Non-Chlorotic Frogs
This portion of the study was directed toward finding physio
logical differences between chlorotic and non-chlorotic frogs. Frog
blood, tadpole respiratory rates, and histological sections of liver
were the main subjects of study.
Methods of study
Hemoglobin determinations. -- The amounts of hemoglobin pre
sent in anuran blood was determined by the acid-hematin test (Cohen and
Smith, 1919). A Heilige hemoglobinometer was used for this purpose,
since much of the work was done under field conditions.
Blood was collected from a large blood vessel or the heart
into heparinized capillary tubes. From these, 20 microliter samples of
blood were transferred to a graduated test tube which contained j milli
liter of 1 per cent hydrochloric acid solution. After mixing the
contents of the tube, they were allowed to stand for ten minutes in
order to insure complete hemolysis of the red blood cells. Then distilled
water was added until the color of the solution matched the color stand
ard of the hemoglobinometer. The calibration mark at the meniscus of
the fluid indicated the hemoglobin concentration of the blood in grams
per hundred milliliters.
Red blood cell counts. Blood samples obtained as for hemo
globin determinations were diluted and counted by standard techniques
using a Spencer Bright Line hemocytometer. Four counts were made for
each individual and the average was used as the red blood cell count.


TABLE 4 (Continued)
LOCATION, HABITAT AND DATE
33- 4.8 km W of Bog Walk (9 VIII)
34. Kellitts (17V)
Streams of Puerto Rico
35- El Yunke habitat of Leptodacty1 us
a 1bilabrus (6 V)
TEMPERATURE
OXYGEN(ppm)
HARDNESS(ppm)
J2
24.5
8.6
-
7
22.0-22.5(2)
8.2-8.4(2)
80-130(2)
6.5-7-0
21.0
3.1
20
6


113
TABLE 2. OCCURRENCE OF GREEN PIGMENT AMONG METAMORPHOSING NEOTROPICAL
ANURAN
TADPOLES
SPECIES
BONE
SOFT TISSUES
PLASMA
BILE
PSEUD 1 DAE
Pseud is paradoxus
P
P
P
P
HYLIDAE
Hyla brunnea
P
P
-
P
Hyla dominicensis
P
P
-
P
Hy la hei1prini
P
P
P
P
Hyla lichenata
P
P
-
-
Hyla marianae
A
A
-
-
Hyla septentriona1is
P
P
-
-
Hyla vasta
P
P
-
-


58
hind limbs, but were in metamorphosis on March 4. Mertens found Hyla
domin?censis in bromeliads in the coniferous forests, but did not
state that they bred there. Noble (1923) indicated that he found tad
poles of this species at 2500 meters elevation in wheel ruts. Lynn
(1958) found tadpoles of this species in a small pool at the head of a
stream.
I found this species only in stagnant water: in a shaded pool
in a dry stream bed, in a flooded wiregrass pasture exposed to the sun,
in wheel ruts of a muddy road which were also exposed to the sun, and
in concrete-1ined pits which had been used for tanning hides. In the
latter two habitats, the water contained so much solid material and algae
that the tadpoles could not be seen except when they came to the surface
to take air. The constant agitation of the water by the dozens or
hundreds of tadpoles appeared to be largely responsible for the turbidity.
When the flooded pasture was visited in May, a large cluster
of tadpoles was noted in one flooded corner. After tadpoles from the
cluster came to the surface, they ordinarily returned to the group.
This behavior continued until my presence caused the group to disperse.
It is of particular interest since the temperature of the water was
above 32 C. and the oxygen concentration averaged 0.7 parts per million.
Brattstrom (1962) has shown that such aggregations of dark tadpoles
absorb more radiant heat than do isolated individuals. He found water
temperatures near such groups to be higher than water temperatures at
a distance. He reasoned that the resultant higher body temperature
caused an increase in metabolic rate, which decreased the time required
for development.


TABLE 1 (Continued)
FAMILY AND SPECIES BONE SOFT PLASMA
TISSUES
Centrolenella pulveratum P P
Centrolene1 la reticulata green eggs
Centrolene1ia spinosa P -
112
BILE SOURCE
Dunn, 1931
Starrett, i960
Savage, 1967
Barrio, 1965a
Centrolene1 la vanzolinii P


TABLE 1 (Continued)
FAMILY AND SPEC IES
PLASMA
BONE SOFT
TISSUES
Phyllomedusa sauvagii
Phrynohyas venulosa P
Phrynohyas venulosa(4) P P
Sphaenorhynchus
aurantiacus (5) P P
Sphaenorhynchus dorisae P
Sphaenorhynchus habrus P
Sphaenorhynchus orophilus P
A
P
P
P
T rachycepha1 us
nigromaculatus P
Leptodacty 1 idae
Eleutherodacty1 us
(several species) A
Leptodacty1 us sp.(2) A
Leptodacty1 us
pentadacty 1 us(1) A
Centrolenidae
A A
A A
Centro lene 1 la albomaculata P
Centrolenel la
fleischmanni
A
-
Centro lene 1 la
fleischmanni
green eggs
-
Centrolenella qranulosa
P
Centrolenella ilex
P
-
Centro 1 ene 1 la
prosoblepon
P P
-
BILE SOURCE
Barrio, 1965a
Barrio, 1965a
Coin, 1957
Coin, 1957
Coin, 1957
Cochran, 1955
P
P
Savage, 1967
Dunn, 1931
Starrett, I960
Savage, 1967
Savage, 1967
Dunn, 1931


131
TABLE 10.
FROGLET RESPI
RATORY RATES MEASURED
IN WARBURG RESPiROMETERS
SPECIES
WEIGHT(mq)
OXYGEN CONSUMED
(25 C.)
(ul 0?/q/hr)
OXYGEN CONSUMED
(35 C.)
(ul 0?/q/hr)
Hyla brunnea
1 12
167
482
120
101
548
128
279
61 1
135
134
395
215
152
406
231
211
471
237
159
459
286
119
331
Four days beyond
metamorphos Is
Hyla brunnea
225
549
Hyla lichenata
282
487


TABLE 4.
WATER CHARACTERISTICS OF TADPOLE HABITATS
LOCATION. HABITAT AND DATE
TEMPERATURE
OXYGEN(ppm)
HARDNESS(ppm)
pH
Bromeliads of Jamaica habitat of
Jamaican tree frogs and tadpoles
1. Rose Valley (25 X, in sun)
24.0-28.0(6)
-
-
5-7
(25 X, in shade)
23.0-26.0(17)
-
-
5.5-7
(14 V, in sun)
25-5-28.5(4)
1.4-2.4(4)
20
5
(14 V, in shade)
25.5-25.7(3)
1 .2-2.4(3)
40
-
(10 VI11 in sun)
27.2
2.7
-
5-5
(10 VIII, in shade)
24.8-25.3(3)
3.0-3.2(3)
-
6.5-7.0
2. 10 km NW of Troy (16 V)
24.0-26.1(3)
1.6-1.9(3)
50
5-5
3. 6 km S of Mandevilie (8 VI1 I)
22.9-23-7(4)
2.3-6.6(4)
-
5.6
4. Point Hi 11 (17 V)
22.5
1. 1
-
5.5
5- Cavalier District (27 X)
25.0-27.0(8)
-
8-40
5-6
6. 1.6 km W of Wakefield (27 X)
25.0-25.5(3)
-
10
5-5-5
7. 6.4 km NW of Juan de Bolas (27 X)
(without dust on leaves)
24.0
-
16-30(3)
5-5-6.5
(with dust on leaves)
24.5
-
70-90(3)
7.0
\


127
TABLE 7. AVERAGE DIMENSIONS OF FROG RED BLOOD CELLS
FAMILY AND SPECIES AVERAGE CELL AVERAGE CELL
LENGTH +
2 S.E.
WIDTH +
2 S.E
Pseudidae
'Pseudis paradoxus
13.2 +
0.73
12.0 +
0.30
Bufnidae
Bufo typhonius
16.3 +
0.92
12.9 +
0.39
Hy1idae
*Hyla boesemani
16.3 +
0.63
11.9 +
0.38
Hyla eqleri
17.4 +
1.48
12.6 +
0.62
Hyla qeoqraphica
15 8 £
0.30
9.9
1.74
Hyla leucophy1 lata
16.7 +
0.61
13.4 +
0.44
Hy 1 a maxima
18.3 +
0.48
11.0 +
0.33
Hyla minuta
15.1 +
0.40
11.7 +
0.43
*Hyla misera
14.0 +
0.54
12.4 +
0.48
*Hyla punctata
17.9 +
1.05
12.8 +
2.00
Hyla rubra
13.9
0.43
11.4 +
0.24
'Osteocepha lus taurinus
13.1
0.36
10.8 +
0.40
Phyllomedusa bicolor
17.0 +
0.73
13.7 +
0.66
Phyllomedusa hypochondria 1is
20.6 +
0.89
14.5 +
1.29
*Phrynohyas venulosa
12.8 +
1.11
10.8 +
0.27
'Sphaenorhynchus aurantiacus
21.6 +
0.47
13.2 +
0.39
Leptodacty1idae
Leptodactv1 us pentadactv1 us
17.2 +
0.42
12.4 +
0.33
Figures represent average for twenty-five cells from a single
of each species.
PRODUCT OF CELL
LENGTH x WIDTH
158.4
210.3
194.0
219.2
156.4
223.8
201.3
176.7
173.6
229.1
158.5
141.5
232.9
298.7
138.2
285.1
213.3
i ndi vidua1
* Chlorotic species


29
When green pigmentation was present in tadpoles, usually
it made its appearance at the onset of metamorphosis and reached a
peak upon the completion of metamorphosis. All of the species of Hyla
listed in Table 2 are from the West Indies and they present similar
appearances at the completion of metamorphosis, except for Hyla marianae.
Most of their skeletons were green, particularly the limb bones and
vertebrae; the green pigmentation was not so pronounced where the marro;
maintained its hemopoietic function; melanin pigmentation had developed,
especially on the dorsal side; pigmentation of the ventral side was not
well developed in Hyla brunnea, Hyla 1 ichenata, or Hyla marianae, all
of Jamaica (stomach contents could be seen through the skin); a white
substance, presumably guanine, had developed in the skin of the other
species, which are usually found in more exposed habitats than the
Jamaican species; green pigmentation of soft parts was particularly
intense in the region of the throat and pectoral girdle; the small size
of most of these species of tadpoles made it difficult to determine
the color of the plasma. In Surinam, the tadpoles of Pseudis paradoxus
were found to be darkly pigmented, externally by melanin and internally
by biliverdin and other pigments. The plasma of this species is green
prior to metamorphosis. The coiled intestines of these gigantic tadpoles
are packed with green plant materials and fill most of the body cavity.
During the present study, a young Rana heckscheri collected
near Gainesville, Florida, at the completion of metamorphosis was found
to have gray-green bone marrow. This appeared to be a stage which
represented degenerating red marrow of the tadpole. This color appeared


LIST OF TABLES
Table Page
1 Occurrence of Green Color in Tissues and
Fluids of Adult Anurans 108
2 Occurrence of Green Pigment Among Metamor
phosing Neotropical Anuran Tadpoles 113
3 Locality Data of Tadpole Habitats 114
4 Water Characteristics of Tadpole Habitats 118
5 Iron and Phosphate Concentrations (Parts
Per Million) in Tadpole Habitats 123
6. Blood Characteristics 124
7 Average Dimensions of Frog Red Blood Cells 127
8 Respiratory Rates of Tadpoles in Warburg
Respirometers at Different Temperatures 128
9 Tadpole Respiratory Rates at 25 C.
(Measured by Jones Method) 129
10Frog let Respiratory Rates Measured In
Warburg Respi rorneters 131


90
similar to those which had iron deficiency. Very little movement of
the tadpoles was evident when the water temperature in their containers
was 16.8 C. Lynn (19^0) suggested that the lack of large bromeliads
above 1600 meters in the Blue Mountains may be the reason for Hyla brunnea
not being found at such altitudes. Low temperatures at these altitudes may
also be a factor which prevents this species from moving to higher
elevations.
Liver histology. -- Study of liver sections of ten species of
frogs from Surinam is of interest (Figures 17_20). No dead or completely
degenerated tissue was seen in these slides but there were several indi
cations of conditions or activities which may cause changes in bile
excretion. There did not appear to be any obstructions to the bile
duct or gall bladder in the majority of individuals examined, but those
with green tissues and plasma usually had very concentrated bile in the
gall bladder, judging from its dark blue color. A resume of liver histo
logical characteristics follows:
Pseudidae
Pseudis paradoxus -- adult male liver apparently normal
except for very heavy pigmentation (the liver appeared black
before dissection);
-- adult female as in male, with possible
cell fragments in blood of liver;
-- tadpole liver structure apparently nor
mal; much less pigmentation than in adults; cell fragments
in blood of the liver; many immature erythrocytes present in
circu lating blood.
Atelopodidae
Atelopus sp. -- male some liver cells appear enlarged and
have rarified cytoplasm; some pyknotic nuclei present; other
wise apparently normal.
Hy 1 idae
Hyla geographica -- adult male veins of liver partly
occluded; considerable pigment present in liver cells.


TABLE 3- LOCALITY DATA OF TADPOLE HABITATS
LOCATION AND HABITAT DATA
Bromeliads of Jamaica habitat of Jamaican
tree frogs and tadpoles
I Rose Va 1 ley
2. 10 km NW of Troy
3. 6 km S of Mandevi1 le
4. Point Hill
5. Cavalier District, St. Andrew
6. 1.6 km W of Wakefield
7. 6.4 km NW of Juan de Bolas
Bromeliad of Puerto Rico habitat of no
known tadpoles
8. El Yunke Biological Station
Ponds and standing waters of Haiti
Hyla dominicensis
habitats of
DATE
ELEVATION AIR
(Meters) TEMPERATURE
25 X 1965
490
-
14 V 1966
490
28.5-33.5
10 VIII 1966
490
26.4-31.2
16 V 1966
750
26.0-27.5
8 VI11 1966
550
24.7-28.2
17 V 1966
600
25.0
27 X 1965
600
26.7
27 x 1965
350
28
27 x 1965
600
24
6 V 1966
800


36
Temperature readings of tadpole habitats were taken with the
thermistor component of the oxygen analyzer wherever possible; other
wise, they were made with standard mercury thermometers, graduated
from -10 C. to 100 C.
Ferrous and total iron, and phosphate concentrations of water
were measured by colorimetric methods using a Hach portable colorimeter
and the appropriate techniques (Hach Chemical Company, no date).
Salinity, alkalinity and hardness were measured by titrametric methods,
using the appropriate kits and techniques developed by the La Motte
Chemical Company. Water pH was measured to the nearest half unit by
means of pHydrion paper.
Observations of Frog Habitats and Behavior
During this study, frogs were collected in a variety of
habitats from tropical rain forest to open, grassy areas to residential
districts. Although eggs and tadpoles were found in or near the environ
ments of the adults of their respective species, the larval environs
were more uniform in appearance. Most of the tadpoles were found in
standing water, but several species of tadpoles were found in flowing
streams. While the temperature and dissolved substances of these
habitats were found to vary considerably, one may satisfactorily divide
tadpole habitats into flowing stream and standing water types. Within
the latter type, special consideration will be given to the bromeliad
microhabitat of Jamaican tree frogs. Ecological data collected during
this study are presented in Tables 3~5-


80
phosphorus per hundred grams) would decrease iron absorption, due to
the poor absorption of iron complexed with phosphates (Cantarow and
Schepartz, 1957)- The animals were kept in glass-distilled water,
rather than spring water, in order to reduce the iron concentration
further.
Two matched groups of twelve tadpoles were placed in small
glass bowls which had been acid-cleaned, rinsed and filled with glass-
distilled water. During the course of the experiment, the water was
changed twice a week and fresh egg yolk was added immediately thereafter.
An excess of food was always available to both groups. The only differ
ence in treatment between the two groups was that ferrous sulfate
(reagent grade) solution was added to the water of the first group.
Individuals of the first group appeared to develop normally, as did the
four larger individuals of the second group. Of the eight smaller tad
poles of the second group, all became pale yellow in color and five were
dead after two weeks. At this time, ferrous sulfate was added to the
water of the second group. After three days, the second group of tad
poles became the same light brown color as those of the first group.
The darker pigmentation appeared to be the result of an increase in hemo
globin. These observations were taken to indicate that limitation of
iron intake of rapidly growing tadpoles resulted in an anemia, which was
relieved by addition of more iron. It is also possible that trace
elements such as copper or cobalt may have been responsible for hemoglobin
formation in both groups. In either case, the anemia did not cause develop
ment of green pigment in the skin of any of the tadpoles.


67
for development was not determined. Three adult males were collected
from this locality and all had green bones, soft tissues and plasma
(Table 1) .
Dowling (i960) found the tadpoles of this species in the Arima
River of Trinidad. She found the jet black tadpolesabout 2 to 5
centimeters in length--swimming in schools. These dense aggregations
of tadpoles, as many as 177 in a group, did not keep to sun or shade.
Neither did the individuals in masses appear to feed; tadpoles observed
feeding were scattered in shallow water.
Hyla vasta. -- Noble (1923) found this giant tree frog in the
northern coastal mountains of Hispaniola and also at 2500 meters elevation
in the central range which runs east to west. Mertens (1939) also
found it in forested areas in the central range, but did not consider
it exclusively montane, since he collected the species at 220 meters in
one locality. It has been found by the present writer and others near
Furcy in southeastern Haiti at 1350 meters elevation. In this unforested
and agricultural area the adults take cover in crevices and under roots
and sod which overhang the stream bank. I also heard a small chorus in
a ravine near the Riviere Froid at an elevation of 200 meters.
The life history of the giant Hispaniolan tree frog was first
recorded by Noble (1923). In the rain-soaked northern mountains of the
Dominican Republic, he found Hyla vasta along the streams which flow
through the rain forest. He stated: "After the sun has set, the giant
tree frog, Hyla vasta, leaves his hiding place among the tree tops and
descends to some rocky ravine. There flattened out on a mossy boulder


13
of intercalary cartilages in the several families of tree frogs as
an example of parallelism (Goin, 1961; Griffiths, 1963; Lynch and
Freeman, 1966). Goin and Goin (1962a, p. 231) stated, "The extra
joint thus provided allows the last phalanx, with its adhesive disc,
to be placed flat against the surface regardless of the position of
the foot--an obvious advantage to the climbing form."
No unusual conditions such as high rates of hemolysis were
noted by Barrio (1965a)- He suggested that the chlorotic conditions
which he witnessed were due to formation of early bile pigment.
Hemolytic conditions are not recorded for Amphibia, but might be
expected during metamorphosis when tadpole hemoglobin is replaced by
frog hemoglobin (McCutcheon, 1936). Varela and Sellares (1938) noted
a rapid decrease in the red cell count of Bufo arena rum at the end of
the breeding period, a change which indicates rapid hemolysis.
It appears that no satisfactory explanation of chlorosis has
been published. However, presence of biliverdin in the chlorotic species
studied by Barrio (1965a) implicates hemoproteins, particularly hemo
globin. Through hemoglobin one might expect involvement of various
structures and functions related to gaseous exchange.
Respiratory Studies
Embryonic and larval respiratory structures
Considerable research has been done on respiration as one of
the most important processes of living organisms. Much of it has been
directed toward micro-organisms, the mammalian species, and animals


32
The degree of pigmentation is no more helpful than its presence or
absence. Of the species collected in Surinam, Hyla punctata, Hyla
crepitans, Sphaenorhynchus aurantiacus and Phrynohyas venulosa had the
highest concentration of green pigment of the frogs collected, but they
certainly do not constitute a homogeneous group of species. Of the
eight species of Hyla that were studied in life in the West Indies,
seven had green pigmentation at metamorphosis or later, but Hyla
marianae lacks green pigmentation in these stages, being orange instead.
There is at least one relationship between chlorosis and frog
phylogeny. Thus far, all chlorotic frogs have been members of only four
families. Most of the green species are members of the family Hylidae,
while the others are included in the Centrolenidae, Pseudidae, and
Pvhacophor i dae. The chlorotic hylids, centrolenids and rhacophorids
are all tree frogs, while pseudids are aquatic for extensive periods.
All four of these families have in common an intercalary cartilage
between the ultimate and penultimate phalanges of the digits.
A close relationship between the four families is unlikely
even though they have the intercalary cartilage, green pigmentation and
tropical distribution in common. Hyperolius, the African genus of
rhacophorid frogs which has green eggs, has a firmisternal pectoral
girdle and diplasiocoelous vertebrae which distinguishes it from chlorotic
frogs of the other three families, which are procoelous, have an arci-
feral pectoral girdle and are Neotropical. There appears to be con
siderable agreement that the presence of intercalary cartilages in


142
Figure 19. Stained section of liver from adult male Pseud is paradoxus.
Note the large amount of pigment present. (200X)
Figure 20. Stained section of liver from adult male Hyla maxima.
Note degeneration of cytoplasm toward the lower right.
(400X)


CONCLUSIONS
1. The green coloration of bones, soft tissues, and plasma
of Neotropical anurans has been shown to be due to the presence of a
green pigment.
2. Identification of the pigment as biliverdin has been
confirmed.
3. The high concentrations of biliverdin have been found only
in frogs which inhabit the Neotropical Region. It appears that this
phenomenon is restricted to tropical tree frogs and to the aquatic frogs
of the family Pseudidae.
4. For the first time, chlorosis has been found in tadpoles
of several species.
5. Incidence of the pigmentation does not appear to be
restricted to either sex or to any particular stage of the development,
although it does not appear until metamorphosis in the tree frogs.
6. There are indications that the pigment concentration is
higher during metamorphosis and at the end of the breeding season, than
at other times.
7- While it appears that the green pigmentation is restricted
to three families of tree frogs and the Pseudidae, this is not considered
to be an indication of relationship between the Pseudidae and the tree
frogs.
104


45
Hvla wIder i. -- This small, green species is said to be quite
common in bromeliads about Mandeville, Jamaica, "and appears to have
a fairly wide range in the central part of the island above 1,000 feet"
(Lynn, 1940). Dr. Albert Laessle (personal communication) found this
species common on Juan de Bolas mountain, from which there is a large
series of specimens in the collection of the Museum of the Institute of
Jamaica. I was unable to collect a specimen there although 1 heard a
call which may have been of this species. The call of this species is
a faint clicking sound, which becomes louder as the frog continues to
call. However, in localities near Moneague and Fishbrook, I traced similar
calls to frogs which appeared to be of the genus Eleutherodacty1 us.
Unfortunately, both of these frogs escaped. During August, 1965, I
collected three adults from the Cockpit Country four miles north of
Quickstep, and a tadpole near Moneague. Another individual was taken
from Crown Lands in the Cockpit Country during May, 1966.
During the present study, adults of this species were found in
large as well as small bromeliads. My experience was similar to Dunn's
(1926) in that this species appeared most often in open woods, usually
with a southerly exposure. While Dunn (1926) collected one Hyla wiIderi
tadpole for each seven of Hyla brunnea, only one Hyla w? Ideri was collect
ed during the present study, along with more than 500 Hyla brunnea
tadpoles. The tadpole was collected from a large bromeliad in one of
several trees in a pasture. It was not darker than the Hyla brunnea
tadpoles collected at the same time, but did appear to have a greater
density of melanophores. The hind limbs on this specimen did not have


25
water from localities in South America, the West Indies, or Southern
Florida. Care was taken to keep the temperature of the fresh tap
water about the same as that which had been removed.
Upon returning to the laboratories in Gainesville, Florida,
the larvae were transferred to culture dishes of appropriate size.
Since Gainesville tap water was found to be lethal to both Hyla brunnea
and Hyla septentr1ona1is, spring water or Holtfreter's solution was
ordinarily substituted. During May, 1966, Holtfreter's solution proved
unsatisfactory and its use was discontinued. At one point, the spring
water appeared harmful and filtered pond water was used in its place.
By avoiding direct sunlight wherever possible and using other
precautions, tadpoles were maintained in the field with few losses
due to excessive temperature. At the University of Florida, the culture
dishes containing the tadpoles were moved from one laboratory to another
in an attempt to maintain a moderate temperature; however, groups of
tadpoles were exposed to water temperatures between known extremes of
16 C. and 28 C.
Thus, with reasonable care, most of the tadpoles were collected
and transferred to Gainesville with little difficulty. A notable except
ion was the Hyla heiIprin? tadpoles, which did not survive more than
three days after capture. Of the other species, most demonstrated an
ability to live for two weeks or more without being fed. However, in
the laboratory and whenever possible in the field, the tadpoles were
fed Gerber's strained foods at regular intervals. Hyla septentriona1is,
Hyla dom?nicensis, Hyla vasta, and Leptodacty1 us albilabrus were fed a


109
TABLE 1 (Continued)
FAMI LY
AND SPECIES
BONE
SOFT
PLASMA
BILE
SOURCE
TISSUES
Atelopodidae
Atelopus sp.
A
A
A
P
Hy 1 idae
Anotheca coronata
P
-
-
-
C. J. Goin
Hy la
a Ibofrenata
P
-
-
-
(Pers. Comm.)
Cochran, 1955
Hy la
albomarqinata
P
-
-
-
Cochran, 1955
Hy la
berthae
-
-
A
-
Barrio, 1965a
Hy la
brunnea
A
A
A
P
Hy la
boesemani (1)
P
P
-
P
Hy la
ca lea rata (1)
P
P
P
P
Hv la
crepitans(2)
P
P
P
P
Hy la
cuspidata
P
-
-
-
B. Lutz, 1954
Hy la
dominicensis
P or A
_
_
_
Lynn, 1958
Hy la
eq leri(2)
A
A
A
P
Hy la
qeoqraphica(1)
P
A
A
P
Hv la
hei1prini (1)
P
P
P
P
Hy la
land formi s (4)
A
A
A
P
Hyla
1 anqsdorffi
P
P
-
-
B. Lutz, 1954
Hv la
leucophy1 la ta (1)
A
A
A
P
Hy 1 a
1ichena ta (1)
P
Hv la
mar ianae (7)
A
A
A
-
Hy 1 a
maxima(3)
P
P
P
P
Hy la
mi sera (1)
A
A
P
P


Dr. James Bohlke, Philadelphia Academy of Natural Sciences; Dr. Doris
Cochran and Dr. James Peters, United States National Museum; Mr. Neil
Richmond and Dr. Clarence McCoy, Carnegie Museum; Dr. Charles F. Walker,
University of Michigan Museum of Zoology; Dr. Robert Inger, Chicago
Natural History Museum; Dr. Walter Auffenberg, Florida State Museum;
Mr. Bernard Lewis and Dr. Thomas Farr, Institute of Jamaica.
1 am grateful to Dr. Thomas Goreau, University of the West Indies,
Dr. Margaret Stewart, State University of New York at Albany, and others
for enlightening discussions. Bonneye Greene, Thomas Quarles, Robert
MacFarlane and John Anderson kindly rendered technical assistance.
Forest service personnel in Jamaica and Surinam made it possible for
me to visit several remote localities. I appreciate the aid rendered
by members of the U. S. Consular Service. Grants for travel expenses
were received from the University of Florida Graduate School, the Sigma
Xi- RESA Fund and the National Science Foundation (GB-3644, to Dr.
Coleman J. Goin).
The services of Rhea Warren, Stephen Bass, Murray de la Fuente
and others who aided in the collection of specimens are appreciated
greatly. Janet Bungay and Geraldine Lennon aided in the preparation of
the manuscript. My wife, Dorothy, assisted in the histological portion
of the study, in preparation of the manuscript, and in giving moral
support.
i i i


153
Stokstad, E. L. R. and J. Koch. 1967- Folic acid metabolism. Physio
logical Reviews 47.: 83-116.
Strawinski, S. 1956. Vascularization of respiratory surfaces in
ontogeny of the edible frog, Rana esculenta L. Zoolgica
Poloniae 327365-
Stuart, L. C. 1951. The distributional implications of temperature
tolerances and hemoglobin values in the toads Bufo mar inus
(Linnaeus) and Bufo bocourti Brocchi. Copeia 1951: 220-229.
Taylor, E. H. 1942. Tadpoles of Mexican Anura. University of Kansas
Science Bulletin 28.: 3 7_55
Umbreit, W. W., R.H. Burrissand J. F. Stauffer. 1964. Manometric
Techniques. Minneapolis: Burgess Publishing Company.
Varela, M. E. and Sellares, M. E. 1938. Variations annuel les du sang
du crepaud Bufo arenarum Hensel. Comptes Rendus Societe de
Biologie J2£: 1248-1249-
Vernberg, F. J. 1955. Hematological studies on salamanders in relation
to their ecology. Herpetologica _M_: 129-133-
Wagenaar, M. 1939- Inquiry into the identity of the green colour from
the spine of the sea pike. Archives neerlandaises de Zoologie
4: 103-105.
Williams, A. W. 1965- Causes and laboratory diagnosis of jaundice.
New Zealand Medical Journal 64: 486-491.
Willstaedt, H. 1941. Zur Kenntnis der grunen Farbstoffe von Seefischen.
Enzimo logia S_: 260.
Wright, A. H. and A. A. Wright. 1949- Handbook of Frogs and Toads of the
United States and Canada, third edition. Ithaca, New York:
Comstock Publishing Company, Inc.
Zweifel, R. G. 1964. Life history of Phrynohyas venulosa (Salientia:
Hylidae) in Panama. Copeia 1964: 201-208.


134
Figure 5. Absorption spectra of bile solutions of Siren lacertina.
200 300 400 500 600
WAVE LENGTH IN MILLIMICRONS
Figure 6. Absorption spectrum of pigment extracted from Hyla
marianae in 70 per cent ethanol.


91
Hvla leucophv1 lata adult female liver apparently
normal; contains many erythrocytes.
Hvla maxima -- adult male fat deposition taking place in
one area; erythrocytes appear to be undergoing disintegration
in liver; relatively little pigment present.
Hyla minuta -- adult male some irregularly shaped nuclei;
considerable pigment present.
Hyla punctata adult male many nuclei present in liver
tissue; these are quite variable in size, shape and stain
ing characteristics; some may be pyknotic; very little
pigment is present in the liver; many unusual cells are
present in blood, perhaps immature erythrocytes.
Osteocepha1 us taurinus -- adult female liver pigmentation
less than normal; odd-shaped nuclei present in liver.
Disease as a Cause of Chlorosis in Frogs
Early in this study, it was assumed that disease was not prin
cipally responsible for the green pigmentation of frogs. Since several
writers had accepted green bones or other green tissues as traits which
were characteristic of certain species, it was not logical to believe that
a bacterial or viral disease would affect all the members of a species or
population in a nearly uniform manner. Neither was it logical that a
genetic disease should be passed from generation to generation with such
regularity. Thus, it was tentatively assumed that the normal metabolism
of chlorotic frogs was responsible for their green pigmentation.
On the other hand, disease cannot be ruled out as a possible
cause of biliverdin accumulation. Individual adult specimens of Hyla
septentriona1? s and Hyla dom? n ? censls may have dark green, light green
or white bones. Here is an example of what one might expect if some of
the frogs were diseased, but such individual variation may also have other
causes.


15
However, there are considerable differences in size of amphibian eggs.
Panton (1952) noted that aquatic eggs of Hyla brunnea were only one-half
millimeter in diameter, whereas those of another frog (probably
Eleutherodacty1 us jamaicensis) that did not develop in water, were
several millimeters in diameter.
The sites of egg deposition and tadpole development appear
related to respiratory processes. Dickerson (1906) believed that the
jelly membranes of amphibian eggs retain heat longer than the surround
ing water. Experiments by Savage (1950), using the eggs of Rana
temporaria, showed that the mean temperature difference between eggs
with jelly envelopes and plain water was 0.63 C. Moore (19^0) suggested
that the compact jelly masses (up to 10 centimeters in diameter) of
Rana syIva tica slow the diffusion of oxygen to the embryos, and that
this becomes critical as the metabolic rate increases in response to a
temperature of about 25 C. However, Savage (1950) pointed out that
water moves freely between individual jelly capsules of Rana temporaria,
thus requiring diffusion of oxygen through only a few millimeters rather
than several centimeters of jelly. Algae associated with jelly membranes
of Rana svIva tica (Dickerson, 1906), Rana aurora (Wright and Wright, 19^9)
and Rana temporaria (Savage, 1961) may affect the respiration of these
frogs' eggs.
Goin and Goin (1962^) noted that amphibian eggs may be laid
singly or in clusters and may be attached to submerged objects, floating
or settled in water, carried by parents, or be laid on land or vegetation.


TABLE 3 (Continued)
LOCATION AND HABITAT DATA
9. 8 km W of Port-au-Prince (flooded
wi regrass pasture)
10. 32 km W of Port-au-Prince (wheel ruts)
11. 4.8 km W of Port-au-Prince
(shaded pool in stream bed)
t
Ponds and standing waters of Grand Cayman,
B.W.I. habitats of Hyla septentriona1is
12. 3-2 km N of Georgetown (pond)
13- Hell (roadside ditch)
14. South Sound (two mangrove swamps)
15- South Sound (four woodland ponds)
16. 4.8 km E of Georgetown (cow well)
17- 8 km E of Georgetown (cistern)
18. Boddentown to East End (three ponds)
19- North Side (woodland pond)
20. North Side (large open pond)
DATE
ELEVATION
(Meters)
AIR
TEMPERATURE
12 V 1966
1
30.2
4 VI11 1966
10
33-2-35.2
10 V 1966
200
26.8
12 V 1966
200
27-2
29
X
1965
4
31.7
29
X
1965
4
30.5
29
X
1965
2
29-5-30.5
29
X
1965
2
30.5-32.2
29
X
1965
2
29.5
29
X
1965
2
28.3
29
X
1965
3
27.2-28.9
30
X
1965
3
27.2
30
X
1965
3
27.7


129
TABLE 9. TADPOLE RESPIRATORY RATES AT 25 C. (MEASURED BY JONES METHOD)
SPECIES WEIGHT(mg) OXYGEN UPTAKE (ul Oy/g/hr)
Hyla dominicens!s
315
262
Hyla dominicensis
481
241
Hyla dominicensis
398
187
Hyla dominicensis
248
216
Hyla dominicensis
226
206
Hyla dominicensis
597
252
Hyla hei1prini
1,203
255
Hyla heilprini
1,444
265
Hyla hei1prini
1,420
277
Hyla vasta
476
203
Hyla vasta
O
CTn
248
Hyla vasta
370
329
Hyla vasta
83
663
Hyla vasta
84
669
Hyla vasta
170
560
Hyla maxima
69
739
Hyla maxima
70
714
Hyla maxima
78
756
Hyla maxima
84
678
Pseud is paradoxus
3,696
53
Pseud is paradoxus
5,948
31
Pseud is paradoxus
4,577
44


GREEN PIGMENTATION
IN NEOTROPICAL FROGS
By
DUVALL ALBERT JONES
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
December, 1967


DISCUSSION
Earlier in this paper, evidence was presented to show that
chlorosis is the result of the accumulation of biliverdin in the tissues.
Therefore, this condition corresponds to human jaundice, which is due to
accumulation of bilirubin. Presumably, the obscure causes of chlorosis
would parallel those of jaundice, which were recently outlined by
Williams (1965).
There appear to be four basic causes of jaundice. Abnormally
high rates of hemolysis cause an increase of bile pigment which may
exceed the liver's capacity of excretion. Defective absorption or con
jugation of bile pigment by the hepatic cells, or obstruction of the
various biliary ducts can cause jaundice. These basic causes of bile
pigment accumulation will be considered in relation to chlorosis of frogs.
High Rate of Red Cell Hemolysis
High rates of hemolysis and bile pigment formation might be
expected in frogs which have a high rate of red cell replacement because
of high erythrocyte counts or short red cell life spans. From Table 6,
one notes that there are chlorotic species with high (Phrynohyas venulosa).
low (Hy la wiIderi), and intermediate red cell counts. Species which lack
the green pigmentation demonstrate a similar range of red cell counts.
Little information regarding frog erythrocyte survival times is available,
but my unpublished data on this subject suggest that there is no significant
93


39
It should be noted that the respiratory rate of Hyla brunnea
tends to drop sharply between 35 and 40 C. (Table 8, Group A), whereas
the respiratory rate of Hyla septentriona1?s does not (Table 8, Group C).
When the Hyla septentriona1is tadpoles were removed from the flasks, they
appeared to be in good condition (heart rate was about three beats per
second). On the other hand, these Hyla brunnea tadpoles were limp after
two hours at 40 C., and one did not recover. This is an indication that
the critical thermal maximum of Hyla brunnea larvae is lower than that
of the Hyla septentriona1is tadpoles. This might be expected when one
considers the higher temperatures to which Hyla septentriona 1 is tadpoles
are often exposed (Table 4). These higher temperatures are largely due
to the lower elevation and greater insolation of Hyla septentr?ona1?s
habitats. Hyla brunnea is more often found in shaded areas and at
greater elevations than are available to Hyla septentriona1is over most
of its range.
Along these lines, it may. be noted that the frog let of Hyla
lichenata appeared in poor condition after being kept at 35 C. for less
than two hours (Table 10). It would appear that Hyla lichenata is more
sensitive to this temperature than is Hyla brunnea at the same stage.
This may account for the fact that while Hyla brunnea is found throughout
the range of Hyla lichenata, the former species extends its range to lower
elevations and more exposed habitats than the latter.
Tadpoles of Hyla brunnea kept in the laboratory were exposed
to low water temperatures during the early morning of November 30, 1965.
This caused some mortality and the surviving tadpoles appeared pale,


18
view of Strawinski's work, more probably, the assoc
iated large operculum). in ordinary ponds, intermediate
as habitats, the arrangements might be expected to be
intermediate also.
There is a great deal of conjecture in all this, but
the microhylids seem to provide examples. Some have such
large gill filaments that they trail in the opercular
cavity in a way quite unlike those of Rana, others have no
gills but have enormous gill fi Iters, for example Glypho-
glossus molossus. Some, such as Hypopachus aquae, are
intermediate and live in ordinary ponds. Rana temporaria
also have a moderate development of gills, filters and lungs,
and lives in ordinary ponds.(P- 56)
Among amphibians, we may turn to the work of Strawinski (1956)
for an estimation of the relative importance of tadpole respiratory
surfaces. In studying the density of capillary networks of Rana
esculenta, he reported that most gaseous exchange of early stages
took place through the skin, with the internal surface of the operculum
also important. It was his belief that the external and internal gills,
and vascularized filtering apparatus were of little importance. As
the lungs developed, their portion of the total respiratory capillaries
increased rapidly, reaching 65 per cent at metamorphosis, the same
proportion as in the adult. Vascularization of the skin also increased
until metamorphosis, when these capillaries accounted for 3^ per cent
of the tota 1.
Blood transport of oxygen
Among vertebrates, blood contained within a closed circulatory
system is largely responsible for the transport of gases in the body.
Hemoglobin pigments, located in red blood cells and responsible for
their color, are the main carriers of oxygen in vertebrates. Exceptions


147
Godlewski, E. 1900. Ueber die Einwirkung des Sauerstoffs auf
Entwicklung und uber den Gaswechsel in den ersten Entwicklung-
stadien von Rana temporaria. Bulletin Internationale de .1 '
Academie des Sciences, Cracovie. pp. 232-255-
Goin, C. J. 1957- Status of the frog genus Sphoenohyla with a synopsis
of the species. Caldasia 8: 11-31.
. 1961. Synopsis of the genera of hylid frogs. Annals of
Carnegie Museum 6: 5-18.
Goin, C. J. and B. W. Cooper. 1950. Notes on a collection of amphibians
from Jamaica. Occasional Papers of the Museum of the Institute
of Jamaica No. 4: 1-9.
Goin, C. J. and 0. B. Goin. 1962a. Introduction to Herpetology. San
Francisco: W. H. Freeman and Company.
and 1962'-. Amphibian eggs and the amphibian environment.
Evolution 16: 364-371-
Goin, C. J. and C. G. Jackson. 1965- Hemoglobin values of some amphibians
and reptiles from Florida. Herpetologica 2J_: 145-146.
Goin, C. J. and J. N. Layne. 1958. Notes on a collection of frogs from
Leticia, Colombia. Publication of Research Division Ross Allen
Reptile Institute J_: 97-104.
Goldberg, C. A. J. 1958. The ferrohemoglobin solubility test. Clinical
Chemistry 4: 146-149.
Gosse, P..H. 1851. A Naturalist's Sojourn in Jamaica. London.
de Graaf, A. R. 1957- A note on the oxygen requirements of Xenopus
laevis. Journal of Experimental Biology 4: 173-176.
Grant, C. 1940. The Herpetology of the Cayman Islands. Bulletin of
the Institute of Jamaica, Science Series. (2): 1-56.
Griffiths, I. 1963. The phylogeny of the Salientia. Biological Reviews
8: 241-292.
Grigg-:, G. C. 1965. Studies on the Queensland lungfish, Neoceratodus
forsteri (Krefft). III. Aerial respiration in relation to
habits. Australian Journal of Zoology J_3_: 413-42!.
Grodins, F. S., A. L. Berman and A. C. Ivy. 1941. Observations on the
toxicities and choleretic activities of certain bile salts.
Journal of Laboratory and Clinical Medicine 27: 181 -186.


66
Thus, water temperatures in the basin ranged from 23.8 C. to 29-9 C.,
while the temperature ofthe water in a pool at the base of the falls
ranged from 23.4 C. to 25-5 C. Iron concentration in the basin was
about three times as great as in the pool (Table 5)- This may be import
ant to young tadpoles, since it was noted that young Hyla brunnea tadpoles
are more sensitive to iron deficiency than are the older ones. Concen
trations of phosphate and pH were similar for both the basin and the
stream pool. Oxygen concentration in the pool ranged from 8.3 to 8.9
parts per million; during the same period, the basin contained 2.0 to
3.1 parts per million of oxygen. Concentration of carbon dioxide in
the basin shortly after it had been in the sun was 24.0 parts per million;
at the same time, the concentration of carbon dioxide in the pool was
10.0 parts per million.
The eggs in the basin were of medium size--2 to 3 millimeters
in diameter, not including the jelly--and rested on the detritus at
the bottom of the basin. It was not known whether the eggs originally
floated on the water surface. The yolk of the eggs was white rather
than yellow and the animal hemisphere was darkly pigmented. Tadpoles
in the basin were observed feeding on the eggs, which filled their ali
mentary tracts. They also came to the surface to take air, but do not
appear to have lungs. When the water temperature was highest, these
tadpoles were observed coming to the surface at intervals of 10 to 25
seconds.
None of the tadpoles observed showed green pigmentation. How
ever, none had we 11-developed hind legs. The length of time required


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EB0NA7PNX_6ZNO5T INGEST_TIME 2015-04-01T19:39:19Z PACKAGE AA00029845_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


71
one wonders if this was the result of extreme effects of temperature
or oxygen upon a pigment which was more generally distributed through
the body. An adult of this species collected from a banana stalk was
bright grass-green on the dorsal surface with white and blue ventrally
and blue and yellow on the flanks. Later the back became quite dark.
This specimen had green bones and green plasma. Noble (1923) described
the dorsal color of a living specimen as "golden green."
Hyla pulehr? 1 ?neata. -- Except for its light-colored longi
tudinal stripes, lack of an interocular bar, and larger size, this small
green tree frog of Hispaniola is quite like Hyla wilderi of Jamaica.
It has been collected from widely scattered points on the island and
Cochran (1941) noted that it seems to be common in certain localities.
From the records of Noble (1923), Cochran (1941), and Lynn (1958), it
seems likely that this species has a distribution similar to those of
Hyla vasta and Hyla he?Iprini: in forests at higher altitudes and in
wooded ravines at lower altitudes. This species appears to be the least
common of the four hylids of Hispaniola; I did not encounter adults or
tadpoles during my visits to Haiti. Noble (1923) and Lynn (1958) found
adults calling from leaves in the vicinity of streams. Lynn (1958)
described its call as a faint clicking like a small telegraph instrument,
similar to that of Hyla wiIderi.
Noble (1927) stated that Hyla pulehr i 1?neata young develop in
water as do the other hylas on the island, but apparently he has given no
further details. There are no known specimens of Hyla pulehr i 1 ineata
tadpoles in the major museums of eastern North America (including the


16
There are few indications of respiratory rates of embryos being corre
lated with these types or sites of egg deposition. Panton's observation
(1952) that a set of large eggs, probably those of E1 eutherodacty1 us
i amaicens ?s, fai led to develop when placed in water may indicate that
embryos which develop terrestrially receive insufficient oxygen when
subjected to the reduced oxygen tension of water. Noble (1931) suggested
that large eggs undergoing rapid development require more oxygen than
they can obtain in water when surrounded by the egg capsule.
As the embryos develop and hatch, other respiratory structures
are formed and behavioral patterns related to respiration arise. Organs
of external respiration in tadpoles are the skin, the internal surface
of the operculum, external and internal gills, the vascularized surface
of the food filtering apparatus, and the lungs.
The most evident respiratory structures to appear in about
the time of hatching are the external gills. These may be small as
in Hyla vasta or enlarged as in Hyla rosenbergi tadpoles, although both
species dwell in tropical streams (Noble, 1931). Dunn (1926) noted
the reduction of gill size in Jamaican hylids, which breed in water
which collects in bromeliads. Likewise, tadpoles of the genus Hoplophryne.
which live in water that collects in banana leaves, also have reduced
gills (Noble, 1929). Terrestrial embryos of the genus Eleutherodacty1 us
may or may not have external gills, their highly vascular tails serving
as important respiratory organs (Lynn, 1961). Among the more unusual
are the gills of some Gastrotheca which are large and bell-shaped (Noble,
1931) .


46
green pigment. Tadpoles of Kyla wi1deri have been collected in March,
Apr!1 and August (Lynn, 1940).
The tadpole of Hyla wilderi has poorly developed gills,
similar to those of Hy1 a brunnea, and like other Jamaican tadpoles, its
body is depressed, and has no fin.
Dunn (1926) found that the specimens which Barbour (1910) con
sidered pale green young of Hyla brunnea were actually Hy1 a w1 lderi.
In fact, Dunn believed that the non-ossified head and green bones of
adult Hyla wilderi represented a neotenic condition because of its resemb
lance to the young of Hyla brunnea and related forms. He suggested that
Hyla wilderi, as well as Hyla lichenata and Hyla marianae arose from a
frog of the Hyla brunnea type through sympatric speciation, because of
differential growth rates.
Hyla marianae. -- Of the Jamaican hylids, least is known of
Hyla marianae. Dunn (1926) originally considered these yellowish-green
or greenish-brown frogs to be the young of Hyla brunnea, but then recognized
them to be of a different species, which he described. The range of this
species appears to be the most restricted of the Jamaican hylids. Most
of the specimens have been collected in or near the Cockpit Country of
west-central Jamaica; two specimens came from Hollymount, Mt. Diabolo,
further to the east (Lynn, 1940; Goin and Cooper, 1950). This area has
a combination of high elevation (above 400 meters), greater rainfall and
less disturbance of habitat than surrounding areas. Limestone cliffs
covered by lianas and hillsides covered with jagged, honeycomb limestone
usually form the steep sides of the "cockpits." Vegetation on these


100
Defective Bile Pigment Conjugation
Although it is possible for small amounts of anuran biliverdin
to be enzymatically conjugated as a glucuronide (Noir, Rodriguez Garay
and Royer, 1965), Barrio (1965a) found that the bili verdn of frogs
reacts as if unconjugated. Thus, a deficiency or inhibition of this
conjugating enzyme system probably would not be important in the green
pigmented frogs.
Disturbed Bile Pigment Excretion
Excretion of bile pigment may be blocked either within the
liver or in the ducts leading away from the liver.
Intrahepatic cholestasis may be a constitutional disease which
is often familial. Dubin-Johnson syndrome (Dubin, 1958) is a chronic,
intermittent, benign form of jaundice in which the liver is often greenish
black and microscopically shows much pigment which is probably lipofuscin.
In these characteristics it resembles the situation seen in Pseud? s
paradoxus, which has a black liver in the adult.
Drugs can cause an intrahepatic blockage of bile due to their
effects upon hepatic cells. Important among these are certain steroid
hormones and chlorpromazine, as well as others (Sherlock, 1964). It is
possible that frog liver is affected by hormones since Bachmann, Goin and
Goin (1966) have shown that there is an increase in polyploidy of frog
liver cells with the onset of the breeding season. Steroid hormones are
known to increase polyploidy in other forms. Thus, it seems possible that
tropical frogs may have an extended breeding season which affects the
liver adversely. Evidence of this was seen in the liver of Hyla maxima


LITERATURE CITED
Agner, K. 1941. Verdoperoxidase. A ferment Isolated from leucocytes.
Acta Physiologica Scandinavica 2_:Supplement 8.
Allen, J. R., L. A. Carstens, and B. E. Olson. 1967. Veno-occlus ive
disease in Macaca sped osa monkeys. American Journal of
Pathology 0: 653 667-
Andrew, W. 1965- Comparative Hematology. New York: Grue and Stratton.
Arhelger, R. B., J. S. Broom, and R. K. Boler. 1965- U1trastructura1
hepatic alterations following tannic acid administration to
rabbits. American Journal of Pathology 46: 409434.
Atlas, M. 1938. The rate of oxygen consumption of frogs during
embryonic development and growth. Physiological Zoology.11:
278-291
Babak, E. 1907a. Uber die funktionelle Anpassung der ausseren Kiemen
bei Sauerstoffmange 1. Zentralblatt fur Physiologie 2_1_: 97-
Babak, E. 1907^- Untersuchungen uber die Warmelahmung und die Wirkung
des Sauerstoffmange1s bei Rana fusca und esculenta. Zentralblatt
fur Physiologie 2 1: 6.
Bachmann, K., 0. B. Goin and C. J. Goin. 1966. Hy 1 id frogs: polyploid
classes of DNA in liver nuclei. Science 154: 650-651.
Barbour, T. 1910. Notes on the herpetology of Jamaica. Bulletin of
Museum of Comparative Zoology 2_: 273-301.
. 1914. A contribution to the zoogeography of the West Indies.
Memoirs Museum of Comparative Zoology 44: 209-259-
- 1931* Another introduced frog In North America. Copeia 1931: 140.
. 1937. Third List of Antillean Amphibians and Reptiles. Bulletin
of Museum of Comparative Zoology 82_: 77-166.
Barbour, T. and C. T. Ransden. 1919- The herpetology of Cuba. Memoirs
Museum of Comparative Zoology 471-213.
Barrio, A. 1965a. Cloricia fisiolgica en batracios anuros. Physis 25:
137-142.
143


TABLE 1 (Continued)
FAMILY AND SPECIES
BONE
SOFT
PLASMA
B! LE
SOURCE
TISSUES
Hyla nasica
-
-
P
-
Barrio,
1965a
Hyla pulchella andina
-
-
P
-
Barri0,
1965a
Hyla pulchella riojana
P
Barrio,
1965a
Hyla pulchella cordobae
.
_
P
Barri0,
1965a
Hyla pulchella pulchella
_
A
Barrio,
1965a
Hyla pulchella prasina
.
.
A
Barrio,
1965a
Hyla phrynoderma
-
-
A
-
Barrio,
1965a
Hyla punctata(2)
P
P
P
P
Hyla punctata
P
-
P
-
Barrio,
1965a
Hyla raniceps
-
-
A
-
Barrio,
1965a
Hyla rubra(4)
A
A
A
P
Hyla septentriona1 is P
or A
A
P or A
P
Hyla siemersi
-
-
P
-
Barrio,
1965a
Hyla squa1irostris
-
-
A
-
Barri0,
1965a
Hyla trachytorax
-
-
A
-
Barrio,
1965a
Hyla vasta P
or A
-
-
P
Hyla wavrin!
P
-
-
-
Rivero,
1961
Hyla wilderi(4)
P
P
-
-
Osteocepha1 us taurinus(2)
P
P or A
P or A
P
Phyllomedusa bicolor(l)
A
A
A
Phyllomedusa helenae
-
green eggs
-
-
Starrett, I960
Phy1lomedusa
hypochondria 1 is (5)
A
A
A
P


40
The jelly mass from which the tadpoles hatch remains in the
bromeliad reservoir longer than do the tadpoles themselves. Panton
(1952) has suggested that the presence of the jelly reduces evaporation
and moderates temperature changes. The low pH of the water probably
prevents the rapid decay of the jelly mass or infertile eggs. In his
micro-limnological study of Jamaican bromeliads, Laessle (1961) found
the following ranges of readings in five bromeliads which contained
eggs or tadpoles of Hyla brunnea: dissolved oxygen, 0.03 2.3 parts
per million; dissolved carbon dioxide, 23.0 41.0 ppm; pH, 4.0 4.5;
temperature, 23.0 25.0 C. Additional measurements made during the
present study increased these ranges to: dissolved oxygen, 0.03 2.7 ppm
pH, 4.0 6.5; temperature, 23.0 28.0 C.
During the present study, the contents of a number of bromeliad
reservoirs were poured into waterproof containers. When jelly masses
without eggs were present, they were not firm and had a tendency to
separate into capsules 1-2 centimeters in diameter. These probably repre
sent capsules which contained four to six eggs each, as mentioned by
Dunn (1926). Within the bromeliad, the mixture of water and jelly has
the consistency of glycerine, as mentioned by previous writers. While
I have noted tadpoles, especially small ones, moving about near the
surface on several occasions, I have also noted larger ones moving verti
cally within the reservoir. The larger individuals tend to remain under
leaf fragments when the reservoir is exposed to the sun. They come to
the surface at intervals to take air and then return to lower depths.
This respiratory behavior apparently was not observed previously, since


132
Opt ica1
Dens ity
Figure 1.
Optica1
Density
Figure 2.
Absorption spectrum of blllverdln In 5 per cent hydro
chloric acid methanol solution.
Absorption spectrum of liver extract from Hvla
septentriona11s in aqueous solution.


150
Manwell, C. 1966. Metamorphosis and gene action I. Electrophoresis
of dehydrogenases, esterases, phosphatases, hemoglobins and
other soluble proteins of tadpole and adult bullfrogs. Compar
ative Biochemistry and Physiology 17: 805-823.
McCutcheon, F. H. 1936. Hemoglobin function during the life history of
the bullfrog. Journal of Cellular and Comparative Physiology
8: 63-81.
Me 1 ichar, V. K. Polacek arid M. Novak. 1962. The relationship between
bilirubin concentration and the level of non-esterified fatty
acids in the blood of new-born infants. Biologia Neonatorum
4: 94-101.
Mertens, R. 1939- Herpetologische Ergebnisse einer Reise nach der Insel
Hispaniola, Westindien. Abhandlungen senckenburg natur
Gesellschaft 449: 1-84.
Miranda-Ribeiro, A. 1926. Notas para servirem ao estudo dos gymnobat-
raquios (Anura) brasileiros. Archivios Museo Nacional do
Rio de Janeiro 2_7: 7- (From Barrio, 1965a)
Moore, J. A. 1940. Adaptive differences in the egg membranes of frogs.
American Natura 1ist 4: 89-93-
Nakajima, H., T. Takemura, 0. Nakajima, and K. Yamaoka. 1963- Studies
on heme aIpha-metheny1 oxygenase I. The enzymatic conversion
of pyridine-hemichromogen and hemoglobin-haptoglobin into a
possible precursor of biliverdin. Journal of Biological
Chemistry 28: 3784-3796.
Negus, V. 1965- The Biology of Respiration. Baltimore: Williams and
Wi1kins Company.
Neill, W. T. 1958. The occurrence of amphibians and reptiles in salt
water areas, and a bibliography. Bulletin of Marine Science of
the Gulf and Caribbean 8_: 1 -97-
Nisimaru, Y. 1931. Bile pigment formation in the liver from hemoglobin.
American Journal of Physiology 97: 654-657-
Noble, G. K. 1923. In pursuit of the giant tree frog. Journal of the
American Museum of Natural History 2_: 104-116.
. 1927. The value of life history data in the study of evolution
of the Amphibia. Annals of the New York Academy of Science 30:
31-128.
. 1929- The adaptive modifications of the arboreal tadpoles
Hoplophryne and the torrent tadpoles of Staurois. Bulletin
of the American Museum of Natural History 8: 291-334.


hypochondria 1is was found on leaves of bushes and small trees at the
edge of wooded areas. The large, yellow eggs of Phy1lomedusa hypo
chondria 1 is are probably attached to leaves above standing water, into
which the larvae drop after hatching. Movements of these pond-type
tadpoles are rather slow.
Sphaenorhynchus aurantiacus. -- This small green hylid was
collected from two similar localities in Surinam. In both habitats,
the frogs called from cattails in standing water, 50 centimeters or more
in depth. In a locality near Domberg, the frogs were found at an inter
section of two drainage canals in which there was considerable aquatic
vegetation. Along these canals were some shrubs and small trees. The
second locality, south of Paramaribo, was a small cattail marsh, perhaps
30 meters in diameter and surrounded by forest except where it emptied
into a roadside canal. Goin (1957) noted that all of the known species
of this genus select areas with still waters as their breeding habitats,
and that they call from the water, or from floating or emergent vegetation.
Lutz (195^) stated that members of this genus lay eggs on leaves, but their
development seems to be unknown.
Goin (1957) specifically noted that certain members of this
genus have green bones (Table 1). From the description of color in
other structures, it appears that green pigmentation of tissues is a
general and striking characteristic of the genus Sphaenorhynchus.
Hy la dominicensis. Of the hylid species on Hispaniola, Hyla
dominicensis is certainly the most common and has the widest distribution.
It is very similar to Hyla brunnea, the most common and widespread of


47
hillsides varies from sparse, scrubby growth, to saplings which shade
an understory of terrestrial bromeliads, to large trees. The areas of
sparse vegetation usually are higher on the hills and have a more
southerly exposure than do the areas of denser vegetation. Dunn took
his frogs from "wild pines" in rather thick woods. 1 collected one male,
two females and four metamorphosing tadpoles of this species from small,
compact bromeliads along the border between a sunny clearing and a thick
growth of saplings on the slope above. Hyla wiIder? and Hyla brunnea
adults were collected from the same site.
Three of the four Hyla marianae tadpoles were in the same small
bromeliad, but between different sets of leaves. All of them were
engorged with eggs. The tadpole mouthparts had been lost and the mouths
of these tadpoles were quite wide. Although the adults of this species
(28-38 mm) were slightly more than half the length of Hyla brunnea
(56-57 mm) adults, the tadpoles of Hyla marianae appeared to be about
the same size as Hyla lichenata tadpoles at metamorphosis. Like the
tadpoles of Hyla brunnea, those of Hyla marianae were quite active and
tended to move upward at metamorphosis. The color of these tadpoles was
brown. The adults which 1 collected could change from orange to brown
and back. The bones of adults and tadpoles were orange with no trace
of green in the soft tissues.
Pond habitats
During the course of this study, a relatively large number of
frog species waie found breeding in bodies of standing water. In addition
to the bromeliad habitat already considered, such environments included:


73
distributed throughout the Neotropical Region, but appear to be restricted
to certain habitats or families of frogs. As Barrio (1965a) noted,
green pigmentation is found in only three families of South American
frogs. Hylids and centrolenids are primarily arboreal following
metamorphosis, the period of life when they appear most likely to have
green pigmentation. Pseudids are generally considered to be highly
aquatic as adults. Although 1 found relatively little green pigment
in some adult Pseudis paradoxus paradoxus in Surinam, Barrio (1965a)
found high concentrations of biliverdin in Pseudis paradoxus platens is,
Lysapsus lime 11 us and Lysapsus mantidacty1 us adu1ts. While green
pigmentation appears in tree frog tadpoles with metamorphosis, it
preceded metamorphosis by an extensive period in Pseudis paradoxus
paradoxus. Thus, the pseudids differ from hylids and centrolenids in
both the type of habitat and in the time of onset of pigment development.
It seems logical to assume that the hyperbi1 iverdinemia of pseudids has
a cause which is different from that of the two tree frog families. In
addition, the point should be made that terrestrial and fossorial anurans
toads largely--have not been reported to have biliverdin in their tissues.
Since distributional and ecological factors have not yielded
a satisfactory explanation for the accumulation of green pigment, it
appears logical to turn to physiological studies.


48
ponds, pools, puddles and wheel ruts filled with water; roadside ditches
and drainage canals; flooded, grassy meadows; and cattail marshes.
These waters varied widely in size, amount of exposure to sun and wind,
and in water characteristics (Tables 4 and 5)- Temperatures at mid-day
ranged to 40 C. in shallow open waters, but generally were less than
35 C. when exposed to direct sunlight for only a few hours; where direct
sunlight did not reach the pools, temperatures remained under 30 C.
Oxygen concentrations varied from less than one part per million to ten
parts per million and appeared to be affected by air currents and
photosynthesis as well as temperature.
The pH of standing surface waters in the West Indies appeared
to average somewhat higher than the pH of water in bromeliads in the
region. This may be due to the fact that the surface strata of most of
the West Indies consist of limestone. At a locality near Juan de Bolas
in Jamaica (Table 4, Habitat 7), the pH of water in three bromeliads
without an obvious layer of dust on the leaves ranged from 5-5 to 6.5;
three bromeliads near the road had a heavy covering of dust on the
leaves and the pH of the water in the reservoirs was 7- Presumably the
dust from the limestone gravel of the road reduced the acidity of water
in bromeliads when it was washed into the central reservoirs by rain
or dew. Other water characteristics do not show definite patterns
(Tables 4 and 5)
It is difficult to make a distinction between exposed and
shaded pool habitats. There may be greater difficulty in assigning a
particular species to one or the other. Hyla dominicensis and Hyla


42
simplified. Finally, the mouth has only a single row of teeth about
it (Dunn, 1926), since these are not needed for scraping food from
surfaces.
The length of time required for development from fertilization
to metamorphosis is not known since neither Lynn (1940) nor I was able to
follow a single group through the entire period. From Lynn's work, nearly
a week passes between fertilization and hatching. After hatching, at
least a month probably passes before metamorphosis is completed. While
tadpoles have been kept in captivity for more than six weeks, it is likely
that metamorphosis could be completed approximately six weeks after ferti
lization under optimal conditions.
At the time that the forelegs penetrate the opercular fold,
the green pigmentation was always quite obvious. Jarring of the bromeliad
or container in which such tadpoles were present caused them to climb up
ward very rapidly. They have no difficulty in climbing out of a bromeliad
reservoir, a water glass, or a plastic bag. Once they leave the water,
the tail is resorbed in twenty-four to thirty-six hours. This is in
agreement with the finding of Schreckenberg (1956) that there is intense
thyroid secretory activity and sudden release of colloid from the thyroid
gland at the stage of tail resorption. The dark green pigmentation of
soft tissues in the guiar region remains for two weeks or more after
metamorphosis and then gradually fades away in the living animal. Green
pigmentation of bones remained as long as the young frogs lived, or about
six weeks after metamorphosis, in the laboratory. In nature, no immature


128
TABLE 8. RESPIRATORY RATES OF TADPOLES IN WARBURG RESPIROMETERS
AT DIFFERENT TEMPERATURES
SPECIES AND GROUP WEIGHT(mg) OXYGEN UPTAKE (ul 02/g/hr)
15
20
O
LA
CM
O
O
0
40
Group A
Hvla brunnea
170
236
303
364
403
275
264
399
483
319
280
306
428
547
646
286
253
00
CA
460
390
Group B
Hvla brunnea
152
41
108
133
252
113
147
186
260
123
152
225
Group C
Hvla septentriona1is
101
232
365
409
421
152
239
352
536
574
21 1
273
425
506
497
219
260
317
395
371
Group D
Hvla septentriona1is
179
108
131
273
190
82
88
138
201
113
132
217
191
106
117
257


152
Rivero, J. A. 1961 Salientia of Venezuela. Bulletin of the Museum
of Comparative Zoology 126: 1-207.
Rodriguez Garay, E., B. Noir, and M. Royer. 1965- Biliverdin pigments in
green biles. Biochimica et Biophysica Acta 1001 411-417-
Ruud, R. T. 1959. Vertebrates without blood pigment: A study of the
fish family Chaenichthyidae. Proceedings of the XVth Inter
national Congress of Zoology, Section 6: 526-528.
Savage, J. M. 1967. A new tree frog (Centrolenidae) from Costa Rica.
Copeia 1967: 325-331.
Savage, R. M. 1950. A thermal function of the envelope of the egg of
the common frog, Rana temporaria temporaria Linn., with
observations on the structure of the egg clusters. British
Journal of Herpetology 3.: 57-
. 1961. The Ecology and Life History of the Common Frog (Rana
temporaria temporaria). London: Pitman and Sons, Ltd.
Schmidt, W. J. 1919. Vollzieht sich Ballung und Expansion des Pigmentes
in Melanophoren von Rana nach Art amoboider Bewegungen oder
durch intraze11u la re Kornchenstroming? Biologische Zentralblat
39: 140-194.
. 1920. Uber das Verhalten der verschiedenartigen Chromatophoren
beim Farbenwechse1 des Laubfrosches. Archiv fur mikroscopische
Anatomie3.: 414-455-
. 1921. Uber die Xantholeukosomen von Rana esculenta. Jena
Zeitschrift 2 (N- s- 50): 219-228.
Schnedorf, J. G. and T. G. Orr. 1941. The effect of anoxemia and oxygen
therapy upon the flow of bile and urine in the newbutalized
dog. 11. Its possible relationship to the hepatorenal syndrome.
American Journal of Digestive Diseases _8: 356-358.
Schreckenberg, M. G. 1956. The embryonic development of the thyroid
gland in the frog, Hyla brunnea. Growth 20_: 295-313-
Sherlock, S. 1964. Jaundice due to drugs. Proceedings of the Royal
Society of Medicine 881.
Starrett, P. 1960. Descriptions of tadpoles of Middle American frogs.
Miscellaneous Publications of the Museum of Zoology, University
of Michigan (1 10): 5~37-
Stejneger, L. 1905- Batrachians and land reptiles of the Bahama Islands.
In: G. B.Shattuck, editor, The Bahama Islands. New York:
MacMillan Company.


68
in midstream, he rests for hours, seemingly enjoying the cool mists
which arise from the torrents. So closely does the frog resemble the
moss and lichen of his surroundings that he would rarely be observed
were it not for his big shiny eyes, which are conspicuous even when
closed, the lower eyelid being trans1ucent"(p. 56).
While Noble (1923) found that the skin secretion of this species
caused inflammation and irritation to his hands, I did not find this
to be true of the frogs collected from Furcy. Neither the boys who
caught them, nor I, after handling them, experienced any discomfort.
Noble (1927) described the early development and habits of
the young as follows: "Hyla vasta laid its eggs in little basins in the
gravel and stones on the edge of the pools in the mountain torrent
(one observation). Six days after hatching, the larvae made their way
out of one of the basins over wet stones into the torrent pool. As
they grew older they developed better stream lines than the tadpoles of
Hyla dominicensis. They were equipped with more rows of teeth. The
mouth was larger and better adapted to holding on to rocks in the stream.
Its tail was thicker and more muscular than that of the stagnant-pool
tadpole"( p. 108) .
He also reported that the eggs of Hyla vasta are pigmented and
are stuck to rocks at the bottom of the basin, and that they give rise
to dark tadpoles with small external gills, which do not rise to the
surface. The temperature of the stream he recorded as 21.7 25.6 C.,
averaging 23-7 C.


79
of the tissues. This treatment was discontinued after a relatively
short period of time because it di d not appear practical to continue.
Attempts to decalcify bone or otherwise change its pigmentation in vivo
were unsuccessful.
Induction of iron deficiency anemia in tadpoles
Iron deficiency is the most common cause of anemia in man and
may become pronounced during pregnancy or during periods of growth. In
one of its more serious forms (chlorosis) iron deficiency anemia causes
the skin to become green. During the nineteenth century, this was found
regularly among girls and young women, particularly those with menstrual
disorders.
Little is known about iron metabolism of frogs. Iron is
absorbed in the duodenal region of the intestine and incorporated into
eggs and hemoglobin (Brown, 1964). However, it seemed possible that a
lack of iron could be a cause of green pigmentation In frogs, as it is
in man. The main difficulty in running an experiment to test this idea
in frogs or most tadpoles is that their intake of iron cannot be regulated
easily. Live insects constitute the normal diet of frogs, and most tad
poles feed upon plant life and other organisms. Tadpoles of Hyla
brunnea are different in this respect, since they feed upon frog
eggs, usually those of their own species. These tadpoles develop normally
when they are fed Gerber's strained egg yolk, and kept in spring water.
The strained egg yolks were known to contain 3-07 milligrams of iron
per hundred grams (Gerber Products Company, 1965). It was thought that
the high concentration of phosphates in egg yolk (290 milligrams of


124
TABLE 6. BLOOD CHARACTERISTICS
FAMILY AND SPECIES
HEMOGLOBIN
RED CELL
HEM0GL0BI!
(Gram Per Cent)
COUNT
PER CELL
(P i cograms'
Microhy1idae
Elachistocleis ovale
8.0
597,500
134
13.0
805,000
161
9-0
725,000
124
Pseudidae
Pseudis paradoxus
7-0
887,500
79
10.0
782,500
128
9.0
767,500
117
8.0
833,000
96
Buronidae
Bufo typhonius
7.5
505,000
149
6.0
460,000
130
Hy1idae
'*H v 1 a calca rata
7.0
517,500
135
*Hyla crepitans
8.0
962,500
83
*Hyla qeoqraphica
8.0
682,500
117
*Hyla heilprini
11.4
400,000
285
Hyla lanciformis
5.5
320,000
172
6.5
345,000
188
7.5
477,500
157


103
pigmentation of the skin appears to have advantages over the green color
produced by the physical arrangement of different types of chromato-
phores. It appears to be long lasting and apparently does not change
according to light, temperature, or hormonal conditions. As long as the
frog remains associated with green vegetation, this should not be parti
cularly disadvantageous. Perhaps the efficiency of energy utilization in
the pigmented forms provides a greater advantage. Since the green pig
mentation results from diffusion of an apparently harmless waste product
into skin and other organs, little or no metabolic energy is required
for the coloring process. In addition, chromatophores may be reduced
without selective disadvantages, thus decreasing the nutrients and energy
needed to maintain chromatophores.
Disadvantages where chlorosis is the normal condition are not
readily apparent, but Lutz and Lutz (1938) stated that Sphaenorhynchus
orophilus loses its green color when not in good health, a situation
opposite to that which would be expected if the green indicated a diseased
condition.


62
the sky was overcast during much of the day so that it appeared unlikely
that the high oxygen content was due to photosynthesis. At the time
of measurement, a strong breeze was blowing and may have been respon
sible for the presence of so much oxygen in the shallow pool. Although
the tadpoles were moderately active at the high temperature, they rose
to the surface infrequently, and then as often to feed from the surface
film as to take air.
The majority of my observations on Hyla septentriona1?s were
made during two visits to Grand CaymanJuly 5-8, 1965 and October 28-30,
1965. At the time of the July visit, most of the island had been without
rainfall for more than a month and was quite dry. Adult Hyla septentrio
nal i s were found only in the vicinity of Georgetown; they became quite
active following a shower on the evening of July 7- Tadpoles were not
found in any of the cow wells inspected, but a number were collected
from a small cistern which was well shaded. The cistern was located
behind an abandoned church 6.4 kilometers southeast of Georgetown and was
about 50 meters from the shoreline.
During the October visit, tadpoles were still present in this
locality. Considerable rain had fallen during the two weeks prior to
my visit so that a number of temporary pools of rain water had collected.
Tadpoles were collected from several small woodland pools near South Sound,
from a small shaded pond and from an open pond near North Side. The
woodland pond and pools were less than 20 centimeters deep; the open pond
was about 50 centimeters in depth.


26
mixture of peas, carrots, and spinach. Hyla brunnea, Hyla iichenata,
and Hyla marianae were fed strained egg yolk. Tadpoles were maintained
satisfactorily for two months or more on these diets. Of those on the
above diets, individuals of all species but Hyla septentriona1is reached
metamorphosis under laboratory conditions. Following metamorphosis,
the young frogs were kept in covered jars, and were fed termites and
other insects.
Adult frogs were collected by hand, using a flashlight at
night to locate them at their calling locations, or by searching for
them in their resting sites during daylight hours. They were transported
in jars or in plastic or cloth bags. In the laboratory, they were kept
in jars or terraria. Crickets or other insects were used to feed the
adult frogs at approximately weekly intervals.
In addition to live specimens, hundreds of preserved frogs
and tadpoles were surveyed for unusual characteristics. Most of these
specimens were of West Indian hylids and were obtained from the museums
mentioned under Acknowledgments.


130
TABLE 9 (Continued)
SPECIES
WEIGHT(mq)
OXYGEN UPTAKE (ul 0?/q/hr)
Pseudis paradoxus
3,013
35
Pseudis paradoxus
6,782
41
Pseudis paradoxus
5,806
36


76
were readily available in Florida, none of which had green bones. Four
methods were utilized in attempts to increase the concentrations of
biliverdin in the experimental animals. Phenylhydrazine hydrochloride
was used to destroy the red blood cells in the frog, thus causing its
hemoglobin to be degraded into biliverdin more rapidly. The second
method was the injection of additional hemoglobin, presumably of mammalian
origin, again to increase the rate of biliverdin formation. Third was
the direct injection of biliverdin into the animal. The fourth method
was to induce an iron-deficiency anemia, such as that known to cause
chlorosis in man.
Pheny 1 hydrazine hydrochloride administration
Since biliverdin formation was induced in Bufo arena rum by
means of phenylhydrazine injection (Cabello, 1943), this method was used
in an attempt to develop green bones in species which were known not
to have green bones.
Two adult specimens of Bufo terrestris, along with ten Hyla
cinerea were given intraperitonea1 injections of phenylhydrazine hydro
chloride in water in October 28, 1964. The dose was at the rate of
1 microgram of phenylhydrazine hydrochloride per gram of body weight of
the frog. On November 19, 1964, each of these frogs was given a second
dose of 2 micrograms of phenylhydrazine hydrochloride per gram of body
weight. During the period of the experiment these frogs were fed twice
a week. The animals were observed daily until December 9-
Injection of phenylhydrazine hydrochloride caused destruction
of red blood cells and formation of biliverdin, but like the other