INFORMATION TO USERS

This material was produced from a microfilm copy of the original document. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the original submitted.

The following explanation of techniques is provided to help you understand markings or patterns which may appear on this reproduction.

1. The sign or "target" for pages apparently lacking from the document photographed is "Missing Page(s)". If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting thru an image and duplicating adjacent pages to insure you complete continuity.

2. When an image on the film is obliterated with a large round black mark, it is an indication that the photographer suspected that the copy may have moved during exposure and thus cause a blurred image. You will find a good image of the page in the adjacent frame.

3. When a map, drawing or chart, etc., was part of the material being photographed the photographer followed a definite method in "sectioning" the material. It is custGmary to begin photoing at the upper left hand corner of a large sheet and to continue photoing from left to right in equal sections with a small overlap. If necessary, sectioning is continued again — beginning below the first row and continuing on until complete.

4. The majority of users indicate that the textual content is of greatest value, however, a somewhat higher quality reproduction could be made from "photographs" if essential to the understanding of the dissertation. Silver prints of "photographs" may be ordered at additional charge by writing the Order Department, giving the catalog number, title, author and specific pages you wish reproduced.

5. PLEASE NOTE: Some pages may have indistinct print. Filmed as received.

Xerox University Microfilms 300 North Zeeb Road Ann Arbor, Michigan 48106 74-10,990 KOLSON, Richard Arthur, 1927- BEHAVIOR OF THE BULLFROG (RANA CATESBEIANA) WITH SPECIAL EMPHASIS ON VOCALIZATION. The Ohio State University, Ph.D., 1973 Zoology

University Microfilms, A XEROX Company, Ann Arbor, Michigan

0 1974

RICHARD AUTHUR KOLSON

ALL RIGHTS RESERVED

THIS DISSERTATION HAS BEEN MICROFILMED EXACTLY AS RECEIVED. BEHAVIOR OF THE BULLFROG

(RANA CATESBEIANA)

WITH SPECIAL EMPHASIS ON VOCALIZATION

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

By

Richard Arthur Kolson, B.Sc., M.Sc.

* * * * *

The Ohio State University 1973

Dr. Donald J. Borror Dr. Theodore A. Bookhout Approved by Dr. Loren S. Putnam Dr. Walter C. Rothenbuhler Adviser // Department of Zoology TABLE OF CONTENTS

Page ACKNOWLEDGEMENTS ...... ii

VITA ...... iii

LIST OF T A B L E S ' . iv

LIST OF ILLUSTRATIONS...... V

INTRODUCTION ...... 1

MATERIALS AND METHODS ...... 7

Construction of the Experimental Pond . . . 7 The Pond The Observation Building Fencing the Area Maintenance of the Area Bullfrogs Used During the S t u d y ...... 18 Capture of Specimens Marking of Specimens Maintenance of Specimens Instrumentation of Experimental Area .... 31 Barometric Pressure Rainfall Wind Maximum-Minimum Temperature Weather Description Temperature Humidity Index Sound R e c o r d i n g ...... 40 Recordings from Natural Ponds Recordings from the Experimental Pond Sonagram Terminology Use of a. Ceramic Model for Behavior Studies 51

RESULTS, ANALYSES, AND DISCUSSION ...... 60

Self Maintaining Behavior ...... 60 Hibernation and Emergence Daily Activity Hunting and Food Seeking Behavior Influence of Sex and Age RESULTS, ANALYSES, AMD DISCUSSION (Continued)

Vocal Behavior...... 100 Types of Vocalizations Produced The Effects of Temperature on Vocalization The Call Energy Index (CEI) Territorial Behavior ...... 145 Maintenance of Territory Playback and Model Presentation Mating B e h a v i o r ...... 154

CONCLUSIONS ...... 157

SUMMARY ...... 169

LITERATURE CITED ...... 176 ACKNOWLEDGEMENTS

Although there are many people who have helped

me, I am especially indebted to Dr. Donald Borror, whose

knowledge and advice were indispensable.

Even though he is not here to read this I wish

to express my deep gratitude to the late Arthur M.

Vorys. He was more a friend than a landlord and never

hesitated to allow me to construct my experimental area

on his property.

I also wish to extend special thanks to Dr.

Walter Rothenbuhler and Dr. Kraig Adler for their

encouragement and contagious enthusiasm.

My sincere thanks to Dr. C. Benjamin Meleca for

insuring that I had the time to complete the final

writing.

And finally, my wife Nancy, who was and is my

severest and most trusted critic. VITA

December 5, 1927 . . . Born— Cleveland, Ohio

1946 ...... Willoughby Union High School, Willoughby, Ohio

1947-1950 ...... Employed as chemist at Barium and Chemicals, Inc., of Willoughby, Ohio

1951-1953...... Army service. Medical Basic Training; transferred to Army •Chemical Corps, Edgewood Arsenal, Maryland

1957 ...... B.Sc., The Ohio State University, Columbus, Ohio. Major: Pathogenic Bacteriology

1961-1962 Research Fellow, Wildlife Research Unit, The Ohio State University, Columbus, Ohio

1962 ...... M.Sc., The Onio State University, C o l u m b u s O h i o . Thesis title: Pathogenesis of Leptospirosis in some Small Mammals Following Experimental Infection

1962-1970 Graduate Teaching Assistant and Associate, Department of Zoology, The Ohio State University, Columbus, Ohio

1970-present Research Specialist, Introductory Biology Program, The Ohio State University, Columbus, Ohio LIST OF TABLES

Table Page

1. Vocalizations produced by the Bullfrog ...... 101-102

2. Parameter Estimates for the "Mating C a l l " ...... 124

3. Parameter Estimates for the "Territorial C a l l " ...... 125 LIST OF ILLUSTRATIONS

Figure Page

1. Completed pond excavation ...... 10

2 . Elevation of the observation building and enclosure ...... 17

3. Top view of the experimental area .... 17

4. The experimental pond in 1969 ...... 20

5. Collection data and identification form . 24

6. Young male bullfrog with bead marker . . 27

7. Thermocouple and sunshade . 35

8 . Positions of thermocouple on pole and f l o a t ...... 39

9. Float in w a t e r ...... 39

10. Experimenral area with parabolae reflectors set-up for sound recording ...... 43

11. Vocalisation terminology ...... 49 12. Ceramic model of bullfrog used during playback and model presentation exoeriments ...... 53

13. Wooden slide block ...... 55

14. Arrangement of suspension wires, control strings and p u l l e y s ...... 55

15. The ceramic model floating in the "low posture"...... 59

16. The ceramic model floating in the "high posture" and being approached by a female bullfrog ...... 59 v 17. Relationship of air and mud temperature of experimental pond, August 1958 to August 1970 ...... 63

18. . Various temperatures of the experimental area for a typical late spring day (May 21, 1 9 5 9 ) ...... 72

19.. Comparison of the relative humidity of the experimental area with that of the adjacent f i e l d ...... 80

20. Various temperatures of the experimental area for a typical fall day (October 8, 1 9 5 8 ) ...... 95

21. Juvenile cluck followed by adult pop . . 105

22. Adult pop followed by two non-pulsatile territorial call notes ...... 105

23. Pulsatile territorial call note followed by adult p o p ...... 105

24. Adult male growl followed by male pop . . 109

25. Adult male g r o w l ...... 109

25. Adult male g r o w l ...... 109

27. Adult alarm call or s q u a w k ...... Ill

28. Two non-pulsatile call notes followed by four burps ending in a p o p ...... Ill

29. Two juvenile c l u c k s ...... Ill

30. Female release calls ...... 115

31. Female release calls ...... 115

32. Male clasping calls ...... 115

33. Male orowl followed by 3 female release c a l l s ...... 117

34. Four female release calls followed by male growl ...... 117

vi 35. Male growl followed by 2 female release calls ...... 117

35. Mating call, June 10, 1959 . . . 120

37. Mating call, June 18, 1953 . . . 120

33. Mating call, June 21, 195 9 . . . 120

39. Territorial call, August 11, 1969 120

40. Territorial call, August 14, 1959 120

41. Territorial call, August 19, 1969 120

42 . Mating call, July 5 , 1970 . . . . 122

43. Mating call, July 21, 196 7 .. . . 122

44. Mating „all., July 28, 195 5 . . . 122

45. Territorial call, August 26, 1969 122

AS Territorial call, August 28, 1959 122

47. Territorial call, September 1, 19? 122

48. Non-pulsatile terri tor.i al call 128

49. Mon-nulsatile territorial call .128

50. Mon—uulsati1e territorial call • 128 si; Female territorial call ...... 131

52. Female territorial call ...... 131

53. Female territorial call ...... 131

Comparison call notes nor vocal nation for natural anc e r imental nond - 137

55 . Graph of Call energy pattern for April 26-August'* 14, 1°70 . . 141

vii INTRODUCTION

The most familiar sounds of spring are probably those of newly returned songbirds. Other typical spring sounds, although enjoyed by fewer persons, are those of frogs and toads. In April the ponds of central Ohio are sites of deafening choruses of spring peepers, (Hyla crucifer), while by late May, few, if any, individuals can be heard. Each of the Ohio species of frogs and toads has a more or less distinct time of breeding, and chorusing is associated with it. There are some species, however, which continue to call long after mating has occurred. In Ohio these are the green frog, (Rana clamitans). and the bullfrog. (Rana ca tesbei ana).

With such species as the wood frog, (Rana sylvatica), the spring peeper, (Hyla crucifer), the striped chorus frog, (Pseudacris niqrita), the cricket frog, (Acris crepitans), the American toad, (Bufo americanus), and the Fowler's toad, (Bufo woodhousii), it is easy to correlate their vocalizations with breeding activities as they cease to call once mating has been accomplished. However, in the case of the green frog and the bullfrog, the explanation for calls heard

1 throughout the summer and as late as the last of August becomes somewhat more involved.

The classic work on frogs and toads by Dickerson

(1905) noted that certain vocalizations were produced by these animals only at certain times or under certain conditions. Bogert (1960) suggested that anuran calls could be placed into broad classes such as mating calls, territorial calls, release calls, warning calls, and distress calls. This idea is brought out by Capranica

(1955) in his recent book where he stated (p. 7) "the facts are that most anuran species utter distinct sounds and seem to possess an organized, though small, vocabulary." In Capranica's (1968) laboratory study of the vocal repertoire of the bullfrog (p. 302) he stated that most of the work on anuran vocalization has been directed to the mating call, and that "Inasmuch as other calls are produced generally at lower sound levels and less frequently in the field, we have less information about the signal characteristics of these calls and the biological significance they convey in anuran sound communication." The work of Capranica (1965, 1968) has been the most exacting as to laboratory analysis of various sounds produced by the bullfrog. However,

Capranica's work has been on bullfrogs kept in the laboratory and under extremely confined conditions.

Field studies which have included observations on vocalizations of bullfrogs under natural conditions have been limited to the work of Wiewandt (1969) and Emlen

(1968).

When this study was begun in the summer of 196 7

the objective was to follow Capranica's suggestion to do

a field study to confirm his laboratory work and expand his findings. Environmental factors had to be included,

as these obviously play as vital a role as social inter­

action in influencing the bullfrog's behavior. As a consequence this study became far more complicated than was originally anticipated, involving more than just vocalization and expanding into a broad study of the bullfrog in its natural environment.

There have been two approaches to the study of anuran behavior, studies conducted in the laboratory and those carried out in the field. The laboratory approach has certain definite advantages that are obvious to any field worker crouching in a blind and sweltering under a layer of mosquito repellant. Animals confined to a laboratory can be observed under controlled conditions and are usually available to the observer under pleasant circumstances. Instruments can be set up permanently and without risk of being damaged by adverse weather.

Three of the more recent laboratory studies which demon­ strate this have been made by Capranica (196 5, 1968) and

Bergeijk (1967). A considerable amount of valuable information comes from such observations. However, laboratory studies, although very important, must be given validity by observations in the field. The problem is clearly stated by Tinbergen (1965, p. 62) "....if we want to understand fully how external stimuli help control an animal's behavior, we must do more than investigate what they can respond to— we must find out what at any given moment they actually do respond to."

This brings us to the second approach, or field study. A number of excellent field studies which typify this approach have been made on the bullfrog by Emlen

(1968), Durham and 3ennett (1963), Brode (1959), Martof

(1953a, 1956), and George (1940). Field work is tremendously difficult, not only because of the incon­ venience, but also from the standpoint of interpreting the data. There is the staggering complexity of uncon­ trolled environmental factors— direct solar radiation, thermal radiation from the atmosphere, the environment and the animal, conduction and convection to and from the animal, direct and reflected light, barometric pressure, humidity, wind velocity, rainfall, etc. In addition to these factors are the equally complex social interactions. In spite of the tremendous com­ plexity of factors to which the animal is subjected in natural surroundings, it is here that the final answers must be found.

When I began this study I approached the problem in much the same way that most field workers do. This entailed traveling to a pond, setting up observational equipment, working for a few hours, then returning home.

I quickly found that areas which at first seemed ideal, such as ponds and wildlife areas, proved impossible to use due to the constant threat to animals and equipment from children, hunters, and animal predators. Often there were disturbances during observation periods by neighborhood children playing or fishing, and it was common, upon returning to a pond, to find marked individual bullfrogs dead along the banks with their hind legs removed. Perhaps the most serious drawback was the limited time I could spend at the pond. Time consumed in traveling to and from study sites (in my case as much as two hours), time to set up and take down equipment, and the cost of such trips limited the observation period and the number of field trips that could be made weekly.

I felt that what was needed was time— more concentrated observation time of the bullfrog in its natural environ­ ment. To eliminate the frustration of disturbance there was the ne for a private experimental area. I wanted the opportu.i:;. fy to study the bullfrog unobserved in a small section of a pond— not just a few hours a week but several hours a day, at night, during various weather conditions and in relative comfort from heat, cold, rain, etc. The answer was a small experimental pond on the property where I lived. MATERIALS AND METHODS

Construction of the Experimental Pond

The Pond

The pond site was located on the rental property

of Arthur M. Vorys, 6995 Havens Corners Road, Blacklick,

Franklin County, Ohio. The exact location of the pond

is 40°01' N. lat., 82°49' W. long., and is at an ele­ vation of 980.007 feet. The pond was located on a high

point of ground that sloped gently to the west.

Selection of the site and dimensions of the experimental pond were determined by three factors: (1) the land was being leased— consequently a limited area was avail­

able for such use on the approval of Mr. Arthur H.

Vorys, the owner; (2) a building, once used for a chicken coop and in very good condition, was present and avail­ able; and (3) money was not available to have a pond dug, and the amount of earth that could be removed by hand tools would naturally limit the size of the pond.

Construction of the experimental area was begun during the summer of 196 7. I decided that the existing building could be used for an observation building, and

that the entire south wall would make up the north side of the enclosed area. The pond area measured 8 feet wide, 3 feet deep, and 13 feet 6 inches long, its great­

est length extending east and west. From the north side,

the bottom extended south for 4 feet of its 8 feet width and then sloped steeidily upward to the south bank. The

sloping south side was to permit planting of emergent water plants such as cattails and water plantain. The entire area was calculated to have a volume of approx­

imately 5 cubic meters or to contain 13 76 gallons of water. This size required the removel of 5.8 cubic yards of earth. Figure 1 shows the completed excavation for the pond.

The bulk of the sides and bottom consisted of hard packed yellow clay. On completion of the excavation the sides and bottom were further packed with a heavy tamp. This was to reduce the amount of leakage once water was introduced.

The experimental pond was designed to be as natural an area as possible considering the limited time for its establishment. I wanted it to be literally a

"piece" of a naturally occurring pond. In order to quickly establish the necessary plants and microflora various materials had to be introduced. To achieve this a trip was made to the numerous ponds at the Delaware

Wildlife Area some 40 miles to the north. The materials 9

Fig. 1. — Completed pond excavation 10

F ig . 1 11

obtained were 40 gallons of pond muck and various aquatic

plants. The heavy dark muck was flowed into the deep

side of the pond.

During the late summer of 1968 the following

plants were identified using Fassett's Manual of Aquatic

Plants (1940): cattail (Typha anqustifolia var,

elonqata), leafy pondweed (Potamoqeton foliasus), large-

leaf pondweed (Potamoqeton amplifolius), duckweed (Lemna minor and Lemna perpusilla), water milfoil (Myriophyllum

sp.), and water plantain (Alisma plantaqo-aquatica).

One of the many problems to be solved was that of maintaining an adequate water level. The observation building had a roof area of 288 square feet which would provide slightly over 154 gallons from a half inch rain­

fall. The problem was that the roof slanted away from

the pond. To correct this a system of removable down-

spouting was constructed which directed rainwater into

the pond. To supplement this source of water, particularly during times of low rainfall, rainwater was obtained from

a cistern on the property and carried in 50 gallon drums

to the pond. This water was poured into a funnel shaped portion of downspouting that entered the pond through the fence area. Consequently it was not necessary to enter

the area to add water and the amount of disturbance was

thus lessened. From June 6 to October 23, 1968, a total 12

of 3150 gallons of cistern water was added. The following

summer was a wetter season, and from May 6 to October 24,

1959, a total of only 1500 gallons was added. Rainfall

for this period amounted to 8259 gallons, making the

total water entering the pond during the summer of 1969

to be 9769 gallons. These figures furnish some idea of

the water requirements of even so small an area.

The Observation Building

The building required some minor modifications.

A window was put in on the west end of the south wall

and was hinged to permit its being opened up and inside.

A second pair of windows, one above the other, was

located about 5 feet east. Both were hinged to permit

opening into the building. The westernmost window was

covered with a two-way mirror plastic sheet measuring

21 by 18 inches. This material was purchased from the

Edmund Scientific Company, 600 Edscorp Building,

Barrington, New Jersey 08007. The upper and lower windows to the east were each covered with heavy brown

wrapping paper except for a 10-inch by 12-inch area

covered by the two-way mirror material. The lower window

extended almost to the ground level; it was used both for observation and for photography. For this purpose a vinyl window shade was installed on the outside bottom

sill in such a way as to permit being drawn up from 13 inside the building- The edges of the shade traveled in channels and this along with spring tension kept the surface wrinkle free and taut. A 3-inch hole was cut so as to be at about the center of the shade when it was drawn up. With this arrangement the shade could be pulled up and hooked from inside the building in the morning. The window could then be opened at any time during the day to permit photography. Both color and black and white photographs were taken to record various activities. A 35 mm H-l Pentax with telephoto lenses ranging from 135 mm to 400 mm was used for the photo­ graphy. Black and white pictures were taken on Tri-X

Pan and developed for an ASA of 1200 using Acufine developer. Color photographs were made on Hi-Speed

Ektachrome and developed for an ASA of 400.

Fencing the Area

To prevent the loss of experimental animals by predation or wandering the entire area was fenced. This consisted of a five-foot-high fence of heavy 3-inch mesh chicken wire which extended 8 to 10 inches below ground level. Over this fence was placed a second covering of light 1-inch mesh chicken wire which was closely attached with staples at the posts and with clips elsewhere. A problem quickly arose upon introducing the first bull­ frogs to the area. I had neglected to consider predation 14 from the air, and after having lost several specimens to the local owls, the area was closed in at the top as well as at the sides. This was accomplished by stretch­ ing heavy wires across the. top using turnbuckles to obtain the required rigidity. Over these were placed lengths of 1-inch chicken wire, each length being clipped along its edge to the next until the entire area was covered. The chicken wire was then clipped to the supporting wires to further anchor it. Using this technique eliminated the need for upright supporting poles.

Once the area was fenced and ready for use, the bank areas were filled with topsoil and planted with grass.

In order to reduce the disturbances to the experimental animals brought on by movements about the yard and adjacent land, a series of plants was tried to produce a natural wall just outside the fence. In the summer of 1968 a planting of tall, sunflower-like plants with the common name "Mexican sunflower" (Tithonia rotundiflora) was quite successful. The following summer both castor beans (Ricinus communis) and common pokeweed

(Phytolacca americana) were tried. The pokeweed provided the best cover with excellent growth even in poor soil.

Figure 2 shows a side elevation of the observation building and pond, and Figure 3, a top view of the entire experimental area.

Maintenance of the Area

Throughout 1968, 1969 and 1970 the experimental area required constant maintenance. Early in the spring while the ground surface was still frozen the old growth of cattails was cut down to the surface of the pond.

This made removal of dead material easy as the frozen surface could simply be raked clean of old stalks and leaves. Frost damage to the fencing was also repaired at this time. With the first warm weather the banks were reseeded wherever the grass had worn thin. Through­ out the summer the grass was clipped regularly in certain areas, while permitted to grow in others. The reasons for this were: (1) clipped areas permitted good obser­ vation of frogs, (2) clear areas and overgrown areas provided natural boundaries on the banks which proved important to the establishment and maintenance of terri­ tories, and (3) overgrown areas provided dark, cool, disturbance-free retreats for the frogs during the heat of the day. The pattern of grass clipping was to cut all of the north bank, half of the east bank and about half of the south bank to about 2 inches in height.

Vegetation on the southeast corner, the west half of the south bank, and all of the west bank were permitted to 16

Fig. 2. — Elevation of the observation building and enclosure

Fig. 3. — Top view of the experimental area 17

Fig. 2,

POKE WEED

• u y r +) fT, M C M GHASS TEMPERATURE THERMOCOUPLE „. l ,| 1 I

.mATivE HUMiomr thermocouples

«Q - FLOAT t c m »c »AT|U»1B THERMOCOUPLE J ~ . li- S U N SHIELD

f WGH GRASS -CHICKEN WIRE fENCE

‘ ^.WATER INLET

PERMANENT GUTTER ^ R E M O * RLE GUTTER

-K-

RELATIVE HUMtCHTT AHO h MAX.—MIN. THERMOMETER grow.

The pond also required constant care. Cattails

rapidly sending out new underwater lateral growths v/ould

have quickly filled the entire pond. These new shoots

were regularly removed, keeping the cattail growth

relatively constant. Duckweed and submergent plant

growths were removed at intervals to keep the water clear

below and at the surface.

When the plants were well established 6 small

common shiners (Notropis cornutus frontalis) were placed

in the pond, both to test the conditions of the pond and

to eliminate any mosquito problem that might arise. In

April of 1970 small fish were observed in the pond.

Figure 4 shows the pond as it appeared in 1969.

Bullfrogs Used During the studv

Capture of Specimens

The first bullfrogs were placed in the pond on

April 28, 1968. All were given letters of the alphabet

starting with A. On June 14, 1968, 6 specimens were

permanently settled in the new experimental area. These

were designated as L, M , N, P, Q, and R. Frogs L, N,

and P were adult breeding males, while M, Q, and R were

adult females. Female M was particularly interesting in

that she was a blue specimen captured at the Delaware Fig. 4. — The experimental pond in 1969 20

Fig. 4. Wildlife Area.

One method used was to collect at night using a powerful flashlight. The technique was to slowly wade along the pond edge until a suitable specimen was ob­ served. With one hand the light was held steadily on the frog, while the other hand was used to quickly grasp the animal behind the fore limbs. Capturing frogs in this manner often failed because of the great difficulty in approaching close enough to seize the animal. The use of a long handled net proved very useful when the frogs were easily disturbed. The success of the method depends on the fact that a frog sitting along a pond edge will generally leap toward the water when disturbed.

The technique was to cautiously place the net between the frog and the water, then take a sudden step forward, causing the anirnal to leap— usually directly into the net. This method proved quite successful in the early spring although later in the season the algae became so thick as to interfere with the movements of the net.

Finally the use of a plug casting outfit proved quite efficient for capturing males early in the breeding season. As the breeding time approaches the large adult males tend to move out from the shore in the evening and begin to establish their territories (Emlen, 1968). I found that around sunset was the best time to plug cast 22 for specimens, as the large breeding males had already-

taken up their territorial positions by this time. The use of a very small plug or so-called ’’popper," dropped slightly beyond a male and moved in slow jerking move­ ments past it, almost always resulted in the bullfrog's readily taking the bait. The tiny hook caused only slight injury to the jaw area which quickly healed with­ out complications.

Marking of Specimens

Each frog collected was weighed and sexed, and natural markings.which occurred on the head were recorded.

In addition to their letter designations, all frogs were permanently marked for field identification. All of this information was then recorded on a standard form. Figure

5 shows the collection data for bullfrog

Many marking techniques are available for use with amphibians, several of which were given in

Woodbury (1956). The methods used up to the time of this writing have been jaw tagging (Raney, 1940), toe clipping

(Martof, 1953a), banding, scarification of the abdomen, tattooing (Kaplan, 1958, 1959), and the use of colored elastic waist bands (Emlen, 1968). These marking techniques presented two problems. First, some of the techniques required recapture of the animals for iden­ tification. Second, none could be used tor identifying 23

Fig. 5. — Collection data and identification form 24

FROG No. and SEX 9(vJ)

UtJPl«M«MTCD 'ZfJ*.

O

- ft* A/ w S c M

RIGHT SIDE LEFT SIDE

LEFT SIDE RIGHT SIDE BEAD MARKING

MEASUREMENTS

HEAD (1) Width at widest angle of jaws: 5 3 (2) Head length (posterior tympanum to snout): (3) Tympanum: (a) widest diameter: (b) narrowest diameter: (4) Intertympanal width: 3 0 >n»>. BODY (1) Length from vent to snout: /-40 mr*-

LEG (1) Tibial length: GS (2) Length of foot: io4 WEIGHT (in grams): 245-5 qms- COLOR CODE DESIGNATION:

Fig. 5. an individual as it floated in the water with only the top of its head visible. The latter problem was probably the more significant as this is usually the way the frog is seen under natural conditions. Finally there was the ever present concern that the marking technique might affect the normal behavior of the individual. With all of this is mind a new marking technique was devised. By this method the frog was marked with a tiny plastic rocaille bead measuring

2 x 2.5 mm-which was attached to the dermal plicae at the top of the tympanum by a single ligature of fine nylon filament. The young male bullfrog in Figure 6 has been marked with a single red bead. For the purpose of easy recognition, three colors were chosen— red, white, and blue. By the Rheinhold color classification

(Kornerup and Wanscher, 1962) the red bead was class­ ified 10A8 and the blue, 23A6. The beads can be used singly or in pairs, and on one or both sides of the head.

A total of 93 combinations could be obtained by this method. Using more than two beads together is not recommended due to the possible difficulties which could^ arise both in attaching them and in quickly recognizing the color combinations.

The bead was attached by use of a small needle and very fine nylon filament which were both soaked for 26

Fig. 6. — Young male bullfrog with bead marker

28

about 15 minutes in a saturated solution of phenylmercuric nitrate. Phenylmercuric nitrate was used as a disin­

fectant because of its very low solubility and toxicity yet high germicidal powers, (Weed and Ecker, 193 3;

Birkhaug, 1933); the low solubility insured long lasting effects after the frog was released into the pond.

The needle, threaded with the nylon filament, was entered just posterior to the eye below the dermal fold

lying at the highest point of the tympanum and directed

toward the midline of the head. The tissue here is predominantly connective tissue and consequently there is very little, if any, bleeding. As the needle emerged on the opposite side of the fold carrying the thread with it, it was simply slipped off the thread, leaving the latter in the tissue. Onto one of the free ends of the nylon thread a single bead or pair of beads was threaded.

Two square knots were then tied tightly enough only to bring the beads snugly to the surface of the skin. Any excess thread beyond the last knot was cut off and a hot needle was lightly touched to the nylon, fusing the ends permanently. Excessive heat should be avoided during melting of the knot to prevent cutting through the thread. At the time of this writing the frogs have been carrying their identification beads for about one and one-half years and are still readily recognizable. 29

Maintenance of Specimens

Once the frogs were in the pond the next problem that arose was the maintenance of an adequate food supply. This was particularly important during the first season as a natural food supply would be at a minimum until the experimental area became established. Early in May of 1968 I was able to collect June beetles

(Phyllophaqa sp.) by setting out floodlights around my home and garage. The lights were directed against white siding which attracted large numbers of these insects.

Every evening throughout their active period the insects were collected until the night temperature chilled sufficiently to reduce their numbers. The beetles were put into 2-pound coffee cans lined with paper toweling to absorb moisture, then put into the refrigerator.

About 3 times a week approximately 40-50 beetles were fed to the frogs during May and June. The feeding was done around dusk through an open window of the observation building. The window was opened early in the day to prevent disturbance. To avoid conditioning the frogs, actual appearance at the window v/as avoided.

Supplementing this source of food were various small insects that continually moved into the area from the surrounding fields. Later in the season small grass­ hoppers began entering the area: the larger grasshoppers 30 were collected and placed in the enclosure. Seining trips were made to local small streams, providing crayfish which were placed in the pond.

With the approach of the first fall season in

1968 I became very concerned that the frogs obtain adequate food to insure fat and glycogen storage for the coming winter. Frogs are notoriously obstinate in accepting anything other than live, moving food. To avoid the nuisance of establishing colonies of mice or obtaining meal worms I devised a new technique which proved to be both convenient and efficient. A long flexible bamboo fishing pole with a light fish line attached v/as used to present the food to the frogs.

After experimenting with various foods such as worms and hamburger, the ideal food was found to be beef kidney.

Kidney is very easily attached to the line, having just the right density and consistency.

Frogs appear to be quite selective in taking artifically presented food as anything too soft or too hard is quickly rejected: chunks of beef kidney were always readily accepted. A very simple, effective technique was developed for attaching the kidney to the line. Using a large needle the line was literally sewn into the kidney by passing the threaded needle back and forth through the chunk of meat making 3 stitches. The 31 needle was then slipped off the line and the end of the

line pulled gently into themeat. This method held the bait quite firmly while allowing the frog to easily

remove it.

One of the observation windows was opened early

in the day and as evening approached the frogs could be

fed as they entered the pond from the tall grassy areas.

As the pole was very long it was completely unnecessary

to appear near the open window and the frogs never seemed

to associate the open window with feeding. Once having

taken food in this manner, the frogs quickly learned to recognize the kidney chunk as "food," and would often leap several inches out of the water or off the banks to

seize the meat in mid-air. About one half a kidney was fed every 4 to 5 days.

Instrumentation of Experimental Area

In order to have as complete an environmental record as possible a large number of environmental factors were recorded daily. A meteorological data form was filled out each day. Four to five observations were made daily consisting of:

Barometric Pressure

Barometric pressure was recorded daily from 32

August 12, 1958, to January, 1970, and from April 21,

1970, to August 27, 1970.

Rainfall

A daily record of rainfall was kept for each season from early spring until freezing weather.

Wind

Both wind direction and estimated speed were observed daily. An anemometer was not available, there­ fore the wind speed was estimated by use of the Beaufort

Wind Velocity Scale. Wind direction and velocity were plotted on standard map forms along with the positions of the frogs— especially early in the spring when little cover was available. A wind vane was constructed which, through a system of simple gears, transmitted the wind direction into the observation building where it could be read off a compass rose.

Maximum-Minimum Temperature

The daily maximum-minimum temperatures were recorded from a Taylor maximum-minimum thermometer.

Weather Description

A brief daily description of the existing weather conditions was recorded. Cloud cover was estimated in

0.2 increments along with a cloud classification for 33

each observation.

Temperature-Humidity Index

The temperature-humidity index was included more

as a means of indicating the kind of day to the observer

than as to the effect it might have on the frogs.

In addition to this information a Speedomax

type G recorder was installed.during the summer of 1968

to monitor various environmental factors until early

December of that year.

The Speedomax type G recorder was equipped with

16 thermocouple leads. Three of the leads were used to monitor the north, east, and southeast corner of the area.

The north and east thermocouples were set at 2 inches

above the ground surface, and each was shaded by use of

an aluminum louver measuring 8 by lb inches and bent

into a semi-circle. Three heavy wire legs were attached

to allow each shade to be adjusted to the best height.

When completed, each sunshade formed a half circle measuring 9 inches across the open end and 6 inches deep.

(See Figure 7.) This technique of shielding the thermo­

couples proved very effective, providing shade while

allowing free movement of air. The thermocouple in the

southeast corner was placed 2 inches above the ground in deep grass and weeds, and, as the growth was quite high here, shielding was unnecessary. Fig. 7. — Thermocouple and sunshade

36

A fourth thermocouple wire was threaded through the chicken wire roof of the enclosure ending within an aluminum louver sunshield 5 feet above the south bank.

Two more wires were passed together along the roof finally dropping to 5 inches from the ground on the south bank. Here one thermocouple read the air tempera­ ture while the other, sealed into a cotton wick which dipped into a water reservoir, read the "wet bulb" temperature. An aluminum sunshield protected both thermocouples. This device gave readouts of relative humidity values at approximately every 8 minutes. The values obtained from this hygrometer were compared to those obtained with a sling psychrometer v/ith no appreciable difference noted between the two.

Installed on the roof of the observation build­ ing was a weatherproof Leeds and Northrup photocell. By means of jumper wires the Speedomax was set to give readings from the photocell at 2 minute intervals, thus producing an accurate record of the gross light conditions from dawn to dusk. Light values obtained from the Speed­ omax were converted to footcandles by simultaneous chart readings and light values taken with a Gossen Lunasix light meter equipped with an invercone. The Lunasix is sensitive to light values ranging from 0.016 to 32,000 footcandles, providing far more accuracy than the Leeds 37 rose or fell. Small lead weights positioned the depth

• . ... of the float so as to keep one thermocouple 2 inches below the surface and the other 1 inch above the water level. In addition to the temperature function this apparatus was equipped with a yard stick and marker, which provided continual information as to the pond depth. The entire assembly was painted with a green enamel to preserve the wood and to keep it harmonious with the surroundings. Figures 8 and 9 show the construction of the entire apparatus.

All 14 of the various thermocouple leads were arranged in such a way as to avoid cluttering the area.

Whenever possible the wires passed along the fence, coming into the area only at that place where they were needed. At the point where all 14 wires were brought into the building a l-mch hole v/as drilled and fitted with a 4-inch length of rubber garden hose. The wires were bundled together and passed through this hose on into the interior of the building and connected to the

Speedomax terminals.

The last two available terminals of the Speedomax were connected to a short thermocouple lead which went to a constant temperature water bath. The water bath v/as equipped with an accurate Centigrade thermometer calibrated to 0,1° C., and served as a standard which 38

Fig. 8. — Positions of thermocouple on pole and float

Fig. 9. — Float in water 39

OEPTH INDICATOR C POND BOTTOM D MOO W O O O DOWEL WATE«_SURfACt

ClAY

Fig. 8 ICAO WEIGHTS

STYtOWAM FLOAT

Fig. 9. 40 could be checked at any time against the thermocouple reading.

During the following season the Speedomax was not available, but a Micromax recorder was obtained and installed in February of 1969. The entire pond had to be rewired for this instrument as the Micromax used iron- cor.stantin thermocouples rather than the copper-constantin wire used by the Speedomax. During 1969 the same areas of the pond were monitored with the exception that the photocell was not available and the southeast corner was not monitored. In addition to the former thermocouple sites, a pair of leads recording relative humidity were placed 9 feet west of the fenced area and at a height of

6 feet. This set of thermocouples was used to compare relative humidity values obtained from the south bank of the experimental pond with the humidity outside the enclosed area.

Sound Recording

Sound recording consisted of the following categories: (1) recordings of bullfrog vocalizations occurring at various times in the season in natural ponds, (2) recordings of bullfrog vocalizations from individuals within the experimental area at various times during the season, and (3) recordings which 41

monitored the frequency at which vocalization occurred

during each evening.

Three tape recorders and two parabolic reflectors

were used throughout the study. They were: (1) Sony

905A, (2) Martel 301D, (3) Uher 4000 Report L, and (4)

an 18-inch and a 24-inch parabolic reflector.

Figure 10 shows the method used for setting up

the parabolic reflectors.

Recordings from Natural Ponds

Bullfrog activity was studied throughout the

active season not only in the experimental area, but

also at natural ponds. This was to observe how the

conditions and activity in the small experimental pond

compared with those of a large natural pond.

A large farm pond located 1700 feet south of the

experimental area was frequently visited for comparison.

Other natural ponds observed were at the Delaware Wild­

life Area and particularly those near the Olentangy Wild­

life Experiment Station on the Area. In addition to

these a large pond at Blendon Woods Park was often

visited.

During visits to the various pond sites a number

of observations would be taken. Continuous tape record­

ings for 1 to 2 hours were used to study frequency of vocalisation and patterns of chorusing. Tape recordings 42

Fig. 10. — Experimental area with parabolic reflectors set-up for sound recording. 43

\b ‘ ’i * '-‘V i. **■

•!» .'A%\

/ •>?:>• V. -tv(W ., ■ , . \i \\ > ' l-^SSSr.

W ^ \ }\\ < • - - \ l i

' • vi *.\u \

h M'

J-r. 1

Fig. 10. 44 were made of individuals for use in sound analysis.

Observations were made of behavioral activities and pond conditions such as depth, water temperature and plant growth.

Recordings from the Experimental Pond

Two kinds of tape recordings were made at the experimental area. The first recorded the vocalizations of both males and females during the various times of the season while the animals were engaged in various activities. These recordings were made on the Uher 4000

Report L at 7 1/2 inches per second which was capable of recording frequencies of 20,000 cps. These recordings were used for analysis with a Kay Vibralyzer. The second class of recordings monitored the frequency with which the animals vocalized throughout the evening, night, and early morning hours. These recordings were made v/ith a Sony 905A recorder equipped with a built-in voice activation mechanism. The microphone was used with a 24-inch aluminum parabolic reflector mounted on one of the window sills of the observation building.

As the reflector was only 15-20 feet from the pond, vocalizations associated with mating or territory easily activated the recorder inside the building. By means of an electric timing switch the recorder was turned on automatically each afternoon at 1600 hours and off again at 0600 hours the following morning. An indication of time intervals was added to each taping session by the simple means of a small alarm clock which was set to ring at 1800 hours, reset to ring at 2000 hours, again at 2200 hours, and finally for 0600 hours.

To protect the alarm clock from the weather it was placed into a small coffee can with a plastic cover, and the can was set on the window sill close to the parabolic reflector. The microphone was also completely sealed into a plastic bag to protect it from moisture. Record­ ings were made every night and under all kinds of weather conditions, ranging from fair weather to severe thunder­ storms. Each morning this tape recording was reviewed and the number of vocalizations and the calls per vocal­ ization during each time period were tabulated.

Sonaqram Terminology

During the organization of this study I became aware that terminology concerning the sounds produced by animals and in particular those produced by frogs varied widely depending on the author. In the hope of resolving some of this confusion, I devised a system of terminology that is partly mine and partly from the work of others.

I decided to refer to a sound or complex of sounds produced by a bullfrog during some natural activity as either a "vocalization" or a "call." A frog being pithed will produce a sound, but I contend that

such a sound should not be referred to as a vocalization,

at least under the conditions I have stipulated. This

use of the term vocalization or call is consistent with

the terminology used by Capranica (1968), Emlen (1968),

and Duellman (1970), although Duellman also refers to a

call or vocalization as a "call-group." I prefer not to use call-group as there are calls which consist of a

single sound. Duellman's call-group would be comparable

to my referring to a series of "call-notes." Included

in my terminology is the word "note" or "call-note." I

have considered a note to be a single sound unit which

can be simple or complex. I find this definition to be

the same as used by Borror (1961) and Duellman (1970).

As an example, consider a typical vocalization of an

adult male heard in early July. It could be phonetically

represented as: R-rum, da-da-dum, da-da-da-dum, da-da-

da-dum, r-r-rum. Here we see a typical five call-note

vocalization. The first note is represented as "r-rum,"

the second as "da-da-dum," the third as "da-da-da-dum,"

and so on. Each note begins on a low pitch, and with a

sort of grunting effort, which then accelerates to a

higher pitch except at the end of the vocalization where

the pitch falls. Some notes of this vocalization are

simple v/hile others consist of a sort of stuttering 47 repetition of sound which Duellman (1970) and I refer to as ’’pulses.” Vocalizations heard in the late fall often consist of a single note with none of the stuttered repetition of sound. This type of call by my terminology would be referred to as a "one-note vocalization."

The sonagram shown in Figure 11 is a typical adult male vocalisation heard in mid-July. This sona- grara represents a five call-note vocalization or call.

Indicated on the sonagram are the fundamental and dominant frequencies. When air passes over the vocal cords it causes them to vibrate at a frequency which is dependent on the tension of the vocal cords. The sound produced when a frog calls is generally made up of a series of frequencies called harmonics. The lowest pitched of these harmonics is referred to as the funda­ mental. That harmonic resonating with the greatest intensity is called the dominant frequency and is usually apparent as the darkest harmonic on the sonagram. Accord­ ing to Duellman (1970), the fundamental is far less variable within a given species than the dominant frequency. This is because the fundamental depends on the tension of the vocal cords, while the dominant frequency is affected by the resonating chamber— the vocal sac. Apparently a partially inflated vocal sac Fig. 11. — Vocalization terminology FREQUENCY IN KILOCYCLES PER SECOND

HARMONICS N Cl NI MJR PUISES MAJOR NOIE CAll ONE 50

produces a lower dominant frequency than a fully inflated

one.

Indicated on the sample sonagram of Figure 11

is a series of very obvious thin dark vertical lines.

These I have called major pulses. They represent the

stuttering quality of a typical adult male call. The

third note of this particular vocalization is broken up

into 7 of these major pulses. A close examination of

such a sonagram would also show a series of very fine vertical lines particularly detectable in the lighter

areas above the fundamental and the dominant frequency.

These, which I call minor pulses, have a repetition rate

of approximately 90-100 per second, probably representing

the actual vibrations of the vocal cords. The funda­ mental is approximately 150 cycles per second, the dominant frequency is about 300 cycles per second, and

the level of the upper harmonics reaches about 1400

cycles per second. The longest note has an approximate duration of 0.9 second with the call note interval

averaging 0.4 second. This entire vocalization lasted

about 5.9 seconds. Patterns produced by the higher

harmonics are highly characteristic of the individual

and are almost comparable to a "voice print." 51

Use of a Ceramic Model for Behavior Studies

One of the purposes of this study was to observe bullfrogs interacting with each other. To produce a situation which would result in such social interactions, a ceramic model of a bullfrog was used. This model could be manipulated from within the observation building.

To maintain realism, a model was chosen which was a fairly accurate representation of a male bullfrog both in size and in coloration. Made of ceramic material, the model was completely unaffected by total and lengthy submersion in water. Figure 12 shows the model sitting in the grass on one of the pond banks. To control the movements of the model in a realistic manner while remain­ ing entirely hidden, I arranged a system of strings and pulleys. These control strings were arranged to enter the observation building, allowing control of the model from within the building. A two-way mirror set into the window permitted both manipulation of the model and observation of the responses of the bullfrogs. A full description of the construction of the model controls follows.

A twelve-gauge wire was stretched across the entire length of the pond approximately five feet above the surface of the v/ater. A small, oblong piece of wood measuring 1x1x4 inches was threaded onto this 52

Fig. 12. — Ceramic model of bullfrog used during playback and model presentation experiments 53

Fig. 12. 54 suspension wire by means of a 1/4-inch hole drilled the entire length of the block. The suspension wire was drawn taut by turnbuckles at each end of the wire.

These were secured to posts which were part of the experimental area fence.

Three small screw-eyes were screwed into the wood slide, one at each end below the drilled slide hole and one in the center of the block on its bottom surface.

The diagram in Figure 13 shows the construction df the block. A very thin nylon monofilament (4-pound test nylon monofilament line) was tied to the ceramic model at a point which gave the best balance and maintained the model in a natural position. The free end of the nylon filament was passed through the lower eye in the slide block. The filament was then brought into the observation building through a small hole bored through the wall above the center window. Next it was necessary to attach control strings to each end of the slide by means of the screw-eyes fixed there. Each control string was then passed along parallel to the suspension wire to a small pulley fixed to the same post that held the suspension wire. The left and right control strings were then brought back on an angle to a point opposite the entrance hole in the observation building. At this point each string passed through a second pulley fixed 55

Pig. 13. — Wooden slide block

Fig. 14. — Arrangement of suspension wire, control strings and pulleys 56

NYLON MODEL LINE

SUSPENSION WIRE

.LEFT C O N T R O L LINE

Fig. 13 cri o

SUSPENSION WIRE

LEFT C O N T R O L LINE

M O O E L U N E

MODEL

RIGHT CONTROL LINE

Fig. 14. to the chicken wire roof of the enclosure. Each control

string then passed into the observation building.

Because the wooden slide could be moved all the way to

either end of the suspension wire, it became possible,

with a little practice, to make the model hop out of the

water onto either the east or the west bank or into the

water from either bank. This was accomplished by pulling

on either the left or the right control string while

pulling at the proper time on the nylon model-line as

the model reached the bank. With practice it became

quite easy to swim the model under water and to bring it

to the surface, either in a real bullfrog's territory or

into confrontation with a bullfrog. The model could be moved along the surface of the water or suddenly dived

to the bottom. Figure 14 shows the arrangement of the

suspension wire, control strings and pulleys, while the

photographs in Figures 15 and 16 show the model as it

appeared from a "frog's-eye-view," and was approached by one of the female bullfrogs. 58

Fig. 15. — The ceramic model floating in the "low posture"

Fig. 16. — The ceramic model floating in the "high posture" and being approached by a female bullfrog 59

Fig. 16. RESULTS, ANALYSES, AND DISCUSSION

Self Maintaining Behavior

Hibernation and Emergence

Observations of environmental conditions were begun on August 4, 1968, and terminated in the fall of

1970. During the late falls of these years the continued absence of bullfrogs from the banks or water surface was taken as an indication of hibernation. All the bullfrogs had begun hibernation by December 4, 1968, and by November 28, 1969. Hibernation appeared to be a gradual process with numerous periods of temporary inactivity. These were always associated with cold periods that resulted in temporary freccing of the pond surface. With the return of milder temperature the thin ice covering would thaw and bullfrogs would reappear at the water surface, their heads coated with mud, indicat­ ing that they had been buried in the mud layer of the pond bottom.

As in hibernation, emergence was a gradual process. The frogs appeared during mild periods in early spring only to disappear temporarily during colder periods. The first appearance of the bullfrogs in the

60 61 spring of 1969 occurred on March 7, and the following spring on March 2.

Figure 16 gives the maximum daily temperatures of the mud bottom at a depth of 6 inches, and the air temperature at 5 feet above ground level. The horizontal bars labeled "A" represent times during which the pond surface was frozen over. During the two winters the pond surface froze to a maximum thickness of 5 inches by late

January. Bars labeled "B" indicate times during which the bullfrogs were active at the water surface or on the banks.

An interesting relationship can be seen between the times of hibernation and emergence and the air-mud temperatures. In this pond, at least, hibernation and emergence are roughly correlated with that time when the air and mud temperatures are equal. This occurred at about 6-7° C. for hibernation and 5-6° C. for emer­ gence. Based on observations made at a nearby pond, these results may not be an accurate representation of emergence. There was a large natural pond approximately

1700 feet south of the experimental pond. This natural pond was on private property and available for only limited use. The pond measured roughly 220 feet by

140 feet and was approximately 15 feet deep near the center. This pond was utilized to compare the activity 62

Fig. 17. — Relationship of air and mud temperature of experimental pond, August 1968 to August 1970 MAXIMUM DAILY TEMPERATURES O O oe

-10 20 30 40 10 0 - -l A B- - U SP OT NOVI E IJN fB A 1AR A I U JUl AGI E I C I O DC JN FB I FEB I JAN I DEC I NOV I OCT I &EP I AUG I l U J I JUN I MAY I 1~APR MAR I fEB I I JAN DEC I V O N I OCT I SEP I AUG 98 16 I 1970 I 1969 I 1968 10 HDD i. 17. Fig. in OOC EID PN FROZEN POND PERIODS A PROS ULRG ACTIVE BULLFROGS PERIODS B 'T' R A M a PR' IMY JN J I U I AUG I l JU I JUN I I"MAY U TEMPERATURE MUD I TEMPERATURE AIR u>

of the bullfrogs in a natural area with the experimental

one. During the spring of 1970 there were no bullfrogs

observed in the large pond as late as March 20, although

the bullfrogs had been out on the banks of the experi­ mental area for almost two weeks. There is a definite relationship between air and water temperatures, and

emergence and hibernation. Willis et al. (1956) clearly

pointed this out by listing the emergence dates for bull­

frogs in the United States. Their data showed a definite northward progression in times of emergence.

Time of emergence for the bullfrog in central

Ohio according to Walker (1946) is as early as April 6.

Wright (1914) claimed that emergence in the area of

Ithaca, New York, could be expected when maximum air

temperature reached 20-24° C. and water bottoms were at

least 14-18° C. Ryan's (1953) observations in the Ithaca area seemed to roughly agree with this. He noted emer­ gence dates of April 18, 1950, and April 9, 1951, with maximum air temperatures on these two dates of 21° C. and 13° C., respectively. He pointed out that although the temperature on April 9, 1951, reached only 13° C., it had risen to 18° C. on the previous day. For the area of central Missouri, 'Willis (1956) gave emergence dates of March 23, 1950, and March 28, 1951, v/ith maximum air temperatures of 19° C. and 22° C., respectively.

I have found that there is a general lack of detailed information on the bullfrog in central Ohio.

Consequently I have on occasions considered the obser­ vations made by various workers on the green frog (Rana clamitans) in evaluating and planning my studies on the bullfrog. The green frog is similar to the bullfrog in that it is highly aquatic and rarely found far from water. My observations lead me to believe that the time of emergence and hibernation are similar for the green frog and the bullfrog. This would probably be most likely if the comparison was with juvenile bullfrogs whose size compares favorably with the average adult green frog.

Size of the individual seems to have some relation to emergence and hibernation and will be discussed shortly.

Comparing the emergence data for the bullfrog with that for the green frog, I found that Martof (1956) gives the last two weeks in March as emergence time for the green frog at Ann Arbor, Michigan. His only comment as to temperature was that the daily maximum temperature should have risen above 16° C. for 3-4 days. Brenner

(1969) presented data on the emergence of the green frog which I found most significant because it involved the temperatures of both air and mud. His studies were made at two sites— Granville, Ohio, which is approximately 66

19 miles east of my experimental area, and Greenville,

Pennsylvania. Brenner gave emergence data for March 28,

196 5, at Granville, Ohio, with an air temperature of

13.2° C. and mud temperature of 9.0° C. His data for

March 23, 1966, at Greenville, Pennsylvania, showed the

air temperature to be 9.7° C. and the mud temperature to be 5.6° C.

Emergence dates for the bullfrogs in my pond for

four years (1969-1972) were March 7, 2, 12 and 4. I

observed that the frogs tended to emerge much earlier

than those in a large natural pond 1700 feet to the

south of the experimental area. (This natural pond is

spring fed and is about 15 feet deep.) For example, in

1970 the bullfrogs made their first appearance in the

experimental pond on March 2, while in the large natural

pond no bullfrogs were seen until the first week in

April. I was unable to compare the two areas for

initiation of hibernation.

There appears to be even less information on

hibernation in bullfrogs than on emergence. Walker

(1946) noted simply that active individuals were seen in

central Ohio as late as October 10. 'Willis et al. (1956)

claimed that bullfrogs in the central Missouri area

probably began hibernation by October 23 when the air

temperature fell below 16° C., and that no individuals were seen after November 14,. Comparing hibernation data of the green frog to the bullfrog , Martof (1955) stated only that hibernation occurred in the last half of October or the first half of November. Brenner

(1959) was more precise, giving hibernation dates of

November 17, 1964, for Granville, Ohio, with an air temperature of 4.2° C. and a mud temperature of 3.9° C.

His data for Greenville, Pennsylvania, gives a hiber­ nation date of October 25, 1965, with an air temperature of 4.4° C. and a mud temperature of 6.6° C. The hiber­ nation and emergence data for my pond is given in

Figure 17.

There appears to be a relationship between the size of the individual and hibernation and emergence.

For four consecutive years I attempted to observe whether juveniles emerged before adults. On one occasion I found that the first frogs to appear were juveniles, but the following season both juveniles and adults appeared together. If there is a warm period during the first days in March both juveniles and adults may emerge to­ gether. On the other hand, my observations indicate that during colder springs the juveniles will precede the adults. This apparent ability to function more efficiently at lower ambient temperatures can be seen even when both young and adult animals are observed during early spring days. I have seen adult bullfrogs at the pond edge, partially out of the water and pressed to the substrate, so inactive that I could actually touch them. Their only reaction was to retract their eyes into their heads. On this same day I watched juveniles actively moving about the pond. On my approach to the pond area the young individuals on the banks would leap into the water. The entire picture was one of active, well coordinated juveniles, and adults which, although emerged, were still at a very reduced level of activity. They were totally unable to respond to food, and often were unable to respond to the approach of the observer. These observations are corroborated by Willis et al. (1956) who also noted that young bullfrogs appear­ ed earlier in the spring than adults, and that larger adults tended to disappear from the ponu first in the fall. Martof's (1956) study on the green frog in

Michigan produced the same observation. He suggested that it might be due to a surface-volume relationship with smaller individuals being able to warm more readily than the larger adults.

There are apparently other factors involved in emergence and hibernation besides environmental tempera­ ture. Brooks (1918), working with Rana pipiens, found that the colder the temperature of the water, the more the frogs tended to be at the bottom. Frogs tend to

react to light either positively or negatively depend­

ing on the environmental temperature (Torelle 1903,

Brooks 1918, Martof 1953b). Martof (1962) found that

the ability of Fowler’s toad to react to light was

greatly affected by the ambient temperature. At 10° C.

the reaction time was doubled with less than 10% of the

625 animals tested moving toward the light. Positive

or negative phototaxis could play an important role in

hibernation and emergence, and might be related to pond

depth if we consider the amount of light that could

reach the bottom of a deep pond as opposed to a shallow

one. In addition to this, Holzapfel (1937) claimed that hibernation was an intrinsic function of the animal

itself and not fundamentally dependent upon the environ­ ment. A more complete answer will require turtner study.

Daily Activity

Response to Air and Water Temperature.— The daily

activity of the bullfrogs was greatly influenced by the various temperatures of their environment. During the

early days of spring the animals engaged in very little

activity other than basking on the warmest bank. During

the first week or two after emergence the frogs exhibited

a faint rosy hue which was most noticeable on the belly

and chest areas. This was particularly apparent in the 70 females due to the lack of throat pigmentation. At this time the animals would leave the pond in the early morning, moving onto the bank which was the most protect­ ed from the wind. They would sit for hours in a very low, flattened position, practically motionless, except for the slow buccal breathing movements.

Various temperatures of the experimental area are shown for a typical late spring day in Figure 18.

On this day (May 21, 1969) the sky was clear and sunny with the air temperature reaching 21° C. There was a light northeast wind of 1-3 m.p.h., gusting to 4-7 m.p.h.

After dark the weather remained clear, but became quite chilly with the temperature falling to near 8° C. The temperatures shown in the graph were of the air tempera­ ture at 5 feet, the air temperature of the north and east banks at approximately 2 inches, and the water tempera­ ture at a depth of 2 inches. On this day the behavior of the bullfrogs was typical for this time of year. By

0800 hours E.S.T. they had taken basking positions along the north bank which was the most protected area due to the observation building's making up this side of the enclosure. By about 1700 hours the frogs had left the banks and entered the water along the bank edges.

Throughout the remainder of the evening, until it became too dark to observe them, the bullfrogs moved out into 71

Pig. .18. — Various temperatures of the experimental area for a typical late spring day (May 21, 1969) DEGREES CENTIGRADE ro o O

III

ZL 73 deeper water and submerged farther and farther as the air temperature fell.

The activity of the frogs during the hot days of summer was quite different. As was mentioned in Materials and Methods, the grasses and weeds of certain areas with­ in the enclosure were permitted to grow while others were constantly clipped to a height of 2-3 inches. By early summer the southeast and southwest corners had become densely overgrown. These areas became cool refuges for the bullfrogs during the hot days. The bank tempera­ ture commonly reached 35-38° C. while the deep grassy areas were consistently 5-7° C. cooler* The animals normally left the pond in the early morning hours, immediately entering the overgrown areas, and did not reappear until dusk. They apparently sat there— not actively hunting— probably moving only to snap at an occasional insect which wandered by. Here the picture was of maintaining water balance through avoidance of intense sunlight and heat.,

With the setting of the sun a different behavior pattern was consistently observed. The frogs would begin

to appear out of the weeds generally in a regular

sequence of individuals: the same individual would be

the first to enter the pond each evening, followed later by the number two individual, and so on. The frogs did 74 not all enter the pond in rapid succession; 15-20 minutes or even an hour might be required before the next bull­ frog would appear. Perhaps some individuals required a greater reduction in light intensity,.temperature, or whatever factors controlled this behavior before they would leave the weeds. However, in most cases the sequence of individuals was orderly. This behavior probably reflects at least in part the physiological distinctiveness of each individual.

A puzzling type of behavior occurred as each frog left the overgrown areas. Each one would consistently move toward the gate at the northwest corner of the enclosure. Tall weeds and plants screened much of the west side of the enclosure and the entire north side con­ sisted of the observation building wall. Therefore, only the nortnwest gate was totally open to the setting sun.

The bullfrogs would hop along the banks or swim the entire length of the pond to eventually end up sitting in a group at the northwest gate— all facing the setting sun. During these periods there was absolutely no aggres­ sive behavior in spite of individuals sitting so close as to be touching. This ritual was repeated each evening provided that the sky was clear: on cloudy, overcast evenings the response was absent or only briefly present in one or two individuals. As soon as the sun set the 75

animals would begin moving back into the pond to begin

their nightly hunting activities. This peculiar

attraction to the setting sun might be involved with the

effects of the long wavelengths of light on the pineal body. In a study of photostimulation of the pineal body

of Rana temporaria, Dodt and Heerd (1962) noted that

light of longer wavelengths produced excitation of the

pineal nerve fibers.

Response to Wind.— -There are two effects of moving air; a cooling effect due to increased evapora­

tion, and the physical pressure exerted by the air move­ ment. Both of these appeared to influence the behavior of the bullfrogs.

The cooling effect resulted from the evaporation of moisture from the animals' body surfaces. The degree ox cooxinu xs ut^penoenu on the humxdxty and nmbxcnt axr

temperature. Consequently the daily humidity and air

temperature greatly influenced whatever effects the wind had.

The behavioral response to the cooling effect of

the wind was various postures assumed by the frog as it

sat on the bank. There were two extremes of posture that

the frogs would assume. One was a low, crouching or flattened attitude. In this position the ventral

surfaces were in close contact with the substrate. The 76 other was a high, stiff-legged, sitting position with the forelimbs fully extended. This results in the belly, chest and throat areas being completely elevated above the ground surface. The posture could vary between these extremes. During days when the humidity was relatively low for the pond area (65%) and the wind speed reached 13-18 m.p.h., the bullfrogs would assume the low, flattened position. When wind speeds reached

25-31 m.p.h., the bullfrogs left the banks and entered the pond. On calm days and days when the humidity was

85% or higher, the high sitting stance was assumed.

The second behavioral response was apparently the result of wind pressure. As mentioned previously, the frogs left the banks when the wind speed reached approximately 25-31 m.p.h. Once in the pond the frogs would avoid the strong gusts by either taking cover under overhanging grasses of the windward banks or by moving into the dense cattail growth. They would sub­ merge low in the water and usually face away from the wind.

An additional effect of wind pressure appeared to be on vocalization, for on nights when winds reached

13-18 m.p.h. vocalization ceased. This observation held for natural ponds as well as the experimental area.

On days when strong gusting winds were from the 77 north the bullfrogs were often seen on the north bank.

This was because the north side of the enclosure was the observation building wall, which provided an excellent wind break.

The overall effect of wind was that of suppress­ ing the normal activity of the bullfrogs. The degree of this suppression thus seemed to be related to the wind speed, and probably the relative humidity and air temperature. The two most pronounced effects are probably an increased cooling effect due to the more efficient evaporation of moisture from the body surface, and an actual physical pressure due to the wind. This latter effect seems to become significant only at wind speeds above 25 m.p.h.

Response to Humidity.— The moisture content of the air probably played an important role in temperature regulation by evaporation, but other than this under the natural conditions of the experimental area any effect of humidity was not detectable.

The relative humidity was continually monitored within the experimental area and outside the area in the adjacent field. These observations clearly illustrated the effect on the moisture content of the air produced by even a small body of water. The humidity was almost always higher in the experimental area than in the field; 78

Figure 19 shows that it averaged 10 to 14% higher in the

pond area. The values shown here represent the highest

relative humidity reached during the day.

The thermocouples which recorded the relative

humidity in the experimental area were located 5 inches

above the ground on the south bank. During the night

hours the humidity generally rose to 100% in the pond

area. If the high values maintained even during the

daylight hours on the banks are considered, the air at

about two inches above the water must have been

continually saturated.

Observations of the postures assumed by the bullfrogs under various conditions of temperature and

humidity agree with the very extensive and excellent

study of the behavioral regulation of temperature in

the bullfrog by Lillywhite (1970). He pointed out that

the bullfrog is capable of maintaining body temperatures

within rather narrow limits comparable to other ecto-

thermic vertebrates. Relative humidity must play a vital role in an organism whose moist skin is often

subjected to the intense heat of the summer sun. Noble

(1931) commented that frogs were better able to use

their digestive and nervous mechanisms on rainy days

than on dry, sunny ones. This greater efficiency might be reflected in the typical nocturnal hunting behavior 79

Fig. 19. — Comparison of the relative humidity of the experimental area with that of the adjacent field. £bily maximum) PERCENT RELATIVE HUMIDITY to C* o " CJ Ot to 08 15 JULY THRU 31 AUGUST 1969 F iq . 1 9 . 81 of the bullfrog.

Response to Severe Weather.— In this category are falling barometric pressure, thunder, lightning, and rain ranging from light sprinkles to heavy downpours.

One of the most obvious factors that is consist­ ently attendant with severe weather is falling barometric pressure. The effects of barometric pressure on the behavior of amphibians is unknown. During his study on

Acris crepitans, Jackson (1952) claimed that no corre­ lation could be found between barometric pressure and calling. Jackson unfortunately took his barometric readings 4 miles from his observation site. All of my barometric readings were taken at the experimental pond and consisted of at least 4-5 readings daily. Although

I was unable to correlate behavior patterns specifically to a changing barometric pressure I believe that baro­ metric pressure is involved in the animals' behavior.

Perhaps the problem is that each factor, such as reduced light intensity, humidity, barometric pressure, etc., is assumed to affect the animal in a specific way. Even if this is true what is more significant is how the animal responds to the complex of these factors. In reality the animal is never subjected to a single factor but rather to a continually varying complex of factors.

Consequently more accuracy is obtained by considering 82

the animals* response to all the factors which together

constitute what we loosely refer to as weather.

Bullfrogs continued to call during a gathering

storm seemingly without being affected by distant low-

rolling thunder or occasional flashes of lightning. The males would continue to call during light rain. If the

thunder increased with an oncoming storm, the vocaliza­

tions would cease. Just prior to the breaking of a

severe thunderstorm all activity in the experimental

pond area would cease. During very heavy rain the frogs

would leave the banks and submerge with half the tympanum beneath the water.

Loud sounds in the area could either stimulate

or inhibit calling activity depending on the quality of

the sound. I have often noticed how the sound of a

small propeller driven airplane would stimulate the bull­

frogs to call, while the deep roar of a jetliner would

totally suppress calling. This was observed at natural

ponds also. Beilis (1957) observed this same effect

during his study on Pseudacris.

A peculiar activity took place during times of

steady, moderately heavy rain. At these times both sexes

of bullfrogs became hyperactive. The animals would move

out of the pond and up to the edges of the enclosure.

They would move along the fence, periodically leaping up 83 against the fencing or actually climbing it— at times they reached six to eight inches off the ground before falling back. There was a continual probing and pushing of the snout into the meshes of the fencing, often resulting in minor injuries to these areas. The overall effect appeared to be a drive to move away from the pond area. As the rain would begin to lessen this activity also subsided. The frogs would finally turn away from the fence, facing back toward the pond. They would sit quietly for perhaps 20-30 minutes, then'finally return to the pond or to other usual patterns of activity.

It is extremely difficult to determine whether this activity is the result of a single stimulus or several. For example, I arranged a garden hose so that

I could direct a spray of water up and over the obser­ vation building, producing an artificial rain in the experimental area. This experiment was conducted on a clear, sunny day, and on a dull, overcast one during midsummer. In both experiments the frogs showed none of the typical hyperactivity associated with rain storms.

Naturally my experiments did not include a falling or low barometric pressure and, perhaps more important, the effects of lightning. Possibly under conditions of ionization coupled with lowered light levels and the tactile effect of rain these become a powerful stimulus inducing the peculiar behavior observed.

Apparently this hyperactivity is not restricted

to the bullfrog, for small green frogs (Rana clamitans) would often be found in the pond on mornings following

an all night rain. On several occasions during a steady rain, green frogs attempting to enter the enclosure were detected from the sounds of their repeated leaping against the fence. Frequently juvenile bullfrogs would enter the area under those same conditions. Those frogs which entered the pond area in this manner were naturally small individuals due to the small size of the fence mesh. Hyperactivity associated with rain appeared to be unrelated to any particular month during the active season.

Several workers have also observed this relation­ ship between activity and rainy weather. The earliest reference that I found was in the classic study, The

Frog Book (Dickerson, 1906). On page 229 Dickerson briefly described a chance observation of a bullfrog migrating during a long-continued heavy rain. 3ragg

(1940) noted that the Great Plains toad (3ufo coqnatus) appeared to breed only after rain. The breeding activities of the spadefoot toad (Scaphiopus holbrookii) also appeared to be initiated only by heavy rainfall

(Walker, 1946). 31anchard (1930) found that the spotted 85

salamander (Ambystoma maculatum) migrated to the breeding

ponds only during rainy nights. Hyperactivity during

rain but not associated with breeding activity was noted

by Martof (1953b) in the green frog and by Raney (1940)

in the bullfrog.

Hyperactivity in the bullfrog brought about

during rainstorms would provide an efficient dispersal

force. During heavy rains some bullfrogs would tend to

move away from the pond, reducing overpopulation. As

hyperactivity seems to increase after dark the migrating

frogs would be relatively safe from dessication as they

traveled through the rain soaked fields. To what degree

this activity would increase predation during such

migrations is open to speculation.

My observations indicate a greater sensitivity

of the female bullfrog to severe weather conditions. I

have repeatedly observed that females tended to be far more active during such periods than the males. This

observation was supported by the fact that almost all of

the bullfrogs and green frogs entering the experimental

area during rainstorms were females. These frogs were

also juveniles, but this was more likely the result of

the small mesh of the fence screening out the larger

individuals. 86

Hunting and Food-Seeking Behavior

Hunting or food-seeking activity of adult bull­

frogs appeared to be restricted to the dark hours, while

the juveniles hunted predominantly during the daytime.

The adults left the pond each morning and retreated to

the overgrown areas which abutted the pond. Possibly

some feeding may have taken place there, but if so, it

was probably more the result of an insect's coming with­

in the immediate reach of an individual than through

active hunting on the part of the frog. On several

occasions adults v/ere observed partially concealed by

the heavy growth. These individuals remained in a low,

crouching position during the day unless the sunlight

eventually filtered through to them when they would

withdraw deeper into the shade of the higher grasses

and weeds. The juveniles never appeared to enter the

overgrown areas even on extremely hot days. At one

time there were six immature bullfrogs in the pond, and

all of them were continually active throughout the

daylight hours. This same behavior of juveniles was

frequently observed at natural ponds. Young bullfrogs

spent the daylight hours hunting among the cattails and

occasionally hopping onto the banks after small insects.

This difference between the juvenile and the adult may

be the result of a greater efficiency of the juvenile to 87 thermoregulate its body temperature. Lillywhite (1970) found that the mean body temperature of adults during the daylight hours was 31.1° C., while that of the juveniles was 29.6° C. The advantage of this ability of the young frog to actively hunt throughout the day is obvious. As the evening approached the roles were reversed with the adults actively moving about the pond area in pursuit of food, while the juveniles became more quiescent. Although young individuals were observed at night they appeared to remain most of the time back in among the heavy cattail growth.

During the heat of the day the adult bullfrog prefers to sit quietly in a shaded spot. My impression of the hunting bullfrog was an alert, well-coordinated animal— quick to learn and act. The bullfrog appears to depend entirely on sight during hunting, and its ability to detect small moving objects is truly remarkable.

Lettvin et al. (1959) demonstrated light receptors in the frog's retina thay they called "bug detectors,” pointing out that the responsiveness of the frog to such stimuli was produced by peripheral rather than central filtering. Probably the bullfrog simply does not recognize a motionless insect as food. Von Uexkull

(1934) proposed this concept in reference to jackdaws hunting grasshoppers. He suggested that perhaps the 88

"death-feigning" reaction of insects is an adaptation which literally makes the insect invisible to its predator. Once the bullfrog recognizes its prey the response can be quite dramatic. Bullfrogs often leap out of the water to seize an insect, and leaps of 3 feet high are common (Davidson, 1963).

The bullfrog seems to be able to learn quickly.

During one experiment some June beetles (Phyllophaga sp.) were thrown into the pond from one of the obser­ vation building windows. After the frogs had consumed these a small dark brown wood chip was thrown into the pond. The chip was immediately seized and just as quickly rejected. Only two more such experiences were required before the frogs would no longer respond to this stimulus. The next batch of beetles was also refused until their movements in the water finally triggered a response in one animal. The others quickly followed suit and returned to feeding. This response of one frog to the activity of another was very common. It could be likened to a young chick finding a worm, and suddenly being pursued by a dozen other chicks, which had reacted to specific rr.ovements of the first chick rather than to the worm. I also observed what appeared to be anticipa­ tion during feeding sessions using kidney chunks. The third afternoon after two feeding sessions the frogs 89 gathered at the north bank opposite the feeding window at the time I had fed them on the previous two occasions.

This behavior was seen thereafter on numerous occasions.

Another example of anticipation was seen during periods when a flood lamp was used in the pond area to attract night-flying insects. This lamp was clamped to a short stake in the north bank and could be turned on or off from within the observation building. The reflector was angled upward at about a 60° angle. A cover made from galvanized window screening was fitted over the front of the reflector. This screening became hot enough to partially disable insects that struck it in their attrac­ tion to the light. These insects would then fall into the water below where the bullfrogs quickly learned to find them. In a matter of 2-3 evenings the frogs began sitting in the water beneath the lamp even before it was turned on. These sessions invariably resulted in numerous encounters between individuals, and the most aggressive animals soon began to have a front row seat.

This hierarchy could be seen in the arrangement of their positions during the anticipation of the coming hunting session. Bergeijk (1967) also observed this same reaction of anticipation in a laboratory colony of bullfrogs.

The bullfrog has a wide variety of foods ranging 90 from small insects to birds. My observations indicated that the predominant food source was insects, which confirms the work of Korschgen and Moyle (1955), who found that insects made up.32.5% of the total food con­ sumed by the bullfrog, and Frost (1935).

The bullfrog is a very impressive pond predator, particularly when one considers that an average bullfrog in nature^ can live to be 8-10 years old (Durham and

Bennett, 1953). At the time of this writing one of the males still in the experimental pond is 8-9 years old.

Oliver (1955) claimed that bullfrogs in captivity have survived 16 years. When the learning capacity that I have observed in this animal, coupled with its voracity, is considered, a bullfrog of 3-4 years hunting experience must represent a formidable predator.

The bullfrogs in the experimental area began feeding about 3-4 weeks after emergence and continued up to the first iveek of October. The adult males had ceased to hunt or to respond to offered food by late

August, while the adult females and juveniles were actively feeding into late September or early October.

During the breeding season the adult male ceases to feed. During this period I was unable to observe any hunting by the male, and offered food was ignored.

Thus the entire energy of the male is directed toward 91 territorial behavior with continual aggressive move­ ments and vocalization. Males were commonly observed during this time calling about once every minute.

The females continued to hunt and to respond to food to within a few days of ovulation. A day or two prior to ovulation the female would become more secretive and during the activity of artificial feeding would sub­ merge or leave the pond for the seclusion of the high grasses. Immediately after ovulation and fertilization both sexes showed a definite avoidance of feeding areas or offered food. Normal feeding began again late in

July and continued until about the first week in

October. A description of a feeding session on

October 7, 1959, has been given under the heading

Influence of Sex and Age.

The degree of responsiveness to food could be evaluated at this time by artificial feeding of kidney chunks. Feeding sessions usually took place in the early evening around 1800 hours. At this time the adults were generally out of the pond and in the deep grassy areas. Merely plopping the bait up and down a few times in the water was usually sufficient to bring the first bullfrog out of the weeds. The bullfrogs learned very quickly to associate the sound of water splashing with feeding. The splashing of the first bullfrog as it seized the meat attracted the others who quickly leaped or slid into the pond and swam toward the activity. I have observed time and again that the hunting activity of one bullfrog will stimulate another. On numerous occasions the movement of one frog would cause a second or even third animal to leave the bank or to turn suddenly and swim toward the activity. In several such instances these animals could not possibly have seen what had attracted the first frog.

Influence of Sex and Age

As fall approached the frogs tended to leave the pond later in the morning and returned earlier in the evening. They gradually began spending less and less time in the overgrown areas, basking in the weakening sunlight durinq the day. By early October the bullfrogs were entering the pond at around 1800 hours E.S.T. At sunset as the air temperatures began to fall rapidly, the animals would retreat deep into the cattails and out of the wind. Throughout the evening they would submerge deeper and deeper with the water reaching more than halfway up on the tympanum. When the surface water temperature was about 6° C. the bullfrogs would typically submerge until about half of their tympanum was exposed.

The body was sunk deep into the water, while the hind legs appeared to be totally extended, relaxed, and 93 hanging almost straight down. The frogs generally retreated to the bottom when the surface temperature of the water approached 5° C. Age and sex may affect temperature tolerance and in addition temperature tolerance may vary among individuals.

During the chill evenings of fall the bullfrogs tended to enter the pond earlier, slowly submerging deeper and deeper as the evening grew colder. Although no actual data were taken the amount of body surface exposed could be correlated to the air temperature. As the animal rested at the surface of the water the exposed parts would be affected not by the air temperature at

5 feet, but by the temperature of the air layer at 1-2 inches above the water surface. This air layer was kept at a higher temperature by the slow loss of heat from the water surface.

Figure 20 clearly shows this buffering effect of the warm surface water and gives the maximum air and water temperatures for October 8, 1968. This was a sunny autumn day with no wind, the air temperature rising to 19° C. and falling to a chilly 4° C. The temperatures represented in the graph are for the air temperature at 5 feet, the air temperature 1 inch above the water, and the water temperature at a depth of

2 inches. 94

Fig. 20. — Various temperatures of the experimental area for a typical fall day, (October 8, 1968) o to o DEGREES CENTIGRADE O l

HOURS E.S.I

Fig. 20, S6 96

Several immature bullfrogs of both sexes seemed more tolerant to low temperatures than the adults.

Although I cannot say for certain that the juveniles

emerged before the adults they definitely remained

active later in the season. During the cold windy days

of spring or late fall the juveniles were active in the

water when the adults were either deeply submerged and motionless or nowhere to be seen. On several occasions

during late October trips to the ponds at Delaware,

Ohio, great numbers of juvenile bullfrogs were observed

along the pond edges although few adults were seen.

Throughout the study adult females seemed more

tolerant to low temperatures than the adult males. This

was the case in both the experimental pond and the

natural ponds at the Delaware Wildlife Area and Blendon

VJoods Park. On October 14, 1968, during a field trip to

the Delaware ponds, the preponderance of females was

striking. Judged from their size, these were young

females still not fully grown. The proportion of females

to males was estimated as 10 to 1. This same trend could be seen in the experimental area. Bragg (1940) has also

noted the apparent greater sensitivity of the male toad,

Bufo coqnatus, to low temperatures.

Another indication of the temperature tolerance

of female bullfrogs was observed during feeding sessions 97

in the fall. During the autumn of 1968 the latest date

that I was able to obtain a feeding response occurred on

October 7. On this evening two females were on the bank and one was in the water along with a single male. By repeatedly dangling the bait in front of the male, I was finally able to obtain a response: he made a slow un­ coordinated movement toward the food with a partly opened mouth. Having missed the food, the animal submerged and was not seen for the rest of the feeding session. Upon the presentation of food at the center of the pond, the two females on the bank dived into the water and quickly swam toward the bait. The first lunge that each female made in seizing the food was quite strong and well co­ ordinated. Each subsequent lunge became weaker and less accurate. By the time each female had succeeded in taking 2 to 3 pieces of meat, their lunges had become weak bobbing movements made with mouths closed. The air temperature at 5 feet was 10° C., the air temperature at

1 inch above the water surface was 11.1° C., and the water temperature at a depth of 2 inches was 12.2° C. On

October 26, 1968, a feeding session was attempted but neither males nor females would respond in any way to the food presented to them. The air temperature at this time at 1 inch above the water surface was 8.4° C., and the water temperature 2 inches deep was 10.5° C. 98

These results are comparable with observations of bullfrogs in natural ponds. By late September bullfrogs did not appear to be feeding although both sexes were

seen on the banks and along the pond edges.

Another measure of temperature tolerance was demonstrated by the ability of the animal to engage in aggressive behavior. As fall approached the males began

to show a decided loss of aggressiveness, while the females remained highly aggressive. Females would actively leap at or on territorial intruders, and produce the low growl vocalization. In contrast to this the males seemed to lose all their aggressiveness. In fact, a behavioral change took place where the males seemed to become tolerant of all other bullfrogs— both male and female. This not only appeared to be a tolerance, but an actual attraction to other bullfrogs. Of course, v/henever a male approached a female, he was quickly driven away by the aggressive female. The result was that I would often see 2-3 males sitting quietly

together.

There are some possibilities as to the adaptive

significance of low temperature tolerance in the adult female and juvenile bullfrog. In the case of the juveniles, the indications are that they emerge from hibernation first and quickly become active. This 99 probably has at least something to do with the greater ease of warming in smaller individuals due to the surface/volume relationship. The juvenile is active during the early spring, the daylight hours of summer and on into late fall. This observation was also made by

Jenssen and Klimstra (1966) on the green frog and by

Lillywhite (1970) on the bullfrog. All of these times are such as to allow the juvenile a considerable amount of time to feed without interruption or competition from adults. I was unable to observe any aggressive behavior between juveniles. This would also permit greater free­ dom of movement for immature bullfrogs in the pond area.

In the case of the adult female the low temperature tolerance would be of great value from three standpoints.

First it would insure that the female would be free from competition in the late fall from adult males. This in turn would insure a longer feeding time in the fall to build up fat stores which would probably have considerable survival value. Lastly, as the adult female remains highly aggressive into late fall, this would provide for significant competition between females, which might result in the survival of the more aggressive females. 100

Vocal Behavior

Types of Vocalizations Produced

The various sounds produced by the bullfrogs were recorded as often as 'possible during this study.

At each recording session careful observations were made

so that both the vocalizations and the accompanying

activity were recorded. In this manner a series of vocalization types was recognized. Some of these

seemed distinctly associated with specific activities, while others occurred during a number of activities.

Sounds were recorded from both sexes and from adult and

juvenile animals.

I have given these ten vocalization types the

following descriptive names: (1) an explosive "pop,"

(2) a low "growl," (3) a soft "burping" sound. (4) a

female release call, (5) a male clasping call, (6) a male non-pulsatile territorial call, (7) a male pulsatile territorial call, (8) a female territorial

call, (9) an alarm call or "squawk," and (10) a

juvenile "cluck." Table 1 is a compilation of the most

pertinent information from these vocalizations.

The Explosive Pop

This sound is produced by both adult males and

females, although males produce it much more frequently.

It can be described as a loud, explosive "sputt" or 101 TA3LE 1

VOCALIZATIONS PRODUCED BY THE BULLFROG

Vocalization Adult Adult Juvenile Time Distance Male Female Period Between Calling Frogs

Pop + + --- April- 2-4 feet Sept.

Growl + + --- May- 1 foot Sept. or less

Burp + ”7 --- Angust- 1 foot Sept. or less

Female —— + ——— June- clasping Release Julv Call

Male + — —— _—— June- clasping Clasoing July Call*

Male Mon- + May- 20-25 Pulsatile August feet Territorial Call

Male + June- 25-50 Pul sat.ile July feet Territorial Call

Female --- + --- A.ugus t- 1-3 Territorial Sept. feet Call

Alarm + + + April- --- Call Sept.

Juvenile ------+ April- --- Cluck Sept. 102

TA3LE 1— Continued.

Vocalization Fundamental Duration Motes Function c.p.s. in per Seconds Call

Pop 200 0.05 aggression

Growl 200-300 1.3-6.0 many aggression

Burp 250 0.02 agaression (?)

Female 300 0.4 non- Release receptive Call female (?)

Male 150 0.6 (?) Clasping Call

Male Non- 100-150 0.65 7-10 maintenance Pulsatile of Territorial territory Call

Male 2 00 1.0 5-7 maintenance Pulsatile of Territorial territory, Call mating (?)

Female 500 0 .6-1.0 aggression, Territorial territory Call

Alarm 550 0.3 warning Call signal (?)

Juvenile 500 0.05 (?) Cluck 103

"pop" of brief- duration. I have carefully observed the conditions under which it was produced in both the experimental area and at various natural ponds, and found it to be always associated with aggressive behavior. This vocalization was typically heard during encounters involving territory and competition for food.

A typical situation which regularly produced this type of vocalization is described under the heading Playback and Model Presentation on pages 148-154.

Wiewandt (1959) referred to a "bonk," which, from his description, could be the same vocalization as the pop I describe. Em.len (1968) described a vocaliza­ tion that he called a "hiccup," which again may be this same pop that I have observed.

Sonagrams of typical pops show that this vocal­ ization has a very brief duration of approximately

0.05 second. The frequencies of this call generally reach 1400 cycles per second (cps), and may go somewhat higher in some individuals. The fundamental appears to be about 200 cps. Examples of pops are shown in Figures

21 thru 23. Figure 21 represents an adult female call on July 5, 1970. Figures 22 and 23 are both from adult males. Figure 22 was from August 4, 1959, and Figure 23 from August 14, 1969. Aside from slight differences due to anatomical variations in individuals, this 104

Fig. 21. — Juvenile cluck followed by adult pop

Fig. 22. — Adult pop followed by two non-pulsatile territorial call notes

Fig. 23. — Pulsatile territorial call note followed by adult pop 105 FREQUENCY IN KILOCYCLES PER SECOND

TIME IN SECONDS Fig. 23. 106

vocalization seems to be the same at all times of the

active season. The sound appears to be quite similar

whether produced by male or female, although there is a

tendency for it to be louder when produced by a male.

The Growl

This is a vocalization associated strictly with

aggressive behavior. It is produced by both sexes, and

to my knowledge is restricted to adults only. Both the

male and female calls are similar except that the male

call tends to be louder and can have a longer duration.

The growl is actually a very low intensity call. I have

sat along the banks of many ponds for hours, and have

rarely heard this call under normal hearing conditions.

However, on several occasions while tape recording using

a 24-inch parabolic reflector and monitoring the record­

ing v/ith earphones, I picked up bullfrog growls during

interactions. In the experimental area it was a fairly

common vocalization, occurring much more frequently at

night than during the daylight hours. This vocalization

is produced only when the frogs are quite close to each

other. The usual distance observed when this call was being produced was about 6-3 inches. If this is always

the case then the low intensity would certainly not

detract from the effectiveness of the call.

From my observations the duration of this 107 vocalization varied from 1.3 to 6.0 seconds. The sonagrams in Figures 24 thru 26 are all male growls and were made during aggressive encounters involving competition for food. The three growls were produced by different males on three different dates. In Figure 24 the growl is followed in 2 seconds by a loud pop. In all of these growls the pulse rate is 40 per second.

The Alarm Call or Squawk

This vocalization is produced when a bullfrog is frightened and leaps from the bank into the water. Frogs often leap into a pond or stream without producing this vocalization, so it may be assumed that the alarm call is not a sound produced involuntarily as the result of muscular contractions associated with leaping. This vocalization is produced by both sexes and by juveniles as well as adults. The sonagram in Figure 2 7 is of an alarm call or squawk produced by an adult female. In this sonagram the fundamental is 550 cps and the dominant frequency appears at 1200 cps. This call is quite brief, averaging about 0.3 second in duration.

The Burp

This is a very low, soft sound. I have been able to detect it only during aggressive encounters, and I have heard it produced only by males. This call is so quiet 108

Fig. 24. — Adult male growl followed by male pop

Fig. 25. — Adult male growl

Fig. 26. — Adult male growl FREQUENCY IN KILOCYCLES PER SECOND

io » b O b .u . O' 00 o 10 u - _ 1_ I I _A_ I 1 infill. I ■ — *■

*} H- H- H- tQ iQ iQ • m ro M NJ z cn • c/i m on Z C7 C/1

M O VO 110

Fig. 27. — Adult alarm call or squawk

Fig. 28. — Two non-pulsatile call notes followed by four burps ending in a pop.

Fig. 29. — Two juvenile clucks Ill I I* ll.fcl IO IO X O' 0D O H* • w o id X. &e*- M x o> 00 o X . . . U f, , . l ...... I FREQUENCY IN KILOCYCLES PER SECOND H- !V> 03 iO U' TIME IN SECONDS 112

that I doubt it would be heard by an observer at a large natural pond. The burp usually preceded or immediately

followed the pop, although I have heard it following a

territorial call. Figure 28 was a particularly fortunate recording as this sound is difficult to record and quite unpredictable in occurrence. Of all the bullfrog vocal­

izations the burp has the briefest duration, lasting for only about 0.02 second. The fundamental is approximately

250 cps with the highest harmonics reaching about 900 cps. The call note interval is about 0.8 second. The

sonagram in Figure 28 begins with two non-pulsatile call notes followed by four burps and ends in a pop.

The Cluck

This vocalization appears to be produced only by juvenile animals. t have not been able to determine if both sexes or only a particular sex produces this call.

The call does not appear to be associated with aggressive behavior. It was most frequently heard as juvenile animals would be moving about during the day.

The cluck is somewhat like the pop but much softer and mellower. These two calls can be compared in Figures 28 and 29. The cluck has two rather distinct areas of intensity while in a loud pop the harmonics are less distinct. Calls Produced During Clasping

The Female Release Call.— According to Capranica

(1968) the release call is made by both sexes; however,

I was able to observe this only with the females in my experimental area. As the name implies this call is produced apparently with the result that a clasping male will release the individual producing the sound.

This vocalization is a rather soft mellow sound, which can be crudely simulated by saying "urk, urk, urk, urk." The fundamental is approximately 300 cps. In

Figures 30 and 31 there are obviously slight differences between these two examples. The vocalizations in these two sonagrams were made by different females while being clasped. The difference in each sonagram could be a reflection of the anatomical differences of these two individuals. It could also be the result of the female's being clasped and struggling while producing the vocal­ ization. However the distinctive patterns of the individual are apparent and can readily be seen by com­ paring Figures 30 and 31 with Figures 33-35. All of the vocalizations in Figures 3 3-35 were produced by the same frog whose calls are shown in Figure 31.

The Male Growl.— This vocalization, which has already been discussed, was heard quite commonly during the clasping of females by males. Usually the growl 114

Fig. 30. — Female release calls

Fig. 31. — Female release calls

Fig. 32. — Male clasping calls 115 FREQUENCY IN KILOCYCLES PER SECOND

TIME IN SECONDS Pig® 32® 116

Fig. 33. — Male growl followed by 3 female release calls.

Fig. 34. — Four female release calls followed by male growl.

Fig. 35. — Male growl followed by 2 female release calls. FREQUENCY IN KILOCYCLES PER SECOND 118 would occur simultaneously with the female release call.

This growl was quite similar to growls produced in other aggressive interactions except that the pulses tended to be less consistent. In the territorial growl the pulses were consistently 40-42 per second. In the three growls of Figures 33-35, the pulse rates vary from 38-46 per second. These differences may be due to the physical pressures occurring during clasping.

The Male Clasping Call.— This vocalization was heard only a few times and recorded even fewer. It was a very distinctive sounding call, quite different from any other produced by the male bullfrog. A sonagram of this call can be seen in Figure 32. The fundamental is approximately 150 cps with the dominant frequency being

250 cps. Each call note had a duration of about 0.6 second with a call note interval of from 0.3 to 0.7 second. None of the call notes showed major pulses.

The Male Pulsatile Call— Territorial Call or Mating Call?

In Figures 36-47 sonagrams of twelve calls are presented. Figures 36-38 and 42-44 were produced by males whose behavior, such as vocalization and mainten­ ance of territory, presented the picture of typical breeding males. These vocalizations were all produced between June 10 and July 28 and represent what has been called the mating call. 119

Fig. 36. — Mating call, June 10, 1969

Fig. 37. — Mating call, June 18, 1968

Fig. 38. — Mating Call, June 21, 1969

Fig. 39. — Territorial call, August 11, 1969

Fig. 40. — Territorial call, August 14, 1969

Fig. 41. — Territorial call, August 19, 1969 FREQUENCY IN KILOCYCLES PER SECOND 1.4 1 1 . . . S-^ 4' 0 6 6 8- 0 6 2- 4 2- ' ' - - - * i i. 38. Fig. i® 36.Fig® i« 37,Fig« T — ' ng in t a m

s l l a c -if? x >irf ' K- •rV ", tr T~“ ~ ‘T m-. m T 6 •w -■> i IE N SECONDS IN TIME 14- t 1 .4- . . . j . &• 6 8 O- 2 0 4- - - c«1 - - •wwi . f f i ?f -ii : ,i ■ ■ - : V 'f I f , , >S-lJ4 K Fig. g. 41 . ig F Fig., 1 H'ill -ii* TERRITORIALCALLS ' ?■■■ ' ■ 43 i 1 1 * 40. 39. | I J * , . Hf HM m .1 : « ! ! : ' 1 ro i-* o 121

Fig. 42. — Mating call, July 6, 1970

Fig. 43. — Mating call, July 21, 1967

Fig. 44. — Mating call, July 28, 1955

Fig. 45. — Territorial call, August 26, 1969

Fig. 46. — Territorial call, August 28, 1969

Fig. 47. — Territorial call, September 1, 196 FREQUENCY IN KILOCYCLES PER SECOND 1.4- 1 1.0- 1.4- . 8 0 - 8 - 6 4-i 2- - - lifiS ; i i ; 4/' .4 > $ ' ng in t a m Fig c 42 i. 43. Fig. i. 44. Fig.

s l l a c ' IE N SECONDS IN TIME ERTRA CALLS TERRITORIAL i. 45,Fig. Figo i. 47. Fig. :! 46, r r 6 S — t

! 122 123

The other six sonagrams seen in Figures 39-41

and 45-47 are from vocalizations ranging from August 11

to September 1. I have called these vocalizations

territorial calls as they occurred long after mating.

These vocalizations all possess the highly

pulsatile "stuttering” sound which has often been

described phonetically as "jug-o-rum." The pulsatile

call is the loudest of all the calls produced by the

bullfrog and can be heard for considerable distances.

I often found that these calls were clearly audible for

a distance of one-quarter mile.

The pulsatile call heard during June and July

has generally been considered a mating call. However,

this same type of vocalization can be heard during late

summer. An examination of Figures 36 thru 4 7 will verify the fact that, aside from variations represent­

ing individual anatomical differences, these vocaliza­

tions appear to be very similar. Schiotz (1971),

Wiewandt (1969), Capranica (1963), and Bogert (1960) all

suggested the possibility that the bullfrog may have a

single vocalization that serves the function of both a mating call and a territorial call. I am in complete

agreement with this.

In Tables 2 and 3 I have compared the character­

istics of 100 pulsatile calls. Fifty of these were made TABLE 2

PARAMETER ESTIMATES FCR THE "MATING CALL"

Characteristic Sample Rangc- Std. Std. 95% Confidence Mean Dev. Error Interval 00 CM 0"

Fundamental 250.0 19.63 ' • 203.0-214.2 208.6 198.0- \ (c.p.s .) o O (Tv o OJ ■—i l . Duration 0.96 0.7- 1.35 0.18 0.03 • (sec.)

Major Pulse Rate 6.41 4.0- 9.0 0.97 0.14 6.13- 6.69 (per sec.)

Call Note Interval 0.47 0.4- 0.55 0.20 0.14 0.20- 0.74 (sec.)

n-50 vocalizations made during June and July. 124 TABLE 3

PARAMETER ESTIMATES FOR THE "TERRITORIAL CALL"

Characteristic Sample Range Std. Std. 95% Confidence Mean Dev. Error Interval

Fundamental 216.2 197.0-300 26.23 3.71 208.8-223.6 (c. p. s. ) \

Duration 0.90 0.70- 1.20 0.11 0.02 0.86- 0.94 (sec.)

Major Pulse Rate 6.80 5.0- 8.8 0.97 0.14 6.52- 7.08 (per sec.)

Call Note Interval 0.43 0.40- 0.50 0.04 0.006 0.42- 0.44 (sec.)

n=50 vocalizations made during August and September. 126

during the mating season and fifty were made in late

summer. The 95% confidence intervals for the character­

istics of the vocalizations made during and after mating

indicate that these are the same call.

Probably when the female bullfrog’s blood- hormone concentration reaches a critical level, she will

respond to the pulsatile call and mating will occur.

After ovulation the female's blood-hormone level falls,

and this same vocalization functions solely as an

advertisement of territory.

As a consequence of this clearer understanding of the pulsatile call I will refer to it from now on

as the male pulsatile territorial call.

The Male Non-Pulsatile Territorial Call

The non-oulsatile territorial call is heard as regularly as the pulsatile form— during the spring,

summer and into the fall. The noticeable difference in this call from the pulsatile form is that it lacks the energy of the pulsatile territorial call. The funda­ mental and dominant frequencies are often somewhat lower, the call notes tend to be of shorter duration, and the interval between call notes tends to be a little longer. In the examples shown in Figures 48-50 the fundamental ranges from 100 to 150 cps, and the dominant frequencies vary from 225 to 275 cps. The call notes 127

Fig. 48. — Non-pulsatile territorial call

Fig. 49. — Mon-pulsatile territorial call

Fig. 50. — Non-pulsatile territorial call FREQUENCY IN KILOCYCLES PER SECOND 1 1 . . . .4« . .8* 0 6 4 2 * * - *

■ — t' r "r J - 1 1 i — i. 50. Fig. SECONDS IN TIME i. 49. Fig. i. 48. Fig. 3 ' ■£**- 0 4 '

' 5 .-ft; 128 129

shown here have an average duration of about 0.65 second, with the intervals averaging about 0.5 second.

Although this vocalization sounds quite loud to

the ear it lacks the greater duration and the unique vibratory quality of the pulsatile territorial call.

The non-pulsatile territorial call rarely extends for more than 7-10 call notes, but I have heard pulsatile

territorial calls with as many as 18 call notes.

The Female Territorial Call

In addition to the growl and pop the female bull­ frog produced a vocalization strictly associated with territorial activity. This same vocalization was des­ cribed by Capranica (1968), and consequently I have used his terminology and called this vocalization the female territorial call.

I found this vocalisation to consist of a single short growl ending in a brief call note with an increased inflection. The sonagrams in Figures 51-53 show a duration of 0.6-1.0 second, although I have heard this call when it lasted for about 1.5 seconds. The intro­ ductory growl has a pulse rate of 34 per second. The fundamental is about 500 cps and the harmonics appear to reach at least 1500 cps.

This vocalization was definitely associated with aggressive behavior, and almost always accompanied 130

Fig. 51. — Female territorial call

Fig. 52. — Female territorial call

Fig. 53. — Female territorial call 131 FREQUENCY IN KILOCYCLES PER SECOND

TIME IN SECONDS F ig , 5 3 , 132 by the female leaping on the intruder.

The Effects of Temperature on Vocalization

In the area of Columbus, Ohio, bullfrog vocal­ ization usually begins in-early April and extends into late August. In the experimental area calling persisted into late September, but this was an abnormal situation as calling is generally ended by mid-August in natural ponds. This unusually long calling period in the experimental area was probably brought about by arti­ ficial feeding.

The fact that temperature affects the physio­ logical processes, particularly those of amphibians, has been established for some time. In his classic work, The Biology of 'the Amphibia, Noble (1931, p. 431) pointed out that the amphibian is literally the slave of its surroundings. A considerable amount of work has been done on the effects of temperature on the calling rate of the genera Bufo and Pseudacris by Bragg (1950,

1948, 1943, 1942, 1940) and Jackson (1952). Both of these workers found very definite relationships between the call rate of the amphibians studied and the ambient temperature. Beilis (1957), working with Pseudacris and Bufo, also found highly significant correlations between call frequency, call duration and ambient temperature. Therefore, the bullfrogs I observed could 133 be expected to show a definite correlation between the quality of their vocalizations and the ambient temperature.

During the cooler- nights of early fall, I observed that territorial calling was more easily dis­ turbed than would have been the case a month earlier.

Consequently I was able to observe that when calling had been suppressed by a local disturbance the character of the call was modified when the next vocalization occurred. This effect was most noticeable in the duration of the call. For example, a calling pattern of

6-note vocalizations, when disturbed, would result in the next vocalization having perhaps 12-15 call notes.

After a time the original call duration would be re­ established. The lower the ambient temperature the more pronounced this effect became. Of course as the fall progressed the evening temperatures became sufficiently cool to reduce calling at first to the early evening hours; finally it ceased entirely. The latest vocalization was heard September 21, 1969.

I believe that what I observed in the increased call duration following suppression of vocalization is what Sherrington (1906) called "Reflex Rebound." He found that once inhibition was removed from a reflex it returned at a higher intensity than had occurred 134

previously.

The effects of the air and water temperature on vocalization become most evident during the late summer.

In central Ohio the bullfrog continues to call until the

latter part of August or early September. Based on observations made during 1968 and 1969 the minimum day­

time air temperatures for this period, at both the experimental pond and the Delaware Wildlife Area, aver­ aged about 13-17° C. However, nighttime temperatures commonly fell to between 6 and 7° C.

At this time of year calling typically began near sunset with each vocalization averaging 5-6 strong call notes(see Materials and Methods, pp. 45-50 for discussion of terms). As the evening progressed and the air and water temperatures fell, the frequency of calling gradually lessened. Vocalizations began to have fewer call notes and the intensity or loudness of each vocalization was greatly reduced. On chilly nights when the air temperature had fallen to around

7° C. the vocalizations were often a single call note v/hich could best be described as a low groan.

The Call Energy Index (CEI)

During this study a total of 4854 pulsatile and non-pulsatile territorial vocalizations was recorded and tabulated as to the numbers of call notes in each 135 vocalization. For each 14 hours of monitoring, a standard tabular form was used and the frequencies of one-note calls, two-note calls, etc., were recorded.

Each of these categories could then be totaled for any given period and then calculated as a percentage of the total number of vocalizations occurring in that period.

This same procedure was used for several natural ponds at the Delaware Wildlife Area. In the case of natural ponds a total of 1374 vocalizations was evaluated. The smaller number here as compared to that of the experi­ mental area was due to the difference of working in one's backyard as opposed to driving 40 miles to Delaware.

To present this information graphically, two time periods were chosen, one to include the approximate breeding time of the bullfrog, and the other represent­ ing vocalizations of the fall season well after breeding had ended. June and July were chosen for the first period, while August represented the fall season. The results of these observations are given in Figure 54.

The breeding period of June-July in the natural ponds showed a predominance of either 1 call note or 5-5 call note vocalizations. In the calling at natural ponds in August, 1 note vocalization represented about 51% of all the vocalizations produced, while the 5-5 note calls fell from 21-22% to 8-10%. The interesting finding is 136

Fig. 54. — Comparison of call notes per vocalization for natural and experimental ponds. PERCENT OF ALL TERRITORIAL VOCALIZATIONS AL OE PERCALL NOTES VOCALIZATION 54. 4 5 . g i F une-j y eprmetlpond). d n o p ental (experim ly ju - e n ju august ju n e - ju ly (n a tu ra l p o n d ) ) d n o p l ra tu a (n ly ju - e n ju V

nat al ) d n o p l ra tu a (n 137 138 that the pattern of vocalizations in the experimental area for June-July was almost the same as that for the natural ponds in August. I believe that this indicates that the calling in the experimental area was more of the pattern involving territorial defense rather than mating.

Using this technique of evaluating vocalization patterns by simply counting the number of vocalizations was not a very reliable measure. What was needed was a more accurate reflection of the amount of energy a particular bullfrog was utilizing in the social inter­ actions during the given time period. This would more precisely reflect the intensity of any bullfrog's involvement. A count of the total number of call notes per given time period is a better representation of this intensity of interaction. I have called these values the "Call Energy Index." This approach still lacks accuracy as it does not include such vocalizations as pops or growls, but rather only pulsatile and non- pulsatile territorial calls.

In mid-April of 1970 the males in the experi­ mental pond began calling. After a week or two all the males except one had become practically silent. The one which became the dominant animal in the area continued to call throughout the season and into late 139

August. Of course throughout the season the other males did produce growls and pops, but rarely territorial calls, and these were always pulsatile and consisted of

1-3 call notes of weak intensity. Consequently I was able to accurately follow the calling pattern of one dominant male. This calling pattern is given in

Figure 55. Two very interesting results were observed.

The first was that the CEI (Call Energy Index) rose sharply prior to the first ovulation and then dropped just as sharply. However, in a few days it once again rose very rapidly and again fell precipitously immedia­ tely after the second ovulation. This male was observed in amplexus with both of the females that ovulated, showing that a male bullfrog will fertilize more than one female during a season. In comparison, the CEI for the previous year— 1969— had peaked practically overnight in the same manner as 1970, although ovulation did not take place in 1969.

My observations from both the experimental pond and the natural ponds indicate that the calling of a male bullfrog will either stimulate or suppress the calling of other males in the immediate area. Such induction of vocalization has been observed by others.

Wiewandt (1969) and Capranica (1955) described the induction of calling in bullfrogs by playback techniques. Fig. 55. — Graph of Call Energy pattern April 26— August 14, 1970. CALL ENERGY" INDEX 0.0 0.5- 1.5—, 10 26 A JN JL AUGUST JULY JUNE MAY PI TR 1 AGUT 1970 UST AUG 14 THRU APRIL 6 2 55. 5 5 . g i F

14 141 142

Jenssen and Preston (1968) showed that the green frog would respond to playback of its recorded call. The calling of the spring peeper (Hyla crucifer) was shown to be definitely stimulated by other calling males

(Jones and Brattstrom, 1961). The genus Pseudacris was also shown by Beilis (1957) to respond to its own voice.

I found that whether the calling bullfrog would stimulate or suppress other bullfrogs depended on the intensity of its vocalization and its proximity to other bullfrogs. In his study of the chorus structure of the

Pacific tree frog (Hyla regilla), Foster (1967) iden­ tified these same two requirements. He also noted that without such stimulation calling was infrequent and irregular. A comparison of the CEI from the experimental area and from natural ponds appears to corroborate this.

During the study at the experimental pond I found that the maximum CEI value reached in late June was approx­ imately 1.5. In a large natural pond where two males were calling on July 11, 1965, the CEI was 2.23 for each of the animals. More significant were the values from a second smaller natural pond measuring roughly 75 feet by 50 feet. In this pond on July 28, 1965, there were

8 males— which I designated "A" through "H"— calling.

Analysis of the monitoring tape showed the CEI values to be: "D"— 7.45, "B" — 5.74, "A"— 4.46, "C"— 4.17, 143

"E"— 2.54, "F"— 2.06, "G"— 2.06, and "H"— 0.14. Except

for bullfrog "H" these values are all dramatically

higher than those observed for the single male vocaliz­

ing in the experimental pond. This clearly indicates

the effect of calling males hearing other males. I

believe this information also shows a social hierarchy

an the chorusing of bullfrogs. Bullfrogs "D" and "B"

were very loud callers, while vocalizations from "F,"

"G," and "H" were much weaker.

A relationship existed between feeding sessions

and the rate of vocalization. After every feeding

session the amount of vocalization sharply increased.

The most reasonable explanation for this would be that

during these feeding times the bullfrogs were forced

into a competitive situation due to the way the food

was presented to them. Manning (1957) referred to the

work of Hinde (1954) and Lorenz (1950). Hinde found in

his studies with the chaffinch "mobbing” response that

the stronger the chaffinch was calling at the time, the

longer it continued calling when the original stimulus

was removed. Lorenz referred to this as "reaction momentum," which implies that time is required for the

animal to slew down after it is once aroused. The

neuro-physiologist refers to this phenomenon as "after­

discharge . ” 144

Although I was unable to obtain a quantitative measure, the bullfrogs in the experimental pond seemed to produce vocalizations of lower intensity or loudness than those in natural ponds. The pulsatile territorial call can easily be heard from 300 to 500 yards from its source and even farther. However, I could not detect such vocalizations from my area at distances between

100-150 yards. One possibility is that the intensity is related to the proximity of other bullfrogs. In the experimental pond calling bullfrogs were never more than 10 feet apart, and often were only 3-4 feet apart.

Bullfrogs observed in natural ponds at the Delaware

Wildlife Area were calling from 25-50 feet apart. This possibility of a distance-intensity relationship in vocalization may be reflected in another way. If the four most commonly heard vocalisations were arranged in order of their intensity, the loudest vocalization would be the pulsatile territorial call followed by the non-pulsatile territorial call, pop and growl. With each of these vocalizations the distance between the frogs involved becomes less and less. For example, under optimal conditions of a large natural pond, the bullfrogs would probably be 25 to 50 feet apart during pulsatile territorial calling. This distance would be reduced to about 20-25 feet for the non-pulsatile 145

territorial call. A bullfrog producing a pop might be

anywhere from 2-4 feet from another bullfrog, while the

growl was never heard from individuals more than a foot

from each other, and usually they were a matter of only

inches apart.

Territorial Behavior

Maintenance of Territory

Effects of Vegetation.— During the spring of

1959, clipping of the grass along certain areas of the banks was begun as soon as the grass was about 3 inches

high. The untended areas were permitted to slowly grow

into tall weeds and grasses. Soon the frogs began to

take up various positions along the banks. These positions were well established by early May. Once

these territories had become established the clipped

areas were permitted to become overgrown. Throughout

this period of about 4 weeks the behavior of the frogs was carefully observed. The frogs began to shift their

territories and this tendency continued throughout the

4-week period. By the end of the 4 weeks the normally clipped area had become so deep in grasses and weeds

that observation of the frogs had become difficult.

Consequently, the experiment was terminated and these

areas once again were trimmed and clipped to their 146 original appearance. By the following morning, the frogs had taken up their original territories exactly as they were when the observations were begun. In some instances this reorientation was so exact that certain frogs returned to precisely the same spot on a bank.

The fact that a bullfrog could recognize a particular area in its environment was strikingly apparent. This observation was also made by Oldham

(1967) with Rana clamitans, and by Dole (1968) with Rana pipiens. How the bullfrog can recognize an area is still not completely understood. Olfaction possibly plays a role. As long ago as 1914, Risser showed that amphibians did have an olfactory sense. Martof (1962) demonstrated that Bufo and Pseudacris had definite olfactory ability. Oldham (1967) was able to show that olfaction probably played a significant role in orien­ tation of Rana clamitans. However, my observations indicate that the bullfrog probably utilized visual clues to a great extent. Not only could the bullfrog recognize its area, but apparently could remember it for a con­ siderable period of time. How long this memory could persist is unknown. In a natural pond the environmental appearance is continually changing as the season wears on. However, there would be certain landmarks such as rocks, stumps, logs, etc., that would remain constant at least throughout a season. The significance of this changing environment was that it seemed to influence

the territorial boundaries of the bullfrogs. As favor­ ite spots on the banks became overgrown the frogs would no longer occupy them and these sites therefore shifted.

This continual mobility of the territorial boundaries of the bullfrog was also observed by Wiewandt (1969).

Vocalization and Fighting.— Vocalization played a definite role in the maintenance of territory. This was very evident during June and July when breeding was taking place. The breeding males could be found at least 25-30 feet out into the large natural ponds.

These males produced pulsatile and non-pulsatile terri­ torial calls at regular intervals averaging about once every minute. Intruders were immediately challenged with loud non-pulsatile calls. If this did not suffice to drive the offender away the defending bullfrog would slowly swim toward the other bullfrog. At this time the defending animal's lungs would be inflated and he would be floating in the "high posture" (Playback and Model

Presentation Experiments, p. 148). When the distance between the two bullfrogs v/as reduced to 2-4 feet, the defender would produce 1-2 loud pops.

Bullfrogs were often seen fighting. During thes encounters the animals would seize each other in a face 148 to face embrace, accompanied by violent wrestling, butting, and shoving. At these times their struggling was very Similar to a swimmer treading water, and the pair would rise up out of -the water exposing much of their bodies. Vocalization during these battles was confined to lov; growls.

During the evenings of late fall females were also observed fighting and vocalizing to maintain terri­ tory or as the result of competition for food. The female bullfrog never engaged in the violent wrestling described above. Aggressive behavior in the females was confined to the low Female Territorial Call and the

Growl, accompanied by leaping on the intruder. Females would attack males or other females with equal vigor.

T-.-1 -1 * * _ _1 _ *1 T*>------J__J ? X" J. cty JJdUK. dl III J. r i CbCii w o 0 1.011 j-j * u w A > ^

During the latter part of August of 1969, I con­ ducted the following series of experiments: (1) the playback of various calls to male and female frogs in the experimental area, (2) a series of controlled encounters during which a model bullfrog was presented under various conditions to both sexes of bullfrogs in the test area, and (3) a number of controlled encounters during which both the model bullfrog and various recorded vocalisations were presented simultaneously and in as natural a way as possible— within the 149

limitations of the equipment used and my understanding

of what was natural to the bullfrogs in the area.

The first experiments were tried using playback

only. As far as the frogs in the experimental area were

concerned there were practically no visible effects to

playback alone. Males were not stimulated to call nor

were they attracted to the sound source. Females

appeared to be totally indifferent to the broadcast

vocalizations. Prior to the playback, bullfrogs that

were visible from the observation window were observed

as to posture and rate of buccal breathing. During and

shortly after playback there were no postural changes,

although the buccal breathing rate often slowed greatly

at the moment of broadcast.

The results of tha second and third experiments

can be discussed together. The effects of both the visual presentation of the model alone and in combination with playback varied greatly. Three vocalisation types

were used during the playback experiments; (1) the

explosive pop, (2) a two-call note vocalization, and

(3) a five-call note vocalization typically heard in the

late summer.

Actual quantitative evaluation of the separate

effects of visual presentation as opposed to visual

presentation associated with playback was not possible. 150

In both cases the responses varied widely depending on

the state of agitation of the bullfrogs.

If no natural calling had taken place for an

hour or more prior to model presentation, or the comb­

ination of model and playback, there was generally no

response. On the other hand, if one of the males had

called recently, within 15-20 minutes of model presenta­

tion— alone or with playback, one of the following

situations would occur. The male having recently called

would generally swim toward the model and would assume

a "hanging position" (hind legs hanging vertically down

into the water), and then submerge. The other possibil­

ity was that he would inflate his lungs and assume a

floating position. This inflated position was usually

followed by 1-3 pops, after which he might swim away or

he might procude a vocalisation of from 3-5 call notes.

As was mentioned much earlier the frogs were

often fed either insects or chunks of kidney. These

feeding sessions always resulted in numerous aggressive

interactions. These were characterized by vocalizations

and actual body contact such as nudging, wrestling, or

even biting. Following a feeding session the animals

always responded quickly to the model even without play­ back. Both sexes were aggressive toward the model just

as both sexes were aggressive during their natural 151

encounters. If the model was presented within about

one half hour after feeding the more aggressive animals

of both sexes would quickly respond. The same male that

dominated the vocalizations in the experimental area was

also the most aggressive in competition for food. This

male, designated "L," was generally quicker to respond

to the model than the most aggressive female of the

area, designated "M.11 Male "L" would react to the model

from a greater distance. The maximum distance from which "L" responded to the model was approximately

7 feet. For female "M" this distance was much shorter, being approximately 3 feet. The usual reaction of male

"L" to the appearance of the model v/as to first swim rapidly toward it, approaching to within about 1 1/2 to

2 feet of it. If the model was allowed to remain at the surface the male wouia then inflate ana promptly call. This call or vocalization would usually consist of 5-6 notes. If the model still remained at the surface the male would next paddle slowly up to it

(about 10-12 inches away) and call again. Loud explosive pops might be interspersed in this call pattern. He would then lungs at the model, seizing it with his fore­ limbs while literally butting the model with his lowered head, driving the intruder down into the water. The head was depressed with the eyes deeply retracted, the snout being used to butt the model. Another factor that 152 seemed to influence the response was the manner in which the model was presented. If it were brought up from beneath the water with only the head showing above water, the bullfrog generally turned quickly toward the intruder and often leaped on the model without any preliminary vocalizations. A low growl or pop would then follow.

If the model were brought up and floated so that most of the upper body was exposed, the bullfrog's approach was somewhat more cautious. Vocalizations would precede the attack— both pops and 1-4 note non-pulsatile territorial calls— but the attack would always follow.

If the dominant male was at such a distance that he either did not see the model or would not react to it, a nearby dominant female would approach the model: on some occasions even two females would approach. The female would produce a low "sputt" sound (probably equivalent to the louder and sharper pop sound of the male) and then a low growl. This was often followed by an active attack much like that of the male. Juvenile bullfrogs never participated in such activity and appeared oblivious to both the model and the playback.

The results obtained through playback alone were disappointing as the bullfrogs in the experimental pond did not respond to playback of pops, pulsatile or non- pulsatile territorial calls. Actually there was a 153

response as seen in the altered rate of buccal breathing,

but none so far as aggressive behavior was concerned.

Possibly this failure to respond to playback in the same

way that Wiewandt (1969) had observed was in part due

to the experiment's being conducted in August rather than

in June or July when the animals are more territorial.

On the other hand, the results with the ceramic

model were very dramatic, and here even without playback

I was able to obtain vigorous responses from both sexes.

The highly mobile model that I used seemed to be superior

to Wiewandt's stationary one. The use of a model to

determine the territorial boundaries is probably only

roughly quantitative. The frog's response varies with

its age, sex, the time of year, its state of arousal

and the environmental conditions. Consequently, a precise measure may not be possible. I repeatedly ob­

served that once a bullfrog attacked the model, the

second appearance of the model would elicit a strong response although the model might be 2-3 times farther from the frog as on the first occasion. After several

such encounters the response would wane, but this varied

also. This was the point at which the additional

stimulus of a playback vocalisation would re-establish

a strong response. The bullfrogs reacted differently

to the model depending on whether it appeared with only 154 the head above v/ater or with its upper surface exposed.

When the model simulated an inflated bullfrog floating at the water surface the response was less immediate and more cautious. This observation seems to corro­ borate Emlen's findings (1968) that a "high" posture functions directly as a threat. I have consistently observed this high inflated posture both in males and in females that were defending a territory in the water.

Mating Behavior

During the summer of 1970 both females "R" and

"M" ovulated and both egg masses proved to be fertil­ ized. There is no doubt then that successful breeding took place in the experimental area. On several occasions a female would be seen being clasped by a male, but these encounters invariably were short lived with the two animals separating quickly. On the day that female "R" ovulated (June 28, 1970) I noticed that the most dominant male called during the entire day, which was quite atypical for the males in the experi­ mental area. Although calls were heard occasionally during the day the bulk of the vocalizations occurred after dark. During the clasping of both females the male was heard and observed to produce at least two vocalizations. These have already been described as 155 the male growl and the male clasping call. I was fortunate in being able to obtain tape recordings of these vocalizations, often along with female release calls.

However, other than for these few observations, very little interaction between males and females which could be interpretated as mating behavior was observed.

The two sexes arrive at a state of excitement, through the activity of the nervous and endocrine systems, where the female will tolerate the contact of amplexus. How long clasping persists in nature is still not fully known. During his study of the life history of the bull­ frog, George (1940) noted repeated unsuccessful attempts to observe the course of mating under natural conditions.

Willis et al. (1956) resorted to observing spent gonads to determine when breeding had occurred. My observa­ tions indicate that clasping and ovulation may be of relatively short duration— perhaps minutes and surely not hours. Exactly how the male "knows" when to shed his sperm to fertilize the eggs at the moment of ovulation is also unknown.

Prior to the day of ovulation all vocalization by the male occurred after dark, but on this day calling was heard throughout the daylight hours. The most unique finding was that the male called during amplexus, producing the growl and what I called the male clasping call. Immediately after ovulation the female was no longer responsive and attempted to avoid the male. On the other hand, the male remained aroused for about

30 minutes after fertilization. During the next two days vocalizations from the male fell sharply and then began to rise again. Why this male came into a breeding condition a second time is unknown. After the second fertilization the arousal was not repeated. These few observations point out the need for a considerable amount of study on the mating behavior of these animals. CONCLUSIONS

Emergence dates for the bullfrogs in the experi­ mental area were in the first week of March on four consecutive years. This is quite early and was probably due to the combination of the shallowness of the pond and its being well sheltered from the wind. The evidence indicates that pond depth is an extremely important factor in emergence. In a near-by deep natural pond emergence was in mid-April. Possibly the bullfrog does not hibernate at the bottom of deep ponds, but rather along the edges at water depths that warm more quickly.

In central Ohio hibernation may not occur until the first week in December depending on the weather.

During the four years of my observations I had the im­ pression that the final signal for hibernation was the permanent freezing of the pond surface.

Air and water temperature had a profound effect upon the daily activity of the bullfrog. Much of the diurnal activity of the adult bullfrog involves the control of body temperature.

As scon as the grasses and weeds of the banks

157 158 began to grow the bullfrogs preferred to retreat into this shade during the day. Apparently this shade provided adequate protection from the sun and effectively replaced the shuttling activity observed by Lillywhite

(1970).

At the time of this writing there is no explan­ ation for the bullfrog's behavior involving the setting sun. The immediate impression given is that the animals were attracted to that area of the sky having the great­ est intensity of light. There must be more to the explanation than this as the frogs did not respond so strongly on evenings when the sunset was a brassy color as they did when it was a deep orange or red. Perhaps the longer wavelengths are involved.

The daily activity of late fall was entirely concerned with feeding and maintenance of territory.

The bullfrogs tended to enter the pond earlier than during the summer as the air temperature dropped more quickly by late afternoon at this time of year. The air layer 1-2 inches above the water surface became an important buffer zone at this time, insulating the bull­ frog from the chill night air. This was because the surface water of the pond is warmed during the day and this warm water slowly loses its heat to the air layer immediately above it. Figure 20, which shows the temp­ 159 eratures of air and water throughout the day and night of October 8, 1968, demonstrates how this zone typically developed. For example, this zone ranged from 4-7° C. warmer than the ambient air temperature at 2200 hours.

This permits the bullfrog to remain active until the early morning hours during the late fall.

Both the sex and age of the bullfrog appeared to influence the animals' tolerance to low temperatures.

Adult bullfrogs of both sexes tended to be active during the night hours, while the juveniles were active during the daylight hours. This separation of adults from juveniles permitted the immature bullfrogs to hunt for food without interruption from the adults. The adult female and the juvenile appear to be more tolerant of low temperatures than the adult male bullfrog. By late October males could no longer be seen at the experimental pond or at natural ponds in the area.

Feeding sessions in late August and early September showed that the ability of adult males to respond to the stimulus was greatly inhibited, while the adult females and juveniles responded normally. This effect­ ively removes the highly aggressive males from the scene allowing the females and immature bullfrogs to feed into the late fall. This may be of vital importance to the female for fat storage and subsequent development of 160

eggs. Because the adult female remains aggressive into

the late fall this may also present an opportuntiy for

selection of the most aggressive females. They would

tend to be more successful in competing for food, and with consequent fat storage, better chances of survival during the winter and egg development in the spring.

The effect of temperature on vocalization was

apparent. This is not surprising and is in complete agreement with numerous studies by other workers who have consistently observed that the physiological processes of frogs are affected by temperature. This suppression of calling by lowered temperature often resulted in vocalizations with greater call intervals and a greater number of call notes per vocalization.

This appears to be a demonstration of Sherrington1s

(1906) "reflex rebound" phenomenon.

The general effect of wind, when sufficiently strong, was to suppress normal activity. This suppression of activity is probably brought about by an increasing cooling due to facilitated evaporation, and to the actual physical pressures of the wind when higher speeds are involved.

Although the effects of changing barometric pressure could not be verified, many months of obser­ vation has left me with the impression that barometric,- 161 pressure does influence the behavior of the bullfrog.

I make this comment in the belief that a simple state­ ment of negative results might influence others to ignore what may prove to be a significant factor in the environment of the bullfrog and perhaps other animals.

The observation of hyperactivity associated with rain storms was studied only in a very limited way.

Hyperactivity rnay be the result of atmospheric ioniza­ tion brought about by electrical discharge during storms. The fact that hyperactivity occurs under natural conditions was confirmed on numerous occasions. Hyper­ activity may serve as a dispersal mechanism, stimulating frogs to move considerable distances without the threat of dessication due to the heavy rains that are always associated with such activity. Limited evidence indicates the possibility that the adult female bull­ frog may show a greater degree of hyperactivity, but much work is needed to confirm and expand the subject in general.

The relative humidity of the experimental area was consistently 10-14% higher than outside the enclos­ ure. This higher humidity influences the degree of evaporation of moisture from the skin of the bullfrog.

Cooling of the skin and loss of body fluids would be 162 reduced under such conditions.

Observations clearly showed that the -bullfrog has the ability to recognize and remember landmarks in its environment. This capacity must play a significant role in the maintenance of territory. Reaction to the ' environmental appearance by the bullfrog may also account for the observed shifting of the territory throughout the season. As weeds grow and the water level rises or falls, the environmental appearance also changes. The bullfrog appears to establish new terri­ torial boundaries at least partially based on these changes. Migration and emigration of bullfrogs also appears to shift the social hierarchy with consequent effects on territorial boundaries.

On emergence in the spring there was no feeding for the first 3-4 weeks. Both sexes began their normal activities as soon as the weather began to warm, but the adult males ceased to feed around mid-June— the beginning of the breeding season in central Ohio. Throughout the breeding season, from mid-June to mid-July, I was unable to observe adult males feeding.

The principal diet of the bullfrog is apparently insects which the bullfrog probably hunts entirely by sight. The bullfrog is capable of detecting a small insect at 5-3 feet— providing that it is moving. During 163 the activity of hunting the bullfrog quickly learns to associate the splashing of other frogs as they captured insects with food and quickly moves toward such areas of activity. Anticipatory behavior was often seen during feeding sessions, and a social hierarchy was evident at these times.

Adult males ceased to hunt or to respond to food by late August or early September. Adult females and juveniles continued active hunting into late September or early October. This difference seems to be due to the greater sensitivity of the male to lower temperatures.

In the area of Columbus, Ohio, vocalization in the bullfrog begins in early April and extends into late

August. During this investigation the development of a

Call Enerqy Index (CEI) based on the number of single call notes per minute appeared to be a valuable measure of several features. The CEI can be used to (1) study call patterns throughout the season for any particular environmental situation, (2) indicate the order of social hierarchy, and (3) indicate the degree of auditory stimulation among males, v/hich may reflect the effects of population density. Call Energy Index studies for the experimental and natural areas showed that there are at least two basic call patterns: (1) a mating season call pattern, and (2) a postmating season 164 call pattern. In both of the call pattern types the vocalizations involved are the pulsatile and the non- pulsatile territorial calls. The mating season call pattern of June-July is quite distinct from the post­ mating season call pattern of July-August.

The call pattern for the experimental pond took the form of the postmating season call pattern during both the mating season and the postmating season. Once mating has been accomplished the activity in a natural pond centers on hunting and the maintenance of territory for that purpose. Although territory is involved in both mating and hunting the call patterns are uniquely different. I believe that the unnaturally close con­ finement of the bullfrogs in the experimental area resulted in a situation where the bullfrogs were pre­ sented with an excessive amount of territorial chall­ enges. Consequently the call pattern for the experi­ mental pond v/as entirely of the form seen during terri­ torial defense. In spite of the absence of the mating season call pattern, mating did take place in the experimental area, although only one male was involved.

In 1970, after several days of vocalization by the males in the test area, one male became the dominant animal, and his vocalizations quickly suppressed all others. This agrees with the findings of Jenssen and 165

Preston (1958) who found that playback would suppress calling if sufficiently loud. The situation of a single male dominating the vocalization in the experimental area is possibly the consequence of the small size of the pond. Under natural conditions vocalizing males would probably be at least 15-20 feet apart. In the test area males were forced to hunt and to maintain territory at distances of 5-8 feet and often much closer. In this situation any male that became the dominant animal would tend to suppress the vocalization of the others.

The vocalization of the bullfrog may be stimul­ ated or suppressed by sounds audible to the animals. As an example, the sound of an airplane may stimulate a chorus to develop in a pond that had been silent for some time, or it might suppress a chorus already in progress. The effects of such external stimuli depend upon the state of the bullfrog perceiving it, and the quality of the sound itself.

During the course of this study ten call types were identified. Of these ten vocalizations only two are not involved in some form of aggressive behavior.

These two are identified as the Alarm Call or Squawk produced by a frightened frog as it leaps into the pond, and the Juvenile Cluck, which is produced while the 166

immature animal is moving about the pond. The other

vocalizations, designated as the Pop, the Growl, the

Burp, the Female Release Call, the Male Clasping Call,

the Male Non-Pulsatile Territorial Call, the Male

Pulsatile Territorial Call, and the Female Territorial

Call, are all produced during aggressive interactions

and are described and discussed in depth in the body

of this work.

As the mating season approaches the adult male bullfrog becomes preoccupied with the maintenance of

territory. The height of arousal appears to be marked by this territorial defense extending into the daylight hours. How long this peak of activity persists in a natural pond was not determined, but in the experimental

area it consisted of a single day, followed by ovulation.

At the present there is no knowledge as to whether ovulation under natural conditions occurs after dark, during the daylight hours, or both. In the single instance of the experimental area, ovulation occurred

in the late afternoon or early evening while still daylight.

During the height of the mating acitivty vocal­

ization by the breeding males reached its peak. The

Call Energy Index (CEI) for the experimental area rose very sharply just prior to ovulation and mating— then 167 fell sharply after fertilization. I have no explanation for the relationship of the CEI to mating other than it probably reflects the hormone level in the breeding male.

During amplexus the male bullfrog vocalizes, producing both the Growl and a vocalization that I call the Clasping Call. The function of these calls, if any, is as yet unknown. In the test area a single male fertilized two females at about a two-week interval.

.How this compares to a natural pond situation must await further investigation.

The presentation of a ceramic model of a bull­ frog along with playback of pre-recorded vocalizations had varied results. The use of the pre-recorded vocal­ izations without the model proved for the most part unsuccessful. The reason for this might lie in the fact that these experiments were carried out in August.

At this time of the year vocalization is rapidly dimin­ ishing and it is very likely that response to vocaliza­ tion is also lessening. In the latter part of summer or early fall the bullfrog's response to other bullfrogs is dependent more on visual than auditory input. Con-

3 e ouen* 11 y,-17 r e s t.o o n s e to the model w a s d r a rn a tic.

The posture a bullfrog assumed in the water reflected aggressiveness or the lack of it. The inflated, 168

floating posture always elicits either an aggressive

response or retreat depending on the social status of

the bullfrog being confronted. This is the so-called

"high" posture. The "low" posture with only the head

above water, if not a submissive posture, is at least one that does not produce an aggressive response.

Attacks on the model were quite violent with

the model usually being leaped upon, butted, and pushed beneath the water. Both sexes responded aggressively

to the model although the males were more forceful in

their attacks. These responses to the model were the

same as those observed between live bullfrogs both in

the experimental area and natural ponds. Juveniles never responded to the model. SUMMARY

A small pond was built for the study of the bullfrog on the property I was leasing. This was under­

taken to maximize the observation time, and to permit observations under as wide a range of environmental conditions as possible.

An area measuring 14 feet by 21 feet with a small pond, 13 1/2 feet by 8 feet by 3 feet deep, was entirely enclosed with fencing. The entire north side of the enclosure consisted of a pre-existing building measuring 12 feet by 24 feet. The pond was planted with cattails, pondweeds, duckweed, etc., and grasses were planted on the surrounding banks. In the spring of 1968 three adult male and three adult female bull­ frogs were introduced to the experimental area. During the course of the next three years numerous green frogs and immature bullfrogs entered the area and took up residence with the adult bullfrogs.

Because the experimental area was literally in my back yard permanent instrumentation could be in­ stalled. The environment was monitored daily for baro­ metric pressure, rainfall, wind direction and speed,

159 170

cloud cover, light intensity, relative humidity, maximum and minimum temperatures, and temperatures of

the banks and water.

Vocalizations of the bullfrogs were recorded on

a Sony 905A, Martel 301D, and Uher 4000 Report L, using bare microphones and an 18-inch and a 24-inch parabolic reflector.

Two types of recordings were made. The first

type was of vocalizations produced during observation periods, and the second was of vocalizations that were

automatically recorded by a sound actuated microphone that recorded all sounds produced from 1600 hours of an evening to 0500 hours of the following morning

(E.S.T.).

The result of this investigation was an ethogram for the bullfrog.

In the central Ohio area bullfrogs emerge in the first week of March to mid-April. The first bull­ frogs to emerge appear to be juveniles. The bullfrog does not feed for 3-4 weeks after emergence, but when feeding commences the diet is principally insects. The

adult males and females are active during the night, spending their daylight hours secluded in high weeds

surrounding the pond. The juveniles are active during

the day, and become inactive after dark. This difference 171

in periods of activity is probably a reflection of the

greater ability of the juvenile to control its body

temperature during the heat of the day. The juvenile

inactivity after dark is most likely due to suppression by the larger, aggressive adults, hunting at this time.

The adult female and the juvenile appear to be more

tolerant of low temperature than the adult male. Adult females and juveniles are actively hunting in late

August or early September, when few if any males can be found. Hibernation appears to begin much later than might be expected, and may start as late as the first week in December. The lack of males in late August does not necessarily imply that they have hibernated for the season. The adult males at this time will begin to retreat to the bottom of the pond and remain there for an hour or so, but then return to the surface again.

By late August they have ceased to feed, and apparently spend their time retreating to the pond bottom and returning briefly to the surface until hibernation is fully underway. All the bullfrogs seemed to be in hibernation only after the pond surface was permanently frozen.

Vocalization in the bullfrog in central Ohio begins in early April and continues into late August.

During this investigation a value called the "Call 172

Energy Index” (CEI) was developed based on the number of

single call notes produced per minute. These single

call notes could either be a vocalization in themselves

or part of a longer, more complex vocalization referred

to as either a pulsatile or a non-pulsatile territorial

call. When the Call Energy Index for the calling of

several bullfrogs is studied for any given period of

time, it can reveal the presence and order of a social

hierarchy. Further study of the CEI indicates that

vocalization of one male can either stimulate or suppress

the calling of another. The important factor here

appears to be the distance between the caller and the

receiver (listener). This is probably further compli­

cated by visual stimulation. The CEI tends to increase

as the distance between vocalizing males decreases.

However, if this distance becomes too small, vocal­

ization is inhibited. Consequently the CEI can reflect

population density. Study of the distribution of call

notes per vocalization during the active season shows

that there are two basic calling patterns— a mating and

a territorial calling pattern.

During this investigation ten call types were

identified. These are: (1) the Pop, (2) the Growl,

(3) the Burp, (4) the Female Release Call, (5) the

Male Clasping Call, (6) the Male Non-Pulsatile 173

Territorial Call, (7) the Male Pulsatile Territorial

Call, (8) the Female Territorial Call, (9) the Alarm

Call or Squawk, and (10) the Juvenile Cluck. Except

for the Alarm Call and the Cluck all of these vocal­

izations are associated with some form of aggressive behavior.

The Male Pulsatile Territorial Call appears to

serve two functions. First it is involved in terri­

torial defense, and second, in June and July when the female frog's blood-hormone level reaches a critical point, this same vocalization is interpreted by the female as a mating call. Both males and females were highly territorial, and used vocalization to help maintain their territory.

As the mating season approached the males ceased to eat, seemingly becoming totally preoccupied with territorial defense. Mating took place in the experi­ mental pond with two of the three females laying fertilized egg masses. The dominant male fertilized both of these females at about a two-week interval.

During amplexus the male produced a vocalization which was distinct from other vocalizations and consequently it was designated the Male Clasping Call. What its function is, if any, is unknown. Just prior to ovula­ tion the dominant male's CEI rose very sharply, and then 174

fell immediately following fertilization. This occurred

in both matings.

The use of prerecorded vocalizations in playback

experiments failed to produce pronounced results, al­

though as this experiment was performed in August these

results are not surprising. The bullfrogs did respond

dramatically to a ceramic model of a bullfrog. This

model was capable of being manipulated from within the

observation building, and such activities as hopping

onto or off a bank, swimming at the surface or beneath

it, suddenly surfacing in a specific bullfrog's terri­

tory, and assuming floating postures, could be simulated

with reasonable accuracy. An excellent opportunity to

observe aggressive behavior was provided by the presen­

tation of the model in these various ways to both male

and female bullfrogs.

A bullfrog with inflated lungs, floating with most of its body at the water surface, represented a highly aggressive posture. On the other hand a bull­

frog with only its head at the water surface showed a

lack of aggression. These findings corroborate those

of Emlen's (1958) in which he referred to these as

"high" and "low" postures. When the model was pre­

sented to a bullfrog in the high posture it was often quickly attacked, while the low posture often escaped 175 immediate attack. Further investigation with model presentation will doubtlessly prove valuable. LITERATURE CITED

Beilis, Edward D. 1957. The effects of temperature on Salientian breeding calls. Copeia, 1957, 2:85-89.

Bergeijk, William A. Van. 1967. Anticipatory feeding behavior in the bullfrog (Rana catesbeiana). Animal Behavior, 15:231-238.

Birkhaug, Konrad E. 1933. Phenyl mercuric nitrate. Jour. Inf. Dis., 53:250.

Blanchard, Frank N. 1930. The stimulus to the breeding migration of the spotted salamander, Ambystoma maculatum Shaw. Amer. Nat., 64:154-167.

Bogert, C. M. 1960. The influence of sound on the behavior of amphibians and reptiles, pages 137- 320. In W. E. Lanyon and W. N. Tavolga (eds.) Animal sounds and communications. A.I.B.S. Publication #7, Washington, D. C. 443 pp.

Borror, Donald J. 1961. The analysis of animal sounds, pages 26-35. In W. E. Lanyon and W. N. Tavolga (eds.) Animal sounds and communications. A.I.B.S. Publication #7, Washington, D. C. 443 pp.

Bragg, Arthur N. 1940. Observations on the ecology and natural history of Anura. I. Habits, habitat and breeding of Bufo coqnatus Say. Amer. Nat., 74:322-349 and 424-438.

Bragg, A. N. 1942. Observations on the ecology and natural history of Anura. X. The breeding habits of Pseudacris streckeri Wright and Wright in Oklahoma including a description of the eggs and tadpoles. Wasmann Collector, 5(2):47-62.

Bragg, A. N. 1943. Observations on the ecology and natural history of Anura. XVL. Life history of Pseudacris clarki (Baird) in Oklahoma. Wasmann Collector, 5(4):129-140.

176 177

Bragg, A. M. 1948. Observations on the life history of Pseudacris triseriata (Wied) in Oklahoma. Wasmann Collector, 7(4):149-168.

3ragg, A. N . 1950. Frequency of sex calls in some Salientia. Researches on the amphibia of Oklahoma. University of Oklahoma Press. Pages 117-125. '-:

Brenner, Fred J. 19S9. The role of temperature and fat deposition in hibernation and reproduction in two species of frogs. Herpetologica, 25:105-113.

Brode, William E. 1959. Territoriality in Rana cl amitans. Herpetologica, 15:140.

Brooks, Eleanor Stabler. 1918. Reactions of frogs to heat and cold. Amer. Jour. Physiology, 46:493-501.

Capranica, Robert P.. 196 5. The Evoked Vocal Response of the Bullfrog. Research Monograph #33, The M.I.T. Press, Cambridge, Mass. 110 pp.

Capranica, R. R. 1968. The vocal repertoire of the bullfrog (Rana catesbeiana). Behavior, 31:302-325.

Davidson. Treat. 19-33. Bullfrog ballet filmed in flicrht. National Geographic, -Tune 1963, 123(6):791-799.

Dickerson, Mary C. 1906. The Frog Book. Originally published by Doubleday, Page and Co., 1905. New Dover edition, 1959, by Dover Pub. Inc., Mew York, N. Y. 2 50 pp.

Dodt, Eberhard and Ewald Heerd. 1952. Mode of action of pineal nerve fibers in frogs. Jour. Neuro­ physiology , ? 5 (3):405-4 29.

Dole, Jim W. 1968. Homing in leopard frogs, Rana piglens. Ecology, 49(3):385-399.

Duellman, William F . 1970. The Hy1id Frogs of Middle America. 42 7 pp. Monograph of the Museum of

NntuiTc1.! Hi_■'ur>ity The M niv-ifS3 o-p K^nsss Ho. 1. 178

Durham, Leonard and G. W. Bennett. 1963. Age, growth, and homing in the bullfrog. Jour. Wildl. Manage., 27(1):107-123.

Emlen, Stephen T. 1968. Territoriality in the bull­ frog, Rana catesbeiana. Copeia, 1968, 1:240- 243.

Fassett, Norman C. 1940. A Manual of Aquatic Plants. McGraw-Hill Book Company, Inc. New York and London. 382 pp.

Foster, Woodbridge A. 1967. Chorus structure and vocal response in the Pacific tree frog, Hyla reqilla. Herpetologica, 23(2):100-104.

Frost, S. W. 1935. The food of Rana catesbeiana Shaw. Copeia, 1:15-18.

George, Ira D. 1940. A study of the life history of the bullfrog, Rana catesbeiana, at Baton Rouge, Louisiana. Unpublished Ph.D. dissertation, The University of Michigan. 96 pp.

Hinde, R. 1954. Factors governing the changes in strength of a partially inborn response, as shown by the mobbing behavior of the chaffinch (Frinqella coelebs). I. The nature of the response, and an examination of its course. Proc. R. Soc. B., 142:306-331.

Holzapfel, Ruth A. 1937. The cyclic character of hibernation in frogs. Quarterly Review of Biology, 1937, 12(l):63-84.

Jackson, Albert W. 1952. The effect of temperature, humidity, and barometric pressure on the rate of call in Acris crepitans Baird, in Brazos County, Texas. Herpetologica, 8:18-20.

Jenssen, Thomas A. and W. D. Klimstra. 1966. Food habits of the green frog, Rana clamitans, in southern Illinois. Amer. Midi. Nat., 76(1): 169-182.

Jenssen, Thomas A. and William B. Preston. 1968. Behavioral responses of the male green frog, Rana clamitans, to its recorded call. Herpetologica, 24(2):181-182. 179

Jones, Jeanne and Bayard H. Brattstrom. 1961. The call of the spring peeper, Hyla crucifer, in response to a recording of its own voice. Herpetologica, 17(4):246-250.

Kaplan, Harold. 1958. Marking and branding frogs and turtles. Herpetologica, 14:131-132.

Kaplan, Harold M. 1959. Electric tattooing for permanent identification of frogs. Herpetologica, 15:126.

Kornerup, A. and J. H. Wanscher. 1962. Reinhold Color Atlas. Reinhold Publishing Corp., New York. 224 pp.

Korschgen, Leroy J. and D. L. Moyle. 1955. Food habits of the bullfrog in central Missouri farm ponds. Amer. Midi. Nat., 54:332-341.

Lettvin, J. Y., Maturana, H. R., McCulloch, W. S. and W. H. Pitts. 1959. What the frog's eye tells the frog's brain. Proc. Inst. Radio Engrs., 47:1940-1951.

Lillywhite, Harvey B. 1970. Behavioral temperature regulation in the bullfroq, Rana catesbeiana. Copeia, 1970, 1:158-168.

Lorenz, K. Z. 1950. The comparative method in studying in n a t e b d h ctvio r p a tte irn S v Syiup# Soc# Exp# Biol., 4:221-268.

Manning, Aubrey. 1967. An Introduction to Animal Behavior. Addison-Wesley Pub. Co. 208 pp. Reading, Mass., Menlo Park, Calif., London, Ontario.

Martof, Bernard S. 1953a. Territoriality in the green frog, Rana clamitans. Ecology, 34(1):165-174.

Martof, Bernard. 1953b. Home range and movements of the green frog, Rana clamitans. Ecology, 34(1): 529-543.

Martof, Bernard. 1956. Factors influencing size of populations of Rana clamitans. Amer. Midi. Nat., 56(1):224-245. 180

Martof, Bernard S. 1962. The behavior of Fowler's toad under various conditions of light and temperatures. Physiological Zoology, 35(l):38-46.

Noble, Kingsley G. 1931. The Biology of the Amphibia. 1954 by Dover Publications, Inc., New York. 5 77 pp.

Oldham, R. S. 1957. Orienting mechanisms of the green frog, Rana clamitans. Ecology, 48(3):477-491.

Oliver, J. A. 1955. The Natural History of North American Amphibians and Reptiles. D. Van Nostrand Company, Inc. Princeton, N. J. 359 pp.

Raney, Edward C. 1940. Summer movements of the bull­ frog, Rana catesbeiana Shaw, as determined by the jaw-tag method. Amer. Midi. Nat., 23:733-745.

Risser, J. 1914. Olfactory reactions in amphibians. Jour. Exp. Zoology, 16:617-652.

Ryan, R. A. 1953. Growth rates of some Ranids under natural conditions. Copeia, (2):73-80.

Schiotz, Arne. 19 71. Evolution and mating call: ecological considerations. Herpetological Rev., 3(1):11 Feb., 1971.

Sherrington, Sir Charles. 1906. The Integrative Action of the Nervous System. Scribner's, New York.

Tinbergen, Niko. 1965. Animal Behavior. Life Nature Library, Time Inc., New York. 200 pp.

Torelle, Ellen. 1903. The response of the frog to light. Amer. Jour, of Physiology, 9:466-488.

Uexkull, Jakob von. 1934. A stroll through the v/orld of animals and men, pages 5-80. In C. H. Schiller (ed.) and K. S. Lashley. 1957. Instinctive Behavior. International Universities Press, Inc., New York. 328 pp.

Walker, Charles F. 1946. The amphibians of Ohio. Ohio State Mus. Sci. Bull., 1(3):1-109. 181

Weed, L. A. and E. E. Ecker. 1933. Phenyl mercuric compounds. Their action on animals and their preservative values. Jour. Inf. Dis., 52:354-363.

Wiewandt, Thomas A. 1969. Vocalization, aggressive behavior, and territoriality in the bullfrog, Rana catesbeiana. Copeia, (2):276-285.

Willis, Yuell L., Moyle, D. L., and T. S. Baskett. 1956. Emergence, breeding, hibernation and transformation of the bullfrog, Rana catesbeiana, in Missouri. Copeia, 1956, 1:30-41.

Woodbury, Angus M. 1954. Uses of marking animals in ecological studies: marking amphibians and reptiles, pages 6 70-6 74. In symposium given at an ecological session of the AAAS meeting at Berkeley, California, Dec. 27, 1954, and reported in Ecology, 37(4):665-689.

Wright, A. H. 1914. Life Histories of_ the Anura of Ithaca, New York. Carnegie Inst., Washington, D. C., Publication #197. 98 pp.