STUDIES ON THE ACOUSTICAL BEHAVIOR AND TAXONOMY

OP THE THEE CRICKETS (ORTHOPTEEA: OECANTHMftE)

OF THE EASTERN UNITED STATES

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State U n iv e rs ity

By

THCMAS JEFFERSON WALKER, J R ., B. A ., M. S.

******

The Ohio State University 1957

Approved hy

Deparmnem; o r Ziooj-ogy ana Entomology ACKNOVnjaXSMENTS

Many persons contributed to the research reported in this

dissertation. I wish to acknowledge their share in any merit the work may hare.

Special acknowledgment is due my adviser, Donald J. Borror, who

has given support, advice, and constructive criticism in all stages

of the work. The knowledge and enthusiasm of Richard D. Alexander

started me on this work, and discussion and field work with him have

been a continuing source of stimulation. George Potor, Division of

Biophysics, gave generously of his time and skill in constructing the

electronic apparatus which produced the imitation tree cricket calls

used in some of the bdiavioral studies. Thanks are due Basil Pames,

Department of Physics, for the design of the apparatus used in study­

ing the periodicity of singing. I am indebted to Henry Wave Shaffer,

Department of Physics, for advice concerning the mechanism of sound

production in tree crickets.

Many taxonomists were obliging in allowing me to examine their

collections and in giving me the benefit of their experience in

namenclatorial problems. Particularly helpful were Ashley B. Gurney,

U. S. National Miseum, who helped in locating and obtaining type

specimens; J. N. Khull, Ohio State University Museum, who brought

specimens from southern Texas; and Edward S. Thomas, Ohio State

Miseum, who informed me of an unusual form of tree cricket he had

found on willows in several areas in Ohio. B. B. Pulton, T. H.

ii iii

Hubbell, and J. A. G. Rehn gave patient advice concerning nomencla- torial questions.

Alvah Peterson and F. W. Fisk were generous in the loan of equipment used in the study of the effect of physical factors upon singing behavior.

Finally I wish to thank the Graduate School, who made the last year of this work free from alien worries by supplying an ample research grant. TABLE OF CONTENTS

INTRODUCTION...... 1

SOUND PRODUCTION AMONG TREE CKCCKETS...... 3

Occurrence of Sound Production ...... 3

Methods of Recording and A n aly sis ...... « ...... 6

Mechanics of Sound Production ...... 8

Structures Associated with Sound Production ...... 8

Nature of Tegminal Movement ...... 10

Determination of P itc h ...... 16

Variations in the Calling Song ...... 26

Major C ategories ...... 26

Physical C haracteristics ...... 28

Effects of Environmental Factors ...... 34

Temperature ...... 34 H u m id ity ...... 63 Air Currents ...... 68 L i g h t ...... 71 S o u n d ...... 72

Effects of Intrinsic Factors ...... 72

Individual V ariation ...... 72 Population V ariation ...... 75

Relation of Sound Production to the Diurnal Cycle ..... 78

Field O bservations ...... 79

Laboratory Observations ...... 79

TAXONOMY OF THE OEGANTHINAE OF EASTERN UNITED ST A T E S ...... 89

Introduction ...... 89

Plan of Presentation ...... 91

Key to the Oecanthinae of Eastern United S ta te s ...... 94

iv V

Genus N eoxabea ...... , ...... 99

Neoxabea bipunctata (De Gear) ...... 99

Genus Oecanth.ua ...... 107

The rileyl Group ...... ,...... 108

Oecanth.ua exclamation!s D avis ...... 109

O ecanthus n iv eu s (De G e e r ) ...... 115

Oecanthus rileyl Baker ...... 12?

The latipennls Group ...... 158

Oecanthus latipennls R ile y ...... 139

O ecanthus v a r ic o r n is F. W a l k e r ...... 145

The nigricornis Group ...... 152

Geographical Distribution .... 154

Ecological Distribution .. 160

Seasonal Distribution ...... 165

Calling Songs ...... 168

M o r p h o lo g y ...... 175

Hybridization and Crossing Experiments ..... 189

Summary of Distinguishing C haracters ...... 191

Oecanthus nigrleornls F. W alker ...... 193

Oecanthus celerinictus n. sp ...... 195

Oecanthus argentinus Saussure ...... 197

Oecanthus quadrlpunctatus Beutenmuller ..... 200

Oecanthus pini Beutenmuller ...... 202 vi

BEHAVIORAL EFFECTS OF THE CALLING SONG...... , 204

Introduction * 204

Synchronous Singing in the Snowy Tree Cricket ...... 204

M e t h o d s ...... 206

R e s u l t s ...... 208

Significance of Synchronous Singing ...... 213

Specificity in the Response of Females to Calling Songs of the M ales ...... 213

M e t h o d s ...... 214

Results ...... 220

Basis of Specificity Among Species Producing Continuous T rills ...... 223

Specificity in Species Producing Broken T rills . 228

Specificity in 0. rileyl ...... 228

Discussion ...... 230

Summary ...... 230

APPENDIX: DISTRIBUTION RECORDS OF THE OECANTHINAE OF THE EASTERN UNITED STATES AND CANADA ...... 232

REFERENCES CITED ...... 269

AUTOBIOGRAHIY...... 280 LIST OF TABLES

T able

I. Effect of mutilation of the tegraina upon the calling song of Oecanthus nigricom is

II. Effect of mutilation of the tegmina upon the calling song of Oecanthus latipennis

III. Laboratory recordings made in the study of effects of temperature upon calling songs

IV. Effect of high and low humidities upon the calling song of an individual of £. rileyi

V. Effect of air currents upon the calling songs of three tree crickets

71. Fulton’s (1926b) classification of the eastern Oecan- thinae and changes indicated by the present study

VII. T rill characters of Neoxabea bipunetata; field re c o rd in g s

VIII. T rill characters of Oecanthus exclamationls; field re c o rd in g s

IX. T rill characters of Oecanthus niveus; field recordings

X. Frequency of occurrence of chirps with various numbers of pulses in the song of 0, r i l e y i

XI. Measurements of 0. varicom is from Hidalgo and Cameron Counties, Texas

XII. Antennal markings of the nigricom is group, pini excluded

XIII. Measurements of members of the nigricom is group

XEV. File characters of the nigricom is group

37. Usefulness of various characters in identifying the members of the nigricom is group

XVI. Response scores and chi square values relating to specificity of response in females of species found in weedy fields viii

Table

X 7II. Response scores and chi square values relating to specificity of response in females of tree- dwelling species 222

XVIII. Response scores and chi square values showing the effect of temperature on the response of females of 0. quadripunctatus to certain sounds 225

xrx. Response scores and chi square values showing the response of 0. nigricom is females to the songs of three species singing at similar pulse rates 227

XX. Response scores and chi square values showing the response of 0, exclamationls females at 76° I*, to continuous and broken trills 227

2 2 1 . Response scores and chi square values showing the response of 0. rileyl females at 75° 7. to certain artificial and natural sounds 230

2 2 1 1 . Distribution of Neoxabea bipunctata in the eastern United States 233

XXIII. Distribution of Oecanthus exclamation!s in the eastern United States 236

2X17. Distribution of Oecanthus niveus in the eastern United States 238

XXV. Distribution of Oecanthus rileyi in the eastern United States and Canada 243

2X71. Distribution of Oecanthus latipennls in the eastern United States 247

2 2 7 II. Distribution of Oecanthus nigricom is in the eastern United States and Canada 250

XXVIII. Distribution of the willow form of Oecanthus n ig r ic o m is 255

XXIX. Distribution of Oecanthus celerlnietus 255 ix

Table gagg

ZEC. Distribution of Oecanthus argentinus in the eastern United State's 257

m rr. Distribution of 0ecanthU3 quadripunctatus in the eastern United States and Canada 261

XXXII. Distribution of Oecanthus pini 267 LIST OF ILLUSTRATIONS

Plate Page

I. Fig. 1. Structural details of functional file, 0. 9 la tip e n n ls

Fig. 2. Tegmina in singing position, £. nigricornis, posterior view

Fig. 3. Areas of tegmen removed in mutilation experiments

II. Fig. 4. Audiospectrogram of calling song of 0. 12. quadripunctatus

Fig. 5. Sections of above song

Fig. 6. Sections of oscillator note of 3300 cps

Fig. 7. Effect of pulse rate on pulse interval, 0, a rg e n tin u s . 64° F.

Fig. 8. Effect of pulse rate on pulse interval, £. argentinus, 74° F.

Fig. 9. Effect of pulse rate on pulse interval, 0. argentinus. 88° F.

III. Fig. 10. Song of 0. latipennls — file unaltered 15

Fig. 11. Same as above — file cut through near lateral end

Fig. 12. Song of 0. latipennls — outside edges of tegmina cut off — 76° F.

F ig . 13. Same a s above — 60° F.

Fig. 14. Song of 0. nigricomis — tips of tegmina cut off

Fig. 15. Section of song in Figure 12

Fig, 16. Section of song in Figure 13

Fig. 17. Section of song in Figure 14

x x i

Plate 2SS®. IV. Major categories of calling songs 27

Pig. 18. Continuous trill, Oecanthus quadrlpunctatus

Pig. 19. Broken trill, Oecanthus exclamationis

Fig. 20. Regular chirp, Oecanthus rileyi

V. Pig. 21. Effect of temperature on pulse rate, 0. 43 exclamationis, laboratory recordings

VI. Pig. 22. Effect of temperature on pulse rate, 0. niveus. 44 laboratory recordings

VII. Pig. 23. Effect of temperature on pulse rate, 0. rileyi, 45 laboratory recordings

VIII. Pig. 24. Effect of temperature on pulse rate, 0. 46 latipennls, laboratory recordings

IX. Pig. 25. Effect of temperature on pulse rate, 0_. 47 nigricornis, laboratory recordings

X. Pig. 26. Effect of temperature on pulse rate, 0. 48 nigricom is, willow form, laboratory recordings

XI. Pig. 27. Effect of temperature on pulse rate, 0. 49 celerinlctus, laboratory recordings

XII. Fig. 28. Effect of temperature on pulse rate, 0. 50 argentinus, laboratory recordings

XIII. Pig. 29. Effect of temperature on pulse rate, 0. 51 quadrlpunctatus, laboratory recordings

xrv. Pig. 30. Effect of temperature on pulse rate, 0. pini, 52 laboratory recordings

xv. Pig. 31. Relationship of pulse rate and pitch, 0. 54 exclamationis, laboratory recordings

Pig. 32. Relationship of pulse rate and pitch, 0. n iv e u s , laboratory recordings Plate

m . Pig. 33. Relationship of pulse rate and pitch, 0. rileyi, laboratory recordings

Fig. 34. Relationship of pulse rate and pitch, 0. latipennls. laboratory recordings

XTO. Fig. 35. Relationship of pulse rate and pitch, 0. nigricom is, laboratory recordings

Fig, 36. Relationship of pulse rate and pitch, 0. nigricom is, willow form, laboratory recordings

27111. Fig. 37. Relationship of pulse rate and pitch, 0_. celerinictus, laboratory recordings

Fig. 38. Relationship of pulse rate and pitch, 0. argentinus, laboratory recordings

XDC. Fig. 39. Relationship of pulse rate and pitch, 0. quadrlpunctatus, laboratory recordings

Fig. 40. Relationship of pulse rate and pitch, 0. plni, laboratory recordings

22. Effect of temperature on tr ill characters in 0. exclama­ tionis and 0. niveus

Fig. 41. Trill rate

Fig. 42. Average trill duration

Fig. 43. Per cent of singing time spent in trilling

2EE. Fig. 44. Effect of temperature on chirp rate, 0_. riley i, laboratory recordings mi. Fig. 45. Number of pulses per chirp, 0, r i l e y i , laboratory recordings

Fig. 46. Per cent of singing time spent in chirping, 0. rileyi, laboratory recordings mu. Fig. 47. Testing boxes as used to determine the effects of humidity upon calling songs

Fig. 48. Apparatus controlling humidity of air entering compartments in Figure 47 xiii

Plate

ZOT, Effect of relative humidity on pulse rate 67

Fig. 49. 0. nigricomis

Fig. 50. 0. celerlnictus

Fig. 51. _C. quadrlpunctatus

XXV. Fig. 52. Equipment for deteimining the effects of air 70 c u rre n ts

Fig. 53. 0. rileyi male, at rest; dorsal view

Fig. 54. £. rileyi male, singing; posterior view

XXVI. Fig. 55. Effect of individual variation on pulse rate 74

Fig. 56. Effect of age of individual on pulse rate

XXVII. Fig. 57. Effect of temperature on pulse rate, first and 76 second generations, 0. argentinus, Franklin Co., Ohio, laboratory recordings

XXVIII. Fig. 58. Apparatus for making a continuous record of a 82 cricket’s singing periods

Fig. 59. Method of lighting during periods of continuous light

xxrx. Fig. 60. Circuit diagram of monitoring device 83

Fig. 61. Graph made by monitoring device

xxx. Fig. 62. Singing periodicity of an individual of £. 85 rileyi in continuous dark and continuous light

XXXI. Fig. 63. Singing periodicity of an individual of 0. 87 nigricom is in continuous dark and continuous light

XXXII. Fig. 64. Distribution of N. bipunctata 103

Fig. 65. Distribution of 0. exclamationis m i n . Fig. 66. Effect of temperature on pulse rate, N. 105 bipunctata, field recordings

Fig. 67. Relationship of pulse rate and pitch, N. bipunctata, field recordings xiv

Plate Page

XZXIV. Pig. 68. Effect of temperature on pulse rate, 0. 113 exclamationis. field recordings

Fig. 69. Relationship of pulse rate and pitch, 0. exclamationis and niveus, field recordings

XXXV. Fig. 70. Distribution of 0. niveus 122

Fig. 71. Distribution of 0. rileyi

XXXVI. Fig. 72. Effect of temperature on pulse rate, 0. 125 niveus. field recordings

Fig. 75. Relationship of pulse rate and pitch, £. rileyi, field recordings

XXXVII. Variations in the song of _0. rileyi 137

Fig. 74. Usual song — all 8-pulse chirps

Fig. 75. Alternation of 5- and 8-pulse chirps

Fig. 76. Aberrant chirps

XXXVIII. Fig. 77. Distribution of 0. latipennis and 0. 143 v a r ic o m is

Fig. 78. Effect of temperature on pulse rate, 0. latipennis. field recordings

XXXIX. Fig. 79. Number of teeth and length of right file, N. 151 bipunctata, rileyl group, latipennis group

XL. Fig. 80. Distribution of 0. nigricornis 155

Fig. 81. Distribution of £. nigricornis, willow form

XLI. Fig. 82. Distribution of 0. celerinictus 156

Fig. 83. Distribution of 0. argentinus

XLII. Fig. 84. Distribution of 0. pini 157

Fig. 85. D istribution of jO. quadrlpunctatus xv

Plate gage XLIII. Fig. 86. Seasonal distribution of the field-dwelling 166 species of the nigricom is group, Columbus, Ohio, 1956

Fig. 87. Effect of temperature on pulse rate, nigricomis group, regression lines, laboratory reco rd in g s

XLI7. Fig. 88. Effect of temperature on pulse rate, 0. 172 eelerinictus, field recordings

Fig. 89. Effect of temperature on pulse rate, 0. argentinus, field recordings

Fig, 90. Effect of temperature on pulse rate, 0. quadripunctatus, field recordings

Fig. 91. Effect of temperature on pulse rate, 0. pint, field recordings

2LV. Fig. 92. Relationship of pulse rate and pitch, 174 comparison of species

XL7I. Fig. 95. Antennal markings of the nigricomis group 177

XLYII. Fig. 94. Number of teeth in right file, 0. nigricom is 182 Franklin Co., Ohio

Fig. 95. Femoral-tibial joint of left hind leg of holotype of 0. eelerinictus

Fig. 96. Eggs of 0. quadripunctatus

Fig. 97. Eggs of 0. eelerinictus

XLYIII. Fig. 98. Number of teeth and length of right file, 186 nigricom is group

XL'DC. Fig. 99. Apparatus for studying synchronism 207

Fig. 100. Apparatus for playing tape loops and for studying the response of females to sound XTi

Plate Page

L. Pig. 101. Responses of singing individual of £. rileyi 209 to random chirps

Pig. 102. Response of individual singing at 190 ch/m to recorded song of 166 ch/m

Pig. 103. Response of individual singing at 187 ch/m to recorded song of 242 ch/m

LI. Pig. 104. Synchronism of £. rileyi with recordings 212

Pig. 105. Diagram of setup used to produce artificial calling songs

LII. Comparison of natural and artificial songs 216

Pig. 106. Oecanthus quadripunctatus

Pig. 107. A rtificial continuous trill

Pig. 108. Oecanthus rileyi

Pig. 109. A rtificial pulsed chirp

Pig. 110. A rtificial pulseless chirp

n i l . Pig. 111. Diagram of cage showing zones used in recording 217 the positions of the crickets

Pig. 112. Response of 0. nigricornia females at 70° and 80° P. to artificial, 4000 eps continuous trills of different pulse rates INTRODUCTION

Tree crickets have received much attention because they are loud, persistent musicians and are abundant in the trees, shrubbery, and fields about man’s habitations as well as in undisturbed woodland.

Certain species have achieved notoriety among agriculturists because of damage to field crops and orchards from feeding and oviposition; however, today's organic insecticides have made them of minor economic importance.

There are two areas in which the present knowledge of tree crickets is particularly inadequate. The first and most fundamental area is that of classification and nomenclature. Currently the tree crickets of the United States east of the hundredth meridian are relegated to eight species, one of which is split into three sub­ species. Studies of the singing behavior, biology, morphology, and distribution of the eastern tree crickets have convinced me that there are at least eleven species involved instead of eight. Studies of the original descriptions and type specimens pertaining to names available for North American tree crickets have shown that of the eight specific names now in use, two are junior synonyms of older names, and one is applied to a species other than that of the type specim ens.

The second area selected for detailed investigation was that of acoustical behavior. The songs of eastern tree crickets were found to be species specific and hence good taxonomic characters. The effects of physical factors upon the singing cf tree crickets were

1 studied. Although, the song of the snowy tree cricket has received more attention than that of any other species of insect in this connection, no detailed study under controlled conditions has been made previously for any species. Neighboring snowy tree crickets are known to sing in exact unison; the mechanism of this synchronism was investigated for the first time.

Surprisingly little is known of the function of insect song. In

Orthoptera the song of the solitary male is known to attract sexually

responsive females, but whether or not the females respond to the

songs of males other than those of their own species has not been

dsnonstrated. If a specificity in response does exist, there arises

the question as to on what the discrimination depends. These ques­

tions wdre investigated using six species of tree crickets.

In this dissertation, I w ill begin with a discussion of the

taxonomy, biology, and song of the species of eastern United States,

and conclude with a report of experiments on the behavioral signifi­

cance of the calling songs. SOUND PRODUCTION AMONG TREE CRICKETS

Occurrence of Sound Production

Sound production and reception are integral parts of tie repro­ ductive behavior of tree crickets. Only the adult males produce sounds, the females having no specialized sound-producing apparatus.

The males stridulate in at least three situations. The most common situation exists when the male is out of contact with any female and is beyond tactile contact with any other male. In this situation, sound production may continue for hours with only brief pauses; the sound produced is generally known as the "calling song." A chorus of th e se songs forms a m ajor p o rtio n o f th e in s e c t n o ise h e a rd in summer and fall in eastern United States. Sexually responsive females are known to be attracted to this type song.

Vihen a female comes in contact with a male, the male ceases producing the calling song (if he is doing so) and begins a charac­ teristic courtship behavior. This behavior has been described by a number of authors (e*&. Houghton,1909a; Jenson 1909a; Hancock 1911;

Pulton 1915). Fundamentally the behavior consists of producing a characteristic song and backing toward the female. In approximate alternation with the sound-producing periods, the male takes a fiim hold on the substrate and shakes violently, largely in an up-and-down motion. Sounds produced in this situation are known as "courtship songs." Eventually the female mounts the male, and a spermatophore is passed. During the transfer the female arches her abdomen down­ ward, and the male arches his abdomen upward. The cerci of the male 4 on either side of the female's ovipositor serve as guides during the insertion of the barbed capillary tube of the spermatophore into the vagina of the female.

Having received the spermatophore, the female remains mounted and begins to feed on the secretions of the male's metanotal glands.

Occasionally the female w ill stop feeding and start to move away.

The male immediately begins a behavior similar to courtship behavior, and the female mounts again and continues feeding. Sounds made in this situation are called "posteopulatory songs." The female con­ tinues feeding on the secretions of the glands for a half hour or longer, the male exhibiting his characteristic behavior each time she starts to leave. While the female feeds, the sperm pass from the spermatophore into the seminal receptacle. Finally the female leaves the vicinity of the male. She then arches her abdomen underneath her thorax, removes the spermatophore, and eats it.

The above description of mating behavior differs from the various published reports cited above in two respects. First, there is no mention in previous accounts of the shaking behavior which roughly alternates with both courtship and posteopulatory singing. I observed this behavior in Oecanthus exclamationis and in all members of the 0. nigricomis group. This behavior pattern is similar to one described by Busnel, Pasquinelly, and Dumortier (1955) for Ephippiger spp. (Orthoptera, Decticinae); however, in the latter case females as well as males exhibited the behavior. The second point of difference concerns the time at which the spermatophore is transferred. Jensen

(1909a) and Houghton (1909a) reported that the spermatophore was passed before the female began feeding on the secretions of the meta- notal glands. Hancock (1911) and Fulton (1915), however, reported that the female feeds fifteen minutes to a half hour before passage of the spermatophore. In my own observations the spermatophore was usually passed before feeding, and never was there extensive feeding before copulation.

M. C. Busnel (1954) lists an additional situation in which tree crickets produce sound. Working in France with the European tree cricket, Oecanthus pellucens Scopoli, she found that when one male entered the area in which another customarily sang, the resident male would emit sounds she called "songs of warning." Singing in this situation was not observed in any of the species occurring in eastern

United States. When two males met, one male would usually leave, but there would be no sound exchange. Occasionally if a female was also in the vicinity, a male under laboratory rearing conditions would perform courtship behavior toward another male. Rarely the courted male would start to feed on the secretions of the metanotal glands, but this always elicited a violent withdrawal of the male doing the c o u rtin g .

In sunmary, song is characteristic of all phases of the repro­ ductive activity of male tree crickets. In this study three types of singing, based on the situation in which produced, are recognized:

(l) calling (male out of contact with female); (2) courtship (male in contact with female, prior to transfer of spermatophore); (3) post- copulatory (male in contact with female, spermatophore already attached to female). Methods of Recording and Analysis

Before discussing sound production among tree crickets further,

I think it desirable to describe the instruments which were used in the recording and analysis of the sounds.

All recordings were made on magnetic tape at a tape speed of fifteen inches per second. In the laboratory, recordings were made with a Magnecorder FT63-A tape recorder mechanism and a Magnecorder

PT63-J recording and playback amplifier (Fig. 100) (Magnecord Inc.,

1101 South Kilbourn Avenue, Chicago 24, Illinois). In the field, recordings were made with a Magnemite 610E recorder (Amplifier

Corporation of America, 398 Broadway, New York 13, N. Y.). A few field recordings were made with a later model of the same machine, the Magnemite 610EV (Fig. 99). The Magnemite is provided with an

asternal flywheel; however, tria l recordings showed that there was little difference in "wow” or flutter and no difference in tape speed whether the machine was used with or without the flywheel. Since the

flywheel was an inconvenience, especially in weedy fields and tangled undergrowth, it was not used.

The tape speed of the Magnemite varied some from one period of

recording to the next, and no system was used whereby the variations

could be determined from the recording itself. As a result, field

recordings are not entirely reliable for comparison with recordings made at other times or or other machines. Deviation from the speed

of the Magnecorder was plus two per cent when the Magnemite 610E was

tested in the fall of 1955. No further tests were made until the fall of 1956 after a series of recordings were made in which my voice was noticeably distorted when the recordings were played on another machine.

A test at that time showed a deviation of plus twelve per cent. Before that time recordings of my voice had sounded more natural, so previous­ ly the deviation must have been less than this. From a comparison of recordings made in the field with those made in the laboratory, it seems that the deviation was fairly constant at plus five per cent during most of the field work in 1956.

Most recordings were made with a D33 or D33A dynamic microphone, but in the fa ll of 1955 some recordings were made with a D22 dynamic microphone (American Microphone Co., 370 South Fair Oaks Avenue,

Pasadena 1, California). The response curves of these microphones are essentially flat (plus or minus two decibels) within the range of frequencies encountered. For field recordings the microphone was sometimes mounted in a 24-inch aluminum parabolic reflector. This allowed recordings to be made of crickets high in trees or deep in tangled weeds. However, most field recordings and all laboratory recordings were made with the microphone held a few inches from the

singing cricket.

Recorded sounds were analyzed with a Vibralyzer (Kay Electric

Co., 14 Maple Avenue, Pine Brook, New Jersey). Borror and Reese

(1953) give a summary of the use of this machine in sound analysis.

In brief it is an electronic device which produces two sorts of

graphs. One is an audiospectrogram, in which frequency is displayed

along the vertical axis, time along the horizontal axis, and intensity

is indicated Dy the darkness of the pattern (Fig. 4). The second is a section, which for any preselected point in time shows frequency on the vertical axis and amplitude on the horizontal axis (Fig, 5).

Mechanics of Sound Production

Structures Associated with Sound Production

Female tree crickets have narrow tegmina wrapped closely around the abdomen, but the males have broad tegmina which far surpass the abdomen in width (Fig. 53). It is with these highly modified tegmina that the males produce their sounds. On the undersurface of the right tegmen near the base is a vein which is at the bottom of a furrow and hence lies in a plane below that of the surrounding wing surface

(Fig. l) . This vein bears on its underside a series of teeth and is known as the file . The file is ,85 to 1.93 millimeters long depending on the species and is approximately perpendicular to the longitudinal axis of the body (Fig. 2). The teeth are not symmetrical in cross-

section but project mesad; the number varies from 17 to 62 and is of

taxonomic significance. The right tegmen has a sclerotized area along

the inner margin near the base. This area is called the scraper and

engages the file during sound production. There is also a scraper on

the right wing and a file on the left; however, these structures do

not engage and are nonfunctional since the right tegmen ordinarily

overlaps the left. In a few preserved specimens the left tegmen

overlaps the right, but it is doubtful that they functioned in this

position in life, since the scraper on the right wing is not fully

sclerotized while the scraper on the left wing is. Specimens with 9 PLATE I STRUCTURAL DETAILS OF FUNCTIONAL FILE Q. LATIPENNIS

SEGMENT OF FILE NEAR LATERAL END

POSTERIOR VIEW OF VENTRAL VIEW LATERAL END OF FILE OF FILE TEETH

LATERAO

F ig . 1

TEGMINA IN SINGING POSITION AREAS OF TEGMEN REMOVED IN Q NIGRICORNIS. POSTERIOR VIEW MUTILATION EXPERIMENTS

MOVEMENT OF -TEGMINA OUTSIDE EDGE CUT OFF

TECMEN ' ",7 7 - -TIP CUT OFF LINCTUflCO' '

\ -A S

L E F\ T \ \> RIGHT WING \ WING

NONFUNCTIONAL FILE FUNCTIONAL FILE

FUNCTIONAL SCRAPER

F ig . 2 F ig . 3 10 wings in reversed position when dead have been tape recorded producing normal sounds when alive, so the uncharacteristic position of the wings may be the result of contortions while dying.

Before starting to sing, the male cricket elevates its tegmina until they are approximately perpendicular to the longitudinal axis of the body (lig, 54). This automatically brings the outer portions of the tegmina (the lateral fields) into nearly the same plane as the inner portions (the dorsal fields) (Fig. 2). When the wings are at rest, the lateral fields of the tegmina are folded underneath the dorsal fields. Sound production is accomplished by opening and closing movements of the tegmina while they are held in an elevated p o s itio n .

Nature of Tegminal Movement

The exact nature of the tegminal movement during sound production was eocplored in a series of experiments. The first question to be answered was whether sound is produced as the tegmina open, as the tegmina close, or both. A clue to the answer to this question is found in the nature of the sound produced. Analyses of tape record­ ings of the songs of tree crickets show that all are alike in consist­ ing of series of uniformly spaced pulses of sound (e.£., Fig. 4).

Each pulse of sound must correspond to a period of movement in which the file and scraper are engaged. If the file and scraper were to engage on both the opening and closing of the tegmina, two pulses of sound would be produced for each cycle of wing movement. If the file and scraper were to engage only during opening or only during closing, 11

one pulse of sound would be produced for each cycle of wing movement.

With a stroboscope, the complete wing movements of a singing male were found to occur 2460 times per minute in one case (Oeeanthus

quadripunctatus at 76° F.) and 2940 times per minute in another (0. argentinus at 76° F .). Analysis of recordings made simultaneously

showed pulse rates of 2430 and 2910 per minute respectively. The

differences in the pulse rates determined with the Vibralyzer and the wing-motion-cycle rates determined with the stroboscope can be accounted for by errors in calibration of the two machines. This

experiment therefore indicates that with each back-and-forth movement of the wings only one pulse of sound is produced. The structure of

the sound pulses substantiates this finding, for each pulse is iden­ tical to its neighbors (Fig. 4). If sound were produced alternately by opening and closing of the wings, adjacent pulses would probably be different while alternate pulses would be identical. A pulse produced by an opening motion would differ from one produced by a

closing motion because the teeth of the file are asymmetric in section and they are more widely spaced at one end of the file than the other.

The next question to be answered is whether the closing or

opening motion is the sound-producing phase of the wing movement.

Nothing has been published concerning this question in tree crickets,

but Fierce (1948) in work with the field cricket, Acheta pennsylvanicus

(Buxmeister), concluded that this species produced sound with the

closing motion. He based his conclusion on motion pictures in which he observed that the wing-opening motions for a series of three pulses were not evenly spaced while the wing-closing motions were, there KILOCYCLES PER SECOND AUDIOSPECTROGRAM OF CALLING SONG OF Q. QUADRIPUNCTATUS Q. QUADRIPUNCTATUS OF SONG CALLING OF AUDIOSPECTROGRAM BP®® g. 4 . ig r F. ° 4 6 E ° 4 7 88° E 88° g 7 ig. F g 8 ig. F g. 9 . ig F g 5 ig. F ETOS F BV SONG ABOVE OF SECTIONS FET F US RT O PLE INTERVAL PULSE ON RATE PULSE OF EFFECT , 2 USSSCN, CPS 0 0 3 3 PULSES/SECOND, 32 E, ° 0 7 0.5 0 .10 .05 . ARGENTINUS 0. IE N SECONDS IN TIME IE N SECONDS IN TIME AE I I HATE g 6 ig. F SECTIONS OF OSCILLATOR OSCILLATOR OF SECTIONS OE F CPS 0 0 3 3 OF NOTE 4 PULSES/SECOND /4 3 6 4 4 PULSES/SECOND /4 3 5 6 1 PULSES/SECOND 31 1.5 i

12 2.0 13 being a longer pause between opening and closing during the first pulse than the two subsequent ones. Since the three pulses of sound produced were evenly spaced, he concluded that they were the results of the wing-closing motions.

Several features of the file in tree crickets furnish clues to the direction of wing movement during sound production. First, there is a knob at the lateral end of the file (Jig. 1) which would act as a stop for the scraper moving along the file during the closing motion. Second, the teeth of the file are slightly farther apart at the lateral end than at the mesal end. In the next section it will be shown that the tooth-strike rate largely determines the pitch of the song. There is frequently a slight down slurring in pitch during each pulse of sound (Fig. 4). This would be expected if the scraper moved at a uniform speed and engaged the teeth of the file during closure of the wings, for at any given wing velocity, tooth-strike rate, and hence pitch, would be lower the farther the file teeth were apart. Of course a slowing of wing movement toward the 8nd of the sound-producing motion could also account for the down slurring of the pulses. A third pertinent feature of the file is that the teeth project mesad (Fig. 1), so passage of the scraper along the file in the lateral direction would require more energy and result in a greater jarring of the wings than passage in the mesal direction. If

Pierce (1948) wrs correct in his deduction that the field cricket produces sound only during the closing motion, then the file is moving against the "grain" of the file teeth. Such a relationship definitely exists in the true katydid, Pterophylla camellifolia 14

(ITabricius). In this species the wing movement is slow enough to be correlated by eye with production of sound, and E. D. Alexander

(personal communication) stated that he had determined that sound is produced when the scraper moves against the direction of projection of the file teeth. If the same relationships exist in tree crickets, the sound is produced during the closing of the wings. Three features of the file, therefore, give circumstantial evidence for considering each laterad movement of the scraper responsible for the production of a pulse of sound.

In an effort to get more evidence on this question, one end or the other of the files of sixteen tree crickets was burned with a hot needle or cut with a razor blade. It was hoped that irregularities at the beginning or end of a pulse could be correlated with corres­ ponding damage to the file; however, only one specimen sang after treatment. This was a specimen of Oecanthus latipennis in which the file had been cut through at a point eight teeth mesad of the knob.

There were 42 file teeth in all. Audiospectrograms of the song of this cricket show that normal pulses usually alternate with pulses in which the latter one third is of a different nature than the first two thirds (Figs. 10 and 11). The aberrant pulses, taken by themselves, would seem good evidence that the break in the file was near the end of the pulse and hence the scraper was moving laterad. The normal- appearing pulses complicate this interpretation, but perhaps they can be considered as results of strokes in which the two segments of the file lay in the same plane, so that the break caused no interruption in the stroke. KILOCYCLES PER SECOND i H g. 15 . ig F ETO O SN SCIN F OG ETO O SONG OF SECTION SONG OF SECTION SONG OF SECTION g 10 ig. F g u ig. F g. 12 . ig F g. 15 . ig F g. 14 . ig F jrnmeiwtiBWiroiro N IUE 2 N IUE 3 N IUE 14 FIGURE IN 13 FIGURE IN 12 FIGURE IN AE S BV-IE U TRUH ER AEA END LATERAL NEAR THROUGH CUT ABOVE-FILE AS SAME OG F .NGIQNS- IS F EMN CT OFF CUT TEGMINA OF TIPS - NiGRICQRNIS 0. OF SONG * — f r - OG F Q. OF SONG OG OF SONG 0.25 t F EMN CT F. ° 6 7 - F F O CUT TEGMINA OF 05" SAME AS A B O V E -6 0 ° F. ° 0 -6 E V O B A AS SAME Ol AIENS-USD EDGES -OUTSIDE LATIPENNIS g 16 ig. F IE N SECONDS IN TIME IE N SECONDS IN TIME AE III HATE 0.5 FL UNALTERED -FILE » ■ « i, 17 Fig, 0.75 15 .1 15 16

In summary, all the available evidence points to the scraper engaging the file and producing a pulse of sound with each closing movement of the wings. On the opening movements, the scraper and file do not engage, or else the scraper runs along the file without

encountering enough resistance to cause the wings to vibrate.

Determination of Pitch

Cricket songs, in contrast to the songs of many Orthoptera, are characteristically nearly pure in pitch. That is, each song has a dominant frequency rather than being a noise-like combination of many frequencies. A section of a tree cricket song compared with the section of an electronically produced pure frequency (Pigs. 5 and 6 ) shows how the song of a tree cricket approximates a pure frequency.

Pigure 12 is a section of a noise-like song. Another feature regard­ ing the pitch of tree cricket songs is that at higher pulse rates

(wing-stroke rates) the pitch produced by a given species is usually higher. In some species this increase in pitch is almost a linear function of pulse rate (Pig. 32); in others, the increase does not continue linearly but appreciably lessens as the pulse rate reaches higher values (Pig. 39). Thus in connection with pitch there are two features to explain •— the production of a pure pitch and the relation­ ship of pitch with pulse rate.

Basically the production of a pure frequency can be accounted for in two ways. One theory is that the tegmina act as simple harmonic vibrators that vibrate once with each tooth impact. In such a situa­ tion the tooth-strike rate is a direct determiner of the frequency of 17 the song. The other theory is that the tegminal membranes are resona­ tors with characteristic frequencies of their own and that the nature of the membranes determines the pitch of the song.

Three workers have published their ideas concerning the determi­ nation of pitch in the Oecanthinae. In each case the conclusions were somewhat different from my own. Pierce (1948) concluded that both methods of frequency determination operate in the Oecanthinae. In the songs of Oecanthus nigricornis, 0 . rileyi. and 0_. pini. he found two frequencies, one which he thought was probably due to tooth- impact rate; the other, to wing resonance. In _0. quadripanetatus he observed a single frequency which he attributed to tooth-impact rate.

In nigricornis and pini the two frequencies found by Pierce seem to have been determined at different times and can be attributed to different pulse rates. In rileyi he found two frequencies produced simultaneously. In specimens with damaged tegmina I have recorded songs with more than one frequency, and such a condition perhaps accounts for Pierce’s results. Most of Pierce's frequency determina­ tions are much lower than comparable ones obtained with the

Vibralyzer. Since he admits that in the frequency range of tree cricket songs he could not be sure whether his apparatus was measuring the fundamental or a harmonic of the insect’s song, in most cases he probably determined a subharmonic instead of the fundamental.

Basquinelly and M. G. Busnel (1954) thought that the stability and purity of the pitch of the song of Oecanthus pellucens is best accounted for by considering the tegmina to be resonators with well- defined constants. They proposed that the two tegmina act as prongs 18 of a tuning fork and that when the tegmina move apart during a single vibration, the scraper is able to slip by one tooth of the file. The subsequent impact with the next tooth reinforces the next vibration of the wings. In their theory, then, the resonant frequency of the wing membranes determines the tooth-impact rate. This theory must be rejected for several reasons. First of all, experiments in which the wing membrane was punctured did not result in a change in the pitch of the song. The resonant frequency of the wing is certainly changed when the membrane is damaged, so no role of this frequency is indicated. Secondly, if the resonant frequency of the wings determines the pitch, changes in pitch \vith changes in pulse rate are difficult to account for. There is nothing in the structure of the wings to suggest the means of radical changes in resonant frequency. Finally, the wings are highly damped, that is, there is much resistance to their movement through the air. Any frequency other than the driving frequency would be expected to be of low intensity. For the resonant frequency to control the tooth impact rate would require considerable action to be initiated by the resonating tegmina, and their highly damped state would make this unlikely.

The theory that the tooth-impact rate is the primary determiner of the pitch of the song agrees with all the evidence available. The tegmina are highly damped simple harmonic vibrators driven by the im­ pacts of the scraper upon the teeth of the file. The action of the tegmina may be compared to that of the sounding board of a piano. The tegmina and the sounding board vibrate at whatever frequency they are 19 driven, and the result is a louder sound at the same frequency as the driving frequency.

Some experiments in which the tegmina were mutilated in various ways bear out the assumption that tooth-impact rate normally deter­ mines pitch. Specimens of £. nigricornis and 0. latipennis which had previously been recorded in the laboratory were mutilated in the following ways (Fig. 3): (l) tegmen punctured, (2) tip of tegmen cut off, (3) outside edge of tegmen cut off, (4) functional file or functional scraper removed. In one group of each species the right tegmen was mutilated, and in another group of each species the left tegmen was mutilated. After a recording had been made of the song of an individual with one tegmen mutilated, the other tegmen was mutilated in the same manner as the first, and further recordings were made. The results of these experiments are listed in Tables I and II. There are two generalizations that may be made immediately.

In no case did an individual without a file or scraper produce a

detectable song, and in no case in which a cricket sang with mutilated wing(s) was there a deviation from the normal pulse rate of that

species at that temperature (for an idea of the Influence of tempera­

ture on pulse rate, see Figs. 21 to 30), The importance of the file

and scraper in sound production is confirmed by the first generaliza­

tion, and the lack of influence of the resonance of the wings on the wing-stroke rate is confirmed by the second.

Upon closer scrutiny of the results, there are some additional

conclusions to be drawn from these experiments. First, at lower

temperatures, and therefore lower wing-stroke rates (Figs. 21 to 30), 20

Table I. Effect of mutilation of the tegmina upon the calling song of Oecanthus nigricornis. Dashe 3 indicate that cricket did not s in g .

T _ Effect on Treatm ent vidualS •> or. ^'Pulse ^ Pulse Htch .

Right tegmen punctured A 76 normal regular normal

Left tegmen punctured B

Both tegmina punctured A 71 normal regular normal Both tegmina punctured A 7 5 | normal regular normal Both tegmina punctured A 80 norm al re g u la r normal

Tip of right tegmen cut off C 75& normal regular normal Tip of right tegmen cut off D 76 normal regular normal Tip of right tegmen cut off D 80 normal re g u la r spread

Tip of left tegmen cut off E 70 normal i r r e g . normal Tip of left tegmen cut off F 76 norm al re g u la r normal

Tips of both tegmina cut off C 84 norm al irreg. spread

Outside edge of right teg. cut off G — ------

Outside edge of left teg. cut off H 7 5 j normal i r r e g . s c a t­ te re d Outside edges of both teg. cut off H — — — — — —

Functional file cut out I ------

Functional scraper cut out J 21

Table II. Effect of mutilation of the tegmina upon the calling song of Oecanthus latipennis. Dashes indicate that cricket did not sin g .

E ffe c t on Treatm ent in d i­ Temp. P u lse P u lse v id u a l ° F . Bate Length P itc h

Bight tegmen punctured A 7 5 | normal regular normal

Left tegmen punctured B 76 normal regular spread

Both tegmina punctured A 76 normal regular norm al Both tegmina punctured A 80 normal regular normal

Tip of right tegmen cut off C 76 normal regular norm al

Tip of left tegmen cut off D 75 normal regular spread

Tips of both tegmina cut off D 65 normal re g u la r spread Tips of both tegmina cut off D 76 normal regular sp read Tips of both tegmina cut off C 76 normal regular spread

Outside edge of right teg. cut off E 75-| normal regular normal

Outside edge of left teg. cut off F 76 normal irreg. normal

Outside edges of both teg. cut o ff F 60 normal regular normal Outside edges of both teg. cut o f f F 65 normal irreg. normal Outside edges of both teg. cut o f f F 75g normal irreg. norm al Outside edges of both teg. cut o ff E 60 normal regular normal Outside edges of both teg. cut o f f E 76 normal irreg. scat­ te re d Functional file cut out G — —------

Functional scraper cut out H — ------22 the effects of mutilations upon the song produced were less than at higher temperatures (and higher wing-stroke rates). For instance, in £. latipennis with the outside edge of both tegmina cut off, at

76° F. (and 53jjs- pulses per second) there was a great mixture of frequencies and an irregular pulse duration (Figs. 12 and 15). The same individual singing at 60° F. produced a song normal for that temperature (30|- pulses per second, 2400 cycles per second) (Figs. 13 and 16). These results may be explained as follows. The ability of a damped vibrator (the tegmen) to vibrate at a driven frequency (tooth- impact rate) and none other decreases as the frequency increases. The higher the driving frequency, the greater is the stress upon the vibrator. The mutilated wings were weakened and could not always stand the stress of the higher frequencies at the higher temperatures; then instead of acting as simple harmonic vibrators, they vibrated in a complex fashion and produced a mixture of frequencies.

Second, in their influence upon the pitch of the sound produced,

the treatments rank from least to greatest in this order: ( 1 ) punc­

tured wing membrane, (2) tip of wing cut off, (3) outside edge of wing cut off. On the basis of the determination of pitch by tooth-

impact rate, these results are easily interpreted. The ability of a vibrator (the tegmen) to vibrate at a particular driven frequency

(the tooth-impact rate) and none other varies with the rigidity of the vibrator. A rigid vibrator w ill vibrate at the driven frequency over

its entire area, whereas a non-rigid vibrator will vibrate at dif­

ferent frequencies in different portions of its area. The treatments

rank in the same order as to effect upon rigidity as to effect upon 23 pitch.. Puncturing the membrane had little effect upon the rigidity of the wing because only one of the supporting veins was cut. Cutting the tip from the wing disrupted the arched veins supporting the distal part of the dorsal field. The effect upon pitch was usually the introduction of additional frequencies with the maintenance of the same dominant frequency as before (Pigs. 14 and 17) — minor areas of the wing vibrated at other than the tooth-impact frequency. Cutting the outer edge from the wing reduced the rigidity greatly, because during sound production this edge and the dorsal field are in slightly different planes. The structural effect of such an arrangement may be illustrated by trying to hold upright from the base a vertically creased and a non-ereased sheet of paper. The creased sheet is rigid enough to remain vertical, but the non-creased sheet collapses. The effect of removing the outside edge seems to have been greater in 0 . nigricornis,.in which the dorsal field is narrower and more dependent upon the support of the outside edge than in 0. latipennis. In both species it is noteworthy that only at temperatures of 76° F. or below did crickets mutilated in this manner sing. When the effects of this sort of mutilation were apparent, the singing was never long sustained.

The irregularity of the pulse duration may have resulted from non- uniform vibration of the wing3 sometimes causing the file and scraper to cease contact sooner than usual.

The acceptance of tooth-impact rate as the determiner of pitch makes same additional relationships between pulse rate and pitch evident. At higher pulse rates (wing-stroke rates) the wings would be expected to move faster, resulting in more teeth being struck per 24 unit time and the pitch rising. The pitch does rise with increasing

pulse rate (Figs. 31 to 40); however, in no species does it rise

steeply enough to result in a doubling of pitch with a doubling of wing-stroke rate. For instance in 0. latipennis (Fig. 34) an increase

in pulse rate from 25 to 50 pulses per second (a factor of 2.00)

results in an increase in pitch from 1950 to 2975 cycles per second,

a factor of only 1.53.

Since pitch is determined by tooth-impact rate, with increasing pulse rates there must be some change in the cycle of wing motion.

One mechanism that would account for the observed relationships is

for the sound-producing part of the cycle of wing motion (the closing)

to increase in velocity less rapidly than the non-sound-producing part

(the opening). In other words, the proportion of the wing-motion

cycle spent in closing (sound production) would increase with increases

in pulse rate. Audiospectrograms of the songs of the same individual

o f 0 . argentinus singing at three different pulse rates show this to

be the case (Figs. 7, 8, 9). At 31 pulses per second, the pulse is

.53 of the duration of the wing-motion cycle; at 46 3/4 pulses per

second, the pulse is .65 of the duration of the cycle; at 65 3/4, the

pulse is .76 of the duration of the cycle.

The number of file teeth struck during a pulse was calculated

for each of the above three songs by using the following formula:

Number of teeth struck * duration of sound pulse in seconds x pitch

in cycles per second. The calculated numbers were 46, 51, and 50

respectively. Since the individual concerned had only 49 teeth in

its file, the duration of the pulses of sound must, in the latter two 25 instances, have been slightly less than measured. The exact determi­ nation of pulse duration is difficult because within limits an increase in the marking intensity of the Vibralyzer w ill cause an increase in the apparent pulse length. Furthermore, the surroundings in which a recording was made can influence the apparent sharpness of the transition from sound to silence. For Figures 7, 8, and 9 , a l l factors in recording and analysis were kept constant, so comparisons among the audiospectrograms are meaningful, although absolute values must be slightly different from those indicated.

Another mechanism that could account for an increase in pulse rate without a similar increase in tooth-impact rate is the striking of a smaller proportion of the file teeth. The increased pulse rate would result from the traversing of less distance while maintaining the same speed. Analysis of recordings of 0. nigricornis made at 80 pulses per second and above, when increases in pulse rate no longer result in marked increases in pitch, showed no trend toward a striking of fewer teeth. However, accurate determination of the number of teeth struck per pulse was impossible because of the difficulties pointed out above in determining pulse duration. Nevertheless, it seems safe to infer that if this mechanism comes into play, it is on a minor scale compared to the first one.

In conclusion it should be pointed out that there are very definite upper lim its of pulse rate if the first mechanism is the only one that operates, for as the ratio of the sound duration to the total duration of the cycle approaches its lim it ( 1 . 0 0 ), the speed of the tegmina in opening must approach infinity. These lim its have not 26 been reached in my observations, so there is no need at present to postulate further mechanisms to account for the relationships of pitch and pulse rate.

Variations in the Calling Song

iSajor Categories

The calling songs of the tree crickets of eastern United States

are alike in that they consist of sequencies of pulses (corresponding

to wing strokes). On the basis of the continuity of the pulse

sequences, the species can be divided into these three categories

(Figs. 18, 19, 20): (l) Species in which the pulse sequence continues

without interruption for minutes at a time. This type of calling song

will be referred to as a continuous trill and is characteristic of

Oecanthus latipennis. _0. varicornis (probably), 0 . nigricornis, 0_.

c e l e r in i c tu s . _0 . argentinus, 0_. quadripunctatus, and £. pini. ( 2 )

Species in which the pulse sequence is discontinued and begun at

irregular intervals of from one to several seconds. This type of

calling song is designated a broken trill and is characteristic of

Neoxabea bipunctata, 0. exclamationis, and _0. niveus. (3) Species

in which the pulse sequence is broken into uniform bursts with uniform

intervals. This type of calling song is called a regular chirp and is

found in only one species of eastern tree cricket, 0 . r i l e y i . FREQUENCY IN KILOCYCLES PER SECOND 2 4- 0 0.6 0.4 0.2 0 10 0 10 0 ^ - 0 o“ lt lt mti im ilftk illtk

g 18 ig. F m m g. 19 . ig F g. 20 . ig F •itiiiii AO CTGRE O CLIG SONGS CALLING OF CATEGORIES MAJOR EATU QARPNTTS 5 F. 75° QUADRIPUNCTATUS OECANTHUS 0.2 2 i i i l i i l l l l i i l i EATU ECAA1NS 3 R 73° EXCLAMAT10NIS OECANTHUS

EATU RLY 75°E RILEYI OECANTHUS OTNOS TRILL CONTINUOUS 10 EUA CHIRP REGULAR RKN TRILL BROKEN i IE N SECONDS IN TIME . 0.6 0.4 0 m am m m m am m 4 0 .4 itii t if i iiit m m i i j » % % » * » » ! AE IV HATE .....

l ~ i mm t i i i i i k i i i m i 0 6 0 2 0 2 20 0.8 0.8 l l i l i l i k i l l l l l i 1 8 27 30 0 3 30 1.0 i.o 28

Physical Character!sties

Before the calling songs of tree crickets can be further classi­ fied and eventually used in distinguishing species, it is necessary to examine their more detailed structure and to determine what varia­ tions in song exist within a species. The following characteristics apply to calling songs and will be used in discussing their varia­ tions, For each characteristic, the definition, applicability, and accuracy of measurement are indicated.

Pulse Rate. While a trill is being produced, the wings maintain a regular rhythm in their movement, and this results in a uniform series of sound pulses, corresponding to the closing portions of the wing-motion cycles. Pulse rate is easily measured and is very useful in distinguishing species. In this dissertation it is expressed as pulses per second —- hereafter abbreviated as p/s. Pulse rate was measured to the nearest 4 p/s by making an audiospectrogram of a little more than two seconds of song and counting the pulses in a distance equivalent to exactly two seconds. The only source of appreciable error in pulse rate measurement was failure to play the recording being analyzed at the exact speed at which it was made.

With the Magnecorder, tape speed was 1.0 per cent slower when measured in the fall of 1956 than when measured in the fall of 1955.

Since the Magnecorder was used to play recordings into the Vibralyzer, when recordings made with the Magnecorder in 1955 wgre played into the

Vibralyzer in 1956, the pulse rate was 1.0 per cent slower than what the insect had produced. The majority of the recordings was made and 29 analyzed in 1956, so error from this source should in most cases be less than one per cent. More serious errors are encountered when recordings made with the Magmenite are considered. Variations in the tape speed of this recorder are discussed on pp. 6-7. The pulse rates of field recordings made in 1955 were multiplied by 1.020 in order to correct for the difference in the speed of the Magnemite and the speed of the Magnecorder. Pulse rates of field recordings made in 1956 prior to October were multiplied by 1.053, while those of field recordings made in October were multiplied by 1.136 to correct for the increased deviation in speed of the Magnemite during that month. The pulse rates determined from field recordings are obviously less re­ liable than those determined from laboratory recordings. Variations in the speed of the Vibralyzer could also be a source of error in determining pulse rate, but no variations were discovered when the machine was calibrated in 1956 and 1957.

Measurement of pulse rate in the song of Oecanthus rileyi pre­

sents difficulties not encountered in other species. Ordinarily each chirp of this species consists of eight pulses. Frequently these pulses are spaced non-uniformly — the first two pulses, the third

through fifth pulses, and the sixth through eighth pulses being

slightly separated from each other (Fig. 76). The "true" pulse rate,

that is, the pulse rate homologous to that measured in the songs of

trilling spedies, is probably the rate during the delivery of each of

the three subgroups within the eight-pulse chirp. However, this rate

is impractical to determine accurately since the entire chirp lasts less

than one-fourth second, and the subgroups are of course even briefer. so

Therefore, a somewhat arbitrary method of pulse rate determination was used for 0. riley i. An audiospectrogram was made of a chirp so that

1.0 millimeter on the graph was equivalent to 2.35 milliseconds. The

distance between the beginning of the second pulse and the beginning

of the eighth pulse was measured to the nearest one-half millimeter.

This gave a measure of the time required to produce six pulses, and

the pulse rate in p/s was calculated from this figure.

Pulse Duration and Interval. Pulse duration is the time which

elapses during the production of one pulse of sound. Pulse interval

is the time that elapses between the end of one pulse and the begin­

ning of the next. Because of the conditions in which the recordings

were made and the lim itations of the method of time analysis, no

attempt was made to make an extensive study of these characteristics.

The discussion on p. 25 indicates in more detail the difficulties in

their measurauent and summarizes the sorts of variation which exist.

Pitch. Ordinarily a tree cricket song has a strong dominant

frequency or pitch. This frequency varies with the pulse rate, but

a particular frequency at a particular pulse rate may be helpful in

distinguishing species. Pitch was measured by taking five Vibralyzer

sections within a two second sample of the recorded song. Each sec­

tion was measured to the nearest 100 cycles per second (hereafter

abbreviated cps), and the value obtained most frequently was considered

to be the pitch of the song. The same correction factors which were

applied to pulse rate were also applied here. The calibration of

frequency determinations was made originally with a Model 20QA audio

oscillator (Hewlett Packard, Palo Alto, California). Later this was 31 checked against a more stable standard, a Stroboeonn (C. G. Conn Ltd.,

Elkhart, Indiana) and found to be five per cent high. Consequently the frequency determinations had to be multiplied by another correction factor to make them agree with the more absolute standard.

T rill Bate. In apecies with broken trills, the average number of trills per minute during an observation of at least five trills is called the trill rate. In determining the trill rate of a recording, the time from the beginning of the first complete trill to the beginning of the last complete trill was measured, and the trills during that period were counted. Pauses longer than ten seconds between trills were classed as breaks in the singing and were not included in calculations of trill rate. Variations in tape speed are sources of error here as in pulse rate determination, and the same correction factors were applied. T rill rate may vary widely even in the same individual while under constant external conditions.

T rill Duration and Interval. In species with broken trills the durations of the trills and the intervals between trills were measured to the nearest one-fourth second by playing recordings of the songs at one-fourth normal speed and noting the beginnings and ends of trills

to the nearest second. Variations in tape speed have the same effect here as in other measuremehts, and the same correction factors were applied. Trill duration and interval are of little use in distin­

guishing the species within the major song categories.

Chirp Hate. The rate of chirping of 0. rileyi is expressed in

chirps per minute, which w ill hereafter be abbreviated as ch/m.

Chirp rate was determined by measuring the time to the nearest 0.1 32 second (the nearest one-fifth second at one-half speed) from the beginning of the first chirp to the beginning of the last chirp in a series of fifty chirps or more. The same correction factors wbre applied here as were applied to pulse-rate values. Chirp rate is fairly uniform for all individuals at the same temperature.

Chirp Duration and Interval. Chirp duration was calculated from the pulse rate by finding the duration of a pulse plus its interval

(p7s^ “^tiplying by 8 (number of pulses in a typical chirp). It is expressed to the nearest millisecond. Since this procedure makes chirp duration include the time of the silent opening motion of the first wing stroke, it should give a higher value than would be ob­ tained by measuring the chirp on an audiospectrogram. However, in rileyi the interval between pulses is quite short compared to the dura­ tion of the pulses, so error from this source is slight. This method of determining chirp duration has the advantage of giving more consistent results than the method of direct measurement, which is subject to the same difficulties encountered in measuring pulse duration. In a comparison of the two methods, chirp duration was measured directly and on the basis of pulse rate from six audiospec­ trograms showing chirps with definite beginnings and ends. The deviations of the calculated chirp durations from the measured ones were -3, 0, *7, -7, -4, *2 milliseconds out of 150, 219, 136, 146,

143, and 148 milliseconds respectively.

Chirp interval in milliseconds was found by subtracting the chirp duration from . The usual correction factors were applied to ch/m both chirp duration and chirp interval. Intensity. No study was made of the intensities of the songs of the various species, the principal reason being that no instrument was available that would measure intensity accurately and in recognized units. M. 0. Busnel (1954) reported intensities for 0. pellucens of

47 to 70 decibels when measured in the axis of the microphone at a distance of 10 centimeters. While no objective data on intensity are available for North American species, it is perhaps worthy of report that the loudness of the calling songs of the various species observed. in this study varies somewhat in proportion to the area of the tegmina.

On this basis, 0. latipennis has the loudest song and 0. celerinictus the softest. Another aspect of intensity deserves mention here, partly because of conflicting accounts in the literature. Williams (1945) described observations on 0 . argentinus in which he noted that highly directional sound waves were sent out perpendicular to the plane of the elevated tegmina. In other words, the song was much louder directly in front of and behind the singing cricket than in other positions at the same distance. Pasquinelly and M. C. Busnel (1954), however, indicated that the songs of crickets are loudest at each side of the insect. They compared the situation to that of a tuning fork, but even the analogy is wrong because the planes of loudest souncj. from a tuning fork are approximately at 45° rather than 90° to thelpline passing through the prongs. My own o b se rv a tio n s on eastern ! tr e e crickets unquestionably confirm Williams's conclusions. In the field, a cricket that is very loud from before or behind may be scarcely audible at the side. The periodical changes in orientation that a 34:

singing cricket makes may be related to this directionality of sound propagation.

Effects of Environmental Factors

The effects of environmental factors, especially temperature, upon the songs of insects have aroused interest for many years.

However, before the recent advances in recording and machine analysis

of sounds, only the few songs with slow, regular rhythms were suitable

for detailed study. With the Vibralyzer it is an easy matter to study variations which cannot be measured by the human ear. The effects of

temperature, humidity, air currents, light, and sound upon the nature

of the calling songs of eastern tree crickets were observed in this

study. Each of these environmental factors w ill be considered sepa­

rately in the following discussion.

Temperature

It has long been known that the chirp rate of the snowy tree

cricket (Oecanthus riley i, formerly misidentified as 0 . niveus) varies with changes in temperature. Same knowledge of this effect evidently

existed in folklore before it came to the attention of naturalists.

The first mention in scientific writing was by Brooks (1882) who

stated that she had seen in the Salem Gazette this formula for esti­

mating the temperature by the number of cricket chirps per minute:

’’Take seventy-two as the number of strokes per minute at 60° tempera­

ture, and for every four strokes more add 1°, and for every four

strokes less deduct the same.” This can be expressed by the formula T = .25 N + 42, where T_is the temperature in degrees Fahrenheit and

N is the number of chirps in one minute. This manner of expressing

the relationship of temperature and chirp rate is the reverse of the

customary way of relating an independent variable (temperature) and a

dependent variable (chirp rate); however, the literature has emphasized

the approximation of temperature by the determination of chirp rate,

so this form is used. Brooks made twelve observations which roughly

confirmed the formula. Although she failed to identify the species

of cricket concerned, it was doubtless 0. riley i. Dolbear (1897) was

the next to report on the "thermometer cricket,” and he too failed to N - 40 identify the singer. He proposed a formula T = 50 ■* — ----- , w hich i s

easily reduced to T = .25 N + 40. Bessey and Bessey (1898) made

observations at Lincoln, Nebraska, and definitely identified their

thermometers as snowy tree crickets. They listed 54 observations

made at temperatures from 54^° to 83° F. From these they derived the W - Q? formula T = 60 ♦ — -■■■ ■ — , which reduces to T - .21 N * 40.4. Edes 4.7 (1899) published some temperature data on rileyi taken by Walter Faxon

and him in New England. If Edes had proposed a formula to fit the

data, it would have been approximately T = .25 N ■* 38. In 1907, Shull

published on the effect of temperature on the song of rileyi. He

found a general correlation but emphasized that the relationship was

far from exact.

Fulton (1925) reviewed previous work and presented considerable

new data. Working in Oregon he found two "races” (probably distinct

species) of snowy tree cricket, which differed in song and other

aspects of their biology. At any given temperature, race "A" had a 36 higher chirp rate than race "B.” Kace "A" corresponds to the rileyi of the East. In observations made in the field and in the laboratory on the effects of temperature on the songs of the two races, Fulton found that each showed an essentially linear increase in chirp rate with temperature. The data for race "A" approximately fit the formula

T = .21 N + 37.6, while the data for race "B” approximately fit the formula T = .34 N ♦ 39. At temperatures over 80° F. (four observa­ tions), chirp rate in race "A” increased faster than predicted by the form ula.

Since Fulton’s work, four w riters have presented new observations on the effects of temperature on the chirp rate of the snowy tree cricket. Allard (1930a) gave data collected at Clarendon, Virginia; between 55 and 70° F. his observations fall approximately along the line T = .20 N ■* 43. Matthews (1942) made observations at Detroit,

Michigan. His data indicate a gradual increase in rate of increase of chirp rate with increasing temperature. Hallenbeck (1949) gave a formula equivalent to T = .23 N * 42, but he did not indicate in what geographical locality he made his observations. Alexander (1956) in observations made chiefly in Ohio and Illinois found an essentially linear relationship between 96 ch/m at 58° and 184 ch/m at 78° F.

That the songs of all Grthoptera change with changes in tempera­ ture has not been so long recognized. Fulton (1925), in discussing the song of _ 0 . rileyi, was the first to indicate such a general rela­

tionship. He stated, ”This relation of wing movement to temperature

is present in all singing insects which the writer has observed.”

Various authors, working with one or a few species of insects, have 37 substantiated the slowing of song with lowering temperature. Hayward

(1901) reported on Pterophylla cam elllfolia (Fabricius); Allard (1929b) reported on Orchelimum agile De Geer; Pulton (1931, 1933), Nemobius spp.; Pierce (1948), Nemobius fasciatus (De Geer); Hallenbeck (1949),

Acheta sp.; and Prings and Prings (1957), Neoconocephalus ensiger

( H a r r is ) .

The most significant contribution to this field is the recent work of Alexander (1956), who studied most of the singing Orthoptera of eastern United States. With the aid of tape recordings and audio-

spec trogrems, he concluded that there was ”a significant variation in

song characters with temperature change” in every species which he

observed at two or more different temperatures. For thirteen species

representing seven genera and six subfamilies, he had enough observa­

tions over a wide enough range of temperatures to show essentially

linear relationships between temperature and wing-stroke rate.

Included among these species were Oecanthus nigricornis, 0. quadri-

punctatus, and £. rileyi.

Whether the relationship between temperature and change in wing-

stroke rate is generally linear or not has been subject to some debate.

No mechanism has been hypothesized that would cause one to expect a

linear relationship, yet nearly all data available fall along a

straight line for each species. The exceptions to this relationship

are of two sorts. In some cases the top of the temperature response

curve is less steep than the mid-section, that is, the rate of increase

in song speed with increasing temperature diminishes at above-average

temperatures. For example, the song of £. latipennis above 85° P. has 38 a lower pulse rate than would he expected on the basis of a linear relationship (Fig, 24). Fulton's (1925) data on 0. rileyi showed an increase in rate of increase in chirp rate at temperatures above 80° F., and Matthew's (1942) data for the same species showed an increasing rate of increase throughout the singing range (55-85° F.). However, the chirp rate data for rileyi collected under controlled conditions in this study (Fig. 44) show no such trend; in fact, in terms of pulse rate (Fig. 23) the rate of increase may decrease slightly at higher temperatures.

The second sort of exception to a linear relationship between song speed and temperature is found in cases in which the bottom of the temperature response curve is less steep than the mid- 3e c tio n , that is, the rate of decrease in song speed with decreasing temperature diminishes at below-average tengjeratures. Such a relationship was found by Frings and Frings (1957) in Heoconocephalus ensiger below

60° F. Frings and Frings fitted their data to a logrithmic relation­ ship; however, their data above 60° F. fit a linear relationship just a s w e ll.

A hypothesis that would account for all the evidence available is that if insects could be stimulated to sing at all temperatures within their range of survival, changes in song in relation to temp­ erature would describe sigmoid curves. However, most insect species normally sing only at temperatures corresponding to the linear portions of the curve. A few species occasionally sing at temperatures above or below the range in which there is a linear response. The only denial of an effect of temperature upon the song of a trea cricket has came from the Stench worker, R. G-. Busnel (1954).

He stated that he found no consistant variation in the song of 0_. pellucens Scopoli with changes in temperature; however, he made his judgement on the basis of a study of pitch rather than pulse rate.

Variations in temperature do not always result in uniform variations in pitch (Rig. 35). When the pulse rate of pellucens is studied, it will probably prove to change with tanperature as it does in all other species of insects which have been so studied.

In exploring the effects of temperature upon the calling songs of eastern tree crickets, I made 515 recordings of 154 tree crickets in a controlled temperature room. All but two of the eleven eastern species were successfully recorded at several temperatures. The room could be held thermostatically at temperatures from 58° to 100° F.

At any setting, the temperature was constant *1° F. The crickets to be tested were confined in individual compartments (2-g- x 5 x 5 inches) in wooden testing boxes consisting of five such compartments (Fig.

47). Each compartment had a glass back which admitted light. The lid of each compartment had a corked one-inch-diameter hole, the cork being removed before the cricket within was recorded. Ventilation was provided by two holes in the front of the compartment and by

spaces between the lid and the walls. A few crickets were recorded while confined in individual transparent plastic rearing cups. These cups were 3g inches in diameter and 2^- inches deep; the lids were provided with 2-§-inch-diameter, gauze- or screen-covered vents. In

each compartment and cup, there were a small dish of ground dog food and a four-dram vial of water with, a dental wick inserted through, the cork. The compartments were furnished with strips of corrugated cardboard, 4 x 8 inches, which served as roosts. At least once a week, bits of lettuce were added and occasionally live aphids were given to the crickets. Under these conditions some individuals remained alive two months and longer with no apparent ill effects.

Comparisons of calling songs recorded in the laboratory with those recorded in the field showed no consistent differences.

The lights in the room were controlled by a time switch, so that they were on 16 hours and then off 8 hours during each 24 hours.

Recordings were made during the dark period with the Magnecorder.

The microphone was held in the open hole in the lid of the compartment in which a cricket was singing, and a recording of at least 15 seconds was made. The temperature of the air of the compartment was then measured to the nearest degree Fahrenheit with a copper-constantan thermocouple and a potentiometer (Leeds and Northrup Co., 4901 Stentor

Avenue, Philadelphia 44, Pennsylvania) (Pig. 47). In some experiments an attempt was made to read temperatures to the nearest one-half degree Fahrenheit, but this proved impractical. Fahrenheit was used because the potentiometer read directly in that scale. The humidity was not controlled, but it was measured in each compartment with an

Amineo-Dunmore electric hygrometer (American Instrument Co., Inc., S il­ ver Spring, Maryland) (Fig. 47) and was found to vary between 50 and

75 per cent relative humidity. During each recording period, every

cricket which was found singing was recorded. Afterward the tempera­ ture of the room was increased or decreased five or ten degrees, and 41 before the next recording period the crickets had at least eight hours

in which to reach equilibrium with the new room temperature. Record­

ings were made over as wide a range of temperatures as possible for

each species. None sang above 100° F ., and the cooling unit would not

cool the roam below 58° F. (and usually not that low).

Recordings obtained in this -manner were analyzed for the charac­

teristics described above (pp. £8-33). Table III lists for each

species the number of individuals recorded, the localities in which

they were collected, the number of recordings made, and the range of

temperatures in which recordings were made. As previous work indi­

cates, in all species the pulse rate (wing-stroke rate) is drastically

affected by temperature (Figs. 21-30). Between 60° and 80° F ., the

pulse rate nearly doubles in each of the ten species tested. The

data for each species show an essentially linear effect of temperature

upon pulse rate. Accordingly, the regression lines of pulse rate on

temperature were calculated and are drawn in on each of the graphs.

Only in _0. rileyi and 0. latipennis is there any suggestion of a

deviation from a linear relationship. In these two species (Figs. 23

and 24) the points at each end of the regression line are all less

than the expected values. This would indicate a slight decrease in

rate of pulse-rate increase with increasing temperature; however,

additional recordings at the extremes of temperature w ill be necessary

before this possible deviation from a linear relationship can be

shown to be typical for these species rather than a matter of chance.

Another characteristic which varies with temperature in all

species studied is the pitch of the song. If the pitch is determined 42

Table III. Laboratory recordings made in the study of effects of temperature upon calling songs.

Ho. No. Temp. L o c a lity S p ecies In d iv . Recordings Range °F . 0. exelamationis Franklin Co., Ohio 4 15 70-87^

0 . n iv eu s E rie C o ., Ohio 2 3 69-79 Franklin Co., Ohio 4 14 63-84 T o tal 6 17 63-84

0 . r i l e y i E rie C o ., Ohio 3 5 66-88 Franklin Co., Ohio 14 32 64-88 T o ta l 17 37 64-88

0. latipennis Franklin Co., Ohio 15 57 58-09

0. nigricornis ty p ic a l Franklin Co., Ohio 14 40 61-89| slow -trilling Carrol Co., Ohio 1 7 62-94 Franklin Co., Ohio 2 7 61-76 T o ta l 3 14 61-94 willow form Franklin Co., Ohio 3 23 60-95 Shelby Co., Ohio 12 21 64-89 T o tal 15 44 60-95

0. celerinictus Pike Co., Ark. 10 39 61-93 Dyer C o ., Tenn. 1 1 76 Brunswick C o ., 7a. 1 3 75-89^ Southampton Co., 7a. 3 14 67-88 T o ta l 15 57 61-93

0. argentinus Union Co., Ind. 1 2 69-80 S c o tt C o., Ky. 1 1 80 Franklin Co., Ohio 17 71 59-94 Ross Co., Ohio 1 12 67-88 Dyer Co., Tenn. 12 42 60-93 T o ta l 32 128 59-94

0. quadripunctatus Union C o ., In d . 4 11 63-88 Franklin Co., Ohio 23 73 61-93 Shelby Co., Ohio 1 1 76 H a lifa x C o ., 7 a . 1 1 80 T o ta l 29 86 61-93

0 . p in i Wake C o., N. C. 1 1 80 Fairfield Co., Ohio 3 19 604—93 T o ta l 4 20 60g-93 100 90

40 PULSES PER SECOND 80 60 70 50 0 2 30 - EFFECT OF TEMPERATURE ON PULSE RATE RATE PULSE ON TEMPERATURE OF EFFECT 60 AOAOY RECORDINGS LABORATORY . EXCLAMAT10N1S 0. ER E FAHRENHEITDEGREES IT V Fgr 21 Figure FIATS V, 0 80 70 ORE F MATERIAL OFSOURCE ERSIN LINEREGRESSION FRANKLINO CO., OHIO Y = 2.030 X - 75.44 - X 2.030 Y = 90 43 PULSES PER SECOND 100 40 0 2 60 80 90 50 30 70 FET F EPRTR O PLE RATE PULSE ON TEMPERATURE OF EFFECT 60 AOAOY RECORDINGSLABORATORY ERE FAHRENHEITDEGREES AE 1 Fgr 22 Figure 71, HATE 0 80 70 . NIVEUS 0. SOURCE REGRESSION LINE FRANKLINO ERIE • = .8 X73.97 - Y 1.889 = or or CO., material 90 OHIO CO., OHIO

44 PULSES PER SECOND 100 40 50 80 60 90 20 30 70 EFFECT OF TEMPERATURE ON PULSE RATE RATE PULSE ON TEMPERATURE OF EFFECT 60 AOAOY RECORDINGS LABORATORY 23 Figure VII, S T A H ER E FAHRENHEITDEGREES 70 . RILEYI 0. o o 80 SOURCE OF MATERIAL MATERIAL OF SOURCE ERSIN LINE REGRESSION FAKI CO., OHIO FRANKLIN O RE O, OHIO CO., ERIE • a .4 X 43.94 - X 1.243 a Y 90 45 PULSES PER SECOND lOOr 80 70 50 40 60 90 30 20 EFFECT OF TEMPERATURE ON PULSE RATE RATE PULSE ON TEMPERATURE OF EFFECT 60 AOAOY RECORDINGS LABORATORY 24 Figure VIII, E T A H ER E FAHRENHEITDEGREES 70 . LAT1PENN1S 0. ° 80 Q, ORE F MATERIAL OF SOURCE REGRESSION LINE LINE REGRESSION FAKI C. OHIO CO., FRANKLIN O * .9 X 53. 4 .6 3 5 - X 1.397 Y* 90 46 100 40 PULSES PER SECOND 50 60 80 90 20 70 30 FET F EPRTR O PLE RATE PULSE ON TEMPERATURE OF EFFECT SOURCE OF MATERIAL MATERIAL OF SOURCE 60 O FRANKLIN CO., OHIO OHIO CO., FRANKLIN O ARL C. OHIO CO., CARROLL • V AOAOY RECORDINGS LABORATORY ERE FAHRENHEIT DEGREES L T I, iue 25 Figure ELATE IX, 70 Q. NIGRICORNIS 80 REGRESSION LINES LINES REGRESSION TYPICAL SLOW-TRILLING 187 - 78. 4 .4 8 7 - X 1.897 « ? - .7 X 60.38 - X 1.477 Y - 90 47 0 4

PULSES PER SECOND 100 0 5 0 3 60 20 80 90 70 - - - EFFECTS OF TEMPERATURE ON PULSE RATE RATE PULSE ON TEMPERATURE OF EFFECTS SOURCE OF MATERIAL MATERIAL OF SOURCE 0 0 90 0 8 70 60 FAKI C. OHIO CO., FRANKLIN O HLY O, OHIO CO., SHELBY • . IRCRI. LLOW FORM W O L IL W NICRICORNIS. 0. AOAOY RECORDINGS LABORATORY ERE FAHRENHEIT DEGREES PLATE 26 Figure PLATE X, on io s s e r g e r L DATA ALL SHELBY CO, OHIO OHIO CO, SHELBY OHIO CO, FRANKLIN 190 - 61.00 - X 1.960 Y> Y » 1.776 X - 66.14 66.14 - X 1.776 Y» LINES

48 100 40 90

20 PULSES PER SECOND 30 60 80 50 70 FET F EPRTR O PLE RATE PULSE ON TEMPERATURE OF EFFECT ORE F MATERIAL OF SOURCE 070 60 IE O, ARKANSAS CO., PIKE A DE C. TENNESSEE CO., DYER □ SUHMTN O AND CO. SOUTHAMPTON B RNWC C. VIRGINIA CO., BRUNSWICK AOAOY RECORDINGSLABORATORY ERE FAHRENHEITDEGREES . CELERIN1CTUS 0. 27 Figure ELATE XI,

80 REGRESSION LINES LINES REGRESSION IE O, ARKANSAS CO., PIKE DATA ALL OTAPO C. AND CO. SOUTHAMPTON RNWC C. VIRGINIA CO., BRUNSWICK - 180 - 78. 2 .5 8 7 - X 1.860 - Y * .8 X 81.10 - X 1.884 * Y = .7 X 79.01 - X 1.873 = Y 90 49 lOOr

40 PULSES PER SECOND 50 30 80 60 20 70 90 - ' ' FET F EPRTR O PLE RATE PULSE ON TEMPERATURE OF EFFECT SOURCE OF MATERIAL MATERIAL OF SOURCE 60 NO C. INDIANA CO., ▲UNION □ DYER CO., TENNESSEE TENNESSEE CO., DYER □ OHIO CO., FRANKLIN O SOT O KENTUCKY CO. SCOTT ■ OHIO CO., ROSS • AOAOY RECORDINGS LABORATORY 28 Figure XII, S T A H ER E FAHRENHEIT DEGREES 0 80 70 . ARGENTINUS 0. REGRESSION LINES LINES REGRESSION Y R O TENNESSEE CO, DYER DATA ALL RNLN O, OHIO CO., FRANKLIN Y - 1.420 X - 57.86 57.86 - X 1.420 Y - = .8 X 88 .8 4 5 - X 1.383 = Y 7 .8 9 -5 X 1.439 Y = 90 50 51

H A T E XIII, Figure 29

EFFECT OF TEMPERATURE ON PULSE RATE 0. QUADRIPUNTATUS LABORATORY RECORDINGS 100

90 SOURCE OF MATERIAL O FRANKLIN CO., OHIO ▲ UNION CO., INDIANA • SHELBY CO., OHIO B HALIFAX CO, VIRGINIA 80 REGRESSION LINES ALL DATA ? « 1.094 X - 43.22 FRANKLIN CQ, OHIO 0 70 ? • 1.093 X - 43.05 UNION CQ, INDIANA

U 6 0

u

3 50

40

30

20

60 70 80 90 DEGREES FAHRENHEIT lOOr 40

PULSES PER SECOND 80 90 30 20 EFFECT OF TEMPERATURE ON PULSE RATE RATE PULSE ON TEMPERATURE OF EFFECT 60 AOAOY RECORDINGS LABORATORY 30 Figure XIV, E T A H ERE FAHRENHEIT DEGREES 70 . PIN! Q. 80 ERSIN LINE REGRESSION SOURCE OF MATERIAL MATERIAL OF SOURCE • FAIRFIELD FAIRFIELD • WK C. N. CAROLINA NO. CO., WAKE U 90 CO., OHIO OHIO 53 in the manner proposed in the discussion on pp. 16-18, changes in pitch are caused by changes in wing velocity during the time the scraper and file are engaged. If these changes were exactly dependent upon changes in wing-stroke rate, the changes in pitch with tempera­ ture would exactly parallel the changes in pulse rate. Such a con­ dition does not exist, and this is at least in part due to unequal increases in the speed of sound-producing and non-sound-producing portions of the wing-motion cycle with increases in temperature (pp.

24-25). Figures 31 to 40 show pitch plotted against pulse rate for each of the ten species studied. In Fig. 34 the dotted line indicates the theoretical rise in pitch with wing-stroke rate that would occur if the velocity of all parts of the wing-motion cycle increased with temperature at the same rate as the pulse rate. Hie solid line, drawn in by eye, indicates the actual trend of the observations on the relationship between pulse rate and pitch. All species are alike in that the actual rise in pitch is less than the theoretical. In most species there is a steadily decreasing rate of increase in pitch with increasing pulse rate. In 0. niveus, _0. celerlnietus, £. argen-

tinus (Figs. 32, 37, and 38), and possibly others, this decrease in rate of increase is exceedingly slight, and the relationship approaches

a linear one. On the other hand, the pitch of the song of the typical

foim of 0. nigricomis rises with increases in pulse rate up to 75 p/s

but rises slightly or not at all with further increases in pulse rate

(Fig. 35). Pitch was plotted against pulse rate rather than tempera­

ture because some variation due to error in temperature measurement

could be eliminated by this procedure. Pitch is not caused by pulse 2.0 4.0 3.0 4.0

KILOCYCLES .5 2 PER SECOND KILOCYCLES PER SECOND 3.0 2 . 0 - Fig. EAINHP F US RT AD PITCH AND RATE PULSE OF RELATIONSHIP EAINHP F US RT AD PITCH AND RATE PULSE OFRELATIONSHIP SOURCE o r MATERIAL MATERIAL r o SOURCE ORE F MATERIAL OF SOURCE O FRANKLIN FRANKLIN O O. EXCLAMATIONIS. LABORATORY RECORDINGS 50 50 NVU. AOAOY RECORDINGS LABORATORY NIVEUS. & CO., OHIO D O CD 60 USS E SECOND PER PULSES USS E SECOND PER PULSES 0 70 60 PLATE X7 PLATE 70 00 0 8 0 8 oo oo o o 90 90 54 100 100 2.0

KILOCYCLES PER SECOND3.5 4.0 2.0 KILOCYCLES PER SECOND 4.0 3.0 USS E SECOND PER PULSES 3 3 . g i F EAINHP F US RT AD PITCH AND RATE PULSE OF RELATIONSHIP EAINHP F US RT AD PITCH AND RATE PULSE OF RELATIONSHIP 040 4 30 10 0 o o . AIENS LBRTR RECORDINGS LABORATORY LATIPENNIS. Q. 0 8 0 4 . IEI LBRTR RECORDINGS LABORATORY RILEYI, O. yy oo USS E SECOND PER PULSES XVI E T A H 50 eo ORE O MATERIAL OF SOURCE. SOURCE OF MATERIAL MATERIAL OF SOURCE O FRANKLIN FRANKLIN O O FRANKLIN FRANKLIN O 080 8 70 0507 0 6 CO., CO., OHIO OHIO 55 56

PLATE X m

RELATIONSHIP OF PULSE RATE AND PITCH O. NIGRIC0RNI3. LABORATORY RECORDINGS

5.0 SOURCE OF MATERIAL SLOW-TRI l LING • CARROLL CO., OHIO a • FRANKLIN CO, OHIO § 4 .5 TYPICAL u O FRANKLIN CO., OHIO LJ m a. u 0.4.0 uV) u >- u 0 3 .5

3.0

3 0 4 0 5 0 60 70 8 0 9 0 F ig. 35 PULSES PER SECOND RELATIONSHIP OF PULSE RATE AND PITCH Q. NIGRICORN IS. WILLOW FORM, LABORATORY RECORDINGS

5.0 SOURCE OF MATERIAL O FRANKLIN CO., OHIO • SHELBY CO, OHIO

u oo

5.4.0 00

3.0

4 0 50 60 70 8 0 90 100 F ig . 36 PULSES PER SECOND .5 4 .5 2 .0 4

KILOCYCLES PER SECOND 3.0 KILOCYCLES PER SECOND 4.5 3.0 3.5 5.0 g, 37 , ig P Fig, 38 Fig, EAINHP F US RT AD PITCH AND RATE PULSE OF RELATIONSHIP EAINHP F US RT AD PITCH AND RATE PULSE OF RELATIONSHIP 0 4 Q. CELERINICTUS, LABORATORY RECORDINGS Q. RETNS LBRTR RECORDINGS LABORATORY ARGENTINUS. 0 5 PLATE XVTII PLATE USS E SECOND PER PULSES USS E SECOND PER PULSES 07 0 6 70 60 0 6 0 5 ORE F MATERIAL OF SOURCE ORE F MATERIAL OF SOURCE a a OTAPO C. AND CO. SOUTHAMPTON a PIKE A SOT O KENTUCKY CO, SCOTT ■ OHIO CO, ROSS • a O NO C, INDIANA CO, UNION A DYER FRANKLIN FRANKLIN DYER CO, TENNESSEE TENNESSEE CO, DYER BRUNSWICK CO., CO., ARKANSAS TENNESSEE CO., CO., OHIO OHIO VIRGINIA 90 80 57 KILOCYCLES PER SECOND KILOCYCLES PER SECOND 4 5 - 5 4 . - 4.0 4.5 4.0 .5 2 g. 39 . ig F Fig. 40 Fig. EAINHP F US RT AD PITCH AND RATE PULSE OF RELATIONSHIP EAINHP F US RT AD PITCH AND RATE PULSE OF RELATIONSHIP 30 0 3 . UDIUCAU. AOAOY RECORDINGS LABORATORY QUADRIPUNCTATUS. Q. 50 0 4 . I . AOAOY RECORDINGS LABORATORY I. PIN O. 0 4 oo USS E SECOND PER PULSES USS E SECOND PER PULSES XLX E T A H 50 SECOND 60 60 SOURCE OF MATERIAL MATERIAL OF SOURCE ORE F MATERIAL OF SOURCE 4 UNION UNION 4 HALIFAXB O • SHELBY SHELBY • AK C, N CO, AAKE Q FRANKLIN FRANKLIN 70 70 CO., CO., CO., INDI ANA CO., OHIO OHIO VIRGI NIA VIRGI OHIO OHIO 0 8 0 8 58 59 rate, but variations in both wing-stroke rate and in velocity of the tegmina while file and scraper are engaged are determined by temperature.

I n ,0 . exclamationis and 0 . niveust which produce broken trills, no correlation xvas evident between temperature and (l) trill rate,

(2) trill duration, or (3) trill duration/trill duration ♦ interval.

Figures 41, 42, and 43 show the data for these factors plotted against temperature. Data frcm field recordings are also included, so as to increase the number of observations. If temperature has any effect upon these characteristics it is extremely slight or irregular.

The extensive literature on the effect of temperature on the chirp rate of £. rileyi was reviewed above. Data collected in this study are shown in Figure 44. As other authors have indicated, the relationship between temperature and chirp rate is linear.

While the usual number of pulses per chirp in 0 . rileyi is eight, chirps containing fewer pulses are frequently produced and occa­ sionally chirps with more pulses are produced. Figure 45 shows the average number of pulses per chirp in the first fifty chirps of each recording made in the study of the effect of temperature upon the song of riley i. The deviations from the average condition are apparently not related to temperature.

The time from the beginning of one chirp to the beginning of the next becomes less with increases in temperature, since the chirp rate is increasing. Likewise, the duration of an eight-pulse chirp becomes less with increases in temperature, since the pulse rate is increas­ ing. The interval between chirps also becomes less with increases in EFFECT OF TEMPERATURE ON TRILL CHARACTERS CHARACTERS TRILL ON TEMPERATURE OF EFFECT PER CENT SECONDS TRILLS PER MINUTE 100 0 4 20 80 30 20 60 I 0 g 41 ig. F g. 3 ERE FAHRENHEIT DEGREES 43 . ig F 50 f ^ 2- l N . XLMTOI AD . NIVEUS 0. AND Q. EXCLAMATIONIS IN R3 IL RECORDING FIELD • o o XLMTOI Q. NIVEUS EXCLAMATIONIS AOAOY RECORDINC LABORATORY cent

of V RG TIL DURATION TRILL AVERAGE 0 7 0 6

singing PLATS XX RL RATE TRILL t

i e tim A ▲ A

t n e p s a a IL RECORDING FIELD

AOAOY RECORDING LABORATORY ▲ ▲

in

80 lli g in l il r t 60 90 CHIRPS PER MINUTE 240 200 220 0 4 0 6 20 00 0 8 0 8 FET F EPRTR O CIP RATE CHIRP ON TEMPERATURE OF EFFECT 60 AOAOY RECORDINGS LABORATORY H A T E X U , Figure Figure 44 , U X E T A H ERE FAHRENHEIT DEGREES 0 7 . B1LEYIQ. 0 8 SOURCE OF MATERIAL MATERIAL OF SOURCE ERSIN LINE REGRESSION O FRANKLIN FRANKLIN O • ERIE ERIE • » 560 X 184.53 - X 0 6 .5 4 » Y O., CO 90 OHIO CO., OHIO 61 62 PLATE XXII NUMBER OF PULSES PER CHIRP Q RILEYI. LABORATORY RECORDINGS Q.

10 a

< 5 60 8 5 70 75 80 85 90 DEGREES FAHRENHEIT

PER CENT OF SINGING TIME SPENT IN CHIRPING O. RILEYI. LABORATORY RECORDINGS 4 6

4 4

bJ u 4 2 a . U a

4 0

3 8 100 120 140 180 ISO 200 CHIRPS PER MINUTE — 1------1______I------1------L___ ------l - 60 65 70 75 80 85 F i g . 46 . APPROXIMATE TEMPERATURE C°F) 63 temperature, and it decreases more rapidly than chirp duration. Con­ sequently the proportion of the total time of a "chirp cycle" which is taken up by the chirp itself (as opposed to the interval) becomes greater with increases in temperature. This relationship is shown in

Figure 46, in which chirp duration/chirp-cycle duration is plotted against chirp rate. The line represents the theoretical relationship as calculated from the regression lines for chirp rate and pulse rate.

H um idity

In attempting to explain some of the variations noted in the chirp rates of individuals of 0, rileyi at the same temperature,

Shull (1907) suggested that humidity might affect the chirp rate. He had no apparatus for determining humidity, but he cited some data which he felt might be explained by assuming that high humidity causes a decrease in the rate of stridulation. Fulton (1925), working with the same species, confined one individual in a screen cage and another in a lamp chimney plugged with cotton and containing fresh green leaves. The relative humidity of the room was 54 per cent; in the lamp chimney, condensation indicated 100 per cent relative humidity.

The cricket in the chimney and the one in the screen cage chirped at the same rate whether they sang alone or in unison. Fulton concluded that the effect of humidity could not be greater than individual differences between crickets. Allard (1930a) determined humidities as well as temperatures in his field observations of the chirp rates of 0. rileyi. He recorded humidities from 78 to 100 per cent and found no correlation between humidity and chirp rate. 64

What evidence there is indicates that humidity does not affect chirp rate in 0 . rileyi; however, effects of low humidities have never been tested nor have the effects of humidity upon other song characteristics and other species been investigated. Consequently, I undertook to determine whether humidity had any influence upon the nature of the calling song in tree crickets.

The humidity experiments were done in the same controlled tem­ perature room as were the experiments with effects of temperature.

As before, the test crickets were confined in individual compartments in wooden testing boxes. This time, however, the humidity of the compartments was controlled by passing a stream of dried or dampened air into each compartment through a hole in its front (Fig, 47). The lids were sealed with masking tape so that the only vent was a small hole in the front of each compartment. Bach day half of the compart­ ments were kept at 93 per cent relative humidity or higher and the

other compartments were kept at 13 per cent relative humidity or lower.

The air which supplied the high-humidity ccmpartments was first bubbled through several bottles of water, and the air which supplied

the low-humidity compartments was first passed through a 36-inch-long,

1-inch-diameter glass tube containing anhydrous calcium chloride (Fig.

48). Recordings were made by placing the microphone against the small vent hole, so humidity conditions within the compartment were not

disturbed. After each recording, temperature and humidity were checked with the thermocouple and electric hygrometer. When the element of

the hygrometer had been inserted into the compartment through the

hole in the lid, the cork was replaced, and the humidity allowed to HATE XXIII

Fig. 47. Testing boxes as used to determine the effects of humidity upon calling songs. A. Electric hygrometer with sensitive element in a compartment. B. Potentiometer with thermocouple in a compartment.

Fig. 48, Apparatus controlling humidity of air entering compartments in Figure 47. A. Compressed air source. B. Pint mason jar. C. Pinchcock controlling air pressure. D. Air line to high- humidity compartments. E. Air line to low-humidity compartments. F. Bottles of water. G. Tube containing anhydrous calcium chloride. 66 regain equilibrium before the reading was taken. After each day’s recording period, the air lines were switched so that high-humidity compartments received low-humidity air and vice versa. The tempera­ ture of the room was set at 75° F .; however, the temperatures within the compartments varied from 73§-° to 76^° during the several runs of the test. At first the low-humidity compartments were about 1° F. colder than the high-humidity compartments, largely because of the

cooling effect of evaporation from the wicks of the watering vials.

The evaporation from the vials was not sufficient to raise the humidity above 13 per cent in the low-humidity compartments, but in

later experiments the vials were removed before low-humidity runs to

prevent the cooling effect.

Data at both high and low humidities were obtained for several

individuals of each of three species — 0 . nigricomis. 0 . celsrinic-

tus, and O', quadripunetatus. The effect upon pulse rate in these

species is indicated in Figures 49, 50, and 51. In these graphs

pulse rate is indicated in terms of the amount of deviation from the

expected values as calculated from the formulas for the regression

lines obtained in the study of temperature effects. It should be

recalled that the recordings upon which these lines are based were made at 50 to 75 per cent relative humidity. Therefore, if there is

any difference among pulse rates of calling songs at high, medium, and

low humidities, it would show up in the graphs. No consistent dif­

ference is evident, so any effect of humidity on the pulse rate is

slight. A similar analysis of the data showed no effect of humidity

on the pitch of songs of these three species. FET F EAIE UIIY N US RATE PULSE ON HUMIDITY RELATIVE OF EFFECT DEVIATION FROM EXPECTED PULSE RATE (PULSES PER SECOND) ♦ 5r +5 r +5 ♦ 5 fg 5 7 7 7 77 76 75 74 'fog. 51 og. 49 . g fo 50 . g o f

E. UIIY R IHR RL HMDT 1% O LOWER OR 13% HUMIDITY REL. o HIGHER OR % 3 9 HUMIDITY REL. • 1 o o 8 O O 0 o *- • ------4 INDIVIDUALS, UNION CQ, INDIANA; 7, FRANKLIN CO, OHIO CO, FRANKLIN 7, INDIANA; CQ, UNION INDIVIDUALS, 4 * * 8 o o o o

8 INDIVIDUALS, FRANKLIN FRANKLIN INDIVIDUALS, 8 7 INDIVIDUALS, PIKE PIKE INDIVIDUALS, 7 - 4 ® 8 o Q. QUADRIPUNCTATUS ERE FAHRENHEIT DEGREES CELERINICTUS Q Q. NIGRICQRNIS XXIV E T A H I

AVERAGE OEVIATION FROM EXPECTED EXPECTED FROM OEVIATION AVERAGE AVERAGE DEVIATION FROM EXPECTED EXPECTED FROM DEVIATION AVERAGE AVERAGE DEVIATION FROM EXPECTED EXPECTED FROM DEVIATION AVERAGE CO., I ______I ______o +.89 P/S P/S +.89 o • +.77 P/S P/S +.77 • • - . 6 2 P/S P/S 2 6 . - P/S • 5 .2 - o • +.35 P/S P/S +.35 • o + .77 P/S P/S + .77 o ARKANSAS CO., OHIO I I ______l ------I 1 ------1 ------1 ------1 ----- CO 1 ----- 68

Table IV. Effect of high and low humidities upon the calling song of an individual of 0, r i l e y i .

i H el. Expected Expected p /s ch/m cps D ate 0E. Humid. p /s ch/m 27 Aug. 7 # 49 48 3 /4 162 155 2600

28 Aug. 74iV 95 50 48 3/4 164 155 2600

29 Aug. 7 3 | 11 48s 47 i 160 151 2600

30 Aug. 75 96 4&| 49 t 158 157 2700

31 Aug. 74 6 48§ 48 158 153 2700

Only one individual of Q_. rileyi sang during the humidity tests.

The data for this individual are listed in Table IV. No consistent differences exist in its song at high and low humidities. All chirp rates are higher than expected, but this is independent of humidity, and similar cases w ill be discussed under individual variation.

Air Currents

Allard (1930a) suggested that wind might affect the chirp rate

of Oecanthus rileyi by increasing evaporation from the cricket,

thereby cooling it and lowering the chirp rate. In a later article

Allard (1930b) reported on an experiment in which he placed a snowy

tree cricket on some raspberry shoots in his sleeping room. When the

cricket began to sing, he directed a small electric fan toward the

shoots. In seven trials, when the fan was turned on, the chirp rate

consistently increased from about 172 to about 185 ch/m. When he

directed the fan toward a cricket singing in a shrub outdoors, ha 69 produced no change in chirp rate.

In an effort to determine what effects win*! had upon the nature of the calling song in tree crickets, I placed individuals in the controlled-tamperature room in cylindrical cages 5 inches in diameter,

8f inches ta ll, and made of 14-mesh screening (Fig. 52). When a cricket began to sing, a recording was made, and then a small electric fan was directed toward the singer, and another recording was made.

The temperature within the cage was measured with a thermocouple before and after the fan was turned on. The air velocity was deter­ mined with a Hastings Air-Meter (Hastings Instrument Co., Inc.,

Hampton, Va.) (Fig. 52) by inserting the probe through a hole in the top of the cage after the recording with the fan turned on had been made. The air velocity was measured at the place in the cage where the cricket had been singing. When the fan was off, there was an air velocity of about 15 feet/minute in the cage caused by the room's temperature-control apparatus.

In tests involving ten crickets, only three sang in the wire cages and continued to sing when the fan was turned on. The results of the analyses of the songs of these three individuals, of three species, are listed in Table V. In no case was the song with the fan turned on significantly different from the song with the fan off.

The minor differences recorded in Table V are no greater than differ­ ences between analyses of successive portions of a song delivered under constant conditions. Allard's (1930b) results with 0_. rileyi indoors may perhaps be explained on the basis that the air the fan circulated was slightly warmer than the air about the cricket with F ig. 52. Equipment fo r determining the e ffe c ts of a ir currents A. Small electric fan. B* Screen cage containing test cricket. C. Hastings Air-Mater.

Fig. 53. 0. rileyi mals, at Fig. 54. 0. rileyi male rest; dorsal view. singing; posterior view. 71

Table V. Effect of air currents upon the calling songs of three tree crickets. Relative humidity approximately 60$.

Wind S p ecies °F. p /s cps ch/m ft/m in

0 . nigricornis 77 15 67 i 3800 ------77 300 67^ 3800 —_

0 . latipennis 72 15 46-| 2800 —- 72 350 47 2800 ----

0 . rileyi 76 15 50g- 2700 166 77 200 5 l | 2700 169

the fan off. Allard did not indicate that he checked for local temperature variations between the two treatments.

L ight

Light intensity is important in determining when a tree cricket sint;a, and experiments with this phenomenon w ill be discussed in a later section. Whether light intensity influences the nature of the calling song is another question and w ill be dealt with here.

No controlled experiments were performed, yet the facts avail­ able strongly indicate that changes in light intensity do not cause appreciable changes in the calling songs of tree crickets. The best evidence in this connection is that recordings of members of the nigricornis group made in the field in bright daylight fall into the same temperature-pulse-rate-pitch patterns as do recordings made in the field at night (and as do recordings made in the laboratory).

Another bit of evidence, incidental to studying the mechanics of sound 72 production, is that the pulse rates and pitches do not change when singing crickets are viewed with a stroboscope.

Sound

Singing males of Oecanthus rileyi respond to certain sounds and sound patterns in their environments in such a way that their chirp rates (but not their pulse rates) are increased or decreased. The most frequent effect of such responses is synchronous singing in neighboring individuals. This phenomenon w ill be discussed in detail in the section dealing with behavioral effects of the calling song.

No work has been done as to the effects of sounds upon the nature of the calling song of other species.

Effects of Intrinsic Factors

While much of the variation in the calling songs of species of tree crickets can be ascribed to variations in the surroundings of the singer (especially to variations in temperature), there remains a considerable residue of variation which must be attributed to factors that are within the cricket. Intrinsic factors may be classed in various ways; in this discussion I have dealt with them under the major headings of individual variation and population variation.

Individual Variation

Dissim ilarities in calling songs between individuals of a local species population are sometimes independent of the immediate environ­ ment. For instance, if one listens to a number of individuals of a 73 species found in the same area at the same time and singing under similar environmental conditions, one can pick out consistent differ­ ences between the songs of individuals. Bessey and Bessey (1898) were the first to describe this sort of variation in the songs of tree crickets. They noted that at any given temperature certain individuals o f 0 , rileyi always chirped a few chirps per minute faster than certain other individuals. Shull (1907) confirmed this phenomenon in rileyi as regards chirp rate and also indicated that certain individuals sing at a consistently higher pitch than others. Shull tried to relate individual variation in chirp rate with wing length but found no correlation. Fulton (1925) emphasized the importance of slight indi­ vidual variations in accounting for the discrepancies in relating temperature and chirp rate in 0 . r i l e y i .

In every species in this study in which several individuals were recorded under a variety of conditions, consistent differences among the songs of individuals were discovered. For example, Figure 55 shows the relationship of pulse rate to temperature in two individuals

each of 0. argentinus and 0. quadrlpunctatus. In each species the

individuals were collected in the same place at the same time and were recorded under the same conditions; nevertheless, there are con­

sistent differences between the songs of the individuals of the same

species. These differences must be results of differences in genetic

constitution and/or major differences in previous environment. They may be compared to the deviations from average size one finds in a

local population of an insect. PULSES PER SECOND PULSES PER SECOND 0 6 0 4 30 0 4 0 3 50 0 5 60 70 g. 55 . ig P g. 56 . ig P FET F NIIUL AITO O PLE RATE PULSE INDIVIDUAL VARIATION ON OF EFFECT FET F G O IDVDA O PLE RATE PUL^E INDIVIDUAL ON OF AGE OF EFFECT . QUADRIPUNCTATUS Q. 0 6 60 _L ARGENTINUS Q NIIUL OLCE A NYMPH, AS COLLECTED INDIVIDUAL JULY IDVDA N. 2 NO. INDIVIDUAL I NO. ■ INDIVIDUAL O COLLECTED 2 6 SEPT. 1956 1956 SEPT. 6 2 COLLECTED FRANKLIN , 6 5 9 1 FRANKLIN FRANKLIN o 80 To 70 -© CO., OHIO PLATE XX7I PLATE ERE FAHRENHEIT DEGREES ERE FAHRENHEIT DEGREES CO., OHIO 90 0 8 O □ - QUADRIPUNCTATUS Q- D NIIUL O 2 NO. INDIVIDUAL I NO. • INDIVIDUAL O AE F RECORDING OF DATE COLLECTED AUGUST AUGUST COLLECTED FRANKLIN AUGUST 8 - 2 A I 30 SEPTEMBER 0 I -3 AUGUST O 15-19 □ ' a 90 CO., OHIO 1 9 5 6 , , 6 5 9 1 74 75

A second possible source of consistent individual variation not related to immediate environment is the age of the singer. The data collected in this study, however, reveal no consistent changes in the nature of the calling song with age. For example, Figure 56 shows the results of analyses of recordings made of the same individual of 0. quadrlpunctatus over a period of nearly two months. The relationship between pulse rate and temperature did not change noticeably during this period.

A third type of individual variation not correlated with immediate environment is due to temporary differences in physiological state, such as fatigue or a full midgut. While such differences probably cause slight variations in song, in these studies they are inseparable from variations due to errors in instrumentation (_e.jz., inaccurate determination of temperature and variations in recorder speed).

Population Variation

It is possible for the nature of the calling song in tree crickets to vary between populations of a species that are separated in time or in space. Information bearing upon both these aspects was gathered during this study.

The two generations of 0_. argentinus in Ohio are populations separated in time. The first generation matures about the first of

July and disappears in early August. The second generation reaches adulthood in September and lives until frost. No significant dif­ ferences were found in the calling songs of these populations. Figure

57 shows the relationship of pulse rate to temperature in laboratory PULSES PER SECOND 40 80 0 5 60 20 30 70 - EFFECT OF TEMPERATURE ON PULSE RATE RATE PULSE ON TEMPERATURE OF EFFECT 60 Q ARGENTINUS. FRANKLIN FRANKLIN ARGENTINUS. Q FIRST AND SECOND GENERATIONS GENERATIONS SECOND AND FIRST AOAOY RECORDINGS LABORATORY 57Figure XXVII, E T A H ERE FAHRENHEIT DEGREES 080 70 08 REGRESSION LINES LINES REGRESSION ORE F MATERIAL OF SOURCE CO., SCN GENERATION SECOND O IS GENERATION FIRST GENERATION FIRST □ EOD GENERATION SECOND * .3 X 59.88 - X 1.437 Y * OHIO 90 76 77 recordings of individuals of the two generations. The regression lines for the two sets of data are so nearly alike that they cannot be separated on the graph.

Geographical variation is the usual term for variations in popu­ lations separated in space. Stilton (1925) discussed geographical variation in the song of £. rileyi and pointed out that on the basis of the data of Edes (1899), Bessey and Bessey (1898), and his own, riley i chirps more slowly in New England than in Nebraska, and more

slowly in Nebraska than in Oregon. Stilton described a test in which he placed males of rileyi from Ohio, Arizona, and Oregon in the same room at 71° E. Within half an hour the following rates were recorded:

Ohio, 130 ch/m; Arizona, 140 ch/m; Oregon, 155 ch/m. From this evi­

dence, it seems that there is a clinal increase in chirp rate in (D. rileyi from east to west.

During this study, recordings were made of tree crickets from localities scattered throughout most of eastern United States, yet no major geographical differences in the calling song of any species were disclosed. The details of the analyses of recordings representing

different geographical areas will be discussed in the species write­ ups; however, two instances in which a large number of laboratory

recordings are involved bear mentioning here. Extensive recordings

of calling songs of individuals of 0. argentinus from Dyer County,

Tennessee, and from Franklin County, Ohio, showed no significant dif­

ferences (Figs. 28 and 38). Likewise, individuals of 0. celerinictus 78 from eastern Virginia and from western Arkansas produced calling songs that did not differ significantly (Figs. 27 and 37).

Relation of Sound Production to the Diurnal Cycle

In terms of their daily periodicity in singing, species of tree crickets may be separated into those which sing almost entirely at night and those which sing during both day and night. Those which ordinarily sing only at night are Neoxabea bipunctata, Oecanthus exclamationis, _0. niveus, 0 . rileyi, 0. latipennis, and _0. pini. The species which commonly sing during the day as well as at night are 0 . nigricornis, 0 . celerinictus, £. argentinus, and 0 . quadripunctatus.

No infomnation is available as to the singing periodicity of _0.

•garlcomis.

The distinction between the two groups is not absolute. The species which characteristically sing during day as well as at night sing more persistently at night; and on rare occasions, the night- singing foims sing sporadically during the day.

It is noteworthy that all species which sing only at night, with the exception of 0 . latipennis, are characteristically arboreal. 0 . latipennis is found in low shrubs or on coarse weeds and brambles, particularly along woodland edges. The four species which sing both day and night are typically inhabitants of weedy fields. They are closely related species, belonging to the so-called nigricornis group; however, one of the arboreal forms, 0 . pini, also belongs to the nigricornis group yet is a night-singing species. 79

The factors responsible for daily periodicity in singing are difficult to determine but the indications are strong that light intensity is the primary factor.

Field Observations

Field observations of the time at which 0. niveus and 0. rileyi began singing in the evening showed a better correlation with light intensity than with temperature or humidity, but they did not eliminate the possibility that a falling temperature and a rising humidity were also determining factors in the beginning of song. No systematized observations were made of the time at which any species ceased singing in the early morning. In general, the number of individuals singing diminishes as the night proceeds; but if the night is warm, some singing continues until dawn. On three occasions a chorus of 0 . r i l e y i was heard singing in the early morning when the sky was beginning to lighten. Tree crickets do not sing at temperatures much below 50° F ., and most singing ceases before this temperature is reached. In the fall, therefore, the usual nightly singing period becomes shorter, even though the nights are longer. The inhibition of singing at night by low temperatures may account for the increased frequency with which night-3inging species sing during the day in late fall.

Laboratory Observations

Observations on the singing periodicity of various species under laboratory conditions furnish further evidence as to the dominant role of light intensity. In the controlled-temperature room all diurnal 80 variations in temperature and humidity w ere eliminated, yet the ten species studied maintained the same song periodicity in relation to the light factor as they had in the field. The room was lighted with fluorescent lights which were turned on and off hy a time clock. The daily dark period was set at 8 hours, leaving a 16-hour light period.

These periods were approximately synchronized with field conditions in m id-3ummer.

That the tree crickets maintained the same periodicity in the controlled-temperature room as they did in the field is circumstantial evidence as to the role of light intensity in the determination of periodicity, hut it does not eliminate the possibility that the crickets have a built-in diurnal rhythm which operates in spite of environmental variations. In the latter part of the summer of 1956, to check on this possibility, I adjusted the dark period in the controlled-temperature room so that it began at 11 am and ended at

7 pm each day. The dark period in the room therefore corresponded to a light period in the field. The crickets adjusted immediately to the

"upside-down” day and sang in relation to the time the lights went on and off rather than in relation to outside light conditions. Thus, any intrinsic rhythm is secondary to the external rhythm of light in te n s ity .

In order to determine what sort of singing periodicity would be

exhibited in lone tree crickets in continuous light or continuous dark, I conducted a series of experiments in which single crickets were confined in a room isolated from all insect noises. The tempera­

ture and humidity were not controlled, but the room was a basement one 81 with thick concrete walls and with no openings to the out-of-doors.

Thermograph records taken during the experiments showed no daily fluctuations, although there were long-term rises and falls of tem­ perature corresponding to hot and cool spells out-of-doors.

To obtain a continuous record of the singing activity of the test cricket, I constructed a device which made daily graphs of the cricket's song periods. The apparatus is pictured in Figure 58. The cricket was confined in a battery jar into which a crystal microphone projected. When the cricket sang, a minute electric current was produced in the microphone, and this was amplified in a three-stage amplifier so that it became strong enough to operate a relay. The relay opened and closed a circuit which activated an electromagnet which moved a pointer. The pointer was part of a modified hydro­ thermograph with the drum set to revolve once per day. The am plifier was designed by Basil Parnes, Department of Physics. Figure 60 is a

diagram of the electronic circuits of the apparatus. Figure 61 shows

the nature of the graph produced. In reading the lines on the graphs as periods of singing and periods of silence, I did not record as a period of silence any break in singing of less than an hour. This

criterion was used because the major periods of singing were the ones

in question.

Only continuous light and continuous dark were used in conjunction with the monitoring apparatus. Continuous dark was produced by

placing a light-proof cardboard box over the battery jar containing

the cricket. Continuous light was obtained by placing fluorescent

lights on each side of the battery jar (Fig. 59). The light intensity HATE XXVIII

Fig. 58. Apparatus for making a continuous re card of a cricket’s singing periods. A. Battery jar in which cricket is confined; in practice, the jar is several feet from the other units of the apparatus. B. Crystal microphone. C. Am plifier and relay. D-F. Magnet power supply. D. Transformer* E» Capacitor. F. Rheostat, G. Modified hydrothermograph; magnet pulls lower pointer down during singing periods of cricket; drum revolves daily.

F ig . 59. Method of lig h tin g during periods of continuous light. INPUT ( . MG C 2. 430 20.0 OHMS C» 30 R|0 I MEG 4.7 R( ? 47000 OHMSR? I t 20 OM 2 1. 000 13.0 , C 2000 OHMS 2Rt . 1000OHMS R. .3 AU6 U 6A i. eo rig. i. 61 Fig. ICI DARM F OIOIG DEVICE MONITORING OF DIAGRAM CIRCUIT R P MD B MNTRN DEVICE MONITORING BY MADE GRAPH 2 4 -HOUR RECORD OF SINGING OF AN INDIVIDUAL INDIVIDUAL AN OF SINGING OF RECORD -HOUR 4 2 7 20.0 C7 4.0 .I 400R-0.0047 0.I 1(00 OHMS C AW S.0 MTD 130 V F . IEI N OTNOS DARK CONTINUOUS IN O. RILEYI OF 400 400 430 130 HC MOVES WHICH MAGNET POINTER AE XX 83 XXIX HATE HOURS 6v ‘ 5Y3GT IO A.C. IIOv » •fife-; 6.3v 20 5v 2 2 a a »IIOv AC. I 84 inside the jar was 125 foot candles as measured with a Model 603 illumination meter (Weston Electrical Instrument Corp., Newark, New

Jersey). In pilot experiments in 1955, the room lights were used, and the light intensity inside the battery jar was 19 foot candles.

Figure 62 shows graphically the results of a series of tests with an individual of 0. rileyi. The test cricket had been singing regularly during each 8-hour dark period in the controlled-temperature room. It was placed in continuous dark at the same time that the daily dark period began. For three days it sang during periods roughly equivalent to the dark periods in its fomer environment.

Then its time of singing became more irregular, but there were s till long periods of singing followed by long periods of silence. When the cricket was exposed to continuous light, it sang only twice and did not sing again until it was exposed to darkness.

Two other experiments with night-singing species gave sim ilar results. In one with 0. rileyi in September 1955, the test cricket was first exposed to continuous light (19 foot candles). In a 5§-day period, it sang only once and then for only 30 minutes. Within five minutes after it was placed in continuous dark, it began to sing and

continued singing for over 7 hours with only minor interruptions.

For the next three days in continuous dark it alternated silent

periods of about 16 hours (18, 16, 15) with singing periods of about

8 hours ( 6 , 5, ll). The singing periods were not continuous, but were

broken by periods of silence as long as an hour and a half. In August

1955, when an individual of 0 . niveus was kept in continuous darkness, 85 H A T E XXX, Figure 62 SINGING PERIODICITY OF AN INDIVIDUAL OF Q. RILEYI IN CONTINUOUS DARK AND CONTINUOUS LIGHT SINGING SILENT (PERIODS OF LESS THAN ONE HOUR NOT INDICATED) PREVIOUS LIGHT REGIME 2-18 AUGUST 1956 LIGHT

CONTINUOUS DARK 16-17 AUGUST

LS-2Q

2 2 - 2 3 aa- 24

2 * jl2S

CONTINUOUS LIGHT (125 F.C.)

2 7 - 2 6 AUGUST L .26- 29- 2 9 - 3 0 30-31 L 31 AUGUST - I SEPTEMBER I - d f ^ = = . 2 - 3______

CONTINUOUS DARK

3 SEPTEMBER______86 the first four days it sang only two times for a total of 6 h o u rs.

Then for the next three days it alternated periods of sporadic singing (9, 10, and 6 hours) with long periods of silence (14, 20,

16 h o u rs ).

These experiments indicate that individuals of 0 . niveus and 0 . rileyi have an imperfect singing periodicity independent of the

immediate environment. Whether or not this rhythm is determined by previous conditioning during adulthood is unknown. Exposure to continuous darkness reveals this rhythm; exposure to continuous light

suppresses it.

Two tests with the monitoring apparatus were made on species which commonly sing during daylight as well as at night. The results

of a test with an individual of 0 . nigricom is are diagrammed in

Figure 63, About the only indication of a trend is that there was

slightly more frequent singing during the periods of continuous dark­

ness than during the period of continuous light. In a similar test with 0. quadripunctatus (17 September to 6 October 1956), the cricket

sang sporadically during two periods of continuous darkness but did

not sing at all in two three-day periods of continuous light.

In the controlled-tsmperature room, many individuals of day-

singing species were never heard singing during the light period,

though they sang regularly during the dark period. On the other hand, some individuals were frequently heard singing during the light

period and also were heard during the dark period. Such individual

variation may occur in the field and would account for the reduced

volume of song of these species during the day. An alternative is 87 ELATE n n , Figure 63 SINGING PERIODICITY OF AN INDIVIDUAL OF 0. NIGRICORNIS IN CONTINUOUS DARK AND CONTINUOUS LIGHT r I SILENT CPERIODS OF LESS THAN ONE HOUR NOT INDICATED)

H SINGING

PREVIOUS LIGHT REGIME 2 3 AUGUST - 7 SEPTEMBER 1956 LIGHT

CONTINUOUS DARK

7 SEPTEMBER

7 - 8

9-10

CONTINUOUS LIGHT CI25F.C.)

10 SEPTEMBER

10-11 I

' 11 - 12 I .... 12-13

CONTINUOUS DARK

13 SEPTEMBER that all individuals sing during the day hut at a reduced frequency.

Observations of marked individuals in the field will be necessary to settle this point. TAXONOMY OF THE OECANTHINAE OF EASTERN UNITED STATES

Introduction

The Oecanthinae are distinguished from all other Gryllidae by their slender bodies, almost horizontal heads, and very slender posterior legs. The tegmina of the male are broad and nearly trans­ parent. Chopard (1951), probably the foremost writer on the Gryllidae of the world, classes the group as a full family, the Oecanthidae.

Being more of a lumper than a splitter at the higher taxonomic levels,

I have elected to follow the older usage.

Of the three genera placed in the Oecanthinae, Oecanthus is world-wide in distribution, Xabea is found only in Australia and the

Malay Archipelago, and Neoxabea is restricted to the New World. One species of Neoxabea and ten species of Oecanthus are known to occur in the United States east of the hundredth meridian.

The most recent taxonomic work on the Oecanthinae of the United

States is that of B. B. Fulton. In 1915, Fulton published "The Tree

Crickets of New York," which contains detailed accounts of the mor­ phology and biology of eight species. In 1925, he described two

"races” of Oecanthus rileyi in Oregon, and in 1926 in "The Tree

Crickets of Oregon," he considered two other species found in the

West. In a second paper in 1926, Fulton discussed geographical variation in the nigricornis group and presented a key to the North

American Oecanthinae.

Fulton's classification of the Oecanthinae is the one which is accepted today. Table 71 lists the species which occur east of the

89 90

Table 71. Fulton's (1926b) classification of the eastern Oecan­ thinae and changes indicated by the present study.

Fulton’s Classification Proposed Classification Reason for Change

Neoxabea bipunctata Neoxabea bipunctata ------(De Geer) (De Geer)

Oecanthus exclamationis Oecanthus exclamationis ~ - Davis Davis

Oecanthus angustipennis Oecanthus niveus Recognition of F itc h (De Geer) senior synonym

Oecanthus niveus Oecanthus rileyi Baker Correct identifi­ (De Geer) cation of type of n iveus

Oecanthus latipennis Oecanthus latipennis ------R ile y R ile y

Oecanthus califom icus Oecanthus varicornis Recognition of Saussure (Texas F. W alker senior synonym m a te ria l)

Oecanthus n. nigricom is Oecanthus nigricom is _ _ _ F. W alker F. W alker

Oecanthus celerinictus Undescribed sp e c ie s

Oecanthus n. argentinus Oecanthus argentinus Elevation to S aussure S aussure specific rank

Oecanthus n. quadripunc- Oecanthus quadripunc- Elevation to tatus Beutenmuller tatus Beutenmuller specific rank

Oecanthus pini Oecanthus pini - - - Beutenmuller Beutenmuller 91 hundredth meridian according to Pulton’s (1926b) key. The table also lists the changes in classification that are indicated on the basis of this study. Of the eleven species of Oecanthinae recognized in this study, only five correspond in name to species recognized by

Pulton. Of the six species names which do not agree with Pulton's classification, one is of an undescribed species, two are elevated from a subspecific to a specific rank, two are formerly unrecognized senior synonyms, and one is a replacement for a name based on mis- identified type specimens.

Plan of Presentation

In the treatment of each species the following topics are dealt with in order.

Synonymy and Bibliography. In the synonymy of each species, the oldest specific name available is listed first, followed by all others applied to the species in chronological order. The type locality and the disposition of the type specimen(s) for each species are given.

Por each specific name, different combinations containing it are listed in chronological order. A comma immediately after the specific name in an entry indicates that the name was used for the species as a result of a misidentification or an error in spelling and has no nomenclatorial status for that species.

The bibliography of the species is part of its synonymy. The principal publications which have appeared using each synonym are cited. References dealing solely with distribution or control and references which are merely paraphrases of the work of others are 92 omitted. An attempt is made to make a complete listing of references containing nomenclatorial discussions, new information on biology or morphology, or descriptions of the song. Samuel H. Scudder (1901) in his "Alphabetical Index to North American Orthoptera" lists all known references to North American Oecanthinae published prior to 1901.

Habitat and Life History. The habitats in which each species was heard or collected are described. Comments on habitat in the literature are reviewed if they differ from personal observations or if they concern areas other than those in which personal collections were made. The life histories of the eastern species of tree crickets are similar in many respects, and Fulton and others have given excellent accounts of both the general life history of tree crickets and of the details pertaining to individual species. Consequently the discussions in the species write-ups here are confined to details not in the literature and to the seasonal occurence of the adult forms.

Some background information on life history is necessary in understanding the discussions. The eggs are laid in holes bored in the stems of woody or herbaceous plants. All eastern species over­ winter in this stage. In central Ohio, the eggs hatch in late l.Iay or early June. The nymphs and adults are omnivorous, but their rather weak mandibles seem to lim it them to soft foods. In the field I have seen various species feed upon aphids, flower parts, tender leaves, and fruits. In the laboratory all species were successfully maintained on a diet of ground dogfood, lettuce, and aphids. Cannibalism in the sense of one individual attacking and eating another living individual was not observed; and if it occurred, it was infrequent. However, any 93

dead member of a culture was fed upon by its fellows. In experiments with rearing techniques, £. nigricom is was successfully reared from

egg to adult on ground dogfood alone, on lettuce alone, and on aphids

alone. The individuals on the diet of ground dogfood, however, were

stunted in comparison with the ones on lettuce or aphids or with the

ones on dogfood in combination with lettuce and/or aphids.

Tree crickets pass through five nymphal instars. Reproductive

activity begins soon after the last molt. Within 24 hours of the last molt, a female of _0 . quadripunctatus mated in a laboratory culture.

Males of £. quadripunctatus were not observed in courtship for two

days after molting, but copulation was observed on the third day.

Copulation occurs repeatedly during adulthood. In laboratory cultures,

on several occasions confined pairs of the nigricom is group copulated

twice within a two-hour period. There was some indication that older

females were less susceptible to the courtship maneuvers of the male,

but systematized observations are needed to settle this pdint.

In Ohio there is only one generation a year in all species ex­

c e p t 0 . argentinus, in which there are two. farther to the south

there seem to be two or more generations per year in £. quadripunctatus,

0 . celerinictus, and perhaps others.

Distribution. The known geographical distribution of each species

is plotted by counties on a map. Tables showing the sources of all

records and the seasonal distribution of the adult are appended to the

text of this dissertation.

Discussion of Song. A review of the previous descriptions of the

song of each species is followed by a discussion of the data on song 94 gathered in this study. Only the calling songs are discussed, and their use in taxonomy is stressed. Originally I planned to discuss courtship and postcopulatory singing. Recordings of 506 songs of this type were made, but they have not yet been analyzed. Preliminary work indicates that courtship and postcopulatory songs do not differ signi­ ficantly from calling songs in pitch and pulse rate. As a rule, courtship and postcopulatory songs are brief, irregular bursts of sound. In species in which the calling song is a continuous trill, the contrast between calling songs and courtship-postcopulatory songs is great. In species in which the calling song is a broken trill, the contrast is much less apparent (at least this is the situation observed in 0. exclamationis). Only once was 0. rileyi heard to produce sound during courtship, and in this instance the sound was similar to a single chirp.

Key to the Oecanthinae of Eastern United States

Based on Morphological Characters

(Modified from Fulton 1926b)

1. Distal portion of hind tibiae armed with several long spines

and numerous small teeth; first antennal segment frequently

marked with black ventrally, and without a prominent

tubercle on distal border...... Oecanthus 2

1*. Hind tibiae armed with terminal spurs only; first antennal

segment not marked with black ventrally, and with a small

prominent tubercle on distal border near middle ......

Neoxabea b ip u n c ta ta (De Geer) 95

2(l). Ventral face of first antennal segment with a broad, white-

or ivory-colored swelling on inner edge, ornamented with

black; tooth interval (length of file/number of teeth)

39-51 microns ...... rileyi group 3

21. Ventral face of first antennal segment without a swelling on

inner edge; tooth interval 23-37 microns ...... 5

3(2). Black mark on first antennal segment a round or oval spot;

second segment with a similar spot; width of dorsal

field of male tegmina nearly half of length; more than

35 teeth in f ile ...... _0. rileyi Baker

3'. Black mark on first antennal segment not round or oval;

second segment with an elongate black mark; width of

dorsal field of male tegmina less than four-tenths of

length; less than 30 teeth in f ile ...... 4

4(3'). Mark on first antennal segment club-shaped, broadest

proximally; pronotum with no median streak; length of

tegmina of male 12-15 mm ...... 0 . exclamationis Davis

4'. Mark on first antennal segment curved or J-shaped, with

proximal end curved toward the inner side; pronotum

usually with a darker median streak; length of tegmina

of male 10-12 mm ...... niveus (De Geer)

5(2’). First antennal segment unmarked with black, or with a narrow

dark line along the inner edge; frons and basal segments

of antennae mottled with reddish-pink; width of dorsal

field of tegmina of male about half of length; subgenital plate of female with a broad notch posteriorly, one-forth

to one-half as broad as widest part of plate ......

latipennis group 6

First antennal segment marked with black, usually with more

than a narrow line along the inner edge; no reddish-pink

on head or antennae; width of dorsal field of tegmina of

male rarely oyer four-tenths of length; subgenital plate

of female with a narrow notch posteriorly, not more than

one-fifth as broad as widest part of plate ......

nigricomis group 7

Antennae reddish-pink proximally, fading out near eighth

segment; first and second antennal segments without

distinct dark markings; in dried specimens, hollow on

side of terminal segment of maxillary palp covers less

than distal half of segment; file with more than 38 teeth

0 . latipennis Riley

Antennae with basal two segments and dorsal-proximal area

of third segment reddish-pink; remainder of third segment

and succeeding segments black but gradually fading from

about the tenth segment to the pale distal segments;

first and second antennal segments usually each with a

narrow dark stripe on inner edge; in dried specimens,

hollow on side of terminal segment of maxillary palp

covers distal half or more of segment; file with less

than 38 teeth ...... 0. varicornis F. Walker 97

7(5'). Ventral faces of first two segments of antenna black or each

with a pair of heavy black marks which may be confluent;

infuscated area fading proximally from the outside mark on

the first antennal segment; venter black; appendages

infuscated or black; pronotum usually with three longi­

tudinal dark strip es ...... 0. nigricomis F. Walker

7*. Ventral faces of first two segments of antenna with black

marks but with no infuscated area fading proximally from

the outside mark on the first antennal segment; venter

usually light, sometimes infuscated; appendages usually

light; tibiae, tarsi, and antennae occasionally

infuscated; pronotum usually with no stripes, occasionally

with one or two faint longitudinal strip e s ...... 8

8(7'). Head, pronotum, le g s , and antennae lig h t to medium brown,

sometimes reddish-brown; venter dull brown; tooth

interval 31-35 m icrons ...... 0_. pini Beutenmuller

8'. Hot marked with brown; tooth interval 23-31 microns. . . . 9

9(8’). Marks on second antennal segment confluent, or space between

marks less than one-third width of inside mark; length of

inside mark on first antennal segment less than 4 times

w id th ...... 0_. argentinus Saussure

9'. Marks on second antennal segment never confluent; space

between marks more than one-third width of inside mark;

length of inside mark on first antennal segment sometimes

more than 4 times width ...... 10 98

10(9’) Outer mark on second antennal segment distinctly lighter

than inner mark and/or at least one third shorter; outer

mark on first segment usually round, hut occasionally

with a lighter transverse tail; hind femur and tibia

with no dark markings; file with more than 48 teeth . .

0. quadripunctatus Beutenmuller

1 0 *. Outer mark on second antennal segment as dark as inner mark

and nearly as long; outer mark on first segment never

round, always equally dark in all portions; hind femur

with black mark at apex on each ventral-lateral margin;

hind tibia with 2 ( r a r e ly 1 ) proximal transverse black

bands on the posterior face; file with less than 48

t e e t h ...... 11

1 1 ( 1 0 * ).Length of inside mark on first antennal segment usually

more than 4 times width; pronotum never with stripes;

space between marks on second antennal segment sometimes

greater than 1 1/3 times width of inside m ark ......

_0 . celerinlctus n. sp.

1 1 *. Length of inside mark on first antennal segment usually

less than 4 times width; pronotum usually with a faint

median longitudinal stripe and occasionally with faint

marginal longitudinal stripes; space between marks on

second antennal segment never greater than 1 1/3 times

width of inside mark. 0_. nigricomis Walker, willow form 99

Genus Neoxabea

The genus Neoxabea was proposed by Kirby in 1906. Since Kirby listed only Neoxabea bipunctata (De Geer) as belonging to the genus, bipunctata becomes the type species of the genus by monotypy. Both

Neoxabea and Xabea are distinguished from Oecanthus in having no spines on the hind tibiae proxlmad of the apex. No reference to a morphological distinction between Neoxabea and Xabea has been found in the literature. The accepted usage is to place species from

Australasia in Xabea and species from the New World in Neoxabea.

Neoxabea b ip u n c ta ta (De Geer)

The Two-Spotted Tree Cricket

Figures 64, 6 6 , 6 V, 69, 79, 92

Gryllus bipunctatus De Geer 1773, p. 523 (type locality, Pennsylvania;

type, a female in the De Geer Collection, Naturhistoriska

Riksmuseum, Stockholm, Sweden). Scudder 1862, p. 432 (synonymy);

Scudder 1868, p. 31 (synonymy); Scudder 1901, pp. 123, 132 (lists

four additional references).

Oecanthus bipunctatus (De Geer). Serville 1831, p. 135. P. Walker

1869, p. 93 (synonymy); Saussure 1874, p. 462 (synonymy);

Saussure 1897, p. 255 (synonymy); Scudder 1900, p. 90 (synonymy);

Scudder 1901, p. 210 (lists ten additional references).

Xabea bipunctata (De Geer). Riley 1881, p. 61. Scudder 1901, p. 340

(lists five additional references).

Neoxabea bipunctata (De Geer). Kirby 1906, p. 76 (original indication

of genus Neoxabea). Allard 1910a, p. 38 (habitat, song); Fulton 100

1915, p. 44 (biology, morphology); Rehn and Hebard 1916, p. 300

(habitat); Blatchley 1920, p. 727 (habitat, synonymy); Morse

1920, p. 413 (habitat, song); Davis 1926, p. 41 (habitat, song);

Fulton 1926b, p. 59 (key); Allard 1929a, p. 571 (song); Fulton

1932, p, 62 (habitat, song); Hebard 1934, p. 254 (habitat);

Hebard 1938, p. 102 (habitat); Alexander 1956, p. 168 (biology,

discussion and analysis of song).

Oecanthus punctulatus, Fitch (erroneous subsequent spelling of

bipunctatus De Geer 1773) 1856, p. 415 (morphology). Scudder

1868, p. 55 (synonymy); Scudder 1901, p. 214 (lists one addi­

tional reference).

Oecanthus formosus F. Walker 1869, p. 94 (type locality, Mexico;

type, a female in the British Museum, London, England). Scudder

1901, p. 211 (lists one additional reference).

The type of bipunctata was loaned to me by Dr. Rene Malaise of the Naturhistoriska Riksmuseum, Stockholm. It is a female in good condition except for a broken ovipositor and a missing left hind leg.

It is the sole specimen pinned under the name Gryllus bipunctatus in the De Geer Collection. It bears this label written by Dr. Malaise:

"Gryllus bipunctatus De Geer holotype."

Oecanthus formosus Walker is listed as a synonym of N. bipunctata because Kirby (1906), who had the holotype at hand, placed it as such.

Walker’s description of formosus definitely places it in Neoxabea, and nothing in the description prevents it from being bipunctata. 101

David R. Ragge of the B ritish Miseum writes (personal comnmnicatioxi) that the type of formosus is "not in good condition."

Habitat and Life History

Neoxabea bipunctata occurs on deciduous trees and in tangled undergrowth. In central Ohio I found it most abundantly in tangles of wild grape vines along forest borders and in two unsprayed apple trees adjacent to woodland. I also observed it on elm, sugar maple, wild cherry, and ash in woodland.

Allard (1910a) stated that at Thompson's Iiills, Georgia, it

"prefers the dense leaf-canopy of grape-vines," but he also found it

"in low trees near dwellings." Fulton (1915) stated that King collected bipunctata near Cedar Point, Ohio, "on oak, willow and wild grape vines at a height of five or ten feet. The trees were along the forest border or standing isolated at the edge of open, sandy areas." Rehn and Hebard (1916) found that in the southeastern United

States "the species is extremely retiring, living only in the densest tangles of heavy forest undergrowth...." At Wingina, Virginia, Davis

(1926) found it in apple trees in a garden. Fulton (1932) stated that in North Carolina it is found in vines, bushes, and trees. Hebard

(1938) indicated that in Pennsylvania, N. bipunctata "prefers vine tangles in openings of woodlands and along forest borders." Alexander

(1956) collected it in Ohio from hackberry, elm, sassafras, and apple.

He stated that "it usually sings from 10 feet or higher in trees." 102

In central Ohio the first adults appear at the end of July and

some individuals evidently live until frost; however, the species becomes noticeably scarcer in early September and was not found in

Franklin County in 1956 after 14 September though other species were

s t i l l common. I n th e S outh, th e only evidence o f more than one

generation a year is the record at Brownsville, Texas, of adults from

8 May to 8 August (Table XXII).

Distribution

The distribution of bipunctata corresponds to the deciduous

forest region of eastern North America (Fig. 64; Table XXII). The

species is also recorded from Teocelo, Vera Cruz, Mexico, by Rehn

(1902b) and from San Marcos, Nicaragua, by Baker (1905); however,

another species may be involved.

In addition to the records mapped in Figure 64, there are two

other records from the eastern United States that should be mentioned.

Blatchley (1920) reported N. bipunctata from M ississippi, and a

specimen in the U. S. National Museum is labeled central Missouri.

Discussion of Song

The song of Neoxabea bipunctata is difficult to distinguish by

ear from the songs of the two other species which produce broken

trills — Oecanthus exclamationis and 0. niveus. Allard (1910a)

(Thompson's M ills, Georgia) stated that its notes are "deeper,

stronger, and richer" than those of niveus. Davis (1926) (Wingina,

Virginia) noted, "The song is continued for some time longer than in H A T S x m i

DISTRIBUTION OF N. BIPUNCTATA DISTRIBUTION OF Q. EXCLAMATIONIS

o LITERATURE RECORD o LITERATURE RECORD • SONG RECORD • SPECIMEN EXAMINED

• SPECIMEN EXAMINED .103

F ig . 64 F ig . 65 104 exclamationis and considerably longer than in angustipennis- [= niveus").

The intervals or stops are very short, in fact on very warm evenings it is almost continuous... .*' Fulton (1932) (North Carolina) found that in bipunctata and exclamationis the trills are usually 15-30 seconds long with very brief intervals, while in niveus the trills are seldom more than 10 seconds long with intervals of one to several seconds.

In distinguishing songs in the field, I had no better success than Fulton in separating bipunctata and exclamationis; however, analysis of recordings revealed song differences which make possible positive identification of recordings.

Fewer recordings were made of the song of bipunctata than of any other species of tree cricket which occurred in areas visited during this study. Five adult males were kept in the laboratory for over a month, but only one sang, and he was not recorded. Field recordings definitely assignable to bipunctata were difficult to obtain because the crickets usually sang from perches out of reach. As a result, the analysis of the details of the song of this species must be based upon seven recordings, three of which were made by R. D. Alexander.

F ig u re 66 shows the effect of temperature upon pulse rate as evidenced in the five recordings for which temperature data are available.

Compared with the songs of other species which produce broken trills, the song of bipunctata at a given temperature has a higher pulse rate. By determining pulse rate from a tape recording, the species may be definitely identified if the temperature datum for the recording is available. In the field the song of bipunctata may H A T E XXXIII 105 EFFECT OF TEMPERATURE ON PULSE RATE 61. BIPUNCTATA FIELD RECORDINGS

100

90

u

KEY TO LOCALITIES O FRANKLIN CO, OHIO • LICKING CO, OHIO 50 « FRAMtLIN CO, OHIO (RECORDED BY R. D. ALEXANDER!

REGRESSION LINE 40

30 60 70 80 90 DEGEES FAHRENHEIT F ig. 66

RELATIONSHIP OF PULSE RATE AND PITCH BIPUNCTATA. FIELD RECORDINGS

KEY TO LOCALITIES O FRANKLIN CO., OHIO • LICKING CO, OHIO C FRANKLIN CO, OHIO (RECORDED BY R. D. ALEXANDER) WILLIAMS CO, OHIO (RECORDED BY R D ALEXANDER) u i

7 0 8 0 90 100 no Fig. 67 PULSES PER SECOND 106 sometimes be distinguished by ear from niveus on this basis. At ordinary singing temperatures, the trill of bipunctata sounds like a low-pitched whine with no easily detectable pulsations in intensity.

In niveus, on the other hand, the trill has easily detectable pulsa­ tions. I was unable to separate bipunctata and exclamationis by ear on this basis except when both were singing at the same time and a direct comparison could be made.

Pitch is of no help in distinguishing in the field the song of this species from other broken trills; however, on the basis of the relationship of pitch to pulse rate, recordings of bipunctata can usually be separated from those of niveus even without temperature data. They cannot be so separated from those of exclamationis.

Figure 67 shows the correlation of pitch and pulse rate in bipunctata.

Figure 69 shows hov; the two other species which produce broken trills compare with bipunctata in this respect. At a given temperature (but not at a given pulse rate), bipunctata has a slightly higher pitch than either exclamationis or niveus. This is contrary to Allard's

(1910a) observations, but slight temperature differences could change the relationships of the pitches of the three species.

The data on the analysis of the trill characters of bipunctata as shown in field recordings are given in Table VII. Comparison with sim ilar analyses of trill characters in exclamationis and niveus

(Tables VIII and IX) makes it apparent that there are no consistent differences among the trill characters of the three species. However, bipunctata usually produces trills of longer duration than does exclamationis or niveus, and the intervals occupy a smaller portion 107

Table VII. Trill characters of Neoxabea bipunctata; field recordings. RDA = recording made by R. D. Alexander.

T rill Duration T r i l l Temp. No. o f T r i l l s L o c a lity ( seconds) P er Cent °F. T r i l l s ■ A lin . Range A ver. o f T o tal

Franklin Co., Ohio 75 9 l - 6 § 3.95 14.0 83.1

Franklin Co., Ohio 64 6 2-13g- 6.14 9.5 87.4

Franklin Co., Ohio (RDA) 74 30 1-3 3 /4 1 .6 6 19.0 47.4

Williams Co., Ohio (RDA) ? 3 4 - 1 8 | 9.03 ------Williams Co., Ohio (RDA) ? 3 2-|-5 3 /4 3.95 ------

Average (of averages) 4.95 14.1 72.6

of the total song. Usually the song of bipunctata can be separated

in the field from that of niveus on the basis of trill length, but it

cannot be reliably separated from exclamationis. One factor which

confuses identification on the basis of the trill characters is that when individuals of species which produce broken trills begin to sing

or when they are disturbed, they sing more irregularly than they do

under other circumstances.

Genus Oecanthus

Serville described the genus Oecanthus in 1831 and included in it

three species — itallcus Latreile (■ pellucens Scopoli), bipunctatus

De Geer (now placed in Neoxabea), and niveus De Geer. To my knowl­

edge, the type species of Oecanthus has not been designated. Follow­

ing the recommendations of the International Commission of Zoological 108

Nomenclature, italicus is here selected as the type species of the genus. The reasons for this selection are ( 1 ) italicus is indigenous from the viewpoint of Serville, the original author of the generic name, (2) italicus is probably the only species which Serville actually studied before proposing the genus, and (5) italicus is placed first in Serville’s list of species belonging to Oecanthus.

Oecanthus is distinguished from the other genera of the

Oecanthinae by numerous small teeth and spines on the distal portion of the hind tibiae.

In the discussion of the genus Oecanthus. the species are divided into three groups as proposed by Pulton (1915) — the rileyi group, the latipennis group, and the nigricom is group.

The rileyi Group

The rileyi group is characterized by a broad swelling on the inner edge of the ventral face of the first antennal segment. The area between the eyes is usually tinged with yellow, and the file teeth average farther apart than in the other groups. To determine average tooth interval, the length of the file was measured along a straight line parallel to the main axis of the file, and the result was divided by the total number of teeth. In the rileyi group the teeth were 39 to 51 microns apart, while in the other groups the file teeth were 23 to 37 microns apart. (See Pigs. 79 and 98 for the results of file measurement in all species of eastern Oecanthinae.)

In the rileyi group the calling song is a broken trill or a regular 109 chirp. The members of this group in the eastern United States live chiefly in bushes and trees and lay their eggs singly in bark.

Specimens of the rileyi group are most easily separated by the markings on the first two antennal segments. In exclamationis, there

is a longitudinal mark on the proximal segment and a shorter mark on the second segment, and the two together resemble an exclamation point {!). In niveus the mark on the proximal segment is J-shaped.

In rileyi there is a round or oval dot on each of the first two antennal segments.

Oecanthus exclamationis Davis

Davis's Tree Cricket

Figures 19, 21, 31, 41, 42, 43, 65, 68, 69, 79, 92

Oecanthus exclamationis Davis 1907, p. 173 (Type locality, Staten

Island and New Jersey; types, 4 female syntypes at the Staten

Island Institute of Arts and Sciences, Staten Island, New York).

Fulton 1915, p. 30 (biology, morphology, song); Morse 1919, p. 30

(habitat); Morse 1920, p. 407 (biology, egg, song); Davis 1926,

p. 40 (habitat, song); Fulton 1926b, p. 59 (key); Allard 1929a,

p. 581 (song); Fulton 1932, p. 62 (habitat, song); Hebard 1934,

p. 252 (biology); Hebard 1938, p. 101 (habitat); Fulton 1951, p.

93 (habitat, song); Alexander 1956, p. 159 (biology, discussion

and analysis of song).

Oecanthus niveus, Fitch (in part: variety "a") (not De Geer 1773)

1856, p. 413. 110

I. C. G-. Cooper, Curator of the Davis Collection, Staten Island

Institute of Arts and Sciences, furnished information on the types of

exclamation!s. He wrote (personal communication) that there are four

specimens of Oecanthus exclamationis in the collection marked "Co-

Type . 11 There is one each from Staten Island; Cranford, Hew Jersey;

Farraingdale, N. J .; and lianasquan, N. J. All are females.

Fitch (1856) described a variety "a" of what he called Oecanthus

niveus. (Actually Fitch’s niveus was rileyi, not De Geer's niveus.)

This variety had "the black dots on the under side of the two first

joints of the antennae lengthened into short stripes." He was evi­

dently referring to a specimen of what Davis later named exclamationis,

but his designation has no status in nomenclature because a single

letter is not available as a specific name.

Habitat and Life History

Oecanthus exclamationis is the most completely arboreal of the

species studied. It is usually heard in the crowns of deciduous trees

and is difficult to collect. Occasionally it is found in young trees

in fence rows, and I collected the species in greatest numbers in a

solitary, unsprayed apple tree along a field road. I also collected

the species from elm, hackberry, wild cherry, red oak (Horth Carolina),

and catalpa (Virginia).

Fulton (1915) stated that on Long Island it is most common on

bur oak but is also found on black jack oak. Davis (1926) (Wingina,

Virginia) noted that it is "often found in post oaks, etc., but may I l l occasionally be collected in bushes growing in moist situations.”

Hebard (1938) (Pennsylvania) indicated that it is more strictly arbo­ real than other species of tree crickets and "prefers oaks, maples, and other deciduous trees." Pulton (1951) (North Carolina) stated that it lives mainly in the tops of forest trees.

In central Ohio exclamationis matures about the first of August and is heard until frost. After the middle of September the species is noticeably less abundant. There is no evidence of a second generation in the South (Table XXIII).

Distribution

The distribution of 0_. exclamationis (Pig. 65; Table XXIII) is very similar to that of N, bipunctata, except the former seems to range farther west. At the Philadelphia Academy of Sciences, I examined specimens which were apparently exclamationis from west Texas

(Brewster County), Arizona (Cochise, r.ojave, and Santa Cruz Counties), and California (extreme east San Bernardino County).

Discussion of Song

The difficulties in distinguishing the calling song of 0_. exclamationis from that of Neoxabea bipunctata were discussed under the latter species. The song of exclamationis may be distinguished in the field from that of niveus in the same ways as the song of bipunctata may be distinguished from niveus — the trills average longer with shorter intervals, and the pulse rate is higher, making the pulsation difficult to detect. 112

Fulton (1915) found that the pitch of the song of exclamationis

"is the lowest of any of the species studied and reaches only to the second B above middle C." He also noted that "the beginning of each note is comparatively weak, but the sound increases in volume and slightly in pitch and continues uniformly until it abruptly ends."

Davis (1926) indicated that the song of exclamationis is louder than niveus, continues longer, and the stops between the periods are s h o r te r .

The relationship of temperature and pulse rate in field record­ ings (Fig. 68) agrees with the data accumulated from laboratory recordings. The pulse rate will distinguish exclamationis from any other species if the temperature is known. Recordings from North

Carolina and Virginia do not differ significantly from those made in

Ohio at similar temperatures.

Pulse rate and pitch retain the same correlation in field record­ ings (Fig. 69) as they had in the laboratory. The relation of pulse rate and pitch w ill usually separate recordings of exclamationis from those of niveus but not from those of bipunctata. Fulton (1915) was correct in assigning exclamationis to a lower pitch than niveus (he had not heard bipunctata) at a given temperature; however, the dif­ ference is not great enough to be used in the field by a person with ordinary pitch discrimination. The pitch is 2750 for exclamationis compared with 2950 cps for niveus at 75° F. Fulton's pitch determina­ tion (960 cps) was evidently the third subharmonic instead of the fundamental. KILOCYCLES2.5 PER SECOND 3.5 3.0 2.0 Fig. 30 E T LOCALITIESKEY TO Q. NIVEUS - A PIKE CO, ARKANSAS a JNS O GEORGIA CO,O JONES B O FRANKLIN CO, OHIO DYERP TENNESSEE CO, O AIKEN CO, CAROLINA SO A B HOLMES CO, FLORIDA SHELBY • CO, OHIO 69 ONSHI PULSE RT AD PITCH AND RATE E S L U P F O IP H S N IO T A L E R HOLMES CO, MISSISSIPPI RAPIDES AND VERNON PARISHES, UNION CO, INDIANA LOUISIANA . XLMTOI AD IES FED RECORDINGS FIELD NIVEUS, O. EXCLAMATIONIS AND 40 UJ u 70 0 100 80 40 90 30 20 FET F EPRTR O PLE RATE PULSE ON TEMPERATURE OF EFFECT g 68 ig. F 50 B 60 R C C R E 3S 10N LINE TOR TOR LINE 10N 3S E R C C R □ AB ORY RC DI S G IN RD RECO Y R TO A R BO LA USS E SECOND PER PULSES HATE ERE FAHRENHEIT DEGREES Q. IL RECORDINGS FIELD EXCLAMATIONIS 0 70 60 7 0 C 2 0 80 70 KEY LOCALITIES TO - Q. EXCLAMATIONIS A C TT OL O C A L IT IE S A A □ WAKE □ NORTH CAROLINA CO, O »A P P O M A T T O X oW A K■ EC > F O R .,N A O N R K T L H INC C A O R ,O O H L IO I N A

FRANKLIN OHIO CO, APPOMATTOX CO, VIRGINIA CO., VIRGINIA 60 90 90 113 114

Table VIII. T rill characters of Oecanthus exclamationis; field recordings. EDA = recording made by R. D. Alexander.

Temp. No. o f T rill Duration T r i l l s T r i l l L o c a lity °F T r i l l s ( seconds) P er P er Cent Range A ver. M inute o f T o tal Wake Co. , No. C aro lin a 79 14 3/4 -1 3 $ 3.44 1 2 .6 65.4

Franklin Co., Ohio 73 17 o 2.57 15.8 61.0

Franklin Co., Ohio 73 11 l£ -5 £ 2.69 16.6 67.2

Franklin Co., Ohio 75 9 4.10 10.3 63.5

Franklin Co., Ohio (EDA) 70 8 2 3/4-4-2 2.76 18.2 75.5

Appomatox Co. , V irg in ia 76 36 3.00 18.6 50.4

Average (of averages) 3.10 1 5 .4 63.8

The increase in volume and the slight increase in pitch at the

beginning of each trill, noted by Fulton (1915) in exclamationis, are

characteristic of the beginnings of trills in all species, and in my

observation, they are no more apparent in exclamationis than in niveus

and bipunctata. The phenomenon must be due to the wing velocity

being less than normal for the first several strokes of a trill. The

result is that both pitch and pulse rate are less than normal for a

fraction of a second.

Table VIII lists the trill characters as shown in field recordings.

In the laboratory, trills were usually more irregular in length than

in the field. Occasionally trills produced in the laboratory were

longer than any heard in the field. One individual at 80° P. sang

for more than 66 seconds without pause, and another at 76° F. sang

for 35 seconds. 115

Oecanthus niveus (De Geer)

The Narrow-Winged Tree Cricket

Figures 22, 32, 41, 42, 43, 69, 70, 72, 79, 92

Gryllus niveus De Geer 1773, p. 522 (type locality, Pennsylvania;

type, a female in the De Geer Collection, Naturhistoriska

Riksmuseum, Stockholm, Sweden). Scudder 1901, pp. 126, 129,

133 (lists six additional references).

Oecanthus niveus (De Geer). Serville 1831, p. 135 (original descrip­

tion of genus Oecanthus). Scudder 1867, p. 309 (song); Scudder

1874 (in part: "night song"), p. 366 (song); Riley 1881, p. 60

(song); (?) Ayers 1883, p. 228 (embryology, parasite); (?) Graber

.1889, p. 133 (embryology).

Oecanthus niveus var. £. angustipennis Fitch 1856, p. 413 (type

locality, New York; type, a male, destroyed).

Oecanthus angustipennis Fitch. Walker 1869, p. 116 (original refer­

ence to angustipennis as a species); Davis 1889, p. 80 (song);

McNeill 1889, p. 102 (song); McNeill 1891, p. 6 (song); Hart

1892, p. 34 (morphology); Scudder 1893, p. 67 (song); Beutenmuller

1894b, p. 251 (habitat, song, synonymy); Beutenmuller 1894c, p.

270 (habitat, song); Scudder 1900, p. 90 (synonymy); Faxon 1901,

p. 183 (song); Scudder 1901, p. 210 (lists 16 additional refer­

ences); Blatchley 1903, p. 450 (biology, sang, synonymy); German

1904, p. 64 (biology, song); Forbes 1905, p. 218 (food); Kirby

1906, p. 74 (synonymy); Allard 1910a, p. 33 (habitat, song);

Allard 1911a, p. 28 (biology, song); Allard 1911b, p. 155 (song); 116

Allard 1912, p. 460 (song); Davis 1914, p. 204 (habitat, song);

Parrott and Pulton 1914, p. 447 (detailed account of biology,

morphology, and song); Pulton 1915, p. 27 (detailed account of

biology, morphology, and song); Gloyer and Pulton 1916, p. 3

(vector of fungus); Rehn and Hebard 1916, p. 296 (habitat); Pox

1917, p. 234 (habitat); Morse 1919, p. 30 (habitat); Blatchley

1920, p. 717 (biology, song); Morse 1920, p. 406 (biology, egg,

song, synonymy); Hebard 1925, p. 37 (habitat); Snodgrass 1825,

p. 440 (song); Davis 1926, p. 40 (biology, song); Pulton 1926 b,

p. 59 (key); Allard 1929a, p. 571 (song); Pulton 1932, p. 64

(habitat, song); Strohecker 1937, p. 233 (habitat); Cantrall

1943, p. 45 (habitat, song); Pulton 1951, p. 93 (song); Priauf

1953, p. 99 (biology); Alexander 1956, p. 159 (biology, discussion

and analysis of song).

At present the name Oecanthus niveus is universally used for the snowy tree cricket (_0 . rileyi), but examination of the type material has revealed that the name must be applied to the narrow-winged tree cricket instead.

Dr. Rene Malaise of the Haturhistoriska Riksmuseum, Stockholm, obligingly sent De Geer’s type series of niveus to me when I wrote him for information concerning it. The series consists of three females, and Dr. Malaise assured me that they are the ones, and the only ones, that De Geer used in preparing his description. All three are definitely narrow-winged tree crickets, the species at present referred to as angustipennis. Each specimen bears this label in Dr. 117

Malaise’s writing: "Gryllus niveus De Geer .'1 In addition one has a

numbered red label attached by Dr. Malaise as a means of identifying

the series. The specimen with the numbered label has at least one of

each pair of appendages, and I have designated it the lectotype and

placed a red lectotype label on the pin.

The present misuse of the name niveus is a result of failure to

consult the type material, the original description being insufficient

to accurately place the species. In 1841, Harris described what he

thought was niveus as possessing "a minute black dot on the under

sides of the first and second joints of the antennae." These dots are

peculiar to the snowy tree cricket. Later authors seem to have ac­

cepted H arris’s erroneous concept of niveus without question, although

some authors prior to 1900 lumped several species (always including

riley i) under the name niveus (e_.£., Scudder 1862; Packard 1881).

The presently used name of the narrow-winged tree cricket,

angustipennis, needs some discussion, even though it is a secondary

synonym of niveus. In 1856, Pitch published an extensive commentary

on what he called the "White Flower-Cricket, Oecanthus niveus, Degeer."

Under this label Pitch evidently included all Worth American Oecanthi-

nae which he had seen except Oecanthus nigricornis (which he called

fasciatus) and Neoxabea bipunctata (which he called Oecanthus

punctulatus). The species which Pitch evidently thought was typical

niveus was actually rileyi. Pitch concluded his detailed description

of "niveus" (actually rileyi plus other species) by describing seven

"varieties." He designated four of these varieties with a letter only 118

(a_, Jb, c, and _d). He gave Latin names in addition to letter designa­

tions to the other three varieties — "e_. d isco lo ratu s"f. fuscipes,"

and angustipennis . 11

Fitch's description of angustipennis, based on a single specimen,

is as follows: "The male with wing covers a third narrower and some­ what shorter than usual, with the wings protruding like tails from

under their tips."

The first author to use angustipennis as a species name was F.

Walker (1869), who merely listed it as having been described by Fitch.

Davis (1889) was the first to apply the name to material of his own.

He wrote that he considered angustipennis and rileyi distinct because

their structure is not the same and their stridulation is different."

Hart (1892) was the first to mention the characteristic antennal markings of this species in conjunction with the name angustipennis.

Finally Beutenmuller (1894b) pointed out that Fitch's description of

angustipennis was too brief for recognition of the species and that

the description could refer as well to quadripunctatus, for instance,

as to the narrow-winged tree cricket. He concluded that since Fitch's

types of Oecanthus "have been destroyed" and correct identification

can never be ascertained, angustipennis should "be retained for the

species so well known to us by this name."

Fitch's angustipennis is thus fixed as a secondary subjective

synonym of De Geer's niveus. The fate of Fitch's other two named

varieties is not so easy to determine.

Both discoloratus and fuscipes were cited as species by F.

Walker in 1869, but no other author has cited them as such. Davis 119

(1907) thought that fuscipes was ’’likely" Oecanthus nigricom is, but concluded that the two names "cannot be placed." On the basis of my study of eastern tree crickets, including material from New York

(Pitch's home state), I cannot definitely place either of Fitch's names.

Fitch’s description of discoloratus reads, "The whole of the head, the first joint of the antennae, the breast and abdomen of a brownish clay color." Any specimen of a tree cricket which had been stained with fluids from other specimens or which had been left in a killing jar overnight could fit this description.

Fitch’s description of fuscipes reads, "One or both the hind legs more or less tinged with blackish." 0 . nigricomis is the only species in New York which ordinarily has blackish hind tibiae, but

Fitch placed specimens of nigricomis in his fasciatus. Melanistic

individuals are known in other species; for instance, Allard (1929a)

found a "black or melanistic individual of the snowy-tree cricket

chirping in the usual manner near Washington, D. C." Fitch’s specimens

could well have been slightly melanistic individuals of any species

of Oecanthus occurring in New York.

Since neither of Fitch's names can be placed definitely on the

basis of his description and the types have been destroyed, I propose

that both be classed as nomina dubia.

Habitat and Life History

Oecanthus niveus is found in a slightly wider range of habitats

than are exclamationis and bipunctata. It is not only characteristic 120 of deciduous trees and tangled undergrowth but is also occasionally found on herbaceous plants. In central Ohio I collected niveus from apple, ash, elm, haclcberry, and wild cherry. I also found it on grape Tine, Virginia creeper, and honeysuckle, sometimes within two feet of the ground. At Cedar Point, Ohio, I collected niveus nymphs in abundance from cocklebur and blackberry growing about an isolated ash tree in a field. In ^airfield County, Ohio, I collected a male from a four-foot pine tree growing in an abandoned field. In several southeastern states (South Carolina, Georgia, northern Florida,

Tennessee, M ississippi, Louisiana, and Arkansas), I heard it frequently in oaks of various species and seldom elsewhere.

Beutenmuller (1894c) (New York) indicated that niveus "inhabits the higher parts of different kinds of forest and fruit trees.”

Blatchley (1903) (Indiana) wrote that it "frequents the borders of groves and especially ironweeds in open pastures.” Allard (1910a) found that at Thompson's M ills, Georgia, it prefers the foliage of tall oaks, and in New England, it prefers low shrubs and the tangles of vines and sweetfero in pastures. Davis (1914) (Florida) collected niveus "among the golden rods and other low plants by the side of the road" and also found it in small oaks and other trees. Fulton (1915) concluded that about Geneva, New York, the species is confined "to woody plants, either trees or large bushes.” In North Carolina,

Fulton (1932) found niveus in "various trees" and sometimes "in herbs and bushes under trees." Strohecker (1937) (Chicago area) reported niveus from black oak, oak-hickory, and beech-maple forests. Cantrall 121

(1943) (George Reserve, Michigan) found it characteristic of the deciduous-arboreal stratum” and in upland woods in the "tall shrub

stratum" but not in the "low shrub-terrestrial stratum." Friauf (1953)

(northern peninsular Florida) reported it from dwarf oaks and other sh ru b s.

Oecanthus niveus in central Ohio matures about the same time as exclamationis — the first of August or a little later. It remains common until frost. A lone fourth instar nymph of niveus collected on September 11, 1956, in Franklin County, Ohio, must have been

either a member of a partial second generation or a retarded individual of the first generation.

There is better evidence of more than one generation per year in

the southern states. I found singing adults of the species in

M ississippi, Louisiana, and Arkansas on 14-16 June. There are records

of adults in November in North Carolina and in October in Georgia.

Blatchley (1920) recorded adults in March, May, June, October,

December, and January at Dunedin, Florida (near Tampa).

Distribution

Oecanthus niveus occurs throughout the deciduous forest and

southeastern evergreen forest regions of eastern United States

(Fig. 70; Table XXIV). Records not indicated on the map are Minnesota

(Lugger 1897) and eastern Nebraska (Blatchley 1920). The only record

from west of the hundredth meridian is by Scudder and Cockerell (1902),

who reported the species from K esilla, Donna Ana County, Hew Mexico. H A T S XXXV DISTRIBUTION OF Q. NIVEUS DISTRIBUTION OF 0 RILEYI

O •.

o LITERATURE RECORD o LITERATURE RECORD • SONG RECORD • SONG RECORD • SPECIMEN EXAMINED • SPECIMEN EXAMINEO M CO CO 123

Discussion of Song

The means of distinguishing the song of niveus from those of the other specie 3 producing broken trills — bipunctata and exclamationis — have been discussed under these species.

There are numerous descriptions of the song of niveus in the literature, and even though from widely separated localities they agree in essential characteristics. hcNeill (1889) (Illinois) found thet the song "lasts only from three to five seconds with an equal interval between trills." Faxon (1901) (New England) wrote that the song consists of "a trill of several second's duration, succeeded by a short pause," suggesting "the spring note of the toad, heard afar off." Allard (1910a) (Thompson’s M ils, Georgia) reported that the trills are short, abrupt, and followed by about the same intervals of rest. He thought that the trill did not maintain a uniform pitch, but died away in a slightly lower key. Fulton (1915) (Kew York) found that the song had a pitch varying from C# to D$, two octaves above middle C (1088 to 1216 cps), "depending on temperature and somewhat on individual variation." "Each trill continues from one to five seconds, but most commonly it lasts for about tiro seconds. The periods of rest vary more and nay be from one to eight seconds or longer. The total number of notes per minute varies, even from one minute to the next and is generally not more than ten, but may occasionally run as high as fifteen. On one occasion a specimen alone in a cage was observed to trill continuously for a minute or more." Snodgrass (1925) (Washington, D. C.) reported that the song 124 consists of purring sounds usually prolonged about two seconds and separated by intervals of the same length. Hilton (1932) (North

Carolina) found that the song consists of trills over one second long and seldom more than ten seconds long and rest periods of from one to several seconds. Cantrall (1943) (Michigan) described the trills as lasting from one to six seconds.

When the pulse rate data obtained from the analysis of field recordings are plotted against temperature (Fig. 72), it becomes apparent that recordings made in the Southeast differ slightly from those made in the Midwest. All data from the former area fall below the line obtained from laboratory recordings of Ohio individuals. Ihe data from recordings made in the field in the Midwest fall nicely along the calculated line. When pitch is plotted against pulse rate the same separation occurs (Fig. 69), with the recordings from the

Southeast being higher in pitch at a given pulse rate than those from the Midwest. If the deviation of the recordings made in the Southeast had been a result of too high tape speed, the pitch would have been lo w e r than predicted at a given pulse rate; therefore, the deviation is real and not an artifact. More data are needed before the variation can be further described.

Two observations on pitch mentioned in the review of previous song descriptions do not agree with the data of the present study.

T rills did not die away in a slightly lower key as reported by Allard

(1910a), and no pitch of less than twice Fulton's (1915) determina­ tions was encountered. HATE XXXVI 125 EFFECT OF TEM PERATURE ON PULSE RATE Q. NIVEUS FIELD RECORDINGS 100

H1 1 TO LOCALITIES O FRANKLIN CO, OHIO 90 A PIKE CO, ARKANSAS • JONES CO, GEORGIA ■ RAPIDES AND VERNON PARISHES, LOUISIANA • SHELBY CO., OHIO 80 ■ HOLMES CO, FLORIDA A UNION CO., INOIANA • HOLMES CO, MISSISSIPPI • AIKEN CO, SOUTH CAROLINA Q o DYER CO, TENNESSEE LrccRESSION LINE FOR LABORATORY O70 RECORDINGS u (SPECIMENS FROM toUJ OHIO! u 6 0 CL <0 UJ to =j50 a. SOUTHEASTERN UNITED STATES 40

30

20 50 60 70 80 DEGREES FAHRENHEIT F ig . 72

RELATIONSHIP OF PULSE RATE AND PITCH O. RILEYI, FIELD RECORDINGS

3 .0 o z uo UJ ^ 2 .5 a aUJ

UJ KEY TO LOCALITIES -J 2.0 O>- O FRANKLIN CO., OHIO u o A UNION CO, INDIANA O o ■ HOKE CO., NORTH CAROLNA D BATH CO,' VIRGINIA _ l ______|______■ i______!______I------I------1 20 30 40 50 60 F ig . 73 PULSES PER SECOND 126

Table IX. T rill characters of Oecanthus niveus; field record­ ings. RDA = recording made by R. D. Alexander.

T rill Duration T r i l l s T r i l l Temp. No, o f L o c a lity ( seconds) P er P er Cent °F. T r i l l s Range Aver. M inute o f T o ta l

P ike C o ., A rkansas 77 9 1&-3& 1 .9 4 21.5 62.6

Pike Co., Arkansas 77 7 2^-9 4.61 11.3 78.1

Jones Co., Georgia 66 6 3 /4 -3 i- 1.90 2 5 .8 73.5

Piatt Co., 111. (EDA) 65 5 2'o--5 3.95 8 .8 52.5

Union Co., Indiana 69 40 £-4 3/4 1.58 32.9 78.0

Rapides Par., Louisiana 74 20 1_Qjl 3.42 15.0 76.9

Vernon Par., Louisiana 74 12 l | - 8t 4.31 13.3 85.9

Carrol Co., Ohio (RDA) 66 16 2.69 17.3 69.7

Franklin Co., Ohio SOl- 9 2-3 3 /4 2.57 9.6 37.0

Franklin Co., Ohio 66 11 i - S 3 /4 2 .0 2 14.7 44 .8

Franklin Co., Ohio 69-g- 23 3/4-2 1.31 1 1 .8 23.1

Franklin Co., Ohio 69-1- 6 1 3 /4-2-q- 2.05 19.5 60.1

Franklin Co., Ohio 74 6 i - i 'i 1.27 15.3 26.2

Franklin Co., Ohio 75 9 i - 3 i 1.64 26.6 65.8

Franklin Co., Ohio (RDA) 71 10 2§-3§- 3.39 8.7 44.5

Shelby Co., Ohio 76 21 i - l i 1.04 24.2 38.0

Shelby Co., Ohio 77 25 3 /4 -2 ± 1.42 19.3 41.2

Average (of averages) 2.42 17.3 56.3 127

The trill characters of niveus in the field are listed in Table

VII. One feature of the trills which is not apparent from the tabular data is that the trill duration and interval are frequently quite regular for ten or fifteen trills. Since a trill of irregular duration is probably the primitive song type (Alexander 1956), the

singing of more regular trills is probably a specialization that can eventually result in a song such as that of rileyi in which trill

(chirp) duration and interval are held constant within narrow lim its.

On the other hand, the song of Keoxabea bipunctata frequently approaches a continuous trill, and further specialization in this direction would lead to a song like that in the latipennis and nlgricornls

g ro u p s.

Oecanthus rileyi Baker

The Snowy Tree Cricket

Figures 20, 23, 33, 44, 45, 46, 53, 54, 61, 62, 71, 73, 74, 75, 76, 79, 92, 101, 102, 103, 104, 108

Oecanthus rileyi Baker 1905, p. 81 (type locality, mountains near

Claremont, California; type, a male in the Pomona College

Collection, Claremont, California).

Acheta nivea, Harris (not De Geer 1773) 1833, p. 582. Jaeger and

Preston 1854, p. 159 (biology).

Oecanthus niveus, Harris (not De Geer 1773) 1841, p. 124. Fitch 1856,

p. 404 (morphology, song); Walsh 1866, p. 128 (food, song);

Walsh 1867, p. 54 (food, oviposition); F. Walker 1869, p. 93

(synonymy); Saussure 1874, p. 458 (synonymy); Harris 1883, p. 154 (song); Caulfield 1888, p. 60 (song); Davis 1889, p. 80

(song); McNeill 1889, p. 102 (song); Murtfeldt 1889, p. 130

(rearing); McNeill 1891, p. 6 (song); Blatchley 1892, p. 141

(biology, song); Scudder 1893, p. 65 (song); Beutenmuller 1894c, p . 269 (habitat, song); Lugger 1897, p. 359 (song); Saussure

1897, p. 253 (synonymy); Bessey and Bessey 1898, p. 263 (effect of temperature on song); Edes 1899, p. 935 (effect of temperature on song); Scudder 1900, p. 91 (synonymy); Faxon 1901, p. 183

(song); Scudder 1901, p. 212 (lists 82 additional references);

Blatchley 1903, p. 446 (biology, song, synonymy); Houghton 1903, p. 150 (rearing); Garman 1904, p. 64 (song); Houghton 1904, p,

61 (rearing); E. Walker 1904, p. 253 (biology, song, synonymy);

Forbes 1905, p. 217 (song); Felt 1906, p. 602 (biology); Kirby

1906 (synonymy); Shull 1907, p, 213 (rate of chirping, synchro­ nism); Jensen 1909a, p. 26 (courtship); Jensen 1909b, p. 26

(oviposition); Parrott 1909, p. 125 (oviposition); Allard 1910a, p. 33 (habitat, song); Allard 1910b, p. 352 (effect of tempera­ ture on song); Allard 1911a, p. 28 (song); Allard 1911b, p. 155

(song); Parrott 1911, p. 216 (oviposition); Allard 1912, p. 460

(song); Parrott and Fulton 1913, p. 177 (oviposition); Parrott and Fulton 1914, p. 430 (detailed account of biology, morphology, and song); Filton 1915, p. 22 (detailed account of biology, morphology, and song); Parrott, Gloyer, and Fulton 1915, p. 535

(biology, vector of fungus); Gloyer and Fulton 1916, p. 3 (vector of fungus); Rehn and Hebard 1916, p. 295 (habitat); Allard 1917, p. 411 (synchronism); Allard 1918, p. 548 (synchronism); Blatchley 1920, p. 714 (biology, song); Morse 1920, p. 404 (biology, song,

synonymy); Lutz 1924, p. 357 (bearing, song); Fulton 1925, p.

363 (detailed account of biology and songs of two "races”);

Snodgrass 1925, p. 439 (song); Fulton 1926a, p. 10 (biology,

song); Fulton 1926b, p. 59 (key); Wakeland 1927, p. 5 (biology,

song); Fulton 1928a, p. 446 (synchronism); Fulton 1928b, p. 555

(synchronism); Allard 1929a, p. 571 (synchronism and effect of

temperature on song); Hebard 1929, p. 309 (hp'Atat); Allard 1930a,

p. 131 (effect of external conditions on song); Allard 1930b, p.

347 (effect of air currents on song); Browne 1931, p. 633 (biolo­

gy) ; Johnson 1931, p. 117 (cytology) Fulton 1932, p. 66 (song);

Fulton 1934, p. 264 (synchronism); Hebard 1935, p. 78 (habitat,

song); Hebard 1938, p. 101 (habitat); Lutz 1938, p. 342 (song

analysis); Urquhart 1941b, p. 30 (habitat); Frost 1942, p. 202

(effect of temperature on song); Cantrall 1943, p. 45 (biology,

song); Pierce 1948, p. 140 (song analysis); Hallenbeck 1949, p.

256 (effect of temperature on song); Alexander 1956, p. 186

(biology, discussion and analysis of song).

Oecanthus niveus is established as the name of the snowy tree cricket, but its use for this species cannot be continued because the type specimens of niveus are not the snowy tree cricket, but the narrow-winged tree cricket (see the discussion under the latter s p e c ie s ) .

There is only one name, riley i, which remains available as a specific name for the snowy tree cricket. The species Oecanthus 130 rileyi was described by Baker, in 1905, "from one male taken in the mountains near Claremont, California.” This specimen is in the

Pomona College Collection and is on loan to the U. S. national Museum.

In the fall of 1955, I examined the type and could find no differences between it and specimens of the snowy tree cricket from the East. The situation is complicated, however, by the occurrence of two "races” of the snowy tree cricket on the West Coast (Eulton 1925, Smith 1930).

The two races have not been separated yet on the basis of morphology, but their song, egg-laying behavior, and habitat are quite different although they occur in the same localities. One form, "Race A," seems to correspond to our eastern snowy tree cricket, while the other form, "Race B," is probably a distinct species.

If Baker’s specimen is Race B, when Race B is described as a distinct species, rileyi w ill be its name; and Race A, including the eastern populations, w ill have to be given a new name. On the other hand, if Baker’s specimen is Race A, rileyi can continue as the name for the eastern form of the snowy tree cricket. Other closely related

species of tree crickets with different songs have been separated on the basis of the structure of the file, and it is possible that Race A and Race B can be separated on the same basis.

Habitat and Life History

The snowy tree cricket is most abundant in settled areas, where

it occurs in shrubbery and vines about houses, in unsprayed fruit

trees, and in the tangles of vines and trees that grow in neglected 131 fence rows. It is also found in wooded areas, but usually in shrubby undergrowth and along the edges of clearings.

About Columbus, Ohio, the species was common in planted shrubs.

It was collected in ivy growing on the walls of buildings, in unsprayed apple trees, in bushes and trees growing at the edges of woods, and in fence rows. It was recorded from, elm, haw, and wild cherry and from Virginia creeper growing on ash. In Erie County,

Ohio, it was found singing on blackberry shoots in an abandoned field.

At Cedar Point, Ohio, it was found in full chorus on rank cocklebur, blackberry, and sumac growing about a solitary large ash tree at the edge of a field. Hone was heard in the tree. In Hoke County, Horth

Carolina, it was found singing in small red oaks growing in an abandoned field. On Warm Springs Mountain, Virginia, it occurred in small oaks and shrubby undergrowth.

Allard (1910a) (Thompson's M ills, Georgia) reported rileyi in low oaks in woods by the roadside and in woody undergrowth in the mountains. Pulton (1915) (Geneva, Hew York) found it "most abundantly in apple orchards" and "more or less common in plantings of other fruit trees and in raspberry plantations, shrubberies, vines and bushy fence rows." Occasionally he heard it among forest trees. Rehn and

Hebard (1916) (mountains of the Southeast) stated that the species is

"both dendrophilous and thamnophilous." Morse (1920) (Hew England) found it "in dense tangles of shrubbery and vines or upon the branches of trees." Hebard (1938) (Pennsylvania) reported that it "prefers vines, apple trees, and bushes." Urquhart (1941b) (Ontario) found it 132 in an open clearing in a hardwood forest. Cantrall (1943) (George

Reserve, Michigan) found it in the "deciduous-arboreal stratum" and in the "tall-shrub stratum" of oak-hickory woodland. Alexander (1956) reported it in bushes and small trees between a damp forest and a large permanent marsh at Bemis Woods in DuPage County, Illinois.

In Oregon, where both Race A and Race B occur, Fulton (1925) reported Race A (seemingly corresponding to the snowy tree cricket of eastern United States) as almost completely arboreal, commonly "on prune and apple and in the native growths of white ash and Gary Oak."

He also found it in cherry, maple, and poplar trees and reported that

"it is usually more abundant among the higher branches and could be heard singing in the tops of quite large trees." The only berry bushes he found it in "were ta ll, coarse blackberries growing under trees." Fulton reported Race B as "preeminently a bush inhabiting form ...very common on loganberry and raspberry."

Oecanthus rileyi matures about the middle of July in central

Ohio, The earliest record, July 7, is of an individual in ivy growing on the south-facing wall of a brick building. Some indivi­ duals live until frost, although the species is noticeably scarcer by late September. There is no indication of a second generation of rileyi in any part of its range.

Distribution

0 . rileyi ranges farther north but does not occur as far south in the eastern states as do the other arboreal species of tree crickets

(Fig. 71; Table XXV). East of the Mississippi River, the species 133 extends southward along the mountains into northern Georgia (Allard

1910a). Two personal records from the sand h ill area of North Carolina do not fit in with the pattern of distribution suggested by the other records. Dr. B. B. Fulton, who has worked extensively with the sing­ ing Orthoptera of North Carolina, said (personal communication) that these records were the only ones from North Carolina that he knew of other than in the mountains in the west. Since rileyi lays its eggs in the bark of many cultivated shrubs, there is a possibility that it was recently introduced into the area; however, the trees in which it was found were not near any planted site.

The species ranges all the way to the West Coast and is found farther south, west of the Mississippi River. It is known to occur in New Mexico (Hebard 1925, 1935), Arizona (B. B. Fulton Collection and Ohio State University Entomological Collection = BBF, OSU),

Colorado (B3F, Hebard 1929), U tah (BBF, OSU), Nevada (BBF, OSU),

Idaho (BBF, Wakeland 1927), C a lifo rn ia (BBF, Browne 1931), and Oregon

(Fulton 1925). Records from Washington and B ritish Columbia (Hebard

1925) could represent Race B instead of Race A.

Reports of the species from Cuba and central Mexico (Fulton 1915) require confirmation.

In the eastern United States, records from Minnesota (Lugger

1897) and New Jersey (Fulton 1915) are not indicated on the distribu­ ti o n map. 134

Discussion of Song

Tlie calling song of 0 . rileyi is easily distinguished from all

other insect sounds in eastern United States. It is the only regular

chirp which is low-pitched and musical and issues from trees and

h u sh es.

There has heen general agreement as to the nature of the song of

riley i, yet probably more has been published on it than on the song

of any other Orthopteran. One reason is that the song of rileyi is more pleasing to the human ear than most insect noises. McNeill

(1889) reported that Burroughs teimed the song "a rhythmic beat,” while Thoreau spoke of it as ’’slumbrous breathing" and the "intenser

dream" of crickets, and Hawthorne described it as "audible stillness"

and declared "if moonlight could be heard it would sound like that."

Allard (1911a) found that a chorus of rileyi "induces a peculiar,

indescribable psychic state — an intermingling of sadness and

reposeful meditation."

Most of the attention to the song of riley i. however, has been

the result of two more tangible phenomena — the effect of temperature

upon chirp rate and the occurrence of synchronous singing among

neighboring individuals. The former was discussed in a previous

section (p. 3 4 ), and the latter w ill be dealt with in the section on

behaviorial effects of tree cricket calling songs.

There remains to be discussed the pitch of the song, the number

of pulses per chirp, and the relation of the present data on effect

of temperature to the data of others. The data on the song of rileyi 135 in this study are almost entirely from individuals from Franklin

County, Ohio. Recordings from Hoke County, North Carolina; Bath

County, West Virginia; Erie County, Ohio; and Union County, Indiana, do not deviate from the Franklin County averages by more than the usual range of individual variation. Field recordings do not differ from laboratory recordings except in being more variable.

The pitch of the song of rileyi has been quantified by six authors. Three (Fulton 1915; Wakeland 1927; Matthews 1942) used aural methods and gave values which are probably subhaimonics rather than the fundamental since they are lower than those determined by machine analyses. All three writers noted a rise in pitch with a rise in temperature. Three authors have used electronic devices to determine pitch. Lutz (1938) found a pitch of 1700 cps at 59° F .;

Pierce (1948) obtained simultaneous values of 1222 and 2125 cps; and

Alexander (1956) reported frequencies of 2500 to 3000 cp3. These values agree with the data obtained in this study (Fig. 73) except for Pierce's loxver value, which perhaps was due to a damaged wing.

The number of wing strokes involved in each chirp was first

studied by Fulton (1925) (Oregon), who simply counted the pulses in the chirps of crickets singing at 48° F. In Race B he observed four pulses per chirp and in Race A only three. Lutz (1938) pictured machine-made graphs of two chirps, one containing seven pulses, and the other eight. Pierce (1948) (New England), using machine analysis,

found that each chirp consisted of "a train of about eight pulses."

Alexander (1956) reported chirps containing two to ten pulses with

the most frequent number being eight. Table X shows the results of 156

Table X. Frequency of occurrence of chirps with various numbers of pulses in the song of 0. riley i; based on the analysis of 3062 chirps in 63 songs of 27 individuals. Parentheses indicate that only one chirp of that number of pulses was observed.

Number of Pulses P er Cent Most Frequent P er Chirp O ccurrence Spacing of Pulses

1 < .1 w

2 . 6 XX

3 .3 XX X

4 .8 XX XX

5 11.2 XX XXX

6 2 .1 XX XXX X

7 7.7 XX XXX XX

8 76.4 XX XXX XXX

9 .5 XX XXX xxxx

10 .2 XX XXX XXX XX

11 .3 XX XXX XXX XXX

12 < .1 (xx XXX xxxx xxx)

counting the pulses in the first fifty chirps of 57 recordings. Pulse

counts for six recordings with less than fifty chirps are also

in c lu d e d .

The spacing of the pulses within a chirp is usually not entirely uniform — each chirp begins with a group of two pulses closely

followed by consecutive groups of three (Figs. 74 to 76). Table X

indicates the most frequent pulse spacings in the different lengths

of chirps. Chirps consisting of a whole number of groups are more HATH XXXVII 137 VARIATIONS IN THE SONG OF Q. RILEY1

USUAL SON G -ALL 8-PULSE CHIRPS 75° F

Q 3r oZ u UJ if) 2 - cc 0 0.5 5 2.0 UJ F ig. 74 a if) Ul ALTERNATION OF 5 - AND 8-PULSE CHIRPS 7 4 ° F u > u 3- 1 , o |U U _J m 2- 1 0 0.5 1.0 1.5 2.0 F ig. 75

ABERRANT CHIRPS

fNORMAL) 8-PULSE CHIRP 5-PULSE CHIRP

60°E

0 .1 .2 .3 0 .1 .2

6-PULSE CHIRP 7-PULSE CHIRP

60°F.H H ■ i t « ■ -£ at*. V''.VT.. .,

.2 0 .2 .3

(NORMAL) 8-PULSE CHIRP PULSE CHIRP

80°r ii*4 #♦*!♦§

.2 .1 .2 Fig. 76

TIME in SECONDS 138

common than chirps containing one pulse more or less than a whole

group. For instance, 2-pulse chirps are more frequent than 1- or 3-

pulse chirps, and 5-pulse chirps are more frequent than 4- or 6-pulse

c h irp s .

Individuals frequently produce nothing but 8-pulse chirps for minutes at a time, the infrequent chirp lengths being produced prin­

cipally by individuals in discontinuous song. Occasionally an

individual w ill alternate 8- and 5-pulse chirps for several hundred

chirps (Fig. 75). Fulton's (1925) report of consistent production of

3-pulse chirps in rileyi in Oregon differs from any observations made

in Ohio. Fulton made his counts on individuals singing at 48° F.,

but while very low temperatures may result in a greater proportion of

short chirps, never has consistent production of short chirps been

reco rd ed .

The effect of temperature upon chirp rate in Ohio individuals

may be summarized by the regression line ch/m = 4.560 T - 184.53.

Compared with the results of studies in other localities, the Ohio

data supplement Fulton's (1925) observation of a clinal variation in

chirp rates from east to west. The Ohio rileyi are slower than the

ones studied by Bessey and Bessey (1898) in Nebraska but faster than

those studied by Edes (1899) and others in New England.

The latipennis Group

Living or recently killed specimens of the latipennis group from

the eastern United States are easily recognized by the red or reddish-

pink color of the frons and basal segments of the antennae. The red 139 pigment fades with time, but members of the group can s till be recog­ nized by an unswollen first antennal segment in conjunction either with tegmina about half as broad as long in the male or with a broadly notched subgenital plate in the female.

In the latipennis group the calling song is a continuous trill, at least in those species which have been studied. The usual habitats are scrubby trees, tangled thickets, and brambles. During oviposition the female places several eggs in the central pith of a stem through a single opening in the outer woody-layer. The eggs are arranged in two groups, one above and one below the opening.

There are only two species of the latipennis group in the eastern

United States. 'They may be separated (p. 96) on the basis of the antennal coloration, the structure of the terminal segment of the maxillary palp, and the number of teeth in the file.

Oecanthus latipennis Riley

The Broad-Winged Tree Cricket

Figures 1, 10, 11, 12, 13, 15, 16, 24, 34, 77, 78, 79, 92

Oecanthus latipennis Riley 1881, p. 60 (type locality, Missouri (?);

type, a male in the U. S. Rational Museum, Washington, D. C.).

Riley 1886, p. 182 (song, biology); Murtfeldt 1889, p. 130 (biolo­

gy); Beutenrauller 1894c, p. 272 (habitat, song); Lugger 1897, p.

363 (habitat); Scudder 1900, p. 90 (synonymy); Scudder 1901, p.

211 (lists 18 additional references); Blatchley 1903, p. 445

(biology, song); Forbes 1905, p. 217 (oviposition); Felt 1906, p.

303 (egg parasite); Kirby 1906, p. 75 (synonymy); Allard 1910a, 140

p. 36 (habitat, song); Allard 1910b, p. 356 (song); Allard 1911b,

p. 155 (song), Allard 1912, p. 461 (song); Pulton 1915, p. 41

(detailed account of biology, morphology, and song); Rehn and

Hebard 1916, p. 299 (habitat); Fox 1917, p. 239 (habitat);

Blatchley 1920, p. 725 (biology, song); Hebard 1925, p. 37 (habi­

tat); Snodgrass 1925, p. 439 (song); Fulton 1926a, p. 17 (mor­

phology); Fulton 1926b, p. 60 (key); Allard 1929a, p. 571 (song);

Johnson 1931, p. 117 (cytology); Fulton 1932, p. 63 (habitat,

song); Hebard 1934, p. 254 (habitat); Hebard 1938, p. 102 (habi­

tat); Lutz 1938, p. 342 (song analysis); Fulton 1951, p. 93 (song);

Alexander 1956, p. 184 (biology, discussion and analysis of song).

Riley’s original description of latipennis was based on fifteen specimens from Missouri, one male from Alabama, and one male from

Columbus, Texas. Riley’s statement that the Texas specimen had

’’black mark3 on the lower surface of the basal joints of the antennae" indicates that it was of a species other than latipennis (probably v a r ie o m is ) . I have not examined the type, but presumably it is one of the Missouri specimens.

Habitat and Life History

Oecanthus latipennis is common in the vines, brambles, and coarse weeds that grow along woodland edges, fencerows, and roadsides.

It also occurs on shrubs and low trees, particularly scrubby oaks.

It is seldom found more than four feet off the ground but occasionally is taken ten feet or higher in the leafy branches of trees. 141

In central Ohio I found it in greatest abundance in Tine en­ tangled coarse weeds and brambles along the edges of woods and thickets. It was also collected on coarse weeds (such as horseweed and teasel) in abandoned fields with no trees or woody vines near by.

In thickets it was found on saplings of various deciduous trees and on grapevine and Virginia creeper. In Virginia and Kentucky I collected it on small oaks in dry open woods, and in western Tennessee

I collected a single specimen from ten feet up in the branches of an elm growing in a wash in an open pasture.

Allard (1910a) (Thompson's M ills, Georgia) found that latipennis

"prefers thickets of asters, goldenrods, and brambles in low grounds."

Pulton (1915) reported that VI. T. Davis found it common among the oak scrub of Long Island, Staten Island, and New Jersey. Rehn and Hebard

(1916) (the Southeast) stated that it "seems to prefer low oaks and oak shoots in woodland." Vox (1917) (Virginia) collected latipennis in an old stubble field, roadside thickets, and in dry open woods on a mountain summit. Blatchley (1920) (Indiana) stated that "it is found mostly on shrubs and vines along fence rows, roadsides, and especially in thickets along the borders of streams." Hebard (1925,

1934) (South Dakota, Illinois) found that it "lives in trees, bushes, vines and ta ll weeds, but shows a very decided preference for oaks, particularly of the scrub type." Pulton (1932) (North Carolina) stated that it occurs mostly in bushes and vines in or near woodland.

Oecanthus latipennis matures later than any other tree cricket.

In contral Ohio the earliest record is August 14. Even much farther

south it does not mature earlier. Pulton’s earliest record at Raleigh, 142

Worth Carolina, is 22 August. The majority of the adult population

seems to live until frost. There is no evidence of a second genera­

tion in any part of its range.

Distribution

Both southward and northward, Oecanthus latipennis is more

restricted in its distribution than the eastern members of the rileyi

group (Tig. 77; Table X3C7T). State records not indicated on the map

are Alabama (Riley 1881), Oklahoma (Troeschner 1954), Missouri (Riley

1881), eastern Nebraska (Hebard 1925), Minnesota (Lugger 1897), and

Michigan (Tulton 1915). The limiting factor northward may be the

length of the growing season, since latipennis takes longer to mature

than other species of tree crickets. There are no records from west

of the hundredth meridian.

Discussion of Song

The loud, low-pitched character of the song of latipennis dis­

tinguishes it from the songs of other eastern species which produce

continuous trills. At 70° T. the song of latipennis is about 2800

cps, while the songs of the nigricom is group (excluding pini) range

from 3300 to 3600 cps. Oecanthus pini sings at about 3100 cps at

70°, but its habitat is different from latipennis, so there is little

chance of confusing the songs of the two in the field. The song of

latipennis has few irregularities, and this factor combined with low

pitch and high intensity gives it the clear bell-like quality which

has often been noted. HATE XXXmi 143 DISTRIBUTION OF 0. LATIPENNIS AND O. VARICORNIS

o l i t e r a t u r e r e c o r d ® SONG RECORD • SPECIMEN EXAMINED VARICORNIS F ig . 77

EFFECT OF TEMPERATURE ON PULSE RATE O. LATIPENNIS, FIELD RECORDINGS 50 -

Z40

REGRESSION LINE FOR LABORATORY RECORDINGS 30 KEY TO LOCALITIES u o FRANKLIN Cd, OHIO ■ GRAYSON AN0 LYON COUNTIES* KENTUCKY • LICKING CO., OHIO 20 ■ a CHESTERFIELD CO, VIRGINIA DYER CO, TENNESSEE

SO 60 70 80 DEGREES FAHRENHEIT Fig. 78 144

All the descriptions of the song in the literature agree as to its physical character, except Pulton's (1915) pitch determination

(1152 to 1216 cps), which must have been of a subharmonic rather than the fundamental. Lutz's (1938) determination of 2700 cps at 45 p/s agrees perfectly with the present data.

Tbere is some disagreement as to the daily periodicity of singing in latipennis. Fulton (1915) reported that in southern Ohio "the males began singing in full chorus at dusk and continued throughout the night"; he later (1932) stated that in North Carolina latipennis sings "at night only." However, Allard (1910a) (Thompson's M ills,

Georgia) reported that it "sings by day and late into the night during moonlight nights," and Blatchley (1920) (Indiana) described the "day note" of the male and then quoted another description, "probably of the night song." I have not observed latipennis singing before dusk, but E. S. Thomas (personal communication) stated that he had observed it doing so occasionally. Probably Allard's and Blatchley's observa­ tions were not representative, and the species has essentially the same daily periodicity throughout its range.

In the relationship of pulse rate to temperature, recordings of latipennis made in the field in Virginia, Tennessee, and Kentucky agree with recordings made in the field and in the laboratory of Ohio individuals (Fig. 78). Likewise the pitch relationships show no significant variation. 145

Oecanthus varicom is F. Walker

The Different-Horned Tree Cricket

F ig u res 77, 79

Oecanthus varicom is F. Walker 1869, p. 94 (type locality, Mexico;

type, a male, lost or destroyed). Kirby 1906, p. 75 (synonymy);

Davis 1907, p. 174 (morphology).

Although varicornis was originally described in Catalogue of the

Specimens of Dermaptera Saltatoria ... in the Collection of the

B ritish Museum, David R. Ragge, Department of Entomology, B ritish

Museum, w rote (p e rso n a l com m unication), ’’U n fo rtu n a te ly th e type material of Oecanthus varicom is .., has never been in our collection here, so I cannot help you .... I should imagine that this material

is now lost or destroyed; if it still exists I cannot imagine where

it might be." In a later letter he wrote that there were no specimens

of 0. varicom is determined as such by Walker in the collection, and

that the type had "quite likely" been returned to the collector by

Walker and that he (Ragge) did not think there was any chance of it

being located now.

The original description of varicomis is sufficient to place

material in the species; however, the prevailing concept of the

species does not agree with Walker’s description. Since it is not

available in most libraries, I w ill quote in full the original English

description (F. Walker 1869) (there is also a briefer Latin descrip­

tio n ) . 146

Male. Pale testaceous, smooth, shining. Head elongated, a little narrower than the fore border of the prothorax. Eyes rather small, hardly convex. Palpi slender, filiform; third joint longer than the second. Antennae full twice the length of the body, black towards the base, testaceous at the base. Prothorax narrower in front; sides straight. Legs very slender; hind tibiae minutely serrated on each side and with a few short slender spines beyond the middle. Wings pellucid. Eore wings very broad, extending much beyond the abdomen; veins very pale testaceous; mediastinal vein with fourteen oblique branches; eight near the base, approximate to each other; the other six remote from each other. Hind wings brilliantly iridescent, extending much beyond the fore wings; veins white. Length of the body 6 lines. The colour of the antennae and the broader fore wings distinguish this species from GE. niveus. a.. Mexico. Prom Mr. Shuckard’s collection.

In early April, 1957, J. N. Knull brought me ten tree crickets he had collected on 25 March in several localities in Hidalgo County,

Texas (just west of Brownsville). They matched Walker’s description in every respect except the general color was light green rather than

’’pale testaceous”; however, in older specimens, from Brownsville, in the Ohio State University Entomological Collection, the green has faded to the color described by Walker. The number of branches of the mediastinal vein (subcosta) was not constant, but varied from 13 to 16 in the right tegmina of five males. The basal branches of the media­ stinal vein numbered 7, 8, 8, 8, and 8 in the five specimens, while the more distal branches numbered 6, 6, 8, 6, and 7 respectively.

The antennal coloration of the fresh specimens was particularly striking and justified the name given the species by Walker. The first and second segments and the dorsal proximal area of the third were light in color and suffused with red. The remainder of the third segment and the immediately succeeding segments were black. Prom 147 about the tenth segment the antennae became gradually lighter, and the distal portion was dusky green. The green and red pigments are faded in older specimens, but the contrasting light and dark areas remain evident.

In the recently killed specimens the area between the eyes and between the antennae was mottled with reddish pink. There was a narrow black line on the inner ventral margin of the first and second antennal segments. Of 22 specimens from Hidalgo and Cameron Counties,

Texas (the 10 specimens in the author's collection and 12 older specimens in the Ohio State University Entomological Collection), all had antennal lines, though some were faint. The hind femur had a short narrow black longitudinal line on the dorsal surface near the distal end, and in some specimens it had other dark markings. The tegminal membranes were transparent except for scattered areas of faint infuscation. The measurements of these 22 specimens are given in Table XI.

These specimens comply with Walker's summary in that they may be distinguished from what he called niveus by the "color of the antennae and the broader fore wings." In regards to what Walker called niveus,

Dr. Ragge wrote that a male from Jamaica identified as niveus by

Walker has tegraina measuring 11.4 mm by 5.1 mm and antennal markings like rileyi.

Even though the Texas specimens agree with Walker's description, they cannot be definitely assigned to varieornls unless specimens of the same species occur in Mexico, the type locality of varicom is

Dr. Ashley B. Gurney of the U. S. National Museum wrote (personal 148

Table XI. Measurements of 0. varicom is from Hidalgo and Cameron Counties, Texas i Fronotal width taken at posterior margin, but impossible to measure uniformly because the lateral margins spread to varying degrees during pinning and drying. Tegralnal width measured at the widest portion of the dorsal field of the right tegmen. All measurements in m illim eters.

Fronotum Right Tegmen Length Length of w W L of Hind Ovipositor W 1 L w Femur 12 m ales extrem es 2 .4 2 .7 .96 12.6 6.0 .47 4.2 8 .1 2 .9 3 .2 1.21 1 4 .1 6.9 .53 5.2 10.0

average 2.67 2.88 1.082 13.33 6.67 .501 4.65 9.15

10 females extrem es 2 .4 2 .0 .80 10.8 4.5 8 .3 5.7 2.7 2 .9 1.21 13.0 5 .4 9 .8 6.4

average 2.59 2.29 .912 11.54 5.02 9.07 6.12

communication) that he had from both northern and southern Mexico specimens of tree crickets which evidently belong to the varicornis of this paper.

The name varicornis has been correctly applied to tree crickets of the United States only once before. Davis (1907) called specimens collected at Brownsville, Texas, and in southern Arizona varicornis.

saying the species "is of the same size as OE. latipennis. Riley, and the head, as in that species, is also coloured pink, but in all but two examples examined there is a single narrow black line on each of the first two antennal joints. These two joints are light-coloured, and are generally pink; the succeeding ten or twelve are black, and

the remainder gradually shade off and are of a lighter hue." In spite of Davis's comments, the name varicornis is still used to refer to a species -which does not fit Walker's original description.

Saussure (1874, 1897) was originally responsible for the m isidentifi- cation, and Rehn (1904b) and others perpetuated the mistake. While at the Academy of Natural Sciences of Philadelphia, I saw material from Mexico and Central America which Rehn had identified as varicom is on the basis of Saussure*s concept. The antennal markings were like those of 0. exalamationis. and no part of the antenna was black as described by Walker. The wings were proportionally narrower than P.

Walker's niveus rather than wider. The pronotum was If to 2 times as long as wide. Saussure (1874, 1897) did not give the width of the male tegmina of his varicom is nor did he describe the antennae other than to mention the black lines on the first two segments; however, a figure showing the pronotum of his' (1897) varicornis demonstrates that his specimens could have been the same as Rehn's and could not have been like those here assigned to varicom is. That Saussure did not have specimens of Walker's varicom is is further evidenced by his

(Saussure's) statement (1874) that P. Walker's Oecanthus peruvianus was probably synonymous with varicom is since the two were distin­ guished on the basis of antennal color, which "depend de la dessicca- tion des individus."

Not only is the name varicom is applied to an inappropriate

species, but another name has come into use for the species that is

correctly called varicom is. The specimens from southern Texas fall within the range of variation described by Pulton (1926a) for

Oecanthus califom icus Saussure, and Pulton's (1926b) key to the 150

Oecanthinae of North America places the Texas specimens in californicus.

I hare not seen the type of californicus, hut Saussure*s (1874) original description indicates that it differs from the Texas speci­ mens of varicornis in having shorter (11.5 mm) and proportionally wider (width/length * .54) tegmina and no dark lines on the basal segments of the antennae. Saussure mentioned no contrasting lighv snd dark portions of the antenna.

Whether Saussure*s californicus is a species distinct from

Walker’s varicom is is a question which I cannot at present answer.

Specimens similar to californicus and varicom is from the United

States west of the Great Plains are generally classified as Oecanthus c a lif o r n ic u s , even though th e re i s c o n sid e ra b le v a r ia tio n among specimens as to coloration and proportions. If there is more than one species involved, californicus may not be a synonym of varicom is but an available name for another species. Whatever the case, it is safe to apply the name varicom is instead of californicus to the

Texas specimens since varicom is is the older name.

Habitat and Life History

Professor Knull informed me that he collected the specimens of varicorri s by beating bushes and the limbs of trees in wooded areas near the Rio Grande River.

The inclusive dates of collection for the 22 specimens studied are 24 March to 22 Kay. 40 NUMBER OF TEETH 45 25 20)- 20)- 30 35r 15- . 10 . 14 . 1.8 1.6 1.4 1.2 1.0 0.8 UBR F ET AD EGH F IH FILE RIGHT OF LENGTH AND TEETH OF NUMBER ______• • o BPNTT, IEI RU, AIENS GROUP LATIPENNIS GROUP, RILEYI BIPUNCTATA, N o*> 1 ______LTPNI GROUP LATIPENNIS # A . m O o I ______RLY GOP (FRANKLIN OHIO) CQ, GROUP RILEYI ° RILEYI . 0 , D O D CD EGH F IE N MILLIMETERS IN FILE OF LENGTH I ______H A T S XXXDC, Figure 79 Figure XXXDC, S T A H I ______t ...... —I ey k ______AA A A A A

O NIVEUS O. * O LTPNI COHIO, KENTUCKY, O. LATIPENNIS * COHIO) £J. BIPUNCTATA o 0 VRCRI (IAG C, TEXAS) CO, (HIDALGO VARICORNIS 0. o o t AA Q. ■ i ■ ■ A A A

EXCLAMATION 13

s e i c e p s TENNESSEE) ------1— 151 152

Distribution

Except for its occurrence in extreme southern Texas and in

Mexico, the range of this cricket cannot be described until its relationship with what is now called californicus is determined.

Discussion of Song

There are no published descriptions of the song of this species.

Since 0. californicus in Oregon (Fulton 1926a) and £. latipennis produce continuous trills, the song of varicornis is probably also such a trill.

The file of varicornis has fewer teeth than the file of lati­ pennis (Fig. 79). In the rileyi and nigrlcomis groups the species with fewer file teeth have higher pulse rates at a given temperature.

If this same relationship holds in the latipennis group, varicomis has a higher pulse rate than latipennis.

The n ig r lc o m is Group

Members of the nigrlcornis group are distinguished by unswollen first antennal segments ventrally marked with black in conjunction with narrow tegmina in the male or a narrowly notched subgenital plate in the female. None of the group has red pigment on the head or an te n n a e .

The calling songs of the nigrlcom is group are continuous trills and are higher in pitch at a given temperature than the songs of the rileyi and latipennis groups. Females of the nlgricornis group 155 oviposit in rows in the stems of plants and make a separate puncture for each egg.

The number of eastern species in the nigricom is group is subject to controversy. At present most orthopterists accept Pulton's (1926b) breakdown of the group into two species, one of which is subdivided into three subspecies. The three subspecies that Pulton recognized —

Oecanthus nigricornis nigricom is. 0. n. quadripunctatus. and £. n. argentinus — were originally described as species, but Pulton could find no means of consistently distinguishing them. He concluded his discussion with this statement: "Until better characters are discovered for separating the three tree crickets of this group, it seems advis­ able to consider them as subspecies.”

In this study I have found biological and morphological charac­

ters which indicate that Pulton's three subspecies should be treated as full species and that there is an additional species which is common in the Southeast.

Two other forms, of uncertain taxonomic status, were also observed; both would have been classified as nigricornis nigricom is by Pulton. One form was pointed out to me by E. S. Thomas and differs

from typical nigricornis in several morphological characters and in

its occurrence on shrubs and trees near water instead of on coarse

old-field weeds. No areas in which the habitat of this form adjoins

the habitat of the typical form have been located and studied, so I

cannot say whether it is an ecotype (a variation within the species

nigricomis correlated with a particular type of habitat) or a 154 distinct species (a population reproductively isolated from typical nigricom is). In this paper it is called the "willow form of nigrlcom is."

The other form of uncertain status produces a trill with a slower pulse rate than that of the trill of the most-frequently collected form of nigricom is. There are no clear-cut morphological differences in the song forms, although the structure of the file will separate some specimens. In this paper the two forms are referred to as

"typical nigrlcomis" and "slow-trilling nigricom is." The use of nigricom is without a qualifier means that the song form was not distinguished. The possible taxonomic relationships of the two song forms w ill be discussed in the section dealing with the calling songs of the nigrlcomis group.

Rather than proceeding immediately with a species-by-speeies presentation as in the other groups, I will first discuss all the members of the nigrlcom is group in relation to a number of factors which are useful in differentiating the species.

Geographical D istribution

Figures 80 to 85 show the known distribution of the five species of the nigricom is group. Tables XXVTI to 2GQCII give the sources of the records plotted on the maps. Additional information on the distribution of each form is given below.

In addition to the localities marked on the map, nigricom is i s known from c e n tr a l M issouri (U. S. N a tio n a l Museum) and Aweme,

Manitoba (Fulton 1926b). Hebard (1928) listed specimens from two ELATE XL DISTRIBUTION OF 0. NIGRICORNIS DISTRIBUTION OF O. NIGRICORNIS. WILLOW FORM

/>

o l i t e r a t u r e r e c o r d • TAPE RECORDINC OF SONG • SPECIMEN EXAMINED • SPECIMEN EXAMINED • SPECIMEN AND RECORDING Of SONG • SPECIMEN AND RECORDING Or SONG 155 r i g . so Fig. 81 H A T E XLI DISTRIBUTION OF Q. CELERINICTUS DISTRIBUTION OF O ARGENTINUS

LITERATURE RECORD • TAPE RECORDING OE SONG • TAPE RECORDING OT SONG • SPECIMEN EXAMINED • SPECIMEN EXAMINED '* • SPECIMEN AND RECORDING Or SONG • s p e c i m e n a n d r e c o r d in g or s o n g 156 Fig. 82 Fig. 83 ELATE XLII DISTRIBUTION OF Q. PINI DISTRIBUTION OF Q QUADRIPUNCTATUS

o l i t e r a t u r e r e c o r d O LITERATURE r e c o r o • SONG RECORD • SPECIMEN EXAMINED 'V-V • s p e c i m e n e x a m in e d • SPECIMEN AND RECORDING OF SONG • SPECIMEN AND RECORDING OF SON 157 Fig* 84 Fig. 85 158 counties in Montana, but these records are far outside the range indicated by the other data and require confimnation. £. nigricornis and quadripunctatus occur farther north than other members of the ni­ gricom is group, but while quadripunctatus occurs far to the south also, nigricom is is restricted to the north and extends southward as far as North Carolina and Tennessee only along the Appalachian Moun­ t a in s .

The distribution of the willow form of nigricornis is inadequate­ ly known. S. S. Thomas, who first observed the form and called it to my attention, collected it on willow in several parts of Ohio. My own records are also from Ohio. In early October 1957, I visited willows at several sites in Dyer County, Tennessee, but did not find this form.

Records in the literature which may refer to the willow form of nigrlcom is are Cantrall (1943), George Reserve, Michigan; Hebard

(1938), southeastern Pennsylvania; and Blatchley (1920), Indiana.

0. celerinictus is largely restricted to the Southeast but occurs northward along the Atlantic coastal plain at least as far north as

Delaware. On two trips from Tennessee to Ohio, I made regular collections along the route and did not find celerinictus north of

Ohio County, Kentucky. There is a slight overlap in the ranges of celerinictus and nigricom is in central Kentucky. Typical specimens of both species were collected at the same site in Muhlenburg County by sweeping goldenrod, mixed with fleabane and dev/berry, on the edge of a drainage ditch. The westward limits of this species are uncer­ tain. Southward, it is not known from peninsular Horida. 159

_0. argentinus is largely a western species but extends eastward into central Ohio, Kentucky, western Tennessee, and M ississippi. A major portion of the range of this species is prairie, and it is possible that it has extended its range eastward in recent years as a result of agriculture making new areas favorable to it. It ranges to the West Coast (California, Oregon, Washington — B. B. Fulton

Collection, U. S. National Museum) and southward into Mexico (Ense­ nada — Ohio State University Entomological Collection).

0. quadripunctatus is the most widespread species of the nigrlcom is group in the eastern United States. It occurs from Maine and Ontario southward to extreme southern Florida. Its westward lim its are uncertain, since another species is perhaps confused with quadripunctatus in the area west of the Great Plains.

The distribution of 0. pini is poorly known because it usually remains high in pine trees and is difficult to collect. It does not seem to occur in the southeastern evergreen forest region. Horse

(1920) wrote that it occurs "as far south as Georgia” but did not substantiate his claim. On a trip through South Carolina, Georgia, northern Florida, and Alabama, 14 to 16 September 1955, I did not hear or collect pini. The species may of course have been evident earlier in the season. 0. pini may occur as far west as Minnesota; at least Alvah Peterson (personal cummunication) reported that he heard continuously trilling tree crickets in the pine trees at Itasca

State Park. 160

Ecological Distribution

Only two of the forms in the nigricornis group (pini and the willow form of nlgricom is) occupy habitats which are largely non­ overlapping with those of other forms. The remaining forms of the nigricornis group are frequently found intermingled upon the same herbaceous plants, although between some forms there is a degree of ecological separation.

Oecanthus pini is found almost exclusively in the crowns of pine trees but also occurs on the closely related tamarack. The principal pines from which it has been reported'are pitch pine and scrub pine

(Pinus rigida Mill, and P. virginiana H ill.). I collected it from young pines growing in abandoned fields in North Carolina and Ohio and by sweeping the vegetation underneath older pines in North

Carolina and Virginia. It was heard singing from the crowns of pines in the above-mentioned states and in eastern Maryland. 0. pini was described by Beutenmuller (1894a) from specimens collected from pine in Connecticut. Pulton (1915) reported it from pine in central

Pennsylvania, southeastern New York, and in New Jersey. Procter

(1946) collected it from pine in Maine, and Cantrall (1945) found it on tamarack in lower Michigan.

The willow form of nigricornis is known only from the shrubs and trees growing along the edges of lakes and streams. I found it on sandbar willow (Salix interior Rowlee) along the Olentangy and Scioto

Rivers in Pranklin and Delaware Counties, Ohio. At Lake Lorsmie

State Park, Shelby County, Ohio, it was abundant on sandbar willow 161 and dogwood (Cornus drummondi Meyer) growing at the edge of the lake and in pure stands of willow growing in low places near the lake. I observed a few individuals in multiflora rose along the lake and one individual on a lower branch of a large cottonwood tree. Dr. Thomas collected a series of this form from Salix at Sardinia, Ohio. In the literature there are at least three reports of the habitat of nigri­ cornis which may be based on observations of the willow form. Cantrall

(1943) reported nigricornis in the George Reserve, Michigan, only from dogwoods and shrubby Salix in the wet shrub-zone habitat. Hebard

(1938) stated that ”a decidedly atypic condition [of nigricornis*] is very abundant in the high weeds of the marshes along the Delaware

River in southeastern Pennsylvania,” and Blatchley (1920) (Indiana)

I’ecorded nigricornis from ”ta ll weeds and bushes along borders of lakes and ponds.”

Oecanthus nigricornis is usually found upon coarse weeds and brambles. As far as is known, slow -trilling nigricornis occurs on the same plants as the typical form. In Ohio, eastern Indiana, and

Kentucky I observed nigricornis in greatest abundance on goldenrod

(Solidago spp.) and giant ragweed (Ambrosia trifida L.). It was also found on many other coarse weeds — for instance, wild lettuce, blackberry, milkweed, sweetclover, and fleabane — in the more fertile old-field situations. The association of nigricornis with habitats of this sort has been noted by observers in other parts of its range:

Procter 1946, Maine; Allard 1911a, New England; Pulton 1915, New York;

Pulton 1932, mountains of North Carolina; Urquhart 1941b, Ontario;

Hancock 1911, Michigan; Blatchley 1920, Indiana; Pulton 1926b, Iowa. 162

08can.th.us celerinictus is most commonly found in neglected

fields on plants which are not so coarse as the ones upon which

nigricornis is typically found. However, it does occur on the heavier weeds and brambles and has even been collected from tree seedlings

growing in open situations. On a trip through Delaware, eastern

Maryland, and Virginia in late July 1956, I collected celerinictus most frequently on common ragweed (Ambrosia arteaisiifolla L.). I

also found it on wild carrot (Daucus carota L.), mature corn, soybeans,

lamb’s quarter, cocklebur, and fleabane. On a trip through North

Carolina, South Carolina, Georgia, northern Florida, and southern

Alabama in mid-September 1955, I collected celerlnictus commonly on

goldenrod, common ragweed, Andropogon, and other roadside herbs. It

was occasionally found in seedling trees and on blackberry. In mid-

June 1956, on a trip through M ississippi, Louisiana, and Arkansas, I

collected the species on daisy fleabane (Brigeron annuus (L.) Pers.),

common ragweed, Johnson grass, sassafras and postoak seedlings,

blackberry, and various other v/eeds, coarse and otherwise. In early

October 1956 in western Tennessee and Kentucky, I found the species on

cotton, daisy fleabane, goldenrod, blackberry, dewberry, and Johnson

g ra s s .

Oecanthus argentinus is found on essentially the same type plants

as celerinictus, except that it is less frequently taken on woody

plants such as tree seedlings and blackberry. In Ohio I found it in

greatest abundance on wild carrot, but it was also common on alfalfa,

sweet clover, timothy, and daisy fleabane. It often occurred in the 163 same fields as nigriconas but was usually on finer-stemmed plants in the more open areas. In mid-June 1956 in western Tennessee, M issis­ sippi, Arkansas, and Missouri, I found it with greatest consistency on daisy fleabane but also collected it from common ragweed, mullein, cotton, Lactuca, Johnson grass, Eupatoria, and other herbs. In western

Tennessee and Kentucky in September 1956, it was common on cotton, goldenrod, and daisy fleabane. In the areas in which their ranges overlap, argentinus and celerinictus were frequently found on the same plants.

In early July 1956, solitary males of argentinus were heard singing from ten to twenty feet up in deciduous trees (ash, maple, crab apple, sycamore) on the Ohio State University campus and in the surrounding residential areas. No immature stages could be found in the vicinity of the singers, so they evidently had flown from other areas. The singing ceased in about a week. Most tree crickets are able fliers, and there may be a fairly wide dispersal of argentinus males at the beginning of the singing season. Additional evidence for such a phenomenon is a tape recording of a solitary argentinus male from nine feet up in a red oak in an oak woods in Rapides Parish,

Louisiana, on 15 June 1956. The females may also disperse in this manner, but they would rarely be collected.

Records of the habitat of argentinus in the eastern United States are scarce in the literature. Pulton (1926b) reported the species on

horseweed and sunflower at Sioux City, Iowa; Alexander (1956) found

it in Illinois on alfalfa; and Beach (1938) collected it from ragweed

a t Law rence, K ansas. 164

Oecanthus quadripunctatus seems to be a common inhabitant of a greater range of plants than are the other species found in old fields.

In central Ohio and eastern Indiana, quadripunctatus commonly occurred on the same plants as both argentinus and nigricornis. In recently abandoned upland fields, quadripunctatus and argentinus would be common on wild carrot, timothy, and other fine-stemmed weeds, and if there were coarser weeds in low places or around the edges of the field, both quadripunctatus and nigricornis would be found there. In weedy red clover fields, adults of all three species could be found in late fall, although nigricornis would be restricted to the coarser weeds. In general quadripunctatus was most abundant on grasses and fine-stemmed weeds.

In South Carolina, Georgia, northern Florida, and southern

Alabama in September 1955, quadripunctatus and celerinictus were found in the same habitats, but quadrlpunctatus seemed slightly more prevalent on roadside grasses than celerlnictus.

In M ississippi, Louisiana, and Arkansas in mid-June 1956, quadripunctatus was frequent on the grasses and weeds of abandoned land, and singing males were often observed a few feet up in scrubby saplings, principally oak. The species was found with both celeri­ nictus and argentinus; however, in only one case — a weedy field dominated by common ragweed in the bottoms of the L ittle River, Sevier

County, Arkansas — were the three found together.

Various w riters have reported upon the habitat of quadripunctatus

in other areas of its eastern range. Fulton (1915) (New York) found

that "upland fields abounding in medium sized weeds such as aster, 165

sweet clover, daisy, golden rod, ragweed and especially wild carrot ...

form the favorite environment of this species.” Morse (1920) (New

England) reported it frctn "weedy jungles of pastures and along wood­

land edges composed of asters, goldenrod, everlasting, St. John’s-wort,

Joe-Pye-weed, etc. Also common on wild carrot in mowing lands, among

raspberry and blackberry bushes, in young birch thickets, in bush-

grown pastures, etc." Fulton (1926b) (Iowa) reported it from ragweed, horseweed, and sunflower; and Alexander (1956) (Illinois) found it on

ragw eed.

Seasonal Distribution

The members of the nigricornis group vary in the time of their maturity and in the number of generations per year. An example of

this variation is that of the species living in weedy fields in

central Ohio. 0. argentinus matures about the first of July, and the

adult population has nearly disappeared by the time quadripunctatus

and nigricornis mature. In mid-September a second generation of

argentinus matures, and from then until frost adults of all three

species can be found. While adults of quadripunctatus and nigricornis

appear almost simultaneously, late instar nymphs of nigricornis are

common for a while after those of quadripunctatus have all matured.

Figure 86 depicts the seasonal distribution of the three species in

1956 as derived from frequent observations and weekly collections in

two weedy vacant lots in Upper Arlington, Ohio (Franklin County).

Since early instar nymphs are difficult to identify, a portion of the 166 H A T E XLIII SEASONAL DISTRIBUTION OF THE FIELD-DWELLING SPECIES OF THE NIGRICORNIS GROUP, COLUMBUS, OHIO, 1956

JAN re B MAR APR MAY JUNE JULY | AUG NOV DEC

£1 NIGR CQPNI3

NYMPHS/ / / / ' /

Q. ARGENTINUS

O QUADRIPUNCTATUS

1 i ...... i...... F ig . 86

EFFECT OF TEMPERATURE ON PULSE RATE NIGRICORNIS GROUP REGRESSION LINES, LABORATORY RECORDINGS

100

90

80

70

LJ60

40 FIELO-OWELLmC SPECIES 1 SOUTHEASTERN UNITED STATES 2 NORTHEASTERN UNITEO STATES 3 WESTERN TENNESSEE AND KENTUCKY, MISSISSIPPI, 30 LOUISIANA, ARKANSAS 4 OHICL INDIANA, ILLINOIS, IOWA

60 70 80 90 DEGREES FAHRENHEIT Fig. 87 167 nymphs from each collection was reared to verify the identification of the remaining material.

Qecanthus nigricornis evidently has but one generation per year throughout its range, but quadripunctatus probably has at least two generations farther to the south. Adults of quadripunctatus have been collected in Florida as early as 8 May and as late as 4 November. In

Arkansas the inclusive dates are 16 June to 28 October. Qecanthus argentinu3 has at least two generations in most areas, but in the northernmost parts of its range there is time for only one generation on the basis of the duration of a generation in central Ohio.

Alexander (1956) and I noted only one generation in Columbus, Ohio, in 1955, but imuediately south and west I collected adults in October of that year. Beach (1938) reported two generations at Lawrence,

K ansas.

■2* celerinictus, the other old-field species of the nigricornis group, evidently has at least two generations throughout its range.

Fulton (1951) at Raleigh, North Carolina, recorded two well defined generations per year of Mquadripunctatus,w but at least part of his

observations must have been of celerinictus. Ashmead (1894) I’elated that eggs collected on cotton in M ississippi on 3 August hatched 6

August. His description of antennal markings suggests that he was

dealing with celerinictus.

In M ississippi, Louisiana, and Arkansas in June 1956, collections

of nymphs and adults showed that argentinus matured a little before

celerinictus and quadripunctatus. These latter two species evidently 168 completed development at more nearly the same time, but the collections suggest that quadripunctatus was slightly ahead of celerinictus.

Adults of Qecanthus pini have been collected as early as 24 July and as late as 28 October in Ohio. One fifth instar nymph and six adults were observed in low pines on 13 August 1956 in Fairfield

County, Ohio. Near Haleigh, North Carolina, adult pini have been collected from 10 July to 22 September, but Fulton (1951) reported that he had not heard singing after 10 September. 0. pini has but one generation per year.

L ittle is known of the seasonal distribution of the willow form of nigricornis. The earliest record for the adult is 18 July (Ottawa

County, Ohio) and the latest record is 13 September (Shelby County,

Ohio). S. S. Thomas collected late instar nymphs and adults at

Sardinia, Ohio on 6 August. There is probably but one generation per y e a r.

Calling Songs

All of the calling songs of the nigricornis group are continuous

trills, and with the exception of pini and possibly the willow form

nigricornis the species commonly sing during the day as well as at

night. At a given temperature the songs of the nigricornis group are

higher in pitch than those of the other groups.

Laboratory experiments on the effect of temperature upon pulse

rate in the songs of the nigricornis group were described in an earlier

section (p. 34). In Figure 87, the regression lines of pulse rate on 169 temperature for all of the fbrms are plotted on the same graph. When these lines are compared, it becomes apparent that in general those species which occur in the same habitats in the same geographical areas have widely different pulse rates. An exception is argentinus and slow -trilling nigricornis, which have similar pulse rates yet may occur together in Ohio fields in late September and October. However, they are usually segregated ecologically (see p. 162).

The relationship of the slow -trilling individuals of nigricornis to typical nigricornis is an interesting problem. I was not aware of their presence until late in 1956, so the information concerning them is scanty. Of thirteen individuals collected from coarse weeds in a vacant lot in Upper Arlington (Franklin County, Ohio), 21 and 23 August

1956, and recorded in the laboratory to determine the effect of temperature on pulse rate, two were slow trillers. Another specimen, collected by R. D. Alexander in Carroll Comity, Ohio, 16 September

1956, on willow, proved to have the same type song. It did not have the morphological attributes of the willow form but had the coloration and dimensions of old-field nigricornis. In 54 recordings of these

14 individuals, there were no intermediates in pulse rate among the two groups (Fig. 25).

Nine individuals from another vacant lot, less than a mile from the one mentioned above, were recorded in the laboratory while producing courtship and postcopulatory songs. Determination of pulse rate in these songs showed that seven were slow trille rs and one a typical singer. The remaining specimen appeared intermediate on the 170 basis of a single recording; however, the temperature data were not as carefully taken for these recordings as for those of the calling songs.

The slow singers did not differ from the others in coloration, markings, or dimensions; however, examination of the file structure of recorded specimens and others disclosed that specimens of nigri- corais from Franklin County, Ohio, fell into two groups as to number of teeth in the file. The grouping of specimens on the basis of file teeth is correlated with the grouping on the basis of song, but there is a slight overlap (Fig. 94). The specimen with the apparently intermediate song was unfortunately not preserved, and its number of

file teeth is unknown. Tooth counts of specimens from Scott County,

Kentucky (38, 40, 37, 38, 37, 37 teeth); three counties in eastern

Indiana (42 , 41, 40 , 38 , 37 , 42 , 36 , 5<0 teeth); and Carroll County,

Ohio (47_, 43, 43, 47_ teeth) indicate that a similar dichotomy in the nigricornis population may exist elsewhere.

In experiments with the response of female tree crickets to the

calling song of the male (see below), it was discovered that test

females of nigricornis (from the field which had mostly typical

trillers) would respond to the pulse rate of the typical song but not

to a pulse rate as low as that of the slow song (Fig. 112). That

there are females which respond to the slow song was not demonstrated.

Other than showing the need for more study of nigricornis, the

facts stated above suggest that (l) there are two reproductively

isolated, sympatric populations, i_.e_., species; or (2) there are

hybrids between nigricornis and same other species, and the hybrids 171 have a slower song than nigricornis but are otherwise nearly indis­ tinguishable from typical nigricornis; there may or may not be female hybrids of the same origin.

In the willow form of nigricornis, the laboratory recordings of

the songs of Shelby County individuals deviated slightly in pulse rate

from those of the Franklin County individuals (Fig. 26); however, the

deviation may have been due to individual rather than population variation, since only three Franklin County specimens were recorded.

Recordings made in the field in Shelby County agree perfectly with

those made in the laboratory.

In 0. celerinictus, the laboratory-recorded songs of specimens

from Virginia have essentially the same pulse rate relationships as

those from Arkansas (Fig. 27). Field recordings from localities

scattered throughout the range of the species shot/ similar relation­

ships (Fig. 88). The field recordings at the higher temperatures were made in bright sun, and the greater scattering of the points based on

these recordings is probably due to less accurate temperature

determination.

In 0. argentinus, laboratory recordings of individuals from Ohio,

Indiana, Tennessee, and Kentucky had similar pulse rates (Fig. 28).

Field recordings from these states and from Arkansas and Missouri fall

along th e same li n e (F ig . 8 9 ).

Laboratory recordings of quadripunctatus from Ohio, Indiana, and

Virginia show similar pulse rates (Fig. 29), and field recordings from

these states and six states in the Southeast fall into the same

pattern (Fig. 90). HATE XLIV 1' EFFECT OF TEMPERATURE ON PULSE RATE EFFECT OF TEMPERATURE ON PULSE K 0. CELERINICTUS Q. ARGENTINUS ' ill b PF.COPDi r JGS

' TO LOCALITIES rRAN«L'N CO, OHIO ■ D»ER. OB'ON, AND ROB* M COUNTIES, T{ NNTSStf • BENTON, St VIE B, AND ‘’i*f < " ht i FRANKLIN, NICHOLAS, CRAVES, i ' MUMltNBERO COUNTIES, KEN’ * BUTLER CO, MISSOURI 80 I UNION CO, INDIANA » FAIRFIELD CO, OHIO I PIATT CO, ILLINOIS Q l RAPIDES PARISH, LOUISIANA Oo 70

_j50 KEY TO LOCALITIES o W0RCC5TCR CO, MARYLANO

* PIK E ANO SEV IE R CO U N TIES, ARKANSAS o OVER ANO OBION COUNTIES, TENNESSEE ■ RAPIDES AND VERNON PARISHES, LOUISIANA a AIKCN AND CHESTERFIELD COUNTIES, SOUTH CAROLINA ■ HOLM ES C O , FLORIOA • BUTLER CO, MISSOURI 20 o RECORDCD BY p D a .L/a n DE 60 70 80 9 0 70 8 0 90 GREES FAHRENHEIT EGREES FAHRENHEIT Fig, F ig . 8 #

EFFECT OF TEMPERATURE ON PULSE RATE EFFECT OF TEMPERATURE ON PULSL RATt 0 QUADRIPUNCTATUS Q P J_N I

I IUIl.1 N MAPI* I V-N «'ll> ' 'I. ill ' -I • I N I lj(. A » I :>* /I UN'**,, AND MAINDI S PABlS«(S, TO l_ O L A I •',[S ■ mmMANA ■ A A «,E CO NORTm C' • urn wf *> ' o, M ississippi ' r A lR f It LD CO, 0**10 . UADlSON, I * ARK, f,«E I NE. «Of KING, AND h o c k in g rn , Ohio MU I B» < OUN I if S. OHifl BATH CO, VIRGINIA • UNION < II, INDIANA i III I PAM < O, U I INCUS ■ I UANKI IN (II, OHIO i >11)1 Ul M » I I (JHiQA > I ONI S I I), Gt ORGIA > AIKfN ill. M ill'll lARO,iNA o 'o , (uiK lN VIN I II, VIRGINIA u

-RCGRtSSiON LINE FOR LABORATORY ' '5 0 RECORDINGS ( A

3 0

O “CCOROEO BY R D ALEXANDER 20 O °l.CORDED H' ■ AHR[ i ’ 90 70 8 0 00 Fig* 90 F ig. 91 173

Laboratory (Tig. 30) and field (Fig. 91) recordings of pini reveal no differences in pulse rate relationships between the individuals from Ohio and those from North Carolina,

The relationship of pitch to pulse rate in laboratory recordings of the various forms of the nigricornis group was plotted in Figures

35 to 40, and analysis of field recordings showed similar relation­ ships. Figure 92 is a composite showing the lines for all forms as drawn by eye for each group of data. The lines for the other species of the Oecanthinae are included for comparison. It is evident that same recordings of species of the nigricornis group can be distinguished from others on the basis of the relation of pulse rate and pitch.

This character is useful even when the temperature datura of a recording is lacking or inexact.

A comparison of the relationships of the members of the nigri­ cornis group in Figure 92 with their relationships in the composite graph for temperature effects on pulse rate (Fig. 87) reveals that in general the faster trilling forms have a lower pitched song at a given pulse rate than the slower trilling forms. The result of this correlation is that at a given temperature the songs of the various forms are similar in pitch. For instance at 70° F. all forms except pini have songs pitched between 3300 and 3600 cps, and pini has a song of about 3100 cps.

Field identification of the calling songs of the nigricornis group is difficult at best and is never certain except between the extremes or when different forms are singing simultaneously. Fulton

(1915), Norse (1920), and Cantrall (1943) pointed out that the song of KILOCYCLES PER SECOND .0 4 .5 4 5, .5 5 2.0 .5 3 .5 2 .0 3 5.0 20 30 EAINHP F US RT AD PITCH AND RATE PULSE OF RELATIONSHIP 40 OPRSN F SPECIES OF COMPARISON 50 AS L, gur 92 re u ig F XLV, HATS USS E SECOND PER PULSES 60

70 80 \A 80

100 174 175 pini is recognizable by its low pitch. A few authors have reported that in some instances they were able to separtae the song of quadripunctatus from that of nigricornis. Faxon (1901) reported that quadripunctatus had a song "clearer in tone" than nigricornis; and Horse (1920) thought that the song of quadripunctatus vms a little lower in pitch and often softer than that of nigricornis. Alexander (1956) found the trill of quadripunctatus to be slower and lower-pitched than the trills of nigricornis and argentinus. He also observed that the song of "quadri- punctatus" (actually celerlnictus) in Virginia and North Carolina was more like that of nigricornis and argentinus than of quadripunctatus in Ohio. Allard (1911a) had previously noted that "the stridulations of 0. quadripunctatus in New England have always seemed louder and lower-toned to the writer than the weaker and shriller trilling of the same species [celerinictus?! in Northern Georgia."

M orphology

There are a number of morphological characters which can be used in distinguishing a large proportion of the members of the nigricornis group. The markings on the antennae and legs and the body coloration are the most convenient morphological characters, but it is hard to control the personal element in their interpretation, and they are occasionally deceptive, particularly in teneral specimens. Tne length of the file and the number of file teeth are useful characters in some instances, and one species can be reliably identified on the basis of the structure of the egg. The length of the tegmina, pro- notum, hind femora, and ovipositor are of some value. Only a few 176 aspects of the morphology of the members of the nigricornis group have been investigated, and the number of specimens studied has been small. Fore detailed work should give a better evaluation of the characters in use and should reveal additional useful characters.

The black markings on the ventral surface of the first two segments of the antennae are particularly useful characters in identifying preserved specimens of the nigricornis group. Fulton

(1926b) divided specimens of the nigricornis group into thirteen classes on the basis of antexmal markings; however, his classification does not reveal differences, which exist, between some of the forms recognized in this study.

By classifying the first and second segments separately and by modifying and supplementing Fulton's classification, I was able to eliminate some of the difficulties in classifying specimens on the basis of their antennal markings. Figure 93 illustrates the 16 cate­ gories set up for the markings on the first segment and the 13 cate­ gories set up for the markings on the second segment. The markings of 0. pini w ill not always fit into this classification, since frequently the outer elements of the pattern are missing. Otherwise, specimens of pini fall into the C or D series. Since pini is readily identifiable by other characters, it was not felt necessary to compli­ cate the classification by its inclusion. The following synopsis describes each category in such a manner that much of the subjective element in classifying the markings can be eliminated. HATE XL7I, Figure 93 ANTENNAL MARKINGS OF NIGRICORNIS GROUP

FIRST SEGMENT

SECOND SEGMENT 178

Synopsis of Classification of Antennal Markings (See Figure 95)

F irst Segment C lass Yentral surface of first segment with, extensive infuscation

Almost entirely black; marks obscure LA

Marks distinct but confluent 2A

Space between marks al/3 average width of longitudinal mark 3A

Space between marks >l/3 average width of longitudinal mark 4A

Yentral surface of first segnent with infuscation confined to the edges of the black markings

Outside mark transverse, lateral portion of mark narrowed but no less dark than mesal portion

Average width of longitudinal mark £ of length

Longitudinal and transverse marks confluent IB

Longitudinal and transverse marks connected by infuscation 2B

Space between marks =l/3 average width of longitudinal mark 3B

Space between marks >l/3 4.2/5 average width of longitudinal mark 4B

Space between marks >2/3 4l l/3 average width of longitudinal mark 5B

Average width of longitudinal mark of length

Space between marks >1/3 =2/3 average width of longitudinal mark 4G

Space between marks >2/3 i l 1/3 average width of longitudinal mark 5C

Space between marks >1 l/3 - 2 times average width of longitudinal mark 6C

Space between marks >2 43 times average width of longitudinal mark 7C 179

Outside mark a dot, sometiraes with a lateral transverse tail of a lighter color

Space between marks >2/3 =1 1/3 average width of longitudinal mark 5D

Space between marks > 1 1/3 £ 2 times average width of longitudinal mark 6D

Space between marks >2 £3 times average width of longitudinal mark 7D

Second Segment

Ventral surface of second segment with extensive infuscation

Almost entirely black; marks obscure 1A

harks distinct but confluent 2A

Space between marks $1/3 width of inside mark 3A

Space between marks >l/3 width of inside mark 4A

Ventral surface of second segment with infuscation confined to the edges of the black marks

Outside mark neither less dark nor l/3 shorter than inside mark

Marks confluent IB

Marks just touching 2B

Distinct gap between marks; $1/3 width of inside mark 3B

Least distance between marks >l/3 =2/3 width of inside mark 4B

Least distance bettveen marks >2/3 £l l/3 width of inside mark 5B

Least distance between marks >1 l/3 £2 times width of inside mark 6B

Outside mark either less dark or at least 1/3 shorter than inside mark

Least distance between marks >1/3 $2/3 width of inside mark 4D 180

Least distance between marks >2/3 £,1 1/3 width of inside mark 5D

Least distance between marks >1 1/3 = 2 times width of inside mark 6D

Only a small portion of the m aterial examined was classified on the basis of these categories; however, the results listed in Table

XII are representative of material collected in the eastern United

S ta te s .

There are taxonomically significant markings on the tibiae and the tips of the hind femora in addition to those on the antennae.

The tibial markings are one or two transverse dark bands on the proximal portions of one or more of the three pairs of tibiae (Fig.

95). In nigricornis these mearkings are usually obscured by the other

dark pigment of the tibiae, but in teneral specimens the markings are

2-2-2 (two bands each on the fore, middle, and hind tibiae). In the willow form of nigricornis, the tib ial markings are not obscured and are usually 2-2-2. In 2 specimens out of 22, the markings on the fore

tibiae were missing (0-2-2). In celerinictus the usual condition is

2-2-2, but occasionally the fore tibiae w ill have one band or none

(1-2-2; 0-2-2), and rarely the middle and/or hind tibiae w ill have

but one band (e_.£., 0-2-1; 1-1-1; O -l-l). 0. argentinus is highly

variable in its tibial markings (0-0-0 to 2-2-2), but the most

frequent condition is 0-0-0. _0. quadripunctatus usually has no

tibial markings, but 2 specimens of 104 were 0-2-0. 0. pini usually

has no markings, but occasionally the posterior tibiae have single

bands (0-0-1). 181

Table XII. Antennal markings of the nigricornis group, pini excluded. Sources of material are as follows: willow form of nigri­ cornis, Ohio (31); nigricornis, song form undetermined, Ohio (51), Indiana (11), Kentucky (4), Maine (1); celerinictus, South Carolina (26), Georgia (21), Kentucky (15), Tennessee (ll), Alabama (10), Florida (8), North Carolina (3); argentinus, Kentucky (54), Ohio (34), Indiana (15), Tennessee (12); quadripunctatus, Ohio (52), Indiana (34), Florida (8), Georgia (4), Virginia (3), South Carolina (2), Kentucky (l).

Willow n i g r i ­ c e l e r i ­ arg en ­ q u a d ri­ C ategory Form o f c o rn is n ic tu s tin u s p u n ctatu s nigricornis 31 in d iv . 67 in d iv . 94 in d iv . 115 in d iv . 104 in d iv .

F irst Segment 1A 4 2k 27 3A 1 27 4A 9 IB 57 2B 47 3B 1 8 4B 10 3 5B 15 1 4C 4 5C 3 60 6C 27 7C 2 5D 1 43 6D 58 7D 3

Second Segment LA. 2 2A 19 3A 30 4A 1 11 IB 82 2B 24 3B 1 1 9 4B 16 4 16 5B 13 55 1 6B 22 4D 2 5D 91 6D 1 10 182 PLATS XL7XI

NUMBER OF TEETH IN RIGHT FILE Q- NIGRICORNIS. FRANKLIN COv OHIO

| SONG NOT RECORDED

HU t y p i c a l f o r m

SLOW-TRILLING FORM

40 42 4 4 49 NUMBER OF TEETH P ig . 94

EGGS OF e g g s o f O. QUADRIPUNCTATUS Q- celerinictus FEMORAL-TIBIAL JOINT OF LEFT DICKSON CO., HIND LEG OF HOLOTYPE OF CHESTERFIELD CO, TENNESSEE SOUTH CAROLINA ' £ CELERINICTUS

I MU

Fig. 95 VJ Fig. 97 183

There may he five black marks at the apex of the hind femur — a single transverse dorsal mark sometimes broken at the middle, a pair of lateral marks, and a pair of ventral marks (Fig. 95). The femora of nigricornis are always fully marked, but in the willow form of nigricornis and in celerinictus the lateral and ventral marks are sometimes faint. In most specimens of argentinus the femora are unmarked, but sometimes the lateral and ventral marks are evident.

In quadripunctatus the femora are never fully marked, but in 9 specimens of 104 there were weak lateral marks, and in 3 of these specimens there were ventral marks also. In pini the femora may be unmarked, partially marked, or completely marked, but the pigment is dark brown instead of black.

The pigmentation of the pronotum and abdomen has taxonomic significance. In nigricornis the pronotum usually has one mesal and two lateral dark stripes. Sometimes the lateral stripes are lacking and sometimes the mesal stripe is lacking, but only rarely is there none. The willow form visually has a faint mesal stripe and occasional­ ly has lateral stripes also. None of the other species have blackish pronotal stripes, although pini has corresponding areas of darker brown. The abdomen of nigricornis is nearly always black on the venter and dark on the lateral areas of the tergum. In the willow form of nigricornis the venter is sometimes black, but usually it is merely darkened. The tergum is correspondingly lighter and sometimes is not darkened at all. In celerlnictus and argentinus the abdomen is either without dark pigment or else the venter and tergum are 184 partially dardened. In quadripunctatus there is never any darkening, and in pini the color pattern is like nigricornis but with brown pigment instead of black,

A final color character is the color of the tibiae, tarsi, and distal portions of the antennae. In nigricornis these are black or dark, in the willow form they are black to light, in pini they are brown, and in the other species they are light.

The members of the nigricornis group vary considerably in size.

Table XIII lists the results of measuring the length of the pronotum, tegmen, hind femur, and ovipositor. Qecanthus pini and nigricornis are \isually larger than the other forms, but there is some overlap.

The ovipositor of the willow form of nigricornis is longer in propor­ tion to the other dimensions than in any other form.

The structure of the file offers a reliable means of distin­ guishing some of the males of the nigricornis group. If the right tegmen is removed and mounted ventral surface up on a microscope slide in a suitable medium (e_.t£. Hoyer’s), the file teeth are easily counted with a compound microscope. The length of the file is deter­ mined by measuring in a straight line, along the principal axis, from the first tooth to the last. Figure 98 shows the length of the file and the number of teeth in 326 specimens of the various forms of the nigricornis group, and Table XIV summarizes the data. _0. pini and typical nigricornis have the file teeth more widely spaced than the other forms. In the other forms the average tooth interval is about the same, so the length of the file and the number of teeth are nearly in direct proportion. 0. quadripunctatus can usually be 185

Table XEII. Measurements of members of the nlgricornis group. Sources of material are as follows: nigricornis, Ohio (34), Indiana (11), Kentucky (4), Maine (l); willow form of nigricornis, Ohio (31); celerinictus, South Carolina (15), Georgia (13), Florida (7), Alabama (6), Tennessee (5), Kentucky (4); argentinus, Kentucky (15), Ohio (15), Indiana (13), Tennessee (7); quadripunctatus, Ohio (20), Indiana (17), Florida (7), Georgia (3), Kentucky (1), South Carolina (1), Virginia (1); pini, North Carolina (4), Ohio (4), Virginia (1). All measurements in m illimeters.

Length of Length of Length of Length of Pronotum Tegmen Hind Femur Ovipositor Mean Range Mean Bange Mean Range Mean Range nigricornis 25 males 2.60 2 .3 12.11 11.1 .8.67 7.7 2.9 13.5 9 .5 25 females 2.63 2 .3 12.18 10.6 8.77 8.0 5.87 5.2 3.0 13.5 10.3 6.3 willow form of nigricornis 21 males 2.35 2.1 10.51 9.2 7.78 7.0 2.9 11.6 9.0 10 females 2.50 2.2 11.65 10.7 8.22 7.8 6.36 5.8 2.9 13.2 8 .8 6.6 celerinictus 25 males 2.24 1.8 10.83 9.6 7.70 6.9 2.6 12.7 8 .4 25 females 2.25 1.9 10.42 8 .5 7.85 6.5 5.01 4 .2 2 .4 12.0 8.5 5 .4 a rg e n tin u s 25 males 1.98 1.8 10.96 10.0 7.08 6.4 2.2 12.6 7 .6 25 females 2.06 1.9 10.58 9.2 7.37 6.7 4.62 4 .3 2 .3 11.8 8.2 4.9 quadripunctatus 25 males 2.34 2.2 11.96 11.1 8.22 7 .4 2 .5 12.8 8.8 25 fem ales 2.34 2.1 11.21 9 .5 8.07 7 .3 4.85 4 .5 2 .5 12.8 9.7 5 .4 p in i 7 m ales 2.82 2.6 13.39 12.3 9.53 8 .4 2.9 14.2 10.2 2 fem ales 2.70 2.7 12.75 12.2 9.55 9 .5 6.35 6.3 13.3 9.6 6 .4 40 NUMBER OF TEETH 45 35 50f~ r 5 5 60 60 r U E O TEH N LNT O RGT FILE RIGHT OF LENGTH AND TEETH OF BER NUM '

. 10 . 14 1.6 1.4 1.2 1.0 0.8 1" a afnB o TYPICAL o a asflnBB B 1 - • ELUl • ■ a a a IB k » t t EGH F IE N MILLIMETERS IN FILE OF LENGTH a A AE LII Fgr 98 Figure XL7III, HATE a ahSU.1L a

D " WILLOW ■ FORM NIGRICORNIS OF " D° “ iAmB t r n o | a a | o n r t iAmB A * 8 0 0 8 { * * 8 &> 8 C O NIGRICORNIS GROUP t E T FORMS TO KEY o § mtA J jy^»2noo o o

t a| B aa| atU a no A o o 000 o a Sin aa a O O O o O O 2 CD A n NIGRICORNIS n n oo 00 oi • oofi 080 o Stxoc&o Stxoc&o §o *°o8 • 8 8*o°|o §8o O □D o § dBo 0

oo o B • • • B o oo Oj • OOj8 AAA A A A A A A A A A A A A OAO ftfl d d ftfl O O A A A a A • && && & A A A A A A • a a ARGENTIo NUS &

QUADRIPUNCTATUS A A A A A • • CELERINICTUS SLOW-TRILLING o • A A A A A A A

A 186 187 Table XTV. Vile characters of the nigricornis group. Sources of material are as follows: willow form of nigricornis, Ohio (20); typi- cal nigricornis, Ohio (13), Kentucky (l); slow-trilling nigricornis, Ohio (8); nigricornis, song form unknown, Ohio (49), Indiana (8), Ken- tucky (6); celerinictus, Arkansas (16), Tennessee (12), South Carolina (10), Georgia (9), Florida (5), Kentucky (5), Louisiana (5), Alabama (3), North Carolina (3), Virginia (3), Maryland (2), Missouri (l); argentinus, Ohio (36), Tennessee (26), Arkansas (12), Indiana (8), Oklahoma (6), Missouri (4), Kentucky (1); quadripunctatus, Ohio (12), Louisiana (8), Florida (6), Indiana (5), Mississippi (4), Georgia (3), Virginia (3), South Carolina (2), Kentucky (l); pini, North Carolina (5), Ohio (5). Tooth interval was calculated by dividing the length of the file by the number of teeth.

Number Length of File Tooth Inter­ Tooth Count o f (millimeters) val (microns) In d iv . Mean Range Mean Range Mean Range 0. nigricornis willow form 20 36.4 32-43 0.998 0.87-1.16 27 .'4 25-29

typical 14 39.S 36-45 1.195 1 .1 0 -1 .3 5 30.5 28-33

slow -trilling 8 46.2 42-51 1.304 1.14-1.39 28.2 25-30

song form unknown 63 4 0 .4 35-50 1.170 0 .9 3 -1 .3 9 29.0 25-32

0. celerinictus 74 41.2 35-49 1.086 0.93-1.30 26.4 23-31

0. argentinus 93 47.5 42-53 1.264 1.09-1.45 26.6 23-30

0. quadripunctatus 44 56.2 51-62 1.479 1.31-1.67 26 .3 24-29

0. pini 10 49.4 46-53 1.608 1.50-1.74 32.6 31-35

separated from most of the other forms either by the length of the file or the number of teeth; however, the tooth count is the more dependable character. The remaining forms have less pronounced differences in their file structure but can sometimes be distinguished on this basis. Slow-trilling and typical nigricornis have not been reliably separated morphologically, but the file structure can be used in many instances. 188

A comparison of the file structure data in Table XTV with the pitch and pulse rate relationships (Fig. 92) reveals several correlations. The two forms which have the most widely spaced teeth

(pini and typical nigricornis) have the most rapid decrease in the rate of increase of pitch with increases in pulse rate. These two forms also have lower pitched songs at a given pulse rate than forms with similar numbers of teeth but with closer spacing. Of the forms with lesser tooth intervals (willow form of nigricornis, slow-trilling nigricornis, celerinictus, argentinus, and quadripunctatus), the songs of the ones with the longer files and more teeth are lower pitched at a given pulse rate (Fig. 92) and have slower pulse rates at a given temperature (Fig. 87).

Fulton (1926b) reported that in Hew York, Ohio, Iowa, and

Florida the eggs of quadripunctatus were distinct from the eggs of the other members of the nigricornis group. The cephalic end of a tree cricket egg is a white cap-like structure with regular rows of minute projections. In quadripunctatus this structure is more conical than in other members of the nigricornis group (Figs. 96 and 97). By dissecting gravid females preserved in alcohol, I obtained quadri­ punctatus eggs from Ohio, Indiana, Kentucky, Tennessee, Louisiana, and eastern Texas and eggs of other forms from these and other states.

The eggs of quadripunctatus were always distinct, although there was some variation in the acuteness of the cap-like structure. 189

Hybridization and Crossing Experiments

Interspecific copulation under field conditions was never observed, nor has it been reported in the literature. Only one individual was encountered that appeared to be hybrid in its morphology. A male collected in a weedy field in Franklin County, Ohio, 20 September 1956, had the coloration of a teneral specimen of nigricornis but did not darken with age. A count revealed 59 file teeth, a number encountered only in quadripunctatus. Ho song was produced.

In the summer of 1955, laboratory crosses were made between females of nigricornis, argentinus, and quadripunctatus and males of each of the three species. The females laid eggs in sections of goldenrod and ragweed stems, which were then placed in individual screen cages outdoors. The eggs from neither the crosses nor the controls hatched; and upon examination in June 1956, they were found to be desiccated. The critical drying probably occurred in the warm weather of late summer and early fall. In the field the stems containing eggs are still alive during this period, and the eggs are kept moist by the plant juices.

In October 1955, 105 wild carrot stems showing tree cricket oviposition scars were collected in the field and placed under 20 different temperature and 2 moisture regimes to determine how eggs might be hatched in the laboratory. The temperature treatments were constant temperatures and weekly alternating temperatures of 0°, 40°, and 80° IT. for various periods of time. After treatment the eggs were held at 80° F. Three stems of the five from each temperature 190 treatment were kept separate and submerged in water for ten minutes each. week. As a control, five stems were placed in a wire cage o u td o o rs.

In only 3 of the 41 treatments did any of the eggs hatch. During the week of 5 to IS April, two nymphs hatched from stems dipped in water each week and kept at 80° S’, continuously; during the same week, one nymph hatched from stems dipped in water each week and alternated between 0° and 80° S’, for four weeks before being held at 80° S’. ; and on 11 June, five nymphs hatched from the stems kept outdoors. The unhatched eggs were examined and found to be desiccated. All were quadripunctatus, including the unhatched ones in the stems from which the nymphs had come.

These tests demonstrated that if tree cricket eggs are to be hatched in the laboratory, care must be taken to maintain the water balance and extended or extreme cold periods are not always essential.

In the simmer of 1956 I ran a series of tests in which a male and a female of each of two forms of the nigricornis group (4 indivi­ duals) were placed in a 6-inch diameter battery jar with a corrugated

cardboard roost for several periods of two hours and observed. The

sexes were kept separate except during observation, and virgin females were used to eliminate effects of conditioning during previous matings.

The following pairs of forms were tested: nigricornis and the willow form of nigricornis (5 replicates), nigricornis and celeri­

nictus (1 replicate), nigricornis and argentinus (6), nigricornis and

quadripunctatus (5), willow form of nigricornis and pini (2), 191

celerinictus and argentinus (5), celerinictus and quadripunctatus

(3), argentinus and quadripunctatus (7), quadripunctatus and p in i ( l ) .

Copulation was completed at least once by most of the females

during each two-hour observation period. Matings between like forms were much more frequent than cross matings except in the case of

nigricornis and the willow form of nigricornis. Here the numbers of

each type of mating were about equal. In every pair of species

except those involving pini, cross copulation was observed at least

once. More exact data cannot be presented because the records of the

experiments have been lost.

That cross copulation can occur between forms does not indicate

whether hybrid offspring can be produced. In field crickets, for

example, cross copulation occurs in the laboratory, but in most cases

no offspring are produced (Alexander 1956).

Summary o f D is tin g u ish in g C h a ra c te rs

Table XV is a summary of the more useful characters which have

been discussed as means of distinguishing the members of the

nigricornis group. 192 Table XV. Usefulness of various characters in identifying the members of the nigricornis group. The symbol _0 indicates that the character is of little use; X indicates that is is useful in some situations; XX indicates that it is useful in most situations; and XXX indicates that the character is 99$ or more effective.

P a irs o f Characters Evaluated Forms to be D is trib u . Song Marks & Color S ize F ile Eggs Distinguished Geog. Ecol. PI s CBS ant. legs body Typical & slow -tril­ ling nigricornis 0 0 XXX XXX 0 0 0 0 XX 0 Typical & willow nigricornis 0 XX 0 X XX X X X X 0 Typical nigricornis & celerinictus XXX 0 0 X XXX XX XXX X 0 0 Typical nigricornis & argentinus X X XXX XX XXX XX XXX XX 0 Typical nigricornis & quadripunctatus X 0 XXX XXX XXX XXX XXX X XXX XXX Typical nigricornis & p in i 0 XXX XXX X XXX XXX XXX 0 XXX 0 S l.-tr. nigricornis & willow form 0 XX XXX XXX XX X X X XX 0 Sl.-tr. nigricornis & celerinictus XXX 0 XX XX XXX XX XXX X X 0 Sl.-tr. nigricornis & argentinus X X 0 0 XXX XX XXX X 0 0 S l.-tr. nigricornis & quadripunctatus X 0 XXX 0 XXX :cxx XXX X XX XXX S l.-tr. nigricornis & p in i 0 XXX XXX 0 XXX XXX AuuX. 0 XX 0 Willow nigricornis & celerinictus XXX XX X 0 XX XX X 0 0 Willow nigricornis & argentinus 0 XX XXX 0 XXX -vxw XX X XX 0 Willow nigricornis & quadripunctatus X XX XXX XXX XXX XXX XX X XXX XXX Willow nigricornis & n in i 0 XXX XXX X XXX XXX XXX X XXX 0 0. celerinictus & a rg e n tin u s X 0 XXX 0 XXX X 0 0 X 0 0. celerinictus & quadripunctatus X 0 XXX XX XXX XXX 0 0 .aXX XXX 0. celerinictus & p in i X XXX XXX 0 X XXX XXX X x

Oecanthus nigricornis F. Walker

The Black-Homed Tree Cricket

Typical and Slow-Trilling, Figures 2, 14, 17, 25, 35, 49, 63, 80, 86, 87, 92, 93, 94, 98, 112

Willow Form, Figures 26, 36, 81, 87, 92, 93, 98

Oecanthus nigricornis F. Walker 1869, p. 93 (type locality, Illinois;

type, a female in the British Museum, London, England). Thomas

1870, p. 206 (quotation of original description); Saussure 1874,

p. 461 (synonymy); Beutenmuller 1894b, p. 250 (biology, song,

synonymy); Beutenmuller 1894c, p. 270 (biology, song); Scudder

1900, p. 90 (synonymy); Fbtxon 1901, p. 183 (song); Scudder 1901,

p. 212 (lists ten additional references); Forbes 1905, p. 218

(food); Felt 1906, p. 699 (habitat); Houghton 1909a, p. 274

(courtship); Houghton 1909b, p. 114 (rearing); Parrott 1909, p.

126 (oviposition); Allard 1911a, p. 28 (habitat, song); Allard

1911b, p. 155 (song); Parrott 1911, p. 216 (oviposition); Parrott

and Fulton 1913, p. 177 (biology, oviposition); Parrott and

Fulton 1914, p. 453 (detailed account of biology, morphology,

and song); Fulton 1915, p. 35 (detailed account of biology,

morphology, and song); G-loyer and Fulton 1916, p. 15 (vector of

fungus); Rehn and Hebard 1916, p. 297 (habitat, morphology);

Caesar 1919, p. 62 (vector of fungus); Morse 1919, p. 20 (habi­

tat); Blatchley 1920, p. 719 (biology, song); Morse 1920, p. 409

(biology, morphology, song); Severin 1920a, p. 28 (biology);

Severin 1920b, p. 1 (biology); Hubbell 1922b, p. 12 (habitat);

Johnson 1922, p. 759 (spermatogenesis); Snodgrass 1925, p. 440 194 (song); Allard 1929a, p. 571 (song); Johnson 1931, p. 117

(biology, cytology); Urquhart 1941b, p. 30 (habitat); Procter

1946, p. 41 (habitat); Pierce 1948, p, 136 (song analysis);

Alexander 1956, p. 178 (biology, discussion and analysis of

song); Alexander 1957, p. Ill (effect of song on female).

Oecanthus nigricornis nigricornis P. Walker. Pulton 1926b, p. 60

(biology, key, morphology); Fulton 1932, p. 63 (habitat, song);

Hebard 1934, p. 253 (habitat, morphology); Hebard 1938, p. 101

(habitat); Cantrall 1943, p. 52 (biology, song).

Oecanthus fasciatus, Fitch (not De Geer 1773) 1856, p. 414. Scudder

1868, p. 55 (synonymy); Davis 1889, p. 80 (song); IIcKeill 1889

(in part), p. 102 (song, synonymy); IvIcNeill 1891 (in part), p. 6

(song); Blatchley 1892, p. 143 (biology); Hart 1892 (in part),

p. 33 (morphology); Scudder 1893, p. 66 (song); Lugger 1897 (in

part), p. 361 (song); Scudder 1900, p. 90 (synonymy); Scudder

1901, p. 211 (lists 13 additional references); Blatchley 1903,

p. 450 (biology, song, synonymy); S. Walker 1904, p. 254 (habi­

ta t, song, synonymy); Hancock 1905, p. 1 (biology, morphology,

song); ICirby 1906, p. 75 (synonymy); Jensen 1909a, p. 25 (biolo­

gy); Hancock 1911, p. 379 (biology, song); Jensen 1911, p. 65

(structure of spermatophore); Fox 1915, p. 301 (habitat).

Oecanthus niveus, Scudder (not De Geer 1773) 1862 (in part), p. 432.

Walsh 1867 (in part), p. 54 (oviposition); Riley 1869 (in part),

p. 138 (morphology); Riley 1873 (in part), p. 120 (food, song);

Scudder 1874 (in part), p. 365 ("day song," oviposition); Riley 195

1881 (in part), p. 60 (synonymy); Scudder 1894 (in part), p. 4

(oviposition).

David Ragge, of the B ritish Iluseum, wrote (personal communication)

that the holotype of nigricornis "is in reasonable condition and is labelled as type." The original description leaves no doubt as to

the identity of the type specimen. IJhether the type is of the typical or slow-trilling form (if the two forms exist among females) is a moot question, but it does not seem to be the willow form.

The name fasciatus was applied to this species as a result of

Pitch’s misidentification of De Geer's Gryllus fasciatus (now placed

in Hemobius) .

The name niveus was used for this species by some of the early w riters because they thought there was but a single species of T-Torth

American Oecanthus. The use of niveus was continued by later writers

who mistakenly ascribed the characteristic oviposition scars of

n ig ric o rn is to the snowy tre e c ric k e t, then known as n iv eu s.

Oecanthus celerinictus n. sp.

The Fast-Calling Tree Cricket

Figures 27, 37, 50, 82, 87, 88, 92, 93, 95, 97, 98

Oecanthus fasciatus, Ashmead (not De Geer 1773) 1894, p. 25 (biology).

Oecanthus quadripunctatus, Allard (not Beutenmuller 1894b) 1910a (in

part), p. 36 (biology, song). Allard 1911a (in part), p. 28

(habitat, song); Rehn and Hebard 1916 (in part), p. 296 (habitat); 196

Fox 1917 (in. part), p. 234 (habitat); Alexander 1956 (in part),

p. 180 (discussion and analysis of song).

Oecanthus nigricornis quadripunctatus, Fulton (not Beutenmuller 1894b)

1932 (in part), p. 63 (habitat, song); Fulton 1951 (in part),

p. 93 (biology, song).

Types: Holotype a male taken by the author from common ragweed

(Ambrosia artem isiifolia L.) near Lugoff, Kershaw County, South Caro­ lina, 14 September 1955. Allotype a female with the same data. The types are in the author's collection but will be deposited in the

U. S. National Museum, Washington, D. C.

Diagnosis: This species is easily distinguished from other members of the nigricornis group by its characteristic antennal markings, its lack of dark markings on the pronotum, the black markings on the tibiae and hind femora, and its rapid song. The

specific name, celerinictus, refers to the fast calling song.

Description of holotypic male: Similar to quadripunctatus in

form, size, and background color. A11TENKAE: Pale throughout. Basal

segment with longitudinal black mark six times long as wide; transverse black mark widest medially, separated from longitudinal mark by width

of latter (class 5C, Fig. 93). Second segment with two longitudinal black marks separated by width of inside mark; outside mark slightly wider and one-fifth shorter than inside mark (class 5B, Fig. 93).

HEAD: No dark markings; sensory area on terminal segment of maxillary

palp three-fifths of total length of segment. THORAX: No dark

markings; length of pronotum, 2.3 mm; width posteriorly, 2.1 mm. 197

ABDOMEN: Venter slightly darkened. LEGS: Fore tibia with two

narrow transverse black bands on outside surface, just proximal of

tympanic openings, middle and hind tibiae with similar but heavier

bands; hind tibia with proximal band darker than distal band (Fig. 95).

Hind femora 7.8 ram long, marked with black at apex (Fig. 95). TEGKEHA:

Length of dorsal field of right tegmen, 10.4 mm; width, 4.2 mm.

Description of allotypic female: Dark markings like those of

holotype. Length of pronotum, 2.4 mm; width, 2.1 mm; length of hind

femur, 8.3 ram; length of ovipositor from bottom of mesal notch in

subgenital plate, 5.0 mm; length of right tegmen, 11 mm.

This species has been called quadripunctatus by previous authors,

although Allard (1911a) and Alexander (1956) detected song differences

between the " quadripunctatus" of the Southeast and the quadripunctatus

of the North. Most of the references to quadripunctatus in the

Southeast probably pertain to celerinictus, since it is more frequently

collected than the former species.

Oecanthus argentinus Saussure

The Western Tree Cricket

Figures 7, 8, 9, 28, 38, 55, 57, 83, 86, 87, 89, 92, 93, 98

Oecanthus argentinus Saussure 1874, p. 460 (type locality, uncertain;

type, a female in the Museum d’Histoire ITaturelle, Geneva,

Switzerland). Saussure 1897, p. 253 (synonymy); Scudder 1900,

p. 90 (synonymy); Kirby 1906, p. 74 (synonymy); Alexander 1956,

p. 175 (biology, discussion and analysis of song). 198

Oecanthus nigricornis argentinus Saussure, Fulton 1926a, p. 13

(biology, song); Fulton 1926b, p. 60 (biology, key, morphology);

Hebard 1929, p. 309 (habitat); Hebard 1934, p. 253 (morphology,

habitat); Beach 1938, p. 303 (biology, cytology); Y/illiams 1945,

p. 4 (song).

Oecanthus nigricornis, Caudell (not F. Ualker 1869) 1902, p. 91.

Oecanthus rehnii Baker 1905, p. 82 (type locality, Stanford University,

Santa Clara County, California; type, a male in the Pomona

College Collection, Claremont, California).

Oecanthus pini, Forbes (not Beutenmuller 1894a) 1905, p. 218 (food).

Oecanthus quadripunctatus. Hubbell (not Beutenmuller 1894b) 1922a (in

part), p. 55.

The identity of Saussure's argentinus is by no means certain.

The principal disparity in the present placement of the species is that Saussure (1374) stated that the type was collected at La Plata,

Argentina. If this locality is correct, it is extremely unlikely that Saussure*s argentinus is the same as the western tree cricket.

However, Saussure, in 1397, reported argentinus from Texas and

Mexico and stated that "this species was described upon two [only one type is mentioned in the original description and only one is pre­ served at GenevaJ specimens labelled as from Argentina; but the locality requires confirmation, to judge from the fact that numerous specimens have been found in the central parts of America."

I have not seen the type, but Charles Ferriere of the Museum

d'Histoire ITaturelle, Geneva, wrote (personal communication) that 199

"the type is not in good condition; both, antennae, except first segment, and the legs, except left hind leg, are broken.” None of the color characters, markings, or structural features described by

Dr. Ferrilre w ill separate the type of argentinus from the western tree cricket. The markings on the basal segment of the antenna are class 2B (Fig. 93). The dimensions given by Saussure in his original description and by Dr. Ferriere indicate that the type is slightly larger than the specimens of western tree cricket which I have measured. Saussure gave the length of the pronotum of the type as 2.6 mm; however, Dr. Ferriere measured it as 2.4 mm. The maximum pronotal length I have found in the western tree cricket is 2.3 mm.

Until more information is available, argentinus should continue as the name of the western tree cricket, and Saussure's original type locality designation should be classed as questionable.

In the event that the name argentinus should prove to apply to another species, the western tree cricket would become Oecanthus rehnii Baker. The type of rehnii is on loan to the U. S. National

Museum, and when I examined it in 1955, I found no means of distin­ guishing it from what is here called argentinus. Only the basal segment of each antenna is present, and the markings are class 2B. 200

Oecanthus quadripunctatus Beutenmuller

The Four-Spotted Tree C ricket

Figures 4, 5, 18, 29, 39, 51, 55, 56, 85, 86, 87, 90, 92, 93, 96, 98, 106

Oecanthus quadripunctatus Beutenmuller 1894b, p. 250 (type locality,

West Woodstock, Windham County, Connecticut; types, a male in

the Harvard Museum of Comparative Zoology, Cambridge, Hassachu-

setts, and a specimen, sex unknown, in the IT. S. national

Museum, Washington, D. C.). Beutenmuller 1894c, p. 271 (synony­

my); Scudder 1900, p. 91 (synonymy); Faxon 1901, p. 183 (song);

Scudder 1901, p. 214 (lists five additional references);

Blatchley 1903, p. 452 (biology, song, synonymy); E. Walker 1904,

p. 255 (synonymy); Forbes 1905, p. 218 (food); Felt 1906, p. 699

(habitat); Kirby 1906, p. 75 (synonymy); Jensen 1909b, p. 25

(oviposition); Allard 1910a (in part), p. 36 (biology, song);

Allard 1911a (in part), p. 28 (habitat, song); Parrott 1911, p.

216 (biology); Allard 1912, p. 461 (song); Parrott and Fulton

1913, p. 177 (oviposition); Parrot and Fulton 1914, p. 456

(detailed account of biology, morphology, and song); Fox 1915,

p. 301 (habitat); Fulton 1915, p. 32 (detailed account of biology,

morphology, and song); Behn and Hebard 1916 (in part), p. 296

(habitat); Fox 1917 (in part), p. 234 (habitat); Morse 1919, p.

30 (habitat); Morse 1920, p. 410 (biology, song); Hubbell 1922b,

p. 12 (habitat); Wakeland 1927, p. 19 (song); Johnson 1931, p.

117 (cytology); Udine and Pinckney 1940, p. 81 (egg parasites); 201

Procter 1946, p. 41 (habitat); Pierce 1948, p. 146 (song analy­

sis); Alexander 1956, p. 180 (biology, discussion and analysis

o f song).

Oecanthus nigricornis var. quadripunctatus Beutenmuller. Houghton

1909a, p. 274 (courtship); Houghton 1909b, p. 114 (synonymy).

Oecanthus nigricornis quadripunctatus Beutenmuller. E. Walker 1910,

p. 356; Blatchley 1920, p. 722 (biology, synonymy); Snodgrass

1925, p. 440 (song); Pulton 1926b, p. 60 (biology, key, morpholo­

gy); Hebard 1928, p. 304 (synonymy); Allard 1929a, p. 581 (song);

Hebard 1929, p. 309 (habitat); Smith 1930a, p. 13 (parasite);

Pulton 1932 (in part), p. 63 (habitat, song); Hebard 1934, p. 253

(habitat); Hebard 1935, p. 78 (habitat); Hebard 1938, p. 101

(habitat); Cantrall 1943, p. 40 (biology, song); Pulton 1951 (in

part), p. 93 (biology, song); Priauf 1953, p. 98 (biology).

Oecanthus niveus, Pitch (not De Geer 1773) 1856 (in part: "variety c"),

p. 413. Scudder 1893 (in part), p. 65 (song).

Oecanthus fasciatus, McNeill (not De Geer 1773) 1889 (in part), p. 102

(song). McNeill 1891 (in part), p. 6 (song); Hart 1892 (in part),

p. 33 (morphology); Lugger 1897 (in part), p. 361 (song).

Oecanthus forbesi Titus 1903, p. 260 (type locality, Urbana, Illinois;

date of collection, September; type, a male in the Illinois

Natural History Survey Museum, Urbana, Illinois). Porbes 1905,

p. 217 (key); Kirby 1906, p. 75 (synonymy: listed as

questionable).

Oecanthus nigricornis, Hebard (not Walker 1869) 1925 (in part), p. 150. 202

Since I have not examined the cotype of quadripunctatus at the

National Museum, I am not prepared to designate a lectotype. The type locality given ahove is that of the eotype in the Harvard Museum and may not he the same as the locality of the other cotype.

I have not seen the type of forhesi, hut Blatchley (1920) and

o th ers place i t as a synonym of quadripunctatus.

Oecanthus pini Beutenmuller

The Pine Tree Cricket

Figures 30, 40, 84, 87, 91, 92, 98

Oecanthus pini Beutenmuller 1894a, p. 56 (type locality, Woodstock,

Windham County, Connecticut; types, two males in the Harvard

Museum of Comparative Zoology, Cambridge, Massachusetts, and one

female and two a d d itio n a l specimens in the U. S. N ational Museum,

Washington, D. C.). Beutenmuller 1894c, p. 271 (habitat, song);

Scudder 1900, p. 91 (synonymy); Scudder 1901, p. 214 (lists two

additional references); Pelt 1906, p. 698 (habitat); Kirby 1906,

p. 75 (synonymy); Pulton 1915, p. 39 (detailed account of biology,

morphology, and song); Rehn and Hebard 1916, p. 298 (habitat,

morphology); Morse 1920, p. 412 (biology, morphology, song);

Pulton 1926b, p. 58 (key, morphology); Allard 1929a, p. 571 (song);

Johnson 1931, p. 117 (cytology); Pulton 1932, p. 63 (habitat,

song); Hebard 1938, p. 101 (habitat); Cantrall 1943, p. 14

(habitat, song); Procter 1946, p. 41 (habitat); Pierce 1948, 203

p. 143 (song analysis); Hilton 1951, p. 93 (biology, song);

Alexander 1956, p. 182 (biology, discussion and analysis of

song).

Oecanthus nigricornis, Hehn and Hebard (not Vlalker 1869) 1910a, p. 649.

Hot having seen two of the cotypes of pini in the U. S. National

Museum, I am not prepared to designate a lectotype. BEHAVIORAL EFFECTS OF THE GALLING- SONG

Introduction

The sounds produced by insects may have varied effects upon other members of the same species and upon other species, Alexander

(1957) has recently summarized the types of behavioral sequences which may be associated with insect sounds.

In this study, only two aspects of the effects of tree cricket sounds were investigated — the response of males of rileyi to the calling songs of other males, resulting in synchronous singing; and the response of females to the calling song of the male, resulting in approach to the male.

Synchronous Singing in the Snowy Tree Cricket

Although synchronism is known in the songs of other insects

(Fulton 1954), the synchronism of the snowy tree cricket has been known longer and discussed more fully than any other example. McNeill

(1889), in what is apparently the first mention of the synchronous singing of rileyi in scientific literature, wrote of "the full chorus of the night-song whose vibrations in exact unison produce that

’rhythmic beat’ as Burroughs has happily phrased i t . . . . ”

Most later writers agreed that snowy tree crickets in a restricted area chirp in unison (e_.,g., Dolbear 1897, Forbes 1905, Fulton 1915), but Edes (1899) and Lutz (1924) thought that the effect was an illusion, and Shull (1907) found ”exact synchronism to be compara­ tively rare.” Allard (1917) was the first to point out that 204 205 synchronous chirping in rileyi is not likely an illusion since certain other intermittent rhythmic trillers do not give a similar illusion when singing together. It was Fulton (1928a) who finally demonstrated experimentally that synchronism in rileyi is a real phenomenon. He removed the front tibiae, containing the tympana, from one of too groups of caged males which had been singing syn­ chronously. After the operation, the controls continued to sing in unison, but the amputees each sang without regard to the others, and the result was chaotic.

Besides demonstrating the genuineness of synchronism, Fulton's experiment showed that auditory stimuli were essential in the phe­ nomenon. Allard (1929a) reported that by an imitation of the notes of rileyi, "I have led chirping individuals to speed up their rate noticeably, in order to synchronize their chirps with my mimicry."

Both Fulton (1925) and Allard (1930a) noted that synchronism in

the field was more perfect at higher temperatures.

While the phenomenon has been adequately' demonstrated and is

self-evident to those who have listened to colonies of rileyi on warm

evenings, there has been no previous work on the means by which

synchronism is effected. In an effort to discover what mechanisms

are responsible for synchronous singing among neighboring males of

rileyi, I ran a series of tests in which recorded sounds were played

to males singing in a controlled temperature room. 206

Methods

The sounds to be played to the singing males were first recorded on magnetic tape, and loops were made by splicing the ends of seven- foot sections of tape to their beginnings. In loops with recordings of regular chirps, care was taken to join the tape at an interval between chirps and in such a manner as to maintain an unbroken chirp rhythm. The uniformity of the chirp intervals was checked by making audiospectrograms. A loop of tape made in this manner, threaded into the llagnecorder, and passed over a reel as shown in Figure 100 could be played continuously for as long as required. Each loop was marked to indicate the direction in which it should be played and was stored on a reel in a labeled envelope.

The speaker used in these te s t s was a Jensen HP-302 "super­ tweeter," which was mounted with a Jensen A-402 crossover network with a.15-ohm resistor taking care of the low-pass output. The intensity of the test sounds was controlled by ear. Preliminary experiments revealed that intensity (above a minimum) was not a c riti­ cal factor in the response of singing crickets to most of the test sounds.

Three males of rileyi were confined in individual 6-inch diameter battery jars (Fig. 99). A crystal microphone projected into each jar, and a fourth microphone picked up the test sound and oral commentary. Usually only one or two of the test crickets sang during a night's recording period. When more than one sang at the same time, perfect synchronism was maintained; however, before any HATE XLIX 207

Figure 99. Apparatus for studying synchronism. A. Speaker. B, C, D. Crystal microphones projecting into jars. E. Electronic mixer. F. Microphone for oral commentary. G. Earphones for auditing sig n a l. H. 610EV tape recorder.

Fig. 100. Apparatus for playing tape loops and for studying the response of females to sound. The cage and speaker are in the foreground; behind thorn is the tape recorder with a loop of tape in p la ce. 208 test sound was played, all but one of the singing crickets were silenced by shaking the jars. Otherwise a modification in the singing behavior of one cricket might be a response to the song of another instead of a response to the test sound. The signals from the microphones were combined in an electronic mixer and recorded with the 61037 tape recorder. The recordings were later audited at normal and reduced speeds, and audiospectrograms were made.

R e su lts

To test the response of a singing cricket to isolated chirps, a loop was played which contained four randomly spaced rileyi chirps recorded at the same temperature as that at which the test cricket was singing. Analysis revealed that the test crickets had two sorts of responses to single chirps (Rig. 101). If the test chirp fell so that it ended during a chirp interval of the test cricket’s song, the cricket would not produce the next chirp as soon as it would other­ wise, i,.e., the cricket lengthened its chirp interval. On the other hand, if the test chirp fell so that it ended during a chirp of the test cricket’s song, the cricket would cut short its chirp (usually from 8 pulses to 5 or 2) and make the following chirp interval shorter than normal.

It can be seen that the first sort of response reduces the chirp rate, the second sort of response increases the chirp rate, and either sort of response would cause an out-of-time cricket to fall in time with another cricket chirping at the same rate. ELATE L L yi A m m »#i' k t t <4 Ml Ml P1PP

U*tM Uu* MMtt * « i f f M l M l m nt #«

6 7 8 10 II SEC Fig. 101. Responses of singing individual of 0. rileyi to random chirps, a . Lengthening of interval. B. Shortening of chirp and following interval.

sign al M* M* H i MM M* n H r r| r # J* s arcricket > 2 3 4 5 <36C. Fig. 102. Response of individual singing at 190 ch/m to recorded song of 166 ch/m.

M.| (li( .k jt tkfck i ^ k n i ir *■* >4 sign al ff^ ir 1 *1 l » 'h M l Hit 4i 'I*1 ’ liK< »V|H tMt* l W iiA l|M ilk crick et

1 1 » 5

It Hi \J*I i1 11 M I • » * i, i <» ,

10 II SEC Fig. 103. Response of individual singing at 187 ch/m to recorded song of 242 ch/m, 210

The next step was to determine whether sim ilar responses occurred when a normal sequence of chirps (rather than isolated chirps) was played to the test crickets. In order to permit differentiation of the test sounds from the sounds of the cricket, artificially produced calling songs at a slightly higher pitch than the normal song were used for the test sounds. The method of production of artificial songs is explained in the next section, and Figures 108 and 109 are audiospectrograms of natural and artificial rileyi songs. Tests with natural songs produced the same results as the artificial songs as far as the ear could detect.

Figure 102 illustrates the response of a test cricket to a lower chirp rate than the one at which it is singing. Since the test chirps end in the intervals of the test cricket’s song, the cricket lengthens its intervals and thereby sings synchronously with the recorded song.

The test cricket does not ordinarily alter its chirp (as opposed to the chirp interval); however, fewer chirps of less than the normal eight pulses are produced.

Figure 103 illustrates the response of a test cricket to a higher chirp rate than the one at which it is singing. Since the test chirps end before the cricket's chirps do, the cricket produces 2- and 5-pulse chirps (instead of the normal 8-pulse chirps) and shortens the chirp intervals. The result is synchronism with the test record­ ing. It can be noted in Figure 103 that the first chirp of the test song occurred during an interval in the cricket’s song, and the cricket lengthened that particular interval, but the subsequent test 211 chirps always ended during the cricket's chirps, and the result was

shorter intervals (and shorter chirps).

A series of tests were made in which crickets singing at various

temperatures were subjected to five different chirp rates. Figure

104 shows the results of the tests in terms of whether or not the

test cricket synchronized with the chirp rate of the recording. It

is evident that the test crickets synchronized with chirp rates both above and below the normal chirp rate at normal chirp rates from 105

to 190 ch/m (64° to 84° F.). It is also evident that in response to

the high and low chirp rates, the crickets quickened their songs to a

greater extent than,they slowed them. One reason that synchronism

did not occur with chirp rates 50 ch/m below normal and lower, is

that the intervals of the test song became so long that after each

recorded chirp the test cricket chirped twice before the next recorded

chirp intervened, i_.e_., the cricket chirped at double the chirp rate

of the test song.

In other tests the effects of sounds unlike the normal chirps

of rileyi were investigated. A loud, continuous note of 4000 cps

caused a decrease in chirp rate — a cricket singing at 144 ch/m

sloxved its song to 120 ch/m when subjected to such a signal. The same

result was obtained with a pulsed signal of 50 p/s and 4000 cps. In

either case the intensity of the signal had to be near the pain level

(for a human being) before the slowing was pronounced. In another

test, a cricket, formerly singing at 165 ch/m, synchronized with an

artificially produced, pulseless chirp of 151 ch/m (see Fig. 110 for

an audiospectrogram of such a chirp). CHIRP RATE OF RECORDING + 100 + + 150. 0 5 -1 -100 0 +5 150 ■50 F i g . 104 104 . g i F 6 200 240 4 2 0 0 2 160 0 2 1 0 8 YCRNS O O IEI IH RECORDINGS WITH O. OFRILEYI SYNCHRONISM E Qo o o Q KEY NIIUL O I NO. O INDIVIDUAL O NIIUL O 3 ° 3 NO. INDIVIDUAL □ LC SMOS INDICATE SYMBOLS BLACK IGA O STP SD O PRODUCE TO USED SETUP DIAGRAM OF RECORDING YCRNS WITH SYNCHRONISM NIIUL O 2 NO. INDIVIDUAL ABORAT MULATOR R O T A L U IM T S Y R TO A R O B LA O 3000 OR 4 0 0 0 CPS 0 0 0 4 OR 3000 R O T A L IL C S O AUDIO O O IC O SONG OF PITCH CHIRP RATE, DURATION, RIIIL ALN SONGS CALLING ARTIFICIAL DETERMINING i. 105 Fig. E C R U O S N INTERVALAND o o DETERMINING EOE RECORDING WAS PLAYED BEFORE BIAS a • ° o ° O CD * 3 0 ______212 1 2 I L E T A H CHIRPS/MINUTE o . * A o o # O UI OSCILLATOR O AUDIO AE RECORDEDTAPE E D O T N E P O DETERMINING 2 0 -1 0 0 CPS 0 0 -1 0 2 PULSE RATE T U P T U O o B o 213

Significance of Synchronous Singing

Lutz (1924) concluded his discussion of the occurrence of synchronism in rileyi with the question, "what would the crickets gain at any rate by chirping in unison?'* This is an important question and one that can be answered, at least in part, with the knowledge available.

It w ill be demonstrated in the next section, that sexually responsive rileyi females are attracted by the chirp rhythm of the song of the male. If a colony of males were singing asynchronously, the chirp rhythm would be obscured to a female at same distance from the individual singers, and she would not respond. However, since the colony chirps in unison, a female outside the colony may react to the colony's chirp rhythm until she is close enough to some particular cricket to be influenced by its song. Synchronism also results in a colony producing a higher maximum sound intensity, just as persons in step make louder noises collectively than persons out of step. The result of this greater intensity is a greater "sphere of influence" for the colony.

Specificity in the Response of Females

to Calling Songs of the kales

Previous workers have demonstrated that the calling songs of male crickets cause the approach of sexually responsive females.

Regen (1913) sho?/ed this to be the case in Gryllus campestris L., and 214

Busnel and Busnel (1954) attracted females of Oecanthus pellucens

Scopoli with songs of the male played through a high fidelity speaker.

Since there may be several species of tree crickets and many other species of singing insects producing sounds simultaneously in a given habitat, the question arises as to whether a female of a given species of tree cricket w ill respond to any insect sound, to any tree cricket sound, or only to the sound of its own species.

The question of specificity in response to sound has not been adequately answered in other groups of singing insects, so any information on specificity in tree crickets is of general interest.

Tree crickets are particularly well suited for such a study because the songs are relatively simple, song differences are restricted to a few characteristics, and several groups of species occur abundantly in close association in the same habitats.

Methods

The tests pertaining to specificity of female tree crickets to the calling songs of the males were made in the laboratory using sounds of two origins — natural and artificial. Both type3 were first recorded on magnetic tape and made into loops like tliose used in the study of synchronism.

A rtificial tree cricket calls were produced by a device (Fig.

105) designed and constructed by Dr. George Potor, Jr., Division of

Biophysics, Ohio State University. The device was basically a pentode

amplifier operating at 4000 (or 3000) cps, screen-grid modulated at

20-100 cps. This was grid biased by a variable DC source to that 215 portion of the plate-charaeteristie curve which allowed passage of bursts of 4000 (or 3000) cps signal. The grid bias could be further switched on and off by a laboratory stimulator to provide chirps of variable frequency and duration. Pulseless chirps could be produced by cutting out the 20-100 cps audio oscillator. Figures 106 to 110 show audiospectrograms of artificial tree cricket calls compared with genuine ones.

The Jensen EP-302 speaker was used, and the intensity of the test signal was standardized for each set of experiments with the YU meter of the Kagnemite 610EY tape recorder. The original standard was that reading produced on the YU meter by a naturally singing tree cricket at one inch from the microphone. Thereafter, sounds to be tested were adjusted to give the same reading with the microphone one inch from the speaker. Later the Kagnemite was calibrated with a General

Radio Company Type 1551-A sound-level meter. At one inch from the outside edge of the speaker the maximum intensity was 93 db. Since the characteristics of the Kagnemite as a sound-level meter vary with the voltage of its batteries, some variation from this intensity must have occurred during the several months of the tests. However, this

Magnemite was seldom used except as a sound-level indicator, so the variation should not have been great. At any rate, for the tests carried out on any one night or on consecutive nights, the intensities were well standardized.

The tree crickets to be tested were confined in a cage (Figs.

100 and 111) with inside dimensions 1* x 1’ x 3». The sides, bottom, and ends were made of 16-mesh plastic screening. The top consisted FREQUENCY IN KILOCYCLES PER SECOND 4 2 4 2 2 j 2 OPRSN F AUA AD RIIIL SONGS ARTIFICIAL AND NATURAL OF COMPARISON 3 5 0 Fig. 110 0 0 2 110 0 Fig. i i 6 CIP/IUE A PULSES/SECOND 'A 0 5 CHIRPS/MINUTE 168 CHIRPS/MINUTE 162 106 ig. F 6 CHIRPS/MINUTE 168 g 107 ig. F g 108 ig. F g. 109 . ig F wWrfli wwWWIrlfwl EATU QARPNTTS F. ’ 0 8 QUADRIPUNCTATUS OECANTHUS ! l l l l l i l l RIIIL OTNOS TRILL CONTINUOUS ARTIFICIAL RIIIL USLS CHIRP PULSELESS ARTIFICIAL EATU RLY 7° F 75° RILEYI OECANTHUS RIIIL USD CHIRP PULSED ARTIFICIAL LII E T A H IE N SECONDS IN TIME 4 0.6 .4 0 l l t t e f l l l ft ft ft ft ftftftftft l l l l l l l l l 4 5 PULSES /SECOND PULSES 5 4 A PULSES/SECOND 'A 4 4 48'A 48'A PULSES/SECOND 0.8 llilltiM t l l l l l l l l l I I H l t D 216 1.0 217 PLATE LIII

ENO DISTANT FROM SPEAKER

STARTING ZONE

+ 2

SPEAKER AT THIS END

F ig. I l l , Diagram of cage showing zones used in recording the positions of the crickets. The finely dashed lines indicate the boundaries of the zones, and the figures in the various zones repre­ sent the scoring system.

+20 — 70* F. - 80* F.

0Cin O Vi +10

Ul

"5. 45 SO 60 65 70 7580 65 PULSES PER SECOND Fig. 112. Response of £. nigrlcorais females at 70° and 80° F, to artificial, 4000 cps continuous trills of different pulse rates. The encircled points are those for which chi square an alysis showed a response significant at the 1# level. The arrows on the lower scale indicate the average pulse rates of the songs of males at the two temperatures* 218

of three closely fitting panes of glass with the middle pane removable.

The cage was constructed so that the supporting members were outside

the glass and screening, leaving the inside surfaces free from

obstructions. The acoustical properties of the cage were far from

ideal, since standing waves were set up which caused an uneven

intensity gradient from the speaker. The results of the tests indi­

cate that the crickets were nevertheless able to orient to the sound

source. In future work a cage of improved design should be used.

The crickets in these experiments were virgin females, obtained

by rearing late-instar female nymphs to maturity in isolation.

Tirgin females were used because they had no previous experience in

responding to sounds and because they were more likely to be sexually

receptive. All crickets used were from Franklin County, Ohio.

For each series of experiments the signals to be tested were

selected, and two or three replicates were set up with the treatments

randomized in each. At the beginning of each treatment ten virgin

females of the test species were placed in the central section of the

cage in a darkened room in which the temperature was held nearly

constant (± 1° F.). Previously the crickets had been marked with

dots of colored paint on the pronota so that individuals could be

recognized. The speaker was placed in the center of one end of the

cage with the forward edge of the speaker one inch from the screening

(Fig. 100). From treatment to treatment the end at which the speaker

was placed was randomized. The signal was played for ten minutes.

At the end of this period a flashlight was used to determine the

positions of the crickets in respect to the seven arbitrary zones 219

indicated in Figure 111. If a cricket was on the line of division

between two zones, it was considered as in the zone farther from the

starting zone. After their positions were recorded, the crickets were returned to the starting position, and the next signal was played.

The results of the tests were evaluated in two ways — an

arbitrary scoring system and a chi square analysis. The scoring

system involved giving the seven cage zones used in recording the

position of the crickets values of •* 3 to - 3 (Fig. 111). With this

system random movement on the part of the crickets would be expected

to give an average score of 0. A maximum positive response for ten

crickets would produce a score of about ♦ 25 and a corresponding

negative response would be expected to yield a score of - 25. The

maximum score is estimated at 25 instead of 30, because ten crickets, would not remain within a four-inch diameter circle; never were more

than five observed within the circle at once. In the early stages of

experimentation the ♦ 3 and - 3 zones were not used in recording the

positions of the crickets; therefore, the maximum score was + 20.

Before these scores were included in the tables in this section, they

were m ultiplied by 5/4 to make them more comparable with later scores.

Since the scoring system was arbitrary, the scores were not

susceptible to ordinary tests of statistical significance. Therefore,

a chi square test was also applied to the results. For this test the

crickets which were outside the starting zone at the end of the test

period were classed into two groups — plus, if they had moved in the

direction of the speaker, and minus, if they had moved away from it.

If the movement was non-oriented, the plus group should be equal in 220 number to the minus group. With the chi square test, the observed distribution could then be compared with the expected one to see if the observed deviated from the expected by more than could be accounted for on the basis of chance within certain probability lim its. Since the chi square analysis takes into account only the direction of movement and not the extent, it does not utilize all the available information and is inferior to the scores as a basis of comparison between tests. It does, however, give an objective measure of the significance of the response in each test.

R esu lts

Specificity of Responses to Natural Sounds

The first problem investigated was whether female tree crickets are attracted to the calling songs of other species of tree crickets occurring in the same habitat. Table X7I shows the results of experi­ ments with females of _0. quadripunctatus and nigricornis, which occur in weedy fields. Table XVII shows the results of experiments with females of 0. rileyi and exclamationis, two tree-dwelling species. In each case a positive phonotaxis by the females to the calling songs of the male of the same species was demonstrated. No such responses were demonstrated to calling songs of other species occurring in the same habitats. It is of course possible that some response existed, but th at it was too small to be demonstrated by the technique used.

In any event there can be no doubt that female tree crickets show a 221

Table XVI. Response scores and chi square values relating to specificity of response in females of species found in weedy fields. Double asterisk indicates significance at the lf° le v e l.

Recordings Tested 0. quadripunctatus 0. argentinus 0. nigricornis at 74° P. at 74° P. a t 750 p . 37 3 /4 p /s 444 p /s 6 4 i p /s 5600 cps ______5500 cps ______5700 cps ______

_0. quadripunctatus females, 74° P.

R e p lic a te 1 *19 ♦4 R e p lic a te 2 +21 *1

A verage +20 *2

Chi Square 10.00** 0.82

0_. nigricornis females, 75* P.

Replicate 1 +1 ♦3 +17 Replicate 2 ♦2 ♦2 +19 Replicate 3 -3 0 +15

Average 0 +2 ♦17

Chi Square 0.25 1.29 19.55** 222

Table XVTI. Response scores and chi square values relating to specificity of response in females of tree-dwelling species. Double asterisk indicates significance at the IjS level.

Recordings Tested 0. rileyi at 75° F. 0. niveus 0. exclamationis 162 ch/m a t 74° p . a t 76° F. 50e p /s 53 p /s 80 p /s 2600 cps 2600 cps 2600 cps

0. rileyi females. 75° F .

Replicate 1 ♦10 ♦4 +3 Replicate 2 ♦10 ♦4 ♦1 Replicate 3 +14 0 -1

Average ♦11 ♦3 +1

Chi Square 10.70** 1.92 0.20

0. exclamationis females, 75^ F.

R e p lic a te 1 ♦1 -2 ♦6 Replicate 2 0 +5 ♦12 Replicate 3 -1 -3 ♦6

Average 0 0 ♦8

Chi Square 0.00 0.17 7.36** 223 more pronounced positive response to calling songs of their own males than to those of males of other species singing in the ssme habitat.

Basis of Specificity Among Species Producing Continuous T rills

If female tree crickets respond to the calling songs of their own species and not to similar songs of other species in the same habitat, the question arises as to what is the basis of this specifi­ city. The continuous trill, the simplest type of song studied, was the first to be investigated to determine what component was responsible for differential responses among females of different species.

The study of the calling songs of species producing continuous trills revealed that pulse rate is the characteristic which differs most widely among species occurring in the same habitat. Thus it seems a logical hypothesis that the female’s discrimination is based on a response to certain pulse rates. The functioning of insect tympanal organs supports this contention, since their response depends on the pattern of sound intensity changes (amplitude modulation pattern) so long as the carrier frequency is within the range audible to the insect (Pumphrey 1940). The pulses in a tree cricket trill are just such amplitude modulations.

The hypothesis that pulse rate is the song component resulting in specificity of response is complicated by the fact that the pulse rate for each species varies with temperature. Different species sing at identical pulse rates if their temperatures are properly adjusted. Therefore, if pulse rate is the basis for specificity of 224 response of the female, the female must respond to different pulse rates at different temperatures.

To test the role of pulse rate in the response of females of a species, two sorts of recordings were used — (l) recordings of songs of the species made at different temperatures and therefore having different pulse rates and pitch and (2) recordings of artificially produced trills having different pulse rates but the same pitch.

Table XVTII shows the results of an experiment, of this nature with females of quadripunctatus. It is apparent that the crickets responded to markedly different degrees to the same signals at different tempera­ tures. At 70° If. there was a statistically significant response to the calling song of the male at 70° T. and to an artificial trill of

32^- p/s; there was no such response to the song of the male at 80° If. or to an artificial trill at 44^; p/s. At 80° F. the situation was reversed, with the females showing a response to the last two signals but not to the first two. Any role of pitch in the specificity of response seems to be eliminated by the responses to artificial trills of controlled pitch.

In a test at 74° 3P., again with quadripunctatus females, a positive phonotaxis was demonstrated to an artificial trill of 36-3;- p/s but not to an artificial trill of 44^ p/s. Tbese pulse rates correspond to those of males of quadripunctatus and argentinus at that temperature. This test substantiates the importance of pulse rate in explaining the data for quadripunctatus females in Table XVI.

In order to determine the extent to which females of a species respond to different pulse rates at given temperatures, 0_. nigricornis 225

Table RFIII. Response scores and chi square values showing the effect of temperature on the response of females of _0. quadripunctatus to certain sounds. Double asterisk indicates significance at the 1% le v e l.

Recordings Tested 0. quadripunctatus A rtificial Trill at 70° F. at 80° F. 53 3/4 p/s 45 p/s 32g- p/s 44-i- p/s 3600 cps 4200 cps 4000 cps 4000 cps

0. quadripunctatus females, 70° F., 22 August 1956

R eplicate 1 +14 +2 +18 -1 R eplicate 2 +21 +8 +12 +7

Average +18 +5 +15 +3

Chi Square 9.31** 1.00 8.33** 0.11

0. quadripunctatus females, 80° F., 23 August 1956

Replicate 1 -3 +18 0 +21 Replicate 2 +11 +20 -1 +12

Average +4 ♦19 0 +16

Chi Square 0.50 9.00** 1.00 1 1 . 00** 226 females ( from a field containing mostly typical singers) were tested at 70° and. 80° F. with, artificial trills of varying pulse rates.

Figure 112 shows the results of these tests. Each point on the graph represents two tests (using ten crickets each). Although the data are insufficient to give a smooth curve of response' at each tempera­ ture, it i s obvious that the ability to respond to given pulse rates varies w ith temperature and that the pulse rate producing maximum response s .t a given temperature approximates that produced by the male of 1^1x6 species at that temperature. It is also evident that at a given -fcemperature there is a range of pulse rates to which females of a species will respond.

B efore pulse rate can be singled out as the sole factor resulting in specificity of response in species producing continuous trills, another possibility must be evaluated — the possibility of the occurrence of elements in the song of one species that inhibit or repulse tlie females of other species. This possibility is easily

checked t>y selecting recordings of different species singing at the

same p u lse rate (and hence at different temperatures) and seeing whether females of one species will respond to all songs. This was

done u sin g nigricornis females (from a field containing mostly typical

singers) a t 70° F. and recordings of argemtinus and quadripunctatus

at pulse rates similar to typical nigricornis at 70° F. The results

are lishedL in Table 2CDC. The responses of the nigricornis females

to the s o n g s of a l l th re e sp ecies were n early e q u al, so th e re seems

to be no inhibitory element in the song of one species to the females

of another. The results of tests with the speaker turned off are 227

Table XIX. Response scores and chi square values showing the response of 0. nigricornis females to the songs of three species sing­ ing at similar pulse rates. Double asterisk indicates significance at the 1 fo le v e l.

______Recordings Tested ______0. nif 4ricornis 0. argentinus quadripunctatus No at 70° F. at 80° F. at 89° F. Sound ______55 p /s 5500 cps 56j: p /s 5900 cps 55^ p /s 4400cps ___ £. nigricornis females, 70° F.

R eplicate 1 414 411 412 -1 Replicate 2 +9 *16 410 -1

Average 418 414 ♦11 -1

Chi Square 8.55** 10.89** 8.00** 0.55

Table XX. Response scores and chi square values showing the response of CD. exclamationis females at 76° F. to continuous and broken trills. Double asterisk indicates significance at the level.

Recordings Tested 01. exclamationis a t 76° F. Artificial Trill ______80 p /s ______2600 cps ______80 p /s 4000 cps Broken T rill (continuous) B ursts of 2-2?;- sec. Continuous Trill Intervals of x-rr sec. ______

Replicate 1 *16 417 419 Replicate 2 415 418 417

Average 416 413 418

Chi Square 9.00** 8.00** 9.00** 228 given in Table XIX to i l l u s t r a t e scores produced by movement w ithout relation to sound. In other tests with no sound, scores as high as

+4 were obtained.

Specificity in Species Producing Broken Trills

Among species with broken trills, pulse rate varies widely, and one would suspect it to be the basis for specificity of response.

However, another factor which might be of importance in inducing a positive phonotaxis in the females of these species is the occurrence of discontinuities in the song. To test the effect of this latter factor, females of 0. exclamationis were played both broken and continuous trills of their own species. A continuous trill was produced by making a loop from a recording of a pulse sequence over five seconds in length.

The results are shown in Table XX, and it is apparent that the females responded at least as well to a continuous trill as to a broken trill. The data of a further test shown in Table XI indicate that 0. exclamationis females respond positively to an artificial trill of the same pulse rats as the natural song, even though the pitch is quite different. It therefore seems that the factor responsible for specificity in species producing broken trills is the same as in species producing continuous trills, i_. e ., pulse rate.

Specificity in 0. rileyi

In the song of 0. rileyi there are two regular rhythms — the pulse rhythm and the chirp rhythm. It is conceivable that either 229 pulse rhythm or chirp rhythm alone induces positive phonotaxis in the female. Since the two are inseparable in the natural song of rileyi, artificially produced signals were used to test whether one or both were responsible. Three artificial signals were played to females at

75° F.: (1) continuous trill of 48^- p/s (pulse rate similar to natural rileyi but no chirp), (2) regular, pulseless chirp at 168 ch/m

(chirp rate similar to natural rileyi but no pulse), (3) regular, pulsed chirp at 48-^ p/s and 168 ch/m. The pitch was imperfectly controlled — 4000 cps in (l) and 3000 cps in (2) and (3) — because more suitable recordings were not available at the time. However, in work with other species there was no indication of a role of frequency differences within this range. Tor instance, _0. exclama­ tionis females responded as well to an artificial trill pitched at

4000 cps as they did to the natural trill at 2600 cps (Table XX).

Table XXI shows that the crickets did not respond significantly to the continuous trill but showed a positive phonotaxis to the pulseless chirp. Tnis indicates that chirp rhythm is essential in the response of the female of 0. riley i. Since pulsed chirps pro­ duced more pronounced phonotaxis than pulseless chirps, the pulse rhythm must have a supplementary effect. The prime importance of the chirp rhythm makes the response quite specific, since 0 . rileyi is the only chirper with a uniform rhythm in its habitat.

The range of chirp rates to which rileyi females will respond at a given temperat\ire was not determined. It is possible that in western United States, females of Fulton's Race A and Race B respond only to the chirp rates of their respective males. 230

Table XU. Response scores and chi square values showing the response of 0 . rileyi females at 75° F. to certain artificial and natural sounds. Single asterisk indicates significance at the 5$ level. Double asterisk indicates significance at the ifo le v e l.

______Recordings _Tested______A rtificial A rtificial Pulse- A rtificial Pulsed 0. rileyi Trill less Chirp Chirp at 75° F. 48^ p/s 48^ p/s 5ChJr p/s 168 ch/m 168 ch/m 162 ch/m 4000 cps ______5000 cps ______5000 cps ______2600 cps

Replicate 1 +6 +5 +20 +16 Replicate 2 -1 +15 +20 +20 4 00 Average +2 +10 +20 t—*

Chi Square 0.40 5.40* 16.00** 16.00**

D iscussion

These experiments demonstrate that the response of the female to the calling song of the male is a factor in the reproductive isolation of species of tree crickets. This being the case, it becones evident why song characters are valuable in separating closely related, synpatric species: unlike the commonly used morphological characters, some song characters are directly related to the biological distinct­ ness of the populations concerned.

Summary

Experiments in which caged virgin females of four species of tree crickets were subjected to recorded natural and artificial sounds suggest these generalizations: (1) Sexually responsive females show a positive phonotaxis to the calling songs of the males of their own species but not to the songs of other species occurring in the same habitat.

(2) In species with continuous trills, pulse rate is the deter­ mining factor in the response of the female to the calling song.

The specific pulse rate to which the female exhibits greatest response varies with temperature in the same manner as does the pulse rate in the song of the male.

(3) In species with broken trills, females are equally attracted to continuous trills of the same pulse rate, and pulse rate is still the determining factor in the response of the female.

(4) In species with regular chirps, the pulse rate does not determine the response of the females, since they will respond to pulseless chirps. The response is to a regular series of sounds of longer duration than the pulses. However, the pulse rhythm may strengthen the response of females to the chirp rhythm. APPENDIX: DISTRIBUTIONAL RECORDS OF HIE OECANTEIHAE

OF HIE EASTERN UNITED STATES AKX) CANADA

The following tables give the sources of the distribution records for each species and the inclusive dates during which the adult has been reported in each county. Published records of distribution are included only i f they are deemed r e lia b le . The sources o f published records are indicated by references to the bibliography of this dissertation. Records based upon personal determination of specimens in collections are identified by the following abbreviations: BBF —

B. B. Fulton Collection, Entomology Museum, North Carolina State

College, Raleigh, North Carolina; MCZ — Museum of Comparative Zoolo­ gy, Harvard University, Cambridge, Massachusetts, VI. L. Brown, Asso­ ciate Curator of Insects; OSM — Ohio State Museum, North High and

15th Avenue, Columbus, Ohio, E. S. Thomas, Curator of Natural History;

OSU — Ohio State University Entomological Collection, Ohio State

U n iv e rsity , Columbus 10, Ohio, J . N. K null, C urator; FAS — Academy of

Natural Sciences of Philadelphia, 19th and the Parkway, Philadelphia

3, Pennsylvania, J. A. G. Rehn, Curator of Insects (only a portion of the Oeeanthinae in this collection were examined); PSU — Entomology

Collection, Department of Zoology and Entomology, Pennsylvania State

University, University Park, Pennsylvania; RD/1 — collection of R. D.

Alexander, Museum of Zoology, University of Michigan, Ann Arbor,

Michigan; TTW — author’s collection; USNM — United States National

Museum, W ashington 25, D. C ., Ashley B. Gurney, C urator of O rthoptera

(only a portion of the Oeeanthinae in this collection were examined).

232 233

Records based upon the song of the species are identified by the notation TJW-s, if the record is a personal one, and by RDA-s, if it is based upon a tape recording made by R. D. Alexander.

Table XXIX. Distribution of Neoxabea binunctata in the eastern United States.

STATE Source of Record Inclusive Dates County (locality) ALABAMA Madison (Huntsville) PAS

ARKANSAS Polk PAS 21 July

CONNECTICUT Fairfield (Hew Canaan) Morse 1920 14 Aug. Middlesex (Portland) Morse 1920 11 Sep. Hew Haven (New Haven) HCZ 30 Aug.

DISTRICT OF COLUMBIA Washington USHM

GEORGIA Charlton (Billy’s Island) Rehn & Eebard 1916 July Fulton (Atlanta) USNM 16 Aug. Jackson (Thompson M ills) USNLi Rabun Rehn & Hebard 1916

ILLINOIS Champaign (Urbana) Hebard 1934 22 Ju ly 31 Ju ly Cook (R iv ersid e, e tc .) Hebard 1934 21 Aug. 11 Sep. DuPage (Glen Ellyn) Hebard 1934 7 Sep. Rock Is la n d (Rock Islan d ) McNeill 1891 Aug. St. Clair (Mascoutah) Hebard 1934 Scott (Manchester) Hebard 1934 23 Aug. Vermillion (M iliary) Hebard 1934 29 Ju ly 7 Aug. White (Grayville) Hebard 1934 Will (Joliet) Hebard 1934 31 Aug.

INDIANA Marion B latchley 1920 Vigo B latchley 1920 3 Aug. 27 Aug. 254

Table XHI (cont.)

STATE Source of Record Inclusive Dates County (locality) IOWA Clayton Froeschner 1954 Henry Froeschner 1954 Johnson (Solon) PAS 12 Aug. S tory Eroeschner 1954 Van Buren Froeschner 1954

KANSAS M orris BEE 20 July

MARYLAND Kent (Chestertown) Rehn 8; Hebard 1916 25 Aug. Montgomery (Cabin John, etc.) USNM 19 July 29 Sep.

MASSACHUSETTS Middlesex (Holliston) KCZ 2 Sep.

NSW JERSEY Cape May (Cape May) PAS 11 Aug. 23 Aug.

NEW YOKE Nassau (Sea Cliff) IvICZ Richmond (Staten Island) Davis 1889 Rockland (RamapoKts.) Fulton 1915 Suffolk (Orient) USNM 1 Sep.

NORTE CAROLINA A-she (Jefferson) BBF 7 Sep. Anson Brimley 1938 Ju ly Avery (Grandfather M t.) BBF 15 Aug. Franklin (Louisburg) BBF 1 Aug. 18 Aug. Granville ('Oxford) BBF 23 Aug. Haywood (Black lit.) PAS Aug. Madison (Walnut) PAS 20 Aug. McDowell (Buck Creek Gap) BBF 22 Sep. Wake (Raleigh) TJ17 8 Aug.

OHIO Adams QSM 10 Aug. A.shland OSM Astabula (Jefferson) OSU Champaign OSLI 6 Sep. Cuyahoga (Chagrin Falls) OSM 17 Oct. Delaware OSLI Erie (Cedar Point) BBF; OSU 21 Aug, Fairfield (Sugar Grove) OSU 235 Table XXII (cont.)

STATE’ County (locality) Source of Record Inclusive i Dates OHIO (c o n t.) F ranklin 03M; OSU; TJN 30 Ju ly 14 Sep. Fulton (Wauseon) OSU 10 Sep. Greene OSLI 18 Oct. Hamilton (Cincinatti) OSM 30 July 10 Aug. Hocking (Neotama) OSU 17 Aug. 23 Oct. Knox OSU; PAS 1 Aug. 22 Sep. Licking TJW 14 Sep. Mercer PAS 12 Aug. Montgomery QS6-I 6 Oct. Perry (Crooksville) OSU 1 Sep. Seneca (Tiffin) OSU 26 A*ug. Shelby TJW 13 Sep. Tuscarawas o a i 28 Sep. Union cm 4 O ct.' Williams (Mud Lake) EQA-s 19 Aug.

PEI©STLVANIA Adams (Arendtsville) PSU 21 July Allegheny (Jeannette) Kebard 1938 Bedford (Sulpher Sprs.) U3NM Centre (State College) PSU 4 Aug. 10 Sep. Clinton (Tamarack) PSU 9 Sep. Delaware PAS; U3NM 23 Aug. 26 Aug. Philadelphia PAS 9 Sep.

TENNESSEE Benton (Camp Mack M orris) TJW 27 July

TEXAS Cameron (B row nsville) OSU 8 May 8 Aug. Dallas (Dallas) MCZ 23 Ju ly V ic to ria (V icto ria) USNM 22 June

VIRGINIA Buckingham Davis 1926 F a irfax MCZ; USHM 1 Aug. 11 Aug. Nelson Davis 1926; USNM 21 July 12 Aug.

WISCONSIN Richland USNM 18 Aug. 256

Table XXIII. Distribution of Oecantlius exclamationls in the eastern United States.

STATE Source of Record Inclusive Dates County (locality) ARKANSAS Polk PAS 21 Aug.

CONNECTICUT New Haven- (New Haven) Morse 1920; PAS 20 Aug. 11 Oct.

DELAWARE New C astle (Nev/ark) OSM 26 Aug.

DISTRICT OP COLUMBIA Washington USNM

GEORGIA Bartow (Cartersville) USNM 21 July Pulton (Atlanta) USNM 13 July 22 Ju ly

ILLINOIS Champaign (Urbana, etc.) Hebard 1934; USiSl 29 July 18 Oct. Clay (Clay City) Hebard 1934 22 Aug. Jersey (Grafton) Hebard 1934 26 Aug. Kankakee (S t. Anne) PAS 20 July Macon (Decatur) Hebard 1934 11 O ct. Union (Aldridge) Kebard 1934 12 Aug. Washington (Dubois) Hebard 1934 25 Aug.

IOWA Henry (lit. Pleasant) BBP; Proeschner 1954; PAS 22 Aug. 24 Oct. LOUISIANA Rapides TJW

MARYLAND Montgomery HcAtee & Caudell 1917; USNM 14 July 29 Sep. MICHIGAN Charlevoix (High Island) USNM 29 Sep.

MISSOURI Greene (Williard) PAS 8 Aug. Taney (Hollister) Blatchley 1920 Wright (Lit. Grove) PAS 16 Aug. 237 Table XHCII (cont.)

STATE Source of Record Inclusive Dates County (locality) HEW JERSEY Camden (Cibbsboro) PAS 4 Aug. Cape May (Wildwood Jc .) PAS 8 Aug. Monmouth (Farm ingdale, e tc .) Davis 1907 Union (Cranford) Davis 1907

MBS'; YORK Hew York (Central Park) BBF; OSU; PAS 12 Aug. 8 Sep. Richmond (Statin Island) 3SF: USM.' 12 Aug. 21 Sep. Suffolk (Orient) U3IM 1 Sep. Tompkins (Ithaca) USiJM

WORTH CAROLIHA Alamance Brimley 1938 Buncombe (Swannanoa) OSK 2 Sep. Burke (Linville) Brimley 1938 H arnett Brimley 1938 Madison (Walnut) PAS 20 Aug. Wake BBF; Fulton 1951 30 June 15 .Aug. Watauga (Blowing Rock) Brimley 1938

OHIO Adams GSM 30 July 20 Oct. A llen OSK 27 Aug. 15 Oct. Ashland 0SL1 Athens (Athens) OSM 14 Oct. Champaign OSM Aug. Defiance CSM 2 Aug. Delaware OSM 6 Oct. E rie Blatchley 1920; OSK; OSU 12 Aug. 12 Sep. F a ir f ie ld OSK 6 Sep. 10 Oct. F ran k lin OSM; RDA-s; TJW 28 July 24 Oct. Greene OSM 18 Oct. Guernsey OSM 30 Sep. Hocking qsl : 13 Oct. Licking BBF 25 Aug. 6 Oct. Lucas OSH 4 Aug. Kadison ostvl 27 Aug. M ercer OSK 21 Sep. Muskingum OSK 12 Oct. Ottawa OSK 15 Aug. 4 Sep. Paulding OSK 5 Sep. 1 Oct. Pike OSK 3 Oct. Union OSM; OSU 4 Oct. 9 Oct. Washington OSK 14 Aug. 28 Sep. 238 Table XXIII (cont.)

STATE Source of Record Inclusive Dates County (locality) OKLAHOMA Noble PAS 31 July

PENNSYLVANIA Centre (State College) PSU 22 Aug. 3 Oct. Luzerne (Dupont) USNM 15 Aug. Philadelphia (Chestnut Kill) Hebard 1934; PAS 9 Sep. 13 Sep.

TENNESSEE Montgomery (Clarksville) B latchley 1920 Washington (Johnson City) B latchley 1920

VIRGINIA Amelia (Amelia) USNM 8 Aug. Appomatox TJW 23 July Buckingham Davis 1926 July Aug. C raig (McAfee’s Gap) PAS 8 Sep. Fairfax (Falls Church) USNLI 22 Oct. Nelson (Wingina) Davis 1926 28 July Aug. Roanoke (Catawba R idge), PAS 8 Sep.

WEST VIRGINIA Kardy (Sugar Knob) USNM

Table XXIV. Distribution of Oecanthus niveus in the eastern United States.

STATE Source of Record Inclusive Dates County (locality) ARKANSAS Pike (Daisy State Park) TJW-s 16 June

CONNECTICUT Hartford (Hartford) Rehn 8c Hebard 1916

DELAWARE New Castle OSM; USNE 16 Aug. 23 Aug. Sussex (Ellendale) OSM 27 Aug.

DISTRICT OF COLUMBIA Washington USNM Aug. 3 Nov. 239 Table XXIV (cont.)

STATE Source of Record In c lu siv e Dates County (locality) FLORIDA. Alachua (G ainesville) Davis 1915 26 Sep. Brevard (LaGrange) Davis 1914 10 Sep. Kardee (Zolfo Springs) USNM 15 July Holmes TJW-s 15 Sep. Jefferson (Konticello) Davis 1915 4 Oct. Levy(Gedar Keys) USNM June Pinellas (Dunedin) B latchley 1920 Polk (Lakeland) Rehn & Hebard 1914 10 Nov. Putnam (Welaka) F ria u f 1953

GEORGIA Chatham ( I s le o f Hope) Rehn L Hebard 1916 3 Sep. Decatur (Spring Creek) Rehn & Hebard 1916 29 July Dougherty (Albany) Rehn & Hebard 1916 F u lto n (Buckhead) Rehn & Hebard 1916 Jackson (Thompson’s Mills) Allard 1910a; USNM 29 June Oct. Jones TJYf 14 Sep. M itchell (DeWitt) Rehn & Hebard 1916 25 Ju ly Rabun (Pinnacle Peak) Rehn & Hebard 1916 20 Aug. Stephens (Toccoa) Rehn & Hebard 1910b 15 Aug.

ILLINOIS Alexander (Cache) Hebard 1934 Champaign (Urbana) USNM 10 Oct. 24 Oct. G allatin (Shawneetown) Hebard 1934 Jackson Hebard 1934 Macon (Decatur) Hebard 1934 Mason (Havana) Hebard 1934 McHenry (Algonquin) Hebard 1934 P ia tt HQA-s 4 Sep. Pope Hebard 1934 Pulaski (Kamak) Hebard 1934 Rock Island (Moline) McNeill 1S91 29 Sep. V erm ilion Hebard 1934 All records Hebard 1934 14 Aug. 16 Oct.

INDIANA Crawford (Wyandotte) USNM Aug. Floyd Blatchley 1920 LaPorte (Sraith) S trohecker 1937 Lawrenc e B latchley 1920 Putnam B latchley 1920 Union TJW 10 Aug. Vigo Blatchley 1920 240 Table XXIV (cont.)

STATE County (locality) Source of Record In clu siv e Dates IOWA Boone Froeschner 1954 Cedar Froeschner 1954 Henry (lit. Pleasant) BBF; PAS 17 Sep. 1 Oct. Tohnson Froeschner 1954 Linn Froeschner 1954 Sioux Froeschner 1954 S tory (Ames) BBF 27 Aug. Woodbury Froeschner 1954 All records Froeschner 1954 8 Aug. 11 Nov.

KANSAS Brown (Fairview ) Is e ly 1905 Douglas BBF Sedgwick (Wichita) USNM 7 Sep.

KENTUCKY Fayette (Lexington) Garman 1904 13 Sep.

LOUISIANA Orleans (New Orleans) USNM Rapides TJW 15 Tune Vernon TJN-s 15 Tune

MARYLAND Baltimore (Idlewilde) usnk 12 Aug. Montgomery USNM 30 Aug. 11 Nov.

MASSACHUSETTS B arnstable (Buzzard’s Bay) Pierce 1948 5 Sep. Hampshire (Amhearst) USNM 5 Sep. 26 Oct. Middlesex (Cambridge) USNM 3 Oct. Norfolk (Wollaston) USNM 15 Aug. Suffolk (Boston) Morse 1920 Worcester (Oxford) Allard 1911b Sep.

MICHIGAN Livingston (George Reserve) C a n tra ll 1945 6 Aug. 21 Sep. Washtenaw B latchley 1920

MISSISSIPPI Holmes (Holmes Co. S t. Pic.) TJW-s 14 Tune Tishomingo (Inka) PAS 14 Tuly

MISSOURI Lawrence (Pierce City) USNM 10 Oct. Taney (Hollister) BBF 5 Sep. 241 Table XXIV (cont.)

STA.TE County (locality) Source of Record In clu siv e Dates NEW JERSEY B urlington Rehn 1902a 13 Aug. Camden (Clementon) Rehn 1904a 13 Aug. M orris (Uendham) u sr n i 12 Oct.

NEW YORK Ontario (Geneva) BBF; USNM 24 Aug. 18 Oct. Orange (West Point) USNM 2 Sep. 10 Sep. Richmond Davis 1895 Seneca (Bast Varick) PSU 25 Aug. Wayne (Clyde) USNM 6 Sep.

HOKES CAROLINA. Alleghany (Roaring Gap) BBF 30 Aug. Avery (Grandfather Kt.) BBF 17 Aug. Buncombe (Sulphur Springs) Rehn & Hebard 1910a 29 Sep. Harnett (Spouting Springs) 3BF 15 Aug. Jackson (Balsam) Rehn & Hebard 1916 15 Sep. McDowell (Marion) BBF 22 Sep. Moore (Southern Pines) Rehn & Hebard 1916 Nov. Polk (Saluda) Rehn A Hebard 1916 Randolph (Ashboro) BBF Stanly (Badin) BBF 3 Sep. Surry BBF; Rehn & Hebard 1916 19 Aug. 16 Nov. Wake BBF; Fulton 1951 21 Ju ly 8 Nov.

OHIO Adams OSM 13 Aug. A llen OSM 19 Oct. Belmont (Bellaire) OSU 29 Aug. C a rro ll RDA-s 14 Aug. Champaign OSM Aug. Clermont OSM 20 Sep. Columbian OSM 10 Sep. Coshocton OSM 31 Oct. Delaware OSM 1 Sep. 6 Oct, E rie OSM; TJiT 20 Aug. 1 Sep, F a ir f ie ld OSM; TJW 13 Aug. 14 Oct. F ran k lin OSM; OSU; TJ1V 13 Aug. 27 Oct, Greene GSM 18 Oct Hamilton (Cincinatti) OSU 13 Oct, Hardin OSM 16 Sep Highland OSM 18 Sep. 20 Oct Hocking (Neotoma) OSM 6 Aug. 5 Nov Holmes OSM 14 Oct Jackson OSM 2 Sep 242 Table LQCEV (cont.)

STATE County (locality) Source of Record In clu siv e Dates OHIO (c o n t.) Lake OSLI 14 Aug. Lawrence OSM 27 Aug. L icking BBF; OSM; TJW 25 Aug. 10 Nov. Logan OSM; OSU 24 Aug. 29 Aug. Lucas OSM 16 Oct. Madison OSM 27 Aug. 5 Sep. Mercer OSM 7 Sep. Muskingham OSM 17 Sep. Noble OSM 28 Oct. Ottawa OSM 7 Aug. 4 Sep. Paulding OSM 23 Sep. 30 Sep. Pickaway OSM 10 Oct. 14 Oct. Pike OSM 14 Oct. Portage OSM 2 Sep. Ross OSM 10 Aug. S cioto OSM 13 Sep. Shelby TJW 13 Sep. Summit OSM 21 Aug. 8 Sep. Tuscarawas OSLI 28 Sep. 29 Sep. Union OSM 4 Oct. Y/ashington OSM 28 Sep. Wayne (Wooster) GSM 11 Aug. W illiam s OSM 29 Aug. 2 Nov.

PENNSYLVAHIA. Centre (State College) PSU 12 Aug. 4 Oct. Delaware (Newtown S q .) USNM 23 Aug. Lancaster (Lancaster) PSU Luzerne (Dupont) USNM 14 Aug.

SOUTH CAROLINA Aiken TJW 14 Sep. Florenc e (F lorence) Rehn A Hebard 1916 6 Sep.

SOUTH DAKOTA Yankton (Yankton) Hebard 1925 27 Sep.

TENNESSEE Dyer (Dyersburg) TJW-s 23 S ep. Hamilton (Chattanooga) USNM 19 Sep. Montgomery (C larksville) Blatchley 1920 Shelby (Memphis) PAS 16 Sep. ----(Great Smoky Mts.) OSM 20 Sep. 243 Table 20CEV (cont.)

STATE County (locality) Source of Record In clu siv e Dates TEXAS Bexar (San Antonio) u sm : 1 July Colorado (Columbus) USNM Dallas (Dallas) USNM Galveston (Dickinson) PAS 20 July G ille sp ie OSU 29 June Kerr (Kerrville) PIAS 17 Aug. Robertson (Iieame) PAS 14 Aug.

VIRGINIA Albemarle (Charlottesville) Fox 1917 17 July Arlington (Rosslyn) USNM 30 Oct. Buckingham Davis 1926 30 July F a irfax HcAtee & Caudell 1917; USNM 17 Aug. 11 Nov. Loudoun (Bluemont) USNM 31 Aug. Nelson (Wingina) Davis 1926 Prince William (Occoquan) usnm 28 Sep. P rin cess Anne (Cape Henry) USNM 25 Aug.

WEST VIRGINIA Greenbrier (White Sulpher Spr) USNM 25 Sep. N icholas OSM 10 July

Table XXV. Distribution of Oecanthus rileyi in the eastern United States and Canada.

STATE County (locality) Source of Record Inclusive Dates CONNECTICUT Hartford (Windsor) Rehn & Hebard 1916

DELAWARE New Castle (Eockessin) Houghton 1903

GEORGIA Jackson (Thompson's Mills) Allard 1910a

ILLINOIS Champaign (Urbana) Hebard 1934 Cook Matthews 1942 23 Aug. 244 Table 2CCV (cont.)

STATE County (locality) Source of Record Inclusive Dates ILLINOIS (c o n t.) DuPage Alexander 1956 Lake Hebard 1934 Lawrence (St. Francisvilie) Hebard 1934 Marion (Centralia) Hebard 1934 Mason Hebard 1934 McHenry Hebard 1934 McLean (Normal) Hebard 1934 Piatt (lionticello) Hebard 1934 All records Hebard 1934 2 Aug. 16 Oct.

INDIANA Boon USNM Aug.

IOWA Boone Froeschner 1954 Cedar Froeschner 1954 C lark Froeschner 1954 Dickinson (Lake Okoboji) BBF 17 Aug. Emmet Froeschner 1954 G uthrie Froeschner 1954 Henry Froeschner 1954 Mahaska Froeschner 1954 Polk Froeschner 1954 Scott (Davenport) M cNeill 1889 23 Ju ly S tory Froeschner 1954 Woodbury Froeschner 1954 All records Froeschner 1954 29 Ju ly 30 O ct.

KANSAS Brown (Fairview ) I s e ly 1905 Sedgewick I s e ly 1905 Ju ly Oct.

KENTUCKY Fayette(Lexington) Garman 1904 11 Sep.

MAINE Cumberland (Portland) Morse 1920

Essex (Salem) Brooks 1882 30 Sep. 17 Oct. Middlesex (Cambridge) Faxon 1901 Worcester (Oxford) Allard 1910b, 1930a Aug. Oct.

MICHIGAN Berrien (Benton Harbor) Hubbell 1922b 4 Sep. Livingston (George Reserve) Cantrail 1943 24 July Oct. 245 Table XX7 (cont.)

STATE County (locality) Source of Record Inclusive D ates MICHIGAN (c o n t.) Washtenaw (Ann Arbor) Cantrall 1943 21 July Wayne (D e tr o it) M atthews 1942 12 Aug. 10 Sep.

NEBRASKA Lancaster (Lincoln) Bessey & Bessey 1898 15 Aug. 3 Sep. HEW YORK Ontario (Geneva) BBF, Parrot 1909 20 July 18 Oct. Richmond Davis 1895

NORTH CAROLINA Alleghany (Roaring Gap) BBF 20 Aug. Avery (Cranberry) Brimley 1938 Aug. Hoke TJW 10 Aug. Moore TJW-s 10 Aug. onro A lle n OSM 29 Ju ly 5 Sep. Champaign o a i 20 J u ly Aug. E rie OSU; TJW 26 Ju ly 1 Sep. F a ir f ie ld OSM 28 J u ly F ra n k lin OSM; TJW 7 Ju ly 23 O ct. Geauga OSU 19 Aug. H ardin OSM 19 Sep. Henry OSK 18 Aug. Holmes OSM 24 Aug. Lake osi'.: 14 Aug. L ick in g BBF; OSM 25 Aug. 25 Sep. Logan OSM 16 Aug. L orain e OSM; TJW 26 July 22 Aug, Lucas OSM; OSU 4 Aug. 27 Aug, M adison OSM 29 Aug. M arion OSK 31 Aug. 9 Oct, M ercer OSM 7 Sep Muskingum OSM - 7 Oct Ottawa OSK Ju ly 4 Sep Ross OSM 10 Aug. Union OSLI 29 J u ly W ashington OSM 14 Aug. W illiam s OSLI 29 Aug.

ONTARIO Essex (Leamington) Walker 1904 7 Aug. Huron (Goderich) Walker 1904 19 Aug. Kent (Chatham) Walker 1904 10 Aug. 246 Table XXV (cont.)

STATE Source of Record Inclusive Dates County (locality) ONTARIO (c o n t.) Lamb to n (S arn ia) Walker 104 12 Aug. Prince Edward Urquhart 1941a 27 Ju ly 21 Aug. York (Toronto) Walker 1904 Aug. 13 O ct.

mTt®YLVANIA Centre (State College) PSU 3 Aug. 30 Aug.

SOUTH DAKOTA Lincoln (Canton) Hebard 1925 26 Aug. 27 Aug. Yankton (Yankton) Hebard 1925 27 Sep.

TEXAS I I Dallas (Dallas) USNM 27 0 0 Uvalde (Uvalde) OSU 30

■VEILIONT Lamoille (Mt. Mansfield) usnl : 24 Aug.

VIRGINIA A rlin g to n A llard 1930a; USNL.: 13 Aug. 7 Sep. Augusta USNM 20 Aug. Bath (Warm Spr. Mt.) TJW 11 Aug. Carroll (H illsville) BBF 25 Aug. Nelson (Wingina) D avis 1926 Aug.

WEST VIRGINIA Grant (Gormania) TJW-s 11 Sep. Preston (Aurora) usnl :

WISCONSIN Fond du Lac (Fond du Lac) Hebard 1934 247

Table 2XVI. Distribution of Oecanthus latipennis in the eastern United States.

STATE County (locality) Source of Record Inclusive Dates ARKANSAS Washington (Fayetteville) MCZ 5 Sep.

DISTRICT OF COLUMBIA W ashington OSU; USNL! 17 Aug. 19 Aug.

GEORGIA Chatham (Isle of Eope) Rehn & Fannin (Ivlorganton) usnm 8 Sep, Fulton (Buckhead) Rehn & Hebard 1916 Jackson (Thompson's H ill) A llard 1910a Jones TJW 14 Sep. Meriwether (WarmSprs.) Rehn & Hebard 1916 Pickens (Jasper) Rehn & Hebard 1916 Richmond (Augusta) Rehn & Hebard 1916 Stephens (Toccoa) Rehn & Hebard 1916 W hitfield (Dolton) Rehn & Hebard 1916

INDIANA Putnam Blatchley 1903 Vigo Blatchley 1903 10 Aug. 1 O ct.

ILLINOIS Adams (Quincy) Eebard 1954 Alexander (Olive Branch) Hebard 1954 Edgar (Borton) Hebard 1934 Jackson Hebard 1934 Jersey (Grafton) Hebard 1934 Lawrence (Lawrenceville) Hebard 1954 Marion (Odin) Hebard 1934 Massac (Metropolis) Hebard 1934 P iatt (Sangamon Twp.) RDA-s 4 Sep. Pope Hebard 1934 Pulaski (Kamak) Hebard 1934 Union Hebard 1954 Vermillion (Hilliary) Hebard 1954 W hite McNeill 1891 6 O ct. All records Hebard 1934 31 July 13 Oct.

IOWA Henry (Mt. Pleasant) BBF 12 Sep, M uscatine Froeschner 1954 4 Sep. S to ry Froeschner 1954 Aug. 248 Table XZVT (cont.)

STATS Source of Hecord Inclusive Dates County (locality) KANSAS Douglas (Lawrence) Hebard 1925

KENTUCKY C aldw ell TJW 25 Sep. Fleming TJW 25 Sep. F ra n k lin TJW 15 O ct. Craves TJW 25 Sep. Grayson TJW 15 O ct. Greenup (Fullerton) TJW Hopkins TJW 25 Sep. Lawrence TJW Lyon TJW 25 Sep. Mason TJW 26 Sep. P ike TJW

MARYLAND Montgomery McAtee & Caudell 1917; USNM 18 Aug. 31 Oct. MISSISSIPPI Washington (Leland) BBF 16 Sep.

NEW JSHSSY Burlington (Medford) Rehn 1902a 13 Aug. Ocean (Lakehurst) BBF 14 Sep. 20 Sep. Somerset (Warrenville) OSM 30 Aug.

NSW YOHK Kichmond Davis 1895 Suffolk (Calverton) Hebard 1925

NOHTH CAROLINA A lleghany (R oaring Gap) BBF 30 Aug. 11 Sep. B eau fo rt BBF 12 O ct. Cumberland (Fayetteville) Rehn & Hebard 1916 9 Sep. Graham 03M 21 Aug. G u ilfo rd Rehn & Hebard 1916 Halifax (Weldon) Rehn A Hebard 1916 Harnett (Spouting Sprs.) BBF 11 Sep. Pender (Burgaw) BBF 10 Sep. Polk (Saluda) Rehn A Hebard 1916 Randolph (Asheboro) BBF 3 Sep. Wake BBF; Fulton 1951 • 22 Aug. 20 Nov. Wayne (G oldsboro) Rehn & Hebard 1916 Yadkin BBF 20 Sep. 249 Table IGCVT (cont.)

STATE Source of Record In c lu s iv e D ates County (locality) OHIO Adams OSM 24 Sep. 19 O ct. A thens OSLI 22 Aug. 6 Sep. Belmont OSU 29 Aug. C a rro ll OSM Champaign OSM 6 Sep, C lin to n OSM Columbiana OSLI 9 Sep. Delaware OSLI 1 Oct. 7 Oct. B rie OSLI 30 Aug. F a ir f i e ld OSLI 31 Aug. 16 O ct. F ra n k lin OSLI; TJW 14 Aug. 13 Oct. G a llia OSU Greene OSLi 30 Sep. Hocking BBF; OSLI; OSU; TJW 20 Aug. 12 Nov. Jackson OSM; OSU 1 Sep. 2 Sep. Lawrence OSM 27 Aug. L ick in g BBF; OSLI 19 Aug. 5 Sep. M adison OSM 27 Aug. M arion OSM 17 O ct. Meigs OSM; OSU 13 Sep. 16 Sep. Muskingum OSil 17 Sep. Noble OSM 28 O ct. Ottawa OSM P e rry OSU 25 Aug. 7 Sep. Pickaway OSM 12 Aug. 25 O ct. S c io to OSM; OSU 15 Aug. 13 Sep. Tuscaraw as OSLI 28 Sep. W ashington OSM 28 Sep.

PENNSYLVANIA. A llegheny Hebard 1938 Philadelphia (Philadelphia) Hebard 1938

SOUTH CAROLINA Aiken TJW 14 Sep. Charleston (Ashley Junction) Rehn & Hebard 1915 Florence (Florence) Rehn & Hebard 1916 6 Sep. Pickens (Clemson) BBF 27 Sep.

SOUTH DAKOTA Yankton (Yankton) BBF 27 Sep.

TENNESSEE D ickson TJW 8 O ct. Dyer (Dyersburg) TJW 24 Sep. Sevier (Gatlingburg) OSLI 17 Sep. 250 Table XXVI (cont.)

STATE County (locality) Source of Record Inclusive Dates VIRGINIA Appomatox (Appomatox) LICZ 6 Sep. Bath (Hot Springs) Fox 1917 6 Sep. Brunswick TJW 13 Sep. Buckingham Davis 1926 Chesterfield TJW 13 Sen. Dinwiddie (Petersburg) Rehn & Hebard 1916 Essex (Tappahannock) Fox 1917 21 Aug. 12 Nov. Fairfax HcAtee & Caudell 1917 14 Sep. 4 Oct. Nelson (Wingina) Davis 1926 Orange (Orange) Rehn & Hebard 1915 Princess Anne (Bayville) Rehn cc Hebard 1916 Spotsylvania (Fredericksburg) Rehn & Hebard 1916

Table S H I, Distribution of Oecanthus nigricom is in eastern United States and Canada.

STATS County (locality) Source of Record Inclusive Dates CONNECTICUT Hartford (Hartford) USUI 17 Sen. New London (LIystic) B ritto n 1927

DELAWARE New Castle (Newark) USUI; Houghton 1909 a 25 Sep,

ILLINOIS Champaign Forbes 1905; Hebard 1934 2? Aug. 28 Aug. Coles (Charleston) Hebard 1934 Cook Hebard 1934; USUI Cumberland (Neoga) Forbes 1905 27 Sep. DuPage HDA-s 9 Sep. 15 Sep. Edgar (Borton) Hebard 1934 G-allatin (Shawneethown) Hebard 1934 Hardin (Elizabethtown) Hebard 1934 Jackson Hebard 1934 Jo Daviess (Galena) u sm i 22 Sep. Johnson (Cypress) Hebard 1934 Lake Hebard 1934 251 Table 3DC7II (cont.)

STATE Source of Record Inclusive Dates County (locality) ILLINOIS (dont.) Lawrence (Lawrenceville) Hebard 1934 Madison (Alton) Hebard 1934 Marion (Centralia) Hebard 1934 McHenry (Algonquin) Hebard 1934 Piatt (Centerville) RDA-s 4 Sep. Pope Hebard 1934 Rock Island (Milan) Forbes 1905 27 Sep. White (Norris City) Hebard 1934 Whiteside (Fulton) Hebard 1934 Winnebago (Rockford) Hebard 1934 All records Hebard 1934 12 Ju ly 29 Sep. All records Forbes 1905 10 Aug. 10 Oct.

INDIANA F ra n k lin TJW 2 Oct. Lake (Hammond) Blatchley 1903 18 Sep. Tippecanoe (Lafayette)„ USNM 20 Aug. Union TJW 2 O ct. Wayne TJW 2 O ct.

IOWA Dickenson (Lake Okoboji) Fulton 1926b 18 Aug. Mahaska (Oskaloosa) Fulton 1926b / 4 Sep. Muscatine (Muscatine) USM 11 Sep. Story (limes) BBF 13 Aug. 10 O ct. Woodbury (Sioux City) Fulton 1926b 16 Aug.

KANSAS Brown (Fairview) USNL! 25 Aug.

KENTUCKY Bourbon TJW 26 Sep. F ra n k lin TJW 16 O ct. H arriso n TJW 16 Oct. Mason TJW 16 O ct. M uhlenberg TJW 15 Oct. R obertson TJW 16 Oct. S c o tt TJW 16 O ct.

MAINS Cumberland (Brunswick) Morse 1920 Hancock (Mt. Desert Island) Procter 1946 6 Sep. 16 Sep. L in co ln TJW 30 Aug. Somerset (Hoxie) Morse 1920 252 Table KXYII (cont.)

STATE County (locality) Source of Record Inclusive Dates MARYLAND Montgomery (Glen Echo) USNM 19 Sep.

MASSACHUSETTS Franklin (E. Northfield) USNM Hampshire (Oummington) USNM 2 Sep. Middlesex (Reading) USNM 21 Aug. Norfolk (Blue Hill) Henshaw 1900 Worcester (Oxford) Allard 1911a; USNM Aug. Oct.

MICHIGAN Berrien (Lakeside) Hancock 1911; PAS Aug. 2 Oct. Lenawee (Adrian) USNM Livingston (George Reserve) Cantrall 1945 21 Aug. 4 Sep. Wayne (D e tro it) USNM Sep.

MINNESOTA Hennepin (Minneapolis) OSM 1 Aug.

NEBRASKA Cuming (West P o in t) USNM Sep.

NEW HAMPSHIRE Cheshire (Jaffrey) Henshaw 1900 Grafton (Franconia) Morse 1920

NEW JERSEY Burlington (Moorestown) F3U 16 Aug. Cumberland (Manumuskin) Rehn 1902a 20 O ct.

NEW YORK Albany (Kamer) F e lt 1906 Aug. Sep. Cattaraugus OSM 19 Aug. 21 Aug. Erie (Buffalo) OSM Greene USNM Aug. Ontario (Geneva) Parrott 1911; USNM 26 Aug. 1 Oct. Orange (West Point) USNM 2 Sep. 10 Sep. Seneca (East Yarick) PSU 25 Aug.

NORTH CAROLINA Ashe (Jefferson) Rehn & Hebard 1916 Sep. Avery BBF; Rehn & H. 1916 28 Aug. 21 Sep. Buncombe BBF; Rehn & H. 1916 Aug. 21 Sep. Haywood (W aynesville) BBF 14 Sep. Watauga (Yalle Crucis) BBF 10 Sep. 253 Table XfflI (cont.)

., , Source of Record Inclusive Dates County (locality) ______CHIO A lle n OSM 14 Aug, 7 Sep. A stab u la OSM 14 Aug. 20 Aug. Belmont OSU 29 Aug. B u tle r TJW 25 Sep. 2 Oct. C a rro ll TJW 16 Sep. Champaign OSM 21 Aug. 4 Oct. Clermont OSU 30 Aug. Darke OSM 5 Sep. Delaware OSM 13 Sep. 7 Oct. E rie OSU; TJW 25 J u ly 14 Sep. F a ir f i e ld OSM; OSU 12 Sep. 16 Oct. F ra n k lin OSU; TJW 27 Ju ly 12 O ct. F u lto n OSM 1 Sep. G a llia USNM 25 Aug. Geauga OSM 26 Aug. Guernsey OSM 30 Sep. H am ilton OSM; OSU 13 Oct. 16 Oct. H ardin OSM 16 S ep. Hocking OSM; OSU 4 Sep. Huron OSM 5 Sep. 15 Sep. Jackson OSM 5 Sep. Lake OSM 14 Aug. L icking BBF; OSM 18 Aug. 5 Sep. Logan OSLi; OSU 16 Aug. 17 Sep. L oraine GSM 8 Sep. 4 Oct. Lucas OSM; OSU 11 Aug. 7 Sep. M adison OSM 27 Aug. 5 Sep. M arion OSLi 17 O ct. M ercer OSLi 2 Sep. 7 Sep. Miami OSM 26 Aug. Montgomery TJW 2 O ct. Morrow OSLI 20 Aug. Muskingum OSM 17 Sep. Ottawa OSM; OSU; USNM 7 Aug. 11 Sep. Pickaway OSLI 14 Sep. 14 Oct. P ortage OSLI 15 Aug. 7 Sep. P reb le TP,7 8 Sep. 2 O ct. R ichland OSM 27 Aug. Ross OSM 10 Aug. Shelby 031,1 8 Sep. S ta rk OSM 23 Sep. Summit OSM 7 Sep, Tuscarawas OSM 9 Sep. 28 Sep. W arren OSLi; TJW 11 Sep. 2 Oct. W ashington oai 25 Sep, Wayne OSM 11 Aug. 254 Table XXVEI (cont.)

STATE Source of Record I n d u s ive s D ates County (locality) OHIO (c o n t.) W illiam s OSM 29 Aug. 7 Sep. Wood OSM; OSU 31 Aug. 6 Sep. Wyandot OSLI 15 Sep. 26 Sep.

ONTARIO Bruce (Bruce Peninsula) Walker 1904 23 Aug. 24 Aug. Huron (Goderich) Walker 1904 19 Aug. Kent (Chatham) 'Walker 1904 10 Aug. Lamb to n Walker 1904 12 Aug. 13 Aug. Ni pis s in g Walker 1904 23 Aug. 12 Sep. Prince Edward (Picton) Urauhart 1941a 7 Aug. Simcoe OSM 21 Aug. York (Toronto) Walker 1904 Aug. Sep.

PENNSYLVANIA Centre (Blaclc Hoshannon) PSU 4 Sep.

SOUTH DAKOTA Brookings Fulton 1926b 17 Aug. 18 Sep. Lincoln (Canton) Fulton 1926b 27 Aug. Yankton (Yankton) Fulton 1926b 27 Sep.

TENNESSEE Carter (Roan Mt. Station) Rehn & Hebard 1916 3 Sep.

VERMONT Windsor (Woodstock) Morse 1920

VIRGINIA Fairfax (Arlington) USNM 6 Sep. Highland (Monterey) Fox 1917 1 Sep. Loudoun (Bluemont) USNM 31 Aug.

WEST VIRGINIA R itc h ie TJW Tucker (Canaan Valley) USNM 21 Aug.

WISCONSIN wane BBF 1 Aug. 22 Aug. 255

Table XXVTII. Distribution of tlie willow form of Qecanthus nigricomis.

STATE Dates County (locality) Source of Record In clu siv e OHIO Brown (S ardinia) OSLi 6 Aug. Delaware (Scioto River) TJW 6 Sep. Franklin (Olentangy River) TJW 5 Sep. Ottawa (Bay P oint) OSLI 18 July Shelby (Lake Lorareie) OSLI; TJW 9 Sep. IS Sep.

Table XXIX. Distribution of Qecanthus celerinictus.

STATE County (locality) Source of Record Inclusive> Dates ALABAMA Butler (Greenville) FAS 3 Aug. Covington TJW 15 Sep. Crenshaw TJW 15 Sep. Dallas (Selma) PAS 9 Sep. Mobile (Springhill) PAS 25 Aug.

ARKANSAS Pike (Daisy State Park) TJW 16 June 17 June S co tt TJW 17 June S eb astian TJW 17 June S ev ier TJW 16 June

DELAWARE Sussex TJW 21 July

FLORIDA Holmes TJW 15 Sep. J ackson TJW 15 Sep.

GEORGIA Baker TJW 15 Sep. Bibb (Macon) PAS 31 July Decatur (Bainbridge) PAS 6 Sep. Dooly TJW 15 Sep. Dougherty TJW 15 Sep. Houston TJW 15 Sep. Thomas (Thom asville) PAS 23 June 256 Table 30CEX (cont.)

STATE County (locality) Source of Record Inclusive Dates GEORGIA (co n t.) Worth. TJW 15 Sep.

KENTUCKY Caldwell TJW 15 Oct. Graves TJW 25 Sep. 15 Oct. Eickman TJW 25 Sep. 15 Oct. Hopkins TJW 15 Oct. Lyon TJW 15 Oct. M arshall TJW 25 Sep. 15 Oct. Muhlenberg TJW 15 Oct. Ohio TJW 15 Oct.

LOUISIANA. Desoto TJW 15 June Rapides TJW 15 June Vernon TJW 15 June

MARYLAND W orcester TJW 21 July

MISSISSIPPI Adams (Natchez) PAS; TJW 13 Sep. Grenada TJW 14- June Harrison (Pass Christian) PAS 23 Aug. Panola TJW Warren (Vicksburg) PAS 13 Sep.

MISSOURI B u tler TJW IS June Mississippi TJW

NORTH CAROLINA Columbus (Lake Waccsmaw) PAS 8 Sep. H alifax (Weldon) PAS 24 July Haraet TJW 10 Aug. Moore (Southern Pines) PAS 10 Sep. New Hanover (Winter Park) PAS; TJW 5 Sep. 7 Sep. Wake (Raleigh) PAS; TJW 8 Aug. 22 Sep. Wayne (Goldsboro) PAS 25 July

SOUTH CAROLINA Aiken TJW-s 14 Sep. Charleston (Charleston) OSM 25 Sep. Chesterfield TJW 14 Sep. Florence (Horence) PAS 6 Sep. Kershaw TJW 14 Sep. 257 Table XXIX (cont.)

STATS County (locality) Source of Record Inclusive i Dates SOUTH CAROLINA, (co n t.) Lexington TJW 14 Sep. Ri cbland(Columbia) PAS; TJW 28July 14 Sep.

TENNESSEE Campbell OSK 10 Ju ly C arro ll tjw 8 Oct. Dyer TJW 23 Sep. Gibson TJW 8 Oct. Humphreys TJW 8 Oct. Obion TJW 25 Sep. 15 Oct. Robertson TJW 6 Oct. Shelby TJW 14 June

TEXAS Bowie TJW

VIRGINIA Brunswick TJW 22 Ju ly Nansemond TJW 22 Ju ly Northampton TJW 22 July Spotsylvania (Fredericksburg) PAS 20 Ju ly

Table XXX. Distribution of Qecanthus argentinus in the eastern United States.

STATE County (locality) Source of Record Inclusive Dates ARKANSAS Benton TJW 17 June Crawford TJW 17 June Montgomery TJW 17 June S co tt TJW 17 June Sebastian TJW 17 June S evier TJW 16 June Washington TJW 17 June

ILLINOIS Adams (Quincy) Hebard 1934 Alexander Hebard 1934; RDA 16 June Champaign (Urbana) Hebard 1934 258 Table XXX (cont.)

STATE County (locality) Source of Record Inclusive Dates ILLINOIS (co n t.) Clark (Casey) Hebard 1934 Edwards (Albion) Hebard 1934 Gallatin (Shawneetown) Hebard 1934 Hardin Hebard 1934 J ackson Hebard 1934 Johnson (Vienna) Hebard 1934 Lawrence Hebard 1934; TJW 19 June Madison (Alton) Hebard 1934 Marion (Centralia) Hebard 1934 lias on Hebard 1934 Montgomery (Hillsboro) Hebard 1934 P ia tt EDA 5 Sep. Pope Hebard 1934 P u lask i Hebard 1934 Saline (Harrisburg) Hebard 1934 Union Hebard 1934 Y^abash (Mt. Cam el) Hebard 1934 Washington (Dubois) Hebard 1934 White (Grayville) Hebard 1934 All records Hebard 1934 27 June 20 Sep.

INDIANA F ran k lin TJW 2 Oct. Greene TJW Knox TJW Union TJ7 10 Aug. 2Oct.

IOWA Franklin and Wright (Dows) BBF 9 July Fremont (Hamburg S t. Pk.) BBF 30 Ju ly Pottawattamie (Council Bluffs) BBF 31 July Woodbury (Sioux C ity) F ulton 1926b 16 Aug.

KANSAS Atchison BBF 9 July Chautauqua BBF Douglas BBF; Hubbell 1922a 4 July 24 Sep. Leavenworth BBF 30 June Montgomery (Independence) USNM Sedgwick (Wichita) USKLI 18 July 27 July Sumner BBF

KENTUCKY Bourbon TJW 14 July 26 Sep. Caldwell TJW 15 Oct. Fleming TJW 14 July 259 Table 2GQC (cont.)

STATE County (locality) Source of Record Inclusive Dates KENTUCKY (co n t.) F ran k lin TJW 15 July 16 Oct. Graves TJW 25 Sep. 15 Oct. Hart TJW 6 Oct. Hickman TJW 25 Sep. 15 Oct. Hopkins TJW 15 Oct. Lyon TJW 15 Oct. M arshall TJW 15 Oct. Mason TJW 14 July N icholas tjw 15 July 14 July Ohio TJW 15 Oct. S co tt TJW 14 July 16 Oct.

LOUISIANA Rapides TJW-s 15 June

MINNESOTA Big Stone OSM 20 Joig. 5 Sep. Ottertail (Battle Lake) PAS 8 Aug. Stevens (Morris) PAS 6 Aug.

MISSISSIPPI Panola TJW 14 June

MISSOURI B u tler TJW 18 June Greene PAS; TJW 18 June 9 July Howell TJW Lawrence TJW 18 June Mississippi TJW 18 June St. Charles (St. Charles) PAS 19 July Shannon TJW Stoddard TJW 18 June Texas TJW Webster TJW Wright TJW

NORTH DAKOTA Barnes (Talley City) PAS 21 Aug.

OHIO Adams (Spencer Knob) OSK Brown TJW 15 July B u tler TJW 2 Oct. Clermont OSM 19 Oct. F a ir f ie ld TJW 12 July Fayette TJW 15 July 260 Table XXX (cont.)

STATE County (locality) Source of Record Inclusivei Dates OHIO (c o n t.) F ran k lin 03M; TJVJ 5 July 25 Oct. G a llia OSH 26 June Highland TJVJ 15 Ju ly Hocking OSM 4 July Knox TJVJ 26 July L icking TJVJ 26 Ju ly Madison TJVJ 2 Oct. Ross TJVJ 24 July Vinton TJVJ 15 Oct.

OKLAHOMA Comanche (F ort S il l) Hubbell 1922a 27 Sep. 10 Nov. Payne (Stillwater) USM! Sequoyah tjv ; 17 June

TENNESSEE Dickson TJVJ 8 Oct. Dyer TJVJ 11 June 14 Oct. Gibson TJVJ 8 Oct. Obion TJVJ 25 Sep. 15 Oct. Robertson TJVJ 6 Oct. Shelby TJVJ 14 June

TEXAS Cameron (Brow nsville) USHvl Cooke USNM 20 Sep. Dallas (Dallas) u sm : 19 May 21 Sep. Duval (San Diego) USNLI 15 May 12 Aug. Frio (Pearsall) USNLI 30 Nov. Lamar (P aris) u sm i 5 June McLennan (Uaeo) USNLI 30 Oct. Menard (Menard) USNM 27 Oct. Tarrant (Fort Worth) USNM Travis (Austin) USNLI Uvalde (Uvalde) 03U 23 May Victoria (Victoria) USNM 2 June 261

Table XKKE. Distribution of Oecantbus quadripunctatus in the eastern United States and Canada.

STATE Inclusive Dates County (locality) Source of Record ALABAMA Houston (Dothan) PAS 6 Sap. Lee (Opelika) PAS 2 Aug. Montgomery (Montgomery) FAS 8 Sep.

AHKAKSA3 Garland (Hot Springs) USHH 28 Oct. M ille r TFT 16 June S evier Tjyj 16 June

DELAWARE Kent (Bombay Iiook) OSM 2 Aug.

DISTRICT OP COLUMBIA Washington OSU; USMi 16 Aug. Oct.

FLORIDA Alachua (Ga ine svi1le ) BBF 27 July 4 Hov. C h arlo tte RDA 9 May Columbia (Lake City) USHLi 2 July Duval PAS 13 Aug. 28 Sep. Franklin (Carrabelle) PAS 2 Sep. Hernando (Brooksville) RDA 10 May Highlands (Childs) PAS 6 Aug. Holmes TJW 15 Sep. Jackson TFT 15 Sep. Lake RDA 4 May Levy(Cedar Key) FAS; USIM 4 June 15 Aug. Marion (Ocala) PAS 19 Sep. Monroe (Cape Sable) RDA 8 May Suwannee (Live Oak) PAS 26 Aug. Walton (DeFuniak Sprs.) PAS; TJW 30 Aug. 15 Sep. Washington TFT 15 Sep.

GEORGIA Bibb (Macon) PAS 31 July Bryan (Groveland) PAS 21 Sep. Charlton (Sevan Mile) FAS 23 Sep. Chatham ( I s le of Hope) PAS 3 Sep. C risp TJW 15 Sep. Decatur (Bainbridge) PAS 5 Sep. Dooly TFT 15 Sep. Dougherty (Albany) PAS 1 Aug. Habersham PAS 5 Aug. 6 Sep. Jones TFT 14 Sep. 262 Table XXXI (cont.)

STATE Source of Record Inclusive Dates County (locality) GEORGIA (co n t.) Rabun (Raven Rock lit.) PAS 4 Sep. Richmond (Augusta) PAS 29 Ju ly Thomas (Thomasville) PAS 21 Sep. Worth TJW 15 Sep.

ILLINOIS DuPage RDA 9 Sep. 16 Sep. McHenry (Algonquin) PAS 9 Sep. McLean (Normal) EDA 1 Oct. Pope (Williams l i t . ) PAS 11 Aug.

INDIANA Bartholomew BBP 25 Aug. P ran k lin TJW 2 Oct. Tippecanoe Pox 1915 20 Aug. Union TJW 10 Aug. 2 Oct. Wayne TJW 2 Oct.

IOWA Dickenson (Lake Okoboji) Pulton 1926b 18 Aug. Mahaska (Oskaloo sa) Pulton 1926b 4 Sep. S tory (Ames) Pulton 1926b 13 Aug. 10 Oct. Woodbury (Sioux C ity) Fulton 1926b 16 Aug.

KANSAS Brcrvm (Hiawatha) USNM Aug.

KENTUCKY Caldwell TJW 15 Oct. P ran k lin TJW 16 Oct. Greenup (Fullerton) TJW 6 n u g . H arrison TJW 16 'Oct. Lawrence TJW 6 Aug. Mason TJW 16 Oct. Pike TJW 6 Aug. Robertson TJW 16 Oct. S co tt TJW 16 Oct.

LOUISIANA Calcasieu (Lake Charles) IAS 10 Aug. Desoto TJW 15 June Rapides TJW Vernon TJW 15 June Table XXXI (cont.) 263

STATE County (locality) Source of Record Inclusive Dates MAINE Cumberland (Brunswick) Horse 1920 Hancock P ro c te r 1946 30 Aug. Somerset (Hoxie) Horse 1920

MARYLAND Allegbany TJW

MASSACHUSETTS Hampshire (Cummington) USNLI 1 Sep. Middlesex (Sherbom) B latchley Nantucket Morse 1920 Norfolk (Blue Hill) Henshaw 1900 Worcester (Oxford) Allard 1911a; USNLI Sep. Oct.

MICHIGAN B errien Hubbell 1922b 30 -Aug. 9 Sep. Livingston (George Reserve) C a n tra ll 1943 24 July 8 Sep.

MINNESOTA O tte r ta il PAS Aug. S co tt PAS 1 Aug. ’Washington (St. Paul) AS 20 Aug.

MISSISSIPPI Adams TJW 15 June Forrest (Hattiesburg) PAS 11 Sep. Kinds (Jackson) E/13 12 Sep. Holmes (Holmes Co. S t. Pk.) TJW 14 June Itawamba (Pulton) PAS 14 July Lauderdale (Meridian) PAS 10 Sep. Monroe (Hamilton) PAS 15 Ju ly Montgomery (Ylinona) PAS 15 Sep.

MISSOURI St. Louis (Glen Echo) USNLI 19 Sep. 26 Sep.

NEW HAMPSHIRE Rockingham (Seabrook) USNM

NEW JERSEY Atlantic (Reega) jtriio 29 Aug. M ercer TJW Ocean (Stafford’s Porge) PAS 16 Sep. Bassaic (Great Notch) USNLI 8 Sep.

NEW YORK Chatauqua 0SLI 18 Aug. 264 Table XHXE (cont.)

STATE Source of Record Inclusive Dates County (locality) NEW YORK (c o n t.) New York (Central Park) USNL: 8 Sep. Ontario (Geneva) BBF; Parrott 1911 17 Aug. 26 Sep. Orange (West Point) USHH 31 Aug. 9 Sep. Rensselaer (E. Greenbush) OSU 9 Sep. Tompkins (Ith aca) PAS 11 Sep. Westchester (Hosholu) PAS 20 Aug.

NORM CAROLINA Buncombe (A sheville) PAS 24 Sep. Haywood (Black I t s . ) PAS Aug. Jackson (Balsam) PAS 15 Sep. Hoore (Southern Pines) PAS 28 Sep. New Hanover PAS 7 Sep.

OHIO Adams OSU 30 July 18 Sep. A llen OSIvI 20 Aug. 5 Sep. Ashland OSLI 10 Aug. As tab u la OSU; OSU 14 Aug. 20 Aug. Athens OSU 6 Sep. Belmont OSU; OSU 29 Aug. 12 Sep. Brown TJW 5 Oct. C a rro ll TJW 16 Sep. Champaign OSU; OSU 24 July 14 Oct. C lark OSU 2 Oct. Clermont OSU; OSU 18 Aug. 4 Sep. Columbiana OSU 9 Sep. Cuyahoga OSU 15 Aug. Darke OSU 5 Aug. Defiance OSU 13 Aug. Delaware OSU 13 Sep. 14 Oct. E rie OSU; OSU; TJW 26 Ju ly 16 Sep. F a ir f ie ld 031!; RDA-s; TJW 8 Aug. 11 Oct. F ran k lin OSU; TJW 30 July 2 Nov. F u lto n OSU 31 Aug. Geauga OSU 16 Sep. Greene OSU 30 Sep. Guernsey 0SC.i 23 Aug. Highland OSU; TJW 10 Aug. 5 Oct. Hocking OSU; OSU; TJW 10 Aug. 15 Oct. Huron cm 15 Sep. Jackson OSU 29 Aug. Lake OSU 19 Aug. Licking OSU; TJW 25 July 25 Sep. Logan OSIvI 16 Aug. L orain OSU; TJW 26 July 8 Sep. 265 Table LCCCE (cont.)

STATE Source of Record Inclusive Dates County (locality) OHIO (c o n t.) Lucas OSLI 13 Aug. 6 Sep. Madison OSLi; TJW 27 Aug. 2 Oct. L ari on OSM 17 Oct. Meigs OSLI; OSU 18 July 13 Sep. Mercer OSLI 2 Sep. 7 Sep. Miami OSLI 27 Sep. 2 Oct. Montgomery OSM; TJW 26 Aug. 2 Oct. Noble OSM 28 Oct. Ottawa OSM 18 July 4 Sep. Paulding OSM 30 Ju ly Perry OSM 8 Aug. Pickaway OSM 12 Aug. 14 Sep. Portage OSM 2 Sep. P reble OSM; TJW 5 Sep. 2 Oct. Ross OSM; TJW 10 Aug. 15 Oct. Sandusky OSM; OSU 24 July 4 Aug. S cioto OSM 16 Aug. Seneca OSU 27 Aug. Shelby TJW 13 Sep. S tark QSEvI 24 Aug. Summit OSLI 7 Sep. 8 Sep. Tuscarawas OSM 28 Sep. Union OSLI 29 July 28 Aug. Vinton TJW 15 Oct. Warren TJW 2 Oct. Washington OSM; OSU 16 Ju ly 13 Aug. Williams OSM 7 Sep. Wood OSM; OSU 31 Aug. 6 Sep.

ONTARIO Kent (Chatham) Walker 1904 10 Aug. Lamb ton Walker 1904 13 Aug. 'Welland Blatchley 1903 4 Sep. York (Toronto) Walker 1904 Sep,

PENNSYLVANIA Bucks (Cornwells) PAS 7 Sep. Centre (State College) PSU 3 Aug. 14 Aug, Cumberland (Carlisle) USNLI 8 Aug. Lancaster (Lancaster) PSU Aug. M ifflin (Lewistown) Udine & Pinckney 1940 Montgomery (Hickorytown) Udine & P. 1940 Perry (Duncannon) Udine c: P. 1940 Philadelphia (Mt. Airy) PAS 18 Sep. Washington (Washington) Udine & P. 1940 266 Table XXXI (cont.)

STATE County (locality) Source of Record Inclusive> Dates RHODE ISLAND Washington (Saunderstovm) PAS Sep.

SOUTH CAROLINA Aiken TJW 14 Sep. Charleston (Charleston) OSM 25 Sep. Lexington TJW 14 Sep.

SOUTH DAKOTA Brookings Fulton 1926b 17 Aug. 18 Sep. Hughes (Pierre) Fulton 1926b 10 Sep. Lincoln(Canton) F ulton 1926b 27 Aug. Yankton Fulton 1923b 29 Aug. 27 Sep.

TENNESSEE Anderson OSM 23 Sep. Dickson TJW 8 Oct.

TEXAS Bowie TJW 16 June

VEHLIOHT Lamoille (Mt. Mansfield) USNM 24 Aug. Windsor (Woodstock) Morse 1920

VIRGINIA Bath TJW 11 Aug. Carroll (Hillsville) BBF 25 Aug. Dickenson TJW 6 Aug. H alifax TJW 11 Aug. Wythe (W ytheville) PAS 5 Sep.

WISCONSIN Dane BBF 13 -Aug. 17 Aug. Jackson USNM 14 Aug. Richland (Lone Rock) PAS 10 Aug. 267

Table JGQCEI. Distribution of Oecanthus pini.

STATS Source of Record Inclusive Dates County (locality) CONNECTICUT Windham (Woodstock) USNM Sep.

DISTRICT OF COLUMBIA Washington USNM

MAINS Hancock P ro cte r 1946 30 Aug.

MAHHAHD C alv ert (Plum P oint) USNM 9 Aug. Washington (Sharpsburg) USNM 18 Sep. Worcester (Snow Hill) TJW-s 21 July

MASSACHUSETTS Barnstable (Cape Cod) Morse 1919 Dukes (West Chop) Morse 1920 Esse:; (Gloucester) Fulton 1915 Plymouth (Wareham) Morse 1920

MICHIGAN Livingston (George Reserve) BBF 24 Aug. Oakland (Milford) Cantrall 1943 3 Aug. 14- Sep.

NEW JERSEY Ocean (Lakehurst) BBF 19 Sep.

HEW YORK Albany (Earner) F elt 1906; USNM 27 July 6 Sep. New York (Central Park) USNM 14 Aug. Richmond (Staten Island) Fulton 1915 S uffolk BBF 11 Aug. 24 Sep.

NORTH CAROLINA. Caldwell (Lenoir) TJW- 3 7 Aug. Wake BBF; TJW 10 Ju ly 22 Sep.

OHIO Adams OSM 30 July 7 Aug. Athens BBF; OSM 28 Oct. Columbiana Hebard 1938 E rie Hebard 1938 F a ir f ie ld OSLI; TJW 13 Aug. 18 Sep. Hocking OSLI; OSU 24 July 30 Sep. Lake OSLI 18 Aug. P erry OSM 19 Aug. 268

Table XXXII (cont.)

STATE Source of Record County (locality) Inclusive Dates OHIO (c on t.) Washington OSM 28 Sep.

PEI'fflSYLYAHIA Columbia (Bloomsburg) BBE 5 Sep.

VIRGIMA Alleghany TJYJ-s 23 Ju ly Amherst TJW-3 23 July Bath TJVi-s 11 Aug. Halifax TJW 11 Aug. llelson (VJingina) Davis 1926 28 July Rockbridge TJW-s 23 July REFERENCES CITED

Alexander, Richard D. 1956. A comparative study of sound production in insects, with special reference to the singing Orthoptera and Cicadidae of the Eastern United States. Ph.D. Dissertation, Ohio S ta te Univ. 529 p. 180 f .

1957. Sound production and associated behavior in insects. Ohio Jour. Sci. 57(2): 101-113. 13 f.

Allard, E. A. 1910a. LIusical crickets and locusts in north Georgia. Proc. Eht. Soc. Wash. 12(1): 32-43. 6 f.

1910b. Some New England Orthoptera observed in late October. Ent. News 21(8): 352-357.

1911a. The musical habits of some New England Orthoptera in September. Ent. News 22(1): 28-39.

1911b. The stridulations of some eastern and southern crickets (Orth.). Ent. News 22(4): 154-157.

1912. Variation in the stridulations of Orthoptera. Ent. News 23(10): 460-462.

1917. Synchronism and synchronic rhythm in the behavior of certain creatures. Amer. Natural. 51(607): 438-446.

1918. Rhythmic synchronism in the chirping of certain crickets and locusts. Amer. Natural. 52(622-3): 548-552.

1929a. Our insect instrumentalists and their musical techniques. Ann. Rept. Smithsonian Inst. 1928: 563-591. 16 pi.

1929b. The last meadow katydid; a study of its musical reactions to light and temperature (Orthoptera:Tettigoniidae). Trans. Amer. Ent. Soc. 55: 155-164.

1930a. The c h irp in g r a te s of the snowy tr e e c ric k e t (Oecanthus niveus) as affected by external conditions. Canad. Ent. 62(6): 131-142. 3 f.

1930b. Changing the chirp-rate of the snowy tree-cricket Oecanthus niveus with air currents. Science 72(1866): 347-349.

Ashmead, William H. 1894. Notes on cotton insects found in M ississippi. Insect Life 7(1): 25-29.

269 270

A y e r s , Howard. 1083. On the development of Oecanthus niveus and its parasite, Teleas. Mam. Bost. Soc. Nat. Hist. 3(8): 225-281. 41 f. 8 pi.

Baker, C. F. 1905. Second report on Pacific slope Orthoptera. Invert. Pacif. 1: 71-83.

Beach, Edith Penfield. 1238, Effect of X-ray upon the snowy tree cricket, Oecanthus nigricornis argentinus. Trans. Kans. Acad. Sci. 41: 303-315. 51 f.

Bessey, Carl A . , and Edward A. Bessey. 1898. Further notes on thermo­ m eter c ric k e ts . Amer. N atu ral. 32(376): 263-264. 1 f .

Beutexunuller, William. 1894a. Description of a new tree-cricket. Tour. N. Y. Ent. Soc. 2(2): 56.

1894b. Notes on some species of North American 'Orthoptera, with descriptions of new species. Bull. Amer. Kus. Nat. Hist. 6(11): 249-252.

1894c. Descriptive catalogue of the Orthoptera found within fifty m iles of New York C ity . B u ll. Amer. Hus. N at. H ist. 6: 253-316-.-

Blatchley, W. S. 1892. The G-ryllidae of Indiana. Proc. Ind. Acad. Sci. 1891: 126-144.

1903. The Orthoptera of Indiana. 27th jinn. Hep. Ind. Dept. Geol. and Nat. Resources 123-471. 2 pi. 122 f.

1920. Orthoptera of northeastern America. Nature Pub. Co.: Indianapolis, Ind. 784 p. 246 f. 1 pi.

Borror, Donald I., and Carl R. Reese. 1953. The analysis of bird songs by means of a vibralyzer. Wilson Bull. 65(4): 271-276.

Brim ley, C. S. 1938. The in se c ts of North C arolina. N. C. Dept. A gr.: R aleigh. 559 p.

Britton, W, E. 1927. Twenty-sixth report of the state entomologist of Connecticut, 1926. Conn. Agr. Exp. Sta. Bull. 285: 161-283. 16 pi. 11 f.

Brooks, Margarette W. 1882. Influence of temperature on the chirp of the cricket. Popular Sci. Monthly 20: 268.

Browne, Ashley C. 1931. Oecanthus niveus (DeGeer) in some California fruits. Mon. Bull. Dept. Agric. Calif. 20(10-11): 633-643. 271

Busnel, Karie-Claire. 1954. Etude des chants et du comportement acous­ tique d1Oecanthus pellucens male. Annales des Epiphyties, fascicule spfecial consacre au colloque sur 1*acoustique des Orthopteres 175-202. 9 f.

Busnel, Marie-Claire, and Rene-Guy Busnel. 1954. La directivite acous­ tique des deplacements de la femelle d*Oecanthus pellucens Scop. Ib id . 356-564.

Busnel, Renb-Guy. 1954. Sur certains rapports entre le moyen d'infor­ mation acoustique et le comportement acoustique des Orthopteres. Ib id . 281-303. 8 f .

Busnel, Ren&-Guy, Franqois Pasquinally, and Bernard Dumortier. 1955. La tremulation du corps et la transmission amc supports des vibrations en resultant corame moyen d*information a eourte portee des Ephippigeres male et femelle. Bull. Soc, Zool. France 80(1): 18-22. 8 f.

Cantrall, Irving J. 1943. Ecology of the Orthoptera and Demaptera of the George Reserve, Michigan. Hus. Zool. Univ. 1'ich. llisc. Pub. 54: 1-182. 10 p i. 2 maps. 3 f .

Caesar, Lawson. 1919. Insects as agents in the dissemination of plant diseases. 49th Ann. Rept. Ent. Soc. Ontario 1918: 60-66.

C audell, A. K. 1902. Rotes on O rthoptera from Oklahoma and Indian territory with descriptions of three new species. Trans. Amer. Ent. Soc. 28: 83-91.

Caulfield, F. B. 1888. A sketch of Canadian Orthoptera. 18th Ann. Rept. Ent. Soc. Ontario 1887: 59-72. 4 f.

Chopard, Lucien. 1951. A revision of the Australian Grylloidea. Records of the South Australian lluseum 9(4): 397-533. 89 f .

Davis, William T. 1889. List of the Orthoptera found on Staten Island. Ent. Amer. 5(4): 78-81.

1895. Insects at Watchogue and Beulah Land, Staten Island, N. Y. Jo u r. N. Y. E nt. Soc. 3(3): 140-143.

1907. A new tree cricket from Staten Island and New Jersey. Canad. Ent. 39(5): 173-174.

1914. Notes on Orthoptera from the east coast of Florida with d e sc rip tio n s of two new species of B elocephalus. Jour. I T . Y. E nt. Soc. 22(3): 191-205. 272

Davis, William T. 1915. List of the Orthoptera collected in northern Florida in 1914 for the American liuseum of Natural History, with descriptions of new species. Jour. N. Y. Ent. Soc. 23(2): 91-101. 4 f .

1926. An annotated list of the Derroaptera and Orthoptera in mid­ summer at Wingina, Virginia, and vicinity. Jour. IT. Y. Ent. Soc. 34(1): 27-41.

De Geer, Charles. 1773. Ilemoires pour servir a l'histoire des insectes. Tome III. Stockholm. 696 p. 44 pi.

Dolbear, A. E. 1897. The cricket as a thermometer. Amer. Natural. 31(371): 970-971.

Edes, Hobert T. 1899. Relation of the chirping of the tree cricket (Oecanthus niveus) to temperature. Amer. Natural. 33(396): 935- 938. 1 f .

Faxon, W. 1901. The habits and notes of the New England species of Oecanthus. Psyche 9: 183.

Felt, Ephraim P. 1906. Insects affecting park and woodland trees. Vol. 2. N. Y. State IIus. llem. 8 : 602- 603 , 698-700. Illus.

Fitch, Asa. 1856. Third report on the noxious and other insects of the state of New York. Trans. IT. Y. State Agrie. 16: 321-507.

Forbes, S. A. 1905. Tree-crickets. Oecanthinae. 23rd Rept. State Ent. 111. 215-222.

Fox, Henry. 1915. Notes on Orthoptera and Orthopteran habitats in the vicinity of Lafayette, Indiana. Proc. Ind. Acad. Sci. 1914: 287- 321

1917. Field notes on Virginia Orthoptera. Proc. U. S. Nat. Kus. 52(2176): 199-234.

Friauf, James J. 1953. An ecological study of the Demaptera and Orthoptera of the Welaka area in northern Florida. Ecol. Ilonogr. 23(2): 79-126. 17 f.

Frings, Hubert, and liable Frings. 1957. The effects of temperature on chirp-rate of male cone-headed grasshoppers, Neoconocephalus ensiger. Jour. Exp. Zool. 134(5):

Froeschner, Richard C. 1954. The grasshoppers and other Orthoptera of Iowa. Iowa State Coll. Jour. Sci. 29(2): 163-354. 123 f.

Frost, 3. V/. 1942. General entomology. LlcGraw-Hill Book Co., Inc.: New York. 524 p. 406 f. 275

Fulton, Bentley B. 1915. The tree crickets of Hew York: life history and bionomics. New York Agr. Exp. 3ta. Tech. Bull. 42: 1-47. 6 pi. 21 f.

1925. Physiological variation in the snowy tree-cricket Oecanthus niveus De Geer. Ann. Ent. Soc. Amer. 18(5): 363-383. 6 f.

1926a. The tree crickets of Oregon. Oregon Agri. Exp. 3ta. Bull. 223: 1-20. 8 f .

1926b. Geographical variation in the nigricornis group of Oecanthus (Orthoptera). Iowa. St. Coll. Jour. Sci. 1(1): 43-61. 3 f.

1928a. A demonstration of the location of auditory organs in certain Orthoptera. Ann. Ent. Soc. Amer. 21(3): 445-448.

1928b. Sound perception by insects. Sci. Hon. 27: 552-556.

1931. A study of the genus Nenobius (Orthoptera: Gryllidae). Ann. Ent. Soc, Amer. 24(2): 205-237. 5 f.

1932. North Carolina’s singing Orthoptera. Jour. Elisha Hitch. Sci. Soc. 47: 55-69.

1933. Inheritance of song in hybrids of two subspecies of Nemo- bius fasciatus (Orthoptera). Inn. Ent. Soc. Amer. 26(2): 368- 376. 1 f .

1934. Hhythm, synchronism, and alternation in the stridulation of Orthoptera. Jour. Elisha Hitch. Sci. Soc. 50(l/2): 263-267.

1951. The seasonal succession of Orthopteran stridulation near Haleigh, North Carolina.. Jour. Elisha Hitch. Sci. Soc. 67(1): 87-95. 2 f .

Garman, II. 1904. On an injury to fruits by insects and birds. Ky. Agr. Exp. Sta. Bull, 116(1): 63-78.

Gloyer, N. 0., and 3. 3. Fulton. 1916. Tree crickets as carriers of Leptosphaeria coniotiiyrium (Fckl.) Sacc. and other fungi. N. Y. Agr. Exp. Sta. Tech. Bull. 50: 1-22. 4 pi.

Graber, Veit. 1889. Vergleichende Studien uber die Keimhullen und die Rdckenbildung der Insecten. Denkschriften k. Akad. Wissenschaften Wien. 55: 109-162. 8 pi.

Ballenbeck, Cleve. 1949. Insect thermometers. Nat. Ilist. 58: 256-259, 285-287. 6 f.

Hancock, Joseph L. 1905. The h a b its o f th e s tr ip e d meadow c ric k e t (Oecanthus fasciatus Fitch). Amer. Natural. 39(457): 1-11. 5 f. 274

Hancock, Joseph L. 1911. Nature sketches in temperate .America. A. C. HcClurg & Go.: Chicago. 451 p. illus.

Harris, Thaddeus W. 1833. Insects in_ Report on the geology, mineralogy, hotany, and zoolog 3r of Massachusetts. Press of J. S. and C. Adams: A m herst. 692 p.

1841. A report on the insects of llassachusetts injurious to vegetation. Folsom, Wells, and Thurston: Cambridge. 459 p.

1883. A treatise on some of the insects injurious to vegetation. New Edition. Orange Judd Co.: New York. 640 p. 278 f.

Hart, Charles A. 1892. On the species of Oecanthus Serv. Ent. News 3(2): 33-34.

Haywood, Poland, 1901. The katydid’s call in relation to temperature. Psyche 9: 179.

Hebard, Morgan. 1925. The Orthoptera of South Dakota. Proc. Acad. Nat. Sci. Phil. 77: 33-155.

Hebard, Morgan. 1928. The Orthoptera of Montana. Proc. Acad. Hat. Sci. Phil. 80: 211-306. 2 pi.

1929. The Orthoptera of Colorado. Proc. Acad. Nat. Sci. Phil. 81: 303-425. Map. 1 pi.

1934. The Dermaptera and Orthoptera of Illinois. Bull. 111. Nat. Hist. Surv. 20(3): 125-279. 5 pi. 167 f.

1935. Orthoptera of the Upper Pdo Grande Valley and the adjacent mountains in northern New Mexico. Proc. Acad. Hat. Sci. Phil. 87: 45-82. 3 f.

1938. Where and when to find the Orthoptera of Pennsylvania, with notes on the species which in distribution reach nearest this state. Oecanthinae. Ent. News 49(4): 101-102.

Kenshaw, Samuel. 1900. Hew England Orthoptera. Psyche 9(294): 119.

Houghton, C. O. 1903. An unusual injury by the snowy tree-cricket (Oecanthus niveus DeGeer). 15th Rpt. Del. Agr. Exp. Sta. ISO- 152. 1 p i.

1904. An unusual injury by the snowy tree-cricket and notes on its feeding habits. Ent. Hews 15(2) : 57-61.

1909a. Observations on the mating habits of Oecanthus. Ent. Hews 20(6): 274-279. 275

Houghton, C. 0. 1909b. ITotes on Oecanthus. Canad. Ent. 41(4): 113-115.

Hubbell, Theodore E. 1922a. Notes on the Orthoptera of Worth Dakota. Qcc. Papers Hus. Zool. Univ. Hieh. 113: 1-56.

1922b. The Denaaptera and Orthoptera of Barrien County, Michigan. Occ. Papers Hus. Zool. Univ. Hi ch. 116: 1-77.

Isely, P. B. 1905. notes on Kansas Orthoptera. Trans. Han. Acad. Sci. 19: 238-249.

Jaeger, B., and E. C. Preston. 1854. The life of North American insects. Sayles, H iller, and Simons: Providence. 204 p. 31 f.

Jensen, J. P. 1909a. Courting and mating of Oecanthus fasciatus Harris. Canad. Ent. 41(1): 25-27. 1 f.

1909b. Observations on the oviposition of Oecanthus quadripunctatus Beutennuller. Ent. Hews 20(1): 25-28. 1 pi.

1911. The structure and systematic importance of the spermato- phores of crickets. Ann. Ent. Soc. Amer. 4(1): 63-67. 1 pi.

Johnson, E. H. 1922. Peripheral migration of a centriole derivative in the spermatogenesis of Oecanthus. Science 56(1461): 759-760.

1931. Centrioles and other cytoplasmic components of the male germ cells of the Gryllidae. Zeitschr. Hiss. Zool. 140(1): 115- 166. 6 pi. 4 f.

Kirby, V7. E. 1906. A synonymic catalogue of Orthoptera. Vol. 2. British museum: London. 562 p.

Lugger, Otto. 1897, Grasshoppers, locusts, crickets, cockroaches, etc., of Minnesota. Minn. Agr. Exp. Sta. Bull. 55: 91-387.

Lutz, Prank E. 1924. Insect sounds. Bull. Amer. Hus. Hat. Hist. 50(6): 333-372.

1938. The insect glee club at the microphone. Hat. Hist. 42(5)*.' 338-345, 378. 21 f.

Matthews, Howard D. 1942. On the stridulations of insects. Science 95(2465): 324-325. 2 f.

McAtee, IT. L., and A. N. Caudell. 1917. Hirst list of the Denaaptera and Orthoptera of Plummers Island, Maryland, and vicinity. Proc. Ent. Soc. Hash. 19: 100-122. 1 pi.

McNeill, Jerome. 1889. Notes upon Gryllus and Oecanthus. Ent. Amer. 5(5): 101-104. 276

McNeill, Jerome. 1891. A list of the Orthoptera of Illinois. I. Psyche 6(177): 3-9.

Morse, Albert P. 1919. A lis t of the Orthoptera of New England. Psyche 26(2) : 21-39.

1920. Manual of the Orthoptera of New England, including the locusts, grasshoppers, crickets, and their allies. Proc. Boston Soc. Nat. Hist. 35(6): 197-556. 99 f. 20 pi.

M urtfeldt, Mary E. 1889. The carnivorous habits of tree crickets. In s . L ife 2(5): 130-132.

Packard, A. S., Jr. 1881. Insects injurious to forest and shade trees. Bull. U. S. Ent. Com. 7: 1-275. 100 f.

Parrott, P. J. 1909. Tree crickets and injury to apple wood. Jour. Econ. Ent. 2(2): 124-127.

1911. Oviposition among tree-crickets. Jour. Econ. Ent. 4(2): 216-218. 1 p i.

Parrott, P. J ., and B. B. Pulton. 1913. Notes on tree crickets. Jour. Econ. Ent. 6(2): 177-180. 15 f.

1914. Tree crickets injurious to orchard and garden fruits. N. Y. Agric. Exp. Sta. Bull. 338: 417-461. 10 pi. 9 f.

Parrott, P. J ., V/. 0. Gloyer, and B„ B. Pulton. 1915. Some studies on the snowy tree-cricket with reference to an apple bark disease. Jour. Econ. Ent. 8(6): 535-541.

Pasquinelly, Francois, and M.-C. Busnel. 1954. Etudes preliminaires sur les mecanismes de la production des sons par les Orthopteres. Annales des Epiphyties, fascicule special consacre au colloque sur 1'acoustique des Orthopteres. 145-152. 6 f.

Pierce, George W. 1948. The songs of insects. Harvard Univ. Press: Cambridge, Hass. 329 p. 243 f.

Procter, Uilliam. 1946. Biological survey of the llmnt Desert Region. Part 711. The insect fauna. V/istar Inst, of Anat. and Biol.: Philadelphia. 566 p. illus.

Pumphrey, R. J. 1940. Hearing in insects. Biol. Rev. Cambridge Philos. Soc. 15(1): 107-132. 10 f.

Regen, Johann. 1913. Uber die Anloekung des Weibchens von Gryllus campestris L. durch telephonisch ubertragene Stridulationslaute des l&nnchens. Pfliiger's Archiv fur Fnysiologie 155: 193-200. 1 f. 277

Rehn, James A. G. 1902a. Records of New Jersey and Pennsylvania Orthoptera. Ent. News 13(10): 309-316.

1902b. A contribution to the knowledge of the Orthoptera of Mexico and Central America. Trans. Amer. Ent. Soc. 29: 1-34.

1904a. Notes and records of New Jersey Orthoptera. Ent. Nev/s 15(10): 325-332.

1904b. Notes on Orthoptera from northern and central Mexico. Proc. Acad. Nat. Sci. Phil. 56: 513-549.

Rehn, James A. G., and Morgan Hebard. 1910a. Preliminary studies of North Carolina Orthoptera. Proc. Acad. Nat. Sci. Phil. 62: 615-650.

1910b. Records of Georgia and Florida Orthoptera, with the de­ scriptions of ope new species and one new subspecies. Proc. Acad. Nat. Sci. Phil. 62: 585-598. 2 f.

1914. Records of Dermaptera and Orthoptera from west central and southwestern Florida, collected by William T. Davis. Jour. N. Y. E n t. Soc. 22(2): 96-116.

1916. Studies in the Denaaptera and Orthoptera of the Coastal Plain and Piedmont region of the southeastern United States. Proc. Acad. Nat. Sci. Phil. 68: 87-314. 3 f. 3 pi.

Riley, Charles Y. 1869. First annual report of the noxious, beneficial and other insects of the state of Missouri. Ellwood Kirby: Jefferson City, Ho* 181 p. 98 f. 2 pi.

1873. Fifth annual report on the noxious, beneficial, and other insects of the state of Missouri. Regan & Carter: Jefferson City, Mo. 160 p . 75 f .

1881. Dept. Interior, U. S. Entomological Commission, Bui. 6, General Index and Supplement to the 9 reports on the insects of Missouri. Gov. Print. Office: Washington. 177 p.

1886. Orthoptera. The Standard Natural History 2: 167-203. 1 pi. 43 f . S. E. Cassino & Co.: B oston.

Saussure, Eenri de. 1874. Recherches zoologiques pour servir a l'his- toire de la faune de l'Am&rique Centrale et du Mexique. 6th part. Imprimerie ImpSriale: Paris. 3: 293-516. 2 pi.

1897. Biologia Centrali-Americana. Insects. Orthoptera. 1: 251-255. 278

Scuddsr, Samuel H. 1862. M aterials for a monograph, of the North American Orthoptera, including a catalogue of the known New England species. Jour. Bost. Soc. Nat. Hist. 7: 409-480.

1867. Notes on the stridulation of some New England Orthoptera. Proc. Bost. Soc. Nat. Hist. 11: 306-313.

1868. Catalogue of the Orthoptera of North America described previous to 1867. Smithsonian Misc. Coll. 189: 1-89.

1874. The distribution of insects in New Hampshire. Rept. of the Geologist on the Geology of New Hampshire 1: 331-380.

1893. The songs of our grasshoppers and crickets. 23rd Ann. Rept. Ent. Soc. Ontario 1892: 62-78. 19 f.

1894. Biological notes on American Gryllidae. Psyche 7(213): 3-5.

1900. Catalogue of the described Orthoptera of the United States and Canada. Proc. Davenport Acad. Nat. Sci. 8: 1-101. 3 pi.

1901. Alphabetical index to North American Orthoptera described in the eighteenth and nineteenth centuries. Occ. Papers Bost. Soc. Nat. Hist. 6: 1-436.

Scudder, Samuel E., and Theodore D. A. Cockerell. 1902. A first list of the Orthoptera of New Mexico. Proc. Davenport Acad. Sci. 9: 1-60. 4 pi.

Serville, J. G. Audinet. 1831. Revue methodique des Insectes de l ’ordre des Orthopteres. Ann. Sci. Nat. 22: 28-65, 134-167, 262-292.

Severin, H. C. 1920a. Eleventh annual report of the State Entomologist of South Dakota for the period ending June 30, 1920. S. Dak. State Coll., Brookings. 40 p. 5 f.

1920b. The black-horned tree-cricket. S. Dak. State Ent., Brookings. Circ. 19: 1-4. 1 f.

Shull, A. Eranklin. 1907. The stridulation of the snowy tree-cricket (Oecanthus niveus). Canad. Ent. 39(7): 213-225. 2 f.

Smith, Leslie M. 1930. The snowy tree cricket and other insects injurious to raspberries. Calif. Agric. Expt. Sta. Bull. 505: 1-38. 16 f.

Snodgrass, R. E. 1925. Insect musicians, their music, and their instruments. Smithsonian Rept. 1923: 405-452. 35 f.

Strohecker, H. E. 1937. An ecological study of same Orthoptera of the Chicago area. Ecology 18(8): 231-250. 1 f. 279

Thomas, C. 1870. A methodical table of the crickets. Amer. Ent. & B o ta n is t. 2(7): 206-207.

Titus, E. S. G-. 1903. A new Oecanthus from Illinois. Canad. Ent. 35(9): 260-261.

Udine, E. J ., and J. S. Pinckney. 1940. Some egg parasites of Oecanthus quadrlpunctatus Beut. and of a species of Orchelimum. Proc. Pa. Acad. Sci. 14: 81-84.

Urquhart, F. A. 1941a. An annotated list of the crickets and grass­ hoppers (Orthoptera: Saltatoria) of Prince Edward County, Ontario. Univ. Toronto Studies, Biol. Series 48: 116-119.

1941b. An ecological study of the Saltatoria of Point Pelee, Ontario. Univ. Toronto Studies, Biol. Series 50: 1-91. 71 f.

Wakeland, Claude. 1927. The snowy tree cricket, its injury to prunes and methods of combating it. Univ. Idaho Agr. Exp. Sta. Bull. 155: 1-29. 19 f.

Walker, E. M. 1904. The crickets of Ontario. Canad. Ent. 36: 142-144, 181-188, 249-255. 1 pi.

1910. The Orthoptera of western Canada. Can. Ent. 42 : 269-27 6, 293-300, 333-340, 350-356. 3 f.

Walker, Francis. 1869. Catalogue of the specimens of Dermaptera Saltatoria and supplement to the Blattariae in the collection of the British Museum. British Museum: London. 224 p.

Walsh, B. D. 1866. Answers to correspondents. Marion Hobart, 111. Pract. Ent. 1(12): 126.

1867. Habits of the tree-cricket (Oecanthus niveus). Pract. Ent. 2(5): 54. 2 f.

Williams, M illard. 1945. The directional sound waves of Oecanthus nigricomis argentinus, or a violinist listens to an insect. E n t. News 56(l): 1 -4 . AUTOBIOGRAPHY

I, Thomas Jefferson Walker, Jr., was bom in Dyer County,

‘Tennessee, July 24, 1931. I received my precollege education in the qoublic sclxools of Dyersburg, Tennessee. I attended the University of

Tennessee a t Knoxville for my undergraduate training and received the

Bachelor o f Arts degree in 1953. With the aid of a National Science

ZFoundation fellowship, I began graduate work in entomology at Ohio

State University and was awarded the degree of Master of Science in

1954. W hile completing the requirements for the degree Doctor of

Philosophy, I held in succession the positions of Graduate Assistant,

A ssistant, and Research Assistant in the Department of Zoology and

Entomology, Ohio S tate University.

280