Comparative study of the acoustical behavior of Phaneropterinae (Orthoptera, Tettigoniidae)

Material Information

Comparative study of the acoustical behavior of Phaneropterinae (Orthoptera, Tettigoniidae)
Spooner, John Dewey, 1935- ( Dissertant )
Walker, Thomas J. ( Thesis advisor )
Hetrick, Lawrence A. ( Reviewer )
Carr, Archie ( Reviewer )
Monk, Carl D. ( Reviewer )
Creighton, John T. ( Reviewer )
Place of Publication:
Gainesville, Fla.
University of Florida
Publication Date:
Copyright Date:
Physical Description:
viii, 105 leaves : ill. ; 28 cm.


Subjects / Keywords:
Broadcasting industry ( jstor )
Buzzes ( jstor )
Female animals ( jstor )
Heart rate ( jstor )
Loudspeakers ( jstor )
Singing ( jstor )
Sound ( jstor )
Sound intensity ( jstor )
Species ( jstor )
Ticks ( jstor )
Dissertations, Academic -- Entomology and Nematology -- UF
Entomology and Nematology thesis Ph. D
Insect sounds ( lcsh )
Orthoptera ( lcsh )
bibliography ( marcgt )
non-fiction ( marcgt )


Much progress has been made recently In describing Orthopteran sounds and in explaining their biological significance. However, some areas remain unexplored. Perhaps the most important of these is the nature and significance of the acoustical behavior of many species of Phaneropterinae. The acoustical behiavior of most species in this subfamily differs from that of almost all crickets and other Tettigonidae (except Conocephalinae) in that they produce more than one type of sound in solitary situations. A solitary situation exists when a singer is out of contact, except in some cases of acoustical contact, with other conspecific individuals. Also the females of several species of Phaneropterinae are known to produce sound which functions in intraspecific communication, a phenomenon with no known parallel in other Tettigoniidae....Gryllluae and Tettigoniidae make sound by rubbing together a file and a scraper at the tegminal bases. Complicatedness of solitary singing may be measured in terms of the tegminal movements involved. In simple singing the singer opens and closes his tegmina in the same manner each time he does so« The result is a series of similar pulses of sound, i.e. a phrase. The number of pulses in a phrase and the pulse repetition rate are usually characteristic of the species involved. Species which produce more than one kind of sound, involving more than one kind of terminal movement, in solitary situations may be said to exhibit complicated singing behavior. I have recordings, some of which were made by other workers, of all 20 species of Phaneropterinae known from Florida. I have made extensive observations of the solitary repertoire of 18 species. Four of these exhibit simple sound production sad the remaining 14 species exhibit varying degrees of complicated singing. Generally, complicated singing by solitary males m of two classes. In one class different kinds of sound are produced at different times and in no fixed sequence. For instance, males of Scudderia texensis produce three strikingly different sounds at different times and in no predictable sequence (Spooner, 1964). In the second class different kinds oi sound are produced consecutively in stereotyped sequences. In this group is Ablycorypha uhleri , which produces the most involved sequence of sounds known for any insect (Alexander, 1960). The song in this species may last 40 seconds or longer and involves gradual and sudden changes in intensity, pulse rate and pulse duration. The fourteen species reported here with complicated singing generally fall into one or the other of these two groups. Two have acoustical behaviors somewhat Intermediate between the two classes. Very little is known about the biological significance of any kind of complicated singing. Some workers (Riley, 1874; Fulton, 1933; Grove, 1939; Alexander, 1960; Spooner, 1964) have observed females of certain species answering certain conspecific male sounds by producing a short lisp or tick. My work (Spooner, 1964) with Scudderia texensis is the only investigation reported which reveals the behavioral significance of sounds of a species with complicated singing. The objectives of this paper are 1) to describe the sounds of several species whose sound production has been heretofore unreported, 2) to present the results of numerous experiments with seven species investigating the behavioral significance of their sounds, and 3) to suggest how complicated singing could have evolved.
Thesis (Ph. D.)--University of Florida, 1964.
Includes bibliographical references (leaves 102-104).
General Note:
General Note:
Statement of Responsibility:
by John Dewey Spooner.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
000432019 ( AlephBibNum )
37441066 ( OCLC )
ACJ1523 ( NOTIS )


This item has the following downloads:

Full Text






August, 1964


Several persons are due thanks for their assistance during the research

reported in this dissertation and in the preparation of the dissertation. Par-

ticular gratitude is due Dr. Thomas J. Walker, my Supervisory Chairman,

for his encouragement, advice and criticisms. Thanks are given to the

several persons who brought me katydids which they collected in the field

and which were used in some experiments. Appreciation is extended to Dr.

Lawrence A. Hetrick, Department of Entomology; Dr. Archie Carr, Depart-

ment of Biology; Dr. Carl D. Monk, Department of Botany; and Dr. John

T. Creighton, Department of Entomology, who served as members of the

Supervisory Committee.

Grateful thanks are extended to my wife, Joyce, for patience during

the research and for typing the dissertation.

ACKNOW LEDGMENTS ............................................. If

LIST OF TABLES ..................................... ... ......... v

LIST OF ILLUSTRATIONS..........................................vil


METHODS AND MATERIALS ..................................... ....4

Species Involved In Experiments................................ 15
Inal cdderla stiata ......... ...... ... ..... .......... .. 15
Microoentrarn ilombifollum .............................. 23
MNwteguminw ~noda ................................. 26
ScJadderia ceata sad uealdderia f gr~ a .................. 30
Amnhlyvcorvj a ti rldaa ................... .............. 40
Amblvcorrvna obl~eit[oUia ................................ 48

Species Not Involve J in -xpernments ................. ............52
Anmblvcorvpha cair ........................... ........ 52
Amblvdorrha rotagdifoli ............................... 55
Amblvcoryphla uhle]r an.] Amblveorva nwer bhlier ........... 58
Aratn rias.isait' ......... ............................ 01
lnsp u.iderira walicar ........ ........ ................. .... 63
ILUcroceatrurm retiduory .................................. 65
PlixM PLg A ...................... ................... 66
Scudloia curv.-iaiat laicauda ............................. 67
Stilisaochl.ora coul~ n aa. ................................... 63
TurMilla rosra ....i ..................................... 70

DISCUSlION A rD CO .CLUEIONS... ..... ........................... .. 73
Kinds of response to sound stimuli ..............................73
lutensity of response to sound stimuli.......................... 74
Sexual maturation of adults ..................................... 74
Stimulus situation for sound production ......................... 75
Necessity or acoustical interactions In pair formation.............. 76
Specificity of acoustical communication systems .................. 77

Table of Contents (continued) Page

Importance of toothatrike rate ................................ 81
Increase in Intensity during songs ............................. 82
Complicatedness of sound production ............................ 82
Movements involved in pair formation ......................... 85
Evolution of complicated soiun production ..................... 8G

SUMMARY ............. ............... .......... ... ......... ... 95

APPENDIX ......... ....... ....... .... .. ... ......... ...... ...... 97

LITERATURE CITED ...............................................102

BIOGRAPHY ..................................................... 105


Table Page

1 Results of aalysie of lisp-tick squences from three males
of Inscudderta etr ta ........ ......................... 17

2 Results of analysis to determine the timing of answering ticks
by females of laseadderia strigIf ............................ 19

3 Results of analysis of lisp-tick sequenses from three males of
blotezumiRA mo ....................................... 27

4 Results of analysis of senagrams to determine the nming of
answering tieks of females of Montesumina modeta............ 28

5 Results of analysis of pulsed phrases of solitary males of
Scudderia cuMan and S. furcata............................. 33

6 Results of analysis to determine the timing of the female tick
after male limps for ecudderia gonsat mand S. furcata........... 36

7 Results of analysis of solitary songs of Amblycory'ha floridman
recorded at 24.20 C ....... ........... ................ 42

8 Results of analysis to determine the timing of the female tick
after recorded male songs of AmbJ~ o B floridia ......... 43

9 Results of experiments to determine which part of the male
sound sequence is Lmportant in evoking the tick reepemse from
females of Amblycorypha florldja ......................... 44

10 Results of analysis of the songs of some males of Amblycorypha
carinata................................................... 54

11 The results of the analysis of the songs of the "fast" and "slew"
clicker forms of AeFnbhr yaa ro~ tiai .......................... 57

12 Results of analysis of t thewo songs of Arethoua pLlangum...... 62

List of Tables (continued)



13 Re3ults of analysts of songa of lascudderin valkeri ............. 64

14 Results of analysis of songs of three males of Mlcrocontrum
retinerve ....................................................... 66

15 Results of analysis of songs of Turpila rostrata ............. 71


Figure Page

1. Buzzes of three Amblycorypha flordana males showing
variations In frequency spectrums ........................... 98

2. Single lisp of five species of Phaneropterinae ................. 98

3. Single click of Inscudderia str ata .......................... 98

4. Short lisp and long lisp of Montezumina modest with
answering female ticks.................................... 08

5. Pulsed phrase of Scudderia cuneata ......................... 99

G. Pulsed phrase of Scud.ieria furcali ..........................93

7. Many-pulsed phrase of Scudderia fjr,... ..................... 99

8. Two clicks and a buzz of Amblycoryphz i- l P'ana............. 99

9. Single phrase of Amblycorypha o'lo.nlfolia ilth answering
female tick ............................................... 99

10. Single phrase of Amblycoryplh cariata .................... 99

11. Diagram of typical complete song pattern of the 'fast clicker"
Amblycorypha rotuindifolia ................................. 100

12. Eight basic pulse groups of fast clicker Amblycorypha
rotundifolia............................................. 100

13. Two basic pulse groups of "slowelfocer' Amblycorypha
rotundifolla ........................................... .100

1i1. Tick-lsp? song of Arethasa Rphalagim .................... 100

15. Pair of paired tlcis of Phrixa maya .........................101

16. Single lisp of Stillnochlora coulonla v ....................... 101

Uist of Illustrations (continLed)

Figure Page

17. Lisp-tick phrase of Turmilia rostrata .........................101

18. One phrase of ticking song of Tupilia rostrat ................ 101

19. Part of a lirping song of Turptla rostrata .................... 101


MueI progress has been made recently in deseriting Qrthoptram sounds

and in explaining their biological slgniicaace. However, smem areas remain

unexplored. Perhaps the most important of these is the nature and significance

of the acoustical behavior of many species of Phaneroptertnae. The acoustical

behavior of most species In this subfamily differs from that of almost all

crickets and other Tettigoall ai. (except Conocephalinae) in that they produce

more than one type of sound in solitary situations. A solitary situation exits

when a singer sl out of contact, secept in some cases of aceastical contact, with

other conapecific ianlviduals. Also the females of several species of Phenerop-

terlnae are known to produce sound which functions in Intraspecleic communication,

a phenomenon with no known parallel in other Tettigonlidae. (See Alexaoaer,

1960. for a comprehensive review of pound communication systems In Orthoptera.)

Gryllldae and Tettigenldee make sonim by rubbing together a file and a

scraer at the tognilnal bases. Complicatedness of solitary saiging may be

measured tn terms of the tegminal movements involved. In simple signing the

singer opens and closes his tegmina in the same manner each time he does so.

The result is a series of similar pulses of souan, I.e. a phrase. The number

of pulses in a phrase anJ the pulse repetition rate are usually characteristic of

the species involved. Species which produce more thn one kind of sound, In-

volving more than one lind of tegminal movement, in solitary situ-atiem may be

said to exhibit complicated singing behavior.



I have recordings, some of which were made by other workers, of all 20

speckls of Phaneropterinan Inown io lroFloi a. i have made extelaivo ouser-

vations of the solitary reprtoir of 18 species. ozor of these exhibit simple

sound praJuctioa aid the remaining 14 .jecion exhibit varying degrees of

complicated singing.

Generally, complicated singing by solitary males is of two classes. In

one class different kinds of sound ara proauc-J at different times -,nd in no fixed

seqcOe. For instance, malso of Scud.jeria t:c:nsis produce three strliungly

different sounds at iJicereat times and in no prJlictable acqlence (Spooner, 1964).

In the second class different kinds of couinJ are pro!luc!d consecutively in stereo-

typed sequences. Li this gro..3 is Amblycoryrha ahlcri, which produces the most

involved sequence of sounds know for Way iect (Alaxawkr, 1960). The ncong

of this species may last 40 seconds or longer a J involves gradual and sudden

chage In intensity, pulke raLo and pulse dlaiation. The fourteen species

reported here with corplicateJ singing generally fall into one or the other of

these two groups. Two havo acoustical behaviors somewhat intore-i:Aato bo-

tween th two classes.

Very little is klown, about the biological significance of any of co.m-

plicated singing. Somo workers (Riley, 1874; Fulton, 1933; Grove, 1959;

Alexanr, 19C0; Spooner, 1904) have observed females of certain species

answoring certain coTspccific male by producing a short lisp or tick.

My work (Spooaer, 1304) with Scujdena tcxeusls Is the only iavestigaUon reported

wlich reveals the behavioral significance or somuns of a species with cucplicated

singing. The objectives of this paper are 1) to describe the sounds of several


species whose sound production has ben heretofore unreported, 2) to present

the results of numerous experimats with seven aspects invetlgatLag the be-

havioral sipnflcance of their soe ds, and 3) to suggest bow compliosted singing

could have evolved.


Observations were made in the field to determine the acoustical behavior

of the different species in natural situations. These observations were com-

pared with those made in the laboratory.

The individual katydids used in this investigation were collected in the field

as late instar nymphs or as adults. The adults were caged individually in cubical,

screened cages, four inches on a side and with metal bottoms. The nymphs were

caged together by species in 12-inch x 12-inch x 16-inch screened cages and

allowed to mature. When the nynlph3 transformed to adults, each individual was

enoed separately as noted above for collected adults. In all cases the katydide

were fed dog biscuit (Purina Dog Chow, Ralston Purina Company, St. Louis,

Missouri), water and occasionally some lettuce. Since Inscudderia wall:eri

feeds almost exclusively on the foliage of pond cypress, Taxodiumn .isticum

nutans (Ait.) Sweet, a few sprigs of pond cypress were fed to this species daily

ia a idltion to dog biscuit and water. Likewise, Inscudderia strigat was fed

sprigs of Hypericum fasciuelatumr Lam. With tho exception of Turpilla rostrata

and Arcthaea phalanfium all the other species of ;atydids studied are apparently

general feeders and lived on the dog biscuit and water diet for long periods. T.

rostrata apparently feeds on mangroe, which does not grow near Gainesville,

Florida,the investigation site. T. rostrata always died after a few days In the


laboratory. Nothing Ls known about the food habits of A AbL. MflS g ldividuatl

are not very common around GO esrville, and are collecetd oly by accldet

whbe sweeping Is deae in relatively open, dry, weedy areas. Iadivlduals of

A L. uad die within a few days In the laboratory.

All eaged indtvidale were kept in an alr-conditioed laboratory la which

the temperature was maintained at about gS0 C. On certain days the temperature

fltetated very slowly from about 340 C to aboet 270 C.

Lights were kept buraing contiLously so that Indlvidual katydids could be

placed in darlmess at say time to reeord thikr seads. Thee katydide are moct-

ly nocturnal oingers and often they could be Induced to sing by this maneaver.

Continuous light did not seem to inhibit the aceautieal behavior of any species

for more than a couple of days. Light has no effect upon the nature of the aeund

produced, but may well determine whieh type of soumd la produced or whether

somud is prodmoed at all (personal obeervatlo and Walker, 1963). For Instance,

the characteristics of the fast-pulsed song or slow-puled song of Soadderia

tremMas are not altered by either light or darkness, but light Iteasity deem

determine to some degree which SOng Il produced in natural sititMens (pooner,


Field recordings were made sing either a Mag"emite 610E (Amplifier

Corporation of America. New York, New York) or a naegra HI PH (KUadl kl,

PadeK-LasManns, Switzerland) porMble tape recorder and a microphae center-

ed in a 24-inch parabolic reflector. Labratory recording were made uing the

Nagra III PH recorder or an Ampex 361-P tape recorder (Ampse Corporation,

Redwood City, California). In all cases a dynamic micrephoe (Model DP3A,

American Microphone Compny, Bweheu, MlehiFa, or Model MI-404I-E,

Type 88A, Radio Corporation of America, Camden, New Jersey) and low-print

tape (Scotch No. 131, Minnesota Mining and Manufacturing Company, St. Paul,

Minnesota) at 15 inches per second were used. Tape speeds were checked

periodically and varied less than t per cent throughout the investigation. In

the field, temperature was measured Immediately after each recording with a

mercury thermometer held as near the singer as practical. In the case of

Stilpoochlora couloninna which cings from treetops in hardwood forests, this

was sometimes as much as 100 feet away. The actual temperature in which the

insect sang could have been several degrees different from the measured temp-

erature. In the laboratory, temperature was measured with a mercury ther-

mometer immediately after each recording and usually within a few inches of

the singer. A few temperature readings measured three to four feet away from

the singer were taken as valid since the air was continuously circulated withLn

the laboratory and thermometers in different positions in the laboratory showed

insignificant variation after calibration and correction. Fringe and Frings (1962),

Walker (1962) and Spooner (1964) show the effect of temperature on the nature of

the sounds produced. At higher temperatures, terminal movements are faster.

The sounds of individual katydids were recorded in the laboratory whenever

individuals sang. It was necessary to place some individuals in low intensity

light and others in darkness to induce them to sing. Whatever the situation

the microphone was held close to the singer's cage and the input level of the

recorder was adjusted so that the VU meter read between -10 and -7. Acoustical

interactions between idliaiduals were recorded by placing their cages close

together in froat of the microphone. Several recordings were made of females

answertag recorded male samds.

The. ageds of each species were analyzed by making audapeetrographs

(soongrams) with a IKy Sana-Graph (Kay Electric Company. Pine Brook,

New Jersey). The Sona-Graph used will analyze frequencies from about 100

cps to 9500 cpa. Beeamse the saeuds of Phaneropterinme contain frequencies

greater thma 9S9 eps, recordings were played at one-half speed into the Sena-

Graph, reduseig the freqencies Il the recordings of the natural smetnd by ene-

half, I.e. to a range which could Ie graphed by the Soaa-Graph. The struteral

unit of the sounds of these katydids, the pulse, graphs as a verteal bar, the

width correspading to the duration of the pulse and the height corresponding to

the rage of freqaeacies present In the sound (see asagrams displayed in Figures


At least me s ram was made of each kind of sound recorded from each

species. Certain species produce a single-pulsed liRp as a characteristic sound.

Saes the lisp duration is Imparta t in eliciting species-specific responses, ten

soaagramn were made of the first ten lisps of each recording. All lisps were

graphed if less than ten had been recorded. In cases in which two or more

different resordgs of the same ladivtidal were available more than ten sea-

grame of the liHp of that individual were made. When possible, ten oaeagrams

of the male-female acoustical sequenee were made for each female. In specie

which have characteristic pulse rate in certain sounds at leat two soaagrams

were made of each recording at timings of five seconds ad ten second from

the beg aling of the recording.


Thn soniarams were analyei with respect to lisp .aratioas, pulse rates,

pulses per phrase, frequency spectrums, etc. Time was mc.isurc- in inches with

a Bruning No. 21-18P scale (Charles Eru-ing Company,Inc ,"ew York, Newv York)

estimatci to the nearest 0.01 inch and. converted to seconds by multiplying by

0.0976 scconds per inch, the speed of the Sona-Graph dram surface. This Is

Zescntially the same as cstirinlting to the nearest 0.001 second. Variations In

methodl of recorina and analysis are presented under the discussions of indi-

vidual species. The frequency spectrums were determined by comparing sona-

grams of the sount:s wiith sonagrams of pure frequencies from a Hewlett-Packard

Diodel 201C (Hewlett-T'-a!;Lr.J Company, Palo Alto, California) audio oscillator.

This method has certain limitations. For instance, tie response of the micro-

phones ued to record the nonmd decreased rapidly to freqJuncies above 15,000 cps

(manufacturer's specifications and our own calibration). The sound of many

of the species cc'i cts-ed herein contain frcquencles well above 15,000 cps so

that the comparison of relative intensities of freqIrMncles displayed by the oons-

grams is not vallJ. No doubt much higher Intensities of frequcncices from 15,000

to 20,000 cpa are present in most sounds than are indicated throughout the figures

shown in this paper. Another limitation lies in the inability of the Sona-Graph

to graph freqr:enciss higher than 19,000 cps at its normal drum speeds while

using convenient tape rccordor speeJ3. Certain sounds undoubtcJly have cub'stan-

tial Ittontitlcs of sounm at about 10,000 cpa, as evidenced by the abrupt tcrmina-

tion of any markings at 19,000 cps when graphing certain sounds. See, for

example, the tick-lisp rong of Aretlhea p!npalng u (Fig. 14).

Cne gets a stror.gly biased idea of the frequency spectrum of a song when


one looks at a single sonagrarn of one phrase of a song. There are differences

In dominant frequencies In the songs of different individuals of the same species.

Sometimes within a single song there are changes in dominant frequencies from

one phrase to the next in a sequence of closely spaced phrases and often even

between successive pulses within one phrase. Figure 1 Illustrates nicely such

differences between individuals and differences between successive phrase

within a single song.

A study of sonagrams pictured by Alexander (1960) Indicates that he may

have had difficulty in interpreting the freqauecy spectrum of the sounds of cer-

tain species. Perhaps his equipment was inadequate to handle frequeneles char-

acteristic of some tettigonild sounds. Generally, the frequencies shown in the

sonagrams he displays are low, in comparison to my own, and some seem to be

completely erroneous. For instance, his sonagrame of the song of Amblyeorypha

uhleri indicate: strongly dominant frequencies from 4000 to 7000 cps with almost

no frequencies above 7000 cps. My experience with A. uhlri is that the most

dominant frequencies of that species' song range from 8000 to 14,000 cps with

a spread of less-lntense frequencies below and above the dominant range. Other

of Alexander's sonagrams show similar discrpeancles, but to a lesser enteet..

Thus, one should use caution in Interpreting the frequency spectrum presented

in any single sonogram of a sound. If the whole range of frequencies displayed

is considered, a better idea of what frequencies really may be present In the

sound will be obtained.

To determine the funtton of the various souade made in solitary situations,

copies of recorded natural seeds were played to individually caged, virgin females


and to males of diffcrlin age and experience. Virgin females were used because

females may not be responsive to the sounds of conspecific males once they have

copulated (Spooner, 1964), so much time might have been lost by working with

females of unknown age and experience. Males apparently copulate more than

once because they resume their acoustical activities some time after copulation

(Grove, 1959, and personal observation).

In studies of responses to broadcast sounds the response arena, cylindrical

cage illustrated by Spooner (1964), was used. The response arena had a half-

inch plywood frame with an inside diameter of 42 inches. Tie entire inside sur-

face was covered with tightly drawn bronco wire scrcealng. The distance be-

tween top and bottom screens was four inches. The top screen was easily re-

movable for the introduction or removal of test individuals. Sixteen equal

sections were delineated by strings attached beneath the bottom screen. The

four corners of the original four-foot-square piece of plywood, from which the

bottom of the arena was made, were left intact to serve as loudspeaker supports.

Single kinds of sound or combinations of different kinds of sound were

broadcast to test Individuals using the playback system of the Ampex 351-P

recorder, a Krohn-Ilte Model 310AB Band-Pass Filter (Krohn-l tc Company,

Cambridge, Massachusetts), an Eico HF-14 amplifier (Electronic Instruments

Company, anc., Long Island City, New York), and a University Model T-202

loudspeaker (tweeter University Loudspeakers, White Plains, .Now Yori:)

which had been modified by removing the sphere in front of the diaphram. The

band-pass filter was set to filter out all frequencies below 5000 cps, the range

including most extraneous noises In the recorltngs, and to pass all (requencies


above 5000 cps. The sounds broadcast were copies of original recordings of

natural sounds made at the same temperature as that maintained in the laboratory.

Continuous-play loops were made, so that the same sound was repeated at pre-

determined intervals. Some of these same loops were broadcast to virgin females

to record the sequence of male sound and answering female ticks. Sonagrams

of the copied sounds were indistinguishable from sonagrams of original record-

Ings. The intensity of the sounds broadcast was measured by supporting the

loudspeaker vertically G.6 inches above the microphone (Type 98B99, General

Radio Company. West Concord, Massachusetts) of a sound level meter (Type

1551-B, General Radio Company indicates the sound pressure level at Its

microphone in terms of a standard reference level of 0. 0002 mlcrobars at

1,000 cps) set on the "A" weighting. Because Spooner (1904) found that the fe-

male of Scudderta takmnsl responds differently to one conspecific male sound

depending on the intensity at which she receives It, three levels of intensity

were broadcast to test individuals. The highest intensity broadcast was deter-

mined from singing males by inserting a three-wire cord between the micro-

phone and the sound level meter and holding the microphone about two inches

dorsal to a singing male. Because readings this obtained were found to be

characteristically 5 decibels (db) lower than measurements of the same sounds

when the microphone was connected directly to the sound level meter, I added 5 db

to each measurement of the intensity of sounds produced by singing males. The

intermediate intensity broadcast-in some cases the lowest intensity -was 50

db. The laboratory had a standing low-frequency noise level of 48 db, so the

50 db readings may have been somewhat in error. Nevertheless, results should


be comparable because all saund level measurements of sounds broadcast were

made in the same manner at the same spot. The lowest Intensity broadcast

was not measurable with the soun' level meter anl just loud enough to be dis-

tinct about five feet away.

All of the experiments were conducted in a small laboratory, 8.5 feet x

11.0 feet, adjoining the large laboratory in which most of the recording was

done. The temperature throughout the two rooms wan generally uniform. Test

individuals could be introduced into the arena and tested after a short aijust-

ment period usually 10 minutes, but sometimes longer. To allow enough

light to track test individuals, during each test a Westinghouse 7 1/2-watt red

light bulb was burneJ in a white, porcelain receptacle on the floor beneath the

center of the arena. During each test I sat behind a writing sten anJ noteJ the

position of the teat individual for the entire test perioJ. A 7 1/2-wAtt reJ light

illuminated the writing stand but was completely shielded from the arena.

Each test consisted of fivo minutes of silence followed by five minutes of

broadcast sound (except in some special cases which are explained later). Each

test was repeated at least four times, i.e. the same loop was broadcast to the

same Individual during four test periods. For each repetition the speaker

position was changed to a different corner of the arena. Repllcations consisted

of playing tha same loop to different individuals, so in some cases only two

replications were possible, e.g. only two virgin females were available. In

other cases four replications were possible. The only females used in any tests

were those which gave positive reactions-tliced -to the female ticl-lnducing

sound of the species conerned. Males used In the tests were those which sang

readily in the laboratory.

Thbse general procedures were followed during the entire course of the

Investigation. Deviations from the above outline were necessary at times, and

such deviations will be noted uader the discussitoa of the individual species.

The original recordings made during ths investigation can be obtained from

the Library of Insect Boedae, Department of Entomology, University of Florida,

Gainesville, Florida.



The following is a species-by-species account of observations of the sing-

ing behaviors, descriptions of the physical characteristics of the sounds, and

the results of numerous experiments to determine the function of the sounds.

I have experimental data for only the first seven species. For one reason or

another for instance, no responsive females were available, or individuals

of certain species would not sing in the laboratory no experiments were

conducted on the remaining species, but possible functions of their sounds are

discussed later.

For an account of ecological situations and geographical distributions of

most of the species, see Alexander (1956). Species not discussed there will

be briefly discussed here.

Certain terms are used in this paper in describing certain kinds of sound.

These terms are largely subjective but reflect differences in the method of

moving the stridulatory apparatus. It was pointed out earlier that the basic

unit of sound of these katyJids is a pulse of sound which corresponds to a

single stroke of the tegmlna. The songs of the katydids described herein contain


pulse groups of varying durations and varying pulse repetition rates. A group

of several pulses delivered nl rapid succession is called a phrase. Phrases

have pulse rates of several pulses per second. A group of pulses that are

delivered slowly generally more slowly than one per second is not con-

sidered a phrase. In this case the individual pulses are functional information

carrying units at least in the species investigated. In this latter case pulses

are of two Lnds. Cae kind is called a tick. A tick i iinslantnAneous arn in'nlves

striking only a few teeth of the stridulatory file at a fast rate (1-10 toothetrikes -

usually 1-3). The second kind of pulse which may be delivered at a very slow

rate is called a lisp and involves striking a larger number of teeth over a great-

er Interval of time, the interval of time pulse duration being species-specific.

Another kind of sound not fitting into any of the above categories is called a clici.

Clicks are usually 2-pulsed sounds, the two pulses being tick-like and different from

each other with respect to either intensity or duration, or both. The meanings of

other terms used In the text should be self-explanatory.

Species InvolveJ in Experiments

Inscadderia strlgta (Scudder)

Adults of Inscudderia strigata start appearing about the second week of

July in Alachua County. Florida and shortly thereafter may be collected in large

numbers from the tope of Ifypericum fasciculatum bushes. They are seldom

found elsewhere.

Three distinctly different kinds of sound are made by solitary males of

I. trigata, none of which has previously been described. The two semads

commonly heard from solitary males are lisps (Fig. 2a) and ticks which are

usually alternated in each acoustical performance. Lisps, delivered 1.3 2.2

sec. apart, variable throughout, are alternated with 1 7 ticks usually 5 6.

The number of ticks is loosely correlated to the time interval between successive

lisps. The lengths of the series vary; the recordings on hand contain 12 33 lisps.

The intensity of the sounds and the lisp rate in each series increases slightly

during the first 2 3 lispe. The number of ticks is often 1 2 initially; 1 2

ticks are added each time until the singer produces the characteristic 5 6.

Often the ticks appear in pairs, but the tick rate is seldom constant. The series

is usually terminated with about a dozen ticks.

Seldom is a male of this species completely isolated from other conspecific

individuals. Their host plant often grows in isolated patches. hen strigata

has been found in one of these patches, it usually Is abundant, as many as four

or five having been collected in areas as small as a three-foot square. I have

observed the acoustical behavior of such natural congregations, where as many

as 50 Individuals may have been involved, on more than ten different nights and

on three different days in mid-morning. The .~oustlcal activity appears to be

the same whether in daylight or darkness. Sound production in congregations

differs from solitary singing in that singing males interact. When one male

starts a series of lisps and ticks, others for several feet around "join-in"

during the ticking with their own tic s, so that there is an almost regular al-

ternation of lisps by one individual and ticks by many. 'hen the lisper reaches

the end of his series, another male usually begins lisping immediately. The

result is that sometimes there is almost continuous lisping and ticking for long


periods (not timed I have been at locations for over an hour where large

numbers sang with only occasional pauses of a few seconds). It seems that

more coatinaeus occurs whea large numbers of individuals are con-

gregateJ. Since tha nymphal stages are as congregated as the adults, sound

does not seem to be a congregating mechanism for strigata, except possibly

In the case when females are acoustically active (disenused later). Table 1

shows the results of the analysis of sonagrams of lisp-tick sequences from

three males. The overall average lisp duration is 77 msec. By reducing the

Table 1. Results of analysis of lisp-tiLc sequences from three males
of Inecadderta strA (Time In mllliseonde.)

No. lisps Li(p duration Delay till first tick
Indlv. C analyzed x x Range X

021-2 26.0 10 65 3 483-500 464
021-2 25.8 10 73 3 521-719 600
021-7 26.5 10 84 6 882-721 599
021-10 25.8 10 77 6 392-800 524

tape speed and, when necessary, by playing sounds into the Sona-Graph while

the Sona-Graph drum turned at reproduce speeo, it was possible to sprBad

pulses of sound sufficiently to count the marks which apparently corresponded

to the number of teeth oi the strlalatory file strjc:; In producing the pulses.

Six lisps of individual 021-2 (eee Table 1) analyzed In this manner averaged 48.5

(range = 45 55) teeth utruek. Eight ticks from the same e nle averaged 7.5

ragee 4 10) teeth struck. The two other individuals truck 1 5 teeth per tick.

The least commonly produced sound heard only oace in the field, at

night, an] only three or four times in the laboratory, in darkness is a low


Intensity, two-pul, eid r7,c. (F:g. 3). This sound is repeated in series and is

produced at tines of relative acoustical inactivity that is, periods when

males sing only occasionally. Such periods are few; the males of this species

are noisy almost continuously after becoming sexually mature. Only one series

consisting of 18 clicks was tape recorded. It was made in the laboratory at

25.50 C. The clicks in tils recording varied from 1.5 2.2 see apart and

averaged 1.7 sec. apart. The second pulse of each click is much more intense

than the first pulse, and the tooth strike rate of the second is greater than that

of the first. Only three teeth are struck in the second pulse, whereas 3 5 are

struck in the first pulse. The delivery rate of the two pulses within a click for

the one recording averages 9.0 pulses per sec. with a standard deviation of 0.2

pulses per sec. In calculating the pulse rate, time was measured from the be-

ginning of the first pulse to the beginning of the second pulse.

The functions of these sounds were not readily revealed by field observa-

tions so a number of individuals were placed about in the laboratory for obser-

vations. The observed laboratory acoustical behavior of males conformed to

that seen in the field except wn virgin females were responsive. Virgin fe-

males answered the lisps with a tick immediately after each lisp. The ticking

of the males became very erratic and more intense when a female was answer-

ing. Ticls from females were emitted very shortly after each lisp when no

males were ticking. Females also answered recorded lisps with or without thz

alternating male ticks. Table 2 shows the lick delay timlu, of three females.

In no case was the female tic!k delayed until the shortest timing of a male tick.


Table 2. Results of analysis to determine the timing of answering ticks by
females of Jeudderia strigata (Time in milliseasnd.)

Souree of Indiv. No. of repoaes Tick delay
isp mansered female oC analyse Rage ax

Reeordtag 021-8 6. 8 9 80-137 105 90
S021-8 88.0 8 61-119 93 19
S021-12 25.8 10 48-117 73 21
021-13 24.8 10 103-147 132 13

Sgling Male CS1-8 236. 1 112
S021-12 28.0 10 61-116 80 18
P 021-13 29.0 1 119

A riss of ieperlmmesa w designs to aseseta the signlficance of the

differat ounds. ContLnuous-ply loops were made of a liip, a limp with ticks,

ticks alone e sped as they are in normal ster f E ad a click. By using different

length. of leader tape to speed the sends the rat of repetition of the broadoaet

somde was controlled. The lisp nas broadcast mea ewry 1.5 ec. as was the

series of ticks inee this wa aliset the medal ad mam rate of at&ernattle by

malen. Clicks were breedctet at rat of moe evry 1.7 sec. The latedaty of

the seuds of five males had bem previously meaared at 60, 03, 84, 6G ad

06 db, so each aomd ws played at S0 aMd 66 db. TM lips were played at a

barely awlible imntmety also (se method for Intensity meauuremet).

The main function of the lisp appears to be the s~tmelattei of females to tick.

Three maleM were tested In the eserimental arena, and all tieked after the lIap

in whatever arrangement or Intensity the lisp w broideast. Females may move

toward lips at tiam In artre beecae Mfth of the e.partim~ e het ales oriented

toward the leadspeker on at least one oaeasten whn lieps were turned on not

all to the saae iStlansty of sound and two of them moved toward bhe leadepakbr


a short llstCance on one occ.lson each. Theso rmoemonto occurred. at fL'e bcgn-

ning of test series, the females having been in darcnss for at least ten minutes

witho::t hanvng hearJ any malo sounds. During sacceedlng tests the females either

re ain motionless or twaled in a random pattern, always t1 g after lisps.

The ticks and clic' ha no visible effect on the test females at any In-

testy. These apparently function only in male interactions.

The only reaction giveu- by the five teot males to lisps, or lispa and ticlc,

was ticling at the time they would have ticked Ln natural situations. Either no

movement or random movements were seen. N:o orientation movements were

over seen. Often when the recorded soniads were turned off at the end of a test,

the test male would give a series of lisps and ticks of his own. Very low in-

tensity lisps almost always cause] the test males to start their own lisps ad

ticks during the tests, whereas they only tickLo in alternation with high -iton-

sly lisps. Such coatrasting behavior suggested that high intensity lisps and

ticks may have an Inhibitory eect upon male lisping. Three series of tests -

lisp alone, ticL~ alone, and lisps and tic:s,. all at 65 db were ran to test this

idea. A group of' ix males were placed on a table In the experimre:ntatioa room

and observe as som s were broadcast to them. The males were allowed to

start their own- staSing, and then the recorder was turned on. The result was

the inhibition of male lisp production by the recorded lisps or ticks only when

a pulse of recorded so-ind preceded and overlapped the time the singer woaul

havo begun a lisp. This would be enough to effect the inhibition of l g of all

but one male in a congregation of singing individuals, for in the groups

there was a moial refractory period of 1.5 sec. at 250 C between seccessivo

lisps of a singer. This modal value was usually the minimal value, but not

always. Another male, in order to start lisping, would have had to abandon the

somewhat longer refractory periods between his initial lispe in order to intereede

the already-singing male. The tick rate of the terminal ticks of tM seri ob-

served always decreased. Tbus, there was greater, opportunity for a new singer

to tart a lisp-tick series at the end of another male's series than at any other


The results of tests of the click sound are inconclusive but suggestive.

All but three males had died by the time those tests were beg-n. The clicks

were broadcast only st 55 db (arbitrarily selected). During each of these teats

the males either remained motionless or moved erratically away from the speak-

er a short dtiatece. One male moved away in five out of eight tests; the second

male moved away in one out of five tests, ad the third male moved away In

two out of four tests. Such movements, compared to no movemesae at all during

the silent part of the tests and no movements a opposed to oriented movement

during other tots, eertatl l saggesttat the click sounds function in male spring.

Not emeugh obeervations were made of the dielly cycle of acoustical activity to

rule out the possibility of a particaar time of day ai which clicking is promimeat.

Spooner (1964) found that males of Seuddiria texemis prodaee a low itmeetty

ticking soud only during the evnig twilight. Ticking it 8l. tminte fucticuo

In male pacing by eausiag the males to move Mlaeticilly as long ae tes re-

ceive the ticking shove a certain intensity, or stil ticking is equal in internal

all around.

The female tick attracted males to asmwering females. Became of the

difficulty in proJucing a simulated female tick at the proper time after a male

lisp it was impossible to use the same experimental technique to determine

the function of the female tick. Caged females were place J, one at a time, on

the corners of the arena in place of the loudspeaker. When the test males lisped

the females answered. While continuing to lisp and tick, test males oriented and

went directly to the answering female. To avoid the possibility of a chemical or

visual stimulus causing such orientation of test males, the female was placed

just outside the darkened room where she was out of sight but could still hear

the test male. Her answer thus broadcast at the corner of the arena through

the loudspeaker induced the same kinds of reaction from the test males.

Since in natural situations very few males are lisping at any one time, it

occurred to me that non-lisping males may go to females answering lisping

males. A recorded lisp and answering female tick were broadcast to test males.

No male went to the speaker. A responsive female was placed on the corner of

the arena and lisps were broadcast from a nearby table. The female answered

the recorded lisps, and all four test males went to the female. The test males

usually started their own series of lisps and ticks when the female answered

so that the female answered the test males. The strldulatory file was removed

from one male In order to silence him. H;e still went all the way to the female

on each of several tests when the female answered the recording. When the

female was placed outside the experimentation room and her answer to non-

test males was broadcast to test males, test males still wont to the corner

where the sound was emitted.

One field observation supports the above data. I was standing in the


midst of a group of siaging males one night anev imitated a 6bmale tick by strik-

ing my flagernala together at about the proper delay tinstg after the lisps of

the one lisping male. Ilot only did the lisper orient aid start moving toward me,

bet a number of "bystanders" did also. One male abat three fea ti front of me

almost fell off his perch when he turned suddenly after my first dimalated tick.

Mloreessran rtoebtfolia (Qm rwe)

MleroeanIWA rbrhi n male proeduo two distinct kinds of msamd in

solitary situatiees liaps (Fig. id) and ele. Bth hve bee desoeied by

Allard (J1928), Falan (1932, 1933), Ales der (1966, 190), mad others. The

account by Grove (19M) is the mot oomprbsaelmve o the acoustical activity

of this species. Alesader (1960) shems semnrams of both sengs recorded

at 650 F (18.3C). TW1 spetee is chin y arboreal ad thereby difficult to

coUllt in numbers. Only three lndviduals, one female ad two males, were

Staedld the laboretry.

Beth limps ad tieks ma be heard at any time of day or night, although more

maoodical entivity Is apparent at night. In meet intasneae lispe ad tieks are

isolated accomplishments, bhavng no eartat relatim ae to other. Bat

sometimes a male may be heard to give a couple of liepe in rapid .uceesion and

follow ap with a series of ticks. Seeh behavior ia the eamupteo rather the the

rule ad usually oceurm when a male beasme aosom really native Awr a period

of allenee.

Apparently btb the Ilip aMd ticks ar made mn tim eclMing Aemlnes of the

tgmifa. A single lip (Fig. 3d) imvolver on cloM of the tegmtin at a rapid


rate. Soniagsms of four lisps (laboratory recording, 260 C) from one male

showed 20 24 toothstriv.cs per lisp. The lisps in this recording were produced

2.0 4.1 sec. apart. The modal and mean rate oL delivery was one every 3.0

sec. Ten lisps from the same recording ranged from 22 to 30 meec. duration

with an average of 25 msec. (sx = 3 msec). These figures indicate a toothstriko

rate during the lisp of about 872 per sec. This is quite a contrast to the tooth-

strike rate of the ticking sound in which the in livid-:al pulses of sound (single

ticks) correspond to individual toothstrikes. This phenomenon has been noted

by several of the above authors. I have recorded in the laboratory three tick-

ing series from each of two males. Cne male was recordeJ at 250 C, the othbr

at 25.50 C. The average tick rate in the middle of each of the three series for

the 250 C individual was 8.5 ticis per see. The average tick rate for the 25.50 C

individual was 8.8 ticks per sec. These males produced 22 34 ticks per series.

Grove indicates that males protuco 28 32 ticks In a series, and Alexander (1956)

says a series consists of 15 30 ticks.

* The sound necls no detailed experimentation to resolve its function.

Both Allard (1.23a) nnd Fultor. C033) noted that females produce a low intensity

tick after a series of male ticks and that males go to females when such acaus-

tical interaction takes place. Alexander (1060) describes the acoustical inter-

action between a male ani female caged near each other but out oi sight from

each other. He says the [c-male's response was so precisely timeJ that her

tick seeded almost a part of the male tick series. Grove and Alexander both

noted that after the fermale answering tick, the answered male would often pro-


duce an irregular shuffling sound. Grove conjectured that this may serve to

confuse the location of the female, for he observed that "llttealag" males would

produce the shuffling sound after having heard a male-female sequence, that

those males would often move toward females answering other males, and that

these males may reach a female and copulate with her without having made a

single swund themselves. I eoaducted no experiments with a recorded male-

female acoustical sequence because no recording of the seqance was made

before the female died. The functlen of the lisps has been the subject of much

conjecture. Grove saw that males In oages woald jump about when one male

started a series of lisps, aad on this evidence he postulated that the IHops ea-

hibited a territorial fiuctiom. Certainly if males were this irritable when lisps

are produced, they would tend to move away from the soned. Grove also sagget-

ed that the liap may serve to keep responsive females in the vicinity of a lisp-

ing male, and Alexander (1960) seemed to favor this idea.

I conducted three series of tests of the lisps with the one female I had -

one seri sack at 55, 75, and 90 db. The lisps were broadcast every three

second during the acoustical part of the tests. At 55 db the female went immed-

iately to the loud speaker in all fear tests. At 75 db the female usually turned

toward the loudspeaker, but only went toward it two times and mth only part-

way. At 90 db she did not move. The female made no tik to any lisp at any

Inteaslty. Tb tests indicated that low intesity limps were female attracting.

Two weeks after the above tests I repeated the whole experiment with the same

female and she gave a similar performance 1.e. no reepesee to high- tntensty

Imps but Immediate orientation a d movement toward the loadepeaker at low

intensity. This 13 a reasonable expectation since females of Scudderia toxensis

move toward a certain conepeclfic male sound oly when it is received at low

intensities (Spooner, 1964). This also explains why Grove never saw this function

of the lisps in his caged individuals. Females close by males receive the lisps

at too high intensity to be responsive.

Monto.umina modesta (Brunocr)

Very little is known about Montezumina modcsta. I have collected it in both

a sand-hill community and in a cypress head. It was equally abundant in both

extremes of comnmunity-type. However, I have found it only where there is some

shrub or tree cover. Nothing has been reported previously of its acoustical


Sound production of M. modest is unique among the phaneropterines

stadied in terms of complication of solitary singing. In nature this species sings

primarily in late afternoon and early twilight. In late twilight and darkness only

occasional sounds from males and females can be heard. It is difficult to de-

termine the acoustical activity of a group of these katydids in nature, so I will

describe what I heard from one group (number uniaown) observed aurally on

about ten different days. As the sun began to sink behind the trees, but not be-

low the horizon, large numbers of lisps were prominent and these were answered

by ticks from both males and females. series of lisps from different individuals

varied from about 10 to about 35 In number. Lisps were pace 0.5 1.0 sec.

apart, and the ticks from both males anJ females came an instant after the lisps.

No movement was ever observed from any eirnin Individual, but few individuals


were observed due to their cryptic coloration anJ because o: the way they perch

underneath leaves andl on the [nuer branches of shrubs. Solitary males not only

lisp but most often produce a very low intensity tick immediately after each lisp

in a maaner such that a lisp-tick is suggestive of what sound coulJ be proAdced

in one opening and closing of the togiina. Males responding to limps of other

males may produce single ticks or 3 4 rapidly delivered ticks.

Laboratory observations were instructive. Uy placing caged males and

females on the table in the experimentation room and by leaving the door open

for the only light source, twilight conditions were simulated-, and caged individuals

sang readily. I found that the male. had two characteristic lisps (Fig. 4). In

most lisp series the first few lisps were usually "short" lispu delivered on

the average about two-thirds second apart (stop watch timings) at 250 C. The

terminal lisps of most Usp series were usually "long" lisps delivered about one

second apart at 250 C. Table 3 shows the results ol the analysis of sonagrams

Table 3. Results of analysis of liep-tick sequences from three males of
Molateumlna modeeta. (Time In milliseca ds.)

No. End of lis;" Beginning of
Type lisps SI ratMln to tick lap to tick
of lisp Indlv. oC analysed Itnge T a Reage X Range T

Short 041-9 25.0 20 15-23 20 2 48-98 75 70-118 95
041-10 25.5 10 18-22 20 1 -G-91 81 83-111 101
041-11 25.0 10 13-20 17 2 45-60 51 64- 79 68

Long 041-U 25.0 12 30-39 35 2 37-93 58 73-123 99
041-10 25.5 3 27-30 28 2 56-67 59 86- 85 87
041-11 25.0 10 23-31 23 3 36-96 68 66-119 45


of the lisp-tick sequences of three males. Short and long lisps produced

without the following ticks measured the same lisp durations as those shown

in Table 3. Average lisp durations for the short and long lisps were about 19

msec. and 30 meec. respectively. For those isps analyzed there was only one

case of overlap between the extremes of lisp durations, but this involved two

individuals. Each individual had distinct and non-overlapping ranges of duration

for the two lisps.

Table 4 shows the results of analysis of sonagrams to determine the timing

of the female tick response. If time is measured from the beginnings of the U ss,

Table 4. Results of analysis of sonagrams to determine the timing of
answering ticks of females of Montezumina modest.

"No. of End of lisp Begining of lisp
Type of lisp responses to female tic: to female tick
responded to LIdiv. oC analyzed R 7 R 7 e.

Short 041-7 25.7 1 38 59
041-12 25.0 10 40-50 44 3 58-65 61 2
041-13 25.0 10 37-16 42 3 53-62 58 3

Long 041-7 25.7 2 15-24 20 4 50-54 52 2
041-12 25.0 7 14-?5 18 4 53-65 61 4
041-13 25.0 10 28-35 82 2 61-68 65 2

then the female tick response timing is essentially the same about 60 msec. -

to both types of lisp. Pcjpon3e delays measured from the ends of the lisps are

not the same. This phenomenon suggests to me that the females may only cecond-

artly nsvwer long lisps after having answered a series of short lisps. Inda d, fe-

male responses to long iUs~s were not nearly as vigorous (see discussion of inten-


sity of respease is a later section) as they were to short lisps. Test females

would answer short lisps with strong, loud ticks, and would continue answering

after the male started producing long lisps (see Fig. 4), but they did not answer

all the long limps. Usually the females stopped responding before the enJ of a

series of long lisps, an. If short lisps dlJ not begin a seriesthe females some-

times did not answer at all. Results from experiments with these sounds also

indicate that female ticing alter a long lisp is not of great Importance.

The procedures outlined in the general methods for testing response to

sounds were not ssable with this species without moJification. I was able to

get responses from virgin females to recorded lIaos only after filtering all fre-

quencies below 15, 000 cpe from the sounds (see discussion on frequency differen-

ces). Two females were tested In the arena with lisps filtered in this manner

and broadcast at 50, 55 (the maximum measured from a male), anJ 60 db. I

reasoned that 60 db near the speaker would be 55 db or less at the test individual.

The test females almost invariably aswereJ the short lisps at all three intensi-

ties. They moved very little toward the short lisps, although they often turned

Immediately and oriented towarJ the loudapeaker. This behavior suggested that

the long lisp may have been female attracting, and experience with other katydids

showed that low intensities of the female-attracting sounds are important in

eliciting response. Thus the initial tests with long lisps were at 50 cb. The

result was no reistion at any time. At 55 db both females went immediately

to the loudspeaker if they were near the loudspeaker when the sound was turned

on. If they wore on the opposite side of the arena from the loadspeaker when the

eound was turned. on, they gave no response. At 60 db they always Ownt immedl-


ately to the speaker. During these tests with long lisps only occasional answer-

ing tickls were made by the females. So, females go to high intensity long lisps.

What, then, gets the sexes together from long-range in areas where the popu-

lations are low? I placed a male in the arena, a female in a small cage in place

of the loudspeaker and leit the door open for a twilight effect. The male san'

series after series, and the female answered almost all his lisps. ie heard

her certainly, for he would turn and orient immediately toward the female after

her first answer. But he never moved toward her. By covering the female's

cage with several layers of cellucotton it was possible to muffle her answers

until they were barely audible to me. At this intensity the male went rapidly

to the female at whatever position I placed her. I then alternated (. 4 times

each) placing the muffler over the female and removing it. The melo naver

moved toward the female when she was uncovereJ aaJ he always moved toward

her when she was muffled. Those experiments indicate that it is the males

which are attracted toward distant females in nature, anJ that the females make

the final move to bridge the gap between the sexes by moving toward high in-

tensity long lisps.

Sculierta cuneata Morse
ScuddJern farcata Brunner

Scudderia cuneata and ScudJeria furcata are best discussed together

because of their very close relationship, overlapping geographical ilttribations,

and similar sound productions. Alexanlcr (1960) pointed out the problem of

understanding how hetorocpcciic males and females of cuncata, furcata and


S. fasciata 3eutnnmuller, a third species which overlaps geographically with

furcata and possibly with c-ista maintain reproductive isolation when the

sonj patterns of the males of the three species are apparently identical. He

posed the idea that the timing of the responses of the females may he different.

Both cuneeta and furesta are common in Alachua County, Florida, but for

the most part they occur in different habitats and adults arm preset at different

times of the year. Although both frequent shrubby woods more than completely

open habitats, cuneata is generally present In hydric to mesic situations, and

farcat usually frequents zeric to meic situations. Cuneata has one generation

per year, adults being found from the middle of July intil early November, and

firoata has two generations per year, adults being foun from early May until

November, but with reduced numbers in August prior to the maturation of the

second generation. Thus, only those individuals of the two species found con-

currently in mesic situations and occasionally in more xeric or hydric lit-

natlons have the potaetality for confusion in pair formation. The presentation

below shows that cnmeata mad fureat have sounds sufficiently different to allow

reproductive isolation by specificity of response of males and females to con-

specific somnda.

The acoustical behavior of solitary males of both ounesta and furcata

sl so similar that only a trained ear oan usually detect which species is pro-

ducing which sound in mixed populations. Without the opportunity for compar-

[son, it is almost impossible to distinguish which species Is siaaing when only

one species occurs. Both species produce single-pulsed lisps (Figs. 2b and 2c)

whish they reiterate a few seconds apart in series of three or four. Differet


series are spaced anywhere from 1 30 minutes apart. A second sound pro-

duced by both species but much less often than lisps is a short phrase (Figs. 5

and 6) in which the pulse rate is quite fast but slow enough to aurally detect its

pulsing nature. The pulsed phrase is repeated at a rate of one every 4 5 sec.

to one minute.

By slowing recordings to one-half speed and using a stopwatch, I determined

the lisp rate for the two species at 250 C. The lisp rate for cuneats recordings

varied from one every 1.7 3.0 sec. and averaged. one every 2.3 sec. The

lisp rate for furcata recordings was slower, ranging from one every 2.4 4.2

sec. and averaging one every 3.3 sec. The lisp duration of cmeata lips at 2&o C

determined from 23 recorded lisps involving four males ranged from 12 to 25

msec. anJ averaged 16 msec. with a standard' deviation of 3 msec. The lisp

duration of 18 furcata lisps at 250 C involving four males ranged from 55 to 70

msec. and average 75 meec. with a standard deviation of 9 msec. Thus the

lisp durations are distinct for the two species. I recorded several pulse 1 phrases

from four cuncjta males and a few pulsed phrases Irom two males of furcata.

The results of the analysis of the phrases are shown in Table 5. There are not

enough data to determine whether there is or Is not a difference in pulse rates

between the two species. The average pulse rate for cuneata between 25. ? C

and 25.09 C was 35.0 pulses per sec. whereas the pulse rate for urcata at

250 C was 35.6 pulses per sec. One dificroncc between the pulsed phrases of

the two species may be the phrase duration which is reflected by the number

of pulses in the phrases. Cuneata pro-'uceJ 2 4, almost always four and

never more than four, pulses per phrase. Exporlmental evidence from both

Table 5. Results of analysis of pulsed phrases of solitary males of
Scuddoria cuneata anJ S. furcata. (Time In milliseconds.)

No. phrases PuHie per second
Species Indiv. 0 C sampled X s

cuneata 064-2 25.8 5 36.5 .2
"064-6 25.2 9 32.1 1.4
034-6 27.4 1 37.5
054-7 25.7 3 36.5 .7
064-7 25.9 4 88.7 .4
084-8 28.0 3 42.8 .6
064-12 27.2 2 89.0 .4

furcata 063-6 25.0 2 36.0 .3
S063-9 25.0 4 35.3 .4

species (discussed later) also suggests that phrase duration may be important

In eliciting responses. Other authors (Allard,1910b, Fulton, 1952; Pierce, 1948;

bad Riley, 1874) hasv described souads of more than one pulse from furcata, but

not In detail.

The only difference between day and night acoustical activity appears to

be increased singing at night. Both lisps and pulsed phrase can be beard at

any time of day, but I have no Information whether the pulsed phrase are more

freqgsBt than the liaps at some specific time of day. The descriptions pree@trd

above agree tn general with those of other authors who hare deeoribed furoata

sounds from saral Impresstons, except hoau of Allard (1911) and Cantrall (1948)

who observed more singing from futrAt in the afternoon than at night.

I have heard only lisps and pulsed phrases from solitary males of either

cuneata or furcats. Yet Cantrall (1943), Fulton (1930),ad Plerse (1948) have

described a very low intensity ticking sound from fure ta. Cantrall whrd the


ticks once and says they were barely auliible live feet away. Single 'tisps"

were proJuced every 2.5 see. Fulton observed ticking from males of furcata

in late afternoon in Cre;on, the ticks being omitted 2 3 sec. apart "to a rate

too rapid to count". He addeJ that females occasionally produce a similar but

somewhat fainter sound. Pierce's observation may have been made in either the

laboratory or the field, but I suspect it was in the laboratory for he obtained

"'several records" from (urcata, and this could be done more easily in the lab-

oratory. I have never heard any ticking from solitary males of caneata or fur-

cata, nor from congregatcJ males of either species where no responsive females

were present. However, in the laboratory where virgin females pro:lcedn ans-wer-

ing ticks after the lizpp o: males, both cuieati and furcab males produced ticks

irregularly in the manner described by Fulton for furcata. In the presence of

responsive females rr.le ticks were heard at almost any time except during the

time whoe a lisp was b.'lng ma'!e and the time a female tie!; was expected ImmJei-

ately after a lisp, males usually produced a fow rapidly-delivered ticks and

stopped before the time of the female tick. After the time of tec female tick,

whether a female tickled or not, males ticke.] Irregularly at a slow rate. .'hcn

responsive females were removed from the room males continaed to tick for a few

minutes slowly by themselves, or in response to lisps. After a few mtnuts

with no responsive females around, male ticking always subside Those

"created" situations were alternately repeateJ a number of times, and always

the results were the same; male ticking only In the presence of a responsive

female. Fulton made no mention of the number of diflcrent observations he

made of male Possibly Fulton noticed it only once In nature as did


Cantrall. 11 so, then It is possible that the presence of responsive females

may have precipitated the ticling that Fulton Lad Cantrall heard. If fierce

actually did his recording in the laboratory, then he may possibly have had a

responsive female present, thue simulating a situation like that in my lab-

oratory. Of course, it is possible that fircata from other locations may pro-

duce a ticking sound In solitary situations similar to that of Scudderia texeasis

(Spooner, 1984).

As mentioned above, cuneata or Cureata males were almost always silent

at the time when a female answering tick was expected. This agrees with the

behavior of Scudderia tmenuA in which species the males produce a "slow-

pulsed song" which the females answer at about a one-seco d delay at 250 C

(Speater, 1964). Listening males of texenals produce loud ticks during and

immediately after the slow-pulsed song but are silent at the timing of the

female tick. Occasionally a male of punesat or furcata ams be hear.l to pro-

duce a single tick after his own lisp at about the timing of the female tick. Sona-

graphic analysis shows that the mean delay of the tick after the lisp in a male

lisp-tick sequence Is longer than the mean delay of a female tick after a lisp

(see Table 8). However, the range of delays of male ticks and female ticks

overlap for each species. Six lisp-tick sequences (three each from two males)

of cumeata ranged from 343 to 440 maec. delay before the tick, and averaged

399 msec. with a standard deviation of 3G msec. Twenty-mne lisp-tick sequences

from one furcata male ranged from 1230 to 1544 msec. delay before the tick and

averaged 13G8 meec. with a standard deviation of 58 mesc.

Table 6 shows the differences in timing of the female response to the lisps


Ta~jI 6. Results of analysiS to determine the timing of the female tick
after male lisps for Scud.,oria euncata and S. f reata. (Time in

Source of Isdlv. Sample Tick delay
Species lisp ansv.cred female OC size Rango X sa:

cuneata recording 064-4 25.0 16 332-383 31 15
04-8 25.0 10 325-408 3GS 23
064-10 25.0 10 270-399 325 40

furcata siging male 0G3-5 24.8 10 840-1422 1054 174
063-12 25.0 3 1110-13B4 1235 137
recording 003-14 25.5 10 1059-120G 1147 61

of specific males. In the laboratory, response was charccteritically specific

for most of the adult life of the katydids. The distinctHio difference between

the lisps of the two species is apparently lisp duration. By standing near fe-

males of cuneata and furcata at the same time I have evoked responses from

one or the other species but seldom from both to the same single stimulus -

by flipping my thumb across the corner of a piece of paper at different rates.

The only difference between the "lisps" I thus produced was In lcnghii. Short

"lisps" evoked ticks from canoata females; long lisps" evoked ticks from

farcata females but, at spoctes-spocific timings. Not all such artificial

lisps evoked responses. Some were uncontrollably too short or too long for

either species and consequently no response occurred at all. After having been

kept in the laboratory for a long perloj without being allowed to copulate, females

would sometimes begin answering the lizp. of ot-her species, but even then at

Esecles-speciflc timings and with less vior (intensity of response discussed

later) than the lisps of their own males.

Controlled caperimonts to determine the functions of the somuns of these


two species wera conducted according to the general methods outlined earlier.

A maximum of 58 db was measured from lisping oaneata males, and a maximum

of G2 db was measured from lisping furcata males. Thus, lisps of cmeests were

broadcast at three inteasltles barely audible, 50, and 60 db and, lisps of

fuerats were broadcast at three laetanstlie barely audible, 50, aad 65 db.

Cuneata lisps were broadcast at a rate of one every two seconds: furotalisps:

one every three seconds. The maximum intensity of the pulsed phrases was

measured at 66 db for cuemta! and 78 db for furoata, so these sounds were

broadcast at intensities of barely audible, 50, and 65 db for cumeta; and barely

audible, 55, and S0 db for furoata. Pulsed phrases for each species were broad-

cast at a rate of one every six seconds.

The only consistent response given In the whole series of tests with indl-

viduals of both cuneta, and urcata was female ticidag to conspecific male lisps.

Other responses were so ineoesiatent between suecessive tests with single indi-

viduals and between individuals that no defioate aoncluions can be drawn from

the data. A summary of the data follows. Two of the four cunesta test females

oriented toward and moved toward recorded eounstta l ps in about on-half of

the tested at 50 db. These same two only ticked to the same lisps at lower or

higher intensitles. One of those and another cuneats female always answered

a easmnta four-plsed phrase and went to the loadspesair riLadea tiag the phrase.

The same female which responded in the above two types of tests answered and

went to a cuneata two-pulsed phrase. Also she went tower a six-palsed phrase

of furcatain five of eight tests at 80 db bat never answered at 80 db. At 55 db

this female answered the furoata six-pulsed phrase a few times and went to the

loudspeakor in four of four tests.

Furcata females vcre equally inconsistent. The six test females usually

made no movements during tests other than those involved in producing ticks.

All six did no more than to tick to lisps at low and high intensities, but three

of the six oriented and made strong movements toward the loudspeaker in about

one-half of the tests when lisps were played at 53 cb. Only one furcata female

answered the six-pulsed fureata phrase used in the tests, this at 55 db and

only at two different times. At barely auJible and at 80 db intensities no re-

sponse was made by any furcata female to the six-pulsed phrase. At 55 db,

however, five of the six females oriented toward the loudspeaker immediately

when the sound was turned on and in more than one-half of the tests the females

moved to.vard the speaker. ITo furcata female ever answered the cuneata foir-

pulsed phrase, but one female in two different tests at 55 cd oriented rinune.iately

toward the loudspeaker when the cuneata four-pulsed phrase was turned on.

Cueata and fur.'ata females consistently answered conspecific male lisps

at species-specific timings. Cuncata females consistently answered cuneata

four-pulsed phrases and sometimes answered furcata six-pulseJ phrases, whereas

furcata females seldom answered any pulsed phrase. Cuneata and furcata females

often oriented and .vot toward medium intensity conspecific lisps and pulsed phrases.

CuOeata females were in"riscriminate in respon lng to pulsed phrases of cither

cuneata males or furcata males.

Neither fast nor slow ticking ha1 apparent edlect on any test females of

either species. However, in tests of fast ticking or slow ticli-n test males of

both species almost always btrned away when the sound was turned on and


erratically moved away from the sound. Ticking sounds from males of either

species are apparently alike and both repel heterospecfic males as well as con-

specific males.

Another sound, not yet mestloned and heard from fureata male in sit-

nations where females were answering and several males were ticking loudly,

is a high Intensity, many-pulsed phrase (Fig. 7) which decreases continually

in intensity and palse rate. This sound had no apparent effect on test males and

females of furoata. It may be a mechanism to release "nervous stress" in males

in such a situation, for a male geneally ticks more softly and at a slower rate

after producing such a sound.

It would be advantageous for males that hear females answering other

males to locate those females by going to ticks produced at a timing specific

for that species. Since Grove found that males of Mtcrocentrun rhemblfolium

sometimes locate females which are answering other males, and since I found

that males of Inacuddera striata can locate their females in this manner, I

waist to know if S. cuneta and S. furoata exhibit the same behavior. Three

males of each species were tested. A test consisted of placing a caged female

at the loudspeaker position and broadcasting lisps from nearby. The males of

both species went rapidly to the female of their own species and made so reac-

tion at all to the heterospecific female answer, thus proving that males are

able to locate conspecfic females answering other conspecific males. An

imperative test is to determine experimentally if the timing of the female tick

is important in attracting males to females. Casual observationa slgget that

It Is.

Amblycoryph floridana Rehn aid Hebard

Amblycorypha floridanr is abundant throughout Florida from early June

through July in almost any lushly vegetated area. Solitary males produce a

sequence of several clicks followed by a buz: (Fig. 8), and the sequence is

usually repeated several times (3 25 repetitions) in each series. In this

text I will refer to one sequence of clicks and a buzz as an entire song (ES).

When starting to sing after having been quiet for a while, a male usually clicks

at a slow rate initially and increases the click rate through perhaps a dozen

clicks before producing the buzz. In succeeding ES sequences only four or

five clicks are usually made before each buzz. The time interval between

ES sequences is usually decreased during the first three or four sequences.

At the end of a series of ES eeqJences the interval between buzzes increases.

Usually a series of clicks (perhaps a dozen or more) ends a series.

In normal singing the two clicks just before the buzz are produced in more

rapid succession than are the preceding clicks, and the buzz follows the last

click with almost no break. Thus I have arbitrarily divided the ES sequence

into three parts Part I: the initial 3 4 clicks, given at a rate of 2 3 per

sec., Part II: the tivo clicks just ahead of the buzz and given in more rapiJ

succession than the initial iclics, and Part III: the buzz.

Most often floriJaaa is found in congregations in favorable habitats, and

in such congregations the predominant sound produced is clicking. When several

males are singing cloco together only ona male will be making both clicl.s anJ

buzzes during any one period. The other males usually cllcl" loudly until Lth

one male producing ES seq'jonces stops buzzing. Then anther male will click


and buzz while his neighbors click. Buzzing evidently inhibits sound production

from the males. A male of Atlantlcus glaber Rehn and liebard present in the

laboratory at the same time, would cause floridama males to completely cease

sound production with its characteristic, high intensity, high frequency buzz.

I could often cause the same effect by rally productag a loud hiss. I also

stopped single males from buzzing or clicking by broadestting a recorded

floridana buzz just ahead of and during the time when the singing males would

have made a sound.

Completely solitary males of floriiana sometimes abbreviate the clicking

part of the ES sequences and produce only the initial 1 2 clicks, pause, and

then buzz. Usually In the abbreviated sequenaee the click just before the buzz

Is present (although certain indllvlduals consistently omit even this cliche Thus,

instead of the usual click, click, click, click, click-click-buzz, the abbreviated

sequence is click, -, -, click-buzz. As Indicated above, such

siaging occurs only when a male is acoustically tislated from other males. Such

isolated males eccur early in the measoe and in oet-of-the-ordlaary habitats. I

have noticed that some of these solitary males fly about, singing a few weqnces

from each perch.

By slorwin down recordings and using a stopwatch, I determined that the

ES sequence in the middle of a series Is repeated every 1.4 2.5 sec., depend-

ing on the individual. The usual rate was about one every two seconds. In an

experiment to determine the effects of temperature on orthopteran sounds T.J.

Walker (personal commtmicatlon) found that the pulse rate of the bzas of three

males of floridana from Alachaa Coanty,Florida, averaged 45.2 pelose per sec.


at 250 C (calculated from regression formula). I recorded at 24.20 C one series

of ES scqtcnces from each of seven males from different localities in Florida

and made two sonagrams of each recording. Some of the results of the analysis

of the sonagrams are shown in Table 7.

Table 7. Results of analysis of solitary songs of Amblycorypha floridana
recorded at 24. 2 C.

Location of Average No. Average No.
collection Pulses pulses per soc. sec. between
in Florida Indiv. per buzz bu;.es clic;.s beginnings of hu-cs

Pcnsaccl.1 Bay 008-24 7-8 34.4 50.3 2.0
006-25 7 85.9 56.8 2.0
00-26 6 3.6 46.4 1.4
006-27 9-10 38.9 55.1 1,7
006-29 9 38.1 49.6 1.6
Salt Springs
(?"-rior Co.) 0OG-30 9-11 49.0 G3.7 2.1
G inew.iio G60-31 10 42.8 57.7 2.6

Those individuals from the Pensacola DBy area had somewhat slower pulse

rates than the central Florida individuals. It 1i clear that the click pulse rate

is greater than the buzz pulse rate. It is interesting also that the toothstrike

rate (not shown) in the click pulses is slower than the toothstril:e rate in the

buzz pulses. Most often the clicks are composed of three pulses. :owevor,

the first pulse is ofte very weak and barely shows up on the sonagrams. Prob-

ably this first pulse corresponds to the initial opening of the tegmina. If so, the

clicks consist of two cycles of opening. and closing the togmina. The relatioaisip

bht-ween floridana clicks and certain sounds of two closely rolate.l species

is discussed later.

Females awer each ES sequonac shortly after the buzz. I recorded


several sequences of recordJe ES and answering female tick from each of four

females from different Florida localities. I rmade sosagrams of the first ten of

these male ES-female tick sequences for each female; the results are tabulated

in Table 8. It seems that southern Florida females may have a longer delay period.

Table 8. Results of analysts to determine the timing of the female tick
after recorded male songs of Amblycorypha floridas. (Time in millisecmds.)

Location of collection Tick delay
In Florida Indiv. 0C X

Sanford 006-19 25.0 133 19
Gaineaville 006-20 25.0 139 11
Salt Springs (Marioa Co.) 006-21 25.0 143 10
Long Pine Key (Dade Co.) 006-22 M6.5 161 27

To determine what part of the ES sequence was important In evoking ticks

from females I broadest recorded parts of the ES sequence Parts I, U, and

II singly and in different combinations, to three females and noted their re-

speses. The loudspeaker was placed one foot away from individually caged

females, tested one at a time, and the sounds were broadcast at 55 db. Each

part or combination of parts of the ES sequence were broadcast 50 times. Two

series of tests were made with each female. The first series involved broad-

casting the same randomized s equence of successive loops of tape to each fe-

male. The seeosJ series involved broadcasting a different rmdomtzed se-

quence of loops of tape each time a different female was tested. On the loops

which haJ part of the EB sequence deleted a blank piece of tape was inserted to

keep the remaining parts ef the ES sequence normally spaced. Each female

was tested twice but not twice In succession. The results of thee tets are


presented in Table 9. In no case was the response to any part of the ES sequence

Tablo 9. Results of experiments to determine which part of the male sound
sequence is important in evoking the tick response from females of Amblycorypha
floridana. (Numbers in the columns represent t!ih number of different repetitions
of the sound to which the fcmal reEponicd. Thb. order of the column headings
reflects in no way the order the different sounds in the experiment were played.)

Test Idiv. Part of ES seq'tence played
series female I iI I I-II I-I1 n-mr CS

1 003-20 7 13 26 2G 40 1 48
10 17 19 24 33 44 45
1 003-21 0 3 9 11 21 8 33
0 6 7 5 31 1 46
1 003-23 0 0 2 0 41 30 50
7 5 35 1 50 48 49
2 OC-20 5 0 7 8 19 23 33
7 0 3 7 18 12 32
2 003-21 2 0 4 5 21 15 31
4 0 6 7 10 13 34
2 003-23 7 0 9 7 22 15 47
9 1 11 19 23 30 50

as great as to the ES sequence, and different combinations of clicks and buzzes

evoked greater response than either clicks or alone. It scoms that c hcl:s

and buzzes combined in sequence are Lmportant in cilciting tike responses from

females. In the field females have been heard responding readily and consistently

totlie abbreviated LE desrclboJ earlier. A logical test in these series of experiments

would have been to play a simulated abbreviated ES and to have compared the re-

sults to those above. I feel that response would have been a matter of deleting a

few clicks of Part I but not the first clic;. Thus, a female listanin3 to an abbrvv-

lateJ ES sequence would hear the beginning anJ be primed for the end of the

sequence; her response might be as if she hoard the whole CS sequence. One

argument against this idea is the fact that female response to Parts I, II, or I


and II was not delayed to a timing as if Part II had been present in those tests.

I made several recordings of one female's answer to a recording of Parts I and

II. The ensuing analysis of sonagrams showed a significantly delayed response

but net enough delay to have allowed a Part IIl to have been completed.

Other experiments were conducted in the response area to determine other

functions of the sounds oomposing the ES sequence. Sounds were broadcast at

80 db (the highest intensity measured from singing males), 55 db, and barely

audible. During the acoustical part of the tests, each sound was repeated

every two seconds.

The first series of areas experiments was. to test movement to the ES

sequence. Three males were tested. No males behaved differently (other than

acoustically) during the aceastical parts of the tests than they did during silence

no matter what the intensity of the broadcast sound. Low intensity ES's stimu-

lated males to sing, but they were inhibited from buzzing by high intensity ES's.

This is discussed above.

The three test females ticked consistently to barely audible ES's, but

either never moved or continued moving in a random pattern when the sound

was turned on. All three females answered almost all ES's at 55 db, but in these

tests they were apparently stimulated to move about. The test females usually

remained motionless or moved very little and randomly during alleaee. Immed-

iately after the 55 db ES recording was turned on all three females started

moving, or if they were already moving at the start of the recording, they

Increased their speed of movement ooaedeeably. However, they moved about

without orienting in relation to the loudspeaker. ES sequmeaes at 80 db always


caused an immediate orientation by all three females toward the loudspeaker

whether they had been still or randomly moving. Usually they proceeded di-

rectly toward the loudspeaker, although two females in two tests each went only

part-way toward the loudspeaker and simply remained oriented toward it for the

rest of the test period.

To get a better comparison of the responses at 55 and 80 db, I set the

volume controls on the recorder so that the maximum intensity broadcast at the

loudspeaker was 80 db when the volume-control knob on the loudspeaker was set

to allow a maximum intensity. Then I turned the lot dspocher volume-control

knob so that the sound played was 55 db. Thus I was able to carry out a 15-minute

test in which five minutes were silent, five minutes were at 55 db, and five minute

were at 80 db. All that was necessary to change from 55 db to 80 db was to turn

the loudspeaker volume-control knob at the 10 minute mark of the tests. The

three test females exhibited the same kinds of reactions in these combined tests

as they had done to separate tests of the 55 db and 80 db ES sequences. Almost

invariably the three kinds of movement exhiblted: by the test females to the thrca

parts of the tests were 1) no movement, 2) random movement, and 3) orientoJ

movement towarJ the speaker. It is Interesting that the females ticked after the

ES sequence at barely audible and 55 db intensities but at 80 db they ticked only


The next step was to determine which part of the ES sequence was important

in causing the above reactions. Any part or combination of parts of the ES sequence

that involved clicks caused random movements in all three females at 55 db.


The more clicks present in a particular test the greater was the amount of move-

ment, e.g. Fart I elicited only slight response. At 80 db the same sounds, those

with clicks, evoked strong orientation and movement toward the loudspeaker in

almost every test. Part III, the buzz, evoked no response from any female during

any area test at any intensity. Since these were the same females listed in Table

9, one would respect some ticking response to Part III. These females were about

two weeks oldor by the time those latter tests were conducted; I.e. age may

have been a factor in the difference In behavior. Part In combined with and

precededr by clicks evoked ticks from the test females in the arena in about the

same proportions as described earlier for the experiment to determine which

parts of the r~S sequence wear important in aliciint 1:c:.s fro-m cfeimales.

This species appears to be another (see discussion of Montezumina

modest) in which the females bridge the final gap between themselves and

the males. Unfort-mately, Amblycorypha floridana was one of the first species

with which I experimented. At that time I was strongly biased with the idea

that phaberopterine males go to close rage answering females, so I decided

to set aside the above presented data until much more time could be devoted

to experiments with floridwiaa. The reeelt was that all my floridana malee died

before I realized, In my work with M. modeeta, that it was possible that in

some species males move toward females answering at long range and that fe-

males make the final movements in getting the sexes togEther. Logical experi-

ments now to be performed are those testing male responses to different inten-

sities of answering female ticks. I hypothesize that males move toward low

intensity female ticks. A logical follow-up to all these experiment would be


to turn a male nad a female loose In a room and observe whtch sex goes toward

the other and to what extcat.

I made one c5servation which may have bearing here. SDeore coadicting

any experiments, I opened the cae of a f.malo floridana and placed a singc,

caged noldan male In the same room 12 feet away on another table. The male

sang several series of ES sequences. The fLialo answered almost every EL,

climbed out of her cage, and orlened toward tLh male. After a few minutes

she climbed dova, walked immediately across the table, and leandJ over the

table edge toward the singing male. Later, I released the male 12 feet away

from the cagaJ female but on the same table. The male oriented inimeditcly

toward the answering female but did not move until after he had proJuced

several ES sequences. Eh route toward the female he would stop an. styldulate

loudly for a few seconds. The closer he came to the female, the greater num-

ber of clicks ho would produce between buzzes. When ho flaally reached the fe-

male's cage, he climbed around over it proJuciag nothing but clic!.s; the female

ticked occasionally. I watched this last scene for only 2 3 minutes before

recaging the. male. These observations support the hypothesis that females

make the final movements In pair formation in this spe.ices. Experir..3LatiMn

along these lines imperative.

Amblycorypha oblongifolla (De Geer)

AmblycoryMha oblonifolia was not known from Florida before this study.

T.J Walker an:-I collected one female from Liberty County (western Florida)

early In June, 1932, thus extnJllng the limits of itsa lown southern distribution.


This particular female wa vary resposive to the sng of a single obMgeit*

male collated by T.J. Wxlker from Werklay CoMuy, Sm1th Crolina.

oSeI lay male of "iml4 prcodse a sort, leud, oamplB s*ead,

whblob I repe ed at v ylag IMeral (Fig. 9). Plawe (194) made an electronic

menlyai ad Almmader (1966, 1960) made an sadtwepatrgrapbto analyst of

the sound. My a alyli agree gerally with that of Alammer (1966), who

sys the sead Is prodneed only at nigM, ad tht "differvt ntdiavtdlate I a

nolawy sully sita a few mIdaes, altreaMg their eMops wIth sme or two

otSer ladividuals, thr n me set for a Saw lieN .... Cobimw thu siag

in burst, spared by ltervalas i which no laIlivtia l a sinlSg. In the

laboFratry my sagle mal sa on oly in darikas ad at soradle Irtervals,

preduaetg 10 20 phrases ech Utme. The phraes wer usually spaced 4 7

see. part.

Aleimder (IOW) ays the oerp ceteaas 2, 3, or 4 pules . .. The first

pulse Is toger ad differnl from the aeta In the dtIrp, giving at impresta n

of speeding up." Ye, wo somagrm he seows (Alatnder, 19661) aeShats a

shiMt, low in sity (compared wth the rat of the sa*d) pal. jet sheid of the

leaer, mere slwly dikdered pls that he ells t* first plke. Alost wery

ae of te smegrame of the phases by te masl hi my laboratory in this

inltal, sort prip e u~ e oh i lower In Ittm ulty tb the m 1rMatag plges. Tld

pulse s probably nd on the latil spaaig streoi of the sgwmala ad s ma-

fiwmeonal. Thsadfm, I shall rest A lmtmer's dytema of mmberbi the paless.

My soagrama leoed aessetllly IBM the Shm by APimder; a wwere tW-

paleil with the flt beiag lager the S u o nlrw. O3y in a few af ap menugrams


does the toothstrike rate during the first pulse increase towarJ the end of the

pulse as indicated. Actually some phrases have decreasing tooth-

strike rates while others have very regular toothstrike rates during the first

pale. By measuring the time interval from the end of the long first pulse to

the end of the last third) pulse I obtained a pulse-rate value for the last two pulses

from six phrases at 250 C. These averaged 20.0 pulses per sec. with a stan-

dard deviation of 0.5 pulses per see. The average total duration of these phrases

was 199 mssc. The average number of toothstrikes per pulse for the three

pulses comprising the phrase were 15, 10, 9 respectively. It seems that the

greater duration of the first pulse Is due to a combination of closing the tegmina

more slowly (assuming the pulses are made on the closing strokes of the tgmilna)

and striking more tcth.

Unfortunately, the male died before I could conduct any experiments to

determine his reactions to recorded conspecific soa-ndn. Just before he died I

noted that the male was stimulated to sing in alternation with broadcast phrases

of any Intensity up to 110 db. I did not measure the maximum intensity emitted

by this male, but 100 db recordings soin.ded comparable to the sounds from the

male. Long before the male died I placed him in the arena and put the caged

female on the corner (loudspeaker position). The female answered his phrases

and immediately the male went straight to the female in every test. Thus, it

seems that males go to females at least at close range.

The female In the laboratory tic:.ed after almost every phrase that the

male made. The average delay at 250 C of the female tic!; after the end of

six male phrases was 205 msec. with a standard deviation of 13 msec.


In order to determine which part of the complex phrase was important in

elloiting the ticks from the female, I divided the phrase into two parts Part I:

the first, long pulse, and Part UI: the final two similar palm. Surprisingly,

both Part I and Part II evoked ticks from the female as well as the whole phrase

(ES). Since the Intensity is relatively uniform throughout the phrase, the only

difference in playing the ES in reverse was the sequenee of arrangement of the

structural components of the ES phrase. The female gave no ticks to such a

sound. However, Part I broadcast backwards ellcitt~ as much response as

did the ES when broadcast forward. Part II backward evoked no response.

Why? I set up a series of tests to determine if this observation could be

treated. These tests involved ES forward and backward, Part I forward and

backward, Part I forward and backward, Part I forward plus Part II backward,

Part I backward plus Part IE forward, Part I backward plus Part U backward,

Part II forward plus Fart 1 forward, Part n forward plus Part I backward, and

Part II backward plus Part I forward. Sound was broadcast every six seconds

at 100 db. The sequence of broadcasting the different combinations of asoad was

changed in every series of tests. Five series of tests were made covering a

period of about a week.

The female responded to every sound broadcast by emitting ticks, except

to these sounds in which Part II was broadcast in reverse. What was it shout

the two terminal pulses broadcast in reverse that rendered thea Incapable of

evoking the tick response from the female? I can only conjecture. Tt may be the

necessity of shortening the pulse lengths in seoGMsivw pulses.

In arena tests to determine whet movements might be invoked by these


sounds, the female answered high intensity ES's readily, but she answered very

few below 80 db. The female made no oriented movements to 95 100 db ES's,

but she always oriented toward 75 80 db ES's and, in about one-half of the tests,

moved toward the loudspeaker a short distance, remaining oriented and answering

a few of the phrases after stopping her movement. The female died overnight

after the above tests and before tests could be conducted at lower intensities.

These limited studies indicate that the female may have been attracted toward

low intensity ES's and would not have made any ticks to low intensity ES's.

Before more definite conclusions can be made concerning the acoustical be-

havior of this species, data are needed involving several individuals.

Species Not Involved In Experi:netnI

The following species were not involved in cxpctrimentatioa to determine

the functions of their sounds. The sounds of some of these species have never

been reported, so the following presentation includes descriptions of known

sounds and, when known, descriptions of the singing behaviors.

Amblycorypha cartnata Rehn and Hsbard

Amblycorypha carinata has been located at only one small place near

Gainesville, Florida. It occupies the undergrowth of a long-leaf pine flatwoods

and the population density is low. It often occurs with Amnblycorypha floridana.

Sound production by this species has not previously been described. :'um-

erous observations of the single population inricate that it is a night singing species.

Solitary males usually produce L t.o-pulsed phrase (Fig. 10) which is repeat-

ed about every two seconds in series of varying numbers. In the field I heard


one male produce 67 phrases in a continuous srtes. Males apparently produce

the same seoud whether they are completely solitary or close to, but not touch-

ing each other. Sometimes two males may alternate phrase regularly but

such alternation is probably happenstance. Infreqmutly, males produce a series

of three-pulsed phrases.

The pulses composing the phrases described above apparently correspond

to single openings and closings of the tegmina, one opening and closing pro-

ducing a single pulse on the closing stroke. However, close inspection re-

veals, in certain sonagrams, very brief pulses which probably are made on

the opening strokes. Usually one of these brief pulses is the initial sound in

each phre (see Fig. 10), and evidently is made an the initial opening stroke

of the tegpnla. These brief pulse are not included in my counting of the nem-

ber of pulses in a phrase.

I have recorded at least one series from each of several males from

different localities. The results of songraphlc analysis of ten phrases of each

of these recordings are shown in Table 10. The analysts was made only of two-

pulsed phrases except for Individual 006-19 from whih individual only three-

pulsed phrases were recorded. A recording of a series of three-pulsed phrase

from individual 006-8 was made, and the results of analysis of these phrases were

similar to those depicted for 006-8 two-paoied phrases. There were bainSial

differences In pulse duration between the first and soced pulses of certala In-

dividuals. A bias in the calculated pulse rates for those Individuals would have

resulted if I had measured the time Interval from the beginning of the first pjse

to the beginning of the second pulse the time spent in the first omplm eyeOSe


Table 10. Results of analysis of the songs of some males of Amblycorypha

Average No. between
Pulses per toothstriles beginnings
Location of second pulse no. of phrases
collection Indlv. oC sx 1 2 3 secondn JE

Gwinnett Co., Ga. 006-3 25.0 15.7 .6 16 14 1.5
000-4 25.0 17.0 .3 22 15 2.3
006-5 25.0 17.2 1.0 8 9 1.1
Alachua Co., Fla. 006-7 25.5 18.2 .4 18 18 1.7
006-8 23.5 18.0 .3 14 14 1.5
006-10 24.8 18.0 .5 17 17 16 2.9
Jackson Co., N.C. 006-9 25.0 19.6 .4 9 11 2.6

of closing and opening the tegnilna or from the end of tha first pulse to the end

of the second pulse. Therefore, I measured both these time intervals and divided

the averaged time interval into one hoping to obtain truer measure of the pulse


There appear to be small variations in pulse rates between lnri viltuals

from different localities. One striking difference between the songs of different

Individuals irrespective of locality is the number of teeth structh per pulse.

Another difference Is tho toothstriko rate; generally the individuals which btraci;

fewer teeth struck them at a slower rate than did the Indivi-lals which struck

more teeth.

No responsive females of carinata were collcctLd, so I do not know what

structural unit of sourd the females answer assuming they do produce sound

in the response to male sounds or at what timing the females answer. The

moat likely unit for a female to respond to is the lioivldual phrase.


Amwircb aft flodtps A. obleuaifolia, aadA. al are closely

Related and tamaemisiz have referred to the in different nseamee u s species

or euSqexsm (Ikoh ad Nuierd, 1905, 1814b; IMslksley, ]IO; Fttie, 1932).

The three speoM are very similar to each other morpholoqgally, ln habitat

pryfaences, ad in seaed pedwotlon. The seeing atsllhfme mry *t be apper-

eat at first gliee, but ela ispeetion *o a click of (FiP. 8), a

phase of oMWlS (rig. 9), aud A phrase of cariata (Fig. 10) reveals

strildog amlaritiesw. All samds here a Initial, low intensity pulse whieh

probably eorresponds to t apmalag st-rke of th tEbagna. Each is a short

sound of two or three pules. Tie biwest diffirrame amern the three senad

are between paso rats sad palus duratits ad all wire probably derived by

modiftiation of one baic sound.

Amb&yceryphi rofa tfela (Scadier)

DeseripUo of the sowds of Amblyearypih ra ndifelia by JdffeM t author

(e.g. Allard, 1911 1911 ad SIWkdd, lI3S) were Inemilatent l onfuwelg

ail Alwmader (1960) shoed tlht the wer two qspece lwvolved, spqrableonly

by ther S*og and partly by their geegraphal d1sirtblaM~. HI described a

emmofsn "fIaler" Iad a seutbee clickerr" wnieb marhap "'abat M20 miles across

the AppwaIeMa Mem~latis." HIe se lbe sag palter in ol ek eo l es two specte

'Ia osuwis ad ifrerible. The sug of the reler 1 cam lposd of grf l of

similar puales wbkih become pregs esivly lewag, Lther tSerminfing i a

tagle, long pulse gramp %ally fellei rd by on to ise shart pale grqps. AU

of te pules I t tMis sg we alike, ad wtk samrtM six to sgt teethstikes.


In the clicker, on the other hand, the successive pulse groups in the song are of

about the same length though there is a slight reduction in the rate of pro.!uction

as the song progresses. Each pulse group is in itself an irreversible pattern

composed of three or four pulses, of which the last is much longer (contains

more toothstrikes) than the first two or three.

"To the human listener, these songs appear to bear no relationship to

each other. However, a closer examination reveals that they have many similar

structural characteristics. Each song is composed of groups of pulse groups,

and the structure of the Individual toothstrikcs appears to be identical. Further-

more, the songs are about the same length, they are produced in chorus in the

two species in the same way, and they arc proJLLceJ at intervals of similar

length in the singing of lone males."

The clicker has been found in Liberty and Jackson Counties, Florida.

Individuals of the clicker kept in cases in the laboratory during this study sang

much like the pattern indicated by Alexander but usually grouped their basic

pulse groups. Figure 13 shows two basic pulse groups of the clickor. Note the

increased duration at the end.

A third species of rotundifolla has been discovered by T.J. Walker

(personal communication). I call it the "iast clicl-er" to distinguish it from

Alexander's "slon cliclker.:' The fast clicker occurs farther south than the other

two species, the southernmost collection having been made at Gainesville. Florida.

Both fast and slow clickers occur together, perhaps not completely overlapping in

habitat preference, in Liberty County, Florida. Wherever found, the fast clicker

has been very sparsely populated, whereas comparatively large numbers of the


slow clicker may be found in favorable habitat.

The song pattern of the fast clicker Is dtagramed in Figure 11. The click

rato basic phrase rate Is much faster than the corresponding click rate -

basic phrase rate of the slow clicker. Also there are fewer pulses in a single

phrase, or click, of the fast clicker (compare Figs. 12 and 13:). Since the only

differences between the fast and slow clickers are the differences in basic phrase

lengths reflected by differences in the number of pulses per phrase and by

differences in the pulse rates within the basic phrases and the basic phrase

repetition rates, I think such terminology (fast and slow clicker) is logical

even though It is not In line with the definitions presented at the beginning of

this section. Alexander has already proposed the term "clicker" and a change

of that species' "name" would only add confusion. A summary of the analysis

of recordings of the songs of the fast and slow clickers is presented in Table

11. On two occasions individual 001-5 almost exactly doubled his click rate In

Table 11. The results of the aaenlysi of the songs of the "fast" aad
"Slow" clicker forms of Amlycerypha rotundifolia.

Seng Location of Clicks per Pulses per
form collection Indiv. o C second click

Fast clicker Alachua Co., Fla. 005-6 25.8 10.2 3
S Leon Co., Fla. 006-4 25.8 10.4 3
Stanley Co., T..C. 005-5 25.8 9.8 4

Slow clicker Liberty Co., Fla. 001-5 26.0 2.6 4-8
"Jackson Co., Fla. 001-6 25.0 2.0 5
001-7 25.0 2.3 4-5
S" 001-8 25.0 1.9 4
001-9 25.0 2.4 5

the middle of a series of clicks. Sometimes the slow ollc:kers produced a single-


toothstrike tick about mid-way between successive clicks for several clicls in

a series.

It would be instructive to know where the female tick comes into the sequence

for each of these species. Slow clicker females would almost have time enough

to "squeeze in" a rapid response just behind each click, but fast ticl or females

would not. The most logical conjecture is that females of both species answer

each group of basic phrases the basic pulse group of the rattler. If a male

hears a female answer one of the initial short units there would be no need to

produce the long, loud sequence, which could be detrimental by possibly allo'w-

ing predators to locate him by sound (see Walker, 1964). If a female tick is

not heard by a male after his initial, short, phrase groups, then the longer,

louder series may serve to attract females toward the male from a distance.

ArAjycery;,l'j uileri Stal
Amblycorypha near uhlerl Stal

Amblycorypha near uhlerl is an described: species from the coastal

plains of couthoa3tcrn Unite I States. 1.. D. Alexan.ler and T.J. V.'alker were

among the first to recognize its distinctiveness from ahlcrl. Both occur in

the same or similar habitats open, wee-ly situations such as abandoned fields

or open woods. Around Gainesville near uhlcri matures 3 4 v~c;s carrier

than uhlori. Generally near uhlerl Is larger than uhleri.

Allard (1912), Fulton (1932), and Alexand-er (105G, 190G) have described

the sound production of uhleri and all reports generally agree. A night singer,

uhleri males sing from porclics 3 4 feet off the ;round. Often males are

found. in local, loose congrogations, ;ieiig p:ac'1 only a few feet apart,


In fields where there may be a much greater inhabitable area. Males appar-

ently retain a single singing perch for an entire evening of singing except

possibly in cases where they have had an acoustical interaction with a female.

Congregated males may sing only in spurts, so that there are periods of singing

by several males interspersed by intervals of silence.

Alexander (19G0) diagrams a typical song pattern of uhleri from North

Carolina and shows sonagrams of different parts of the song. The following

quote from Alexander (1956) adequately describes the song. "The calling song

of Uhler's katydid is a soft sound, audible only a few yards away, and is prob-

ably the most complicated sequence of sounds produced by any American orthop-

teran. It lasts up to 40 or 50 seconds and contains 3 or 4 distinct phases which

are consistently repeated in the same sequence in every song . . The song

begins with a rapid, tsip-i-tsip-i-tslp-1-tsip which continues for about 7 11

seconds, and Involves pulses delivered at a rate of about 12 per second (prob-

ably 0 wingstrokes). This merges abruptly into a slower delivery of about 7

pulses per second which lasts only about 2 seconds, and ends with an abruptly

louder, rattly phrase which dies away in intensity. This phrase, which contains

7 or B pulses delivered at a rate of about 10 per second, slowing at the end, is

repeated anywhere from 3 or 4 to 10 times at Intervals of 1/2 to 6 3/4 seconJs

(In the songs recorded). Often one or more of these phrases is preceded by

a few soft ticks, apparently caused by striking Individual file teeth. At least

4 different kinds of pulses are involved in the 3 different phases of this song,

and there are both gradual and abrupt chasge in intensity, and ia speed of wing

motion. "


The song of near uhlert is very similar to that of uhlerl. Four parts,

homoloous to the fair parts of ulleri's song (Alexan.ler says three parts), are

involved and produced in the same sequence. But each part is much briefer

in the song of near uhlori, and the their, part is produced only once. Thus, near

uhleor's complete Eorig lasts 7 10 see. at 25 C. Ofioa, when first starting

a series of songs, inar uhleri produces only parts one and two for several

seqainces before producing a complete song. For Instance, a series or songs

may have a sequence, denoted by parts sung, of 1-2, 1-2, 1-2, 1-2-3-4, 1-2-3-4,

1-2-3-4, .. etc. Singing males of near uhlerl behave quite differently from

those of uhlerL They do not usually congregate, and single males fly from perch

to perch singing a short series of songs from each porch.

In his temperature experiments, T.J. Walker (personal communicauc-)

has found that each species has a characteristic pulse rate for each part of the

song and the pulse rates of one species are different from the pulse rates of the


Recently T.J. Walker conducted somo irellminary experiments in an

effort to determine the behavioral significance of the different parts of the song

of these two species (personal communication). lie fo'mn that no one particular

part of the song by itself would evoke the tick response from conspecific females,

and that females of the two species respond at strilingly different timings. In

tests with the entire song, uhlori females ticked about 7.5 see. after the end

of part II. This delay allowed the production of two phrases of Part III before

the female tick. Yet, apparently the timing was in relation to the initial Parts

I and II because the female ticl; dalay was characteristically about 7.3 sec.


after the initial Parts I and II whether or not one or two Part II's were produced.

Ln OtWr tests the results were not completely decisive as to function, but

the data did allow some strong suggestions to be made. It sems as though

ulIl females may go to singing males from close range. Near uhleri females

probably do not move toward males at any time. This agrees with the singing

behaviors of the male of the two species. Uhleri males are stationary for long

periods. Thus, uhleri females would be able to move to a single singing male

easily. On the other hand, near uhlari males' behavior of moving from perch

to perch would make it difficult for conspecific females to reach a singing male.

Certain combinations of parts of the songs different combination for each

species are especially important in eliciting the female tic!. More work is

needed concerning the behavior of these species.

AragLiae phaelu Scadder

This species has been collected In Alaesna County, Florida, in very dry

Iabitats generally old fields and tarkey oak woodlands.

No one has previously described phalorEt's seItd production. I have

recorded two different songs from individually caed males. One song (the elixg

og) consists of a series of 11 12 liapy pules. The second song (the tick-

Igip 8g; Fig. 14) is a complex one involving Uicks and lisps. Every tick-

Ilap song contains two sequences a. ticks followed by a lisp. Tlek-lilp songs

are usually produced 3 6 sec. apart in series cotainltg e many as a doaen

songs. Table 12 shows the results of sonsgraphic a l sal of recordiags of

the two songs of this specees at room temperature. On.M one typical living


Taulc 12. Results of analysis of the two songs of Arethaea phlaniurm.

Ticks per Lisps per No. ticks lisp
Type of No. songs second second before lisp duration
song Indiv. oC analyzed s X s, 1 2 (msec.)

Ticl:-lisp 011-2 24.S 1 4-5.2 3.0 7 12 54
011-2 25.0 5 .14.1 .1 3.1 .2 8 11 55
011-6 25.5 6 45.3 .3 2.8 .1 7 13 50

Ilsping 011-1 27.0 1 S.0 23
011-6 25.5 3 4.9 .4 22

song was recorded (individual 011-1); the three songs indicated for individual

011-0 were brief, containing 4, 2, and 3 pulses respectively, so that its lisp-

rate is not to be taken as typical. The lisp rate of the tick-lisp soag is the rate

at which the two lisps were given. There was no significant difference in the

number of teeth struck bctw.ecn individuals or between! types of lisps within an

individual. Therefore the shortened lisps of the lisping song are a result of

increased rate of Logninal closure. This is evidenced by visually comparing

the spacing of toothstriicos in tle lisps of the two songs. The ticks of the tict:-

lisp song are evidently produced by in-Jividual ocnings and clostau s of the

tegmina during which a single tooth is struck. Tie evidence for this is the

very uniform tick rate and the fact that, in some cases, there are low intensity

pulses followvln cac tick which probably are the result of striking the stridu-

latory apparatus when opening the torgmna. At times only this low intensity

pulse is present, indicating that the singer may have "missed" on his closing

stroke (see the gap which shoulJ correspond to the sixth tick in the second

sequeice of Figure 14).


Nothing is known about the signing behavior in nfaral sitlttons. I haw

heard a male sgingn out-of-deors only on one oeemada for a piote of about

five minutes. It was night and the sager produced] eel tick-liip esoa.

I can only conjecture a iet the funetlonal sigifileee o these soande.

Perhaps the lilng song elicits female ties similar to t"e female reupoase

to ticking In Mlcroeanir n rbn flam., for Lastoe and the lisp-tick soeg

attracts or spaces females and males rPepctlvely.

IaJlaG S a litoerd

Thim spocie s is umlly foarl arguee Gamaewile ki ceypres head as la

oEuderiSiWm~ (dtcesmed earler in this selaen). It is my feels g tnat

Uth two species am aleely r fied. Both are resrelred feeders en very

resinous veitatlon (f ont leaves of ri n M f ilin d rfE rl

on teves of TwiaOijujm zm and beo we merphkllgicslly etmilar.

f they are closely related, one would eqwet their soue d production to be siml-

lar as wva mshem for SaMlMar W and S. .taI ta, for the thr e Am~e ory-

p2l'., el L9agMli&E ariMtI, aid orldMAs, for Amciuroiry.s afcrl and A,

near s9 l and for th AB~lYa lsra t tl grop. In Meh of lese speole

complemek the elego relaeenhips among *Is speelwe wooe obvious by their

marpheioloal similarttimes nd by their semetical behavior.

Only se sornd hs ever ben h rd from caged or wild intvidwals of

Ianscuddgeri walker by anyone in our laboatory. I eail tie s ond a ip, &al-

though In reality It is a two-pulsew somd (see Fig. Sb and dersflu from the

definition of 9 lisp. LHoeIWr, the two prices upprt to be masM in At MBsB


opening and closing of the tegmina, the low Intensity first pulse being produced

on the opening stroke. Hence, my justification for calling this sound a lisp.

Four males were recorded in the laboratory and the analysis of those recordings

is shown in Table 13. The lisp duration is qaite long whether one considers

the second pulse or the entire lisp.

Table 13. Results of analysis of songs of Inocudderia walkcri. (Time in

No. lisps Pulse duration Total time
Indiv. 0 C analyzed 1 2 of sound

022-1 26.6 2 51 130 204
022-2 27.8 2 50 156 223
'22-3 26.5 3 40 151 217
022-4 25.0 2 49 164 251

Uttle is known of the singing behavior of wal.erl. Individuals sing from

the tops of the cypress trees that they frequent and proJuco sueh low intensity

lisps that it is nearly impossible to make good oba.rvations whcn the general

community resounds with the sounds of many other species of singing Orthoptcra .

Usually walker produces about four lisps in a soag, these spaced 1.8 2.7

aee. apart. Different songs are repeated a few nunutes apart.

The kinship betweoonL-v'.i andl triga, as far as acoustical behavior Is

concerned, is not strengthened substantially by this aarlysis. Possibly the an-

estral species was a simple "llsper ," and these two species have evolved their

distinctive sounds very recently. Based on the supposed closo relationship

betwv.een walkori and striata, I aspect that females of wal;.cri tick in response

to the lisp probably with a greater delay than females of striaa tic: to lisps

or thlir males.

Microcontrum retinerve Burmeliter

This species abounds in hydrIc and magic hardwood forests of southeastern

United States. It is a most difficult species to study because it inhabits and

sings from treetops. A collector is most likely to catch adults of retinervo

around street lights through wooded areas, for the katydid often seems to be

attracted to light.

A typical song of retteervo males around Gainesville, Florida, is a series

of loud, brusque phrases with decreasing numbers of pulses in each phrase.

Seven Is the maximum number of pulse per phrase that I have heard. Typically,

a male produces a sequence of 5-, 4-, 3-, 2-, 2-, 2-pulsed phrases, the ter-

minal two-pulsod phrases often alternated with two-pulsed phrases from another

individual. Singing males fly from perch to perch singing one to a few songs

from each perch.

Since captured males never sang in the laboratory, I had to rely upeo analysis

of field recordings made by T.J. Walker. Fortunately, he obtatied three good

recordings on one evening, all at 240 C. Thus, the analysis of these recordings

should be comparable to analyses of recordings of other species made at lab-

oratory temperatures (about 250 C). In my analysis I used only those phrases

which contained three or more pulses. The summary of the analysis of these

songs Is presmeted in Table 14.

I have no exact data on the Interval between successive songs, but obser-

vations Indicate songs may be repeated 2 3 times per minute.


Table 14. Results of analysis of songs of three males of Microcentrum
retinerve at 240 C.



No. phrases Pulses per sec. Average No seconds
analyzed b sx between phrases

1 12.8 2.4
5 13.8 .2 1.9
3 14.4 .2 1.6

The song of retinerve has been described by Allard (1910a, 1928a) and

Alexander (1956). Alexander (1980) presents a sonagram of a complete song

of retinerve from Ohio. Both Allard's and Alexander's descriptions vary from

mine. Allard says the usual number of phrases is three and rarely four (Alexander,

1966, thinks Allard's observations are probably Irom around Washington, D.C.).

On the other hand, Alexander (1956) says the song at Raleigh, N. C. and in Ohio

consists of two and rarely three phrases, each phrase composed of 2 7 pulses.

He says common scquences are 7 5, 6 4, 5 3, and 4 3 pulses per phrase.

Comparisons of the geographical variations in pulse rate at the same temperature

would be a valuable adjunct to these d.ita. I asspect that the pulse rate is function-

ally important and the phrase length is not.

I suspect that females answer every phrase produced in a song immediately

after the phrase. Since males move around at frequent intervals, I imagine that

females do not locomote toward the singing males, but that males move the entire

distance separating them from females once they have heard an answoring female


F-hrixa maya Saussure and Pictet

This specles Is known from the United Statos only from three collections.


-- -


Blatc'ilel (13SW) ya that Davis (1914) and P eardJ (1915 ) each took a aeagi

specimen in Brickell's Ihammock, Miami, Florida. T.J. Walker made the third

collection of six inJdvidvils in 1B60 nar Flamingo, Florida. and made a short

laboratory reeordfig of its mg.

Only one kind of sound was heard from the male recorded sad hat may bu

described as a pair of paired ticks or a a pair of clicks (Fig. 15). Every phrase

ha. the same appearuce and the phrases are repeated 2.0 3.5 ee. apart.

Analysis of Walker's recording (eentaletng 12 phrases) ave a pulse rate within

tick pairs of 11.6 pulsea pr sec. with a atmadard deviation of 0.7 pulses per

sec. The pairs of ticks within the phrases aid a repetition rate of 4.5 pairs

per sec. and a standard deviation of n.3 pairs per see.

The umMsoal rates of repetition of the ticks within the phrases made this

sound comnple within Itself. Eaeh tick probably corresponds to a aligle closure

of the ta niaa, so in prouhcting each phrase the wingstreke rte changes twice.

I will not hazard a Iaes as to the function of the somnd, alhtagh I do tMhnk

that single phrase would be the unctlUol l Iait. The saod is so low in inen-

slty that it probably fiu tions only at less rang (the stridulatory apparatus of

this species is very degenerate).

Senddiari curylo riat .taoed* BrauIer

This species is foaud in the pins flatwood of Alachea Ceauty, Florida.

Inhabiting the harbaesis ad shrob strata. I ha e wver seen l at d where

its population density was more than labot two main per aere, aetbated by

llstening to singing divldabil, mad generally the dem appears kles tam this

figure. I have made very few attenlve observatioems o its singing behavlor


although I have casually heard it night-after-night during June and July lor two

years. It seems that the males are quite stationary for long periods while

singing. I have heard only one song from laticaida, and workers in our lab-

oratory have not reported any song from this species different from the one

I have heard. It is produced only at night.

A typical song is a short series of phrases with each succeeding phrase

usually containing one pulse more than the preceding phrase just the reverse

of the pattern described for retincrvc. The usual sequence of

phrases heard contains 2-, 3-, 4-, 5-pulses per phrase. I have heard males

"count" from one to seven never more than seven in some songs. Often

a male will repeat phrases of a certain number of pulses producing a seence

like 2-, 3-, 4-, 4-, 5-, 5. Different songs are produced several minutes


Of the recordings on hand only two were made at 250 C. One recording

contilns only one phrase with ive pulses delivered at a rate of 5.7 pulses per

sec. The second recording contains four phrases with pulse rates of 4.5, 4.7,

4.8, and 4.8 pulses per see. Intervals between phrases varied from 3.5 to 5.5

sec. In his temperature experiment T.J. Walker obtained a pulse rate of 5.8

pulses per sec. at 250 C (calculated from the rogresclon formula).

Several authors (e.g. Cantrall, 1943; Fulton, 1032; Rehn and Iebard,

1914a) have described the song of Scudderia olrvieana curvicn.tda which is a

northern subspecies of S. cuvicaa. Their descriptions generally agree with

those given above; Alexander (9156) gives a slightly slower pulse rate, and


says that 4-. 5-. and 6-pulsed phrases are rarely heard. Fulton (1951) de-

scribed a song similar to that described above. He thinks his description is

of latieasua's song. Some of the writers describe a single-pulsed 'tsick" or

'brwzl" produced in tie daytime. This could be a single pulse of the song

described here.

The two males involved in the temperature experiment exhibited some

interesting acoustical interactions. When one male asng sogs like that described

above, it often stimulated the other male to do the same thing, but not syn-

chronously. At other times one male either one would give loud, slowly

delivered ticks after each phrase produced by the other. Perhaps intensity of

the sound received has something to do with what response is elicited.

I would suspect tat females answer each phrase of the song.

Stilaechlaer cotlemians (Samseure)

This large katydid is not known north of Alachua County, Florida. It is

generally a tree-top dweller in hardwood forests.

No males of coaleniane have ever sung in our laboratory, and we have

never seen a wild male singing. However. T.J. Walker and I have heard, In

San Feiake Hamnmoek, near Gainesville, and in hammocks of southern Florida,

a long, loud, course Uisp (Fig. 16), which we Esppose is made by males of

couleaip a.

Only one field recording suitable for analysis wae made and tlt at 19.50 C.

The five lisps of the recording were 197, 211, 248. 219, and 224 rmec. duration

(average 220 meec). Two other recordings with more than one lip wee made

at 200 C and, although not suitable for analysis, I could measure the time inter-

vals between lisps. Successive lisps were produced 6.5 20.0 sec. apart, but

usually 13 15 see, apart. I believe each lisp of this species functions as a

complete song.

Turpilia rostrata (Rehn and liobard)

This species has been found only in the sibtroplcal hammocks and man-

grove swamps of southern Florida chiefly in the latter. fond, It has

usually been nurmcrous.

Three songs are known from rostrata males: a ticKin~ song, a lispin. Eon,

and a liEp-tie, son. TickiLg songs are produced during thU evening twilight.

As darkness sets in, ticking gives way to iLe liLping song, anJ still later the lisp-

tick song becomes prominent. Late at night one may hear the lisping songs and

the lisp-tick songs with alxout eqial rates of occurrence. Certain individuals may

produce the ticking song late at night. At times one may hear the lisp-tice song

and lisping song produced in a regular soqlence.

The ticking song (Fig. 18) is a series of phrases irregularly spaced 0.3 -4.0

sec. apart and composed of 1 5 ticks usually 2 3 ttlcs per phrase. The

tick rate is surprisingly tuiform within the tick phrases (see Table 15). The

number of tlc:;s per phrase, however, is completely unpredictable. A series of

phrases of a ticking song may contain 2-, 3-, 3-, 3-, 1-, 4-, 2-, 2-, 4-, 1-,

etc. pulses per phrase. The length of a ticking song is indefinite; ticking pro-

cce:ls more or lss continuously until a lisping song Is produce-i later n fie



Table 15. Results of analysis of songs of Turpilla rostrata.

Imsp duration
Kindj :o. phrases Ticks see. IaspD/sec. (maec
of ong Indiv. OC analysed x 'x X Sx

Ticktag 071-2 26.0 10 20.0 2.1
Lisping 071-2 20.0 1 2.5 30 2
Iaping 071-3 25.0 4 3.4 .5 31 2
iasp-tick 071-8 24.5 5 23.1 .7 2.4 .3 35 2
2 malem 071-11
in mage & -1225.0 5 26.2 2.5
in cae & -12

The lisping ong (Fig. 19) consists of a slaglo phrase of 5 11 lisps pro-

duced at a regular rate until the last 2 3 lisps, at which time the lisp rate

slows down. The lisp rates indicated In Table 15 are calculated from the first

5 7 regularly delivered lips of each song. Intervals between songs may vary

from one to several minutes.

Only one recorlung (laboratory) near 250 C was obtaliu oc the lisp-ticA

song (Fig. 17). hI this recording the singer (individual 071-8 in Table 15) added

a series of lisps to each of the five lisp-tick phrases, so that the Intervals -

35 G5 seconds between successive lisp-tick phrases were not typical. Another

recording, made at 23.50 C In the field between 10 and 11 o'clock at night, had

two songs of 5 md 7 lisp-tick phrases respectively. Intervals between successive

phrases in those songs varied from 4.0 to 8.3 sec. and averaged 5.7 sec.

No responsive females were ever collected at least no captured fomala

ever gave any apparent response to any test so I do not know to what song, or

soend, the females would answer or be attracted. By the very nature of the tick-

ing song, and on the basis that talking in certain other species results in male


spacing, I suspect that the ticking In rostrata functions in male spacing. This

idea is supported by the fact that the two males caged together in a four-inch,

cubical cage produced many tickling sounds which were erratic and intense. At

times the tempo from the two males :woall lessen and the ticks pro.iuced in those

quieter periods were at a tick rate similar to the tic: rates of individual 071-2

(field recording) who was assumed to be out of physical contact with another male

while ticking. Contrary to this, however, is the fact that males in arena tests

never made any movements during any tests. If ticlJng actually does function

in male spacing then one would assume the lisping songs to fiction in male-

female relationships. Indeed, I think that once the problems with rostrata are

worked out, we will see at't females answer individual lisps close after the

lisps. As to whether females move toward singing males, I will not guess.

It could be that the almost regular rate of delivery of the lisps could function

in this capacity when received at low inteusltles.


There Is no instance of any speels of saying Orthoptera harng "learned"

its ornd repertoire by liUteling to ether members of the spuoes. IndividMual

which hatch and mature in the spring of each year never bear the sounds made

by their parts. Yet, they produce sotuds Ideatieal to thkse ao the parete.

Individuals reared in iselation do likewlee. Alxauoier'o (169b) stastment.

'there is no 'culture' in cricket signalling" is applimble to katydido also.

Kiads of resonse to mound stinull

When an individual hears a sound it may exhlbt any one of several behavior

patterns. It may do nothing different from what itwas dolag before it heard

the sound. This Ie typical of the response give to most kharospeelflc somids.

Conspecific sounds functioning in intraspeetle coaalo~atlon usually evoke

kidnes or taxes. Kinetic eactionsc are evidened by a k*tydid's Mtersug to move

when it hears a sound, its continued random movement frequent tur~ng a

long as the sound is repeated, and its stopping soon after the sound seans. Such

kiness were typical of males of certain species In tests of ticking sede. In

natural populations such movement could result in the spacing of individuals as

as result of kinetioally moving in respese to the ticidng seuade. There may

also be Interspecific interaction in this respects, mlnoe many espelte with tick-

ing sounds in their repertoire often oceur toLgther.

Tactic resetions, orientation toward roses of stitU, we embibited



when individuals are attracted to a sound. Such reactions may be more strictly

termed telota.xes, for the orientations are directed toward single sources of

sound when many sounds are present.

Intensity of response to sound stimuli

Different intensities of response from different individuals to the same

sound and within a single individual to the same sound at different times have

been noticed. At one time an individual may casually walk toward a sound if it

is an attracting sound or produce one tick if it is a tick eliciting sound. At other

times an individual responding to an attracting sound may alternately lean over on

one side and then the other, holding up the front leg on the high side as if to more

fully expose the auditory tympanum on that leg to the sound. Between successive

alternations of "leaning and listening" the katydid may run a few steps toward the

sound. In the same context, females often produce, in response to tick eliciting

sounds, not one, but two, three, or more ticks in rapid succession.

Sexual maturation of adults

Every species studied showed a surprisingly long delay between the time

of molting to the adult stage and the time of attaining sexual maturity, as evi-

denced by the beginning of sound production by males and responsiveness to

conspecifie sounds by females. In almost every species studied 5 7 days

passed after the final molt before the insect became acoustically active. These

insects would be excellent ones in which to study various hormone concentrations

after their final molts.

During the first few days of the seasons in which the different species

attain adulthood, certain individuals are often found far from their normal


habitats. Sometimes an individual may be heard singing in out-of-the-ordinary

habitats all during the eason. Such observations Indicate that individuals of the

species involved may disperse during the adult period prior to seal maturity.

If they do Indeed fly around, It Is obviously one way of intermixing the genetic

material of different populations.

Stimulus situation for sontd production

Almost nothing is known about what external stimuli are important In

Inducing an Individual to produce sound. Only in a few cases are they known; for

Instance, certain male sounds are absolutely necessary to evoke ticking responses

from females or other males. The chief problem lies in determining the nature of

the stimulus situation for apontaneously producing different sounds In those species

which produce more than one kind of sound in solitary situations. In many cases

- e.g. In most species of Amblycorypha low light intensity is required for

sound production. In other cases e.g. In Scadderia toxansis and Turpllla

restrata intermediate light Inteasittes stimulate certain kinds of sound pro-

duction. In still other cases e.g. S tammeis certain sounds may be pro-

duced principally by day, other sounds principally by night. In a few cases -

e.g. S. furcata and S. camitsl the same sounds are produced day and night.

But what causes a male to produce one kind of sound for a period and suddenly

change to a second kind of sound, whether the different kinds of sound are Isolated

accomplishments one from another or are produced in a regular sequence?

Undoubtedly, there must be a change within the linger, because, so far as it is

known, external cues do not change at rates which could be correlated to the

rates of change of the kinds of sound produced. I doubt that this question will

soon be answore:.,

n'fl.e,!tty or acoZtlrl!nt :ltra"t'on9 In paIr Mrzratlon

If a sipcles of C'rthoptora I.1l no intrinlso mochan!sa to al-! in the form-

atfi of sexually r3aTpr.slvo pairs, the chance of males ard far.alos of sparsely

populated 5pocies corning t L nature at tines when both aro sexually

r.aturo andl ac:l ally roesonsWvro vwud h a1 liMt. In ouier words. casual palr-

ing of malas a; toamales would ( not seam tn proim.c a very o fective broe'ling

p.rlatlman unless tho ropuiuatloi slalty were B!i rt.' the i.Av'. Anl .s of the

oe concern' wcro active. f-oynlatons with lIw C~.entlos, as is the

coo with nimy clof rhanoropterrina. res be very effective

Ipnmlator.s if prviled W"t a nwochwrtni to aid la pair farjustio (ieas orf n-

ilaliy those of R. P. -raL3ler prranson aomnt'-r!cation). Several scr'

mocha.nissm aro o!=oti for "Ifferent kIn Is of Insects for instance, attracton

t, spocif- host plants, a':- the plhromone of certain Lcpidoerto n ma

other oInocts. The aco-Luical systems of .uL ing insects fa.cticn in this

capacity, I apparently very offnciently.

Tl-.e thil'-; h'!a- occurrol to re that neon t!al Inleractiors may be nec-

essary for copulalion to occur In oomre, ,rhaps raiy, species. I know of

only one bit of evidence to support each a- a'loa if t'ore aro publlsaho re-

cor 'a along these line., I havo not foInd them. On two occastous I riacod an

accoustically acrivetnale and a res onslve female of ScMlderla teaensis in atennal

contact an a ta'ile ,t in order to observe coq,)ltiio close at hanm. The only

source of li~t was a nearby 7 1/2-watt rod light. On both occasions tU :ale

awr female circled1 oach other slowly, each foolUng the other with Its antennae.


After 1 2 minutes of such behavior the two separated. A few minute after

the separation the male produced his slow-pulon song, and the female answered

with a tick. In both Lnstancee the male turned Immedlately toward the female,

lowered the Intensity of his selds, and produced another slow-pulsed ong.

The female answered and he moved toward her rapidly. The semqenae of low

intensity, alow-pulsed song and female tick was repeated two more times and

finally the male went straight to the female, moving the last 7 8 taehes with-

out saiging. On both occasions a brief play of antannea was followed by the male's

turning around, raising his wings somewhat, aad the female mounting him from

the rear. Copulation resulted In the first observasion, but In the second obser-

vation I accideatly disturbed the pair and the male flew to the other side of the

darkened room. But almost immediately bw began to sing qgain and flew toward

the female when she anwered. After the male flew three times and did not

alight on the table with the female, I picked him up and placed him in siMad

contact with the female. Thil time they saienated emch ethe brelrly, the

male turned around, raised his wings, and the female meomted. Thee, even

though I had just handled the male they mated. Thees obserl Lom ertatly

suggest thatjn S. moanau at least, mating may be the end behavior of a sequenee

of patteras involving acoutIcal lter etwos.

Specifleity of Moustical comuqMlatlm 5ryem

The daerlptions of the songs of the different specie pre ted here clearly

show the specificity of the soang of each species, pWtienllrly the ses involved

in male-female interactions. This Is smetly what oe wauld esp*et, sRes aria-

tation and movement toward snd is the prtmlay Maehelamm of erngeleg maeis


and females together. Not only are the male songs specific for each species,

but the timing of the female response is charactcritlcally specific among those

species whero similarities between heterospecifie male songs are close enough

to cause confusion among females of the different species concerned. Thus, the

preventing of matings betwcon males and females of different species is equally

as significant a function of the acoustical communication systems as their role

in pair formation. Alexander (19I2a), while discussing criclket taxonomy, ad-

vises that "it is an e.xonsive proccJure to brinj together sexually responsive,

compatible males and females, and the mechanisms involved should be highly

specific and efficient."

In the case of those species in which the males produce lisps, only one

pair of species produce lisps which are nearly identical. A second pair produce

lisps which are similar in duration but different in other respects. A third pair,

involving one species of the second pair, produce lisps which are similar enough

to cause possible confusion at times among responsive females. The three

species-pairs will be discussed in order.

r.alos of Inscudderia strirata and of Scurderia furcata produce lisps of

77 msee. and 75 msec. (average values see data presented earlier for each

species). The frequency spectrum for each species is practically the came and

the females answer the lisps in normal acoustical interactions. For the most

part, adults of i stri.ata, having only one generation per year, occur in the

latter part of July and in August. S. furcata on the other hand, has two gen-

erations per year, the break between generations occurring during the peak

population of I. strizata. Nevertheloss, at times, adults of both species are


present at the same time, I have collected adults of S. ftroWa within a few feet

of Hypericun fasciculatum bashe containing l. L adults. No doubt, fe-

males in these situations sometimes answer beterospecific male lisps. Yet,

no deleterton results should come from such inability of females to discriminate.

Males ofL. jrtal have been shown to go to females answering other coaspoeiflo

male; perhaps males of flrcata do also. The timings of the female tiek re-

sponse in the two species are very specific ad nen-overlapping. A tick with

the wrong timing should elicit no orientation toward the tick by a male hearing

It. Under thee clrcumstaBces I ee no reason why there should be say eon-

fusion between these two species in nature.

The second speciel-pair with similar lisp are Malzumina models and

Microoentrum rhoabifollua~. The leog Usp of i. modSty averages 31 maec.

duration and the liep of M. rhombtfoliam Is 26 nmee. The range of variations

overlap at 28 29 mrec., so there may be some eanfulom. However, the

esnnds function entirely differently in the two specta Females of M. models

are attracted to high intensity edaspeelfc i1p, but Lea Ms of ipembiebem

are attracted to lew Intaulty conspelfic liaps. Differences in the spacing of

males ad females may remlt in females moving toward the wrong nund at

times, possibly sometime even raeuting in contact between hetroop.eific

males ad females.

Two striking differanee between the llaps of the two pieces may serve to

prevent wasteful ependitures of the time and mesg. M. mgjata lisps gener-

ally have much higher freqnsete (dominant frageAusiee 12 18 kc) tha. M

rhombtfolum lisep (8 12 ko). Them diferwies may be nogh to alew


discriminatory rcspo.ises on the part of the females. Evidence for this is the

fact that DM. modest females would not respond to recorded male lisps until

afterI had filtered all sounds below 15,000 cps. Actually this just increased

the relative intensities of hiMher froeqoncies over lower frequencies. More

work in tills respect should clear up those questions. The second big difference

between the lisps of these two species is in lisp rate. M. modest ong lisps

are delivered at a rate of bout one per second, whereas M. rlonbifoliuni

lisps are produced 2 3 see. apart. It may be that at least minimum refrac-

tory periods are necessary to elicit proper responses from the females. i.e. M.

rhlmbifoliun females may not be able to respond to lisps produced faster than

two per second.

The third species-pair with similar lisps is Scudderia cuncata and, again,

MIonto ina modest. The lisps of S. oaneata average 16 mIsec. duration, and _l.

modosta short lisps average 19 msoc. duration. The ranges, however, are broad-

ly ovorlapping, but the differences in female timing are characteristic and the

ranges of variation of the male timing do not overlap. The dominant frrceqcnclos

present in the two species broadly overlap. Also, the rate of lisp production is

slower in S. cuneata than In modest. These differences may contribute to

the females' ability to discriminate between the lisps of the two species.

The importance of refractory periods has been raised. Li the laboratory

where males of different species were lisping at the same time, I observed that

a male could seemingly lisp too rapidly to get any response from listening con-

spocific individuals. I have already mentioned that I cold rib my thumb across

the edge of a piece of paper and cvo:e tick responses from females of several


different species, depending on th E drtion of the "lmp" I produeed. At time,

I could eoke several suemestive ticks from certain females. If I speeded-up

the rate at w r h I "liaped ," :'thu fhtuls would sep maewing. They would

often reamm answering when I prodaeed "limp" at ti minimal Mb p eate char-

aoitMesti ftr the species iUw d. As important deflesin In this evidiene is

not knowing what the actual cdar tt of the "'lsp" I prodmed may hare been.

By tape reordiag the latem~tie thi should be an easy quetiae to amwer.

impourtaoe toof l trhb e Me

h species which isp, there Is a pomelblity that differeaess Ln teethatrike

rates within the I ap may serve o discriminatory mse far nldividuale rt pen-

Adve to oonpeMcflc lsps. I alyzed se ral lips from embh species ad fond

that even thegh the number of teEth struck per lisp in etch species wa clearly

different, those were only smalH difeneW e in toeethetrike a te eoept betwen

the liap of Itnedda i trIi and Seld a f~g al, which hrae lape with

idmtioal durationas ad frequmey spectrum. Th 1. strp Ulspp Mnalyzed

lad a toothtrike rate of 690 per sec. compared to 800 tootluetrle per meC. for s.

.ar Even with oace diateet dfforeees I deobt that teethttrtle rats will

be baund to be Important in allowed diMerLminatien between Hips, beesmm the

insect auditory system is net belieed to be abk to oncods such dclrekssm.

Furthermef, if toothltrike rat were ipeiata, I should ha aebtsated no re-

sponse from artificial HMps prodded by rauli ng my dlimb ages the ee nw of

a pieee of paper. I Yf then wes n0to tato IU rel. Finally, teshttrikie aHsI

Mten vary within single Haps and fau eer lisp to tie anl.

Increase In intensity cluring aonm s

Some species increase in intensity of found production toward the end of

certain parts of songs, of certain songs, or of song sequences. l.nown for this

are Scudderia texensis slw-pulsed song: gradual increase; S. furesta -

increase In successive lisps; and Amblycorynha uhleri and A. near ulleri -

increaso during Part I. One obvious advantage, at least in those species which

repeat the basic functional unit of sound successively, of Increasing the intensity

between successive units is to allow individuals successively farther froa the

sound producer to hear the sound. But why increase intensities within functional

units' 1 havo some scanty evidence that such increases toward the end ray be

functional and therefore potentially useful in species isolation. A recording of

one phrase of a S. texensis slow-pulsed song that I have does not exhibit any

ciun;e in intensity from beginning to end. Females never answered this re-

cording although they consistently answered a recorded phrase with similar

pulse rate but with Increasing intensity towar.1 the end. Similarly, texersis

females wold answer a "phrase" I proJucoJ by scrubbing my finger back and

forth across a piece of paper, only so long as my phrases gradually increased

in intensity. S. flrcata females were loss finicky but seemed to answer artificial

' lisps" (produced by my thumb) which terminally increased in intensity.

Complicatedness of sound pro;.iction

Complicatedness of sound pro actionn Is measured in terms of the terminal

movements involved in producing the different kinds of soin cf solitary situa-

tions. The simplest kind of katydid souid Is a lisp ani Is m.ade in a single open-

ing and closing of the tegmina. Ettlpaochlora couloniana is the only species for

which a single-pulsed lisp (probably made on the cle~ ng stoke of the tegmina)

is the only known solitary seend. eauM e ari mnakesus a two-palsed lisp

by producing soand on both the opening and closing trohe of a single opening

and closing of the totgina.

Another kind of simple singing involves repetition of one kind of tbeinal

movement in producing single phrase. Amblycerphka es*,t produces two-

pulsed (or three-pulsed) phrase, the pulse of which are idda1sl. The sepa-

rate phrase. of a Microcentru retierv song contain differing numbers of palees,

but the tegmins movements Lnvlved n iprodueolg each pulse of seh phrase are

identical. Tbh ticks in the song of Phrian i j a are identical, and within tick-

pairs the pulse rate is alwrys the eam "Ratter" Ai~Myeaoypba radifeM

belong in this greap.

Complieated sound production is of foar classes. Te first class involves

an increase in intnsity of each succesive pulse in a phrase, the pulses beiag

otherwise identical. Inereses of inteasity in successive pulse. re pirel that

the siger engage the stridalmtory appears t~ border in se- h swoesiave sound

producing stroke. Se ddWia g uarvalW lajaa makes only this kind of sound.

The second class of compliomednes of sound productive involves the pro-

ditein of drastically different kinds of ound from time to time in no fixed se-

quenee. Four of the species estdied are group hIre. Thy are Mkaroesantri

rheaiHtfolhka the Uctling sag ad the lisping song Swaeldeia ofal -

lsping song and song with pulled piswase; S. fhga Isplg song and sag

with pulsed phrases; and twml.i fast-pa~i d goa., slw-palsed song, Aad

ticking sonds. The different sends of each of Ithe pcief htae no eonetunt


relation one to another and each functions indepondo3tly. Tegminal movemcnts

are Identical in producing each pulse of a given kind of sound (except sometimes

in Intensity) but are different from one kind of sound to the next.

The third class of complication involves the producing of certain sounds

independently as in the second class, and producing other sounds (or the same

sounds that were produced independently) in a stereotyped pattern. Producers

of this class are Arethaea phalanglum lisping song and tick-lisp song (lisps

of the latter are different from lisps of the former); Inscudderia striata -

clicking sound and lisp-tic! song; and Turpilia rostrata lisping song, ticking

song, and lisp-tick song (the lisp and ticks of a lisp-tick phrase are identical

to isolated lisps and ticks but are produced in a regular sequence), and lisp-tick-

lisp song. This class of complicatedness bears special significance to recon-

struction of the evolution of complicated singing discussedd later).

The fourth and last class of complicatedness Involves producing two or

more idnds of solitary sound in a stereotyped sequence. Eight species discussed

in this paper are found in this class. They are Amblycorypha floridana -

regularly repeated sequence of cllck:s and buzzes; A. oblon~~ folia regular

sequence of one long pulse plus two short pulses; A. near rotundtfolla: "slow-

clicl:er" series of phrases containing a regular sequence of several short

pulses and one long pulse; A. near rotundifolia: ''slo clicker'" series of

phrases containing a regular sequence of two or three short pulses and one long

pulse; A. uhleri regular sequence of phrases, each with characteristic pulse

rates, pulse durations, ani pulse intensities; and lMontezumina modest -

regular sequence of a series of short lisps followed by a series of long lisps.


This classifloation may be modified as more types of complletedMeas are


MovementL involved In pair formation

Pair formation among the seven species of Pbanerepterlaae that I have

stadied in detail does not always involve the same Idnds of movement on the

part of males and females. In fact, the relative kinds mad amoants of move-

ment of males and females of different speoes can be put Into three categories.

In one category the male produces the female tick elicitor, the female ticks but

does not move, and the male moves all the distance to the female. Isseaddeoia

trIgata seems to beloeg in this group. Evidence to date places both SoadeMrl

furcata and S. cau ta here. The male goes to the female answering male lisps.

In the second category the male produce one sound which attracts the female

toward the male hut net l the way to the male. Then the male produose a

second sound which evokes ticks from the female. The female ticks attract

the male, which moves the final distance separating the male from the female.

Scudderia teaxnsls aad Miorocentrum rbembifolum belong in this category.

The third category includes theme species in which the male produces the female

tick elicitor, the female ticks, the male moves part-way to the female, then the

male prodmees a smeond sound whieh stimulates the female to mnv the final

distance toward the male. Monteauml e modest and Amblycory ha fleridama

belong in this group.

There Is some evidemee to allow us to predict which caegory Amblcerypnb

uhleri and A. near uhleri belong to. A. uhlri female preobalygo toward A.


uhlorl males from close range, whereas A. near uhleri females evidently do

not move at all, letting the conspeclfic males move the entire distance.

Amblycorypha oblong~olia may fit into either of the first two categories.

Before this species can be placed definitely, more must be known about the

movements of both the males and females.

It is very Interesting that no phanoroptcrine species is known to have fe-

males which are silent and which move all the way to the males. I predict that

at least one phaneropterine will be found to exhibit ouch behavior. Another

category which one would expect is one in which the males produce a female tick

eliciting sound, the females answer, and both the males and females move

toward each other until contact is made. The obvious disadvantage to this

kind of system is the difficulty of homing in on a moving source of sound. Since

the Phaneropterinae inhabit coarse vegetation and fly toward attracting con-

specific sounds, it is not likely that many, if any, cases of this category will

be found. Such a system could function well only where males and females

could move relatively slowly, walking or running, towarJ one another In a

straight line.

Evolution of complicated soun] production

Up to this point the central theme has been the description of different

kinds of sound within and between species and of how the sounds operate. Ad-

mittedly, very little has been found out, but I think enough is known to permit

some tentative conclusions as to how complicated singing behavior In the

Phaneropterlnae evolved.

When the ancestors of the Phaneroptcrinae diverged from the stock which


gave rise to other groups of Tettigeniidae, it was already a strong singer, prob-

ably producing a callingg" song which operated at a distance. Alexander (1962b)

surmises that the calling function (attraction of females to singing males) Is

so widespread and so similar among Grylldae and Tettigoalidae that It must have

appeared before these two families became separate evolutionary lines. Although

we are not concerned here with the origin of sound production, we must consider

how the female attracting function arose, for it could have been that sounds with

other functions arose in a similar manner.

Alexander (1962b) explains why the first acoustical signal of the ancestral

tsttigoatid was almost certainly a mediator of courtship, operating at cloe-range,

and he suggests that the calling function arose as an outgrowth of the original

courtship function (based on evidence he has collected in work with many species

of crickets). This would have involved "Increasing rhythmlcity, intensity, and

duration of the original courtship song beesae these characteristics enhanced

the courtship funetlon tself, through increasing constancy, range, sad re-

dundancy. Eventually, through just this kind of clnge, this song must have

become operative at such distances that it was sometimes advantageous for the a

male to be triggered into trldalatlon without contact with the female, and some-

times advantagems for the female to be attracted by hearing the sound when

she was not otherwise in contact with the male. In this way the calling function,

in the approximate form that it assumes today, could have evolved." This line

of reasoning seems valid. The next step would be for either the males "to

develop structurally different signals, with slightly different effects, for the

two situations" (close-rage or at a distance or any other two different ritdatils-);


"or (perhaps originally) for the female to respond differently to the original signal

that served both calling and courtship, depending on whether or not she was in

contact with the male through senses other than auditory. In all likelihood these

changes did take place in many cases, with the resulting development of two

separate signals." Thus, in today's Gryllidae there are two characteristic sig-

nals between males and females, functional at long- and close-range respectively,

or if only one signal is present it functions in the female attracting capacity,

the courting function being effected by females feeding upon dorsal glands of

the male.

At this point it is well to interject that in most crickets only one sound,

the female-attractor, is functional at long range. Courtship activities Involve

males and females in contact through senses other than auditory e.g. tactile,

olfactory, visual. In the Phaneroptcrinae there Is no known instance of any

sound production characteristic of males and females in intimate contact

through whatever senses may be involved. Thus, all sounds of these katydids

operate from a distance with the exception of sounds produced when males come

into physical contact with one another. This is in another context, however.

We are at present considering male-female interactions.

I have stated that the phaneropterine ancestors, after having become a

separate evolutionary line, probably had a functional long-range female attract-

ing song. But I doubt that these early katydids made sounds which served a

courtship function as exists in many moJern Gryllldao. If they did, there should

be at least a few species today which retained the behavior. I really do not see

why there are not some, even if the trait has secondarily evolved. More likely,


this subfamily diverged from other tettigoalid groups in whioh feeding upon dorsal

glands of the male by the female was the principal courtship activity. nl fact, the

original Tettgnalidae may have diverged from a tettigeulid anesetor in which

dorsal feeding was characteristic. In the few cases where mating has been ob-

served among Phmeropterinae, "licking" of the male doreum by the females has

always been observed.

Whatever the ancestry, the point is that the original Phaseroptertnae prob-

ably did not make a courtship sound. Somewhere very early in phnasropterine

divergence courtship activitie may have begun to include wing-Jerkldg, or some

such activity, on the part of both malee and females or simply on the part of the

female to the calling song when In Intimate contact with the male. It is certainly

reasonable to suppose that females have the ability, or could evolsvethe ability,

to move their tegmlna in the same mamer as the males do is stridulation. Indeed,

Huber (1962) has shown that much of the nervous an-l muscular system necessary

for stridulation is contained in the female (from Aleander, 1962b). Wing-Jerk-

ing, or other comparable signals, could have fuitioaed initially as a visual

stimulus, but almost assuredly sound would have ben involved the males with

their etrldulatory structures and the females simply by iaeidaetally rubbing their

wings together. The courtship could have involved alterations of signals by the

males and females, the females "anewatlIg" only in response to the male signal,

acoustical or otherwise. This kiad of courtship activity could Involve more and

more sound In the signals and allow the two sexes to qippCt jnd move toward '

one another from close range without having seen one another. Once say kind

of orientation by males toward ounmd produced by females took plae even


at very close rare the durio w%- opon fuI gjrlaj r and grcatLr s~para!iol.

he necessary prercq.tdsiLo ltr me eparation (as in Alexrnder's Jisacasuln of

the e'olutiou of the calling song) would have bcn for it to have been an a:.vantage

for the males to be triggered into projicing the female itfng-jcr.; cllcitng soutLi,

rsllting in hLo femalo ticd as it is herein termed, anj fur it to have bee an ad-

v.La e for the cfmales to answer the sound before having had any other-'OLh -

acoustical contact with the males. buch seisctive a'vantago is obvious a

sexual responsive male, if separated from an urmknown, .seoally respon3l'v

female by a relatively short distance could by-pas producing the usual

attractor, which may continue for long perlo.s by solitary males, andi tLiOid-

lately learn of the femalo'o presence and proceeJ to court her. Retantio. of

the female attractor is Obviously a:vantaeoura. From here on all sorts of

separ pat ways could be U res~ltia in malea roving towar.L femalos

and females mving toward males n ral ifrent conte t, giving ua the

categories outlinedd iunier the preceding subheading.

row did the tic!in s0smua sounds t3at fimctioa in male-male iter-

nations evolve? Alexa.n:er (b1J2b) postulats that aggressive sounds in

crickets ttLue involved in male-male irteractions appoa nr as outgroivtW

of the calling function. .As is that the calling: sog ad agg-resaive

sigala of species which havo both in their ac-astical reportoir. are very

similar, and that the calling song functions like tbe aggressive signals. although

to a lesser extent, in interactions between males. Among the Pliaeropterinaa

the involved in male-female lateractioan bear little reseibblao to the

somdls proJi.ced in interactions between males except the ticking sounds of


Microcentrum rhombifolium, bht here the resemblance is superficial; rhombifol-

lum's ticks are regularly produeed in series of 15 34 mingle toothetrikee in-

volving a single, slow closing of the tegmina. Ticks Involved in male-male

Interactions usually vary in the intensity from one pulse to the next, are very

erratic, and are delivered at rates dependent tpoq the intensity of the stimulus

causing their production. A single tick usually involve a complete wiagstroke.

The first Licking probably arose In situations where male came into physloal

contact with one another. Evidence for this is that more species produce ticks

in this situation than In any other. Generally males contact one another phys-

ically when they are mutually attracted toward a female. Usually each males

push eseh ether around with their front legs. In esch a situation any slight move-

ments of the tegmina may have been the result of the excited state of one of the

males Involved. Such behavior would have been advantageous if the terminal

movements tended to repel the other male to any degree whatever. Males not

in physical contact could continue producing female attracting sounds or female

Lick elicitors. The genes which contributed to the tgmina-flipping and the re-

action to it, whatever movement may have been involved, weuld have tended to

have been conserved more often than not. To have mnbeqmently involved

sound in this case ticks in each terminal movements Oes almost without

saying. Onee ticks were made during physical eocourters as males moved

toward females, It would have been a decided advantage for certain males if

they sometimes produced ticks after female-oriented seage of other males. Both

the male which ticked and the male repelled by the ticks would rave beflsMad

by not having come Into contact and coieteqtetly herlag wasted time by the