Temporal variation in natural and sexual selection of male calling behavior in the field cricket Gryllus Rubens

MISSING IMAGE

Material Information

Title:
Temporal variation in natural and sexual selection of male calling behavior in the field cricket Gryllus Rubens
Physical Description:
vi, 89 leaves : ill. ; 29 cm.
Language:
English
Creator:
Velez, Manuel J
Publication Date:

Subjects

Subjects / Keywords:
Crickets -- Behavior   ( lcsh )
Sexual selection in animals   ( lcsh )
Natural selection   ( lcsh )
Zoology thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Zoology -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 2004.
Bibliography:
Includes bibliographical references.
Statement of Responsibility:
by Manuel J. Velez.
General Note:
Printout.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 003101104
sobekcm - AA00004967_00001
System ID:
AA00004967:00001

Full Text










TEMPORAL VARIATION IN NATURAL AND SEXUAL SELECTION OF MALE
CALLING BEHAVIOR IN THE FIELD CRICKET GRYLLUS RUBENS













By

MANUEL J. VELEZ


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

UNIVERSITY OF FLORIDA


2004














ACKNOWLEDGMENTS

I am indebted to my advisor, H. Jane Brockmann, for her invaluable support

in the design and analysis of this research project. I am also thankful to Colette

M. St. Mary, Benjamin Bolker, Marta L. Wayne, and Thomas J. Walker for their

helpful feedback on my research. My deepest gratitude goes to Laura Sirot,

Suhel Quader, Kavita Isravan, Rebecca Hale, and other colleagues at the

Department of Zoology for their support in the development and analysis of this

project. I am also thankful to George Casella for giving me statistical advice.

My deepest gratitude goes to the University of Florida Beef Teaching Unit

and Environmental Landscape Horticulture Education Lab for allowing me to

conduct field research in their pastures.

This work was funded by an NSF Graduate Predoctoral fellowship. My PhD

committee consisted of Drs. H. Jane Brockmann, Colette M. St. Mary, Benjamin

Bolker, Marta L. Wayne, and Thomas J. Walker.














TABLE OF CONTENTS
Paae

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

ABSTRACT ....................................................... ....................................

CHAPTERS

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

2 VARIABLE SELECTION FOR MALE SONG IN THE
FIELD CRICKET GRYLLUS RUBENS .....................................5.......5

Introduction .......................................................................................... 5
Methods ..................................................... .............................. 9
Study Species............................... ...... ................ 9
Fly Abundance and Activity............. ................................ 10
Field Recordings of Individual Males................................. 10
Night Surveys .................................................................. 12
Results........................................................ 13
Fly Abundance and Activity ................................................13
Field Recordings ................................................ 14
Night Surveys...................................................................14
Discussion ................................................................................... 15

3 SEASONAL VARIATION IN FEMALE PHONOTAXIS
TO MALE CALLING SONG OF THE FIELD CRICKET
GRYLLUS RUBENS................................................................... 26

Introduction ...............................................................................26
Methods.................................................................................. 29
Phonotactic Experiment ......................................................... 29
Source of Male Calls............................................ 32
Fly Activity ...................................................................... 33
Results........................................................ 33
Phonotactic Experiment....................................... 33
Fly Activity ...................................................................... 34
Discussion ..............................................................................35









4 VARIATION IN SEXUAL SELECTION FOR MALE
SONG IN THE FIELD CRICKET
GRYLLUS RUBENS ................................................................. ...46

Introduction ............................................... ..... 46
Methods ............................................................................. 50
Pitfall Trap Experiment .......................................................50
Source of Male Calls........................................... 53
Data Analysis ................................................................... 53
Results............................................................................. ....54
Discussion .............................................................................. .. 55

5 GENERAL MODEL: A TEST OF THE FACTORS
AFFECTING MALE CALLING DURATION IN THE FIELD
CRICKET GRYLLUS RUBENS ..................................................... 63

Factors Affecting Male Calling Duration ......................................... 63
Model: Predicting Seasonal Patterns in Male Calling Duration..............64
Male Calling Strategies ......................................................64
Calculating Fitness of Male Calling Strategies ........................65
Seasonal Changes in Mating Benefit (M)...............................66
Seasonal Changes in Probability of Escaping Fly
Parasitism (E) ............................................................... 67
Model Results: Seasonal Changes in Fitness of
Male Calling Strategies. ................................................ 69
Comparison of Optimal Seasonal Strategies with
Observed Calling Durations............................... .......69
Variations of the Model.......................................................71
Results of Model Variations............................................. 71
Relative Importance of Natural versus Sexual Selection ..........72
Implications of the Model...................... ......73

LIST OF REFERENCES........................................ .....................82

BIOGRAPHICAL SKETCH ................................................................... 89














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

TEMPORAL VARIATION IN NATURAL AND SEXUAL SELECTION OF MALE
CALLING BEHAVIOR IN THE FIELD CRICKET GRYLLUS RUBENS

By

Manuel J. V61ez

August 2004

Chair: H. Jane Brockmann
Major Department: Zoology

A primary goal of sexual selection theory is to understand how natural and

sexual selection interact to favor variation in sexually-selected traits. The goal of

this dissertation is to examine variation in male calling behavior of the field cricket

Gryllus rubens and how temporal changes in natural and sexual selection may

favor this variation. The results of this dissertation are important because they

show that temporal variation in natural and sexual selection may favor variation

in male sexually-selected traits.

Male calling song in G. rubens attracts females. It also attracts gravid

females of the parasitoid fly Ormia ochracea. These female flies deposit larvae

on and around calling males, killing infected crickets within 7-10 days. In northem

Florida, 0. ochracea is abundant in the fall and rare in the spring, and fly

parasitism rates are higher in the fall than in the spring. This temporal variation in

natural selection could favor different male songs in the fall than in the spring. I








evaluated this possibility by examining seasonal variation in male calling

behavior. I found that fewer males sing each night in the fall than in the spring

and that fewer males sing at times when parasitoid flies are most active.

Because seasonal changes in parasitoid fly abundance may change the

costs and benefit of choice, fall and spring females may behave differently

towards calling males. I evaluated this hypothesis by observing the phonotactic

behavior of fall and spring females. I found that fall females are less attracted to

male songs than spring females at times when flies are most active.

Seasonal changes in sexual selection may change the seasonal benefits

males derive from singing. I explored this hypothesis by examining whether the

benefits of singing change seasonally and whether such changes are correlated

with seasonal differences in male calling duration. The results of a playback

experiment show that long duration calls attract more females in the fall, where

call durations are long, than in the spring, where call durations are short.

Temporal changes in natural and sexual selection may favor seasonal

variation in male G. rubens calling behavior. To evaluate how such fluctuating

selection may affect seasonal patterns in calling duration, I developed a model

evaluating the fitness of different calling strategies in the fall and spring given

seasonal changes in natural and sexual selection. By varying these selective

forces independently and comparing the fitness of calling strategies to observed

calling durations, I found that temporal changes in sexual selection may have a

greater effect on calling duration than changes in natural selection.













CHAPTER 1
GENERAL INTRODUCTION


A primary goal of sexual selection theory is to understand how natural and

sexual selection interact to favor variation in male sexually-selected traits

(Andersson 1994). Crickets are ideal subjects for the study of sexually selected

traits for several reasons. First, males produce calling songs to attract females

for mating, and females select mates based on the quality of their songs (for

review see Loher & Dambach 1989). This means that male song is likely to be

influenced by sexual selection. Second, many song characters are highly

variable among males (Cade & Wyatt 1984, Cade 1991, Walker 1998,2000).

Third, song characters can be manipulated independently from male phenotype

by tape recording male calling songs and editing the recordings. This provides a

means of isolating different song characters and testing for their separate effects

on female choice (Hedrick 1986). Fourth, females choose males overtly: they

approach males and mount them. This female behavior offers the opportunity to

separate the effects of female choice from male-male competition. Additionally,

because females approach speakers broadcasting male songs in the same way

they would approach calling males, female choice can be studied without the

problem of males adjusting their song in response to females, and females

modifying their preference in response to changes in male song. Collectively,

these characteristics make crickets model subjects for studies on sexual








selection (Cade 1979, Crankshaw 1979, Boake 1984, Hedrick 1986 & 1988,

Simmons 1988).

The goal of this dissertation is to examine variation in male calling

behavior of the field cricket Gryllus rubens and how temporal changes in natural

and sexual selection may favor this variation. The results of this dissertation are

important because they show that temporal variation in natural and sexual

selection may favor variation in male sexually-selected traits.

The field cricket G. rubens is well suited for studying the effects of natural

selection on calling song traits. In addition to attracting females, the calling song

of male G. rubens also attracts gravid females of the parasitoid fly Ormia

ochracea (Walker 1986, 1993). These female flies deposit larvae on and around

calling males, killing infected crickets within 7-10 days. In northern Florida, 0.

ochracea is abundant in the fall and rare in the spring (Walker 1986), and fly

parasitism rates are higher in the fall than in the spring (Walker & Wineriter

1991). This seasonal variation in the costs of calling suggests that natural

selection on male calling song varies temporally. This temporal variation in

natural selection can result in different calling songs in different seasons. I

examine this hypothesis in the second chapter of this dissertation by exploring

whether male G. rubens exhibit seasonal variation in calling behavior and

whether this variation is associated with seasonal changes in parasitoid fly

abundance.

Gryllus rubens is also well suited for studying the effects of temporal

variation in sexual selection. Even though females do not sing, Walker and








Wineriter (1991) have shown that females can also be parasitized. If females are

infected as they approach calling males, then the costs of mate choice may be

higher in the fall when flies are abundant than in the spring when flies are rare.

These seasonal changes in the costs of mate choice may cause females to be

less attracted to male calling song in the fall than in the spring to reduce the risk

of fly parasitism. This temporal variation in female behavior may alter the

strength of sexual selection seasonally. I evaluate this hypothesis in the third

chapter of this dissertation by examining female attraction to male calling song at

different times of the year.

Seasonal variation in sexual selection may alter the seasonal benefits that

males derive from singing. As a result, such variation may favor seasonal

changes in male calling behavior. I evaluate this hypothesis in the fourth chapter

of this dissertation by examining whether the benefits of singing change

seasonally and whether such changes are correlated with seasonal differences in

male calling duration.

I conclude this dissertation with a discussion of how temporal changes in

natural and sexual selection should affect seasonal variation in male G. rubens

calling duration. I test these effects by calculating the fitness of different calling

strategies in the fall and spring given seasonal changes in natural and sexual

selection. By varying natural and sexual selection independently, I estimate the

relative importance of natural versus sexual selection in favoring variation in male

calling duration.





4

By studying male calling behavior and the benefits that males derive from

singing when flies are rare and abundant, this dissertation examines how

temporal variation in natural and sexual selection may interact to favor variation

in male sexually-selected traits.












CHAPTER 2
VARIABLE SELECTION FOR MALE SONG IN THE
FIELD CRICKET GRYLLUS RUBENS


Introduction


Conspicuous male traits such as bright colors, large size, or elaborate

behavior evolve through the joint action of sexual and natural selection

(Andersson 1994). Sexual selection favors males with conspicuous traits since

these traits are advantageous in mate acquisition through either female choice or

male-male competition (Darwin 1871). Natural selection, on the other hand, acts

against males with conspicuous traits if these traits lower male survival.

Conspicuous traits may lower survival by being energetically costly or by making

males more vulnerable to predators. Therefore, conspicuous male traits evolve

depending on the relative strength of sexual and natural selection. This

hypothesis predicts that when the costs of natural selection increase relative to

the benefits of sexual selection, then males should become less conspicuous. In

this study, I evaluate this prediction by examining the male calling behavior of the

field cricket Gryllus rubens in two seasons that differ in the risk of parasitism from

acoustically orienting flies.

Male crickets produce calling songs by rubbing their raised forewings

together, which attracts conspecific females for mating (Loher & Dambach 1989).

In some species, the same calling song also attracts parasitoid flies which

deposit larvae on and around calling males, killing infected crickets within 7-10








days (Cade 1975, Walker 1986, 1993). This fly parasitism risk has been used to

explain variation between species in the nightly calling duration of male field

crickets (Orthoptera; Gryllidae). Cade and Wyatt (1984) found that male G.

texensis, which is parasitized, spend less time calling per night than male G.

veletis, G. pennsylvanicus, and Teleogryllus africanus, which are not parasitized

by phonotactic flies. They argue that fly parasitism has selected against long

nightly calling in species that are parasitized. They also found that the distribution

of male G. texensis nightly calling duration is highly skewed because some

males do not call. It is thought that males that do not call or males that sing less

search for females so there is some expectation of success for these males,

particularly when female densities are high. Cade and Wyatt (1984) attribute the

skew distribution of nightly calling duration in G. texensis to the higher rates of

parasitism in this species.

However, species differences in time spent calling per night may be due to

species differences in the energetic costs of singing. The song of G. texensis is

made up of sound pulses organized into trills. In contrast, the songs of G. veletis

and G. pennsylvanicus are composed of sound pulses grouped into chirps with

quiet periods between them. Because trilling songs may be more energetically

costly than chirping songs (Prestwich & Walker 1981, Prestwich 1994),

differences in time spent calling per night between the trilling G. texensis and the

chirping G. veletis and G. pennsylvanicus may be explained by energetic

differences in song production. Collectively, these findings cannot separate the








effects of fly parasitism risk and energetic costs of calling on nightly calling

duration in male crickets.

Fly parasitism risk has also been used to explain variation in the timing of

male calling in crickets. French and Cade (1987) found that male G. texensis call

more around dawn than in the early evening. They hypothesized that dawn

calling is common in G. texensis because parasitoid flies are less active at dawn.

However, G. veletis and G. pennsylvanicus also call more around dawn (French

& Cade 1987), and they are not infected by parasitoid flies. Furthermore, in all

three species, male calling around dawn coincided with increased female

movement and mating activities (French & Cade 1987). Therefore, it is not clear

whether male G. texensis call more around dawn because flies are less active at

that time or because females are more sexually receptive around dawn.

Testing the effect of fly parasitism risk on nightly calling duration and

timing of calling in male crickets under laboratory conditions is difficult because it

is possible that crickets do not respond directly to flies but rather to

environmental cues associated with fly activity. If so, then examining male calling

behavior in an experimental arena with and without flies will not reveal the effect

of parasitoid flies on behavior. One solution to this problem is to study male

calling behavior in populations with and without parasitoid flies. This is the

approach that was used by Zuk et al. (1993) and Simmons et al. (2001).

However, any correlation between male calling behavior and fly abundance may

also be explained by environmental or other differences between the two

populations (for example, populations without flies may be at higher latitudes








than those with flies). This limitation of working with geographically distinct

populations is overcome when studying a population that is exposed to seasonal

changes in fly abundance. This is the case for the field cricket G. rubens. The

calling song of this species attracts gravid females of the parasitoid fly Ormia

ochracea (Walker 1986, 1993). In northern Florida, 0. ochracea is abundant in

the fall and rare in the spring (Walker 1986), and fly parasitism rates are higher in

the fall than in the spring (10% versus 0%; Walker & Wineriter 1991). These

seasonal differences in fly abundance and rates of parasitism suggest that the

costs of calling are higher in the fall than in the spring. This seasonal variation in

natural selection serves as a natural experiment to examine the effect of fly

parasitism risk on nightly calling duration, number of males calling and timing of

calling in male crickets.

I hypothesize that seasonal changes in natural selection favor seasonal

changes in male calling behavior in G. rubens. Accordingly, I made the following

predictions. First, I predicted that fewer males would call in the fall than in the

spring because some males do not sing (personal observation) and

experimentally muted males have lower rates of fly parasitism than non-muted

males under high fly densities (Walker & Wineriter 1991). Second, I expected

that males should spend less time calling per night in the fall than in the spring

because the number of parasitoid flies attracted to broadcasts of cricket songs

increases with broadcast duration (Cade et al. 1996). Third, I expected that fewer

males would call in the early evening in the fall than in the spring because

parasitoid fly activity is highest in the early evening (Cade et al. 1996). I tested








these predictions by recording the calling duration of individual males under field

conditions and conducting nightly surveys of singing males.

Finding that variation in male calling behavior in G. rubens is correlated

with seasonal changes in natural selection from the risk of fly parasitism would

be important because it would help us understand how males deal with the

conflicting demands of attracting females while avoiding parasitism. Because

males of many animal species have mating signals that attract both females and

predators and/or parasites (reviewed by Burk 1982 and Magnhagen 1991), the

results of this study may have important implications for many systems in which

males or females attract mates with conspicuous displays.



Methods


Study Species


The field cricket Gryllus rubens is found in lawns, pastures, and along

roadsides in southeastern United States (Alexander 1957). Unlike most other

species of field crickets, male G. rubens produce a trilling song in which sound

pulses are repeated and grouped into long pulse trains (Doherty & Callos 1991).

This trilling song attracts females who approach and mount males. It also attracts

gravid females of the parasitoid fly Ormia ochracea (Walker 1986, 1993).








Fly Abundance and Activity


To confirm that 0. ochracea flies are more abundant in the fall than in the

spring and that flies are most active in the early evening, I monitored noctumal

changes in fly activity in a pasture at the University of Florida Beef Teaching Unit,

Alachua Co., FL that was 1-km from the transects used in the night surveys. To

monitor fly activity, I used a slit trap (Walker 1989) with an electronic sound

synthesizer that broadcasts calls of G. rubens (4763 Hz carrier frequency, 50.0

pulses per second, 50% duty cycle; continuous trill at 106 dB, 15 cm above the

speaker). I activated the synthesizer an hour before sunset and counted the

number of flies captured at the trap 3, 7, 11, and 13 h after sunset. I carried out

this procedure in the fall from 20 September to 17 November 2000 (N=8) and

spring from 3 April to 23 May 2001 (N=6).



Field Recordings of Individual Males


To study seasonal changes in the number of calling males and time spent

calling by males per night, I recorded the calling duration of individual males

under field conditions. For these recordings, I collected adult males from a field at

the University of Florida Environmental Landscape Horticulture Education Lab,

Alachua Co., FL, using a sound trap with an electronic sound synthesizer

(Walker 1986). Males were weighed to the nearest 0.001 g and placed

individually in square plastic containers measuring 13 x 13 x 9 cm (I x w x h) with

wire screen lids for 10 days to ascertain that they were not parasitized by 0.

ochracea. Each container had an inch of sand, an inverted egg carton shelter,








and Purina Cat Chow and water ad lib. All crickets were held under ambient light

and temperature in a greenhouse less than 1 km from the collection site. After 10

days, I transferred the containers to the field and laid them out in a straight line

keeping a distance of 15 m between containers. This distance was chosen so

that males could not hear the singing of other males being recorded. Each plastic

container was set in a square aluminum pan measuring 20 x 20 x 4 cm (I x w x h)

with an inch of water to prevent fire ants from invading the containers. Next to

each container, I placed an Optimus model CTR-115 voice-activated cassette

tape recorder. I switched on all tape recorders an hour before sunset and allowed

them to run for 16 h. All tape recorders were checked hourly to replace any tapes

that were full. Two to four males were recorded each night. I electronically

monitored 20 males over nine nights in the spring (April/May 2001) and 23 males

over nine nights in the fall (September/October 2001). Each male was monitored

only once. Fall and spring field recordings did not differ in the minimum

temperature recorded each night (Meanfa1i SE = 15.441.35 C, Meanspmng SE

= 15.591.64 C, P=0.943). There was also no seasonal difference in the

maximum temperature recorded each night (Meanfa i SE = 30.060.86 C,

Meansping SE = 28.161.37 C, P=0.260).

I calculated the nightly calling duration for each male by listening to the

tapes with a stopwatch. If the male stopped calling for more than 2 sec, then the

stopwatch was stopped until the male resumed calling. The nightly calling

duration was the total time spent calling during the 16 h that males were

monitored.








Night Surveys


I studied seasonal changes in the number of calling males and the timing

of calling by conducting night surveys of singing males in a pasture used to graze

cows at the University of Florida Beef Teaching Unit. This pasture was divided

into 10 130 x 20 (I x w) m parallel transects. A night survey consisted of a 16 h

sampling period beginning an hour before sunset in which I walked in a straight

line through the center of each of five transects randomly selected at 4 h intervals

counting the number of calling males. At the start of each 4 h interval, I recorded

the temperature in each transect 2 cm above the soil. On a given night, the same

five transects were surveyed four times during the 16 h survey period. Each

transect was surveyed two to three nights each season. Fall surveys were

conducted from 23 September to 17 November 2000 and spring surveys from 3

April to 23 May 2001 because these are the dates when flies were shown to be

abundant and rare, respectively (Walker 1986).

To test the prediction that fewer males will call in the early evening in the

fall than in the spring, I performed a Repeated Measures ANOVA on the number

of calling males in the night surveys. This variable was transformed to achieve

normality by adding 0.375 to the variable and taking the square root of this sum

(Ott 1993). This transformation achieved normality and validated the sphericity

assumption (variance-covariance matrix of dependent variable is circular in form)

of the Repeated Measures ANOVA. The terms season (fall and spring) and time

(4 h interval) were used as factors. The average temperature across the four time

intervals in each transect was used as a covariate. Only one night survey per








transect was used in this analysis. For each transect, I chose the first night

survey where the average temperature recorded in the four time intervals fell

between 15 and 25 C. This restriction was used to limit the effect of extreme

temperatures on male calling behavior. For this analysis, I assumed that

transects were independent between seasons because the lifespan of males is

likely to be shorter than the time elapsed between the fall and spring night

surveys.

Data were analyzed using SYSTAT 6.0 and SPSS 11.5. Results are

presented as means standard error.



Results


Fly Abundance and Activity


Ormia ochracea flies were much more common in the fall than the spring.

In the fall, a mean (SE) of 17.6 6.2 flies/ night were captured at the slit trap

(N=8). In contrast, no flies were ever captured in the spring (N=6). In the fall, the

number of 0. ochracea flies captured was highest in the first few hours after

sunset (Fig. 2-1). Flight activity then steadily declined throughout the remainder

of the night but showed a small increase around dawn (i.e., 13 hours after

sunset).








Field Recordings


A higher proportion of males sang in the spring, 15/20 (75%), than in the

fall, 10/23 (43%) (Fisher's exact test: P=0.030). Overall, fall males did not differ in

nightly calling duration (44.5 min 13.8) from spring males (31.8 min 9.1) (t-test,

Nfati=23, Nspnng=20, t=0.765, P=0.449). However, when males that did not sing at

any time during the 16 h test period were excluded from this analysis, then fall

males that sang spent more time calling (102.3 min 20.3) than spring males

that sang (42.4 min 10.9) (Fig. 2-2; t-test, Nfa,=10, Nspng=15, t=2.60, P=0.020).

Although fall males were heavier than spring males (Fig. 2-3; t-test, Nifal=23,

Nspnng=20, t=3.693, P=0.001), male mass did not have a significant effect on

nightly calling duration for males that sang in either the fall (Fig. 2-4; N=10,

R2=0.090, P=0.399) or the spring (Fig. 2-4; N=15, R2=0.002, P=0.885).




Night Surveys


More males were observed singing in the spring than in the fall across all

sampling periods in all surveys (Fig. 2-5; Repeated Measures ANOVA, season,

F1,18=8.769, P=0.008). Time of day had a significant effect on the number of

calling males (Fig. 2-6; Repeated Measures ANOVA, 4 h interval, F3,54=7.319,

P<0.001). More males sang in the first few hours after sunset in the spring than

in the fall (Fig. 2-6; Repeated Measures ANOVA, time*season, F3,54=3.776,

P=0.016). These results came from a model that excluded temperature as a

covariate. Temperature was excluded because a preliminary model with season








and time as factors and temperature as a covariate showed that temperature did

not have a significant effect on the number of calling males (Repeated Measures

ANOVA, temperature, F1,17=0.058, P=0.812).



Discussion


This study has shown that seasonal changes in male G. rubens calling

behavior are correlated with seasonal changes in the risk of fly parasitism. The

probability of encountering parasitoid flies on a given night is higher in the fall

than in the spring. In the fall, this probability is highest in the early evening and

then declines steadily throughout the night. This suggests that the natural

selection costs of calling per night are higher in the fall than in the spring, and

that these costs decrease within a night in the fall. As predicted by these

seasonal changes in the risk of fly parasitism, a lower proportion of males sang in

the fall than in the spring and fewer males sang in the early evening in the fall

than in the spring. These findings are consistent with the hypothesis that natural

selection affects the evolution of sexually selected male traits. When the natural

selection costs of conspicuous traits increase relative to the benefits, then males

behave less conspicuously.

However, other factors besides the risk of fly parasitism change

seasonally and may explain the seasonal variation in the number of calling

males. These factors include differences in density, differences in the proportion

of parasitized males, and differences in male size between seasons. Differences

between seasons in female behavior will be discussed in chapter 3. Fewer








males may have sung in the fall than in the spring because fewer males were

present. Several lines of evidence argue against this possibility. First, Veazey et

al. (1976) found that G. rubens densities in northern Florida are higher in the fall

than in the spring. Second, my field recordings of individual males showed that a

smaller proportion of males sang in the fall than in the spring. Alternatively, fewer

males may have sung in the fall than in the spring because a higher proportion of

males were parasitized. Parasites have been shown to decrease the ability of

males to sing (Cade 1984, Kolluru et al. 2002). However, my field recordings

show that this cannot entirely explain the pattern because healthy fall and spring

males differed in calling behavior. Finally, fewer males may have sung in the fall

than in the spring because spring males were larger than fall males. This is

possible because male size has been shown to be correlated with calling

duration in sagebrush and ground crickets (Sakaluk & Snedden 1990, Forrest et

al. 1991). However, this cannot be an explanation for the seasonal differences in

the number of calling males because fall males were heavier than spring males.

The combined evidence suggests that fewer males sang in the fall than in the

spring because the risk of fly parasitism is higher in the fall than in the spring.

More than half of the males that were recorded in the fall did not sing over

the 16 h period they were monitored. If females approach calling males, how do

these silent males get mated? It is likely that these males walk or fly around

searching for females. This mating strategy has been described in several

species of field crickets (French & Cade 1989, Hissmann 1990, Zuk et al. 1993,

Hack 1998). Cade (1979) proposed that searching may be an alternative mating








tactic. He hypothesized that males using this tactic may have lower mating

success on any given day than calling males. However, because searching

males may also have lower rates of fly parasitism, they may have higher survival

and longer lifespans than calling males. This type of alternative mating tactic has

been described for guppies. Males usually display their bright color patterns to

females by performing sigmoid displays in which males quiver their bodies in an

S-shape (Baerends et al. 1955). Some males, however, do not perform these

displays and instead attempt a sneak copulation by performing a gonopodial

thrust. This mating tactic may be favored by natural selection because it is less

conspicuous to predators than sigmoid displays (Endler 1987). This hypothesis

has been supported by studies showing that males switch from sigmoid displays

to gonopodial thrusts when the risk of predation is high (Endler 1987, Magurran &

Seghers 1990, Reynolds et al. 1993, Godin 1995). Taken together, studies on

crickets and guppies suggest that when the natural selection costs of

conspicuous male traits increase, males may use alternative mating tactics.

Males using these tactics are favored because even though they may have

reduced mating success, they may have higher survival when compared to

males using the more conspicuous tactic. Additionally, the success of searching

males may increase as density increases.

As predicted, I found that more male G. rubens sang in the early evening

in the spring than in the fall when flies were more active. French and Cade

(1987) studied the timing of male calling in G. texensis, G. veletis, and G.

pennsylvanicus. Their results showed that male calling increased at dawn in field








populations of all three species. Male G. rubens in this study did not increase

calling around dawn. One possible explanation for this finding is that the small

surge of parasitoid flies around dawn at my study site may have increased the

natural selection costs of calling at those times. Consistent with this interpretation

is the fact that the populations of G. texensis that French and Cade (1987)

studied did not experience a surge in fly activity around dawn, and the

populations of G. veletis and G. pennsylvanicus were not parasitized by 0.

ochracea (Cade & Wyatt 1984). Collectively, these studies suggest that the.

timing of male calling in crickets has responded to natural selection from

parasitoid flies.

I predicted that males should spend less time calling per night in the fall

than in the spring because the number of parasitoid flies attracted to broadcasts

of cricket songs increased with broadcast duration (this study, Cade et al. 1996).

The field recordings of individual males found the opposite result: fall males that

sang spent more time calling per night than spring males that sang. Bertram

(2002) also found the same seasonal pattern in male G. texensis. One factor that

may explain these results is density. As density increases, males are expected to

call less and search more because the gains from searching increase (Alexander

1961 & 1975, Otte 1977, Cade 1979). This effect has been demonstrated in

experimental studies by showing that male calling decreases with density (Cade

& Wyatt 1984, Cade & Cade 1992). However, seasonal differences in density

cannot explain the seasonal differences in nightly calling duration from this study





19


nor the Bertram (2002) study because cricket densities at both sites were higher

in the fall than in the spring (Veazey et al. 1976, Bertram 2002).

One possible explanation for the seasonal differences in nightly calling

duration in G. rubens and G. texensis is sexual selection. If sexual selection for

nightly calling duration is stronger in the fall than in the spring, then this seasonal

difference in mating benefit may explain why fall males that sang spent more

time calling than spring males that sang. I evaluate this hypothesis in chapter 4.

















Q.



CL
4-
U')




00








S am

-r-r




E

E






/.-m 2




00.0
p0~ n~deD sa!U o jeqwnu ue0 0





LL 8







21









CN


z m S8
-
- "-


co



Sto





0r
z c












S-r-
un E-


5, C
SEV


00) -




S04)

CL






0) C
uff
_j ____ | ________ | ________ o *
o ^ Q o m @



























I I I I I I


I I I I I I


(6) sseuw aleGB


cn


04
1








8(D
4) D




E w

E u






cu.c

0 .-
0 2





4u





)m

C'd
CM
."*
U- (


P.- co
C; Co
6 o


C0
d d

















IU'

Z?
SII I
%. w 7


xo




I I 0
S0
co



o (b



0 a2
0 E



x m
0 x


o x x




x 0 c
X c -0
LOl
ox



CM (2
S0 x


Xx 0


XT..
X OC,
X OwS


















t

c
0)






5





(D
0




_ (U
E

C =
0


a,








10 to






0)"


._C
















c-

.-


.2m






s Jc
0)


o








a) is
.I

\ II







CT
., / o o I





LI."|












CHAPTER 3
SEASONAL VARIATION IN FEMALE PHONOTAXIS TO MALE CALLING SONG
OF THE FIELD CRICKET GRYLLUS RUBENS


Introduction

Males with sexually-selected traits are often at higher risk from predators

and parasites (reviewed by Burk 1982 and Magnhagen 1991) than those without.

Understanding how these risks affect female choice is important because

theoretical models on the evolution of female preferences show that choosiness

decreases when the risks of choice increase relative to the benefits (Parker

1983, Kirkpatrick 1987, Pomiankowski 1987, Real 1990, Iwasa et al. 1991). This

is because females that choose conspicuous males may also experience

increased risk of predation or parasitism because of their association with these

showy males. A decline in choosiness will reduce the strength of sexual selection

on male traits, which will cause males to become less extravagant. Therefore,

males with sexually selected traits evolve to be less conspicuous when predation

or parasitism is high not only because less conspicuous males survive better, but

also because the benefits of conspicuousness through female choice are

reduced. Despite the potential importance of these risks to females, few studies

have examined how predation risk affects female choice (although see Dill et al.

1999). Similarly, while many studies have investigated female choice for males

differing in parasitism levels (Borgia & Collis 1989, Hoglund et al. 1992, Houde &

Torio 1992, Fitzgerald et al. 1994, Buchholz 1995, Kraaijeveld et al. 1997, Lopez








1998), no study has evaluated how parasitism risk affects female choice. In this

study, I hypothesize that risk of parasitism decreases female attraction to males

with sexually-selected traits. I evaluated this hypothesis by examining female

phonotaxis to male calling song in the field cricket Gryllus rubens in two seasons

that differed in the risk of parasitism from acoustically orienting flies.

Female G. rubens approach calling males and mount them for mating.

The same male vocalizations that attract female crickets also attract gravid

females of the parasitoid fly Ormia ochracea. Female flies deposit larvae on and

around calling males, killing infected crickets within 7-10 days (Walker 1986,

1993). Even though females do not sing, Walker and Wineriter (1991) collected

females during the fall of 1990 and found 10% parasitism (3 of 10) and a similar

rate, 9%, among males (1 of 11). It is likely that females acquire parasites by

interacting with calling males that have been attacked by 0. ochracea that same

night. Several lines of evidence support this idea. First, Walker and Wineriter

(1991) experimentally muted males and found that unmuted males under high fly

densities were not parasitized by 0. ochracea (0 of 10) whereas non-muted

males were (7 of 13). This suggests that calling is associated with fly parasitism.

Because females do not sing and males only sing calling songs to attract females

for mating (Loher & Dambach 1989), females are likely to be parasitized as they

approach calling males. Second, 0. ochracea flies deposit larvae around calling

males. Females may become parasitized by walking over these larvae which

adhere to them (Cade 1975, 1979, 1984). These scattered larvae may also

explain how male and female G. firmus are parasitized by 0. ochracea (Walker &








Wineriter 1991); G. firmus co-occur with G. rubens but their calls do not attract

parasitoid flies (Walker 1986). This apparent link between female searching

behavior and fly parasitism make G. rubens an ideal species to test the

hypothesis that risk decreases female attraction to males with sexually-selected

traits.

Parasitoid flies may decrease female G. rubens attraction to male calling

song to minimize the risks of choice. In northern Florida, 0. ochracea is abundant

in the fall and rare in the spring (Walker 1986, Chapter 2), and fly parasitism.

rates are higher in the fall than in the spring (10% versus 0%, Walker & Wineriter

1991; chapter 2). This seasonality in fly abundance and parasitism rates

suggests that fall females may incur higher risks of choice than spring females.

Such costs may be greatest in the early evening because 0. ochracea flight

activity is highest just around dusk and then steadily declines (Cade et al. 1996,

Chapter 2). These nocturnal changes in parasitoid fly activity may favor females

that are less phonotactic to male calls in the early evening. This would be

consistent with laboratory experiments on gobiid fish, crickets, and guppies,

which have shown that female choosiness decreases when predation risk

increases (Forsgren 1992, Hedrick & Dill 1993, Godin & Briggs 1996). Because

0. ochracea flies are only active after dusk in the fall and not the spring (Chapter

2), I hypothesized that fall females would be less phonotactic to male calling

song in the early evening than spring females. I evaluated this hypothesis by

conducting a phonotactic experiment on fall and spring females shortly after

dusk. In this experiment, females were given a choice between a silent speaker








and a speaker broadcasting male calls. I predicted that fewer females would

choose the broadcasting speaker in the fall than in the spring, and fall females

would spend less time near the broadcasting speaker than spring females. The

potential finding that fall females are less attracted to male calling song in the

early evening than spring females would be important because it would reduce

the strength of sexual selection on males in the fall and help explain why fewer

males sing around dusk in the fall than in the spring (Chapter 2).



Methods

Phonotactic Experiment

In this experiment, I quantified and compared among seasons the

phonotactic behavior of wild-caught females for male calling song at the time

when parasitoid flies are most active. Gryllus rubens exhibits a wing dimorphism:

males and females are either long-winged and capable of flight or short-winged

and flightless. Both morph types were used in this phonotactic experiment in the

proportion that they were available. I used two methods of collecting crickets;

some were collected by turning over objects and debris and others were

captured using sound traps (Walker 1986). I used the two methods because I

collected very few long-winged individuals by turning over objects and debris and

sound traps captured long-winged individuals exclusively. All collections were

made at the University of Florida Organic Garden, Alachua Co., FL.

Females were weighed to the nearest 0.001 g and housed individually in

square plastic containers measuring 13 x 13 x 9 cm (I x w x h) with wire screen








lids. Each container had an inch of sand, an inverted egg carton shelter, and

Purina Cat Chow and water ad lib. All containers were held under ambient light

and temperature in a greenhouse less than 1 km from the collection site. After 10

days (the number of days necessary to ensure that the crickets were not

parasitized by 0. ochracea), I brought females to the lab and tested their

preference for male calling song. Parasitized females were not used in these

experiments. The actual sample size for the fall was Nong-wne=24 and Nho-

,nged=16. The sample size for the spring was Nongnged=23 and NshortngWd=17.

Phonotactic tests were conducted in a rectangular arena (Fig. 3-1)

measuring 1.0 x 0.3 x 0.3 m (I x w x h). The floor of the arena was made of

compressed foam, and the walls consisted of metal 2 mm mesh wire screen To

prevent the crickets from climbing out of the arena, I taped a 5 cm wide paper

strip along the wire screen walls 15 cm from the floor. On the outside at each end

of the arena, I placed a Radio Shack model 277-1008 mini-audio amplifier

speaker. These speakers were connected to a Sony model D-EJ721 portable

CD-Walkman. Prior to a test, a female was placed in the center of the arena

under an inverted egg carton shelter. I covered this shelter with a plastic cup and

allowed the female to acclimate for 5 min. After that time, I lifted the plastic cup

with a string attached to it and gave the female 5 min to choose between a silent

speaker and a speaker broadcasting a male calling song. Output of the

broadcasting speaker was set at 72dB, 15 cm in front of the speaker, using a

Briel & Kjaer model 2219 sound level meter. This sound level falls within the








natural variation of sound intensities of calling males (M.J. V61lez, unpublished

data).

I observed females in the arena using a 40 watt red light. Each female

was tested only once two to three hours after sunset within the four-hour period

when 0. ochracea flies are most active (Chapter 2). For each test, I recorded the

latency to leave the shelter, which end screen the female contacted first (criterion

for female choice), and the time spent within 10 cm of the broadcasting speaker.

The floor of the arena was lined with paper. Between tests, this paper was

replaced, the metal wire screen walls were wiped with paper towels, and the

position of the broadcasting speaker was randomized. No end biases were

detected. I examined female phonotactic behavior in the laboratory from 12 June

to 5 July (N=40) and 10 October to 6 December 2001 (N=40) because these are

the times of year when flies are rare and abundant, respectively (Walker 1986).

The temperature in these tests did not differ between the two times of the year

(Meanal, SE = 22.580.06 C, Meanspng SE = 22.660.05 C, P=0.463).

This phonotactic experiment was conducted in the absence of 0.

ochracea flies because I assumed that females do not detect flies directly. I

based this assumption on the finding that male crickets in the presence of flies do

not sing differently than solitary males (S.M. Bertram, unpublished data).

I analyzed the time that females spent within 10 cm of the broadcasting

speaker using a general linear model with model: season (fall, spring), collection

method (turning over objects and debris, sound traps), morph (long-winged,

short-winged), mass, season*collection method, season*morph, season*mass.








The P-value reported for each variable is the value after all other variables were

taken into account. Data were analyzed using SPSS 11.5. Results are presented

as means standard error.



Source of Male Calls

Male G. rubens produce a trilling song in which sound pulses are repeated

and grouped into long pulse trains (Fig.3-2; Doherty & Callos 1991). Field-

collected males were individually recorded in the lab using an Optimus model

CTR-115 sound-activated cassette tape recorder. Ten males were recorded in

the fall and 10 males in the spring. Each male was recorded for at least five

minutes. The recordings of each male were digitized and divided to make four

one-minute calls. This generated 40 fall calls and 40 spring calls. The digitized 1-

min calls were burned onto CD discs, and one of these calls was played back

repeatedly to each female for 5 min during the preference test. Each female

received a different call. Fall females were tested with fall calls and spring

females with spring calls. The temperature at which males were recorded did not

differ between the fall and the spring (Meanari SE = 21.881.33 C, Meanpi +

SE = 22.750.80 C, P=0.584). This temperature also did not differ from the

temperature during the female phonotactic tests (Meanrec SE = 22.310.76 C,

Meanphon SE = 22.630.040C, P=0.684).








Fly Activity

To verify that 0. ochracea flies are more active in the early evening in the

fall but not the spring, I monitored fly activity in a pasture at the University of

Florida Beef Teaching Unit, Alachua Co., FL. To monitor fly activity, I used a slit

trap (Walker 1989) with an electronic sound synthesizer that broadcasts calls of

G. rubens (4763 Hz carrier frequency, 50.0 pulses per second, 50% duty cycle;

continuous trill at 80 dB, 15 cm above the speaker). I activated the synthesizer

an hour before sunset and counted the number of flies captured at the trap 4 h

later. I carded out this procedure from 5 June to 12 July (N=20) and from 10

October to 26 November 2001 (N=20) because these were the times when

females were tested in the phonotactic experiment.



Results

Phonotactic Experiment

Fewer females chose the broadcasting speaker in the fall (20/40; 50%),

than in the spring (30/40; 75%) (Fisher's exact test: P=0.013). Fall females also

spent less time within 10 cm of the broadcasting speaker (Fig. 3-3; ANOVA,

season, F1.72=9.736, P=0.003) than spring females. Furthermore, fall females

took longer to leave the shelter than spring females (Fig. 3-4; t-test: Nfall=40,

Nspnng=40, t=3.115, P=0.003). In general, upon leaving the shelter, fall females

tended to walk around it and after some time, many of them went to the

endscreen with the silent speaker and walked around that area. In contrast,

spring females tended to leave the shelter and walk directly to the endscreen








with the broadcasting speaker. Many of them climbed the metal screen wall

facing the speaker walking back and forth in front of the broadcasting speaker.

The time that females spent within 10 cm of the broadcasting speaker was

not significantly affected by collection method (ANOVA, collection method,

F1,72=0.227, P=0.635) or morph type (ANOVA, morph, F1,72=0.009, P=0.926).

Similarly, the time that females spent within 10 cm of the broadcasting speaker

was not significantly affected by the interaction between season and collection

method (ANOVA, season*collection method, Fi.72=0.208, P=0.650) or the

interaction between season and morph (ANOVA, season*morph, F1,72=0.124,

P=0.726).

Female mass had a significant effect on the time that females spent within

10 cm of the broadcasting speaker (ANOVA, mass, F1,72=5.231, P=0.025). This

effect differed between the seasons. In the spring the time that females spent

within 10 cm of the broadcasting speaker decreased with female mass but not in

the fall (Fig. 3-5, ANOVA, season*mass, Fi.72=8.219, P=0.005).



Fly Activity

More flies were captured in the early evening per night in the fall than in

the spring (Meanfai SE = 2.60.8, Meanspong SE = 0.20.1, Mann-Whitney U

test, U=317, nfa11=20, nspdng=20, P<0.001).








Discussion

This study demonstrates seasonal changes in female G. rubens

phonotactic behavior to male calling song. Parasitoid flies are most abundant in

the fall and most active in the early evening (Chapter 2). In the spring, flies are

rare. Because females may be parasitized as they approach singing males, I

predicted that fall females should be less phonotactic to male calling song in the

early evening than spring females. The results of this study support this

prediction. Fewer females chose the speaker that was broadcasting male calls in

the fall than in the spring, and fall females spent less time near the broadcasting

speaker than spring females. These results are consistent with the hypothesis

that risk of parasitism decreases female attraction to males with sexually-

selected traits.

Few studies have examined the effect of risk on female choice for sexually

selected traits. French and Cade (1987) examined female mating behavior in G.

texensis housed in large outdoor arenas. This field cricket is also parasitized by

0. ochracea (Cade 1975). Their results showed that females mated more

frequently around dawn, which coincided with the time when parasitoid flies were

less active (Cade et al. 1996). Similarly, Hedrick and Hill (1993) examined female

preference for long-bout calls in G. integer in the presence and absence of cover.

In this species, females prefer long-bout calls over short bout calls (Hedrick

1986) all else being equal. However, when females were given a choice between

a long-bout call played across open space and a short-bout call played through

100% cover, females did not express a preference for long-bout calls, suggesting








that the increased predation risk associated with open space reduced the

attractiveness of long-bout calls. Similar results have also been found in sand

gobies and guppies (Forsgren 1992, Godin & Briggs 1996). Typically, females of

these two species prefer to mate with brightly colored males. However, this

preference disappeared when females had to choose in the presence of a

predator. Collectively, these results are consistent with the findings of this study

and support the hypothesis that when choice is risky, the intensity of sexual

selection declines.

In this study, females were collected using two methods; turning over

objects and debris, and using sound traps. The use of sound traps may have

introduced a bias since individuals that were captured using this method were

initially attracted to male calling songs. However, this possibility was not

supported because collection method and the interaction between season and

collection method did not have significant effects on the time that females spent

within 10 cm of the broadcasting speaker. Another bias that may have affected

the results of this study is that different numbers of long- and short-winged

females were collected in the fall and spring. However, again, this was not

supported because morph type and the interaction between season and morph

type did not significantly affect the time that females spent within 10 cm of the

broadcasting speaker. Collectively, these results suggest that seasonal changes

in female phonotactic behavior were not due to seasonal changes in the effects

of collection method or morph type.








The time that females spent near the broadcasting speaker decreased

with female mass in the spring but not the fall (Fig. 3-5). This seasonal difference

in the interaction between female mass and phonotactic behavior may be due to

seasonal differences in the mating experience of females. Female crickets mate

multiply (reviewed by Loher & Dambach 1989). Their mating experience has

been shown to affect their phonotactic behavior. Lickman et al. (1998) presented

females differing in mating experience with multi-stimulus songs. Their results

show that virgin females exhibited more directional movement towards male

songs than mated females. Virgin females may be more phonotactic than mated

females since this will increase the chances that they will mate at least once and

secure some sperm. In this study, if heavy females in the spring were old and

mated, and light females were young and virgin, then this relationship may

explain why female phonotaxis decreased with female mass in the spring. This is

possible because G. rubens has two pulses of adults in the fall and spring, and

two smaller cohorts of adults in the summer and winter (T.J. Walker, unpublished

data). At the time that spring females were collected from the field (15 May to 25

July 2001), the summer cohort was starting to emerge as adults. Therefore, I

could have collected both mated spring females, and virgin summer females. In

contrast, because at the time that fall females were collected (1 October to 25

November 2001), no females from the winter cohort had yet emerged. This

would mean that my fall sample may have primarily consisted of mated fall

females. This could explain why phonotactic behavior did not decrease with

female mass in the fall. It is important to note, however, that even when this








seasonal interaction between female phonotactic behavior and female mass is

taken into account, season still has a significant effect on female phonotactic

behavior.

Fall females took longer to leave the shelter during the phonotactic

experiment than spring females. This seasonal difference in female behavior may

be interpreted as fall females behaving more cautiously around male calls than

spring females in order to reduce fly parasitism or other sources of increased

risk. This explanation is consistent with a study on G. integer in which males

differing in song conspicuousness were presented with a novel, potentially

dangerous predator stimulus (Hedrick 2000). Males with more conspicuous

songs took longer to emerge from a shelter after the predator stimulus than

males with less conspicuous songs. Hedrick (2000) interpreted these differences

in latency to leave the shelter as males compensating behaviorally for their

conspicuous mating displays. This interpretation is consistent with the findings of

this study for females and suggests that future work should examine how

parasitism risk affects female searching behavior.

Fall females took longer to leave the shelter than spring females. Could

this difference in behavior account for the shorter time that females spent within

10 cm of the broadcasting speaker in the fall than in the spring? This is possible

because time spent in the shelter is time that could not have been spent near the

broadcasting speaker. However, some evidence suggests that this is unlikely.

The mean time that fall females spent in the shelter was 98 sec (Fig. 3-4). This

means that fall females spent on average 202 sec outside of the shelter. This








time is much greater than the mean time that spring females spent within 10 cm

of the broadcasting speaker, 121 sec (Fig. 3-3). These results suggest that

seasonal differences in female latency to leave the shelter do not fully explain the

seasonal variation in the time that females spent within 10 cm of the

broadcasting speaker.

This study has important implications for understanding the evolution of

sexually-selected traits. Male G. rubens sing to attract females for mating. This

behavior, however, also attracts parasitoid flies in some populations, in some

seasons, and at some times of the night. Differences in male calling behavior

between populations, seasons, or time of the night have been interpreted as due

to the natural selection pressure, parasitism. However, if parasitism risk also

decreases female attraction to male calls, then males may derive fewer benefits

from calling. In this study, I have found that fall females are less attracted to male

calling song in the early evening than spring females. This reduction in the

strength of sexual selection, as well as the increased natural selection costs for

males, will weaken selection for calling in the early evening for fall but not spring

males. Therefore, fall and spring males should differ in the timing of calling

behavior. Fall males should not sing in the early evening, but spring males

should sing at that time. This prediction has been confirmed in G. rubens

(Chapter 2): surveys of calling males in the field showed that fewer males sing in

the early evening in the fall than in the spring. This correlation between female

behavior and male calling behavior suggests that future studies investigating

variation in sexually selected traits should examine not only how risk affects





40

natural selection on males with sexually selected traits, but also how risk affects

female choice for males with sexually selected traits, and hence the benefits of

conspicuous traits.












I-
(A 0
U,

LCZ


O U
C,,



























r P



















ww
o












f
r







































,'*












m00
r













,,(


")






Z ) n
'c




'0r





I
'0




'0





'r *

o <
-


(zH~f Aounbej



















Em
Co



z





4-
0o









4-D
r- 0













a 0
Ca 0
w

Cu IT








E 0)










to o o o 0 .EO
CV) NM 04 t

(oes) je)eads 6ui seopeojq jeau juads awil ueayy C o

a) c)



i T .a 5






44















1-


0

SE
ilt

4)




----------------- "|
(,D



C)) CE












C ) o _O 0 o


0
e,
O (A
iEl













4-
E.



o M 0


o 0r o 2
Eil




O M
o--o-----p-Si
oo ..-
0 0 mO
o oo


00 v x
o .-- 5,
0 0 0 0 to gQ-
o o -t a
o x x oXX Xx) E c -Q)
0 x xox ,, 4M
xx xE ,0
-x x|
x xo "- ,
x x 0 I
x -,)


S ooo o E-



(O 8 uO O 8 0 ,

(oes) jejeeds 6ul seopeojq jiteu lueds ow111 -D c













CHAPTER 4
VARIATION IN SEXUAL SELECTION FOR MALE SONG
IN THE FIELD CRICKET GRYLLUS RUBENS


Introduction

A primary goal of sexual selection theory is to understand how natural and

sexual selection interact to favor variation in male sexually-selected traits

(Andersson 1994). Crickets represent an ideal system in which to study this

question for several reasons. First, calling song is highly variable among males

(for review see Bertram 2002). Such variation can be easily quantified by

individually recording males and digitally analyzing their songs. Second, male

song attracts conspecific females for mating. Typically, females approach singing

males and mount them (Loher & Dambach 1989). This suggests that male calling

is affected by sexual selection. Finally, male vocalizations also attract parasitoid

flies (Cade 1975, Walker 1986, 1993). Female flies deposit larvae on and around

calling males, killing infected crickets within 7-10 days (Walker 1986, 1993). This

means that male calling may also be affected by natural selection. Collectively,

these conditions make crickets an ideal system to study how natural and sexual

selection interact to favor variation in male sexually-selected traits.

One of the most variable song characters in male crickets is nightly calling

duration or the total time spent calling per night. Using an artificial selection

experiment, Cade (1981) demonstrated that much of the variation in nightly

calling duration in the cricket, Gryllus texensis, was due to genetic differences








among males. Subsequently, Cade and Wyatt (1984) found that nightly calling

duration was highly variable among male G. texensis, G. pennsylvanicus, G.

veletis, and Teleogryllus africanus. Similarly, Cade (1991) determined that nightly

calling duration was also highly variable among male G. integer and G. rubens.

Despite the potential role of sexual selection in explaining variation in male

cricket song, most studies examining variation in nightly calling duration have

ignored the effect of sexual selection (Cade 1984, Cade & Wyatt 1984, French &

Cade 1989, Sakaluk & Snedden 1990, Forrest et al. 1991, Cade & Cade 1992,

Gray 1997, Wagner & Hoback 1999, Kolluru et al. 2002, but see Bertram 2002).

The goal of this study is to examine whether seasonal variation in male G.

rubens nightly calling duration is correlated with seasonal variation in sexual

selection.

In northern Florida, male G. rubens that sing spend more time calling in

the fall than in the spring (Chapter 2). Three factors have been used to explain

variation in male cricket calling behavior: male condition, parasites, and density.

Nightly calling duration should increase with male condition because males in

good condition are better able to bear the energetic costs of calling than males in

poor condition (Sakaluk & Snedden 1990, Forrest et al. 1991, Gray 1997,

Wagner & Hoback 1999). However, this possibility is not supported by the

available data for G. rubens. Although fall males are heavier than spring males,

nightly calling duration does not increase with male mass (Chapter 2). Similarly,

parasitized males should call less than unparasitized males because parasites

decrease the ability of males to sing (Cade 1984, Kolluru et al. 2002). However,








this hypothesis does not explain seasonal variation in G. rubens calling since

unparasitized fall and spring males differed in nightly calling duration (Chapter 2).

Finally, males at high densities should sing less than males at low densities

because as density increases, the costs of defending territories increase as well

as the chances of encountering a female by searching without calling (Alexander

1961 & 1975, Otte 1977, Cade 1979, Cade & Wyatt 1984, French & Cade 1989,

Cade & Cade 1992). But, differences in density cannot explain the longer calling

duration of fall G. rubens because male densities are higher in the fall than in the

spring (Veazey et al. 1976). Taken together, it is unlikely that the three most

common explanations can account for the seasonal changes in male G. rubens

nightly calling duration.

The same male G. rubens vocalizations that attract females also attract

gravid females of the parasitoid fly Ormia ochracea (Walker 1986, 1993). These

females deposit larvae on and around calling males, killing infected crickets

within 7-10 days. If parasitoid flies attack males more in the spring than in the fall,

then this seasonal difference in parasitism risk may explain the seasonal

differences in nightly calling duration. However, for G. rubens this is not an

explanation because 0. ochracea in northern Florida is abundant in the fall and

rare in the spring (Walker 1986, Chapter 2) and fly parasitism rates are higher in

the fall than in the spring (10% versus 0%, Walker & Wineriter 1991). Therefore,

seasonal changes in parasitism risk cannot explain the seasonal changes in

male G. rubens calling duration.








Males that sing may have longer calling durations in the fall than in the

spring because seasonal changes in photoperiod and temperature cause fall and

spring males to differ in nightly calling duration. This possibility is supported by

rearing studies showing that photoperiod and temperature may affect calling

song parameters. Walker (2000) reared male G. rubens under different

photoperiods and temperatures to examine whether developmental conditions

affect pulse rate. He found that temperature influenced pulse rate, but

photoperiod did not. Similarly, Olvido and Mousseau (1995) reared striped

ground crickets (Allonemobius fasciatus) under two environments differing in

photoperiod and temperature. Their results showed that chirp rate, chirp duration,

interchip interval, pulse number, and carrier frequency are affected by rearing

environments. Despite this work, no study has ever found a significant effect of

photoperiod on male nightly calling duration in crickets. Bertram and Bellani

(2002) reared male G. texensis under fall-like and spring-like photoperiods and

found that rearing photoperiod does not significantly affect nightly calling duration

in males. This suggests that seasonal differences in male G. rubens nightly

calling duration are not due to seasonal changes in photoperiod. However, the

effect of temperature during development on nightly calling duration has not been

tested in any cricket species.

Male G. rubens may exhibit seasonal changes in nightly calling duration if

the benefits of singing change seasonally. These benefits may vary seasonally if

sexual selection operates differently in the two seasons. Accordingly, I

hypothesized that males sing longer in the fall than in the spring because the








benefit of singing increases with calling duration in the fall but not in the spring. I

evaluated this hypothesis by using pitfall traps fitted with speakers broadcasting

male calling songs of different duration. I counted the number of females

captured at the traps in the fall and spring. and predicted that the number of

captured females should increase with calling duration in the fall but not in the

spring. This prediction is important because it would be the first evidence to

suggest seasonal variation in sexual selection. Such variation could favor

variation in the sexually-selected trait, male calling duration.



Methods

Pitfall Trap Experiment

I examined seasonal variation in sexual selection for nightly calling

duration from 14 April to 12 June and 23 September to 5 November, 2002

because these are times when males sing for short and long durations,

respectively (Chapter 2). During both fall and spring, female and male G. rubens

occur in two forms, long- and short- winged. The two morphs differ in their ability

to fly as well as in other life-history characteristics such as the onset of

reproduction (Roff 1984, Zera & Rankin 1989, Holtmeier & Zera 1993). Females

of both morphs were used in this study in the proportions in which they occurred.

All experiments were conducted in a field at the University of Florida

Environmental Landscape Horticulture Education Lab, Alachua Co., FL. I chose

this field because male G. rubens sing all year long at this site.








Tests for seasonal variation in sexual selection for nightly calling duration

were conducted using five pitfall traps (Fig. 4-1) laid out in a straight line 10 m

apart. Each trap was made by excavating a hole in the ground and placing a 2.4

L plastic bucket measuring 17 x 15 cm (height x diameter) in the hole so that the

rim was level with the ground. Over each bucket I placed a plastic funnel 10 cm

tall with an upper diameter of 18 cm and a spout diameter of 6 cm. The funnel

rested on the rim of the bucket and the spout was 7 cm above the bottom of the

bucket. I placed a 33-cm square PVC tube frame (1.9 cm diameter) with four 7-

cm legs over the bucket-funnel trap. The frame rested on the ground surrounding

the bucket-funnel trap and supported a metal grill top with 1.5 cm spaces

between the bars. In the center of this grill top, I positioned a Sony model D-

EJ721 portable CD-Walkman with a Radio Shack model 277-1008 mini-audio

amplifier speaker lying flat over the CD player. The speaker Ipointed upward and

rested 9 cm over the center of the trap. Crickets that were attracted to calls

played on the CD player, flew (long-winged morph) or walked (short-winged

morph) into the bucket-funnel trap where they were collected hourly. Five

different treatments of male calling duration were used, one for each pitfall trap.

The treatments consisted of broadcasting male calling songs for 0, 1, 2, 3, or 4

hours. These calling durations fall within the natural variation of time spent calling

by males (Chapter 2). Output of the broadcasting speakers was set at 80dB,

measured 15 cm above each speaker, using a Bruel & Kjaer model 2219 sound

level meter. This sound level falls within the natural range of sound intensities of

calling males (M.J. V6lez, unpublished data).








Each experiment began an hour before sunset when I activated the CD

player over the pitfall trap set to broadcast for 4 h. An hour later, I turned on the

CD player over the pitfall trap set to broadcast for 3 h. This continued until all CD

players broadcasting songs had been activated. The outcome of this staggering

the onset of the calling was that all CD players finished broadcasting male calling

songs three hours after sunset. This pattern was used because individual

recordings of males under field conditions show that calling start times are

inversely related to calling duration (M.J. Velez, unpublished data). The pitfall

trap with the 0-h treatment was handled like the pitfall trap set to broadcast for 4

h: a CD player was placed over the trap an hour before sunset. However, this

player did not broadcast any songs. This treatment controlled for the number of

non-phonotactic females that fell into a trap. All pitfall traps were checked hourly

to count the number of captured females from one hour after the experiment

started to the end of the experiment three hours later. Trapped females were

removed from the field to ensure that they were not sampled in more than one

treatment or night. Any trapped males or juveniles were released back into the

field. Each night constituted one replicate of the experiment. All treatments within

a replicate used the same male calling song, but different replicates used

different songs. Between replicates, I randomized the position of the treatments.

Randomization was achieved by generating all possible combinations of

treatments and picking one at random each night. This experiment was

conducted on 35 nights in the spring and 25 nights in the fall.








Source of Male Calls

Male G. rubens were collected from the field using sound traps (Walker

1986). They were housed individually in square plastic containers measuring 13

x 13 x 9 cm (I x w x h) with wire screen lids. Each container had an inch of sand,

a few pieces of Purina Cat Chow (changed daily) and water ad lib. All containers

were held under ambient light and temperature in a greenhouse less than 1 km

from the collection site. After 10 days (the number of days necessary to ensure

that the crickets were not parasitized by 0. ochracea), I brought the males into

the lab and individually recorded each for at least 5 min using an Optimus model

CTR-115 sound-activated cassette tape recorder. All of the males used in these

recordings were unparasitized and long-winged because short-winged males

could not fly into the sound traps used to collect crickets. Ten males were

recorded in the fall and 10 males in the spring. The recordings of each male were

digitized and used to make four 1 min calling songs. This generated 40 fall and

40 spring calling songs. These songs were burned onto CD discs and played

back repeatedly in the pitfall trap experiment for the duration of the treatment. In

the fall, I used fall calling songs over pitfall traps, and in the spring, I used spring

calling songs. The mean temperature at which males were recorded was 23.10C

and did not differ between fall and spring calls (t-test, P=0.702).



Data Analysis

A generalized linear model with a Poisson distribution and a logit link was

used to regress the calling duration of a treatment on the number of females








captured in that treatment's pitfall trap. This model was used because the data

could not be transformed to achieve normality. Because treatments within a

replicate were not statistically independent, I used a mixed model to analyze the

data. In this model, calling duration was a fixed effect, and replicate number was

a random effect. Fall and spring data were analyzed separately. The data were

tested for overdispersion to validate the use of the Poisson model.

If females search for mates at different times of the night in the fall and

spring, then such seasonal differences in female behavior could affect the

relationship between calling duration and number of trapped females. To

evaluate this possibility, I compared the capture time to the nearest hour of each

female in the pitfall traps during the fall and spring using a Mann-Whitney U test.

Data were analyzed using R 1.7.1 (Dalgaard 2002) and Systat 6.0

(Wilkinson 1997). Descriptive results are presented as means + standard error.



Results

The number of captured females increased with treatment calling duration

in the fall but not in the spring (Table 4-1; Fig. 4-2; fall: t=4.532, df=99, P<0.001;

spring: t=0.419, df=139, P=0.676). Females were captured in all calling duration

treatments in the spring and fall with the exception of the 0-h treatment in which

no female was captured in either the fall or the spring (Fig. 4-2). Sixteen females

arrived at the pitfall traps in the fall (niong=12, nshot=4), and 10 in the spring

(ng=6, nshort=4). Fall and spring females did not differ in the time at which they

were captured in the pitfall traps (Fig. 4-3; nfa,=16, nsping=10, Mann-Whitney U








test=56.5, P=0.189). This non-significant result remains even when the number

of females captured in the first two hours are compared with the number of

females captured in the last two hours in the fall and spring (X2=0.005, df=1,

P=0.940)



Discussion

Male G. rubens that sing spend more time calling in the fall than in the

spring (Chapter 2). This is surprising because there are more phonotactic,

parasitoid flies present in the fall (Walker 1986, Chapter 2) so one would expect

males to call less to avoid parasitism. Furthermore, as I argued in Chapter 3,

females should be less attracted to calling males in the fall because they too are

parasitized by these flies (Walker & Wineriter 1991). The results of the pitfall trap

study provide an explanation: males that sing longer are more likely to attract

females during the fall but not during the spring (Table 4-1). What might explain

this result?

If fall females search for mates in the early evening and spring females

search for mates in the late evening, then these temporal differences in female

activity may explain why the number of captured females in my pitfall traps

increased with calling duration in the fall but not the spring. This is possible

because of the way I set up the experiment: the long calling duration treatments

began earlier in the evening than the short calling duration treatments.

Consequently, fall females, preferring to search for mates in the early evening,

would have a higher probability of encountering a pitfall trap with a long calling








duration whereas spring females, preferring to search for mates in the late

evening, would have an equal probability of encountering a long or short calling

duration treatment. However, the results of this study do not support this

explanation. Fall and spring females did not differ in the time at which they were

captured in the pitfall traps. This suggests that male calling duration increases in

the fall but not the spring because long duration calls are more effective at

attracting females in the fall but not in the spring. This pattern is consistent with

seasonal variation in sexual selection.

Female G. rubens densities in northern Florida are higher in the fall than in

the spring (Veazey al. 1976). If females choose males randomly with respect to

calling duration, then males with long duration calls should have a higher

probability of attracting females than males with short duration calls. Because

this probability should increase proportionally with density, seasonal changes in

female density cannot explain the seasonal changes in the success of long

duration calls. Instead, the seasonal differences in the success of long duration

calls suggest that fall females do not have a preference for calling duration and

spring females prefer short duration calls. Theoretical models on the evolution of

female preferences show that choosiness decreases when the risks of choice

increase relative to the benefits (Parker 1983, Kirkpatrick 1987, Pomiankowski

1987, Real 1990, Iwasa et al. 1991). Parasitoid flies are abundant in the fall and

rare in the spring (Walker 1986, Chapter 2). Fall females may not have a

preference for calling duration because the costs of choice due to fly parasitism

may be high in that season. Even though females do not sing, they are








nonetheless parasitized to the same levels as males are (Walker & Wineriter

1991). In contrast, females are more phonotactic in the spring than in the fall

(Chapter 3). In an experiment giving females a choice between a silent speaker

and a speaker broadcasting a male calling song, fewer females chose the

speaker that was broadcasting male calls in the fall than in the spring, and fall

females spent less time near the broadcasting speaker than spring females. This

may mean that less singing is required to attract females in the spring than in the

fall. This seasonal variation in sexual selection is consistent with the seasonal

changes in the success of long duration calls.

The results of this study are consistent with recent work showing within-

and between-population variation in the female preference functions of field

crickets. Wagner et al. (1995) sequentially presented female G. integer from a

single population with songs differing in the number of pulses per trill, inter-trill

interval, or the proportion of missing pulses within a trill. These presentations

revealed that individual females differed significantly in their preference functions

based on the number of pulses per trill and inter-trill interval. Furthermore, these

preference functions were highly repeatable within females. Similarly, Simmons

et al. (2001) examined geographic variation in the female preference function for

the long chirp portion of male Teleogryllus oceanicus calling song. Their results

show that the female preference function for the amount of long chirps differed

between populations. The findings from these studies suggest that selection can

act on female preference functions. The outcome of such process may favor

variation in male sexually-selected traits.








To evaluate the general importance of temporal variation in sexual

selection as a mechanism favoring variation in male sexually-selected traits,

future studies should examine how variation in female choice is affected by

factors that change the costs and benefits of choice, such as predation,

parasitism, sex ratios, and population densities. If any of these factors fluctuates

in time, then the costs and benefits of choice as well as the resulting female

choice patterns may also vary temporally. In turn, this temporal variation in

female choice could act to favor variation in sexually-selected traits















m
-- o






0 Rc


U, V r \0
m a ,) v






.0 0.--
E E~E E
-5 (a -- ,_ .=


E,..me m aE
S.'o m E ;

0.E
E >c"


= a c -a L E


o E o. ,.
Sm -0 G- w m
E1-c.- wi-- 0



4D M *S c (0 (0 E
ME 3 :w i |8



4 E(- o Eo
I L)- 3 m Es .C




a. $ h it cm x Q
LLm















c

0 C a














w '0- 0
'm- ca





S c


o2S E







SE -G *


















^ *^ c
-c0 ^ ^S


4- V V





I co 0 )





m, cr,










I-I
Lo
w ) ()
V
2 10 '- CD C0
I*) 6 6 0 6



0 CD 04










I, r
cn cI S s


I.n co .























































,n co N C 0
0 0 0 0 6


4-1
C
co




oo
* 0


CU
0
'-


T- CM O
6 o o o


4)
18|
^|.|
CL
4)



S 2










OW

E~~
C
CI3
m










c >
E 02=












cc
E a0 ;
E! r-E

0 o C
LL


cc x

2 mB
i EL';C
? w *o c
oc



E 9E o~ .a



"w 2

icl2
S~IEO
aE-EsS






3l,l
u. n5.


1B!u /saeleUwa peJnmde jo o eqwnu leo, ueeal






62











III
Co;

z E)
r: x
Ccc

C6JO
i oi
a-) 0







E 40-
_) 0-- -'



E(E
cc
00
I -- i -- i i -- -- i i -- i -4-- 0i-cU 1


r- -6


0 0o
0 0



m L cO*O E
0



CD C CD
E 0
E E

00.0

04T
Prrr

E 4L

>'-
0 C D N 0 (q co CD co 0CI)

C c E
Aouenboij (D
oE

0,) 'ea
SIfB
4J *
iT 0.
t-I













CHAPTER 5
GENERAL MODEL: A TEST OF THE FACTORS AFFECTING MALE CALLING
DURATION IN THE FIELD CRICKET GRYLLUS RUBENS


Factors Affecting Male Calling Duration

Fall and spring generations of the field cricket Gryllus rubens in northern

Florida experience different selective pressures and social environments. Males

of this species sing which attracts females for mating. Their calling song also

attracts gravid females of the parasitoid fly Ormia ochracea (Walker 1986, 1993).

These female flies deposit larvae on and around calling males, killing infected

crickets within 7-10 days. In the fall when parasitoid flies are abundant (Walker

1986, Chapter 2), and cricket densities are high (Veazey al. 1976), calling songs

of long duration attract more females than short duration calling songs (Table 4-

1, Fig. 4-2). This is quite different from the spring when parasitoid flies are rare

(Walker 1986, Chapter 2), cricket densities are low (Veazey al. 1976) and calling

song duration does not affect female attraction rates (Table 4-1, Fig. 4-2). In this

chapter, I examine how seasonal changes in fly parasitism, female visitation

rates, and cricket density should alter the costs and benefits of calling for fall and

spring males. I evaluate these effects by building a model that calculates the

fitness of different calling durations in each season. I then use this model to

predict observed seasonal patterns in calling duration. Additionally, I

independently vary the effects of fly parasitism and female choice on male calling

duration within each season to evaluate the relative importance of natural versus








sexual selection in favoring seasonal patterns in male calling duration. This

model is valuable because it is the first to evaluate simultaneously the effects of

fly parasitism, female attraction rates, and cricket density on the male character,

calling duration. This model also provides insights into how different selective

pressures and social environments can interact to favor variation in a male trait.



Model: Predicting Seasonal Patterns in Male Calling Duration

Male Calling Strategies

In this model, males can adopt one of four calling strategies: (1) call for 1

h, (2) call for 2 h, (3) call for 3 h, and (4) call for 4 h. These calling durations fall

within the natural range of time spent calling by male G. rubens under field

conditions (Fig. 2-4; Cade 1991). I assume that males adopt one calling strategy

and use it throughout their lifetime. This assumption is supported by the finding

that the mean time spent calling per 24 h under laboratory conditions does not

change with male age in G. texensis, G. pennsylvanicus, G. veletis, and

Teleogryllus africanus (Cade & Wyatt 1984). Furthermore, Cade (1981) found

that some of differences between male G. texensis in calling duration are due to

genetic differences among males.

Individual field recordings of male G. rubens show that some males do not

call (Fig. 2-4). These males may obtain mates by walking around searching for

females. This alternative male tactic has been documented in several species of

field crickets (French & Cade 1989, Hissmann 1990, Zuk et al. 1993, Hack 1998).

I did not include this behavior in my model because little is known about the








mating success and parasitism risk of walking males. Cade (1979) proposed that

walking males may have decreased mating success on any given day, but they

may have higher survival rates and longer lifespans than calling males because

they do not attract as many parasitoid flies. This hypothesis, however, remains

untested and hence I exclude non-calling as a strategy in my model.



Calculating Fitness of Male Calling Strategies

Male calling attracts conspecific females and parasitoid flies. Each calling

strategy carries a mating benefit, M, and a probability of escaping fly parasitism,

E. Each day, males mate before they face fly parasitism. The daily fitness of

each calling strategy F1 is defined as:

Fi = E|Mi + (1-Ei)MI,

where i is the calling strategy.

If a male sings and is not parasitized on a given night, he accrues the

mating benefit of that day and sings again in the next day. If a male sings and is

parasitized on a given night, he accrues the mating benefit of that day. After fly

parasitism, the infected male has one more day to carry out his calling strategy

and receive the corresponding mating benefit. After that day, the infected male

does not get any further reproduction. This restriction is meant to mimic the

temporal dynamics of the effects of fly parasitism on male calling duration.

Kolluru et al. (2002) experimentally infected male T. oceanicus with 0. ochracea

larvae and monitored the proportion of time spent calling by males each night

after fly parasitism. Their results show that the proportion of time spent calling by








males remained unchanged the day after fly parasitism but declined sharply to

almost zero after the second day.

I calculated the lifetime fitness of each calling strategy using the formula:
d=60
1" Mi(d+l)(1-Ei)EI d-1
d=1

where d equals day and the length of the breeding season is 60 days. This

formula individually adds the daily fitness of each strategy over the 60 days of the

breeding season for all possible outcomes (i.e., male is parasitized on the first

day, male avoids fly parasitism on the first day and is parasitized on the second

day, male avoids fly parasitism on the first and second day and is parasitized on

the third day, etc.). Fitness was calculated using R 1.7.1 (Dalgaard 2002).




Seasonal Changes in Mating Benefit (M)

Long duration calling songs are more likely to attract females than short

duration calling songs in the fall but not the spring (Table 4-1; Fig. 4-2). This

seasonal pattern in female attraction suggests that the mating benefit of singing

should increase with calling duration in the fall but not the spring. I estimated the

mating benefit of the different calling strategies in each season using data from a

field experiment carried out in the fall and spring in which I counted the number

of females captured at pitfall traps fitted with speakers broadcasting male calling

songs differing in duration (1 h, 2 h, 3 h, and 4 h; Fig. 4-2). Because the number

of captured females increased with calling duration treatment in the fall, I used

the mean number of captured females per night for each treatment to estimate








the mating benefit of each calling strategy in the fall (Table 5-1). Because there

was no significant relationship between calling duration treatment and number of

captured females in the spring, I assumed that all calling strategies have a similar

mating benefit in the spring. I estimated this mating benefit by calculating the

mean number of captured females per night for all calling treatments combined

(Table 5-1). These seasonal changes in mating benefit with calling duration

indicate that males that call more attract more females in the fall but not in the

spring.

Female density may also affect the mating benefit of the different calling

strategies. Males should call less at high densities because females may be

easier to locate without calling and calling territories may be more expensive to

defend at high than at low densities (Alexander 1961 & 1975, Otte 1977, Cade

1979). Because cricket densities are higher in the fall than in the spring, males

should call less in the fall than in the spring. This potential seasonal effect of

density on the mating benefit of the different calling strategies is already taken

into account in the empirical estimates of seasonal changes in mating benefit due

to differences in female visitation rates. This is because the pitfall traps used to

collect these data captured females at their natural densities in each season.



Seasonal Changes in Probability of Escaping Fly Parasitism (E)

Calling males attract parasitoid flies (Cade 1975). Walker and Wineriter

(1991) experimentally muted G. rubens males by removing the right tegmen.

They found that muted males have lower fly parasitism rates (0/10) than unmuted








males (7/13) under field conditions at high fly densities. This finding suggests that

the probability of escaping fly parasitism should decrease with calling duration.

To estimate this relationship more precisely, I used data collected in the fall on

the arrival times of flies captured at a slit trap broadcasting male G. rubens

calling song (Chapter 2). The number of flies arriving at the trap decayed rapidly

with time (Fig. 2-1). These data can be used to estimate the probability of

escaping fly parasitism for different calling strategies because male calling start

times are inversely related to calling duration. To achieve this, I fitted a negative

exponential equation to the data on fly arrival times using a generalized mixed

linear model with a Poisson distribution. I used this equation to estimate the

probability of at least one fly infecting a male of each calling strategy. To

accomplish this, I used data collected in the fall on the probability of at least one

fly arriving at a slit trap broadcasting male G. rubens calling song for 4 h (Chapter

2). I set this probability as the y-intercept in the negative exponential equation I

derived earlier and used this equation to calculate the probability of at least one

fly infecting a male that sings for 1 h, 2 h, and 3 h. The outcome of these

manipulations is that the probability of escaping fly parasitism decreased with

calling duration in the fall (Table 5-1). This relationship between the probability of

escaping fly parasitism and calling duration indicates that males that call more

have lower survival rates and shorter lifespans than males that call less. This is

because the cumulative probability of escaping fly parasitism in this model

decreases more rapidly for long calling durations than for short calling durations








over the 60 days of the breeding season. The end result is that this model

incorporates the tradeoff between calling duration and life span for fall males.

Walker (1986) estimated that fly abundance in the spring is 1 % of the fly

abundance in the fall. This means that the probability of escaping fly parasitism

should be higher for all calling strategies in this season. I estimated this

probability for each strategy by increasing their counterpart fall probability by

99%. This assumes that there are 99% fewer flies in the fall than in the spring, fly

parasitism risk increases linearly with parasitoid fly abundance, and the

probability of escaping fly parasitism decreases with calling duration in the same

way as it does in the fall (Table 5-1).



Model Results: Seasonal Changes in Fitness of Male Calling Strategies

The fitness of calling strategies increased with calling duration in the fall

but not the spring (Fig. 5-1). In the fall, the strategy 'call 4 h' had the highest

fitness, and 'call 1 h' had the lowest fitness (Fig. 5-1). In the spring, the strategies

'call 1 h' and 'call 2 h' had the highest fitness, and 'call 3 h' and 'call 4 h' had the

lowest fitness (Fig. 5-1).



Comparison of Optimal Seasonal Strategies with Observed Calling Durations

Even though parasitoid flies are more common in the fall than in the spring

(Walker 1986, Chapter 2), male G. rubens spent more time calling in the fall than

in the spring (Fig. 2-5). This finding agrees with the predictions of my model. The

fitness of calling strategies increased with calling duration in the fall but not the








spring. The probability of escaping fly parasitism decreased with calling duration

in both the fall and spring. In the fall, however, this increased cost of singing with

calling duration was offset by a disproportionate increase in the mating benefit

with calling duration. This resulted in calling strategies with long calling durations

obtaining higher fitness than calling strategies with short calling durations. In

contrast, the increased cost of singing with calling duration in the spring was not

offset by the mating benefit because this benefit did not change with calling

duration. This resulted in calling strategies with short calling durations achieving

higher fitness than calling strategies with long calling durations.

A comparison of the mean calling duration of male G. rubens in the fall

and spring with the optimal seasonal calling durations predicted by my model

reveals that males sang about as much as expected in the spring (0.71 h vs. 1 or

2 h) and less than expected (1.71 h vs. 4 h) in the fall. One possible explanation

for this discrepancy is that my model did not include the energetic costs of

calling. Such costs are known to increase linearly with calling duration in trilling

species of crickets like G. rubens (Prestwich & Walker 1981). Fall males may

have not sung as much as predicted because it may be more costly to sing in the

fall than in the spring. This could occur if temperatures are higher in the fall than

in the spring. It is clear that if these seasonal changes in the energetic costs of

calling were incorporated into my model, then the predicted calling durations

would be lower than the current ones in the fall.








Variations of the Model

To evaluate the relative importance of natural versus sexual selection in

favoring seasonal patterns in male calling duration, I independently varied the

effects of fly parasitism and female visitation rates on male calling duration within

each season. In the first variation of the model, I drastically reduced fly

parasitism by setting the probability of escaping fly parasitism to 0.999 for all

calling strategies in the fall and spring (Table 5-2). I then examined the fitness of

the different calling strategies given only seasonal variation in the mating benefit

(Table 5-2). This would be the situation, for example, in a population with few

flies, which may be found in the northern extent of the G. rubens range. In the

second variation of the model, I held the mating benefit constant for all calling

strategies in the fall and spring by assigning all strategies the average mating

benefit for both seasons combined (Table 5-3). I then evaluated the fitness of the

different calling strategies given only seasonal variation in the probability of

escaping fly parasitism (Table 5-3). This would simulate the effect of a population

in which there was no relationship between calling duration and female visitation

rates.



Results of Model Variations

In the model with reduced and constant fly parasitism, and variable mating

benefit, the results are the same as when fly parasitism was included: the fitness

of longer duration calling strategies increased in the fall but not the spring (Fig. 5-

2). In the fall, the calling strategy 'call 4 h' had the highest fitness, and the








strategy 'call 1 h' had the lowest fitness (Fig. 5-2). In the spring, all calling

strategies had the same fitness (Fig. 5-2). In the model with variable fly

parasitism and constant mating benefit (Fig. 5-3), however, the results are not

the same as when mating benefit varied: the fitness of calling strategies 'call 1 h'

and 'call 2 h' were higher than the fitness of strategies 'call 3 h' and 'call 4 h' in

both the fall and the spring



Relative Importance of Natural versus Sexual Selection

Reducing fly parasitism and holding it constant did not affect the general

pattern of the original model: the fitness of longer calling duration strategies

increased with calling duration in the fall but not the spring in both models. This

finding suggests that sexual selection may be more important than natural

selection in favoring seasonal patterns in male calling duration. Male G. rubens

calling durations may have been higher in the fall than in the spring because long

duration calls were more effective at attracting females in the fall than in the

spring. This implication is significant because it suggests that variation in sexual

selection may be more important than variation in natural selection in shaping the

evolution of male traits.

Even though sexual selection may be more important than natural

selection, nonetheless sexual selection and natural selection clearly interact to

shape seasonal patterns in male G. rubens calling duration. In the model with

variable fly parasitism and constant mating benefit, short calling strategies had








higher fitness than long calling strategies. This suggests that natural selection

can modulate the effects of sexual selection on male calling duration.



Implications of the Model

The conclusions of this model enhance our understanding of how natural

and sexual selection may interact to favor variation in sexually selected male

traits. In G. rubens, male song attracts both females and parasitoid flies. Signals

that attract both mates and predators or parasites are common in species where

males have sexually selected traits (for review see Burk 1982). The outcomes of

my model show that in systems like this one, sexual selection may be more

important than natural selection in favoring variation in male traits, but that

natural selection can modify the effects of sexual selection.

Seasonal variation in calling duration has been observed in other Gryllus

species. Male G. texensis sing to attract females for mating and their calling song

also attracts 0. ochracea (Cade 1975). The risk of fly parasitism in male G.

texensis may increase with calling duration because the number of 0. ochracea

flies attracted to broadcasts of G. texensis song increases with calling duration

(Cade et al. 1996). Because 0. ochracea is more abundant in the fall than in the

spring, the risk of fly parasitism should be higher in the fall than in the spring.

Cade (1989) studied female G. texensis attracted to male calling song in late

summer and early fall. He found that the number of females attracted to male

song increased with calling duration. Unfortunately, Cade (1989) did not examine

female G. texensis attraction to male song in the spring. However, if female G.








texensis attraction to male song does not increase with calling duration in the

spring, as in the closely related species G. rubens, then G. texensis would be

similar to G. rubens in the seasonal changes in mating benefit and probability of

escaping fly parasitism. Accordingly, my model would predict that male G.

texensis should spend more time calling in the fall than in the spring. Bertram

(2002) found evidence for this prediction, although she had a different

explanation for the cause. This finding supports the general conclusions of this

model in explaining seasonal variation in male G. rubens calling duration.

This model is valuable because its formulation is general enough to

explain variation in any sexually selected male trait that attracts both females and

predators or parasites. Such tradeoffs in sexually selected traits are well known

(Burk 1982; Godin & Briggs 1996). For example, this model could be used to

evaluate geographical variation in male color pattems of guppies (Poecilia

reticulata). Male guppies show extensive geographical variation in their

conspicuous color pattems. Their bright colors are preferred by females (Kodric-

Brown 1985, Houde 1987, Long & Houde 1989) but at the same time make the

males more vulnerable to predation (Endler 1987). This variation has been

explained by a balance between female preference for conspicuous males and

natural selection against conspicuous males (Endler 1983). However, studies

over the last two decades have shown that female preferences also vary with

predation levels (Breden & Stoner 1987, Stoner & Breden 1988, Godin & Briggs

1996; Brooks & Endler 2001). Female preference for conspicuous males

disappears when predation is high. This means that the geographical variation in








male guppy color patterns may be due to differences among populations in

sexual selection as well as natural selection (Dill et al. 1999).

My model could help examine the relative importance of natural versus

sexual selection in favoring geographical patterns of male color patterns in

species like guppies. The strategies of the model could be different areas of

conspicuous body colors or different levels of bright coloration. The mating

benefit of the different color areas could be estimated from studies examining

female preference for male color patterns in low and high predation areas.

Similarly, the probability of escaping predation in geographic areas with different

color strategies could be estimated from studies examining the survival of males

under low and high predation levels. Once these parameters are calculated, they

could be manipulated using my model to assess the relative importance of

natural versus sexual selection in favoring geographical variation in male guppy

color patterns. If the gain in fitness through sexual selection (female preference)

shows an accelerating effect whereas natural selection (risk of predation) is

linear with color brightness, then the results will be similar to my model with

Gryllus: the effect of sexual selection has a stronger effect on driving variation in

male color variation than natural selection. This application shows how my

model could be used to examine variation in any sexually selected male trait that

attracts both females and predators.



















- I -


a)- -.
_C O _C P-1 r0
0) 01 0O qc o M co D
CO) ci i C


4)
E
c


















.L
0



.







(0

i4

= "i=









0V)
(0 "














Is



Co





i-E


-I 9* 4 4


I 9. 4


S S I I -


CD
a co




Lo




0
a-


4 9


















0o



Zt


E

I
Da,

M
'4-
0
o













cc
0
a,
CD
4-I




















CD
0
a


0,
a,
0

C






o








0 -W
t-x
0







a,
c




0)
(0-
iL


sseul4g ewUle!l


I I I

















- I I -


0) I 10)0
*L 0) 0) 0) 0)
(0 c ; 0 _


.4-.

E
CC


r>~
a-
Ei i



0'a

II
0 5




co m








CLU
(D
Sa,









cr~i
(0n >
'or
45 =
0)


E .S



ZS,







ri~c
HIC
.r c m


(0 cZE
Ci
M@c0

m 0
cc m
|| F


I I 4 4


1 4. 4


* I i -


1U.
0)
rr




4- .=
S'E
UE
.-0)
0-
P 2


2
n.
a.


- E



cc E
M: j
o)



Ou,
Co

S rE


4 t I 4. 4


















- Y -


4 4 4 4


j )
E
CU



8
E




|E
ECL
0) C


Ea
0 -

















CL -0
-0
CU.


0CU


















L6
Io


yam


4. I*4.


- 4 4. 4.


a a U -


0)

U,-


5C
C .F-


I-t

a.


4 4 4















0) *r
E..




4--
* i










ED-
>1






L oi
07



6 cc







) E
-M 0-
00 6 0 :.
SU e 0
(= (L)
---^ 1 I |
s
^J^ l









.t































LIIZ
















I III


sseulg ew!iejI1l


I


II


0



c,
I..
0)
(D= O



4co
0)



E
c
S 2







Cc'
4-.

L6.
iI


^1~(
^ 2


0.00


*go,

.uC0
(0.J
wEE







ct:a,
|



0)

US)






LLU
|12

lle
5.E
~ (0r
Q^> 0
u5?


-M














LIST OF REFERENCES


Alexander, R.D. 1957. The taxonomy of the field crickets of the eastern United
States (Orthoptera: Gryllidae: Acheta). Ann. Entomol. Soc. Am., 50,
584-602.

Alexander, R.D. 1961. Aggressiveness, territoriality, and sexual behaviour in field
crickets. Behaviour, 17, 130-223.

Alexander, R.D. 1975. Natural selection and specialized chorusing behaviour in
acoustical insects. In: Insects, Science and Society. (Ed. by D. Pimentel),
pp. 35-77. Academic Press, New York.

Andersson, M. 1994. Sexual selection. Princeton: Princeton University Press.

Baerends, G.P., Brouwer, R., & Waterbolk, H.T. 1955. Ethological studies on
Lebistes reticulatus (Peters), I. An analysis of the male courtship pattern.
Behaviour, 8, 249-334.

Bertram, S.M. 2002. Temporally fluctuating selection of sex-limited signaling
traits in the Texas field cricket, Gryllus texensis. Evolution, 56, 1831-1839.

Bertram, S.M. & Bellani, R. 2002. The influence of photoperiod on temporal mate
signaling patterns in the Texas field cricket, Gryllus texensis. J. Insect
Behav., 15,607-615.

Boake, C.R.B. 1984. Male displays and female preferences in the courtship of a
gregarious cricket. Anim. Behav., 32, 690-697.

Borgia, G. & Collis, K. 1989. Female choice for parasite-free male satin
bowerbirds and the evolution of bright male plumage. Behav. Ecol.
Sociobiol., 25,445-454.

Breden, F. & Stoner, G. 1987. Male predation risk determines female preference
in the Trinidad guppy. Nature, 329, 831-833.

Brooks, R. & Endler, J.A. 2001. Female guppies agree to differ: phenotypic and
genetic variation in mate-choice behavior and the consequences for
sexual selection. Evolution, 55, 1644-1655.








Buchholz, R. 1995. Female choice, parasite load and male ornamentation in wild
turkeys. Anim. Behav., 50, 929-943.

Burk, T. 1982. Evolutionary significance of predation on sexually signaling males.
Florida Entomol., 65, 90-104.

Cade, W.H. 1975. Acoustically orienting parasitoids: fly phonotaxis to cricket
song. Science, 190, 1312-1313.

Cade, W.H. 1979. The evolution of alternative male reproductive strategies in
field crickets. In: Sexual Selection and Reproductive Competition in
Insects. (Ed. by M.S. Blum & N.A. Blum), pp. 343-378. Academic Press,
New York.

Cade. W.H. 1981. Alternative male strategies: genetic differences in crickets.
Science, 212, 563-565.

Cade, W.H. 1984. Effects of parasitoids on nightly calling duration in field
crickets. Can. J. Zool., 62, 226-228.

Cade, W.H. 1989. Nightly and hourly rates of attraction of flying field crickets,
Gryllus integer, to conspecific song. Can. J. Zool., 67, 2540-2542.

Cade, W.H. 1991. Inter- and intraspecific variation in nightly calling duration in
field crickets, Gryllus integer and G. rubens (Orthoptera: Gryllidae). J.
Insect Behav., 4, 185-194.

Cade, W.H. & Cade, E.S. 1992. Male mating success, calling and searching
behavior at high and low-densities in the field cricket, Gryllus integer,
Anim. Behav., 43,49-56.

Cade, W.H., Ciceran, M., & Murray, A.M. 1996. Temporal patterns of parasitoid
fly (Ormia ochracea) attraction to field cricket song (Gryllus integer). Can.
J. Zool., 74, 393-395.

Cade, W.H. & Wyatt, D.R. 1984. Factors affecting calling behaviour in field
crickets, Teleogryllus and Gryllus (age, weight, density, and parasites).
Behaviour, 88, 61-75.

Crankshaw, O.S. 1979. Female choice in relation to calling and courtship songs
in Acheta domesticus. Anim. Behav., 27, 1274-1275.

Dalgaard, P. 2002. Introductory Statistics with R. Springer, New York.

Darwin, C. 1871. The descent of man, and selection in relation to sex. London:
Murray.








Dill, L.M., Hedrick, A.V., & Fraser, A. 1999. Male mating strategies under
predation risk: do females call the shots? Behav. Ecol., 10, 452-461.

Doherty, J.A. & Callos, J.D. 1991. Acoustic communication in the trilling field
cricket, Gryllus rubens (Orthoptera: Gryllidae). J. Insect Behav., 4, 67-82.

Endler, J.A. 1983. Natural and sexual selection on color patterns in poeciliid
fishes. Environ. Biol. Fishes, 9, 173-190.

Endler, J.A. 1987. Predation, light intensity and courtship behavior in Poecilia
reticulata (Pisces: Poeciliidae). Anim. Behav., 35, 1376-1385.

Fitzgerald, G.J., Foumier, M., & Morrissette, J. 1994. Sexual selection of an
anadromous population of threespine sticklebacks: no role of parasites.
Evol. Ecol., 8, 348-356.

Forrest, T.G., Sylvester, J.L., Tiesta, S., Smith, S.W., Dinep, A., Cupit, T.L.,
Huggins, J.M., Atkins, K.L., & Eubanks, M. 1991. Mate choice in ground
crickets Gryllidae Nemboiinae. Florida Entomol., 74, 74-80.

Forsgren, E. 1992. Predation risk affects mate choice in gobiid fish. Am. Nat.,
140,1041-1049.

French, B.W. & Cade, W.H. 1987. The timing of calling, movement, and mating in
the field crickets Gryllus veletis, G. pennsylvanicus, and G. integer. Behav.
Ecol. Sociobiol., 21, 157-162.

French, B.W. & Cade, W.H. 1989. Sexual selection at varying population
densities in male field crickets, Gryllus veletis and G. pennsylvanicus. J.
Insect Behav., 2, 105-121.

Godin, J.G-J. 1995. Predation risk and alternative mating tactics in male
Trinidadian guppies (Poecilia reticulata) Oecologia, 103, 224-229.

Godin, J.G-J. & Briggs, S.E. 1996. Female mate choice under predation risk in
the guppy. Anim. Behav., 51, 117-130.

Gray, D.A. 1997. Female house crickets, Acheta domesticus, prefer the chirps of
large males. Anim. Behav., 54, 1553-1562.

Hack, M.A. 1998. The energetic of male mating strategies in field crickets
(Orthoptera: Gryllinae: Gryllidae). J. Insect Behav., 11, 853-867.

Hedrick, A.V. 1986. Female preferences for male calling bout duration in a field
cricket. Behav. Ecol. Sociobiol., 19, 73-77.








Hedrick, A.V. 1988. Female choice and the heritability of attractive male traits: an
empirical study. Am. Nat., 132, 267-276.

Hedrick, A.V. 2000. Crickets with extravagant mating songs compensate for
predation risk with extra caution. Proc. R. Soc. Lond. B, 267, 671-675.

Hedrick, A.V. & Dill, L.M. 1993. Mate choice by female crickets is influenced by
predation risk. Anim. Behav., 46, 193-196.

Hedrick, P.W. 1986. Genetic polymorphism in heterogeneous environments: a
decade later. Ann. Rev. Ecol. Sys., 17, 535-566.

Hissmann, K. 1990. Strategies of mate finding in the European field cricket
(Gryllus campestris) at different population densities: a field study. Ecol.
Entomol., 15, 281-292.

Hoglund, J., Alatalo R.V., & Lundberg, A. 1992. The effects of parasites on male
ornaments and female choice in the lek-bredding black grouse Tetrao
tetrix. Behav. Ecol. Sociobiol., 30, 71-76.

Holtmeier, C.L. & Zera, A.J. 1993. Differential mating success of male wing
morphs of the cricket, Gryllus rubens. Am. Mid. Nat., 129, 223-233.

Houde, A.E. 1987. Mate choice based upon naturally occurring color pattern
variation in a guppy population. Evolution, 41, 1-10.

Houde, A.E. & Torio, A.J. 1992. Effect of parasitic infection on male color pattern
and female choice in guppies, Behav. Ecol., 3, 346-351.

Iwasa, Y., Pomiankowski, A., & Nee, S. 1991. The evolution of costly mate
preferences. II. The "handicap" principle. Evolution, 45, 1431-1442.

Kirkpatrick, M. 1987. The evolutionary forces acting on female mating
preferences in polygynous animals. In: Sexual Selection: Testing the
Alternatives. (Ed. by J.W. Bradbury & M.B. Andersson), pp. 67-82. Wiley,
Chichester.

Kodric-Brown, A. 1985. Female preference and sexual selection for male
coloration in the guppy. Behav. Ecol. Sociobiol., 17, 199-205.

Kolluru, G.R., Zuk, M., & Chappell, M.A. 2002. Reduced reproductive effort in
male field crickets infested with parasitoid fly larvae. Behav. Ecol., 13,
607-614.








Kraaijeveld, A.R., Emmett, D.A., & Godfray, H.C.J. 1997. Absence of direct
sexual selection for parasitoid encapsulation in Drosophila melanogaster.
J. Evol. Biol., 10, 337-342.

Lickman, K., Murray, A.M., & Cade, W.H. 1998. Effect of mating on female
phonotactic response in Gryllus integer (Orthoptera: Gryllidae). Can. J.
Zool., 76, 1263-1268.

Loher, W. & Dambach, M. 1989. Reproductive Behavior. In: Cricket Behavior and
Neurobiology. (Ed. by F. Huber, T.E. Moore, & W. Loher), pp. 43-82.
Comell University Press, Ithaca.

Long, K.D. & Houde, A.E. 1989. Orange spots as a visual cue for female mate
choice in the guppy (Poecilia reticulata). Ethology, 82, 316-324.

Lopez, S. 1998. Acquired resistance affects male sexual display and female
choice in guppies. Proc. R. Soc. Lond. B, 265, 717-723.

Magnhagen, C. 1991. Predation risk as a cost of reproduction. TREE, 6,
183-186.

Magurran, A.E. & Seghers, B.H. 1990. Risk sensitive courtship in the guppy
(Poecilia reticulata). Behaviour, 112, 194-201.

Olvido, A.E. & Mousseau, T.A. 1995. Effect of rearing environment on calling-
song plasticity in the striped ground cricket. Evolution, 49, 1271-1277.

Ott, R.L. 1993. An Introduction to Statistical Methods and Data Analysis. 4th ed.
Duxbury Press, Belmont.

Parker, G.A. 1983. Arms races in evolution-an ESS to the opponent-independent
costs game. J. theor. Biol., 101, 619-648.

Pomiankowski, A. 1987. The costs of choice in sexual selection. J. theor. Biol.,
128,195-218.

Prestwich, K.N. 1994. The energetic of acoustic signaling in anurans and
insects. Amer. Zool., 34, 625-643.

Prestwich, K.N. & Walker, T.J. 1981. Energetics of singing in crickets: effect of
temperature in three trilling species (Orthoptera: Gryllidae). J. Comp.
Physiol. 143, 199-212.

Real, L. 1990. Search theory and mate choice. I. Models of single-sex
discrimination. Am. Nat., 136, 376-405.








Reynolds, J.D., Gross, M.R., & Coombs, M.J. 1993. Environmental conditions
and male morphology determine alternative mating behavior in Trinidadian
guppies. Anim. Behav., 45,145-152.

Roff, D.A. 1984. The cost of being able to fly: a study of wing dimorphism in 2
species of crickets. Oecologia, 63, 30-37.

Sakaluk, S.K. & Snedden, W.A. 1990. Nightly calling durations of male
sagebrush crickets Cyphoderris strepitans: size mating and seasonal
effects. Oikos, 57, 153-160.

Simmons, L.W. 1988. Male size, mating potential and lifetime reproductive
success in the field cricket, Gryllus bimaculatus (De Geer). Anim. Behav.,
36, 372-379.

Simmons, L.W., Zuk, M., & Rotenberry, J.T. 2001. Geographic variation in female
preference functions and male songs of the field cricket Teleogryllus
oceanicus. Evolution, 55, 1386-1394.

Stoner, G. & Breden, F. 1988. Phenotypic differentiation in female preference
related to geographic variation in male predation risk in the Trinidad guppy
(Poecilia reticulata). Behav. Ecol. Sociobiol., 22, 285-291.

Veazey, J.N., Kay, C.A.R, Walker, T.J., & Whitcomb, W.H. 1976. Seasonal
abundance, sex ratio, and macroptery of field crickets in northern Florida.
Ann. Entomol. Soc. Am., 69, 374-380.

Wagner, W.E.Jr., Murray, A.M., & Cade, W.H. 1995. Phenotypic variation in the
mating preferences of female field crickets, Gryllus integer. Anim. Behav.,
49, 1269-1281.

Walker, T.J. 1986. Monitoring the flights of field crickets (Gryllus spp.) and a
tachinid fly (Euphasiopteryx ochracea) in north Florida. Florida Entomol.,
69, 678-685.

Walker, T.J. 1989. A live trap for monitoring Euphasiopteryx and tests with E.
ochracea (Diptera: Tachinidae). Florida Entomol., 72, 314-319.

Walker, T.J. 1993. Phonotaxis in female Ormia ochracea (Diptera: Tachinidae), a
parasitoid of field crickets. J. Insect Behav., 6,389-410.

Walker, T.J. 1998. Trilling field crickets in a zone of overlap (Orthoptera:
Gryllidae: Gryllus). Ann. Entomol. Soc. Am., 91,175-184.

Walker, T.J. 2000. Pulse rates in the songs of trilling field crickets (Orthoptera:
Gryllidae: Gryllus). Ann. Entomol. Soc. Am., 93, 565-572.






88

Walker, T.J. & Wineriter, S.A. 1991. Hosts of a phonotactic parasitoid and levels
of parasitism (Diptera: Tachinidae: Ormia ochracea). Florida Entomol., 74,
554-559.

Wilkinson, L. 1997. Systat 6.0 for Windows. Prentice Hall, Upper Saddler River,
New Jersey.

Zera, A.J. & Rankin, M.A. 1989. Wing dimorphism in Gryllus rubens: genetic
basis of morph determination and fertility differences between morphs.
Oecologia, 80,249-255.

Zuk, M. Simmons, L.W., & Cupp, L. 1993. Calling characteristics of parasitized a
and unparasitized populations of the field cricket Teleogryllus oceanicus.
Behav. Ecol. Sociobiol., 33. 339-343.














BIOGRAPHICAL SKETCH

Manuel J. Vl61ez was born in San Juan, Puerto Rico, on February 4, 1974.

He is the son of Manuel J. V61ez Jr., a college professor of invertebrate zoology

at the University of Puerto Rico, and Alsacia Bosch de V61ez, a social worker and

housewife. He grew up in San Juan, where he attended elementary school at

Colegio Nuestra Senfora de la Providencia, and middle and high school at

Colegio San Ignacio de Loyola. After graduating from high school, he left San

Juan and attended Comell University, where he earned a B.A. in biology with a

concentration in neurobiology and behavior. While at Comell University, he

conducted behavioral research both in the field and lab in a wide range of taxa

including fish, birds, and mammals. After graduating from Comell University, he

attended the University of Florida, where he earned an M.S. degree in zoology.

His master's thesis explored parental care strategies in the cichlid fish Aequidens

coeruleopunctatus. After he obtains his Ph.D. in zoology, he will attend Boston

University School of Law, where he will study intellectual property.








I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope
and quality, as a dissertation for the degree of Doctor of Philosophy.


fi. Jane Brockmann, Char
Professor of Zoology

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope
and quality, as a dissertation for the degree of Doctor of Philosophy.


Colette M. St. Mary ,
Associate Professor of Zoology

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope
and quality, as a dissertation for the degree of Doctor of Philosophy.


Benjaon Bolker
AssistOlt Professor of Zoology

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope
and quality, as a dissertation for the degree of Doctor of


"Marta L. Wayne
Assistant Professor of Zoology

I certify that I have read this study and that in my opinion it conforms to
acceptable standards of scholarly presentation and is fully adequate, in scope
and quality, as a dissertation for the degree of Doctor of philosophy.


Thomas J. W ler
Professor Emeritus of Entomology and
Nematology

This dissertation was submitted to the Graduate Faculty of the Department
of Zoology in the College of Liberal Arts and Sciences and to the Graduate
School and was accepted as partial fulfillment of the requirements for the degree
of Doctor of Philosophy.








This thesis was submitted to the Graduate Faculty of the Department of
Zoology in the College of Liberal Arts and Sciences and to the Graduate School
and was accepted as partial fulfillment of the requirements for the degree of
Doctor of Philosophy.


August 2004
Dean, Graduate School













LD


























UNIVERSITY OF FLORIDA

3 1262 08554 2628