Female Mate Choice and No Detected Predation ... - Oxford Academic

BEHAVIOR

Female Mate Choice and No Detected Predation Risk in Relation to the Calling Song of Oecanthus niveus (Gryllidae: Oecanthinae) JUDITH XOCHITL PONCE-WAINER

AND

RAUL CUEVA DEL CASTILLO1

Laboratorio de Ecologõá, Unidad de Biologõá, Tecnologõá y Prototipos, FES Iztacala, Universidad Nacional Auto´ noma de Me´ xico, Apartado Postal 314, Me´ xico 54090, Edo. de Me´ xico, Me´ xico

Ann. Entomol. Soc. Am. 101(1): 260Ð265 (2008)

ABSTRACT Oecanthus niveus (Gryllidae: Oecanthinae), a tree cricket species, inhabits the Tehuacań Valley, Mexico, year-round. Solitary males radiate calling songs that attract receptive females. The calling song is composed of discrete trains of Þve pulses with regular intertrain intervals. The courtship and mating behaviors are similar to other tree cricket species. In this study, we evaluated 1) the relationship between male body size and train duration, and the train period of its calling song; 2) female choice by exposing females to contrasting trains and train periods of the calling songs; and 3) the predation risk associated with the song patterns preferred by females. We hypothesized that 1) there is a signiÞcant relationship between male body size with train duration and the train period of calling songs, 2) females prefer song patterns associated with larger males, and 3) that a high predation risk is associated with these patterns. In line with our hypotheses, we found a signiÞcant relationship between male body size and song patterns. Females preferred calls with a train duration corresponding to a male 11% larger than an average male of the population. Nevertheless, no increase in invertebrate predation was detected in a Þeld experiment for this song pattern. The train duration of the song is a reliable indicator of male size that does not have associated predation costs. Females may obtain direct or indirect beneÞts, or both, as a result of their mate choice preferences. KEY WORDS mate choice, cricket, calling song, predation, Zapotitlań Salinas

A goal of sexual selection theory is to explain the evolution of structures and behaviors used in male courtship of many animal species (Darwin 1859). By mating with winners of maleÐmale contests or by actively selecting mates, females can obtain direct (e.g., greater production of eggs or enhanced rearing of offspring) or indirect beneÞts, or both. With indirect beneÞts, each generation of females may produce more attractive sons that can achieve greater mating success, or highly selective females may produce more viable offspring by mating with males who possess viability (good genes) and attractiveness alleles that are inherited by the sons and daughters (Andersson 1994). Songs can constitute a signal that females may use to assess potential mates. Cricket calling song is a reliable indicator of male condition because it is variable among individuals and energetically costly (Gerhardt and Hubert 2002). In several cricket species, song patterns are related with male body size and in general, large males produce song patterns that are preferred by females (GreenÞeld 2002). However, these songs can also attract acoustically orienting insect predators or parasitoids; thus, the song can represent a compromise between attractiveness and 1

Corresponding author: e-mail: [email protected].

avoidance of detection by predators (Zuk and Kolluru 1998, Simmons et al. 2001). O. niveus inhabits the Tehuacań Valley, Mexico, year-round. The highest population density occurs from October to December (48 crickets per ha) (Romero-Mata 2007). Body size variation has been observed in both sexes (as estimated by the foreleg length, size ranges from 5.5 to 9.1 mm in both sexes). Females show positive phonotaxis to the song of their co-speciÞcs males (Walker 1957). Sexual behavior in O. niveus is similar to other Oecanthus species (Brown 1999). Males initiate calling at dusk, between 1800 and 1900 hours, and end ⬇5Ð 6 h later. Males have been observed on Mirabilis oblongifolia (Gray) and Wigandia urens (Ruiz & Pavo´ n) Kunth plants (R.C.C., personal observation). Females oviposit on Montanoa sp., M. oblongifolia, and W. urens (R.C.C., personal observation). Larva and adults are preyed on by different spider species, including Peucetia viridans Hentz (Araneae: Oxyopidae) (Romero-Mata 2007) and a hemipteran. Moreover, tachinid ßies have been observed in the habitat. These parasitoid ßies attack orthopterans, and some species use host songs to locate hosts (Cade and Wyatt 1984, Walker 1993). However, before this work, we did not know whether ßies or other insect predators use sound to locate O. niveus.

0013-8746/08/0260Ð0265$04.00/0 䉷 2008 Entomological Society of America

January 2008

PONCE-WAINER AND DEL CASTILLO: MATE CHOICE AND PREDATION RISK ON O. niveus

The song patterns of orthopteran species that inhabit Neotropical areas, and their interaction with acoustically orienting predators, remains mostly unknown. In this work, we describe the characteristics of the calling song of the males of Oecanthis niveus (Gryllidae: Oecanthinae), its relationship with male body size, female preferences for male song patterns, and the predation risk associated with the song patterns. We hypothesized a relationship between song patterns and male body size. Moreover, we expected that females preferred song patterns associated with larger males, and that these songs had a higher predation risk.

Materials and Methods Capture and Rearing of Crickets. Cricket males used to quantify variability in song pattern were captured in the Þeld as nymphs between March and May 2006 on the banks of the Salado River, Zapotitlań Salinas. This site is a xerophytic shrubland located in the southern part of the state of Puebla, Mexico (18⬚ 16⬘ 59⬙ N; 97⬚ 26⬘ 51⬙ W) with an annual precipitation of 412 mm. In the laboratory, larvae were placed in groups of 10 in transparent plastic jars (⬇2,800 ml), and they were given cricket chow (FlukerÕs Cricket Farm Inc., Baton Rouge, LA) and water (in vials); food and water were supplied ad libitum. Insects were maintained at 22 ⫾ 1⬚C and on a photoperiod of 12:12 (L:D) h. Females and males were placed in individual plastic containers (459 ml) at maturity. Thus, they were separated physically but not acoustically. Ten days after reaching sexual maturity, the songs of the males were recorded (see below). Calling Song and Male Body Size. Male crickets were placed individually in a cylindrical 1-liter mesh plastic container. The recording session started after a 2-min acclimation period. We recorded calling songs using a Radio Shack electret condenser microphone; 33-3013 (70 Hz to 16 kHz) connected to a Marantz PMD 222 tape recorder and TDK MA90 metal bias cassettes. The microphone was located 5 cm away from a male, at room temperature (22 ⫾ 1⬚C). Recordings of cricket songs were analyzed by digitizing 10-s song samples from each male. Samples of the songs were digitized (22-kHz sampling rate) and analyzed using Avisoft SASLab Pro software. We measured the following calling song components: 1) frequency of maximum sound energy (peak frequency), 2) duration of Þve-pulse trains, and 3) the train period, from the beginning of one train to the beginning of the next (Fig. 1). Nine measurements were made per song character of each male cricket. A repeated measures analysis of variance (ANOVA) was performed to evaluate the repetitivity in train and train period durations. In addition, the intraclass correlation and the correlation among the three song parameters were estimated. Because both variables were repetitive (see Results), their average values were used in subsequent analyses. These and further analyses were performed using JMP 5.1 software (JMP 5.1, SAS Institute, Cary, NC).

261

To estimate the relationship between male body size and song patterns, the length and width of the right forewing and the right hind femur length of the males were measured using a digital caliper (Mitutoyo Corporation, Tokyo, Japan) to the nearest 0.1 mm. A principal components analysis (PCA) was used to summarize the variation of the three morphological variables. Before the analysis variables were log transformed to improve their linear relationship. We tested the relationship between the Þrst PC and the song components (Þve-pulse train duration, train period, and peak frequency) by linear regression. Song Patterns, Female Preferences, and Predation Risk. To analyze female choice for song patterns, two synthetic songs were generated using the recording of the song of a male with an average body size (foreleg, 8.2 mm; train duration, 0.177 ms; peak frequency, 2.37 kHz) as the template. The song was digitally manipulated using the software SASLab Pro, version 4.3 (Avisoft Bioacoustics, Berlin, Germany) at a sampling rate of 22.05 kHz. One synthetic song had a 250-ms Þve-pulse train duration sequence, which corresponds to the emission of a small male (foreleg, ⬇8 mm). The other song had a Þve-pulse train duration of 100 ms, which corresponds to a large male song pattern (foreleg, ⬇9.1 mm). The songs were recorded using Avisoft SASLab Pro software, connected to a Marantz SM 222 tape recorder, and TDK MA90 metal bias cassettes. The calls were broadcast simultaneously using two Panasonic RQ-CR07V walkmans plugged to Radio Shack 19-318 speakers with intensity adjusted to 70-dB SPL, measured at 30 cm from the source (TES-1350 Sound Level Meter). This intensity approximated the natural calling song of O. niveus under Þeld conditions. In line with the protocol of Brown et al. (1996), single females were placed in the center of a circular plywood arena (2 m in diameter), divided into 12 equal pie-shaped sections with drawn lines (Fig. 2). Calls were broadcast through one of the two speakers at opposite sides of the arena. Females were allowed to move freely about the ßoor of the arena until they crossed a line 1 m from the center. The sections from which the female exited and the time taken to exit were recorded. If the female exited from one of the three sections closest to a speaker she was judged to have shown positive phonotaxis. Trials in which the female did not exit the arena within 5 min were stopped and the female was run through the trial again at a later date. The side of the arena from which each song was broadcasted was randomized in each trial. We explored the potential for acoustically oriented predators activity the last week of September and October 2006. The synthesized songs used in the previous experiment were broadcast for 14 nights at two traps. The sampling period was divided in two intervals of seven nights, with an interval of one month in between. Each trap consisted of a cylindrical 30-cmdiameter plastic jar with sticky Tanglefoot placed on top of and around a Radio Shack 19-318 speakers situated on the bottom of the jar. Each speaker was connected to a Panasonic RQ-CR07V Walkman, and placed on a branch of W. urens plant at 1.5 m from the

262

ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA

Vol. 101, no. 1

Fig. 1. (a) Oscillogram of calling song of males of O. niveus showing a series of three trains composed of Þve pulses each. Duration of a Þve-pulse train (A) and period of the Þve train (B) of calling songs were analyzed. (b) Spectrogram. (c) Power spectrum of the O. niveus calling song.

January 2008


263

Table 1. Principal component analysis of three morphometric characters of males of O. niveus from Zapotitlán-Salinas, Puebla, Mexico, in the reproductive season of 2006 Variable

PC 1

PC 2

PC 3

Forewing length Forewing width Hind femur length % of total variance Eigen value

0.597 0.558 0.576 70.716 2.121

⫺0.160 0.786 ⫺0.596 16.851 0.505

⫺0.785 0.263 0.559 12.433 0.373

Eigen-vectors of the three principal components and percentage of explained variance is provided (N ⫽ 21).

Results

Fig. 2. Schematic diagram of the mate choice arena. Lines were drawn to divide the arena into 12 equal pieshaped sections. Calls were broadcast through one of the two speakers at opposite sides of the arena. Females were allowed to move freely about the ßoor of the arena until they crossed a line 1 m from the center (see methods).

ground (Fig. 3). The traps were separated by 4 m. The songs were broadcasted at 70 dB, at 30 cm. Broadcasts began at 1900 hours and ended at 0600 hours. After the night sampling, captured insects were collected and placed in recipients with alcohol 70% to be identiÞed in the laboratory. In addition, from July 2005 to August 2006, we went to Zapotitlań Salinas 1 d at the end of each month, to catch and bring adult males to the laboratory to determine whether they had parasitoids.

Calling Song and Male Body Size. The calling song of O. niveus in Zapotitlań-Salinas is composed of discrete Þve-pulse trains followed by regular intertrain intervals (Fig. 1a). The peak frequency is ⬇2.25 kHz. The Þve-pulse train duration and its period were consistent (F ⫽ 0.814; df ⫽ 7, 30; P ⫽ 0.58). Intraclass correlation for the three parameters was signiÞcant (rI Þve-pulse train ⫽ 0.958, rI Þve-pulse train period ⫽ 0.999, rI peak frequency ⫽ 0.999). The Þve-pulse train and train period were negatively and signiÞcantly correlated (r ⫽ ⫺0.518, P ⬍ 0.033), whereas the correlations between train duration with peak frequency (r ⫽ 0.340, P ⫽ 0.18), and between train period with peak frequency were not (r ⫽ ⫺0.050, P ⫽ 0.85). The PC 1 was considered an index of male body size because the Eigen-vectors of three morphological characters were positive (Table 1). The second and third PCs were discarded from the analysis because their Eigen values were lower than 1 (Quinn and Keough 2002). There were negative and signiÞcant relationships between male body size and Þve-pulse train duration (F ⫽ 15.558; df ⫽ 1, 19; P ⬍ 0.001; Fig. 4a) and peak frequency (F ⫽ 5.293; df ⫽ 1, 19; P ⫽ 0.033; Fig. 4b). However, the relationship between body size with train period was not signiÞcant (F ⫽ 2.513; df ⫽ 1, 19; P ⫽ 0.13). Song Patterns, Female Preferences, and Predation Risk. Females showed signiÞcant differences in their phonotaxis preferences (␹2 ⫽ 9, df ⫽ 1, P ⫽ 0.003). From 42 females, 27 were attracted to the large male song, whereas nine females were attracted to the small male song, and six did not respond. None of the adult males (n ⫽ 187) and nymphs (n ⫽ 391) brought to the laboratory were parasitized. Moreover, a number of insects such as moths, ants and ßies were found in the acoustic traps, but no parasitoids or other potential acoustical predator. Discussion

Fig. 3. Drawing of an acoustic trap. The trap consisted of a cylindrical 30-cm-diameter plastic jar with sticky Tanglefoot placed on top of and around Radio Shack 19-318 speakers situated on the bottom of the jar.

The song patterns of orthopteran species that inhabit Neotropical areas, and their interaction with acoustically oriented predators remains mostly unknown. In Zapotitlań Salinas, males of O. niveus produce highly repetitive discrete Þve-pulse trains fol-

264

ANNALS OF THE ENTOMOLOGICAL SOCIETY OF AMERICA r2= 0.42

0.4

a)

Train duration (s)

0.35 0.30 0.25 0.20 0.15 0.10 0.05 -4

-3

-2

-1

0

1

2

3

Body Size 1 (PC 1) r2= 0.20

2.45

b)

Peak frequency (kHz)

2.40 2.35 2.30 2.25 2.20 2.15

Vol. 101, no. 1

that parasitoid ßies have other speciÞc prey, or that the frequency of tree cricket songs may not be detected by the parasitoids. In addition, we must consider that the acoustic trapping only covered a small fraction of the time that the tree crickets can be found in the Þeld, and the attacks could occur at any other time. Moreover, in the sampling area there were singing males that could distract the parasitoids from the acoustic traps. The last sampling period coincided with the peak of maximum cricket population density. Nevertheless, the tree crickets could be temporally avoiding the phonological peak of parasitoids (Zuk 1993, Cade et al. 1996). However, we may discard this possibility because we did not Þnd evidence of parasitization in the Þeld-exposed crickets. In any case, it is still possible that potential larger predators, such as mice and lizards could detect crickets by their songs, and we were not able to detect them with the traps used in this study. If the calling song of males of O. niveus in Zapotitlań Salinas is not a direct target of acoustic predators, the evolution of the song patterns can be mainly modeled by female mate choice. However, the energetic costs associated with the sound emission can reduce the malesÕ expected survival (Hunt et al. 2004), and these costs can still constrain the evolution of song patterns that are preferred by females. The possibility to detect larger acoustic predators and energetic constraints that can be modeling the evolution of O. niveus song patterns in Zapotitlań Salinas and other populations in central Mexico will be evaluated in the future.

2.10 -4

-3

-2

-1

0

1

2

3

Body Size 1 (PC 1)

Fig. 4. Relationship between O. niveus male body size (PC 1) with (a) male Þve-pulse train duration and (b) peak frequency. Both regressions were signiÞcant (see Results).

lowed by regular interpulse intervals. Females were attracted to short pulse durations produced by large males, suggesting a directional preference of male song patterns. Because pulse duration is a reliable indicator of male size, the females can evaluate their potential mates through male song. The females may obtain direct or indirect beneÞts as a result of their mate preferences. If male body size and song patterns have genetic variance, females may gain in producing attractive sons that can achieve greater mating success (Charlesworth 1987). Moreover, females may receive direct beneÞts as result of their choice. In O. nigricornis large males have elevated protein levels in their glandular gift secretions (Bussie` re et al. 2005), and the females increase their fecundity and lifespan as a result of feeding on these secretions (Brown 1997). In several cricket species, the song may represent a compromise between attractiveness and avoidance of detection by predators (Zuk et al. 1998, Mu¨ ller and Robert 2002, Bertram et al. 2004, Bertram and Warren 2005). However, we did not Þnd evidence of acoustically oriented predator attacks on the Zapotitlań Salinas cricket population. Different alternative explanations may be possible; including the possibility

Acknowledgments We thank Darryl Gwynne and Juan Nun˜ ez-Farfań for valuable suggestions to improve the manuscript quality. Valuable suggestions on the manuscript also were made by two anonymous reviewers. The Universidad Nacional Auto´ noma de Me´ xico supported this work, through Programa de Apoyo a Proyectos de Investigacõón e Innovacõón Tecnolo´ gica (PAPIIT) grant IN IN216203 (to R.C.C.).

References Cited Andersson, M. 1994. Sexual selection. Princeton University Press, Princeton, NJ. Bertram, S. M., and P. S. Warren. 2005. Trade-offs in signalling components differ with signalling effort. Anim. Behav. 70: 477Ð 484. Bertram, S. M., S. X. Orozco, and R. Bellani. 2004. Temporal shifts in conspicuousness: Mate attraction displays of the Texas Þeld cricket, Gryllus texensis. Ethology 110: 963Ð 975. Brown, W. D. 1997. Courtship feeding in tree crickets increases insemination and female reproductive life-span. Anim. Behav. 54: 1369 Ð1382. Brown, W. D. 1999. Mate choice in tree cricket and their kin. Annu. Rev. Entomol. 44: 371Ð396. Brown, W. D., J. Wideman., M.C.B. Andrade, A. C. Mason, and D. T. Gwynne. 1996. Female choice for an indicator of male size in the song of the black-horned tree cricket, Oecanthus nigricornis (Orthoptera: Gryllidae: Oecanthinae). Evolution 50: 2400 Ð2411. Bussie`re, L. F., H. A. Basit, and D. T. Gwynne. 2005. Preferred males are not always good providers: female choice

January 2008


and male investment in tree crickets. Behav. Ecol. 16: 223Ð231. Cade, W. H., and D. R. Wyatt. 1984. Factors affecting calling behaviour in Þeld crickets, Teleogryllus and Gryllus (age, weight, density, and parasites). Behaviour 88: 61Ð75. Cade, W. H., M. Ciceran, and A. M. Murray. 1996. Temporal patterns of parasitoid ßy (Ormia ochracea) attraction to Þeld cricket song (Gryllus integer). Can. J. Zool. 74: 393Ð395. Charlesworth, B. 1987. The heritability of Þtness, pp. 21Ð 40. In J. W. Bradbury and M. B. Andersson [eds.], Sexual selection. Testing the alternatives. Wiley, New York. Darwin, C. 1871. The descent of man and selection in relation to sex. John Murray, London, United Kingdom. Gerhardt, C. H., and F. Hubert. 2002. Acoustic communication in insects and anurans. Chicago University Press, Chicago, IL. Greenfield, M. D. 2002. Signalers and Receivers. Mechanisms and evolution of Arthropods communication. Oxford University Press, Oxford, United Kingdom. Hunt, J., R. Brooks, M. D. Jennions, M. J. Smith, C. L. Bentsen, and L. F. Bussie`re. 2004. High-quality male Þeld crickets invest heavily in sexual display but die young. Nature (Lond.) 432: 1024 Ð1027. Mu¨ ller, P., and D. Robert. 2002. Death comes suddenly to the unprepared: singing crickets, call fragmentation, and parasitoid ßies. Behav. Ecol. 13: 598 Ð 606. Quinn, G. P., and M. J. Keough. 2002. Experiment design and data analysis for biologist. Cambridge University Press, Cambridge, MA.

265

Romero-Mata, A. 2007. Estructura poblacional y preferencia de oviposicio´ n de Oecanthus niveus (Orthoptera: Gryllidae) en Zapotitlań Salinas, Pue. B.S. thesis, Facultad de Ciencias, Universidad Nacional Auto´ noma de Me´ xico, Me´ xico. Simmons, L. W., M. Zuk, and J. T. Rotenberry. 2001. Geographic variation in female preference functions and male songs of the Þeld cricket Teleogryllus oceanicus. Evolution 55: 1386 Ð1394. Walker, T. J. 1957. SpeciÞcity in the response of female tree crickets (Orthoptera, Gryllidae, Oecanthinae) to calling songs of the males. Ann. Entomol. Soc. Am. 50: 626 Ð 636. Walker, T. J. 1993. Phonotaxis in female Ormia ochracea (Diptera: Tachinidae), a parasitoid of Þeld crickets. J. Insect Behav. 6: 389 Ð 410. Zuk, M. 1993. Calling characteristics of parasitized and unparasitized populations of the Þeld cricket Teleogryllus oceanicus. Behav. Ecol. Sociobiol. 33: 339 Ð343. Zuk, M., and G. R. Kolluru. 1998. Exploitation of sexual signals by predators and parasitoids. Q. Rev. Biol. 73: 415Ð 438. Zuk, M., J. T. Rotenberry, and L. W. Simmons. 1998. Calling songs of Þeld crickets (Teleogryllus oceanicus) with and without phonotactic parasitoid infection. Evolution 52: 166 Ð171. Received p11 April 2007; accepted 5 September 2007.