1 playbacks Katie Ehlers Department of Biology, University of Wisconsin Madison ABSTRACT In many animal species, females do not mate randomly but rather they choose males based on their courtship sign als. reproductive success. It was thus predicted that crickets would both respond to playbacks in an attempt to compete with the singing male and also alter their song in some way when singing competitively. Two different playbacks each treatment to determine its affect of their signal. Results show that the crickets did in fact change their song with respect to the number of elements per song and the rate of the song. It was concluded that crickets respond the same to one cricket as they do to many crickets; the number of cric kets used for a given playback did not affect the RESUMEN En varias especies animales, las hembras no escogen pareja al azar si no que escogen machos basadas en las seales de cortejo. En grillos, el canto puede no s er utilizado nicamente para la atraccin de pareja sino tambin para reducir el xito reproductivo de los competidores. Es as que predije que los grillos respondern a playbacks para competir con los machos cantando y tambin alterar su canto para aumen tar la competitividad. Dos playbacks diferentes fueron creados, uno con el canto un grillo solitario y el otro con un coro de grillos, y los individuos recibieron cada tratamiento para determinar si afect su seal. Los resultados muestran que los grillo s cambian su canto con respecto al nmero de elementos por canto y la tasa de canto. Se concluye que los grillos responden de la misma manera a cantos individuales o grupales, el nmero de grillos usados en los playbacks no afecta la probabilidad de respu esta. INTRODUCTION Insect songs generally consist of rhythmic sequences of relatively similar brief sounds, and with signals (Gogala 1995). In crickets, songs are only produced by males, and females analyze these songs not only to determine the species identity of the singer, but also to evaluate his quality (Pollack 2000). These male calls are species specific and ordinarily serve to attract conspecific femal es from a distance (Alexander 1962). The sexually receptive females will respond to these male acoustic calls by moving towards the calling male (Leonard and Hedrick 2009). It has been suggested that females prefer songs that are more energy demanding to produce. This is genetic quality (Pollack 2000). For example, it was found that female crickets tended to prefer
2 courtship songs with a high tick rate (amo unt of chirps) and long high frequency tick duration (how long the high frequency chirp lasted) (Rantala 2003). There are several strategies that crickets employ to compete with other males. Over singing, one such strategy, is where one insect produces it s normal call over a second caller. Masking, on the other hand, is where the insect produces a sound that muddles a competing these, are often part of a co mplex male male interaction and have been described for many insects, including crickets (Bailey et al. song, another tactic might be the use of acoustic signals to distract the searching female ( Bailey et al. 2000). Although this final strategy is not as popular as the other two, it is possible that male crickets are not attempting to compete with another male but instead, the individual is In most acoustic insects, the responding male changed the rate of similar song elements when exposed to playbacks (Bailey et al 2006). Based on this finding, predictions were made that a playback of another cricket call or mul tiple cricket calls would elicit a response in each group of competing males. It was believed that crickets would respond to a single call by increasing their rate. crickets songs should mean more competition. Therefore, given that the most inform ative part of and spacing between them, and the organization of these sounds (Popov et al 1974, in Pollack 2000), it was believed that acoustical analysis This study looks at how signals from other males change the song that any given individual makes. It was hypothes and treatment songs with regard to the rate of the songs and the frequency of different sounds pulses, but not the quality of the sounds elements. This is because the crickets are attempting t o consequently increasing their own likelihood of finding a mate. Finally, the prediction was made that more crickets would respond to the cricket chorus than would re because more individuals singing means more competition. MATERIALS AND METHODS Study Site This study was conducted in the Biological Station in Monteverde, Costa Rica. The unidentified species of crickets used in this study w ere hand collected from the forest surrounding the station. The crickets stayed in the lower lab unless they were receiving treatments. On nights when data were to be collected, they were kept together in the computer lab closet while awaiting treatment because it was quieter here than other areas of the station. For recordings, the crickets taken to closet in the attic, the darkest and quietest recording area that could be found.
3 Methods In November 2009, ten individuals were collected. These cricket s were placed in plastic Ziploc food storage containers that were housed in the lower lab of the Biological Station. The cricket houses contained one cricket, a small jar lid to hold water, and food, which alternated between dog food and lettuce. After the crickets were collected, the two different treatments, individual playback and the cricket chorus, were created using the songs of the crickets that had been gathered. One cricket was used in the individual playback while ten crickets were used in the cricket chorus. The playbacks were modified in Raven Version 1.2.1, an acoustics program. Each playback was divided into five different parts with each part lasting 55 seconds. The first, third, and fifth sections had the songs of crickets, were overla pping sections, while the second and the fourth parts were left blank, non overlapping sections. The different partitions in the playback were used to see if there was a difference in overlapping and non overlapping calls for the crickets. After each play back was created, they were transferred to an iPod so that they could be played to each individual. Crickets were recorded starting after 6:30pm. This time was chosen because after a few nights of observation, it was found that all of the crickets were calling after this time each night. recorded using a Marantz PMD660 digital recorder. Every recording was denoted by a different track number, and these track nu mbers were documented as to which individual they corresponded to. The initial recording would serve as the control for each individual. treatment was random in that som e received the individual playback first while others received the cricket chorus first. When it was their turn to receive a treatment playback, the cricket was placed in the attic and given between five to ten minutes to recover from being moved to a new the cricket responded to as well as any additional observati ons, were also noted. Acoustic Analysis After each individual underwent both treatments, acoustic analysis was performed using Raven. In Raven, different songs were selected that would serve as the data points. A song was defined as all of the chirps that were sung together. To be considered part of a given song, a chirp has to be within one second of the chirp that was last sung. Each song was composed of different elements. Each element represented a unique sound, or chirp. Within each song, thr ee elements were chosen for further analysis. The first and the last sound of each song along with an arbitrary sound in the middle of the song were selected in Raven, and data for each of these points was recorded. Thus for each song, three different da ta points were calculated. Using Raven, the high frequency, low frequency, delta frequency, and delta time were able to be computed for each element. Also, with the help of Raven, the total duration of each song along with the number of elements in each song were noted, and this information was used to calculate the rate of each song. With the control song, two of the recorded songs were randomly chosen. For each song, the three elements were chosen and the high frequency, low frequency, delta frequenc y, duration,
4 and number of elements per song were calculated. With the treatment tests, the same variables were calculated for each of the three elements per song. In this case, however, a total of four songs were chosen from each recording with two song s coming from non overlapping sections of the playback, and two songs coming from overlapping sections. Finally, the variables for the control data were averaged together and the treatment data were also averaged together to determine which parts of the s ongs were being changed, if any, when the males were exposed to playbacks. This was done so that there would only be one value for each variable per individual gh averaging all of the high frequency determined using Raven. RESULTS Individuals were not more likely to respond to the cricket chorus than they were to respond to the individual playback, or vice versa. To find figure this out, a chi squared test wa s run. The finding was that there was no significant difference between the two treatments being employed ( 2 = 0.202, df = 1, p = 0.653). Figure 1. Relationship between the treatment used and the response. As shown here, it is evident that there wa s no statistically significant difference in responses between the two treatments. When comparing the two treatments, it can easily be seen that similar numbers of the individuals responded to both treatments. Before (control) and after playback songs (c ombination of data for both treatments) were compared. Fewer elements were recorded after playback than before (paired t test, F = 2.135, p = 0.0307, df = 9). A similar decline was also found between the before and after average rate (paired t test, F = 2.568, p = 0.0151, df = 9). With both the average number of elements and the average rate, the before values were higher than the after values. Thus, the control songs were both faster and contained more elements than the songs recorded after the treatme nts were conducted.
5 Figure 2. Change in the average number of elements due to treatment. The number of elements before treatment was given was significantly higher than the number of elements after treatments were conducted. As shown here, for all ind ividuals, except three, had a larger before average number of elements. Individual 8 is missing because he did not respond to either treatment. Figure 3. Change in the average rate due to treatment. The rate of the before tre atment for each individual was significantly faster than the rate after treatment, which is clearly evident by this graph as all individuals have at least a slightly higher before rate than the after treatment rate. Note that Individual 8 is missing his a fter average rate because he did not respond to either of the playbacks. Finally, statistic analysis was done to see if there was a relationship between any of the variables (high frequency, low frequency, delta frequency, delta time, duration, number of elements, and rate) and whether or not the crickets responded only to the cricket chorus. Individual 8, however, was not included in these tests as he did not respond to either treatment. After running a t test for each of these variables, a significant result was discovered between the delta frequency and the response (F = 9.17, df = 7, p = 0.0192).
6 Figure 4. Relationship between response and average delta frequency with cricket chorus as the treatment. The maroon color rep resents no response while the lavender color signifies a response. As shown here, the individuals that respond tend have higher delta frequencies than did no t respond to either treatment. Additional Observations While recording crickets, it was noticed that some of the crickets jumped when the playbacks occurred. Notes were made as to the treatment received and the number of jumps. Interestingly, two of th e crickets, Individuals 10 and 12, jumped, and they both occurred when they were responding to the chorus. An additional observation was made that Individual 8 was the only cricket not to respond to either treatment. DISCUSSION Since there was no signif icant difference in response to the different treatments (Figure 1), it can be concluded that while crickets respond to other crickets, it does not matter if it is one or many. Instead of being influenced by the number of males they are competing against, a more important factor in determining whether or not males would respond to a competing individual could be the distance that the male is from another competitor. Therefore, other factors, such as the distance a male is to another singing male, must make the males more likely to respond rather than the number of males the individual is competing against. Future studies should study the effect of to other males. When comparing the before treatment with the after treatment, it was discovered that cricket songs were simpler and slower after hearing another cricket. They had less chirps spread out over a longer period of time. (Figures 2 and 3, respectively). These data shows that aspects finding that other acoustic insects adjust their calls in response to a neighbor (Schatral et al 1991). Disagreement arises, however when looking at the finding that there would be more elements in a song when the males were competing against another male (Rantala 2003). It is
7 plausible that the crickets are backing down from an extreme cricket that they deem as too competitive to ev en bother competing against him. Further research should be done to see if When looking a t which individuals responded to the cricket chorus and which ones did not, it was found that there was a significant difference in the average delta frequency with non responders having lower average delta frequencies (Figure 4). This demonstrates that c rickets with a smaller range between its high and low frequencies are less likely to respond to playbacks. These findings could relate to dominance relationships in male crickets. It would make sense that dominant males, who would presumably receive more mating opportunities than subordinate males, would have a larger variation between the high and low frequency points in their calls, which could be a desirable trait that females prefer. It has been found that female crickets favor louder calls, which ar e songs in which the upper part of the frequency spectrum occupies higher frequencies (Schatral et al 1991). There is possibly something about the fitness of the dominant male that enables him to increase his delta frequency while weaker males are not ab le to do this. It is also probable that having a wider range of frequencies while they sing, it could possibly mean that they are more likely to reach higher parts of the frequency spectrum. Finally, it could be also be reasonable to suggest that most of the responders to the cricket chorus were dominant since dominant males would be more likely to respond to a playback of many crickets due to the fact that they are naturally more aggressive than subordinate males. The jumping of Individuals 10 and 12 ca n be explained by the fact that they are most they were competing against in the playback. A study has shown that in male crickets, fights are started by an an tennal touch between the two males, which then leads to antennal fencing among other things. It was also demonstrated that after a fight, the loser, or the subordinate individual, changes his behavior from aggression to avoidance towards not only the domi nant individual, but to other individuals as well. This depression of agonistic behavior can be prolonged if an individual suffers three sequential loses (Iwasaki 2006). Therefore, it could be possible that some individuals were severely subordinate indi viduals, such as Individual 8, while other subjects were so highly dominant, as was the case with Individuals 10 and 12, that they were jumping around ready to fight when they felt that they were being challenged by the calls in the playbacks. Another pos sibility, however, is that the males who jumped were trying to get away from the super cricket. There are still many areas with cricket calls that are still not completely understand that warrant further research. For example, an experiment similar to the one done here could be combined with a test of dominance to see if losing or winning a fight makes you less likely to respond to a playback. It could be predicted that the winners would be more likely to compete against a playback while the subordinate i ndividuals would be less likely to respond. Another area where further research is needed is to see which aspects of female preference are the greatest, e.g., if females prefer a louder call or short chirps better. Finally, a repeated run of this experim songs, which could have caused discrepancies in the data.
8 ACKNOWLEDGEMENTS I am grateful to the Biological Station and staff for providing me with both the space to perform my project and the crickets the used in it. I thank Alan Masters for his constant wisdom and support, even all the way from Istanbul. I would also like to express my gratitude to Yimen Araya, who taught me how to record my cricket songs, analy ze them, and run my statistical tests. This project would not have been possible without the help of Pablo Allen Mange for his assistance and remarkable cricket catching abilities, since I would not have had very many subjects without him. Finally, I wou ld like to thank the other staff, Anjali Kumar and Jos Caldern Ulloa (Moncho), and the other CIEE students for their guidance and support throughout this entire experience. LITERATURE CITED Alexander, R. D. 1962. The role of behavioral study in cri cket classification. Systematic Zoology. 11: 53 72. Bailey, W., Macleay, C., and T. Gordon. 2006. Acoustic mimicry and disruptive alternative calling tactics in an Australian bushcricket ( Caedicia ; Phaneropterinae; Tettigoniidae; Orthoptera): does matin g influence male calling tactic? Physiological Entomology. 31: 201 201. Bailey, W. and Field, G. 2000. Acoustic satellite behavior in the Australian bushcricket Elephantodeta nobilis (Phaneropterinae, Tettidoniidae, Orthoptera). Animal Behaviour. 59: 3 61 369. Gogala, M. 1995. Songs of four cicada species from Thailand. Bioacoustics. 6: 101 116. Iwasaki, M., Delago, A., Nishino, H., and H. Aonuma. 2006. Effects of Previous Experience on the Agonistic Behaviour of Male Crickets, Gryllus bimaculatus Zoological Science. 23: 863 872. Leonard, A. and A. Hedrick. 2009. Single versus multiple cues in mate discrimination by males and females. Animal Behaviour. 77: 151 159. Pollack, G. 2000. Who, what, where? Recognition and localization of acousti c signals by insects. Neurobiology of behaviour. 10: 763 767. Popov, A., Shuvalov, V., Svetogorskaya, I., and A. Markovich, eds. 1974. Acoustic behavior and auditory systems in insects. Abh Rheinisch Westfal Akad Wiss. 53: 281 306. Rantala, M. and R. K ortet. 2003. Courtship song and immune function in the field cricket Gryllus bimaculatus Biological Journal of the Linnean Society. 79: 503 510. Schatral, A. and W. Bailey. 1991. Decisions during phonotaxis in the bushcricket Requena verticalis : do f emales change direction to alternative call? Ethology. 88: 320 330.
xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xlink http:www.w3.org1999xlink xmlns:xsi http:www.w3.org2001XMLSchema-instance
leader 00000nas 2200000Ka 4500
controlfield tag 008 000000c19749999pautr p s 0 0eng d
datafield ind1 8 ind2 024
subfield code a M39-00317
Alteracin de las canciones de los grillos debido a la competencia a travs de las grabaciones
Alteration of crickets songs due to competition via playbacks
In many animal species, females do not mate randomly but rather they choose males based on their courtship signals. With crickets, calls may not be used soley for attracting a mate; they can also be used to reduce a competitors reproductive success. It was thus predicted that crickets would both respond to playbacks in an attempt to compete
with the singing male and also alter their song in some way when singing competitively. Two different playbacks were created with one having a single crickets song and the other having a cricket chorus, and individuals received each treatment to determine its affect of their signal. Results show that the crickets did in fact change their song
with respect to the number of elements per song and the rate of the song. It was concluded that crickets respond the same to one cricket as they do to many crickets; the number of crickets used for a given playback did not affect the individuals likelihood to respond.
En muchas especies de animales, las hembras no escogen pareja al azar si no que escogen a los machos basadas en las seales de cortejo. En los grillos, los cantos no se pueden usar solamente para atraer a una pareja sino que tambin para reducir el xito reproductivo de los competidores. Es as que predije que los grillos respondieran a las grabaciones para competir con los machos cantando y tambin alterando su canto para aumentar la competitividad. Dos grabaciones diferentes fueron creadas, uno con el canto de un grillo solitario y el otro con un coro de grillos, y los individuos recibieron cada tratamiento para determinar si afecta su seal. Los resultados muestran que los grillos cambian su canto con respecto al nmero de elementos por canto y la tasa de canto. Se concluye que los grillos responden de la misma manera a cantos individuales o grupales, el nmero de grillos usados en las grabaciones no afecta la probabilidad de respuesta.
Text in English.
Sound production by insects
Monteverde Biological Station (Costa Rica)
Produccin de sonido por los insectos
Estacin Biolgica de Monteverde (Costa Rica)
Tropical Ecology Fall 2009
Ecologa Tropical Otoo 2009
t Monteverde Institute : Tropical Ecology