Evolution: Please provide concise explanation. thanku

the decrease in resistance because of an Phys. Soc. Jon. 64, 3626 (1995). tional Science and Technology (CREST) of JST, and increase of current in the path with the 13. T. Mori et al., ibid. 66, 3570 (1997). by a Grant-in-Aid for Center of Excellence Research from the Ministry of Education, Science, Sports and higher current. 14. K. Ogawa, W. Wei, K. Miyano, Y. Tomioka, Y. To- The induction of a conducting path in kura, Phys. Rev. B, in press. Culture of Japan. The work at JRCAT was supported 15. Y. Okimoto, Y. Tomioka, Y. Onose, Y. Otsuka, Y. by The New Energy and Industrial Technology De- the absence of light is clearly triggered by the Tokura, Phys. Rev. B, in press. velopment Organization. presence of the first path, because at con- 16. Supported in part by the Core Research for Evolu- 6 March 1998; accepted 7 May 1998 stant temperature Vind decreases with higher current flow through the conducting path and with closer proximity to the first path. The distance across which a metallic path Call Duration as an Indicator of Genetic Quality can be induced can be much greater than the in Male Gray Tree Frogs distance at which a change of reflectivity due to the formation of metallic patches can Allison M. Welch,* Raymond D. Semlitsch, H. Carl Gerhardt be detected. The trigger voltage is distinctly dependent on temperature, whereas the The "good genes" hypothesis predicts that mating preferences enable females to select width of a conducting path is not. mates of superior genetic quality. The genetic consequences of the preference shown The successful visualization of the local by female gray tree frogs for long-duration calls were evaluated by comparing the photoinduced I-M transition by the depicted performance of maternal half-siblings sired by males with different call durations. Off- differential reflection technique opens the spring of male gray tree frogs that produced long calls showed better performance during way for a variety of further experimental larval and juvenile stages than did offspring of males that produced short calls. These studies. Our observations already indicate data suggest that call duration can function as a reliable indicator of heritable genetic that the requirements for creating the tran- quality. sition and maintaining the transition are fundamentally different. With respect to ap- plications, the local I-M transition is a tool for switching the resistivity of a material by The "good genes" model of sexual selec- through increased fitness of offspring. This many orders of magnitude in a controllable tion predicts that some attributes of male prediction can be tested by evaluating the and observable way. The generation and re- courtship displays advertise genetic quality. relation between paternal call duration and moval of one or more conducting paths at Preferences for such attributes should allow the genetic quality of offspring. arbitrarily chosen spots of a sample is per- females to mate with high-quality males Male gray tree frog advertisement calls formed by the appropriate choice of external and thereby benefit indirectly through en- consist of rapidly repeated pulses. In dense parameters and monitored with visible light. hanced quality of offspring (1). Although choruses and in response to playbacks, males These features suggest an application of the the good genes hypothesis has been tested tend to increase call duration by increasing Downloaded from http://science. sciencemag.org/ on February 7, 2020 local photoinduced I-M transition in the several times (2), few studies have provided the number of pulses per call (1 1, 12). None- construction of optical switching devices. In direct genetic evidence supporting this hy- theless, some males consistently produce the experiment, a gap of 150 um between pothesis (3). Only one such study involved longer calls than others in the same acoustic the electrodes was chosen to simplify the a species in which females cannot benefit environment (7, 12-14). Although long imaging, and a regulated dc power supply was directly from their choice of mates (4). calls are usually produced at slow rates, used for experimental convenience. With a Because selection for direct benefits such as thereby keeping aerobic metabolic costs rel- gap width of 25 um, however, the applied courtship feeding or parental care should atively constant (1 1, 14), males that produce voltage could be reduced to the order of 1 V, overwhelm any selection for indirect (ge- long calls spend less time calling per night which can be provided by ordinary power netic) benefits (5), the role of good genes (11) and attend fewer choruses per season supplies. selection in the evolution and maintenance (8) than males that produce short calls. Long of female preferences is best tested in spe- calls thus appear to impose higher nonaero- REFERENCES AND NOTES cies in which females do not benefit directly bic costs than short calls. Call duration may, 1. A. J. Millis, Nature 392, 147 (1998). from mate choice. therefore, be an honest indicator of male 2. Z. Jirak, S. Krupicka, Z. Simsa, M. Dlouha, S. Female gray tree frogs (Hyla versicolor) genetic quality. Vratislav, J. Magn. Magn. Mater. 53, 153 (1985). strongly prefer male advertisement calls of We tested whether call duration indicates 3. Y. Tomioka, A. Asamitsu, H. Kuwahara, Y. Mori- moto, Y. Tokura, Phys. Rev. B 53, R1689 (1996). long duration in laboratory experiments (6, heritable genetic quality by using maternal . H. Yoshisawa, H. Kawano, Y. Tokura, ibid. 52, 7). In the field, females freely initiate mat- half-siblingships (half-sibships) to compare 13145 (1995). ings with calling males and do not always the performance of different males' off- 5. A. Asamitsu, Y. Tomioka, H. Kuwahara, Y. Tokura, Nature 388, 50 (1997). choose the first male encountered (7). Be- spring while experimentally controlling for 5. Y. Morimoto, H. Kuwahara, Y. Tomioka, Y. Tokura, cause males do not defend oviposition sites, all maternal effects. Maternal half-sibships Phys. Rev. B 55, 7549 (1997). offer nuptial gifts, or contribute parental care were generated by artificially crossing each V. Kiryukhin et al., Nature 386, 813 (1997). (8, 9), and no difference has been found in female with two males that had been giv- B. K. Miyano, T. Tanaka, Y. Tomioka, Y. Tokura, Phys. Rev. Lett. 78, 4257 (1997). fertilization success as a function of call du- ing calls of distinctly different durations in 9. W. Eberle et al., Appl. Phys. Lett. 68, 3329 (1996). ration (10), there are no apparent direct the same social environment (Table 1). O. W. Pretti, in Nonlinear Physics of Complex Systems. benefits of a female's mate choice. We there- Thus, within each maternal half-sibship, Current Status and Future Trends, J. Parisi, S. C. Muller, W. Zimmermann, Eds. (Springer-Verlag, Ber- fore predicted that females selecting mates one sibship was sired by a male with calls of lin, 1996), pp. 341-352. with long calls should benefit indirectly longer duration than the male siring the 11. N. Balkan, B. K. Ridley, A. J. Vickers, Eds., Negative other sibship. Because call duration varies Differential Resistance and Instabilities in 2D Semi- Division of Biological Sciences, University of Missouri, with chorus density, males' calls must be conductors, NATO Advanced Research Workshop, II Ciocco, Lucca, Italy (NATO ASI Series, Overijse, Columbia, MO 65211, USA. assessed in the same social context in order Netherlands, 1992). *To whom correspondence should be addressed. E-mail: to be validly compared. Thus, in 1995 we 12. Y. Tomioka, A. Asamitsu, Y. Morimoto, Y. Tokura, J. awelch@biosci.mbp.missouri.edu selected nine sets of two males that had 1928 SCIENCE . VOL. 280 . 19 JUNE 1998 . www.sciencemag.orgREPORTS been calling within 2 m of each other, and Missouri. among response variables (25), the main in 1996 we selected six sets of two field- Because the relative performance of dif- effect of call duration was significant at caught males that had been calling simul- ferent genotypes can vary significantly with both food levels in 1996 and showed the taneously in a small captive chorus. The environmental conditions (18), we reared same trend at both food levels in 1995 mean difference in call duration between the resulting tadpoles at two food levels, (Table 2), with offspring of males with long the long- and the short-caller in each set thereby creating an unfavorable and a favor- calls showing a general performance advan- was 10.1 pulses per call in 1995 and 15.8 able growth environment in which to com- tage over offspring of males with short calls. pulses per call in 1996 (Table 1); in labo- pare the performance of offspring (19). Com- The probability of obtaining these four mul- ratory experiments, female H. versicolor parison of our results with those from field tivariate results that independently support routinely base preferences on differences of studies indicates that our high food treat- the same directional hypothesis was calcu- as few as 2 pulses per call (15). Further- ment was a realistic approximation of con- lated as P = 0.0008 (Table 2) with the use more, the average call durations of individ- ditions encountered in nature (20). Tadpoles of a combined probability test (26). The ual males classified as long-callers did not from the crosses (1995, n = 538; 1996, n = multivariate tests therefore support the hy- overlap with the average call durations of 384) were raised individually in containers pothesis that offspring performance is pre- individuals classified as short-callers (16). filled with 1.0 liter of water in the laboratory dicted by paternal call duration. Long- and short-callers did not differ in at the two food levels; 15 tadpoles per family The specific benefits realized by off- body mass. External artificial fertilization were reared at each food level in 1995 and spring of long-callers differed among our allowed unambiguous assignment of pater- six tadpoles per family in 1996 (21). To experimental environments (Table 2). Be- nity, and rotation of the egg-stripping of assess offspring performance, we used several cause variation in the quality of the growth each female between the pair of males variables (22) that are important determi environment is predicted to influence the eliminated the possibility of effects of fer- nants of fitness in anurans, predicting future relation between larval growth and devel- tilization order (17). In 1995, each of nine survival and age and size at maturity, which opment (27), this difference in responses gravid females was artificially crossed with influence lifetime reproductive success (23). among environments is not unexpected. a different set of males to generate nine Offspring of males with long calls always The consistency of the general benefit re- maternal half-sibships. In 1996, each of performed significantly better than or not alized by offspring of long-callers in our 11 gravid females was artificially crossed significantly differently from offspring of experimental environments suggests that a with at least one set of males to generate males with short calls (24) (Table 2). In general performance advantage may be ap- 16 maternal half-sibships. All frogs were multivariate analyses where responses were plicable in other environments as well. collected near a pond in Boone County, combined to account for correlations Overall, these results provide strong evi- dence that males with long calls relative to those of other males in the same social Table 1. Average calling performance of sires exhibiting long versus short calls. For each 1995 male, environment sired offspring of significant- approximately 25 consecutive calls were analyzed from field recordings. For each 1996 male, at least 20 ly higher phenotypic quality than males Downloaded from http://science.sciencemag.org/ on February 7, 2020 min of consecutive calls were analyzed from digitally collected data with short relative call durations. We can Call duration attribute these observed phenotypic differ- Year Performance Calling effort* ences to differences in paternal genetic Pulses per call Seconds contribution, because our comparison of maternal half-sibships controls for mater- 1995 Long-callers 28.3 1.74 0.214 nal genetic contributions and maternal Short-callers 18.2 1.05 0.188 effects. Thus, our results demonstrate that Difference 10.1 + 4.91 0.69 + 0.36+ 0.026 + 0.040# 1996 Long-callers 31.5 1.41 0.092 relative call duration reliably reflects ge- Short-callers 15.7 0.68 0.082 netic quality in H. versicolor. Our data Difference 15.8 + 4.6+ 0.72 + 0.24 0.010 + 0.048 suggest a genetic correlation between sire call duration and offspring performance, *Calling effort was measured as duty cycle-the proportion of time during which the individual was producing sound. +P 0.05; paired t test. which implies that each trait has a heri- cable basis. The preference for long calls should, therefore, enable females to select Table 2. Relative performance of offspring of males exhibiting long versus short calls. A shorter larval high-quality mates and benefit indirectly period is interpreted as better performance. For all other variables, larger values indicate better perfor- through increased fitness of offspring. Be- mance. NS, not significant. Dashes indicate data not collected in 1995 cause female H. versicolor do not gain direct benefits from their choice of mate, 1995 1996 the indirect genetic benefits we have doc- Parameter High food Low food High food Low food umented suggest good genes selection as a probable explanation for the evolution Larval growth NS Long >shortt Long >> short Long > short+ and maintenance of the female preference Larval period Long >> short NS Long > short NS in this species. Metamorphic mass NS Long > short NS NS Larval survival Long > short NS NS NS Postmetamorphic NS Long > short REFERENCES AND NOTES growth 1. A. Zahavi, J. Theor. Biol. 53, 205 (1975). MANOVA A = 0.96, df = 3, A = 0.81, df = 3, > = 0.90, df = 4, A = 0.71, df = 4, 2. R. D. Howard, H. H. Whiteman, T. I. Schueller, Evo- P = 0.0887 P = 0.0590 P = 0.0143 P = 0.0216 lution 48, 1286 (1994); R. D. Semlitsch, Behav. Ecol. Sociobiol. 34, 19 (1994); B. D. Woodward, Am. Nat. Combined probability x2 = 26.67, df = 8, P = 0.0008 128, 58 (1986); S. L. Mitchell, Evolution 44, 502 test (1990). 3. K. Norris, Nature 362, 537 (1993); B. C. Sheldon, J "NS = P > 0.05; univariate ANOVA (24). +Long >> short = P short = 0.05 > P > 0.01. Merila, A. Qvarnstrom, L. Gustafsson, H. Ellegren, www.sciencemag.org . SCIENCE . VOL. 280 . 19 JUNE 1998 19291. Evaluate the strength of evidence provided in the paper that long calls impose a significant cost to males, as expected for \"honest indicators\" of male quality. What kinds of experiments could be done to bolster the \"honest indicator\" idea? 2. What is main experimental value of the maternal half-sib design used in these experiments. 3. Strong good genes selection should result in the removal of lower fitness genotypes. Assuming the conclusions of this paper are sound, can you propose an explanation why there might be so much variation in male genetic quality? 4. What is your evaluation of the strength of support in Table 2 for the idea that long-call males sire higher quality offspring