A Day Gecko Darkens its Body Color in Response to Avian Alarm Calls

Current Herpetology 32(1): 26 33, February 2013 2013 by The Herpetological Society of Japan doi 10.5358/hsj.32.26 A Day Gecko Darkens its Body Color in Response to Avian Alarm Calls RYO ITO 1,2 *, ISAMI IKEUCHI 1,3, AND AKIRA MORI 1 1 Department of Zoology, Graduate School of Science, Kyoto University, Sakyo, Kyoto 606 8502, JAPAN 2 Present address: Primate Research Institute, Kyoto University, Kanrin 41 2, Inuyama, Aichi 484 8506, JAPAN 3 Present address: Takinocho 2 19 3 202, Nagaokakyo, Kyoto 617 0817, JAPAN Abstract: Rapid body color change of animals in response to environmental stimuli has at least three biological functions: predation avoidance, thermoregulation, and intraspecific communication. We tested the hypothesis that Phelsuma kochi, a Madagascan giant day gecko that normally has a bright green body color, darkens its color to maximize its level of background matching so as to evade predation. Because recent studies revealed that some lizard species are able to eavesdrop on avian alarm calls and respond with antipredator behavior, we conducted a playback experiment of avian alarm calls to examine whether P. kochi recognizes alarm calls and changes its body color in response to them. We played back alarm calls and songs of a syntopically occurring passerine bird, Terpsiphone mutata, and white noise to free-ranging geckos. The geckos changed their body color quicker and darker in response to alarm calls than songs, and they tended to keep their dark coloration for a longer duration after the playback of alarm calls than that of songs or white noise. This result suggests that P. kochi is able to eavesdrop on alarm calls of syntopic birds and respond by darkening its body color to reduce its conspicuousness to predators. Key words: Eavesdropping; Gekkonidae; Phelsuma kochi; Body color change; Antipredator behavior; Madagascar INTRODUCTION The ability to change body color in response to environmental stimuli has been reported in numerous animals such as fish (Mäthger et al., 2003), amphibians (Camargo et al., 1999), reptiles (Stuart-Fox et al., 2006), crustaceans * Corresponding author. Tel: +81 568 63 0554; Fax: +81 568 62 9552; E-mail address: ryo.j.h.ito@gmail.com (Hemmi et al., 2006), and cephalopods (Hanlon, 2007). Various biological functions of this physiological color change have been documented, such as predation avoidance, thermoregulation, and intraspecific communication (Sherbrooke et al., 1994; Stuart-Fox et al., 2006; Stuart-Fox and Moussalli, 2008). Phelsuma kochi is a diurnal gecko distributed over the northwestern and western coasts of Madagascar (Glaw and Vences, 2007). This species is a relatively large, arboreal gecko

ITO ET AL. GECKO COLOR CHANGE AGAINST BIRD CALL 27 (total length reaches up to 200 mm) that often perches on tree trunks and branches to ambush small arthropods (Ikeuchi et al., 2005). It normally has bright green body color and is very conspicuous, at least to human eyes, when it perches on trees. Phelsuma geckos are known for their ability to quickly darken body coloration when stressed (Glaw and Vences, 2007), but the biological function of this has not been investigated systematically. Here, we tested the hypothesis that P. kochi changes body color to reduce conspicuousness as an antipredator tactic. To elicit antipredator responses in the geckoes, we used auditory stimuli. This is because this gecko is known to have well-developed ears (Wever, 1978) and because recent studies revealed that some species of iguanian lizards are able to eavesdrop on avian alarm calls and respond to them by increasing vigilance against nearby predators (Vitousek et al., 2007; Ito and Mori, 2010). Specifically, we examined whether P. kochi darkens its body color in response to alarm calls of a passerine bird, the Madagascar paradise flycatcher. Therefore, this experiment tested the hypothesis that P. kochi is able to eavesdrop on and recognize avian alarm calls as well as the hypothesis that its body color change has an antipredator function. MATERIALS AND METHODS Study site and subjects We conducted a field experiment in the dry forest of Ampijoroa, Ankarafantsika National Park, in northwestern Madagascar (16 15'S and 46 48'E) in 2005. The forest vegetation consists of a deciduous canopy 10 15 m high and a fairly sparse understory. It is split by road Route Nationale (RN) 4 into an eastern and a western part. Lake Ravelobe (area ca. 30 ha) is situated along the eastern side of RN 4. Our main study area was along the lake and its immediately surrounding areas. At least 86 avian, 47 reptilian, nine amphibian, and 20 mammalian species inhabit Ampijoroa forest (Mizuta, 2005; Mori et al., 2006; Ito et al., unpublished data). Phelsuma kochi is a common lizard in the animal community of Ampijoroa forest (Mori et al., 2006). The Madagascar paradise flycatcher, Terpsiphone mutata, is a passerine bird, widely distributed in Madagascar, and inhabits Ampijoroa forest syntopically with P. kochi (Mizuta, 2005). This bird communicates vocally and often emits alarm calls while mobbing raptors such as Frances s sparrow hawk (Accipiter francesii), a predator of both T. mutata and P. kochi (Rene de Roland and Thorstrom, 2003; Hasegawa et al., 2009). The flycatcher and the day gecko do not have direct ecological interactions such as predator-prey, resource competition, and host-parasite interactions. Data collection We conducted the playback experiment between 1000 h and 1600 h in December, when P. kochi and T. mutata are reproductively active (Mizuta, 2002a; Ikeuchi et al., 2005). The playback experiment was conducted to test whether P. kochi distinguishes songs of T. mutata from its alarm calls that were emitted against nearby predators. We employed three types of sound stimulus: alarm calls, songs, and white noise. In preparation for the playback experiment, we recorded alarm calls of an individual of T. mutata when it mobbed a Frances s sparrow hawk. Songs of T. mutata were selected from three sources, which were collected from three males. We used a Sony digital audio player (TCD-D10) to record songs and calls and edited them to make 60 s stimulus each, using the sound editing software Avisoft-SASLab Light (Fig. 1). We used three sources of white noise and created 60 s control stimulus (Fig. 1) to test the effects of unnatural sounds. Prior to the experiment, we searched for geckoes by slowly walking along paths and trails. Upon finding a gecko perching on a tree trunk or branch, we started a playback experiment. We put a Yuasa speaker (YSP-588B) and a Sony digital video camera 3 m from the individual. To acclimate the gecko to the presence of the observers, we waited for 1 min before the initiation of the experiment after setting up the equipment. Because we played

28 Current Herpetol. 32(1) 2013 FIG. 1. Sonograms of sound sources used in the playback experiment. Alarm calls of Terpsiphone mutata (A); songs of T. mutata (B); white noise (C). FIG. 2. Schematic sequence of the playback experiment. Each experiment was composed of three trials (alarm call, song, and white noise trials), and each trial consisted of three periods (baseline, playback, and silence periods). The order of trails within an experiment was randomized. back all of the three auditory stimuli (alarm calls, songs, and white noise) to each gecko, each experiment was composed of three trials (Fig. 2). In each trial, we played back a CD of one of the three types of sound stimulus to the individual with a Panasonic portable CD player (SL-CT500). Each trial was separated by an interval of 60 to 180 s. Each trial was composed of three periods; a baseline period: a 60 s duration immediately before a sound stimulus was played back; a playback period: a 60 s duration when a sound stimulus was played back; and a silence period: a 120 s duration immediately after playback was turned off (Fig. 2). In the playback period, each sound stimulus was turned up to the maximum volume in the first 10 s and turned down to the minimum volume in the last 10 s. Mean maximum amplitudes of each sound stimulus, A-weighting, at 3 m away from the speaker were 80.2, 81.9 and 71.2 dba for alarm calls, songs, and white noise, respectively. The order of playback stimuli was randomized. The body color change of the gecko was recorded by the video camera throughout the experiment. During the experiment, to reduce possible observer s effects, we did not gaze at the subjects directly and stayed motionless behind

ITO ET AL. GECKO COLOR CHANGE AGAINST BIRD CALL 29 the video camera as much as possible. We stopped experiments if any birds uttered during the trials. To avoid using the same individuals multiple times, we conducted experiments at least 50 m apart from each other, which is larger than home range diameter of P. kochi (Ikeuchi et al., 2005). Most subject individuals eluded capture after the experiments. We did not use small individuals and conducted the experiment using 10 adult individuals. Variables and analyses To digitize the sequential brightness change, we divided the video records of each trial into 1/8 s sections and calculated mean brightness of the dorsal body for each section during the experiment, using NIH Image (a public domain image processing and analysis program from National Institute of Mental Health of USA). Because some individuals did not have intact neck skin, and others were missing their tails, we analyzed color changes of only the dorsal part of the body. Brightness is presented in 256 grades: 0 indicates darkest, and 255 indicates brightest. We recorded three types of variables in each trial: start latency, recovery time, and degree of darkening. We compared each variable among the three sound stimuli. Start latency is the time from the beginning of the playback period to the time brightness decreased to the minimum brightness during the preceding baseline period (MinBB) (Fig. 3). If brightness did not decrease to MinBB, we recorded start latency as 60 s. Thus, the range of start latency is 0 to 60. Recovery time is the time from the beginning of the silence period to the time the brightness rose up to mean brightness during the preceding baseline period (MeanBB) (Fig. 3). If brightness did not recover in 120 s, we stopped the trial and recorded recovery time as 120 s. If start latency was 60 s (i.e., brightness did not decrease to MinBB during the playback period) or if brightness returned up to MeanBB during the playback period, we recorded recovery time as 0 s. Thus, the range of recovery time is 0 to 120. The degree of FIG. 3. Schematic explanation of recorded variables. (a) Start latency: the time from the beginning of the playback period to the time brightness fell down to the minimum brightness of the baseline period. (b) Recovery time: the time from the beginning of the silence period to the time brightness rose to mean brightness of the baseline period. (c) Degree of darkening: difference between mean brightness during the baseline period and the minimum brightness of the playback period. darkening is the difference between MeanBB and the minimum brightness during the playback period (Fig. 3). If the minimum brightness during the playback period was not darker than MeanBB, we recorded the degree of darkening as 0. Thus, the range of the degree of darkening is 0 to 255. RESULTS In general, geckos darkened their body color in response to the playback sounds but recovered during the subsequent silence period (Fig. 4). Figure 5 depicts a typical body color change from the baseline to playback periods. Start latency was significantly shorter for alarm calls than for songs (Fig. 6A; paired t- test; t=4.29, P=0.0013). There were no significant differences in start latency between alarm calls and white noise (t=1.17, P= 0.264), and between songs and white noise (t=1.43, P=0.181). Recovery time in white noise trials was significantly shorter than that in alarm call and song trials (Fig. 6B; paired t- test; former, t=5.02, P<0.001; latter, t=3.00, P=0.013). Recovery time tended to be longer for alarm calls than songs, but it fell short of

30 Current Herpetol. 32(1) 2013 darkening between alarm calls and white noise (z=1.07, P=0.320), and between songs and white noise (z=0.44, P=0.700). DISCUSSION FIG. 4. Examples of temporal patterns of body color change of Phelsuma kochi against alarm calls (A), songs (B), and white noise (C). the significant level (t=2.07, P=0.065). The degree of darkening was larger in alarm call trials than in song trials (Fig. 6C; Wilcoxon signed rank test: z=2.31, P=0.019). There were no significant differences in the degree of Our results showed that individuals of P. kochi changed their body color quicker and darker in response to alarm calls of the flycatcher than to its song. In addition, they tended to keep their dark coloration longer after hearing alarm calls than after hearing songs or white noise. These results support the hypotheses that P. kochi is able to discriminate the alarm calls from the songs and that the gecko changes its body color as an antipredator response. The absence of the statistical significance, albeit nearly significant, in recovery time between alarm calls and songs may have at least two mutually non-exclusive reasons. First, it may have been caused by the fact that the sound pressure level of songs was slightly (1.22 times) higher than that of alarm calls. Louder sound may be perceived by the gecko as a threatening stimulus, which may lead to a prolonged antipredator reaction (see Jones and Jayne, 2012). Second, uncontrolled environmental factors during the silence period, such as auditory stimuli from nearby freeranging birds, may have affected recovery time. Actually, in one song trial, several birds flew by and uttered alarm calls while we were measuring recovery time, and the test individ- FIG. 5. Body color change of Phelsuma kochi in response to alarm calls. Before the playback of alarm calls (A); minimum (darkest) brightness during the playback of alarm calls (B).

ITO ET AL. GECKO COLOR CHANGE AGAINST BIRD CALL 31 FIG. 6. Comparisons of starting latency (A), recovery time (B), and degree of darkening (C) among three sound stimuli in the playback experiment. Values are given as mean±se for (A) and (B). Box plots are shown for (C): The boxes indicate the first to the third quartiles; thick lines across the boxes show medians; bars on and from the edge of the boxes show upper and lower adjacent limits. Dots out of the upper and lower adjacent limits indicate outliers. ual, which had almost recovered its normal body color, darkened again. Although start latency and darken degree did not significantly differ between alarm call and white noise trials, the sequential pattern of the brightness change against white noise was obviously different from that against alarm calls. The geckos changed their body color to white noise as rapidly and dark as to alarm calls, but the body color quickly returned to the normal range (up to MeanBB) in white noise trials. Actually, recovery time in white noise trials is nearly 0, which means the geckos started reverting to their normal color even while the white noise was still audible. This strong but short reaction to white noise may indicate that the geckos responded to a novel (artificial) sound with color change but immediately considered it as a non-threatening sign so as to recover the body color. Talling et al. (1996) suggested that a novel sound is an arousing stimulus that initially activates the defense mechanisms in animals. Several descriptive reports and a few recent experimental studies suggest that rapid body color change of geckos may serve as functions of intraspecific communication and predation avoidance by background matching (Zaidan and Wiebusch, 2007; Vroonen et al., 2012). Although P. kochi with darkened body color in our experiment was not completely cryptic on the tree trunk surface, the darker body of the gecko was much less conspicuous than its normal body color, at least to human observers. Natural predators of P. kochi, such as birds and snakes, would have the color discrimination ability different from that of humans, but achromatic discrimination could be similar to humans (see Sillman et al., 1997; Siddiqi et al., 2004; Stuart-Fox et al., 2008). Thus, the observed body color change of P. kochi would be effective to reduce the probability of being detected by predators. Experimental study to test the effects of darker body color to evade predation is desired. At present, two species of iguanian lizards (Amblyrhynchus cristatus and Oplurus cuvieri) are known to be able to eavesdrop on avian alarm calls and respond with antipredator behavior (Vitousek et al., 2007; Ito and Mori, 2010). Our study is the third report of reptiles that eavesdrop on heterospecific alarm calls. Interestingly, two of them (O. cuvieri

32 Current Herpetol. 32(1) 2013 and P. kochi) are observed in Ampijoroa forest. In this forest, more than 160 species of terrestrial vertebrates coexist (Mizuta, 2005; Mori et al., 2006; Ito et al., unpublished data), constituting complex interspecific interaction networks that include predator-prey relationships and resource competition both within and between vertebrate classes (Mizuta, 2002b; Mori and Randriamahazo, 2002; Nakamura, 2004; Nakamura et al., 2004; Hasegawa et al., 2009). Whether this species-rich, ecologically complex community has facilitated the local evolution of the heterospecific eavesdropping by lizards or eavesdropping by lizards is a more ubiquitous phenomenon that has been unexplored is a challenging question worthy for pursuing. ACKNOWLEDGEMENTS We are grateful to M. Nakamura, T. Mizuta, H. Takahashi, H. Sato, M. Hasegawa, B. Razafimahatratra, T. M. Randriamboavonjy, T. Randrianarisoa, F. D. Hanitrininosy, and zoological researchers working in Ampijoroa forest for their assistance in the field, to T. Mizuta for his technical help in recording and editing birds songs and calls, and to F. Rakotondraparany and H. Rakotomanana for their help in arranging and conducting this research. We also thank the staff of Madagascar National Parks for their cooperation in conducting this research. Experimental procedures adhered to the guidelines of Madagascar National Parks. This research was financially supported by Grants-in-Aid for the International Scientific Research (B) (nos. 17405008 and 24405008) from the Ministry of Education, Culture, Sports, Science, and Technology, Japan, and in part by the Global COE Program (A06) of Kyoto University. The field study was conducted with permission from the Ministry of the Environment and Forests, Madagascar, through Madagascar National Parks. LITERATURE CITED CAMARGO, C. R., VISCONTI, M. A., AND CASTRUCCI, A. M. 1999. Physiological color change in the bullfrog, Rana catesbeiana. Journal of Experimental Zoology 283: 160 169. GLAW, F. AND VENCES, M. 2007. A Field Guide to the Amphibians and Reptiles of Madagascar. Third Edition. M. Vences and F. Glaw Verlag GbR, Köln. HANLON, R. T. 2007. Cephalopod dynamic camouflage. Current Biology 17: 400 404. HASEGAWA, M., MORI, A., NAKAMURA, M., MIZUTA, T., ASAI, S., IKEUCHI, I., RAKOTOMANANA, H., OKAMIYA, T., AND YAMAGISHI, S. 2009. Consequence of interclass competition and predation on the adaptive radiation of lizards and birds in the dry forest of western Madagascar. Ornithological Science 8: 55 66. HEMMI, J. M., MARSHALL, J., PIX, W., VOROBYEV, M., AND ZEIL, J. 2006. The variable colours of the fiddler crab Uca vomeris and their relation to background and predation. Jounral of Experimental Biology 209: 4140 4153. IKEUCHI, I., MORI, A., AND HASEGAWA, M. 2005. Natural history of Phelsuma madagascariensis kochi from a dry forest in Madagascar. Amphibia-Reptilia 26: 475 483. ITO, R. AND MORI, A. 2010. Vigilance against predators induced by eavesdropping on heterospecific alarm calls in a non-vocal lizard Oplurus cuvieri cuvieri (Reptilia: Iguania). Proceedings of the Royal Society. B, Biological Sciences. 277: 1275 1280. JONES, Z. M. AND JAYNE, B. C. 2012. The effects of sound on the escape locomotor performance of anole lizards. Journal of Herpetology 46: 51 55. MÄTHGER, L. M., LAND, M. F., SIEBECK, U. E., AND MARSHALL, N. J. 2003. Rapid colour changes in multilayer reflecting stripes in the paradise whiptail, Pentapodus paradiseus. Journal of Experimental Biology 206: 3607 3613. MIZUTA, T. 2002a. Breeding biology of the Madagascar Paradise Flycatcher, Terpsiphone mutata, with special reference to plumage variation in males. Ostrich 73: 67 69. MIZUTA, T. 2002b. Predation by Eulemur fulvus

ITO ET AL. GECKO COLOR CHANGE AGAINST BIRD CALL 33 fulvus on a nestling of Terpsiphone mutata (Aves: Monarchidae) in dry forest in northwestern Madagascar. Folia Primatology 73: 217 219. MIZUTA, T. 2005. The Ampijoroa forest, home to the Rufous Vangas. p. 15 40. In: S. Yamagishi (eds.), Social Organization of the Rufous Vanga. Kyoto University Press, Kyoto. MORI, A., IKEUCHI, I., AND HASEGAWA, M. 2006. Herpetofauna of Ampijoroa, Ankarafantsika strict nature reserve, a dry forest in northwestern Madagascar. Herpetological Natural History 10: 31 60. MORI, A. AND RANDRIAMAHAZO, H. J. A. R. 2002. Leioheterodon madagascariensis (Madagascar Menarana Snake). Diet. Herpetological Review 33. 57. NAKAMURA, M. 2004. Predation by Eulemur fulvus fulvus on eggs of Ploceus sakalava sakalava (Aves: Ploceidae) in Ankarafantsika, Madagascar. Folia Primatology 75: 376 378. NAKAMURA, M., OKAMIYA, T., AND YAMAGISHI, S. 2004. The diet of adult and nestling Sickle-billed Vangas Fauculea palliata, a species endemic to Madagascar. Journal of the Yamashina Institute for Ornithology 35: 155 158. RENE DE ROLAND, L. A., AND THORSTROM, R. 2003. Accipiter spp., Goshawk and Sparrowhawks. p. 1091 1094. In: S. M. Goodman, and J. P. Benstead (eds.), Natural History of Madagascar. University of Chicago Press, Chicago. SHERBROOKE, W. C., CASTRUCCI, A. M. D. L., AND HADLEY, M. E. 1994. Temperature effects on in vitro skin darkening in the mountain spiny lizard, Sceloporus jarrovii: a thermoregulatory adaptaion? Physiological Zoology 67: 659 672. SIDDIQI, A., CRONIN, T. W., LOEW, E. R., VOROBYEV, M., AND SUMMERS, K. 2004. Interspecific and intraspecific views of color signals in the strawberry poison frog Dendrobates pumilio. Jounral of Experimental Biology 207: 2471 2485. SILLMAN, A. J., GOVARDOVSKII, V. I., ROHLICH, P., SOUTHARD, J. A., AND LOEW, E. R. 1997. The photoreceptors and visual pigments of the garter snake (Thamnophis sirtalis): a microspectrophotometric, scanning electron microscopic and immunocytochemical study. Journal of Comparative Physiology A 181: 89 101. STUART-FOX, D. M. AND MOUSSALLI, A. 2008. Selection for social signaling drives the evolution of chameleon colour change. PLoS Biology 6: 22 29. STUART-FOX, D. M., MOUSSALLI, A., AND WHITING, M. J. 2008. Predator-specific camouflage in chameleons. Biology Letters 4: 326 329. STUART-FOX, D. M., WHITING, M. J., AND MOUSSALLI, A. 2006. Camouflage and colour change: antipredator responses to bird and snake predators across multiple populations in a dwarf chameleon. Biological Journal of the Linnean Society 88: 437 446. TALLING, J. C., WARAN, N. K., WATHES, C. M., AND LINES, J. A. 1996. Behavioural and physiological responses of pigs to sound. Applied Animal Behaviour Science 48: 187 202. VITOUSEK, M. N., ADELMAN, J. S., GREGORY, N. C., AND STCLAIR, J. J. H. 2007. Heterospecific alarm call recognition in a non-vocal reptile. Biology Letters 3: 632 634. VROONEN, J., VERVUST, B., FULGIONE, D., MASELLI, V., AND VAN DAMME, R. 2012. Physiological colour change in the Moorish gecko, Tarentola mauritanica (Squamata: Gekkonidae): effects of background, light, and temperature. Biological Journal of the Linnean Society 107: 182 191. WEVER, E. G. 1978. The Reptile Ear: Its Structure and Function. Princeton University Press, Princeton. ZAIDAN, F. AND WIEBUSCH, P. L. 2007. Effects of temperature and illumination on background matching in Mediterranean geckos (Hemidactylus turcicus). Texas Journal of Science 59: 127 136. Accepted: 10 December 2012