BREVIA can provide an entirely additional benefit to crypsis (difficulty in visually detecting prey) (2) and functions even in the absence of crypsis, when the prey is in clear view. Although the role of predator cognition in the evolution of warning coloration is widely studied (3), researchers studying the evolution of John Skelhorn,1 Hannah M. Rowland,2 Michael P. Speed,2 Graeme D. Ruxton1* other aspects of adaptive coloration have concenrganisms are under strong selection to Selenia dentaria) as prey. Before testing, chicks trated on understanding how predators’ sensory sysavoid detection by predators and to cap- were divided into nine groups, each containing tems have influenced the evolution of prey coloration ture prey, and understanding how animals’ eight individuals. Birds in all groups received four (2). Our results show that predators’ cognitive stratvisual appearances are adapted for these purposes 2-min trials in which they were placed in the ex- egies (recognition and identification), rather than continues to pose interesting questions for evolu- perimental arena individually. The items placed in their sensory capabilities, are the selective force tionary theory (1). Although the function of cryp- the experimental arena with them during these trials driving the evolution of masquerade and raise the sis (avoiding detection) (2), aposematism (warning differed among groups. Three groups encountered possibility that predator cognition may be a more important selective agent than previously realized. Masquerade is more widespread than betterstudied aspects of adaptive coloration such as aposematism and is found in a large number of species from a wide range of taxa. Plants from the genus Lithops look remarkably like stones; stick insects Fig. 1. (A) The time in seconds (mean T SE) taken to make a first peck at the test stimulus (Brim indicates brimstone; Thorn, early resemble twigs; the Amthorn;andTwig,hawthorntwig).Thelatencytoattacktheteststimulidifferedamongourexperimentalgroups(Kruskal-Wallistest;c2 =56.10,P< azon fish Monocirrhus 0.001, df = 8). Specifically, birds trained with branches (Br) took significantly longer to attack the test stimuli than did birds in the control groups polyacanthus is visually trainedonmanipulatedbranches(Man)ornothing(No)(Kruskal-Wallistest;c2 =47.39,P<0.001,df=1).SamplesizeisN=8ineachgroup.(B) almost indistinguishable Thetimeinseconds(meanT SE)thatbirdshandledpreyfor inthetesttrial.Handlingtimes (4)differedamongourexperimentalgroups(Kruskal- from leaves, and birds Wallis test;c2 = 35.34, P < 0.001, df = 5). Birds trained with branches took significantly longer to handle caterpillars than didbirds in the control from the family Nyctibiidae bear an uncanny groups trained on manipulated branches or nothing (Kruskal-Wallis test; c2 = 31.16, P < 0.001, df = 1). Sample size is N = 8 in each group. likeness to tree stumps. coloration) (3), and mimicry (resembling a defended a hawthorn branch complete with leaves (hawthorn Therefore, masquerade appears to have evolved organism) (4) are intensively studied, one aspect of Crataegus spp. being a common host plant for the on multiple occasions, and its ecological imporadaptive coloration has been almost completely two caterpillar species). Three groups encountered tance requires further investigation in light of our ignored: masquerade. Masquerading organisms a manipulated hawthorn branch that had been experiments. appear to closely resemble inedible and generally bound in purple cotton thread to change its visual References and Notes inanimate objects such as twigs, leaves, stones, and appearance without influencing its physical struc1. G. D. Ruxton, T. N. Sherratt, M. P. Speed, Avoiding bird droppings. Individuals using this defensive ture or odor. The final three groups experienced an Attack: The Evolutionary Ecology of Crypsis, Warning strategy are assumed to avoid predation or gain empty arena. The test stimulus differed among Signals and Mimicry (Oxford Univ. Press, New York, groups given the same previous experience: one access to prey by being misidentified as either 2004), pp. 23–25. 2. M. Stevens, S. Merilaita, Philos. Trans. R. Soc. London Ser. B inedible objects by their predators or as innocu- group received a single Brimstone larvae, another 364, 423 (2009). ous objects by their prey. There is currently no a single Early thorn larvae, and the remaining 3. J. Mappes, N. Marples, J. A. Endler, Trends Ecol. Evol. 20, group received a single hawthorn twig (7). empirical evidence to support this theory (5, 6). 598 (2005). Birds with prior experience of twigs took Demonstrating that organisms benefit from 4. M. P. Speed, Anim. Behav. 46, 1246 (1993). 5. J. A. Endler, Biol. J. Linn. Soc. Lond. 16, 25 (1981). masquerade is methodologically challenging: It is longer to attack both species of twig-resembling 6. J. A. Allen, J. M. Cooper, J. Biol. Educ. 19, 268 (1985). difficult to determine whether a predator has de- caterpillars, and handled them more cautiously, 7. Materials and methods are available as supporting tected and misidentified an individual (that is, mas- compared with birds that had either no experimaterial on Science Online. querade) (5, 6) or whether it has simply failed to ence of twigs or experience only of twigs whose detect the prey item (which would be crypsis) (2). visual appearance had been manipulated by bind- Supporting Online Material www.sciencemag.org/cgi/content/full/327/5961/51/DC1 For predators to misidentify a masquerading prey ing them in colored thread (Fig. 1). Our results show that masquerade functions Materials and Methods item as the object that it closely resembles, the Figs. S1 to S3 predator must have previous experience of that to promote misidentification of the masqueradobject. By manipulating predators’ previous expe- ing organism. The caterpillars in our experiment 14 September 2009; accepted 27 October 2009 rience of the putative model but keeping their ex- were treated differently by predators with previ- 10.1126/science.1181931 posure to the masquerader the same, it is possible ous experience of twigs in ways that are likely to 1 Division of Ecology and Evolutionary Biology, Faculty of to determine whether predators are misidentifying lead to antipredatory protection: increased latency Biomedical and Life Sciences, University of Glasgow, Glasgow masquerading prey as their models or simply failing to attack and more cautious handling. Most powerG12 8QQ, UK. 2Division of Population and Evolutionary to detect them. We used domestic chicks (Gallus fully, this effect occurred at close range, entirely out Biology, School of Biological Sciences, University of Liverpool, gallus domesticus) as predators and putative twig- of context, on a visually contrasting substrate and Liverpool L69 7ZB, UK. resembling caterpillars (the Brimstone moth, Opis- in an empty arena with no other prey or objects *To whom correspondence should be addressed. E-mail: thograptis luteolata, and the Early thorn moth, present. We therefore can conclude that masquerade [email protected]

Masquerade: Camouflage Without Crypsis

www.sciencemag.org

SCIENCE

VOL 327

1 JANUARY 2010

Downloaded from www.sciencemag.org on March 12, 2010

O

51