Infrared Reflectance in Leaf-Sitting Neotropical Frogs Abstract. Two members of the glass-frog family Centrolenidae (Centrolenella fleischmanni, C. prosoblepon) and the hylid subfamily Phyllomedusinae (Agalychnis moreletii, Pachymedusa dacnicolor) reflect near-infrared light (700 to 900 nanome- ters) when examined by infrared color photography. Infrared reflectance may confer adaptive advantage to these arboreal frogs both in thermoregulation and infrared cryptic coloration. Many arboreal members of the glass- frog family Centrolenidae and tree-frog family Hylidae are green, and thus cryptically colored when viewed in vis- ible light (400 to 700 nm). Infrared color photography (/) reveals that two cen- trolenids (Centrolenella fleischmanni. C. prosoblepon) and two phyllomedusine hylids (Agalychnis moreletii, Pachyme- dusa dacnicolor) also reflect light in the near-infrared region (700 to 900 nmj. This is, to our knowledge, the first report of infrared reflectance in neotropical frogs. Since photosynthetic leaf surfaces also reflect infrared, these animals are virtually indistinguishable from the leaves on which they sit. both in visible and near-infrared light ranges. All other North American frogs so examined [Bufo debilis, B. boreas (2), B. coniferus; Rana pipiens (2). R. palmipes, R. cates- beiana; Hyla cinerea. H. squirella, H. euphorbiacea, H. chuneque, and H. cy- anomma] absorb infrared light and stand out sharply against foliage (Fig. 1). Cott (i), using black and white in- frared film, found that the Australian tree-frog Hyla coerulea (=Litoria cae- rulea) reflects infrared light. Litoria cae- rulea, A. moreletii. and A. (=Pachyme- dusa) dacnicolor all contain a newly dis- covered red pigment in unusual melanosomes (4). Bothfleischmanni and prosoblepon groups of Centrolenella contain a purple pigment in their chro- matophores (5). Whether these two skin pigments are identical, or play any role in infrared reflectance, has not been de- termined. There are two likely functions for in- frared reflectance in leaf-sitting frogs, (i) Although the near-infrared is not heat (6), photons of these wavelengths will lose energy as heat if they are absorbed by the skin. Thus, the ability to reflect infrared may play a physiological role in thermoregulation by preventing exces- sive heat gain, (ii) Infrared reflectance may conceal frogs from predators with infrared receptors (3). Little research has been done on near-infrared sensitivity, and supportive evidence is sparse. Both the eyes of birds and the pit organs of snakes may act as near-infrared light re- ceptors. In pigeons and chickens, the sensitivity maxima of the eyes are shifted toward longer wavelengths than those of humans (7), and the tawny owl responds to infrared light (900 nm) (8). Visual sensitivity extending just into the near-infrared would allow birds to see most green frogs on green leaves, al- though centrolenids and phyllomedu- sines would remain camouflaged. Boid and crotaline pit organs are usually inter- preted as thermal detectors, adaptations for nocturnal predation on warm-blood- ed prey (9). In diurnal snakes, however, these receptors may be used to detect frogs that act as infrared sinks among leaves that are reflecting light of these wavelengths. The facial pits of crotaline snakes are directionally sensitive and may allow infrared depth perception (10). Many species of birds and snakes are known to eat frogs and forage in their Fig. 1. A comparison of the color characteristics of a hylid and a centrolenid frog in a conven- tional (top) and an infrared (bottom) color photograph. Although both frogs match the green leaf in light ranges visible to man, only Centrolenella fleischmanni (top frog) reflects near-infrared light. This allows it to blend with foliage both in the visible and near-infrared ranges of light, unlike Hyla cinerea (bottom frog), which absorbs infrared and is distinguished from the leaf surface in an infrared photograph. diurnal retreats. Predation by birds and snakes may have selected for infrared cryptic coloration in tropical leaf-sitting frogs. PATRICIA A. SCHWALM* PRISCILLA H. STARRETT Department of Biological Sciences, Allan Hancock Foundation, University of Southern California, Los Angeles 90007 ROY W. MCDIARMID Department of Biology, University of South Florida, Tampa 33620 References and Notes 1. Kodak Infrared Ektachrome film has a sensitivi- ty extending to about 900 nm [Kodak Publ. N-17 (1974)]. The three emulsion layers are sensitized to green, red, and infrared light rather than to blue, green, and red. By placing a yellow filter over the camera lens, blue light is excluded, re- sulting in a shift of colors in the photograph. Green objects appear blue, red objects appear green, and infrared appears red. 2. H. L. Gibson, W. R. Buckley, K. E. Whitmore, J. Biol. Photogr. Assoc. 33, 1 (1965). 3. H. B. Cott, Adaptive Coloration in Animals (Methuen, London, 1940), pp. 9-11. 4. J. T. Bagnara and W. Ferris, Copeia, 3, 592 (1975). 5. P. H. Starrett and J. M. Savage, Bull. South. Calif. Acad. Sci. 72, 57 (1973). 6. Radiation in these wavelengths (700 to 900 nm) can be emitted by very hot objects, such as the sun or lamp filaments, but it is not heat per se. Similarly, the near-infrared can be reflected or emitted as luminescence by objects that are not hot themselves [Kodak Publ. M-28 (1972)]. 7. R. Granit, Sensory Mechanisms of the Retina (Oxford Univ. Press, London, 1947), pp. 294- 296. 8. F. L. Vanderplank, Proc. Zool. Soc. London 1934, 505 (1934). 9. P. H. Hartline, in Handbook of Sensory Physi- ology , vol. 3, part 3, Eleclroreceptors and Other Specialized Receptors in Lower Vertebrates, A. Fessard, Ed. (Springer-Verlag, New York, 1974), pp. 297-312. 10. R. C. Goris and S. Terashima, J. Exp. Biol. 58, 59 (1973). 11. We thank N. L. Nicholson and P. M. Shugar- man for technical help and suggestions; J. A. McNulty, R. J. Wassersug, and S. Arnold for critical discussions. Permits to collect frogs in Oaxaca. Mexico, were provided to R.W.M. by M. L. Cossio Gabucio and Dr. A. Landazuri Or- tiz, Department of Conservation, Mexico. Sup- ported by biomedical sciences support grant RR07012-09 and biomedical research support grant 5 S07 RR07012-10 from the Division of Re- search Resources, Bureau of Health Professions Education and Manpower Training, National In- stitutes of Health, to P.A.S. and P.H.S. * Present address: Committee on Evolutionary Biology, University of Chicago, Chicago, 111. 60637. 18 November 1976; revised 17 January 1977 REPRINTED FROM ECENCE196:1225-1227 10 JUNE \9rH