Infrared camouflage in leaf-sitting frogs: a cautionary tale on adaptive convergence

Many cryptic green animals match leaves in invisible near-infrared (NIR) wavelengths. This observation is an enduring puzzle because animals do not see NIR light, so NIR background matching is unlikely to contribute to visual camouflage. Two alternative explanations have been proposed-infrared camouflage (i.e. matching the temperature of the background) and thermoregulation-but neither hypothesis has been experimentally tested. To test these hypotheses, we developed bilayer coatings that mimicked the reflectivity of green leaf-sitting frogs with high NIR (HNIR) or low NIR (LNIR) reflectance. Under a solar simulator in the laboratory, agar model frogs with LNIR reflectance heated up more quickly and reached higher temperatures than those with HNIR reflectance. However, when placed in a tropical rainforest (natural habitat of leaf-sitting frogs), HNIR and LNIR models did not significantly differ in the similarity of surface temperature to the adjacent leaves or in core temperature, thus failing to support the infrared camouflage and thermoregulation hypotheses, respectively. The lack of difference between treatments is probably due to the limited exposure of frogs to direct solar radiation in their natural habitats. We propose an explanation for NIR background matching based on specific mechanisms underlying green coloration and translucence in frogs and caution against assuming adaptive convergence.