Coevolution between bats and katydids in the neotropics

Research project P 14257 Coevolution between bats and katydids in the neotropics H. RÖMER 06.03.2000 This study combines field and laboratory investigations of acoustic behavior, neurophysiology and genetics in two groups of neotropical katydids (Orthoptera) with contrasting predation pressure by different guilds of bats. The overall aim is to relate the patterns of veriability of several traits between species, and between Pseudohylline and Phaneropterine katydids, to the underlying selective processes which generate these. The focus of the investigation is on the comparison of several traits in species of two subfamilies of neotropical katydids, where we test the hypothesis that the Pseudiphylline species, due to predation pressure by gleaning and passively listening bats have evolved a reduced sensitivity to airborne sound and increased substrate vibration sensitivity. This may be advantageous for detecting conspecific signals in a noisy environment. By contrast, Phaneropterine species are subject to predation by aerial hawking insectivorous bats, and have evolved sensory and central nervous traits that enable them to escape from a bat attack in flight. This includes increased ultrasonic sensivity of certain interneurones. We also focus on the consequences of these adaptations for intraspecific acoustic communication, and how environmental gradients in predation pressure and other biotic and abiotic factors influences the evolution of traits. For a representative species of Pseudophylline katydid, we will examine the genetic consequences of fragmentation into subpopulations due to predation by bats, which forces the species to live in isolated patches of a bromelid. The sensory consequences, as well as those for gene flow between subpopulations, will investigate.

We investigated the predator-prey relationship between insect-eating bats and their prey, long-horned grasshoppers, on the tropical Barro Colorado island in the Panama canal. Some bat species hunt their flying prey in open areas around the canopy region of the forest, using ultrasound echolocation calls, whereas another group of bat species (gleaning bats) searches for prey in the forest understorey, but they do use the sound signals of their prey to home in and catch katydids. Twenty species of the subfamily of Pseudophyllinae live in the forest understorey and are subject to predation by gleaners; 46 species of the subfamily of Phaneropterinae live in the canopy and are hunted by aerial insectivorous bats. Members of the different katydid species exhibit a number of traits which enable them to increase their survival chance in the presence of bats. Phaneropterinae, which are good flyers, have ears which are very sensitive, and nerve cells within the CNS which respond strongly to bat ultrasound. This enables them to detect echolocation calls of their predator at large distances and to initiate evasive flight responses. By contrast, the different selection pressure on species of Pseudophyllinae has resulted in another set of adaptations: their ears are less sensitive, and the CNS does not respond very well to bat ultrasound, but they are two orders of magnitude more sensitive to the low-frequency vibration of the substrate (leaves of plants), on which they live. This is adaptive, because these understorey katydids have evolved a suite of behavioural characters, which increases their chance of survival to passively listening bats: they hide themselves in spiny bromeliads during the day and actively choose individual plants in a field of more than hundred, which serve the protective function best. Males of these species also reduce the "on-time" of their airborne sound signals, and in addition often drum with several body parts on the substrate, which can be detected by females on the same plant, with no risk involved that bats could detect this vibratory signal. We also found out, that these alternative air-borne sound and vibration signals are used in a proportion depending on the risk of predation: Under full moon conditions, when the light is bright enough that visually hunting predators can be successful, they significantly reduce their signaling or use more vibratory signals compared to new moon nights. The close relationship between bats as predators and katydids is also observed in the fact, that when katydids are less abundant during full moon nights, some bat species stay a significant portion of the night in their roosts and show little flight activity compared to new moon conditions. We further investigated the trade-off between signaling for mates with reduced "on-time" and detection of such signals by the nervous system, because many species and individuals of crickets, grasshoppers and frogs produce a high level (70 dezibel) of background noise in the nighttime rainforest. Similar to what humans experience at a cocktail-party when they selectively listen to particular sound sources in the background noise, katydids appear to be able to detect the calls of their mates at a reasonably high rate.