In contrast to Ormia, the European species Therobia leonidei (Mesnil 1964) (Diptera: Tachinidae) is known to infect a variety of bushcricket species (Leonide 1969). All of the known bushcricket host species produce short chirps less than 1 s in length, repeated after pauses remarkably longer than the chirps themselves. Therefore they might be harder to detect than singing crickets. The broad tuning of the ear (Stumpner and Lakes- Harlan 1996) reflects the broad frequency spectrum of the songs of its hosts (Leonide 1969), all of which have their peak song frequencies in the range of 12±28 kHz (Heller 1988). Among the known hosts of are several species of the genus Poecilimon Fischer, 1854 (Tettigonioidea, Phaneropteridae) (Lakes-Harlan and Heller 1992; Lehmann 1998), which differ in singing behavior (Heller 1988). Poecilimon females and experimentally muted males were never infected (Lakes-Harlan and Heller 1992), showing that song production is necessary for parasitism.
Results of the experiments described above showed that the choice of a bushcricket host by the phonotactic tachinid T. leonidei was related to song structure. They parasitized the polysyllabic species P. mariannae to a greater extent than the monosyllabic P. veluchianus. Female flies parasitized more than three times (3.8/3.3) as many P. mariannae males as P. veluchianus males. This preference for the species with polysyllabic songs was independent of whether it was the local resident or the introduced species. Therefore, local adaptations of the fly to its natural host species or to ecological factors are an unlikely explanation for the constancy of this preference in both experiments. All additional factors (SPL, frequency of song, body size) that were measured showed no differences between the two species. Although it is possible that other unknown factors are involved in the preference for P. mariannae as a host, song structure is the most likely explanation for this preference of the fly. [1]