Mid-flight prey switching in the fringed-lipped bat (Trachops cirrhosus).

While foraging, eavesdropping predators home in on the signals of their prey. Many prey signal from aggregations, however, and predators already en route to attack one individual often encounter the signals of other prey. Few studies have examined whether eavesdropping predators update their foraging decisions by switching to target these more recently signaling prey. Switching could result in reduced localization errors and more current estimates of prey location. Conversely, assessing new cues while already in pursuit of another target might confuse or distract a predator. We tested whether fringed-lipped bats (Trachops cirrhosus) switch prey targets when presented with new cues mid-approach and examined how switching and the distance between simulated prey influence attack accuracy, latency, and prey capture success. During nearly 80% of attack flights, bats switched between túngara frog (Engystomops pustulosus) calls spaced 1 m apart, and switching resulted in lower localization errors. The switching rate was reduced, and the localization advantage disappeared for calls separated by 3 m. Regardless of whether bats switched targets, attacks were less accurate, took longer, and were less often successful when calls were spaced at larger distances, indicating a distraction effect. These results reveal that fringed-lipped bats attend to cues from non-targeted prey during attack flights and that the distance between prey alters the effectiveness of attacks, regardless of whether a bat switches targets. Understanding how eavesdropping predators integrate new signals from neighboring prey into their foraging decisions will lead to a fuller picture of the ways unintended receivers shape the evolution of signaling behavior.

WAIS test-retest reliability for a clinical out-patient sample.

WAIS test-retest reliabilities were calculated for a clinical out-patient sample with testing intervals varying from 1 to 10 yr. There was no relationship between test-retest interval and the stability of test scores and the correlations between IQs were quite satisfactory (Full Scale IQ = .897, Verbal IQ = .906, Performance IQ = .876). Reliabilities remained high even when the sample was divided by diagnosis into organic, neurotic, personality disorder, and schizophrenic subgroups.