Potent and affordable video and computer systems for automatic data acquisi
tion are becoming increasingly important in behavioural neuroscience. It ha
s remained challenging, however, to acquire data from small and fast-moving
animals, such as insects in flight, due to the limited spatial and tempora
l resolution of the systems currently available. Our research on free-flyin
g insects motivated the development of new methods in the context of two di
fferent experimental settings. First, the position and precise body axis di
rection of honey bees approaching a food source were automatically measured
. Second, the flight trajectories of a phonotactic parasitoid fly homing in
on its cricket host were recorded in 3D. We used pan-tilt cameras, i.e. ca
meras with moveable optics, to follow the animal's path with a close up ima
ge. Novel methods were developed for image acquisition and position measure
ment using pan-tilt cameras, as well as calibration and data evaluation in
3D world coordinates. The innovations of this system comprise: (1) Acquisit
ion of images in high spatial detail over large observation areas. (2) Imag
e acquisition at a field rate of 50 Hz PAL. (3) Free positioning of the cam
eras for 3D acquisition. (4) Computation of the flight path in 3D world coo
rdinates. We illustrate the capabilities of the system with data obtained f
rom a calibration object as well as from the behaviour of unrestricted, fre
e-flying flies and bees. Potential applications in behavioural neuroscience
and the psychophysics of sensory perception are briefly discussed. (C) 200
0 Elsevier Science B.V. All rights reserved.