Although the ability to use the Earth's magnetic field for long distance orientation and navigation has been demonstrated in many animals, the search for the appropriate receptor has not yet finished. It is also not entirely clear whether the use of magnetic field information is restricted to specialists like migrating birds, or whether it is a sense that is also suited to short distance orientation by avian species. We successfully trained nonmigratory zebra finches in a four-choice food-search task to use the natural magnetic field as well as an experimentally shifted field for short distance orientation, supporting the view that magnetic field perception may be a sense existing in all bird species. By using a conditioning technique in a standard laboratory animal, our experiments will provide an ideal basis for the search for the physiological mechanisms of magnetic field perception.

Background: Zebra finches can be trained to use the geomagnetic field as a directional cue for short distance orientation. The physical mechanisms underlying the primary processes of magnetoreception are, however, largely unknown. Two hypotheses of how birds perceive magnetic information are mainly discussed, one dealing with modulation of radical pair processes in retinal structures, the other assuming that iron deposits in the upper beak of the birds are involved. Oscillating magnetic fields in the MHz range disturb radical pair mechanisms but do not affect magnetic particles. Thus, application of such oscillating fields in behavioral experiments can be used as a diagnostic tool to decide between the two alternatives.

SUMMARYMost birds have laterally placed eyes with two largely separated visual fields. According to studies in pigeons laterally eyed birds move their eyes independently in most situations, eye coordination just occurred during converging saccades towards frontal stimuli. Here we demonstrate for the first time that laterally eyed zebra finches show coordinated eye movements, regarding direction and amplitude. Spontaneous and visually elicited movements of the two eyes were recorded simultaneously, using a newly developed eye tracking system. We found that, if one eye moves in a certain direction, the other eye simultaneously performs a counter-movement in the opposite direction. Based on these data we developed a hypothesis of how laterally eyed birds cope with the situation in which the left and right eye simultaneously obtain images with different content. We suggest that the counter-movements maintain the spatial relationship of the two visual fields. 'Oculospatial constancy', as we call it, facilitates the combination of the left and right visual percept on the level of peripheral or unattended viewing, and the localization of appearing stimuli within the whole visual field. As soon as two visual stimuli simultaneously appear in the left and right visual field, the birds decide on one stimulus and direct the fovea of the appropriate eye towards it for high resolution analysis, the other eye simultaneously performing a counter-saccade. This leads to the assumption that, in contrast to simultaneous peripheral perception with two eyes, the processing of foveal information is possible only for one eye at one time.

BackgroundThe primary visual cortex of mammals is characterised by a retinotopic representation of the visual field. It has therefore been speculated that the visual wulst, the avian homologue of the visual cortex, also contains such a retinotopic map. We examined this for the first time by optical imaging of intrinsic signals in zebra finches, a small songbird with laterally placed eyes. In addition to the visual wulst, we visualised the retinotopic map of the optic tectum which is homologue to the superior colliculus in mammals.Methodology/Principal FindingsFor the optic tectum, our results confirmed previous accounts of topography based on anatomical studies and conventional electrophysiology. Within the visual wulst, the retinotopy revealed by our experiments has not been illustrated convincingly before. The frontal part of the visual field (0°±30° azimuth) was not represented in the retinotopic map. The visual field from 30°–60° azimuth showed stronger magnification compared with more lateral regions. Only stimuli within elevations between about 20° and 40° above the horizon elicited neuronal activation. Activation from other elevations was masked by activation of the preferred region. Most interestingly, we observed more than one retinotopic representation of visual space within the visual wulst, which indicates that the avian wulst, like the visual cortex in mammals, may show some compartmentation parallel to the surface in addition to its layered structure.Conclusion/SignificanceOur results show the applicability of the optical imaging method also for small songbirds. We obtained a more detailed picture of retinotopic maps in birds, especially on the functional neuronal organisation of the visual wulst. Our findings support the notion of homology of visual wulst and visual cortex by showing that there is a functional correspondence between the two areas but also raise questions based on considerable differences between avian and mammalian retinotopic representations.

The eyes of zebra finches are placed laterally, the foveae are looking into different directions. It is unlikely that the birds are able to process different images from both eyes simultaneously. A neural mechanism might therefore be necessary to guide the birds' attention to one of the two eyes and to reduce the processing of information of the other. Previous studies revealed that information from the ipsilateral eye is indeed suppressed on its way to the telencephalon by the activity of the contralateral eye. It has been suggested that two nuclei of the tecto-thalamic tract, nucleus subpraetectalis and nucleus interstitio praetecto subpraetectalis, are a central part of such a suppressive mechanism. Using electrophysiological recordings, we investigated the influence of these two nuclei and nucleus rotundus on the processing of binocular visual information by treating the nuclei with picrotoxin or electrolytic lesions. Deactivation of inhibitory neurons within SP/IPS leads to a significant increase of the ectostriatal responses to ipsilateral and bilateral stimulation, the responses to contralateral stimulation remain unaffected. Lesioning SP/IPS does not alter the responses to visual stimuli. Treatment of nucleus rotundus with picrotoxin increases contralaterally and bilaterally, but not ipsilaterally evoked responses. A wiring diagram is presented which interprets these findings.