Temporal ventriloquism effect in European starlings: Evidence for two parallel processing pathways.

The brain constantly has to interpret stimuli from a range of modalities originating from the same or different objects to create unambiguous percepts. The mechanisms of such multisensory processing have been intensely studied with respect to the time window of integration or the effect of spatial separation. However, the neural mechanisms remain elusive with respect to the role of alerting effects and multisensory integration. We addressed this issue by choosing a test paradigm where we could manipulate potentially alerting stimuli and simultaneously activating stimuli independently: We measured the temporal ventriloquism effect in European starlings by using the temporal order judgment paradigm with subjects judging the temporal order of the lighting of 2 spatially separated lights. If spatially noninformative acoustic stimuli were added to the visual stimuli the performance improved when the 2 visual stimuli were flanked by acoustic cues with a small time-offset compared to synchronous presentation. Two acoustic cues presented with asymmetric offsets showed that this effect was mainly driven by the cue trailing the second visual stimulus, while an acoustic cue leading the first visual stimulus had less effect. In contrast, 1 singleton acoustic cue prior to the first visual stimulus, without a second acoustic cue, enhanced performance. Our results support the hypothesis that the first stimulus pair with the leading sound activates alerting mechanisms and enhances neural processing, while the second stimulus pair with the trailing sound drives multisensory integration by simultaneous activation within the temporal binding window. (PsycINFO Database Record

Azimuthal sound localization in the European starling (Sturnus vulgaris): III. Comparison of sound localization measures.

Sound localization studies have typically employed two types of tasks: absolute tasks that measured the localization of the angular location of a single sound and relative tasks that measured the localization of the angular location of a sound relative to the angular location of another sound from a different source (e.g., in the Minimum Audible Angle task). The present study investigates the localization of single sounds in the European starling (Sturnus vulgaris) with a left/right discrimination paradigm. Localization thresholds of 8-12° determined in starlings using this paradigm were much lower than the minimum audible angle thresholds determined in a previous study with the same individuals. The traditional concept of sound localization classifies the present experiment as an absolute localization task. However, we propose that the experiment presenting single sounds measured localization of the angular location of the sound relative to a non-acoustic spatial frame of reference. We discuss how the properties of the setup can determine if presentation of single sounds in a left/right discrimination paradigm comprises an absolute localization task rather than a localization task relative to a non-acoustic reference. Furthermore, the analysis methods employed may lead to quite different threshold estimates for the same data, especially in case of a response bias in left/right discrimination. We propose using an analysis method precluding effects of response bias on the threshold estimate.

Effect of head turns on the localization accuracy of sounds in the European starling (Sturnus vulgaris).

Long signal durations that represent closed-loop conditions permit responses based on the sensory feedback during the presentation of the stimulus, while short stimulus durations that represent open-loop conditions do not allow for directed head turns during signal presentation. A previous study showed that for broadband noise stimuli, the minimum audible angle (MAA) of the European starling (Sturnus vulgaris) is smaller under closed-loop compared to open-loop conditions (Feinkohl & Klump, 2013). Head turns represent a possible strategy to improve sound localization cues under closed-loop conditions. In this study, we analyze the influence of head turns on the starling MAA for broadband noise and 2 kHz tones under closed-loop and open-loop conditions. The starlings made more head turns under closed-loop conditions compared to open-loop conditions. Under closed-loop conditions, their sensitivity for discriminating sound source positions was best if they turned their head once or more per stimulus presentation. We discuss potential cues generated from head turns under closed-loop conditions.