Estudo do comportamento de rastreamento visual na coruja buraqueira

Abstract

Under naturalistic viewing conditions, acquisition of visual information is constantly disrupted by transient events such as eye blinks and saccadic gaze shifts. Their inclusion into any realistic account of visual processing is of paramount importance. In the present study, we used an ethological approach to examine the temporal dynamics of eye blinks, head movements and gaze stabilization in the diurnally active owl Athene cunicularia. Owls have negligible eye movements such that head positioning can be taken as a reliable indicator of eye fixation. Our results are based on 17 hours of video recorded camcorder with frame rate of 25 Hz and 960x600 pixels, between 9:00 a.m. and 6:00 p.m., from a total of nine adult owls housed in outdoor aviaries. Digital movies were analyzed on a frame-by-frame basis. Four behavioral categories were considered for analysis: (i) “gaze” (instance where the eyes and head did not move for at least 40 ms i.e. two video frames); (ii) “blink” (complete transient closure of the eyelids); (iii) “gaze-shift” (head motion); and “out of view” (when neither eye was visible to the observer). Seven animals were observed during three circadian periods: morning (Mn), afternoon (An) and afternoon before eating (Ae). In addition, two animals were observed while “resting” (R) and when performing a two-alternative forced-choice discrimination task, a behavioral situation referred as “task” (T). For each behavioral category and state, we merged the data obtained from all animals. Behaviors could unambiguously be coded for about 77% of the total recording time. Gaze was by in large the most frequent behavior with significant difference between groups Mg x An and R x T (t test, p >0.05). The distribution of gaze durations was broad, ranging from 80 to 25,600ms, and negatively skewed with median values of 1840 ms (mode: 480ms) for resting owls (RO) and 780 ms (mode: 240ms) for discriminating owls (DO). Blink episodes lasted on average 235 ± 45 ms (arithmetic mean ± SD; RO: 290 ± 64 ms; DO: 180 ±25 ms). Mean saccadic gaze-shift duration was 161± 144 ms (RO: 260 ±160 ms; DO: 150±130 ms). RO blinked significantly more often than DO (median: 7.45/min versus 3.05/min, Wilcoxon rank sum, P < 0.001). Independently of the behavioral state of the animal, we also found that around 85% of blinks were concomitant with head movements. However, the fact that most head movements exhibited by DO were not accompanied by blinks as they were seen to be in RO suggests that blinks were not simply triggered by reflexive mechanisms involving corneal stimulation. Our results provide evidence for task-dependent interaction between blink and head movement control mechanisms, a phenomenon that has also been reported in primates.