Exemplary sequence of six trials. Units at the axis represent seconds elapsed within each trial. The frequencies of the sample tones (open squares) and test tones (closed squares) either were the same (trial 2) or they were different (trial 1, and trials 3–6). Each trial contained up to six sample stimuli and a final test stimulus presented with one delay chosen at random from the three different delays of a delay set. The delays between sample stimuli were chosen randomly from the same delay set.
Individual performance of five birds as a function of the delay between the last sample and the test stimulus. Top row: Mean percentages of all responses. Hit or miss can be scored if sample and test stimuli differ in frequency; correct rejection or false alarm can be scored if sample and test stimuli frequency are matching. Middle row: Mean performance ( s.d.) for each of the five starling individuals. Individual performances are expressed as percent correct responses, i.e., , where and are the number of correct responses in nonmatching (i.e., hits) and matching trials (i.e., correct rejections), respectively. The number of incorrect responses in nonmatching (i.e., misses) and matching trials (i.e., false alarms) are and , respectively. The dotted line indicates the threshold criterion of 75% correct responses and the average auditory memory persistence time of the individual starling (i.e., the delay at which the performance dropped to 75%) is given in each panel. Bottom row: Mean performance for the five starlings for three of the six different sample stimulus repetitions (1, 3, and 5 sample stimuli).
Mean performance of five starlings [percent correct responses, i.e., ] as a function of the number of repetitions of the sample stimulus (1–6). The number of correct responses in nonmatching (i.e., hits) and matching trials (i.e., correct rejections) are and , respectively. The number of incorrect responses in nonmatching (i.e., misses) and matching trials (i.e., false alarms) are and , respectively. Only data from delay sets for which performance for all five starlings was determined are shown.
Mean performance ( s.d.) of the five starlings for frequency separations smaller than one octave and equal to or larger than one octave as a function of the different delays. The number of correct (i.e., hits) and incorrect responses (i.e., misses) in nonmatching trials are and , respectively. Only delay conditions in which hit rates for all five starlings were acquired are shown. Differences between the two frequency difference categories were significant for all different delays.
Mean performance for different animal species expressed as percent correct responses for each delay. Starlings with long auditory memory persistence time (AMPT) were grouped into the lAMPT group, and starlings that showed a comparatively short AMPT were grouped into the sAMPT group. Since only one sample stimulus was presented before the test stimulus in the compiled studies, only data from trials in which the starlings were presented with one sample stimulus were included.
The delay sets with which subjects were tested, the number of subjects tested , and the maximum number of sample stimuli presented in trials of that delay set . With increasing delay set number fewer sample stimuli were presented before the test stimulus to prevent single trials from becoming too long. When performance of a starling was still above 75% correct for the longest delay of a set, the next delay set was tested. Thus, as the delay was increased, fewer starlings were tested.
Individual auditory memory persistence times in seconds of five starlings (An, Gi, Li, Ma, and Ru) for the two different frequency sets (frequency set 1: 1100, 1386, 1746, 2200, 2772, ; frequency set 2: 1235, 1556, 1960, 2470, 3112, ) presented in the first and second experiments.
Article metrics loading...
Full text loading...