78 
THE WILSON JOURNAL OF ORNITHOLOGY • Vol. 123, No. 1, March 2011 
III! till W -llll Illl II 
FIG. 2. A. Oscillogram (above) and spectrogram (below) of a sequence of the characteristic song of the Cuban Tody. B. 
Detailed spectrogram (left) and power spectra (right) of the song marked in A with an asterisk. 
Statistical Analysis .—Statistical analysis was 
performed using Statistica Version 7.0 (StatSoft 
Inc. 2004). The data sets were normally distrib¬ 
uted (Kolmogorov-Smimov test, P > 0.05) and 
we used parametric statistics (one-way ANOVA). 
Statistical differences between several mean 
values were analyzed with a Newman-Keuls 
post-hoc test (identical letters indicate no statis¬ 
tically significant difference between means). All 
analyses were conducted using mean values per 
individual and the level of significance was a = 
0.05. 
A multivariate discriminant function analysis 
(DFA) was performed using SPSS Version 16.0 
(SPSS, Chicago, IL, USA) with individuals as 
cases and number of notes, note duration, interval 
between notes, initial frequency, peak frequency, 
final frequency, and interval between songs as 
parameters. We used DFA to generate linear 
combinations of variables for assigning cases to 
their pre-determined groups (Quinn and Keough 
2002 ). 
RESULTS 
Sounds Emitted. —The characteristic song ( tot- 
tot-tot ) of the Cuban Tody was present in >98% 
of the recordings in each province studied. It was 
emitted by single birds in most cases that 
alternated singing with feeding behavior. The 
song was repeated at time intervals of 0.04- 
15.62 sec and included 3—23 notes. Each note was 
downward frequency-modulated and often had a 
