Marco Gamba, Chiara De Gregorio, Daria Valente, Longondraza Miaretsoa, Teresa Raimondi, Olivier Friard, Valeria Torti and Cristina Giacoma

A typical feature of speech and music is rhythm. For instance, in music, the rhythm is produced
by making particular portions in a sequence standing out from others. These portions can be
longer, louder or higher in pitch. Primate songs are elaborate displays where members of a
social group can engage in duets and choruses. We investigated Madagascar’s indris (Indri
indri) songs, aiming to understand the extent to which rhythmic features may vary across
individuals and the sexes. Similarly to what we find in music, indris can emit units with variable
duration and pitch. We found that phrases with different duration showed different inter-onset
intervals. Namely, phrases with two or three units showed longer inter-onset intervals than
those with two units, determining a different rhythm in those type of phrases. We found a
significant effect of sex, where males showed longer inter-onset intervals than females.
Moreover, the higher the number of phrases an indri emitted during a song, the shorter were
the inter-onset intervals within a phrase. The duration of inter-onset intervals within a phrase
positively affected the duration of inter-onset intervals between phrases. We have also found
that the coefficient of variation of both inter-onset intervals was significantly higher in females
than in males. The frequency of the units emitted during the indris’ song also changed
dramatically. We found that the median pitch of males was higher than females. Our findings
highlighted the complexity of the indris’ song. The rhythmic structure differed within and
between phrases of two, three or four units. The interval between onsets decreased significantly
during a phrase and differed between phrase containing a different number of units. Differences
in the phrases’ rhythmic structure suggest that indris may regulate timing, unit duration, and
interval duration. The indris share this ability with other nonhuman primates. In agreement
with findings on the chimpanzees, the indris appear to adjust the timing of their emissions
during the song. The ability to adjust utterances of the song has emerged when investigating
contextual variation in the song’s structure and may play an essential role in maintaining social
interaction within and between groups. We demonstrated a remarkable difference between the
sexes, with females showing shorter inter-onset intervals than males, although the lack of
dimorphism would predict little differences in the vocal apparatus’s size. Previous works
suggested that there might be a link between sexual dimorphism in call frequency and social
system. Indeed, polygynous primates showed more pronounced sexual dimorphism in
fundamental frequency compared to monogamous species. These studies suggested that size
differences in the vocal air sacs may explain these differences, a fascinating hypothesis to test
the indris, which we know have dorsal vocal sacs in both males and females. The pitch of units
emitted in phrases showed a tendency towards overall sex differences, where males showed
higher fundamental frequencies than females. Our knowledge of the ‘singing primates’ is still
limited, and although they show elaborate vocal display, it is still unclear whether learning
processes may play a role in their vocal production.