14

Birdsong is degraded as it propagates through the habitat. This affects its use as a signal in communication networks, depending on song type as well as sender and receiver location. Nest holes constitute special receiver locations to females of many hole-nesting species. Although such locations are frequently used by females during their fertile period, for instance at dawn when information gathering from singing males may be essential, the conditions for receiving sounds inside nest holes are currently unknown. We investigated these conditions with a sound transmission experiment, in which great tit songs were broadcast in a deciduous forest and rerecorded from both outside and inside a nest box. Several aspects of sound degradation encompassing signal attenuation and distortion were quantified.

Attenuation was strongest inside the nest box for both song and background noise. However, the signal-to-noise ratio, which affects song detection and discrimination, was only slightly reduced inside the nest box. Signal distortion, which has implications for both information transfer and ranging, provided conflicting results for the two types of song notes, the highly modulated buzz notes and the less modulated pure notes. These results provide the first evidence that entering nest boxes complicates song reception conditions for female songbirds. We suggest that this may ultimately affect vocally mediated information gathering in the network.

When separated from their mother and other group members, guinea pig Cavia porcellus pups emit distinctive high pitched whistles. To determine if these vocalisations are individually distinctive, we recorded the whistles of isolated guinea pig pups, 8 to 10 days old, and subjected their acoustical parameters to discriminant analysis. The results of the reclassification accuracy were higher than random assignment, indicating the existence of individual differences. Individual pup vocalisations did not differ from one another by any single acoustic parameter, but by a set of parameters. Individual recognition of such isolation calls by mothers could play an important role in facilitating reestablishment of contact.

Song structure and singing behaviour in two nesting populations of Willow Warbler Phylloscopus trochilus acredula of different density were studied in Moscow and Kostroma Regions, Russia, from 1994-1997. In total 520 songs of 9 males from the low-density population and 566 songs of 12 males from the high density population were analysed. Singing periods in the high density Willow Warbler population (11 males, 758 min.) were longer (58% of the time) than in the low density population (13 males, 866 min.) (58%). In the high density population, 51% of singing time consisted of “singing duels” (singing in gaps between the songs of a neighbouring male, and singing during the singing of a neighbouring male), while in the low density population only 7% did. The initial phrases of Willow Warbler songs were significantly longer in the high density population, with up to 9-13 elements, while there were no occurrences of this length in the low density population. Songs as a whole were longer and more variable in the high density population. Thus changes in structure, singing activity and behaviour are linked to male-male interaction and density of the nesting community.

There is a need to improve the assessment of causes and consequences of vocal behaviour. The lack of descriptions of the context of call or song production comes from the complexity of its definition. The context is composed of numerous physical and social parameters and therefore its analysis should be multi-dimensional. Classical approaches involve a relatively subjective data reduction. This is due to a pre-selection of the parameters which may be potentially involved, as the analysis becomes rapidly complicated when the number of factors increases. This paper describes a helpful method that allows a much wider range of potential parameters to be explored and displayed visually. The parameters selected for subsequent analysis are those indicated by the display as most relevant and are therefore not arbitrary.

We applied this method to observations of events (external or internal to the group) or behaviours (including postures) preceding and following call production. Calls of Campbell’s monkeys Cercopithecus campbelli present call types composed of several stereotyped sub-types, raising the question of whether or not such variations have a functional meaning for animals. We present three examples of the application of this method to describe in detail the context of production of a given call type, to detect specific temporal sequences of production and to discriminate between structurally quite similar sub-types.