13

Acoustic researches were made in the Miramare Marine Reserve (Gulf of Trieste, Adriatic Sea, Italy) and just outside it in order to evaluate interferences between fish acoustic communication and man-made under- water noise due to the passage of outboard engine boats. Average sea ambient noise was recorded both inside and outside the protected area because in the reserve the passage of boats is intermittent, while outside the boat traffic is intense. Then observations in tanks and in the field were made on the territorial agonistic behaviour of the red mouth goby Gobius cruentatus, a benthic fish present in and out the protected area. Two different fish sounds were recorded: a train of distinct knocks and a drumming sound, a less regular spaced train. The frequency range is from 50 Hz up to 800 Hz with highest amplitude below 200 Hz. Meanwhile the boat noise was recorded from the smaller distance between engine source and fish inside (200 metres) and outside (10 metres) the preserved area. Finally all of these recordings were compared in a frequency - P.S.D. graph. The result of the comparison suggests that in the Reserve the acoustic communication in G. cruentatus is undisturbed and that this approach is a useful method to assess and prevent the impact of man-made noise in a protected area.

Song learning in birds is a process of memorising a song model and subsequently trying to match its own vocal output with the memory formed. Thereby, each syllable including the silent interval preceding it are copied as a single unit (Williams and Staples 1992). Zebra finch song consists of a chain of syllables repeated several times making up a bout, and is preceded by introductory notes (Sossinka and Boehner 1980). Birds raised with no access to adult song models develop a song (untutored song) that differs from normal song (Williams 1993), but is initiated by a string of introductory notes (Price 1979). Therefore, the question arises whether the introductory note is genetically determined. To answer this question we recorded and quantified a total of 376 introductory notes of song from normal (n=8) and untutored birds (n=6). The duration of the introductory note (syllable length), as well as the preceding silent interval (syllable period length) and the frequency of energetic maximum were analysed. Syllable length differs slightly between normal (51+/-2 msec) and untutored birds (67+/-7; p=0.053). Syllable period length, however, was significantly longer in untutored song (183+/-13 vs. 122+/-14 msec, p=0.02). Therefore, we conclude, that the length of an introductory note is genetically fixed, while introductory note period length is a learned feature of song (supported by the Deutsche Forschungsgemeinschaft SFB 515 to B.E. N.-B.).

The Humboldt Penguin Spheniscus humboldti is a South American penguin that lives on the pacific coast from the 6th to the 33rd south parallel. It nests in colonies where each pair builds a true nest in a hole dug in the ground and covered with stones. The acoustic behaviour of this specie has never been studied in depth, despite the fact that it is one of the most common penguins housed in captivity. The first aim of the research, carried out at Genova Aquarium since March .98, was to collect as many vocalisations as possible and give a formal description of each acoustic behaviour recorded, completed with spectrograms and images. Vocalisations were recorded with two micro- phones (dynamic cardioid MD7000) connected to a S-VHS video-recorder and a DAT recorder. The video-recorder correlates the sound signals with images collected by a S-VHS video-camera set in front the penguin exhibit. Four acoustic categories were collected and will be described with the visually observable behaviours associated. Some preliminary hypotheses about the function of these vocalisations will also be part of the poster presentation.

Whitethroats Sylvia communis have very elaborate song element repertoires. The song repertoire reflects male quality in terms of age and wing length, and high-repertoire males were more likely to mate than low-repertoire males. In this experiment the meaning of different motif song repertoire sizes to unmated males was investigated by playback experiments. Each male was presented with three different playback stimuli: high repertoire, low repertoire, and low repertoire with songs elongated by repeating the last elements of each song to the length of high repertoire songs. The difference in the total number of elements presented between low and high repertoire playback arose since the main part of the variation in motif songs were presented in the last part of the song. Approach measures, time spent near speaker and song activity in response to playback will be examined in relation to the different playback stimuli. Furthermore, the response to the different playbacks will be considered in relation to the males' own song element repertoire in the post-playback period. The risk of losing a competition for territory or attraction of females may depend on the repertoire size of the male relative to that of the intruder's repertoire size.

Ultrasonic communication has been reported in mice in three different social contexts: mother-infant, male-female and female-female social interaction. These signals belong to the same frequency range, around 70 kHz. While in the first two cases ultrasounds have been interpreted as infantile signals that inhibit aggression, no clear functional explanation has been suggested for female-female interaction. As ultrasounds are usually associated with intense sniffing on the part of the emitter, we tested the hypothesis that these vocalisations can be associated with information concerning food quality eaten by a conspecific. We measured the amount of ultrasonic calls emitted during the first three minutes of a female-female interaction. A preliminary analysis showed that most of the vocalisations were emitted by the resident animals. Moreover, the quality of the food eaten by the female intruder affected the amount of ultrasonic calls uttered by the resident; in fact, female fed on highly palatable familiar food elicited a greater amount of calls. This data suggests that animals are able to detect the characteristics of the food eaten by their conspecifics and show more interest is individuals fed on highly palatable food.

Sound production has been widely described for territory holding fish, including cichlids, during reproductive and agonistic activities. Males of the African mouth-brooding cichlid Oreochromis mossambicus defend territories where they dig pits to attract females. Sound emissions, a series of pulses, were studied in this species to determine their association with agonistic and courtship acts. Focal observations on visual and acoustic behaviour were carried out for 8 territorial males of various sizes belonging to 5 mixed sex groups. The number of times each behavioural act occurred with and without sound emission was scored and tested for dependence. Agonistic and courtship episodes were also quantified and correlated with sound production rate. Courtship episodes were categorised into low (either 'tilt' or 'lead' occur) and high (either 'tail wagging' or 'quivering' occur) rank episodes. Sound production was significantly associated with the courtship acts 'tail wagging' and 'quivering' in all fish and with 'pit circling' and 'dig' in larger males. Courtship episodes were positively correlated with sound emission rate, though only for high rank episodes in small individuals. Sound production was not significantly associated with agonistic behaviour.

Many species of fish are vocal and several studies have shown that these sounds emitted by fish have a communicative function. One group of marine fish, the Gadidae or codfish, contains a large number of vocal species. Codfish calls are relatively simple and species-characteristic. However, one gadoid, the haddock, produces a range of different calls in different social contexts. Both male and female haddock produce short sequences of repeated 'knocks' during aggressive encounters. But during reproductive behaviour male fish produce calls that vary in their characteristics as courtship proceeds. We have recently re-examined the repertoire of sounds produced by a small group of spawning haddock, and have attempted to relate the sounds made by males to the different patterns of behaviour shown by the fish. We have also characterised the sequences of behavioural acts that lead up to spawning. Haddock sounds were classified into single and multiple 'knocks'. Each 'knock' was composed of two low frequency pulses. Multiple 'knocks' were further split into 5 different categories based on sound duration and 'knock' repetition rate. The behavioural acts Follow, Flaunt, Solitary Display and Lateral Display were significantly associated with sound production. Sounds became longer and showed higher 'knock' repetition rates as the male fish became more aroused and came closer to spawning. It is suggested that the sounds serve to bring male and female fish together, in the same part of the ocean, and that sounds also play a role in synchronising the reproductive behaviour of the male and female fish.

In most bird species, Acoustic Communication Signals (ACSs) fulfil various functions and form the major part of the intra-specific communication system. The first function to be fulfilled by any animal communication system is the Species-specific Recognition (SSR). In the few animal groups that developed ACSS, these signals generally carry that information. This is particularly the case of birds where, except or a few species that replaced sound signalling with a visual one, a given ACS permits instant SSR. Such ACS should be called the "functional song'', whose definition becomes: ACS carrying the information for SSR. Identification of the "functional song'' is more objective than any previous definition of song, but is not always evident. Our definition implies firstly that ACSs associated with territorial defence and mating, even when similar to the "functional song'', need to be properly distinguished for they fulfil subsequent and fully distinct functions. It implies also that such "functional song'' provides the code message necessary for efficient SSR: this is clear in species with a genetically determined song, but becomes more complicated when learning is involved, reaching a challenging situation in the case of species with versatile or imitative song. This exposition will be illustrated by examples taken from our researches on Brazilian avifauna.

The king penguin Aptenodytes patagonicus is a highly colonial seabird breeding in large and dense colonies on the seashore of subantarctic islands. The ability to recognise mates, parents or chicks is particularly important in seabird colonies, where nest-sites are densely packed. To be fed, a king penguin chick must identify the call of its parents. To study the individual recognition and the coding and decoding features in the call, we played-back to the chicks different experimental signals. All responses obtained were compared with the response obtained by the parent's call. Firstly, we played back experimental signals made from artificially modified recordings of natural signals. Different features of the signals were modified independently of each other in order to determine their importance. Finally, we investigated the recognition process using a series of progressively more complex synthetic signals calculated accordingly to the importance of the parameters that we wanted to investigate.

Hippolais polyglotta and H. icterina are two bird species whose parapatric distributions slightly overlap. The sympatric zone has moved during last decades because of the breeding range expansion of the former species and of the regression of the latter. Both species hybridise and evidence of introgression for morphological characters has been observed in an old zone of sympatry. Only the receding species H. icterina exhibited a morphological shift, suggesting that Melodious warblers asymetrically introgressed into Icterine warblers. Song analyses were carried out on males from sympatric and allopatric areas. This study aimed at assessing the congruence of morphological and vocal patterns of variation within the hybrid zone. 80th species are interspecifically territorial so that competition for space could be a major component of song evolution. In addition, song is learned in oscines. Thus, the patterns of variation of morphology and signal could differ to some extent. A parallel evolution was found for Icterine warblers, suggesting an influence of hybridisation. However, a vocal shift occurred in sympatric H. polyglotta where no morphological evidence of hybridisation was observed. Hybridisation alone cannot explain song variations in Hippolais warblers. Epigenetic and other evolutionary mechanisms are likely involved in producing the observed pattern. Finally, the convergence in signals in sympatry could increase the occurrence of hybridisation and partially explain the dynamics of the hybrid zone.