Marine bioacoustics is an increasingly important tool for studying coastal ecosystems, as it allows both biological sounds and anthropogenic noise sources to be monitored within the underwater acoustic environment. This thesis used hydrophone-based passive acoustic monitoring to describe the soundscape of selected sites in the Aeolian Archipelago over a three-year study period, with particular attention to vessel-related anthropogenic noise and cetacean acoustic activity. Recordings were manually analyzed through aural inspection and spectrogram visualization. Acoustic events were classified into biological and anthropogenic categories, and the analysis included event counts, event rates, spatial and temporal distribution, day/night comparison, duration and frequency characteristics, and potential co-occurrence between cetacean vocalizations and boat noise. The results showed a clear dominance of anthropogenic acoustic events, mainly represented by boat noise, which was the most frequent sound subtype in the annotated dataset. Anthropogenic noise was not evenly distributed among sites and showed a higher event rate during daytime, suggesting a relationship with the daily rhythm of human maritime activity. Cetacean vocalizations were detected only in the Salina dataset from 2023, with 301 events concentrated on three recording dates and mainly during evening and nighttime hours. Isolated whistles represented the majority of cetacean detections, whereas whistle trains were less frequent but showed greater variability in duration and frequency range, indicating episodes of more intense vocal activity. A small proportion of cetacean events temporally overlapped with boat noise, suggesting possible conditions of acoustic masking, a potential reduction in communication space (active space), acoustic plasticity, or simply the shared use of the same area by cetaceans and vessels. Overall, this study provides a first acoustic characterization of selected coastal areas of the Aeolian Archipelago. In this context, passive acoustic monitoring proved to be an effective tool for describing cetacean acoustic activity, the temporal distribution of anthropogenic noise, and the potential interactions between these two components. However, the findings also confirm that PAM primarily documents acoustic occurrence, and that a more comprehensive ecological interpretation requires integration with complementary methods such as visual observations, vessel-tracking data, and calibrated acoustic measurements.
La bioacustica marina rappresenta uno strumento sempre più importante per lo studio degli ecosistemi costieri, permettendo di monitorare sia le componenti biologiche sia le sorgenti di rumore antropico presenti nel soundscape subacqueo. Questa tesi ha utilizzato un approccio basato sul monitoraggio acustico passivo tramite idrofoni per descrivere il contesto acustico di siti selezionati dell’Arcipelago Eoliano nell’arco di tre anni, con particolare attenzione al rumore antropico associato alle imbarcazioni e alla presenza acustica dei cetacei. Le registrazioni sono state analizzate manualmente attraverso ascolto e ispezione degli spettrogrammi, classificando gli eventi acustici in categorie biologiche e antropiche. L’analisi ha incluso il conteggio degli eventi, il calcolo degli event rate, la distribuzione spaziale e temporale, il confronto giorno/notte, le caratteristiche di durata e frequenza, e la possibile co-occurence, ovvero una sovrapposizione temporale tra vocalizzazioni dei cetacei e rumore di barca; quindi, situazioni in cui vocalizzazioni dei cetacei e rumori antropici erano presenti nella stessa registrazione o in stretta associazione temporale. I risultati hanno mostrato una netta prevalenza degli eventi antropici, principalmente rappresentati dal boat noise, che costituiva la componente dominante del dataset annotato. Il rumore antropico non risultava distribuito uniformemente tra i siti e presentava un event rate maggiore durante il giorno, suggerendo una relazione con il ritmo giornaliero delle attività marittime. Le vocalizzazioni dei cetacei sono state rilevate esclusivamente nel dataset di Salina del 2023, con 301 eventi concentrati in tre date e prevalentemente nelle ore serali e notturne. Gli isolated whistles rappresentavano la maggior parte delle vocalizzazioni, mentre i whistle trains, meno frequenti, mostravano maggiore variabilità in durata e range di frequenza, indicando episodi di attività vocale più densa. Una piccola parte degli eventi cetacei risultava associata temporalmente al rumore di barca, suggerendo possibili condizioni di masking acustico, una potenziale riduzione dello spazio di comunicazione (active space), fenomeni di plasticità acustica oppure un semplice uso condiviso dell'area da parte di cetacei e imbarcazioni. Nel complesso, lo studio fornisce una prima caratterizzazione del soundscape di alcune aree costiere dell'Arcipelago Eoliano, evidenziando come il monitoraggio acustico passivo rappresenti uno strumento efficace per descrivere l'attività acustica dei cetacei, la distribuzione temporale del rumore antropico e le potenziali interazioni tra le due componenti. Tuttavia, i risultati confermano anche che il PAM documenta principalmente l'occorrenza acustica degli eventi e che una piena interpretazione ecologica richiede l'integrazione con osservazioni visive, dati sul traffico nautico e misure acustiche calibrate.
Ascoltando il mare: studio idrofonico del rumore antropico e dell'attività acustica dei cetacei nell'Arcipelago delle Eolie
URBINATI, MANUEL
2025/2026
Abstract
Marine bioacoustics is an increasingly important tool for studying coastal ecosystems, as it allows both biological sounds and anthropogenic noise sources to be monitored within the underwater acoustic environment. This thesis used hydrophone-based passive acoustic monitoring to describe the soundscape of selected sites in the Aeolian Archipelago over a three-year study period, with particular attention to vessel-related anthropogenic noise and cetacean acoustic activity. Recordings were manually analyzed through aural inspection and spectrogram visualization. Acoustic events were classified into biological and anthropogenic categories, and the analysis included event counts, event rates, spatial and temporal distribution, day/night comparison, duration and frequency characteristics, and potential co-occurrence between cetacean vocalizations and boat noise. The results showed a clear dominance of anthropogenic acoustic events, mainly represented by boat noise, which was the most frequent sound subtype in the annotated dataset. Anthropogenic noise was not evenly distributed among sites and showed a higher event rate during daytime, suggesting a relationship with the daily rhythm of human maritime activity. Cetacean vocalizations were detected only in the Salina dataset from 2023, with 301 events concentrated on three recording dates and mainly during evening and nighttime hours. Isolated whistles represented the majority of cetacean detections, whereas whistle trains were less frequent but showed greater variability in duration and frequency range, indicating episodes of more intense vocal activity. A small proportion of cetacean events temporally overlapped with boat noise, suggesting possible conditions of acoustic masking, a potential reduction in communication space (active space), acoustic plasticity, or simply the shared use of the same area by cetaceans and vessels. Overall, this study provides a first acoustic characterization of selected coastal areas of the Aeolian Archipelago. In this context, passive acoustic monitoring proved to be an effective tool for describing cetacean acoustic activity, the temporal distribution of anthropogenic noise, and the potential interactions between these two components. However, the findings also confirm that PAM primarily documents acoustic occurrence, and that a more comprehensive ecological interpretation requires integration with complementary methods such as visual observations, vessel-tracking data, and calibrated acoustic measurements.| File | Dimensione | Formato | |
|---|---|---|---|
|
Urbinati Manuel tesi. PDF-A.pdf
non disponibili
Dimensione
1.83 MB
Formato
Adobe PDF
|
1.83 MB | Adobe PDF |
I documenti in UNITESI sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/20.500.12075/28158