RESUMO
Passive Acoustic Monitoring (PAM) is emerging as a solution for monitoring species and environmental change over large spatial and temporal scales. However, drawing rigorous conclusions based on acoustic recordings is challenging, as there is no consensus over which approaches are best suited for characterizing marine acoustic environments. Here, we describe the application of multiple machine-learning techniques to the analysis of two PAM datasets. We combine pre-trained acoustic classification models (VGGish, NOAA and Google Humpback Whale Detector), dimensionality reduction (UMAP), and balanced random forest algorithms to demonstrate how machine-learned acoustic features capture different aspects of the marine acoustic environment. The UMAP dimensions derived from VGGish acoustic features exhibited good performance in separating marine mammal vocalizations according to species and locations. RF models trained on the acoustic features performed well for labeled sounds in the 8 kHz range; however, low- and high-frequency sounds could not be classified using this approach. The workflow presented here shows how acoustic feature extraction, visualization, and analysis allow establishing a link between ecologically relevant information and PAM recordings at multiple scales, ranging from large-scale changes in the environment (i.e., changes in wind speed) to the identification of marine mammal species.
RESUMO
BACKGROUND: The Mediterranean subpopulation of fin whale Balaenoptera physalus (Linnaeus, 1758) has recently been listed as Vulnerable by the IUCN Red List of threatened species. The species is also listed as species in need of strict protection under the Habitat Directive and is one of the indicators for the assessment of Good Environmental Status under the MSFD. Reference values on population abundance and trends are needed in order to set the threshold values and to assess the conservation status of the population. METHODS: Yearly summer monitoring using ferries as platform of opportunity was performed since 2008 within the framework of the FLT Med Network. Data were collected along several fixed transects crossing the Western Mediterranean basin and the Adriatic and Ionian region. Species presence, expressed by density recorded along the sampled transects, was inspected for assessing interannual variability together with group size. Generalized Additive Models were used to describe density trends over a 11 years' period (2008-2018). A spatial multi-scale approach was used to highlight intra-basin differences in species presence and distribution during the years. RESULTS: Summer presence of fin whales in the western Mediterranean area showed a strong interannual variability, characterized by the alternance of rich and poor years. Small and large groups of fin whales were sighted only during rich years, confirming the favorable feeding condition influencing species presence. Trends highlighted by the GAM can be summarized as positive from 2008 to 2013, and slightly negative from 2014 to 2018. The sub-areas analysis showed a similar pattern, but with a more stable trend during the second period in the Pelagos Sanctuary sub-area, and a negative one in the other two sub-areas. Our findings further confirm the need for an integrated approach foreseeing both, large scale surveys and yearly monitoring at different spatial scales to correct and interpret the basin wide abundance estimates, and to correlate spatial and temporal trends with the ecological and anthropogenic drivers.
RESUMO
This study assesses vessel-noise exposure levels for Southern Resident Killer Whales (SRKW) in the Salish Sea. Kernel Density Estimation (KDE) was used to delineate SRKW summer core areas. Those areas were combined with the output of a regional cumulative noise model describing sound level variations generated by commercial vessels (1/3-octave-bands from 10â¯Hz to 63.1â¯kHz). Cumulative distribution functions were used to evaluate SRKW's noise exposure from 15 vessel categories over three zones located within the KDE. Median cumulative noise values were used to group categories based on the associated exposure levels. Ferries, Tugboats, Vehicle Carriers, Recreational Vessels, Containers, and Bulkers showed high levels of exposure (Leq-50thâ¯>â¯90 dB re 1⯵Pa) within SRKW core areas. Management actions aiming at reducing SRKW noise exposure during the summer should target the abovementioned categories and take into consideration the spatial distribution of their levels of exposure, their mechanical and their operational characteristics.