Passive acoustic monitoring of Yelkouan and Scopoli's shearwaters on Malta and Croatia

All data and scripts uploaded and described here are related to the manuscript being submitted under the title: "Applying a passive acoustic monitoring method for burrow-nesting seabirds on Mediterranean islands". To respect the journal’s anonymity requirements, all names in scripts have been removed as have authors to this repository. Autonomous recording units (ARUs) and detection algorithm All ARUs used in the study were Wildlife Acoustics SM4s (Maynard, USA), recording in stereo and at 24kHz. They were placed at shearwater colonies in Malta and Croatia (Lastovo and Vis archipelagos) between 2018 and 2025. Recordings were first used to train a DNN algorithm to detect shearwater vocalisations in two classes, Yelkouan shearwater (Puffinus yelkouan) and Scopoli's shearwater (Calonectris diomedea). The algorithm is available at: https://anonymous.4open.science/r/Seabirds-detection-DEC5 The test set used for evaluation of the algorithm is available in this repository. We then applied the DNN at 0.75 proabbility threshold to the whole dataset of recordings. The script: "YESHdetections_prep_unrestricted_recordingperiod.R" links the detection segments to each recording for Malta. The script "ARU_detectionsfromalgorithm_Croatia.R" carries out the same process for recordings from Croatia. Recordings without any detections receive a value of zero. Respective detection and recording tables loaded by the scripts are found in the folders Malta and Croatia. Detection segments with start and stop time per recording were then calculated as call rates by taking duration in seconds of each shearwater class and dividing by number of recordings each night per ARU. Abundance data In addition to passive acoustic monitoring, we use Capture-Mark-Recapture data from Malta, analysed using a Jolly-Seber model, code in script “YESH_JS_2018_2025.R”; and occupied nest counts from Croatia. Abundance data is provided in the respective folders. Raw capture-mark-recapture and nest search data is available upon reasonable request. The scripts "CMR_ARU_Trends.R" and "ARU_detectionsfromalgorithm_Croatia.R" test the relationship between abundance and call rates using GLMMs for Malta and Croatia respectively, as the first question made in the study. In the same scripts and using GLMMs together with pairwise post-hoc comparisons, we use the call rates to answer two more questions: The second question made was whether call rates during the breeding season vary over time, both across years and within breeding seasons. The third question was whether any temporal trends in shearwater call rates can be explained by conservation measures, specifically rodent control, carried out at some of the colonies. Acknowledgements The development of the algorithm was funded through LIFE Artina, with co-financing from the GO Co-Financing Fund (NCF), managed by the Malta Council for the Voluntary Sector (MCVS) and supported by the Ministry for Inclusion and the Voluntary Sector (MIV). Conservation actions were possible due to several EU LIFE projects. In Malta the LIFE Yelkouan shearwater project initiated rodent control at Rdum tal-Madonna during 2007-2010, while LIFE Arċipelagu Garnija allowed for expansion to further colonies in 2018 and LIFE PanPuffinus! funded the conservation action until 2025. LIFE Artina implemented rodent management actions in Lastovo archipelago. Data collected in 2021 emanated from the EU funded project EMFF 8.3.1 under the European Maritime and Fisheries Fund 2014–2020 with a total cost of €1.6 million in public eligible EMFF funds (75% EU and 25% MT). The goals of this European Maritime and Fisheries Fund (EMFF 2014–2020) project was to devise a holistic approach toward marine monitoring and develop a comprehensive database of data collected about the Maltese waters. The copyright of such data is the property of the project in line with the provisions of GF/Admin/48/2020 and will remain vested in the project in question. 2021 data was provided by the ERA, Malta.