A quantitative analysis to detect mass mortality events (wrecks) in Procellariiformes

This is the authors' abstract. We don't add key points for this paper.

Mass mortality events (MMEs) or ‘wrecks’ represent sudden and large‐scale mortality episodes often associated with extreme environmental conditions. Despite their ecological importance, MMEs remain poorly documented in the South Atlantic, particularly for Procellariiformes, a group highly vulnerable to climate‐driven anomalies. Detecting such events in this region is challenging because baseline mortality levels are unknown, limiting the ability of monitoring programs to identify abnormal mortality episodes. We developed a quantitative framework to detect and characterize seabird MMEs by integrating time‐series decomposition, changepoint analysis and kernel density estimation. The framework was applied to 8 years (2017–2024) of standardized beach monitoring data covering approximately 3320 km of the Brazilian coastline. This approach allowed us to identify anomalous mortality peaks, estimate the temporal duration of mortality episodes and delineate their spatial extent and hotspots along the coast. We detected seven previously unreported MMEs impacting 3779 stranded Procellariiformes. Events occurred primarily during austral autumn in northeastern and during austral spring in southeastern–southern regions, revealing recurrent seasonal and spatial patterns. Cory's Shearwater Calonectris borealis , Manx Shearwater Puffinus puffinus and Atlantic Yellow‐nosed Albatross Thalassarche chlororhynchos were the most affected species, with juveniles disproportionately represented. Kernel density analysis identified recurrent hotspot regions of mortality along the Brazilian coast. We propose mortality thresholds in an early‐warning system to guide monitoring efforts during abnormal stranding events. Synthesis and applications . Our study provides the first systematic assessment of seabird mass mortality events in the South Atlantic and introduces a transferable statistical framework for detecting MMEs in other taxa and regions lacking baseline mortality information. By identifying recurrent temporal anomalies, high‐risk areas and vulnerable species, we could establish quantitative indicators of MMEs. This approach offers a practical tool for monitoring programs to improve early detection of mortality events and to implement rapid response protocols.