Analysis of Fin Whale Lunge-Feeding in Southern California Using Multisensory Biotags

• Main Author: Bogan, Leah Kathleen
• Other Authors: Širovič, Ana
• Format: Thesis
• Language: unknown
• Published: 2021
• Subjects: Biotags; Accelerometer; Kinematics; Autonomous; Lunge-feeding; Fin Whale; Balaenopteridae; Generalized Additive Modeling; Balaenoptera physalus; Fin whale
• Online Access: https://hdl.handle.net/1969.1/194388

Balaenopterids are among the largest animals to have lived on earth, yet they are often the most elusive to research. Despite their size, we are still discovering new populations. As technology and the sciences converge, advancements in instrumentation are meeting the challenges where whale study and ocean research intersect. Multisensory bio-logging tags are at the forefront of research innovation able to customize a suite of sensors for remote observation of animals in extreme environments. Biotag data translate to behaviors that enable quantification of vital statistics and inform on individual and population health. Balaenopterids have a distinct feeding behavior termed lunge-feeding which exhibits a unique energetic signature. Quantifying these lunges provides information on dive efficiency, metabolic rates, feeding ecology etc. For this study, fin whale (Balaenoptera physalus) lunging depth was analyzed from 24 biotags deployed from 2010-2018 in southern California for 247 hrs. of recorded data. A generalized additive modeling framework was used to test whether lunge depth (deep, greater than 135m or shallow, less than 135m) was dependent on the time of day (day or night by way of hour), season (spring, summer, or fall) and region (Inshore North, Inshore Central, Offshore). There were distinctions found in depth of lunges over the course of 24hrs with deep lunges occurring primarily during the day and shallow dives at night, likely following a diurnal prey migration pattern. Seasonal distinction in frequency and depth of lunges was also observed, with feeding-lunge depth and frequency increasing from spring, through summer, peaking in the fall. Standardization of rapid analysis using machine learning could lead to improved predictions of whale aggregations based on these feeding behaviors. Correlation of feeding whale density with krill aggregation has the potential of producing real-time density probability predictions of whales, based on the more easily monitored real-time krill densities through ...