Effective conservation and management are paramount to long-term recovery
of endangered species. Achieving recovery relies on knowledge of
their behaviour and habitats. Neophoca cinerea (Australian sea lion) is an endangered
species endemic to Australia. This study focused on three key questions
aimed at improving its management near Perth (Western Australia): testing
a sea lion identification tool, describing haul-out patterns, and identifying
human disturbance sources. N. cinerea whisker spot patterns as a tool for individual
identification were tested using Chamfer distance-transform. Patterns
contained sufficient information to reliably (99%) identify individuals in populations
of 50, matching 90% correctly when testing known captive animals
photographed at 90°. Off-angle photographs resulted in 48% correct matches.
Resighting in the wild proved unfeasible in this study. However, resights of
four scarred N. cinerea at Carnac and Seal Islands (the main study sites) confirmed
returns and visitations to both islands. To describe haul-out patterns,
generalized additive models were applied to hourly counts between 0800h-
1600h. N. cinerea numbers followed 17-18 month cycles, inversely aligned
with the breeding cycle. During non-breeding seasons, hauled-out numbers
increased throughout the day, and were associated with air temperature and
tide. Research investigating human disturbance indicated that all human activity
types elicited responses, which varied between islands depending upon
stimulus types (vessel types, people), ranges to stimuli, and activities at each
islands. People at close range elicited most elevated responses, including aggression
and retreating. This occurred mainly at Carnac Island since direct
beach access is allowed. Significant rates of lower-level disturbances were
also elicited by vessels within close proximity, including paddlers and those
undertaking noisy activities. Resulting recommendations include further restriction
and enforcement in approach distances allowed. Also, ongoing monitoring
of abundance and behaviour is required for long-term trend estimation.
Because of high variability in haul-out behaviour, surveys undertaken at comparable
times are recommended.