Deep-Sea Research I 55 (2008) 519–531 Uncommon or cryptic? Challenges in estimating leopard seal abundance by conventional but state-of-the-art methods Colin Southwell a,Ã , Charles G.M. Paxton b , David Borchers b , Peter Boveng c , Tracey Rogers d , William K. de la Mare a,e a Department of the Environment, Water, Heritage and the Arts, Australian Antarctic Division, Channel Highway, Kingston, Tasmania 7050, Australia b School of Mathematics and Statistics, University of St. Andrews, North Haugh, St. Andrews, Fife KY16 9SS, Scotland c Alaska Fisheries Science Center, NOAA Fisheries, 7600 Sand Way Point N.E., Seattle, WA 98115, USA d Australian Marine Mammal Research Centre, University of Sydney/Zoological Parks Board of NSW, Mosman, New South Wales, 2088, Australia e CSIRO Marine and Atmospheric Research, Cleveland Laboratories, P.O. Box 120, Cleveland, Queensland 4163, Australia Received 1 August 2007; received in revised form 14 January 2008; accepted 17 January 2008 Available online 6 February 2008 Abstract The method traditionally used to estimate pack-ice seal abundance employs sighting surveys from ships or aircraft to estimate the number of seals hauled out on the ice, combined with studies of haul-out behaviour to estimate the proportion of time spent on the ice. Application of this approach has been improved in recent times by developments in survey methodology and satellite technology that theoretically allow biases in the estimation of hauled-out abundance and haul- out behaviour to be accounted for that previously could not be addressed. A survey using these conventional but state-of- the-art methods was undertaken in the summer of 1999/2000 off east Antarctica between longitudes 641E and 1501E to estimate the abundance of leopard (Hydrurga leptonyx) and other pack-ice seal species. Because they are either uncommon or very cryptic, very few leopard seals were encountered despite a large survey effort. This presented challenges in both application of the methods and analysis of the data. Abundance estimates were derived using a number of plausible predictive models. The model considered as the most reliable returned best estimates of 7300 and 12,100 for definite and definite plus probable leopard seal sightings, respectively, with 95% confidence intervals of 3700–14,500 and 7100–23,400. These estimates are likely to be negatively biased and should be treated as minimum estimates only. r 2008 Elsevier Ltd. All rights reserved. Keywords: Abundance; Distribution; East Antarctica; Haul-out; Model-based inference; Leopard seal; Pack-ice 1. Introduction With a wide distribution throughout the pack-ice of the Southern Ocean and a diverse diet of krill, fish, squid, penguins and seals (Oritsland, 1977), the leopard seal (Hydrurga leptonyx) may have a significant, direct and simultaneous top-down effect on multiple trophic levels in the pack-ice food web. Siniff and Stone (1985), for example, infer a strong influence on crabeater seal population dynamics from observation of a high proportion of crabeater seals with extensive scarring attributed to leopard ARTICLE IN PRESS www.elsevier.com/locate/dsri 0967-0637/$ - see front matter r 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.dsr.2008.01.005 Ã Corresponding author. Tel.: +61 3 6232 3450; fax: +61 3 6232 3449. E-mail address: colin.southwell@aad.gov.au (C. Southwell).