Boluses: an effective method for assessing the proportions of cephalopods in the diet of albatrosses

The method of collecting and analyzing boluses to characterize the cephalopod diet of albatrosses has been used in many diet studies. However, no study has validated this method. We compared boluses and stomach samples from Gray-headed Albatrosses (Thalassarche chrysostoma) and Black-browed Albatrosses (T. melanophris) to (1) study the consumption and diversity of cephalopods in these species, (2) investigate biases associated with each sampling method, and (3) estimate the number of samples needed to characterize these albatross's cephalopod diet. We found that collection and analysis of boluses is a simple, efficient, and noninvasive method for assessing the cephalopod diet of these albatross species, but it is inadequate for characterizing the more easily digestible dietary components, such as fish and crustaceans. Both boluses and stomach samples showed that the two albatross species fed on cephalopods of similar sizes and from the same families (Ommastrephidae, Onychoteuthidae, and Cranchiidae). Furthermore, the main prey species (Martialia hyadesi, Kondakovia longimana, and Galiteuthis glacialis) and the total number of cephalopod species consumed (18-24 species) were the same for both albatrosses. To include all cephalopod species, using a sample-randomization technique, a minimum of 61 and 43 boluses were needed for Gray-headed and Blackbrowed albatrosses, respectively; but to adequately describe the diversity and size frequency of the main prey species, 82 and 371 boluses would be needed. Received 4 October 2003, accepted 30 March 2005. Key words: albatrosses, Cephalopoda, diets, food sample methods.