“…To overcome these problems, studies of modern 'death assemblages' (groups of organisms that were not necessarily associated during their lives but were recovered together after death) have been compared with archaeological faunas (e.g. Osborne, 1983;Smith, 1996aSmith, ,1996bSmith, , 1998Smith, , 2000Hellqvist, 2004;Smith et al, 2005;Kenward et al, 2012). Also, multivariate statistics have been used to define 'indicator groups' of taxa by their recurrent associations in the archaeological record (Carrott & Kenward, 2001;Kenward & Carrott, 2006;Smith, 2012).…”