5. Reliability of Archaeological Evidence and Gaps in Collections

An important aim in developing the Archaeological Fish-Bone Images archive was to identify major gaps in Australian fish reference collections that currently limit our ability to interpret important aspects of Sydney regional prehistory and historical archaeology. Researchers want to discuss long-term changes to the natural environment in the Sydney region, the importance of fish and fishing to Aboriginal people before British colonisation in AD 1788 and the impact of colonialism on the way people lived in Sydney, including their relationships with each other and with Sydney's maritime environment. Historical evidence can tell only some of these stories (e.g. Karskens 2009). Material evidence recovered through archaeology is also essential and any apparent differences and contradictions between archaeological and historical evidence merit particular attention.

The Australian Museum (2009) lists over 140 families representing at least 580 fish species for Sydney harbour alone, yet only 33 native Australian fish taxa have been reliably reported from local archaeological sites (Table 5). Many of these have only been identified to genus, Family or a more general category rather than exact species. Several fishes described in documents as significant to Aboriginal people in the late 18th century are absent from the archaeological record (Attenbrow 2010), while some fishes found in colonial-period sites have not yet been identified in pre-AD 1788 Aboriginal sites and vice versa (Colley in prep.).

How accurately do such patterns of data represent the reality of prehistoric and historical fishing practices? We know the quality of Australian fish bone data, as elsewhere, can be strongly influenced by site-formation processes, differential preservation and sampling methods, including on-site sieving to recover small bones (e.g. Colley 1987; 1990). Small fish bone samples are common for Australian excavations. For example, only 304 identified fish bone fragments were recovered from Mt Trefle (Vaucluse, Sydney) by Attenbrow and Steele (1995, 52), which is one of the best reported Aboriginal sites. Also relevant are limited skills in specialist fish bone identification in Australia, given the small size and different profile of the profession (Ulm et al. 2005) compared to Britain, Europe and North America, where expert in-depth studies of fish remains are more common. Gaps in fish reference collections for Sydney regional fishes are also crucial. It is impossible to identify many fish remains from archaeological sites in Sydney, New South Wales and elsewhere in southern and south-eastern Australia because this basic research infrastructure is lacking.

Interpretation of archaeological, historical, ecological and taxonomic information (Table 1) suggests at least 68 fish taxa not yet identified from Sydney archaeological sites are potentially important for regional prehistory and historical archaeology. Based on criteria noted above, 'more important' fishes are those which are (or were): more abundant or frequent either year round or seasonally; considered better eating (excellent, good or fair); of greater interest to anglers; of higher commercial value in the 19th century; remarked on more often in the literature and/or described as having some other particular historical or cultural significance. Lower priority fishes also feature in literature but less significantly so, and they are often described as being rare or regarded as poorer quality food and not particularly sought after.

For 21 of these 68 'missing' taxa (Table 6) adequate modern reference specimens are available in ANU and USyD collections, and images of selected bones are now accessible online in the AFBI archive, yet none have been reported from a Sydney archaeological site. While technically possible, it seems unlikely all these fishes are genuinely absent from the archaeological record. Many taxa in Table 6 are species (e.g. Shortfin Eel, Anguilla australis and Longfinned Eel Anguilla reinhardtii), for which members of the same family (Anguillidae) or genus (Anguilla sp.) have been identified archaeologically (Table 5). It is common in Australian archaeozoology to identify fishes only to Family or genus. Depending on physical preservation and the exact skeletal parts involved (which are variably diagnostic to species) it may be possible to identify more archaeological fish remains to species through further examination of existing collections should this be useful for research.

Table 7 lists a further 47 fish taxa not yet identified from any Sydney archaeological site. Bones of most of these fishes were not available in ANU or USyD reference collections, except for a few lower priority taxa deliberately omitted from the ABFI archive at this stage for resource reasons. Some poisonous fishes (Families Diodontidae, Ostraciidae and Tetraodontidae) are listed in Table 7. None of these were mentioned in any documents or historical accounts reviewed so far. However, they have distinctive and robust bones and archaeological examples were previously identified from Aboriginal shell middens in Tasmania (Colley and Jones 1987), where they were interpreted as stomach contents of other fish or marine animals or as representing non-food use. There is no obvious reason why similar bones should not be found in Sydney archaeological sites.

Most of the fishes listed in Table 7 are considered likely to be found in Sydney archaeological sites, but without reference specimens it is impossible to tell. The high and medium taxa in particular should inform further research designed to investigate fishing. The information in Table 7 also provides a guide to which fishes should be given priority when expanding physical and digital reference collections and archives in future.


© Internet Archaeology/Author(s)
University of York legal statements | Terms and Conditions | File last updated: Tue Nov 2 2010