As evident in Table 3.4a and Table 3.4b, few tissue samples collected for TBT analysis recorded values above the detection limit during both the dry and wet season monitoring (ie 3-12% and 0-3%, respectively). Of the tissues analysed, only 12% of samples in December 1998 and 3% of samples in January 1999 had TBT above the detection limit. Only 3% of tissue samples had detectable TBT in August and it was below analytical detection limits (20 mg Sn kg-1 ) in all biota collected during the September 1999 trawl.

TBT was only detected in tissues from biota trawled in Southern Waters and Deep Bay. Samples collected in Southern Waters showed less contamination than those collected from Deep Bay with the Gizzard Shad and Mantis Shrimp recording highest tissue TBT concentrations. In the wet season sampling, TBT was only detected at concentrations above analytical detection limits in shrimp tissues collected from Southern Waters. All other trawl stations (ie Mirs Bay and Port Shelter) did not record TBT at concentrations above the analytical detection limit. Analytical detection limits for TBT may need to be reduced, therefore, if this contaminant is to be recorded in the tissues of local species more frequently.

Highest TBT concentrations were recorded in Mantis Shrimp (Oratosquilla interrupta) tissues collected from Deep Bay. In December 1998, TBT values ranged from 21-33 mg Sn kg-1 in Mantis Shrimp collected in Deep Bay waters. In the wet season (August and September 1999), however, TBT was not recorded at levels above the detection limit in any Deep Bay biota. The maximum TBT concentration of 32 and 21 mg Sn kg-1 were detected in the Gizzard Shad (Clupanodon thrissa) sampled in Deep Bay during December 1998 and January 1999, respectively. These two phylogenetically distinct species utilise different ecological niches as the Gizzard Shad is a pelagic fish whereas the Mantis Shrimp is a burrowing species. The main exposure route to the Mantis Shrimp is likely to be through TBT adsorbed to sediments. The TBT exposure routes to Gizzard Shad are likely to be through the direct uptake from water (bioconcentration) and from contaminated food (bioaccumulation).

In the dry season, TBT values of 22-23 mg Sn kg-1 were recorded in Sebasticus mamoratus (Rockfish) trawled from Southern Waters. All of the other species collected during the dry season, recorded mean values below the analytical detection limit, ie less than 20 mg Sn kg-1. TBT was recorded less often in biota collected during the wet season. The Cotton Shrimp (Metapenaeus joyneri) and the Mantis Shrimp (Oratosquilla interrupta) were the only biota to show detectable TBT in August 1999 and concentrations were recorded at 14 mg Sn kg-1 (mean value of five replicates) and 12.4 mg Sn kg-1 (mean value of five replicates), respectively(23) . All other tissue samples from Southern Waters were below analytical detection limits for TBT during the wet season trawling.

Levels of TBT in tissue were below the analytical detection limit for all of the species sampled from Mirs Bay and Port Shelter during both dry season (ie during December 1998 and January 1999) and wet season monitoring (ie during August and September 1999). This may indicate an absence, or very low levels, of inputs which would reduce bioavailability and hence tissue contamination.

An evaluation of the environmental significance of the values of TBT recorded from marine life in these Environmental Baseline Surveys is hampered by a lack of comparable studies and also by temporal variation in the target species analysed. Information gathered from both the dry and wet season surveys have been compared from other surveys in Hong Kong such as the EM&A programme for CMP IV and are discussed further in Section 4.2.3.