This distinction, however, is often indistinguishable when one goes through environmental monitoring reports. The issue of sample preservation may also be considered. Would data from samples that are preserved and those that are not preserved be comparable?
For contaminants in marine sediments, the issue of sample pre-treatment also pervades. Data on trace metals in sediments, for instance, may be derived from total digests or partial digests of sediments; size-fractionated (e.g., ＜63μm) or analyzed regardless of size fractions; and may consist of sections of cores or integrated samples. Moreover, when data are eventually reported, these range from total concentrations, levels normalized to some size fraction, normalized to the element aluminum, or expressed as enrichment factors.
3. Data may not be reliable, hence, incomparable. This is perhaps one of the most problematic issues that confront the need to effectively compare and exchange data on marine pollution in the region.
The reliability of data is a function of and is no better than all the efforts that lead to the final result (Kramer 1994) the “chain” in data acquisition. Briefly the links in the chain include: sampling, sample pre-treatment (e.g., filtration/centrifugation, addition of chemicals, preservation), transport conditions, storage, sample treatment, instrumental analysis, calculation, and evaluation of results. Many chemists in the laboratories that “generate the data”, do not appreciate the problems associated with maintaining the integrity of samples before analysis. In my limited experience, I suspect that all of the minor and major steps prior to laboratory analysis, if not subjected to the same rigors of quality assurance and quality control that might be adopted in the laboratory, can seriously compromise data.
Even among the chemical parameters that are relatively easy to determine (e.g., Category 1), obtaining accurate results remains a problem for some of these analytes. To illustrate, let me cite a few recent examples. In the mid-1990s, the IOC conducted an inter-laboratory comparison for the determination of nutrients in seawater. While many laboratories did very well, some grossly underestimated or overestimated the levels of nutrients in the samples provided (Figures 2 and 3). This was for an exercise where all of the other steps in the information gathering chain had been isolated and the principal concern was directed at analytical capabilities. One wonders what the results would have been if the exercise started with the sampling step.
At the recent end-of-program conference of the ASEAN-Canada CPMS-II held in 1998, Hugdahl and Watson (1999) reported on the results of the 1997/98 marine pollution monitoring inter-laboratory study. This activity was participated in by 24 laboratories in ASEAN that determined all or some of the following parameters: trace metals in sediment, tissue, and seawater; polycyclic aromatic hydrocarbons in soil and marine sediment; organochlorine pesticides in soil; and selected nutrients in simulated seawater. Some of the results of this activity are shown in Figures 4-7. The overall assessment made was that the performance varied greatly among laboratories and test parameters. Hugdahl and Watson (1999) also recommended that “regional laboratory data could not feasibly be combined in a single database for temporal or spatial trend analysis without the application of rigorous data acceptance criteria.”
4. Certified reference materials are expensive and may be difficult to acquire. Associated with the problem of acquiring reliable data is the use of reference materials to check the reliability of test procedures. The use of these reference materials is widely recognized as an important quality assurance component in marine pollution monitoring programs (Waldichuck et al., 1987).
