Temporal trends of organochlorines in Northern Europe, 1967-1995. Relation to global fractionation, leakage from sediments and international measures.

The time trend monitoring of organochlorine pollution was carried out in Sweden since the late 1960s. This report presents data on concentrations of DDT, PCB, HCHs and HCB in biota samples collected and analysed annually. All the matrices and compounds studied show a significant decrease over time. The data cover severely polluted Swedish marine and fresh water in southern Sweden as well as locally unpolluted waters in remote northern Arctic regions of Sweden. A total of 13 time series representing different locations and species are presented for the different pollutants. The period studied covers the time when pollution was serious as well as the time of recovery. All monitoring activities were carried out at the same laboratories over the entire study period, which means that comparability over time is good in the sets of data presented. The various time trends show a convincing agreement with trends and annual change over time, although the concentrations differ between the species and locations investigated, the highest concentrations being in the south. Since the annual changes are normally similar regardless of locations and species, spatial variations in concentrations remain over time, although concentrations are lower today. The onset of changes in concentrations over time can be related to international measures or other circumstances that lowered releases into the environment. Similarities in the annual changes, as well as the time when changes began, are discussed with respect to suggested hypotheses on the fate of the investigated organochlorines. It was not possible to verify that the oxygenation of anoxic sediments mobilised old pollution in Baltic sediments. Neither was it possible to conclude that eutrophication has caused a measurable effect on the rate and timing of the decreases. Finally, long-range transport to Arctic regions seems to be due more to a one step transport than to the 'Grass-hopper' effect. The comprehensive database used, clearly shows how important it is to have datasets big enough to describe between-year variation before attempting to evaluate the time trend. In addition, if between-year variation is not known, it is then also difficult to evaluate spatial variation on the basis of single year observations.