Carbon fluxes in a marine trout cage farm in the Gullmar Fjord, western Sweden, were measured to investigate how much of the carbon supplied to the farm was recovered in harvest, how much was lost to the environment, and the properties and fate of this environmental loss. The measured fluxes included fish food, juveniles, harvest, fish loss (death and escape), sedimentation from the cages, and benthc release measured with diver-operated flux chambers and a gas collection unit in situ. Carbon mass balances for the farm, based on the measured fluxes (flux method), were constructed for each of 2 consecutive growing seasons. Another mass balance (accumulation method) was based on the total carbon input with food and juveniles to the farm since it was started, the removal of carbon with harvested fish and fish loss, and the recovery of carbon in the sediment originating from the farm after 7 growing seasons. Some 21 to 22 % of the total carbon input to the farm was recovered in harvest, fish loss constituted 1 to 3 %, and 75 to 78 % (or 878 to 952 kg C per tonne of fish produced) was lost to the aquatic environment. On a seasonal basis and of the carbon input to the farm, solute release from the cages (probably CO2 produced during fish respiration and excreted urea) removed 4 to 49 %, sedimentation of faeces and excess food removed 29 to 71 %, flux from the farm sediment of dissolved and gaseous carbon (total carbonate [C,], methane and dissolved organic carbon) transferred 2 to 6 O/o back to the overlying water, and 23 to 69 % was accumulated in the sediment. (Ranges of values represent inter-seasonal variability.) The long-term (7 seasons) sediment accumulation of carbon amounted to 18 'Yo of the total carbon input to the farm. Of the carbon deposited on the sediment surface 3 to 20 '/o was released back to the overlying water seasonally. C, dominated the annual benthic fluxes. Loss to the environment of dissolved carbon [the sum of solute release from the cages and benthic flux) amounted to between 6 and 55 % of the carbon input to the farm (or 9 to 71 'X of the total environmental loss) on a seasonal basis, and 5 8 % (or 76% of the total environmental loss) on a long-term basis. This study constitutes the first step in an assessment of the eutrophication caused by the fish farm.

Phosphorus fluxes were measured in a marine fish cage farm in the Gullmar Fjord, western Sweden. The measured fluxes included fish food, juveniles, harvest, fish loss (dead fish and escapers), sedimentation from the cages and benthic solute release (in situ). Two different types of mass balances were constructed. The flux method was based on seasonal input of fish food and juveniles, sedimentation and removal by harvest, fish loss and benthic fluxes Mass balances according to the flux method were constructed for each of 2 consecutive growing seasons The accumulat~on method was based on total input of phosphorus through fish food and juveniles, removal by harvest and fish loss, and net accumulation in the s e d m e n t over 7 growing seasons. Both types of mass balances gave similar results. Of the total phosphorus input to the farm, 17 to 19 % was recovered in harvest, fish loss constituted 1 to 4 %, and 78 to 82 % was lost to the environment. The environmental loss of phosphorus for each ton of fish produced was 22.4 k g (1985), 19.6 kg (1986) and 21.9 k g (1980 to 1986). Of the loss to the environment, 34 to 41 O/ O was in dissolved form and 59 to 66 % was accumulated in the sediment. O n a seasonal basis, the benthic flux transferred 4 to 8 % of the sedimented phosphorus back to the overlying water This constituted about 1 O/o of the total phosphorus content in the sediment derived from the fish farm.

Gasliquid chromatography provides a method of extreme sensitivity for the detection and determination of organochlorine pesticides residues. A single retention-time cannot, however, be regarded as a reliable identification of an unknown compound. A "clean-up" procedure is described by which interference is virtually eliminated, and a technique is proposed in which the change in retention-time after simple chemical reactions provides confirmation of the identity of a pesticide.