188
minor local sources. Concentrations in the river Elbe - generally the most important
source of pollutant input to the German Bight - are relatively low, and thus also the
influence on chlorpyrifos contamination is low.
A comparison of these findings with literature data is difficult because LOD in most
studies are much higher (>ng/L) than in this survey, and chlorpyrifos is mostly reported
to be below the LOD.
It was found in several studies (e.g. Dabrowski et al., 2002) that concentrations of
chlorpyrifos in the lower ng/L range for water and low pg/kg range for sediments
occurred in river runoff following application of the pesticide. These studies were
carried out mainly in areas with intensive use of the insecticide - in the US, China, and
southern Europe.
Atmospheric transport and deposition is well documented. Apart from atmospheric
transport after spraying applications, volatilisation and escape to the atmosphere has
been suggested (Nhan et al., 2002). Strong evidence for the importance of atmospheric
long-range transport is provided by Gabarino et al. (2002), who observed high
concentrations of chlorpyrifos (70 - 80 ng/L) in arctic snow in Alaska.
The fate of chlorpyrifos in the marine environment was investigated by Nhan et al.
(2002) and Kale et al. (1999) in 14 C-chlorpyrifos experiments in microcosms.
Considerable degradation was observed in the marine environment, with TCPy the main
degradation product. Although rapid sorption onto sediment was observed, only 1-2 %
of the initial amount of 14 C-chlorpyrifos was detected in sediment by the end of the
experiment. Accumulation in fauna and flora reached maxima of 5.8 and 2.2 %,
respectively. The balance of radioactivity suggests that the main loss of the insecticide
from the system occurred through volatilisation and escape to the air. The persistence
half-life in the microcosms was calculated at 5 days. Similar half-lifes of 3.5 and 20
days in pond water are reported by Racke (1992).
Sorption to sediments did not lead to significant enrichment in the sediments
investigated. Therefore, sediments should not be the primary monitoring matrix. This is
