Matthiessen, P., Arnold, D., Johnson, A. C., Pepper, T. J., Pottinger, T. G., Pulman, K. G. T., Matthiessen, P., Arnold, D., Johnson, A. C., Pepper, T. J., Pottinger, T. G., and Pulman, K. G. T.
Most studies of hormonal activity in rivers have focused on inputs from sewage treatment works (STW), and their consequences for endocrine disruption in fish. It is possible that livestock is contributing to this hormonal activity in rivers. This study represents a search for evidence of steroid hormone contamination in streams associated with livestock farms. The majority of the 10 sites selected were streams running through dairy farms, although some examples of beef, sheep and pigs were included. Passive water samplers (Polar Organic Chemical Integrative Samplers – POCIS) were deployed up- (control) and down-stream of the farms for 3 to 10 weeks (mean = 39 days) during the period November 2004 to January 2005. At one site, water samples were also taken automatically during rainfall events. All samples were solvent-extracted. Total oestrogenic activity in concentrates of the extracts was analysed using the Yeast Estrogen Screen (YES) calibrated against 17β-oestradiol (E2), while oestrone (E1), E2 and 17α-ethinylestradiol (EE2) were analysed by liquid chromatography–mass spectrometry (LC–MS/MS). Stream water from the entirety of only one rainfall event was sampled directly, but this revealed background activity (E2 equivalents) of 0–0.3 ng/l, rising to a transient peak of 9.4 ng/l. Average oestrogenic activity at this site as estimated from the POCIS samplers was 1.8–2.7 ng E2 equiv./l. Estimated average oestrogenic activity across all sites (with one exception) lay in the range 0– 26.5 ng E2 equiv./l (mean = 2.0 ng/l; S.D. = 5.1), based on the POCIS samples. The outlier was 292 ng/l, and this could not be specifically linked with livestock rearing. 92% of monitoring stations (at least one on each farm) contained some oestrogenic activity, and activity was higher at downstream sites in 50% of cases. Although no EE2 was detected analytically in any stream, E1 and E2 were almost ubiquitous, with E2 equivalents ranging from 0.04 to 3.6 ng/l across all sites. Furthermo