Pesticides were detected in 11 (78%) of the boreholes, and in 8 (57%) at concentrations exceeding the PCV on at least one occasion. These findings were complemented by data from deeper boreholes in which concentrations exceeding the PCV were also detected. Pesticide occurrence could be related to landuse; mecoprop and isoproturon were frequently detected at several locations reflecting the arable landuse, dicamba at a golf course, and atrazine, propazine and terbutryn near a railway. Temporal variations have emerged at individual sites and across the study area as the result of extended and frequent monitoring.1

Diffuse pollutants such as pesticides pose a significant threat to groundwater quality. Following the wet autumn and winter of 2000-2001 an upturn in trace pesticide concentrations was detected in blended water drawn from part of the Triassic Sandstone aquifer. A groundwater sampling and monitoring programme was undertaken to assess whether this increase would continue. Over a period of 18 months, 190 groundwater samples were collected from the upper part of the aquifer and analysed for 40 pesticides and for nitrate. A total of 25 pesticides were found of which 16 were detected more than once. Positive pesticide detections were in excess of the EU maximum admissible concentration (MAC) for individual substances of 0.1 µg l-1 on 33% of occasions. The most commonly detected pesticides generally have higher agricultural application rates and/or relatively greater solubilities. Very high concentrations of mecoprop (up to 7.1 µg l-1) were consistently found at two of the sampling sites. Analysis of mecoprop enantiomers suggested more than one source for this contaminant. High concentrations of atrazine (up to 1.4 µg l-1) were also detected at three sites and high concentrations of isoproturon (1.2 µg l-1) were found where very high mecoprop concentrations were also present. Nitrate concentrations exceed the EU MAC of 11.3 mg l-1 N at four sites. The spatial and temporal distribution of pesticides obtained from the monitoring network shows no clear trends for prediction of future concentrations. Nitrate-N concentrations and pesticide detections show no clear relationship, suggesting different source, transport or degradation pathways.

The Corallian limestone of northern England (UK) is widely exploited for water supplies and exhibits the karstic phenomena of sinking rivers, conduit development and groundwater velocities of several kilometres per day. To test a number of model-derived source protection zones and elucidate contaminant transport mechanisms in the aquifer, three tracer tests were conducted from a set of swallow-holes draining the River Derwent toward public water supply wells in the eastern part of the aquifer. Tracers used included: Enterobacter cloacae (bacteriophage), Photine C (optical brightener), sodium fluorescein (fluorescent dye) and sulphur hexafluoride (dissolved gas), the varying properties of which make them suitable analogues for different types of potential contaminant. Observed tracer transport times and arrival patterns indicate that tracer transport occurs through karstic channels embedded in a network of primary fissures which exert control over tracer concentrations once initial tracer plumes have passed. A dipole flow system is observed between the swallow-holes and the closest abstraction well, whilst previously modelled source protection zones do not accurately reflect either groundwater velocity or those areas of the aquifer supplying the wells. These findings imply that managing such aquifers for potential contamination should rely upon empirical tracer evidence for source-protection zone modelling.