University of Wisconsin Digital Collections
Link to University of Wisconsin Digital Collections
Link to University of Wisconsin Digital Collections
Ecology and Natural Resources Collection

Page View

LeMasters, Gary S.; Doyle, Douglas J. / Grade A dairy farm well water quality survey
[DNR-058] (1989)

Methods,   pp. 3-10 PDF (430.5 KB)

Page 5

Potential Non-sampling Errors
Time Dependency
A special challenge associated with the sampling design is the time
dependency of groundwater recharge and leaching of pesticides and NO3-N
through soils. It was impossible to test all wells in a short time frame. The
laboratory could process twenty-five samples per week and samples cannot be
stored for more than seven days, so the samples had to be collected over a six
month period. The effects of time dependency may have been confounded by
the summer long drought of 1988 and are difficult to quantify without repeated
sampling procedures. Retesting wells would improve the explanatory power of
this pilot study.
Laboratory Detection
The analytical method used in this study has a detection level of
approximately 0.15 ug/l for the pesticides and approximately 0.5 mg/l for NO,-
N. Therefore, if the concentration of the pesticide in question was less than
0.15 ug/l it was not detectable and was considered a zero reading. As a result,
the actual mean concentration of a pesticide in wells on Wisconsin Grade A
dairy farms and the proportion of wells with any pesticide or NO-N may be
underestimated. However, the detection level is quite low and is probably less
significant as a health standard than as a statistical concern.
Water Collection Site
In addition to selecting a random sample of Grade A dairy farms, it was
necessary to specify which well the water collector would sample if a dairy
farm had two or more wells. We specified that the water would be drawn from
the well that supplies water to the two compartment wash sink in the milk
house that is required of each Grade A milk producer. If there was more than
one well the water would be sampled from a tap connected to the well most
often used in the milk house.
There are two problems with this procedure. First, any filters attached to the
supply line before the tap could affect the concentrations of pesticides and NO,-
N in the sample. However, no research is available on the effects of water
treatment on these compounds. Secondly, in the case of multiple wells accessed
through the milkhouse, the water collector had to judge which well was most
often used in the milkhouse. These potential sources of bias could not be
controlled without a more stringent experiment.

Go up to Top of Page