available in some electronic format, but it may need to be verified and adjusted for scale. Some desired information is perhaps not readily available and expensive to obtain during the years of the study, such as: actual pesticide usage, actual timing of application, agronomic measures, crop types, weather (esp. daily precip, temp, wind), and hydrologic (stream discharge) data. Perhaps this detailed data collection should be for selected subsets of CWSs, and the more general the ancillary data collected for all sites.
It is important that historical records on pesticide monitoring at each selected CWS be obtained and used to augment the sampled data. Estimates of pesticide use for each watershed will also need to be obtained. USDA should be consulted on how they may be able to help provide use data. These data collection costs need to be estimated and included in the program budget prior to implementation of the monitoring program. A significant portion of the funding for the project may need to be allocated up front to cover the costs of obtaining and analyzing this information.
9.) We have defined our population as CWSs with pesticide use in their watersheds, based on the assumption that runoff and near field spray drift are the major routes of loading to the water supply.
Is this a reasonable assumption and should we monitor facilities that do not have
pesticide use in the watershed?
In general, defining the population of pertinent CWS by pesticides used in their watersheds is appropriate. It should be realized that there are a number of situations where pesticides are used or introduced into the watershed beyond that which is reported in agricultural use statistics
the data that would be used to estimate pesticide use for the watersheds. These situations
include illegal use, non-agricultural use (e.g. urban use, forestry, aquaculture, roadways), manufacturing/distribution, and atmospheric deposition. Chemical loading from groundwater discharge to streams is also a component, but this would be most significant from within the watershed.
In the FIFRA SAPs experience, and from their review of the scientific literature on pesticide use in forestry and aquaculture, long-range atmospheric deposition generally can be considered insignificant for the purpose of this survey. While pesticide release during manufacturing/distribution could likely be considered insignificant for the purpose of this survey, as well, the locations of pesticide manufacturing facilities should certainly be understood in relation to sampling locations. They might be purposefully excluded or included but their potential impact shouldn't be allowed to confound the results. Illegal use is always an unknown. Non-agricultural use statistics generally do not exist, as is reported in the Agenecy background document. Roadway use could be important for some watersheds, but the use varies greatly from state to state. Urban use of pesticides may have a substantial effect on their stream concentrations. As expected, USGS studies suggest the environmental fate of agricultural and urban use pesticides differ.