Using the previous example of hydrogen sulfide, production and subsequent degassing in the discharge,
this problem may be considered as a seasonal problem and sampling would be concentrated prior to
and during the anoxic period. Using techniques for determining sample size then allows for the
determination of a reasonable distribution of sampling effort with the available number of samples. In
the case of elevated metals concentrations, this is also a seasonal event but is also related to reservoir
operations and sampling effort may focus on critical periods (e.g., low flow or maximum
concentrations) or on total loads (more frequent measurements during increased discharge). For
episodic events, i.e., a fish kill, intensive sampling may be the best approach and frequency is again a
function of sample size (often limited by resources).
Other considerations for sampling frequency include the response of the water resource to
changes in pollutant source, the magnitude of the minimum amount of change required for detection
(MDC), the system variability, and the probability of detecting a trend (statistical power) (USEPA
1991). MDC is defined as the minimum change in a water quality parameter over time that is
considered statistically significant. The MDC can be estimated from historical records to aid in
determining the required sampling frequency and to evaluate monitoring feasibility (Spooner et al.
1987). An example of sample allocation is provided later in this chapter in the tools section.
4. Site Location
Where to sample is also a function of the perceived problem and sample frequency (or
monitoring program) designed to identify the source of the problem and may be a single site, established
fixed locations, or several stations determined in the field as in an intensive study.
For fixed locations, representativeness (discussed in more detail later in this chapter) and accessibility
are important considerations. The sites selected must represent conditions for the lateral transect of the
section or include variability estimates and must be accessible for sample and data collection and
instrument calibration and maintenance. The use of established USGS gaging stations is a good choice
for correlation to discharge and possible access to a secured site. While bridges, boat ramps, and
recreation sites are often easily accessed, representativeness at these sites should be evaluated,
particularly if the flow is highly variable (due to islands, bridge supports, abutments, etc.) or channel
morphometry has been modified (e.g., protected or dredged areas for boat launching or recreation).
Other considerations include changes in local geology which may effect certain chemical constituents,
changes in land use, differences in riparian conditions, and locations of groundwater, tributary, and point
sources. All of these factors could contribute to changes in longitudinal distributions and can be
addressed with appropriate site location (e.g., upstream and downstream comparisons between sites).
A final note on site location is to mention that sampling from a habitat perspective may be a useful
approach. For instance, if the fisheries or benthic community is the focus of the study, then sampling in
a variety of habitats at several locations may be a better approach than more stations distributed
longitudinally or more frequent sampling at selected sites. Site location is a characterization process
determined by the objectives of the study.