Lakes and Reservoirs
The lake or reservoir reference system should be similar in basin shape, size, and
and control lakes
hydraulicdetention times to the treatment
should mix and stratify similarly. Depending on the monitoring objective, other
nontreatment factors such as land use, habitat, and water chemistry may be
A cross-sectional study of several lakes within the same region (more than one
treatment or more than one reference lake or both) may be monitored to increase
the chances that the impact will be detected. Carpenter (1989) discusses the
importance of sufficient treatment strength and the advantage of using a network
of multiple treatment and reference lakes for impact assessment. Climatic factors
influencing the entire network of lakes can be tracked to improve the detection of
For monitoring localized problems on large lakes or reservoirs, a bay or tributary
arm with similar morphometric and hydraulic characteristics may be used as a
reference site; however, careful definition of differences between sites and the
area1 extent of treatment effect must be determined.
Number of Samples
The time between samples or the sampling interval and the number of sampling
events or years of monitoring are key elements of the sampling design.
For monitoring the state of biological variables, the length of the life cycle may
determine the sampling interval. Level macroinvertebrate and fish sampling
occurs generally one to four times a year, with timing adjusted for flows or
reproductive cycles. Level I lake monitoring for water column chemical constitu-
ents may be every 14 days depending on the time of year and the objective. Level
I grab sampling for stream chemical constituents may be every 7 to 14 days,
monthly, or seasonally, depending upon the objective.
Monitoring at regular intervals increases the chance that the monitoring program
can detect a trend. Sampling should be repeated within a year for systems where
the temporal variability is estimated for the year or season and for a measure of its
variability (i.e., mean and coefficient of variation). The extent of repeated
sampling within a year is initially specified by the monitoring program objective
and planned statistical analysis to test the null hypothesis. Consideration should
also be given to the seasonal changes and to the life cycle for biota. Minimum
Monitoring at regular
sampling frequency may be two times the length of the life cycle for some biota.
intervals increases the
Spooner et al. (1987a) developed a method to calculate the MDC in water quality
chance that the monitoring
variables for three RCWP projects. The method was applied to fecal coliform data
for Tillamook Bay, Oregon; total phosphorus and fecal coliform data in Snake
program can detect a
Creek, Utah; and suspended sediment concentrations in Rock Creek, Idaho
subwatersheds. The effect of the MDC with changes in the sampling interval, the
explanatory variable, and the total number of sampling events can be determined.
The concept of the MDC is illustrated in Figure 4.3 for a two-year, four-year, and
a lo-year sampling scheme. Note the decrease in the magnitude of change in
suspended sediment concentration required to detect statistical significance as the
number of years of monitoring increases.