hydrologic factors affecting changes in stream phosphorus concentrations. Stream
discharge and estuarine salinity may be used to explain either increases in fecal
coliform counts due to transport in runoff or decreases due to die-off from high
salt concentrations. Temperature determines solubility of dissolved oxygen (DO),
making it an important explanatory variable for DO. Biochemical oxygen demand
(BOD) can deplete DO and may be an explanatory variable for a DO monitoring
program. Both suspended solids and chlorophyll a can affect Secchi depth
transparency, making them appropriate explanatory variables. For monitoring
trout abundance, one explanatory variable may include the percent fines in
substrate sediment because substrate composition affects reproductive success.
Similarly, the area of undercut banks is a measure of hiding cover to escape from
prey. Relevantevents that could affect monitoring results, such as droughts,
floods, and storms, or fishery management and harvest, should be tracked and
documented. The last row of Table 3.2 lists a generic primary water quality
variable y and a land treatment variable x to show that land treatment variables
should be measured along with the water quality explanatory variables.
Table 3.2. Example Primary Variables and Explanatory
Variables for Trend Monitoring.
Variable
Primary Variable
Total phosphorus
Antecedent precipitation, stream
discharge, water table depth
Fecal coliform
Stream discharge, estuarine salinity
Dissolved oxygen
Water temperature, biochemical oxygen
demand
Secchi depth
Suspended solids, chlorophyll a
Percent fines in sediment, area of
Trout abundance
undercut banks
Water quality variable y
Land treatment variable x
One approach to identifying appropriate explanatory variables is through a
statistical analysis of a historical data set. Explanatory variables should be selected
because they measure factors in the ecosystem that are thought to effect the
primary variable(s) of concern. A check should be made to assure appropriate
selection by verifying that the selected explanatory variable and the primary
variables are statistically correlated (e.g., using linear regression techniques).
For some monitoring programs, variables or metrics may be summarized and
combined into an index. An index contains less information and therefore less
explanatory power than the original data, but it may be more easily used and
understood by the public or the decision-makers. Indices are chosen for their
ecological meaning and ability to summarize information on community struc-
ture, function, or response to pollution.
3.4
Previous Page
