Environmental Sampling Design - Fundamentals of Environmental Sampling and Analysis - page 51

in Chapter 4. Surface waters of various types can be very heterogeneous both
spatially and temporally as a result of flow and stratification, making it difficult to
collect representative samples (Keith, 1990). Stratification is a common problem
in oceans, deep lakes during the stratified seasons (summer and winter), and
slow-flowing streams ( 5 m). Stratification also occurs when two streams of

water merge, such as at the outfall of a wastewater discharge point. Estuarine
water may present special challenge because strata move up rivers unevenly
(Keith, 1990).

>

Representative Biological Samples

The heterogeneity of biological samples presents unique challenges because of the
difference in species, size, sex, mobility, and the tissue variations. Consultation with
a trained ecologist or biologist is often recommended. For example, the selection of
an inappropriate species may introduce an error, which can be minimized by
selecting a species that is representative of the habitat and whose life cycle is
compatible with the timing of the study. Migratory or transient species should be
avoided. Tissues for residual chemical analysis should be well homogenized.
Ideally, tissue homogenates should consist of organisms of the same species, sex,
developmental stage, and size, since these variables affect chemical uptake. Fur and
shell should be removed from tissue, as they cannot be practically homogenized as a
whole (EPA, 1997).
Keep this in mind that, in addition to the physical variations discussed above,
''representativeness'' always depends on the project objective. This means that
samples are representative in one case, but the same samples will not be
representative for another purpose. For illustration, consider Figure 3.4, a site
map for a dry lagoon formerly fed by a pipe discharging wastewater. The analytical
results of soil samples drawn from randomly located sites A, B, and C may be
representative if the objective is to address whether the pipe released a particular
contaminant. However, these data are not representative if the objective is to
estimate the average concentration in the entire old lagoon. For that estimation,

Area where outfall from
pipe was expected to
accumulate

Discharge pipe Discharge pipe Discharge pipe

B B B

A A A H H H

C C C

G G G

F F F

D D D

Figure 3.4 Sampling for an old lagoon to
illustrate the dependence of sample
representativeness on the project objective
(U.S. EPA, 2002)

Old lagoon
(now dry)

Old lagoon
(now dry)

Old lagoon
(now dry)

E E E