WSP10787

• File Number: 8170
• Description: Arkansas Basin Water Quality Issues
• State: CO
• Basin: Arkansas
• Water Division: 2
• Date: 9/1/1981
• Author: Bureau of Reclamati
• Title: Heavy Metals Pollution of the Upper Arkansas River - Colorado - and its Effects on the Distribution of the Aquatic Macrofauna
• Water Supply Pro - Doc Type: Report/Study

and substrate in front of the net are then washed
by hand or disturbed and the .Ioosened organ-
isms float into the net. This type of sampler is
most effective when the substrate is small
enough to be worked by hand, and the flow
velocity is strong. Depth should be less than
about 300 mm, except for the traveling-kick
method which is most effective in 1 50 to
600 mm of water.

Grab samplers may be used in deep, slow-
flowing rivers. Most grab samples are limited to
substrates of fine sands or mud. Gravel or larger
stones may inhibit the action of the sampler by
not allowing the jaws to close properly IHella-
well, 1978 [721 ).

Dredge and shovel samplers are useful in
shallow as well as deep waters, provided the
substrate is only coarse gravel or finer material.
Usinger and Needham (19561 (1541 fitted a
dredge sampler with steel tines along the leading
edge to assure that large stones could not enter
and damage the net.

Cylinder and box samplers enclose an area of
substrate, and the organisms within are re-
moved. Hynes 119711 185J suggests the use of
a hand net for dipping the organisms out.

Core samplers operate best on sediments that
are relatively fine and compacted. Penetration
depends on the weight and sharpness of the
coring apparatus and the density of the
substrate, Hamilton et al. (19701 167]
developed multiple coring units which may take
several cores simultaneously.

Techniques for evaluating the colonization of
aquatic macroinvertebrates have been described
by many investigators (Hester and Dendy, 1962
(73); Mason et aI., 1967 1103); Anderson and
Mason. 1968 (31; Arthur and Horning, 1969
(5); Hilsenhoff, 1969 [74]; Kreis and Smith,
1970 [93]; Dickson et aI., 1971 [37); Fullner,
1971 (59]; Jacobi, 1971 (86); Beak et al..
1973 (6); and Bergerson and Galat. 1975
(101).

Meier et al. (1979) [109) found that the
exposure time for maximum colonization of an
artificial substrate was 39 days. The procedures
used in this type of study require a variety of
substrates such as hardboard or porcelain
plates, limestone or porcelain spheres. sticks.

bark, concrete. styrofoam, and stones. These
artificial substrate samplers should be con-
sidered experimental techniques and do not truly
represent the macroinvertebrate populations
found in the stream. However, they are ex-
tremely useful in difficult sampling situations
and pollution studies because they provide a
uniform substrate. Many environmental factors
affect the performance and efficiency of arti-
ficial substrate samplers as they do other sam-
pling procedures.

Number and Frequency of Macroinvertebrate
Samples

True quantitative sampling is difficult in aquatic
systems because of variations in the environ.
ment and the resulting "patchiness" ii,e"
clumped distributionl in animal populations.
Totally random sampling uses a grid of the study
area on which sampling locations are selected
using random numbers. Stratified random
sampling uses the grid and random number
approach to a selected habitat type. A third type
of systematic sampling involves the transect
method or sampling at regular intervals.

The number of samples needed to characterize
the aquatic macrofauna also varies widely.
Gaufin et al. (19561 162) concluded that three
samples were needed to provide at least 50 per-
cent of the total number of species. Needham
and Usinger 119561 1119] reported that at least
two samples would be needed to be 95-percent
confident that at least one Plecoptera,
Ephemeroptera, Tricoptera. and Diptera would
be present in a sample. Thorup (19701 1150)
concludes that 20 samples are necessary to give
reliable estimates for the dominant species.
Chutter and Noble (19661 (26\ have shown
that a large increase in the number of samples
result in a small improvement in sampling effi-
ciency. Chutter (19721 (25) concluded that the
allowable error of the population mean was
40.5 percent. Cuff and Coleman (1979) (30)
suggest that substantial gains in precision in
benthic sampling may be obtained with the
same number of samples by increasing the
number of stations and reducing the number of
grabs per station to an optimum of one.

Buikema and Benfield (1979) (15) stress the
importance of life history in interpreting
biological information. and Resh (1979) (129J
suggests that sampling at monthly intervals may

6