9598

• UCREFRP Catalog Number: 9598
• Author: Lubinski, K., J. Wiener and N. Bhowmik.
• Title: Regulated Rivers Research and Management.
• USFW Year: 1995.
• USFW - Doc Type: \
• Copyright Material: NO

FISH MONITORING
CHIC
M
N N
n
U)
? o
0
_.- ....... .- -
se
OF
-- dec
10 15 20
# Runs
NICK
M
N N
O.
? O
O

10 15 20
# Runs
O
co
2)
N N
a
V7
? o
0
co
A)
lV N
a
U)
yk O
O
10 15 20
# Runs
271
Figure 5. Selected species accumulation curves generated from 20 electrofishing runs in inflow zones. Reservoirs sampled in supplementary
study in 1992 were Chickamauga, Kentucky, Nickajack and Watts Bar. Note asymptote is not achieved before 15 runs, suggesting that
samples with only 10 runs are not adequate to characterize fish assemblages present at a site
for RFAI scores across months (Figure 3). If the measurement error had been high relative to the observed
monthly variability, we would conclude that the RFAI is too sensitive to slight differences in species com-
position or abundances within the fish assemblage. On the contrary, the high monthly variability suggests
a source of variability other than measurement error.
Resampling at the level of the electrofishing run by choosing 10 of 20 available runs provides an estimate of
sampling error, or variation in scores resulting from random variability among run locations. The measure-
ment error is generally less than the sampling error for the autumn 1992 supplemental data set (Figure 4).
The availability of 20 electrofishing runs allowed the construction of species accumulation curves (Figure 5).
These curves indicate the rate at which new species are being added to the sample for each electrofishing run.
A sampling plan which extends to the asymptote of the curve will provide a more reliable sample than a plan
that terminates before the asymptote. In each reservoir, new species were being added to the sample at the
15th run. This indicates that the standard sample of 10 runs is insufficient to generate a representative sample
of the fish assemblage. We conclude that much of the overall variability in index scores results from inadequate
sampling at a site.
DISCUSSION
A successful RFAI can be used to assess resource condition and even diagnose the factors responsible for
degradation. In addition, it can track reservoir condition following management actions. As a monitoring
tool, it may alert managers to problem areas that require further, more detailed investigation. To be
successful, however, a multimetric index must overcome three challenges: (1) the index must incorporate
the relevant biology so that it provides a meaningful measurement of the status or condition of the fish
assemblage; (2) the samples on which the index is based must accurately represent critical and relevant
attributes of the sampled fish assemblage; and (3) the index scores must be a sensitive integrator of biological
responses to stress from human-induced impacts.
KENT
----___-
:.. ...............
10 15 20
# Runs
WATT