9360

• UCREFRP Catalog Number: 9360
• Author: Bundy, J. M. and K. R. Bestgen.
• Title: Evaluation of the Interagency Standardized Monitoring program Sampling Technique in Backwaters of the Colorado River in the Grand Valley, Colorado.
• USFW Year: 2001.
• USFW - Doc Type: Fort. Collins.
• Copyright Material: NO

ISMP backwater sampling evaluation
where largemouth bass density was number of fish per 10 m2. Habitat complexity was not a
significant effect in this model. Re-ordering the logistic equation to yield probability of
detection given an investigator-chosen largemouth bass density was of the form:
1

py =
1 + exp(1.781 - 3.065(largemouth bass density))
(2)
where largemouth bass density was number of fish per 10 m2. This relationship assumes average
backwater depth and habitat features for backwaters in the Colorado River in the Grand Valley.
Solutions to the logistic regression equation over a range of bass abundance values suggested that
the probability that the ISMP technique would detect bass in a backwater was relatively low even
when bass were abundant (Fig. 10). For example, a 50 % probability of detection was achieved
only after largemouth bass abundance exceeded 0.58 fish per 10 m2 surface area in an average
size backwater. Assuming that the average backwater surface area in the Grand Valley in 1997
and 1998 was 718 m2, this equates to an abundance level of 42 bass to achieve a 50 % probability
of detection. The positive intercept in this relationship suggested that there was a small
probability of detecting bass with ISMP sampling when estimated abundance was low.
A similar logistic regression analysis conducted for green sunfish suggested that detection
probabilities of ISMP sampling increased as green sunfish density increased (p < 0.0001).
Habitat complexity was not a significant effect in this model, which had the form:
logit Y= -1.5935 + 2.7317(green sunfish density),
(3)
-13-