7332

• UCREFRP Catalog Number: 7332
• Author: Valdez, R. A., et al.
• Title: Final Report Habitat Suitability Index Curves for Endangered Fishes of the Upper Colorado River Basin.
• USFW Year: 1987.
• USFW - Doc Type: \
• Copyright Material: NO

I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I

tributaries). Since it could not be determined from the database if their
presence was associated with spawning, the data were used for nonspawning
adults. For the Colorado squawfish, partitioning by time of year was made
along the same lines of reasoning. Data on size 2 (25-150 mm) fish were
di vided into March-June vs. July-November in the Green River because of
perceived differences in habi tat use. Size 5 (400+ rom) squawfish were
partitioned into April-June vs. other times of the year for data from the
Colorado, Green, Yampa, and White rivers because the experts had observed large
squawfish using large backwaters and quiet riverside habitats during spring
months, which are not available at other times of the year. The experts noted
that the majority of habitat data collected on the three target species in the
upper basin are from March through October, and that no habitat data are
available for the winter months. Studies are currently being conducted with
radiotelemetered Colorado squawfish and razorback sucker in the Green and Yampa
rivers to address this data gap.

7. Time of Day. Although the experts acknowledged apparent or perceived
differences in habitat use by some life stages by time of day, in only one case
was the request f!lade to partition data by this criterion. A special
statistical test was performed on habitat data collected from radiotelemetered
adult Colorado squawfish from the Yampa River. For the period April-June, the
test groups were 6:00 a.m. - 6:00 p.m. vs. 6:00 p.m. - 6:00 a.m. (day vs.
night). For other times of the year, the test groups were 5:00 a.m. - 9:00
p.m. vs. 9:00 p.m. - 5:00 a.m.. The experts felt that, except for ~hese data,
there are insufficient collections taken randomly over the 24-hour period to
allow for an unbiased analysis. Most habitat data were collected between 6:00
a.m. and 10:00 p.m..

8. Sample Design. Two issues were identified as options for partitioning
data by sample design; these were randomized vs. nonrandomized sampling, and
multiple vs. single measurements of depth, velocity, substrate on
radiotelemetered fish. The experts agreed that data collected from only one
specific location in a nonrandomized fashion should not be included in HSI
curve development because the use of such data could bias the results heavily
toward the habitat sampled and its associated depth, velocity, and substrate
measurements. Regarding radiotelemetry data, the experts agreed that only
depths, substrates, and mean column velocities at the location of the fish
would be used, since not all of the database contained the additional
measurements taken 1 meter out and 1 meter in from the fish location. The
experts chose to develop curves on the basis of the fish location measurements
and, at Workshop *2, compared the velocities taken at the fish location with
those taken 1 meter on each side.

9. Habitat Parameters. The experts agreed by consensus that, as a
minimum, HSI curves would be developed for depth, velocity, and predominate
substrate type (SUB 1) for each partition of data identified in each size class
of each species. To provide the user of these preliminary curves with insight
to the distribution of these parameters by habitat type, the experts requested
associated histograms of habitat type; one histogram was to accompany each set
of depth, velocity, and substrate criteria. Although the experts identified
temperature as a fourth habitat parameter of importance, it was noted that
temperature cannot be considered as microhabitat in the traditional sense,

15