7610

• UCREFRP Catalog Number: 7610
• Author: Wick, E. J. and J. A. Hawkins.
• Title: Colorado Squawfish Winter Habitat Study, Yampa River, Colorado, 1986-1987
• USFW Year: 1987.
• USFW - Doc Type: Final Report Year 1.
• Copyright Material: NO

•
8
Data analysis
•
Data was transferred from data sheets to database files on an IBM
compatable computer using fields identified in Appendix B. Additional data
fields were added to the data base to aid in data stratification, calculations
and compatability with other databases. Negative velocities were converted to
positive values in calculation of descriptive statistics and on histograms. •
Each field observation of a fish consituted one record within the
database. There are 1332 records within the database "master" file. These
records were compressed into a "working" file of 317 records. Each record in
the "working" file contains the time interval that the fish spent in each
location and the microhabitat measurements from that location. A particular
map location could have several entries depending on the number of times a •
fish frequented it during an observation period. Also, several radiotagged
squawfish could use the same location during an observation period.
Based on study design, data were partitioned into two groups. The long
term (24 hour) partition consisted of one 24 hour observation on each of 10
fish. The short term (2.5 hour) partition contained 2.5 hour observations
conducted biweekly on each fish. Target group observations scheduled at •
different times were combined with non-target observations which usually
occurred during daylight hours. When simultaneous observations were being
conducted on additional fish (other than the fish scheduled for observation)
the additional data was entered as extra data. Extra data were removed for
micro-habitat analysis since there was not an equal chance of collecting it on
all fish and because of potential spiking of histogram intervals where several •
fish were observed together at the same site location (Bovee 1986).
The long term (24 hour) data partition was used to examine behavioral
differences between morning, afternoon, evening, and night time partitions.
The analysis was conducted to investigate temporal differences in depth,
velocity, substrate, and habitat type use.
Determination of differences among the time periods was conducted by •
visual examination of histograms of the microhabitat parameters and comparison
of summary statistics of means, variances, and range of depth and velocity.
Statistical tests were not used in this comparison because of a lack of
independence in the data.
Measurements were compared between the 24 hour and 2.5 hour data sets to
determine if the two sampling techniques would produce similar results and if •
combining the data sets was warranted. Measurements were also compared between
the main, inner, and outer holes using 2.5 hour, 24 hour, and.combined 2.5 and
24 hour data sets.
Comparisons of "in" and "out" measurements were made to examine
variability of habitat near fish locations and determine if there was
selection of certain habitat parameters by the fish. •
This preliminary data analysis approach was taken to enable comparisons
between data collected in this study with that collected on a similar study on
the Green River.
•
0