7019

• UCREFRP Catalog Number: 7019
• Author: Hawkins, J. A.
• Title: Age and Growth Of Colorado Squawfish From the Upper Colorado River Basin, 1978-1990.
• USFW Year: 1991.
• USFW - Doc Type: Fort Collins, CO.
• Copyright Material: NO

4

sides of equation (I) giving:

log10 W = IOg10 0 + b IOg10 L,

(2)

where 10glo 0 is the intercept and b is the slope of the line. These constants are estimated by
either least-squares (ordinary major axis) or geometric-mean (standard major axis) regression.
The geometric-mean method is often recommended because weight is not truly independent of
length and both are subject to errors or fluctuations (Ricker 1973, 1975a and 1975b; Anderson
and Gutreuter 1983; Bolger and Connolly 1989), but this method has been considered
inappropriate and difficult to interpret (Sprent and Dolby 1980; Cone 1989). Least-squares
regression is commonly used due to its "better understood" statistical properties (Bolger 'and
Connolly 1989; Jolicoeur 1975). An excellent review of methods for determining the
weight-length relationship was provided by Le Cren (1951).

: ,

Fish from each of the four major rivers (Colorado, Green, White, and Yampa) were
grouped by month of capture, and a regression was calculated by the least-squares method using
equation (2). An analysis of co-variance between months was performed on regression equations
using SAS statistical packages (SAS Institute. Cary, N.C.) for micro-computer.

The regression equation of weight and length provides a good description of a population
or subgroup, but a comparison between groups is not easy because two parameters (0 and b) must
be considered (Bolger and Connolly 1989). Often only slopes are compared, with larger slopes
indicating faster growth and therefore better condition, but slopes should be compared only when
intercepts are equal (Bolger and Connolly 1989). If slopes of different subgroups are equal, as
determined by analysis of covariance, then intercepts are good indicators of the condition of each
group (Le Cren 1951; Bolger and Connolly 1989). .

Condition indices

Condition factor is derived by rearranging equation (1) into:

0- W. .
-[}i

(3)

A condition factor often used is Fulton's condition factor (K):

K = Jf.,-. * 10 n
L.;J ,

(4)

where n equals 2, 3, 4, or 5 depending on units of measure. The scaling constant (n) converts the
result to a mixed number for better comprehension (Anderson and Gutreuter 1983; Cone 1989).
K represents growth of an idealized fish with isometric growth where weight is proportional to
length cubed. K is easy to calculate and does not require knowledge of the weight-length
relationship, but problems with K are many. K will vary for the same fish depending on whether
metric or English measurements are used (Anderson and Gutreuter 1983). It assumes that growth
is isometric (b=3). Fish of different sizes cannot be compared because K will increase with length
if b for the population is greater than 3. Anderson and Gutreuter (1983) suggested limiting
comparisons to similar lengths. Cone (1989) cautioned that each stratum (e.g. sex, strain, growth
stanza) to be compared should be cheCked to confirm isometric growth.

Although used regularly, Fulton's condition was determined inappropriate because it does
not account for allometric growth. Instead, relative condition (Kn) was calculated for individual
fish from each river with the formula:

Kn = W b* 100
oL '

(5)