9726

• UCREFRP Catalog Number: 9726
• Author: Marsh, P.C., C.A. Pacey and B.R. Kesner.
• Title: Decline of the razorback sucker in Lake Mohave, Colorado River, Arizona and Nevada.
• USFW Year: 2003.
• USFW - Doc Type: Transactions of the American Fisheries Society
• Copyright Material: YES

1254
1600
1400
1200
1000
800
z 600
400
200
MARSH ET AL.
FIGURE 1. Number of unmarked and marked razor-
back suckers captured during annual springtime sam-
pling in Lake Mohave, Arizona and Nevada, 1991-2002.
Data are from calculations of annual single-census pop-
ulation estimates (Chapman modification of the Peterson
method; Ricker 1975). The solid line with triangles rep-
resents M, the number of fish marked during the marking
or previous year. The thick line is the trend line from
the regression of M versus time. The dotted line with
squares represents R, the number of marked fish captured
in the sampling or current year that were originally
marked in the marking year.
number of fish marked in 1991 and not necessarily
to a catastrophic event. After a decade of sampling,
the population estimate for 2001 is only 13% of
the 1992 estimate. Data indicate a significant
downward trend over time for the razorback sucker
population (R2 = 0.67; Figure 1). An apparent neg-
ative trend in R was not significant (R2 = 0.16).
Although N* generally gives an unbiased esti-
mate, statistical bias in the marking year 1999
needs to be addressed, because an R of fewer than
4 fish makes bias probable (there was only one
capture in 2000 of the fish marked in 1999). Fur-
ther, the condition of an unbiased estimate of N*
if (M + C) > N* is not met for any of the years,
and therefore N* has a negative bias, which results
in underestimation of the number of fish (Ricker
1975).
For the PVA, R was estimated through linear
regression as -0.27, U2 as 0.036, and the average
population growth rate (A-) as 0.78 (95% CLs =
0.69, 0.87). The population growth rate value pre-
dicts a decrease in the average of population
curves. The analysis of variance results indicated
that regression analysis was nonsignificant (P >
0.05), and therefore PVA was suitable for the data.
Mean time (years) to extinction was estimated to
be 14.65 (95% CLs = 9.84, 20.395), 12.10 (95%
CLs = 8.13, 16.84), and 6.19 (95% CLs = 4.16,
8.62) for thresholds of 50, 100, and 500 individ-
uals, respectively (Figure 2). Median time to ex-
tinction, which is the time at which extinction
I
0.9
0.8
c 0.7
06
4
01,
> 03
0.2
U a
r
5 10 15 20 25 30
Yews
- - 50 indviduals --- -100 individuals -500 mdividuels
FIGURE 2.-Cumulative probability of extinction for
the razorback sucker population in Lake Mohave over
time. Estimates were calculated from N*, the annual sin-
gle-census population estimate for the marking year
2002 (Chapman modification of the Peterson method;
Ricker 1975). The dotted line at the y-axis value of 0.5
represents the median time to extinction. The broken
line at 0.95 represents the point at which the probability
of population persistence is 5%.
probability reaches a value of 0.5, was similar to
the mean (Figure 2). The CDF also showed that a
5% probability of persistence would be attained
for thresholds 50, 100, and 500 individuals in ap-
proximately 20, 17, and 9 years, respectively. The
probability of ultimate extinction for all thresholds
combined could be realized within 24 years from
now.
Discussion
The results of our study predict that the Lake
Mohave population of adult razorback suckers is
rapidly dwindling toward functional extirpation.
Marsh and Minckley (1992) previously found no
detectable change in annual mark-recapture pop-
ulation estimates over the 5-year period, 1986-
1991. Similarly, from 1974 to 1981 and from 1981
to 1991, there was no evidence of change in ra-
zorback sucker catch per unit effort (Minckley
1983; Bozek et al. 1984; Marsh and Minckley
1992). Population estimates have consistently de-
creased over the last 11 years. Although fishing
effort remained relatively constant from 1991
through 2002, total razorback sucker catch de-
creased, while the proportion of marked captures
slightly increased (Figure 1). This result is con-
sistent with a decline in overall abundance due to
mortality. As predicted in Figure 2, time to ex-
tinction could be as short as 6 years for remaining
razorback suckers. Only in both the marking and
sampling year 2000 was C equal to M, which was
probably due to the absence of captures of fish
marked in 1999. This was likely a result of mark-
ing only 33 of a possible 244 fish, which also was
reflected in R from the 2000 sampling year.
M
1991 1992 1993 1994 1995 19% 1997 1998 1999 2000 2001 2002
Year