|
1253
<br />NOTES
<br />TABLE 1.-Annual single-census population estimates (Chapman modification of the Peterson method; Ricker 1975)
<br />of razorback suckers captured during annual springtime sampling in Lake Mohave, Arizona and Nevada, 1991-2002.
<br />Abbreviations are as follows: N* is the annual single-census population estimate for the marking or previous year, CL
<br />stands for confidence limit, M is the number of fish marked during the marking year, C is the combined number of
<br />marked and unmarked fish captured in the sampling or current year, and R is the number of marked fish captured in
<br />the sampling year that were originally marked in the marking year.
<br />Mark-
<br />ing
<br />year
<br />
<br />N* 95% CL
<br />
<br />Lower Upper
<br />
<br />M
<br />
<br />C
<br />
<br />R
<br />Sampling
<br />year
<br />2001 2,698 1,573 5,081 284 122 13 2002
<br />2000 2,872 1,965 4,392 239 310 26 2001
<br />1999 8,161 1,458 81,601 33a 239 1 2000
<br />1998 4,506 2,627 8,489 226 2576 13 1999
<br />1997 5,355 3,190 9,735 310 240 14 1998
<br />1996 6,678 4,780 9,661 657 344 34 1997
<br />1995 9,322 7,049 12,653 645 706 49 1996
<br />1994 13,517 10,281 17,774 988c 696 51 1995
<br />1993 16,932 13,549 21,162 1,222° 1,0650 77 1994
<br />1992 20,853 17,060 25,491 1,5024 1,317 95 1993
<br />1991 44,333 30,118 68,415 709 1,5600 25 1992
<br />Mean 12,293 619 620
<br />(SD) (12,124)
<br />Total 6,808 7,192
<br />I Does not include 211 fish that were released unmarked.
<br />b Includes 211 fish that were released unmarked.
<br />0 Includes one fish that was released with an unknown PIT tag mark.
<br />d Includes five fish that were released with unknown PIT tag marks.
<br />e Includes 33 fish that were released unmarked and 5 fish released with unknown PIT tag marks
<br />therefore, a "trap-happy" response is not likely,
<br />and a "trap-shy" response would only serve to
<br />increase population estimates. Multiple captures of
<br />the same marked fish during the same sampling
<br />year were removed from analysis because our pri-
<br />mary concern was capture of marked fish over the
<br />course of a year and not within the same sampling
<br />year. The final assumption can be adjusted for tag
<br />loss if it is known (Seber 1973). Although PIT
<br />tags are considered permanent marks (Prentice et
<br />al. 1990), there are instances when they become
<br />defective, are lost, or are overlooked during scan-
<br />ning, all of which bias the estimate. There is little
<br />information on retention and long-term losses of
<br />PIT tags, although in some instances tag loss may
<br />be significant (Buzby and Deegan 1999). Care was
<br />taken to ensure all injected PIT tags were func-
<br />tional by scanning newly marked fish prior to their
<br />release.
<br />Population viability analysis was performed on
<br />N* to determine the mean and median times to
<br />extinction for the razorback sucker population in
<br />Lake Mohave. First, linear regression was used to
<br />determine the suitability of PVA to estimate the
<br />population parameters R and 62. We then calcu-
<br />lated the cumulative distribution function (CDF)
<br />of conditional extinction time, and viability curves
<br />were generated from the adjusted Peterson esti-
<br />mate of the population size from marking year
<br />2002 for extinction thresholds of 50, 100, and 500
<br />individuals. Calculations followed the methodol-
<br />ogy outlined in Morris and Doak (2002) and were
<br />based on 95% confidence.
<br />Our study needed to meet four major PVA as-
<br />sumptions in order to estimate razorback sucker
<br />population viability. First, capture data represent
<br />estimates of total population size based on an as-
<br />sumption that all adult razorback suckers have
<br />equal opportunity for capture. Second, there were
<br />a sufficient number of years with capture data, such
<br />that year-to-year variation in capture data reflected
<br />environmentally driven variations and not study
<br />error (our study had 11 years of data, with rela-
<br />tively constant fishing effort each year). Third,
<br />there have been no extreme environmental chang-
<br />es, either positive or negative (Lake Mohave is a
<br />relatively stable, closed environment). And fourth,
<br />razorback sucker population growth rate was not
<br />affected by density and has not changed as the
<br />population size has decreased (population growth
<br />rate is assumed constant and negative, as there is
<br />mortality and no recruitment).
<br />Results
<br />Razorback suckers have declined more than
<br />94% over the past decade (Table 1). The greatest
<br />decline occurred between 1991 and 1992; how-
<br />ever, this decline may be attributed to the small
|