9727 - Laserfiche WebLink

Home Browse Search

550 MARSH ET AL. TABLE 1.-Summary of numbers of razorback Suckers released as juveniles and later captured during annual March censuses in Lake Mohave, Arizona-Nevada, 1992-2002. The total number of fish captured by release year includes multiple captures of the same fish in different years. Release Number year released 1993 1994 1995 Number captured by capture year 1996 1997 1998 1999 2000 2001 2002 Combined 1992 131 4 7 2 3 3 1 3 1 2 4 30 1993 1,342 1 3 3 4 3 2 0 2 1 4 23 1994 2,167 0 6 7 4 4 4 5 3 33 1995 1,497 20 13 22 16 11 18 15 115 1996 3,076 5 14 41 26 26 21 133 1997 7,271 2 8 11 12 4 37 1998 7,657 14 16 39 20 89 1999 15,993 2 12 32 22 68 2000 7,159 28 17 45 2001 11,404 1 1 2002 81a 0 Total 57,778b 5 10 5 33 31 45 88 83 163 111 574 a Number released through April 1. b 295 records with duplicate data not included. Five records with unknown release not included. divided into three categories (<_300 mm, 301-350 mm, and >350 mm); the intermediate size-class contains the size release target. Two other factors considered in the analysis were capture status (i.e., captured or not captured in the census year), and census year (i.e., 1999-2002). To eliminate the confounding factor that larger fish were more prone than smaller fish to capture in the first years following release, we imposed an approximate 3- year delay to allow releases to attain a catchable size (e.g., release TL data from repatriates cap- tured in 2002 would include only fish released be- fore January 1, 1999). This delay also allowed fish to experience the majority of size-based posts- tocking mortality and attain lengths at which mor- tality was assumed to be asymptotic with respect to size. We used Program MARK to fit the data to a mark-recapture model and derive maximum like- lihood estimates for apparent survival ((D) and re- capture (p) rates (White and Burnham 1999). This program uses an iterative method to estimate pa- rameters and allows many conceptual models to be fitted and compared. Age-structured (cohort) and time-structured (Cormack-Jolly-Seber) mod- els were initially considered, and size at release was modeled as a covariate of first-year and sec- ond-year survival and recapture rates. Akaike's in- formation criterion (AIC) was used to compare model fit. Growth was examined by fitting the release and capture data from all repatriate captures to a von Bertalanffy Growth (VBG) curve model. Although reparameterized models of the VBG exist specif- ically for mark-recapture studies (e.g., Francis 1988; Haddon 2001), the main objective of the VBG for our study was to determine the number of years a fish required to reach adult size after release. This was best achieved utilizing an alter- native method that for a given fish (i.e., per unique PIT tag) used its TL at release and at just one capture or recapture, so that multiple recaptures of the same fish across years did not result in addi- tional growth curve influence. Actual age at re- lease was unknown due to cohort mixing at rearing sites, and thus age-0 denotes age at release into the lake while capture age was the number of years at large after release. A best-fit VBG curve is Li = LJ1 - e-K(,-1o)11 where growth parameters Lco (asymptotic length of the growth curve), K (growth coefficient de- scribing rate at which the asymptotic length is ap- proached), and to (theoretical age at length 0) were assessed using the solver routine in Microsoft Ex- cel 2000 to minimize the sum-of-squared residuals (least squares). Male and female growth rates and maximum adult sizes differ (Minckley 1983; Mc- Carthy and Minckley 1987), and thus, gender-spe- cific curves were generated (juvenile or unknown gender data were omitted). Results The repatriate razorback sucker database con- tained 58,073 releases and 822 capture or recapture records from 1992 through April 2002. After the 295 duplicate (same tag number recorded twice) release records were removed, 57,778 release re- cords remained in the database (Table 1). Also, 9 TAI 1973) March fish c; repatri triated captu relew recap captu Thus, fish h An estim, from we re: censu The r( captul could lease Am calcul becau timatt captui four e captui (a rec (1973 bias i seven is app erated recapt Fisl 350 m numb( of the found ysis: 1 with t size-cl rates i